filename
string | cleaned_text
string |
|---|---|
PMC1971868.txt
|
TITLE: c-erbB- oncoprotein expression in primary and advanced breast cancer. AUTHORS: C. Lovekin, I. O. Ellis, A. Locker, J. F. Robertson, J. Bell, R. Nicholson, W. J. Gullick, C. W. Elston, R. W. Blamey ABSTRACT: Immunoreactivity for c-erbB- oncogene product expression has been investigated in patients with breast cancer using the polyclonal antibody 21N. Three series of patients were studied, presenting with primary operable cancer, with stage and with stage disease. Representative tissue sections of each primary tumour were stained using a standard immunoperoxidase technique. Invasive tumour membrane immunoreactivity was assessed and identified in % of patients with primary operable cancer and % in the advanced breast cancer group. The results demonstrate a relationship between poorer survival and oncogene expression in all three patient groups. Patients in the primary operable cancer group with membrane oncoprotein expression had a poorer outcome, % -year survival, compared with those in which membrane expression was absent, % -year survival. The median survival of patients with stage disease with c-erbB- membrane positivity was months compared to months with membrane negativity. In stage disease median survival with membrane expression was . months compared to . months with no membrane expression. In addition in the series of primary cancers a correlation existed between histological grade and membrane immunoreactivity. Multivariate analysis showed histological grade to be a more powerful prognostic factor than c-erbB- protein expression. In conclusion, this study demonstrates, in a large series of patients presenting to one centre, that c-erbB- protein expression is a prognostic indicator in patients with primary operable and advanced breast disease.ImagesFigure BODY:
|
PMC1779275.txt
|
TITLE: Nectin--mediated entry of a syncytial strain of herpes simplex virus via pH-independent fusion with the plasma membrane of Chinese hamster ovary cells AUTHORS: Mark G Delboy, Jennifer L Patterson, Aimee M Hollander, Anthony V Nicola ABSTRACT: BackgroundHerpes simplex virus (HSV) can utilize multiple pathways to enter host cells. The factors that determine which route is taken are not clear. Chinese hamster ovary (CHO) cells that express glycoprotein D (gD)-binding receptors are model cells that support a pH-dependent, endocytic entry pathway for all HSV strains tested to date. Fusion-from-without (FFWO) is the induction of target cell fusion by addition of intact virions to cell monolayers in the absence of viral protein expression. The receptor requirements for HSV-induced FFWO are not known. We used the syncytial HSV- strain ANG path as a tool to evaluate the complex interplay between receptor usage, membrane fusion, and selection of entry pathway.ResultsInhibitors of endocytosis and endosome acidification blocked ANG path entry into CHO cells expressing nectin- receptors, but not CHO-nectin- cells. Thus, under these conditions, nectin- mediates pH-independent entry at the plasma membrane. In addition, CHO-nectin- cells supported pH-dependent, endocytic entry of different strains of HSV-, including rid1 and HFEM. The kinetics of ANG path entry was rapid (t1/ of – min) regardless of entry route. However, HSV- ANG path entry by fusion with the CHO-nectin- cell plasma membrane was more efficient and resulted in larger syncytia. ANG path virions added to the surface of CHO-nectin- cells, but not receptor-negative CHO cells or CHO-nectin- cells, induced rapid FFWO.ConclusionHSV- ANG path can enter CHO cells by either endocytic or non-endocytic pathways depending on whether nectin- or nectin- is present. In addition to these cellular receptors, one or more viral determinants is important for the selection of entry pathway. HSV-induced FFWO depends on the presence of an appropriate gD-receptor in the target membrane. Nectin- and nectin- target ANG path to divergent cellular pathways, and these receptors may have different roles in triggering viral membrane fusion. BODY: BackgroundProductive entry of HSV into host cells proceeds following endocytosis [] or by direct penetration at the cell surface []. The viral and cellular factors that determine which pathway is utilized are not clear. The viral envelope glycoproteins gB, gD, and gH-gL are required for entry by both endocytic and non-endocytic routes [-]. Expression of a cellular entry receptor is required for both penetration at the plasma membrane and for penetration following endocytosis [,-]. Such receptors function individually and can mediate entry into non-permissive cells, such as Chinese hamster ovary (CHO) cells []. The viral ligand for HSV entry receptors is gD [-]. In the absence of a gD-receptor, HSV is still endocytosed by CHO cells, but fails to penetrate the endosomal membrane and is degraded [].The known gD-receptors include nectins, which belong to a subgroup of the immunoglobulin (Ig) superfamily [-]. They are broadly distributed cell-cell adhesion molecules that are components of cadherin-based adherens junctions []. Nectin- and nectin- are ~% identical, and their N-terminal Ig-like variable (V) domains are critical for gD-binding [,-] and for viral entry [,-]. All HSV strains tested to date [,,] are able to utilize nectin- as an entry receptor. Nectin- mediates entry of several laboratory strains and clinical isolates of HSV- and HSV-, including HSV- isolates from the CNS of patients with herpes simplex encephalitis [,]. Amino acid changes in gD at residues , , or confer the ability to utilize nectin- [,,,]. Additional gD-receptors include HVEM, a member of the TNF-receptor superfamily [] and heparan sulfate that has been modified by -O-sulfotransferase- []. Nectin- [] and B5 [] also mediate HSV entry, but their viral ligand(s) is not clear.Following endocytosis from the cell surface, HSV entry into a subset of cell types also requires intracellular low pH [,,,,]. CHO cells expressing gD-receptors are a widely used, well-characterized model system to study pH-dependent, endocytic entry. Inhibitors of endosomal acidification block HSV entry at a step subsequent to endocytic uptake but prior to penetration of the capsid into the cytosol []. It has been proposed that HSV utilizes distinct cellular pathways to enter its relevant target cells []. Alphaherpesviruses undergo pH-dependent, endocytic entry into certain epithelial cells [,,], including primary human epidermal keratinocytes [], yet utilize a pH-independent entry pathway into neurons [,,]. Recently, Whitbeck et al. showed that in vitro binding of HSV to liposomes could be triggered by a combination of receptor-binding and low pH [].Direct study of the membrane fusion activity of herpesvirions has proven difficult. Fusion-from-without (FFWO) is the induction of target cell fusion by addition of intact virions to the monolayer surface in the absence of viral protein expression. Virus-cell fusion during entry and virion-induced FFWO are analogous inasmuch as both involve similar effector (virion) membranes and target membranes. Several syncytial strains of HSV-, such as ANG path, are capable of triggering FFWO []. HSV-induced FFWO is cell type-dependent [], but the receptor requirements of FFWO are not known. In the present study, ANG path is used as a tool to investigate the influence of viral and cellular proteins on the route that HSV takes into cells. The ANG path-CHO cell model system allows examination of the inter-relatedness of gD-receptor usage, HSV-induced fusion, and selection of entry pathway.ResultsHSV- strain ANG path can utilize nectin- or nectin- for entry into CHO cellsFirst we determined that nectin- or nectin- can each function to mediate HSV- ANG path entry into CHO cells. All strains of HSV- and HSV- can utilize nectin- for entry. The HSV- strain ANG path and its parent ANG have alterations in gD at positions and that are predictive of nectin- utilization [,,,]. ANG utilizes both nectin- and nectin- for entry into CHO cells [,]. Monolayers of CHO cells expressing nectin- or nectin- were infected with serial dilutions of HSV- ANG path. As expected, ANG path failed to infect receptor-negative CHO cells (Fig. 1A), but formed syncytia on CHO nectin- and CHO-nectin- cells (Fig. 1B and 1C). Similar results were obtained using a beta-galactosidase reporter assay for HSV entry (data not shown). The ANG path syncytia that formed on CHO-nectin- cells were ~% larger than those that formed on CHO-nectin- cells (Fig. 1B and 1C). The larger plaque size may reflect enhanced entry activity and/or cell-to-cell spread mediated by nectin-.Figure 1Syncytium formation of HSV- strain ANG path on CHO cells expressing nectin- or nectin-. HSV- ANG path was added to wild type CHO (A), CHO-nectin- (B) or CHO-nectin- cells (C) for h. Based on the titration of ANG path on Vero cells, the MOIs were (A), (B), or (C). Visualization of syncytia was facilitated by immunoperoxidase staining with HR50 antibody to HSV. Magnification, ×.HSV- strain ANG path has enhanced plating efficiency on nectin- cells relative to nectin- cellsPlaque-forming strains of HSV such as KOS and KOS-rid1 do not form substantial plaques on receptor-expressing CHO cells. Hence, to determine the plating efficiency of ANG path we employed the syncytial HSV- strain MP [] for comparison. Unlike many other strains, HSV- MP enters receptor-negative CHO cells with low efficiency []. The expression of nectin-, but not nectin-, enhances MP entry [,] in the CHO cell background. MP is not a FFWO strain.The plating efficiency of ANG path on CHO-nectin- cells was approximately two logs greater than on CHO-nectin- cells (Table ). The plating efficiency on CHO-nectin- cells was approximately two logs less than that obtained on Vero cells. MP formed syncytia on wild type CHO cells at reduced efficiency (approximately two logs) as compared to Vero cells (Table ). The presence of nectin- did not enhance MP infection above the CHO cell background, but instead reduced the plating efficiency for reasons that are not clear. MP had a log enhanced plating efficiency on CHO-nectin- cells relative to CHO-nectin- cells (Table ), which is consistent with previous reports. Importantly, as CHO-nectin- cells support MP entry and syncytium formation, the reduced efficiency of ANG path entry is not due to receptor expression levels or some other defect of the CHO-nectin- cells. Also in support of this notion, CHO-nectin- cells are equivalent to CHO-nectin- cells in their ability to support entry of HSV- rid1 []. Together, the results indicate that ANG path can use either nectin- or nectin- for entry into the CHO cell lines, but it utilizes nectin- more efficiently.Table 1Plating efficiency of HSV- syncytial strainsCell typeVeroCHOCHO-nectin-1CHO-nectin-2VirusTiter (PFU/ml)ANG path8. × . × . × 106MP7. × . × . × . × 102Viruses were titered by limiting dilution. Similar results were obtained in at least three independent experiments.HSV- ANG path entry mediated by nectin- or nectin- receptors occurs via distinct cellular pathwaysThe entry of wild type strains of HSV- and HSV- into CHO cells expressing gD-receptors is blocked by agents that affect endosome acidification [], and is consequently considered pH-dependent. Entry of ANG path into CHO-nectin- cells was inhibited significantly by the weak base ammonium chloride (Fig. 2A). Surprisingly, entry of ANG path into the nectin--expressing cells was refractory to inhibition by the low-pH-altering agents. Similar results were obtained with MOIs ranging from . to (data not shown). Thus, ANG path stands out as the only HSV strain known to enter a CHO cell line (CHO-nectin- cells) by a pH-independent pathway. This suggests that nectin- and nectin- direct HSV- ANG path to distinct entry pathways in the CHO cell.Figure 2Dependence of ANG path entry on intracellular low pH and endocytosis. (A) Effect of alteration of intracellular pH on ANG path entry. CHO cells expressing either nectin- or nectin- were treated with the indicated concentrations of ammonium chloride (NH4Cl) for min. HSV- strain ANG path was added (MOI of ) for h in the continued presence of NH4Cl. Cells contain the lacZ gene under the control of an HSV-inducible promoter. Entry was measured as the % of beta-galactosidase activity relative to that obtained in the absence of agent. Data shown are means of quadruplicate determinations with standard deviation. (B) Effect of inhibition of endocytic uptake on ANG path infection. ANG path was added to CHO-nectin- or CHO-nectin- cells ( PFU/well) in control medium (A) or hypertonic medium containing . M sucrose. At min post-infection, medium was removed, extracellular virus was acid-inactivated, and plates were incubated for h. Syncytia were detected by immunoperoxidase staining and quantified. Shown are representative data from at least three independent experiments.ANG path enters CHO-nectin- cells by pH-independent fusion with the plasma membraneThe pH-independence of entry does not necessarily indicate entry at the plasma membrane. For example, entry of Epstein Barr virus into B cells is pH-independent, yet it proceeds via an endocytic pathway [,]. In addition, Milne et al. demonstrated that HSV enters murine melanoma cells by a pH-independent, endocytic pathway []. To assess directly the role of endocytosis, we used cell treatments that selectively block HSV entry by endocytosis. First, we analyzed the effect of high sucrose (hypertonic) medium, which inhibits endocytic uptake of HSV from the plasma membrane, but has no effect on HSV penetration at the plasma membrane []. Treatment of CHO-nectin- cells with hypertonic medium during virus entry inhibited syncytium formation of HSV- ANG path (Fig. 2B). In contrast, hypertonic treatment of CHO-nectin- cells had no inhibitory effect (Fig. 2B), suggesting that ANG path penetrates the CHO-nectin- plasma membrane in a pH-independent, non-endocytic manner. Thus, deposit of the HSV capsid under the plasma membrane of CHO cells can lead to productive entry.CHO-nectin- cells can support either endocytic or non-endocytic entry of HSV depending on the virus strainThe phosphatidyl inositol -kinase inhibitor wortmannin selectively inhibits pH-dependent, endocytic entry of HSV [,,,], possibly at a step involving endosomal trafficking []. To study the effect of wortmannin on ANG path entry, we included the HSV- strain KOS-rid1 [] as a control because it also utilizes both nectin- and nectin- for entry [,]. Wortmannin inhibited rid1 entry into CHO-nectin- cells, but had little inhibitory effect on ANG path entry into these cells (Fig. 3A). Entry of both ANG path and rid1 viruses into CHO-nectin- cells was inhibited by wortmannin in a concentration-dependent manner (Fig. 3B). We also tested treatment of CHO-nectin- cells with monensin, a carboxylic ionophore that inhibits endosome acidification. Monensin inhibited rid1 entry into CHO-nectin- cells as previously reported [], but ANGpath entry was refractory to this treatment (Fig. 3C). These results confirm that nectin- supports a pH-dependent, endocytic pathway for ANG path, and that nectin- supports pH-independent fusion of ANG path with the plasma membrane of CHO cells. As a single cell line, CHO-nectin-, supports distinct entry pathways for two different HSV- strains, this indicates that HSV contains one or more determinants for the selection of entry pathway. Further, receptor-expressing CHO cells can support HSV entry by multiple pathways.Figure 3CHO cell entry pathways of HSV- rid1 and ANG path mediated by nectin-. CHO-nectin- (A, C) or CHO-nectin- cells (B) were treated with the indicated concentrations of wortmannin (A, B) or monensin (C) for min. HSV- strains rid1 or ANG path were added (MOI of ) for h in the continued presence of agent. Entry (beta-galactosidase activity) was measured as in the legend to Figure .Rapid entry kinetics of HSV- ANG path by either endocytic or non-endocytic pathwaysThe kinetics of entry of a single virus strain by two distinct pathways in the CHO cell background was measured. The entry of ANG path mediated by either nectin- or nectin- was rapid, with a t1/ of – min (Fig. ). By min p.i., greater than % of infectious virus had disappeared from the surface of cells regardless of which receptor was present or which pathway was used (Fig. ).Figure 4Kinetics of ANG path entry via distinct entry routes. HSV- ANG path was bound to CHO-nectin- or CHO-nectin- cells in well dishes for h at °C ( PFU/well). Cells were washed with PBS and then shifted to °C. At the indicated times post-infection, extracellular virus was inactivated by treatment with sodium citrate buffer (pH .). Cells were washed with PBS and incubated for h at °C. Cells were fixed in methanol:acetone, and syncytia were quantified by immunoperoxidase staining. Data shown are the mean of quadruplicate samples +/- standard deviation.ANG path virion-induced fusion of CHO cells is mediated by nectin-2ANG path is among the subset of syncytial HSV- strains that cause fusion-from-without. Addition of ANG path to Vero cells at high multiplicity causes rapid cell fusion (FFWO) in the absence of viral protein synthesis [,,]. Receptor-negative CHO cells are an ideal model system to test the role of gD-receptors. Since ANG path utilizes nectin-, but not nectin-, for fusion with plasma membrane during entry, we asked whether nectin- would selectively trigger FFWO when ANG path virions were added to the surface of CHO cells. Fusion-from-without was not detected when ANG path virions were added to receptor-negative CHO cells (Fig. 5A). Similarly, FFWO was not detected when ANG path virions were added to CHO-nectin- cells (Fig. 5B), even after overnight incubation with an MOI of (data not shown). However, by h p.i. in the presence of cycloheximide, ANG path virions induced dramatic FFWO of CHO-nectin- cells (Fig. 5C). Fusion of cells was evident as early as – min p.i. (data not shown). As there was no viral protein synthesis, it is likely that the viral particles themselves triggered the fusion of cells.Figure 5Receptor-dependence of fusion-from-without induced by HSV-. ANG path virions (MOI of ) were added to receptor-negative CHO cells (A) or CHO cells expressing either nectin- (B) or nectin- (C) in the presence of mM cycloheximide. Cells were incubated at °C for h, and were then fixed with methanol and stained with Giemsa. Magnification, ×. Approximately % of CHO-nectin- cells were fused under these conditions. Results with receptor-negative CHO cells (< % fusion) were indistinguishable from CHO-nectin- cells. (D) Antibodies to nectin- block ANG path-mediated fusion-from-without. Anti-nectin- polyclonal antibody R154 or anti-nectin- polyclonal antibody R143 was added to CHO-nectin- cells in well dishes at °C for min. HSV- ANG path was added to the monolayers for °C for h in the presence of a : dilution of antibody. Cells were fixed, photographed, and quantified for ANG path-induced FFWO. Experiments were repeated at least three times with similar results.To demonstrate that nectin- was specifically responsible for triggering FFWO, CHO-nectin- cells were pretreated with antibody to nectin- and assessed for fusion. The anti-nectin- polyclonal antibody R143 inhibited ANG path virion-induced FFWO of CHO-nectin- cells (Fig. 5D). The control anti-nectin- antibody R154 had no inhibitory effect on this fusion process (Fig. 5D). Thus, HSV-induced FFWO depends on an appropriate gD-receptor in the target membrane. The results suggest that the ability of nectin- to mediate rapid, pH-independent entry at the plasma membrane (Fig. and Fig. ) correlates with its ability to trigger rapid, pH-independent FFWO (Fig. ).The HSV- FFWO strain HFEM does not cause detectable nectin- mediated pH-independent fusionWe examined HSV- HFEMsyn to determine whether the entry and fusion phenotypes of ANG path were shared by another strain. Like ANG path, HFEMsyn has a syncytial phenotype and causes FFWO []. Receptor-negative CHO cells were refractory to infection by HSV- HFEMsyn (Fig. 6A). Both CHO-nectin- cells and CHO-nectin- cells supported syncytium formation by HFEMsyn (Fig. 6B and 6C). HFEMsyn utilized nectin- three logs less efficiently than nectin-. HFEM entry into either CHO-nectin- or CHO-nectin- cells was inhibited by both ammonium chloride and monensin (Fig. 6D and 6E), indicating pH-dependent entry in both cell types. ANG path may have a unique determinant that enables entry by fusion with the plasma membrane of CHO-nectin- cells.Figure 6Entry and FFWO activities of HSV- strain HFEMsyn. As in the legend to Figure , syncytium formation of HFEMsyn was determined on wild type CHO (A), CHO-nectin- (B) or CHO-nectin- cells (C). Images represent MOIs of (A, C) or (B). Effect of lysosomotropic agents on HFEMsyn entry. Virus entry into CHO-nectin- or CHO-nectin- cells in the presence of ammonium chloride (D) or monensin (E) was assayed as in the legend to Figure . Cells were infected with HFEMsyn at equivalent multiplicities based on the plating efficiency of HFEMsyn on the respective cell types. Based on Vero cell titer, this corresponds to MOIs of and for CHO nectin- and CHO-nectin- cells, respectively. Receptor-triggered FFWO of HFEMsyn. As in the legend to Figure , HFEMsyn was added to CHO-nectin- (F) or CHO-nectin- cells (G).Unlike ANG path, HFEMsyn triggered detectable FFWO on the CHO-nectin- cells, but not the CHO-nectin- cells (Fig. 6F and 6G). Nectin- can thus trigger HSV-induced FFWO. The results suggest that FFWO does not correlate with plasma membrane fusion during entry. Instead, the ability of a FFWO strain to efficiently utilize a given receptor may correlate with its ability to cause FFWO triggered by that receptor.DiscussionA given animal virus can enter cells by multiple pathways []. HSV can enter its host cells by endocytosis or by direct penetration at the plasma membrane. How a particular pathway is selected is of fundamental importance. CHO cells that express gD-receptors support pH-dependent, endocytic entry of HSV. We identified a laboratory strain of HSV-, ANG path, that can enter CHO cells by pH-independent fusion with the plasma membrane in a receptor-specific manner. Our results indicate that gD-receptors are required for FFWO. Viral determinants, cellular gD-receptors, and the background of the target cell all contribute to the entry route taken by HSV.Host cell determinants of HSV entry pathwayPrevious studies have indicated a role for the target cell in determination of HSV entry pathway [,,,]. Murine melanoma cells are non-permissive for HSV entry. Expression of a gD-receptor results in endocytic uptake of HSV from the cell surface and subsequent pH-independent penetration from an endosome []. In contrast, initial endocytic uptake from the surface of CHO cells occurs independently of the known gD-receptors []. CHO cells may contain unidentified cellular receptors needed for internalization of HSV from the surface. BHK-derived, J cells that express nectin- support pH-independent entry of HSV []. Fusion of nectin- with either carboxy-terminal sequences of epidermal growth factor or with a glycosylphoshatidylinositiol anchor resulted in chimeric receptors that support pH-dependent entry into J cells []. Thus, alternate forms of nectin- can mediate different entry routes.The current study indicates that nectin- and nectin- differ functionally in their ability to target incoming ANG path virions in CHO cells. These receptors interact with distinct yet overlapping regions of gD [,,,,]. In our experimental system, nectin- and nectin- may mediate pH-dependent and pH-independent membrane fusion, respectively. We are currently investigating the receptors and entry pathways that ANG path utilizes in other target cells.Viral determinants of HSV entry pathwayHSV contains one or more determinants for the selection of entry pathway (Fig. ). Candidate determinants include gB, gD, and gH-gL which are essential for entry [-]. Compared to the wild type HSV- KOS strain, the gB [] and gD [] of ANG path have and amino acid differences, respectively. Alterations in gD at positions and [] as well as ectodomain and cytoplasmic tail mutations in gB [,] have been proposed to be important for FFWO activity. The role of these residues in the selection of entry route is currently being evaluated.The composition of the ANG path virion allows direct triggering of fusion by nectin-, at least in the context of the CHO cells tested. One possibility is that ANGpath interaction with nectin- is sufficient to functionally substitute for the combination of nectin- interaction and exposure to intracellular low pH. Analysis of the difference between these receptor interactions may lead to a better understanding of how membrane fusion is triggered during HSV entry. Interestingly, ANG path entry into Vero cells is also unique in that it is highly resistant to inhibition by soluble, ectodomain forms of gD [].Fusion-from-without as a model for membrane fusion during HSV entryA current model of HSV entry posits that gD binding to receptor triggers a cascade of events culminating in fusion [-]. The viral and cellular requirements for HSV entry have been largely recapitulated in a cell-to-cell fusion assay [,-]. In this surrogate assay, transfected cells that express gB, gD, gH and gL on the cell surface are mixed with untransfected target cells. Comparisons of cell-cell fusion with virus-cell fusion must be drawn cautiously. Herpesviral envelopes are derived from internal cellular membranes, not the plasma membrane. It is possible that glycoproteins displayed on the plasma membrane of transfected cells have distinct roles in fusion (i.e., are activated differently) than glycoproteins that are actually incorporated into virions.FFWO is an underutilized model to analyze the membrane fusion activity of HSV particles. Although a high MOI is required to detect FFWO, virus-cell fusion during entry and FFWO have significant similarities. The effector membrane and target cell membrane are analogous for both fusion processes. Furthermore, FFWO, like HSV entry, is gD-receptor dependent.ConclusionTwo members of the nectin family of HSV receptors, nectin- and nectin- can target the same laboratory strain of HSV to endocytic and non-endocytic pathways, respectively. The combination of ANGpath and nectin- at the surface of a CHO cell line triggers rapid, pH-independent membrane fusion that can lead to viral entry or FFWO. An appropriate gD-receptor is required for HSV-induced FFWO. This is similar to the receptor requirement for the membrane fusion processes that accompany viral entry or cell-to-cell fusion. Together, the results indicate that viral factors, in addition to cellular factors such as nectins, contribute to the selection of HSV entry route. This report demonstrates that the ANG path-CHO cell system can serve as a model to study the molecular connections between receptor usage, membrane fusion, and choice of entry pathway.MethodsCells and virusesVero cells (American Type Culture Collection; ATCC; Rockville, Md.) were propagated in Dulbecco's Modified Eagle's Medium (Invitrogen, Grand Island, NY) supplemented with % fetal bovine serum (FBS; Gemini Bio-Products, West Sacramento, Calif.). CHO-K1 cells stably transformed with the Escherichia coli lacZ gene under the control of the HSV ICP4 promoter are designated CHO IEβ8 []. CHO IEβ8 cells stably transformed to express nectin- (M3A cells) or nectin- (M2A cells) [,,] (provided by G. Cohen and R. Eisenberg, University of Pennsylvania) were propagated in complete medium, Ham's F12 nutrient mixture (Invitrogen) supplemented with % FBS, μg/ml puromycin (Sigma, St. Louis, Mo.), and μg/ml G418 sulfate (Fisher Scientific, Fair Lawn, NJ). % of cells expressed nectin- or nectin- on the cell surface as determined by immunofluorescence. Cells were subcultured in non-selective medium prior to use in all experiments.HSV- strains ANG path [] and KOS were obtained from T. Holland, Wayne State University. HSV- MP [] was obtained from ATCC. HSV- HFEM [] and KOS-rid1 were obtained from P. Spear, Northwestern University. Rid1 is a KOS derivative with a Q27P mutation in gD []. Virus stocks were grown and titered on Vero cells.Plaque assayAt – h p.i. culture medium was removed, and cells were fixed with ice-cold methanol-acetone solution (: ratio) for min at -°C and air-dried. Virus titers or syncytium formation were determined by immunoperoxidase staining with anti-HSV polyclonal antibody HR50 (Fitzgerald Industries, Concord, Mass.).Beta-galactosidase reporter assay for HSV entryConfluent cell monolayers grown in well dishes were infected with HSV- and incubated at °C for h. .% Nonidet P- (Sigma) cell lysates were prepared, chlorophenol red-b-D-galactopyranoside (Roche Diagnostic, Indianapolis, In.) was added, and beta-galactosidase activity was read at nm with a microtiter plate reader (BioTek Instruments, Winooski, Vt.). Mean results and standard deviations were calculated for four replicate samples.Inhibition of uptake from cell surfaceHSV was prebound to cells in well dishes ( PFU/well) in culture medium containing mM HEPES and .% BSA at °C for h. Cells were treated with medium containing . M sucrose (hypertonic), or control complete medium for min at °C. Cells were washed with phosphate buffered saline (PBS), and the remaining surface-bound virions were inactivated by sodium citrate buffer (pH .) for min at °C. Cells were incubated in normal medium for h, and then syncytia were quantified.Treatments with lysosomotropic agentsPerformed as reported previously []. Briefly, cells were treated with medium containing ammonium chloride or monensin for min at °C. Virus was added, and cells were incubated in the constant presence of agent for h. Beta-galactosidase activity indicated successful entry.Virion-induced fusion-from-without assayConfluent cell monolayers grown in or well dishes were pretreated with growth medium containing . mM cycloheximide (Sigma) for min. Cell-free supernatant preparations of HSV- ANG path were added to cells at multiplicities from to for up to h at °C in the constant presence of cycloheximide. Cells were rinsed with PBS, and then fixed in % methanol. Monolayers were air dried, and then nuclei were stained with Giemsa.To measure inhibition of fusion by anti-receptor antibodies, cells were chilled to °C for min. Rabbit polyclonal serum against nectin- (R154) or nectin- (R143) (obtained from R. Eisenberg and G. Cohen, University of Pennsylvania) were added for min at °C at a : dilution in culture medium adjusted to mM HEPES. HSV- ANG path was added, and plates were incubated for h at °C in the presence of antibody.Micrographs were taken with a Zeiss Axiovert 40C microscope equipped with a Canon PowerShot G6 digital camera. Digital images were processed with Adobe Photoshop CS2 version .. To quantitate fusion, photomicrographs of random fields from triplicate wells (> cells/well) were scored. The number of nuclei present in clusters of or more divided by the total number of nuclei yielded the % fusion.Competing interestsThe author(s) declare that they have no competing interests.Authors' contributionsMD carried out HSV- rid1, HFEM, and wortmannin analyses. JP carried out fusion-from-without studies. AH performed energy depletion experiments. AN conceived of the study, carried out plating efficiency, kinetics, and lysosomotropic agent experiments, and supervised the work.
|
PMC1762327.txt
|
TITLE: Physiological Mouse Brain Aβ Levels Are Not Related to the Phosphorylation State of Threonine- of Alzheimer's APP AUTHORS: Yoshitake Sano, Tadashi Nakaya, Steve Pedrini, Shizu Takeda, Kanae Iijima-Ando, Koichi Iijima, Paul M. Mathews, Shigeyoshi Itohara, Sam Gandy, Toshiharu Suzuki ABSTRACT: BackgroundAmyloid-β peptide species ending at positions and (Aβ40, Αβ42) are generated by the proteolytic processing of the Alzheimer's amyloid precursor protein (APP). Aβ peptides accumulate in the brain early in the course of Alzheimer's disease (AD), especially Aβ42. The cytoplasmic domain of APP regulates intracellular trafficking and metabolism of APP and its carboxyl-terminal fragments (CTFα, CTFβ). The role of protein phosphorylation in general, and that of the phosphorylation state of APP at threonine- (Thr668) in particular, has been investigated in detail by several laboratories (including our own). Some investigators have recently proposed that the phosphorylation state of Thr668 plays a pivotal role in governing brain Aβ levels, prompting the current study.MethodologyIn order to evaluate whether the phosphorylation state of Thr668 controlled brain Aβ levels, we studied the levels and subcellular distributions of holoAPP, sAPPα, sAPPβ, CTFα, CTFβ, Aβ40 and Aβ42 in brains from “knock-in” mice in which a non-phosphorylatable alanyl residue had been substituted at position , replacing the threonyl residue present in the wild-type protein.ConclusionsThe levels and subcellular distributions of holoAPP, sAPPα, sAPPβ, CTFα, CTFβ, Aβ40 and Aβ42 in the brains of Thr668Ala mutant mice were identical to those observed in wild-type mice. These results indicate that, despite speculation to the contrary, the phosphorylation state of APP at Thr668 does not play an obvious role in governing the physiological levels of brain Aβ40 or Αβ42 in vivo. BODY: IntroductionAlzheimer's disease (AD) is characterized by abnormalities in post-translational processing of two proteins, the amyloid precursor protein (APP) and the microtubule associated protein tau []. A “unifying hypothesis for Alzheimer's disease” has been suggested, according to which the aberration of a single post-translational modification, such as protein phosphorylation, might underlie both pathologies []–[]. Both proteins are phosphoproteins []–[], and cyclin-dependent protein kinase (cdk5) and glycogen synthase kinase 3β (GSK-3β) are examples of protein kinases that can potentially phosphorylate both APP and tau []–[].A recent review summarizes the extensive literature linking protein phosphorylation and APP metabolism []. First messengers, such as neurotransmitters and hormones, impinge upon neurons and direct APP toward the cell surface and away from the trans Golgi network (TGN) and endocytic pathway, and hence away from the Aβ-generating β-/γ-secretase pathway [ibid]. At the cell surface, APP can be processed by a nonamyloidogenic pathway, known as the α-secretase pathway, via a process known as “regulated APP ectodomain shedding” [ibid]. With regard to Aβ generation, the regulated shedding phenomenon is noteworthy because hyperactivation of the α-secretase pathway can lead to relatively greater cleavage of APP by α-secretase(s), thereby reducing or completely abolishing Aβ generation [ibid]. Despite the well-documented observation that the same protein kinase (i.e., protein kinase C) controls both regulated shedding by α-secretase and the phosphorylation state of the APP cytoplasmic tail at serine- (Ser655), the phosphorylation state of APP at Ser655 is not involved in controlling regulated shedding of the APP ectodomain. Primary citations for this summary are included in [].The exclusion of an important role for the APP phosphorylation state in controlling regulated APP ectodomain shedding process raised an obvious question: “What role is played by physiological regulation of the direct phosphorylation of APP?”. First of all, in brain, APP phosphorylation predominantly involves Thr668, not Ser655 []–[]. Phosphorylation at Thr668 is detectable in mature APP (mAPP) but not immature APP (imAPP) [], suggesting that the phosphorylation state of Thr668 may regulate some aspect of APP maturation in the early secretory pathway. In addition to cdk5 and GSK-3β mentioned above, phosphorylation at Thr668 can be regulated by c-jun N-terminal kinases (JNK) [], []. With regard to the physiological role of APP phosphorylation, the phosphorylation state of Thr668 has been reported to modulate outgrowth of neurites in PC12 cells [], [], the tethering to APP of the adaptor proteins FE65 [], [] and X11-L [], and the interaction of APP intracellular domain fragment (AICD) with FE65 [], []. Based on these data, we have proposed a model wherein the phosphorylation state of APP at Thr668 may govern the state of activation of an intracellular signaling cascade across APP that leads to generation and translocation of AICD, analogous to the well-characterized signaling cascade across Notch that leads to generation and translocation of the Notch intracellular domain [].Recently, the phosphorylation state of APP Thr668 has been proposed to control interaction of the APP cytoplasmic tail with the prolyl isomerase Pin1 [], [], []. Interaction of Pin1 with APP Thr668 has been predicted to have important effects on generation of the Aβ peptide that not only accumulates in amyloid plaques but also forms oligomers that have recently been proposed to be the proximate mediators of neurotoxicity []. One report indicated that Aβ levels were decreased in the brains of Pin1-deficient mice [], while another report described the opposite effect, i.e., that Pin1 deficiency increased Aβ generation []. In work unrelated to Pin1, Tsai et al proposed a model wherein the phosphorylation state of APP Thr668 would be pivotal in modulating the subcellular distribution of APP and generation of Aβ [] and/or the amyloidogenic γ-secretase cleavage of APP CTFs []. The potential importance of APP Thr668 phosphorylation was further emphasized in a recent review []. Because all these reports and reviews hinged on attribution of some biological significance to the phosphorylation state of APP Thr668, we investigated in vivo brain APP metabolism and Aβ levels in mutant mice generated by knocking into their genome an APP gene containing a non-phosphorylatable alanyl substitution at position . Here we report characterization of these mice, including the quantification and subcellular distribution of all standard APP metabolites.ResultsGeneration of threonine- phosphorylation-site mutant miceWe constructed targeting vectors in which an alanine (A) was substituted for threonine (T) at position in exon (Fig. 1A). Targeted C57BL/-derived MS12 embryonic stem (ES) cells were injected into Balb/c blastocysts. Chimeras were bred to C57BL/ mice to generate heterozygotes. Fertilized eggs from the heterozygotes were then injected with a Cre-expression plasmid vector to delete the Pgk-neo gene cassette from the germ line via the Cre-loxP system [], and resulting progeny carrying a substituted allele without the Pgk-neo gene cassette were backcrossed to C57BL/ mice for at least generations, and then intercrossing was performed to produce Thr668Ala homozygous mice. Thus, the mutant lines used in this study were coisogenic to the C57BL/ strain. Seven generations of crossing is considered acceptable for mice generated by a coisogenic strategy []. The success of the mutagenesis was confirmed by sequencing (Fig. 1B), Southern blot, and polymerase chain reaction (PCR) analysis (Fig. 1C). We also confirmed the mutation by immunoblot analysis using anti-phospho-APP Thr668 specific antibody (Fig. 1D). There was no anti-phospho-APP Thr668 antibody staining in homozygotes (A/A), while heterozygotes (T/A) had approximately half the level of the staining that was observed in wild-type (T/T). Anti-pan APP-specific antibody staining revealed unaltered expression levels of APP in all genotypes (Fig. 1D)../journal.pone..g001Figure 1Generation of Thr668Ala Knock-in Mutant Mice.(A) The targeting vector (a1), a partial map of the APP gene (a2), the resultant targeted allele (a3), and the knock-in allele after Cre recombination (a4) are illustrated.Filled boxes denote coding sequences of exons and .Shaded parts in exon18 correspond to the ′ non-coding region.A substitution is represented by a dot in exon .B, BamHI; X, XhoI.Probes for Southern blot analysis for screening of targeted ES clones are indicated with small bars in (a2–).PCR primers (E1 and E2) for genotyping of mice are indicated with small arrows in (a2 and a4).(B) Verified sequences from wild-type (T/T), and Thr668Ala mutation homozygotes (A/A).The Thr668 flanking genomic region was amplified using mouse tail DNA as the template by PCR and sequenced.(C) Southern blot analysis for targeted ES cells and PCR analysis for genotyping wild-type and knock-in mouse lines.DNA from G418-selected ES cells was digested with XhoI and analyzed by Southern blotting with a ′ external probe.The -kb and -kb fragments represent wild-type and targeted alleles, respectively. PCR fragments of -bp and -bp represent wild-type and knock-in alleles, respectively.(D) Immunoblot analysis of mouse whole brain.APP was immunoprecipitated with anti-pan APP polyclonal antibody followed by immunoblotting with anti-pan APP antibody UT- or anti-phospho-Thr668-specific antibody UT- [].APP from homozygotes has no immunoreactivity with UT- although the APP expression levels detected by UT- were indistinguishable from those of wild-type mice.Mature APP (mAPP; N- and O-glycosylated form), immature APP (imAPP; N-glycosylated form), and phosphorylated APP (pAPP) are indicated with arrows.Thr668Ala mutation did not affect brain development or structureThe A/A mice had no gross cytoarchitectural abnormalities as revealed in Nissl-stained sections of the hippocampal region from A/A mice (Fig. 2A). Other brain regions were also histologically normal (data not shown). Furthermore, immunohistochemical and/or immunoblotting analyses for synaptophysin (presynaptic marker), microtubule-associated protein (MAP2; dendritic marker), glial fibrillary acidic protein (GFAP; astroglial marker), X11L (APP binding protein), and PSD95 (postsynaptic marker) revealed no abnormalities in A/A mice, even at ages of mo or older (Fig. 2B and C). These results contrast to those from APP-null mutant mice that showed marked decreases of synaptophysin, synapsin, and MAP- [], a reduction in dendrite length [], and gliosis []. The normal phenotype of APP Thr668Ala mice suggests that this mutation does not cause a major loss-of-function of APP../journal.pone..g002Figure 2Normal brain structure and expression and distribution of proteins related to neuron and glia in Thr668Ala mutant mice. (A) Nissl-stained hippocampal sections show no difference between wild-type (T/T) and Thr668Ala mutation homozygotes (A/A).Scale bars represent µm.(B) Immunohistochemical analysis in CA1 hippocampal region of aged (> mo-old) mice.Immunostaining for GFAP (astroglial marker), synaptophysin (presynaptic marker), and MAP- (neuronal dendritic marker) in hippocampal CA1 region of wild-type (T/T) and Thr668Ala mutant (A/A) mice are shown.S.o., stratum oriens; Py, pyramidal cell; Rad, stratum radiatum. Scale bars represent µm.(C) Western blot analysis of APP, X11L, MAP2, synaptophysin, PSD95 (postsynaptic marker), and GFAP from the brains of mo-old wild-type (T/T) and Thr668Ala mutant (A/A) mice.Analysis of APP and its metabolites in APP Thr668Ala knock-in miceThe brains of APP Thr668Ala knock-in mice were analyzed for expression of APP and for steady-state levels of its metabolites (Figs. and ; Table ). APP Thr668 phospho-state specific antibodies were used to establish that the knocked-in protein was not phosphorylated at Thr668 (Fig. and Fig. 4A). Amino-terminal fragments of APP cleaved by α- (sAPPα) and β- (sAPPβ) secretases were quantified and compared with those of wild-type mouse brain (Fig. ). No differences of total amount of sAPP, sAPPα, and sAPPβ were detected between Thr668Ala knock-in and wild-type mice../journal.pone..g003Figure 3APP and sAPP in wild-type and Thr668Ala mutant mouse brain.(Left) APP and sAPP in wild-type (W) and mutant (M) mouse brain. Homogenates of brains taken from -month-old mice were fractionated as described and analyzed by immunoblotting for sAPP and APP holoprotein (22C11), APP (APP/C), phosphorylated APP (pThr668APP), sAPPα, sAPPβ and MAP2B.Protein bands observed in the insoluble fraction (ppt panel) of sAPPβ are non-specific.(Right) sAPP and APP levels are displayed.The densities of the bands from soluble APP (sAPP) were standardized to the densities of MAP2B, and those from APP were standardized to the densities of actin.All were normalized to unity for wild-type mice (.). The bars indicate means±S.D. (N.S.; n = )../journal.pone..g004Figure 4Aging-dependent phosphorylation of APP and APP CTFs and quantification of CTFs in wild-type and Thr668Ala mutant mice brain.(A) APP phosphorylation state in brains of post-natal day (P0), young adult (-month), and aged adult (-month) mice.The upper panel was probed with anti-pan-APP C-terminal antibody G369, and the lower panel was probed with an anti-phospho-threonine--specific antibody.W, wild-type mouse; M, Thr668Ala mutant mouse.(B) APP carboxyl-terminal fragments (APP CTFs) in wild-type (W) and mutant (M) mouse brain.C99 and C89 are products resulting from cleavage of APP by BACE, while C83 results from cleavage of APP by ADAM-/-.PhosphoC99 (pC99), phosphoC89 (pC89), and phosphoC83 (pC83) are all mono-phosphorylated at Thr668, and these peptides are numbered here according to standard APP695 nomenclature.(C) Expression levels of CTFα and CTFβ in middle aged wild-type and Thr668A mutant mouse brain.Various species of CTF are schematically represented at the left and indicated at right with bars.Samples were electrophoresed after treatment with either buffer or λ phosphatase (λ PPase).The amounts were normalized to unity for wild-type mice (.).The bars indicate means±S.D. (N.S.; n = )../journal.pone..t001Table 1Table . Aβ40 and Aβ42 levels in brains from -month-old wild-type and Thr668Ala mutant mice.Aβ40 (pM)p valueAβ42 (pM)p value pM Wildtype ±.±. Mutant ±.76p>.±.84p>.72Results of ELISA for murine Aβ [] are expressed as means±S.E.M. (N.S.; student's t) and are representative of results obtained from three independent assays, using five individual mice of each genotype in each assay.Next, we examined protein levels of APP C-terminal fragments (CTFs) by immunoblot analysis using anti-APP cytoplasmic tail antibodies (Fig. ). The phosphorylation levels of APP and CTFs are relatively lower at the birth (P0), they increase during growth (-mo), and they slightly decrease during aging (-mo). Cleavage by α-secretase generates CTFα, which is composed of the carboxyl-terminal amino acids of APP and is also known as C83. β-secretase can cleave APP at either the peptide bond N-terminal to D1 (preferred) or at the bond N-terminal to E11; hence, CTFβ and CTFβ′ fragments, C99 and C89 [], respectively, are generated, with the preponderant species being CTFβ′/C89 []. Because these CTFs are phosphorylated at Thr668 (Fig. 4B) and generate complicated patterns [], [], we treated samples with lambda protein phosphatase (λPPase) in order to identify and quantify phospho- and dephospho-CTF species precisely (Fig. 4C). Treatment of immunoprecipitates containing these CTFs with λPPase resulted in the appearance of three discrete bands corresponding to dephospho-forms of C99, C89, and C83, all of which were present at identical levels in both wild-type and Thr668Ala knock-in mice.Endogenous murine Aβ40 and Aβ42 levels were also detected in brain. Rodent Aβ is much less prone to aggregation than is human Aβ [], and therefore little, if any, insoluble Aβ is typically detected in the wild-type mouse brain. Nevertheless, we extracted mouse brain Aβ with M guanidine chloride/TBS (the standard protocol for dissolving insoluble Aβ), and we quantified the solubilized Aβ. The levels of Aβ40 and Aβ42 in the brains of wild-type and Thr668Ala knock-in mice were indistinguishable (Table ).We fractionated mouse brains using an iodixanol gradient fractionation system identical to that described in the published cell culture study [] (Fig. ). Using homogenates of wild-type mouse brains, we observed that total mAPP as well as the corresponding phosphorylated form were largely recovered in fractions –, where GM130 (Golgi marker protein, concentrated in fractions –) and EEA1 (early endosomal protein, concentrated in fractions –) were co-distributed (upper panel). Total Thr668Ala mAPP in knock-in mutant mouse brains was identical in levels and distribution except that, as expected, there was no detectable phospho-mAPP (Fig. ). These data do not support the prediction [] that the phosphorylation state of APP at Thr668 modulates its subcellular distribution in brain in vivo../journal.pone..g005Figure 5Hemi-brains from wild-type (upper panel of six blots, labeled “Wild-type”, far left) or mutant (lower panel of six blots; labeled “Mutant T668A”, far left) -month old mice were used for fractionation on iodixanol density-gradients as described ().Equal aliquots (according to volume) were analyzed by immunoblotting with antibodies as specified: anti-early endosome antigen (Transduction laboratories, EEA1, top panel), anti-cis-Golgi matrix protein (Transduction laboratories, GM130, second panel), anti-protein disulfide isomerase (Stressgen, PDI, a marker for the endoplasmic reticulum, third panel), anti-post synaptic density (Transduction laboratories, PSD95, a synaptic membrane marker, fourth panel), anti-pan-APP (G369, fifth panel) and anti-phospho-threonine668 APP (Cell Signaling, bottom panel).DiscussionGeneration, aggregation, and deposition of Aβ are key steps in the pathogenesis of AD []. Aβ is generated in the process of intracellular trafficking of APP, which is type I membrane protein. Following biosynthesis of APP, APP enters the endoplasmic reticulum where the protein is subjected to N-glycosylation (imAPP) and then transported to Golgi and subjected to O-glycosylation (mAPP). In neurons, membrane vesicles containing mAPP may be transported to nerve terminals along the axon using a kinesin-dependent motor system or else may be transcytosed and localized in dendrites []. At least some APP is exposed on the plasma membrane. This cell-surface APP, as well as APP in the trans Golgi network, is conveyed to the endocytic system via clathrin-coated vesicles []. Thus, APP might possibly be cleaved in one or several of compartments; i.e., the constitutive secretory pathway, the plasma membrane, and/or the endocytic system.The short cytoplasmic domain of APP plays an important role in the regulation of intracellular APP trafficking and contains several functional motifs such as -GYENPTY-, where several regulatory proteins interact []. One motif in the cytoplasmic domain of APP that has been proposed to be functionally important is -VTPEER- which forms a type I β-turn and amino-terminal helix capping box structure and contributes to the stability of a carboxyl-terminal helix structure [], []. The phosphorylation of Thr668, situated within this motif, induces significant conformational change in the cytoplasmic domain, changing its interaction with FE65 [], []. Because of these structural phenomena, phosphorylation of APP at Thr668 has been proposed by various groups to control either the metabolism of APP, some APP-related physiological function(s), and/or some FE65-mediated events.Some biological phenomena appear to correlate with APP phosphorylation. Neurite outgrowth in PC12 cells has been reported to correlate with the phosphorylation of APP, and a Thr668Glu substitution remarkably reduced neurite extension response following treatment with NGF []. Membrane proteins, including APP, act to tether FE65 to Golgi membranes, and phosphorylation of APP releases FE65 from its membrane-protein anchor [], []. The released FE65 may translocate into the nucleus and activate gene expression [].However, the role for the phosphorylation state of APP at Thr668 in the regulation of APP metabolism in brain in vivo has not been clarified. As mentioned above, Tsai and colleagues reported that amino acid substitutions at Thr668 alter the intracellular distribution of APP in cultured cells []. Alternatively, a recent model from Lu and colleagues holds that the prolyl isomerase Pin1 interacts with the phosphorylated form of APP at Thr668 , and that the interaction of Pin1 with APP Thr668 has important effects on Aβ levels [; for review, see ]. Aβ levels in Pin1-deficient brains have been studied not only by Lu and colleagues [], but they have also been studied independently by Akiyama and colleagues []. These groups reported directly contradictory results, however, with one group reporting that Aβ levels were decreased in the brains of Pin1-deficient mice, while the other reported that Aβ levels were increased in the brains of Pin -deficient mice [], []. Thus, since four recent and highly visible papers [, ,; rev in ] have emphasized the potential importance of APP Thr668 phosphorylation state in controlling brain Aβ levels, we investigated brain APP metabolism and Aβ levels in mutant mice generated by knocking into the mouse genome an APP gene containing a non-phosphorylatable alanine substitution at position of APP.The levels and subcellular distribution of APP and all its metabolites (sAPPα, sAPPβ, CTFα, CTFβ, Aβ40, Aβ42) from the brains of wild-type and APP Thr668Ala mutant mice were indistinguishable. Hence, contrary to widely publicized models [, , , rev in ], the phosphorylation state of APP Thr668 does not play an obvious role in governing physiological levels of brain Aβ in vivo.Of note, during the final stages of production of this manuscript, a report appeared from Cruz et al [] in which a revised model is presented wherein changes in APP metabolism caused by cdk5 are attributed to increased levels of β-APP-site cleaving enzyme (BACE) in contrast to this group's previous proposal [] that the state of phosphorylation of Thr668 was the most important modulator of Aβ levels and subcellular distribution of APP and its derivatives. This revised model of Tsai and colleagues, focusing on BACE levels rather than APP phospho-Thr668 levels [], is entirely consistent with what we have reported here.It is worth noting that we cannot, of course, rule out the possibility that pathological changes in APP Thr668 phosphorylation state might modulate its function or metabolism. Presently, we favor a model in which APP phosphorylation at Thr668 may regulate either some important physiological function of APP that is not directly linked to the proteolytic processing pathways that control Aβ levels in brain neurons in vivo. Evidence already exists that such functions might well include modulation of neurite outgrowth [], [] and/or intracellular signaling via AICD or FE65 [], [].Materials and MethodsProduction of mutant miceMutant mice were generated by a standard gene knock-in method using MS12 ES cell line derived from mice of the C57BL/ background []. Briefly, a -kb genomic fragment from C57BL/ mice containing exons and was used for constructing targeting vectors. The T668A point mutation (see Figure 1B) was introduced by PCR mutagenesis. For positive selection, the Pgk-neo gene cassette flanked by loxP sites was inserted at the EcoR1 site in the ′ non-coding region of APP-exon . For negative selection, the diphtheria toxin A-fragment gene cassette derived from pMC1DT-A was added to the ′ end of the targeting vector. After transfection of MS12 ES cells derived from the C57BL/ mouse strain and selection with G418, targeted clones were identified by Southern blot analyses. Chimeras, generated by injection of the targeted MS12 ES cells into Balb/c blastocysts, were mated with C57BL/ mice to obtain mutant heterozygotes. The Pgk-neo cassette flanked by loxP sites was excised by injecting the Cre recombinase expression vector, pCAGGS-Cre [] into the pronucleus of heterozygous fertilized eggs. The mouse genotypes were determined by PCR, using tail DNA as the template and the primers (E1: ′-CACATTGATTTCTTTGTGCCTG-′ and E2; ′-TCTGTACAATCATCCTGCAG-′), which anneal to the loxP flanking regions. Fragments of bp and bp were amplified from wild-type and knock-in mutant alleles, respectively.Immunoblot analysesBrains were homogenized in radioimmunoprecipitation (RIPA) lysis buffer containing µM microcystin-LR, µg/ml chymostatin, and µg/ml leupeptin, and the clear supernatants were used for immunoblot analysis. Aliquots of the lysates ( µg protein) were separated on SDS-PAGE (% (w/v) polyacrylamide for APP, X11L, MAP2 and PSD95, or .% (w/v) polyacrylamide for synaptophysin and GFAP, and analyzed by immunoblotting with anti-APP cytosolic domain-specific polyclonal antibody (pAb) G369, anti-phospho Thr668 (pAPP)-specific pAb (Cell Signaling Technology, MA), anti-X11L cytoplasmic domain (-AMFRLLTGQETPLYI- of human X11L)–specific monoclonal antibody (mAb), anti-MAP2 mAb HM- (Sigma, St. Louis, MO), anti-synaptophysin mAb SVP- (Sigma), anti-PSD95 mAb Clone16 (Transduction Laboratories), anti-actin MAB1501 (Chemicon International Inc., CA) and anti-GFAP mAb GF12. (Progen, Germany), respectively. Immunoreactive proteins were visualized with an enhanced chemiluminescence system (ECL; Amersham Pharmacia Biotech, Uppsala, Sweden).APP CTFs were detected by immunoblotting as described [], with modifications as described. Equal aliquots of homogenates ( µg) were separated by electrophoresis in a Tris-tricine gel (% [w/v] polyacrylamide). The separated proteins were transferred to nitrocellulose membranes, the membranes were boiled in PBS for min, and probed with anti-APP cytoplasmic domain antibody (Sigma). Immunoreactive proteins were again visualized with an ECL detection system (Amersham Pharmacia Biotech). For sAPP detection, brains of mice of various ages (∼ month-old and month-old) were homogenized in TBS buffer ( mM Tris-HCl, mM NaCl containing µM microcystin-LR, µg/mL chymostatin, and µg/mL leupeptin) and centrifuged at ,× g for min at °C. The pellets were subjected to two additional cycles of resuspension in equal volumes of TBS and centrifugation for min at ,× g at °C. Samples were then lysed in an equal volume of × RIPA buffer and sonicated. Protein concentrations were quantified using a BCA protein assay kit (Pierce). Aliquots containing µg protein were separated in % (w/v) Tris-glycine SDS-polyacrylamide gels, transferred to nitrocellulose, probed with anti-sAPPβ (kindly provided by T. C. Saido), or anti-sAPPα antibody 2B3 (IBL, Takasaki, Japan) and immunoreactive protein species were detected using an ECL detection system.ImmunocytochemistryAged mice ( mo-old or older) were deeply anesthetized with Avertin and transcardially perfused with physiologic saline and then % (w/v) paraformaldehyde in . M Na-phosphate buffer pH7. at °C for min. The brain was excised and post-fixed with the same fixative at °C overnight and cryoprotected in % (w/v) sucrose/phosphate-buffered saline (PBS). The brains were embedded into OCT compound (Sakura Fine Technical Co. Ltd., Tokyo, Japan) and frozen coronal sections ( µm) were prepared. Immunohistochemical staining was performed using the ABC method (Vector Laboratories, Burlingame, CA). Sections were incubated with .% (v/v) Triton X- in PBS for min at room temperature followed by incubation with .% (v/v) hydrogen peroxide in PBS for min at room temperature to quench endogenous peroxidase activity. The sections were then blocked with % (v/v) horse serum in PBS at °C overnight. After an overnight incubation at °C with anti-GFAP (clone GF12.), anti-synaptophysin (DAKO Corp., CA; clone SY38), or anti-MAP2 (Chemicon International, Inc., CA; clone MAB378) mAbs, the sections were further incubated with horse anti-mouse IgG conjugated to biotin (Vector Laboratories) for h at room temperature, followed by the ABC complex. The peroxidase activity was revealed using diaminobenzidine as the chromogen.Quantification of AβEndogenous mouse brain Aβ40 and Aβ42 were quantified with the ELISA system developed by Mathews and colleagues as described [].
|
PMC2375061.txt
|
TITLE: CALENDAR AUTHORS: No authors listed ABSTRACT: No Abstract BODY: No Body Content
|
PMC2011273.txt
|
TITLE: Epidemiology of melanoma of the eye in the Oxford Region, -. AUTHORS: A. J. Swerdlow ABSTRACT: No Abstract BODY:
|
PMC545208.txt
|
TITLE: Painful Horner Syndrome as a Harbinger of Silent Carotid Dissection AUTHORS: Amit Nautiyal, Sonal Singh, Michael DiSalle, John O'Sullivan ABSTRACT: A painful Horner's syndrome should alert clinicians to the possibilty of a silent carotid dissection BODY: PRESENTATION of CASEA -y-old white female presented to the hospital in July with pain in the left eye and left upper lid ptosis. She did not perceive any difference in perspiration between the two halves of her face. She was a nonsmoker and denied any history of head or neck trauma, or ocular, cardiac, vascular, or neurologic disease. Neuro-ophthalmological examination was normal except for mm of left upper eyelid ptosis (drooping of the eyelid), miosis (constriction of the pupil), and mild enophthalmos (recession of the eyeball into the orbit) consistent with classic left-sided Horner syndrome (Figure ). There was no carotidynia (a neck pain syndrome associated with tenderness to palpation over the carotid bifurcation) or carotid bruit. A chest radiograph obtained to rule out an underlying left apical superior sulcus tumor was normal. Magnetic resonance imaging/magnetic resonance angiography of the brain with cross-sectional imaging of the neck was obtained, which revealed extracranial left internal carotid artery dissection (Figures and ). The patient was treated with unfractionated heparin and coumadin and made an uneventful recovery. The patient was seen in the clinic a few months later and did not have any complications at follow-up.Figure 1Photograph of Patient Showing Left-Sided Horner SyndromeFigure 2Magnetic Resonance Imaging/Magnetic Resonance Angiography of the Neck Showing Left Internal Carotid Artery DissectionFigure 3T2-Weighted Magnetic Resonance Imaging Showing Blood in the Arterial Wall and Narrowing of the Lumen of the Left Internal Carotid ArteryThis is also known as the “crescent sign,” a hallmark of internal carotid artery dissection.DISCUSSIONHorner syndrome—characterized by the constellation of miosis, ptosis, anhidrosis (lack of sweating), enophthalmos, and anisocoria (unequal pupil size)—is present in up to % of internal carotid artery dissections []. Most patients experience neck, facial, and head pain ipsilateral to the lesion because of ischemia or stretching of the trigeminal pain fibers surrounding the carotid arteries []. Ophthalmic manifestations have been reported to occur in up to % of patients with internal carotid artery dissection []. Common findings in descending order of frequency are painful partial Horner syndrome (due to disruption of the third-order neuron oculosympathetic fibers) as seen in our patient, transient monocular vision loss, and permanent visual loss [].De Bray et al. studied the prognosis of cases of isolated Horner syndrome due to internal carotid artery dissection []. They found that % of cases of Horner syndrome due to internal carotid artery dissection were painful. The risk of an early ischemic stroke within the first wk was high (around %) without initial antithrombotic treatment [].Internal carotid artery dissection is a potentially life-threatening condition and carries a substantial risk of disabling stroke []. Carotid dissection is under-recognized as a cause of Horner syndrome and can be missed []. It is important to diagnose dissection because anticoagulation can prevent carotid thrombosis and embolism []. The investigation of choice is magnetic resonance imaging and angiography scan of the head and neck [].The treatment advocated for dissection is anticoagulation for – mo [].Learning PointsPainful Horner syndrome should alert clinicians to the possibility of a silent carotid dissection until proven otherwise [].Magnetic resonance imaging and angiography scan of the head and neck is the imaging modality of choice to look for dissection [].For patients with carotid dissection, anticoagulation with warfarin and coumadin is recommended for – mo to prevent carotid thrombosis and embolism [].
|
PMC1968596.txt
|
TITLE: Classical disseminated Kaposi's sarcoma in HIV-negative patients; an unusually indolent subtype. AUTHORS: I. G. Ron, A. Kuten, N. Wigler, G. Fried, S. Nitezky, M. J. Inbar, J. Dale, S. Chaitchik ABSTRACT: Kaposi's sarcoma is a rare neoplasm of characteristic chronicity. The classical form which occurs most often in elderly men of Eastern European origin, comprises both an indolent, cutaneous type marked by spontaneous regression with prolonged survival, and a rarer, disseminated variant is more fulminant. Seven elderly Jewish patients with classical, disseminated, visceral Kaposi's sarcoma were studied; they were neither homosexual nor drug-abusers. All immunologic parameters were normal and serum tests for HIV antibodies, CMV, and EBV were negative. Five of these patients were treated and four responded well, including two complete remissions. The prolonged survival of these patients (% at years) suggests the existence of an indolent subtype or forme fruste of the usually aggressive form of classical Kaposi's sarcoma. BODY:
|
PMC2241642.txt
|
TITLE: Yarrowia lipolytica vesicle-mediated protein transport pathways AUTHORS: Dominique Swennen, Jean-Marie Beckerich ABSTRACT: BackgroundProtein secretion is a universal cellular process involving vesicles which bud and fuse between organelles to bring proteins to their final destination. Vesicle budding is mediated by protein coats; vesicle targeting and fusion depend on Rab GTPase, tethering factors and SNARE complexes. The Génolevures II sequencing project made available entire genome sequences of four hemiascomycetous yeasts, Yarrowia lipolytica, Debaryomyces hansenii, Kluyveromyces lactis and Candida glabrata. Y. lipolytica is a dimorphic yeast and has good capacities to secrete proteins. The translocation of nascent protein through the endoplasmic reticulum membrane was well studied in Y. lipolytica and is largely co-translational as in the mammalian protein secretion pathway.ResultsWe identified S. cerevisiae proteins involved in vesicular secretion and these protein sequences were used for the BLAST searches against Génolevures protein database (Y. lipolytica, C. glabrata, K. lactis and D. hansenii). These proteins are well conserved between these yeasts and Saccharomyces cerevisiae. We note several specificities of Y. lipolytica which may be related to its good protein secretion capacities and to its dimorphic aspect. An expansion of the Y. lipolytica Rab protein family was observed with autoBLAST and the Rab2- and Rab4-related members were identified with BLAST against NCBI protein database. An expansion of this family is also found in filamentous fungi and may reflect the greater complexity of the Y. lipolytica secretion pathway. The Rab4p-related protein may play a role in membrane recycling as rab4 deleted strain shows a modification of colony morphology, dimorphic transition and permeability. Similarly, we find three copies of the gene (SSO) encoding the plasma membrane SNARE protein. Quantification of the percentages of proteins with the greatest homology between S. cerevisiae, Y. lipolytica and animal homologues involved in vesicular transport shows that % of Y. lipolytica proteins are closer to animal ones, whereas they are only % in the case of S. cerevisiae.ConclusionThese results provide further support for the idea, previously noted about the endoplasmic reticulum translocation pathway, that Y. lipolytica is more representative of vesicular secretion of animals and other fungi than is S. cerevisiae. BODY: BackgroundYarrowia lipolytica is a hemiascomycetous dimorphic yeast, generally regarded as safe (GRAS), which has been used for biotechnological applications. It is able to produce large amounts of several metabolites such as citric acid and to secrete a variety of extracellular proteins (alkaline or acid proteases, RNase, lipases etc.) []. Its good protein secretion capacities have allowed the engineering of powerful heterologous protein expression systems [reviewed in []]. Y. lipolytica is also a conveniently tractable model organism, of which the secretion pathway was studied for several years in our laboratory []. We focused on the early steps of protein translocation in the endoplasmic reticulum [-], on the quality control of protein folding [] and on the glycosylation pathway []. Several genes involved in these steps were cloned and analysed.The results of the Génolevures II sequencing project of four hemiascomycetous yeasts [] allowed us to search for proteins involved in the secretion pathway of Y. lipolytica and we compared them to the proteins of the three other yeasts, Candida glabrata, Kluyveromyces lactis and Debaryomyces hansenii. C. glabrata has become the second most common cause of candidiasis after Candida albicans. C. glabrata is not dimorphic, in contrast to other Candida species, and is phylogenetically closer to Saccharomyces cerevisiae []. K. lactis is less closely related to S. cerevisiae and has the capacity to grow on lactose as a sole carbon source, it has been used for industrial applications [,]. D. hansenii is a cryotolerant marine yeast which grows at salinities up to %. D. hansenii is the most common yeast found in cheese and provides proteolytic and lipolytic activities during cheese ripening []. In this work, we first established the list of proteins, predicted from whole genome analysis, which are potentially involved in vesicular transport in Y. lipolytica. Candidates were identified through BLAST searches against S. cerevisiae protein sequences. We then search for homologues of these proteins in the predicted protein set encoded by the three other genomes. Among the differences observed, we noticed a number of plasma membrane SNARE proteins (three Ssop) in Y. lipolytica compared to the four other yeasts. S. cerevisiae and C. glabrata have two SSO genes whilst in K. lactis and D. hansenii, we detected only one gene. We finally focused on one specific feature of the Y. lipolytica secretory pathway, namely the existence of a Rab4-related protein. In mammalian cells, the GTP binding protein Rab4p is involved in the regulation of plasma membrane protein recycling []. A Rab4-related protein is also found in Schizosaccharomyces pombe and in filamentous fungi such as Neurospora crassa, Aspergillus fumigatus or Phaenerochaete chrysosporium but is absent from S. cerevisiae, Candida albicans [] and the three other hemiascomycetous yeasts. We constructed a strain of Y. lipolytica deleted for the RAB4 gene and analysed its phenotypic pattern.Results and discussionVesicle-mediated protein transport pathwaysThe only membrane that a secretory protein must traverse is the membrane of the endoplasmic reticulum, the transport of the protein to its final destination continues through vesicles which bud and fuse between organelles [for reviews: [,]]. Vesicle budding is mediated by protein coats; vesicle targeting and fusion depend on Rab GTPase, tethering factors and SNARE complexes.Vesicle buddingProtein coats (see Additional file )Endoplasmic reticulum to Golgi transportEndoplasmic reticulum to Golgi transport is mediated by the action respectively of the COPII and COPI coat complexes [[-] for reviews]. The COPII coat is assembled on the endoplasmic reticulum membrane and allows cargo selection and membrane budding []. The COPI complex is involved in retrieval of recycled proteins back to the endoplasmic reticulum []. COP I subunits could also have a role in vacuolar sorting [].COPII coat vesicles (see Additional file -)Vesicle budding is initiated by the activation of the GTPase Sar1p by the endoplasmic reticulum integral membrane guanine exchange factor Sec12p [,]. Sar1p initiates membrane curvature [[], for mammalian Sar1p see []]. The membrane-bound Sar1p-GTP recruits the heterodimer complex Sec23p-Sec24p. These pre-budding complexes are gathered by the Sec13p-Sec31p complex into nascent vesicles [see []: the mammalian Sec13p-Sec31p structure]. The Sec23p subunit activates the hydrolysis of GTP by Sar1p and reverses the assembly process. Sec16p stabilizes the coat against premature disassembly after Sar1p hydrolyses GTP []. Using fluorescence resonance energy transfer to monitor the assembly and disassembly of COPII coat, it was suggested that a kinetically stable prebudding complex was maintained during multiple Sar1p GTPase cycles []. In S. cerevisiae, there are one Sec23p, one Sec23p-related protein, one Sec24p, two Sec24p-related proteins (Sfb2p and Sfb3p) and one Sec12p homologue (Sed4p). In Y. lipolytica and D. hansenii, all the COP II coat components are well conserved and we find two Sec23p-homologues, two Sec24p-homologues but no Sed4p proteins. In C. glabrata, there are two Sec23p-homologues, three Sec24p-homologues and two Sec13p-homologues and in K. lactis we found the same proteins as in S. cerevisiae with the exception of Sfb2p and Sed4p.COP I coat vesicles (see Additional file -)The COP I coat assembles by the same process as COP II complex involving an Arfp-GTPase [for a review about Arf1p: []; mammalian Arf1p: []; mammalian COPI assembly review: []]. All the S. cerevisiae components are conserved in Y. lipolytica, though the Y. lopolytica Sec28p is only weakly related to the S. cerevisiae protein. In K. lactis, the Arf1 protein homologue was not identified but another Arf protein could play the role of Arf1p (see Additional file -).Post-Golgi transportIn the trans-Golgi network, the proteins are sorted to the plasma membrane, the endosomal/vacuolar system or recycled back from the endosome. Coated vesicle adaptors facilitate cargo selection [for reviews: [,]].Adaptor protein complex (see Additional file -)In S. cerevisiae, by homology to the mammalian adaptor protein (AP) subunit sequences, three potential heterotetrameric adaptor protein complexes have been identified []. Each complex is composed of two large (Aplp), one medium (Apmp) and one small (Apsp) subunits. The AP- complex is associated with clathrin-coated vesicles and is involved in retention of late Golgi membrane proteins [] and trafficking to the vacuole []. This complex is alone able to associate with clathrin [,]. Unlike the mammalian AP- complex which associates with endocytic clathrin-coated vesicles, the AP- complex of S. cerevisiae is apparently not involved in endocytosis. The AP- complex is involved in independent clathrin-coated vesicle transport of membrane proteins from Golgi to vacuoles []. In Y. lipolytica, we also identified three potential AP complexes. As in S. cerevisiae, two AP- medium subunits were found, but only two small subunits could be identified which could correspond to the AP- and AP- small subunits. The three other yeasts have the same set of proteins as S. cerevisiae for their adaptor protein complexes.GGA proteins (Golgi-localized, γ ear-containing, ARF-binding proteins) (see Additional file -)GGA proteins are implicated in Golgi to endosomes clathrin-coated vesicle transport and bind to ubiquitin to facilitate this sorting [,]. In S. cerevisiae, the GGA gene is duplicated but in the four yeasts studied, as in C. elegans and D. melanogater, there is only one gene: GGA2 which, evolutionary, is closer to the hypothetical common ancestor [].Synaptojanin-like protein (see Additional file -)The S. cerevisiae Inp53p, a synaptojanin-like protein acts together with the AP- complex in the Golgi to endosome clathrin-dependant pathway which is distinct from the direct Golgi to prevacuolar compartment mediated by GGA proteins [].Retromer complex (see Additional file -) and sorting nexins (see Additional file -)Sorting nexins play a role in retrieval of proteins from the prevacuolar compartment or post-Golgi endosomes and different nexins operate in different classes of endosomes [reviewed in [-]]. The sorting nexins Snx4p, Snx41p and Snx42p are required for the retrieval of the SNARE Snc1p from the post-Golgi endosome; Grd19p and the retromer complex are involved in the retrieval of endosomal SNARE Pep12p from the prevacuolar compartment []. The retromer complex consists of five proteins: Vps5p, Vps17p, two sorting nexins which form a dimer and associate with the complex formed by Vps26p, Vps29p and Vps35p [see [,] for mammalian retromer complex structure].Vesicle targeting and fusionRabGTPase (see Additional file and Additional file -)Rab proteins are small monomeric guanosine triphosphatase (GTPase) which are membrane-associated and cycle between an active GTP-bound state and an inactive GDP-bound protein. These switches regulate all the steps in the secretion pathway. Mammalian Rab proteins belong to the Ras superfamily of GTPase. All the members of this superfamily have conserved nucleotide, phosphate and magnesium binding sequences but the Rab sequences can be distinguish by their C-terminal prenylation site and five Rab-specific regions (RabF) []. Pereira-Leal and Seabra [] have also identified Rab subfamily specific regions (RabSF). They studied the evolution of the Rab family [] and by their analysis, they observed that Rab proteins co-segregating in the phylogenetic trees showed a pattern of similar cellular localisation and/or function. In S. cerevisiae, Ypt1p, Ypt31p/32p and Sec4p are the essential Rab GTPases which regulate the exocytic pathway and Ypt6p, Ypt7p and Ypt51p/52p/53p are involved in the endocytic pathway. S. cerevisiae also has two other Rab GTPase, Ypt10p and Ypt11p which are also present in C. glabrata and we can find Ypt11p in K. lactis. Ypt10p seems to be involved in endocytic function and Ypt11p is required for endoplasmic reticulum inheritance []. In C. glabrata, a Ypt53p homologue could not be identified and in K. lactis it is Ypt32p which was not found. In Y. lipolytica, we can find homologues of the nine S. cerevisiae proteins necessary for the secretion pathway. Ypt10p and Ypt11p are absent but we can find two other Rab-related proteins, Rab2p and Rab4p as in the filamentous fungi []. The analysis of the phylogenic tree (Fig. ) obtained after alignment of several human Rabp, S. cerevisiae and Y. lipolytica Yptp sequences revealed that human Rab1p, Rab2p, Rab4p, Rab5p and Rab11p cosegregate with Y. lipolytica proteins. As N. crassa [] and other filamentous fungi, Y. lipolytica has a large protein secretion capacity and is able to switch from a yeast life cycle to a filamentous form in response to environmental conditions; in this latter form it needs a better capacity of secretion and of recycling plasma membrane material. In mammalian cells, Rab2p has been proposed to regulate the retrograde transport between the Golgi and the endoplasmic reticulum [] and Rab4p is involved in the recycling of plasma membrane proteins [see [,] for a review about recycling pathways]. The comparison of the mouse, rat, human, N. crassa, Schizosaccharomyces pombe and Y. lipolytica sequences (Fig. and Additional file ) shows that the nucleotide, phosphate and magnesium binding regions, the RabF and RabSF sequences are well conserved. We also compared these Rab4p sequences with other proteins of the mammalian Ras superfamily (data not shown) and we identified a sequence GIQYG next to the RabSF4 region, and particularly the tyrosine residue only present in the Rab4p sequences. The analysis with the NetPhos program (CBS prediction server) identified this tyrosine as a potential phosphorylation site. We suggest that this tyrosine could be important in the regulation of Rab4p activity. In order to get more information about this filamentous fungi specificity, we analysed the effects of a deletion of the gene coding for the Y. lipolytica Rab4-related protein. This deletion showed only slight phenotypic changes. The aspect of the colony on rich medium plate was slightly different (Fig. ). The ability to make the dimorphic transition was not impaired but at OD600 of , we quantified that the percentage of cells undergoing a dimorphic transition for the wild type strain was % and % for the mutant strain and the cells in the yeast form had a more spherical appearance in the mutant strain than in the wild type and aggregated more readily (Fig. ). The round morphological aspect is also observed in the Y. lipolytica rac mutant, the Rac protein is another member of the Ras superfamily which is implicated in the induction of the hyphal growth []. The slight differences in the morphology of the mutant strains suggested a potential modification of the wall composition. This was confirmed by the increased sensitivity of the strain to calcofluor white (Fig. ), implicating an increase in chitin composition of the wall. We also observed a decrease in the sensitivity to SDS (Fig. ) suggesting a decrease in the porosity of the wall. These two events are also encountered when the genes coding for a heterotrimeric G-proteins of Aspergillus nidulans are mutated, these mutations in this filamentous fungus confer resistance to the antifungal plant PR- (Pathogenesis-Related) protein []. We suggest that the Y. lipolytica Rab4 protein could be important to recycle the receptor associated with a heterotrimeric G-protein. The mutant Y. lipolytica rab4 strains were able to produce diploids as well as the wild type strain (data not shown) indicating that the recycling of the pheromone receptor, associated with a G-protein, was not impaired. The Y. lipolytica Rab4p does not regulate endocytosis as the incorporation of FM4- was the same as the wild type strain compared to a sls2 mutant strain in which the FM4- incorporation is delayed (Fig. ). Y. lipolytica Sls2p is homologous to the S. cerevisiae Rcy1p which plays a role in the recycling pathway (see "The vesicle-SNARE Snc1p recycling" section below).Figure 1Phylogenetic tree of some human Rabp, S. cerevisiae and Y. lipolytica Yptp. "]" indicates when Y. lipolytica protein sequences are closer to human ones. The tree was obtained with ClustalX program, . version [] and presented with Treeview program, .. version [].Figure 2Ypt4p/Rab4p protein sequences alignment. The figure shows the upper quartile, for the full image, see Additional file . Mus musculus (Mm), Rattus norvegicus (Rn), Homo sapiens (Hs), Drosophila melanogaster (Dm), Neurospora crassa (Nc), Schizosaccharomyces pombe (Sp) and Yarrowia lipolityca (Yl) Ypt4/Rab4 protein sequences alignment was obtained with ClustalX program, . version [] and presented with GeneDoc program, .. version [].Figure 3Colony morphology of Y. lipolytica strains. Wild type (wt) and two independent clones of rab4Δ (-,-) strains were grown as isolated colonies on solid YPD rich medium. Observation (×) of a five days culture by binocular microscopy.Figure 4Disruption of Y. lipolytica RAB4 gene does not impair hyphal growth but affects dimorphic transition. Microscope observation of the wild type (wt) and the mutant (rab4Δ) strains in liquid rich YPD medium exponential growth (OD600:) and stationary phase (OD600:).Figure 5SDS and Calcofluor white sensitivity. The rab4Δ (-,-) mutant strains are more sensitive to calcofluor white (CW) and more resistant to SDS than the wild type (wt) strain.Figure 6Endocytosis in the rab4Δ mutant strain is not impaired. The incorporation of FM4- in the rab4Δ- mutant strain is the same as in the wild type (wt) strain compared to a sls2 mutant strain in which the FM4- incorporation is delayed. Y. lipolytica Sls2p is homologous to the S. cerevisiae Rcy1p which plays a role in the recycling pathway (see "The vesicle-SNARE Snc1p recycling" section). Low panel: Nomarski.Regulation of Rab-GTPase [for reviews see [-]]Rab proteins cycle between cytosolic inactive GDP-bound form and active membrane associated GTP-bound form. The cytosolic form exists in a complex with a GDP dissociation inhibitor (GDI). Post-translational prenylation of the protein is important for its activity and prenylated Rabp is recruited to the appropriate membrane by a GDI displacement factor (GDF) which catalyses the dissociation step. The nucleotide exchange is favoured by the guanine nucleotide exchange factor (GEF). GTP-bound Rabp is then activated and can interact with its effectors. The recycling of the Rab protein is stimulated by the GTPase activating protein (GAP) and the GDP-bound Rab protein is released from the membrane by GDI.Prenylation [[], see [] for mammalian prenylation] (see Additional file -)Most of Rab proteins contain two C-terminal cysteine residues which are isoprenylated with two geranylgeranyl moieties. This reaction is catalyzed by geranylgeranyl transferase II (GGTase II), this enzyme has two subunits, a third subunit, Rab escort protein (REP), is a chaperone.GDP dissociation inhibitor (GDI) (see Additional file -)Gdi1p recycles the Yptp/Sec4p proteins from their target membranes back to the vesicular pool [].GDI displacement factor (GDF) (see Additional file -)In vitro experiments with mammalian proteins identified that Yip3p catalyses the dissociation of endosomal Rab proteins from GDI []. The Yip family (Yip1p, Yip2p, Yip3p, Yip4p, Yip5p and Yif1p) are membrane proteins which interact with prenylated Rab proteins []. In S. cerevisiae, Yip1p has been identified through a two-hybrid screen as a protein interacting with Ypt1p and Ypt31p in their GDP form []. With a similar screen, Yif1p has been identified as a Yip1p-interacting protein []. These two proteins form an integral membrane complex that bind Ypt1p and is required for Golgi membrane fusion by interaction with the Golgi SNARE proteins []. Yos1p (Yip One Suppressor ) associates with Yip1p-Yif1p complex []. This protein was only identified in D. hansenii.Guanine exchange factor (GEF) (see Additional file -)The activation and membrane stabilisation of the Rab protein are accompanied by exchange of the GDP for the GTP, this activity being catalysed by the guanine exchange factor. Each GEF is specific for a Rab protein and seems to be recruited by the activated Rabp playing a role immediately upstream in the secretion pathway [,]. The TRAPP I protein complex binds the COP II vesicles and activates Ypt1p by guanine exchange []. TRAPP II Trs120p-Trs130p subunits join the TRAPP I complex to switch the GEF activity from Ypt1p to Ypt31p-Ypt32p acting in late Golgi [,]. Sec2p [see [] for the crystal structure and for the crystal structure of the Sec2p/Sec4p complex] is a highly efficient guanine exchange factor of Sec4p [], the Rabp essential for exocytosis [[], see [] for Sec4p regulation cascade, [] for Sec2p association with exocyst]. Vps9p is the Ypt51p GEF []. The Ric1/Rgp1p is the Ypt6p GEF [], the BLAST against the S. cerevisiae proteins showed only one subunit in Y. lipolytica, Rgp1p, but by comparison with the D. hansenii protein we could identify a potential Y. lipolytica Ric1p. And Vps39p is the Ypt7p GEF [].GTPase activating protein (GAP) (see Additional file -)The recycling of the Rab protein is favoured by the GTPase activating protein. In S. cerevisiae, eight GTPase activating proteins have been identified and are not specific for one Rab protein in in vitro experiments (Gyp1p, Gyp6p, Gyp7p: []; Gyp2p, Gyp3p, Gyp4p: []; Gyp5p, Gyp8p: []). In the four yeasts, the Gyp3p homologue, Gyp4p was not identified. Gyl1p is Gyp-like protein interacting with Gyp5p involved in the control of polarized exocytosis [], this protein has an homologue only in C. glabrata but not in the three other yeasts.Tethering factors (see Additional file ), [reviewed in [-]]The secretory vesicles are tethered to their target membrane by two classes of molecules: coiled-coil proteins able to form homodimeric complex as long as several times the diameter of the vesicle and large multisubunit complexes.Endoplasmic reticulum-cis-Golgi-networkSeveral factors are involved in the tethering of vesicles to the Golgi, TRAPP complex (see Additional file -), COG complex (see Additional file -) and Uso1p (see Additional file -). TRAPP is associated with the Golgi and two forms of the complex exist: TRAPP I ( subunits) acts in the endoplasmic reticulum to Golgi transport and TRAPP II which contains the TRAPP I subunits together with three other proteins acts in Golgi traffic. In Y. lipolytica only two TRAPP II specific subunits were identified by comparison with the S. cerevisiae protein sequences but for the Trs65p we used the protein identified in D. hansenii to detect a potential Y. lipolytica protein. Both complexes are able to interchange guanine nucleotide on Ypt1p. In vitro, TRAPP I can bind COPII vesicles by binding the coat Sec23p subunit [] and this could be the first event before interaction of the vesicle with its target []. The crystal structure of the mammalian Bet3p, the most conserved TRAPP protein, reveals a dimeric structure with hydrophobic channels and a covalent modification with a palmitate [,], the crystallographic study of the complex Bet3p-Trs33p reveals specific interactions between these subunits []. This subunit could be responsible for the targeting and the anchoring in the Golgi membrane and could direct the other TRAPP components to the Golgi [,]. Trs120p, a TRAPP II subunit, is required for vesicle traffic from the early endosome to the late Golgi []. Trs120p and Trs130p TRAPP II subunits are conserved from yeast to mammals; the Trs65p subunit is conserved only in some fungi and unicellular eukaryotes []. The other tethering factors, Uso1p, a long coiled-coil protein and the COG complex composed of eight subunits in S. cerevisiae are recruited before the last step of membrane fusion. Uso1p and the COG complex also have a function in sorting of endoplasmic reticulum-vesicles containing GPI-anchored proteins [see [] for a review about differential ER exit] and in retrograde vesicular trafficking within the Golgi [reviewed in []]. The Uso1 protein consists of an N-terminal globular head region, a coiled-coil tail which mediates dimerisation and a C-terminal acidic region. The NCBI Conserved Domain Architecture Retrieval Tool has identified in the Y. lipolytica Uso1 protein the first two domains but not the C-terminal acidic region. When we compare the consensus sequence of this region with the Y. lipolytica sequence, only the last residues of this domain are well conserved. The Saccharomyces cerecisiae and the mammalian COG complex are composed of eight subunits, a multiple of four subunits, as one the GARP and the exocyst complexes (see below), which could reflect an interaction with a four-component complex such as the trans SNARE complex []. In Y. lipolytica, only five subunits were identified as probable COG proteins by comparison with the S. cerevisiae proteins. Cog1p, Cog2p and Cog7p were not detected by BLAST searches but in mammals these proteins were identified by their function as their sequence similarity with the S. cerevisiae proteins is low []. Cog2p was found in D. hansenii and its sequence used to make the BLAST search with the Y. lipolytica proteins. This allowed the identification of a potential Y. lipolytica Cog2p homologue, but Cog1p and Cog7p were not found in Y. lipolytica, and in D. hansenii, Cog1p was not found. These proteins probably exist but should be identified by another means.Cis-Golgi-network-Endoplasmic reticulum (see Additional file -)Dsl1p complex is a large complex composed of the peripheral endoplasmic reticulum membrane proteins Dsl1p, Dsl3p (Sec39p) and Tip20p [,]. Dsl1p contains three domains, an N-terminal coiled-coil region of aminoacids which interacts with Tip20p, a central highly acidic region of interaction with Ret2p and Ret1p (two COP I subunits) and a conserved C-terminal sequence which could recruit cytoskeletal elements []. The amino acid N-terminus from the Dsl1p protein identified in Y. lipolytica does not align with the S. cerevisiae sequence. Nevertheless the Y. lipolytica sequence also contains potential coiled-coil regions (as determined by the coiled-coil prediction program, NPS@:Network Protein Sequence Analysis, []). The Tip20p sequence of Y. lipolytica has only % identities with the S. cerevisiae sequence. This could explain the divergent N-terminal sequence of Dsl1p which is involved in the interaction of the two proteins. Dsl1p, Tip20p and Dsl3p are required for the stability of the SNARE complex at the endoplasmic reticulum [].GolgiS. cerevisiae TRAPP II complex (see Additional file -) is composed of ten subunits and could have a role in retrograde transport of Golgi vesicles []. The trs130 mutant (coding for a TRAPP II subunit) displays synthetic interaction with mutation in a COPI subunit (Ret2p) and a deletion of ARF1 (see Additional file -) is implicated in COPI formation [].A role of the COG complex (see above) has also been found in retrograde transport to early Golgi vesicles [].The VFT/GARP complex (see Additional file -) localizes to the trans-Golgi network and is required for retrograde traffic from early endosomes to the Golgi [,]. The S. cerevisiae complex is composed of four subunits [,], it is the effector of Ypt6p and interacts with the SNARE Tlg1p. Only three subunits were identified in Y. lipolytica and D. hansenii but the undetected Vps51p unit also has no homologue in mammalian [] and seems to be a regulatory subunit [] which could be replaced by another protein as suggested by Liewen, et al. []. The structural analysis of the interaction between S. cerevisiae Tlg1p and Vps51p has determined an N-terminal peptide of Vps51p which deletion does not block transport to the late Golgi from endosomes [].Golgins (see Additional file -, -) [reviewed in [,]] are coiled-coil proteins which organize the structure and the trafficking pathways in the Golgi. These proteins have mainly been studied in mammalian cells but in Sacharomyces cerevisiae several homologues have been identified: Uso1p is the homologue of the mammalian p115 required for endoplasmic reticulum-to-Golgi transport; Grh1p, the GRASP65-homologue, a Golgi localized protein component of the spindle assembly checkpoint []; Imh1p, involved in transport between an endosomal compartment and the Golgi, Imh1p contains a Golgi-localization (GRIP) domain that interacts with activated Arl1p-GTP to be localized to the Golgi, this is regulated by Arl3p [[,]] and Arl3p requires the N-terminal acetyltransferase NatC complex and the protein Sys1p to be targeted to the Golgi [] [for a review about Arl proteins see []]; Coy1p, the CASP homologue, a Golgi membrane protein related to Giantin, its deletion in S. cerevisiae restores normal growth to cells lacking the SNARE Gos1p [] and Rud3p, a golgin--related protein, is a Golgi matrix protein that is involved in the structural organization of the cis-Golgi [].Golgi-Endosome, Endosome-Vacuole []The TRAPP II subunit Trs120p is required for vesicle traffic from early endosome to the late Golgi [].The Vps Class C/HOPS complex [reviewed in: [,]) (see Additional file -):The HOPS complex composed of Vps11p, Vps18p, Vps16p, Vps41p with the protein Vps19p are the effectors of Ypt51p in endosomal traffic. In vacuolar transport, the complex seems to recruit the Rab Ypt7p GEF Vps39p which activates Ypt7p. Activated Ypt7p acts on the HOPS complex to promote tethering and binding to the SNARE Vam3p through the interaction with the SNARE-binding protein, Vps33p. Vps33p together with other HOPS complex subunits is found in complex with Vps8p, a hydrophilic membrane-associated protein []. HOPS complex binds phosphoinositides and SNARE Vam7p [].Golgi-Plasma MembraneThe Exocyst (see Additional file -, -)The S. cerevisiae exocyst complex is composed of eight subunits (Sec3p, Sec5p, Sec6p, Sec8p, Sec10p, Sec15p, Exo70p and Exo84p), a quatrefoil complex as in the COG and the GARP complexes []. The activated Rab protein, Sec4p, present on the secretory vesicles, binds the exocyst subunit Sec15p in subcomplex with Sec10p resulting in the association with the other subunits and Sec3p []. Sec3p is the spatial landmark defining the sites of polarized exocytosis []. The localization of Sec3p is mediated by Rho GTPases, Rho1p [] and Cdc42p []. Rho3p plays a role in exocytosis through its interaction with Exo70p ([-]). These Rho proteins also have a role in actin polymerisation. Assembly of the exocyst occurs when the subcomplex associated with the vesicles joins Sec3p and Exo70p on the plasma membrane []. The Sec6p subunit dimerizes and interacts with the SNARE Sec9p, playing a role in SNARE complex regulation []. A cyclical regulatory network contributes to the establishment and maintenance of polarized cell growth []. Bem1p interacts with Sec15p and is involved in the Cdc42p-mediated polarity [].SNARE (Soluble N-ethylmaleimide-sensitive factor Attachment protein REceptor) proteins [see reviews: [,,-] for mammalian SNAREs] (see Additional file and Additional file -)After the tethering of the vesicle close to its target membrane, the fusion of the membranes is initiated through the action of SM (Sec1/Munc18, see below) and SNARE proteins. SNARE proteins share one conserved sequence called the SNARE motif which contains – amino acids that include heptads repeat typical of coiled coils. They contain a C-terminal transmembrane domain or a hydrophobic post-translational modification motif. SNARE proteins associate to form complex undergoing conformational changes. Free SNARE motifs are unstructured and when they are associated in a complex they assemble into elongated four-helical bundles. SNAREs are present on the vesicle and the target membranes and the formation of the complex pulls the membrane close together. SNARE proteins are classified in to subfamilies based on a highly conserved layer of interacting amino acids (three glutamines: Qa-, Qb-, Qc-, one arginine: R-) in the centre of the helix bundle. All complexes contain one copy of each SNARE motif. In S. cerevisiae, twenty-four SNARE-encoding genes have been identified [reviewed in []]. Twenty-two of these genes could be found in Y. lipolytica by sequence homology. As in other fungi [], Vam3p and Spo20p were not detected. Spo20p, which seems specific to S. cerevisiae (also not identified in the three other yeasts), contains both Qb and Qc SNARE motifs and is required during sporulation for the prospore membrane formation [,]. Vam3p SNARE is required for homotypic vacuole fusion in S. cerevisiae and there is no homologue outside the Saccharomycetes, the Saccharomycetaceae, K. lactis, has a Vam3p SNARE, but in D. hansenii, another Saccharomycetaceae as in Y. lipolytica (Dipodascaceae) the Vam3 protein was not detected. The mitosporic Saccharomycetales, C. glabrata, possess a Vam3 protein. Pep12p is a late endosomal Qa SNARE with sequence similarity with Vam3p, that can complement vam3 mutants []. In Y. lipolytica and in D. hansenii, the BLAST searches allowed us to identify the Pep12p SNARE and another SNARE we propose to name Pep12p-like because its sequence is closer to Pep12p than to Vamp3p. This Pep12p-like SNARE could play the role of the S. cerevisiae Vam3p in both yeasts. A specificity of Y. lipolytica is the presence of three SSO genes resulting probably from gene triplication, and this seems unique to Y. lipolytica. The Sso proteins are implicated in the fusion of the secretory vesicles to the plasma membrane and S. cerevisiae Sso1p also has sporulation-specific functions []. The multiplicity of SSO genes in Y. lipolytica could reflect its good secretion capacity and its capacity to induce hyphal growth which needs better recycling of plasma membrane material. On the contrary, only one Sso protein (Sso2p) and one Snc protein (Snc2p) were found in K. lactis and D. hansenii. The SSO and SNC genes of K. lactis have been cloned by complementation in S. cerevisiae of sso2- and snc1Δ snc2Δ sem1Δ mutant strains. The K. lactis Ssop can perform both of the S. cerevisiae Ssop functions and the K. lactis Sncp seems to perform only the S. cerevisiae Sncp functions []. Sft1p was not detected in C. glabrata, Syn8p was not detected in C. glabrata and K. lactis and Nyv1p was not detected in D. hansenii.Regulation of fusionThe Sec1/Munc18 (SM) proteins (see Additional file -) [reviewed in []]The SM proteins confer specificity to membrane fusion through their binding to the N-terminal domain of the Qa SNARE protein. Four SM proteins have been found in S. cerevisiae, Sly1p acting between the endoplasmic reticulum and Golgi [] and Sec1 at the final step of exocytosis, Vps45p and Vps33p playing a role between Golgi and endosomes and between endosomes and vacuole. Each SM protein can bind several Qa SNARE but also other SNARE implicated in the same complex, as has been shown for Sly1p []. In addition one SNARE can bind two SM proteins at the same organelle: Vps45p and Vps33p with Pep12p []. Mso1p, a Sec1p-interacting protein, binds to SNARE complex and plays an essential role for vesicle fusion during prospore membrane formation [,].Vsm1p (see Additional file -)The phosphorylation of SNAREs by the cAMP-dependent protein kinase (PKA) regulates their ability to assemble into functional complexes [,]. Phosphorylation of the Sed5p t-SNARE regulates endoplasmic reticulum-Golgi transport as well as Golgi morphology []. The Ssop phosphorylation allows the binding of Vsm1p, a negative regulator of secretion which prevents the formation of the SNARE complex []. Dephosphorylation of Ssop by ceramide activated protein phosphatase (CAPP) increases its ability to form a complex with Sec9p []. Three S. cerevisiae genes code for a PKA but only one gene could be identified in Y. lipolytica and D. hansenii and two were identified in C. glabrata and K. lactis. The CAPP is composed of two regulatory subunits (Tpd3p and Cdc55p) and one catalytic subunit (Sit4p) which were found by sequence homology in the four yeasts in addition to further one Tpd3p-like in C. glabrata.SNARE recyclingSNARE complex dissociation (see Additional file -):After the membrane fusion step, the trans-SNARE complex becomes a cis-SNARE complex whose dissociation requires the ATPase Sec18p and the soluble NSF-attachment protein (Sec17p) as cofactor [,]. In vacuole fusion, Sec17p may displace HOPS from SNAREs to permit subsequent rounds of fusion []. Two Sec18 proteins were detected in C. glabrata.The vesicle-SNARE Snc1p recycling (see Additional file -)The S. cerevisiae RCY1 gene has been identified with a screen for mutants affected in membrane traficking along the endocytic pathway []. Rcy1p contains an amino-terminal F box and a CAAX box motif in its carboxyl-terminal sequence. The F box region of the protein is required for the recycling of the vesicle-SNARE Snc1p. The CAAX box is required for its localization in polar growth regions. Rcy1p interacts with Skp1p through the F box motif and both proteins form a complex necessary for the recycling function []. Rcy1p is a positive regulator of Ypt6p []. Gyp1p, the Ypt1p GTPase activator is also involved in the recycling of Snc1p [], as well as Snx4p, Snx41p and Snx42p []. The ARF-GAP Gcs1p facilitates the incorporation of the Snc1p into COPI recycling vesicles []. Rcy1p is a downstream effector of Ypt31, 32p [].In Schizosaccharomyces pombe, Pof6p, the Rcy1p-homologue, has been identified through a two-hybrid interaction with Skp1. Both proteins are required for normal septum processing and cell separation [], a function which may also require the exocyst function [].Previous to these works, a mutation sls2-, was isolated in Y. lipolytica that causes synthetic lethality when combined with the conditional lethal mutation in the 7S RNA of the signal recognition particle []. Rcy1p and Pof6p are the homologues of Sls2p.ConclusionThe sequencing of four hemiascomycetous yeasts has allowed us to search for proteins involved in vesicular transport by comparison with proteins identified in S. cerevisiae. The method used does not allow the identification of a protein which does not exist in S. cerevisiae or does not belong to a protein family. To identify new candidates, a list from other organism should be established or experimental approach should be performed. The proteins identified are highly conserved between the five yeasts but we have brought to light several specificities of Y. lipolytica in keeping with its good protein secretion capacities and its dimorphic aspect. In Table , we summarize these differences. Some of the proteins for which we did not find homologues have probably a too divergent sequence to be identified through the BLAST searches and may be identified in the future by functional screens. But, the presence of Rab2p- and Rab4p-related proteins as is found in fungi, a potential role of Rab2-related protein and Ypt1p in vesicular transport between endoplasmic reticulum and Golgi and of Rab4-related protein in addition to Sls2p/Rcy1p in membrane recycling reflect the greater complexity of the Y. lipolytica secretion pathway, which is probably dictated by the necessity to secrete and recycle membrane material needed for its filamentous growth. In this work, we have shown that the Rab4p-related protein could have a role in this membrane recycling since a modification of the aspect of the colony, the decrease of the number of cells undergoing dimorphic transition and the change of the wall permeability of the rab4 deleted strain were observed. The three Sso proteins are also indicative of a large secretion capacity and a study of this specificity would be interesting. As previously shown through the diverse studies of Y. lipolytica, this dimorphic yeast has a secretion pathway closer to the mammalian one than has S. cerevisiae. Its translocation apparatus is largely devoted to the co-translational translocation of nascent peptides through the endoplasmic reticulum membrane as is seen in mammalian cells. C. glabrata and K. lactis are the closest to S. cerevisiae for the proteins listed. We found among the C. glabrata proteins, COPII coat protein Sec13p-homologues and Sec18p-homologues implicated in the SNARE recycling and this seems to be specific to this yeast. For K. lactis and D. hansenii, only one Ssop and one Sncp were detected. Though D. hansenii is classified in Saccharomycetaceae, as is S. cerevisiae, and Y. lipolytica is a Dipodascacae, these two yeasts seem closer at least for the proteins involved in vesicular transport. These results are in agreement with the phylogenetic tree presented by Dujon et al []. In Fig. 7a, we have shown the proportion of the Y. lipolytica protein sequences listed (see Additionnal file : Y. lipolytica column) with the greatest homology to S. cerevisiae, C. glabrata, K. lactis, D. hansenii, Schizosaccharomyces pombe, N. crassa, other fungi, animals and plants sequences (see Additional file ). % of these proteins are closest to fungi and N. crassa. If we exclude animals, plants and fungi (keeping only N. crassa) from this search (Fig. 7b), we observed that % of the Y. lipolytica proteins listed are closer to N. crassa. The last comparison is performed without N. crassa (Fig. 7c): in this case, % of the proteins are closer to D. hansenii. This could reflect their common physiology, the filamentous growth and good secretion ability for Y. lipolytica and N. crassa and their proteolytic and lipolytic activities and high salt tolerance [] for Y. lipolytica and D. hansenii. These results confirm that considering the identified proteins playing a role in vesicular transport, Y. lipolytica is closer to the fungi than to S. cerevisiae. The proteins that are the best conserved between S. cerevisiae and Y. lipolytica, are the SNARE proteins (see Additional file ). In Fig. 8a, we have presented the percentage of the Y. lipolytica proteins with the greatest homology to the S. cerevisiae and animal: % of these proteins are closer to the animal ones, particularly, proteins of the AP complexes, Ypt and Arf proteins, TRAPP and HOPS complexes (see Additional file ), whereas for S. cerevisiae, only % of these proteins are closer to animals than to Y. lipolytica (Fig. 8b, see Additional file ). Koumandou and coworkers [] have analysed the protein sequences of tethering complexes and SM proteins from five eukaryotic supergroups. They conclude that the most recent common eukaryotic ancestor had a complex endomembrane system with COG, Exocyst, Dsl1, GARP tethering complexes which could have originated from one common ancestral complex, TRAPP and HOPS complexes which are independently derived and all four SM protein families represented. The phylogenetic tree presented by Dujon et al [] indicates that Y. lipolytica may be less distant from the last common eukaryotic ancestor and the observation of a good sequence conservation between Y. lipolytica and animal TRAPP and HOPS subunits for example, suggests that these complexes are required for the evolution of multicellular organisms. Similarly, Hall et al. [] showed that Rab4p was an ancient component of the endomembrane trafficking system since it exists, and its recycling function is conserved, in Trypanosoma brucei which belongs to an eukaryotic supergroup separated from that of yeast, fungi and animals. In Y. lipolytica also, the presence and the possible recycling role of a Rab4-like protein was observed while in S. cerevisiae and in the three other hemiascomycetous yeasts, this protein has been lost. These observations indicate that S. cerevisiae has diverged further from the last common eukaryotic ancestor than has Y. lipolytica, as far as vesicle-mediated protein transport pathways are concerned and that Y. lipolytica has retained the complexity of the trafficking system allowing evolution to a multicellular organization. So as has been said for fungi [], we can say that «Yarrowia lipolytica and humans are closer than you think» and that this yeast constitutes an interesting model to study the secretion pathway.Table 1Differences observed for the five hemiascomycetous yeasts.S. cerevisiaeY. lipolyticaC. glabrataK. lactisD. hanseniiCOPIISec24p, Sfb2,3p (Sec24p-related) proteins3 proteins2 proteins2 proteinsSec13p1 protein2 proteins1 protein1 proteinSed4pno hits1 proteinno hitsno hitsAdaptorAps1,,3p2 proteins3 proteins3 proteins3 proteinsGga1,2p1 protein (Gga2p) protein (Gga2p) protein (Gga2p) protein (Gga2p)Sorting NexinSnx41,42p1 protein2 proteins2 proteins2 proteinsYptpYpt7p1 protein1 protein1 protein2 proteinsYpt10pno hits1 proteinno hitsno hitsYpt11pno hits1 protein1 proteinno hitsYpt31-32p2 proteins2 proteins1 protein (Ypt31p) protein (Ypt32p)Ypt51,,53p3 proteins2 proteins3 proteins3 proteinsno hitsRab2,4p-relatedno hitsno hitsno hitsYptp regulationYos1pno hitsno hitsno hits1 proteinGyp4pno hitsno hitsno hitsno hitsGyl1pno hits1 proteinno hitsno hitsCOG complexCog1pno hits1 protein1 proteinno hitsCog2p1 protein1 protein1 protein1 proteinCog7pno hits1 protein1 protein1 proteinArfpArf1,,3p2 proteins(Arf1,3p) proteins(Arf1,2p) proteins (Arf2,3p) proteinsArl1p2 proteins1 protein1 protein1 proteinArl3p localizationNatC complex ( proteins) proteins3 proteins3 proteins3 proteinsGARP complexVps51pno hits1 protein1 proteinno hitsSNARE-QaVam3pPep12p-like1 protein1 proteinPep12p-likeSso1,2p3 proteins2 proteins1 protein (Sso2p) protein (Sso2p)SNARE-Qb,QcSpo20pno hitsno hitsno hitsno hitsSNARE-QcSft1p1 proteinno hits1 protein1 proteinSyn8p1 proteinno hitsno hits1 proteinSNARE-RNyv1p1 protein1 protein1 proteinno hitsSnc1,2p2 proteins2 proteins1 protein (Snc2p) protein (Snc2p)Exocytosis SNARE regulation proteinsTpd3p (CAPP regulatory subunit) protein2 proteins1 protein1 proteinTpk1,,3p (PKA) protein2 proteins2 proteins1 proteinSNARE recyclingSec18p1 protein2 proteins1 protein1 proteinSee Additional file for the list of proteins potentially implicated in vesicular transport.Figure 7The percentage of Y. lipolytica proteins with the greatest homology: a: to Saccharomyces cerevisiae, Candida glabrata, Kluyveromyces lactis, Debaryomyces hansenii, Schizosaccharomyces pombe, Neurospora crassa, other fungi, animals, plants proteins; b: to Saccharomyces cerevisiae, Candida glabrata, Kluyveromyces lactis, Debaryomyces hansenii, Schizosaccharomyces pombe, Neurospora crassa proteins; c: to Saccharomyces cerevisiae, Candida glabrata, Kluyveromyces lactis, Debaryomyces hansenii, Schizosaccharomyces pombe proteins. See Additional file for the list of E-values obtained with BLAST of Y. lipolytica proteins against NCBI eukaryotic sequences.Figure 8Percentages of greatest homology. a: The percentage of Y. lipolytica proteins with the greatest homology to S. cerevisiae and animal proteins; b: The percentage of S. cerevisiae proteins with the greatest homology to Y. lipolytica and animal proteins. See Additional file for the list of E-values obtained with BLAST of the S. cerevisiae proteins against NCBI eukaryotic protein sequences.MethodsStrains and growth conditionsEscherichia coli strains DH5alpha (F'/endA1 hsdR17 (rK- mK+) supE44 thi- recA1 gyrA (Nalr) relA1 Δ (lacIZYA-argF) U169deoR (φ 80dlacΔ (lacZ)M15) was used as host strain for bacterial transformations and plasmid propagation.The Yarrowia lipolytica strain INAG136463 (MatB, scr1 : : ADE1, SCR2, his-, leu-, ura3) was used for the inactivation of RAB4-related gene.Escherichia coli cells were grown in LB medium (% bactotryptone, % yeast extract, .% NaCl), μg/ml ampicillin, °C. Yarrowia lipolytica cells were cultivated either on rich YPD medium (% yeast extract, % bactopeptone, % glucose), °C, or on minimal medium: .% yeast nitrogen base without amino acids (Difco laboratories), % glucose as carbon source, mM phosphate buffer pH ., °C with amino acids required and , mg/ml '-fluoroorotic acid for ura3- strain selection.Gene inactivationDisruption was performed using the two-step «pop-in/pop-out» method []. The disrupted gene was obtained by the deletion of the BstEII-ClaI fragment of the RAB4-related gene cloned between the HindIII-KpnI sites of the p0 vector [].DNA techniquesStandard techniques were used according to Sambrook et al. []. Enzymes were supplied by New England Biolabs. All vectors inserts were checked by sequencing by Genome express (France).Transformation proceduresThe E. coli strains were transformed by the method of Chung and Miller [].Y. lipolytica strain transformations were carried out according to Xuan et al. [].Sensitivity to SDS and Calcofluor WhiteCells of the INAG136463 (wt) and two clones (, ) of the deleted rab4-related strains were grown in YPD medium. μl droplets of serial dilutions of exponential growing cultures of each strain were inoculated on the surface of YPD plates containing . μg/ml, μg/ml, μg/ml Calcofluor White (CW) or . %, .%, .% sodium dodecyl sulfate (SDS).FM4- stainingFor the strains of yeast cells, OD600 units of exponential growth in YPD medium (OD600 .–) were resuspended in μl of YPD containing μM FM4-. Cells were incubated min. at °C and washed three times in ice-cold medium. Cells were resuspended in YPD and incubated at °C. Aliquots were taken at various times and internalization was stopped with mM NaN3 and mM NaF. Stained cells were visualized using fluorescence optics [adapted from []].Informatic analysesHemiascomycetous yeast genome sequences, BLAST searches of vesicular secretion proteins and BLAST results (performed Apr , with ,, sequences) were obtained from the Génolevures web site []. S. cerevisiae sequences were collected from Saccharomyces Genome Database []. BLASTs against protein databases were obtained from NCBI (BLAST with ,, sequences) [] and Infobiogen web site []. Protein analyses were done with NCBI Conserved Domain Architecture Retrieval Tool [], ExPASy Proteomics tools [] and CBS Prediction Servers [].A list of proteins implicated in S. cerevisiae vesicular secretion was made from literature. These S. cerevisiae protein sequences were used for BLAST searches with the Génolevures web site. For protein families such as Rab protein, autoBLAST, which means BLAST of a sequence against its own genome, were made to identify all the members of the family. The protein sequences of the new members were identified by BLAST searches against the NCBI eukaryotic protein sequences.The percentages of proteins with the greatest homology (Fig. and ) were determined by quantification of the best E-values obtained with the BLAST searches against the NCBI eukaryotic protein sequences.AbbreviationsY. lipolytica,Yl: Yarrowia lipolytica; C. glabrata, Cg: Candida glabrata; K. lactis, Kl: Kluyveromyces lactis; D. hansenii, Dl: Debaryomyces hansenii; S. cerevisiae, Sc: Saccharomyces cerevisiae; N. crassa, Nc: Neurospora crassa; SNARE: Soluble N-ethylmaleimide-sensitive factor Attachment protein Receptor.Authors' contributionsDS conceived the study, carried out the molecular genetic studies, the sequence analyses and drafted the manuscript. JMB participated in the sequence analyses. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1Drawing of Yarrowia lipolytica identified proteins coats. PM: plasma membrane, ER: endoplasmic reticulum, RE: recycling endosome, EE: early endosome, LE: late endosome, MVB: multi-vesicular bodies, SV: secretory vesicle.Click here for fileAdditional file 3Drawing of Yarrowia lipolytica identified Ypt/Rab GTPases. PM: plasma membrane, ER: endoplasmic reticulum, RE: recycling endosome, EE: early endosome, LE: late endosome, MVB: multi-vesicular bodies, SV: secretory vesicle.Click here for fileAdditional file 4Full image of Figure 2Click here for fileAdditional file 5Drawing of Yarrowia lipolytica identified tethering factors. PM: plasma membrane, ER: endoplasmic reticulum, RE: recycling endosome, EE: early endosome, LE: late endosome, MVB: multi-vesicular bodies, SV: secretory vesicle.Click here for fileAdditional file 6Drawing of Yarrowia lipolytica identified SNARE and SM proteins. PM: plasma membrane, ER: endoplasmic reticulum, RE: recycling endosome, EE: early endosome, LE: late endosome, MVB: multi-vesicular bodies, SV: secretory vesicle.Click here for fileAdditional file 7E-values. E-values found for BLAST of Yarrowia lipolytica proteins against Saccharomyces cerevisiae, Candida glabrata, Kluyveromyces lactis, Debaryomyces hansenii, Schizosaccharomyces pombe (Sp),Neurospora crassa, other fungi, animals, plants, obtained with NCBI web site. Numbers between brackets indicate the order of best BLAST hits. Fungi: Ashbya gossypii (Ag), Aspergillus clavatus (Ac), Aspergillus fumigatus (Af), Aspergillus nidulans (Asn), Aspergillus niger (An), Aspergillus orizae (Ao), Aspergillus parasiticus (Ap), Aspergillus terreus (Ast), Chaetomium globosum (Chg), Coccidioides immitis (Ci), Coprinopsis cinerea (Cc), Cryptococus neoformans (Cn), Gibberzlla zeae (Gz), Hypocrea lixii (Hl), Magnaporthe grisea (Mg), Neosartorya fischeri (Nf), Neurospora crassa (Nc), Paracoccidioides brasiliensis (Pb), Phaeosphaeria nodorum (Pn), Ustilago maydis (Um). Animals: Aedes aegypti (Aa), Aiptasia pulchella (Ap), Anopheles gambiae (Ang), Apis mellifera (Am), Bombyx mori (Bm), Bos taurus (Bt), Caenorhabditis briggsae (Cb), Caenorhabditis elegans (Ce), Canis familiaris (Cf), Danio rerio (Dr), Drosophila grimshawi (Dg), Drosophila melanogaster (Dm), Drosophila pseudoobscura (Dp), Gallus gallus (Gg), Homo sapiens (Hs), Macaca mulatta (Mam), Mus musculus (Mm), Oryzias latipes (Ol), Pan troglodytes (Pt), Pongo pygmaeus (Pp), Rattus norvegicus (Rn), Strongylocentrus purpuratus (Stp), Xenopus laevis (Xl), Xenopustropicalis (Xt). Plants: Arabidopsis thaliana (At), Brassica oleracea (Bo), Brassica rapa (Br), Hyacinthus orientalis (Ho), Lotus japonicus (Lj), Medicago truncatula (Mt), Nicotiana tabacum (Nt), Oenothera odorata (Oo), Oriza sativa (Os), Pisum sativum (Ps), Solanum chacoense (Soc), Solanum tuberosum (St), Zea mays (Zm). (As Debaryomyces hansenii Vps35p, Snx3p, Gyp2p, Sec20p, Sec18p sequences were absent from the NCBI database when the comparison was done, the e-values were obtained with the NCBI BLAST of the Debaryomyces hansenii protein sequence against Yarrowia lipolytica sequences).Click here for fileAdditional file 8E-values. E-values found for NCBI BLAST of Saccharomyces cerevisiae proteins against Yarrowia lipolytica and animal proteins (see Additional file legend for list of abbreviations).Click here for fileAdditional file 2List of Yarrowia lipolytica genes coding for the proteins potentially implicated in vesicular transport. They were obtained by comparison against Saccharomyces cerevisiae protein sequences, BLAST results come from Génolevures web site, if Candida glabrata, Kluyveromyces lactis, Debaryomyces hansenii protein was found is indicated (see Additional file for the list of the Candida glabrata, Kluyveromyces lactis, Debaryomyces hansenii genes).Click here for fileAdditional file 9List of Candida. glabrata, Kluyveromyces lactis, Debaryomyces hansenii genes coding for the proteins potentially implicated in vesicular transport.Click here for file
|
PMC2738435.txt
|
TITLE: Chikungunya Virus, Cameroon, AUTHORS: Christophe N. Peyrefitte, Dominique Rousset, Boris A.M. Pastorino, Regis Pouillot, Maël Bessaud, Fabienne Tock, Helene Mansaray, Olivier L. Merle, Aurelie M. Pascual, Christophe Paupy, Aurelia Vessiere, Patrice Imbert, Patrice Tchendjou, Jean-Paul Durand, Hugues J. Tolou, Marc Grandadam ABSTRACT: We report the isolation of chikungunya virus from a patient during an outbreak of a denguelike syndrome in Cameroon in . The virus was phylogenetically grouped in the Democratic Republic of the Congo cluster, indicating a continuous circulation of a genetically similar chikungunya virus population during years in Central Africa. BODY: Chikungunya virus (CHIKV), formerly only an anecdotally described arbovirus, is now a worldwide public health problem (). Recently, numerous cases of CHIKV infection have been reported from a major outbreak of febrile illness around the Indian Ocean, which included Comoros, Mauritius, Réunion Island (,), and southern India ().CHIKV is widely distributed in tropical Africa (,) and in Asia (). In Africa, until , the virus was described as endemic, perpetuated through a sylvatic cycle involving wild primates, humans, and mosquitoes of the genus Aedes (,). During the past years, the urban cycle has also tended to play a role in Central Africa (). Nevertheless, although recent serologic surveys suggest a high prevalence of Togaviridae, Flaviviridae, and Bunyaviridae (,), understanding of the circulation and effects of arboviruses in Cameroon remains imprecise. This lack of understanding may reflect confusion between arboviral infections and hyperendemic Plasmodium falciparum infection.We report the first isolation, to our knowledge, of CHIKV in Cameroon. The virus was identified during an outbreak of a febrile syndrome in French soldiers in Douala and in patients from an urban medical center in Yaoundé. We also found evidence of cocirculation of CHIKV and dengue virus (DENV).The StudyIn Douala, Cameroon, sporadic cases of a denguelike syndrome were recorded in French soldiers (patients and ) on April and May , , respectively (Table). From the end of May through the end of July , more cases of denguelike syndrome, which included fever, asthenia, maculopapular rashes, and arthralgia, were observed in Yaoundé. The number of patients who sought treatment at the Yaoundé Medical Center peaked in mid-June . Blood samples were collected from of the patients who visited the medical center. The patients’ ages ranged from to years. The delay between the onset of symptoms and the sampling ranged from to days with a median of days (Table). All but patient lived in Yaoundé, and none of these patients had a history of travel abroad or from Yaoundé. Nine patients were Cameroonian, and all other patients were from other countries; patients were female. A blood sample from a -year-old woman who returned to France from Yaoundé was also received. All patients had negative results for P. falciparum according to rapid test (Core Malaria Pf, Core Diagnostics, Birmingham, UK) and thick smear examination.TableCharacteristics of patients with febrile acute denguelike syndrome, Cameroon, *Patient no.Sex/ageCitySymptom onsetSampling dateDelay, daysIgM†IgG†PCR‡1M/35Douala3 Apr7 Apr4NegNegNeg1M/35Douala3 Apr11 May21Pos CHIKPos CHIKVNeg2§M/36Douala22 May23 May1NegNegPos CHIK3F/54Yaoundé11 Jun12 Jun1NegNegNeg4F/49Yaoundé10 Jun13 Jun3NegNegNeg5F/42Yaoundé11 Jun12 Jun1NegNegNeg6F/41Yaoundé7 Jun7 Jun0NegPos FlaviNeg7M/38Yaoundé15 Jun19 Jun4NegNegNeg8M/30Yaoundé18 Aug19 Jun1NegNegNeg9M/21Yaoundé15 Jun19 Jun4NegNegNeg10F/32Yaoundé21 Jun22 Jun1NegNegNeg11F/22Yaoundé19 Jun26 Jun7Pos CHIKVNegNeg12F/53Imported case20 Jun26 Jun6Pos CHIKVPos CHIKNeg13M/42Yaoundé24 Jun27 Jun3NegPos FlaviNeg14F/42Yaoundé18 Jun28 Jun10NegNegNeg15F/27Yaoundé18 Jun29 Jun11NegPos FlaviNeg16F/43Yaoundé26 Jun30 Jun4NegNegNeg17M/31Yaoundé16 Jun30 Jun14NegNegNeg18F/37Yaoundé22 May30 Jun39Pos CHIKPos CHIKNeg19F/45Yaoundé10 Jun30 Jun20NegNegNeg20M/45Yaoundé22 Jun30 Jun8NegPos Flavi and CHIKVNeg21M/1Yaoundé23 Jun30 Jun7NegNegNeg22F/9Yaoundé10 Jun30 Jun20NegNegNeg23M/4Yaoundé26 Jun30 Jun4NegNegNeg24M/54Yaoundé4 Jul6 Jul2NegNegNeg25F/45Yaoundé14 Jun5 Jul21NegNegNeg26M/48Yaoundé24 Jun6 Jul12NegNegNeg27M/20Yaoundé30 Jun1 Jul1NegNegNeg28M/37Yaoundé4 Jul11 Jul7Pos CHIKPos CHIKVNeg29M/36Yaoundé9 Jul11 Jul2NegNegNeg30M/33Yaoundé28 Jun10 Jul12NegNegNeg31M/32Yaoundé10 Jul12 Jul2NegNegPos DENV32F/38Yaoundé16 Jul17 Jul1NegNegNeg33M/45Yaoundé21 Jul26 Jul5NegNegNeg*IgM, immunoglobulin M; Neg, negative; Pos, positive; CHIKV, chikungunya virus; Flavi, flavivirus; DENV, dengue virus. †CHIKV isolation successful. ‡DENV, West Nile virus, Wesselsbron virus, Rift Valley fever virus, Bunyamwera virus, and CHIKV antibodies tested. §DENV, West Nile virus, CHIKV tested by real-time RT-PCR.Serum specimens were tested for immunoglobulin M (IgM) and IgG antibodies specific for DENV, West Nile virus (WNV), Wesselsbron virus, Rift Valley fever virus, Bunyamwera virus, and CHIKV by IgM-antibody capture (MAC-ELISA) and IgG sandwich ELISA, respectively (). A serum sample was consider positive if the optical density (OD) ratio of viral antigen to uninfected cells was >. The presence of CHIKV, DENV, and WNV genomes was tested for by specific real-time reverse transcription PCR (RT-PCR) (). Virus isolation on C6/ and Vero cells was attempted on samples that were positive by RT-PCR ().The serologic follow up of patient (Table) for a -week period detected seroconversion to a virus antigenically related to CHIKV virus (the OD ratios obtained with the second sample were >) for IgM and IgG. A sample from patient was obtained the day after the onset of symptoms, and no antibodies to all tested arboviruses were detected. However, the specimen was positive by real-time RT-PCR for CHIKV. The patient’s sample yielded CHIKV when cultured, and the envelope gene was partially sequenced (position ,–,, GenBank accession no. Bankit851776). The .-kb sequence genetic analysis did not show any codon deletion or insertion when compared with other African CHIKV sequences available in the GenBank database (,,). A high degree of identity was observed when the sequence was compared with the Democratic Republic of the Congo (DRC) strains isolated in (). Paired identity ranged from % to .% at the nucleotide level and from .% to .% at the amino acid level. The Cameroon isolate displayed a higher nucleotide divergence (paired identity ranging from % to .%) when compared with the Réunion Island strains (,,). However, amino acid sequences were highly conserved (%–.%). The sequence identity among these isolates highlights their common origin and particularly the genetic stability of CHIKV despite the years and the geographic distance from the DRC outbreaks. As shown in the phylogenetic tree (Figure), the CHIKV Cameroon strain clustered with DRC CHIKV strains with a high bootstrap value of . This genotype of CHIKV was closely related to strains from the Central African Republic and the Uganda isolate (,). The close genetic relationship suggests a continuous circulation of a homologous CHIKV population in Central Africa with a high degree of genetic stability. The genetic stability of the Central African CHIKV strains during years, whether associated with epidemic or sporadic cases, highlights the peculiar importance of the few mutations detected in the recent Réunion Island isolates (). This also suggests that the Central African strain CHIKV zone of circulation now includes India (), the Indian Ocean, and Cameroon.FigurePhylogenetic tree of chikungunya virus (CHIKV) based on partial nucleotide sequences (′ extremity of E1/′-UTR, position ,–,). Phylogram was constructed with MEGA program and tree drawing used the Jukes-Cantor algorithm for genetic distance determination and the neighbor-joining method. The percentage of successful bootstrap replicates (, bootstrap replications, confidence probability >%) is indicated at the nodes. The length of branches is proportional to the number of nucleotide changes (% of divergence). Asterisk (*) and arrow indicate the strains isolated in this work. The dark triangle corresponds to viruses clustering together. O’nyong-nyong virus (ONNV) sequence has been introduced for correct rooting of the tree. The GenBank reference no. for the Cameroon CHIKV isolate is EF051584.The phylogenetic tree also illustrates the differences between the Cameroon isolates and the Asian subgroup isolates. Moreover, when compared with Asian CHIKV, including the isolates, the Cameroon strain showed %–.% and .%–.% identity at the nucleotide and amino acid levels, respectively. Despite the similarity, cross-neutralization experiments must be conducted to confirm the protective effect of the Asian CHIKV-based vaccine against Central African strains ().Among patients from Yaoundé, (patient ) had only IgM antibodies specific to CHIKV, while patients and had both IgM and IgG antibodies specific to CHIKV (Table). One patient from Cameroon (patient ) had IgG specific to both CHIKV and flavivirus. Three patients (nos. , , and ), of whom were Cameroonian, had antibodies specific for flavivirus. All samples were negative for WNV and CHIKV by RT-PCR. One sample (from patient ) was positive for DENV; however, no virus was detected by cell culture. These results suggested a cocirculation of CHIKV and dengue virus during the same period, which is consistent with the suspected circulation of dengue virus, CHIKV, and yellow fever virus observed in a study from through in Cameroon ().In Cameroon, as in DRC (), patients were likely infected in urban or periurban centers (Yaoundé, the capital of Cameroon; Douala, a major city). These infections occurred in a context where Aedes albopictus tends to replace indigenous Ae. aegypti in rural and urban Cameroonian environments (). This finding suggests that urban cycles and urban vectors, in addition to the traditional forest-dwelling vectors, may play an important role in the maintenance and amplification of CHIKV in Africa.ConclusionsSince its first isolation in (), CHIKV has been isolated in different Central African countries (,). Until now, only alphavirus strains antigenically suspected to be CHIKV had been isolated from human patients in Cameroon (). Recent serosurvey studies suggested a possible CHIKV circulation in Cameroon (,). Our Cameroon CHIKV isolate confirmed its circulation in this country. Our study suggests a -year continuous circulation of genetically stable and indigenous strains in Central Africa rather than importation of CHIKV from the recent Indian Ocean or Asian outbreaks. Moreover, the genetic stability of the Central African CHIKV highlights the importance of the unique molecular features that was shown in Réunion Island isolates ().
|
PMC2394270.txt
|
TITLE: Constitutive activation of STAT3 and STAT5 is present in the majority of nasopharyngeal carcinoma and correlates with better prognosis AUTHORS: J-R Hsiao, Y-T Jin, S-T Tsai, A-L Shiau, C-L Wu, W-C Su ABSTRACT: Constitutively activated signal transducers and activators of transcription (STAT) factors, in particular STAT1, STAT3 and STAT5, have been demonstrated in a variety of human tumours and cancer cell lines. However, data on the expression of these STATs in nasopharyngeal carcinoma (NPC) are limited. In this study, the expression patterns of STAT1, STAT3 and STAT5 were immunohistochemically examined on the archival specimens from patients with NPC. Staining results of each STATs were then correlated with the clinical parameters and prognosis of these patients. The results showed that constitutive activation of STAT3 and STAT5 was detected in the majority, . and .%, respectively, of the tumour specimens. Furthermore, coexpression of activated STAT3 and STAT5 was found in .% of the specimens. In contrast, constitutive activated STAT1 could only be detected in (.%) cases. Surprisingly, following radiotherapy, patients with constitutive STAT5 activation, or activation of both STAT3 and STAT5, had better disease-free survival and overall survival than those without activated STAT5. To our knowledge, this is the first report providing the overall expression patterns and prognostic significance of specific STATs in NPC. BODY: Signal transducers and activators of transcription (STAT) proteins comprise a family of transcription factors and function as downstream effectors of many cytokine and growth factor receptors (Darnell, ). Signalling pathways mediated by STATs are critical for many normal cellular functions, including embryonic development, organogenesis, immunological interaction, growth, differentiation and survival (Buettner et al, ). During cytokine or growth factor stimulation, STATs are tyrosine phosphorylated, dimerise, and translocate into the nucleus to regulate target gene expression (Darnell, ). The activation duration of STATs under physiological conditions is usually temporary and is tightly regulated by a number of cellular mechanisms, such as tyrosine dephosphorylation, ubiquitin/proteosome-mediated degradation, negative feedback loop mediated by CIS/SOCS/JAB/SSI family of proteins, or inhibition of STAT DNA-binding activity through association with protein inhibitor of activated STAT (PIAS) proteins (Lin and Leonard, ). However, constitutively activated STATs, especially STAT1, STAT3 and STAT5, have been found in a variety of human tumours and cancer cell lines, including blood malignancies and solid tumours (Turkson and Jove, ). In various tumour cell lines, persistent signalling of specific STATs, in particular STAT3 and STAT5, has been shown to stimulate cell proliferation and prevent apoptosis through upregulating a number of target genes, such as c-Myc, cyclins and bcl-x. In contrast, inhibition of constitutively activated STAT3 or STAT5 leads to growth suppression or apoptosis (Buettner et al, ).Nasopharyngeal carcinoma (NPC) is endemic in South-east Asia and closely associated with Epstein–Barr virus (EBV) (Tsai et al, ). Epstein–Bars virus resides in NPC cells in a status called type II latency, with only a few latent genes expressed (Brooks et al, ). One of these latent genes, latent membrane protein (LMP-), has been shown to have transforming potential (Baichwal and Sugden, ; Kaye et al, ) and is implicated in the oncogenesis of various EBV-associated malignancies, including NPC. LMP- can also interact with Janus Kinase (JAK3) and activate STAT proteins (Gires et al, ). In a recent study, the JAK/STAT pathway was shown to play a role in maintaining in vivo latency of EBV (Chen et al, ). Aberrant STAT activation has also been proposed as a necessary and predisposing event for EBV-driven tumorigenesis in immunocompetent individuals (Chen et al, ).Although the presence of constitutive STAT1, STAT3 and STAT5 activation has been demonstrated in NPC tissue (Chen et al, ), data on the levels of expression of these STATs in NPC are lacking. Thus, whether expression of these constitutively activated STATs affects prognosis after treatment remains unclear. This study used immunohistochemical methods to examine the expression pattern of STAT1, STAT3 and STAT5 in biopsy specimens of patients with NPC and correlated the results with clinical parameters of these patients.MATERIALS AND METHODSPatients and specimensA total of NPC patients with complete follow-up records and available archival nasopharyngeal biopsy specimens were recruited in this study. All of the patients, including males and females, were histologically confirmed to have WHO type II (nonkeratinising) or WHO type III (undifferentiated) NPC (Shanmugaratnam et al, ). The clinical status of these patients was determined using the UICC/AJCC staging system (Sobin and Wittekind, ) after reviewing their clinical records and image studies which included chest X-ray, abdominal sonography, bone scan and CT scan. All patients were previously untreated and received radiation treatment with curative intention between and at the Department of Radiation Oncology of National Cheng Kung University Hospital. In all, patients received radiotherapy as their sole treatment modality. The other five patients received concurrent chemotherapy with bolus infusion of cisplatin mg m− day− on the beginning day of weeks , and after the start of the radiotherapy regimen. All patients completed the full-course of radiation therapy within weeks without any interruption and received regular follow-up at our hospital. Among the patients, remained disease-free for a median follow-up period of . months, ranging from . to months. In total, patients suffered from loco-regional relapse and/or distant metastasis after treatment, including nine patients with local recurrence (one also with distant metastasis), eight patients with recurrence in the neck (five also with distant metastasis), two patients with both local and neck recurrence (one also with distant metastasis) and one patient with distant metastasis without loco-regional relapse. All the archival nasopharyngeal biopsy specimens used in this study were obtained from the Department of Pathology of National Cheng Kung University Hospital.Immunohistochemical studySerial -μm histological sections were cut, mounted on glass slides coated with -aminopropyltriethoxysilane, and air-dried overnight at room temperature. The sections were then deparaffinised in xylene and rehydrated in ethanol. Haematoxylin and eosin (H&E) staining was first performed in each specimen to confirm the presence of tumour cells. Endogenous peroxidase activity was then blocked with methanol containing % H2O2 for min. For all sections used in this study, the antigen retrieval procedure was performed by immersing the slides in citrate buffer (pH .) and then heating the slides in a microwave oven for min. The sections were then incubated with primary antibodies using : dilution for STAT1, STAT3 and STAT5 (mouse anti-human IgG1, Santa Cruz, CA, USA) at °C overnight, followed by staining with Universal Immuno-peroxidase polymer (UIP) solution (Simple Stain MAX PO MULTI, Nichirei, Tokyo, Japan) for min. The sections were finally reacted with AEC (-amino--ethylcarbazole) substrate solution (DAKO, Glostrup, Denmark) and then counterstained with haematoxylin before being mounted. Non-human reactive mouse IgG1 (DAKO, Glostrup, Denmark) was used as an isotype negative control.The sections were then observed under a light microscope by a pathologist (IT Jin) who was blinded to the clinical characteristics of the patients. The degree of expression of each STAT was then determined independently with the following rules. Intranuclear staining of tumour cells was considered to indicate the presence of constitutively activated STATs. Using high magnification power (× ), five representative fields in each section were evaluated. In total, tumour cells were counted in each field. In sections containing less than five tumour nests, all tumour nests were evaluated. The percentage of immunoreactive tumour cells with nuclear staining was then calculated. When equal or more than % of the counted tumour cells on one slide showed identifiable nuclear staining, the slide was classified as ‘positive’ for significant constitutive STATs activation. If less than % of tumour cells showed a nuclear staining pattern, the slide was considered to show no significant constitutive activation of STATs and was classified as ‘negative’. Staining results of the specimens were then correlated with the clinical characteristics of the patients.Statistical analysisχ2 test was used to calculate the significance of the relation between expression of each STAT and clinical characteristics. The Kaplan–Meier method was used in the survival analysis. Log-rank test was used to calculate the significance of differences in the survival analysis. Cox's proportional hazards regression model was used to study the influence of covariates on survival time. A probability level of less than . (P<.) was considered to indicate a significant difference.RESULTSConstitutive activation of STAT3 and STAT5 in the majority of NPC specimensBoth STAT3 and STAT5 were significantly activated in a high proportion of the NPC specimens included in this study. Significant constitutive STAT3 activation was noted in (.%) of the samples studied, while significant constitutive STAT5 activation was noted in (.%) of the specimens (Table 1Table 1Patient characteristics and the staining results of STAT3 and STAT5 Constitutive STAT3 activationConstitutive STAT5 activation (+)(−)P-value(+)(−)P-valueNumber of specimens (.%) (.%) (.%) (.%) Clinical parameters AgeMean±s.d. (years).±..±...±..±.. SexMale34150.. Female93 TT11360.31a1360.55a T22512 T300 T450 NN01770.73b1590.84b N1189 N261 N321 StageI740.46c830.98c II2612 III51 IV51 WHO subtypeII1470.. III2911 Treatment modalityR/T only39170.. Concurrent C/T+R/T41 Prognosis after treatment Disease-free ..* Recurrence of disease aT1+T2 vs T3+T4.bN0+N1 vs N2+N3.cStage I+II vs Stage III+IV.*Statistically significant (P<.).). In contrast, constitutive STAT1 activation could only be detected in eight (.%) of the cases. The intensity of intranuclear staining varied among the specimens, and differences of staining intensity were also noted among tumour nuclei within the same histological section. However, strong intranuclear staining, or intranuclear and cytoplasmic staining, was noted in of the specimens classified as ‘positive’ for significant constitutive STAT3 activation (Figure 1A, BFigure 1Staining of STAT1, STAT3 and STAT5 on serial pathological sections of a NPC biopsy specimen. (A) H&E stain. Large, vesicular nucleus with prominent nucleolus (arrow) is noted in majority of the tumour cells. (B) Strong intranuclear (arrowheads) and cytoplasmic staining of STAT3. (C) intranuclear (arrows) and faint cytoplasmic staining of STAT5. Nuclear STAT5 staining is also noted in endothelial cells of a nearby vessel (arrowheads). (D) No nuclear staining of STAT1 in the tumour cells (arrows). Nuclear staining of STAT1 is noted in an infiltrating lymphocyte (arrowhead) (original magnification × ).) and in of the specimens classified as ‘positive’ for significant constitutive STAT5 activation (Figure 1C). In contrast, most of the specimens showed no intranuclear staining of STAT1 protein in tumour cells (Figure 1D).Frequent coactivation of both STAT3 and STAT5Among the specimens classified as ‘positive’ for significant STAT3 activation, were also classified as ‘positive’ for STAT5 activation. Among the remaining specimens that were classified as ‘negative’ for STAT3 activation, were also classified as ‘negative’ for constitutive STAT5 activation (Table 2Table 2Results of STAT3 and STAT5 staining on the NPC specimens Constitutive STAT3 activation (+)(−)TotalConstitutive STAT5 activation(+) (.%) (.%) (−) (.%) (.%) Total ). The proportion of specimens with same STAT3 and STAT5 classification was .% ( out of ), suggesting that STAT3 and STAT5 might be coactivated.Constitutive activation of STAT5, or both STAT3 and STAT5, correlates with better prognosisAmong the 61specimens stained for STAT5 protein, specimens were classified as ‘positive’ and as ‘negative’ for significant constitutive STAT5 activation. As shown in Table , no difference was found in the clinical parameters of these two groups of patients, including age, sex, TNM stage, WHO subtype classification and treatment modalities. However, comparison of treatment outcome among these two groups revealed that patients whose biopsy specimens showed significant constitutive STAT5 activation had better prognosis after treatment (χ2 test, P=.). Patients who had no significant constitutive STAT5 activation on histological sections were more likely to suffer from loco-regional relapse and/or distant metastasis. On the other hand, the STAT1 and STAT3 staining results showed no such correlation (χ2 test, P=. and ., respectively). This result still held true in the survival analysis. Patients whose biopsy specimens showed significant constitutively activated STAT5 also had better disease-free survival and overall survival than those patients who had ‘negative’ results on their biopsy specimens (Figure 2A, BFigure 2Survival of NPC patients as a function of the staining results of STAT5, or both STAT3 and STAT5: (A) Overall survival as a function of results of STAT5 staining, (B) disease-free survival as a function of results of STAT5 staining, (C) overall survival as a function of results of both STAT3 and STAT5 staining, and (D) disease-free survival as a function of results of both STAT3 and STAT5 staining.; log-rank test, P<. and <.). Since activation of both STAT3 and STAT5 was frequently noted in the same specimen, we then explored whether this coactivation could have an impact on treatment outcome. Interestingly, patients who had ‘positive’ staining results for both activated STAT3 and STAT5 ( patients) had significantly better overall survival and disease-free survival than those are ‘negative’ for both staining ( patients) (Figure 2C, D; †log-rank test, P<.). However, the double-positive group did not significantly predict better prognosis than the STAT5 positivity only group, both in overall survival and disease-free survival. On the contrary, the double-positive group significantly predicted better prognosis than the STAT3 positivity only group in overall survival (Figure 2C, ‡log-rank test, P<.), implying that STAT5 is more important in predicting prognosis than STAT3. In figure 2C and D, although the STAT5 positivity only group has a favourable survival curve than the other groups, there was no statistical significance detected. This might be due to the small number of patients in the STAT5 positivity only group (five patients). To clarify whether the better prognostic prediction was due to STAT5 positivity only or due to simultaneous activation of STAT3 and STAT5, multivariate analysis was performed. In the Cox's regression model, only STAT5 was selected to be an independent variable for both overall survival (hazard ratio ., P=.) and disease-free survival (hazard ratio ., P=.), suggesting that STAT5 is the major determinant of better prognosis.DISCUSSIONConstitutively activated STATs have been detected in a variety of human tumours and cancer cell lines, including blood malignancies and solid tumours. Most studies demonstrating constitutive activation of STATs in solid human malignancies have used either Western blot analysis and/or electrophoretic mobility shift assay (EMSA) as the main methodology. However, considering the ‘lymphoepithelioma-like’ nature of NPC (Shanmugaratnam et al, ), applicability of these methods might be confounded by the infiltrating lymphocytes in this tumour and is not suitable under such conditions. Therefore, we decided to use immunohistochemical methods to study the expression patterns of STATs in NPC. The suitability of this strategy is further supported by the fact that, in addition to their presence in lymphocytes (Figure 1D), constitutively activated STATs are also noted in other nontumour cells (Figure 1C).A recent study found persistent STAT1, STAT3 and STAT5 activation in NPC tissue (Chen et al, ). However, insufficient data have been reported to provide an overall picture of the expression of these STATs in NPC. This study has provided the first evidence that significant constitutively activated STAT3 and STAT5 are present in over half of NPC patients (Table ), while STAT1 activation is present in only a minor proportion of these patients. The intranuclear staining intensity of STAT3 and STAT5 was strong in many of the tissue specimens (Figure ), suggesting that STAT signalling is very active in this malignancy. We also noted that constitutively activated STAT3 and STAT5 frequently coexisted in the same specimen. It has been shown that different STATs can be activated by the same ligand and/or intracellular tyrosine kinase (Ruff-Jamison et al, ; Carlesso et al, ; Olayioye et al, ). Simultaneously persistent STATs activation is also noted in a variety of human cancers (Turkson and Jove, ). As many cellular genes regulating cell cycle and apoptosis, such as c-Myc, cyclin D and bcl-x are downstream targets of STAT3 and STAT5 (Buettner et al, ), it is reasonable to speculate that persistent signalling of STATs, particularly STAT3 and STAT5, may play a role in tumorigenesis of NPC.Although STAT1 activation has been reported in some tumours and cell lines, STAT1 activation is associated with tumour suppression rather than proliferation in most conditions (Buettner et al, ). In a recent study, activated STAT1 was also shown to negatively regulate angiogenesis, tumorigenicity and metastasis of tumour cells (Huang et al, ). STAT1 deficiency has also been found in a variety of tumour cell lines and this deficiency is responsible for the lack of INF-γ-mediated tumour suppression effects (Wong et al, ; Abril et al, ; Sun et al, ; Pansky et al, ). Therefore, it was not unexpected that constitutively activated STAT1 was found in only eight (.%) of the NPC specimens. In our study, only one of the patients suffering from disease recurrence had significant STAT1 activation.The presence of constitutively activated STAT5 and coactivation of STAT3 and STAT5 was correlated with better prognosis in this study. Using Cox's regression model, STAT5 was identified to be an independent prognosis-predicting factor. Although STAT3 was found in .% of these NPC patients, it is not an independent prognostic factor for survival. The lack of statistical significance for STAT3 might result from the frequent co-activation of STAT3 and STAT5, or the relative small number of patients in this series. It is also possible that STAT5 is truly the only determinant of better prognosis. The prognostic significance of each STATs in NPC deserves further investigation.In most conditions, activation of STAT3 or STAT5 up-regulates cell cycle progression and antiapoptotic genes in cells. Therefore, recognizing the constitutive activation of STAT5 as a good prognostic factor is out of our expectation. However, in a number of different studies, activated STATs have been reported to play a role in differentiation and apoptosis (Bromberg and Darnell, ). Activated STAT3 has been proposed to facilitate mammary gland involution by inducing extensive epithelial apoptosis through upregulating IGFBP- (Chapman et al, ). Transfection of a constitutively activated STAT5 mutant into an IL--dependent Ba/F3 cell line induces expression of bcl-xL and pim- and renders the cell line factor-independent. However, IL- treatment of the factor-independent cell line resulted in apoptosis within h, or differentiation followed by cell death. This apoptosis might have been due to the concomitant upregulation of JAB/SOCS-/SSI- and p21 by the super-active STAT5 signaling (Nosaka et al, ). Thus, differences in the fate of cells might be determined by the activation intensity and duration of specific STATs and may be cell-type specific. It has been shown that cellular STATs can also be activated through interactions between JAK3 and LMP- protein of EBV (Gires et al, ). LMP- can also upregulate expression of epidermal growth factor receptor (EGFR) (Miller et al, ), which may then modulate expression of specific STATs upon ligand stimulation (Petersen and Haldosen, ; Kloth et al, ; Leong et al, ). Interestingly, in line with our findings, despite the fact that LMP- of EBV is known for its transforming ability and promotes cellular proliferation in various studies (Baichwal and Sugden, ; Kaye et al, ), tumours with LMP- expression tend to have better prognosis in EBV-associated malignancies such as Hodgkin's disease (Montalban et al, ; Glavina-Durdov et al, ) and NPC (Hu et al, ). Regarding the intimate association between STATs and NPC, further clarification of the relation between STAT signalling and the gene expression pattern of EBV may shed light on the pathogenesis of this unique malignancy and might have future therapeutic applications.In conclusion, by using immunohistochemical methods, we demonstrated that constitutive activation of STAT3 and STAT5 was present in the majority of biopsy specimens from patients with NPC. Besides, constitutive coactivation of STAT3 and STAT5 was frequently noted in the same specimen. By Cox's regression analysis, activation of STAT5 is an independent factor that predicts better prognosis in NPC patients after radiotherapy.
|
PMC3004891.txt
|
TITLE: S. Typhimurium sseJ gene decreases the S. Typhi cytotoxicity toward cultured epithelial cells AUTHORS: A Nicole Trombert, Liliana Berrocal, Juan A Fuentes, Guido C Mora ABSTRACT: BackgroundSalmonella enterica serovar Typhi and Typhimurium are closely related serovars as indicated by >% DNA sequence identity between shared genes. Nevertheless, S. Typhi is a strictly human-specific pathogen causing a systemic disease, typhoid fever. In contrast, S. Typhimurium is a broad host range pathogen causing only a self-limited gastroenteritis in immunocompetent humans. We hypothesize that these differences have arisen because some genes are unique to each serovar either gained by horizontal gene transfer or by the loss of gene activity due to mutation, such as pseudogenes. S. Typhi has % of genes as pseudogenes, much more than S. Typhimurium which contains %. As a consequence, S. Typhi lacks several protein effectors implicated in invasion, proliferation and/or translocation by the type III secretion system that are fully functional proteins in S. Typhimurium. SseJ, one of these effectors, corresponds to an acyltransferase/lipase that participates in SCV biogenesis in human epithelial cell lines and is needed for full virulence of S. Typhimurium. In S. Typhi, sseJ is a pseudogene. Therefore, we suggest that sseJ inactivation in S. Typhi has an important role in the development of the systemic infection.ResultsWe investigated whether the S. Typhi trans-complemented with the functional sseJ gene from S. Typhimurium (STM) affects the cytotoxicity toward cultured cell lines. It was found that S. Typhi harbouring sseJSTM presents a similar cytotoxicity level and intracellular retention/proliferation of cultured epithelial cells (HT- or HEp-) as wild type S. Typhimurium. These phenotypes are significantly different from wild type S. TyphiConclusionsBased on our results we conclude that the mutation that inactivate the sseJ gene in S. Typhi resulted in evident changes in the behaviour of bacteria in contact with eukaryotic cells, plausibly contributing to the S. Typhi adaptation to the systemic infection in humans. BODY: BackgroundSalmonella enterica serovar Typhi (S. Typhi) is a human-restricted pathogen that causes enteric fever or typhoid. Salmonella enterica serovar Typhimurium (S. Typhimurium) is considered a broad host range pathogen that causes gastroenteritis in several warm-blooded animals such as calves and humans, but produces a typhoid-like systemic infection in mice [-]. Although the mechanism by which serovar Typhimurium causes gastroenteritis is well studied, less is known about the pathogenesis of the serovar Typhi. One limitation to the study of typhoid fever is the absence of a good animal model. For this reason, although the S. Typhimurium - mouse model has been used to infer S. Typhi important virulence mechanisms by the expression of S. Typhi genes in S. Typhimurium, the information derived from infection of mice is limited mainly because the virulence factors are tested in an heterologous system. Furthermore, S. Typhimurium does not cause typhoid in humans, suggesting that genetic differences between both serovars are crucial for disease development.The evolution of a broad host pathogen, such as S. Typhimurium, to a host-restricted pathogen, such as S. Typhi, might have occurred by (i) the acquisition of genetic material through horizontal gene transfer, (ii) genome degradation (i.e., the loss of genetic information by deletion or pseudogene formation) or (iii) a combination of both of these mechanisms [,]. The acquisition and persistence of DNA segments containing genes with pathogenicity or virulence functions (i.e., pathogenicity islands) will depend on the advantage they confer to the pathogen infectious cycle. Thus, bacteria with a great ability to colonise different environments, such as Pseudomonas aeruginosa, generally have larger genomes than those that survive in restricted niches [].The phenomenon by which a microorganism becomes adapted to its host involves the loss of genetic functions resulting in pseudogene generation, a process termed "reductive evolution". This process has been observed in human-adapted pathogens such as Shigella flexneri, Mycobacterium leprae and Salmonella Typhi [,]. For example, the loss of the ompT gene in Shigella confers a virulent phenotype by allowing bacteria to transmigrate across eukaryotic cells [,]. In the case of Salmonella, some serovars have accumulated mutations that enhance their survival within their respective hosts. For example the poultry-adapted S. Pullorum and S. Gallinarum serovars are non-motile because they have a point mutation in the flgK gene [,]. When S. Enteritidis and S. Typhimurium are isolated from infected poultry, these bacteria are frequently non-motile, suggesting that the niche occupied in birds can select against flagellation []. These non-motile S. Typhimurium strains have been shown to be non-virulent when used to infect mice. Thus, in the S. enterica, the adaptation to a particular vertebrate host seems to drive the loss of virulence factors for some serovars. The result of this adaptation may contribute to the narrowing of the host range and to the development of host specificity [].S. Typhi is an intracellular facultative pathogen that contains over pseudogenes, nearly % of its whole genome [,]. Several of the mutations that gave rise to these pseudogenes occur in systems related to pathogenicity mechanisms. For example, the S. Typhimurium sseJ gene encodes an effector protein regulated by Salmonella pathogenicity island (SPI-) [,]. SPI- regulated genes are related to bacterial intracellular trafficking and proliferation, and encode a protein complex known as the type III secretion system (T3SS). The T3SS mediates the injection of effector proteins from bacteria into eukaryotic cells [-]. These effector proteins modulate the S. Typhimurium endocytic pathway and allow the establishment of bacteria in a specialised vacuole termed the Salmonella-containing vacuole (SCV) []. Late stages of SCV synthesis include the formation of tubular membrane extensions known as Salmonella-induced filaments (Sifs). Sifs are thought to result from the fusion of late endocytic compartments with the SCV and their formation requires at least five SPI--dependent effectors: SifA, SseF, SseG, SopD2 and SseJ [-]. In this context, S. Typhimurium sseJ encodes an acyltransferase/lipase that participates in SCV biogenesis in human epithelial cell lines [,-]. The coordination of SseJ and SifA is required for bacterial intracellular proliferation []. Some studies have shown that SseJ is needed for full virulence of S. Typhimurium in mice and for proliferation within human culture cells [].S. Typhi lacks several effector proteins that are crucial for the pathogenicity of the generalist serovar S. Typhimurium []. The absence of these proteins could contribute to the specificity of the human-restricted serovars, and could play a role in evolutionary adaptation. In S. Typhi, sseJ is considered a pseudogene. In this work, we studied the effect of trans-complementing S. Typhi with the S. Typhimurium sseJ gene and assessed the phenotype in human cell lines. Our results show that the presence of the sseJ gene induces phenotypic changes in cytotoxicity and in intracellular proliferation of S. Typhi in human epithelial cell lines. Our results suggest that the loss of SseJ function contributes to the development of a systemic infection in S. Typhi.ResultssseJ is a pseudogene in S. TyphiTo assess whether the sseJ locus is a pseudogene in the serovar Typhi, we compared the available sequences of S. Typhi Ty2, S. Typhi CT18 and S. Typhimurium LT2 [,,]. We observed that the sequence corresponding to sseJ in S. Typhi is a ' partial remnant of bp, in contrast with the complete ORF found in S. Typhimurium ( bp). In order to corroborate these bioinformatics results, we designed a PCR assay with two sets of primers. The primers SseJ1Tym + SseJ2Tym yield a bp amplicon only when sseJ is complete, while the primers SseJRT1 + SseJRT2 yield a bp amplicon if the ' sseJ locus is present (Figure ). As shown in Table we observed a PCR product with the SseJRT1 + SseJRT2 primers in all the strains tested, including the reference strains (S. Typhi CT18, S. Typhi Ty2 and S. Typhimurium LT2) and S. Typhi clinical strains obtained from Chilean patients (STH collection). Nevertheless, we observed a PCR amplicon with the SseJ1Tym + SseJ2Tym primers only when the S. Typhimurium genomic DNA was used as template, strongly suggesting that the sseJ gene is an incomplete gene (i.e., a pseudogene) not only in the S. Typhi Ty2 and CT18 strains, but in all the Typhi clinical strains tested. To independently assess this hypothesis, we performed a Southern blot using the bp amplicon as a specific probe (Figure ). The annealing of the probe with the EcoRV digested genome of S. Typhimurium yielded a bp fragment, while in S. Typhi, we observed a bp fragment. As shown in Figure our data indicated that the presence of the pseudogene in S. Typhi CT18 is conserved in the S. Typhi clinical collection (STH). Therefore, the sseJ pseudogene seems to be a feature in serovar Typhi that distinguishes it from the serovar Typhimurium. S. Typhi STH007 presents no hibridisation with the probe, showing that this strain presents a larger deletion in the sseJ locus compared with other strains tested. S. Typhi STH2370 showed a slightly larger fragment than the other S. Typhi clinical strains presumably because of point mutations that changed the EcoRV restriction sites. Therefore, serovar Typhi has a genetic mutation in sseJ gene correlating with the previous studies made in strain CT18. We reasoned that the sseJ gene in the serovar Typhi is inactivated.Table 1PCR and Southern blot analysis of sseJ gene in S. Typhimurium vs. S. Typhi isolatesStrainPCR1460 bpPCR127 bpStrainsSerovar TyphimuriumATCC14028s++LT2++Serovar TyphiSTH2370-+STH001-+STH004-+STH005-+STH006-+STH007-+STH008-+STH009-+Ty2-+Figure 1Genomic organization of sseJ in S. Typhi and S. Typhimurium. The figure shows the annealing localization of the primers designed (small arrows), the recognition sites of EcoRV and the sseJ probe hibridisation site (thick black line labelled bp for S. Typhimurium and bp for S. Typhi). The data were obtained from S. Typhi CT18 and S. Typhimurium LT2 genomes, available in public databases www.ncbi.nih.gov.Figure 2Southern blot analysis of sseJ in S. Typhimurium and S. Typhi strain collection. Genomic DNA digested with EcoRV was electrophoresed on an agarose gel and analyzed by Southern. Bands correspond to S. Typhimurium sseJ gene (. Kb) or S. Typhi sseJ pseudogene (. Kb).S. Typhi harbouring the S. Typhimurium sseJ gene exhibits a decreased disruption of HT- polarised monolayersIf the loss of SseJ function in S. Typhi is advantageous for the interaction of bacteria with host cells, we should observe that wild type S. Typhi will present a different behaviour than the S. Typhi harbouring the S. Typhimurium sseJ gene when they are in contact with eukaryotic cells. This hypothesis was first tested by infecting polarised HT- monolayers with the strains under study using a modified transepithelial migration assay that included addition of gentamicin (after h of infection, see Materials and Methods) into the upper chamber (black arrow, Figure ). As shown in Figure the recovered CFU × ml- represented the bacteria which migrated to the lower chamber and survived the presence of the gentamicin that passed through the cell monolayer. If the integrity of the monolayer is disrupted by bacteria, gentamicin will leak through the lower chamber decreasing the recovered CFU × ml-. If the monolayer is not disrupted, the recovered CFU × ml- should remain essentially constant over the same time course. As observed in Figure the recovered CFU × ml- corresponding to S. Typhimurium 14028s presented a slight decline over the time course of the assay (white diamonds), suggesting that the monolayer integrity is not largely affected by bacteria. In contrast, the CFU × ml- of S. Typhi STH2370 recovered from the lower chamber abruptly decreased until they became undetectable, strongly suggesting that the gentamicin leaked into the lower chamber due to a monolayer disruption (black squares). When S. Typhi were complemented with the S. Typhimurium sseJ gene (sseJSTM) (in the pNT005 plasmid, see Materials and Methods), and used to infect the monolayer, we observed that the corresponding recovered CFU × ml- remained essentially constant, marking a sharp difference with the otherwise isogenic wild type strain and highly resembling the S. Typhimurium phenotype (compare the white diamonds and black triangles).Figure 3Cell permeability assay of S. Typhi and S. Typhimurium through H-T29 human cell line monolayers. (White diamonds) S. Typhimurium 14028s, (black squares) S. Typhi STH2370, (black triangles) S. Typhi STH2370/pNT005. The arrow indicates the time at which gentamicin was added. The results represent the average of three independent experiments. Each experiment was performed in duplicate. The values are expressed as the means ± SD of three independent experiments (asterisks represent p < .). The CFU × ml- numbers from infected cells with S. Typhi carrying empty plasmid (pSU19 or pCC1) showed no differences with respect to the wild type strain (data not shown).In order to independently assess whether S. Typhi harbouring the S. Typhimurium sseJ gene shows a decreased disruptive effect toward cultured cell monolayers than the wild type S. Typhi, we measured the transepithelial electrical resistance (TER). TER is a measure of the movement of ions across the paracellular pathway. Measurement of TER across cells grown on permeable membranes can provide an indirect assessment of tight junction establishment, stability and monolayer integrity []. As shown in Figure after h of infection wild type S. Typhi efficiently disrupted the monolayer as inferred by the lower TER measured compared with the control without bacteria. However, when HT- cells were infected with S. Typhi/pNT005, TER values were similar to those obtained with S. Typhimurium 14028s. This result indicates that S. Typhi/pNT005 was less disruptive on the monolayer than S. Typhi wild type, supporting the result shown in Figure . To discard a possible gene dosage effect by the vector copy number, we infected cells with S. Typhi/pNT006 (complemented with a single-copy vector harbouring sseJSTM) and the TER obtained was similar to that of S. Typhi/pNT005. This result demonstrated that the effect on cell permeability was due to the presence of sseJSTM and not to an artifact produced by gene dosage.Figure 4The presence of the sseJ gene in S. Typhi promotes the disruption of the epithelial monolayer. HT- cells were grown in transwells for - days. Polarised HT- cells were apically infected with the wild type S. Typhi or the respective complemented strains. TER h post-infection reported as a percentage of the initial TER value and is expressed as the means ± SD of three different experiments, each performed in duplicate. The percentages of TER values from cells infected with S. Typhi carrying each empty plasmid (pSU19 or pCC1) showed no differences with respect the wild type strain (data not shown).S. Typhi harbouring sseJSTM was less cytotoxic than wild type S. TyphiKops et al. demonstrated that S. Typhi Ty2 causes rapid death of some C2BBe cells in monolayers []. Because cell monolayer permeability may be increased due to cell death during infection, we wanted to assess whether the presence of sseJSTM in S. Typhi contributes to decrease cytotoxicity, as the results of the Figure and strongly suggest. Cell membrane damage due to cytotoxicity leads to the release of cytoplasmic enzymes, and the measurement of lactate dehydrogenase (LDH) release is a well-accepted assay to estimate cell membrane integrity and quantify cell cytotoxicity [,]. Then, the LDH release induced by S. Typhimurium, S. Typhi, S. Typhi/pNT005 or S. Typhi/pNT006 was compared. As shown in Figure we found that wild type S. Typhi STH2370 was the most cytotoxic strain among all bacteria tested. This result suggests that the SseJ effector protein decreased S. Typhi cytoxicity when bacteria interact with human cell lines, resulting in increased cell permeability.Figure 5Analyses of cytotoxicity HT- infected with complemented and wild type S. Typhi strains. HT- cells were grown in transwells for - days. Polarised HT- cells were apically infected with the S. Typhi wild type or the respective complemented strains. Released LDH was measured h post-infection and reported as percentage relative to the S. Typhi wild type. The values correspond to the means ± SD of three independent experiments, each performed in duplicate. The percentages of each S. Typhimurium 14028s, S. Typhi STH2370/pNT005 and S. Typhi STH2370/pNT006, have significantly differences respect S. Typhi STH2370 wild type. LDH release from infected cells with S. Typhi carrying empty plasmid (pSU19 or pCC1) showed no differences with respect to the wild type strain (data not shown).The presence of sseJSTM in S. Typhi increased bacterial intracellular retention/proliferation within HEp- cellsIt has been reported that sseJ contributes to the intracellular proliferation of S. Typhimurium [,]. Moreover, the decreased cell death produced by the presence of sseJSTM in S. Typhi strains (Figure ) may lead to an increased proliferation of intracellular bacteria because of a decreased cytotoxicity. A less cytotoxic pathogen should be retained inside eukaryotic cells over time, allowing an increased bacterial proliferation. If this hypothesis is correct, S. Typhi carrying sseJSTM should exhibit increased CFUs in the gentamicin protection assay (see Materials and Methods). As expected, Figure shows that the presence of sseJSTM yielded a significantly increase in the CFUs recovered from the infected cells compared to the wild type.Figure 6Gentamicin protection assay of complemented and wild type strains of S. Typhi. HEp- cells were grown and infected with the S. Typhimurium 14028s, S. Typhi STH2370 or the respective S. Typhi complemented strains. The recovered CFUs were counted h post-infection. The values correspond to the means ± SD of three different experiments, each performed in triplicate. The CFUs recovered from infected cells with S. Typhi with each empty plasmid (pSU19 or pCC1) showed no differences with respect to the wild type strain (data not shown).DiscussionIn the process of adaptation to humans, genes no longer compatible with the lifestyle of S. Typhi within the host were selectively inactivated. These inactivated genes are called "antivirulence genes" and their loss of function results in the adaptation to a given host []. S. Typhi is a facultative bacterial pathogen that has accumulated a high number of pseudogenes (approximately % of the genome) and over % of them have completely lost their functions [,]. When compared with the genome of free-living organisms, facultative pathogens harbour several pseudogenes and a population structure that promotes the maintenance of the mutations. In this context, S. Typhi represents an intermediate step between obligate bacterial parasites and free living bacteria, exhibiting some genome erosion directed to inactivate and lose detrimental or non-essential functions for their environment (i.e. host) []. Thus, we hypothesized that the loss of some of these genes contributed to the adaptation of S. Typhi to the systemic infection.Our results suggest that the loss of the fully functional SseJ protein in S. Typhi contributed to the adaptation to the systemic infection by increasing bacterial cytotoxicity in epithelial cells. The increased cytotoxicity presented by S. Typhi compared with S. Typhimurium is not only related to the loss of functions, as we showed here with the sseJ pseudogene; but also to the acquisition of new functions. It has been reported that S. Typhi presents a pathogenicity island (named SPI-) that harbours hlyE. The hlyE gene encodes a cytolysin that has proved to be cytotoxic toward different cell types [-]. SPI- is shared by other Salmonella enterica serovars that have been shown to cause systemic infections in humans, but is absent from S. Typhimurium []. In addition, the functional transfer of the S. Typhi hlyE gene to S. Typhimurium promotes deep organ infection in mice []. All this evidence suggests that S. Typhi has been selected for an increased cytotoxicity inside its host in order to perform a successful systemic infection. Thus, an increased cytotoxicity toward the epithelial barrier may guarantee the development of a deeper infection and a decreased retention inside epithelial cells at the bacterial entry point.On the other hand, the presence of the sseJSTM gene in S. Typhi significantly enhances the retention time within epithelial cells and/or the intracellular proliferation as we showed in Figure in agreement with previous reports that indicate that SseJ enzymatic activitycontributes to intracellular replication in host tissues [,]. Accordingly, it is possible that the sseJ loss of function was selected in S. Typhi in order to promote a decreased retention/proliferation of bacteria inside the eukaryotic cells. It is known that the intracellular proliferation is essential for the virulence of S. Typhimurium []. Nevertheless, recent studies revealed that the magnitude of the CD8+ T cell response correlates directly to the intracellular proliferation in Salmonella enterica, showing that a reduced intracellular proliferation limits antigen presentation and development of a rapid CD8+ T cell response, indicating that reduced intracellular proliferation of virulent pathogens may be an important mechanism of immune evasion. []. Accordingly, Salmonella presents several responses directed to downregulate the intracellular proliferation, reinforcing the concept that a state of low proliferation within the host cell is strategy to enhance virulence in a determined niche []. Actually, it has been shown that Salmonella expands its population in the liver by increasing the number of infection foci rather than undergoing massive intracellular growth in individual host cells, where the bacterial spreading from the initial infection foci to nearby cells may be facilitated by inducing cytotoxic effects in the infected cells [,].How sseJSTM reduces the cytotoxicity in S. Typhi is not clear. It is known that the lipid imbalance associated to the presence of lipid alcohols, fatty acid and sterols is related to cytotoxicity and apoptosis [,]. Any process that limits the accumulation of these species is likely to be cytoprotective []. One such process involves the presence of different acyltransferase gene families that generate neutral lipids or steryl esters from these lipid alcohols []. SseJ, that presents glycerophospholipid: cholesterol acyltransferase (GCAT) activity in eukaryotic cells [], might plausibly contribute to the reduction of the lipid-associated cytoxicity. The precise mechanisms underlying this process is unknown, but one possibility is that the presence of sseJSTM in S. Typhi is affecting the lipid remodelling in the infected cells, in turn reducing the cytotoxicity.All our results together suggest that the loss of the sseJ gene in S. Typhi contributed to the adaptation to the systemic infection by increasing the bacterial-induced cytotoxicity and by decreasing the retention/proliferation inside the epithelial cells.ConclusionsBased on our results we conclude that the mutation that inactivate the sseJ gene in S. Typhi resulted in evident changes in the behaviour of bacteria in contact with eukaryotic cells, plausibly contributing to the S. Typhi adaptation to the systemic infection in humans.MethodsBacterial strains, media and growth conditionsThe S. Typhi and S. Typhimurium strains used in this study are described in Table . Strains were routinely grown in Luria-Bertani (LB) medium (Bacto Tryptone g × l-; Bacto Yeast Extract g × l-, NaCl g × l-) at °C, with vigorous shaking, or anaerobically by adding an overlay of μl of sterile mineral oil as a barrier to oxygen prior to invasion assays with cultured human cells. When required, the medium was supplemented with antibiotics at the following concentrations: chloramphenicol μg × ml-, ampicillin μg × ml- and kanamycin μg × ml-. Media were solidified by the addition of agar ( g × l- Bacto agar).Table 2Bacteria strains and plasmids used in this studyStrain or plasmidRelevant characteristicReference or SourceStrainsSerovar TyphimuriumATCC14028sWild-type strain, virulentATCCLT2Wild-type strainS. MaloySerovar TyphiSTH2370Clinical strain, virulentHospital Dr Lucio CórdovaSTH001Clinical strain, virulentHospital Dr Lucio CórdovaSTH004Clinical strain, virulentHospital Dr Lucio CórdovaSTH005Clinical strain, virulentHospital Dr Lucio CórdovaSTH006Clinical strain, virulentHospital Dr Lucio CórdovaSTH007Clinical strain, virulentHospital Dr Lucio CórdovaSTH008Clinical strain, virulentHospital Dr Lucio CórdovaSTH009Clinical strain, virulentHospital Dr Lucio CórdovaTy2Wild-type strainInstituto de Salud PúblicaPlasmidspGEM-TeasyHigh-copy-number cloning vectorPromegapCC1Single-copy vector, F plasmid derivedStratagenepNT002pGEM-Teasy carrying the S. Typhimurium sseJ geneThis workpSU19Medium-copy-number cloning vector[]pNT005pSU19 carrying the S. Typhimurium sseJ geneThis workpNT006pCC1 carrying the S. Typhimurium sseJ geneThis workConstruction of plasmidsThe sseJ PCR product was initially cloned into pGEM-T Easy (Promega) to yield plasmid pNT002, and the presence of the gene was confirmed by PCR amplification and restriction endonuclease assays. The DNA fragment containing the sseJ gene was obtained from pNT002 and cloned into the EcoRI site of the medium-copy number vector pSU19 [] to yield the plasmid pNT005. The presence of the gene and its promoter region was confirmed in all plasmids by PCR amplification and restriction endonuclease analyses. The PCR product was directly cloned in the pCC1 vector according to manufacturer's instructions (CopyControl™ PCR Cloning Kit, Stratagene) to yield the plasmid pNT006. The expression of sseJ gene from each plasmid was confirmed by Western blotting (data not shown).Bioinformatic analysesComparative sequence analyses were made with the complete genome sequences of S. enterica serovar Typhi strains CT18 (GenBank: AL627270.) and Ty2 (GenBank: AL513382), serovar Typhimurium LT2 (GenBank: AE006468.). The sequences were analyzed using the BLAST, alignment, and phylogeny tools available at http://www.ncbi.nlm.nih.gov/ and by visual inspection to improve alignments.PCR amplificationPCR amplifications were performed using an Eppendorf thermal cycler and Taq DNA polymerase (Invitrogen Cat. N° -). Reaction mixtures contained × PCR buffer, . mM MgCl2, each dNTP ( mM), primers ( mM), ng of template DNA, and U polymerase. Standard conditions for amplification were cycles at °C for seconds, °C for min and °C for min seconds, followed by a final extension step at °C for min. Template S. Typhi chromosomal DNA was prepared as described []. Primers SseJ1Tym (CATTGTATGTATTTTATTGGCGACG) and SseJ2Tym (AATCGGCAGCAAAGATAGCA) were used to amplify bp, and were designed from the S. Typhimurium LT2 sseJ reported sequence. The conditions for amplification of bp were cycles at °C for seconds, °C for seconds and °C for min, followed by a final extension step at °C for min. Primers SseJRT1 (GCTAAAGACCCTCAGCTAGA) and SseJRT2 (CAGTGGAATAATGATGAGCT) were designed from the S. Typhimurium LT2 sseJ reported sequence.Southern hybridisationsHybridisation probes for sseJ were generated by PCR amplification and were purified and labelled using the Detector™ Random Primer DNA Biotinylation Kit (KPL). Genomic DNA from Salmonella serovars was prepared as described by Maloy [], cleaved with EcoRV (Invitrogen) and the fragments were resolved on a .% agarose gel. The DNA was then transferred to a nylon membrane and cross-linked by UV irradiation. Hybridisation was performed according to the protocol described in the chemiluminescent system, using a DNA Detector™ HRP Southern Blotting Kit (KPL) and Kodak XAR- film.Cell permeability assayWe used an in vitro assay modified from the method described by McCormick []. Briefly, the colon carcinoma HT- cell line was grown to confluence (- days) on . μm pore-size filters ("transwells", Millicell®, Millipore) with glucose-free RPMI (Gibco). Each transwell was inoculated individually to the apical surface with μl of approximately × CFU ml- of bacterial cultures and immediately incubated for min at °C. After extensive washing with sterile PBS (NaCl .% w/v; KCl .% w/v; Na2HPO4 2H2O .% w/v; KH2PO4 .% w/v), the extracellular bacteria were killed by treatment of monolayers with gentamicin ( μg × ml-). Immediately after gentamicin treatment, the medium from basal compartment of the epithelial cell monolayer was collected and plated for colony forming units (CFU) to assess the number of bacteria that passed through the cell monolayer. The polarisation of cells was confirmed by transepithelial electrical resistance (TER) and transmission electron microscopy (data not shown).Transepithelial electrical resistanceTER was used to monitor changes in epithelial cell culture integrity. TER in HT- enterocytes was studied using an EVOM electrode (World Precision Instruments). The enterocytes were grown to confluence (- days) on . μm pore-size filters ("transwells", Millicell®, Millipore). The electrical resistance readings were recorded after subtracting the average resistance of two membranes in the absence of enterocytes at the beginning of the assay (t0) and h post-infection (t1). Controls included the incubation of the cells with EDTA and Triton X- (% PBS). The reading was expressed as percentages and calculated as follows:%TER(1h)=×(TER t0×TER t1−)We verified the HT- polarisation by TER and transmission electron microscopy.LDH Cytotoxicity AssayCytotoxicity of infected HT- cells was assayed using a lactate dehydrogenase (LDH) Kit (Valtek), which measures the extracellular release of LDH into the media by dead cells, according to the manufacturer's instructions. The absorbance values of treated cells were expressed as a percentage relative to the wild type S. Typhi after correcting for background from media without cells at nm.Gentamicin protection AssayTo measure bacterial invasion, the method described by Lissner [] and modified by Contreras [] was used. Briefly, HEp- monolayers ( × cells/well) were grown at °C in a % CO2/% air mixture in RPMIFS (RPMI medium supplemented with % fetal bovine serum pre-treated for min at °C). The tested bacterial strains were grown anaerobically to mid-exponential phase and then harvested by centrifugation prior to infect the monolayers in -well microtiter plates at a multiplicity of infection of :. After incubation of h to allow bacterial entry into the cells, monolayers were washed twice with phosphate-buffered saline (PBS), and μL of RPMI containing gentamicin ( μg × ml-) was added to each well. The plates were then incubated for h to kill any remaining extracellular bacteria. In the case of the strains carrying vectors, the medium was supplemented additionally with chloramphenicol during the entire assay. The medium was removed and cells were washed twice with PBS. Then, the cells were lysed with sodium deoxycholate (.% w/v, in PBS). The number of intracellular bacteria (CFU at t3) was determined plating onto LB agar plates with chloramphenicol (the strains carrying plasmid) or without antibiotic (the wild type strains). Quantitative invasion assay values were calculated as follows: h post infection index=×(intracellular CFU mL−1at t3×CFU mL−1added)StatisticsAll results are expressed as means ± SD of an individual experiment performed in triplicate. P values were calculated according to Student's t-test, and values p < . or p < . were considered statistically significant.Authors' contributionsAT: designed the studies, performed the experiments and wrote the manuscript; LB: performed the transepithelial electrical resistance experiment, contributing significantly in the development of the other experiments and in the preparation of manuscript; JAF: participated in writing the paper; GCM: designed the studies and participated in the revision of the manuscript. All authors read and approved the final manuscript.
|
PMC1831786.txt
|
TITLE: Photostability of commercial sunscreens upon sun exposure and irradiation by ultraviolet lamps AUTHORS: Helena Gonzalez, Nils Tarras-Wahlberg, Birgitta Strömdahl, Asta Juzeniene, Johan Moan, Olle Larkö, Arne Rosén, Ann-Marie Wennberg ABSTRACT: BackgroundSunscreens are being widely used to reduce exposure to harmful ultraviolet (UV) radiation. The fact that some sunscreens are photounstable has been known for many years. Since the UV-absorbing ingredients of sunscreens may be photounstable, especially in the long wavelength region, it is of great interest to determine their degradation during exposure to UV radiation. Our aim was to investigate the photostability of seven commercial sunscreen products after natural UV exposure (UVnat) and artificial UV exposure (UVart).MethodsSeven commercial sunscreens were studied with absorption spectroscopy. Sunscreen product, . mg/cm2, was placed between plates of silica. The area under the curve (AUC) in the spectrum was calculated for UVA (– nm), UVA1 (– nm), UVA2 (– nm) and UVB (– nm) before (AUCbefore) and after (AUCafter) UVart ( kJ/m2 UVA and kJ/m2 of UVB) and before and after UVnat. If theAUC Index (AUCI), defined as AUCI = AUCafter/AUCbefore, was > ., the sunscreen was considered photostable.ResultsThree sunscreens were unstable after min of UVnat; in the UVA range the AUCI was between . and .. In the UVB range one of these sunscreens was unstable with an AUCI of . after min. Three sunscreens were photostable after min of UVnat; in the UVA range the AUCI was between . and . and in the UVB range between . and .. One sunscreen showed in the UVA range an AUCI of . after UVnat but an AUCI of . after UVart. Five of the sunscreens were stable in the UVB region.ConclusionThe present study shows that several sunscreens are photounstable in the UVA range after UVnat and UVart. There is a need for a standardized method to measure photostability, and the photostability should be marked on the sunscreen product. BODY: BackgroundSunscreens give good protection against sunburn, actinic keratosis and squamous cell carcinoma. The results for preventing cutaneous malignant melanoma (CMM) and basal cell carcinoma are less conclusive [-]. One explanation for this can be that UVA radiation (– nm) plays a role for induction of CMM [] and that it is mainly in the UVA range the photodegradation of the sunscreen occurs. In the present work, commercially available sunscreens, containing organic chemical and/or inorganic chemical filters, have been exposed to natural UV (UVnat) as well as to artificial UV (UVart) in order to study their photostability.Previous studies have shown that some sunscreens lose part of their protection when exposed to UV radiation [-]. Several sunscreen producers claim that their products give good protection against both UVA and UVB radiation; however, the photostability of the product is rarely declared. This is also important for the consumer to know when choosing a sunscreen. Since it has been known for several years that some products may be photounstable, one would have expected a large improvement in the photostability of sunscreen products. Up to now, there is no standard method for determining photostability of a sunscreen [,,]. Neither is there an international standard method for measuring UVA protection, and several different systems are currently in use [-].The aim of this study was to investigate the photostability of commercial sunscreen products after UVnat and after UVart.MethodsSunscreensSeven commercial sunscreens were included, all available on the Swedish market. Three sunscreens contained only organic chemical filters, three sunscreens had a combination of inorganic and organic chemical filters, and one sunscreen contained solely inorganic chemical filters. In Table the photoactive compounds of the sunscreens and the Sun Protection Factor (SPF) of the product are shown.The sunscreen was weighed and placed between two plates of polished fused silica (quartz) with diameter mm and thickness mm. The amount applied was . mg/cm2. The absorbance was too high for proper measurements when the recommended amount of mg/cm2 was applied, causing distortion in the absorption spectrum. For this reason a thinner layer was applied. A previous study has shown that the result were independent whether an application thickness of or mg/cm2 was used [].Light sourcesFor UVA radiation, a UVASUN (MUTZHAS, Germany) was used. The output is mainly between and nm.For UV radiation (including UVB), an Esshå Corona Mini (Sweden), equipped with two fluorescent tubes, Philips TL ( W), was used. This is a broadband radiation source from to nm with a major peak at nm. There are strong mercury peaks at nm and nm.The irradiance at the exposure plane was measured with an International Light IL Radiometer/Photometer using a probe named SED for UVA and a probe named SED for UVB radiation. The fluence rate of the UVA lamp was W/m2 when measured from a distance of cm. Twenty minutes' exposure gave a dose of kJ/m2. This corresponds to the UVA dose that reaches the earth's surface during one sunny summer day in Gothenburg []. We also measured the spectral distribution of the UVB lamp. By combining the spectral distribution with the action spectrum of the probe, the fluence rate of the UVB radiation was . W/m2. Twenty minutes of exposure gave a dose of kJ/m2 UV radiation (including UVB). This corresponds to Standard Erythema Doses (SED) when further weighted by the CIE action spectrum []. This is a much higher dose than normal for one summer day in Gothenburg [] or what has been reported from Denmark []. In spite of that fact, the majority of sunscreens showed good stability in the UVB range. In Table the UV doses reported from the Swedish Metrological and Hydrological Institute (SMHI) are listed.For UVnat, samples were placed horizontally outdoors when the weather was sunny. This was done in early July in Gothenburg (latitude: ° ' N). The total exposure time was min (Table ) with measurements of the absorption spectra before exposure and after min, min and min of UVnat. SMHI measures the global irradiance in many places in Sweden and gives the CIE erythema weighted UV radiation as well (Table ).To eliminate the possibility that the degradation of the photoactive compounds could be caused by a temperature increase, control samples of sunscreen between silica plates were placed on a heating plate for minutes. The temperature was kept at °C ± °C, which was the same as that measured during exposure to the UVA lamp. This is about °C higher than the temperature of the skin. Spectra were recorded prior to and after heating. The temperature did not influence the degradation since the absorption spectra did not change after heating.SpectrometerIn all studies the spectra were recorded by a Cary spectrophotometer (Varian, USA). It is a two-beam spectrophotometer without integrating sphere, which measures the transmission by scanning over the wavelength range of interest. Without integrating sphere the measured absorbance includes also some scattered radiation. Therefore, the spectra of samples with inorganic filters, which scatter light, may show a too high absorbance.Area under the curve index (AUCI)The AUC for UVA, UVA1 (– nm), UVA2 (– nm) and UVB was calculated for each spectrum before (AUCbefore) and after (AUCafter) UVart ( kJ/m2 UVA and kJ/m2 of UV radiation (UVB included) and before and after UVnat. If the AUCI (AUCI = AUCafter/AUCbefore) was >., the sunscreen was considered photostable.The AUC was calculated with the following equation: ∑ λ min λ max A ( λ ) Δ λ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaadaaeWaqaaiabbgeabnaabmaabaacciGae83UdWgacaGLOaGaayzkaaGaeuiLdqKae83UdWgaleaacqWF7oaBcyGGTbqBcqGGPbqAcqGGUbGBaeaacqWF7oaBcyGGTbqBcqGGHbqycqGG4baEa0GaeyyeIuoaaaa@41B6@ where A is absorption and λ is wavelength. It was measured in steps of nm.For UVA λmax = nm and λmin = nm. The same calculation was done for each UV range respectively, before and after UVart and before and after UVnat.Maier et al. used the difference between the spectral transmission before and after a defined UV exposure, ΔT. A product was labeled photounstable if the mean photoinstability was higher than % ( mg/cm2 product was used) []. In our study we chose the AUCI instead. Since we used . mg/cm2 we considered the product photostable if the AUCI was higher than ..ResultsSunscreensThe photostability of the sunscreens tested varies considerably. The photounstable sunscreens start to degrade rather rapidly when exposed to the sun. After min of UVnat, Sunscreens and are unstable (AUCI <.). Sunscreens containing inorganic chemical filters are more photostable in our study than sunscreens with organic chemical filters with the exception of Sunscreens and 5Sunscreens , and are photostable after UVnat; in the UVA range the AUCI was between . and . after min and between . and . in the UVB range. Sunscreen shows in the UVA range an AUCI of . after UVnat but . after UVart.Sunscreens , and are unstable. They show after min UVnat an AUCI between . and . in the UVA range and between . and . in the UVA1 range.Sunscreens , , , and are stable in the UVB region whereas Sunscreens and are not. During exposure, absorption ranges of Sunscreens , and are shifted towards shorter wavelengths (Fig. 1a–c).In Table the AUCI is presented.This is true for all three samples, after both UVnat and UVart. Sunscreen is more unstable after UVart than to UVnat (Fig. ).The spectra were normalized by dividing the maximum value of the spectrum before irradiation by itself, so that the peak value of the spectrum before irradiation was set to .The temperature was higher, during exposure to the UVart than during exposure to UVnat, but the temperature did not influence the absorption. Sunscreens , and were stable both after UVart and after UVnat. Sunscreens and were very little influenced by UV exposure (Fig. 3a–c). In agreement with findings from other studies, sunscreens with the UV filter combination ethylhexyl methoxycinnamate (EHMC) and butyl methoxydibenzoylmethane (BMDBM) were unstable [,,,].DiscussionIn most cases UVnat compared to the UVart gave qualitatively similar results. However, UVnat, with a lower fluence rate than UVart, gave similar yields of degradation. In addition, the fluence rate of the UVAart was higher than that of the UVAnat, which could be expected to degrade the sunscreens faster. But this is not the case, except for Sunscreen . Since the dose of UVAart was higher than UVAnat, Sunscreen probably provides sufficient protection for the consumer.Commercial sunscreens generally have low viscosity in order to be easy to apply. The temperature increase of the samples during UV exposure, especially after UVart, may lower the viscosity further. This may result in reductions of the optical path lengths of the samples. However, this was not the case in our study since samples kept on a heating plate for min at °C showed a similar spectrum before and after heating.Four of the seven sunscreens contain TiO2. If the particles are too small they may lose their scattering effect and, consequently, not give as good protection as larger particles. This may be the case for Sunscreen (Fig. 3a). Several other studies show that inorganic chemical filters are not always photostable [,-]. Our study indicates the opposite, but seven sunscreens are a quite small amount of material, so this finding should be interpreted with caution.When mixed with petrolatum, some sunscreens undergo degradation during exposure to UV radiation, especially in the UVA range []. This is also the case for one of the most frequently used UV filters BMDBM. This compound is included in six of the seven sunscreens studied here (Table ). Our results confirm the findings from other studies that sunscreens containing the combination of EHMC and BMDBM are photounstable, regardless of what other UV filters they contain [,].Some manufacturers of sunscreens claim that commercially available sunscreens are photostable because the photoactive species are in a vehicle that stabilizes them. This claim does not seem to be correct in several cases. There are several studies about how improvement of photostability may be obtained, e.g. with nanoparticle encapsulation of EHMC [], liposphere preparation of BMDBM [] or a combination with diethylhexyl syringylidene malonate and BMDBM []. These findings are very interesting and will hopefully lead to an improvement in photostability in commercial available products.When sunscreens without metallic oxide particles are compared, Sunscreen seems to be more rapidly degraded than BMDBM dissolved in petrolatum. Not only does the UVA protection decline after exposure, but also the UVB protection. EHMC is one of the two UVB-absorbing filters present in Sunscreen , and the only one in Sunscreen . EHMC dissolved in petrolatum is rather photounstable []. The vehicles of Sunscreens and are nearly identical. The UVA-absorbing compound benzophenone- (BZ-) is added in Sunscreen . The presence of this compound may stabilize BMDBM, in agreement with earlier findings []. Another stabilizer that may work is anisotrizine (CAS no --)is []; however, that compound was not included in any of the products in this study. Sunscreen also has a higher SPF. However, a degradation manifesting itself in the UVA1 region should be noted.Sunscreen is photostable but does not contain any metallic oxide particles. This may be due to a vehicle that successfully prevents degradation and/or due to microstructures of the emulsion itself (Fig. 3a). It is interesting to compare this spectrum with that of Sunscreen (Fig. 1c) which, according to the list of contents, includes TiO2 particles but does not show the scattering slope. The size of the particles may be too small ( nm, according to the producer) to influence the absorption spectrum in the visible range. Small particles of TiO2 are expected to give maximal scattering in the UVB or UVC region. Larger particles can cause significant scattering also in the UVA and visible region. It follows that the small nanoparticles cannot give good protection in the UVA region in this case.The peak between and nm in the absorption spectra of Sunscreens , and (Figs. 1c, , 3a) can be attributed to BMDBM. In view of this it should be noted that the UV exposure makes the products react quite differently. In Sunscreen the BMDBM peak almost vanishes totally after min of UVnat, while the peaks in the other two sunscreens are more stable. We suggested above why there could be degradation in the UVA range in Sunscreen despite the presence of TiO2 particles. The reported stabilizing effect of -Methylbenzylidene camphor (MBC) [] does not manifest itself in the case of Sunscreen .The photostable Sunscreen contains, in addition to BMDBM and TiO2, a third UVA absorber, terephthalylidene dicamphor sulfonic acid (TLDCSA), which can stabilize BMDBM. It has also been shown that TiO2 may stabilize ketoprofen and may be used in protecting photounstable species [].Many commercial sunscreens give, according to the manufacturers, good UVA and UVB protection. However, the photostability of the sunscreen in the UVA range is not always adequate. Most sunscreens offer good protection against UVB while the UVA photostability of some products decreases substantially during UV exposure. The potential toxicity of the photoproducts also needs to be investigated further.For the consumer it is very difficult to know what product to choose, since the photostability varies between different brands and the photostability is not marked on the bottle. To know which photoactive compound the sunscreen contains is not good enough. The stability also depends on factors like preservatives, oxygen radical scavengers, and base formulation. It is not reasonable that the ordinary consumer should have knowledge of this. If the product claims to give broadband protection, this protection should remain also after sun exposure. The fact that sunscreens are photounstable has been known for many years. Our study clearly shows that there are still many photounstable products on the market. When buying a sunscreen, the consumer should automatically receive a photostable product.ConclusionThe present study shows that several commercially available sunscreens are not photo stable. Degradation is clearly manifested in the absorption region in the UVA range after solar irradiation. In general, sunscreens with TiO2 particles seem to be more photostable, with Sunscreens and as exceptions. Special focus should be on the commonly used UVA absorber BMDBM. In three out of six sunscreens in our study this molecule was degraded during UV exposure. Stabilizers of BMDBM may work, but not under all conditions. There is a need for a standardized method to measure photostability and the photostability should be marked on the sunscreen product.Competing interestsThe author(s) declare that they have no competing interests.Authors' contributionsHG and NT-W have made contributions to conception of design and interpretation of the data. They carried out the experimental set-up during the absorption studies and wrote the main part of the manuscript.BS carried out some of the absorption studies and drafted the manuscript.AR have made substantial contributions to conception of design, interpretation of the data and drafting and revising the manuscript. AR also participated in the coordination of the study.AJ, JM, OL and A-MW have made substantial contribution to concept of design and drafting and revising the manuscript critically.All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:
|
PMC3072923.txt
|
TITLE: Adverse events associated with acupuncture: three multicentre randomized controlled trials of cases in China AUTHORS: Ling Zhao, Fu-wen Zhang, Ying Li, Xi Wu, Hui Zheng, Lin-hao Cheng, Fan-rong Liang ABSTRACT: BackgroundIn order to evaluate the safety of acupuncture in China objectively, we investigated the adverse events associated with acupuncture based on three multicentre randomized controlled trials (RCTs) to assess the safety of acupuncture, identifying the common types of acupuncture adverse events, and analysing the related risk factors for their occurrence.MethodsThis observational study included patients who received acupuncture from three multicentre RCTs respectively for migraine, functional dyspepsia and Bell's palsy. The patients and their acupuncturists documented adverse events associated with acupuncture after treatment. We collected data about adverse events due to acupuncture treatment from their case report forms. We analysed the incidence and details of the adverse effects, and studied the risk factors for acupuncture adverse events with non-conditional logistic regression analysis.ResultsAmong the patients, patients (.%) suffered at least one adverse event throughout the treatment period. We did not observe the occurrence of serious adverse events. patients with adverse events recovered within weeks through effective treatment such as physiotherapy or self-treatment. A total of patients withdrew because of adverse events. There were types of adverse events related to acupuncture, including subcutaneous haematoma, bleeding, skin bruising and needle site pain. Subcutaneous haematoma and haemorrhage in the needling points were the most common adverse events. Age and gender were related to the occurrence of acupuncture adverse events. The older the patients were, the higher the risk of adverse events was. In addition, male patients had slightly higher risk of an adverse event than female patients.ConclusionsAcupuncture is a safe therapy with low risk of adverse events in clinical practice. The risk factors for adverse events (AEs) were related to the patients' gender and age and the local anatomical structure of the acupoints. AEs could be reduced and mitigated by improving the medical environment, ensuring a high technical level of the acupuncture practitioners and establishing a good relationship of mutual trust between doctor and patient.Trial RegistrationClinicalTrials.gov: NCT00599586, NCT00599677, NCT00608660 BODY: BackgroundAcupuncture has been practiced in China for thousands of years as part of the Traditional Chinese Medicine (TCM). More recently, it has gradually won acceptance in western countries as an alternative or complementary treatment for various conditions, including chronic pain syndrome [-]. Because of its widespread use, the safety of acupuncture is of increasing importance. Most side-effects encountered in acupuncture practice are mild, but life-threatening adverse events such as organ injury have been reported []. In the past few years, AEs associated with acupuncture treatment in foreign countries have been published in the form of systematic reviews or surveys [-], but acupuncture-related AEs that occurred in China have not received attention as many Chinese practicing acupuncturists do not regularly monitor and report AEs. In this paper we report AEs due to acupuncture in cases of patients and analyse the risk factors for their occurrence, based on three multicentre RCTs of acupuncture, to provide an objective evaluation of acupuncture safety in China. This study is expected to be useful for acupuncture practitioners as well as educators and health care decision-making departments.MethodsPatients1968 outpatients came from three multicentre RCTs in China. These three RCTs are "An RCT to treat migraine with acupuncture" (MI-RCT), "An RCT to treat functional dyspepsia with acupuncture" (FD-RCT), and "An RCT of acupuncture and moxibustion to treat Bell's palsy according to different stages" (BP-RCT).Three RCTs were registered respectively (NCT00599586, NCT00599677, and NCT00608660) at the U.S. clinical trial registry (http://clinicaltrials.gov/). According to the criteria of the International Classification of Headache Society, patients who met the diagnosis of migraine with or without aura were included in the MI-RCT study. In the FD-RCT study, patients diagnosed with meal-induced dyspeptic symptoms and epigastric pain according to the Rome Ⅲ Diagnostic Criteria for functional Gastrointestinal Disorders. Patients meeting the diagnosis of Bell's palsy according to Chinese Medicine and Western Medicine were enrolled in the BP-RCT. The three RCTs were completed in hospitals located in five provinces (Sichuan Province, Hunan Province, Hubei Province, Tianjin and Shandong Province). cases of patients were monitored from December to October .DesignThree RCTs were performed in accordance with the principles of the Declaration of Helsinki (Edinburgh version), and trial protocols were approved by the Affiliated Hospital of Chengdu University of Traditional Chinese Medicine review board and ethics committee. The National Clinical Trial Centre of Chinese Medicine, Chengdu, Good Clinical Practice (GCP) Centre in China was responsible for the central randomisation and data management. The inclusion criteria and protocols of the three RCTs have been described in detail elsewhere [-].AE is defined as an unfavourable medical event that occurs during or after the treatment regardless of causal relationship []. Serious adverse effects (SAEs) refers to those that caused hospitalisation, extended duration of hospitalisation, disability, impaired ability to work, death or were life threatening, resulting in events such as congenital malformations in the process of the clinical trials. AE and SAE were defined a priori from the literature and the State Food and Drug Administration (SFDA) in China. AEs include subcutaneous haematoma, minor haemorrhage, serious pain, fainting and local infection, and SAEs include spinal cord injury, punctured organs, convulsions and pneumothorax. All the definitions were expounded in the clinical work manual of each trial.During the clinical trial, physicians and patients were asked to evaluate AEs/SAEs associated with acupuncture and were recorded in the case report form (CRF). After each acupuncture treatment cycle, each patient was asked to complete a questionnaire (see additional file : Adverse Events Questionnaire for Patients) about whether they suffered excessive pain in the needle points, nausea, dizziness, aggravation of illnesses or other discomforts during the treatment, and were evaluated if there was bleeding, haematoma, bruising, acupuncture fainting, lag needle, broken needle, forgotten needle, etc. If any of the above adverse reactions happened, doctors should record them in the AEs reports (see additional file : Adverse Events Reports for Acupuncturist) including the type of AE, when it occurred, how long it lasted, the severity, location, any remedial actions and when they were alleviated, the relevance to acupuncture therapy and whether the patients quitted the trial. Monitors, appointed by the trial organisers, directly tracked AEs and periodically verified the integrity and authenticity of records for quality control [].InterventionsPatients in the three RCTs were blind selected to which acupuncture treatment they received. All enrolled patients received a -week acupuncture treatment and a -month follow-up period. After randomisation, patients received treatments over a period of weeks, and the treatments would be administered once per day for continuous days, with a day rest interval. The treatment courses of the three RCTs were designed according to the acupuncture protocol used in Chinese clinical acupuncture practice. All acupuncture practitioners had undergone either at least years of acupuncture training and were qualified TCM doctors or associate chief TCM doctors with more than years of clinical experiences. Acupoints and manipulation procedures in MI-RCT and FD-RCT were standardised, and BP-RCT was performed using semi-standardised acupoints and standardised manipulation. Details of acupoints selected and manipulation in each trial were shown in additional file (Details of acupoints selected and manipulation in each trial) and acupoints locations was shown in Figure . Sterile disposable one-time-use needles (Hwato Needles, Sino-foreign Joint Venture Suzhou Hua Tuo Medical Instruments Co., China) were used to achieve "de Qi" sensations. Patients in MI-RCT, FD-RCT and group of BP-RCT received electro-stimulation and the other patients were stimulated manually.Figure 1Acupoints Locations. Locations of acupoints used in the three randomized controlled trials.Data monitoring and StatisticsAccording to standard operating procedures (SOPs) for clinical research data management, clinical trial data entry and management were commissioned by the GCP centre of Chengdu using Remote Clinical Data Management Systems (RCDMS). All the results were entered twice and then checked. For inconsistent values we checked the CRFs item by item in order to ensure data accuracy. After data entry, CRFs and data from the database of each trial were randomly selected and checked again to ensure consistency. After the final confirmation, all the data were imported into SPSS .. After the logic programming check, the obviously incorrect data were modified. If the errors of data were from CRFs, the data manager then revised the data according to the answers from the researchers.The SPSS . for Windows (SPSS Inc., Chicago, IL) was used for data analysis. Measurement data were indicated as mean ± SD. The acupuncture AEs related risk factors were analysed by unconditional logistic regression analysis. P value was noted as significant when it was less than ..ResultsGeneral characteristics of patientsA total of cases of patients (MI-RCT , FD-RCT , and BP-RCT ) received acupuncture therapy in the three RCTs. The average age was . ± . years. The general characteristics of patient per trial were shown in Table .Table 1General Characteristics and Distribution of PatientsCharacteristicClinical TrialsTotal(%)MI-RCTFD-RCTBP-RCTGenderMale82173474729 (.)Female3934204261239 (.)MarriageUnmarried134191211536 (.)Married3414026891432 (.)Educational BackgroundPrimary school343990163 (.)Middle school206213405824 (.)University and above235341405981 (.)Acupuncture experienceNot experienced3465884911425 (.)Have experienced1295409543 (.)BMI (kg/m2)< . (.).~. (.).~. (.)Age (year)~ (.)~ (.)~ (.)~ (.)~ (.)AEs and sequent treatments74 patients (.%) experienced AE during the observational cycle. The general information of the patients with AEs was shown in Table . Overall, cases of AEs were recorded. The most common adverse effects were needle-site bleeding (.%) and subcutaneous bleeding (.%) (Table ). All types of AEs observed in this study have been reported previously [,,]. It was worth mentioning that the following adverse events, lag needle, broken needle and forgotten needle, were not reported, and there were no SAEs relating to organ injury or nerve injury caused by malpractice.Table 2General characteristics of patients with AEs associated with acupunctureCharacteristicsClinical TrialsTotal (%)MI-RCTFD-RCTBP-RCTGenderMale611219 (.)Female29101554 (.)MarriageUnmarried6129 (.)Married30102565 (.)Educational BackgroundPrimary school62210 (.)Middle school1231429 (.)University and above1861135 (.)Acupuncture experienceNot experienced25111854 (.)Have experienced110920 (.)BMI (kg/m2)< . (.).~. (.).~. (.)Age (year)~ (.)~ (.)~ (.)~ (.)~ (.)Table 3List of AEs in acupuncture and subsequent treatmentsType of AEsTotal patients with AEs(N = )Physiotherapy(N = )Self-treatment(N = )Others(N = )No treatment(N = )MI-RCT n (%)FD-RCT n (%)BP-RCT n(%)MI-RCT n (%)FD-RCT n (%)BP-RCT n(%)MI-RCT n (%)FD-RCT n (%)BP-RCT n(%)MI-RCT n (%)FD-RCT n (%)BP-RCT n(%)MI-RCT n (%)FD-RCT n (%)BP-RCT n(%)Local reactionsSubcutaneous haematoma6 (.) (.) (.) (.) (.) (.) (.) (.)Minor haemorrhage in needling position23 (.) (.) (.) (.) (.) (.)000000Subcutaneous bruise6 (.) (.) (.) (.) (.)00Prolonged pain at the site of needling01 (.) (.)0000000000Systemic reactionsAcupuncture fainting01 (.) (.) (.) (.)000Abdominal distension01 (.) (.)0000000000Dizziness/vertigo01 (.) (.)0000Leg weakness1(.) (.)00000000000Muscle spasm01 (.) (.) patients (.%) received subsequent treatment for recovery from AEs, including physiotherapy (, .%), self-treatment (, .%) and other effective methods (, .%) (Table ). patients with AEs recovered in weeks after subsequent treatment. One patient with dizziness did not fully recover within weeks but reached full recovery one month after the cessation of acupuncture treatment. Of these patients, three patients suffered either muscle spasm or dizziness or abdominal distension and withdrew from the study. The remaining patients completed the entire treatment process. of these AEs cases were certainly due to the acupuncture treatment, and the remaining cases were probably related to the acupuncture treatment.Logistic regression analysis of AE risk factors related with acupuncture treatmentWe analysed various potential risk factors associated with acupuncture treatment using logistic regression. We set the occurrence of AEs associated with acupuncture as a dependent variable (two categories, "" for no AE, "" for AE), and age (continuous variable), sex (two categories, "" for male, "" for female ), education (multi-categorical variables, "" for primary school education level, "" for high school education level, "" for university education level and above ), history of acupuncture treatment (two categories, "" for without acupuncture experience, "" for with acupuncture experience) and body mass index (continuous variable) were set as covariates. All the variables were adopted into the model for analysis by using forward stepwise selection of independent variables for Wald Probability and Statistics Act for Logistic Regression analysis. Logistic regression equation was: Logit (AE) = -. + . * age + (-. * gender); regarding the relative risk factor (Exp (B)) value terms, with one year increase, the possibility of AEs would increase .- = . times. From the perspective of gender, the possibility of AEs for females is .- = -. times that for males, i.e. the possibility of AEs for females is . times less than for males (Tables and ).Table 4Variables not included in the analysis of the modelScoredfpStep 1VariablesGender ()..044Educational Background1..565Educational Background ()..334Educational Background ()..422Acupuncture experience ()..326BMI2..104Overall Statistics7..192Step 2VariablesEducational Background0..662Educational Background ()..456Educational Background ()..436Acupuncture experience ()..274BMI1..233Overall Statistics3..496Table 5Logistic Regression analysis for variables included in the ModelBS.E.WalddfPExp(B)% C.I. for EXP(B)LowerUpperStep 1aAge0.......034Constant-.....019Step 2bAge0.......034Gender()-.......991Constant-.....023Notes: a. Variable(s) entered on step : age; b. Variable(s) entered on step : gender.DiscussionWith the increasing acceptance of acupuncture in more and more countries, governments and professional institutions, the safety of acupuncture is becoming of key concern in public discussion. Every acupuncture practitioner should objectively and factually report acupuncture AEs. According to the reports from United States, Germany, Britain, Korea, Japan et al [,-], the incidence of AEs ranged from .% to .%, with the most common acupuncture AEs being pain, fatigue, bleeding and haematoma. In this study, in order to ensure real and objective safety evaluation of acupuncture, we used a synchronous mode of observational studies to obtain the first-hand research data. The patients and their physicians were required to complete an acupuncture AE questionnaire and acupuncture adverse reports respectively. They were supervised and spot-checks were made during the whole process to minimize omission and selection bias. Since after-effects of acupuncture are known to exist, the study period to evaluate AEs included the acupuncture treatment process and the three months after treatment.According to the type and frequency of AEs in acupuncture treatment established by Witt [], ecchymoma and haemorrhage in needling position were the most common AEs, with an occurrence rate of more than %. Our results were consistent with the Witt's. Previous literatures indicated there were kinds of AEs, but the incidence and types of AEs found in our study were lower than former records. Comparing relevant literatures, we inferred that acupuncture AEs were associated with following factors.Firstly, acupuncture AEs were related to the patient's own attributes. In this study, according to the results of logistic regression analysis, the occurrence of acupuncture AEs did not have significant correlation to BMI, educational background and acupuncture treatment experience of patients, but was correlated to age and sex. The older the patient, the higher was the risk of AE, and male patients had slightly higher risk of AE than female patients. The analysis of risk factors for acupuncture AEs has only rarely been studied in previous literatures, usually only case reports have been reported. We have done these preliminary studies but still need to collect more demographic information and details to further this investigation.Secondly, the types of acupuncture AEs were connected to the location and anatomical structure of the acupoints. In our study, the most common AEs of acupuncture were ecchymoma and bleeding in the needling position. From the occurrence percentage of the two types of AEs, subcutaneous haematoma accounted for / (.%) in acupuncture treatment for Bell's palsy, and bleeding in the needling position accounted for / (.%) in the therapy of migraine. The treatment plan of peripheral facial paralysis involved Yangbai (GB14), Dicang (ST4), Jiache (ST6), Xiaguan (ST7), Taiyang (EX-HN5), Quanliao (SI18) and other facial acupoints. As facial skin is thin and rich in blood vessels and superficial fascia are composed of loose connective tissue with plentiful vessels, it is easy to cause vessel damage or latent subcutaneous bleeding during needle manipulation. In the treatment of migraine, we chose Fengchi (GB20), Luxi (TE19), and Touwei (ST8) which are located on cranial region of head. As the cranial skin and superficial fascia have abundant blood circulation, and fibrous connective tissue under the superficial fascia and the vessel wall are closely linked, scalp bleeding is common after acupuncture and it is difficult to self-terminate as pressure needs to be applied to stop the bleeding. Thus, mastering the anatomical characteristics of acupoints accurately could not only prevent the occurrence of AEs but also contribute to treating adverse reactions actively and minimize the harm to patients.Thirdly, AEs were related to the medical environment. The patients of three RCTs were from hospitals in five provinces. These medical institutions are all public medical institutions and are top hospitals in China. All patients received acupuncture treatment using disposable needles in order to avoid cross-infection such as needle site infection, the emergence of sepsis and hepatitis, which can be caused by defective needle disinfection [].Fourthly, we inferred that the incidence of acupuncture AEs and acupuncture practitioner's experience and skills are probably and inversely related. In public hospitals of China, acupuncture physician should have obtained qualifications for a doctor's license before treating patients. This means that an acupuncture physician has received at least years university education and at least year clinical experience, not including the period of internship. The education background and working experience of all acupuncture practitioners in these three RCTs were very similar, they had undergone at least years of acupuncture training and were qualified TCM doctors. Additionally, before the beginning of these clinical trials, all the acupuncture practitioners had received SOPs training courses of acupuncture operation plan. Only the acupuncturists who finished the training course and passed the qualification examination were able to take part in the clinical trial. During the process of acupuncture treatment, every trial unit was required to have at least two acupuncturists with more than years of clinical experience, as technical advisors to provide guidance to young practitioners. We made the above inference by comparing the education background and qualifications of acupuncture physician reported in most of the foreign countries [,,] (e.g. Modlock et al. described that midwives were trained in acupuncture according to the guidelines and performed acupuncture treatments five to six times a week) with those in our study. The level of training may be one of the reasons for the difference in frequency and types of acupuncture AEs between China and foreign countries.Lastly, we considered that the mutual trust between doctors and patients could reduce the occurrence of AEs. The level of communication between doctors and patients during the process of clinical trials could be important so that all the patients were able to follow the doctor's instruction, for example, do not receive acupuncture treatment in a fasting state or satiated state. To some degree, such measures could help reduce acupuncture syncope and aggravation of the disease caused by improper care.The limitation of this study was it involved only migraine, functional dyspepsia, and Bell's palsy, without relating to other appropriate disease in acupuncture, and two-thirds data about acupuncture safety in this manuscript were from the trials for efficacy study.ConclusionsOur study confirmed that acupuncture is a safe therapy. The risk factors for AEs were related to the patient's gender and age and the local anatomical structure at the acupoints. AEs could be reduced and mitigated by improving the medical environment, promoting the technical level of the acupuncture practitioners, and establishing a good relationship of mutual trust between doctor and patient. As acupuncture practitioners, we should be alert to the occurrence of acupuncture AEs. Once AEs taking place, we ought to handle and report them with scientific, rigorous and serious attitudes. Acupuncture training institutions, educators and medical decision-making departments should attach great importance to the provision of professional standards and skill training for acupuncture practitioners in order to minimize the risk of AEs.List of abbreviationsRCT: Randomized controlled trial; AE: Adverse event; TCM: Traditional Chinese Medicine; MI-RCT: An RCT to treat migraine with acupuncture; FD: Functional dyspepsia; FD-RCT: An RCT to treat functional dyspepsia with acupuncture; BP-RCT: An RCT of acupuncture and moxibustion to treat Bell's palsy according to different stages; GCP: Good Clinical Practice; SAE, Serious adverse effect; SFDA: the State Food and Drug Administration; CRF: Case report form; SOP: Standard operating procedures; RCDMS: Remote Clinical Data Management SystemsCompeting interestsThe authors declare that they have no competing interests.Authors' contributionsLZ drafted the manuscript. FWZ performed the data analysis. All authors contributed to the further writing of the manuscript as well as read and approved the final manuscript.Supplementary MaterialAdditional file 1Adverse Events Questionnaire for PatientsClick here for fileAdditional file 2Adverse Events Reports for AcupuncturistClick here for fileAdditional file 3Details of acupoints selected and manipulation in each trialClick here for file
|
PMC3001692.txt
|
TITLE: Making sense of health information technology implementation: A qualitative study protocol AUTHORS: Rebecca R Kitzmiller, Ruth A Anderson, Reuben R McDaniel ABSTRACT: BackgroundImplementing new practices, such as health information technology (HIT), is often difficult due to the disruption of the highly coordinated, interdependent processes (e.g., information exchange, communication, relationships) of providing care in hospitals. Thus, HIT implementation may occur slowly as staff members observe and make sense of unexpected disruptions in care. As a critical organizational function, sensemaking, defined as the social process of searching for answers and meaning which drive action, leads to unified understanding, learning, and effective problem solving -- strategies that studies have linked to successful change. Project teamwork is a change strategy increasingly used by hospitals that facilitates sensemaking by providing a formal mechanism for team members to share ideas, construct the meaning of events, and take next actions.MethodsIn this longitudinal case study, we aim to examine project teams' sensemaking and action as the team prepares to implement new information technology in a tiertiary care hospital. Based on management and healthcare literature on HIT implementation and project teamwork, we chose sensemaking as an alternative to traditional models for understanding organizational change and teamwork. Our methods choices are derived from this conceptual framework. Data on project team interactions will be prospectively collected through direct observation and organizational document review. Through qualitative methods, we will identify sensemaking patterns and explore variation in sensemaking across teams. Participant demographics will be used to explore variation in sensemaking patterns.DiscussionOutcomes of this research will be new knowledge about sensemaking patterns of project teams, such as: the antecedents and consequences of the ongoing, evolutionary, social process of implementing HIT; the internal and external factors that influence the project team, including team composition, team member interaction, and interaction between the project team and the larger organization; the ways in which internal and external factors influence project team processes; and the ways in which project team processes facilitate team task accomplishment. These findings will lead to new methods of implementing HIT in hospitals. BODY: BackgroundHospital-based health information technology (HIT) implementation research is needed to identify reproducible strategies to eliminate barriers to HIT use and promote its adoption and integration []. We found few studies of HIT implementation, and this absence may contribute to the slow and inconsistent adoption of HIT observed in hospitals []. This study will address two weaknesses identified in the literature on hospital-based HIT implementation: the absence of evidence about strategies to improve implementation and how to construct and manage project teams tasked with HIT implementation.In this study, we will prospectively examine a multidisciplinary project team as it prepares to implement a HIT system in a tertiary care hospital. Due to a lack of literature on project teamwork specific to HIT implementation, we rely on the general literature about hospital-based project teamwork. We will use sensemaking to explain the social processes embedded in large scale organization change [-], and qualitative methods to achieve the following aims: describe and compare sensemaking across multidisciplinary project teams whose members differ in terms of hierarchical role and discipline; describe how the sensemaking of multidisciplinary project teams changes over time; describe how multidisciplinary project teams' sensemaking influences the actions taken; and identify team member behaviors that facilitate or inhibit sensemaking of a multidisciplinary project team.HIT implementation literature falls into three categories: anecdotal case reports, effectiveness research, and research describing HIT impact in clinical settings. First, the majority of hospital HIT implementation literature is anecdotal and lacks systematic evidence for sound implementation interventions [,]. Second, HIT efficacy studies often discuss lessons learned; however these lessons were explanations of findings, rather than empirical observations [,]. Finally, generalizability of HIT impact studies is hampered by methodological concerns [,]. The majority of studies used retrospective, self reported data, focused mainly on HIT system users, usually physicians, and evaluated a single type of HIT system, such as provider order entry. Thus, best implementation methods remain largely unknown.HIT impact studies identified unanticipated social effects, such as reallocated work [], interrupted work processes [,], altered information exchange, communication patterns, and interpersonal relationships [-], and in some cases, patient harm [,]. Studies have also found that hospital staff member's perspectives about HIT processes for, and outcomes of, implementation varied by organizational identity [], role [], and work unit [], causing variation in action. Care providers who perceived HIT as a threat, resisted using the HIT system [-]. Those who saw HIT as a benefit to patient care, on the other hand, used the system and advocated for system improvements [,]. Thus, care providers varied implementation experiences combined with differing expectations, objectives, and needs may contribute to the slow and uneven adoption of HIT in hospitals.Project teams have not been well studied, even though they are responsible for implementing HIT. Project teamwork is a popular strategy that hospitals use to create change []. To develop solutions and anticipate consequences of change, hospitals populate teams with members with different experience, skill, and knowledge []. Diverse members bring new information to the team and they provide connections with others in the organization [,]. Thus, effective teamwork facilitates collaboration, coordinated effort, and task accomplishment []. Studies show that teams are usually better problem solvers than individuals perhaps because they represent the combined input of all members, or because team member interactions facilitate learning associated with shared expertise and social interaction [-].During HIT implementation projects, hospitals need access to various knowledge and skills to uncover interdependencies and critical expectations, and to determine actions. Research on healthcare project teams noted that diverse membership and positive interpersonal interaction was associated with team innovativeness and positive organizational outcomes [,,-]. However, studies found that diverse perspectives also created conflict that was linked to poor team performance []. It appears necessary to carefully manage relationships between people to achieve benefits of diversity. Interpersonal interaction and diversity of team membership, therefore, are an important focuses of the proposed study; specifically, we focus on sensemaking processes of the team.Theoretical framework: sensemakingSensemaking, a social process of searching for answers and meaning, drives the actions people take []. Sensemaking occurs through verbal discourse between hospital staff members. Whether planned or unplanned, change challenges people's ability to understand what is happening, to anticipate what will happen, and to know what steps to take [,,], suggesting that sensemaking processes may be more important than decision-making processes for successful change. Because HIT implementation in hospital settings does not occur in a linear fashion and includes unpredicted, unexpected outcomes, implementation team members cannot expect to make optimal decisions [,,]. They are forced to make 'good enough' choices and adjust as new information becomes available and understanding of circumstances changes []. When compared to traditional linear, process-focused perspectives on HIT-related change, such as decision making and diffusion of innovation [,], sensemaking may help us to better manage project team actions because it is a process that accounts for new information as events unfold and for social interaction and construction of meaning [-].Research on sensemaking in hospital studies suggests several things. Organizational role, such as nurse, physician, or manager, influences the sense that staff members make of events [-]. The sense that hospital staff members make influences their choices and actions [-]. Through discourse with other staff members, hospital staff members construct the meaning of information and events and shape and reshape their understanding as events unfold and new information becomes available [-]. Project teamwork provides a formal mechanism for enhancing sensemaking. Through dialog, team members share varying perspectives on team tasks, construct the meaning of events as the HIT is implemented, and take action in response to evolving meaning. Through sensemaking, team members define what is happening, jointly revise their understanding, learn, and problem solve, setting the course for HIT implementation [,,]. This view of sensemaking and the review of literature on project teams, thus serves as the basis for our methodological choices. Refer to Additional file for additional reference material used in developing this study.MethodsDesignWe propose a qualitative, longitudinal multiple case study through which we will examine the evolving sensemaking of three multidisciplinary HIT project teams using direct observation and organizational document review. We will follow the activities of these teams throughout the pre-implementation phase of the project. We defined this period as the time between team formation and the first time the HIT is used by hospital staff in the provision of care []. Through our choice of methods, we plan to address the following four weaknesses in prior research on hospital HIT implementation, project teams and sensemaking: retrospective data collection; reliance on self-reported perceptions of HIT implementation; focus on single participant identity; and focus on single work units.Following Institutional Review Board approval, we will contact the Chief Nursing Officer and Chief Information Officer to obtain permission to conduct the study. As an incentive, a consultation summarizing findings of the study with recommendations for future project teams will be provided to the organization and to the case study participants. Following a protocol described by Utley-Smith et al. [], the consultation will serve as a method of disseminating research findings directly to study participants in the form of a summary of research findings and some recommendations related to teamwork strategies for more effective sensemaking. Knowledge participants gain during the consultation may validate study participants' project experience, influence decisions to participate in future projects, and enhance participants' IT implementation skills []. Subjects in prior research have reported that they perceived a direct benefit from such consultations and recommendations [].SettingThis study will be conducted within a single, academic, tertiary care hospital in the southeastern United States. Consisting of beds in nursing units, the hospital has a highly complex, interdependent care environment where changes in care practices, such as HIT implementation, may result in unexpected consequences. The hospital decided to implement an HIT system, an electronic nursing documentation system, in its nursing units. Because of the anticipated impact of this system, the hospital is forming a multidisciplinary project team comprised of nine sub-teams. Each sub-team will be tasked with a different aspect of the HIT project and staffed with a cross-section of clinical disciplines and functional business team members. Using selection criteria described below, each sub-team selected for inclusion will represent a single case. We anticipate that project team members will have little history of working together; thus, the unique knowledge each member brings to the team's tasks may be largely unknown by other members of the team and team management processes will be necessary.Sample selectionSelecting case study teamsPrior research suggests that team members' perspectives on HIT implementation may differ based upon their departmental affiliation, professional training, organizational role, and hierarchical level [,-]. Further, a team's roles and responsibilities may shape the discourse, meaning, and actions taken during the project []. Thus, to capture how sensemaking is influenced by team member diversity, we will select sub-teams of the larger project implementation team for in-depth case study using two criteria: the sub-team has a broad scope of project responsibility within the larger project team and its members have different social identities. Three of nine project implementation sub-teams meet the criteria of broad responsibility and diverse membership and thus will be included in-depth case study: the executive team, the communication team, and the implementation team. The executive team (n = ) will include administrative and clinical executives and directors from multiple departments, and has a broad scope in that it will provide resources for the project and ensure alignment of project goals with organizational strategic goals. The communication team (n = ) will include administrative, clinical, and technical directors, managers, and staff representing many organizational levels, and has a broad scope in that it will produce all organizational communication about the project including minutes, articles, video, and web-based documents. The implementation team (n = ) will include directors, managers, and front-line staff from nursing units, pharmacy, information technology, and hospital education, also representing many organizational levels. This team has a broad scope in that it will collect information about care practices, identify unit level information and care needs, and recommend modifications to the system in support of those needs.The six sub-teams that will not be selected for in-depth case study include the steering committee, the neonatal development team, the psychiatric development team, the device selection team, the training team, and the informatics team. These teams will have narrower scopes of responsibility (e.g., selecting equipment), or their membership will be homogenous (e.g., all psychiatric nurses). To understand how the overall project is unfolding across the nine teams, however, we will collect published minutes from meetings held by the six teams not directly observed to include in analysis of documents. Further, during case study sub-team meetings, an update on the work of all nine teams will be summarized and presented. During the executive, communication, and implementation team meetings, we will capture this information in the field notes. Together, these documents and field notes will provide us with information about events and actions of other teams that we do not directly observe.MeasurementSensemakingSensemaking will be measured qualitatively using direct observations of the executive, communication, and implementation teams; and project document review. This approach will capture multiple perspectives and rich data on HIT implementation events [,,]. We derived a set of sensemaking behaviors from a literature review [,,,,], which we evaluated in a preliminary study, and used to developed an observation guide (Additional file : Appendix A []) intended to capture sensemaking and subsequent actions. We anticipate that through discourse in team meetings, members will share their unique knowledge (e.g., care processes within a department), their perspective on the HIT implementation, and their interpretation of information and events [] that will then direct their actions []. The observation guide will also facilitate documenting the actions the teams plan to take and their anticipated results as well as the teams' reflections on the actions taken. Questions on the observation guide included the following: What information do participants share and how do they share it (e.g., past experience, information from others, hypothetical scenarios)? What interpretations, labels, and conclusions do team members express? What new ideas, decisions or proposed actions will be taken and by whom? What form does the discussion take? And, how do team members interact with each other?The document review guide (Additional file : Appendix B) is designed to capture written discourse where the project team formally records and/or shares information with external constituents about the team's goals and actions taken related to the HIT implementation. Data collection will include date obtained, description of the document, date of event or contact associated with the document, significance of the document, and a brief summary of the contents.Participant demographic dataTeam member demographic data (Additional file : Appendix C) will be obtained by self-report at the time when participants are introduced to the study. Data collected will include current job title, current unit of assignment, tenure in their profession, tenure in the organization, tenure on the unit of assignment, highest educational level, highest educational level in the profession, technology experience, gender, age, and ethnicity/race. These data will be used to explore variations in sensemaking because studies indicate that these are individual characteristics that are likely to influence sensemaking [,,,].Data collection proceduresDirect observationWe will directly observe team meetings and activities (e.g., training sessions) throughout the study. During observations, we will observe and manually record verbal communications between team members, using field notes and jottings []. These notes will be typed directly into a laptop versus being handwritten on paper and transcribed at a later time []. We will also document observations, such as seating arrangement, note passing, and eye rolling. We will audio record the meetings to support the field notes and listen to the tapes to verify that the field notes accurately capture communications; the recordings will not be transcribed verbatim. All data will be tagged with date and time to capture emerging trends. Meetings will generally occur once a month and last approximately to minutes. Direct observations occur during regularly scheduled meetings pose minimal burden to participants. Field notes, jottings, and audio recordings are tantamount to meeting minutes. Electronic field notes will be formatted and imported into AtlasTI.DocumentsDocuments related to the project (e.g., articles) and project records (such as meeting minutes, presentations, policies and procedures, and flyers), will be maintained by the HIT project team in a Lotus notes database, published to the hospital intranet for review by hospital staff members, and published in organizational periodicals and newsletters. These documents are produced by the committee and reflect the way in which they wish to represent their work to external constituents. We will access documents electronically or in printed form from the intranet, and add them to the study database. Document date will be used to facilitate placement in and retrieval from the study database. Once formatted, documents will be imported into AtlasTI and summarized following the guide (Additional file : Appendix B). Documents will serve to corroborate and augment evidence from direct observation, or to contradict observational evidence [,].Participant demographic dataParticipant consent for use of demographic data will be obtained after we provide a review of the nature of the study, participant's role, confidentiality, and the associated risk/benefits of participation. Participants will complete the survey tool described earlier (Additional file : Appendix C). As new team members are added, we will follow the demographic data collection procedure. The survey will take approximately minutes to complete, posing minimal burden to participants. The demographic data will be entered into Microsoft Excel tables and accessed with SAS (v. .) for analysis.Data analysisWe will use qualitative analysis procedures recommended by Crabtree and Miller []. Our research team contains experts in health informatics; organizational cooperation and fairness; and organizational sensemaking and learning. As we develop hypotheses for each research aim, we will conduct research team discussions to uncover bias and propose alternate perspectives on emerging themes.Code developmentFrom the literature on sensemaking, we have developed an a priori set of codes (Additional file : Appendix D). Coding reduces the data so that the data remain manageable, facilitating data clustering and laying a foundation for further analysis []. Through iterative review and ongoing discussion between RK and RA, we will refine the definitions of each a priori code. When a segment of text does not fit an existing code, we will ask, 'What's going on here?' 'What triggered this participant action?' 'What follows this participant action?' 'How might sensemaking explain what is happening?' Through this open-coding technique, we will further develop our codes. We will develop decision rules and definitions to guide the categorization of data, and record these in the electronic codebook [,]. To minimize the loss of meaning that may occur when reducing data, we will record all data transition steps and retain original raw data, including meeting audio recordings, until the study is complete.First, we will read the entire field note or document to get a sense of the whole and create an initial memo to capture our emerging impressions []. In a second reading, we will code units of text that described sensemaking events using our a priori codes. We will then create a second memo, summarizing initial ideas about the field note, documenting areas that need follow up []. Coded units will be sorted into categories and subcategories and analyzed for recurrent themes.To address our research aims, we will use within-case and between-case analyses. To describe and compare sensemaking across multidisciplinary project teams, data will be analyzed for each case study sub-team so that we can gain a rich understanding of the sensemaking of each individual team []. In the cross-case analysis, for this aim, we will organize each team's sensemaking themes into three separate data matrices and compare across teams to establish similarities and differences among the teams. Because the three project teams' members differ in professional and organizational identity, it is likely that the teams will differ in terms of the sense they make of new information and project events. To describe how the sensemaking of temporary multidisciplinary project teams changes over time, we will analyze themes in temporal sequence. Since sensemaking is shaped by experience, it is likely that sensemaking of the project teams' members will shift following significant events. To describe how multidisciplinary project team's sensemaking influences the actions taken by the teams, we will organize the data matrices by the actions of each team in order to identify the antecedents and consequences of these actions. Finally, to identify which team member behaviors facilitate and or inhibit the sensemaking of a multidisciplinary project team, we will use open coding guided by the literature on project teams. Some examples of codes may include respect or openness to ideas. The coded data will be analyzed for themes that explain how team member behaviors either facilitate or inhibit team sensemaking.Assuring rigorWe will use several established strategies to assure confirmability, dependability, and credibility [,] in qualitative data collection and analysis. These are briefly described in Table . We will log all study material in a Microsoft Access database using a date/time/source stamp to facilitate access to these materials. This database will serve as the basis for an audit trail.Table 1Strategies for Ensuring RigorCriteriaStrategies to assure criteria are metConfirmability: unrecognized researcher biases are controlledRK and RA (and later the research team) serve as the check and balance for uncovering assumptions and suggesting rival hypotheses.Member checks will be used to confirm findings.Dependability: candidate performance remains consistent over timeGuides will be used for all data collection.RK and RA will meet bi-weekly to review data collection and refine techniques.An electronic code book will be used to track all data transformations.An audit trail will be established.RK and RA will read and each code at least % of the field notes and compare coding. We will discuss and come to agreement about codes and interpretations.Credibility: results are plausible and authenticTriangulation of data from multiple sources: direct observation (multiple healthcare disciplines and organizational hierarchical levels) and documents.Member checks will be used to confirm findings.DiscussionThis study appears to be the first to prospectively examine a multidisciplinary HIT implementation project team and its sub-teams. Hospitals often form project teams to provide a formal mechanism for sharing different perspectives on events, in this case, an HIT implementation, and developing solutions to implementation issues. The project team in this study represents a huge organizational investment in that more than people will be involved in the project. Rather than using traditional, mechanistic models for studying HIT implementation in hospitals, we propose an innovative perspective--sensemaking--that reveals embedded social processes that shape large scale organization change. Effective sensemaking facilitates team members' understanding of what is happening, their learning, their problem solving, and, ultimately, the actions they take (or do not take) with regard to system implementation []. Prior research linked these activities to successful HIT implementation in non-hospital settings [,,,]. This study will: identify HIT implementation issues within the complex hospital environment and how team members deal with roadblocks and unexpected events; and describe the link between team member social interaction and implementation actions. These findings will lead to new methods of managing multidisciplinary project teams and implementing HIT in hospitals.Strengths and limitationsWe recognize several limitations inherent in our design choices. Our study will be conducted in a single, large-scale academic hospital, thus generalizability of our findings to other types of healthcare organizations may be limited. We will neither interview individual project team members about their project team experience nor will we observe their interactions with people outside of the teams. However, because the sense the project team makes of ongoing implementation events is dependent upon verbal exchanges, we believe the choice to limit our observations to project team activities, such as meetings, will allow us to describe important discourse in sensemaking of HIT implementation. Finally, all project sub-teams will not be directly observed. However, we will note how our selected teams are keeping team members informed of other aspects of the overall project, and we will include documents from excluded sub-teams in our analysis.Through our methodological choices, we aim to enrich the project team and hospital-based HIT implementation literature. Unlike many other studies, ours focuses on the people responsible for HIT implementation and will capture the interpretations and actions of a diverse group of project team members. Rather than relying on participant perceptions of events and potentially unreliable, retrospective data collection methods, the prospective case study design captures:. Key antecedents and consequences of ongoing, evolutionary, social process of implementing HIT.. Key internal and external factors that influence project teams including team composition, team member interaction, and interaction between project teams and the larger organization.. Key ways in which internal and external factors actually influence project team processes.. Key ways in which project team processes facilitate team task accomplishment.The resulting in-depth, rich description of HIT implementation will facilitate determining how sensemaking differs among project teams, how sensemaking develops over time, what information and events teams respond to, what meaning is constructed, and what actions result from that meaning. Thus, this study will make a significant contribution to advancing our understanding of how project teams function within the complex hospital care environment and bring about organizational change.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsRK designed the study and drafted the manuscript. RA and RM guided study design and read and revised the manuscript. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1Additional reference material used in developing the background and significance for the study.Click here for fileAdditional file 2Appendix A (direct observation guide); Appendix B (document guide); Appendix C (participant demographic survey tool); and Appendix D (a priori code list)Click here for file
|
PMC2696424.txt
|
TITLE: The N-glycome of human embryonic stem cells AUTHORS: Tero Satomaa, Annamari Heiskanen, Milla Mikkola, Cia Olsson, Maria Blomqvist, Minna Tiittanen, Taina Jaatinen, Olli Aitio, Anne Olonen, Jari Helin, Jukka Hiltunen, Jari Natunen, Timo Tuuri, Timo Otonkoski, Juhani Saarinen, Jarmo Laine ABSTRACT: BackgroundComplex carbohydrate structures, glycans, are essential components of glycoproteins, glycolipids, and proteoglycans. While individual glycan structures including the SSEA and Tra antigens are already used to define undifferentiated human embryonic stem cells (hESC), the whole spectrum of stem cell glycans has remained unknown. We undertook a global study of the asparagine-linked glycoprotein glycans (N-glycans) of hESC and their differentiated progeny using MALDI-TOF mass spectrometric and NMR spectroscopic profiling. Structural analyses were performed by specific glycosidase enzymes and mass spectrometric fragmentation analyses.ResultsThe data demonstrated that hESC have a characteristic N-glycome which consists of both a constant part and a variable part that changes during hESC differentiation. hESC-associated N-glycans were downregulated and new structures emerged in the differentiated cells. Previously mouse embryonic stem cells have been associated with complex fucosylation by use of SSEA- antibody. In the present study we found that complex fucosylation was the most characteristic glycosylation feature also in undifferentiated hESC. The most abundant complex fucosylated structures were Lex and H type antennae in sialylated complex-type N-glycans.ConclusionThe N-glycan phenotype of hESC was shown to reflect their differentiation stage. During differentiation, hESC-associated N-glycan features were replaced by differentiated cell-associated structures. The results indicated that hESC differentiation stage can be determined by direct analysis of the N-glycan profile. These results provide the first overview of the N-glycan profile of hESC and form the basis for future strategies to target stem cell glycans. BODY: BackgroundDuring the last decade global genomics and proteomics analyses of defined cell populations have revolutionized our understanding of cell biology. Glycomics – the study of global glycan expression profiles – has been predicted to be a next step forward []. Glycans, the carbohydrate units of glycoproteins, glycolipids, and proteoglycans, are capable of great structural variation and their specific molecular structures carry vast amounts of biological information []. It has been estimated that more than half of all cellular proteins are glycosylated [], but little is known of glycan structures in specific cell types. Glycans linked to cell surface proteins and lipids form a dense layer – the glycocalyx – on the extracellular side of the cell surface. The glycocalyx has first-line functions in the communication of the cell and its environment, including both cell-to-cell contacts [,-] and interactions with extracellular matrix components []. In addition, the specific roles of N-glycans involve regulation and control of protein folding [,] and trafficking [].Human embryonic stem cells (hESC) [] provide models for the study of human development and toxicology and have therapeutic potential in regenerative medicine []. To effectively utilize these cells, novel differentiation stage and lineage specific stem cell markers are required. Since glycans are abundant components of the cell surface, reagents that specifically recognize hESC glycans should be useful tools for the identification, isolation, and manipulation of stem cells. In fact, the monoclonal antibodies currently used to define hESC, including the globo-series glycosphingolipid epitopes SSEA- and SSEA-, and the keratanase-sensitive glycoprotein associated epitopes Tra – and Tra –, recognize glycan antigens [-]. Further, the expansion of undifferentiated hESC and the directed differentiation of hESC to specific progeny lineages in cell culture remain problematic. Understanding how cells interact through the glycocalyx with feeder cells and other components of the culture environment may enable rational design of specific culture systems.In the present study, a global analysis of the asparagine-linked glycans (N-glycans) of hESC and cells differentiated from them was performed by mass spectrometric profiling of unmodified glycans. We found that hESC have a characteristic and complex protein N-glycosylation profile. The data provide insight into the glycobiology of hESC and can be utilized as a basis for future studies exploring the role of stem cell glycans.ResultsAnalysis strategyIn order to generate mass spectrometric glycan profiles of hESC, embryoid bodies (EB), and further differentiated cells, a matrix-assisted laser desorption-ionization (MALDI-TOF) mass spectrometry based analysis was performed. We focused on the most common type of protein post-translational modifications, N-glycans, which were enzymatically released from cellular glycoproteins. During glycan isolation and purification, the total N-glycan pool was separated by an ion-exchange step into neutral N-glycans and sialylated N-glycans. These two glycan fractions were then analyzed separately by mass spectrometric profiling (Fig. and ), which yielded a global view of the N-glycan repertoire and allowed comparative analysis of differentiation-associated changes. The present scarce sample amounts did not allow us to purify individual glycan components for structural analyses. However, detailed structural analyses were achieved from the total neutral and acidic N-glycan pools by a combination of proton NMR spectroscopy, specific glycosidase digestions, and MS/MS fragmentation experiments.Figure 1Mass spectrometric neutral N-glycan profile of human embryonic stem cells (hESC). A. MALDI-TOF MS spectrum of neutral N-glycan fraction isolated from a hESC sample. B. Average of relative signal intensities from most abundant neutral N-glycans of four finnish hESC lines (blue columns), embryoid bodies derived from the hESC lines (EB, red columns), and stage differentiated cells (st., white columns). The columns indicate the mean abundance of each glycan signal (% of the total detected glycan signals). Error bars represent the standard error of mean. Proposed monosaccharide compositions are indicated on the x-axis and proposed structures for the major N-glycans are shown as schematic drawings. Gray circle/H: hexose, green circle: mannose, yellow circle: galactose, blue circle: glucose, gray square/N: N-acetylhexosamine, blue square: N-acetylglucosamine, red triangle/F: fucose/deoxyhexose, violet diamond/S: N-acetylneuraminic acid, light blue diamond/G: N-glycolylneuraminic acid. Asterisks indicate known polyhexose contamination that was not included in panel B.Figure 2Mass spectrometric acidic N-glycan profile of hESC. A. MALDI-TOF MS spectrum of acidic N-glycan fraction isolated from a hESC sample. B. Average of relative signal intensities from most abundant sialylated N-glycans of the four hESC lines, EBs, and stage differentiated cells. N-glycan signals carrying structural features associated with either hESC (F ≥ , red circles) or differentiated cells (N = H, gray circles) are indicated below panel B. See the legend of Figure for details and monosaccharide symbols.MALDI-TOF mass spectrometric analysis has been shown to be accurate in relative quantitation of glycan components within complex glycan mixtures []. We and others have found that it is a powerful tool for comparing changes in the glycan composition especially between closely related samples [-]. The present MALDI-TOF mass spectrometric methods have been optimized for relative quantitation of non-derivatized glycans and potentially labile glycan residues such as sialic acids are reliably determined. This was further verified during the present work by comparison of non-derivatized and permethylated glycan samples. The relative proportions of the major glycan signals were similar and no sialic acid or fucose loss was detected (data not shown). Due to the large m/z range of observed glycan ions the percentage of each glycan signal might not represent absolute molar percentage of the total cellular N-glycome. The amounts of especially larger glycans may be underestimated. However, we demonstrated with standard molecules that over the m/z range there was a linear response between the amount of added analyte and its relative signal intensity in the recorded glycan profile (see Additional File , Supplementary Fig. ).Each step in the glycan purification sequence was controlled for reproducibility by mixtures of glycoproteins, synthetic glycans, and glycan mixtures extracted from human cells. We routinely obtained the same mass spectrometric profiles for the standard glycan mixtures after each purification step during the multi-stage purification process [] showing that there was no selective loss of glycans during purification. The yield of the N-glycosidase reaction was over % with model glycoproteins. The robustness of the present method was further evaluated by subjecting human cell samples to blinded analysis by five different persons. The results were highly comparable showing that the present method reliably reproduced the characteristic glycan profile of each cell type (see Additional File , Supplementary Fig. ). The glycan profile was demonstrated to be reproducible also when different dilutions of the same sample were analyzed, demonstrating that the results were not sensitive to the exact amount of cells in each sample. However, in the present analyses the cell amounts were comparable in the whole sample series.We analyzed N-glycan profiles of all the biological materials that were in contact with the stem cells and could potentially contaminate the samples with glycoproteins, including hEF and mEF feeder cells, cell culture media, and culture media supplements. The N-glycan profiles revealed that the employed cell harvesting and washing procedures had been efficient and neither the feeder cells nor the culture media glycoproteins had affected the observed stem cell N-glycan profiles to a marked extent. By comparing the characteristic glycan signals in each potential contamination source, we could calculate that N-glycan purity in the samples was at least %.hESC N-glycan profilesNeutral N-glycans comprised approximately two thirds of the combined neutral and sialylated N-glycan pools of hESC. The relative proportions of the two glycan pools were analyzed by combining corresponding aliquots from the neutral glycan pool and the acidic glycan pool after broad-range sialidase digestion and comparing the combined glycan profile to the separate neutral and acidic glycan profiles (data not shown).The most abundant neutral N-glycan signals detected in the four hESC lines are presented in Figure 1B (blue columns). The similarity of the profiles with the four hESC lines, which is indicated by the minor variation in the glycan signals, suggests that the four cell lines closely resembled each other. For example, of the most abundant glycan signals were the same in every hESC line. The five most abundant signals (H5N2, H6N2, H7N2, H8N2, and H9N2) comprised % of the neutral N-glycan signals and dominated the profile.All the major N-glycan signals in the acidic N-glycan fraction (Fig. 2B, blue columns) contained sialic acid residues. There was more variation between individual cell lines in the most abundant acidic N-glycans than in the neutral N-glycans. However, the four hESC lines again resembled each other and the five most abundant sialylated N-glycan signals were the same in every cell line: S1H5N4F1, S1H5N4F2, S2H5N4F1, S1H5N4, and S1H6N5F1. The most abundant sialylated glycan signals contained the H5N4 core composition and differed only by variable number of sialic acid (S or G) and deoxyhexose (F) residues. These biantennary-size N-glycans together comprised % of the total glycan signal intensity in the acidic glycan fraction.We detected N-glycans containing N-glycolylneuraminic acid (G) in the hESC samples, for example glycans G1H5N4, G1S1H5N4, and G2H5N4. N-glycolylneuraminic acid has previously been reported in hESC as an antigen transferred from culture media containing animal-derived materials []. Accordingly, the serum replacement medium used in the present experiments contained bovine serum glycoproteins. We have recently detected Neu5Gc in N-glycans of hESC and in vitro cultured human mesenchymal stem cells by mass spectrometric N-glycan analysis [].The four hESC lines shared the same overall N-glycan profile and there was only slight cell line specific variation within the profiles. The most common N-glycan signals were the same in all the hESC lines and accounted for circa % of the total detected N-glycans. Similarly, EBs derived from each hESC line produced N-glycan patterns with similar characteristics, regardless of the different starting hESC line.Changes in the N-glycan profile during hESC differentiationA major goal of the present study was to identify glycan structures that would be specific to either stem cells or differentiated cells, and could therefore serve as differentiation stage markers. In order to determine whether the hESC N-glycome undergoes changes during differentiation, the N-glycan profiles obtained from hESC, EB, and stage differentiated cells were compared (Fig. 1B and 2B). The profiles of the differentiated cell types (EB and stage differentiated cells) were clearly different compared to the profiles of undifferentiated hESC, as indicated by non-overlapping distribution bars in many glycan signals. This suggested that differentiation induced the appearance of new N-glycan types while earlier glycan types disappeared.Figure presents the observed sample-to-sample variation in relative abundance of four N-glycan signals: S1H5N4F2, S1H5N4F1, and H9N2 (hESC-associated), as well as S1H5N5F1 (differentiated cell-associated). S1H5N4F2 (Fig. 3A) was the major hESC-specific glycan signal and had high abundance in all the four hESC lines, while its relative amount dropped consistently in all samples from both EB and further differentiated cells. In contrast, S1H5N5F1 (Fig. 3B) that was the major differentiation-associated glycan signal, was highly abundant in all EB and further differentiated cell samples, while it was only a minor glycan in all hESC samples. These two glycan signals could serve as markers of either the stem cells or the differentiated cells, respectively. The relative signal intensities of both S1H5N4F1 (Fig. 3C) and H9N2 (Fig. 3D) were on average slightly higher in hESC than in the differentiated cells, but individual differentiated cell samples differed greatly from the average. These glycan signals may therefore not be useful differentiation stage markers as such, but may however indicate trends in glycan biosynthesis.Figure 3Sample-to-sample comparison of relative signal intensities of four major N-glycan signals. Sample-to-sample comparison of relative signal intensities of four major N-glycan signals in the whole dataset of four hESC lines (FES , FES , FES , and FES ), embryoid bodies derived from them (EB), stage differentiated cells (st.), and two human fibroblast samples from the same cell line that was used as feeder cells in the propagation of hESC. A. Glycan signal S1H5N4F2 is characteristic to hESC. B. Glycan signal S1H5N5F1 is characteristic to differentiated cells. C. and D. The major N-glycan signals S1H5N4F1 (C.) and H9N2 (D.) are abundantly expressed in hESC and show variable expression in the differentiated cell types. Asterisks mark statistical significant differences between sample groups according to one-way ANOVA (p < .).The major differentiation stage associated N-glycan signal types were also visualized in the sialylated N-glycan profiles (Fig. 2B). There was a significant hESC association in the glycans S1H5N4F2 (Fig. 3A), S1H6N5F2, and S1H5N4F3 as well as other glycan signals that contained at least two deoxyhexose residues (F ≥ ; marked with red circles under Fig. 2B). Even minor signals with this structural feature were more abundant in hESC than in the differentiated cell types. Sum of the relative signal intensities of all F ≥ N-glycan signals in Fig. 2B was % for hESC, % for EB, and % for stage differentiated cells, indicating that multifucosylation of sialylated N-glycans was gradually decreased during hESC differentiation. Structural analysis of these glycans is presented in the following section.In contrast, a major group of N-glycan signals which increased during differentiation contained equal amounts of N-acetylhexosamine and hexose residues (N = H) in their monosaccharide composition, e.g. S1H5N5F1 (Fig. 3B). This is consistent with N-glycan structures containing non-reducing terminal N-acetylhexosamine residues. EB and further differentiated cells showed increased amounts of N-glycans expressing such monosaccharide compositions (marked with gray circles under Fig. 2B). Sum of the relative signal intensities of all N = H N-glycan signals in Fig. 2B was % for hESC, % for EB, and % for stage differentiated cells, indicating that this structural feature was drastically increased during hESC differentiation at the EB stage and then further increased in stage differentiated cells. However, these glycan structures were not characterized further in the present study.Of the three sample types, only hESC were grown in the presence of hEF feeder cells, while both EB and further differentiated cells were grown without hEF. Therefore the observed hESC-specific glycans could have been contaminants derived from hEF. Figure clearly shows that this was not true for the major hESC-associated glycan signal S1H5N4F2 (Fig. 3A) that had very low abundance in both analysed hEF samples. Similar examination of other hESC-associated glycans was consistent with this and we could conclude that any potential hEF-derived contamination was too minor to be detected. The major glycan structures that we identified to be associated with either the stem cells or the differentiated cells were not detected in the potential contamination sources. These structures included the major N-glycan signals with monosaccharide compositions indicating complex fucosylation and terminal HexNAc, for example the glycan signals shown in Fig. 3A–B. These signals were also undetectable in the cell culture media and supplements (data not shown).The N-glycan profiles of the differentiated cells were also quantitatively different from the undifferentiated hESC profiles. A practical way of quantifying the differences between glycan profiles was to calculate the sum of the signal intensity differences between two samples. According to this method, the EB neutral and sialylated N-glycan profiles had undergone a quantitative change of % and % from the hESC profiles, respectively. Similarly, the stage differentiated cell neutral and sialylated N-glycan profiles had changed by % and %. Taking into account that the proportion of neutral to sialylated N-glycans in hESC was approximately :, the total N-glycan profile change was approximately / during the transition from hESC to stage differentiated cells. The present data thus indicated that the mass spectrometric profile of the hESC N-glycome consisted of two discrete parts regarding propensity to change during hESC differentiation – a constant part and a changing part. As described above, even the minor signals reflected the differentiation stage associated N-glycan structural features. Therefore the mass spectrometric profile in itself was a specific and sensitive marker of hESC differentiation stage.Statistical analysis of N-glycan profilesTo evaluate statistical differences between hESC and differentiated cell N-glycan profiles, we performed one way ANOVA for each glycan signal (see Additional File ). The analysis indicated statistical significance for the hESC-associated N-glycan signals, e.g. large high-mannose type N-glycan signals H7N2, H8N2, and H9N2, as well as complex fucosylated N-glycan signals H5N4F2, S1H5N4F2, and S1H5N4F3.Factor and correlation analyses were employed in order to find common factors which could explain the observed distribution of N-glycan signals between the different cell types. The analyses were performed separately to the neutral N-glycans and acidic N-glycans (data not shown). Observation of the resulting factors indicated structure type related clustering of glycan signal groups in each factor. For example, major contributing signals in Factor N1 (explaining % of all variation in the neutral N-glycan fraction) reflected the balance between hESC-associated large high-mannose type N-glycans (H7N2, H8N2); and differentiation-associated low-mannose type N-glycans (H2N2, H3N2, H4N2), whereas Factor N2 (%) was dominated by differentiated cell types associated glycan types: complex-type N-glycans with N = H type non-reducing terminal HexNAc (H5N5, H5N5F1) and hybrid-type N-glycans (H5N3, H5N3F1, H6N3). Correlation analysis indicated correlation among the large high-mannose type N-glycan signals (H6N2, H7N2, H8N2) and negative correlation with the small high-mannose type N-glycan H5N2.In acidic glycan factor analysis, Factor A1 (explaining % of all variation in the acidic N-glycan fraction) reflected balance between ) differentiated cell associated hybrid-type N-glycan signals (S1H5N3F1, S1H6N3) and ) hESC associated complex-type N-glycans (S1H6N5F1, S1H7N6F1, S1H8N7F1) and complex-fucosylated glycans (S1H7N6F3). The major contributing signals in both Factors A2 and A3 (together explaining % of all variation in the acidic fraction) were N = H type terminal HexNAc bearing glycans (S1H5N5, S1H5N5F1, S2H5N5F1, S1H5N5F3, S1H6N6F1, S2H6N6F1).In conclusion, the performed statistical analyses demonstrated that N-glycosylation changes during hESC differentiation were regulated so that the glycan biosynthetic changes were consistently similar from cell line to cell line, which was directly reflected in the level of cellular N-glycan profiles and even distinct glycan signals.Structural analyses of the major hESC N-glycansThe N-glycan fractions were further analyzed by proton NMR spectroscopy recorded from N-glycans isolated from a larger sample of hESC grown on mEF. The assigned N-glycan structures are included in Figure (for details see Additional File , Supplementary Fig. and Supplementary Tables and ). In the obtained NMR spectrum of the hESC neutral N-glycans, signals consistent with large Man6-Man9 high-mannose type N-glycans were detected. No evidence of other structures was found since the high-mannose type N-glycan structures could explain all signals in the spectrum. In similar analysis of the sialylated N-glycan fraction, all the detected signals were consistent with biantennary complex-type N-glycans with type N-acetyllactosamine (LacNAc) antennae, α2,- and α2,-linked sialic acids, and α1,-linked N-glycan core fucose modifications. No signals corresponding to type LacNAc antennae or other fucose linkages were detected. It was determined by integration of indicator signals that α2,-linked sialic acids were more abundant than α2,-linked sialic acids.In order to validate glycan structure assignments made based on the mass spectrometric and NMR spectroscopic profiling analyses, we performed enzymatic degradation experiments with subsequent mass spectrometric detection (Fig. ). Comparison of the original and digested N-glycan profiles allowed estimation of the relative proportions of non-reducing terminal monosaccharide residues of neutral N-glycans characteristic of hESCs (Fig. 4A). α-mannose residues were the most common terminal residues in the neutral glycan fraction, occurring in high-mannose type, low-mannose type, and hybrid-type N-glycan signals. β1,-linked galactose residues occurred in hybrid-type and complex-type N-glycans, whereas terminal β1,-linked galactose residues were not detected. Non-reducing terminal β-N-acetylglucosamine, α1,-fucose, and α1,- or α1,-linked fucose residues were also present in minor glycan components (data not shown), but not in the major N-glycan signals presented in Figure .Figure 4Proposed structures for most abundant N-glycan signals detected in the present study. Proposed structures for most abundant N-glycan signals detected in the present study ( neutral and sialylated N-glycans) are based on combined results from MALDI-TOF mass spectrometry, proton NMR spectroscopy (NMR), and exoglycosidase analyses with α-mannosidase (αMan), β1,- and β1,-galactosidase (β4Gal), β-N-acetylglucosaminidase, specific α1,/- and α1,-fucosidases (α3/4Fuc and α2Fuc), and broad-range sialidase (SA). Relative abundances in hESC and EB N-glycan profiles are indicated. Only the positive identifications in 1H-NMR analyses and sensitivity to specific exoglycosidase digestions have been marked. Monosaccharide symbols are as in Figure . Where appropriate, glycosidic bonds have been indicated. Two simplifying assumptions have been made: i) in structures with H ≥ and N ≥ , the proposed structures have been assigned a trimannosyl core structure, and ii) all fucosylated structures have been assigned a core fucose residue.The abundances of the major glycan signals in hESC and EB are indicated in Figure and show distinct features in stem cells and differentiated cells. Most specifically, high-mannose type N-glycans were associated with hESC, while low-mannose type and hybrid-type N-glycans were more abundant in EB. These differences suggest that regulation of N-glycan biosynthetic pathways is changed during hESC differentiation. We are currently studying the potential regulatory mechanisms leading to these N-glycan profile differences (TJ et al., manuscript in preparation). The observation of abundant low-mannose type N-glycans is interesting. We observed comparable amounts of fucosylated and non-fucosylated low-mannose type N-glycans in both hESCs and EBs, while analyses of the culture media revealed that they contained only non-fucosylated N-glycans. This indicates that these structures were produced by the studied cells and they did not originate from culture medium glycoproteins.Analysis of complex fucosylation in N-glycans of hESCAs noted above, there was a significant hESC association in N-glycans containing at least two deoxyhexose residues (F ≥ ; see Fig. 2B). In contrast, glycan signals such as S2H5N4 that contained no deoxyhexose (F = ) were increased in the differentiated cell types. This suggested that sialylated N-glycans in undifferentiated hESC were subject to more fucosylation than in the differentiated cell types. The most common fucosylation type in human N-glycans is α1,-fucosylation of the N-glycan core structure [] and this was also the major type of fucosylation detected in the present NMR profiling. In human N-glycans containing more than one fucose residue there should be other fucose linkages in addition to the α1,-linkage [,], indicating complex fucosylation. The F ≥ structural feature decreased as the cells differentiated, showing that complex fucosylation was characteristic of undifferentiated hESC.Exoglycosidase analysis of the sialylated N-glycan fractions was performed with specific α1,- and α1,/-fucosidase enzymes to characterize the major hESC-specific N-glycan signals with complex fucosylation (Fig. 5A). Based on the sensitivity of the signal S1H5N4F2 towards the employed fucosidase enzymes, it was shown to include a mixture of isomeric N-glycan structures. Nearly half of the structures within S1H5N4F2 carried one terminal α1,-linked fucose residue, while the other half carried one terminal α1,- or α1,-linked fucose residue. α1,/-fucosidase had a larger effect after desialylation (data not shown), indicating that in minor structures there was a fucose residue located subterminal to sialic acid. The majority of the structures contained exactly one fucose residue that was not susceptible to the employed fucosidase treatments, indicating fucosylation in the N-glycan core sequence. This was consistent with the NMR results showing that N-glycan core α1,-fucosylation was the most abundant fucose linkage in hESC N-glycans. All detected desialylated and defucosylated N-glycans of hESC were sensitive to β1,-galactosidase but not to β1,-galactosidase digestion, indicating that the major N-glycan antennae were type LacNAc. Taken together, the results suggested that the fucosylated antennae were either α1,-fucosylated H type i.e. Fucα1-2Galβ1-4GlcNAc or α1,-fucosylated Lex i.e. Galβ1-(Fucα1–)GlcNAc. Due to limited abundance of the minor structure with fucose residue subterminal to sialic acid, we were not able to sequence it, although a candidate structure is sLex i.e. Neu5Acα2-3Galβ1–(Fucα1-)GlcNAc.Figure 5Analysis of major complex fucosylated N-glycans of hESC. A. The major sialylated N-glycan signal with complex fucosylation feature (S1H5N4F2) was subjected to exoglycosidase sequencing with linkage-specific fucosidases and broad-range sialidase. As outlined in the reaction scheme, the signal was shown to be an approximately : mixture of α1,- and α1,-/α1,-fucosylated biantennary complex-type N-glycans with core fucosylation. B. and C. The acidic N-glycan fraction was permethylated and subjected to MS/MS fragmentation. Both S1H5N4F2 (B.) and S1H5N4F1 (C.) produced fragments that supported the structure assignments as indicated in the schematic drawings. D. The studied hESC lines were shown to express four of the known human fucosyltransferases. Minimal glycan acceptor and product specificities for the encoded enzymes are shown in the schematic drawing.MS/MS fragmentation analysis was applied to the most abundant hESC N-glycans with differential fucosylation, S1H5N4F2 (Fig. 5B) and S1H5N4F1 (Fig. 5C). The fragmentation patterns supported the exoglycosidase digestion and proton NMR profiling results and were consistent with sialylated biantennary complex-type N-glycans with predominant core fucosylation. Fucosylated antennae derived from S1H5N4F2 were represented by a single fragment ion at m/z . that fits to both H type and Lex terminal structures (Fig. 5B), while fragment ions corresponding to sLex were not detected.Previously the gene expression profiles of the FES hESC lines have been determined [] and we extracted the information of the expressed fucosyltransferases from this data (Fig. 5D). The hESCs expressed fucosyltransferase genes FUT1, FUT2, FUT4, and FUT8. Of these genes, FUT1 and FUT4 were overexpressed in hESC when compared to EB. As shown in Figure 5D, in N-glycans expressing type LacNAc antennae (Galβ1-4GlcNAc) the functional expression of the corresponding glycosyltransferase enzymes may produce the following fucosylated epitopes: H type (FUT1), Lex and sLex (FUT4), Ley (combined action of FUT1 and FUT4), as well as N-glycan core α1,-fucosylation (FUT8). Structures corresponding to Ley (difucosylated LacNAc) were not observed in the present study. Taken together, the evidence indicated that the major fucosylated epitopes in hESC N-glycans were H type and Lex, and additional minor sialylated and fucosylated structures were most likely sLex. However, based on the present experiments it could not be excluded that minor Ley or type structures could be present in hESC N-glycans.The identified hESC glycans can be targeted at the cell surfaceFrom a practical perspective stem cell research would be best served by reagents that recognize cell-type specific target structures on cell surface. To investigate whether individual glycan structures we had identified would be accessible to reagents targeting them at the cell surface, we performed lectin labeling of three candidate structure types. Lectins are proteins that recognize glycans with specificity to certain glycan structures. Previous studies have extensively described lectin labeling of hESC [-] and we have also initiated a study of lectin and glycan antibody ligands on hESC surfaces (MM et al., manuscript in preparation). In the present study, hESC colonies grown on mouse feeder cell layers were labeled in vitro by fluorescein-labelled lectins (Fig. ). The hESC cell surfaces were clearly labeled by Maackia amurensis agglutinin (MAA) that recognizes structures containing α2,-linked sialic acids, preferably α2,-sialylated LacNAc, indicating that such sialylated glycans were abundant on the hESC cell surface (Fig. 6A). In contrast, the cell surfaces were not labelled by Pisum sativum agglutinin (PSA) that recognizes α-mannosylated glycans and potentially also core fucosylated N-glycans (Fig. 6B). However, PSA labelled the cells after permeabilization (data not shown). Finally, Ulex europaeus agglutinin I (UEA-I) that recognizes fucosylated structures, especially H type , stained hESC surfaces (Fig. 6C). The specificity of the lectin bindings was validated by inhibition with specific glycan inhibitors as described in the Methods section. Interestingly, the mouse fibroblast cells showed complementary staining patterns compared to hESC, suggesting that these lectin reagents efficiently discriminated between hESC and feeder cells. Figure 6D shows the FACS results with UEA-I further demonstrating that hESC were highly positive for complex fucosylated glycans including H type terminal fucosylated sequences. Consistent with the lectin labeling results, the present structural analyses also demonstrated that both complex fucosylated structures including H type and α2,-sialylated LacNAc were abundant terminal structures in hESC N-glycans. The results further suggested that the identified glycan structures could be utilized to select reagents specifically targeting undifferentiated hESC, while not binding to other cell types. Glycans expressed on the hESC surface would be good targets for recognition by specific antibodies, which is subject of ongoing research in our laboratories.Figure 6Lectin staining of hESC colonies grown on mouse feeder cell layers. A. Maackia amurensis agglutinin (MAA) that recognizes α2,-sialylated glycans, preferably in type LacNAc, stained hESC but not feeder cell surfaces. B. Pisum sativum agglutinin (PSA) that recognizes α-mannosylated glycans and core fucosylated N-glycans, stained only mouse feeder cell surfaces. C. Ulex europaeus agglutinin I (UEA-I) that recognizes α1,-fucosylated glycans preferably within H type , stained the hESC colony. Mouse fibroblasts had complementary staining patterns with the lectins, indicating that their surface glycans are clearly different from hESC. D. Fluorescence-assisted cell sorting (FACS) diagrams of UEA-I selected hESC, showing that the majority of hESC were positive for cell surface UEA-I ligands.DiscussionIn the present study, mass spectrometric and NMR spectroscopic analysis methods were applied in the first structural analysis of hESC N-glycan profiles. Previously, the glycosylation of hESC has been studied with lectins and antibodies [-], and a preliminary report has been published on mass spectrometric profiling of mouse embryonic stem cell (mESC) N-glycans []. The objective in the present study was to provide a global view on the N-glycan profile, or a "fingerprint" of hESC N-glycosylation, to structurally characterize the most abundant N-glycan structures of hESC, and to compare hESC N-glycosylation with differentiated cells. The hESC N-glycome was found to be characteristic to the cell type and different from either differentiated human cells or mESCs. The data provided information regarding the most characteristic features of hESC N-glycosylation and demonstrated that a dramatic N-glycan profile change takes place during hESC differentiation.Over one hundred N-glycan signals were detected from each cell type. However, it is important to realize that many of the mass spectrometric signals in the present analyses include multiple isomeric structures and the one hundred most abundant signals therefore represent a larger amount of different glycans. The major N-glycans observed in hESC covered all the major N-glycan classes, namely oligomannose-type, hybrid-type, and complex-type N-glycans [], and were decorated with sialylated and fucosylated antennae with equal complexity in all differentiation stages. This directly demonstrated that stem cell N-glycosylation was already as sophisticated as in the differentiated cells.We found that the hESC N-glycan profile contained both a constant part and a variable part. The variable part was a sensitive indicator of the differentiation commitment. The major glycan types in the constant part were high-mannose type and biantennary complex-type N-glycans. The most characteristic feature of the variable part of the hESC N-glycome was complex fucosylation. In fact, it was found that % of the acidic N-glycan signals detected in hESC were multifucosylated. On the other hand, structurally different glycan structures were favoured by the differentiated cell types. About / of the total N-glycan profile of hESC changed during their differentiation. This demonstrated that during differentiation hESC substantially changed the appearance of their glycocalyx. New N-glycan features emerged in EB and further differentiated cells. These features included additional N-acetylhexosamine residues, potentially leading to completely new glycan epitopes presented on the differentiated cell surface. Such drastic changes in the N-glycome may profoundly alter both cell-cell interactions and the cells' responses to exogenous signals.Glycans perform their functions in cells by acting as ligands for specific glycan receptors [-], functioning as structural elements of the cell [], and modulating the activity of their carrier proteins and lipids []. More than half of all proteins in a human cell are glycosylated []. Consequently, a global change in protein-linked glycan biosynthesis can simultaneously modulate the properties of multiple proteins. It is likely that the large changes in N-glycans during hESC differentiation have major influences on a number of cellular signaling cascades and affect in profound fashion biological processes within the cells.The major hESC-specific N-glycosylation feature we identified was complex fucosylation. Fucosylation is known to be important in cell adhesion and signalling events [,] as well as being essential for embryonic development []. Knock-out of the N-glycan core α1,-fucosyltransferase gene FUT8 leads to postnatal lethality in mice [], and mice completely deficient in fucosylated glycan biosynthesis do not survive past early embryonic development []. Fucosylated glycans such as the SSEA- antigen, a special form of Lex [-], have previously been associated with both mESC and human embryonic carcinoma cells []. However, SSEA- is not expressed by hESC, which has previously been interpreted such that hESC would not express Lex. A recent report has suggested that mESC proliferation and differentiation can be influenced via specific recognition of fucosylated and sialylated glycoconjugates in a mESC line transfected with L1 receptor [].The published gene expression profiles for the same hESC lines as studied here [] have demonstrated that four human fucosyltransferase genes, FUT1, FUT2, FUT4, and FUT8 are expressed in hESC (Fig. 5D), and that FUT1 and FUT4 are overexpressed in hESC when compared to EB (TJ et al., manuscript in preparation). FUT8 encodes the N-glycan core α1,-fucosyltransferase whose product was identified as the major fucosylated epitope in hESC N-glycans by the NMR analysis. The hESC-specific expression of FUT1 and FUT4, encoding for α1,-fucosyltransferase and α1,-fucosyltransferase enzymes, respectively [], correlates with our findings of simple fucosylation in EB and complex fucosylation in hESC. The hESC-expressed enzyme product of FUT2 (Secretor) may primarily be linked to other glycan classes such as O-glycans or glycolipids based on its preference for type , , and chains not detected in N-glycans (MM et al., manuscript in preparation). Interestingly, the FUT4-encoded enzyme is capable of synthesizing both Lex, sLex, and SSEA-, although the capability to synthesize sLex may be low [,]. We detected N-glycan antenna structures consistent with Lex in hESC N-glycans. Consistent with this, Lex has been reported to be present in mESC N-glycans []. N-glycan signals potentially corresponding to sLex were detected in very low amounts in hESC, which is consistent with the reported specificity of the FUT4-encoded enzyme [,]. Our finding of H type structures in hESC N-glycans is a novel feature that differentiates hESC from mESC. However, Wearne et al. [] have already reported α1,-fucosylation in hESC by utilizing UEA-I lectin staining. Significantly, product of the hESC-overexpressed fucosyltransferase FUT1 (H enzyme) is mainly responsible for H type antigen biosynthesis (Fig. 5D). In conclusion, although hESC do not express the specific fucosylated antigen recognized by the SSEA- antibody, they share with mESC the characteristic features of complex fucosylation and expression of the Lex antigen. The functions of these major fucosylation modifications in hESC remain to be elucidated in future studies. The present results suggest that the SSEA- antibody does not recognize Lex when when presented on a biantennary N-glycan antenna.Human embryonic stem cell lines have previously been demonstrated to have a common genetic stem cell signature that can be identified using gene expression profiling techniques [-]. Such signatures have been proposed to be useful in hESC characterization. In the present report we provide the first glycan profile signatures for hESC. The profile of the expressed N-glycans might be a useful tool for analyzing and classifying the differentiation stage in association with gene and protein expression analyses. In the present work we demonstrated that multiple mass spectrometric glycan signals correlated with the differentiation stage of hESC (Fig. ). The present results suggest that N-glycan profiling could be developed into a tool for monitoring hESC differentiation status. Glycan profiling might be more sensitive than the use of any single cell surface marker and especially useful for the quality control of hESC-based cell products []. However, further analysis of hESC glycans may also lead to establishing new glycan structures as stem cell markers in addition to the commonly used SSEA and Tra glycan structures.The present lectin staining experiments demonstrated that specific glycan molecules were abundant on the cell surface of hESC. The cell surface presentation of glycans makes them excellent targets for development of cell type specific recognition reagents. It seems plausible that knowledge of the changing surface glycan epitopes may be utilized as a basis in developing reagents and culture systems that would allow improved identification, selection, manipulation, and culture of hESC and their progeny. The present data allow rational selection and evaluation of glycan-specific antibodies based on knowledge of hESC glycan structures.Venable et al. [] and Wearne et al. [,] have extensively characterized hESC and EB reactivity for different lectins. Their results with MAA and UEA-I lectins confirm the present results about the expression of cell type-specific glycan epitopes on undifferentiated hESC surface. These previous studies may also provide cues for differentiation-associated N-glycan changes that were not structurally characterized in the present study. Venable et al. [] described that N-glycans modified by bisecting GlcNAc residues as detected by Phaseolus vulgaris erythroagglutinin (PHA-E) were enriched in cells that had low or absent SSEA- staining, potentially indicating that such N-glycan structures were early signs of hESC differentiation. Wearne et al. [] also found PHA-E ligands on differentiated hESC. It could be hypothesized that part of the terminal N-acetylhexosamine carrying N-glycans (e.g. N-glycans with N = H structural feature, see Fig. 2B) found in the present study to be associated with differentiation cells, could be modified by bisecting GlcNAc. Although specific N-glycan structural information is hard to extract from lectin binding profiles, the data of these earlier reports support our findings of abundant terminal α-mannose and LacNAc residues, both α2,- and α2,-sialylation, N-glycan core and peripheral fucosylation, as well as the presence of biantennary as well as branched complex-type N-glycans. However, specificities of individual plant lectins towards terminal mono-or oligosaccharide epitopes are usually not well characterized and may have multiple interpretations. Therefore the previous lectin studies gave a useful impression of the terminal monosaccharide epitopes but not exact structural information. The present structural data including larger oligosaccharide structures could have parallels with the prior data, but one should also consider technical factors which may explain differences or potentially cause artificial similarities. These differences include e.g. cell culture conditions, assay techniques, sample preparation, and differences between the studied cell lines. For example, we have previously shown that minor cell culture reagents may cause glycan contamination of stem cells [].By employing rapid purification and direct analysis of non-derivatized glycans we demonstrated mass spectrometric N-glycan profiling of the scarce hESC samples, enabling analysis of samples as small as cells. In many glycomic studies of mammalian cells and tissues the isolated glycans have been derivatized (e.g. permethylated) prior to mass spectrometric profiling [-] or chromatographic analysis []. However, we chose to directly analyze the picomolar quantities of unmodified glycans and high sensitivity was achieved while omitting the derivatization and the subsequent additional purification steps. This straightforward method could be widely applicable to analysis and monitoring of stem cell lines. We have recently applied the same method to human cord blood hematopoietic cells [] as well as human mesenchymal stem cells [].Stem cell glycosylation has been reported to be sensitive to composition of the cell culture medium [,]. In the present study, we analyzed all biological material in contact with the cells to exclude potential contamination sources. Since no cell type in the present study had identical cell culture conditions, we could not exclude the possibility that the observed profile differences were in part influenced by differences in cell culture. However, our data supports the conclusion that the major identified N-glycan structural features were not dependent on changes in either cell culture media or growth surface. hESCs and EBs were grown in the same culture medium except that bFGF was omitted from EB culture, while the major difference between hESC and EB culture was that hESC colonies were grown on feeder cells and EBs in suspension. To analyze if different growth surfaces could affect cellular glycosylation, we have compared N-glycan profiles of hESCs grown on hEF, mEF, Matrigel, and a defined non-animal growth support (MM et al., manuscript in preparation). On all these surfaces, hESCs show an N-glycan profile typical to undifferentiated cells, including abundant complex fucosylation (such as S1H5N4F2, Fig. 3A) and low terminal HexNAc (such as S1H5N5F1, Fig. 3B). This suggests that the identified hESC-associated N-glycan profile features are not sensitive to changes in the growth surface. In addition, human fibroblast feeder cells, grown together with hESC, produce an N-glycan profile missing the key identifying characteristics of hESC glycosylation (for example complex fucosylation, Fig. 3A). Further, stage differentiated cells grown as monolayers in different culture medium expressed the same differentiated cell associated structures as EBs (Fig. , , ).ConclusionHuman embryonic stem cells have a characteristic N-glycan profile which undergoes major changes when the cells differentiate. Information regarding the specific glycan structures may be utilized in developing reagents for targeting these cells and their progeny. Future studies investigating the developmental and molecular regulatory processes resulting in the observed N-glycan profiles may provide significant insight into mechanisms of human development and regulation of glycosylation.MethodsHuman embryonic stem cell linesFinnish hESC lines FES , FES , FES , and FES were cultured as described previously []. Briefly, two of the analysed cell lines were initially derived and cultured on mouse embryonic fibroblast (mEF) feeders, and two on human foreskin fibroblast (hEF) feeder cells. For the present studies all of the lines were transferred on hEF feeder cells and cultured in serum-free medium supplemented with Knockout serum replacement (Gibco). To induce the formation of embryoid bodies (EB) the hESC colonies were first allowed to grow for – days whereafter the colonies were cut in small pieces and transferred on non-adherent Petri dishes to form suspension cultures. The formed EB were cultured in suspension for the next days in standard culture medium without bFGF. For further differentiation (into stage differentiated cells, St.) EB were transferred onto gelatin-coated culture dishes in media supplemented with insulin-transferrin-selenium and cultured for days.For N-glycan profiling, on average cells were collected mechanically from culture on hEF feeder cell layers, washed five times with phosphate buffered saline, and stored frozen until the analysis. In fluorescence-assisted cell sorting (FACS) analyses –% of cells from mechanically isolated hESC colonies were typically Tra – and Tra – positive (data not shown). The differentiation protocol favours the development of neuroepithelial cells while not directing the differentiation into distinct terminally differentiated cell types []. Stage cultures consisted of a heterogeneous population of cells dominated by fibroblastoid and neuronal morphologies. For more detailed structural analyses by NMR spectroscopy and glycosidase digestions, up to million hESC were grown on mEF layers.Glycan isolationAsparagine-linked glycans were detached from cellular glycoproteins by F. meningosepticum peptide:N-glycosidase F digestion (Calbiochem) and purified as described previously []. Briefly, the detached glycans were purified by sequential precipitation/extraction and solid-phase extraction steps with miniaturized chromatography columns of C18 silica, strong cation-exchange resin, porous graphitized carbon, and for the sialylated glycans also microcrystalline cellulose.Mass spectrometry and data analysisMALDI-TOF mass spectrometry was performed with a Bruker Ultraflex TOF/TOF instrument (Bruker Daltonics, Germany) essentially as described []. Relative signal intensities of neutral and sialylated glycan components were assigned based on their relative signal intensities in the mass spectra when analyzed separately as the neutral and sialylated N-glycan fractions [-]. We calculated relative intensities for all detected glycan signals using the Flexanalysis . software (Bruker Daltonics). The present glycan profiles were extracted from the resulting signal lists by removing the effect of isotopic pattern overlapping, multiple alkali metal adduct signals, products of elimination of water from the reducing oligosaccharides, and other interfering mass spectrometric signals not arising from the original glycans in the sample. The resulting glycan signals in the presented glycan profiles were normalized to % to allow comparison between samples. The relative amounts of each glycan signal are expressed as "% of total profile" similarly as previously reported [,,]. The mass spectrometric fragmentation analysis of permethylated glycans was performed using the Bruker Ultraflex TOF/TOF instrument according to manufacturer's instructions.Glycosidase analysisAliquots from the N-glycan fractions were subjected to digestion with α-mannosidase from Jack beans (C. ensiformis; Sigma-Aldrich); β1,-galactosidase, β-N-acetylglucosaminidase, and α2,-sialidase from S. pneumoniae; broad-range sialidase from A. ureafaciens; and β1,-galactosidase, α1,/-fucosidase, and α1,-fucosidase from X. manihotis (all from Calbiochem). Reactions with approximately – pmol oligosaccharide aliquots were carried out by overnight digestion at +°C in μl of mM sodium acetate buffer pH .. The activities of the enzymes in each reaction were optimised such that they had the following substrate specificities in control reactions: α2,-sialidase digested Neu5Acα2-3Galβ1-4GlcNAc (Neu5Acα2-3LacNAc) but not Neu5Acα2-6LacNAc; β-N-acetylglucosaminidase digested GlcNAcβ1-3LacNAc but not GalNAcβ1-4GlcNAc; LacNAc but neither Galβ1-3GlcNAc nor Galα1-3LacNAc were digested with β4Gal; β1,-galactosidase digested Galβ1-3GlcNAc; α1,/-fucosidase digested Fucα1-(Galβ1-)GlcNAc (Lex) but not Fucα1-2Galβ1-3GlcNAc; α1,-fucosidase digested Fucα1-2Galβ1-3GlcNAc but not Lex; and α-mannosidase digested the high-mannose type N-glycans in a standard human N-glycan mixture. Digested glycan fractions were purified for analysis by solid-phase extraction and analyzed by mass spectrometry as described above.NMR methodsThe isolated glycans were purified for the analysis by gel filtration high-pressure liquid chromatography in a column of Superdex peptide HR / (Amersham), with water (neutral glycans) or mM NH4HCO3 (sialylated glycans) as the eluant at a flow rate of ml/min. The eluant was monitored at nm. Oligosaccharide pools were quantified against external standards N-acetylglucosamine and N-acetylneuraminic acid. Prior to NMR analysis the purified glycome fractions were repeatedly dissolved in .% deuterium oxide and dried to omit H2O and to exchange sample protons. The 1H NMR spectra at MHz were recorded using a cryo-probe for enhanced sensitivity [].Lectin bindingFluorescein-labeled lectins used in lectin binding studies were from EY Laboratories. Specificity of binding was controlled by inhibition with mM α3'-sialyllactose (Kyowa Hakko Kogyo, Japan), mM α-D-mannose methyl glycoside (Sigma-Aldrich), and mM L-fucose (Danisco Sweeteners, Finland) for Maackia amurensis agglutinin (MAA), Pisum sativum agglutinin (PSA), and Ulex europaeus agglutinin-I (UEA-I), respectively. Fluorescence-assisted cell sorting (FACS) analyses were performed essentially as described [].Statistical analysisNormalized mass spectrometric N-glycan profile data from hESC (n = ), EB (n = ), and stage differentiated cells (n = ) were imported to Statistica . software (StatSoft). If all or almost all data values were zero, the corresponding signal was removed from the data set. In one way ANOVA with Fisher LSD post hoc analysis and Factor analysis signals having all or most of the values zero in certain cell type were omitted. Whisker box plots with means and standard deviations were developed and screened to have an overall view of the data and to identify mass peaks with variation between different cell lines or differentiation stage. Factor analysis was performed by principal component extraction; factor loadings were Varimax normalized and signals having factor loadings >. and factors explaining >% of variance were included into the model. Pearson correlation analysis was performed and correlations more than . or less than -. were considered significant.AbbreviationsAbbreviations are as follows: bFGF: basic fibroblast growth factor; EB: embryoid bodies; ER: endoplasmic reticulum; F: deoxyhexose; FACS: fluorescence-assisted cell sorting; Fuc: L-fucose; G: N-glycolylneuraminic acid; Gal: D-galactose; GalNAc: N-acetyl-D-galactosamine; GlcNAc: N-acetyl-D-glucosamine; H: hexose; hESC: human embryonic stem cells; hEF: human foreskin fibroblast; H type : Fucα1-2Galβ1-4GlcNAc; ITS: insulin-transferrin-selenium; LacNAc: N-acetyllactosamine; Lex: Lewis x, Galβ1-(Fucα1-)GlcNAc; Ley: Lewis y, Fucα1-2Galβ1-(Fucα1-)GlcNAc; MAA: Maackia amurensis agglutinin; MALDI-TOF: matrix-assisted laser desorption-ionization time-of-flight; mEF: mouse embryonic fibroblast; mESC: mouse embryonic stem cells; N: N-acetylhexosamine; N-glycan: asparagine-linked glycan; Neu5Ac: N-acetylneuraminic acid; NMR: nuclear magnetic resonance; PSA: Pisum sativum agglutinin; S: N-acetylneuraminic acid; sLex: sialyl Lewis x, Neu5Acα2-3Galβ1-(Fucα1-)GlcNAc; St.: further (stage ) differentiated cells; UEA-I: Ulex europaeus agglutinin-I. The schematic representation of oligosaccharides is in accordance with the guidelines proposed by the Consortium for Functional Glycomics and as described in the legend of Figure .Authors' contributionsTS and AH contributed equally to this work. AH performed N-glycosylation analyses and TS analyzed the data and drafted the manuscript. MM, CO, TT, and TO provided the hESC samples. AH, MB, JH, TS, and JS developed glycan analysis techniques for stem cells. TJ analyzed gene expression data. MM and MT performed lectin binding assays. OA and AO performed the NMR analysis. JN participated in writing the manuscript. JS and JL supervised the project and contributed equally to this work. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1Supplementary data. Supplementary data including the following supplementary figures and tables: Supplementary Figures and . Examples of glycan profiling method evaluation. Supplementary Figure and Supplementary Tables and . NMR analysis of neutral and sialylated N-glycans.Click here for file
|
PMC2211757.txt
|
TITLE: Statistical validation of megavariate effects in ASCA AUTHORS: Daniel J Vis, Johan A Westerhuis, Age K Smilde, Jan van der Greef ABSTRACT: BackgroundInnovative extensions of (M) ANOVA gain common ground for the analysis of designed metabolomics experiments. ASCA is such a multivariate analysis method; it has successfully estimated effects in megavariate metabolomics data from biological experiments. However, rigorous statistical validation of megavariate effects is still problematic because megavariate extensions of the classical F-test do not exist.MethodsA permutation approach is used to validate megavariate effects observed with ASCA. By permuting the class labels of the underlying experimental design, a distribution of no-effect is calculated. If the observed effect is clearly different from this distribution the effect is deemed significantResultsThe permutation approach is studied using simulated data which gave successful results. It was then used on real-life metabolomics data set dealing with bromobenzene-dosed rats. In this metabolomics experiment the dosage and time-interaction effect were validated, both effects are significant. Histological screening of the treated rats' liver agrees with this finding.ConclusionThe suggested procedure gives approximate p-values for testing effects underlying metabolomics data sets. Therefore, performing model validation is possible using the proposed procedure. BODY: BackgroundIn life science research many measuring tools emerged in recent years. These tools give a coarse profile of biological classes such a transcripts (transcriptomics), proteins (proteomics) and metabolites (metabolomics). This paper focuses on the field of metabolomics; the comprehensive quantitative and qualitative analysis of all small molecules of cells, body fluids, and tissues. The mix of hypothesis and discovery driven omics-experiments create novel biostatistical challenges noted since combining pattern recognition and body fluid profiling in the early eighties []. Interpreting the multivariate metabolomics results means integrating biological knowledge with possible contributing metabolites.Metabolomics data sets comprise hundreds of metabolites measured in typically tenths of samples. Multivariate statistics on data that have fewer samples than metabolites is cumbersome. Usually there is an experimental design underlying the metabolomics data sets. The obvious technique for analyzing such data, Multivariate Analysis of Variance (MANOVA) [] cannot deal with data that consists of more metabolites than samples.The recent introduction of ANOVA-based extensions of multivariate data analysis methods may open new angles to analyze metabolomics data. These methods aim to analyze designed experiments with more measured metabolites than samples. Among the new methods are ANOVA-principal component analysis (PCA), principal response curves (PRC) and ASCA [-]. All these methods are a combination of PCA and ANOVA. In this paper we will provide a validation procedure for ASCA using a randomization strategy. The models of other ANOVA-based methods may also use this validation procedure.Analysis of variance simultaneous component analysis, ANOVA-SCA or ASCA is a generalized version of analysis of variance for univariate data to the multivariate case []. With this method it is possible to isolate the variation in the data induced by a factor varied in the experimental design. Analyzing this isolated variation with simultaneous component analysis may reveal the relation between the samples and metabolic profile. ASCA successfully helped the quality control in an application of the metabolomics platforms NMR, GC-MS and LC-MS []. In an experiment with toxin dosed animals, ASCA successfully disentangled the effects and helped to visualize the homeostatic capacity of the animals [].The independent factors in the experimental design translate into a mathematical model that associates the factors to the measured metabolites. It is essential to question whether an effect found in the sample reflects the effect of this specific factor in the population or that it is merely a sampling fluctuation. This paper tries to answers that question and to provide a way to validate ASCA models. Experiments in metabolomics typically have few samples and normality and equal variances can neither be assumed nor tested. Therefore, we propose a procedure for validating megavariate effects in ASCA without the common assumptions of normality or equal variance.Section two will define the goal and explains some of the theory of statistical validation. That section also explains the ASCA method by defining the model constraints and the used notation scheme. Some of the essential properties, like orthogonality of effect estimates are explained. Explaining ASCA ends with an example of the ASCA model SCA notation. The section that follows details how to randomize the data given the experimental setup of the study. It also details why not to use jackknife or bootstrap, but why permutations are the way to go. A simple example details the model validation, followed by an explanation of how to randomize the data. In section three a simulated data set will serve as an example to certify the validation procedure. Also in that section an experiment with bromobenzene dosed rats will be analyzed and validated. Finally, the last section gives some closing remarks. Methods and Theory2. Definitions and purposeThe experiments in a metabolomics study often follow an experimental design with varying levels of treatment conditions, also known as factors []. Typically the observed metabolic profiles of two different levels of one factor are not the same. This inequality of levels is due to sampling fluctuations and the effect of the varied factor.. ASCA modelsThis section explains the ANOVA-SCA method. The basis of ASCA is the variation partitioning property of ANOVA that allows estimating the effects of the factors encoded in the experimental design []. ASCA has some desirable properties such as orthogonality of effect estimates. Orthogonal effect estimates suit metabolomics experiment analysis well as it allows unique isolation of effect specific variation. Consider, for instance, the case where the treatment regime consists of metabolite data from two dosage levels and three measured time points. ASCA allows isolating the time effects independently from the dosage effects; it can isolate general aging from drug intervention effects.The variation isolation works as in ANOVA; the preceding example of metabolite data from two dosage levels at three time points translates to a two-way ANOVA design. This design consists of two main effects, time and dosage, and a time dosage interaction effect. The main effects and the interaction effect are all orthogonal; this enables perfect isolation of effect specific variation.In the following text the boldface uppercase characters represent matrices (X), vectors are in lowercase bold-italic (x) and scalars in lowercase italic (x). The experimental data is shown as X (I × J). The I rows contain the samples while the J columns in X describe the metabolite levels within the samples.The following text assumes that X is mean centered, that is, the mean of each column in X is , equation .∑i=1Ixij=∀j MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaadaaeWbqaaiabdIha4naaBaaaleaacqWGPbqAcqWGQbGAaeqaaOGaeyypa0JaeGimaaJaeyiaIiIaemOAaOgaleaacqWGPbqAcqGH9aqpcqaIXaqmaeaacqWGjbqsa0GaeyyeIuoaaaa@3BE2@If the matrix Xδ contains the estimates of an effect, then equation defines the sum of squares (SSQ) of that effect, here shown for effect δ.SSQδ=‖Xδ‖=∑i=1I∑j=1J(xij) MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGtbWucqWGtbWucqWGrbqudaWgaaWcbaacciGae8hTdqgabeaakiabg2da9maafmaabaacbeGae4hwaG1aaSbaaSqaaiab=r7aKbqabaaakiaawMa7caGLkWoadaahaaWcbeqaaiabikdaYaaakiabg2da9maaqahabaWaaabCaeaacqGGOaakcqWG4baEdaWgaaWcbaGaemyAaKMaemOAaOgabeaakiabcMcaPmaaCaaaleqabaGaeGOmaidaaaqaaiabdQgaQjabg2da9iabigdaXaqaaiabdQeakbqdcqGHris5aaWcbaGaemyAaKMaeyypa0JaeGymaedabaGaemysaKeaniabggHiLdaaaa@@Xτ, Xδ and Xτδ represent the isolated variation due to time, dose, and their interaction respectively. Xe contains the individual variation that is not induced by the factors.A general two-way ANOVA model is shown in equation , where τ and δ are the main effects with levels c and d, j is the variable index. An example of -way ANOVA model common in the metabolomics field shows in (equation ) how the variation is composed of time effects, dosage effects, interaction between time and dosage and residuals (equation and ). Each of the effect partitions in equation is orthogonal to the others, equation . This orthogonal property allows for the variation decomposition shown in equation []. The effect estimates are not normalized.xcdj = τcj + δdj + (τδ)cdjAlternatively equation can be written in the matrix form, shown in equation .X = Xτ + Xδ + Xτδ + XeThe variation measured in sum of squares, can be uniquely partitioned into the effect, equation .||X|| = ||Xτ|| + ||Xδ|| + ||Xτδ|| +||Xe||2The orthogonality of the effects is shown in equation .XθTXφ=∀{θ,φ}⊂{τ,δ,τδ,e}:θ≠φ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieqacqWFybawdaqhaaWcbaacciGae4hUdehabaGaemivaqfaaOGae8hwaG1aaSbaaSqaaiab+z8aMbqabaGccqGH9aqpcqaIWaamcqGHaiIicqGG7bWEcqGF4oqCcqGGSaalcqGFgpGzcqGG9bqFcqGHckcZcqGG7bWEcqGFepaDcqGGSaalcqGF0oazcqGGSaalcqGFepaDcqGF0oazcqGGSaalcqWGLbqzcqGG9bqFcqGG6aGocqGF4oqCcqGHGjsUcqGFgpGzaaa@@The SCA estimates the information in the partitions time, dosage and dosage time interaction. The two-way ANOVA style ASCA model (equations , ) give the following ASCA model after SCA (equation ):X=TτPτt+TδPδT+T(τδ)P(τδ)T+TePeT+E MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieqacqWFybawcqGH9aqpcqWFubavdaWgaaWcbaacciGae4hXdqhabeaakiab=bfaqnaaDaaaleaacqGFepaDaeaacqWG0baDaaGccqGHRaWkcqWFubavdaWgaaWcbaGae4hTdqgabeaakiab=bfaqnaaDaaaleaacqGF0oazaeaacqWGubavaaGccqGHRaWkcqWFubavdaWgaaWcbaGaeiikaGIae4hXdqNae4hTdqMaeiykaKcabeaakiab=bfaqnaaDaaaleaacqGGOaakcqGFepaDcqGF0oazcqGGPaqkaeaacqWGubavaaGccqGHRaWkcqWFubavdaWgaaWcbaGaemyzaugabeaakiab=bfaqnaaDaaaleaacqWGLbqzaeaacqWGubavaaGccqGHRaWkcqWGfbqraaa@571A@A more detailed review of the ASCA method properties is found elsewhere [].. Type of resampling to useA way to tackle the problem of validating ASCA models is by using resampling techniques, being jackknife, bootstrapping and permutation tests []. The basic idea will be explained by a univariate analysis of two groups of equal size. Later, this will be generalized to the megavariate case. The standard way of testing the difference between group means, with the underlying null-hypothesis that the population group means are not different, is with a t-test. The ANOVA F-test comes down to a t-test for the two group case. Under the assumptions of normality and equal group variances the t-statistic ist=(x¯−x¯)sp MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWG0baDcqGH9aqpdaWcaaqaaiabcIcaOiqbdIha4zaaraWaaSbaaSqaaiabigdaXaqabaGccqGHsislcuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaOGaeiykaKcabaGaem4Cam3aaSbaaSqaaiabdchaWbqabaaaaaaa@3A46@sp=1n(s12+s22) MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGZbWCdaWgaaWcbaGaemiCaahabeaakiabg2da9maakaaabaWaaSaaaeaacqaIXaqmaeaacqWGUbGBaaGaeiikaGIaem4Cam3aa0baaSqaaiabigdaXaqaaiabikdaYaaakiabgUcaRiabdohaZnaaDaaaleaacqaIYaGmaeaacqaIYaGmaaGccqGGPaqkaSqabaaaaa@3CE6@where x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ and x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@ are the group means, n the num-ber of samples in the groups and s1 and s2 are the group standard deviations []. The pooled standard deviation sp can be calculated easily from s1 and s2 given the assumptions of normality and equal group variances. Actually, sp is the standard deviation of (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@), showing the rationale of the t-statistic: a measure of the devia-tion (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@) in its standard deviation units sp. Including the proper degrees of freedom allows for testing the null-hypothesis of equal group means.The bootstrap and jackknife work by resampling the samples in the groups, keeping the grouping structure intact, estimating from those resamplings the group standard deviations s1 and s2. However, this does not directly give the wanted result, because the value needed for the t-test is the standard deviation of (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@). Assuming normality and equal variances, this value can be calculated from the group variances using equation . This is a reasonable assumption for analytical replicates of a sample, but not directly for the biological variation across subjects. The assumption of equal group variances is questionable in this case. These assumptions cannot easily be tested given the small group sample sizes. Thus, it is not clear how to obtain a standard deviation value for (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@) from the jackknifed or bootstrapped s1 and s2 without making extra assumptions.Permutation tests work directly on the variability of (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@) by randomly permuting class labels and recalculating the group-mean differences. Actually, such permutation tests go back a long way [] as an alternative for t-tests and are now also routinely used in gene-expression data analysis, as for instance Significance Analysis of Microarrays (SAM) [].The standard deviation of (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@) has the squared Euclidean distance in its numerator, the denominator is constant over the permutations. In centered data the squared Euclidean distance equals the sum of squares (SSQ) of (x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIXaqmaeqaaaaa@2F59@ − x¯ MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqaIYaGmaeqaaaaa@2F5B@). Using the SSQ as effect statistic, the generalization from univariate to multivariate follows from summing the univariate SSQ's for all variables.. How to randomize the dataRandomizing or permutation is the uncoupling of the data from the group labels [,]. Take note that in data with a zero mean (equation ) the random sampling expectated value is . Considering the level averages, the randomization procedure tests whether the results with randomized labels are as different from zero as the original result is. The randomization, or permutation, does not change the metabolite values for a sample, but it reassigns each sample randomly to one of the treatment groups.. Model validation exampleThis section gives a detailed example of how the permutation works and how it will help to validate models.In most experimental designs it is important to assess the statistical confidence of the effect estimates. An experiment with two measurement series and three measurements in each series will serve as example for the validation. If these series are a and b, the two series comprise the levels of the effect δ, giving the model shown d in equation . This equation holds for both vectors and matrices, shown here is the matrix form of effect of factor δ. The null hypothesis (H0) is the sum-of-squares (SSQ) associated with the effect of factor δ is zero (Equation ). The alternative hypothesis (H1) states the that SSQ of the effect of factor δ is larger than zero.X = Xδ + XeH0 : ||Xδ|| = ;H1 : ||Xδ|| > 0The chosen distance measure that marks how far the group averages are apart is the squared Euclidean distance. In the hypothetical case with a known population, the factors without effect have an SSQ that is zero. Due to the small sample size, the distance between a and b will never be exactly zero, giving an ||Xδ|| > . The SSQ is by its nature also a distance measure that describes how far the effect levels are from zero. In the univariate context, usually variances of the groups are analyzed. In the multivariate context the analysis focuses on SSQ's. The SSQ also conveniently describes the variation in the data.The measured results for series a are , and , for series b the results are -, - and -. These values satisfy equation . The average of level a is and the average of level b is -, shown in equation .x=[ab]=[−−−],xδ=[a¯a¯a¯b¯b¯b¯],xδ=[−−−] MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieWacqWF4baEcqGH9aqpdaWadaqaauaabeqaceaaaeaacqWFHbqyaeaacqWFIbGyaaaacaGLBbGaayzxaaGaeyypa0ZaamWaaeaafaqaceGbbaaaaeaacqaI1aqnaeaacqaI0aanaeaacqaIZaWmaeaacqGHsislcqaIZaWmaeaacqGHsislcqaI0aanaeaacqGHsislcqaI1aqnaaaacaGLBbGaayzxaaGaeiilaWIae8hEaG3aaSbaaSqaaGGaciab+r7aKbqabaGccqGH9aqpdaWadaqaauaabeqageaaaaqaaiqb=fgaHzaaraaabaGaf8xyaeMbaebaaeaacuWFHbqygaqeaaqaaiqb=jgaIzaaraaabaGaf8NyaiMbaebaaeaacuWFIbGygaqeaaaaaiaawUfacaGLDbaacqGGSaalcqWF4baEdaWgaaWcbaGae4hTdqgabeaakiabg2da9maadmaabaqbaeGabyqaaaaabaGaeGinaqdabaGaeGinaqdabaGaeGinaqdabaGaeyOeI0IaeGinaqdabaGaeyOeI0IaeGinaqdabaGaeyOeI0IaeGinaqdaaaGaay5waiaaw2faaaaa@5EA1@Randomization is uncoupling the group labels from the data and randomly reassigning them. To show the randomization the samples with the ± will switch groups. Level a now has the measured values -, and , while level b has , - and -, equation . xp shows the permuted x and xδp MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieWacqWF4baEdaqhaaWcbaacciGae4hTdqgabaGaemiCaahaaaaa@316E@ its average.xp=[−−−],xδp=[−−−] MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieWacqWF4baEdaahaaWcbeqaaiabdchaWbaakiabg2da9maadmaabaqbaeGabyqaaaaabaGaeGynaudabaGaeGinaqdabaacceGae4NeI0ccbeGae03mamdabaGae03mamdabaGaeyOeI0IaeGinaqdabaGaeyOeI0IaeGynaudaaaGaay5waiaaw2faaiabcYcaSiab=Hha4naaDaaaleaaiiGacqaF0oazaeaacqWGWbaCaaGccqGH9aqpdaWadaqaauaabiqageaaaaqaaiabikdaYaqaaiabikdaYaqaaiabikdaYaqaaiabgkHiTiabikdaYaqaaiabgkHiTiabikdaYaqaaiabgkHiTiabikdaYaaaaiaawUfacaGLDbaaaaa@4C7C@The distance between the averages is much smaller in the randomized set than the averages of the original data, equation .‖xδ‖=∗(±)= and ‖xδp‖=∗(±)= MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaadaqbdaqaaGqadiab=Hha4naaBaaaleaaiiGacqGF0oazaeqaaaGccaGLjWUaayPcSdWaaWbaaSqabeaacqaIYaGmaaGccqGH9aqpcqaI2aGncqGHxiIkcqGGOaakcqGHXcqScqaI0aancqGGPaqkdaahaaWcbeqaaiabikdaYaaakiabg2da9iabiMda5iabiodaZiabbccaGiabbggaHjabb6gaUjabbsgaKjabbccaGmaafmaabaGae8hEaG3aa0baaSqaaiab+r7aKbqaaiabdchaWbaaaOGaayzcSlaawQa7amaaCaaaleqabaGaeGOmaidaaOGaeyypa0JaeGOnayJaey4fIOIaeiikaGIaeyySaeRaeGOmaiJaeiykaKYaaWbaaSqabeaacqaIYaGmaaGccqGH9aqpcqaIYaGmcqaI0aanaaa@5A2B@The distance between series a and b is much larger than any of the SSQ's after randomization. There is no permutation that gives a larger SSQ than the original grouping. The larger distance in the original model suggests a significant difference in the series a and b, equation .‖xδp‖≪‖xδ‖⇒H0 probably false MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaadaqbdaqaaGqadiab=Hha4naaDaaaleaaiiGacqGF0oazaeaacqWGWbaCaaaakiaawMa7caGLkWoadaahaaWcbeqaaiabikdaYaaakiablQMi9maafmaabaGae8hEaG3aaSbaaSqaaiab+r7aKbqabaaakiaawMa7caGLkWoadaahaaWcbeqaaiabikdaYaaakiabgkDiElabdIeainaaBaaaleaacqaIWaamaeqaaOGaeeiiaaIaeeiCaaNaeeOCaiNaee4Ba8MaeeOyaiMaeeyyaeMaeeOyaiMaeeiBaWMaeeyEaKNaeeiiaaIaeeOzayMaeeyyaeMaeeiBaWMaee4CamNaeeyzaugaaa@568B@Randomly reassigning multivariate samples to a group works in the same way as described in the preceding paragraphs for univariate data. The randomization leaves the order of metabolites of the sample unaffected. The SSQ, equation , allows for univariate and multivariate calculation of the sum of squares; thereby forming the generalization to the multivariate case.Repeating the randomization procedure many times gives just as many SSQ values. These values define the reference distribution. When most of the randomization SSQ results are larger than the original group assignment result, the effect SSQ is a sampling fluctuations and H0 is not to be rejected. Finally, the probability value or p-value is defined to be the number of SSQs in the reference distribution that is larger than the original SSQ. So when of the SSQs in the reference distribution are larger than the original SSQ, the probability of finding a larger than original SSQ value is (p-value) / = ..A good estimate on the randomized SSQ distribution needs many randomization iterations. How many randomization iterations are enough for a good probability estimate is difficult to establish before starting, because that largely depends on the data. However, repeating the randomization series should give similar results, this suggests enough iteration. The permutations are a random subset of all the possible permutations []. This approach is also known as Monte Carlo resampling.This method validates the multivariate ANOVA partitioning, not the SCA part of ASCA. The SCA method subsequently describes most variation within each partition.. One-Way ANOVA DesignThe preceding example is an example of a one-way ANOVA design with two levels (equation ). To get a reference distribution one can simply permute the group labels. If the original SSQ is larger than most of the reference distribution the model is considered significant, otherwise it is not. In the preceding example series a is significantly different from series b.. Two-Way ANOVA DesignA two-way ANOVA extends the one-way ANOVA design to two factors. In metabolomics experiments common factor examples are time and a drug intervention.Unlike one-way designs two-way ANOVA designs may also include an interaction term. The interaction captures the relation between two main factors. In a time and drug example, the interaction effect means the drug shows a different at different time points.Each main effect needs to be validated separately, getting the reference distribution for the main effect is the same as for the one-way ANOVA. Getting the reference distribution for the interaction term is a bit more complicated. The best option is to permute the residual samples (equation ) []. Residual samples are samples that have the main effects removed.Xr = X - Xτ - Xδ2. Nested ANOVA DesignNested ANOVA designs are extensions of ANOVA design with another factor nested in the main effect. Some special cases need nested ANOVA models, like experiments that measure one animal at different times. The repeated measuring nests the factor time in the animal. The randomization strategy in such cases only allows for placing the animal time series in other levels of the one-way ANOVA factor. In a nested ANOVA design, the permutable unit is the animal itself [,,].. Related MethodsA widely used method in metabolomics is principal component analysis. This method, however, does not take group structures into account, hindering the analysis of effects. Methods that are more closely related to ASCA are SMART and PRC [,]. However, these methods differ on a key issue, namely orthogonality of effect estimates (equation ). The effect estimates of SMART are not orthogonal, as a result the here proposed validation procedure cannot be used. In PRC the effect estimates are orthogonal up to the deflation of the control condition. The proposed validation procedure can be used in PRC as long as it is used before deflating the control effect.. Experimental environmentThe ASCA algorithm was implemented in MATLAB script code, using The MathWorks MATLAB version . release running on Fedora Core on an Intel Corporation Pentium IV (. GHz) computer.The ASCA algorithm is in the download section of our website []. The validation algorithm can be found there as well. Results & DiscussionThis section shows results to certify the proposed validation method for synthetic data and real world experimental data.The real world experiment deals with toxin-dosed rats []. Various other methods already analyzed this experiment; the results strongly suggest the toxin is affecting the animal. This experiment serves as a real world certification of the suggested statistical validation approach in multivariate data sets.. Examples; certifying the procedures with designed dataThis example study showcases two data sets. The first data set has two effect levels that are significantly different. The second data set has two effect levels that are not significantly different. This is to test the suggested procedure in the simple case with known true statistics.ASCA describes each effect level by the averages of the metabolites in that level. In this example study, ASCA will test if the multivariate average of the first rows is different from the multivariate average of the last rows. When many randomizations give an SSQ that is equally large as the original SSQ, the groups probably do not differ. When only a minor fraction of the randomizations give a larger group distance, the groups most likely differ.In the model the effect δ has two levels (equation ). d In the first data set the first rows are filled with ones and the last rows are filled with zeros. Normal distributed white noise (N(σ = , μ = )) is added to this data. The second data set is filled with zeros and white noise (N(σ = , μ = )) is added to it.Figures 1b &1d show the two example data sets, the rows are individual samples and the columns are the metabolites. The colored cells show each metabolite value of every sample.Figure 1Example study to certify the validation procedure, it consists of one significantly different and one nonsignificantly different data set. Figures A and C show the SSQ reference distribution found by permuting the data. If the red dot is outside most the reference distribution and is on the right side, the group is significantly different. The figures B and D show the data from this example experiment. Careful inspection of figure B reveals the top half differs from the bottom half, it is more yellow and red then the bottom half. The D figure lacks this property.In the true significant example the effect δ is designed to be different. The top half of figure 1b has more red colored cells while the bottom half has more blue colored cells. Figure 1a shows the reference distribution of randomized SSQ's, using a vertical line to show the SSQ of the original grouping.Following the proposed validation procedure, the conclusion is clear: the halves are unlikely to be the same because all the permuted SSQ's are smaller than the original SSQ (p = ., SSQ = .). Conclusion: the difference in levels is significant. This model validation used , randomization iterations taking about minutes of computing time.Repeating the validation procedure on data without a designed difference between the two dosage levels, serves as a negative control. The level averages will differ a little, but these differences are sampling fluctuations.Figure 1d is similar to figure 1b but without the designed differences in the dosage levels. Figure 1d does not show a seeming difference between the top and bottom half. Figure 1c shows the reference distribution for the data set equal level averages.Following the proposed validation procedure, the conclusion is clear: the halves are likely to be the same because many (.%) of the permuted SSQ's are larger than the original SSQ (p = ., SSQ = .). Conclusion: the difference in levels is not significant. This model validation used , randomization iterations taking about minutes of computing time.To test if the proposed validation procedure rejects the H0 in the fraction of the significance threshold, the model from equation was used with different realisations of white noise (σ = , μ = ). With an significance threshold of α = ., of the H0's are expected to be rejected. The number of rejections were in the expected range, given a % confidence interval from a binomial distribution with α = . for n = .. Experimental results: Rats dosed with hep-atotoxicant bromobenzeneIn this experiment there are five groups of rats; a control group, a corn oil (the toxin vehicle) control group and a low, medium and high dosage of bromobenzene. The collected urine from three individual rats of each treatment group is measured on the NRM platform, at , and hours after the toxin administration [,]. The rats are sacrificed after each sampling to collect tissue sample for histology and transcriptomics analysis.One sample from the highest dosage group is missing. To avoid unbalanced ANOVA issues we assume this missing sample equals the average of the two samples collected and measured from that group at the same time point.The main effects and the interaction effect of the -way ANOVA models were tested by the ASCA validation. Here the focus is on the factor dosage and dosage-time interaction. The models are significant, with a drug dose difference p ≤ ., SSQ = . (figure 2a) and dosage-time interaction p ≤ ., SSQ = . (figure 2b). The interaction significance was calculated on the residuals, thus after removing the time and dosage effect (equation ). The not nested experimental design allows the use of a simple two-way ANOVA permutation scheme.Figure 2Validation of the ASCA model for bromobenzene treated rats, validation of the dosage and the dosage-time interaction and the Xδ + Xτδ score plot. This experiment deals with the urine analysis of bromobenzene treated rats, the experimental design includes two types of controls and dosage levels of the hepatotoxicant bromobenzene. The dosage and the interaction models are both significant as is clear from the reference distributions (p ≤ .). Because the dosage and the interaction models are significant they are superimposed and analyzed by SCA. The score plot of the SCA solution is shown. From this plot it is clear by visual inspection that the average dosage levels differ and that the interaction effect exists.The dosage and the interaction effect are significant, combining the dosage and interaction gives a data set that describes all effects that depend on dosage (equation ). SCA helps to reduce the dimensionality of this data set.Xδ+τδ = Xδ + XτδXδ+τδ=T(δ+τδ)P(δ+τδ)T+E MathType@MTEF@@@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieqacqWFybawdaWgaaWcbaacciGae4hTdqMaey4kaSIae4hXdqNae4hTdqgabeaakiabg2da9iab=rfaunaaBaaaleaacqGGOaakcqGF0oazcqGHRaWkcqGFepaDcqGF0oazcqGGPaqkaeqaaOGae8huaa1aa0baaSqaaiabcIcaOiab+r7aKjabgUcaRiab+r8a0jab+r7aKjabcMcaPaqaaiabdsfaubaakiabgUcaRiab=veafbaa@4A15@SCA summarizes the validated toxin and interaction variation. Grouping the scores (T in equation ) according to the factor levels gives figure 2c. The conclusion is the treatment with the hepatotoxicant differs between dosage groups and the dosage responses change over time. Additionally, the results suggest the animals treated with the lowest dosage fully recover or go back to the state of the controls. The animals dosed with the medium dosage need more time, but also go back to the control state. The animals given the highest dosage do not recover to the control state. Histological liver examination revealed extensive damage caused by the bro-mobenzene, corroborating these findings. ConclusionExtending ASCA with a permutation procedure enables validation of ASCA models. Referencing the ASCA models to the permutation based reference distribution gives validation statistics. If the model is significant, the following SCA decomposition describes the validated induced effects.The proposed method gives validation statistics to the ASCA models. ASCA itself allows for summarizing the designed experimental data. Combining ASCA and the ASCA model validation forms a powerful summary of designed experimental data.
|
PMC2360410.txt
|
TITLE: Enhanced repair of DNA interstrand crosslinking in ovarian cancer cells from patients following treatment with platinum-based chemotherapy AUTHORS: P Wynne, C Newton, J A Ledermann, A Olaitan, T A Mould, J A Hartley ABSTRACT: Despite high tumour response rates to platinum-based chemotherapy in ovarian cancer survival is poor due to the emergence of drug resistance. Mechanistic studies in clinical material have been hampered by the unavailability of sensitive methods to detect the critical drug-induced effects in individual cells. A modification of the single cell gel electrophoresis (comet) assay allows the sensitive detection of DNA interstrand crosslinking in both tumour and normal cells derived directly from clinical material. Tumour cells isolated from ovarian cancer patients were treated ex vivo with μM cisplatin for h and crosslink formation and repair (unhooking) measured. No significant difference in the peak level of crosslinking in tumour cells was observed between patients who were either newly diagnosed or previously treated with platinum-based therapy, or between tumour and mesothelial cells from an individual patient. This indicates no difference in cellular mechanisms such as drug transport or detoxification. In contrast, the percentage repair (unhooking) of DNA interstrand crosslinks was much greater in the group of treated patients. At h in the newly diagnosed patient tumour samples, only one gave >% repair and gave <% repair; however, out of treated patient samples gave >% repair and showed >% repair. The estimated median difference (newly diagnosed minus treated) was − (% CI − to −), and the P-value from a Mann–Whitney test was <.. In eight patients, it was possible to obtain tumour samples prior to any chemotherapy, and also on relapse or at interval debulking surgery following platinum-based chemotherapy. In these patients, the mean % repair prior to therapy was . rising to . following treatment. These data demonstrate increased repair of DNA interstrand crosslinks in ovarian tumour cells following platinum therapy which may contribute to clinical acquired resistance. BODY: The lifetime risk of a woman developing ovarian cancer is in and around two thirds of these patients present with advanced disease (Ozols et al, ). The standard first line treatment for ovarian cancer is cytoreductive surgery followed by carboplatin alone or more commonly in combination with paclitaxel (du Bois et al, ). This treatment results in a complete response in the majority of women; however, most responding patients eventually relapse with disease that becomes resistant not only to platinum compounds, but also to a wide range of other chemotherapeutic agents (Salzberg et al, ). The prognosis for women with relapsed ovarian cancer remains poor with a -year survival of % (Colombo et al, ). A greater understanding of the mechanisms underlying drug resistance could lead to measures to overcome it and improve survival.Resistance to chemotherapeutic agents such as carboplatin can be intrinsic or acquired (Perez, ). Intrinsic resistance is present at the time of diagnosis, and the patient fails to respond to first line chemotherapy. Studies in ovarian cancer cell lines have shown that acquired resistance to platinum drugs can be multifactorial, consisting of mechanisms which include altered drug transporter proteins (Kelland, ), increased drug inactivation, for example, by binding of drugs to glutathione (Kelland, ), evasion of apoptosis by mutation of genes such as p53 (Manic et al, ) and enhanced ability to repair DNA damage such as by upregulation of ERCC1 (Ferry et al, ). It is still unclear, however, which of these contribute most to acquired drug resistance in the clinical setting. Studies in clinical material have been hampered by the unavailability of sensitive methods to detect the critical drug-induced effects in individual cells.The cytotoxicity of carboplatin and cisplatin results from the formation of platinum-DNA adducts which include monoadducts, intrastrand crosslinks, interstrand crosslinks (ICLs) and DNA-protein crosslinks (Zwelling et al, ; Comess and Lippard, ). Intrastrand crosslinks constitute the majority (>%) of lesions formed on cellular DNA and these distorting lesions are repaired by nucleotide excision repair (Fichtinger-Schepman et al, ). ICLs, which link the two complementary strands of DNA together, comprise less than % of the total lesions on DNA but are highly cytotoxic and difficult to repair (McHugh et al, ).We have previously demonstrated that a modification of the single cell gel electrophoresis (comet) assay (Spanswick et al, ) can be used successfully in the clinical setting to detect and quantify the levels of ICLs in patient lymphocytes and tumour cells at pharmacologically relevant doses of bifunctional alkylating agents (Hartley et al, ; Webley et al, ; Spanswick et al, ; Corrie et al, ). The method has also been used to measure cisplatin-induced ICLs in vitro (De Silva et al, ). In this study, we have used the method to compare the formation and repair (unhooking) of ICL's following ex vivo exposure to cisplatin in tumour cells and normal cells isolated from ovarian cancer patients who were either newly diagnosed, or had been previously treated with platinum-based chemotherapy.MATERIALS AND METHODSPatient populationEthics approval was gained from the Joint UCL/UCLH Committee on the Ethics of Human Research. Ovarian cancer patients receiving treatment between February and February were recruited to take part in this study. Solid tumour tissue or ascitic fluid was obtained from ovarian cancer patients aged between and years. Samples were obtained at diagnosis, interval debulking surgery (IDS) or at relapse. In some cases, paired samples were obtained at diagnosis and IDS, or at relapse.Preparation of tumour and non-tumour cells from clinical materialAscitic fluid was aliquoted into plastic ml conical tubes and spun at × g for min. Cell pellets were resuspended in Dulbecco's modification of Eagle's medium (DMEM) containing % fetal calf serum (FCS) and mM glutamine, and seeded into large tissue culture flasks. All cells were maintained in a humidified atmosphere with % CO2 at °C. After h, the entire volume of tissue culture medium in each flask, containing unattached cells was transferred into a fresh large tissue culture flask, and DMEM (with FCS and glutamine) was replaced in the original flasks. Non-tumour cells generally attached to the plastic surface within the first hour, whereas tumour cells required a longer period of incubation. Tumour cells also required a longer period of time to detach in response to trypsin, compared to the non-tumour, mesothelial cells. Further purification of the tumour samples was achieved by trypsinisation until the contaminant mesothelial cells were seen to detach, while the tumour cells remained in situ.In sterile conditions, primary tumour was finely dissected and flushed with DMEM containing % FCS and mM glutamine, to produce a single cell suspension, which was seeded into large tissue culture flasks.ImmunocytochemistryAntibodies to CA125 (Novocastra, Newcastle, UK) and AUA1 (Skybio, Bedfordshire, UK) which stain ovarian tumour cells and not mesothelial cells, and to Calretinin (Zymed, Cambridge, UK) and CK5 which stain mesothelial cells but not ovarian cancer cells, were used to differentiate the two cell types in cytospin preparations using standard immunocytochemical techniques. Analysis was performed on the same cell population that was treated with cisplatin.Treatment of tumour and non-tumour cells ex vivo with cisplatinPrimary cultures of tumour and mesothelial cells were trypsinised and seeded at a concentration of × cells per ml into -well plates. Cells were left to attach overnight. Half of the samples were non-drug-treated controls, the other half were drug treated with μM cisplatin (David Bull Laboratories, Australia), diluted in DMEM, for h at °C in a humidified atmosphere with % CO2. The cisplatin was removed and fresh DMEM with % FCS and mM glutamine was added to the samples. Immediately after drug treatment, a drug treated and a non-drug-treated control sample were trypsinised, centrifuged at g for min, then resuspended in FCS with % DMSO. Samples were then frozen in a polystyrene box within a −°C freezer. This procedure was repeated ., , and h after drug exposure.Single cell suspensions were prepared at a cell density of × cells per ml. Cells were treated with μM cisplatin in DMEM at °C, % CO2. After exposure, cell samples were centrifuged at g for min, and then resuspended with DMEM with % FCS and mM glutamine. Immediately after drug treatment, a drug treated and a non-drug-treated control samples were trypsinised, centrifuged at g for min, and then resuspended in FCS with % DMSO. Samples were then frozen in a polystyrene box within a −°C freezer. This procedure was repeated ., , and h after drug exposure.Measurement of DNA interstrand crosslinking using the single cell gel electrophoresis (comet) assayThe details of the modified single cell gel electrophoresis (comet) assay to measure DNA ICLs are described in detail elsewhere (Hartley et al, ; Spanswick et al, ). All procedures performed on the single cell suspension sample were carried out on ice and in subdued lighting. All chemicals used were obtained from Sigma Chemical Co. (Poole, UK) unless otherwise stated. Immediately before analysis, cells were irradiated (. Gy, . Gy min−) to deliver a fixed number of random DNA strand breaks. After embedding cells in % agarose on a precoated microscope slide, the cells were lysed for h in lysis buffer ( mM disodium EDTA, . M NaCl, mM Tris-HCl pH .) containing % Triton X- added immediately before analysis, and then washed for h in distilled water, changed every min. Slides were then incubated in alkali buffer ( mM NaOH, mM disodium EDTA, pH .) for min followed by electrophoresis in the same buffer for min at V (. V cm−), mA. The slides were finally rinsed in neutralising buffer (. M Tris-HCl, pH .) and then in saline.After drying, the slides were stained with propidium iodide (. μg ml−) for min and then rinsed in distilled water. Images were visualised using a NIKON inverted microscope with a high-pressure mercury light source, – nm excitation filter and nm barrier filter at × magnification. Images were captured using an online CCD camera and analysed using Komet Analysis software (Kinetic Imaging, Liverpool, UK). For each duplicate slide, cells were analysed. The tail moment for each image was calculated using the Komet Analysis software as the product of the percentage DNA in the comet tail and the distance between the means of the head and tail distributions, based on the definition of Olive et al (). Crosslinking was expressed as the percentage decrease in tail moment compared to irradiated controls calculated by the formula: where Tmdi is the tail moment of drug-treated irradiated sample, TMcu the tail moment of untreated, unirradiated control and TMci the tail moment of untreated, irradiated control.Statistical analysisStatistical analyses were performed in Minitab version .. Probability plots were observed to determine whether the three variables (percentage decrease in tail moment, the paired difference in tail moment decrease between tumour and mesothelial cells, and percentage repair at h) were normally distributed. If they were, unpaired or paired t-tests were performed, and the mean difference with % CI obtained. If the distributions were not normal, the median was used as the measure of central tendency and the Mann–Whitney non-parametric test used to examine differences between groups. Minitab also provides an estimate of the median difference between newly diagnosed and treated patients, with % CI.RESULTSMeasurement of DNA interstrand crosslinking in ovarian tumour cells treated ex vivo with cisplatin using the single cell gel electrophoresis (comet) assaySamples were obtained from patients prior to any platinum-based chemotherapy (Table 1A) and from patients following platinum-based chemotherapy (Table 1B). In eight cases (patients , , , , , , and ) paired samples were obtained at diagnosis, and at relapse or IDS following platinum-based chemotherapy. Either primary tumour cell cultures from drained ascitic fluid or single cell suspensions from ovarian tumours from surgery were obtained. Immunohistochemistry was used to determine the purity of the tumour cell population. In all cases, the cell sample contained >% tumour cells and in the majority of cases it was >%.Cells were treated with cisplatin for h at μM. This dose was determined from pilot experiments in human ovarian cancer cell lines to give an optimal level of DNA ICLs as determined by the single cell gel electrophoresis (comet) assay. Following h treatment cells were re-suspended in drug-free medium and samples taken for measurement of DNA crosslinking at h (the peak of crosslinking with cisplatin), and h. DNA crosslinking was expressed as the % decrease in tail moment compared to control non-drug treated cells as previously described (Hartley et al, ; Spanswick et al, ). Crosslink response curves for patient samples and are shown in Figure . In the tumour cells from patient (Figure 1A) around % decrease in tail moment is observed at the peak of crosslinking ( h). By h, the majority (>%) of the crosslinks have been repaired or ‘unhooked’ from the DNA. In contrast, in the cells from patient (Figure 1B) although the level of crosslinks at the peak is similar, very little unhooking is observed at h (<%) and the majority of crosslinks persist at h.Peak level of cisplatin-induced DNA interstrand crosslinks in patient tumour and mesothelial cellsThe peak ( h) level of DNA interstrand crosslinking was determined in all patient tumour samples, following treatment with μM cisplatin. The data are presented in Figure with the newly diagnosed patients and those treated with platinum-based chemotherapy shown separately. A high level of crosslinking was observed in all the samples tested with the % decrease in tail moment ranging from to %. The mean level of crosslinking in all samples was .. The percentage decrease in tail moment was normally distributed. The mean difference between the two groups (newly diagnosed minus treated) is −, with % CI − to . The P-value from an unpaired t-test was ., indicating no evidence of a real difference.In many of the samples derived from ascitic fluid, it was possible to isolate mesothelial cells to act as a non-tumour direct comparison within the same patient. Data for the matched samples are shown in Figure . The paired difference between the percentage decrease in tail moment (mesothelial minus tumour cells for each patient) was normally distributed. In the newly diagnosed cases, the mean difference between tumour and control cells was . (% CI −. to .), with a P-value from a paired t-test of .. In the seven treated cases, the mean difference was . (% CI −. to .), P-value of .. Therefore, in each group, there was no evidence of a difference between tumour and mesothelial cells. These data demonstrate that tumour cells and mesothelial cells do not differ significantly in their uptake or cellular metabolism of cisplatin thereby allowing similar levels of DNA damage to occur.Repair of cisplatin-induced crosslinks in patient tumour cellsThe ability of the tumour cells to repair the DNA ICLs produced by cisplatin was determined from the crosslink response curves produced for each patient sample. The level of crosslinking was compared at and h and the % repair at h calculated. These data are shown in Figure . A highly heterogeneous response was observed between the different patient samples ranging from no repair to almost % repair at h. In some samples, the level of crosslinking was even slightly higher at h than at h resulting in a ‘negative’ % repair value. Strikingly, the response in the samples from newly diagnosed patients was generally very different to that in the samples from platinum-treated patients. In the newly diagnosed patients, only one gave a level of repair above % and gave <% repair. In contrast, out of previously treated patients gave >% repair and showed >% repair. The mean % repair was . in the newly diagnosed patients compared to that of . in the treated patients. Percentage repair at h was not normally distributed. The estimated median difference (newly diagnosed minus treated) is − (% CI − to −), and the P-value from a Mann–Whitney test was <.. These results show that the percentage repair was much greater in the group of treated patients.In the mesothelial samples, the repair response was more homogeneous than in the matched tumour samples. The mean % repair in the mesothelial samples was .±. (.±. in the samples from newly diagnosed patients and .±. in the seven samples from treated patients).Repair of cisplatin-induced crosslinking in tumour cells from the same patient before and after platinum-based chemotherapyIn eight patients, it was possible to obtain tumour samples prior to any chemotherapy, and also on relapse following platinum-based chemotherapy. The % repair values for these patients are shown in Figure . In four patients, the second sample was taken at IDS (Figure 5A) and in the other four, the second sample was taken at relapse following a treatment-free interval < months (Figure 5B). These two clinically distinct groups showed similar changes. In five out of the eight patients prior to chemotherapy the tumour cells did not show any repair of crosslinks at h, and <% repair at h was observed in the other three. In contrast, the tumour cells following chemotherapy show extensive unhooking of crosslinks in each case with % repair ranging from . to ., with seven of the eight samples showing >% repair. In these eight patients, the mean % repair prior to therapy was .±. rising to .±. following treatment.DISCUSSIONThe data presented here clearly demonstrate that, in tumour cells isolated from ovarian cancer patients, the peak level of DNA interstrand crosslinking produced by the chemotherapeutic drug cisplatin is very similar (mean .±.), as determined by the single cell gel electrophoresis (comet) assay. This is irrespective of whether the tumour sample was from a newly diagnosed patient, or one who had been treated with platinum-based chemotherapy. This would indicate that any molecular mechanism of drug resistance that has been evoked following chemotherapy does not involve an ‘upstream’ mechanism (e.g., altered drug transport, increased detoxification) which would prevent the drug from reaching its cellular target, DNA. Similarly, tumour cells and mesothelial cells from the same patient do not differ significantly in their uptake or cellular metabolism of cisplatin thereby allowing similar levels of DNA damage to occur.The single cell gel electrophoresis (comet) assay allows DNA interstrand crosslinking to be measured in clinical samples at pharmacologically relevant doses of crosslinking drug. This can be used to measure crosslinking in lymphocytes or solid tumour material where samples are taken following treatment of patients with drugs such as ifosfamide (Hartley et al, ), treosulfan (Corrie et al, ) or antibody directed enzyme pro-drug therapy (Webley et al, ). Alternatively, it can be used to measure crosslink formation and repair in cells isolated from patients and treated ex vivo with drug as in the present study, or as previously demonstrated in myeloma plasma cells treated with melphalan (Spanswick et al, ). In the latter study, myeloma cells from chemotherapy naïve patients were all incapable of repairing melphalan-induced crosslinks at h after the peak of formation. Cells from melphalan resistant patients all showed significant repair ranging from to % repair at h. In the current study, the repair of crosslinking was more heterogeneous in both the newly diagnosed and treated patient populations but the overall trend to increased repair of interstrand crosslinking was clearly evident.It should be noted that repair as measured by the comet assay is really the ‘unhooking’ of one arm of the crosslink to release the covalent linkage of the two strands of the double helix. This is the first step in the complex molecular mechanism of repair of DNA ICLs (McHugh et al, ) and the comet assay cannot determine if the repair process has gone to completion and correctly restored the integrity of both strands of the DNA. Mammalian cells defective in the unhooking step of cisplatin interstrand crosslink repair, as measured using the comet assay, include cells bearing mutations in nucleotide excision repair (e.g., XPB, XPD, XPG, ERCC1 and XPF) and homologous recombination (e.g., XRCC2 and XRCC3) (De Silva et al, ). Cells defective in ERCC1 are highly sensitive to cisplatin and several groups have investigated the influence of ERCC1 on resistance to platinum chemotherapy (Ferry et al, ; Reed, ) and suggest that ERCC1 is a good marker for cellular or clinical resistance to these drugs. In ERCC1 mutant cells, however, the high cisplatin sensitivity observed compared to other mutant cells which are equally defective in the unhooking step of interstrand crosslink repair is most likely due to a defect other than in excision repair (De Silva et al, ).It has previously been demonstrated that the repair of DNA ICLs produced by cisplatin, measured by the technique of alkaline filter elution, was reduced in human lymphocytes from normal volunteers aged around compared to those from volunteers aged around (Rudd et al, ). In the current study, the age range of patients was from to and there was no correlation between age and extent of repair in the newly diagnosed patient tumour samples.Platinum compounds are the most active agents in ovarian cancer treatment and the decision to retreat recurrent disease with platinum is based on clinical observations that have shown the likelihood of response is dependent on the platinum-free interval (Blackledge et al, ; Markman et al, ). The study of ex vivo treatment of tumour samples from women with newly diagnosed and relapsed ovarian cancer has identified biochemical changes in the formation and repair of cisplatin-induced DNA crosslinks that provide new information on some of the mechanisms associated with resistance to platinum in patients with ovarian cancer. Firstly, the ability of cisplatin to form DNA crosslinks is similar in normal (mesothelial) and tumour tissue and similar levels of crosslinking were seen in patients whose tumours were exposed to in vitro cisplatin after a ‘platinum-free interval’ of less than or greater than months. Secondly, in comparison to the platinum naïve group, there were marked differences in the repair of platinum-induced crosslinks in tumour cells removed at IDS or after relapse at a less than or greater than months platinum-free period. In the previously treated group, % showed greater than % repair compared with % in chemonaïve patients.A Cox regression was used to examine the association between progression-free survival and percentage DNA repair in the newly diagnosed patient samples. There was no evidence of an association (the hazard ratio for an increase of percentage point was ., % CI .–., P-value=.). No relationship was therefore evident between repair of ICLs and inherent sensitivity. In the case of treated patients, the number in each category (IDS, platinum-free interval (PFI) > months, PFI < months) was too small to perform the equivalent analysis.The paired samples (Figure ) allow further conclusions to be drawn. In all four paired samples taken pre-treatment and then at IDS there was an increased ability to repair platinum-induced crosslinks after – cycles of platinum-based chemotherapy. This suggests that significant changes in the tumour have either developed, or become evident through selection after as few as three cycles of chemotherapy. In three of these cases, the outcome after chemotherapy and surgery was a complete response, but relapse occurred between and months in all four cases. As previously stated, the unhooking of DNA crosslinks is only one of a number of events leading to repair and contributing to clinical resistance. Even in a larger group of patients, it is unclear whether this early change in the tumour metabolism has clinically meaningful information. Similarly, within the sub-group of clinical ‘platinum-sensitivity’ or ‘-resistance’, it is difficult to draw conclusions about a relationship of repair to progression-free survival on further treatment as the number of patients per group is small and the treatment given at relapse varied. Furthermore, the definition of platinum-sensitivity is a clinical one and represents an empirically defined grouping of patients, based on an observed probability of response to platinum re-challenge. However, there is a consistent pattern within the paired samples.For the four patients with paired samples at relapse/progression, a significant increase in repair was also seen compared to their pre-treatment sample. Three patients were considered too unwell for further treatment at this point but one (patient ), treated with cisplatin and etoposide had a partial response lasting more than months. Whilst this study does not assist the clinical decision process about the choice of therapy for first or subsequent line therapy, it clearly shows that changes in the tumour metabolism of cisplatin-induced crosslinks evolve quickly after platinum-based therapy and that the mechanisms of clinical resistance are likely to involve the repair and processing of DNA ICLs. The early appearance of these differences merits further investigation in a larger number of patients treated with platinum-based therapy to determine any relationship between this enhanced DNA repair and clinical outcome.
|
PMC2771102.txt
|
TITLE: Matrix Metalloproteinase Is Necessary for the Migration of Human Bone Marrow-Derived Mesenchymal Stem Cells Toward Human Glioma AUTHORS: Ivy A W Ho, Kelly Y W Chan, Wai-Hoe Ng, Chang M Guo, Kam M Hui, Philip Cheang, Paula Y P Lam ABSTRACT: Human mesenchymal stem cells (MSCs) have increasingly been used as cellular vectors for the delivery of therapeutic genes to tumors. However, the precise mechanism of mobilization remains poorly defined. In this study, MSCs that expressed similar cell surface markers and exhibited multilineage differentiation potentials were isolated from various donors. Interestingly, different MSC isolates displayed differential migration ability toward human glioma cells. We hypothesized that distinct molecular signals may be involved in the varied tumor tropisms exhibited by different MSC isolates. To test this hypothesis, gene expression profiles of tumor-trophic MSCs were compared with those of non–tumor-trophic MSCs. Among the various differentially regulated genes, matrix metalloproteinase one (MMP1) gene expression and its protein activities were enhanced by -fold and -fold, respectively, in highly migrating MSCs compared with poorly migrating MSCs. By contrast, there was no change in the transcriptional levels of other MMPs. Functional inactivation of MMP1 abrogated the migratory potential of MSCs toward glioma-conditioned medium. Conversely, the nonmigratory phenotype of poorly migrating MSC could be rescued in the presence of either recombinant MMP1 or conditioned medium from the highly migrating MSCs. Ectopic expression of MMP1 in these poorly migrating cells also rendered the cells responsive to the signaling cues from the glioma cells in vivo. However, blocking the interaction of MMP1 and its cognate receptor PAR1 effectively diminished the migratory ability of MSCs. Taken together, this study provides, for the first time, supporting evidence that MMP1 is critically involved in the migration capacity of MSCs, acting through the MMP1/PAR1 axis. Stem Cells ;:– BODY: INTRODUCTIONHuman mesenchymal stem cells (MSCs) are nonhematopoietic adult stem cells with multipotent capacities. The innate tropism of MSCs for tumors, combined with the fact that these cells can be expanded to a clinical scale production with ease, have prompted great interest in using MSCs to deliver antitumor agents to the tumor microenvironment. This is of particular importance to targeting tumor cells with invasive capacity such as glioma cells. However, the mechanism and factors responsible for the tumor tropism of MSCs remain fully elucidated.MSC migration has been postulated to be similar to hematopoietic stem cell (HSC) migration as both cell types reside in the bone marrow. One of the most widely recognized receptor/ligand pairs involved in HSC trafficking is the stromal cell derived factor- (SDF-; also know as chemokine [C-X-C motif] ligand or CXCL12) and its receptor, chemokine receptor four (CXCR4), which are crucial for the homing and engraftment activities of HSC [–] as well as for recruitment of endothelial progenitor cells to sites of ischemic tissues []. However, the role of SDF-/CXCR4 in MSC mobilization is less clear compare with HSCs. In fact, recent studies have suggested that SDF-/CXCR4 axis may not play a major role in MSC migration. Findings from Ip et al. demonstrated that the blocking of CXCR4 receptors had no impact on murine MSC migration []. Furthermore, CXCR4/SDF- does not possess the same migration importance in the ability of HSC to migrate and engraft in immunodeficient animals unless enforced expression of CXCR4 is employed [,]. This finding is recently confirmed in the context of MSCs in which enforced expression of CXCR4 by lentiviral gene transfer was demonstrated to enhance homing of MSCs to bone marrow in vivo []. Clearly, a better understanding of the mechanisms that regulate the migration abilities of MSCs is needed before these cells could be employed as potent agents for delivering therapeutic genes to tumor cells.The signaling cues to mobilize MSCs appear to be dependent on the physiologic/pathologic status of the local environment. For example, the hepatocyte growth factor/c-met signaling pathway has been implicated in MSC mobilization and recruitment to damaged tissues [,]. Under hypoxic stress, MSCs exhibited an increase in the migratory propensity induced by basic fibroblast growth factor [] and vascular endothelial growth factor in the damaged regions []. In the tumor microenvironment, recruitment of MSCs to tumor cells was associated with elevated expression of proinflammatory molecules such as interleukin-(IL) , IL-, and monocyte chemoattractant protein-, which may be mediated by the activation of urokinase plasminogen activator and its receptors on human tumor cells []. At present, it is unclear whether cytokines, chemokines, growth factors, or proteolytic enzymes activate the migratory process of MSCs. This is because many of the studies on MSC migration involved profiling the conditioned medium of MSCs that led to the identification of many candidate factors or targeted knockdown of genes which have been shown to play a role in cellular migration. We have taken a different approach by performing gene expression profiles on batched MSCs that exhibited high migratory activities versus those that migrated poorly. This strategy has allowed us to focus on genes whose products exert a significant influence on the migration of MSCs toward tumor cells. The candidate gene, matrix metalloproteinase one or MMP1, was found to be highly upregulated in highly migratory MSCs that exhibited the tumor trophic property and was chosen for subsequent studies.Cellular migration is a complex process that involves the breakdown of extracellular matrix (ECM) detachment of cells from the basal membrane, migration of cells from original location, survival of cells during the migration process, intravasation into target tissue, and finally, interaction of the migrated cells with the target microenvironment []. The degradation of ECM during the migration process requires the action of proteolytic enzymes such as metalloproteinases (MMPs), which are zinc-dependent endopeptidases []. MMPs are secreted as inactive proenzymes or zymogens that are activated by the cleavage of the prodomain []. Depending on the substrate specificity and structure, they are divided into several subgroups: collagenases (e.g., MMP1), stromelysins (e.g., MMP3; MMP10), matrilysins (e.g., MMP7; MMP26), gelatinases (e.g., MMP2; MMP9), and membrane-type matrix metalloproteinase- (MT1-MMP). In particular, interstitial collagenase (MMP1) has been reported to be involved in the invasion of breast carcinoma []. Stromal-derived MMP1 was recently shown to cleave and activate the G-protein-coupled receptor, protease-activated receptor one (PAR1), leading to activation of intracellular signal that regulates the invasion process in breast cancer cells [].In the present study, the functional role of MMP1 in the migratory activities of various MSC isolates was studied. Targeted knockdown of endogenous MMP1 was shown to inhibit the migration ability of MSCs in vitro. Conversely, exogenous expression of MMP1 in poorly migrating MSCs could reconstitute the tumor trophic abilities of these cells in vitro and in vivo. In addition, the disruption of interaction between MMP1 and PAR1 was found to severely impair the migration ability of MSCs. Taken together, our results showed, for the first time, the functional importance of the MMP1/PAR1 axis in modulating the migration of MSCs toward human glioma.MATERIALS AND METHODSCell Culture and RNAi TransfectionThe Institutional Review Board of National Cancer Center and Singapore General Hospital have approved this study. Isolation and characterization of MSCs was performed as previously described []. Primary glioma was obtained from a brain tumor biopsy of a patient diagnosed with grade IV gliomas. Normal human astrocytes (NHA), ΔGli36, and - cells were cultured as described previously []. Full methods are available as supporting information.RNAi transfection was performed using Lipofectamine RNAiMax (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Stealth negative control (medium GC; Invitrogen) was used as control. In brief, all RNAis were transfected at a final concentration of nM into × cells cultured in a six-well dish (BD Biosciences, Franklin Lakes, NJ, http://www.bdbiosciences.com) according to the manufacturer's protocol.Construction and Packaging of pHGCX-MMP1 Herpes Simplex Virus- Amplicon Viral VectorThe pHGCX-MMP1 Herpes Simplex Virus (HSV)- amplicon plasmid vector which contains the eGFP gene under the control of the viral immediate early promoter (IE4/) was obtained from Dr. Y Saeki (Massachusetts General Hospital, Boston, MA). The MMP1 gene was inserted into the KpnI and NotI site located downstream of the strong CMV promoter. See supporting information Methods for detail information.In Vitro Migration AssayA Modified Boyden chamber assay was used to investigate the in vitro migration of MSCs. MSCs ( × ) were cultured in a -well tissue culture insert with an μm pore size membrane (BD Biosciences). Migration of MSCs across the membrane was subsequently determined by counting the number of propidium iodide-stained nuclei on the underside of the membrane under × magnification. Full method is included in the supporting information Methods.In Vivo Migration AssayMSCs ( × ), suspended in μl of complete medium, were injected into the contralateral hemisphere of ΔGli36 human glioma cells-bearing ( × ) mice days post-tumor implantation (Bregma [, ], . mm lateral, . mm depth). Migration of MSCs from the site of injection was assessed weeks post-MSC cells implantation. Quantification of migrated MSC was performed on single-cell suspension using flow cytometry. See supporting information Methods for detail. All animal experiments were performed according to the guidelines and protocols approved by the Institutional Animal Care and Use Committee at the Singapore General Hospital.Varani Migration AssayCell migration was quantified by recording the number of cells migrated away from the agarose drops using a method described by Varani et al. []. MSCs were resuspended ( × cells per milliliter) in culture medium containing .% low melting point agarose and maintained at °C. Drops of cell suspensions of approximately .– μl were applied to the center of the wells in a -well tissue culture dish. The dish was then placed on ice for minutes to allow the agarose to solidify. The cell-laden agarose droplets were then slowly covered with μl of either glioma-conditioned medium or fresh culture medium. Cell migration was measured daily for days. Each sample was repeated in triplicate and the experiment was repeated twice. Migrated cells were visualized using wide-field microscopy with an inverted microscope (TE300; Nikon, Tokyo, Japan, http://www.nikon.com), and images were acquired on a CCD color digital camera (DXM1200F) using image acquisition software (ACT- v2.; Nikon).Affymetrix GeneChip AnalysisThe Affymetrix cDNA GeneChip Human Genome U133 Plus . Array which composed of , transcripts was used to identify factors that influence the migration of MSC. We compared the gene expression profiles of highly migratory MSCs (MSC-; MSC-) and lowly migratory MSCs (MSC-; MSC-). A total of two independent hybridizations were performed using cells of either passage or . Five microgram of total RNA was converted into double-stranded cDNA using a T7-(dT) primer containing T7 RNA polymerase promoter and Superscript II reverse transcriptase (Affymetrix Inc., Santa Clara, CA, http://www.affymetrix.com). cRNA labeling, hybridizations, washes, and scan steps were performed according to manufacturer's instructions (Affymetrix Inc.). Probe arrays were scanned using the Affymetrix Microarray Suite, and images were imported as CEL files into Partek Genomic Suite (Partek Inc., St. Louis, MO, http://www.partek.com) for analysis. Genes of interest were matched to those in the Affymetrix NetAffix Gene Ontology analysis system. The expression microarray has been submitted to the Gene Expression Omnibus database at http://www.ncbi.nlm.nih.gov/geo/. The accession number is GSE12098.Real-Time Reverse Transcriptase Polymerase Chain Reaction, MMP1 ELISA, and Bioactivity AssayReal-time reverse transcriptase polymerase chain reaction (RT-PCR) was performed as described previously []. Quantitation of pro-MMP1 protein expression and activity were performed using Quantikine human pro-MMP1 ELISA kit (R&D Systems Inc., Minneapolis, http://www.rndsystems.com) and MMP1 Biotrak activity assay system (GE Healthcare UK Limited, Little Chalfont, Buckinghamshire, U.K., http://www.gehealthcare.com) respectively, according to manufacturer's suggestions. See supporting information Methods for details.Statistical AnalysisStatistical analyses were performed using Prism . (Graphpad Software Inc., San Diego, CA, http://www.graphpad.com). Nonpaired parametric data were compared with Student's t-test; for in vivo quantitation of migrated MSCs, paired t-test was used. p values <. were considered statistically significant.RESULTSIsolation and Tumor-Trophic Characterization of MSCs In Vitro and In VivoHuman bone marrow-derived MSCs were isolated as previously described []. Both MSC- and MSC- were positive for CD13, CD44, CD90, and CD105, but were negative for the hematopoietic lineage markers CD34 (hematopoietic stem/progenitor cells), CD45 (leukocyte), and CD49a (monocytes and activated T-cells) (supporting information Fig. S1A). In addition, the abilities of MSCs to differentiate into adipocytes and osteocytes were demonstrated by the presence of Oil Red-O staining and von Kossa staining, respectively (supporting information Fig. S1B). Primary human glioma biopsy, in comparison with glioma cells lines, contains a variety of cells and tissue structures such as ECM, stroma, macrophages, and tumor cells, which closely resemble the in situ tumor scenario. Using a modified Boyden chamber assay, MSC- exhibited significant migration in the presence of either glioma-CM or primary glioma lysates (Fig. 1A). By contrast, the migratory activities mediated by MSC- were greatly impeded compared with MSC-, suggesting that the differential migratory potential is an intrinsic property of MSCs. These results indicate that MSCs are chemoattracted toward soluble factors secreted by primary human brain tumor and glioma cell line. Furthermore, different MSC isolates exhibited differential degrees of tumor tropism in vitro.Figure 1Differential migration properties of MSCs. (A): Migration of MSCs toward conditioned medium from glioma cells and primary glioma lysate was analyzed using a modified Boyden chamber assay. Migration of MSCs was determined by counting the number of propidium iodide-stained nuclei on the underside of the membrane under × magnifications. Bar graph represents number of migrated cells. (B): Coronal section of the mouse brain indicating the injection site (*) of MSC, and the location of the preimplanted ΔGli36 human glioma cells (T). Confocal fluorescent images showed the migration of the MSC- (upper panel), MSC- (middle panel), and normal human astrocytes (lower panel) to the tumor sites. Images were taken days post-CM-DiI-labeled-MSC injection with a confocal system (LSM Meta; Carl Zeiss, Göttingen, Germany) using a ×/. N.A. Plan-Fluor objective. Right panel: flow cytometry analysis of the percentage of CM-DiI cells in the left and right hemisphere. Purple-filled curve indicates the percentage of cells in the left hemisphere; green line indicates the percentage of cells in the right hemisphere. Total number of cells used for fluorescence-activated cell sorting analysis was ,,. Data shown are averages ± SEM, n = . Abbreviations: DMEM, Dulbecco's modified Eagle's medium; FITC, fluorescein isothiocyanate; MSCs, mesenchymal stem cells.To establish if similar behavior was observed in the in vivo setting, both MSC- and MSC- were prelabeled with a red fluorescent vital dye, CM-DiI, followed by implantation into the contralateral hemisphere of glioma-bearing mice (Fig. 1B). Two weeks after injection, with MSCs, these cells had migrated away from its original injection site and were found in the contralateral hemisphere, specifically around the peripheral and within the ΔGli36 tumor region. CM-DiI-labeled-MSC- could not be detected in the original injection site (i.e., the hemisphere opposite to the implanted ΔGli36 cells), thus confirming the tumor trophic property of MSC-. Furthermore, the percentage of CM-DiI-positive MSC- in the glioma-bearing hemisphere was .% ± .% higher than the nonglioma-bearing hemisphere (Fig. 1B). On the contrary, similarly injected CM-DiI-labeled-MSC- and CM-DiI-labeled NHA were not detected at the glioma-bearing hemisphere as shown by both the immunofluorescence image and flow cytometry analysis. Both MSC- and NHA-injected hemisphere exhibited more CM-DiI positive cells in comparison with the glioma-bearing region. In fact, the cells remained near or at the injection site. These results suggested that different MSC isolates exhibited differential migration ability toward glioma cells.Differential Migratory Abilities of MSCs Is Mediated by MMP1To examine if the observed differential migration ability was donor-specific, MSCs harvested from additional eight donors were tested for their migration activities toward glioma. These cells were morphologically and immunophenotypically similar to both MSC- and MSC-. Migration assays were performed four times using cells restricted to either passage or in an attempt to minimize variation due to prolonged in vitro culture. As shown in Figure 2A, differential migration abilities were observed in the various MSC isolates. Based on the percentage of migrated cells, we classified the MSCs into three categories, namely, highly migrating (MSC- > .%), medium migrating (.% ≤ MSC-, , , , ≤ .%), and poorly migrating (MSC-, , , < .%).Figure 2MMP1 is differentially expressed in various MSC isolates. (A): Differential migration of MSCs and real-time RT-PCR analysis of MMP1 transcript in MSCs. Data shown are averages of four replicates of independent experiments. Data shown for real-time RT-PCR are averages of duplicate samples, experiments were performed independently thrice. (B): Expression of MMP1 in μg of conditioned medium harvested from various MSCs. Data shown are averages of duplicate wells from representative experiments performed independently thrice. (C): MMP1 activity was quantitated in conditioned medium harvested from various MSCs. Data shown are averages of duplicate wells ± SEM from representative experiments performed independently twice. (D): Real-time RT-PCR analyses of MMP1, MMP2, MMP9, and MT1-MMP transcripts were performed in highly migratory MSC- and MSC- versus poorly migrating MSC- and MSC-. For comparison purposes, the fold change of each member of the MMP was expressed to the human embryonic kidney cells, . Data shown are averages of duplicate samples and performed independently twice. Abbreviations: MMP, matrix metalloproteinase; MSCs, mesenchymal stem cells; MT1, membrane type .To identify the molecular pathways involved in the differential migration activity of MSCs, gene expression profile between the highly migratory and poorly migratory MSCs were determined by cDNA microarray using Affymetrix GeneChip Human U133 Plus . Array. MSCs from the highly migrating and medium migrating group (MSC- and ) were compared with those from the poorly migrating group (MSC- and ). Using Partek software analysis, genes were found to be upregulated by at least twofold in the highly migrating/medium migrating MSCs. Many of these genes that matched to those in the Affymetrix NetAffix Gene Ontology analysis system belong to chemokines, metalloproteinases, and cell adhesion molecules, as represented in supporting information Table S1. Among these candidate genes, MMP1 was chosen for further study because it exhibited the greatest fold change (-fold) with a significant p value of .. To determine if the migration activity of MSCs could be correlated to the expression of MMP1, quantitative real-time RT-PCR analysis was performed (Fig. 2A). The mean expression values of MMP1 when normalized to 18S were found to be high (between values of and ) in highly migrating or medium migrating groups. On the other hand, MMP1 transcripts were minimally detectable (between values of . and .) in poorly migrating MSCs. To further confirm the differential expression of MMP1 in various MSC isolates, the level of MMP1 protein expression was quantified using an ELISA assay. As shown in Figure 2B and 2C, MMP1 expression (p = .) and activity was significantly higher (p = .) in the highly migrating MSCs as compared with the poorly migrating MSCs. Taken together, these results show that the differential migratory abilities of MSCs are correlated, in part, to the functional expression of MMP1.MMP2, MMP9, and MT1-MMP have been shown to be essential for the invasive capacity of human MSCs []. As such, the mRNA expression of these genes was analyzed in representative cells from each group using real-time RT-PCR analysis. The levels of MMP2, MMP9, and MT1-MMP transcripts were found to be similar between the highly migrating (as represented by MSC- and MSC-) and the poorly migrating cells (as represented by MSC- and MSC-) (Fig. 2D). This finding was consistent with our microarray data in that MMP2, MMP9, and MT1-MMP1 were not differentially expressed between the two groups of cells. Taken together, these results indicate that MMP2, MMP9, and MT1-MMP1 are not critical determinants of MSC migration.Targeted Knockdown of MMP1 RNA Inhibits Migration and Activity in MSCsTo further confirm the critical role of MMP1 in MSC migration, RNA interference assays were performed. Three MMP1-RNAi, namely, MMP1-RNAi-, MMP1-RNAi-, and MMP1-RNAi-, that target different regions on the MMP1 transcript were synthesized. As shown in supporting information Figure S2A, these RNAi showed at least % reduction in the number of migrated cells when compared with control-RNAi (ctrl-RNAi)-transfected cells. To ensure an effective MMP1 RNA inhibition, these RNAi were pooled together and chosen for subsequent studies (supporting information Fig. S2A). To confirm the target specificity of the MMP1-RNAi, the endogenous transcriptional levels of MMP1, MMP2, MMP9, and MT1-MMP in MSC- were determined by real-time PCR. As shown in supporting information Figure S2B, MMP1-RNAi-MSC- had at least a % knockdown of MMP1 mRNA expression, which was not observed in ctrl-RNAi-treated-MSC- cells. In addition, in MSC- cells treated with either MMP1-RNAi or ctrl-RNAi, there was no significant decrease in the transcriptional levels of MMP2, MMP9, and MT1-MMP, thus confirming the specificity of RNAi against MMP1. The targeted knockdown of MMP1 gene resulted in corresponding decrease in the level of MMP1 protein and activity (Fig. 3A). As a consequence, the number of migrated cells from MMP1-RNAi-MSC- was significantly reduced when compared with those that were transfected with ctrl-RNAi or naïve cells. Similar results were obtained in other highly/medium migrating MSCs, namely, MSC-, , , and , which have been transfected with MMP1-RNAi (supporting information Fig. S2C).Figure 3Targeted knockdown of MMP1 RNA inhibits migration and activity in MSCs. (A): Migration of MMP1-RNAi-transfected MSC was analyzed using a modified Boyden chamber assay. Bar graph represents number of migrated cells. Fluorescent images below showed representative images of PI-stained cells. Phase contrast photomicrograph showed the migration of MMP1-RNAi-transfected MSCs performed using the Varani migration assay, as representatively shown as original magnification ×. All images were visualized using wide-field microscopy with an inverted microscope (TE300; Nikon), and images were acquired on a CCD color digital camera (DXM1200F; Nikon) using image acquisition software, ACT- v2. (Nikon). (B): Effect of exogenous MMP1 on the migration profile of poorly migrating MSC-. Bar graph represents number of migrated cells. In all of the above experiments, data shown are averages of triplicates ± SEM from representative experiment. Abbreviations: CM, conditioned medium; Ctrl, control; DMEM, Dulbecco's modified Eagle's medium; MMP1, matrix metalloproteinase one; MSCs, mesenchymal stem cells; RNAi, RNA interference.The role of MMP1 in the migration of MSC- toward human glioma was further investigated using Varani migration assay. MMP1-RNAi-MSC- cells showed an inhibited outward dissemination in the presence of glioma-CM, which was not observed in ctrl-RNAi-MSC- or fresh tissue culture medium (Fig. 3A). These results provide supporting evidence that the migratory activities of MSCs are mediated, at least in part, by MMP1.Impaired Migration of MSCs Can Be Rescued by Exogenous MMP1 ExpressionAs MMP1 is a secreted protein, we next ask whether the conditioned medium derived from MSC- (MSC--CM) could provide the necessary components to induce migration of the poorly migrating MSCs (MSC- and MSC-). Our results showed that MSC--CM could effectively rescue the poorly migratory phenotype of MSC- and MSC- (Fig. 3B and supporting information Fig. S3A, respectively), with an enhanced in migratory activities by at least fourfold. Similarly, addition of exogenous recombinant human MMP1 proteins to MSC- and MSC- could significantly induce the migration of these cells. By contrast, conditioned medium from MMP1-RNAi-MSC- failed to induce migration of MSC- and MSC-.In view of the above observations, a gain-of-function assay was performed to evaluate if overexpression of MMP1 would potentiate the migration of MSC- and MSC- toward glioma-CM. We have previously demonstrated that MSCs could be transduced efficiently by a HSV- amplicon viral vector while retaining its stem cell characteristics []. To further confirm the functional involvement of MMP1 in MSC migration, the gene encoding MMP1 was subcloned into a HSV- amplicon vector (denoted as pHGCX-MMP1) and subsequently packaged into recombinant virions (Fig. 4A). Poorly migrating MSC, MSC-, and MSC- (shown in supporting information Figure) were used for the gain-of-function assay. Poorly migrating MSC- was efficiently infected by pHGCX-MMP1. As shown in Figure 4B, using multiplicity of infection (MOI) of ., approximately % of enhanced green fluorescent protein (eGFP)+ cells could be detected after hours of infection by FACS analysis. ELISA performed on the conditioned medium harvested at hours postinfection confirmed the ectopic expression of MMP1 in pHGCX-MMP1-infected cells in comparison with controls (Fig. 4B). Transwell migration assay subsequently showed a significant increase in the percentage of pHGCX-MMP1-infected-MSC- and - that had migrated across the filter membrane, p < . (Fig. 4C and supporting information Fig. S3B). By contrast, MSC transduced with vector alone did not exhibit significant migration compare with naïve MSC-. Taken together, these results confirm the important role of MMP- in mediating the MSC migration.Figure 4Overexpression of MMP1 restores the migration of MSC-. (A): MMP1 gene was inserted into the multiple cloning site located downstream of the strong CMV promoter on the pHGCX-MMP1 HSV- amplicon vector, which contains the eGFP gene. (B): Top: Percentage of infectivity was determined using FACS for eGFP expression hours postinfection. Flow cytometry FL- height analysis demonstrating the shift in peak of eGFP+ cells at MOI of .. Red line represents mock transduced-MSCs; green-filled peak represents pHGCX-MMP1-transduced MSCs. Bottom: MMP1 expression in pHGCX-MMP1-transduced MSC was determined after hours of infection at MOI of .. Data shown are averages of triplicates ± SEM from representative experiment. (C): Migration of pHGCX-MMP1-transduced-MSC- was analyzed hours postinfection using a modified Boyden chamber assay. Bar graph represents number of migrated cells. Data shown are averages of triplicates ± SEM, experiment was performed independently twice. (D): In vivo migration of pHGCX-MMP1-transduced-MSC- in glioma-bearing mice. Confocal fluorescent images showed the migration of the CM-DiI-labeled naïve MSC-, CM-DiI-labeled naïve MSC-, pHGCX-transduced-MSC-, and pHGCX-MMP1-transduced-MSC- to the tumor sites. Both CM-DiI-labeled naïve MSC- and CM-DiI-labeled naïve MSC- were pseudocolored green. Images were taken days post-CM-DiI-labeled-MSC injection with a confocal system (LSM Meta; Carl Zeiss, Göttingen, Germany) using a ×/. Numerical Aperture (N.A.) Plan-Fluor objective. Sections were shown at original magnification ×. *, Injection site, T, tumor. Abbreviations: amp, ampicillin; bGHpA, bovine growth hormone poly A; CMV, cytomegalovirus; eGFP, enhanced green fluorescent protein; MFI, mean fluorescence intensity; MMP1, matrix metalloproteinase one; MSCs, mesenchymal stem cells; pac, packaging signal.The effect of MMP1 overexpression on MSC tumor-tracking ability was further confirmed in vivo. The highly migrating MSC- was used as positive control and the poorly migrating naïve MSC- was used as a negative control. Fourteen days after injecting MSCs into mice bearing intracranial tumors, the brains were harvested and cryosectioned at μm thickness. Our results showed that the reconstitution of MMP1 in MSC- could rescue the nonmigratory phenotype of MSC- as indicated by the presence of eGFP positive MSCs in the tumor region (denoted as T in Fig. 4D). On the contrary, pHGCX-transduced-MSC- remained close to the injection site (indicated as “*” in Fig. 4D), exhibiting similar characteristics as the naïve MSC-. These results indicated that MMP1 could sensitize poorly migrating MSCs to signaling cue from glioma cells. Taken together, we have clearly demonstrated that MSCs with minimal migratory activities can be reverted by supplementing the cells with recombinant or exogenous MMP1 gene products, which are essential for the tumor-trophic migratory activities of MSCs.Interference of PAR1/MMP1 Axis in Highly Migratory MSCs Completely Abolish the Migratory ActivitiesRecently, the G protein-coupled receptor, PAR1, has been found to be cleaved by MMP1, which promotes breast cancer migration and invasion []. To investigate whether PAR1 plays a role in MMP1-dependent chemotaxis of MSCs, we examined whether the inhibition of PAR1 proteolysis may affect the ability of MSC- to migrate toward glioma-CM. The inhibition of PAR1 activation was performed by incubating MSC- in the presence of an anti-PAR1 monoclonal antibody (ATAP2), which specifically binds to the cleavage domain of PAR1, thus preventing the proteolysis of PAR1 by MMP1. Anti-PAR1 treated MSC- resulted in an approximately % decrease in the number of MSC- migrating to the glioma-CM in comparison with naïve MSC- (Fig. 5A). Conversely, no significant difference was observed in control IgG1-treated cells. This observation was further confirmed by the Varani migration assay in which anti-PAR1-treated MSC- failed to migrate toward glioma-CM (Fig. 5B). Similar experiments were also performed in poorly migrating MSC-. The addition of anti-PAR1 antibodies, but not anti-IgG1, was shown to further reduce the mobility of MSC- toward glioma-CM. The addition of exogenous MMP1 to this anti-PAR1 treated MSC- also did not enhance the migratory activities (Fig. 5C). This suppressive effect on the migration of MSC- was lifted when exogenous MMP1 was added to the cells without the presence of PAR1 antibodies, thus demonstrating that MMP1-mediated MSC migration is directed solely through interaction with its cognate receptor, PAR1.Figure 5Functional MMP1/PAR1 axis mediates MSC migration. (A): Effect of anti-PAR1 blocking antibody on the migration of MSC- was examined using a modified Boyden chamber assay. Bar graph represents number of migrated cells. Data shown are averages of triplicates ± SEM, experiment was repeated independently three times. Photomicrograph showed the representative images of PI+ migrated cells at original magnification ×. (B): Varani migration assay was performed to confirm the effect of PAR1 on MSC migration. Photomicrograph showed migration of MMP1-RNAi-transfected MSCs. Representative images from two independent experiments were shown. Images were shown as original magnification x100. Slides were visualized using wide-field microscopy with an inverted microscope (TE300; Nikon), and images were acquired on a CCD color digital camera (DXM1200F; Nikon) using image acquisition software, ACT- v2. (Nikon). (C): Effect of MMP1/PAR1 interaction on the migration ability of MSC- treated with the various proteins was examined using a modified Boyden chamber assay. Bar graph represents number of migrated cells. Data shown are averages of triplicates ± SEM, experiment was performed independently twice. Abbreviations: CM, conditioned medium; MSCs, mesenchymal stem cells; PAR1, protease-activated receptor one.DISCUSSIONIt is becoming increasing clear that achieving targeted trafficking of stem cells will be critical for effective tissue regeneration in the clinic. In this study, we have demonstrated that MSCs exhibit differential tumor tropisms despite the fact that they were isolated using an identical procedure and cannot be distinguished based on their phenotypic or multipotential characteristics. The isolation of MSCs with different tumor-trophic properties has allowed us to identify the molecular characteristics of migrating cells in tumor microenvironment. In particular, the differential migratory activities of MSCs are mediated, at least partially, by endogenous MMP1 expression.Cell migration assays were performed to determine the gliomatrophic properties of our various MSC isolates. Interestingly, we observed that different MSC isolates exhibited differential migratory activities in vitro, which could be grouped according to their migratory activities into high, medium, and poor. This is unlikely due to higher cell culture confluence as reported by De Becker et al. [], as both highly migrating and poorly migrating MSCs were seeded under a mean cellular density that was considered as high according to the reported experimental conditions. Furthermore, the differential tumor tropisms exhibited by MSCs was clearly demonstrated in intracranial human glioma xenografts (Fig. 1B). At present, we are uncertain if the differential migration activities exhibited by the various isolates of MSCs reflected the bona fide pathophysiological state in which the MSCs are isolated or merely that the pool of MSCs contain subpopulations at different states of differentiation that are indistinguishable by cell surface markers. Nevertheless, the identification of these cells has provided us with an approach to identify distinct signals that are critical to the migration of MSCs by comparing the two MSC populations in comparative gene expression studies. By contrast, many of the previous reports have focused on identifying genes involved with MSC migratory activities by exposing MSCs to different stimuli or tumor CM before gene profile expression assays [,].Using DNA microarray analysis, our results showed that MMP1 is significantly upregulated (∼-fold) in highly migrating MSCs in comparison with the poorly migrating MSCs. This was confirmed by the ∼-fold to ,-fold higher level of MMP1 transcripts in the highly migratory MSCs compare with poorly migratory MSCs using real-time PCR. The differential mRNA expression of MMP1 was further supported by ELISA assay: MSCs with greater migratory activities were shown to secrete substantial amounts of MMP1 proteins into the culture supernatants (Fig. 2B). In addition, the level of active MMP1 was found to be tightly correlated with the migratory activities of various MSC isolates (Fig. 2C). Ideally, the relative amount of MMP1 RNA transcripts, MMP1 protein expression and functional cell migratory activities should be consistent among the various MSC isolates. For example, the top three MSC isolates with the highest migratory activities were MSC- ( ± .), MSC- (. ± .), and MSC- (. ± .). The MMP1 RNA levels were also found to be in the same order (Fig. 2A). However, the level of MMP1 protein expression was found to be highest from MSC-, followed by that found in MSC- and MSC- (Fig. 2B). These interexperimental variations could be explained as each of the assays (RNA, protein, or cell migration) was performed at different times even though special care was taken to ensure that all experiments were performed using cells of passage to . Nevertheless, the overall MMP1 mRNA, protein, and migration profiles were consistent among the highly migrating versus the poorly migrating MSCs.MMPs are generally known for their roles in tissue remodeling by degradation of ECM and basement membrane components. However, MMPs have also been implicated in the activation of growth factors and cytokines by degrading their precursors or inhibitors, thereby adjusting the cancer cells to the tissue microenvironment [,]. MMP1 has been shown to degrade insulin-like growth factor binding proteins (IGFBP)- and - [,]. This in turn modulates the bioavailability of the insulin-like growth factors (IGFs) depending on tissue types and physiologic/pathologic status []. For example, MMP9 induced IGFBP2-IGF2 complex proteolysis resulted in the extracellular release of free IGF2 with positive and biologic effect on astrocytoma cellular growth and migration []. From our microarray results, the level of IGF2 transcripts was the second highest in MSCs with greater mobility in comparison with those that migrate poorly (supporting information Table S1). Thus, it is possible that MMP1 could also act as an IGFBP2 proteinase; the observed elevated level of IGF2 could represent the released, unbound IGF2 that ultimately initiates the mobilization of MSCs. Alternatively, the migration process may also be guided by cytokines, such as IL-, and chemokines, such as CXCL1 and CXCL2, all of which have been found to be elevated in the highly migratory MSC population. The precise signal transduction pathway that governs the migration of MSCs will require further study.Interestingly, even though platelet derived growth factor β (also known as PDGF-BB), MMP2, MMP9, and MT1-MMP were reported to be players mediating the migration and invasion of MSCs in vitro [,], upregulation of these genes in highly migratory MSCs was not observed. Similarly, there was no significant difference in the level of MMP2, MMP9, and MT1-MMP transcripts expression between the highly migratory (MSC- and MSC-) versus poorly migratory MSCs (MSC- and MSC-) when analyzed using real-time RT-PCR (Fig. ). However, the design of this experimental study does not take into account the possible synergistic effect of these MMPs in the regulation of MSC mobilization. Hence, we could not exclude that these MMPs may still play a role in the overall mobilization of MSCs.Recently, Boire et al. reported that PAR1 is a MMP1 receptor that promote invasion and tumorigenesis of breast cancer cells in vitro and in vivo []. Shi et al. showed that blocking PAR1 cleavage and activation using anti-PAR1 antibody could inhibit invasion and chemotaxis of prostate cancer cells []. Using a monoclonal antibody against PAR1, we showed that blocking the MMP1/PAR1 interaction significantly reduced the migration ability of MSCs. The importance of this axis in MSC migration was further supported by the lack of migratory response when recombinant MMP1 was added to cells previously incubated in anti-PAR1 antibody. Furthermore, Western blot analysis showed that the PAR1 expression level was similar in highly migrating versus poorly migrating MSCs (supporting information Fig. S4). Thus, it appears that the level of MMP1 expression and the specific interaction of MMP1 with PAR1 proteins determine the differential migration ability of MSCs. In bone marrow-derived MSCs, the level of MMP1 was reported to be regulated by the hypoxia-inducible factor-1α []. Ectopic expression of the secretory form of fibroblast growth factor- could also induce MMP1 transcription in endothelial cells, thus resulting in enhanced migratory activities []. On the other hand, PAR1 activation in human late endothelial progenitor cells has been shown to enhance the mRNA levels of both SDF- and CXCR4 []. We have not studied the SDF-/CXCR4 axis closely because these genes were not differentially expressed between the highly migrating and poorly migrating MSCs. This finding might explain the observation why only a small subpopulation of human MSC has been shown to express functional CXCR4 receptors []. However, we could not exclude these genes from providing secondary support that may assist in the mobilization of MSCs. At present, the potential role of MMP1 in regulating SDF- expression is unknown; it is also unclear whether MMP1/PAR1 axis acts directly or indirectly on SDF- and/or CXCR4 activation. As such, further research is necessary to dissect the relation between the two pathways. Nevertheless, it is becoming clear that both the ECM and the microenvironment surrounding MSCs are not merely scaffold, but also harbor cryptic biological functions that may regulate the migration, plasticity, self-renewal and pluripotency of MSCs.In conclusion, we report that the migratory activity of MSCs toward glioma is mediated via the MMP1/PAR1 axis in vitro and in vivo. An adequate understanding of the tumor tropism of MSC bears important implication for effective cellular delivery of therapeutic agents for brain tumor therapy.CONCLUSIONOur results highlighted the critical role that MMP1 plays in the process of mobilizing MSCs toward human glioma cells. MSCs expressing low levels of MMP1 fail to response to signaling cues from human glioma cells even though other MMPs are present at levels similar to the highly migrating MSCs. Targeted knockdown of MMP1 or the inactivation of PAR1 that disrupt its association with MMP1 resulted in the lost of MSC migratory activities. Anti-PAR1 antibodies treated MSCs could not migrate even in the presence of exogenous MMP1, suggesting that both MMP1 levels and the specific interaction between MMP1 and PAR1 are important factors that determine the migratory abilities of MSCs. To our knowledge, this is the first comprehensive report on the involvement of MMP1 on the migration of MSCs via the activation of the PAR1 receptor.DISCLOSURE OF POTENTIAL CONFLICTS OF INTERESTThe authors indicate no potential conflicts of interest.
|
PMC2667471.txt
|
TITLE: Segmentation of overweight Americans and opportunities for social marketing AUTHORS: Jane Kolodinsky, Travis Reynolds ABSTRACT: BackgroundThe food industry uses market segmentation to target products toward specific groups of consumers with similar attitudinal, demographic, or lifestyle characteristics. Our aims were to identify distinguishable segments within the US overweight population to be targeted with messages and media aimed at moving Americans toward more healthy weights.MethodsCluster analysis was used to identify segments of consumers based on both food and lifestyle behaviors related to unhealthy weights. Drawing from Social Learning Theory, the Health Belief Model, and existing market segmentation literature, the study identified five distinct, recognizable market segments based on knowledge and behavioral and environmental factors. Implications for social marketing campaigns designed to move Americans toward more healthy weights were explored.ResultsThe five clusters identified were: Highest Risk (%); At Risk (%); Right Behavior/Wrong Results (%); Getting Best Results (%); and Doing OK (%). Ninety-nine percent of those in the Highest Risk cluster were overweight; members watched the most television and exercised the least. Fifty-five percent of those in the At Risk cluster were overweight; members logged the most computer time and almost half rarely or never read food labels. Sixty-six percent of those in the Right Behavior/Wrong Results cluster were overweight; however, % of them were familiar with the food pyramid. Members reported eating a low percentage of fast food meals (%) compared to other groups but a higher percentage of other restaurant meals (%). Less than six percent of those in the Getting Best Results cluster were overweight; every member read food labels and % of members' meals were "made from scratch." Eighteen percent of those in the Doing OK cluster were overweight; members watched the least television and reported eating % of their meals "made from scratch."ConclusionThis study demonstrated that five distinct market segments can be identified for social marketing efforts aimed at addressing the obesity epidemic. Through the identification of these five segments, social marketing campaigns can utilize selected channels and messages that communicate the most relevant and important information. The results of this study offer insight into how segmentation strategies and social marketing messages may improve public health. BODY: BackgroundIt is no longer news that unhealthy eating behaviors and sedentary lifestyles have contributed to the current obesity epidemic in the United States. However, the percent of Americans who are overweight do not form a homogeneous group – attitudes, demographic characteristics and lifestyle choices vary greatly within this subset of the US population. Segmentation theory tells us that a "one size fits all" approach to marketing social change may not meet the needs of all people. Further, marketing research has revealed the importance and effectiveness of tailoring messages and incentives to meet the needs of different population segments. "Social marketing" is defined as "a social change campaign organized by a group which intends to persuade others to accept, modify or abandon certain ideas, attitudes, practices or behavior" []. A social marketing campaign using market segmentation may be one effective tool for helping move more Americans toward healthier weights [].The food industry has used market segmentation of consumers for decades. As early as , Haire segmented consumers based on personality characteristics in order to increase the sales of instant coffee []. Today, more than half a century later, segmentation is still being used to market twenty-first century foods to consumers [,]. Even the dairy industry has engaged in segmentation in an effort to increase sales of dairy products based on research that links the consumption of dairy foods to weight loss []. Segmentation has enabled the industry to target its products toward specific groups of consumers with similar attitudinal, demographic, or lifestyle characteristics.The success of segmentation strategies for food marketing suggests that such techniques may hold promise for identifying ways to change consumer behavior regarding unhealthy food and lifestyles []. Psycho-behavioral segmentation – or segmenting on the basis of what people are doing (i.e., the behavior), and why (i.e, the social and psychological antecedents to the behavior) – has already been employed for health promotion research focusing on alcoholism [] and overall health [,]. In some instances segmentation has even been explicitly tied to social marketing efforts: " a day for better health", for example, is a social marketing campaign that encourages more positive nutrition behaviors among American consumers []. The " a day" campaign helped increase the percentage of Americans consuming five or more servings of fruits and vegetables per day from percent in to percent in []. To achieve this, the campaign recognized and made use of the existence of market segments, both demographic and psychosocial []. Recent reports by the US Department of Health and Human Services and the National Institutes of Health further highlighted the need to identify specific population segments for targeted interventions in the fight against obesity, including efforts to assess how obesity-related knowledge, behavior, and environments may affect consumer behavior [,].Segmentation is used by marketers because it works. Not every individual is a potential consumer of a given product, idea or service, so tailoring messages to specific groups can be more effective than broadcasting to everyone. Consumers are segmented based on geographic location, demographic characteristics, and product use. Contemporary marketers now also employ lifestyle-based and product benefit approaches [,-,].While the techniques of market segmentation have long been used in the field of for-profit marketing, they have only recently been used as tools to help meet the goals of social marketing campaigns. One impediment to widespread adoption of market segmentation strategies is that social marketing segmentation studies related to healthy food behaviors have not shown consistent results across demographic, behavioral and lifestyle variables [,]. A wealth of empirical research has linked demographic characteristics, dietary behaviors, media habits, and psychological variables to overweight status – for a thorough review of this literature, see Jeffery & Utter [], Ball et al. [], or Trudeau et al. []. However, though some studies have linked factors such as socio-economic status, gender, and dietary patterns to overweight [-], others have not [,]. Ultimately, there remains a certain degree of uncertainty in the scientific community about how the energy imbalance leading to increased body weight among Americans is occurring []. This suggests that while previous research can provide some guidance as to the types of variables that should be included in a segmentation study of overweight in the US, health advocates cannot simply interpolate past results to shape social marketing campaigns for changing health behaviors.Social Learning Theory [], the Health Belief Model [], and their offshoots have been proposed as theoretical frameworks suited to the application of market segmentation in studies of consumer health behavior change [,]. Such models typically include both personal and environmental variables []. By way of example, Miles et al. [] examined a mass-media health campaign in the United Kingdom based on both Social Learning Theory and the Health Belief Model. Through the analysis of the results of this campaign a clear indication of market segments emerged including demographic segments characterized by socio-economic status, age, gender, and overweight []. Following the campaign, men reported larger lifestyle changes than women. Those in lower income categories were less likely to be aware of the campaign. Such information can be invaluable for reformulating future campaigns to ensure that they are more effective.In the US, Loughrey et al. used segmentation techniques in the form of audience-profiling to promote the Dietary Guidelines for Americans []. Using national market data, the researchers delineated three market segments based on the Healthy Eating Index: Better Eaters, Fair Eaters and Poor Eaters. In addition to demographic characteristics, beliefs, values, and both food and media habits were used in the segmentation process. Demographic variables explained little regarding differences across segments, and there were no differences between the media habits of the groups. However, dietary choices and attitudes did prove significant. Better Eaters were more likely to take action to eat a healthy diet and were better able to anticipate outcomes of their behaviors. Poor Eaters were less likely to worry about the nutritional content of foods. Fair Eaters fell in the middle of these two extremes. Based on these segments, three distinct message-development strategies were undertaken. It was determined that Better Eater messages might best focus on simple, positive messages to help maintain healthy eating behaviors. Fair Eaters were in need of messages that would precipitate action to change eating behaviors. A highly targeted approach was recommended for Poor Eaters; one which both captured attention and established 'cultural relevance' [].Qualitative focus group research is another approach that has been employed to gain insights into how to communicate appropriate health-related messages to consumers, especially messages based on the Dietary Guidelines for Americans []. Focus groups including segments based on gender, overweight, and age have shown that messages to the public must be inclusive, trustworthy, and not "too markety" []. These results point to the need to further segment the population beyond demographic characteristics. Although messages must be inclusive, they must also be applicable to different sub-segments of the population, each of which may have differing lifestyles or may draw upon different sources of information.Today segmentation is emerging in more modern incarnations that move beyond social marketing. The Internet has enabled marketers to refine segmentation to the level of microsegments, made possible in part because web users' behavior can be tracked more readily and unobtrusively than that of traditional consumers [,]. The Internet also provides the medium to act on these narrow segments through unique offerings that appeal to these narrow segments []. Similarly, political campaigns have recently attributed some of their successes to data mining technology used to identify and segment like-minded individuals and craft uniquely appealing messages to targeted segments about their candidate, dubbed "microtargeting" [,]. Some argue that political marketing [] and accompanying meaningful segmentation represents an avenue that must be traversed in any successful campaign []. Though the implications of these new and expanding marketing channels have yet to be explored in a social marketing context, "microsegmentation" and "microtargeting" may provide opportunities for more effective health behavior promotion in modern societies.An effective fight against obesity must coordinate public and private campaigns against unhealthy food choices. This requires developing and delivering clear, coherent health messages and developing targeted programs on specific segments of the population []. Ultimately, there is emerging evidence that the development of targeted messages based on segmentation of the population holds promise for a social marketing campaign seeking to promote healthier dietary and lifestyle choices [,]. Empirical research in this area remains in its infancy, but available research does provide insights into theoretical foundations and empirical measures appropriate for use in social marketing segmentation studies.This study uses cluster analysis to identify different segments of US consumers based on food choices, activity, food knowledge, overweight, and other environmental variables. The goal of the analysis is to identify distinguishable segments with the U.S. overweight population that can be reached with appropriate messages and media channels aimed at moving them toward more healthy weights.MethodsStudy populationData used in this study are from a national poll, funded by a United States Department of Agriculture Grant. The questionnaire was administered over a two-week period by trained staff using a computer-aided telephone interviewing system (CATI). Each interview took approximately minutes to complete, and up to five callbacks were made to households. A geographically stratified random sampling technique was used. Seven regions of the United States were identified: Northeast; Mid-Atlantic; South; Great Lakes; North Central; South Central; and Pacific. Land-line telephone numbers for each of the seven regions were purchased from infoUSA®. InfoUSA® provides consumer contact lists. They catalog telephone directories to compile extensive, comprehensive databases of consumer information that are updated monthly and sold to market researchers. A random number generator was used within each region to generate the final sample size of . Only adults over the age of were eligible to complete the survey; approval for the study was obtained from the Institutional Review Board at the University. Power calculations reveal that the sample size resulted in a % probability of detecting a % difference in the dependent variable with % power. Demographic characteristics of the sample are provided in Table .Table 1Description of the SampleBehavioral VariablesVariable DescriptionSummary Statistic(n = )Knowledge Know PyramidKnow food pyramid = .83a Read LabelsMost of the time, sometimes, rarely, neverMost of the timebType of Labels Read CaloriesRead calories label = .48a CarbohydratesRead carbohydrates label = .23a FatRead fat label = .48a ProteinRead protein label = .08a SaltRead salt label = .18a Serving sizeRead serving size = .05a SugarRead sugar content = .22aFood Behavior OverweightOverweight = .53a Calories rightMore, less, about right amountLess than neededb Eat wellAlways, sometimes, never choose a healthy dietSometimes choose a healthy dietb Eat weightEating to lose weight = .55a Fast food mealsPercent fast food meals12. (.)c Restaurant mealsPercent restaurant meals12. (.)c Prepared mealsPercent prepared food meals6. (.)c Scratch mealsPercent meals from scratch67. (.)cRisk Factors SmokerSmoker = .17a IllnessIllness limits activity = .24a InsuranceCovered by insurance = .89aActivity TVMinutes of TV123. (.)c ComputerMinutes of home computer52. (.)c ExerciseExercise minutes/day, times/wk = .66a Exercise weightUsing exercise to lose weight = .49aDemographic VariablesgenderGender ( = male).45aeducationLess than HS = .02aHS grad = .08aSome college = .53aBachelors or more = .37aincomeLess than $, = .13a$, to $, = .18a$, to $, = .17a$, to $, = .15a$, or greater = .38achildrenHas children = .42aEmployment employedEmployed = .55a unempUnemployed = .04a retiredRetired = .18aageAge in years45. (.)cRegion neastNortheast = .28a southSouth = .27a midamerMiddle America = .26a westWest = .21aUrban characteristic ruralRural = .32a suburbSuburban = .75a urbanUrban = .25an = 581a dummy variable /1b median reported;c continuous variable, standard deviation in ()MeasuresThe survey instrument was designed to collect information about both personal and environmental characteristics. While no segmentation study can completely measure all the respondent characteristics related to obesity or healthy lifestyle behaviors, our questionnaire attempted to collect data on a wide array of characteristics and behaviors drawn from Social Learning Theory and the Health Belief Model, or found to be significant in other studies [,,]. Socio-demographic characteristics included gender, education, income, children, employment status, age, geographic region and urban/rural residence.Body mass information was collected in such a way as to minimize under-reporting of overweight among survey respondents: respondents were first asked their height, and then the CATI system automatically branched to a single question on weight that corresponded to the CDC standard calculations of body mass index (BMI) by height and weight []. For example, if a respondent was a female with a height of feet inches ( inches), the CATI system branched to a question asking, "Do you weigh less than pounds?" If the respondent indicated "yes," they were classified as not overweight. If they answered "no," they were classified as overweight based on the CDC body mass tables [].The instrument also collected data on leisure time physical activity, including whether respondents obtained the recommended minutes of exercise five days per week (exercise). Sedentary behavior was measured through questions about computer use (computer) and television watching (TV). Health related variables included whether respondents had a chronic illness limiting their activity (illness limits activity), whether they were covered by a health insurance plan (insurance) and whether they smoked (smoker). Question wording was based on the CDC's Behavioral Risk Factor Surveillance System Questionnaire []. A proxy for food knowledge was created based on respondents' food information searching behavior. Respondents were asked how often they read food labels (read labels) and what type of information they looked for on a label. This information was obtained through a series of yes/no questions about respondents' label-reading behavior related to calories (calories), carbohydrates (carbohydrates), fat (fat) protein (protein), salt (salt) serving size (serving size), and sugar (sugar). Respondents were also asked whether they had heard of the food pyramid (know pyramid). MyPyramid was not yet introduced at the time of the survey [].Motivations about diet and exercise were collected based on the questions "Do you get at least minutes of exercise five days a week?" (exercise), "Are you exercising to lose weight?" (exercising weight) and "Are you eating to lose weight?" (eating weight). Respondents were also asked whether they were eating an appropriate number of calories per day (calories right) and whether they "most often choose a healthy diet," "sometimes choose a healthy diet," or "eat pretty much what they want" (eat well). Information on other food behaviors included the number of meals eaten at home (both prepared from scratch (scratch meals) and purchased prepared (prepared meals)), fast food meals (fast food meals), and number of other restaurant meals (other restaurant meals) in the last week. For each individual the total number of weekly meals consumed was calculated and percentage of meals from each source was derived to standardize across respondents. Descriptive statistics are provided in Table .Statistical analysesTwo Step Cluster Analysis using Schwart's Baysian Criteria in the Statistical Package for Social Sciences (SPSS ..) was used to identify clusters of respondents. This technique is an exploratory method used for both continuous and categorical data. Demographic variables were omitted from the cluster analysis, so that cluster membership was driven by respondent behaviors rather than demographic characteristics. Once the clusters were identified, bi-variate tests of association (ANOVA and Chi-square depending on level of measurement) were used to determine whether cluster membership was associated with demographic characteristics.ResultsCluster AnalysisFive clusters were identified based on overweight status, information search, activity level, health indicators, and food behaviors. Three of the five clusters were characterized as overweight, comprising almost three-quarters of the sample. The remaining two clusters (the remaining percent of the sample) were characterized as not overweight. Table describes the characteristics of each of the clusters. The clusters can be summarized as follows:Table 2Behavioral Characteristics of ClustersVARIABLE/CLUSTERCLUSTER :"Highest Risk"(n = )CLUSTER :"At Risk"(n = )CLUSTER :"Right Behavior/Wrong Results"(n = )CLUSTER :"Getting Best Results"(n = )CLUSTER :"Doing OK"(n = )KnowledgeKnow Pyramid (%).....3Read labels (%) Always28..... Most of the time29..... Sometimes38..... Rarely/Never3.....0Type of Label Calories (% that read)..... Carbohydrates (% that read)..... Fat (% that read)..... Protein (% that read)..... Salt (% that read)..... Serving Size (% that read)..... Sugar (% that read).....4Food BehaviorOverweight (%).....5Calories % More than needed9..... % About needed amount66..... % Less than needed14.....4Eat Healthy % Most Often Eat Healthy22..... % Sometimes Eat Healthy50..... % Eat What I Want26....5Eating weight (% eating to lose weight).....6Fast Food Meals (% of meals per week; [#]a). [. per week]. [ per week]. [ per week]. [ per week]. [ per week]Other Restaurant Meals (% of meal per week; [#]a). [ per week]. [. per week]. [ per week]. [ per week]. [ per week]Prepared Food Meals (% of meals per week; [#]a). [. per week]. [ per week]. [ per week]. [ per week]. [ per week]Scratch Meal (% of meals per week; [#]a). [ per week]. [ per week]. [ per week]. [ per week]. [ per week]Risk FactorsSmoker (%).....5Illness (% Illness limits activity).....0Insurance (% Covered by insurance).....2ActivitySedentary Behavior: Television (minutes/day) Computer Use (minutes/day)845476258Exercise (% Exercise regularly).....7Exercising weight (% Exercise to lose weight).....1a The number of meals per week from each category (in []) is an approximation based on the average total weekly meals consumed by cluster members. All have been rounded to the nearest . meals.Overall n = 581Highest RiskNineteen percent of the sample fell into this category. Ninety-nine percent of this cluster is overweight, % read food labels at least some of the time, and a majority (.%) look at fat content. Two thirds believe that they eat "about the right number of calories per day," and more than half "sometimes choose a healthy diet." Eighty percent get little or no exercise, and three quarters are not using exercise to lose weight. Over % reported that "chronic illness limits activity." This cluster watches more television than others (more than / hours per day), and eats almost % of meals in the form of fast food, other restaurant food, and prepared foods.At riskTwenty-two percent of the sample fell into this category. Fifty-five percent of the cluster is overweight. Almost half rarely or never read food labels and % are unfamiliar with the food pyramid. Sixty percent indicate they "eat what they want." Three fourths report getting at least minutes of exercise five days a week, but % are not using exercise to lose weight. Compared to other clusters, more members of this segment are smokers (%) and do not have health insurance (nearly %). This cluster reported eating % of their meals as fast food and eating % of their meals at restaurants or as prepared foods.Right Behavior/Wrong ResultsThirty-three percent of the sample fell into this category. Almost two-thirds of the cluster is overweight. Eighty percent report they "always read food labels," and most of the group looks for a wide variety of label information. Ninety-five percent of respondents in this group know what the food pyramid is. Eighty-five percent report they eat "less or about the right number of calories," and more than half report "choosing a healthy diet." The majority of this cluster is dieting and exercising to lose weight. This group eats a low percentage of fast food meals (%), but a higher percentage of other restaurant meals (%) compared to other clusters.Getting Best ResultsThirteen percent of the sample fell into this category. Less than six percent of cluster members are overweight. Everyone in this group reported reading labels, and % know what the food pyramid is. Almost % of this cluster reported eating fewer calories than they need and not one reported that they "eat what they want". Ninety percent eat to lose weight and % exercise five times per week. This cluster watches less television than the others, and members report eating % of their meals "made from scratch."Doing OKTwelve percent of the sample fell into this category. Within this cluster, % of respondents are overweight. Eighty percent read food labels always or most of the time, but most focus on fat and salt content over other information. Seventy-five percent reported choosing a healthy diet and two-thirds reported eating fewer calories than they need. Members of this cluster are not dieting, and % are not exercising to lose weight. Almost one-third stated that chronic illness limits their activity. Fifteen percent of this cluster smokes. Members of this segment use home computers about one hour per day, but watch the least television of all groups ( minutes per day). They eat % of their meals "made from scratch." This group also eats the least fast food (six percent of meals) compared to the other clusters.Bi-variate AnalysisBi-variate analysis was used to test whether any demographic characteristics are related to cluster membership. Table presents the results of ANOVA and Chi-square analyses. Gender, income quintiles, age, education and region of residence were found to be significantly related to cluster membership.Table 3Bivariate Analysis of Cluster Membership by Demographic CharacteristicsVARIABLE/CLUSTERCLUSTER : "Highest Risk"CLUSTER : "At Risk"CLUSTER : "Right Behavior/Wrong Results"CLUSTER : "Getting Best Results"CLUSTER : "Doing OK"P valueaGender (% female)...... **Education (%). ***Less than high school0.....0High school grad2.....3Some college54.....1Bachelors or more42.....6Income (%). ***Less than $,.....$, to $,.....$, to $,.....$, to $,.....$, or greater40.....3Children (% has children)......216Employment (%)Employed53...... ***Unemployed2......417Retired27...... *Age range (%). ***–.....–.....–.....–.....–..... or more26.....6Region (%). ***Northern27....519South19.....6Middle America16.....4West36.....1Urban characteristic (%).163Rural30.....6Suburban43.....6Urban26.....8a Chi2 statistic; Overall n = * P < .** P < .*** P < .001Almost three quarters of the Getting Best Results cluster are women, as are two-thirds of the Doing OK cluster. Over half of the Right Behavior/Wrong Results cluster are men. With regard to income, half of the Highest Risk cluster had household incomes of above $,. Of those, most were in the highest income group. In contrast, over % of the At Risk group had incomes under $, per year, and over % made less than $,. Of the Right Behavior/Wrong Results cluster, about % had incomes over $,. The Getting Best Results cluster is also a high income group, with almost percent of this cluster reporting incomes over $, per year. The income levels of the Doing OK cluster were more evenly distributed: almost one third reported incomes of between $, and , per year, while another % reported incomes of over $, per year. Age was also significantly related to some clusters' membership: almost half of the Highest Risk and Right Behavior/Wrong Results clusters are between the ages of and . Almost half of At Risk cluster members are between the ages of and . The Getting Best Results and Doing OK clusters have a more even age distribution. Finally, more than a third of the Highest Risk cluster lives in the West.DiscussionThis study demonstrates that five distinct, recognizable market segments can be identified for social marketing efforts aimed at addressing the obesity epidemic. Through the identification of these segments, social marketing campaigns can target individuals in order to more successfully communicate the most relevant and important information. The demographic, behavioral, and knowledge characteristics of the five segments identified here imply some strategies for communicating appropriate information to each of the groups.The findings suggest that Highest Risk consumers may benefit from public service messages that remind them of "mindful eating," food choice, and activity as a part of a healthy lifestyle, as it appears they do not pay much attention to the foods they eat, and they exercise the least. Understanding that physical body image, behavioral self-image and self-efficacy can create perceived barriers to exercise [,], exercise promotion and support for this group may be appropriate to help address the combined challenges of limited exercise experience and current overweight status. This group could also potentially benefit from basic information about calories and nutritional label interpretation: although the Highest Risk group reports reading all label types more than members of the At Risk cluster, it is possible that they are still misinterpreting (or not using) this information when making purchases and eating decisions. Finally, the data suggest that the Highest Risk cluster watches the most television of any cluster, suggesting these information channels may be the best way to communicate messages tailored to this specific group. Because we do not know more about specific television programming consumed by the Highest Risk group, prime-time television programming would be a sensible start to communicate food choice, exercise, and food label-use education messages.Meanwhile, it appears that the At Risk group would benefit most from messages that improve their overall dietary selections. Members of this segment eat almost half their diet away-from-home, and the majority reports they "eat what they want". Social marketing efforts must seek to successfully communicate the relationships between nutritional intake, health and impending risks to this group. At Risk group members must also be informed where to find nutrition information about the foods they are consuming in restaurants. This may be one important factor resulting in higher energy intakes within this segment. The data further show that the At Risk cluster logs more computer time than other clusters; tracing computer "cookies" might thus help in micro-targeting of the At Risk group. Combined, internet and television channels together would have the greatest reach for the almost % of the sample that fell into the At Risk and Highest Risk segments.The Right Behavior/Wrong Results group reported a basic understanding of food knowledge and appears to be reading food labels more than any other segment. However, over % of this segment's members are overweight. Clear, direct messages containing steps that individuals can take to use available nutrition information effectively may work for this group. Given the frequency with which members of this group dine out at restaurants, nutritional messages in restaurant venues might be one way to reach this group in particular. Another possibility, however, is that the Right Behavior/Wrong Results group maintains incorrect perceptions about dietary choices and exercise: members of this group may thus be reporting they are making efforts when in fact their definitions of eating healthy foods and engaging in physical activity do not agree with conventional norms. Communicating specific goals and benchmarks to members of this cluster – along the lines of the " a day for better health" social marketing campaign in the US [] – might help overcome such misperceptions.Finally, the Getting Best Results and Doing OK groups eat the majority of their meals "made from scratch." These groups are of a healthy weight and perhaps just need to have their healthy food behaviors reinforced. However, the Doing OK group also contains the second highest percentage of smokers and has a high percentage of respondents that said chronic illness limited their activity. Members of this group may benefit from broader health messages, as well as specific information on less-intense exercise techniques for individuals who wish to continue an exercise regime in spite of illness. With this segment it is particularly important to acknowledge the risks associated with many non-healthy behaviors such as smoking, which can reduce appetite and thus induce weight loss (resulting in an apparently healthy weight, when the individual in question may not be 'healthy' at all). Further messages to discourage the use of tobacco and other substances to promote weight loss may be of value for this cluster.The main strength of this study is the isolation of key message and media channel segmentation strategies that may help address the current obesity epidemic. The analysis also raises some important areas for future research and may inform future public policy efforts aimed at combating obesity in the US. Nevertheless, as with previous studies, although some demographic differences are significant between the five groups identified here, there are overweight people in all demographic groups [], suggesting that there is no "single message" or "single target" for social marketing efforts. Ultimately, while we can make suggestions on the types of messages that might prove to be useful for each segment, more in-depth research is needed within any targeted segment in order to determine specific message strategies that are likely to be most effective. It must also be emphasized that any analysis based on self-reports of behaviors must be cautiously interpreted, due to the well-documented trend of social desirability bias in respondent self-reports, particularly in food and weight-related research []. Nevertheless this study puts forth a preliminary market segmentation based on channel and message strategies that may help move Americans toward more healthy weights and lifestyles.ConclusionThe results of this study point to segments of the US population that may be responsive to social marketing messages that may help move them toward more healthy weights and behavior patterns. The study is among the first to apply market segmentation techniques to the domain of social marketing for obesity reduction and prevention: as such, the analysis raises some important questions for future research and gives insights into areas for public policy. Market segmentation may allow social marketing campaigns to reach specific audiences with the most effective message through the most effective media. Segmentation analysis facilitates the process of sending consumers the relevant messages specific to their lifestyles and their needs. The food industry has already employed this technique with well-documented success. Social marketing can play an important role in combating obesity and sedentary lifestyles by adopting similar techniques based upon clusters such as those identified through this research.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsJK conducted the analysis and drafted the original manuscript. She was the P.I. on the project grant. TR edited the manuscript, contributed to the literature review and prepared it for publication.
|
PMC2655754.txt
|
TITLE: Correction: Dopamine Inhibits Mitochondrial Motility in Hippocampal Neurons AUTHORS: Sigeng Chen, Geoffrey C. Owens, David B. Edelman ABSTRACT: No Abstract BODY: Figures and appeared out of order. Please view the correct Figure with its legend here:Figure 8Treatment with IBMX has different effects on hippocampal neurons and striatal neurons (A-C). A. Inhibition of mitochondrial movement by IBMX. Mean speed (μm/min) of all directionally moving mitochondria from pooled experiments before and after treatment with IBMX ( μM; n = , paired t-test; p<.). B. Western blot analysis shows that administration of IBMX ( μM) reduces Akt activity in hippocampal neurons, which is represented by decreased levels of phosphorylated serine- of Akt, relative to total Akt, over time. C. Western blot analysis shows that administration of IBMX ( μM) increases Akt activity in striatal neurons, which is represented by decreased levels of phosphorylated serine- of Akt, relative to total Akt, over time.
|
PMC2774947.txt
|
TITLE: Loss of RNA–Dependent RNA Polymerase (RDR2) Function Causes Widespread and Unexpected Changes in the Expression of Transposons, Genes, and -nt Small RNAs AUTHORS: Yi Jia, Damon R. Lisch, Kazuhiro Ohtsu, Michael J. Scanlon, Dan Nettleton, Patrick S. Schnable ABSTRACT: Transposable elements (TEs) comprise a substantial portion of many eukaryotic genomes and are typically transcriptionally silenced. RNA–dependent RNA polymerase (RDR2) is a component of the RNA–directed DNA methylation (RdDM) silencing pathway. In maize, loss of mediator of paramutation1 (mop1) encoded RDR2 function results in reactivation of transcriptionally silenced Mu transposons and a substantial reduction in the accumulation of nt short-interfering RNAs (siRNAs) that recruit RNA silencing components. An RNA–seq experiment conducted on shoot apical meristems (SAMs) revealed that, as expected based on a model in which RDR2 generates nt siRNAs that suppress expression, most differentially expressed DNA TEs (%) were up-regulated in the mop1 mutant. In contrast, most differentially expressed retrotransposons (%) were down-regulated. This striking difference suggests that distinct silencing mechanisms are applied to different silencing templates. In addition, >, genes (% of analyzed genes), including nearly % (/) of genes in chromatin modification pathways, were differentially expressed. Overall, two-thirds of differentially regulated genes were down-regulated in the mop1 mutant. This finding suggests that RDR2 plays a significant role in regulating the expression of not only transposons, but also of genes. A re-analysis of existing small RNA data identified both RDR2–sensitive and RDR2–resistant species of nt siRNAs that we hypothesize may at least partially explain the complex changes in the expression of genes and transposons observed in the mop1 mutant. BODY: IntroductionRepetitive sequences, including transposable elements (TEs) and tandem repeats, comprise a substantial fraction of many eukaryotic genomes. To protect genome integrity TEs are typically transcriptionally silenced via epigenetic mechanisms []–[]. At the core of many of these mechanisms are a variety of small RNAs. Diverse small RNA pathways exist in most eukaroytes producing microRNAs and short-interfering RNAs (siRNAs) that function to negatively regulate gene expression and/or to suppress the activity of transposons. siRNAs derived from double-stranded RNAs via mechanisms that are either RNA-dependent RNA polymerase (RDR) dependent or independent. Various RDRs (e.g. RDR1, RDR2 and RDR6) are functional in different siRNA pathways and most siRNAs are biosynthesized from the heterochromatic loci, a process that generally requires RDR2 and DICER-LIKE3 (DCL3)[]. In Arabidopsis, these heterochromatic nt siRNAs, predominate the small RNA population. These siRNAs recruit chromatin-targeted RNA silencing components to form transcriptionally silent heterochromatin, which is derived mainly from TEs and tandem repeats [], by cytosine methylation and various histone modifications, such as histone deacetylation and histone H3 lysine dimethylation. RDR2 is required for the biogenesis of most of these nt siRNAs [],[]. Thus, RDR2 is a key component of the chromatin-targeted RNA silencing process, which is also called the RNA dependent DNA Methylation (RdDM) pathway.The maize homolog of AtRDR2, mediator of paramutation1 (mop1) [], is required to establish and maintain paramutation at multiple genetic loci []. Paramutation is an epigenetic process in which the interaction of two alleles of a single locus causes a heritable change of one allele that is induced by the other allele. Paramutation was first observed on the red1 (r1) locus in maize by R.A. Brink in the 1950s in which specific weakly expressed alleles can heritably change other strongly expressed alleles to weakly expressed alleles []. In maize, paramutation is also observed in b1 and pl1 loci. The mop1 gene, which encodes the maize version of RDR2, is required for paramutation in b1, r1and pl1 loci, demonstrating that an RNA-dependent mechanism is important for paramutation in maize []. A tandem repeat kb upstream of b1 locus is required for paramutation in b1 locus. This repeat does not share sequence similarity with the b1 coding sequence, but has been demonstrated to physically interact with the b1 transcription start site []. Mutations in mop1 can also result in reactivation of transcriptionally silenced Mutator transposons and can substantially reduce the overall levels of nt siRNAs, demonstrating that in maize RDR2 contributes to the silencing of repetitive elements and plays an important role in the biogenesis of nt siRNAs [],[].Shoot apical meristems (SAMs) are responsible for the elaboration of all above-ground plant organs []. Maintenance of SAM identity and organogenesis are precisely regulated by a complex regulatory network involving various transcriptional factors and signal transduction proteins, as well as epigenetic factors []. In a previously conducted global gene expression analysis of maize SAMs and seedlings we identified ,+ genes as being preferentially expressed in SAMs []. This included ∼ retrotransposons that were substantially up-regulated in SAMs as compared to seedlings despite the fact that mop1 was expressed > fold higher in SAMs than in seedlings. It was not clear at that time why repetitive retrotransposons and mop1, which contributes to the silencing of repetitive elements, would be both significantly up-regulated in SAMs. Given that retrotransposon-derived transcripts represent ∼% of all SAM transcripts (as opposed to ∼.% of seedling transcripts)[], the SAM is an ideal model for studying the effects of mop1 on the accumulation of retrotransposon-derived transcripts.Here we report an analysis of an RNA-seq experiment that detected hundreds of transposons and thousands of the genes that are differentially expressed in mop1 mutant and non-mutant SAMs. This finding suggests that RDR2 plays a role in regulating the expression of not only transposons, but also of genes. Consistent with this observation, RDR2 mutants also exhibited a distinct SAM morphology relative to their wild type siblings, suggesting a role for RDR2 in normal SAM development. As expected based on its role in the RdDM pathway, loss of RDR2 function resulted in the up-regulation of many DNA TEs, retrotransposons and genes. However, some DNA TEs and many retrotransposons and genes were down-regulated in the mop1 mutant.ResultsCollection of SAMs via Laser Capture Microdissection (LCM) and RNA–seqA family segregating ∶ for mop1 homozygotes and heterozygotes was planted in a growth chamber and harvested -days after planting. Individual seedlings were genotyped to identify mop1 homozygotes and heterozygotes (see Materials and Methods). To determine the effects of mop1 on SAM development the ratios of height versus width of mutants (homozygous) and non-mutants (heterozygous) SAMs were compared (Figure ); mutant ratios were significantly smaller than non-mutant ratios (p-value = .; Table S1). SAMs plus leaves at plastochron (P0) and P1 stages were collected via LCM followed by RNA extraction and amplification according to our previously published procedures []. A pooled RNA sample from twelve mutant SAMs and a pooled RNA sample from ten non-mutant SAMs were subjected to Illumina/Solexa sequencing (see Materials and Methods). . million reads from mop1 mutants and . million reads from non-mutants could be uniquely mapped to the Maize Genome Sequencing Project's (MGSP's) B73 reference genome [] (see Materials and Methods)../journal.pgen..g001Figure 1Comparison of SAM morphologies between mop1 mutants and non-mutants.Safranin O/FastGreen stained image of SAMs from (A) a mop1/mop1 mutant and (B) a non-mutant sibling. See Table S1 for a quantitative analysis.Expression of various TEs is differentially affected by loss of RDR2 functionMore than % of the B73 reference genome is composed of TEs []. To ensure genome stability these elements are mostly suppressed via genome defense systems such as chromatin-based silencing, which is guided and reinforced by nt siRNAs []. RDR2 plays an important role in the biogenesis of nt siRNAs []. To test the hypothesis that the loss of RDR2 function in the mop1 mutant would result in the activation of TEs (i.e, an increased accumulation of TE-derived transcripts), the Illumina/Solexa reads obtained from the SAM were aligned to all annotated TEs in the B73 genome (see Materials and Methods). As expected based on the mechanism associated with the RdDM silencing pathway, many DNA TEs (class II) and retrotransposons (class I) were up-regulated in the mop1 mutant (Table ). However, a significant fraction of DNA TEs and retrotransposons were down-regulated (Table ). The most strongly up-regulated DNA TE super-families were Stowaway and Tourist and the most strongly down-regulated DNA TE family was CACTA (Figure S1). Both the Ty1/Copia-like and Ty3/Gypsy-like super-families of retrotransposons include some families that were up-regulated and others that were down-regulated (Figure S2)../journal.pgen..t001Table 1Expression patterns from diverse super-families of TEs.TE classSuperfamilya TotalUpb Downb Up+Downb IRIL3022RIX1811213RLC1534462106RLG18849105154RLX24645112157 Total (%) (%) II CACTA hAT Helitron Mariner MULE PIF/Harbinger Stowaway Tourist Total (%) (%) a RIL, LINE (L1) retrotransposons; RIX, unknown LINE retrotransposons; RLC, Ty1/Copia LTR retrotransposons; RLG, Ty3/Gypsy LTR retrotransposons; RLX, unknown LTR retrotransposons. b Up, number of up-regulated sub-families/families; Down, number of down-regulated sub-families/families; sub-families for class I TEs; families for class II TEs; Up+down, total number of differentially expressed sub-families/families.To assess the effects of mop1 on specific groups of DNA TEs, each DNA TE super-family was divided into families based on a phylogenetic analysis conducted by the Maize Transposon Consortium that defined unique families (i.e., monophyletic clades) []. Of these families, % (/) were differentially expressed in the mop1 mutant (Table ; Table S2). Consistent with our hypothesis, most (%; /) of the differentially expressed DNA TEs were up-regulated in the mop1 mutant. In particular, of the / TE families annotated as Mutator-like elements (MULEs) that were differentially expressed, (%) were up-regulated. This is consistent with the report that silenced Mutator TEs can be reactivated in mop1 mutants []. Similarly, of the / hAT families that were differentially expressed, % (/) were up-regulated. Among the / differentially expressed CACTA families many (/) individual CACTA families were up-regulated. Even so, CACTA DNA TEs as a group were down-regulated by loss of RDR2 function (Figure S1). The up-regulation of one MULE element and the down-regulation of one hAT element (viz., an Ac-like element) were confirmed by quantitative real-time PCR (qRT-PCR) (Figure )../journal.pgen..g002Figure 2Validation of eight differentially expressed genes via qRT–PCR.Eight differentially expressed genes were chosen from the RdDM pathway (ago4a [Chromdb ID: AGO104], ago4b [Chromdb ID: AGO105], ago4c [Chromdb ID: AGO119], ddm1 [Chromdb ID: CHR101], met1 [Chromdb ID: DMT101]), transposons (hAT [a member of the ZM_hAT_8 sub-family; http://www.maizesequence.org]) and MULE [a member of the MULE sub-family DTM_Zm33205; http://www.maizesequence.org]) and a regulator of SAM development (liguleless3 [Gene model ID: GRMZM2G087741; http://www.maizesequence.org]) for qRT–PCR validation. Primers used for qRT–PCR are presented in . Fold change was presented as the relative abundance of transcript in the mop1 mutant/non-mutant. The quantitative fold changes obtained from between RNA–seq and qRT–PCR experiments were significantly correlated (Pearson correlation coefficient was ., r2 = ., p-value = .). A t-test of equal expression between the mutant and non-mutant using the data from four biological replications of qRT–PCR were conducted (p-value ≤., *; p-value ≤., **).A similar analysis was performed on retrotransposons, which have been categorized by the Maize Transposon Consortium into super-families (e.g., RLG, retrotransposon LTR Gypsy), families (e.g., Huck) and sub-families (e.g., Ac186577_1525). % of the unique retrotransposon sub-families were differentially expressed to various degrees in the mop1 mutant (Table S3). Consistent with current understanding of the RdDM silencing pathway, approximately one-third (%) of the differentially expressed retrotransposon sub-families were up-regulated in mop1 mutants as compared to non-mutants. But inconsistent with expectations, the majority (%) of the differentially regulated retrotransposons were down-regulated in the mutant (Table ; Table S3). Some specific sub-families of retrotransposons were mostly up- or mostly down-regulated. For example, within the Ty1/Copia-like super-family, / Ji and / Opie sub-families were up-regulated. Within the Ty3/Gypsy-like super-family, / Flip and / Huck sub-families were up-regulated. In contrast, / Ty3/Gypsy-like Cinful-zeon sub-families and / annotated Ty3/Gypsy-like Prem1 sub-families were down-regulated in the mop1 mutant (Table S3). These distinct responses to loss of RDR2 function suggest that the expressions of different retrotransposon families are regulated via different mechanisms.Many chromatin modification genes are down-regulated in the mop1 mutantTo study the impacts of mop1 on chromatin modification pathways, maize genes annotated as being involved in chromatin modification were downloaded in Dec. from ChromDB []. In total, of these genes can be uniquely mapped to the B73 reference genome (see Materials and Methods). RNA-seq reads that map to these chromatin-associated genes were used to conduct Fisher's exact tests. Nearly % (/) of these chromatin pathway genes were differentially expressed between the mop1 mutants and non-mutant siblings using a % false discover rate (FDR) cutoff, a frequency far higher than observed when considering all genes (%). Approximately ¾ (%) of differentially expressed chromatin genes were down-regulated in the mop1 mutant (Table S4). A wide variety of chromatin-associated genes exhibited differential expression between mop1 mutants and non-mutants, including those affecting various histone modifications, such as histone ubiquitination, methylation, acetylation and deacetylation (Table S4). As the maize homolog of AtRDR2, mop1 is expected to function in the RdDM pathway, which involves the biogenesis of nt siRNAs, de novo methylation of DNA, maintenance of DNA methylation, and demethylation []. Almost all of the genes known to be implicated in the RdDM pathway were down-regulated in the mop1 mutants (Table ). The maize genome contains two DCL3 paralogs, which are involved in the biogenesis of nt small RNAs. One of these (Zmdcl3b) was down-regulated while the other (Zmdcl3a) was up-regulated (Table ). These opposite responses suggest that the two DCL3 paralogs may be functionally distinct. This type of divergent gene expression pattern was also observed for DRM protein in which one maize homolog was down-regulated and another was up-regulated. In addition to DRM protein, AGO4, DRD1, and MET1proteins function in the de novo methylation pathway []. All maize homologs of these genes were down-regulated (Table ). CMT3, MET1, DDM1, HDA6, SUVH4 function in the maintenance methylation pathway []. All maize homologs of these genes were down-regulated (Table ). DNA demethylation is thought to regulate epigenome dynamics in opposition to the RdDM pathway. In Arabidopsis, ROS1 and DME remove DNA methylation []. There are two homologs of DME gene in maize and both were down-regulated as well (Table ). The expression levels of several of genes important for epigenetic silencing (viz., met1, met3, three ago4 paralogs, and ddm1) were tested via qRT-PCR with results that were consistent with those obtained from RNA-seq (Figure ). These observations demonstrate that there is widespread down-regulation of components in the RdDM pathway in the mop1 mutant, suggesting either that MOP1 positively regulates the entire pathway, or that genes involved in chromatin modification and DNA methylation are co-regulated in maize../journal.pgen..t002Table 2Differentially expressed genes in RdDM pathway.Chromdb IDa Gene Nameb log2(FC)c FDRd DCL102 dcl3a .28e-06DCL104 dcl3b −.51e-02AGO104 ago4a −.79e-21AGO105 ago4b −.76e-17AGO119 ago4c −.51e-06CHR127 drd1 −.31e-11DMT101 met1 −.03e-11DMT106 drm1/ −.73e-13DMT102 cmt3a −.03e-40DMT105 cmt3b −.74e-22SDG118 suvh4/kyp −.04e-16HDA108 hda6 −.72e-09CHR101 ddm1 −.58e-04DNG101 dme1 −.04e-12DNG103 dme2 −.79e- a ID used in Chromdb (http://www.chromdb.org). b ID used in general RdDM pathway []. c log2 transformation of fold change as the relative abandunce of transcripts in mutants/non-mutants. Positive value indicates the up regulation and negative value indicates down regulation. d The false discovery rate calculated using Benjamini and Hochberg's procedure [] for the p-value from Fisher's exact test.In addition to mop1, mutations in two other maize genes are known to affect the accumulation of both – nt siRNAs and DNA methylation. rmr1 (required to maintain repression1) encodes a SWI/SNF2 class chromatin remodeling protein. Mutations in rmr1 have dramatic effects on accumulation of – nt siRNAs, maintenance of the repressed state of paramutant genes, and methylation of Mu transposons [],[]. Unlike the related DDM1 orthologs (Table ), expression of rmr1 was not significantly changed in the mop1 mutant. This observation is consistent with the suggestion that RMR1 may act genetically upstream and sometimes independently of RDR2 []. The second cloned gene that affects – nt siRNA accumulation in maize is rmr6. This gene encodes the conserved Pol IV largest subunit (RPD1) and is required for paramutation []. Although expression of this gene was reduced in the mop1 mutant, this change was not significant.Widespread changes in gene expression following loss of RDR2 activityThe finding that SAM morphology differs between mop1 mutant and non-mutants led us to hypothesize that mop1 affects not only the expression of TEs and components in RdDM pathway but also genes important to the development of the SAM. To test this hypothesis the Illumina/Solexa reads were mapped to the “filtered gene set” of maize generated by the MGSP (see Materials and Methods). Among reads that could be uniquely mapped to the genome, . million (%) from mop1 mutants and . million (%) from non-mutants aligned to gene models (Figure S3; Table S5). At least one Illumina/Solexa read from at least one of the two genotypes aligned to , of the , genes in the MGSP's filtered gene set (Table S5). Of these genes, , (% of ,) could be declared to be differentially expressed between the mutant and non-mutant pools at an estimated % FDR []. The ratio of number of genes that were up-regulated in the mop1 mutant to those that were down-regulated was ∼∶ (Table S6). Consistent with our finding that mop1 mutant SAMs differ morphologically from non-mutant siblings (Figure ; Table S1), several key regulators of SAM development, including fasciated ear2 [], terminal ear1-like [], outer cell layer4 [] and liguleless3 (encoding a Knotted class homeodomain protein) [] were differentially expressed (Table ). The differential expression of one of these genes, liguleless3, was validated via quantitative real-time PCR (qRT-PCR) (Figure ). This finding suggests that nt siRNAs play a role in regulating (directly or indirectly) the expression of not only transposons, but also of genes../journal.pgen..t003Table 3Key regulators of SAM developments showing differential expression.Gene IDa log2(FC)b FDRc SwissProt IDd Protein NameE-valuee RefGRMZM2G104925−..00e-04Q940E8Fasciated ear20 [] GRMZM2G085113−..62e-09A9XIW7Terminal ear1-like protein6e- [] GRMZM2G123140−..05e-02B3GW90Putative HD-ZIP IV family transcription factor OCL40 [] GRMZM2G0877410..09e-06Q9SYT6Knotted class homeodomain protein liguleless30 [] a Refer to http://www.maizesequence.org. b log2 transformation of fold change as the relative abandunce of transcripts in mutants/non-mutants. Postive values indicate up regulation and negative values indicate down regulation. c False discovery rate calculated using Benjamini and Hochberg's procedure [] for the p-value from Fisher's exact test. d Protein ID retrieved from SwissProt_Trembl database via blastp (1e- as cutoff). e E-value from blastp search.Re-analysis of existing nt siRNA dataThe analyses described above demonstrate that loss of RDR2 function results in widespread changes in the accumulation of transcripts from DNA TEs, retrotransposons and genes. As expected based on a model in which RDR2 generates nt siRNAs that suppress expression, many TEs and genes were up-regulated in the mop1 mutant. Interestingly, some DNA TEs and many retrotransposons and genes were actually down-regulated in the mop1 mutant, demonstrating new complexities in the regulation of the expression of transposons and genes.To explore this complexity we undertook a re-analysis of nt siRNAs isolated from immature ears of mop1 and non-mutant plants []. Nobuta et al. reported that mop1 mutants accumulated many fewer nt siRNAs (as a proportion of all small RNAs) than do wild-type []. Their analysis treated all nt siRNAs as a group. We extended their analysis by considering the effects of the mop1 mutation on the accumulation of each individual species of nt siRNAs in their data set. As such, our analysis enabled us to identify specific RNA species that make up a significantly greater proportion of the observed reads from one genotype than from the other (see Materials and Methods).Considering the union of mop1 mutant and non-mutant reads, the Nobuta et al. data set contains >.3M distinct nt siRNA species. Many of these are present at very low abundance and as such it would not be possible to detect statistically significant differences between the two genotypes even if such differences exist. Of the % of RNA species for which or more counts were recorded in the union of the two genotypes (,/.3M), we found that % (,/,) of the nt siRNAs were differentially expressed between the two genotypes. Consistent with the report of Nobuta et al., most of the , differentially regulated species of nt siRNAs (,) were down-regulated in the mop1 mutant (Figure S4; Table S7). We term these “RDR2-sensitive nt siRNAs”. Quite unexpectedly, , distinct species of nt siRNAs were significantly “up-regulated” in the mop1 mutant (Figure S4; Table S7). Although some of these may be actually up-regulated, others may simply be less down-regulated in the mutant than are other species of nt siRNAs (see Materials and Methods). We have therefore termed these “up-regulated” species “RDR2-resistant nt siRNAs”.DiscussionRDR2 is an essential component of the heterochromatin silencing pathway in multiple species [],[] and functions in DNA and histone methylation, the biogenesis of nt siRNAs and the silencing of repetitive DNAs []. The maize homolog of RDR2, mop1, was originally identified as a mutant that functions as an epigenetic regulator of a target gene via interactions with upstream tandem repeats []. mop1 is also required for the methylation of the terminal inverted repeats of Mu TEs and for the maintenance of silencing of MuDR transposons []. Based on its mutant phenotypes, it has been hypothesized that the mop1 gene regulates many loci []. To test this hypothesis and to examine the effect of RDR2 on the silencing of TEs in a large, complex genome, we conducted RNA-seq experiments on SAMs of mop1 mutant and non-mutant seedlings. SAMs were selected for analysis because they are responsible for the elaboration of all aerial organs [], they have a complex transcriptome [], and our prior analyses had revealed that multiple retrotransposons and mop1 are all substantially up-regulated in SAMs as compared to seedlings [].Effect of RDR2 on gene expressionWe identified more than , genes whose expression differed between mop1 mutant and non-mutant SAMs. These widespread differences in gene expression are consistent with the multiple developmental defects associated with the loss of mop1 function in mutants [].Over several generations, maize lines that carry the mop1 mutation can accumulate a variety of epimutant phenotypes (Lisch, unpublished data). In this study we controlled for the effects of any segregating epi-alleles by analyzing RNA from pools of mop1 and non-mutant SAMs. However, our discovery that genes involved in a variety of silencing pathways including DNA methylation, histone modification and RNA-mediated silencing, are differentially regulated in the mop1 mutant complicates any facile explanation for the origins of these phenotypes. Unlike rdr2 mutants in Arabidopsis, ddm1 and met1 mutants can have severe effects on plant morphology [], and the maize homologs of both of these genes are down-regulated in the mop1 SAMs. It is not clear whether or not the down-regulation in the maize mop1 mutants of so many genes involved in epigenetic regulation is consequential, but it does suggest that many of the phenotypes that arise in mop1-containing lines over multiple generations may not be the direct result of the loss of RDR2 activity. The same may well be true for at least some of the differences in gene expression that we observed.Effect of RDR2 on the accumulation of TE-derived transcriptsRNA-seq data identified hundreds of DNA TEs and retrotransposons that are differentially regulated in SAMs. Based on a model in which RDR2 generates nt siRNAs that silence DNA TEs and retrotransposons, our expectation was that loss of RDR2 function in mop1 SAMs would result in the up-regulation of DNA TEs and retrotransposons. Although we did observe that the majority of differentially expressed DNA TEs (%) were up-regulated in the mop1 mutant, less than half of all differentially regulated retrotransposons (%) were up-regulated. This suggests that at least some DNA TEs and retrotransposons are silenced via distinct mechanisms.RDR2–dependent silencing of pericentromeric TEsPericentric heterochromatin is rich in TEs in many species, and these sequences are typically heavily methylated and associated with large numbers of nt siRNAs []. Consistent with its role in the RdDM pathway loss of RDR2 function results in the up-regulation of certain TEs, including Huck elements which are members of the Ty3/Gypsy-like super-family of retrotransposons. Fluorescence in situ hybridization (FISH) experiments reveal that although Huck elements are dispersed along all chromosomes they are significantly enriched in the vicinity of, but not in, centromeres []. Indeed, in general Ty3/Gypsy-like sequences cluster in pericentromeric regions across all grass species []. Our observations provide evidence that at least one pericentromeric repeat (i.e., Huck) is transcriptionally silenced via the RdDM pathway.RDR2–sensitive and RDR2–resistant nt siRNAsIn Arabidopsis RNA gel blot experiments, the population of nt siRNAs is almost entirely eliminated in the rdr2 mutant [], indicating RDR2 is required for the biogenesis of nearly all nt siRNA. In the maize mop1 mutant, the population of nt siRNAs is dramatically reduced, but not eliminated []. Via a re-analysis of an existing small-RNA data set we identified >, unique “RDR2-sensitive” and ∼, unique “RDR2-resistant” nt siRNAs that are “down-regulated and “up-regulated” in the mop1 mutant, respectively.RDR2–independent silencing of DNA TEs and retrotransposonsIn contrast to elements such as Huck, the silencing of some types of DNA TEs and retrotransposons (e.g., most Prem1 elements) does not appear to require RDR2, as evidenced by the fact that they are down-regulated in the mop1 mutant. The hypothesis that an RDR2-independent heterochromatin silencing pathway exists in maize is consistent with our previous observation that many retrotransposon are significantly up-regulated (some >,×) in SAMs as compared to seedlings even though mop1 transcripts accumulate in SAMs to a level × higher than in seedlings. On the other hand, because new retrotransposon insertions are quite rare in maize [], we considered the possibility that a significant proportion of the retrotransposon-derived transcripts we detected in SAMs are generated via RDR2 activity itself [], which can produce aberrant non-polyadenylated RNAs. If this were the case, these species would indeed be lost in the RDR2 mutant (along with associated siRNAs). However, because the procedures we used to construct our RNA-seq libraries preferentially target mRNA species this possibility seems unlikely.We therefore considered other RDR2-independent mechanisms for silencing DNA transposons and retrotransposons in a complex genome such as that of maize. Because the expression of many genes is affected by the mop1 mutant, it is possible, for example, that some of these effects could be antagonistic to the direct effects of mop1 on gene silencing. In addition, Lippman et al. [] reported that the epigenetic inheritance of different TEs differed from mutant to mutant in Arabidopsis and proposed the existence of distinct but interacting pathways responsible for transposon silencing via siRNAs and histone modifications. Observations from fission yeast offer a plausible possibility for an RDR2-independent pathway. In this yeast, inhibition of histone deacetyltransferases causes an inherited loss of heterochromatin []. Several genes encoding histone deacetyltransferases (HDACs) were up-regulated in the mop1 mutant. It is possible that enhanced expression of these HDACs could enhance silencing of some TEs. Similarly, the reduction in expression of the maize orthologs of ROS1 and DME1, both of which are required for demethylation of a variety of target genes in Arabidopsis [], could result in the silencing of a variety of genes in mop1 mutants. Hence, our observations in the mop1 mutant of the down-regulation of some DNA TEs and many retrotransposons, enhanced expression of genes in the HDAC silencing pathway, and decreased expression of genes in the demethylation pathway are consistent with the existence of multiple silencing mechanisms, but suggest that these mechanisms can potentially interact antagonistically.Nobuta et al. reported that nt small RNAs are highly abundant in the mop1 mutant [] and suggested that these small RNAs may be the result of an alternative mode of heterochromatic siRNA production that is independent of, and may even be enhanced by, the loss of RDR2. Alternatively, or in addition, the RDR2-independent silencing we observed could be the result of the RDR2-resistant nt siRNAs we identified. As discussed in the Materials and Methods section, these RDR2-resistant nt siRNAs may actually be produced at higher levels in the mop1 mutant. If this were the case, then these RDR2-resistant siRNAs could be responsible for the enhanced silencing of some of the DNA TEs and retrotransposons we observed in the mop1 mutant. If, on the other hand, the RDR2-resistant siRNAs are simply less susceptible to loss of RDR2 function, they would need to be more effective at silencing in the mop1 background to explain the enhanced silencing of DNA TEs and retrotransposons we observed. RDR2-resistant nt siRNAs might, for example, exhibit enhanced repressive activity in response to changes in chromatin structure resulting from loss of RDR2 activity.Potential sources of RDR2-resistant siRNAs include novel combinations of sense/anti-sense transcripts and transcribed inverted repeats. In maize retrotransposons are often present in vast nested arrays []. Enhanced transcription of these nested retrotransposons (due perhaps to loss of RDR2-dependent silencing) could result in the production of novel combinations of sense and antisense RNAs that could be processed into biologically active siRNAs even in the absence of RDR2. Thus, the effects of the mop1 mutant on a given transposon family may be a reflection not just of its sequence, but of the physical distribution of that family within the genome. With that in mind, it is interesting to note that the one family of DNA TEs with a high proportion of up-regulated members (CACTA elements) exhibits a distinct chromosomal distribution relative to the other families of DNA transposons. CACTA elements are significantly more likely to be found in gene-poor, heterochromatic regions of the genome than are all other DNA TEs [].In addition, dosage effect has been reported in the maize Activator/Dissociation (Ac/Ds) and Mutator transposon families []–[], demonstrating that the regulation of the TE transposition is complex; similar complexity likely contributes to our observation of the down-regulation of TEs in the mop1 mutant. For example, some of the changes in expression of TEs and genes observed in this study could be due to epigenetic interactions between TEs and genes. In several cases it has been demonstrated that expression of TEs can reduce expression of nearby genes []–[]. For example, de-repression of an LTR retrotransposon flanking the BONSAI gene in Arabidopsis in a ddm1 mutant background results in epigenetic silencing of BONSAI []. This process involves the production of BONSAI-specific siRNAs. Given our observation that DNA TEs, which tend to be preferentially located in gene-rich regions of the genome are likely to be up-regulated in the mop1 mutants, it is possible that many of the negative effects on gene expression are due to similar interactions between genes and nearby TEs. To more comprehensively analyze the relationship between levels of gene expression, siRNA production, and DNA methylation, it will be necessary to analyze all of these variables in a single tissue. Further, given the number of variables involved, a clear understanding of cause and effect relationships between RDR2 activity and expression will require detailed analyses of individual transposons, retrotransposons and genes.Materials and MethodsGenetic stocks, plant growth conditions, genotyping, and RNA–seqThe mop1- allele used in this study has been described previously []. This mutant is within a mixed genetic background, including both the highly inbred a1-mum2 minimal Mutator line [] and the Mutator line from which mop1- was first derived []. The mop1- mutation in this background was maintained through several generations via sib crosses, self fertilizations, or back-crosses with the a1-mum2 stock. Although the progenitors of this line contained active MuDR elements, these elements were no longer present in the line used for this study, which lacked detectable Mutator activity.This genetic background is distinct from that analyzed by Nobuta et al. []. Importantly, the family used in these experiments is closely related to a mop1 mutant lineage that gave rise to a large number of unique morphological phenotypes not previously observed in mop1 mutant plants (Lisch, unpublished observation). Given this, and given the dramatic differences in TE composition between maize inbred lines, direct comparisons of transcript data between the current data set and that of Nobuta et al. should be treated with caution.A plant having the genotype mop1-/mop1- was crossed to Mop1/mop1- heterozygote and the resulting progeny kernels planted in growth chambers (PGW-, Percival Scientific, http://www.percival-scientific.com). Temperature and light cycles were set as degrees for hours of light and degrees for hours of dark. During the light period the light intensity at the surface of the growth medium was maintained between and umol m− sec−.At -days after planting SAMs were collected using the PALM MicroBeam System (115V Z, P.A.L.M. Microlaser Technologies, http://www.palm-microlaser.com). Plants homozygous and heterozygous for the mop1- mutant allele were distinguished using two pairs of primers: a pair of mop1-specific primers consisting of RDRF3 (sequence: ′-TCTCCACCGCCCACTTGAT-′) and RDRR2 (sequence: ′-ATGGCCAGCAGGGTGTCGCAGAT-′) and a primer pair consisting of the Mutator TIR primer Mu-TIR (′-AGAGAAGCCAACGCCAWCGCCTCYATTTCGTC-′) and the mop1-specific primer RDRF3. Twelve mop1-/mop1- and ten Mop1/mop1- SAMs were used to form mop1 mutant and non-mutant pools. Collected SAM tissues were used for RNA extraction, RNA amplification and synthesis of double stranded cDNAs according to our previous published procedures []. These procedures preferentially target polyadenylated transcripts. Illumina/Solexa libraries were constructed using these double stranded cDNAs following Illumina/Solexa's standard protocol for genomic library preparation. The resulting libraries were sequenced on the Solexa 1G Genome Analyzer at the Michael Smith Genome Sciences Centre (Vancouver, BC, Canada). Each library was sequenced using lanes on a Solexa flow cell. The Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo) accession number for the data used in the paper is GSE16789.Alignments of RNA–seq to the maize reference genome and TEsThe resulting Solexa reads were aligned to the maize B73 reference genome (Release 4a.) (http://www.maizesequence.org) with the short read aligner NOVOALIGN (http://www.novocraft.com) using base sequences. Low quality bases located at the end of reads were trimmed and only reads that mapped uniquely to the genome with a maximum of two mismatches including insertion/deletion (indel) across bases were used for subsequent analyses. The “filtered gene set”, a collection of high-quality gene models developed by the MGSP, was projected onto the B73 reference genome.In addition, the Illumina/Solexa reads were also aligned directly to the DNA TE families and retrotransposon subfamilies. Due to the repetitive property of the TEs, each read is allowed to be mapped to multiple DNA TE families or retrotransposon subfamilies but each read is only counted once within each family or sub-family with same alignment criteria as used for alignments to the reference genome.The chromatin-associated genes were mapped to the maize B73 reference genome using criteria of % identity and % coverage. Reads that uniquely mapped to the reference genome were projected onto each of these chromatin-associated genes allowing us to detect differential expression.Identifying differential expression via a likelihood ratio test and Fisher's exact testTwo statistical procedures to identify differentially expressed genes were compared and evaluated: a likelihood ratio test based on a Poisson model (below) [] and Fisher's exact test. Although the two procedures produced similar p-values (R = .; Figure S5), the Fisher's exact test was more conservative. It was therefore selected for use in this study.The likelihood ratio test analysis generally followed the procedure described in Marioni et al. []. For each gene, the number of reads from the mop1- mutant sample and the non-mutant sample were modeled as independent Poisson random variables with mean λmCm for mutant and mean λnCn for non-mutant, where Cm and Cn denote counts of the total number of mapped reads for the mutant (,,) and non-mutant (,,) samples, respectively. It is straightforward to show that the likelihood ratio statistic for testing the null hypothesis of H0: λm = λn is T = {km log (km/Cm)+kn log(kn/Cn)−k log(k/C)}, where km is the number of mutant reads for the gene in question, kn is the number of non-mutant reads for the gene in question, k = km+kn, and C = Cm+Cn. This statistic is distributed approximately as a chi-square random variable with degree of freedom when the null hypothesis of no differential expression is true. Thus, p-values were obtained by comparing the observed statistic for each gene to the chi-square distribution with degree of freedom. Ideally, sequencing would have been carried our separately for multiple independent biological replications of each genotype so that over dispersion relative to the Poisson distribution could have been assessed and accounted for using an analysis like that proposed by Robinson and Smyth []. Note that our qRT-PCR validation and analysis (discussed below) was based on separate measurements of independent biological replications.Detection of RDR2–sensitive and –resistant nt siRNAsEach species of nt siRNAs was tested whether the proportions in the library between mop1 mutant and non-mutant were significantly different via Fisher's exact test. Because we are able to measure only the abundance of each species in a genotype relative to the total number of reads for that genotype, it is difficult to formally distinguish nt siRNAs that are up-regulated in the mutant from those that make up a significantly greater proportion of the observed reads only because of the absence of many other nt siRNA species in the mop1 mutant, thereby making them proportionately more abundant. For the purposes of this study we therefore carefully define up-regulation to mean that a particular species is significantly more abundant in one sample of reads than in another. It is important to note that this does not necessarily mean that the number of RNA molecules of that particular species increases on a per cell basis.qRT–PCR validation and data analysisPrimer design for qRT-PCR was conducted as described []. RNA samples independent from those used in the RNA-seq experiment were extracted from four biological replications from mop1-/mop1- and Mop1/mop1- ( SAMs pooled within each replicate per genotype) by using the same procedure as the RNA-seq experiment. To prepare the cDNA template, combined oligodT and random hexamers was used to perform reverse transcription reactions at °C for hour with SuperScript III. A reverse transcription without SuperScript III was conducted to control for genomic DNA contamination. qRT-PCR was conducted on an Mx4000 multiplex quantitative PCR system (Stratagene). RNA from a human gene (GenBank accession no. AA418251) was spiked into each reaction as an external reference for data normalization. Genotype-specific Ct values for each gene and control were calculated and then the ΔΔCt was computed. For each gene, ΔΔCt across biological replications was used to conduct a t-test in R (www.r-project.org) [].Supporting InformationFigure S1Overall expression fold changes of mutant versus non-mutant for differentially expressed DNA TEs. In this analysis all members of each differentially expressed super-family were treated as a group. The percentage of reads that match each super-family among all mapped reads in each genotype was calculated and the fold change was computed as the ratio of the percentage of mutant versus non-mutant for each super-family.(. MB TIF)Click here for additional data file.Figure S2Overall fold changes of mutant versus non-mutant for differentially expressed retrotransposons. In this analysis all members of each differentially expressed family were treated as a group. The percentage of reads that match each family among all mapped reads in each genotype was calculated and the fold change was computed as the ratio of the percentage of mutant versus non-mutant for each family.(. MB TIF)Click here for additional data file.Figure S3Distribution of numbers of mapped reads across tested genes.(. MB TIF)Click here for additional data file.Figure S4RDR2-sensitive and RDR2-resistant nt siRNAs in wild-type and mop1 mutants. The log2 transformation of read counts in non-mutant (x-axis) versus mop1 mutant (y-axis) for each species of the , RDR2-resistant nt siRNAs (green dots) and the , RDR2-sensitive nt siRNAs (red dots) were plotted.(. MB TIF)Click here for additional data file.Figure S5p-value comparison between likelihood ratio test and Fisher's exact test.(. MB TIF)Click here for additional data file.Table S1 mop1 mutants and non-mutants have distinct SAM morphologies.(. MB PDF)Click here for additional data file.Table S2List of differentially expressed DNA transposon families.(. MB PDF)Click here for additional data file.Table S3List of differentially expressed retrotransposon sub-families.(. MB PDF)Click here for additional data file.Table S4List of differentially expressed chromatin-associated genes.(. MB DOC)Click here for additional data file.Table S5Alignment of RNA-seq reads to genome and genes.(. MB PDF)Click here for additional data file.Table S6List of differentially expressed genes and related annotation.(. MB DOC)Click here for additional data file.Table S7List of RDR2-sensitive and RDR2-resistant nt siRNA.(. MB PDF)Click here for additional data file.Table S8Primer sequences used for qRT-PCR experiment.(. MB PDF)Click here for additional data file.
|
PMC2585092.txt
|
TITLE: Food consumption patterns in the Waterloo Region, Ontario, Canada: a cross-sectional telephone survey AUTHORS: Andrea Nesbitt, Shannon Majowicz, Rita Finley, Frank Pollari, Katarina Pintar, Barbara Marshall, Angela Cook, Jan Sargeant, Jeff Wilson, Carl Ribble, Lewinda Knowles ABSTRACT: BackgroundThe demographics and lifestyles of Canadians are changing, thereby influencing food choices and food preparation in the home. Although different dietary practices are associated with increased risk of foodborne illness, our ability to evaluate food consumption trends and assess risks associated with foodborne illness is limited by lack of data on current eating habits and consumer food safety practices. The objective of this study was to describe, for the first time, the food consumption patterns in a Canadian-based population from a food safety perspective, in order to establish baseline data on actual food intake of individuals.MethodA cross-sectional telephone survey of , randomly selected residents of Waterloo Region, Ontario, Canada (C-EnterNet pilot site) was conducted between November and March . Food intake was assessed using a -day dietary recall method.ResultsCertain food items were consumed more than others among the same food groups, and consumption of many food items varied by gender and age. Specific foods considered high-risk for the transmission of certain enteric pathogens were significantly more likely to be consumed by males (i.e. unpasteurized juice, bean sprouts, and undercooked meat) and elderly individuals (i.e. undercooked eggs). The majority of households prepared and consumed most meals at home, allocating an average of minutes to prepare a meal.ConclusionBaseline data on actual food intake is useful to public health professionals and food safety risk assessors for developing communication messages to consumers and in foodborne outbreak investigations. BODY: BackgroundFactors that influence food consumption choices among individuals and populations include cultural, social, and economic factors []. As food consumption patterns change over time, public health authorities and the food industry need to monitor the dietary intakes of the population [].In Canada, several sources provide information on food consumption patterns and nutrient intake. Food disappearance data and household food expenditure data have been used to identify national trends in food availability, and illustrate the dynamics of the food supply and consumer demands [,]. However, these data have limitations in that they do not measure individual consumption [-].Individual dietary habits were first assessed by the Nutrition Canada Survey in , providing valuable data on the intakes of several nutrients and the food consumption patterns of various sub-populations within Canada []. Current data on Canadian food consumption is limited to nutritional intake and status [-], or to specific subgroups such as adults [,], native communities [], specific cultural communities [], or rural populations [].Although it is important to monitor the nutritional status of the population, it is equally important to establish baseline data on actual food intake and practices of individuals. To address this, a survey on food consumption patterns from a food safety perspective was initiated in a Canadian community. The objectives of the study were to evaluate the patterns of food consumption in the general population, and to describe demographic factors that relate to the consumption of specific food items.MethodsThis study was part of a broader cross-sectional telephone survey addressing food consumption patterns, knowledge of food safety principles, the prevalence of unsafe food-handling practices, and the prevalence of gastrointestinal symptoms, which was conducted in the Waterloo Region (Ontario, Canada) between November and March []. The Waterloo Region is located in southern Ontario, Canada, and is composed of three urban and four rural municipalities. Waterloo Region was selected since (a) it is the pilot sentinel site for C-EnterNet, the Public Health Agency of Canada's national integrated enteric pathogen surveillance program [], (b) it includes both rural and urban settings, and (c) it has a population size of over ,, with diverse ethno-cultural communities.The expected frequency of specific food consumption and handling behaviours deemed to be of most importance relating to the objectives of the study were used to determine sample size. We assumed expected frequencies of .%, .%, and .% for eating egg dishes with runny yolk, using treated tap water, and washing hands after handling raw meat, respectively, in accordance with previous studies [,]. In addition, we considered an estimated prevalence of gastrointestinal illness of % per month [], since a secondary objective of the overall survey was to evaluate the relationship between food consumption and symptoms of gastrointestinal illness. The final sample size of , was based on a combination of these calculations, using an allowable error of % and a % confidence. Sample size calculations were performed using Epi-Info Version . (CDC, Atlanta).A multistage random-sampling design was used, and the seven municipalities within the study area were sampled proportionally as per the Census []. Households were randomly sampled from a list of residential telephone numbers (SelectPhone Canada™, InfoUSA Inc., version .). The individual in the household with the next birthday was selected to participate in the survey provided they consented. If consent was not provided, the interview was terminated and the call status was recorded as not willing to participate. Respondents were considered eligible if they spoke English, were over the age of months, and resided at a listed telephone number in the study area. Individuals were excluded from the survey if they had traveled outside of Canada within seven days prior to the interview, in order to capture Canada-specific food consumption.Residential telephone numbers were attempted three times, with each attempt on different days and at different times of day. Each call attempt allowed a minimum of five rings. Once an eligible individual at a given household was identified, five attempts were made to contact that person to complete the survey. Any call-backs were scheduled on the day and time requested by the respondent. Proxy respondents were used, in particular, for participants less than years of age who were less likely to recall their dietary intake over the past days and for those between and years of age whose parent or guardian felt the child would not be able to answer the questions themselves. Proxy respondents were not accepted for individuals over the age of .The research protocol was approved by the Human Subjects Committee of the University of Guelph (Guelph, Ontario, Canada). Informed consent was obtained verbally from all participants and from a parent or guardian of all respondents under the age of years.Interviews were conducted by trained interviewers at the Centre for Evaluation of Medicines (St. Joseph's Hospital, Hamilton, Ontario, Canada) using a computer-assisted telephone interviewing (CATI) system. The project manager monitored the first few interviews performed by each interviewer to ensure standardization of the survey tool. An Interviewer Manual was also provided to assist the interviewers with frequently asked questions and key definitions.The survey was developed by modifying questions from existing questionnaires [,], with certain questions kept identical to facilitate future inter-survey comparisons. The survey covered topic areas: (i) number of meals and dining locations, (ii) fruits and vegetables, (iii) dairy and eggs, (iv) alternative protein sources, (v) meats and seafood, (vi) water, (vii) home hygiene and food safety practices and knowledge, (viii) health (gastrointestinal illness in both one week and four weeks prior to the interview), (ix) demographics (age, gender, income, education, ethnicity, place of residence, pet ownership, number of children in household and household size), (x) grocery shopping practices and purchase of retail meat, including beef, chicken and pork, (xi) convenience foods, and (xii) food preparation, including meal preparation time and food handling knowledge and practice. All questions on food consumption addressed the respondent's consumption in the seven days prior to the interview. The interviewer read a list of food items to participants, and respondents were also given the opportunity to name food items they had consumed that were not on the list. Where appropriate, questions covering grocery shopping practices were asked of persons who were most familiar with these practices in the household, who may or may not have been the selected respondent.The questionnaire was pre-tested in a small convenience sample of individuals before applying it to the study population in order to identify questions that were confusing, ambiguous, or misleading, and to estimate the time required to complete the survey (average of minutes). Data were analyzed using Intercooled Stata . for Windows (StataCorp LP, College Station, TX). Individuals who responded "don't know/not sure" or who refused to answer a question were excluded from analysis for that question. The Waterloo Region's population from the Canadian census was used to calculate expected population characteristics [].Prevalences were estimated for the consumption of all food items. Single food items that had less than ten responses were grouped into an 'Other' category for that specific food group. Exact confidence intervals (CI's) were computed for all proportions at the % level. These data were weighted to the Canadian Census by gender. Differences between gender and each food item consumed were tested using the Pearson's chi-square (χ2) test, and Fisher's exact test when the expected frequencies for a contingency table were less than five []. Age was categorized into six groups: ≤ years of age (children), between and years of age (adolescents), between and years of age (young adults), between and years of age (adults), between and years of age (older adults), and ≥ years of age (elderly). Pearson's χ2 test or Fisher's exact test were also used to determine any differences between the age categories and each food item consumed. Statistically significant differences were determined by a two-tailed probability of less than .. Odds ratios (OR's) were used to estimate the strength of association between food consumption patterns and demographic variables (adjusted for age and gender), where males and individuals years of age or older were chosen as the reference groups. Although a large number of food items were included in this survey, we did not adjust for multiple testing. As a result, estimates are likely to have an increased type I error rate and fewer significant differences may exist than reported here.ResultsStudy population characteristicsA total of , of the , contacted and eligible persons consented to be interviewed, yielding a response rate of .%. A total of ineligible individuals were unable to participate because they had traveled outside of Canada in the past seven days. Respondent characteristics are shown in Table . Compared to residents of the study area, respondents were older than the Census population, were more likely to be female, had a higher education level, and had a higher total annual household income.Table 1Demographic comparison showing percent of residents and survey respondents per category (except where noted) in the Waterloo Region, Ontario, Canada, November – March 2006Waterloo Region Residents (n = ,)Survey Respondents (n = ,)Sex Male49..* Female50..*Age (years) Mean---. Median35.. –..* –.. –.. –..* –..* +..*Cultural group† North American---. European---. African---. Mediterranean---. Asian---. Native North American/Aboriginal---. South American---. Austral-Asian---.04Education No high school diploma27.. High school diploma14.. Some college or university18..* College or trade diploma28..* University, graduate, or professional18..*Total household income <$,..* >$, to <$,..* >$, to <$,.. >$, to <$,.. >$, to <$,..* >$,..*Residence‡ City or urban area---. Suburban area---. Town or village---. Rural area---.3Mean household size2..9Median household size---.0Mean # children in household---.* Proportion of respondents significantly different than the Waterloo Region residents, P < .† Survey respondents reported with which cultural group they most identified. No comparable information was available from the Canadian Census.‡ No comparable information was available from the Canadian Census.Food consumptionThe total proportions of each food item consumed by the study population during the seven days prior to the interview are presented in additional file . Of the % of respondents consuming produce, approximately % indicated that some or all of the produce they consumed was organically grown. Of the % of respondents eating nuts, % indicated that some or all of the nuts they had eaten were consumed raw. Of the % of respondents consuming tofu, % were vegetarian. Differences in consumption of food items by gender (adjusted for age) and age (adjusted for gender) are presented in Table and in additional file , respectively, showing only those food items whose differences were statistically significant at P ≤ .. The proportion of respondents who reported that their food consumption patterns were typical of a normal week's food consumption are presented, by food category, in Table .Table 2Odds of a female respondent consuming a specific food item compared to a male respondent (adjusted for age), showing only those food items whose differences were statistically significant (P ≤ .) in the Waterloo Region, Ontario, Canada, November – March 2006Food ItemsOR95% CI†(A) Herbs & Sprouts Bean sprouts0.. < OR < .(B) Fruits Lemon6.. < OR < . Grapefruit2.. < OR < . Kiwi1.. < OR < . Raspberries1.. < OR < . Blueberries1.. < OR < . Pears1.. < OR < . Strawberries1.. < OR < . Grapes1.. < OR < . Unpasteurized Juice0.. < OR < .(C) Vegetables Beans2.. < OR < . Cauliflower1.. < OR < . Peppers1.. < OR < . Fresh celery1.. < OR < . Fresh carrots1.. < OR <. Fresh spinach1.. < OR < . Broccoli1.. < OR < . Lettuce1.. < OR < . Other bulb onion1.. < OR < . Cucumber1.. < OR < .46Deli Salads Pre-bagged, mixed salad greens1.. < OR < . Pasta salad1.. < OR < .(D) Dairy Yogurt1.. < OR < . White milk1.. < OR < . Sour cream1.. < OR < .(E) Cheese Gouda2.. < OR < . Cream cheese1.. < OR < . Cottage cheese1.. < OR < . Marble cheese1.. < OR < . Feta cheese1.. < OR < . Parmesan cheese1.. < OR < .(F) Nuts Almonds1.. < OR < . Pecan2.. < OR < .(G) Meats Ham4.. < OR < . Ground beef1.. < OR < . Steak0.. < OR < . Hamburgers0.. < OR < . Hamburgers not made in the home0.. < OR < . Hamburgers from pre-made uncooked patties0.. < OR < . Lamb0.. < OR < . Pink or undercooked pork*.. < OR < .84CI – confidence interval.*Among respondents that reported eating pork†Exact % confidence intervalsTable 3Percentage of respondents who noted that their reported food consumption patterns, as determined by this survey, were typical of a normal weeks' consumption, Waterloo Region, Ontario, Canada, November – March 2006Food GroupTypical of a Normal Week's Consumption(%)Fruits and vegetables95.19Dairy and egg products95.50Meats91.98Seafood80.28Household dietary habitsIn the seven days prior to the interview, respondents reported eating approximately . evening meals together with the majority of the household members, and spent on average a total of minutes preparing a typical meal within the home. The study population reported eating roughly . meals or snacks per day. Respondents consumed approximately . meals or snacks consisting of leftover food items in the past seven days. Among respondents who ate leftover foods, % consumed the food within one to two days of initial preparation, % within three to four days, and % beyond four days. The number of meals in the past seven days, by dining location, is presented in Table . During an average week, respondents shopped for groceries . times. On average, % of food items purchased were ready-to-eat foods, % of food items purchased were ready-to-cook foods, and % of food items were purchased as basic raw ingredients.Table 4Mean, minimum and maximum number of meals prepared in the past seven days, by dining location, as reported by respondents in the Waterloo Region, Ontario, Canada, November – March 2006Meal LocationMeanStd. Dev.Min.Max.Home25..79060Outside home3.. Fast-food chain restaurant0.. Sit-down restaurant0.. Cafeteria or restaurant with buffet0.. Catered event0.. Pizza parlour or coffee/donut shop or street vender1.. Market deli or salad bar or ready-to-eat food from supermarket0.. Other restaurant0..66018Around % of respondents followed a special dietary practice, of which approximately % followed special diets for various medical conditions, including those prescribed for diabetes. Vegetarianism was reported by % of respondents, while another % of respondents reported following special weight loss diets, and less than % reported following weight gain diets. Religious dietary practices or fasting was reported by less than % of respondents.DiscussionPrevious food consumption surveys conducted in Canada have primarily focused on dietary intake from a nutritional perspective, or have used methods that provide little information on actual food consumption patterns. This study describes the food consumption patterns in a Canadian-based population, from a food safety perspective, presenting baseline data on actual food intake of individuals. The results illustrate that certain food items were consumed more than others among the same food groups, and that consumption of many food items varied by gender and age. The findings also demonstrate high levels of at-home food preparation and consumption.Food itemsThe extensive range of food items reported by survey respondents illustrates the food preferences of the population, and corresponds to the diversity of the Canadian food supply, as evidenced by food disappearance data [] and the abundance of imported products, and foreign produce available throughout the year in grocery stores [].Within each food group, certain food items were consumed more than other items from the same food groups. The observed higher consumption prevalences of certain food items among the food groups are consistent with estimates of food availability within the Canadian food supply and consumer demands []. Comparing food availability data to the reported intake of food in populations is difficult since they represent different levels of dietary information [], however, potential relationships may be identified.Data depicting the amount of food available for consumption in Canada over the past two years [] reveal that bananas, apples, white milk, cheddar cheese, eggs, chicken, and beef were the most available food items among the food groups classified in this study (See additional file ). Lettuce, carrots, and tomatoes were also highly available among fresh vegetables; however, potatoes dominated this commodity []. Potato consumption was quite low among survey respondents, consumed by only .% of the population. Potatoes were not included in the list of food items asked of the respondents because consumption of raw produce was of interest in this study and potatoes are rarely eaten raw. The reported consumption of potatoes likely reflects the consumption of cooked potatoes as opposed to raw potatoes and is most likely underreported as it represents open-ended responses given by respondents. In fact, a higher proportion of respondents reported consuming potato salad (.%) when asked specifically about potato salad consumption. The high prevalence of yogurt consumption is consistent with its availability, which has increased substantially in Canada over the past decade [].GenderThe consumption of certain food items varied significantly by gender. Food items that were more likely to be consumed by males included meat products, unpasteurized juice and bean sprouts, whereas females were more likely to consume more fruits and vegetables, and dairy products. These consumption patterns are consistent with those found in a number of other dietary surveys [-]. Studies conducted in Great Britain also observed gender differences in food preferences, reporting that women consumed more fruits and vegetables compared to men, and had a higher preference for cultured dairy products such as cottage cheese and yogurt [,].The greater consumption of fruits and vegetables and yogurt by females compared to males, appear to be more healthful food choices, suggesting that females may choose to eat certain foods perceived to be healthy and avoid foods considered to be unhealthy (e.g. fat avoidance). The particularly high consumption of lemons by females (compared to males) (OR = ., % CI = ., .) could be the result of adding lemon to water, which may be viewed as a healthful practice or taste preference by females.Previous studies have shown that, in general, women engage in healthier eating patterns compared to men [] and are more motivated to respond to health messages focused on diet []. Motivation to practice healthier diets among women may reflect the greater importance of self image and physical appearance to women compared to men [,]. Such motivating factors have been associated with healthier diets [,] and higher intakes of fruits and vegetables [].Along with the notion of healthier eating patterns among women than men, this study also found that males are significantly more likely to consume food items considered high risk for the transmission of certain enteric pathogens, such as unpasteurized juice, bean sprouts, and undercooked meat. This is consistent with previous studies that show that women are more concerned than men with food safety issues [,]. These findings highlight an opportunity for health and food safety education targeted at males to help increase their awareness of the health risk involved in the consumption and preparation of certain food items. More detailed information on the specific high-risk foods consumed by males in the target population would be useful to help make effective public health campaigns.AgeSignificant differences in food consumption by age group were also observed in this study. In general, fruits, vegetables, eggs, cottage cheese, fish, and nuts were more likely to be consumed among elderly individuals than younger individuals, whereas children and adolescents were more likely to consume dairy products such as milk, ice cream, and cheese. Adults consumed intermediate amounts of most food items compared to the other age categories. However, adults had higher consumption of herbs, yogurt, cheese, tofu and steak. Previous studies confirm that food intake varies among age groups [,,].Specific differences found between age groups that corroborate our findings include greater consumption of milk and dairy products among children compared to older individuals [,,], and increased intake of fruits and vegetables associated with increasing age [,,,]. The finding that older respondents were less likely to consume milk, especially chocolate milk, compared to younger respondents was anticipated since the Canada Food Guide recommends children under years of age consume two to four servings of milk each day for sufficient calcium intake []. Perhaps the higher consumption of chocolate milk among children and adolescents suggests a preference for sweetened beverages, given that recent studies on beverage consumption among children suggest that changes in milk consumption may be the result of sugar-sweetened beverages displacing milk [,].Although soymilk and tofu consumption was generally low among respondents, elderly individuals showed the highest consumption of soymilk, and tofu consumption was highest among individuals between and years of age. While information on current intake of soy is well known among Asian populations [,], where soy foods are considered a staple part of the diet, information is relatively scarce among Western populations. A European study evaluating soy product consumption found that soy dairy substitutes were the most frequently consumed soy foods, although soy product consumption was relatively low.Significant differences were found between age categories and egg consumption. Elderly respondents were significantly more likely to consume eggs, including undercooked eggs, than other respondents. These results differ from other study findings [,], which reported that consuming undercooked eggs was less common among the elderly. The high proportion of elderly individuals consuming runny eggs implies that this age group may be unaware of the health risks associated with eating undercooked eggs. Raw or undercooked eggs have been identified as a major source of Salmonella Enteritidis and Typhimurium infections []. Thus, the higher consumption of undercooked eggs among older respondents is important from a public health perspective because salmonellosis is particularly severe for this age group [,]. Food safety messages could be targeted towards the elderly and include information on health risks associated with eating raw or undercooked eggs, while emphasizing the importance of cooking eggs well to prevent illness.The consumption of chicken nuggets, hamburgers not made in the home, and deli meats were highest among children, adolescents, and young adults compared to older individuals. This suggests that respondents in the younger age categories, or perhaps their parents, may be more interested in convenience food and food prepared outside of the home. Studies on food consumption trends show comparable findings among adolescents and young adults, where the consumption of cheeseburgers, pizza, ham, and salami has increased [,] representative of inexpensive food usually consumed outside of the home. These findings may also serve to explain the general low fruit and vegetable consumption among the younger age groups compared to the elderly age group. Kearney, Hulshof and Gibney [] reported that, with increasing age, there was a decrease in the proportion of eating any meals outside of the home, and younger individuals were more likely to have their lunchtime meal outside of the home. Thus, the higher consumption of deli meats among individuals younger than years may be related to meals eaten at school, where lunches may often consist of deli meat sandwiches.The results presented here highlight differences in food consumption likely due to taste and preference, but differences in dietary choice by age may also be attributed, in part, to cohort effects []. For example, the diet of older individuals may partially reflect their habits in earlier decades. In addition, a variety of health conditions can dictate changes in diet as people age. However, in this study, only a very small proportion (%) of respondents followed a special diet for various medical conditions, of which % were over the age of years.Household dietary patternsOn average, the number of meals reported as being prepared within the home (. meals) within the past seven days was much higher than those meals reported as being prepared outside of the home (. meals), at a restaurant location. In addition, the weekly average number of meals that consisted of leftover food items (. meals) was just slightly below that of the average number of restaurant meals. However, it is unknown whether the leftover meals had originally been prepared within the home. Nevertheless, these observations suggest that households in the study area still prepare the majority of their meals at home.The results presented here are comparable to those found in the Canadian Community Health Survey [] on eating habits, which reported that the majority of patrons of fast-food establishments selected pizza and sandwiches as their preferred meal choice. On average, about half of grocery foods purchased were basic raw ingredients, while half of purchased food items consisted of ready-to-eat and ready-to-cook foods. In addition, the average time spent on evening meal preparation in the home was minutes. These findings may be somewhat unexpected as current consumer trends reveal a decrease in the time spent in meal preparation [,] and increased demand, purchase, and consumption of ready prepared foods and fast-food meals in response to evolving lifestyle changes [,-]. Therefore, despite emerging trends in the food industry, it appears that many households still allocate time for meal preparation.Study LimitationsThis study had several limitations. This study focused on food consumption patterns from a food safety perspective, and consumption of raw food items was of particular interest. As a result, food consumption patterns reported by respondents may not be entirely reflective of their usual consumption patterns since some food items, such as vegetables that are usually consumed cooked (i.e. potatoes), were not included in the survey. Future surveys should incorporate questions on the consumption of both cooked and raw food items.Data on food consumption were subdivided on single demographic variables: gender and age. Demographic characteristics are seldom statistically independent of one another. For example, education and income have been shown to be important determinants in food intake [,,,]. In addition, since food choices are the result of a complex mixture of interacting social, cultural, and other environmental factors [,-], it is possible that the observed differences among gender and age may not be due to these factors alone. Nevertheless, assessing the overall effects of both gender and age provides insight into the food consumption practices in different gender and age groupings necessary to support outbreak investigations and risk assessments, and may prompt further analysis.Consumption of some food items is likely to vary according to season, often based on availability and price [,]. This study only collected data during one season (winter), precluding analysis of seasonal variation. It was shown, however, that the majority of reported diets consumed by respondents in the past seven days were representative of a typical weekly diet. It is possible that seasonal variation in diet may not be large within Canada because of the availability of an extensive variety of foods throughout the four seasons. However, a British Columbia study found that many foods were seasonally consumed with higher consumption of fruits in the summer than in the winter []. Future surveys should attempt to account for seasonality in the survey design by implementing the survey in different seasons or over a calendar year [].Response bias due to factors such as recall, the use of proxy respondents, and social desirability may have affected the results. In an attempt to reduce recall bias, a list of food items was used for the food consumption questions to prompt the respondents' memory. In addition, respondents were given the opportunity to provide answers that may have been omitted in the food list. Although a shorter recall period, such as hours, may have reduced recall bias compared to a seven-day recall, single-day intake data tend to over- or under-estimate the usual intake of an individual []. Multiple days of dietary intake data reduce the effects of day-to-day variability on estimates of usual food intakes [] and we chose to use this method here.Younger children, in particular, are less able to recall their dietary intake than adults [], and a surrogate respondent is often required to obtain such information. Although proxy respondents may not know exactly what a child ate when they were not together, studies comparing reports of subjects and their proxy respondents indicate that proxy information is useful despite possible under-reporting [,]. In this study we attempted to minimize proxy response bias by asking proxy respondents who were most acquainted with the child's daily activities to provide information on the child's diet.Finally, the accuracy of the recall method also depends on the honesty of the respondents. Studies have shown that social desirability, the tendency to overestimate desirable behaviours and underestimate undesirable ones [], biases self-reports of food and nutritional intake [-]. Barros, Moreira and Oliveira [] showed that respondents tended to over-report consumption of foods perceived to be healthy, such as vegetables, and under-report consumption of foods perceived to be unhealthy, such as white bread and beer. Methods to measure and control for socially desirable responses are needed to reduce potential bias in self-reported dietary intake.Although the overall response rate in this survey was low, it was consistent with response rates found in other dietary surveys in Canada [,]. A low response rate suggests the potential for selection bias, potentially limiting external validity of the study if non-respondents were different from survey respondents []. Since we did not collect information on non-respondents, the extent of potential bias could only be made by comparing the survey respondents to the total Waterloo Region Census population. The differences between the survey respondents and the study population were expected and were most likely the result of the sampling method used. In addition, the sampling frame used was a list of residential telephone numbers and the survey was administered in English only. However, of the , eligible respondents, only (%) did not participate due to language or hearing problems. Future studies may consider a sampling strategy to include potential respondents who do not speak English, who do not have telephones, as well as those with unlisted telephone numbers and mobile phone users.ConclusionThis study established baseline data on actual food intake of individuals in a Canadian based population from a food safety perspective. Differences in food consumption patterns were influenced by gender and age, although other important socio-demographic variables may also explain differences in dietary intake. These data demonstrate that certain gender and age groups may have unbalanced diets, which may be cause for concern from a nutritional perspective. Food safety education could be combined with strategies to promote healthier diets with appropriate nutritional guidance. This information is useful for food safety risk assessments and provides a baseline against which future outbreak investigations can compare proportions reported from case-series of infected individuals. Of concern is the consumption of specific foods among certain gender and age groups, suggesting that there may be a need not only to conduct food safety education among all consumers, but to target specific demographic groups that may be at a higher risk of foodborne disease. The majority of households reported preparing and consuming most meals at home. This suggests that consumer awareness of, and responsibility for, proper hygiene and food safety practices in the home is essential.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsSM was the project facilitator and secured project funding. SM, RF, AN participated in the study concept and design. AN was responsible for data analysis, data interpretation and manuscript preparation. All authors provided input into the survey tool, and provided expert advice on data interpretation and discussion of the paper. All authors have approved the final version of the manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:Supplementary MaterialAdditional file 1Total proportion of food items consumed by survey respondents.Click here for fileAdditional file 2Odds ratio of reported consumption of food items by age group.Click here for file
|
PMC2362426.txt
|
TITLE: Medroxyprogesterone acetate inhibits interleukin secretion from KPL- human breast cancer cells both in vitro and in vivo: a possible mechanism of the anticachectic effect AUTHORS: J Kurebayashi, S Yamamoto, T Otsuki, H Sonoo ABSTRACT: Interleukin (IL-) is a multifunctional cytokine. Recent reports suggest that circulating IL- secreted from tumour cells plays an important role in cancer-induced cachexia. Medroxyprogesterone acetate (MPA) has been used as an endocrine therapeutic agent for patients with breast cancer. It has been suggested that MPA decreases serum IL- levels and preserves the bodyweight of patients with advanced breast cancer. However, the mechanisms of action responsible for the anticachectic effect of MPA have not been elucidated. Therefore, the effects of MPA on IL- secretion were studied both in vitro and in vivo using a human breast cancer cell line, KPL-, which secretes IL- into medium and induces cachexia when injected into female nude mice. MPA (– nM) dose-dependently decreased basal IL- secretion into medium, and also suppressed tumour necrosis factor (TNF-α)-induced IL- secretion. Both basal and TNF-α-induced IL- mRNA levels were dose-dependently lowered by MPA. Moreover, intramuscular injections of MPA ( mg kg− twice a week) into nude mice bearing KPL- transplanted tumours significantly decreased serum IL- levels without affecting tumour growth and preserved the bodyweight of recipient mice. These findings suggest that suppression of IL- secretion from tumour cells, at least in part, causes the anticachectic effect of MPA. © Cancer Research Campaign BODY: No Body Content
|
PMC3044744.txt
|
TITLE: Characterizing Associations and SNP-Environment Interactions for GWAS-Identified Prostate Cancer Risk Markers—Results from BPC3 AUTHORS: Sara Lindstrom, Fredrick Schumacher, Afshan Siddiq, Ruth C. Travis, Daniele Campa, Sonja I. Berndt, W. Ryan Diver, Gianluca Severi, Naomi Allen, Gerald Andriole, Bas Bueno-de-Mesquita, Stephen J. Chanock, David Crawford, J. Michael Gaziano, Graham G. Giles, Edward Giovannucci, Carolyn Guo, Christopher A. Haiman, Richard B. Hayes, Jytte Halkjaer, David J. Hunter, Mattias Johansson, Rudolf Kaaks, Laurence N. Kolonel, Carmen Navarro, Elio Riboli, Carlotta Sacerdote, Meir Stampfer, Daniel O. Stram, Michael J. Thun, Dimitrios Trichopoulos, Jarmo Virtamo, Stephanie J. Weinstein, Meredith Yeager, Brian Henderson, Jing Ma, Loic Le Marchand, Demetrius Albanes, Peter Kraft ABSTRACT: Genome-wide association studies (GWAS) have identified multiple single nucleotide polymorphisms (SNPs) associated with prostate cancer risk. However, whether these associations can be consistently replicated, vary with disease aggressiveness (tumor stage and grade) and/or interact with non-genetic potential risk factors or other SNPs is unknown. We therefore genotyped SNPs from regions identified by several prostate cancer GWAS in , prostate cancer cases and , controls from the NCI Breast and Prostate Cancer Cohort Consortium (BPC3). We replicated out of SNPs (P-values ranging from . to −). Two SNPs located near KLK3 associated with PSA levels showed differential association with Gleason grade (rs2735839, P = . and rs266849, P = .; case-only test), where the alleles associated with decreasing PSA levels were inversely associated with low-grade (as defined by Gleason grade <) tumors but positively associated with high-grade tumors. No other SNP showed differential associations according to disease stage or grade. We observed no effect modification by SNP for association with age at diagnosis, family history of prostate cancer, diabetes, BMI, height, smoking or alcohol intake. Moreover, we found no evidence of pair-wise SNP-SNP interactions. While these SNPs represent new independent risk factors for prostate cancer, we saw little evidence for effect modification by other SNPs or by the environmental factors examined. BODY: IntroductionProstate cancer is the most common non-skin cancer among men in industrialized countries, but beyond age, ethnicity and family history, very little is known about its etiology. Observed familial aggregation together with evidence from both twin and epidemiological studies demonstrate a key role for inherited genetic variants [].Genome-wide association studies (GWAS) conducted within the last few years have identified multiple common single nucleotide polymorphisms (SNPs) associated with prostate cancer risk []–[]. However, the function of these SNPs (or the causal variants these SNPs serve as proxies for) remains largely unknown and data describing their correlation with clinical factors or their interplay with other genetic and non-genetic factors are sparse, mainly due to the large sample sizes needed for sufficient statistical power.To this end, we selected SNPs from regions identified in previous GWAS and genotyped them in , prostate cancer cases and , controls within the NCI Breast and Prostate Cancer Cohort Consortium (BPC3). We tested each SNP for association with two strongly predictive clinical factors: Gleason grade and tumor stage. We investigated interactions between SNPs and known and potential environmental risk factors. Finally, we performed exploratory analysis to identify possible pair-wise SNP-SNP interactions.ResultsAssociation between SNPs and prostate cancer riskSubject characteristics are displayed in Table . All SNPs were significantly associated with prostate cancer risk (Table ) and directions of associations were consistent with previous findings []–[], []–[], [], [], []. Although risk estimates varied somewhat between different cohorts (Table S1 and Figure S1), we observed overall no strong statistical evidence for heterogeneity (P>.). Risk effects per allele ranged from . (rs2928679) to . (rs16901979). Carriers of two copies of the rare ‘A’ allele of rs16901979 had a -fold increased risk to develop prostate cancer in this population. The allele frequency of rs16901979 varies widely across ethnicities (Hapmap population frequencies: . in CEU, . in CHB and . in YRI), and thus might explain a part of the differences seen in prostate cancer incidence across populations. Based on p-value, we observed the strongest association for rs4430796 located in HNF1B/TCF2 (OR: . (% CI: .–.), P = .•−) and the weakest association for rs4961199 near CPNE3 (OR: . (% CI: .–.), P = .). In addition, rs266849 near KLK3 was only weakly associated with prostate cancer risk (OR: . (% CI: .–.), P = .). rs266849 was initially identified in a GWAS using controls selected for low prostate-specific antigen (PSA) levels (<. ng/ml) [] and it has been suggested that rs266849 is a marker for circulating PSA levels rather than for prostate cancer risk [], []../journal.pone..t001Table 1Characteristics of the study populations.ATBCCPS-IIEPICHPFSMCCSMECPHSPLCOCasesControlsCasesControlsCasesControlsCasesControlsCasesControlsCasesControlsCasesControlsCasesControls Number of Individuals ,,,,,,,,,,,,, Mean age (sd) .(.).(.).(.).(.).(.).(.).(.).(.).(.).(.).(.).(.).(.).(.).(.).(.) ≤ years (%) ()()()()()()()()(),()()()()()()() > years (%) ()(),(),()()()()()()()()(),()()()() FH + (%) ()()()()NANA201()()()()()()NANA143()() Missing (%) ()() () (),(),() ()()()()()(),(),(!)()() BMI (%) <()()()()()()()()()()()()()()()()–()(),(),()()()()()()()()()()()()()>()()()()()()()()()()()()()()()() Missing (%) ()()()()()()()()()()()()()()()() Height (%) < cm ()()()()()()()()()()()()()()()() – cm ()()()()()()()()()()()()()()()() > cm ()(),()()()()()()()()()()()()()() Missing (%) ()()()()()()()()()()()()()()()() Smoking (%) Never0()() () () () ()()()()()()()()()()()Former0()() (), ()()()()()()()()()()()()()Current1,()()()()()()()()()()()()()()()() Missing (%) ()()()()()()()()()()()()()()()() Alcohol Consumption (%) < g/day816()(),(),(),(),(),(),()(),()()(),(),()()()≥ g/day176()()()()()()()()()()()()()()()() Missing (%) () () () ()()() ()()()()()()()()()() Diabetes (%) () () () () () () () () () () () () () () () () Missing (%) ()()()()()()()()()()()()()()()() High Grade (%) ()NA323()NA75()NA114()NA148()NA206()NA197()NA126()NA Missing (%) (!)NA326($)NA780()NA190()NA27()NA28()NA38()NA2()NA High Stage (%) ()NA409()NA257()NA139()NA93()NA85()NA231()NA245()NA Missing (%) ()NA65()NA506()NA197()NA34()NA15()NA572()NA0()NA1Age is calculated at age of diagnosis/selection as control except for MCCS (at baseline for controls) and MEC (at blood draw for controls).2Defined as Gleason – or WHO grade III.3Defined as Stage CD../journal.pone..t002Table 2Associations between selected SNPs and prostate cancer risk.SNPPosChrGeneN CasesN ControlsMAF CasesMAF ControlsOR HetOR HomOR alleleP-valueRefrs72104862,, EHBP1 ... (.–.). (.–.). (.–.).47E-051rs146561843,, THADA ... (.–.). (.–.). (.–.).02E-0512rs12621278173,, ITGA6 ... (.–.). (.–.). (.–.).001012rs266075387,, – ... (.–.). (.–.). (.–.).18E-052rs4857841129,, EEFSEC ... (.–.). (.–.). (.–.).32E-083rs1702191895,, PDLIM5 ... (.–.). (.–.). (.–.).35E-0512rs1250042695,, PDLIM5 ... (.–.). (.–.). (.–.).002112rs7679673106,, TET2 ... (.–.). (.–.). (.–.).74E-1012rs9364554160,, SLC22A3 ... (.–.). (.–.). (.–.).000862rs1048656727,, JAZF1 ... (.–.). (.–.). (.–.).05E-144rs646565797,, LMTK2 ... (.–.). (.–.). (.–.).52E-072rs151226823,, NKX3– ... (.–.). (.–.). (.–.).52E-0712rs292867923,, SLC25A37 ... (.–.). (.–.). (.–.).003612rs496119987,, CPNE3,CNGB3 ... (.–.). (.–.). (.–.).0124rs1016343128,, – ... (.–.). (.–.). (.–.).54E-192rs7841060128,, – ... (.–.). (.–.). (.–.).43E-183rs16901979128,, – ... (.–.). (.–.). (.–.).55E-115rs620861128,, – ... (.–.). (.–.). (.–.).59E-113rs6983267128,, – ... (.–.). (.–.). (.–.).62E-264rs1447295128,, – ... (.–.). (.–.). (.–.).06E-256rs4242382128,, – ... (.–.). (.–.). (.–.).17E-284rs7837688128,, – ... (.–.). (.–.). (.–.).70E-227rs16902094128,,... (.–.). (.–.). (.–.).78E-0813rs1571801123,, – ... (.–.). (.–.). (.–.).00498rs1099399451,, MSMB ... (.–.). (.–.). (.–.).95E-254rs4962416126,, CTBP2 ... (.–.). (.–.). (.–.).92E-054rs7127900219015011979396940... (.–.). (.–.). (.–.).01E-0712rs1241845168,, – ... (.–.). (.–.). (.–.).42E-079rs793134268,, – ... (.–.). (.–.). (.–.).11E-162rs1089644968,, – ... (.–.). (.–.). (.–.).58E-194rs1164974333,, TCF2 ... (.–.). (.–.). (.–.).55E-0810rs443079633,, TCF2 ... (.–.). (.–.). (.–.).09E-2811rs750193933,, TCF2 ... (.–.). (.–.). (.–.).78E-202rs185996266,, – ... (.–.). (.–.). (.–.).74E-1811rs26684956,, KLK3 ... (.–.). (.–.). (.–.).00852rs273583956,, KLK3 ... (.–.). (.–.). (.–.).05E-062rs575916741,, BIK ... (.–.). (.–.). (.–.).30E-1212rs594557251,,423X NUDT11 ... (.–.).17E-111rs594561951,,412X NUDT11 ... (.–.).27E-.Gudmundsson, et al. ..Eeles, et al. ..Yeager, et al. ..Thomas, et al. ..Gudmundsson, et al. ..Amundadottir, et al. ..Yeager, et al. ..Duggan, et al. ..Zheng, et al. ..Sun, et al. ..Gudmundsson, et al. .. Eeles et al. .Gudmundsson et al 2009The primary analysis in most GWAS assumes an additive increase in risk for each risk allele carried. rs4961199 (P = .) was the only SNP showing nominally significant evidence of departure from additivity in our data. This was not unexpected since rs4961199 was initially identified using a recessive inheritance model [].Replication in non-CGEMS cohortsEleven of the SNPs included in this paper were identified through the Cancer Genetic Markers of Susceptibility (CGEMS) project (http://cgems.cancer.gov/) which has partial overlap with BPC3. Eight of these eleven SNPs were identified through CGEMS stage including ATBC, CPS-II, HPFS and PLCO. We attempted to replicate these SNPs in the other cohorts (EPIC, MCCS, MEC and PHS) which collectively comprise , prostate cancer cases and , controls. Out of these eight SNPs, six replicated (Table S2) and two did not (rs4961199 near CPNE3 (OR: . (% CI: .–.), P = .) and rs4962416 in CTBP2 (OR: . (% CI: .–.), P = .)). Three SNPs (rs4857841 (EFFSEC), rs7841060 (8q24) and rs620861 (8q24)) were identified in CGEMS stage (that in addition to ATBC, CPS-II, HPFS and PLCO included EPIC and MEC). Therefore, we tested these three SNPs in MCCS and PHS only (, cases and , controls). Both rs7841060 (OR: . (% CI: .–.), P = .•−) and rs620861 (OR: . (% CI: .–.), P = .) were associated with prostate cancer risk whereas rs4857841 was not (OR: . (% CI: .–.), P = .). Since we did not replicate rs4857841, rs4961199 and rs4962416 in the non-CGEMS studies, we did not pursue these SNPs in further analysisSNP associations by tumor stage and gradeWe next examined whether any SNP was differentially associated with tumor grade or stage at diagnosis (Table ). A total of , cases were classified as high-stage (stage C or D at diagnosis) and , were classified as high-grade (Gleason grade – or equivalent, i.e. coded as poorly differentiated or undifferentiated). For % of the cases, we did not have information about tumor stage or Gleason grade. The minor alleles of two SNPs in the KLK3 gene (rs266849 and rs2735839), which have been previously associated with decreasing PSA levels [], [], were inversely associated with low-grade disease (Gleason <) (OR: . (.–.) for rs266489, OR: . (.–.) for rs2735839) but associated with increased risk for high-grade disease (OR: . (.–.) for rs266489, OR: . (.–.) for rs2735839). The differences in SNP associations between low-grade and high-grade disease were statistically significant in case-only analysis (P = . for rs266849 and P = . for rs2735839). These results remained significant after Bonferroni correction (P = . for rs266849 and P = . for rs2735839). No other SNP was differentially associated with tumor grade or stage after adjusting for multiple testing../journal.pone..t003Table 3Associations between selected SNPs and Gleason grade and Stage.Odds Ratio1 (% CI)Odds Ratio1 (% CI)SNPGleason <8Gleason –10P case-only testStage ABStageCDP case-only testrs7210481. (.–.). (.–.).. (.–.). (.–.).5rs14656181. (.–.). (.–.).. (.–.). (.–.).34rs126212780. (.–.). (.–.).. (.–.). (.–.).86rs26607531. (.–.). (.–.).. (.–.). (.–.).82rs48578411. (.–.). (.–.).. (.–.). (.–.).53rs170219180. (.–.). (.–.).. (.–.). (.–.).004rs125004261. (.–.). (.–.).. (.–.). (.–.).007rs76796730. (.–.). (.–.).. (.–.). (.–.).49rs93645541. (.–.). (.–.).. (.–.). (.–.).30rs104865670. (.–.). (.–.).. (.–.). (.–.).02rs64656571. (.–.). (.–.).. (.–.). (.–.).02rs15122681. (.–.). (.–.).. (.–.). (.–.).05rs29286791. (.–.). (.–.).. (.–.). (.–.).78rs49611991. (.–.). (.–.).. (.–.). (.–.).70rs10163431. (.–.). (.–.).. (.–.). (.–.).92rs78410601. (.–.). (.–.).. (.–.). (.–.).82rs169019791. (.–.). (.–.).. (.–.). (.–.).33rs6208610. (.–.). (.–.).. (.–.). (.–.).37rs69832670. (.–.). (.–.).. (.–.). (.–.).54rs14472951. (.–.). (.–.).. (.–.). (.–.).06rs42423821. (.–.). (.–.).. (.–.). (.–.).24rs78376881. (.–.). (.–.).. (.–.). (.–.).09rs169020941. (.–.). (.–.).. (.–.). (.–.).74rs15718011. (.–.). (.–.).. (.–.). (.–.).62rs109939941. (.–.). (.–.).. (.–.). (.–.).29rs49624161. (.–.). (.–.).. (.–.). (.–.).88rs71279001. (.–.). (.–.).. (.–.). (.–.).81rs124184511. (.–.). (.–.).. (.–.). (.–.).14rs79313420. (.–.). (.–.).. (.–.). (.–.).33rs108964490. (.–.). (.–.).. (.–.). (.–.).34rs116497430. (.–.). (.–.).. (.–.). (.–.).21rs44307960. (.–.). (.–.).. (.–.). (.–.).47rs75019390. (.–.). (.–.).. (.–.). (.–.).28rs18599621. (.–.). (.–.).. (.–.). (.–.).58rs2668490. (.–.). (.–.).. (.–.). (.–.).18rs27358390. (.–.). (.–.).. (.–.). (.–.).17rs57591670. (.–.). (.–.).. (.–.). (.–.).11rs59455721. (.–.). (.–.).. (.–.). (.–.).57rs59456191. (.–.). (.–.).. (.–.). (.–.).161Odds ratios were estimated using multinomial regression.Association between non-genetic factors and prostate cancer riskWe tested for association between prostate cancer risk and potential non-genetic risk factors including family history of prostate cancer, diabetes, BMI, height, smoking and alcohol consumption. As expected, we observed a strong association between family history of prostate cancer and prostate cancer risk (OR: ., % CI: .–., P = .•−) as well as between diabetes and prostate cancer risk (OR: ., % CI: .–., P = .•−). Adjusting for BMI did not alter the association between diabetes and prostate cancer (data not shown). BMI was inversely associated with prostate cancer risk (OR: . (% CI: .–.) per BMI unit increase, P = .). This association was limited to obese men (BMI >) compared to normal weight men (BMI<) (OR: ., % CI: .–., P = .), and we observed no association for being overweight (OR: ., % CI: .–., P = .). Adjusting for diabetes and smoking attenuated the association between obesity and prostate cancer risk (OR: ., % CI: .–., P = .). The inverse association between BMI and prostate cancer risk was restricted to non-aggressive cases as defined by Gleason grade < and tumor stages A and B (data not shown). Height was not associated with prostate cancer risk, when analyzed as a continuous variable (OR: ., % CI: .–. per cm increase, P = .) or in tertiles (OR: ., % CI: .–., P = .). We observed a non-significant reduced prostate cancer risk among both former smokers (OR: ., % CI: .–., P = .) and current smokers (OR: ., % CI: .–., P = .) compared to never smokers. Adjusting for alcohol consumption or BMI did not change the results (data not shown). Finally, consuming more than g alcohol per day (corresponding to two drinks) was associated with an increased prostate cancer risk (OR: ., % CI: .–., P = .). Adjusting for smoking did not alter this association (data not shown).SNP-environment and SNP-SNP interactionsTo investigate if the associations with family history of prostate cancer, diabetes and BMI were stronger in specific genetic strata, we tested for effect modification by including a SNPxE interaction term in the model. We also tested for SNP effect modification of age at diagnosis (studying the main effect of age is not appropriate since our population comprises of a series of nested case-control studies matched on age). After adjusting for multiple testing, no SNP showed significant statistical interaction with any of the non-genetic factors tested (Table S3 and Table S4). Of note, two SNPs in the 8q24 region (rs620861, P = . and rs6983267, P = .) showed nominally significant interactions with age at diagnosis, with the association being stronger in younger men. These results are in line with previous reports of stronger associations with earlier onset of disease for SNPs in the 8q24 region [], [], []. We observed marginally significant interactions between diabetes and rs10486567 in JAZF1 (P = .) and between BMI and rs10486567 (P = .). This is of particular interest since genetic variation in JAFF1 has been associated with diabetes, albeit not the same genetic variants. In this study, obesity was associated with a reduced risk for prostate cancer. It has been shown that BMI is inversely associated with PSA levels [] and thus, obese men are less likely to get diagnosed through PSA screening. Because BMI was associated with non-aggressive disease, we also looked at possible SNP-BMI interactions stratified by disease aggressiveness but observed no significant interactions (data not shown).To assess if the ambiguous associations between prostate cancer risk and height, smoking and alcohol consumption are due to hidden SNP-environment interactions, we conducted a joint test of the environmental main effect and the SNP-environment interaction effect. This test has proven powerful when the non-genetic effect is limited to a specific genetic stratum []. Across SNPs, the joint test was not significant for either alcohol or smoking after adjustment for multiple testing (Table S5, Table S6 and Table S7). Similarly, standard interaction tests between SNPs and height, SNPs and smoking and SNPs and alcohol consumption were not significant. Exploratory analyses of all possible pair-wise SNP-SNP interactions revealed no excess in significant interactions than expected by chance ( out of tests, Table S8). Furthermore, no SNP-SNP interaction was significant after correcting for multiple testing using a Bonferroni correction (lowest nominal P-value was .). Yeager and colleagues identified a SNP-SNP interaction between rs4242382 and rs620861 (P = .) []. We also observe this interaction (P = .), but not when the analysis was restricted to only MCCS and PHS (P = .).DiscussionIn this study, we set out to examine whether SNPs identified in GWAS to be associated with prostate cancer show variation in risk by disease aggressiveness (tumor stage and grade) and/or interact with non-genetic and genetic factors. All SNPs tested were significantly associated with prostate cancer in the overall analysis. However, the CGEMS project, which included four and six BPC3 studies in its second and third stage stages, respectively, contributed to identification of eleven SNPs investigated in the present study. We tested whether associations for these eleven SNPs could be confirmed in the remaining studies and with the exception of three SNPs, the findings were replicated with risk magnitudes similar to those in the CGEMS analysis. We could not replicate rs4961149 using data from three of the non-CGEMS cohorts. Since rs4961199 was included in CGEMS stage based on its recessive association, we also tested the recessive model in the non-CGEMS studies and observed a non-significant association similar but weaker as compared with CGEMS (OR: ., % CI: .–., P = .).Few of the observed associations differed by disease stage, tumor grade or environmental exposures. The most noteworthy finding was the qualitatively altered association according to Gleason grade for two SNPs near KLK3 (rs266849 and rs2735839), where the minor alleles were associated with lower risk of low-grade disease but higher risk of Gleason – tumors. This was previously observed by Kader and colleagues [] who studied , patients and found a strong association between Gleason grade and rs2735839 (P = .•−). The minor alleles of these SNPs have been associated with lower PSA levels indicating that carriers are less likely to be diagnosed at an early stage through PSA screening [], []. However, we did not observe any difference in the association of these two SNPs by disease stage, suggesting that delayed diagnosis might not fully explain these associations. Interestingly, the significant positive association of these two SNPs with Gleason – tumors support the clinical observations that PSA expression is lower in malignant than in normal prostatic epithelium and is further reduced in poorly differentiated tumors [], []. Together, these results suggest that KLK variation might influence high-grade prostate cancer risk through a yet unidentified pathway or simply as a genetic marker of the probability of a diagnosis of high versus low-grade prostate cancer diagnosis through its influence on PSA levels. To test this hypothesis, we performed case-only analysis based on year of diagnosis to reflect the introduction of wide-spread PSA screening (up to ( men) vs. after ( men)). If the association between Gleason grade and KLK3 variation is due to altered PSA levels, we would expect to see differential associations according to year of diagnosis. We did not observe such differences, however, suggesting that the KLK3-prostate cancer association is not mediated by altered PSA levels. A recent Icelandic study conducted stratified analysis based on year of diagnosis and noticed that the association with prostate cancer was confined to the group of cases diagnosed in or later. These results suggest that the association between the KLK3 locus and prostate cancer is driven by the increasing frequency of PSA testing [].After adjusting for multiple testing, no other SNP was associated with clinical sub-types. Earlier studies had failed to link these SNPs to clinical characteristics [], [], suggesting that these SNPs affect prostate cancer risk overall and not solely for more (or less) aggressive or advanced cancer.We found overall no evidence that these SNPs interact with known or proposed risk factors for prostate cancer including family history of prostate cancer risk, age of onset, diabetes, BMI, height, smoking or alcohol consumption. Studying the interactions between SNPs and diabetes was of particular interest since genetic variation in JAZF1 and TCF2 has been associated with both prostate cancer and diabetes [], [], [], [], []. We did see a borderline statistically significant interaction between rs10486567 in JAZF1 and diabetes, but this particular SNP has not been associated with diabetes risk. A previous study conducted in CPS-II and PLCO found that diabetes did not mediate the association between JAZF1 and HNF1B/TCF2 SNPs and prostate cancer risk [], and we observed no statistical interaction between diabetes and three SNPs in HNFIB/TCF2.We observed no significant associations between prostate cancer risk and smoking or height and only a weak association between prostate cancer and alcohol consumption, even after accounting for the possibility of differences in the effects of these exposures by genotype. A meta-analysis of studies observed that height was positively associated with risk (RR . per cm increment, % CI .–.) but the association was only seen in cohort studies []. A recent large meta-analysis of smoking and prostate cancer incidence found overall no evidence of an association but reported an increased risk when considering number of cigarettes smoked. Moreover, they observed a % risk increase for former smokers []. We did observe a marginal association between alcohol intake and prostate cancer risk. This is in line with earlier results indicating a weak risk increase for men consuming at least grams alcohol per day (OR: . (% CI: .–.)) and for men consuming at least grams per day (OR: . (% CI: .–.)) [].Overall, these results imply that the lack of robust associations between these environmental factors and prostate cancer risk is not due to interactions between these exposures and variation in any of the SNPs assessed in this study. However, the lack of significant interactions does not rule out that gene-environment interactions exist in prostate cancer. All SNPs under study have been linked to prostate cancer through their main effects. Agnostic approaches such as incorporating gene-environment interactions in a genome-wide association study setting might identify genetic variants that only affect risk when acting with other factors. The lack of significant interactions can also reflect the low power to detect only modest interaction effects despite our sample size of , cases and , controls. It is important to note that our results do not rule out small departures from a multiplicative odds model for the joint effect of pairs of individual markers and risk factors, nor does absence of departure from a multiplicative odds model necessarily imply that these genetic loci and risk factors do not interact in some causal manner. Moreover, absence of interaction as defined here does not imply absence of a “public health interaction”, where the benefit from reducing a risk factor in terms of absolute risk reduction differs across genotypes [].This is, to our knowledge, the first large-scale study to explore possible interactions between confirmed prostate cancer susceptibility markers and a broad spectrum of known and possible environmental factors. The SNPs considered in this study show marginal per-allele odds ratios ranging between . and .. It is possible that these odds ratios might be larger in strata defined by other prostate cancer risk factors, not evaluated in this study. It is well recognized that exploring such interactions requires large study populations with well-defined exposure data. With , prostate cancer cases, , controls and prospectively collected data within established cohorts, BPC3 is in a unique position to explore both gene-gene and gene-environment interactions as demonstrated here. For example, in the absence of main effects (which is not the same as assuming no marginal effect and plausibly consistent with modest marginal genetic or environmental effects), the BPC3 has % power to detect an interaction effect of . assuming an allele frequency of % and an environmental exposure with a prevalence of %.As with all studies utilizing environmental exposure data, the present investigation would be expected to have some degree of misclassification in the measurement of those factors. It is possible that alternative modeling of the environmental risk factors or more precise exposure quantification would increase statistical power (e.g. analyzing intensity, duration or pack-years of smoking rather than as never/former/current). However, a critical issue in conducting pooled analysis across studies is to harmonize data. As exposure data gets more refined, there is an increasing risk of discrepancies between cohorts which increases the risk of “misclassification”. Since our study cohorts (MEC exempted) included predominantly men of European ancestry, we were limited in our ability to study other ethnicities.Genome-wide association studies have been particularly successful for prostate cancer. Recently published secondary analysis of GWAS has now added ∼ additional prostate cancer SNPs to those presented here [], [], []. At time of this study, we did not have genotype data for these SNPs in BPC3 and it remains to be seen if they are differentially associated with clinical subtypes or if they interact with non-genetic factors.In summary, we independently replicated the association between prostate cancer risk and SNPs identified in multi-stage genome-wide association studies of prostate cancer. Except for SNPs in KLK3 that were differentially associated with Gleason grade, we did not detect any differentiation in SNP associations according to Gleason grade or stage at diagnosis, two clinical factors strongly predictive of disease outcome. Moreover, we found no strong evidence that these SNPs interact with age, family history, diabetes, BMI, height, smoking or alcohol consumption.Materials and MethodsStudy PopulationThe BPC3 has been described in detail elsewhere []. In brief, the consortium combines resources from seven well-established cohort studies with blood samples collected as follows: the Alpha-Tocopherol, Beta-Carotene Cancer Prevention (ATBC) Study in - [], American Cancer Society Cancer Prevention Study II (CPS-II) in [], the European Prospective Investigation into Cancer and Nutrition Cohort (EPIC – comprised of cohorts from Denmark, Great Britain, Germany, Greece, Italy, the Netherlands, Spain, and Sweden) in [], the Health Professionals Follow-up Study (HPFS) in [], the Multi-Ethnic Cohort (MEC) in [], the Physicians' Health Study (PHS) in [], and the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial in – []. In addition, the Melbourne Collaborative Cohort Study (MCCS) established in – [] recently joined the consortium. Together, these eight cohorts collectively include over , men who provided a blood sample.Prostate cancer cases were identified through population-based cancer registries or self-reports confirmed by medical records, including pathology reports. Except for the MCCS study, the BPC3 consists of a series of matched nested case-control studies within each cohort; controls were matched to cases on a number of potential confounding factors, such as age, ethnicity, and region of recruitment, depending on the cohort. MCCS used a case-cohort design, with a randomly sampled sub-cohort serving as controls. Written informed consent was obtained from all subjects and each study was approved by the Institutional Review Boards at their respective institutions. The IRBs for each study were as follows: US National Cancer Institute and National Institute for Health and Welfare (Helsinki, Finland) (ATBC), The Emory University School of Medicine Institutional Review Board (CPS-II), Ethikkommission - Medizinische Fakultät Heidelberg and Imperial College Research Ethics Committee (EPIC), The Institutional Review Board of Harvard School of Public Health (HPFS), The Cancer Council Victoria Human Research Ethics Committee (MCCS), The Institutional Review Board at the University of Southern California and the Institutional Review Board at the University of Hawaii (MEC), The Human Subjects Committee at Brigham and Women's Hospital (PHS) and NCI Special Studies Institutional Review Board (PLCO).The current study was restricted to individuals who self-reported as being Caucasian. We had genotype data for a total of , prostate cancer cases and , controls. Data on disease stage and grade at time of diagnosis were collected from each cohort, wherever possible. A total of , cases were classified as high-stage (stage C or D at diagnosis) and , were classified as high-grade (Gleason grade > or equivalent, i.e. coded as poorly differentiated or undifferentiated). For % of the cases, we did not have information about tumor stage or Gleason grade.Baseline information of height and body weight, family history of prostate cancer, cigarette smoking status (never, past, and current), alcohol intake (g/day) and information about a pre-existing diabetes diagnosis were collected by self-report. Family history, which was defined as having at least one first-degree family member diagnosed with prostate cancer, was available for all but two cohort studies (PHS and EPIC). For some countries in EPIC, weight and height was measured.Collection and harmonization of non-genetic dataWe collected data on family history, diabetes at baseline, smoking, alcohol consumption, height and BMI for each study. Family history of prostate cancer was dichotomized into “yes” (, subjects) or “no” (, subjects). Age was calculated at age of diagnosis/selection as control except for MCCS (at baseline for controls) and MEC (at blood draw for controls) and further dichotomized into younger or equal to years old or older than years. BMI was calculated based on baseline weight (kg) and height (m) categorized into categories: normal weight (BMI< kg/m2, , subjects), overweight (BMI – kg/m2, , subjects) and obese (BMI> kg/m2, , subjects). Height was analyzed both as a continuous variable and in tertiles (< cm (, subjects), – cm (, subjects) and > cm (, subjects).Smoking was categorized into categories: never (, subjects), former (, subjects) and current (, subjects). Alcohol was dichotomized into never and moderate drinkers (< g/day or two drinks per day; , subjects) or heavy drinkers (≥ g/day or drinks per day; , subjects). Pre-existing diabetes was dichotomized into “yes” ( subjects) or “no” (, subjects).We agreed on a common protocol prior to data collection based on data availability in the studies. Each study was responsible for sending the data in a format as described in the protocol to facilitate data harmonization. We agreed on collecting as detailed information as possible without having to exclude any study due to lack of covariate information (that is, we aimed for the least common denominator for the variables of interest). Inconsistencies or clarifications were handled by a dialogue between the data coordinating center and the individual studies. All studies have published analysis on these variables earlier and details on quality checks can be found in study-specific publications. All statistical analyses were conducted centrally.SNP selection and genotypingWe selected SNPs based on the literature for prostate cancer GWAS (Table ). These include (genomic location in parenthesis): rs721048 (2p15), rs1465618 (2p21), rs12621278 (2q31.), rs2660753 (3q12.), rs4857841 (3q21.), rs17021918 (4q22.), rs12500426 (4q22.), rs7679673 (4q24), rs9364554 (6q25.), rs10486567 (7p15.), rs6465657 (7p21.), rs1512268 (8p21.), rs2928679 (8p21.), rs4961199 (8q21.), rs1016343 (8q24.), rs7841060 (8q24.), rs16901979 (8q24.), rs620861 (8q24.), rs6983267 (8q24.), rs1447295 (8q24.), rs4242382 (8q24.), rs7837688 (8q24.), rs16902094 (8q24.), rs1571801 (9p33.), rs10993994 (10q11.), rs4962416 (10q26.), rs7127900 (11p15.), rs12418451 (11q13.), rs7931342 (11q13.), rs10896449 (11q13.), rs11649743 (17q12), rs4430796 (17q12), rs7501939 (17q12), rs1859962 (17q24.), rs266849 (19p13.), rs2735839 (19p13.), rs5759167 (22q13.), rs5945572 (Xp11.) and rs5945619 (Xp11.). For rs12418451, we used genotypes from either rs12418451 or rs10896438 (r2 = . in HapMap CEU population) and for and rs2928679 we used genotypes from either and rs2928679 or rs13264338 (r2 = . in HapMap CEU population). We did not have genotype data on rs4961199, rs16901979 and rs16902094 for MCCS.Genotyping was performed using the TaqMan assay (Applied Biosystems, Foster City, CA) in five different genotyping laboratories: Core Genotyping Facility at National Cancer Institute, Harvard School of Public Health, University of South California, DKFZ and UK Cancer Research. Blinded duplicated samples indicated no genotyping error. For each autosomal SNP, we tested HWE in the controls in each study separately. All autosomal SNPs were in HWE (P>.).Statistical methodsWe tested the association between prostate cancer risk and each SNP with a likelihood ratio test based on unconditional logistic regression. We adjusted all analyses for study and age at diagnosis or selection as a control in five year intervals using indicator variables. All odds ratios are calculated per copy of the minor alleles (,,) carried. For each SNP, we used Cochran's Q statistic to test for heterogeneity between studies.To estimate odds ratios for high-grade or low-grade disease, we performed multinomial regression with an outcome variable coded as (control), (low-grade) or (high-grade). To test for differential SNP associations between low-grade and high-grade disease, we used a likelihood ratio test based on case-only analysis. We repeated this analysis for high-stage/low-stage disease.We tested for interaction between SNPs and non-genetic factors by conducting a one degree-of-freedom likelihood ratio test of a single interaction term (SNPxE) as implemented in an unconditional logistic regression. When an environmental factor had more than two categories (as is the case for smoking, BMI and height), we used ordinal coding for the interaction term. To explore whether associations with proposed environmental risk factors may have been masked by effect heterogeneity, we performed a joint ( d.f.) test of the environmental main effect and the interaction effect. This test has been shown to outperform the standard marginal test when the environmental effect is restricted to a genetic stratum []. Cohorts with no variability in exposure (such as ATBC and smoking) were excluded from gene-environment interaction analyses. We tested for pair-wise SNP-SNP interactions using a one degree-of-freedom likelihood ratio test of a single interaction term as described for the SNP-environment interaction tests.We tested for dominance deviation from an additive model by including an additional SNP covariate coded as (,,) for (homozygote common allele, heterozygote, homozygote rare allele) respectively. Based on unconditional regression, we performed a one degree-of-freedom likelihood ratio test where the full model was tested against a model only including the SNP covariate with additive coding (,,) as described above. All reported P values are two-sided and uncorrected for multiple hypothesis testing. Analyses were conducted in R [] and SAS version ..Supporting InformationFigure S1Study-specific SNP associations with prostate cancer risk. For rs4961199, rs16901979 and rs16902094 we did not have genotype data from MCCS.(DOC)Click here for additional data file.Table S1Heterogeneity in main effects between studies for selected SNPs.(DOC)Click here for additional data file.Table S2Associations between SNPs identified through CGEMS and prostate cancer risk stratified on CGEMS membership.(DOC)Click here for additional data file.Table S3SNP-Environment interactions for family history of prostate cancer and age at diagnosis.(DOC)Click here for additional data file.Table S4SNP-Environment interactions with diabetes and BMI.(DOC)Click here for additional data file.Table S5Association between height and prostate cancer risk stratified by SNP genotypes.(DOC)Click here for additional data file.Table S6Association between smoking and prostate cancer risk stratified by SNP genotypes.(DOC)Click here for additional data file.Table S7Association between alcohol consumption and prostate cancer risk stratified by SNP genotypes.(DOC)Click here for additional data file.Table S8Pair-wise SNP-SNP interactions that reached a p-value < ..(DOC)Click here for additional data file.
|
PMC2364995.txt
|
TITLE: Complexes of Zinc With Picolinic and Aspartic Acids Inactivate Free Varicella-Zoster Virions AUTHORS: S. Shishkov, T. Varadinova, P. Bontchev, C. Nachev, E. Michailova ABSTRACT: Zn(II) picolinate and aspartate, Zn(pic) and Zn(asp), have been shown to inhibit key steps of the replication of HSV-. In the present study we describe the effect of Zn(pic) and Zn(asp) on the replication of VZV and on the infectivity of free virions. The experiments are done using BHK- cells, a clinical isolate of VZV and Zn-complexes in concentration of μM. When Zn-complexes are present during the whole period of infection, the yield of infectious virus progeny decreases up to %. The infectivity of VZV is completely restored after the removal of zinc. The virucidal effect is manifested at the 2nd h of contact, when % of the virions are inactivated. The results show that both Zn(pic) and Zn(asp) specifically inactivate free VZV virions with no effect on viral replication. BODY: No Body Content
|
PMC1794503.txt
|
TITLE: Hepatocyte growth factor ameliorates dermal sclerosis in the tight-skin mouse model of scleroderma AUTHORS: Tsuyoshi Iwasaki, Takehito Imado, Sachie Kitano, Hajime Sano ABSTRACT: The tight-skin (TSK/+) mouse, a genetic model of systemic sclerosis (SSc), develops cutaneous fibrosis and defects in pulmonary architecture. Because hepatocyte growth factor (HGF) is an important mitogen and morphogen that contributes to the repair process after tissue injury, we investigated the role of HGF in cutaneous fibrosis and pulmonary architecture defects in SSc using TSK/+ mice. TSK/+ mice were injected in the gluteal muscle with either hemagglutinating virus of Japan (HVJ) liposomes containing μg of a human HGF expression vector (HGF-HVJ liposomes) or a mock vector (untreated control). Gene transfer was repeated once weekly for weeks. The effects of HGF gene transfection on the histopathology and expression of tumor growth factor (TGF)-β and IL- mRNA in TSK/+ mice were examined. The effect of recombinant HGF on IL- production by TSK/+ CD4+ T cells stimulated by allogeneic dendritic cells (DCs) in vitro was also examined. Histologic analysis revealed that HGF gene transfection in TSK/+ mice resulted in a marked reduction of hypodermal thickness, including the subcutaneous connective tissue layer. The hypodermal thickness of HGF-treated TSK/+ mice was decreased two-fold to three-fold compared with untreated TSK/+ mice. However, TSK/+ associated defects in pulmonary architecture were unaffected by HGF gene transfection. HGF gene transfection significantly inhibited the expression of IL- and TGF-β1 mRNA in the spleen and skin but not in the lung. We also performed a mixed lymphocyte culture and examined the effect of recombinant HGF on the generation of IL-. Recombinant HGF significantly inhibited IL- production in TSK/+ CD4+ T cells stimulated by allogeneic DCs. HGF gene transfection inhibited IL- and TGF-β mRNA expression, which has been postulated to have a major role in fibrinogenesis and reduced hypodermal thickness, including the subcutaneous connective tissue layer of TSK/+ mice. HGF might represent a novel strategy for the treatment of SSc. BODY: IntroductionSystemic sclerosis (SSc) is a connective tissue disorder of unknown etiology that is characterized by an excessive deposition of extracellular matrix protein in the affected skin, in addition to various internal organs. Currently, no effective therapies for SSc exist []. The tight-skin (TSK/+) mouse is a genetic model for human SSc. Although the phenotypic characteristics of the TSK/+ mouse are not identical to those of human SSc patients, TSK/+ mice produce autoantibodies against SSc-specific autoantigens, including topo-I, fibrillin (fbn-), collagen type , and Fcγ receptors [,]. The gene defect responsible for the TSK/+ phenotype in mice is yet to be definitively identified; however, the fbn- gene has been recently proposed as a candidate locus for this disorder []. The fbn- gene mutation seems to provide an explanation for the embryonic lethality of homozygous tight skin (TSK/TSK) animals. Heterozygous (TSK/+) mice, which have a normal lifespan, exhibit fibrosis and thickening of subcutaneous dermal tissue. Other abnormalities in TSK/+ mice include increased lung collagen content, enlarged air spaces reminiscent of pulmonary emphysema, and, with advanced age, development of progressive myocardial fibrosis and hypertrophy [-].Hepatocyte growth factor (HGF) was originally identified and cloned as a potent mitogen for hepatocytes [,]. It has mitogenic, motogenic, and morphogenic effects on various epithelial tissues, including the liver, kidneys, lungs, and intestines [-]. HGF also shows antiapoptotic activity [] and has a role in suppressing fibrosis in the liver []. HGF might, therefore, induce tissue repair in dermal sclerosis associated with SSc. We recently demonstrated that repeated transfection of the human HGF gene into skeletal muscle induced continuous production of HGF, strongly inhibited acute graft-versus-host disease (GVHD) after allogeneic hematopoietic stem-cell transplantation (HSCT), and protected against thymic damage caused by GVHD in a well-characterized mouse model of GVHD [,]. The present study was performed to examine the therapeutic effect of HGF on tissue fibrosis in TSK/+ mice.Materials and methodsAnimalsTSK/+ and homozygous pallid (pa/pa) mutant mice with a C57BL/ background were obtained from the Jackson Laboratory (Bar Harbor, ME, USA). TSK/+ mice are heterozygous for both TSK and pa gene mutations. TSK/+ mice were produced by mating TSK/+ male mice with pa/pa female mice. TSK/+ progeny were identified by their back-coat and eye colors, in addition to their characteristic loss of skin pliability. To verify the TSK/+ genotype, PCR amplification of a partially duplicated fbn- gene was carried out using genomic DNA from each mouse, as described previously []. C57BL/ and (C57BL/ × DBA/)F1 (BDF1) mice were obtained from the Shizuoka Laboratory Animal Center (Hamamatsu, Shizuoka, Japan). All mice were maintained in a pathogen-free facility at the Hyogo College of Medicine (Nishinomiya, Hyogo, Japan). Animal experiments were performed in accordance with the guidelines of the National Institutes of Health (Bethesda, MD, USA), as specified by the animal care policy of Hyogo College of Medicine.Expression vector and preparation of liposomes containing hemagglutinating virus of Japan (HVJ)Human HGF cDNA (. kb) was inserted into the EcoRI and NotI sites of the pUC-SRα plasmid under control of the cytomegalovirus enhancer-promoter []. HVJ liposomes containing plasmid DNA and high mobility group protein were constructed, as described previously [,]. Briefly, phosphatidylserine, phosphatidylcholine, and cholesterol were mixed at a weight ratio of : . : . This lipid mixture ( mg) plus plasmid DNA (– μg), which had previously been complexed with – μg of high mobility group nonhistone chromosomal protein purified from calf thymus, were sonicated to form liposomes and then mixed with ultraviolet-irradiated HVJ. Excess free virus was subsequently removed from the HVJ liposomes by sucrose-density-gradient centrifugation.Gene transferTSK/+ mice were injected in the gluteal muscle with either HVJ liposomes containing μg of the human HGF expression vector (HGF-HVJ liposomes) or mock vector (untreated control). Gene transfer was repeated once weekly for weeks.HistopathologyTissues were fixed in % buffered formalin and embedded in paraffin. Sections were stained with hematoxylin and eosin and were examined by light microscopy.RT-PCRRNA was extracted using an Isogen (Nippongene, Toyama, Japan) kit, according to the manufacturer's instructions, and cDNA was prepared using . μM random hexamers (Applied Biosystems Inc., Foster City, CA, USA). IL- and TGF-β mRNA levels were quantified by real-time RT-PCR in a total volume of μl for – cycles of seconds at °C and – cycles of minute at °C. Samples were run in triplicate and relative expression levels were determined by normalizing expression levels to that of β-actin. The primer sequences used were as follows:. IL-, CCAGCTAGTTGTCATCCTGCTCTTCTTTCTCG and CAGTGATGAGGACTTGGACTCATTCATGGTGC.. TGF-β1, TGGACCGCAACAACGCCATCTATGAGAAAACC and TGGAGCTGAAGCAATAGTTGGTATCCAGGGCT.. β-actin, TGTGATGGTGGGAATGGGTCAG and TTTGATGTCACGCACGATTTCC.Mixed lymphocyte reaction (MLR) and in-vitro cytokine productionCD4+ T cells and CD11c+ dendritic cells (DCs) were purified from spleen cells using immunomagnetic beads (Miltenyi Biotec, CA, USA). The purity of the CD4+ and CD11c+ cell populations was >% and >%, respectively. CD4+ T cells from TSK/+ (H-2b) mice ( × cells/ml/well) were cultured with irradiated ( Gy) CD11c+ DCs from BDF1 (H-2b × d) mice ( × cells/ml/well) in -well flat-bottomed plates (Falcon Labware, Lincoln Park, NJ, USA). After hours, viable cells ( × cells/ μl/well) were stimulated in -well flat-bottomed plates (Falcon Labware) coated with μg/ml anti-CD3 mAb. After hours, IL- and IFN-γ concentrations in the culture supernatants were measured by ELISA.ELISA for IL-4The levels of murine IL- in culture supernatants were measured by ELISA using antimouse IL- mAb (Genzyme Pharmaceuticals, Cambridge, MA, USA) according to the manufacturer's protocol.Statistical analysisGroup mean values were compared by the two-tailed Student's t-test. A p value of < . was considered statistically significant.ResultsPrevention of scleroderma by HGF gene transfectionWe previously reported that repeated transfection of the human HGF gene into skeletal muscle of allogeneic HSCT recipients reduced the tissue damage and subsequent inflammatory responses caused by acute GVHD [,]. To investigate the possible therapeutic effects of HGF on SSc, young TSK/+ mice were treated with HGF gene transfection. Treatment consisted of once-weekly intramuscular injection of either HGF-HVJ liposomes or control mock vectors for a period of weeks (Figure ). Histologic analysis revealed that HGF treatment of TSK/+ mice resulted in a marked reduction of hypodermal thickness, including the subcutaneous connective tissue layer (Figure ). Skin fibrosis was assessed quantitatively by measuring hypodermal thickness. The hypodermal thickness of HGF-treated TSK/+ mice was decreased two-fold to three-fold compared with untreated TSK/+ mice (Figure ).Effect of HGF gene transfection on pulmonary changesThe effect of HGF gene transfection on pulmonary changes was examined. Control mock-vector-transfected TSK/+ mice all displayed a markedly dilated alveolar space, with fewer alveolar walls compared with pa/pa mice. This TSK/+-associated abnormality in lung architecture was unaffected by HGF gene transfection (Figure ).Effect of HGF on the expression of IL- and TGF-β mRNA expression and productionIL- and TGF-β have been postulated to have major roles in fibrinogenesis in animal models [-]. To clarify whether modulation of IL- and TGF-β has a role in the prevention of sclerosis induced by HGF gene transfection in the scleroderma model mouse, we examined the mRNA expression of these cytokines. HGF gene transfection significantly inhibited the expression of IL- and TGF-β1 mRNA in the spleen and skin but not in the lung (Figure ). We also performed MLR and examined the effect of HGF on the production of IL-. Responder CD4+ T cells from TSK/+ mice were cultured with irradiated ( Gy) CD11c+ DCs from BDF1 mice with or without recombinant HGF. After days' culture, viable cells were stimulated by culture with anti-CD3 mAb for hours and the IL- level in the culture supernatant was assayed by ELISA. HGF significantly inhibited IL- production from TSK/+ CD4+ T cells stimulated by BDF1 DCs (Figure ).DiscussionSSc is an autoimmune connective tissue disease that is characterized by microvascular damage, extracellular matrix deposition, and fibrosis. There is no completely effective treatment for this disease at present. We previously demonstrated that serum HGF levels were significantly elevated in patients with SSc and serum HGF levels correlated to markers of endothelial injury (thrombomodulin) and interstitial lung injury (KL-), suggesting that elevation of serum HGF counteracts the endothelial and interstitial lung injury caused by SSc []. The serum level of HGF is significantly elevated in various diseases, depending on the severity of the disease [-]. However, endogenously induced HGF is not sufficient to repair tissue injuries, and, therefore, supplementation with exogenous HGF is necessary to accelerate the tissue repair process in animal models [,,]. In the present study, we assessed the effect of exogenous HGF on skin fibrosis and the development of pulmonary defects in the TSK/+ mouse model of SSc. Both our present study and other previous studies [] have shown that dermal thickness is similar in TSK/+ and wild-type littermates, but hypodermal thickness, including the subcutaneous connective tissue layer, is significantly increased in TSK/+ mice compared with wild-type littermates. HGF gene transfection of TSK/+ mice for a period of weeks resulted in a marked reduction of hypodermal thickness, including the subcutaneous connective tissue layer. Although the therapeutic effect of HGF is not significant, we also observed the reduction of hypodermal thickness in TSK/+ mice following HGF gene transfection for a period of weeks (data not shown).Although the cause of SSc is unknown, IL- and TGF-β have been postulated to have major roles in fibrinogenesis. In one study, intravenously administered human immunoglobulin decreased splenocyte secretion of IL- and TGF-β, which resulted in abrogation of fibrinogenesis in TSK/+ mice and, consequently, prevented the accumulation of fibrous tissue []. Furthermore, administration of an anti-IL- or anti-TGF-β antibody prevented dermal collagen deposition in TSK/+ mice and murine sclerodermatous GVHD, respectively [,]. In the present study, HGF treatment also reduced expression of both IL- and TGF-β mRNA in the spleen and skin.IL- regulates collagen and extracellular matrix production by fibroblasts [,]. TSK/+ mice exhibiting disrupted genes encoding IL- receptor alpha (IL-4Rα) or IL- lacked skin sclerosis [,], suggesting that IL- has a crucial role in skin sclerosis in TSK/+ mice. A primary source of IL- in vivo is CD4+ T cells [] and a previous study demonstrated that CD4+ T cells were essential to the TSK/+ phenotype, because a lack of these cells prevented development of dermal thickening []. Therefore, we examined the effect of HGF on the generation of IL- from CD4+ T cells. HGF significantly inhibited IL- production from CD4+ T cells stimulated by allogeneic DCs, suggesting that HGF inhibits dermal fibrosis, in part, by inhibiting IL- production by CD4+ T cells.We also observed downregulation of TGF-β1 mRNA expression in TSK/+ mice by HGF gene transfection. TGF-β1 has a role in the induction of fibrosis, and HGF gene transfection inhibited the production of TGF-β1 from macrophage-like cells and fibroblastic cells []. Downregulation of TGF-β1 expression and inhibition of fibrosis by HGF were noted in a rat model of liver cirrhosis [] and a mouse model of chronic renal failure []. Recently, HGF has been shown to downregulate TGF-β1 expression and prevent dermal sclerosis in a murine bleomycin-induced scleroderma model []. The authors observed that HGF gene transfection significantly reduced both the expression of TGF-β1 mRNA and the production of TGF-β1 by fibroblastic cells and macrophage-like cells that infiltrated the dermis. Furthermore, HGF gene transfection prevented the symptoms of not only dermal sclerosis, but also lung fibrosis induced by bleomycin injection.By contrast, HGF gene transfection failed to alter the development of pulmonary abnormalities in TSK/+ mice in our study. The pathologic alteration of the lung structure of TSK/+ mice represents pulmonary emphysema and is not related to hypersynthesis of collagen that is similar to the pulmonary fibrosis associated with SSc []. Apparently, emphysema in TSK/+ mice is not owing to the mutated fbn- gene that is responsible for the occurrence of cutaneous hyperplasia, because transgenic mice bearing a mutated fbn- gene developed cutaneous hyperplasia but did not exhibit pulmonary emphysema []. Furthermore, TSK/+ mice exhibiting disrupted genes encoding IL-4Rα, TGF-β, or IL- lacked skin sclerosis but developed emphysema, indicating that different genes are involved in the development of skin sclerosis and pulmonary emphysema in TSK/+ mice [,]. Furthermore, other studies have shown that the pulmonary pathology remained unchanged in TSK/+ CD4 knockout (CD4-/-) mice [] and TSK/+ mice treated with an anti-IL- antibody []. The dermal and pulmonary components of the TSK/+ phenotype can, therefore, be dissociated in vivo.We used a transgenic approach instead of using a recombinant protein for the following reasons:. Because the half-life of a recombinant protein is quite short, recombinant protein treatment needs an enormous dose of the active form of HGF protein and frequent injections.. Administering a high dose of the active form of the HGF protein could cause adverse effects, such as tumorigenesis in other organs [].. Recombinant protein treatment is very costly.By contrast, the transgenic approach is simple, safe, and cheap and needs much less frequent injections. Repeated weekly injection of HGF-HVJ liposomes achieves a continuous high plasma level of HGF [-].Although further studies are needed to fully explore the effects of HGF on SSc, it is possible that HGF therapy might be a promising strategy for the treatment of SSc.ConclusionHGF gene transfection of TSK/+ mice resulted in a marked reduction of hypodermal thickness, including the subcutaneous connective tissue layer. However, TSK/+-associated defects in pulmonary architecture were unaffected by HGF gene transfection. HGF gene transfection significantly inhibited the expression of IL- and TGF-β1 mRNA in the spleen and skin but not in the lung. Recombinant HGF significantly inhibited IL- production by TSK/+ CD4+ T cells stimulated by allogeneic DCs. HGF might represent a novel strategy for the treatment of SSc.AbbreviationsCD4-/- = CD4 knockout; DC = dendritic cell; ELISA = enzyme-linked immunosorbent assay; fbn- = fibrillin ; GVHD = graft-versus-host disease; HGF = hepatocyte growth factor; HSCT = hematopoietic stem-cell transplantation; HVJ = hemagglutinating virus of Japan; IFN = interferon; IL = interleukin; IL- receptor alpha = IL-4Rα; mAb = monoclonal antibody; MLR = mixed lymphocyte reaction; pa/pa = homozygous pallid; PCR = polymerase chain reaction; RT = reverse transcriptase; SSc = systemic sclerosis; TGF = tumor growth factor; TSK/TSK = homozygous tight skin; TSK/+ = tight skin.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsT Iwasaki conceived the study, participated in the design and co-ordination of the study, and participated in the interpretation of the results. T Imado and SK performed the animal study and histologic analysis. HS participated in the design of the animal study.
|
PMC2777903.txt
|
TITLE: Acute epiglottitis as the initial presentation of pediatric Systemic Lupus Erythematosus AUTHORS: Sirirat Charuvanij, Kristin M Houghton ABSTRACT: We report a case of a -year old girl, who initially presented with acute epiglottitis, sepsis and multi-organ failure. She was subsequently diagnosed as having Systemic Lupus Erythematosus. To the best of our knowledge, this article describes the first case of Haemophilus influenzae type f epiglottitis as the initial presentation of SLE in childhood. BODY: IntroductionSystemic Lupus Erythematosus (SLE) is a periodic, multisystem, autoimmune disease characterized by the presence of antinuclear antibodies. Approximately % of SLE cases are diagnosed before the age of and children often have more severe disease at onset and higher rates of organ involvement than adults []. We report a case of SLE with initial presentation of Haemophilus influenzae epiglottitis. Laryngeal involvement and pneumococcal epiglottitis have infrequently been reported in the adult literature [-]. This is the first pediatric case and the first case of Haemophilus influenzae epiglottitis reported in SLE.Case presentationA -year old fully immunized and previously healthy girl of Filipino origin presented to her local hospital with a one day history of high fever (°C), sore throat, stridor and shortness of breath. She was transferred to BC Children's Hospital and was intubated by an otolaryngologist in the operating room. Subsequently, she was admitted to the ICU and was intubated and ventilated for days. A lateral neck radiograph revealed a profoundly swollen epiglottis and complete airway obstruction (Figure ). The diagnosis of acute epiglottitis was made based on her clinical picture and imaging findings. She was empirically treated with intravenous Cefotaxime. Her throat swab and blood culture were positive for Haemophilus influenzae type f.Figure 1A lateral neck radiograph demonstrates a profoundly swollen epiglottis and complete airway obstruction.On day of admission, she developed multi-organ system dysfunction with anemia, thrombocytopenia, acute renal failure and hypertension (BP / mmHg). Investigations revealed hemoglobin at g/L, white blood cell × /L with lymphocyte of . × /L. Her platelets were ,/mm3. DAT was positive. PT and INR were normal. APTT and dilute Russell Viper time were elevated at sec (.-.) and . ratio (.-.) respectively. Fibrinogen and D Dimer were elevated. Her serum creatinine was umol/L (<). Urinalysis showed - RBC/hpf, protein . g/L with hyaline and granular casts. Treatment included supportive care, antibiotic (Cefotaxime), red blood cell and platelet transfusions.On day of admission, she developed seizures which were felt to be due to severe hypertension (BP / mmHg). Computed tomography (CT) brain revealed multiple areas of abnormal hypodensity in the subcortical white matter of both hemispheres (Figure ). Magnetic resonance imaging (MRI) brain demonstrated extensive T2 signal changes throughout the white matter with diffuse cerebral edema involving both cerebral hemispheres (Figure ). A diagnosis of probable posterior reversible encephalopathy syndrome (PRES) was made.Figure 2CT brain reveals multiple areas of abnormal hypodensity in the subcortical white matter of both hemispheres.Figure 3MRI brain shows extensive T2 signal changes throughout the white matter with diffuse cerebral edema involving both cerebral hemispheres.On day of admission, the rheumatology team was consulted. The main differential diagnosis at that time included SLE, Thrombotic Thrombocytopenic Purpura (TTP) and Wegener's Granulomatosis. A week later, her autoimmune and serologic workup came back positive for anti-nuclear antibody (ANA) at a titer of >: , positive anti-double stranded DNA, lupus anticoagulant, anticardiolipin antibody (ACA IgG was elevated at . MoM) and low complements C3 (. g/L) and C4 (. g/L). She underwent a renal biopsy which showed mesengial lupus nephritis (WHO class II) but due to a poor sample (only glomeruli on light microscopy) the possibility of active proliferative lupus nephritis could not be excluded. Based on her clinical and laboratory findings, she was diagnosed with SLE and treated with a course of pulse Methylprednisolone mg/kg/day for days and subsequently converted to high dose oral prednisone ( mg/kg/day). She received her first Cyclophosphamide infusion ( mg/m2/dose) in the hospital prior to discharge.She was discharged home on prednisone ( mg/kg/day) with a scheduled taper, Hydroxychloroquine, Enalapril, Amlodipine and Clonazepam. She was scheduled to receive Cyclophosphamide infusions as an outpatient. At last assessment in our rheumatology clinic ( months after initial presentation), her disease was in clinical remission. Her anti dsDNA was negative. C3 was slightly low at . g/L. C4 was normal. ENA panel remained negative. She continued to have positive ANA at lower titer of : . She completed courses of monthly Cyclophosphamide infusions and was switched to Mycofenolate Mofetil for maintenance immunosuppressant therapy.DiscussionEpiglottitis is an inflammation of the epiglottis and the adjacent tissues surrounding the epiglottis. The clinical presentation includes abrupt onset of high fever, severe sore throat, dysphagia, "hot potato" voice, drooling and rapid progression of airway obstruction. Several bacteria have been associated with epiglottitis in children. Among them, Haemophilus influenzae type b was responsible for most cases of epiglottitis in pediatric patients in the pre- Hib vaccine era [].Haemophilus influenzae is a pleomorphic gram-negative coccobacillus. There are encapsulated and non-encapsulated forms. After the initiation of Hib protein-polysaccharide conjugate vaccines, most cases of invasive Haemophilus influenzae infection have been attributed to non-type B strains []. Serology type f (Hif) is reported as the most common cause of invasive encapsulated non-b Haemophilus influenzae disease in children []. The risk of invasive Haemophilus influenzae disease include younger and older ages, complement deficiency, hypogammaglobulinemia, sickle cell anemia, functional asplenia, malignancy and HIV infection.SLE is the prototypical multisystem autoimmune disease. Clinical manifestations of SLE are variable and the most common pediatric presentations include arthritis, malar rash, nephritis and central nervous system disease []. Our patient had lupus nephritis, hematological abnormalities, a positive ANA and a positive antibody to double stranded DNA, thus fulfilling four of the revised American College of Rheumatology classification criteria for SLE []. TTP was also initially considered given her thrombocytopenia, anemia, coagulopathy, fever, renal involvement and CNS involvement. TTP associated with SLE has been described [-]. However, her peripheral smear did not show typical microangiopathic changes of TTP and ADAMTS was not performed. Furthermore, her seizure activity was thought to be due to PRES, not TTP or CNS lupus but the clinical presentations have considerable overlap and definitive diagnosis is challenging. PRES is a rapidly developing neurologic condition characterized by headache, decreased alertness, visual loss, seizures, hypertension and lesions in the posterior cerebral white matter. PRES was first reported in by Hinchey []. PRES associated with SLE has been reported in adults and children [-]. Our patient's MRI brain showed multiple hypodensity lesions in the subcortical white matter of both hemispheres, right post-central gyrus and on the left just at the central sulcus in keeping with the effects of hypertensive encephalopathy ("PRES").Apart from the infection risk associated with immunosuppressive therapies, infection in SLE can be secondary to a primary defect in the immune system or alterations in the innate and adaptive immune responses. In addition, splenic dysfunction is reported in SLE patients [,]. One study reported .% of SLE patients with functional asplenia []. These patients seem to have increased susceptibility to encapsulated bacteria including pneumococcal, meningococcal, Haemophilus influenzae and salmonella infection. Complement abnormalities are also reported in SLE patients resulting in defective opsonization of encapsulated organisms []. Even though infections occur commonly in SLE patients, it is rarely the presenting feature of the disease. The presentation of our patient with Haemophilus influenzae type f epiglottitis is unusual.Although our patient did not have morphologic erythrocyte changes, it is possible that she developed splenic dysfunction during her initial clinical presentation of SLE. This may explain an increased susceptibility to Haemophilus influenzae. However, a 99m-technetium sulphur colloid scan and 99m- technetium labeled, heat-damage erythrocyte scan were not performed to support this hypothesis. The other hypothesis is that SLE is induced by Haemophilus influenzae infection. It is reported that bacterial DNA can enhance several of the autoimmune abnormalities observed in SLE and perhaps play a pathogenic role in the induction of SLE []. There is a report of Haemophilus influenzae type f pericarditis and tamponade as the initial manifestation of systemic lupus erythematosus in an adult []. There are no reports of epiglottitis as the initial presentation of SLE in the literature but there are case reports of pneumococcal epiglottitis in adults with SLE and lupus overlap syndrome [,]. Laryngeal involvement in adult SLE ranges from mild ulcerations, vocal cord paralysis, laryngeal edema to necrotizing vasculitis with airway obstruction []. Furthermore, severe upper airway obstruction from cricoarythenoiditis was reported as the presenting manifestation of a SLE flare in an adult [].In conclusion, our pediatric patient presented with Haemophilus influenzae type f epiglottitis and septicemia and was subsequently diagnosed as having SLE. An opsonization defect and splenic dysfunction are most likely responsible for her increased susceptibility to infection with encapsulated bacteria. In addition, it is postulated that severe bacterial infection may play a role in inducing SLE. SLE has been called the "disease with a , faces" as it can present in many ways and it needs to be considered in the differential diagnosis of a sick child with multi-system disease. Clinicians need to be vigilant for infection in children with SLE as they may be at increased risk from their disease, functional hyposplenism and immunosuppressant therapy. Indeed, infection is the leading cause of death in pediatric SLE []. Our clinical practice is to vaccinate all newly diagnosed SLE patients against encapsulated organisms.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsSC and KH participated in drafting the manuscript. All authors read and approved the final manuscript.ConsentConsent was obtained from the parent of the patient for publication of this case report and accompanying images.
|
PMC1761841.txt
|
TITLE: Ethical Issues in Providing Online Psychotherapeutic Interventions AUTHORS: Craig A Childress ABSTRACT: The Internet offers psychotherapists a new communication medium through which they can deliver psychotherapeutic interventions that are appropriate to the medium. Yet online psychotherapy also offers new ethical challenges for therapists interested in providing online psychotherapeutic services. The differences between interactive text-based communication and in-person verbal communication create new ethical challenges not previously encountered in face-to-face therapy. This article will examine the Internet's potential for providing online psychotherapeutic interventions and will review the ethical issues involved with providing interactive text-based psychotherapy. BODY: IntroductionThe Internet provides a new medium for interpersonal communication that holds the potential for delivering forms of psychotherapeutic interventions that are appropriate to the medium. The challenges facing psychotherapists lie in discovering what types of interventions are appropriate to this new medium and in delineating the potential advantages and limitations inherent to this new communication format.Mental heath professionals are already exploring the usefulness of the Internet medium in delivering online psychotherapeutic interventions [-], and several professional associations have developed general guidelines for the online delivery of therapeutic services [-]. Simply practicing within recommended guidelines, however, does not release each individual therapist from the personal responsibility to be aware of, and to independently evaluate, the variety of ethical issues involved in the practice of online therapy. The obligation to act ethically cannot be transferred to an organization, but remains the personal responsibility of each therapist seeking to practice online.The Internet provides several different communication systems, some of which are similar to in-person communication (e.g. video technology), some of which are similar to traditional print media (e.g. web pages), some are hybrids of interactive in-person communication and traditional written-word communication (e.g. email and chat), and some are unique to the Internet medium (e.g. MOOs and MUDS). The ethical and pragmatic challenges facing psychotherapists seeking to use the Internet to deliver online psychotherapeutic interventions will vary depending upon which communication medium is being used.This article will focus primarily on the use of email to deliver online psychotherapeutic interventions. While two-way video technology may someday become widely available, it is unclear whether it will ever gain acceptance as a common means of personal communication on the Internet. Two-way video technology has long been available for telephones, but people have not rushed out to buy video telephones. It is also questionable whether people will feel comfortable having video cameras in their homes. Cameras attached to personal computers may be viewed as an unwelcome intrusion into personal privacy. While interactive real-time video communication holds potential in a variety of health related interventions, the technology may remain limited to large organizational and hospital uses, without widespread dissemination into personal use. Should two-way interactive video become widely accepted and available, then it can be incorporated into delivering therapeutic interventions; and since video is a form of "-to-face" interaction, the ethical issues will, to a large extent, be similar to those encountered with in-person, face-to-face therapy.Currently, however, email provides the backbone of interactive online communication, and it may hold the greatest potential for delivering psychotherapeutic interventions using the Internet []. The most significant feature of email is that it is text-based communication, and this is the source of its greatest strengths and its greatest limitations.Issues Specific to Email Psychotherapeutic InterventionsAs text-based communication, email is asynchronous (not in real-time), which allows the participants to communicate at their own convenience. The asynchronous nature of email can facilitate the client's perception of the therapist's availability, and may provide the client with a more intense psychological holding environment [] than is available through a traditional in-person relationship. The perceived availability of the therapist may also enhance the client's ability to incorporate the therapist's presence into daily life. Rather than waiting for the weekly in-person session to discuss an issue, the client can instead write the therapist an email while the issue is still active, thereby evoking the therapist's psychological presence in the moment.Asynchronous text-based communication also allows the parties involved to carefully consider and edit their communication, which is an advantage over real-time text or in-person communication. Even asynchronous video and audio communication do not offer the advantages of editing afforded by interactive text-communication. Asynchronous text communication actually involves writing small-scale essays, similar to traditional letter writing. The interactive nature of email communication comes from the speed by which these essays or letters are exchanged, which gives the illusion of greater interactivity than might actually be present.Interactive text-based communication also involves the loss of the non-verbal social cues that provide valuable contextual information in conversation and can influence the interpretation of meaning in communication. Miscommunication may therefore be more likely with interactive email communication. The loss of physical social cues may also increase the client's tendency to project personal psychological material onto the blankness of cyberspace communication. This enhanced tendency toward projection of personal material in text-based communication may be helpful in some forms of psychotherapeutic interventions, and it may offer distinct advantages over in-person communication as well as a potential risk for increased miscommunication.Nor does email need to be the sole intervention offered to clients. Email might be fruitfully integrated into a traditional in-person psychotherapy, allowing the client to continue the therapeutic process between in-person sessions with the therapist. The in-person therapist might prescribe certain writing assignments to the client, which can then be emailed to the therapist between sessions. The issues raised in the email can then be addressed during the in-person session. There is some evidence that persons may feel more comfortable self-disclosing through a computer []; and clients might use email communication to broach sensitive issues, such as past experiences of abuse, that they may be unwilling to address during the in-person session. Having once raised the issue in email, it can then be dealt with more extensively during the in-person sessions.Therapists might also incorporate email into cognitive/behavioral interventions. The in-person therapist can prescribe homework involving the monitoring of behavior or thoughts. This daily homework could then be emailed to the therapist, allowing the therapist more timely access to the client's progress. Daily emailing of homework to the therapist might also help motivate and reinforce the client's completion of the assigned homework on a daily basis. It would also allow therapists to monitor the intervention throughout the week; and if corrections to the behavior program are needed, they can be developed and incorporated into the program prior to the next in-person session. Incorporating email into a traditional in-person therapy holds the potential to increase the speed of therapeutic progress, the depth of material discussed, and cost-effectiveness of treatment [].Email clearly holds potential for delivering psychotherapeutic interventions, either by itself or in conjunction with a traditional in-person relationship. This potential has led increasing numbers of psychotherapists to begin exploring the use of email (and chat) to deliver psychotherapeutic interventions []. However, the use of a new communication medium involving interactive text-based communication raises unique ethical questions not previously addressed within the confines of the in-person therapeutic relationship. The delivery of therapeutic interventions solely through the Internet (i.e., not in association with an in-person therapeutic relationship) offers the most problematic ethical issues, and it will be those issues that will be focused on in the remainder of this article.Guiding Ethical PrinciplesEthical Responsibility to Provide ServiceOne of the initial ethical issues involves the responsibility of mental health professionals to provide services to meet the demand of consumers. While reservations may exist regarding the provision of online psychotherapeutic services, if consumers desire such services and if there is a reasonable expectation that online therapeutic interventions can be beneficial, then we have a professional obligation to address this demand. If mental health professionals do not step forward to provide these services, then consumers will be forced by the lack of response from the professional community to seek online therapeutic services from unlicensed and untrained providers. Inasmuch as there appears to be a reasonable expectation that some form of psychotherapeutic intervention can be developed that can appropriately be delivered using Internet text-based communication [,], it is incumbent upon the field of professional psychology to explore the ethical and professionally responsible delivery of online psychotherapeutic interventions. Yet, once we accept the obligation to explore the professional delivery of online psychotherapeutic services, then other problematic ethical issues emerge [-].Do No HarmThe evaluation of the potential harms associated with any treatment intervention needs to be considered within the context of the potential benefits to be accrued from the intervention []. Only by considering both the potential risks and the possible benefits can we appropriately evaluate a proposed intervention. The simple presence of risk does not necessarily preclude the use of an intervention if it is sufficiently justified by the potential benefits. With interventions that have a reasonable likelihood of being beneficial for the client, the important issue becomes for the therapist to understand the nature of the risks, to minimize the risks to the extent possible, to fully inform clients as to the nature of the risks within the context of the possible benefits, and then to allow the clients to make an informed decision regarding their treatment options.In assessing the risks of online psychotherapy it is also important to note that in-person psychotherapy is not without risks. In-person clients can become sexually attracted to the therapist and vice versa; therapists can be incompetent in the delivery of services; the therapist's confidential records are vulnerable to being stolen or viewed by unauthorized persons even if stored in a locked office and a locked file cabinet; miscommunication can occur during in-person therapy; and clients receiving in-person therapy can deceive and mislead the therapist. The issues with the delivery of online psychotherapy (e-therapy) are the extent to which traditional risks are enhanced in text-based communication, the possible emergence of novel types of risks not present in face-to-face therapy, and the degree to which the potential benefits justify the possible risks.The use of email to deliver therapeutic interventions opens several areas of potential risk to online consumers. Clients of online services can be at greater risk for breaches in confidentiality [,]. This increased risk to confidentiality occurs at the therapist's end, at the client's end, in the transmission of information, and in the potential for legal subpoena of records. Therapists using the Internet to deliver therapeutic interventions should evaluate the security of their websites and computers against outside intrusions that would compromise client confidentiality. These intrusions might include high-tech invasions by hackers downloading files from the therapist's computer, to low-tech intrusions involving the inappropriate availability of the client's email to the therapist's office staff or family members. Therapists using the Internet to deliver online psychotherapeutic interventions may wish to consider installing systems which use firewalls, passwords, and backup data storage systems to increase the security of email communications and to protect against the inadvertent loss of clinical files resulting from computer malfunctions.Online consumers of mental health services must likewise consider security issues on their end of the communication. Other persons who have access to the client's email, such as employers or family members, may be able to read stored copies of the client's email or incoming email from the therapist. Additionally, human error in addressing email has sometimes resulted in email being sent to the wrong person. Inadvertently sending private information meant for the therapist to a friend or family member can result in embarrassing and painful situations for the client. Potential online consumers of mental health interventions need to be informed about these potential breaches of confidentiality in order to fully evaluate the possible risks versus the potential benefits of online psychotherapy.Breaches of confidentiality can also occur as email is in transit. The potential vulnerability of email in transit may not, however, accurately represent its actual vulnerability in practice. While email may be intercepted in transit, it is unlikely that individual emails sent between private parties are actually intercepted and read from the incredible volume of email sent each day. Still, this potential breach in confidentiality needs to be understood and evaluated by clients before choosing to engage in online psychotherapy. Encryption technology can improve security of email communication, and online therapists may wish to make encryption of email routinely available to their clients.Online mental health clients also need to consider the possibility that email records may be subject to subpoena. While professional communication with physicians and attorneys is considered legally privileged, it is unclear if this legal protection extends to psychotherapists. The standards for recognition of legal protection of privileged communication may also vary from one jurisdiction to another. Online psychotherapists should consider their policy regarding the disclosure of records in response to legal subpoena, and clients need to be informed about this possible breach of confidentiality.The use of email to provide psychotherapeutic interventions also entails other risks to clients beyond those associated with the confidentiality of communication. For example, the loss of nonverbal cues significantly impedes the therapist's ability to make a full assessment and diagnosis. Important in-person cues, such as flattened or inappropriate affect, characteristics of speech, memory function, or physical evidence of a medical condition that may be associated with the psychological symptoms, are all lost in email communication. An impaired ability to make an adequate diagnosis will adversely affect the ability of online therapists to develop appropriate treatment plans and, as a result, the treatment interventions that are developed may be to the detriment of the client. Online testing may improve diagnostic capabilities [], and gathering a full psychosocial history may be facilitated by online questionnaires; yet the loss of visual and auditory cues will still affect the therapist's diagnostic ability, and the impact of this diminished diagnostic capability needs to be carefully considered. Still, while problems making online diagnoses may limit the scope of issues appropriately addressed in online therapy, some types of online interventions, such as interactive journaling [] or humanistic/existential approaches, may nevertheless be developed that are appropriate for delivery in a text-based format with some populations.The increased potential for miscommunication in text-based therapy may also increase the risk of inadvertently harming clients and perhaps re-traumatizing emotional injuries disclosed during the course of online therapy. Text-based interactive communication is more vulnerable to miscommunication because it lacks the non-verbal cues associated with in-person communication that modify meaning and provide context for the interpretation of meaning. Furthermore, interactive text communication is not the normal means of interpersonal communication for most therapists trained in in-person psychotherapy. Therapists may therefore lack the writing skills needed to express subtleties of meaning through the written word.Working with psychological issues typically involves addressing conflicting client motivations involving a desire for self-disclosure aimed at securing help for painful personal issues, and competing motivations directed toward maintaining interpersonal defenses to preserve self-esteem and prevent re-traumatization of emotional injuries. Interactive text-based communication often sounds harsher than intended. Online miscommunication may result in clients feeling hurt because they perceive the therapist's response as being critical or rejecting. Online clients also do not have the benefit of the interpersonal holding environment offered by the in-person relationship in which to interpret and integrate the therapist's comments, and injured online clients may be more likely to simply withdraw from the relationship into the blankness of cyberspace, taking their injury with them. Since nonverbal feedback cues that might signal the miscommunication, such as the client's body language and facial expression, are not available in email communication, online therapists may often be unaware of the miscommunication and therefore will be unable to address the client's injury.This possibility of emotional injury and re-traumatization may be further exacerbated by the increased self-disclosure and disinhibition associated with online communication [,]. While increased self-disclosure may be helpful in some therapy circumstances, it may also involve clients prematurely moving past defenses designed to protect them against emotional injury and re-traumatization. This may leave them more vulnerable to injury should they interpret a therapist's communication as being critical or rejecting.Clients in online psychotherapy may also be at increased risk of harm if the online intervention is not effective in creating change in the client's life, yet offers enough solace so as to reduce the client's motivation to seek more beneficial in-person therapy. Consumers of online mental health services are at risk in this case not because of a direct effect of the online intervention, but because the online intervention prevents them from seeking treatments that will more effectively address their needs. However, e-therapy may also serve as a convenient and helpful entry into the mental health system for many persons who might benefit from therapy but who are reluctant to begin in-person therapy because of the social stigma associated with psychotherapy, their anxiety of addressing emotional issues, and the physical inconveniences of scheduling in-person therapy sessions. For such persons, the convenience and perceived anonymity associated with computer-mediated communication may encourage them to contact an online psychotherapist. Their initial online therapeutic relationship may help demystify psychotherapy and facilitate their entry into in-person mental health treatment.The ethical practice of e-therapy requires that therapists develop a thorough understanding of all of these issues. Online discussion groups dealing with Internet psychology can help therapists explore some of these issues. Yet, despite the therapist's professional evaluation of the issues involved with providing online therapeutic interventions, the ultimate issue is the degree to which the potential benefits justify the possible risks, and a decision on this issue can only come from a fully informed client. While mental health professionals can decide that the potential benefits associated with the intervention do not justify the risks, the opposite decision, that the benefits do justify the risks, can only be made by a fully informed client. Therapists seeking to provide online psychotherapeutic interventions must, therefore, be informed as to the potential risks so that they can take every possible precaution to reduce or eliminate those risks, and so that they can fully educate potential clients regarding the possible risks associated with e-therapy.Providing Effective InterventionsWhile controversies exist as to what criteria should be used to evaluate the effectiveness of psychotherapy, in-person psychotherapy nevertheless has an extensive history and well-elaborated theoretical frameworks supporting its use. Both history and theoretical frameworks are missing from the practice of interactive text-based therapy, and it is currently unclear to what degree traditional therapeutic orientations and models can be translated into online, text-based communication. Most psychotherapy depends, to a greater or lesser degree, on the development of the therapeutic relationship [-]. However, it is precisely the nature of the therapeutic relationship that is most impacted by text-based communication.The ethical practice of e-therapy requires the therapist to have a clearly delineated model of psychotherapy appropriate to delivery in a text-based format [,]. In the emerging field of online psychotherapy, it would also behoove the ethical practice of e-therapy if therapists remained close to empirically derived support for the interventions used until more experience is gained with regard to the medium of interactive text-based communication. Therapists providing online psychotherapeutic interventions should also contribute to the developing understanding of e-therapy by conducting quantitative and qualitative evaluations of the services they deliver.Practicing Beyond the Boundaries of CompetenceFor psychologists, the Ethical Code of the American Psychological Association [] specifically directs that psychologists should practice only within the area of their competence based on training and experience (Standard .04a); and that where standards for training do not yet exist, psychologists should " reasonable steps to ensure the competence of their work and to protect patients, clients, students, research participants, and others from harm" (Standard .04c; p. ). Psychotherapists trained in traditional psychotherapy need to carefully consider whether they are competent to practice in an interactive text-based format, and to evaluate by what manner and training they achieved their competence in this new communication medium.Interactive text-based communication offers an entirely new communication format that differs significantly from in-person verbal communication. The nonverbal cues that in face-to-face communication provide valuable information that modifies meaning and aids interpretation of the communication are significantly absent in text-based communication. It takes considerable skill to communicate emotion and contextual intent solely through the written word. Text can often sound harsher than intended and, without contextual cues such as tone of voice and body language, text-based communication is more likely to be misunderstood and misinterpreted. With text-based communication there is also a greater likelihood of projective psychological material emerging in the absence of the physical presence which serves to ground in-person communication. Skill in verbal communication does not necessarily translate into skill in written communication, especially interactive text-based communication that involves a series of interpersonal interpretations within each exchange.Without clearly delineated models for text-based psychotherapy, and without training in the subtleties of interactive written-word communication, therapists seeking to provide online psychotherapy need to carefully evaluate their current level of competence to practice in a text-based format.Professional Accountability and the Redress of GrievancesMental health professionals wishing to practice online also need to consider their legal authority to practice in a jurisdiction in which they are not licensed to practice []. This issue extends beyond the legality of their activity to include the rights of clients to redress grievances.The ethical practice of online psychotherapy must provide for the client's ability to redress grievances. Clients should be clearly informed prior to beginning an online therapeutic relationship about the regulatory agencies and professional associations governing the therapist's work []. Still, simply being informed about oversight agencies may not offer online clients an actual ability to redress grievances when the therapist and client may live in separate jurisdictions separated by hundreds or even thousands of miles []. For example, while it may be possible for a client in India to file charges with an ethics board or Attorney General located in the therapist's home jurisdiction of Wisconsin, the practical limitations imposed by distance and the financial resources needed to overcome such limitations may leave the client unprotected in fact.Laws governing the appropriate practice of psychotherapy, such as ordinances governing the release from confidentiality to report child abuse, may also differ from one geographic jurisdiction to another. When the therapist and client live in two different legal jurisdictions with differing laws regarding the practice of psychotherapy, which jurisdiction's laws take precedence and govern the client-therapist relationship?In order to avoid the many problematic legal and professional issues related to practicing psychotherapy online, some therapists may be tempted to define their online work as psychoeducational rather than psychotherapeutic. While some online work can legitimately claim to be primarily educational, therapists treating individual clients across multiple sessions should carefully consider whether their work is primarily educational or therapeutic. One of the central issues in making this distinction is whether the client perceives an individual professional relationship has been established. While it may be tempting to try and circumvent legal and professional liability for online work by defining it as psychoeducational, it creates significant ethical problems if such a definition misrepresents the service. Ethical problems can also arise if the online service being provided is held out as therapeutic on one web page, with a disclaimer of psychoeducational intent located on a separate web page. An ethically appropriate description of the online service must clearly, consistently, and accurately describe the intent of the service and the nature of the professional relationship involved.Informed ConsentThe absence of physical presence also impacts the ability to verify identity. Without the ability to verify identity the issue of treating minors without parental consent becomes problematic. Therapists seeking to practice online must evaluate what steps will be taken to verify the age of clients so as to not treat minors without the knowledge and consent of their parents.In addition, the issue of informed consent is closely related to the issue of disclosure. As discussed earlier, in order to make an informed consent to treatment clients need to fully understand the potential risks and benefits associated with an intervention. Specific risks that clients need to be informed about involve the possibility that inadvertent breaches of confidentiality may occur with online communication, the experimental nature of online psychotherapeutic interventions and the possibility of unknown and unintended consequences, and the potential for miscommunication in text-based communication [].In some ways, however, the Internet offers advantages in developing an informed consent process. Professional web pages allow for multi-faceted and multi-layered discussion of relevant issues which remain constantly available on the Internet for clients to review. Web pages can address issues such as the potential risks involved with online treatment and the theoretical underpinnings of the treatment. The discussion of informed consent through email also allows for a documented record of the informed consent process.Crisis Intervention PlanningOnline psychotherapists need to consider plans for addressing the variety of crises that may present in therapy including suicidal clients, physical and sexual abuse, threats to harm others, and the possible discovery that the client's issues would more appropriately be addressed with intensive in-person therapy or hospitalization. Prior to beginning a therapeutic online relationship, therapists may wish to discuss crisis plans and develop in-person referrals local to the client in preparation for possible future crises. Such crisis planning should include obtaining a verified valid street address and phone number that would allow the therapist to invoke the local police should such an intervention become indicated.Boundary IssuesTherapists interested in providing online interventions also need to consider the possible boundary issues involved with establishing an online therapeutic relationship. For example, with instant message systems clients might be alerted every time the therapist is online and could send the therapist instant messages for chats every time the therapist signs onto the Internet. Clients might also access the therapist's personal web page or sign onto online discussion groups to which the therapist also belongs. In addition, some clients may continue to send the therapist emails after the termination of the relationship, and e-therapists will need to consider their response to such ongoing contact. Some clients may also use the Internet to harass or stalk current or former therapists.ConclusionsThe Internet provides new opportunities to provide beneficial psychotherapeutic interventions with clients. Yet in providing online psychotherapeutic interventions, therapists need to evaluate the degree to which the online clients are informed regarding the potential risks they are assuming, including the risk that because there is little formal research on the process of online therapy, there may arise unforeseen and unanticipated problems. Therapists also need to evaluate their own competency to deliver text-based interventions and the source of this competency in their background and training. Before providing online therapy, mental health professionals also need to develop theoretical models for the interventions being used that are appropriate to delivery in a text-based format.Therapists seeking to provide online interventions also need to become thoroughly familiar with the risks associated with e-therapy and with the professional guidelines being developed for the ethical practice of e-therapy. Online professional discussion groups devoted to Internet psychology may help by offering professional consultation regarding issues related to e-therapy; yet therapists cannot rely entirely on professional guidelines or online consultation, and must actively accept their personal responsibility for fully understanding, considering, and addressing the potential ethical issues involved with online therapy.
|
PMC2376119.txt
|
TITLE: Systematic review for non-surgical interventions for the management of late radiation proctitis AUTHORS: A S Denton, H J N Andreyev, A Forbes, E J Maher ABSTRACT: Chronic radiation proctitis produces a range of clinical symptoms for which there is currently no recommended standard management. The aim of this review was to identify the various non-surgical treatment options for the management of late chronic radiation proctitis and evaluate the evidence for their efficacy. Synonyms for radiation therapy and for the spectrum of lower gastrointestinal radiation toxicity were combined in an extensive search strategy and applied to a range of databases. The included studies were those that involved interventions for the non-surgical management of late radiation proctitis. Sixty-three studies were identified that met the inclusion criteria, including six randomised controlled trials that described the effects of anti-inflammatory agents in combination, rectal steroids alone, rectal sucralfate, short chain fatty acid enemas and different types of thermal therapy. However, these studies could not be compared. If the management of late radiation proctitis is to become evidence based, then, in view of its episodic and variable nature, placebo controlled studies need to be conducted to clarify which therapeutic options should be recommended. From the current data, although certain interventions look promising and may be effective, one small or modest sized study, even if well-conducted, is insufficient to implement changes in practice. In order to increase recruitment to trials, a national register of cases with established late radiation toxicity would facilitate multi-centre trials with specific entry criteria, formal baseline and therapeutic assessments providing standardised outcome data.British Journal of Cancer () , –. doi:./sj.bjc. www.bjcancer.com© Cancer Research UK BODY: Pelvic radiotherapy is an essential component of curative therapy for many cancers of the urological, gynaecological and gastrointestinal tracts. In the treatment of pelvic malignancy there is a need to minimise the risk of chronic radiation injury to normal tissue without compromising the possibility of cure. The gastrointestinal tract and in particular the rectum, is the site most frequently affected. The prevalence of severe toxicity such as fistulation, stricture formation, transfusion-dependent bleeding or secondary cancers after pelvic radiotherapy is unknown but has been estimated to be % at years (Nostrant, ) although this may be an underestimate (Komaki et al, ; Tan et al, ; Denton et al, ). Much more common may be lesser side effects such as diarrhoea, urgency, faecal incontinence and tenesmus which in recent studies have been suggested to impair quality of life considerably (Kollmorgen et al, ; Schultheiss et al, ).Acute radiation proctitis refers to radiation-induced injury of the rectum, during or within months of radiotherapy. Chronic radiation proctitis can continue from the acute phase or begin after a latent period of at least days (median time – months) (Eifel et al, ) and may be more common in those who develop severe acute proctitis (Denham et al, ), and in those with diabetes, inflammatory bowel disease, hypertension or peripheral vascular disease. Most of these risk factors have been defined by studies dependent on suboptimal assessment of rectal toxicity. Other factors which may increase the risk of late complications include tumour stage, and the total dose and fractionation of radiation, although the severity of radiation damage may not be entirely dependent on the radiation dose (Schwarchuk et al, ).Whether a patient as developed chronic proctitis may be assessed in various ways. Changes in symptoms such as irregularity of bowel dysfunction, rectal blood loss or pain may be graded according to criteria stated in systems such as LENT SOMA and the Franco-Italian Glossary (Chassagne et al, ). The criteria for the grade of severity varies with the scoring system used, emphasising the need for a single universally agreed measure, but there are no described specific radiological features which define radiation proctitis (Capps et al, ). Endoscopically, proctoscopy, rigid sigmoidoscopy or flexible sigmoidoscopy may be used but it is recognised that these different methods may not produce reproducible findings between observers (Baron et al, ). Tissue biopsy may be inconclusive. The few small longitudinal studies of ano-rectal physiological parameters following RT that have been published are contradictory (Iwamoto et al, ; Yeoh et al, ).In other inflammatory conditions affecting the bowel, symptoms and measures of disease activity may correlate poorly (Yoshida, ) and there has been little research to determine how endoscopic or radiological appearances relate to bowel function. Severe fibrosis with stricturing may provide incontrovertible evidence of radiation-induced damage but radiation-induced bowel changes are also seen in patients without symptoms and sometimes it can be hard to attribute the patients' bowel habit directly to physical damage caused by radiation to the rectum.New gastrointestinal symptoms may be triggered for many reasons. Firstly, therefore, in the patient who has undergone radiotherapy, it may be difficult to untangle the relative contribution of psychological and physical damage to bowel function. Secondly, when there is severe physical radiation-induced damage present in the rectum, it is possible that patients will also have other sites of bowel damage, perhaps leading to bile salt or fat malabsorption. Thirdly, published series may not represent a true cross section of irradiated patients. There is some data suggesting that the reported number of cases with chronic radiation proctitis and a proportion of patients who seek help for subsequent symptoms is a fraction of the true prevalence (Yeoh and Horowitz, ; Ooi et al, ). Experience from both oncological and gastroenterology practice has long shown that unless directed questions are asked and exact answers pursued that patients may not tell their physicians about their specific gastrointestinal symptoms, particularly if they feel the physician is not in a position to do anything about them. The medical treatment of radiation proctitis is not clearly defined and management is often difficult. This is in part, due to problems establishing the diagnosis and also because a proportion of the biological changes may not be reversible.The aim of this study was to examine systematically the non-surgical treatments that have been proposed for this condition and the quality of the evidence that suggests those treatments may be efficacious.MATERIALS AND METHODSCriteria for considering studies for this review: Types of studies Randomised studies were included preferentially for analysis. Conclusions from the non-randomised, non-controlled data were drawn if there was insufficient evidence from randomised controlled trials.Types of participants Patients must have been diagnosed with a pelvic malignancy and undergone pelvic radiotherapy as part of their treatment schedule, subsequently developing radiation proctitis of any grade, continuing from completion of radiotherapy for more than three months, or occurring more than three months after completion of radiotherapy.Interventions Studies were included which described the use of non-surgical interventions for the treatment of late radiation proctitis and the specific intervention was used either vs a placebo/nothing or against other therapies.Search strategy for identification of studies: Concepts Synonyms for radiation therapySynonyms for the spectrum of lower gastrointestinal radiation toxicityConcepts A and B combined with the Boolean operator ‘AND’ A filter was not used because any type of study was considered. This basic strategy was expanded for text and MeSH terms before being applied to a sequence of databases. Relevant studies were identified and the inclusion criteria were then applied.Databases In order to be as comprehensive as possible search strategies were employed to identify all relevant studies irrespective of language. First, electronic databases were searched. The time frame used was from April back to (Table 1Table 1Databases searched for this review). Secondly, reference lists of identified studies and the relevant chapters of major oncology and gastroenterology text books were searched. Thirdly, groups and individuals likely to hold unpublished data were approached (Table 2Table 2Groups and individuals contacted for further unpublished data related to this topic).Methods of the review: A final list of all potential relevant articles was created in one core database and were assessed independently by two reviewers to determine whether they complied with the preceding inclusion criteria for this the review. Where differences existed they were resolved by consensus and when necessary in consultation with a third reviewer. Included studies were graded according to the criteria used by the NHS executive (Table 3Table 3Grading criteria used by the NHS Executive).Statistics: Dichotomous data is expressed as the odds ratio. Uncertainty in each treatment is expressed using % confidence intervals (% CI). Continuous data, i.e. symptom scores, were converted to the weighted mean differences and an overall weighted mean difference was calculated with standard errors. The Cochrane Review Manager software RevMan . was used for estimation of overall treatment effects/meta-analysis of results. Both fixed and random effects models were used to calculate a weighted average of the treatment effects across the studies under review. Sensitivity analyses as well as including study quality (quality of allocation concealment) also included year of publication, the type of outcome measures, and random and fixed effects models if appropriate.RESULTSUsing the search strategy described, studies were identified, of which met the inclusion criteria, and included three randomised controlled trials (Table 4Table 4Results of literature searches).Studies using anti-inflammatory agentsFirst line treatment of this condition is conventionally with anti-inflammatory agents. Seven studies were identified which used such agents. The highest level of evidence was provided by three randomised controlled trials (grade IC).Kochhar et al, , Grade IC, IndiaRadiation-induced proctosigmoiditis. Prospective, randomised, double-blind controlled trial of oral sulfasalazine plus rectal steroids vs rectal sucralfate (n=) There were females treated for cervical cancer and one male treated with prostate cancer. The mean duration from the completion of radiotherapy was . months. Exclusion criteria were the use of steroids in the last weeks. Symptoms were assessed using an in-house scoring system for diarrhoea, bleeding, tenesmus and endoscopic appearance. According to the score the cases were then graded. Baseline characteristics in both groups were comparable. Patients were randomised to either, rectal prednisolone mg b.d. and oral sulfasalzine mg t.d.s. (n=) or rectal sucralfate suspension g b.d. with oral placebo t.d.s. for sulfasalzine (n=). There were three drop-outs in the anti-inflammatory group and two in the sucralfate group. Treatment was continued for weeks.Eight out of in the anti-inflammatory group showed clinical improvement compared to out of in the sucralfate group. Seven out of in this group showed endoscopic improvement compared to out of in the sucralfate group, and the degree of improvement was greater in the sucralfate compared to the anti-inflammatory group. The odds ratio for clinical improvement was . (% CI .–.). No difference in endoscopic appearance was detected between the sucralfate and the anti-inflammatory groups, odds ratio . (% CI .–.). The response was only reported for the week follow-up period. There was no quality of life assessment.Rougier et al, , Grade IC, FranceRectites radiques: efficate comparee de deux types de corticoides adminstre localement (n=) (Radiation proctitis: comparing the efficiency of local administration of two types of corticosteroids) There were females and three males, treated with radiotherapy for gynaecological malignancies and nine colorectal or anal tumours. All had completed radiotherapy a minimum of months ago. The diagnosis of radiation proctitis was graded on flexible sigmoidoscopy and on the degree of bleeding. Two rectal steroid preparations were compared, mg betamethasone enema b.d. (n=) or mg hydrocortisone acetate mousse b.d. (n=). The outcomes used were bowel activity, rectal bleeding, tenesmus with endoscopic grading. They were assessed clinically and endoscopically at and days. Two patients were lost to follow-up from the betamethasone group, and no treatment related complications were reported.Over the weeks of treatment the endoscopic appearance improved in a greater proportion of the hydrocortisone group ( out of ) than the betamethasone group ( out of ), odds ratio . (% CI .–.). There was no difference in the reduction of bleeding between the hydrocortisone group ( out of ) and the betamethasone group ( out of ), odds ratio . (% CI .–.). The response was only reported for the week follow-up period. Potential bias in this study includes more severe intial disease in the betamethasone group and also that the betamethasone enema was poorly tolerated in out of compared with out of in the hydrocortisone group. There was no quality of life assessment.Cavcic et al, , Grade IC, CroatiaMetronidazole in the treatment of chronic radiation proctitis: clinical trial (n=) Sixty patients with rectal bleeding and diarrhoea were allocated to treatment with metronidazole (× mg orally per day), mesalazine (× g orally per day), and betamethasone enema (daily) or mesalazine and betamethasone enema, but without metronidazole, for year. The efficacy of metronidazole was assessed on the basis of rectal bleeding, diarrhoea, and rectosigmoidoscopy in all patients. The incidence of rectal bleeding and mucosal ulcers was significantly lower in the metronidazole group at weeks (P=.), months (P=.), and months (P=.). There was also a significant decrease in diarrhea and oedema in the metronidazole group, after weeks of treatment (P=.), months (P=.), and months (P=.). One year after treatment, out of of the metronidazole group demonstrated a reduction in the grade of their rectal bleeding compared to out of in the group treated with mesalazine and betmethasone, odds ratio . (% CI .–.). Similarly out of in the metronidazole group compared to out of had experienced reduction in their diarrhoea and rectal erythema, odds ratio . (% CI .–.). The rectal ulceration at year had decreased in out of of the metronidazole group compared with out of of the group treated with anti-inflammatories alone, odds ratio . (% CI .–.). No side effects were reported.Other studiesIn addition, four non-randomised studies were identified. One prospective cross-over trial of six patients, 5ASA enemas vs betamethasone enemas, showed no significant benefit for either treatment (Triantifillidis et al, ). Three series reported the effect of various anti-inflammatory agents: sulfasalazine enemas nightly for – months in a group of four patients (Baum et al, ) were not effective. Oral sulfasalazine, for year, in a group of four patients (Goldstein et al, ) demonstrated an improvement from the baseline symptoms as did daily oral oestrogen and norethisterone for month in one case of refractory radiation proctitis (Wurzer et al, ).Studies using SCFA enemasShort chain fatty acids (SCFA) are the main oxidative fuel of colonic mucosa and their use may be impaired in chronic radiation proctitis. SCFA are predominantly produced in the colon by anaerobic bacterial fermentation of non-absorbed carbohydrates, in dietary fibre. Butyric acid is the most important. SCFA also exert a trophic effect on the colonic mucosa by stimulating the physiological pattern of proliferation and promoting cellular differentiation. In the setting of radiation induced ischaemia, the associated mucosal atrophy may interfere with mitochondrial fatty acid oxidation so supplementation of SCFA in the form of enemas could overcome this deficiency and improve the energy supply to colonocytes. Moreover, the dilator effect of SCFA on arteriolar walls may improve mucosal blood flow. Five reports of four studies were identified in this section. Two were randomised studies.Talley et al, , grade IC (Australia)Short-chain fatty acids in the treatment of radiation proctitis: a prospective randomised, double-blind, placebo-controlled, cross-over pilot trial (n=) Two females and males; prostate, one cervix and two rectum, treated with pelvic radiotherapy a mean period of . months earlier were assessed using an in-house symptom score (six items: rectal pain, rectal bleeding episodes, quantity of blood, days of diarrhoea, number of stools and urgency). Endoscopic and histological scores were also calculated. The patients were randomised to receive either a normal saline placebo enema or a ml enema containing mM of butyrate administered twice a day. Each treatment was prescribed for weeks with a week washout period between treatments. Three patients failed to complete the course. The total symptom score at baseline ranged from – (median .). There was a non-significant improvement in symptom scores on the active treatment, mean score . (range –) compared with . mean score (range –) for those receiving placebo. The results are published in the form of the median value (plus interquartile range) for the average placebo period and active period and the raw data is required from the authors before the results can be dichotomised. There was no quality of life assessment.Pinto et al, , grade IC, (Portugal)Short chain fatty acids are effective in short term treatment of chronic radiation proctitis. Randomised, prospective double blind, controlled trial (n=) Nineteen patients (one male and females) with grade III chronic radiation proctitis were randomised to SCFA enema, ml b.d. for weeks (n=) or a placebo (n=). Baseline characteristics of each group were comparable. The specific outcome variables monitored for a period of months were: adverse effects, number of days of rectal bleeding in the previous week and haemoglobin. The endoscopic appearance was scored for: hyperaemia and neovascularisation, friability, oedema and erosions by two independent assessors who were blinded to the treatment. Biopsies for mucosal DNA and protein content were also measured.One patient from the SCFA arm and two from the placebo arm did not complete the trial. At weeks the patients treated with SCFA had a significant reduction (. to .) in the number of days per week of rectal bleeding (P=.) whereas in the placebo group it fell from . to . (P=.). Haemoglobin levels at the end of the treatment period were higher in the SCFA group, .±. vs .±. g dl− (P=.) and the endoscopic scores had decreased significantly in both groups but were significantly lower in the SCFA group (P=.). Changes in DNA and protein concentration decreased in both groups but significantly more so in the placebo treated group (P=.). In long term follow-up two patients failed to complete the course from the placebo group because of severe bleeding leaving nine in the SCFA arm and five in the placebo group. At the end of the months the endoscopic score was similar in the two groups. No adverse events related to either of the groups were noted and there was no quality of life assessment. As a consequence of the way the data has been reported we have presented this data as a weighted mean difference (WMD) which for endoscopic scores was (−. to .), a statistically non-significant difference between SCFA and placebo. For the number of days of rectal bleeding per week at the end of the treatment period the WMD was − (−. to −.), again a statistically non-significant difference, but perhaps showing a trend to less days of rectal bleeding with the SCFA enema.Other studiesIn addition, two non-randomised series were identified. Mamel et al () (IIIC), in an open study of six cases using SCFA reported a significant improvement in clinical and mucosal response although this is not statistically assessed. Al-Sabbagh et al () (IIC) in their prospective pilot study of seven cases demonstrated a significant improvement for rectal bleeding but not endoscopic appearances. The duration of the response is not stated other than for the treatment period of months in Al-Sabbagh et al (). No toxicity was reported in any of the studies. Quality of life data were not presented.Although there were two trials in this category using the same intervention against a placebo, they are not directly comparable due to differences in the outcome measurements and method of data presentation. Talley et al () was a cross-over study and used a cumulative score encompassing the number of episodes of rectal bleeding per week. Pinto et al () recorded the number of days on which rectal bleeding occurred. Although both sets of cases were scored endoscopically, this is not on the same scale so that the numbers are not comparable. Additionally, regardless of the different scales, because of the absence of individual patient data we were unable to dichotomise and combine the data. Talley et al () reported no benefit from the use of SCFA (but the treatment period of weeks was short), and Pinto et al () also reported no significant improvement either endoscopically or on rectal bleeding. It is not surprising that neither of these trials yielded any significant results. These numbers are small for any type of study but especially for prospective, randomised ones and therefore statistically underpowered.Studies using sucralfate and pentosan polysulphateSucralfate is a highly sulphated polyanionic dissacharide. In this setting, its postulated mechanisms of action include stimulation of epithelial healing and the formation of a protective barrier. Pentosan polysulphate (PPS) is a synthetic derivative of a glycosaminoglycan which is present in the surface of the bladder, vessels and the gastrointestinal tract lining. The ability of PPS to reduce epithelial permeability and prevent adherence has been extrapolated to the large bowel. Eight relevant references were identified addressing the effect of these agents in chronic radiation proctitis. One is a RCT, two prospective open studies, and four studies describe the effect of rectal sucralfate in case reports or small series with one report of the effect of oral sucralfate.Kochhar et al, , Grade IC, IndiaRadiation-induced proctosigmoiditis. Prospective, randomised, double-blind controlled trial of oral sulfasalazine plus rectal steroids vs rectal sucralfate (n=) The strongest evidence for the use of sucralfate comes from the evidence presented in this trial, which compares the use of anti-inflammatories with rectal sucralfate. Criticisms of this study are that the allocation concealment is not clear and there is no explanation for those cases lost to follow-up. The endpoints used to evaluate response, i.e. scores for clinical and mucosal effect, are useful and show that the odds ratio for the beneficial effect of sucralfate on clinical outcomes is . (% CI .–.) and . (% CI .–.) for mucosal response. The duration of follow-up was only weeks.Other studiesIn addition seven non-randomised studies were identified. In two prospective studies Kochhar et al (), followed cases treated with rectal sucralfate, and Grigsby et al () described cases treated with oral PPS, IIC. Both are well conducted with definitive outcomes of response and show a benefit for the use of rectal sucralfate for a period of months and oral pentosan for year, respectively. Of the retrospective reports (four describe the use of rectal sucralfate as a case report (Kochhar et al, ), and series of three, eight and three patients respectively (Ladas and Raptis, ; Kochhar et al, ; Stockdale and Biswas, ). All showed a benefit in both clinical and mucosal outcome, although the majority of these were not graded or scored to demonstrate significant differences. One report outlines the beneficial effect of oral sucralfate in established chronic radiation proctitis in three isolated cases for a minimum of years (Sasai et al, ). Morbidity related to sucralfate was not reported and there was one case of a rash with oral pentosan. The treatment intervals were generally short and none provide any quality of life data.Formalin therapyIn radiation proctitis, vascular telangiectasia and non-healing mucosal ulceration, perhaps due to an underlying obliterative arteritis, may lead to severe recurrent haemorrhage. Formalin may sclerose and seal fragile neovasculature in radiation damaged tissues preventing further bleeding. Application directly to the mucosa produces local chemical cauterisation. It can stop bleeding by sealing the neovascularised telangiectatic spots and ulcers. The success of bleeding control is related to the accurate localisation and application of formalin to all the affected points.We identified references relating to the use of rectal formalin. Three were prospective case series (level IIC evidence) and were retrospective reports (level IIIC). These reports are quite heterogenous (Table 5Table 5Results of formalin reports). None used a control group or quoted any quality of life data. The severity of radiation proctitis is graded in only one report. The technique and the concentration of formalin used vary. The two main approaches use irrigation of formalin (five reports), either .% formalin solution (n=) or % formalin solution (n=). The other method (described in reports) is the direct application of gauze soaked in formalin, and although most reports used % formalin (n=) there was one report that used % solution. None of the reports used an objective scoring system to assess the response to treatment, but two reports described the effect in terms of changes of haemoglobin level. The remainder described the response in terms of the transfusion requirements or the effect on bleeding. Statistical analysis was not presented in any of these reports.In summary, patients were described with a mean follow-up of months. In each of the reports there appeared to be benefit from the use of formalin in chronic haemorrhagic proctitis, but this may be subject to reporting bias. Serious side effects were reported in patients with five cases of anal ulceration, two rectal strictures, two patients with faecal incontinence and two with anal pain. Minor side effects were not frequently reported. Duration of effect cannot be assessed reliably from the data available but appears to be a minimum of months. The absence of quality of life data means the impact of this treatment from the patient's perspective cannot be addressed.Thermal coagulation therapyEndoscopic coagulation with a variety of devices has been reported to be effective for the control of radiation-induced bleeding. The technique generally used is coagulation of focal bleeding telangiectasia rather than the entire friable mucosa. Several treatment sessions are often required. Scarring and re-epithelisation with more normal tissue tend to occur over time.Sixteen relevant studies were identified. Apart from one RCT and one prospective series the remaining are retrospective series or reports. Although these case series all refer to the use of ablative therapy in late radiation proctitis, different types of coagulation probes or lasers are described. Eight look at the use of the YAG laser, four at Argon lasers, three at bipolar cautery and two at heater probes. The results of these studies are described in Table 6Table 6Results of reports for the effects of thermal therapy in late radiation proctitis.Jensen et al, , (USA) grade ICA randomised prospective study of endoscopic bipolar electrocoagulation and heater probe treatment of chronic rectal bleeding from rectal telangiectasia (n=) Selection criteria included pelvic RT at least years earlier, rectal bleeds at least three times per week, anaemia, having failed year of medical therapy and consideration for surgery, and a life expectancy of years. Chronic radiation proctitis was confirmed endoscopically but not graded in cases ( with carcinoma of the prostate and three with cervical cancer). Patients were randomised to treatment with a heater probe (n=) or a bipolar electrocoagulation probe (n=). Treatment sessions were repeated with the same probe till the bleeding resolved and a mean of four sessions per case were administered for both probes. The physicians who subsequently assessed the patients were blinded. All other treatments were discontinued. Assessment was repeateded every – weeks until bleeding stopped and then every – months for year. If new telangiectasia were noted they were retreated by the same probe.In both the bipolar probe and the heater probe the mean fall in severe bleeds per case was statistically significant at P<.. In the months of endoscopic treatment vs months medical therapy the severe bleeding episodes (defined as a bleed provoking and unscheduled hospital assessment) diminished significantly for the bipolar probe ( vs %) and heater probe ( vs %). The weighted mean difference did not show a significant difference between the two treatments . (% CI −. to .). The reduction in the units of blood transfused per case pre- and post-treatment was only statistically significant in the group treated with the heater probe and the weighted mean difference reflected this (−.; % CI −. to −.). The increase in the haematocrit compared with the baseline was statistically significant for both groups and comparing the effects of the two groups the weighted mean difference favoured the heater probe (−.; % CI −. to −.). During follow-up endoscopy there was resolution of the telangiectasia, scarring or epithelial replacement in all cases in both groups. Patient interviews before, and months after treatment revealed that rectal bleeds, tenesmus and their general health had improved. There were no recorded side effects.Other studiesNone of the references in this section state the grade of the proctitis and one study (Silva et al, ) uses a formal score to quantify rectal bleeding pre- and post-treatment allowing an objective comparison to be conducted. Generally, outcome assessments used are presence or absence of bleeding, control of bleeding, haemoglobin pre- and post-treatment and transfusion requirements. The point at which transfusion occurs is rarely described making transfusion requirements and haemoglobin values relative measures. The impression is that thermal coagulation therapy has a useful role in haemorrhagic radiation proctitis that is refractory to other medical treatments. This cannot be statistically supported because of the quality of the reports and the size of the trials.Hyperbaric oxygen therapyHyperbaric oxygen therapy (HBO) has an angiogenic effect and has been shown to cause an eight- or nine-fold increase in the vascular density of soft tissues over air-breathing controls. The subacute and chronic phases of radiation wounds are particularly suited to this form of therapy. HBO acts to stimulate collagen formation at the wound edges through elevation of local tissue oxygen tensions. New microvasculature dependent on collagen matrix, is greatly enhanced in this setting and allows re-epithelialisation to occur.Nine relevant references were identified, reporting treatment in cases that were followed up for a mean period of months (Table 7Table 7Hyperbaric oxygen results). Eight were retrospective series or reports and there was only one prospective observational case series (Williams et al, : level IIC). All of these reports were case series with heterogeneous characteristics. In only two of these reports (Feldmeier et al, ; Gouello et al, ) were there full baseline assessments of the degree of chronic radiation proctitis with a score or a grade of the histological or symptomatic features. Also the pressures of HBO and duration of the treatment sessions varied. The number of treatments depended on degree of the lesion. Assessment of response was usually with a vague description of the resolution of symptoms but not parameters that could be scored and used for statistical analysis. The duration of response was inconsistently recorded. Quality of life data was not recorded in any report. Side effects were reported in eight cases but were largely transient, minor and related to aural barotrauma. Therefore, although the impression is that HBO may be of value for large bowel chronic radiation changes that are refractory to other treatments, the degree of benefit and the cumulative effect or duration of response cannot be quantified because of the methodology and quality of the data.Miscellaneous interventionsOxidative stress is thought to be a major mechanism in the development of chronic radiation proctitis and agents with anti-oxidant properties have been used in an attempt to limit tissue damage in radiation injury. We identified one series investigating the effect of vitamins C and E in the treatment of cases with established chronic radiation proctitis who presented with one or more symptoms including rectal bleeding pain urgency or tenesmus. The severity and frequency of symptoms and a score of the lifestyle impact were used to assess response before and after treatment for year and all reported a sustained improvement in their initial symptoms without side effects (Kennedy et al, ).Strictures of the rectosigmoid junction and rectum are a recognised consequence of late radiation damage. The narrowing results in acute or episodic periods of large bowel obstruction, often decades after the original radiation therapy. Non-surgical dilatation has been attempted. Three relevant studies were identified and included two case reports of the use of a dilator (Triadafilopoulos and Sarkisian, ) or stent (Yates and Baron, ) and one series of four cases which were endoscopically dilated (Johansson, ). Although the strictures are described they are not scored and the absence of a formal baseline assessment and objective response means that the effect which appears to be beneficial in each report cannot be quantified, nor can the duration of response be determined from the data available. Side effects include one case of brief post dilatation bowel pain and another case of perforation. Quality of life issues are referred to in one report where the patients' general health was felt to improve as a result of the treatment (Yates and Baron, ).DISCUSSIONThis review of the literature for non-surgical interventions for the management of late radiation proctitis has identified a number of treatment approaches supported by varying levels of evidence. Six controlled studies are reported, three relate to the comparison of different rectal steroid preparations and the comparison between anti-inflammatory agents and rectal sucralfate, two use short chain fatty acid enemas and one contrasts the effect of bipolar electrocoagulation vs the heater probe. These show that the use of sucralfate may be better than anti-inflammatory agents, which in turn may have greater effect if used with metronidazole. Rectal hydrocortisone may be better than betamethasone.The reports describe slightly different interventions and outcome parameters so that they cannot be compared and a summary statistic cannot be derived. Ironically the wealth of case series although of interest, can only give an impression of effect and cannot be collectively summarised to produce a cumulative response rate or duration of effect. The overwhelming feature of the data presented here is that the majority of references describe one individual's or one centre's experience of a specific intervention administered without comparison to a control or another agent. In many cases, it is not clear how extensively patients have been investigated and whether other causes for their symptoms have been excluded. The reasons for the paucity of controlled studies may relate to the relative rarity of late radiation proctitis and the logistic difficulties that exist in compiling a series large enough to be randomised between therapies.There were a number of specific problems with the details available in many of the identified reports. First, few background details were available with particular respect to the tumour stage and radiation details. Secondly, the method of determining the diagnosis was only occasionally documented. Late radiation proctitis may be less easy to grade accurately on rigid sigmoidoscopy compared to flexible sigmoidoscopy and there is also anecdotal evidence that some of the bowel preparations commonly used for flexible endoscopic examinations may exacerbate the changes seen. None of the endoscopic scoring systems take these factors into consideration but clearly the mode of assessment needs to be recorded for subsequent comparisons. Thirdly, there was rarely a formal baseline assessment of the symptoms and where this was stated the scoring systems were often different so that the grade could not be used as a comparison with other cases. Fourthly, the outcome measures, where used, were variable as was the scoring of the response when performed. The duration of the responses and the side effects of treatments were not always stated. Quality of life scores either before or after the intervention were hardly ever recorded. Deterioration in presenting symptoms may not always relate to failure of the intervention but may be a consequence of tumour or disease progression although this is often difficult to determine as the two scenarios are often indistinguishable. Therefore specific details regarding the method of ascertaining a failed response need to be stated. Finally, interventions were not standardised so that there were substantial variations in dose and administration. These combined factors serve to produce a very heterogeneous cohort of reports that cannot be scored or graded to produce subclasses and objective response rates unless more information is available. A proposed scheme to address all these issues might include:A detailed account of the determination of the diagnosis including background details such as precise radiation prescription and absence of infection.Scored symptoms as in Ulcerative Colitis (UC) with the Colitis Activity Index (Rachmilewitz, ) or the Ulcerative Colitis Scoring System (Schroeder et al, ), before and after the specific intervention.Scored endoscopic appearance (as for UC e.g. Baron score (Baron et al, ) and the Colitis Endoscopic index (Rachmilewitz, )), before and after the specific intervention.Scored histological assessment (as for UC) which should be standardised.The outstanding issue is the degree to which radiation induced damage to the rectum contributes to symptoms in this group of patients and how much comes from residual tumour, psychological factors, co-incident small intestinal damage or other changes within the abdomen and pelvis outside the bowel. Some of the areas that need to be addressed include:How should late radiation proctitis be staged and how should endoscopic, radiological, physiological and quality of life assessments be combined to assess the patient's needs best?Who needs intervention?What treatments are effective?Is there an optimal step up treatment approach?How should patients be followed up after treatments, how long for and what parameters should be measured subsequently?Implications for practice and researchIf the management of late radiation practice is to become evidence based then good quality placebo-controlled studies need to be conducted to support the treatment options recommended. This review suggests that late radiation proctitis is not reported very often by patients to the clinicians who deliver the pelvic radiotherapy and as a result a number of fundamental issues remain to be clarified. First, the true incidence of the disease is not clear. Therefore, physicians caring for patients who have undergone pelvic radiotherapy need to be more aware that this group may develop problematic symptoms which may need detailed questioning to elicit and which may require specialist gastroenterological assessment to characterize in detail. Secondly, there is an urgent need to define clearly the diagnostic criteria and a unified grading system by which late radiation proctitis can be categorised. Without such a system, it is unlikely that meaningful randomised studies, particularly in a multi-centre setting, can be designed.We propose that cases should be enrolled into regional or centralised registers of radiation toxicity or that all such patients should be referred to regional centres with an interest in radiation-induced gut toxicity. In this way terminology, baseline assessments including comorbidity and the documentation of therapeutic effect could be standardised. Interventions could be randomised and outcome data could be pooled to assess the response to treatment objectively. This approach would provide an evidence base of results of different treatments to develop an integrated care pathway for this difficult condition.
|
PMC2002207.txt
|
TITLE: Evaluation of desmin as a diagnostic and prognostic marker of childhood rhabdomyosarcomas and embryonal sarcomas. AUTHORS: P. Dias, P. Kumar, H. B. Marsden, P. H. Morris-Jones, J. Birch, R. Swindell, S. Kumar ABSTRACT: The diagnostic and prognostic relevance of desmin expression in rhabdomyosarcomas (RMS) and embryonal sarcomas (ES) was examined using a peroxidase anti-peroxidase staining procedure. Fifty-nine RMS but only one ES stained for desmin (P less than .). The maximum percentage of desmin containing cells was in RMS compared with only % in ES. Desmin positivity correlated inversely with survival (P less than .) in that RMS with high proportions of desmin positive cells were associated with poorer prognoses than those containing fewer desmin positive cells. If the degree of expression of desmin is related to myogenic differentiation, then our results indicate that poorly differentiated RMS tend to have a better prognosis than the well differentiated tumours. One possible explanation is that the poorly differentiated RMS respond better to chemotherapy than to well differentiated RMS. A multivariant analysis incorporating desmin staining, treatment, histology, age and gender revealed that the two most significant independent prognostic factors were treatment and histology.ImagesFigure BODY:
|
PMC1513519.txt
|
TITLE: T Wave Alternans And Ventricular Tachyarrhythmia Risk Stratification: A Review AUTHORS: Masahiko Takagi, Junichi Yoshikawa ABSTRACT: Sudden cardiac death (SCD) is one of the leading causes of mortality in industrialized countries. Thus, identifying patients at high risk of SCD is an important goal. T wave alternans (TWA) is a new method for identifying patients with lethal ventricular tachyarrhythmias, and is dependent on heart rate. The maximal predictive accuracy is achieved at heart rates between and bpm, so that TWA is usually measured during exercise, phamacological stress, or atrial pacing. It has been shown that TWA has high sensitivity and negative predictive value for predicting SCD after myocardial infarction and is also useful for predicting SCD in patients with nonischemic cardiomyopathy. Although the implantable cardioverter defibrillator (ICD) is now the primary therapy for preventing SCD, it is difficult to identify those patients who are susceptible to lethal ventricular tachyarrhythmias for primary prevention. In the prediction of SCD, TWA can be used as a screening test of appropriate patients for further electrophysiological examination and therapy. BODY: Sudden cardiac death (SCD) is a leading cause of mortality and remains a major clinical challenge []. Concerning therapeutic modalities, tremendous progress has been made in the development of the implantable cardioverter defibrillator (ICD). However, progress in identifying patients at high risk of SCD has lagged behind. Large multicenter trials have shown that electrophysiologic study (EPS) may be useful in identifying patients who would benefit from ICD therapy [,]. Unfortunately, EPS is costly, invasive, and imperfect []. Several noninvasive markers of risk-stratification have been studied and compared with EPS. Left ventricular ejection fraction (LVEF), frequent ventricular premature complexes (VPC) on Holter recording, and ventricular late potentials (LP) are all sensitive, but of low specificity and positive predictive value (PPV) [,]. Measurement of heart rate variability, especially in combination with LVEF, VPC, and LP, has significantly improved risk prediction, but PPV remains low []. A screening procedure that is more sensitive and specific - with high PPV for identifying patients at high risk of developing ventricular tachyarrhythmias - is needed.Recently, assessment of repolarization alternans (T wave alternans [TWA]) in the electrocardiogram (ECG) has been suggested as a predictor of susceptibility to lethal ventricular tachyarrhythmias [,]. Although overt TWA in the ECG is not common [], digital signal processing techniques capable of detecting subtle degrees of TWA (microvolt TWA) have shown that TWA may represent an important marker of vulnerability to ventricular tachyarrhythmias. This review discusses the electrophysiologic mechanisms that link TWA to arrhythmogenicity and the recent clinical data for its prognostic efficacy in predicting lethal ventricular tachyarrhythmias.History of TWATWA was first described in ; it is the variation in vector and amplitude of the T wave that occurs on an every-other-beat basis []. In a review by Kalter in , patients were identified with macroscopic TWA - a frequency of .% []. In humans, macroscopic TWA has been associated with increased vulnerability to ventricular tachyarrhythmias under several pathophysiologic conditions such as myocardial ischemia [-], vasospastic angina [,], marked electrolyte abnormalities [,], hypertrophic cardiomyopathy (HCM) [], the long QT syndrome [-], and the Brugada syndrome [,]. Microscopic TWA was first reported in []. Thereafter, many studies have led to the development of the method for detecting microvolt TWA and to the establishment of a relationship between TWA and susceptibility to ventricular tachyarrhythmias, in humans [,].Pathophysiology of TWAIonic currents and TWAThere is some evidence that TWA is linked to alternations in cellular calcium homeostasis, which significantly influence the action potential duration (APD) []. In the failing heart, electrical remodeling is a recurring feature that has been associated with an increased risk of SCD []. The arrhythmogenesis is due to the functional expression of proteins to control calcium homeostasis.Potassium channels may also play an important role in ischemia-induced TWA. The different sensitivity of KATP channel activation during ischemia between epicardium and endocardium may be linked to TWA at the cellular level [-].Arrhythmogenesis and TWACurrently, the hypothesis regarding the arrhythmogenic mechanisms associated with TWA is based on the concept that heterogeneous prolongation and increased dispersion of repolarization produce reentrant ventricular tachyarrhythmias []. The heterogeneity in dispersion of repolarization results in a : appearance on the surface ECG and provides conduction block in the areas with prolongation of repolarization, which fractionates the wavefront and facilitates reentry. Shimizu et al. [] studied an experimental model of long QT syndrome utilizing an arterially perfused wedge of canine left ventricular wall. When the preparation was paced at a critical fast rate, there was pronounced alternation of APD of mid-myocardial (M) cells, resulting in a reversal of the transmural repolarization sequence leading TWA in the unipolar ECG . The alternation of APD of M cells observed under long QT conditions may exaggerate transmural dispersion of repolarization and develop torsade de pointes. Pastore et al. [] investigated TWA in Langendorff-perfused guinea pig heart using mapping of epicardial APD during pacing. The critical pacing rate induced concordant TWA, which developed to discordant alternans of APD and increased susceptibility to ventricular tachyarrhythmias.Measurement of microvolt TWADetection of microvolt TWA has been made possible by the use of advanced signal processing techniques and high-resolution electrodes to reduce noise. A number of beats, generally , are sampled and a time series of amplitudes of multiple corresponding points on the T wave are analyzed using a Fast Fourier Transform to generate a power spectrum (Figure ). Several frequency peaks correspond to respiratory variation, pedaling (if bicycle exercise is performed), and noise. The presence of alternans is indicated by a frequency peak at . cycles per beat []. The analysis yields two measurements: the alternans magnitude and the alternans ratio. The former represents the magnitude of the alternating variation in T wave morphology compared to the mean T wave; a conventional threshold of . μV is used for significance. The latter is a measure of the statistical significance of the alternans with respect to the standard deviation of the background noise; it is generally required to be greater than , for significance. Furthermore, by definition, TWA must be sustained for more than minute [].TWA is a rate-dependent phenomenon and microvolt TWA could develop in normal subjects at sufficiently high heart rate. It has been shown that the onset heart rate is relatively low in patients with structural heart disease and history of sustained ventricular tachyarrhythmia []. Kavesh et al. [] showed that both TWA and false positive results increase with heart rate. Therefore, an onset heart rate of less than beats per minute is a conventional requisite for positivity.The original study of TWA was performed with atrial pacing to increase heart rate []. Either bicycle or treadmill exercise is now available with the use of high-resolution electrodes and advanced noise reduction algorithms. However, noise, premature beats, rapid changes in heart rate, or prominent beat-to-beat variability of RR intervals, may all mask true alternans []. All of these factors - plus the failure to achieve target heart rate - may result in an indeterminate test of TWA.Clinical studies of TWA for ventricular arrhythmic risk stratificationThe first large clinical study of TWA was performed by Rosenbaum and co-workers [], who observed patients undergoing both EPS and TWA measured during atrial pacing. Over the following months, ventricular tachyarrhythmic events occurred in % of patients with a significant level of TWA, compared with only % of those without. TWA was performed equivalently to EPS as a predictor of ventricular tachyarrhythmic events. Recently, Gold et al. [] reported a prospective multicenter study of TWA, measured with bicycle exercise testing, in patients referred for EPS because of syncope or presyncope, cardiac arrest, ventricular tachycardia, or supraventricular tachycardia. Signal-averaged electrocardiography (SAECG) was also performed at the time of TWA. Follow-up was obtained in patients with a mean duration of days. The predictive value of TWA and EPS for ventricular arrhythmic events was comparable, and better than SAECG . The combination of TWA with SAECG appeared to enhance the predictive value for ventricular arrhythmic events and the results of EPS.TWA in patients with prior myocardial infarctionRegarding the prognostic utility of TWA in patients with a prior myocardial infarction (MI), only a few studies have been reported. TWA, SAECG, and LVEF were measured in patients with recent MI (± days after the onset of the MI) []. A positive TWA test showed the highest sensitivity, relative risk, and negative predictive value (NPV) but also the lowest specificity, PPV, and predictive accuracy compared to SAECG and LVEF. With multivariate analysis, the combination of TWA and SAECG was the most significant predictor. Recently, a larger cohort study consisting of patients who underwent TWA testing (.±. months after the onset of the MI) revealed that TWA predicted SCD or resuscitated ventricular fibrillation (VF) [].The sensitivity, NPV, and risk hazard of TWA for predicting SCD or VF were higher than LVEF, SAECG, and the presence of non-sustained ventricular tachycardia (VT); however, the specificity and PPV remained worse. The utility of TWA for prognosis in patients with recent MI needs to be further clarified.TWA in patients with cardiomyopathyAdachi et al. [] reported a study of patients with dilated cardiomyopathy (DCM) who underwent a TWA test. Analysis of recorded ventricular tachyarrhythmias, including non-sustained or sustained VT, revealed that ventricular tachyarrhythmias were more common in patients with a significant level of TWA (the sensitivity, specificity, and predictive accuracy rates of TWA to predict VT were %, %, and %, respectively).Klingenheben et al. [] studied patients with congestive heart failure, a mean LVEF of ±%, and no history of sustained ventricular tachyarrhythmias. During months of follow-up there were no patients in the TWA negative group that experienced an arrhythmic event or SCD. Multivariate Cox regression analysis revealed that TWA was the only independent predictor of arrhythmic events. A study of patients with DCM, and with arrhythmic events during ± months demonstrated that TWA in a group of patients with an onset heart rate less than beats per minute was the most significant predictor of arrhythmia-free survival (the sensitivity, specificity, PPV, NPV, and relative risk were %, .%, .%, .%, and ., respectively) [].Studies of TWA in patients with other cardiomyopathies are more scarce. Momiyama et al. [] studied patients with HCM. A significant level of TWA was found in % of patients at high risk of ventricular tachyarrhythmias, compared with none of the other patients who were at low risk. This result suggests that TWA may be a useful marker for high risk of ventricular tachyarrhythmias in patients with HCM; however, this finding was based on a small number of patients. The role of TWA for prognosis in patients with HCM needs to be elucidated.TWA in patients with the long QT and Brugada syndromesMacroscopic TWA has been reported in patients with the long QT syndrome [-,]. Prolongation and unstable state of the ventricular action potential may produce the macroscopic TWA and result in the polymorphic VT known as torsade de pointes. The prognostic value of microscopic TWA has not yet been assessed in patients with the long QT syndrome.In patients with the Brugada syndrome, some reports have revealed that intravenous administration of class Ic antiarrhythmic drugs induced macroscopic TWA and resulted in VF [,]. These results suggest that in the Brugada syndrome class Ic antiarrhythmic drugs may accentuate the underlying sodium channel abnormalities, produce an unstable state of repolarization, increase the triggering PVC, and induce VF. On the other hand, Ikeda et al. [] reported a low prognostic value of microscopic TWA in patients with the Brugada syndrome.Limitations of TWAThere are some technical and electrophysiologic limitations of TWA. The technical limitations are: ) TWA cannot be measured in patients with atrial fibrillation, which is a common arrhythmia in patients with structural heart disease; and ) the presence of frequent atrial or ventricular ectopy, excessive motion artifacts, and, in particular, the inability to achieve the target heart rate would render the results of the TWA test indeterminate. An incidence of indeterminate results of up to % is present in most of the published studies. The electrophysiologic limitation is that TWA testing may lose much of its predictive power within a few weeks after onset of MI.PerspectivesThe clinical utility of TWA in evaluating arrhythmic risk stratification appears promising for patients with suggestive ventricular tachyarrhythmias, congestive heart failure or an LVEF of less than %, and a recent MI. The ultimate role of the TWA test as a noninvasive predictor of SCD awaits larger-scale prospective studies. In the near future, better definition of the clinical role of TWA test will be achieved with the results of the ongoing ABCD (Alternans Before Cardioverter Defibrillator) trial, in which ICD implantation will be performed in patients with ischemic cardiomyopathy and abnormal EPS and TWA results.
|
PMC2648945.txt
|
TITLE: Bilateral blunt carotid artery trauma associated with a double lower thoracic spine fracture: a case report and review of the literature AUTHORS: Dimitrios S Evangelopoulos, Michalis Athanasakopoulos, Konstantinos Kokkinis, Dimitrios Korres, Spyros G Pneumaticos ABSTRACT: BackgroundA year old female suffering from a double fracture of the lower thoracic spine, as a result of a car accident, was referred to the Spine Unit of our Department. No neurological deficit was detected during clinical examination. Due to the fracture's instability, posterior stabilization was performed.Case presentationPostoperatively, the patient presented a decrease in Glasgow Coma Scale (GCS) combined with motor and sensory deficit from the right upper and lower extremities. Cerebral ischemia was diagnosed, secondary to distal emboli from the right and left internal carotid arteries bilaterally.ConclusionPatients' neurological deficits eventually resolved after conservative treatment and anticoagulation therapy. BODY: BackgroundAlthough general trauma is not a common cause of stroke, head and neck trauma associated with cerebrovascular injury is listed as the most frequent cause of stroke in adolescents and young adults [,]Injury of the extracranial carotid or vertebral artery, with associated cervical spine fractures, is a rare but a well documented entity. The first report of an internal carotid artery thrombosis by a non-penetrating neck trauma occurred in []. There have been several reports of patients having pathology in one or more arteries as a result of axial fractures after motor vehicle accidents. More specifically, blunt carotid artery trauma represents only % of all carotid artery injuries, but % of reported cases have been associated with severe neurologic deficits [].We present the case of a patient with blunt carotid trauma bilaterally, combined with double fracture of the lower thoracic spine. To our knowledge, such a case has not yet been described in the literature.Case presentationWe present a case of a year old female with a fracture of the 8th and 12th thoracic vertebrae, subsequent to a motor vehicle accident. Upon admission, physical examination revealed a hematoma on the face, and the upper extremities. Glasgow Coma Scale (GCS) was / and the pupils were equal and reactive to light. The neck was not swollen and no hematomas were observed. Carotid arteries were palpable. Airway was unobstructed and breathing was satisfactory. The patient was hemodynamically stable. The abdomen was soft and non tender. During rectal examination the sphincter tone was normal and no blood was detected. Neurological examination did not reveal any focal deficit.Imaging studies revealed fractures of the 8th and 12th thoracic vertebrae. Furthermore, she suffered of a fracture of the distal end of the right radius and a type III acromioclavicular separation of the left side ().Figure 1CT of spinal fracture.Following appropriate preoperative work up, posterior segmental instrumentation was performed for stabilization of the T12 fracture with the use of pedicle screws. Intraoperative neurophysiologic monitoring was utilized, without any change in the baseline signals (Fig. , ).Figure 2AP x-rays post-stabilization.Figure 3Lateral x-rays post-stabilization.Postoperatively, she was lethargic and confused. The patient also developed a right sided hemiparesis. Since it was estimated that the change in the patient's neurological status was of central origin, a new CT-scan of the head was performed as well as a triplex of the carotid arteries. Neither study demonstrated any pathology.Subsequently, an MRI of the head was performed. The later revealed the presence of a cortical infract at the area of the paracentral lobe as well as a high signal intensity at T2W image and low signal intensity at T1W image in the left caudate and left lentiform nuclei, indicating ischemia in the above mentioned anatomical areas. Pathological signal intensity, indicative of ischemia, was also detected at the anterior portion of the right lentiform nucleus. No other lesions were detected in the areas of brain stem, cerebellum or ventricular system.Further evaluation included Digital Subtraction Angiography (DSA) and MR imaging which demonstrated a dissecting aneurysm at the subpetrous portion of the internal carotid arteries bilaterally. No lesions were detected in the endocranial vessels and in the vertebral arteries (Fig , , , ).Figure 4MR angiography.Figure 5MR angiography.Figure 6DS angiography (RCCA, LCCA).Figure 7DS Angiography weeks post trauma (RCCA, RICA).The patient was placed in anticoagulation therapy with intravenous heparin, according to BCI (Blunt Carotid Artery Injury) protocol [], aiming for an APTT of –. Intravenous anticoagulants were gradually substituted by per os administered Coumadin for at least months [,,]. During the hospital stay, the patient's neurological function improved. At her latest follow up, six months post op, the patient did not demonstrate any motor or sensory deficit in her upper and lower extremities with persistence of brisk deep tendon reflexes in the right side. She is fully ambulatory and is able to perform her activities without difficulty.DiscussionAccording to Crissey et al [], four types of craniocervical injuries may produce traumatic thrombosis of the internal carotid artery: a) direct blows to the anterior neck in the elderly leading to fracture of an atheromatous plaque at the carotid bifurcation, b) blows at the side of the head in youngsters leading to hyperextension, rotation or lateral flexion, stretching the internal carotid artery across the C3 vertebra. These are usually a result of a major impact, such as violent assaults or motor vehicle accidents, c) blunt intraoral trauma, mainly concerning children and d) basilar skull fracture leading to thrombosis of the intrapetrous portion of the internal carotid artery. However, neither the mechanisms nor the associated injuries can accurately predict the type of lesions as stated by Zelenock et al. [] since no consistent pattern of associated injuries could be established. According to Yamada et al. [], the delayed neurological deficits are only safe facts leading to diagnosis, since only % of patients present symptoms of transient ischemic attacks or stroke on admission.Since the risk of a delayed stroke is very high, great importance is given by many authors in finding ways of early diagnosis that would allow the early beginning of anticoagulation or interventional therapy [-]. Mears et al. [], in , reported a case of a blunt carotid artery trauma with neurological deficit and a negative CT-scan of the head, as a result of a blunt cervical trauma. This case resembles to our case in the fact that CT-scan of the head was negative. At the same time, Mooney et al. [], presented the case of a delayed hemiparesis following a non penetrating carotid artery trauma.Martin RF et al. [], in , reported a series of eight patients from a ten-year interval who sustained blunt injuries to the carotid arteries. Six of them (%) suffered a hyperextension injury or had a cervical spine fracture or both. From those patients, five were treated conservatively. No patient died and seven of eight made a complete neurological recovery or remained asymptomatic. The author concludes that one should suspect such a lesion when dealing with hyperextension injuries, with cervical spine fractures or with patients whose neurological deficits can not be explained by intracranial trauma. Moreover he proposes the conservative treatment for small intimal flaps or dissections in asymptomatic patients.Cornacchia et al. [] agree that this injury is associated with cervical spine fractures and conclude, that as many as % of the reported cases have permanent neurological deficit. According to the authors, cerebral angiography remains the diagnostic gold standard but other modalities, such as transcranial Doppler and MRI angiography may also be of great importance.Nevertheless, the aggressive angiographic screening in order to prevent the stroke has not been sufficiently tested. Risk factors for BCI (blunt carotid artery injury) remain numerous, non-specific and not always identifiable prior to the neurologic event [,]. Bfill et al. [,] in their series, demonstrated that % of patients with BCI had no associated injuries and were discovered not because they underwent early angiographic screening but because they developed early neurologic symptoms. The authors, in their effort to group most of the risk factors, concluded that risk factors for BCI were younger age, GCS ≤, diffuse axonal brain injury, head injury, petrous bone fracture, Lefortt II or III fracture, chest injury and abdominal injury. The authors believe that the most common mechanism of injury is that of hyperextension and contralateral rotation since in this way the upper portion of internal carotid artery is stretched along the anteriorly placed lateral articular processes and pedicles of the upper cervical spine.Velmachos et al. [], also announced several possible risk factors for BCI, such as basilar skull and midface fractures, cervical fractures, severe blow to head and neck, broken helmet, neck seat-belt contusion etc. Nevertheless, the author believes that a widely applied, resource-consuming screening program may increase the rate of early diagnosis of BCI, but the improvement in outcome will remain uncertain. Therefore, according to the author, a cost-effectiveness analysis is necessary for trauma surgeons to accept such an aggressive protocol as the standard of care.On the other hand, Friedman et al. [], in his study for vertebral artery injury after cervical spine trauma, states that such lesions as determined by MR angiography are common but they remain clinically occult and only a small percentage of patients may suffer devastating neurological complications. According to the author, noninvasive assessment of the vessels by means of MR imaging should be an integral part of the evaluation of the acutely injured cervical spine.ConclusionIn our case, no cervical trauma was present. Apart from the thoracic spine fractures T8 & T12, no other spinal fracture were detected. The patient had a GCS of / upon presentation and no seat-belt sign or other risk factors were present. The only associated injury involved a grade III acromioclavicular separation with mild hematoma involving the right shoulder.To our knowledge, a case of bilateral internal carotid artery injury associated with fractures of the upper thoracic spines' spinal processes and T8 and T12 vertebrae is not described in the literature. It is important however to keep in mind that polytrauma patients may have suffered occult injuries which the treating physician is not always easy to recognize. One has to investigate the cost effectiveness of extensively evaluating such patients with numerous examinations in the absence of clinical findings, prior to developing a protocol as the standard of care for such patients.AbbreviationsCT: Computer Tomography; MR: Magnetic Resonance; BCI: Blunt Carotid Injury; GCS: Glasgow Coma Scale; DSA: Digital Subtraction AngiographyConsentWritten informed consent was obtained from the patient for publication of this case report and any 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' contributionsEDS treated the patient and wrote the manuscript, AM treated the patient and collected the data, KK performed CT-scans, MRI and DS Angiography, KD revised the manuscript, PSG treated-operated the patient and reviewed the manuscript.
|
PMC3063816.txt
|
TITLE: Predicting hospital mortality among frequently readmitted patients: HSMR biased by readmission AUTHORS: Wim F van den Bosch, Johannes C Kelder, Cordula Wagner ABSTRACT: BackgroundCasemix adjusted in-hospital mortality is one of the measures used to improve quality of care. The adjustment currently used does not take into account the effects of readmission, because reliable data on readmission is not readily available through routinely collected databases. We have studied the impact of readmissions by linking admissions of the same patient, and as a result were able to compare hospital mortality among frequently, as opposed to, non-frequently readmitted patients. We also formulated a method to adjust for readmission for the calculation of hospital standardised mortality ratios (HSMRs).MethodsWe conducted a longitudinal retrospective analysis of routinely collected hospital data of six large non-university teaching hospitals in the Netherlands with casemix adjusted standardised mortality ratios ranging from to and a combined value of over a five-year period. Participants concerned patients admitted times in total during the years - . Predicted deaths by the HSMR model over a five-year period were compared with observed deaths.ResultsNumbers of readmissions per patient differ substantially between the six hospitals, up to a factor of . A large interaction was found between numbers of admissions per patient and HSMR-predicted risks. Observed deaths for frequently admitted patients were significantly lower than HSMR-predicted deaths, which could be explained by uncorrected factors surrounding readmissions.ConclusionsPatients admitted more frequently show lower risks of dying on average per admission. This decline in risk is only partly detected by the current HSMR. Comparing frequently admitted patients to non-frequently admitted patients commits the constant risk fallacy and potentially lowers HSMRs of hospitals treating many frequently admitted patients and increases HSMRs of hospitals treating many non-frequently admitted patients. This misleading effect can only be demonstrated by an analysis over a prolonged period, but occurs, in effect, every day of the year. This finding is relevant for all countries where hospitals use HSMR for monitoring and improving hospital performance. The use of 'admission frequency' as additional adjustment variable may provide a more accurate HSMR. BODY: BackgroundIn various countries in the world, risk adjusted in-hospital mortality ratios are currently calculated using routinely collected data. The variations seen in crude mortality for which adjustment is needed, may be attributed to various sources, including variations in: registration data, casemix, quality of care and chance []. After adjustment for casemix, the hospital standardised mortality ratio (HSMR) can be used as a tool for hospitals to analyse their death rates by comparing their riskadjusted mortality with the national average, establishing a starting point for improvement of hospital outcomes []. In some countries HSMRs are also made publicly available for the benefit of for example: the patient, the politician, the hospital manager and the clinician. This public reporting potentially may result in league tables, suggesting that the ranking reflects differences in quality of care. A somewhat more subtle approach is being applied in the UK by combining standardised mortality ratios with other safety indicators, in order to establish hospital safety rankings. These rankings are publicised in the "How Safe is Your Hospital" guide [], compiled by researchers at the Dr Foster Intelligence Unit at Imperial College London. However, publicly comparing and judging hospital performances in this way is heavily disputed [,] at the moment. Some publications mention the Netherlands as one of the countries where HSMRs are also made publicly available [,], yet up to and including the HSMRs have not been publicised with the exception of a few hospitals that voluntarily put their values on the web. Mandatory publication was planned to start in , based on admission data from , and various articles suggest that the quality of the Dutch HSMRs is sufficiently high to justify publication. Heijink et al [] for example state that they did not find evidence that the HSMR cannot be used as an indicator to monitor and compare hospital quality in the Netherlands. However there is some controversy around this subject and the Dutch Ministry of Health recently decided to postpone mandatory publication for at least one year, because there were doubts about the reliability and validity of the current figure []. This decision was based on the outcomes of a study [] conducted by Santeon, a group of six large non-university teaching hospitals in the Netherlands (Additional file : Table S1), confirming indeed why publicly comparing HSMRs may not be a good idea [].One of the findings of this Dutch study [], not being addressed in [], showed large differences between the hospitals with respect to numbers of readmissions per patient, if measured over a prolonged period of time. Currently the HSMR in the Netherlands is not adjusted however for any form of readmission. In the UK the HSMR is adjusted for readmission for acute cases, but not for a prolonged period of time.The association between readmissions and the HSMR is mentioned in a paper by Jarman []. The effects of adjusting for readmissions are described as follows: " ....There is also not much difference between normal HSMRs based on all admissions and those based on only one (e.g., the last) admission in a year. ....". We interpret this finding as follows: If a patient visits the hospital more than once that year, then any one of the admissions would represent the contribution of that patient to the HSMR ratio. Using a limited part of the patient's admission history, would, on average, not make a difference to the HSMR of that hospital. This may suggest that adjustment for readmission would not make sense. Using this method, we question whether a period of one-year is sufficiently long to embrace all the effects of readmission and indeed whether all the risk conditions surrounding readmissions can be properly addressed. A more recent article by Jarman [] however, does address the need to investigate adjustment for readmissions as follows: " ... further improvements to the case-mix model are being evaluated. The numbers of previous admissions within a given time period, which requires the linking of admissions of the same patient, could be of potential use....". In our publication we will share how differences in numbers of previous admissions made an impact upon the HSMRs of the hospitals and what can be done to improve the HSMR model used.In order to allow comparisons to be unbiased, adjustment for the differing risks of patient specific variables, that is the casemix, is necessary. Risk factors used in the adjustment may be related in different ways to the in-hospital risks. Ignoring non-constant risk relationships commits the constant risk fallacy. Attributing the residual (unexplained) variation from case-mix adjusted mortality to quality of care commits the "case-mix adjustment fallacy" [,]. An example of this phenomenon applied to the HSMR model in the UK is described by Mohammed et al []. Here the interaction of two casemix adjustment variables - comorbidity and admission type - with hospitals was considerable. These effects could be explained by differences in clinical coding and admission practices across hospitals rather than by differences in the quality of care. The lesson we learn from this study is that variables being adjusted may mean different things to different hospitals. Since adjustment of variables is 'admission-based', one can ask whether the practice of admitting and readmitting in itself may be prone to the constant risk fallacy as well. In other words, does an admission of a patient mean the same thing to one hospital as it means to another hospital admitting 'the same patient'? More specifically attuned to our study: are the risk conditions the same for a patient being admitted only once compared to admissions of a frequently admitted patient? In order to analyse this, we addressed the following research questions using HSMRs from six Dutch hospitals:. Are there substantial differences in the numbers of readmissions within a given time period between the six hospitals?. Is there a significant association between HSMRs and numbers of readmissions per patient?. Does the casemix change as readmissions increase and how does it change?. How do we adjust for readmission in order to provide more accurate HSMRs and standardised mortality ratios (SMRs) on diagnostic level?MethodsSetting'Santeon' (Additional file : Table S1) is a group of six large non-university teaching hospitals geographically spread over the Netherlands with HSMRs ranging from (favourable) to (poor) over the years - and an overall HSMR value of (table ). This group of cooperating hospitals covers about % of the total Dutch hospital healthcare in terms of the number of admissions.Table 1Admission numbers and HSMR values years - Santeon hospitals and overall value.HospitalOverall valueABCDEFNumber of admissions over period - 20071147147841746322668026133350978418566HSMR value over period - % Confidence Interval( - )( - )( - )( - )( - )( - )( - )* Hospitals are presented in random order and labelled A - F.Statements1. This study did not concern experimental research, neither research carried out on humans, neither any experimental research on animals.. The dataset used in this study concerned a selection of the 'Landelijke Medische Registratie' (National Medical Registration) covering data from the six Santeon hospitals over the years - . Each of the boards of the six hospitals approved the usage of their part of the dataset.HSMR model used in this studyThe HSMR of a hospital is based on the predicted risks of death per admission. For a certain period of time, for example a year, the HSMR is calculated with the formula:The Dutch HSMR model (DHM-) used in this study accounts for at least % of hospital mortality. Here adjustment is made for: age, sex, admission type (emergency or non-emergency), length of stay, year of discharge, socio-economic deprivation, comorbidity and CCS diagnostic group based on ICD- coding. The DHM-, was developed by the Dr Foster Unit in the UK in cooperation with 'Prismant' and the 'PraktijkIndex' in the Netherlands. DHM- is very well described in [] and resembles the UK model. Differences with the UK model concern: the use of day cases (not used in UK) and the use of CCS groups in NL (based on ICD- coding) versus CCS groups (based on ICD- coding) in UK of which in common. Furthermore the UK model adjusts for palliative care, source of admission and for the number of previous emergency admissions; DHM- does not adjust for any of these three. Calculations of the HSMR-results were carried out by Prismant, an independent research and advisory agency for the Dutch health service.Definition of ReadmissionThe term 'readmission' is often used for unplanned readmissions within a limited period of time, for example days, for treatment of the same disease. Planned readmissions are another frequently occurring form of readmitting patients, particularly for chronically ill patients, for example suffering from neoplasms. In these cases treatment schemes may span a prolonged period of time, even a number of years. Furthermore, a patient may visit the hospital for a different disease than during a previous admission, which in a way can be seen as a readmission as well. In our study we considered all these cases as 'readmissions' of the same patient over the study period of five years. We adopted the term 'nth admission' [], according to the definition, applying to admissions that occurred after the 6th admission, where 'nth' is an ordinal number - e.g., seventh, eighth, etc. We used the term also for n = , , ....Identifying the nth admission of a patientRoutinely collected data on hospital admissions in the Netherlands are being collated in the national medical registration LMR ('Landelijke Medische Registratie'). In general, readmissions are not adequately registered in the LMR. In order to be able to identify the value of n of the nth admission of a patient, we used the dates of discharge combined with the patient's identification number.Admission frequencyHSMR-variable data is collected on a 'per admission' basis. The model considers each admission to be an independent stochastic experiment (like repeatedly throwing a dice), separate from previous admissions, for which a risk of death number is predicted and accumulated into the denominator of the HSMR. The admission represents a container or proxy for risk conditions around a single hospitalization of a patient. However, comparing mortality ratios of hospitals for a certain fixed period of time, one might not only be interested in the quality of single admissions, but also in the quality of the end result - was the patient discharged alive after the last admission in a row? The patient is the primary subject of interest for hospitals for whom they bear the responsibility for care. For the purpose of this study we introduce a new variable that may serve as a risk proxy more attuned to the patient, taking into account all the patient's admissions during the study period. Since risk conditions may be linked to how often a patient was admitted in total, we allocated each patient to a distinct patient class P(m), where:Admission frequency m = the number of times a patient was admitted during a fixed period of time(Occasionally we use the number of times a patient was admitted after the initial admission, in which case we apply the term 'readmission frequency' = m-).The second research question formulated tests the capability of the current HSMR model for predicting mortality accurately for these classes. Based on this we draw conclusions on whether the new variable m turns out to be a valuable addition to the HSMR model. This idea was suggested but, as far as known, not further explored by Jarman [].Admission view versus Patient viewThe dataset of all of the admission records can be grouped in different ways or views. We have studied and compared mortality risks based on the following two different views:• Admission view: this view is based on nth admission classes A(n) where A(), A(), etc. represent all admission records for all first admissions, all second admissions, etc. For example each patient in class A() contributed one admission record to class A(), but also one to A() and one to A().• Patient view: this view is based on admission frequency classes P(m) representing all admission records of patients who have been admitted exactly m times over the five-year period. For example: each patient in class P() contributed exactly three admission records to this class and did not contribute any admission record to any other patient view class. In fact, patients with equal admission frequencies are clustered in distinct mutually exclusive classes, in contrast with the admission view where one patient may contribute admission records to many classes. For example every patient contributes one admission record to class A().Calculating mortality figuresWe calculated crude mortality, predicted mortality (based on DHM-) and standardised mortality ratios (SMRs) by applying the HSMR formula for each class A(n) and P(m) for m,n = , , , ... In order to preserve power we have grouped the results of the higher (m > ) classes P(m) as follows: m = , into one group, m = - into one group, m = - into one group, and m > into one group. Similarly for A(n). For the SMRs we also calculated the % confidence intervals. We analysed the association between observed and predicted outcomes, and goodness of fit for both views.ResultsDuring the five-year study period patients accounted for admissions in total. patients (% of total patients) concerned 'first-and-only' admissions. The other patients (% of total patients) who were admitted more than once accounted for admissions (% of total admissions) of which were readmissions (% of total admissions).Variations in admission frequenciesWe analysed the distribution of the admission frequency per patient class P(m) per hospital (table ). For m = hospital B has the highest percentage admissions (% of total), hospital D the lowest (% of total). For P(m > ) hospital D has high scores, for example: % of the admissions concerned patients being admitted more than times (P(m > )), whereas for hospitals E and F this percentage amounts to ,%. We also analysed the inter-hospital variation of the average admission frequency on the level of the main CCS diagnostic groups (figure ). The first three diagnostic groups shown - neoplasms, heart diseases and respiratory diseases - cover two thirds of all readmissions. For these diagnostic groups we calculated the ratio between the highest and the lowest observed hospital average of readmission frequencies. For neoplasms, hospital D (. readmissions) and hospital B (. readmissions) differed by a factor of . For heart diseases and respiratory diseases the ratio between highest and lowest amounted to . and . respectively. Overall (bottom line of figure ) hospital D has the highest average readmission frequency (.) and hospital B and F have the lowest average (.); a factor of difference between the highest and the lowest.Table 2Distribution of number of (re)admissions over Patient view classes m for each of the six hospitals.Percentages of admissions of total admissions per hospitalPatient view classABCDEFAll hospitalsP(m = ).%.%.%.%.%.%.%P(m = ).%.%.%.%.%.%.%P(m = ).%.%.%.%.%.%.%P(m = ).%.%.%.%.%.%.%P(m = , ).%.%.%.%.%.%.%P(m = -).%.%.%.%.%.%.%P(m = -).%.%.%.%.%.%.%P(m > ).%.%.%.%.%.%.%Total100%%%%%%%* Hospitals are presented in random order and labelled A - F.Figure 1Distribution number of readmissions divided by number of patients per hospital per main CCS diagnostic group *. The chart is showing the average number of readmissions. The six hospitals are labelled A - F. Each diagnostic group title also shows readmission sample sizes, for example there were readmissions for neoplasms in total.Mortality per patient view class and per admission view classIn table the total results of mortality calculations are shown from the perspective of the patient view classes. The row of class P(m = ) shows the outcomes for all patients who were admitted exactly once, row P(m = ) for patients admitted exactly twice, and so on. The crude mortality per patient for first-and-only admissions amounts to .%. For m = through to m = we see an average growth in mortality of roughly % per patient view class increment. For m = through to + the crude mortality per patient class does not show growth anymore and stabilizes around %. Table also shows that DHM- predicts a decline in mortality per admission from .% (P(m = )) to .% (P(m > )).Table 3Mortality figures per Patient view class m: crude mortality per patient and per admission, predicted mortality per admission.Observed number ofCrude mortality perDHM- predictedPatient view classpatientsadmissionsdeathspatientadmissionnumber of deathsmortality per admissionP(m = ).%.%.%P(m = ).%.%.%P(m = ).%.%.%P(m = ).%.%.%P(m = , ).%.%.%P(m = -).%.%.%P(m = -).%.%.%P(m > ).%.%.%Total240662418566152276.%.%.%* The numbers represent totals of the six hospitals. The crude mortality per patient stabilises around % after (re)admissions. The predicted mortality per admission increasingly differs (higher) from observed mortality for increasing patient view classes. The statistical significance of this is shown in table by means of corresponding standardised mortality ratios and % confidence intervals.In table the results of mortality calculations are shown from the perspective of the admission view classes. Row A(n = ) shows the outcomes for all of the patients being admitted for the first time. From these patients, were admitted at least a second time, shown in row A(n = ), etc. In this case, going from A(n = ) to A(n > ), a decline of mortality per admission from .% to .% is predicted by DHM-.Table 4Mortality figures per Admission view class: crude mortality per admission and predicted mortality per admission.Admission view classObserved number ofCrude mortalityDHM- predictedadmissionsdeathsper admissionnumber of deathsmortality per admissionA(n = ).%.%A(n = ).%.%A(n = ).%.%A(n = ).%.%A(n = ,).%.%A(n = -).%.%A(n = -).%.%A(n > ).%.%Total418566152273.%.%* The numbers represent totals of the six hospitals. Predicted mortality per admission is in line with observed mortality for all admission view classes. The statistical significance of this is shown in table by means of corresponding standardised mortality ratios and % confidence intervals.In the patient view the number of observed deaths for first-and-only admissions (P(m = )) is clearly higher than predicted and for P(m = ) - P(m > ) lower (table ). In the admission view the differences are smaller (table ).From tables and we can calculate the standardised mortality ratios for both views via the ratio of observed deaths and predicted deaths per category. Table shows the results, including % confidence intervals (p < .). Figure presents a graphical representation of this. In the patient view the SMRs decline from (P(m = )) to (P(m > )) and none of the corresponding % confidence intervals includes the expected overall HSMR value . (% CI: . - .). This shows a significant association between patient view categories and SMRs indicating a lack of model fit. In the admission view however, the SMRs fluctuate between and and all corresponding % confidence intervals include the overall HSMR value . (% CI: . - .) indicating a good fit of the HSMR model for this view, which we will discuss later.Table 5Standardised mortality ratios (SMRs) for patient view and for admission view.Patient ViewSMR per Patient View CategoryLower % CIof SMRUpper % CIof SMRAdmission ViewSMR per Admission View CategoryLower % CIof SMRUpper % CIof SMRP(m = )...4A(n = )...3P(m = )...5A(n = )...6P(m = )...5A(n = )...6P(m = )...8A(n = )...5P(m = , )...5A(n = , )...9P(m = -)...3A(n = -)...6P(m = -)...8A(n = -)...0P(m > )...5A(n > )...1Figure 2Standardized mortality ratios of Patient view classes and Admission view classes including % confidence intervals *. Santeon overall HSMR equals . (% CI: . - .).Casemix risk profile for the nth admissionFinally, we also studied the variation over the nth admission classes of five casemix variables (table ). Year, sex and social deprivation are considered less relevant variables for this. We split the table into patient and admission specific property groups. Variations in casemix for nth admissions for n going from to + can be characterized by: an average age that initially increases from (n = ) to years (n = ) and then gradually drops to years; a steady increase of comorbidity as indicated by a diminishing contribution of the two lowest and an increasing contribution of the higher Charlson indices; a predominance of heart diseases for lower values of n and a predominance of neoplasms for higher values of n; a decrease of emergency admissions; a decrease of lengths of stay (for n > ). As readmissions increase the casemix changes as indicated by the combination of variations of these five casemix variables. These changes are associated with a decline in mortality for n > as predicted by DHM- (table ).Table 6Values/distributions of major casemix variables for which adjustment is done by DHM- per nth Admission for six hospitals.AAAAAAAA(n = )(n = )(n = )(n = )(n = ,)(n = -)(n = -)(n > )Number of admissions2406627577833902185461912412634124835437Patient specific casemix propertiesAverage age at nth discharge (years)........0Distribution of Charlson indices:%%%%%%%%%%%%%%%% and %%%%%%%%, and %%%%%%%%Distribution of Admissions per main diagnostic CCS group:Neoplasms13%%%%%%%%Metabolic & shock etc4%%%%%%%%Heart disease40%%%%%%%%Respiratory disease11%%%%%%%%Gastro-intestinal disease13%%%%%%%%Other main CCS groups21%%%%%%%%Admission specific casemix propertiesDistribution of emergency admissions43%%%%%%%%Average length of stay (days)........8Discussion(Re)admission frequencies show substantial variations between the hospitals. For example: the overall value of the readmission frequency of hospital D equals double the value of hospital B and F. For the main CCS diagnostic group neoplasms, this value for hospital D equals even three times the value of B. We conclude that there are substantial differences in the numbers of readmissions within a given time period between the six hospitals.Patients with higher admission frequencies bear lower predicted risks per admission, which can be explained by shifts in the casemix. For the patient view however, lowering of risks is only partly predicted by DHM-, since the amount of predicted deaths compared to observed deaths turned out to be relatively low for first-and-only admissions and high for readmissions. The corresponding standardised mortality ratio in the patient view is high for class P(m = ) () and low for the other classes (gradually dropping to for P(m > )) compared to the overall average of . In contrast to this we found that the standardised mortality ratios per nth admission view class are approximately as predicted, fluctuating around . DHM- demonstrates quite a fair goodness of fit for the nth admission view classes. This result matches the findings of Jarman [], where no differences in HSMR were detected by picking nth admissions for any value of n. At the same time however, a comparison of predicted and observed deaths for the patient view classes does not demonstrate a good fit for the model. The question now arises why DHM- should be suited to fit the nth admission view and not be suited to fit the patient view? As we will explain below, different admission frequency classes of the patient view may incur risk differences, not detected by the known adjustment variables. In that case, comparing these classes would commit the constant risk fallacy and establishing a model fit along the lines of the patient view would be preferred to a fit along the lines of the admission view.We will demonstrate this point by three examples that show mechanisms, lowering risks for higher admission frequencies and having nothing to do with higher quality of care:Example : Admission policies may increase the number of readmissions without proportionally increasing real risks. One hospital may systematically combine the diagnosis and treatment into a single admission. Another hospital may have an admission for diagnosis and a second one for treatment, being granted a double predicted risk count, most likely without doubling of real risk, but doubling the expected risk.Example : Hospitals may show different treatment practices in the frequency with which chronically ill patients are admitted for the same disease. For example readmission frequencies for the treatment of neoplasms in hospitals B and D differ on average by a factor of . This difference can be explained by differences in balance between inpatient versus outpatient treatments (late versus early adopters of a trend moving from inpatient toward outpatient treatment, for example []). The predicted risk contribution for hospital D may thus be tripled compared to B, but it is not likely that patients of D are being exposed to a tripled real risk. On the contrary, the mortality risk per patient stabilizes around % if admitted or more times (table ). Every incremental admission for chemotherapy further increases the denominator of the HSMR, but on average does not increase the observed mortality, the numerator, further.Example : Patient referrals to tertiary care, frequently occurring in the Netherlands, may cause differences. The transfer of a patient back and forth most often is an advantage for the referring hospital, as they count a double admission, while the hospital to which the patient is being referred is at a disadvantage because they only count for one admission. On top of this the latter hospital has to deal with the risk of conducting a potentially complicated medical procedure. It is unlikely that a patient under these circumstances will experience a doubling of risk in the referring hospital.Another plausible mechanism may be hidden in the physical condition of frequently readmitted patients. If these patients are unexpectedly resilient, they will dominate higher admission frequency groups through natural selection and consequently cause lower undetected risks, compared to lower admission frequency groups. If this hypothesis is valid, it might become visible as well in patient specific casemix properties that are known to us such as age and comorbidity (table ). The average age of frequently admitted patients indeed is decreasing for the highest values of n, indicating a fitter population. The comorbidity is increasing however; a logical consequence of the fact that we applied the notion of readmission for any disease. So patients with more co-morbidity may be admitted more often for the various diseases they suffer from. It also indicates higher vulnerability and in that case would contradict the hypothesis. Although the hypothesis looks appealing, we cannot proof its correctness with the data currently available.Looking back at the admission history of a frequently admitted patient, the additional risk-lowering factors just described, may have come into play already after the patient's first admission. The nth admission view constitutes a cross-section of various patient view groups, each having their own additional risks. Consequently the nth admission view cannot discriminate additional risk differences and is not suited to be used for readmission adjustment. Instead, we think the variable 'admission frequency' of the patient view should be used for this purpose. Patient views are showing why DHM- predicts too high a risk for a frequently readmitted patient, as illustrated by the following case that we observed:Patient × of hospital D contributed . predicted deaths to the denominator of the HSMR through seven successive admissions within six months. A single patient however can maximally contribute a value of - in case of death - to the numerator of the HSMR.Numerous examples alike became available in our study, for example: we found patients in hospital D, each of whom contributed more than predicted death to the denominator of the HSMR due to various readmissions. In total these contributions in hospital D added up to predicted deaths, whereas 'only' deaths were observed in this group. Clearly the number of deaths predicted by DHM- for frequently admitted patients is being overestimated.A final illustration of this phenomenon is shown in figure : a scatter diagram where we plotted the HSMRs as well as the SMRs of the three main CCS diagnostic groups that show the largest variations in readmission frequency - neoplasms, heart diseases and respiratory diseases - against the readmission frequency (see also figure ). In all cases there is a downward trend: higher readmission frequencies corresponding with lower (H)SMRs.Figure 3HSMRs and SMRs of neoplasms, heart diseases and respiratory diseases versus readmission frequency *. Linear regression lines are shown.We conclude that there is a significant association between HSMRs and numbers of readmissions per patient.Since the study involved five consecutive years, we were not able to capture the complete patient view. Patient view sequences that started before and continued to emerge in the period - were truncated. The same happened with sequences that started during the period - and continued in the years thereafter. Consequently the picture will never be complete. For each patient being admitted at least twice, we calculated the time elapsed between the date of the first and the date of the last discharge. For % of these patients, the time elapsed amounted to less than one year and for % less than years, suggesting that the larger part of the effect has been captured.How can HSMRs be adjusted for the effects of readmissions? For that purpose, an additional adjustment variable 'admission frequency,' as used in this study in the patient view, may be applied. After adjustment, the SMRs of the patient view classes in our study will fluctuate around . As a consequence the goodness of fit for the admission view will be lost. However, this does not provide us with a principal problem since the division into admission frequency classes (patient view) is along the lines of identified and distinct risk classes for which adjustment is clearly needed. Nth admission classes turn out to be meaningless in terms of risk differentiation.In particular for the higher admission frequencies, a prolonged measurement period of various years was needed in order to make visible the effects we described. For example the average time which elapsed between the first and last date of discharge for patient view class P(m > ), amounted to . years. This does not mean however that having many patients in class P(m > ), does not have an effect if measured for only one year. Ideally the adjustment for readmission would be based on an 'admission counter' in the file of each patient that is being increased by after every new readmission. Since such counts are not being kept, an approach along the lines of this study, awkward as it might be, will be necessary.A final remark concerns the following: (frequent) readmissions are sometimes taken to be a proxy indicator for poor quality of care. But instead of working against the hospitals with higher readmission frequencies the current HSMR seems in that case to work in favour of those hospitals, underlining the following statement: If HSMRs in the Netherlands ever will be publically reported and used to compare hospitals, then the issues raised in [] need to be resolved and, on top of this, adjustment for readmission will be necessary. If HSMRs and particularly SMRs on diagnostic level are used by hospitals as a starting point for quality improvement [] - a better idea for usage of the HSMR indicator - then adjustment for readmission will be necessary as well, in order to prevent misleading signals to be generated. Hospitals will in that case better be able to avoid falsely acting upon too high SMR values as well as to avoid falsely non-acting upon too low SMR values.ConclusionsThis study has shown that patients admitted more frequently experience a lower risk of dying per admission. The HSMR model used, detects lower risks for higher admission frequency groups. However the observed mortality of these groups demonstrate the real risks to be even lower, indicating differing risk conditions between the groups. Consequently, comparing admissions of patients from different admission frequency groups, as done by the current HSMR model, commits the constant risk fallacy. The study showed substantial variations in the overall distribution of readmission frequencies, up to a factor of , between the six hospitals. As a result hospitals with high admission frequencies experience unadjusted decrease of risks and hospitals with low admission frequencies experience unadjusted increase of risks. These misleading outcomes can only be demonstrated by analysis over at least three years, but is in effect every day. As % of all admissions in this study concerned readmissions, the impact on the HSMR of some hospitals may be substantial. This finding is relevant for all countries where hospitals use (H)SMRs for monitoring and improving hospital performance. Further research to identify how the effects of readmissions may impact the HSMR in other countries is clearly warranted. The accuracy of the HSMR may improve by taking admission frequency as an additional adjustment parameter.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAll authors have contributed significantly to the conception and design of the study. WB collected and analysed the data and produced the first draft of the manuscript. All authors revised the manuscript critically for important intellectual content and gave final approval of the version to be published.Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/-///prepub
|
PMC2362734.txt
|
TITLE: Hypoxia significantly reduces aminolaevulinic acid-induced protoporphyrin IX synthesis in EMT6 cells AUTHORS: I Georgakoudi, P C Keng, T H Foster ABSTRACT: We have studied the effects of hypoxia on aminolaevulinic acid (ALA)-induced protoporphyrin IX (PpIX) synthesis in EMT6 monolayer cultures characterized by different cell densities and proliferation rates. Specifically, after ALA incubation under hypoxic or normoxic conditions, we detected spectrofluorometrically the PpIX content of the following populations: (a) low-density exponentially growing cells; (b) high-density fed-plateau cells; and (c) high-density unfed-plateau cells. These populations were selected either for the purpose of comparison with other in vitro studies (low-density exponentially growing cells) or as representatives of tumour regions adjacent to (high-density fed-plateau cells) and further away from (high-density unfed-plateau cells) capillaries. The amount of PpIX per cell produced by each one of these populations was higher after normoxic ALA incubation. The magnitude of the effect of hypoxia on PpIX synthesis was dependent on cell density and proliferation rate. A -fold decrease in PpIX fluorescence was observed for the high-density unfed-plateau cells. PpIX production by the low-density exponential cells was affected the least by ALA incubation under hypoxic conditions (.-fold decrease), whereas the effect on the high-density fed-plateau population was intermediate (-fold decrease). © Cancer Research Campaign BODY: No Body Content
|
PMC3051686.txt
|
TITLE: ,′-Azanediyldiethanaminium pyridine-,-dicarboxylate AUTHORS: Hossein Aghabozorg, Maryam Saemi, Zeynab Khazaei, Vahid Amani, Behrouz Notash ABSTRACT: The crystal structure of the title compound, C4H15N3 +·C7H3NO4 −, consists of diethylenetriaminium (,′-azanediyldiethanaminium) cations and pyridine-,-dicarboxylate anions, which are linked by N—H⋯O, N—H⋯N and C—H⋯O hydrogen bonds. C—H⋯π interactions are also observed. In the anion, the carboxylate groups are oriented at dihedral angles of . () and . ()° with respect to the pyridine ring. BODY: Related literatureFor general background to proton-transfer compounds, see: Sheshmani et al. ( ▶); Aghabozorg et al. (2008a ▶,b ▶,c ▶); Derikvand et al. ( ▶). ExperimentalCrystal data C4H15N3 +·C7H3NO4 − M r = .29Monoclinic, a = . () Å b = . () Å c = . () Åβ = . ()° V = . () Å3 Z = 4Mo Kα radiationμ = . mm− T = K0. × . × . mm Data collection Stoe IPDS II diffractometerAbsorption correction: integration (X-RED32; Stoe & Cie, ▶) T min = ., T max = . measured reflections3593 independent reflections2523 reflections with I > 2σ(I) R int = . Refinement R[F > 2σ(F )] = . wR(F ) = . S = . reflections200 parametersH atoms treated by a mixture of independent and constrained refinementΔρmax = . e Å− Δρmin = −. e Å− Data collection: X-AREA (Stoe & Cie, ▶); cell refinement: X-AREA; data reduction: X-RED32 (Stoe & Cie, ▶); program(s) used to solve structure: SHELXS97 (Sheldrick, ▶); program(s) used to refine structure: SHELXL97 (Sheldrick, ▶); molecular graphics: ORTEP- for Windows (Farrugia, ▶); software used to prepare material for publication: WinGX (Farrugia, ▶).Supplementary MaterialCrystal structure: contains datablocks I, global. DOI: ./S1600536810054413/xu5129sup1.cif Structure factors: contains datablocks I. DOI: ./S1600536810054413/xu5129Isup2.hkl Additional supplementary materials: crystallographic information; 3D view; checkCIF report
|
PMC1079822.txt
|
TITLE: A gene expression fingerprint of C. elegans embryonic motor neurons AUTHORS: Rebecca M Fox, Stephen E Von Stetina, Susan J Barlow, Christian Shaffer, Kellen L Olszewski, Jason H Moore, Denis Dupuy, Marc Vidal, David M Miller ABSTRACT: BackgroundDifferential gene expression specifies the highly diverse cell types that constitute the nervous system. With its sequenced genome and simple, well-defined neuroanatomy, the nematode C. elegans is a useful model system in which to correlate gene expression with neuron identity. The UNC- transcription factor is expressed in thirteen embryonic motor neurons where it specifies axonal morphology and synaptic function. These cells can be marked with an unc-::GFP reporter transgene. Here we describe a powerful strategy, Micro-Array Profiling of C. elegans cells (MAPCeL), and confirm that this approach provides a comprehensive gene expression profile of unc-::GFP motor neurons in vivo.ResultsFluorescence Activated Cell Sorting (FACS) was used to isolate unc-::GFP neurons from primary cultures of C. elegans embryonic cells. Microarray experiments detected , unique transcripts of which ~, are enriched in unc-::GFP neurons relative to the average nematode embryonic cell. The reliability of these data was validated by the detection of known cell-specific transcripts and by expression in UNC- motor neurons of GFP reporters derived from the enriched data set. In addition to genes involved in neurotransmitter packaging and release, the microarray data include transcripts for receptors to a remarkably wide variety of signaling molecules. The added presence of a robust array of G-protein pathway components is indicative of complex and highly integrated mechanisms for modulating motor neuron activity. Over half of the enriched genes () have human homologs, a finding that could reflect substantial overlap with the gene expression repertoire of mammalian motor neurons.ConclusionWe have described a microarray-based method, MAPCeL, for profiling gene expression in specific C. elegans motor neurons and provide evidence that this approach can reveal candidate genes for key roles in the differentiation and function of these cells. These methods can now be applied to generate a gene expression map of the C. elegans nervous system. BODY: BackgroundThe nervous system is assembled from disparate classes of neurons that together define the overall properties of the network. The specific functions of these neurons are governed by genetic programs that control cell fate []. Thus, a key to understanding the molecular basis for neural function is to establish the gene expression blueprint that orchestrates neuronal differentiation. With its simple, well-defined nervous system and powerful genetics, the nematode C. elegans is a useful model system for addressing this issue. The C. elegans hermaphrodite nervous system is composed of exactly neurons. The morphology and connectivity of each one of these neurons has been defined at high resolution []. In addition, the birth of each neuroblast is embedded in a lineage diagram of every cell division in C. elegans development [,]. The C. elegans genome is fully sequenced and contains ~, predicted genes []. At a fundamental level, the identity of a given class of neuron is defined by a unique combination of these genes. In principle, microarray-based strategies could be employed to establish these cell-specific patterns of gene expression. However, the small size of the nematode has limited access to individual cells for molecular analysis. Here we describe a strategy, MAPCeL (Micro-Array Profiling of C. elegans Cells) that overcomes these obstacles to generate neuron-specific gene expression profiles.MAPCeL exploits recently developed methods of culturing C. elegans embryonic cells. GFP markers for specific classes of neurons and muscle cells are expressed in vitro and can be used to identify the corresponding differentiated cell types. We established that these GFP cells arise at a frequency predicted by their abundance in the intact embryo and display normal morphological, molecular, and physiological characteristics []. For example, a GFP reporter for the unc- homeodomain transcription factor gene is expressed in motor neurons out of a total cells in the mature embryo (Figs. 2D, 5A) []. In vitro, we detected a comparable fraction (~%) of unc-::GFP cells. Moreover, cultured unc-::GFP cells adopt neuronal-like processes and express molecular markers also seen in vivo (Fig. 2D–E) []. On the basis of these results, we have profiled cultured unc-::GFP neurons with the expectation that this approach will provide a comprehensive picture of genes expressed in these motor neurons in vivo.Figure 1MAPCeL strategy for profiling C. elegans GFP neurons. Embryos are isolated from gravid adults and treated with chitinase to degrade the egg shell. Embryonic cells are cultured for hours and enriched by FACS. Amplified, labeled aRNA is hybridized to the Affymetrix C. elegans array.Figure 2Isolation of unc-::GFP neurons by FACS. A. Fluorescence intensity profile of wildtype (non-GFP) cells. Boxed areas exclude autofluorescent7 cells (arrow). B. unc-::GFP cells are gated to exclude propidium iodide-stained (non-viable) cells. C. Light scattering gate for GFP-positive cells (circle) to exclude cell clumps and debris. D. Combined fluorescence and DIC image of unc-::GFP labeled motor neurons in L1 larva. (DA2 is not visible here.) Arrow points to embryo at stage (< min) prior to unc-::GFP expression. E, F. Fluorescence and DIC images of hr culture from unc-::GFP embryos. G. unc-::GFP neurons after enrichment by FACS. Arrow heads point to rare (~%) non-GFP cells. Scale bars are microns.Figure 3Coefficients of Determination (R2) for individual hybridizations. A. Scatter plot of normalized intensity values (log base ) for representative hybridization (DMR30) from all cells (Reference) compared to the average intensity of four Reference hybridizations. B. Scatter plot of representative unc-::GFP hybridization (DM39) compared to the average intensities for all three unc-::GFP hybridizations. C. Results of single unc-::GFP hybridization (DM39) (red) compared to average Reference intensities (green) to identify transcripts showing differential expression. The unc- transcript (arrowhead) is highly enriched (~×) in unc-::GFP neurons D. R2 values for all pairwise combinations of unc-::GFP hybridizations. E. R2values for all pairwise combinations of Reference (i.e. all cells) hybridizations.Figure 4Comparison of Expressed Genes (EGs) in unc-::GFP neurons vs all cells (N2 reference). transcripts are detected exclusively in unc-::GFP motor neurons and transcripts are detected exclusively in the N2 reference data set. transcripts are detected in both data sets.Figure 5GFP reporters validate UNC- motor neuron genes. A. unc-::GFP is expressed in embryonically-derived motor neurons. Colored circles indicate approximate location of UNC- motor neuron soma in newly hatched L1 larva. Transgenic animals expressing GFP reporters for representative UNC- motor neuron genes. Anterior to left. B. acr-. Confocal GFP/DIC projection. acr-::GFP is expressed in RME in the head and in ventral nerve cord (VNC) motor neurons (arrows). C, D. flp-. Confocal DIC/GFP image of head region. (C) and matched confocal GFP projection (D). Note DD1 commissure to dorsal side (white arrow). Arrowhead marks nerve ring. E, F. F29G6.. Arrows point to VNC motor neurons (L2 Larva) (E). Posterior view of VNC showing F29G6.::GFP expression in all VNC motor neurons. Asterisk marks gut autofluorescence (F). G, H. tig-. DIC/GFP image of L2 larva. Note GFP expression in VNC motor neurons (arrows) and in head muscles (arrowhead) (G). Confocal projection of anterior VNC. tig-::GFP is detected in A and B class motor neurons, pharyngeal muscle (arrow) and body wall muscle (bwm) (H). Scale bars are microns in panels C, D, F, H and microns in B, E, G.We describe methods for isolating unc-::GFP-labeled neurons by Fluorescence Activated Cell Sorting (FACS). mRNA from these cells is amplified, labeled and hybridized to the C. elegans Affymetrix Gene Chip. A comparison to microarray data derived from all embryonic cells reveals ~ genes with significantly higher levels of expression in unc-::GFP neurons. The validity of these data is supported by the inclusion of genes known to be expressed in these neurons in vivo and by the generation of new GFP reporters from previously uncharacterized genes on this list. We conclude that MAPCeL offers a reliable strategy for profiling gene expression of a specific motor neuron class. Using this approach, we have provided, for the first time, a comprehensive picture of gene expression in a subset of C. elegans motor neurons. We expect that MAPCeL can now be applied to other C. elegans neurons and thereby link specific neuronal fates with unique combinations of differentially expressed genes.ResultsProfiling strategyunc-::GFP is expressed in embryonic motor neurons; () I5 (pharynx), () SAB (retrovesicular ganglion), and () DA (ventral nerve cord) (Figs 2D, 5A) []. Although each of the motor neuron classes is morphologically distinct, the DA and SAB motor neurons, which constitute the majority (/) of unc-:GFP neurons, also share several characteristics including common presynaptic inputs, anteriorly directed axonal processes, cholinergic activity, and similar defects in unc- mutants [,]. It is therefore reasonable to assume that many of the same genes would be expressed in both of these motor neuron classes and that these could be revealed in microarray experiments.A schematic of our approach to profile unc-::GFP cells is presented in Fig. . C. elegans embryonic cells were cultured for hr to allow differentiation of GFP-labeled motor neurons. unc-::GFP cells are rarely observed in freshly dissociated preparations but constitute about % of all cells after day in culture. The delayed appearance of unc-::GFP cells in culture is consistent with the developmental timing of unc-::GFP expression in vivo; unc-::GFP motor neurons are normally generated after morphogenesis is initiated []. These older embryos are not dissociated by our methods []. Fluorescence Activated Cell Sorting (FACS) is used to isolate enriched (~%) populations of unc-::GFP cells. RNA is extracted, amplified, and labeled for application to the C. elegans Affymetrix Gene Chip.unc-::GFP motor neurons are isolated by FACSIt is necessary to plate freshly dissociated embryonic cells on a solid substrate to promote differentiation and to prevent clumping. Although C. elegans neurons show extensive morphological differentiation on peanut lectin-coated glass they also adhere avidly and cannot be easily removed. We discovered that cells plated on poly-L-lysine coated surfaces also differentiate but can be readily dissociated from substrate by gentle trituration. A fluorescence profile was established for cells from the non-GFP wildtype strain (N2) to identify autofluorescent intestinal cells. Because these cells autofluoresce in both the Propidium Iodide (PI) and GFP channels, they are largely restricted to the diagonal axis of this scatter plot (Fig 2A). PI was added immediately prior to sorting to stain damaged cells (~%). Separate experiments with PI-stained wildtype cells and with cells from unc-::GFP embryos were used to establish sorting gates for PI and GFP-labeled cells, respectively (Fig 2B). As shown in Fig 2C, viable unc-::GFP neurons were simultaneously gated by light scattering parameters. This gate was established empirically to achieve ~% enrichment of unc-::GFP labeled cells (Fig 2G). We typically obtained about , unc-::GFP neurons from each sort. RNA from the equivalent of , unc-::GFP neurons was pooled for each separate microarray experiment. We will refer to microarray results from unc-::GFP marked cells as the "unc-::GFP motor neuron" data set. Reference RNA was extracted from all viable cells sorted from a hr culture of wildtype embryonic cells. Microarray results with this "Reference" data set should reflect transcript levels in the average differentiated embryonic cell.Microarray experiments yield reproducible profilesData obtained from successive hybridizations of two separate arrays with the same labeled probe yielded a coefficient of determination, R2 = . (data not shown). This result indicates that potential differences between individual Affymetrix arrays or hybridization and scanning procedures are not significant sources of error. The overall concurrence of the experimental (unc-::GFP motor neuron) and Reference data is illustrated graphically in the scatter plots shown in panels A and B of Fig . To assess the reproducibility of sample preparation methods (e.g. FACS isolation, RNA extraction, amplification, labeling, etc.), R2 was calculated for each pairwise combination of independent samples. An average R2 of . (n = ) was calculated for the wildtype (N2) reference samples (Fig 3E); average R2 was . (n = ) for the unc-::GFP motor neuron data set (Fig 3D). These values are indicative of highly similar samples and thereby show that our methods are reliable.Detecting Expressed Genes (EGs)Differential hybridization to perfect match (PM) vs mismatch (MM) oligo probes on the Affymetrix chip was used to identify transcripts reliably detected as "present" in the Reference and unc-::GFP motor neuron data sets (see Methods, Additional Files , ). This list was adjusted in two ways for the unc-::GFP motor neuron data set to arrive at a more accurate representation of Expressed Genes (EGs) (Additional Files , ). In the first instance, transcripts that were statistically downregulated in unc-::GFP motor neurons relative to the wildtype reference were removed from the "present" list as these are likely to be detected because they are actually highly enriched in contaminating the non-GFP cells (~%) (Additional File ). Conversely, we included transcripts that were considered enriched according to our statistical methods but originally scored as "absent" on the basis of PM vs MM signals used by Affymetrix MAS . software (see Methods, Additional file ). This second adjustment simply acknowledges that enriched transcripts are clearly expressed and therefore should be scored as "present." We refer to the transcripts in these modified lists as EGs (Expressed Genes). A total of , EGs were detected in the Reference data set and , EGs in the unc-::GFP motor neuron data set (Fig ) (Additional Files , ). Overall , unique transcripts were detected in these experiments or about % of all predicted C. elegans ORFs [] (Additional file ). These results are comparable to microarray data from whole embryos that also detected about half of the predicted C. elegans genes []. Genes that are not detected may be expressed in a relatively small number of cells. This point is substantiated by our finding that EGs in the unc-::GFP motor neuron data set are not scored as present in the Reference data set (Additional file , Fig ). For example, the transcription factor UNC- is normally expressed in a small subset of embryonic neurons including the DAs []. The unc- transcript is enriched in the unc-::GFP motor neuron data set (Table , Additional file ) but is not detected in the Reference (Additional file ). Thus, it seems likely that the overall number of EGs should increase as additional classes of embryonic cells are profiled (RMF, SEV, SJB, DMM, unpublished data).Table 1Expression of promoter-GFP reporters for transcripts enriched in unc-::GFP motor neuron data set. Reporters were examined for expression in DA, SAB and I5 neurons. / reporters showed GFP expression (bold type) in these cells. GFP reporters are listed according to statistical rank. All GFP-positive reporters were visible in embryos (data not shown) but were scored in larval animals to ease neuron identification.RankCosmidGeneProteinUNC- neuronOther cells15F33D4.3flp-13neuropeptideI5ASE, ASG, ASK, BAG, DD, M3, M5, head neurons []17C11D2.6nca-1Ca++channelDADB, VA, VB, head/tail neurons56F09C3.2phosphataseDAVA, VB, VD, DB, intestine, hypodermis98T19C4.5novelno GFP161T23D8.2tsp-7tetraspaninDAall VNC motor neurons, head/tail neurons, touch neurons165CC4.2nlp-15neuropeptideDD, head/tail neurons, body muscles, pharyngeal muscle []210F29G6.2novelDADB, touch neurons, pharyngeal neurons, head neurons215F39G3.8tig-2TGF-βDAVA, VB, DB, body wall muscle, touch neurons, pharyngeal muscle233F55C12.4novelDAVB, DB, DD, AS, VD234E03D2.2nlp-9neuropeptideVA, intestine, head neurons []239ZC21.2trp-1Ca++channelDADB, VA, VB254Y47D3B.2Anlp-21neuropeptideDADB, VA, VB, AS, body muscle, head neurons, intestine []329F36A2.4twk-30K+ channelDAall VNC motor neurons []377C18H9.7rpy-1rapsynDAVD, AS, VB, DB, body muscles593F43C9.4amig-13CUB domainDADB, ant. VNC motor neurons, pharyngeal/intestinal valves, hypodermis []782T05C12.2acr-14nAChRDAVB, AS, DB, DD, HSN, VC4 & , AIY, head neurons, muscles, intestine788T27A1.6mab-9transcrip. factorDADD, DB, VD, AS877K02E10.8syg-1Ig DomainDAVA, HSN and other neuronsTable 2Summary of genes with enriched transcripts in unc-::GFP neurons. Genes are organized into categories according to molecular function (KOG or other description ). Gene families or functional groups with potential functions in neurons are emphasized in this list. Statistical rank is indicated for each transcript.Cosmid NameCommon NameRankKOG (Other description)Axon Guidance and OutgrowthB0273.4aunc-5934Netrin transmembrane receptor unc-5T19B4.7unc-40188Receptor mediating netrin-dependent axon guidanceF41C6.1unc-6616Netrin, axonal chemotropic factorF56D1.4aclr-1754Protein tyrosine phosphataseM79.1aabl-1882Protein tyrosine kinaseF09B9.2unc-115696Actin-binding LIM Zn-finger protein Limatin involved in axon guidanceB0350.2unc-44449AnkyrinC01G10.11aunc-76182Kinesin-associated fasciculation and elongation protein involved in axonal transportK10D3.2unc-(RUN domain protein required for axonogenesis and sex myoblast migration)Wingless SignalingK10B4.6cwn-1216Wnt family of developmental regulatorsY71F9B.5alin-17899Smoothened and related G-protein-coupled receptors (Frizzled Receptor)Acetylcholine Receptor SubunitsK11G12.2acr-266Acetylcholine receptorR01E6.4acr-12292Acetylcholine receptorF21F3.5unc-38399Acetylcholine receptorT05C12.2acr-14782Acetylcholine receptorY110A7A.3unc-63800Acetylcholine receptorF21A3.7212Acetylcholine receptorY105E8A.7lev-10323Cubilin, multiligand receptor mediating cobalamin absorptionT14A8.1ric-3237Unnamed protein (Required for nAChR assembly/trafficking)Ligand-gated Ion ChannelZC196.7glr-5792Glutamate-gated kainate-type ion channel receptor subunit GluR5 and related subunitsT27E9.941Ligand-gated ion channel (glycine/GABA)Y71D11A.5502Ligand-gated ion channel (glycine/GABA)Y46G5A.30snf-5686Sodium-neurotransmitter symporterC09E8.1a578Sodium-neurotransmitter symporterG-proteinsM01D7.7aegl-30124G protein subunit Galphaq/Galphay, small G protein superfamilyF08B6.2gpc-2469G protein gamma subunitF56H9.4gpa-9638G-protein alpha subunit (small G protein superfamily)G-protein Pathway ComponentsF28C1.2egl-10872G protein signaling regulatorsC05B5.7rgs-1486G protein signaling regulatorsF17C8.1acy-1884Adenylyl cyclaseR07E4.6kin-2467cAMP-dependent protein kinase types I and II, regulatory subunitC17F4.6gcy-19213Natriuretic peptide receptor, guanylate cyclaseC50H2.2egl-(Gαo coupled receptor)F57F5.5pkc-1747Serine/threonine protein kinaseF39B2.8928Predicted membrane proteinC24A8.41010STE20-like serine/threonine kinase MSTNeuropeptidesW07E11.3flp-231Unnamed protein (FMRF-like peptide)C18D1.3flp-(FMRF-like peptide)C03G5.7flp-(FMRF-like peptide)F33D4.3flp-1315Unnamed protein (FMRF-like peptide)E03D2.2nlp-9234Unnamed protein (Neuropeptide-like protein)CC4.2nlp-15165Unnamed protein (Neuropeptide-like protein)Y47D3B.2anlp-21254Unnamed protein (Neuropeptide-like protein)F13B12.5ins-(Insulin-like peptide)T28B8.2ins-18767Unnamed protein (Insulin-like peptide)F56A11.5928Uncharacterized Fe-S proteinNeuropeptide Processing and SecretionT03D8.3613Proprotein convertase (PC) chaperone involved in secretion (neuroendocrine protein 7B2)C32E8.7ric-1920Secretory vesicle-associated protein ICA69, contains Arfaptin domainZK897.1unc-31588Ca2+-dependent activator proteinNeuropeptide ReceptorT05A1.1npr- transmembrane receptor (neuropeptide receptor family)F59D12. transmembrane receptor (rhodopsin-like GPCR)T07D4. transmembrane receptor (rhodopsin-like GPCR)Y62E10A.-transmembrane receptor (rhodopsin-like GPCR)K07E8.5452Unnamed protein (FMRF receptor)C35A11.1814Unnamed protein (rhodopsin-like GPCR)ZC84. transmembrane receptor (rhodopsin-like GPCR)F56B6.5uvt- transmembrane receptor (somatostatin receptor)F56A11.5928Uncharacterized Fe-S protein (Neuropeptide receptor activity)Selected C. elegans genes are enriched in UNC- motor neuronsA majority of transcripts in the Reference and unc-::GFP motor neuron data sets show comparable levels of expression (Fig ). Many of these transcripts are likely to encode core functions required in every cell. Other transcripts in this group could be limited to subsets of embryonic cells that include UNC- motor neurons. Genes that are widely expressed in neurons, for example, may not be detectably enriched in unc-::GFP motor neurons in comparison to the Reference because neurons constitute a significant fraction (~%) of all cells in the embryo. To illustrate this point, we note that UNC- (Syntaxin), an integral component of the neurotransmitter release mechanism and therefore expressed in most neurons [,], is detected in the unc-::GFP motor neuron data set but is not enriched (Table , Additional Files , ).As graphically illustrated in the scatter plot shown in Fig. 3C, subsets of genes in the unc-::GFP motor neuron data set are differentially expressed relative to the average expression levels for all cells in the Reference data set (R2 = .). As expected for a gene that is selectively expressed in unc-::GFP neurons, the hybridization signal for the unc- transcript is highly elevated (×) in comparison to all cells. Significant numbers of genes are also under-expressed in UNC- motor neurons relative to other embryonic cells. Transcripts showing ≥ .× fold intensity difference in the unc-::GFP motor neuron vs Reference data sets were defined using SAM statistics at a False Discovery Rate (FDR) of ≤ %. By these criteria genes are enriched (red) in UNC- motor neurons (Additional file ) whereas transcripts are depleted (green) (Additional file ). The threshold of ≥ .× fold was defined empirically. At higher values (e.g. ≥ .×) genes with known expression in these cells were excluded (e.g. acr-, unc-) [,] (Additional file ) whereas, a lower threshold of .× included significantly more false positives (e.g. muscle genes, pat-, sup-) [,].Confirmation of UNC- motor neuron genesInformation gleaned from published literature and from wormbase , identified genes with known expression in embryonic motor neurons that also express unc-::GFP (I5, SAB, DA) (Additional file ). We detect (%) of these genes as EGs of which (%) are enriched. In addition, a significant number of transcripts encoding core neuronal functions (e.g. axon guidance, neurotransmitter signaling, etc.) are detected in the unc-::GFP data set (Table , Additional file ). For example, in addition to UNC- (syntaxin or t-SNARE,) other components of the SNARE complex, SNB- (synaptobrevin or v-SNARE) and SNAP- (Y22F5A.) are detected [,]. We also examined the data set for potential false positives by considering transcripts that are known to be highly expressed in other tissues but not in UNC- motor neurons. For example, in the embryo, the homeodomain protein UNC- is exclusively detected in DD motor neurons. Expression of the GABA vesicular transporter, UNC-, in DD motor neurons depends on unc- function []. UNC- motor neurons are cholinergic and as expected neither of these GABA specific transcripts are present in the unc-::GFP motor neuron data set (Additional file ).The strong representation of ~% of genes known to be expressed in I5, SAB, and DA motor neurons in the unc-::GFP motor neuron dataset indicates that other previously uncharacterized transcripts in this list are also likely to be expressed in these cells in vivo. To test this idea, we evaluated GFP reporter lines for representative genes detected as enriched in the unc-::GFP motor neuron data set (Fig. ). As shown in Table , % (/) of these promoter-GFP fusions show expression in UNC- motor neurons in vivo. Of the GFP reporters not detected in these neurons, one of them, T19C4., fails to express GFP in any cell. This finding could mean that the upstream sequence selected for this construct does not overlap the endogenous T19C4. gene regulatory region. In some cases, cell-specific expression of C. elegans genes depends on distal upstream regions, intronic sequences, or ' domains that would not be included in these ' promoter GFP fusions []. This explanation could also account for the apparent absence of GFP expression in the unc-::GFP motor neurons of the nlp- and nlp- GFP reporters. The validity of this data set is further substantiated by the observation that GFP expression in DA motor neurons is detected even for lower ranking genes (e.g. syg-::GFP, statistical rank = ). Thus, we believe that the transcripts listed in the unc-::GFP motor neuron data set are likely to constitute an accurate representation of genes normally expressed in these cells.We note that the positive GFP reporters shown in Table are not uniformly detected in UNC- neurons: all but one (flp-) are expressed in the DAs, one in I5 and none in the SAB motor neurons. This bias reflects the relative abundance of DA motor neurons (~% or / of unc-::GFP neurons in vivo) in the cells used to generate this data set and thus could indicate that most of the enriched transcripts are also expressed in the DAs. Therefore, results presented below are largely focused on potential gene functions in DA motor neurons.Families of neuronal genes expressed in UNC- motor neuronsHere we describe transcripts detected in the unc-::GFP dataset with an emphasis on genes that are enriched in these cells and therefore likely to encode proteins with important roles in the differentiation or function of UNC- motor neurons (Table ). A comprehensive discussion of gene families from this list can be found in Additional file . Selected examples are presented here. Gene names for enriched trancripts discussed in this section are shown in bold and are listed in Table . All EGs are listed in Additional file .Axon guidance and outgrowthGrowth cone steering and cell migration along the dorsal-ventral body axis in C. elegans depend on the UNC-/netrin guidance cue. The UNC-/DCC receptor mediates an attractive response to UNC-/netrin whereas co-expression of UNC-/DCC with a second UNC- receptor, UNC-, results in repulsion [,]. The UNC-/netrin signal is released from ventral ectoderm [] to repel growth cones expressing both UNC- and UNC-; this interaction is required for normal outgrowth of DA motor neuron commissures to the dorsal nerve cord []. As expected, unc- and unc- transcripts are enriched in UNC- motor neurons. unc-, which is known to be expressed in the I5 pharyngeal neuron, is also elevated []. The CLR- receptor protein tyrosine phosphatase (RPTP) is proposed to inhibit attractive UNC-/netrin responses via interactions with UNC-. In the DA motor neurons, CLR- also promotes UNC-/netrin repulsion by an UNC--independent mechanism []. As predicted by these models, the clr- transcript is elevated in UNC- motor neurons. Relevant to this point, we note that the C. elegans Abelson tyrosine kinase ortholog, abl-, is also enriched. In Drosophila, Abl tyrosine kinase antagonizes the axon guidance role of RPTPs in motor neurons []. It will be interesting to determine if ABL- functions similarly in C. elegans and, in this case, if ABL- works in opposition to CLR- during DA motor axon outgrowth (Fig. ).Figure 6Model of DA motor neuron axon guidance. Ventrally derived UNC-/Netrin guidance cues binds to the UNC-/DCC and UNC- receptor to steer the DA motor axon toward the dorsal nerve cord. The receptor tyrosine phosphatase, CLR-, promotes dorsal motor axon outgrowth via an UNC-/DCC independent pathway []. The transcript encoding the C. elegans ortholog of Abelson tyrosine kinase (ABL-) is enriched in the unc-::GFP motor neuron data set and is proposed to antagonize CLR- activity.We also detected axon guidance effectors unc- and ced- in our microarray array dataset. Genetic approaches have shown that unc- (AbLIM, actin binding protein) and ced- (Rac GTPase) are downstream effectors of UNC- signaling and presumptive links to the cytoskeleton [,].Transcripts for genes with general roles in axon outgrowth are enriched in the unc-::GFP motor neuron data set. These include unc- (ankyrin-like), unc- (novel) and unc- (RUN domain). All three of these genes are highly expressed in the C. elegans nervous system. unc- encodes multiple alternatively spliced transcripts with broad roles in axonal morphogenesis []. UNC- and its vertebrate homologs define a new protein class of unknown biochemical function. In C. elegans, unc- mutants show axon outgrowth and fasciculation defects []. unc- and unc- (serine/threonine kinase) mutants display similar neuronal defects with misplaced processes and enlarged abnormal varicosities []. UNC- (EG) has been proposed to phosphorylate UNC- to regulate vesicular trafficking during axonal process outgrowth [,].Wingless signalingWingless (Wnt) signaling controls multiple developmental processes in the nervous system ranging from cell determination to axon guidance and synaptogenesis [,]. The C. elegans genome contains Wnt genes and Wnt receptors or Frizzled homologs []. One of each, cwn- (Wnt) and lin- (Frizzled), are enriched. Transcripts for other components of the canonical (mig-, mom-, cwn-, dsh-, dsh-, Y73B6BL.) and noncanonical (lit-, mom-, par-, tap-) Wnt signaling pathways are detected as EGs. Thus, UNC- motor neurons are presumptively competent to send as well as respond to Wnt signals. Functions for Wnt signaling in the C. elegans motor neuron circuit have not been defined. Possibilities include the regulation of synaptogenesis as suggested by studies of Drosophila motor neurons which secrete Wnt to control both presynaptic and postsynaptic differentiation at the neuromuscular synapse []. A gradient of Wnt signaling controls cell migration along the AP axis in C. elegans []. Responsiveness to this graded Wnt signal could account for the anterior polarity of DA motor neurons in the dorsal nerve cord as suggested by the recent finding that commissural axonal polarity along the AP axis in the vertebrate spinal cord is dependent on Wnt signaling [].Nicotinic Acetylcholine Receptors (nAChRs)The C. elegans genome encodes at least distinct nAChR subunits []. Two of these, ACR- and UNC- are expressed in DA class motor neurons [,] and are enriched in the unc-::GFP motor neuron data set. Expression of unc- [] and unc- (J.L. Bessereau, personal communication) in ventral cord motor neurons has been previously reported and these are also detected as EGs (Additional file ). acr-::GFP is expressed in neurons (A. Gottschalk and W. Schafer, personal communication), and we have validated the enrichment of acr- with GFP reporters that confirm expression in DA motor neurons (Fig ). In body muscle, UNC- is an essential component of a levamisole-sensitive nACh receptor that also includes UNC-, UNC-, LEV- and LEV- [,]. ACR- may coassemble with UNC-, UNC-, and UNC- to generate a related nACh receptor in UNC- motor neurons (A. Gottschalk and W. Schafer, personal communication). Five additional sets of nAChR subunits are detected as EGs and a so-called "orphan" ligand gated ion channel (LGIC) subunit, F21A3., with significant similarity to the nAChR gene family, is enriched. Despite the diversity of nAChR subunits expressed in UNC- motor neurons and the potentially complex array of resultant receptors, no functions have been directly assigned to nAChRs in these cells []. Although loss-of-function mutations in nAChR subunits that are also expressed in muscle (i.e. unc-, unc-, unc-) result in locomotory defects, gene knockouts of acr- (Y. Jin, personal communication) and acr- (data not shown), which are exclusively expressed in neurons, do not produce obvious effects on motility or behavior. Perhaps the surprisingly large number () of nAChR subunit genes detected in these cells results in overlapping functions that mask defects in single gene knockout mutants. Alternatively, these nAChRs may mediate subtle aspects of motor neuron activity. This idea is consistent with models in which nAChRs act presynaptically to modulate neurotransmitter release [,]. Finally, we detect enrichment of transcripts for proteins RIC- (novel) and LEV- (CUB domain) that mediate nAChR localization [,].Ligand-Gated Ion ChannelsUNC- motor neurons are potentially responsive to additional classes of neurotransmitters. Enrichment of glr- (kainate type ionotropic glutamate receptor subunit) is correlated with its known expression in the SAB motor neurons []. As members of the GABA/Glycine family of ligand-gated receptors, the presumptive anion channels encoded by T27E9. and Y71D11A. are predicted to hyperpolarize UNC- motor neurons and thus inhibit cholinergic activity []. It may be significant that a candidate sodium/chloride-dependent glycine transporter, snf-, is enriched. (C09E8., an outlier in the sodium/chloride-dependent transporter family is also enriched.) In mammalian cells, plasma membrane transporters GLYT1/GLYT2 remove glycine from the synaptic cleft, and in the case of GLYT2, thereby recycle glycine for reuptake into synaptic vesicles []. UNC- motor neurons do not express the GABA/Glycine vesicular transporter, UNC-, however, and are therefore unlikely to release glycine presynaptically []. In this case, the physiological function of the SNF- transporter could mirror that of GLYT1, which is believed to attenuate glycinergic signaling by pumping glycine into a non-glycinergic glial cell []. To date, the potential function of glycinergic signaling in C. elegans has not been explored.G-protein signalingCholinergic motor neuron activity in C. elegans is modulated by G-protein signaling pathways that respond to the neurotransmitters acetylcholine, serotonin (-HT), and dopamine (Fig. ) [-]. In each case, acetylcholine release is either promoted by EGL- (Gαq) or inhibited by GOA- (Gαo). Input to these antagonistic pathways is provided by G-protein coupled receptors (GPCRs). Pharmacological evidence suggests that a muscarinic acetylcholine receptor activates EGL- to enhance ACh release at the neuromuscular synapse [,]. The enriched muscarinic AChRs, GAR- and GAR- could account for this effect [,]. Similarly the enriched -HT receptor, SER-, is a strong candidate for the GPCR mediating the inhibitory effect of serotonin on ACh release from ventral cord motor neurons []. Dopamine may either activate or inhibit ACh release within the same cholinergic motor neuron. Activation depends on DOP- which is enriched in UNC- motor neurons. Inhibition is attributed to DOP-. Expression of DOP- in cholinergic ventral cord motor neurons is reportedly weak and we do not detect the dop- transcript in our data set []. UNC- motor neurons are also potentially responsive to GABA as a transcript (Y41G9A.) encoding a metabotropic GABA type B1 receptor is enriched. GABA dependent effects on cholinergic motor neuron activity have not been previously reported in C. elegans.Figure 7G-protein signaling pathways regulating neurotransmitter release in cholinergic motor neurons. Components shown in bold are enriched in the unc-::GFP motor neuron data set. All others are EGs with the exception of the dop- transcript which is not detected (light gray text). See Table for protein descriptions. Green denotes interactions that promote acetylcholine (ACh) release and red marks steps that inhibit synaptic vesicle fusion. Neurotransmitters are highlighted in gray boxes. This figure adapted from Reynolds et al. () [].Genetic screens for mutations affecting neurotransmitter release have revealed a complex web of interacting components that couple G-protein signaling to synaptic vesicle fusion (Fig ) [,,,] (D. Sieburth and J. Kaplan, personal communication). With one exception (dop-), transcripts for all of the known components of these pathways are either enriched (acy-, egl-, gpc-, kin-, pkc-) or detected as EGs (dgk-, egl-, egl-, gpb-, gsa-, kin-, ric-, unc-). Lack of enrichment of some of these components is consistent with the widespread utilization of G-protein signaling pathways in C. elegans neurons and muscle cells [,]. As noted above, these data have also revealed several additional enriched transcripts with potential roles in G-protein dependent locomotory behavior. egl- encodes an orphan GPCR and rgs- an RGS protein, both of which can regulate goa- signaling in the egg laying circuit [,]. RNAi of F39B2., which encodes a highly conserved but unusual protein with both serine/threonine kinase and -transmembrane domains, results in a locomotory defect [] that could be indicative of a neuromodulatory function in DA motor neurons. A complete list of G-protein signaling components detected in this dataset can be found in Table .Table 3Summary of G-protein signaling genes in unc-::GFP neurons. Genes either known to function in C. elegans motor neuron G-protein signaling pathways or likely to have a role therein (e.g. SER-) are listed here with KOG descriptions. Transcripts are listed as either enriched with statistical rank (top) or as EGs (bottom).Cosmid NameCommon NameStatistical RankKOG DescriptionEnrichedF47D12.1gar- transmembrane receptorY40H4A.1gar-3421Muscarinic acetylcholine receptorF15A8.5ddop- transmembrane receptorY22D7AR.13ser- transmembrane receptorY41G9A.4769GABA-B ion channel receptor subunit GABABR1 and related subunits, G-protein coupled receptor superfamilyM01D7.7aegl-30124G protein subunit Galphaq/Galphay, small G protein superfamilyF28C1.2egl-10872G protein signaling regulatorsF57F5.5pkc-1743Serine/threonine protein kinaseF17C8.1acy-1884Adenylyl cyclaseR07E4.6kin-2467cAMP-dependent protein kinase types I and II, regulatory subunitPresent (EGs)Y69A2AR.1ric-8Signaling protein RIC-/synembryn (regulates neurotransmitter secretion)C16C2.2aeat-16G protein signaling regulatorsB0348.4aegl-8Phospholipase CF52A8.2gpb-2G-protein beta subunitC26C6.2goa-1G-protein alpha subunit (small G protein superfamily)ZK542.2aunc-13Neurotransmitter release regulator, UNC-13C09E10.2adgk-1Diacylglycerol kinaseR06A10.2gsa-1G protein subunit Galphas, small G protein superfamilyNeuropeptide signalingThe C. elegans genome includes a large and diverse array of genes encoding potential neuroactive peptides. GFP reporter studies indicate that these genes are predominantly expressed in neurons. "flp" genes encoding FMRFamide and related peptides (FaRPs) have been described. FaRPs have been shown to modulate a wide array of invertebrate neural functions []. Previously reported expression of flp-, flp-, flp- in the pharyngeal I5 neuron [] (Fig ) is confirmed by their enrichment in the unc-::GFP motor neuron data set (Table I). flp- is also elevated in these cells and additional flps are detected as EGs (Additional Files , ). Specific FaRPs modulate cell excitability (flp-), locomotion (flp-) and feeding behavior (flp-) in C. elegans [,]. The inhibitory action of the FLP- peptide on pharyngeal muscle activity is consistent with its expression in I5 [].The C. elegans genome contains genes encoding predicted insulin-like peptides []. Transcripts for two of these, ins- and ins-, are enriched; ins-, ins- and ins- are present but not significantly elevated relative to other cells. ins- and ins- have been implicated in the DAF- insulin receptor dependent pathways regulating growth, metabolism and lifespan [].A total of genes encoding other potential classes of neuropeptides have also been identified in the C. elegans genome. Three of these neuropeptide-like protein (nlp) genes, nlp-, nlp-, and nlp-, are enriched in UNC- motor neurons (Table ). An additional group of nlp transcripts are detected as EGs (Additional File , ). To date, no functions have been directly assigned to nlp genes in C. elegans [].Our studies have revealed that a surprisingly large number of neuropeptide genes are transcribed in UNC- motor neurons. These results indicate that UNC- motor neurons are likely to exhibit significant neurosecretory activity. This conclusion is consistent with our finding that proteases required for neuropeptide processing and activation [T03D8. (Proprotein convertase (PC) chaperone), egl- (zinc carboxypeptidase) and egl- (subtilisin-like proprotein convertase)] are also expressed in these cells [-]. Other genes with important roles in neurosecretion are also detected. ric- encodes a novel arfaptin-related protein that is believed to function in the Golgi in the generation of neurosecretory granules and may through this activity and subsequent neuropeptide signaling exert an indirect effect on ACh release from conventional synaptic vesicles [,]. Our finding that ric- is highly enriched in cholinergic motor neurons could be indicative of autocrine neuropeptide modulation of ACh secretion at the neuromuscular synapse. Consistent with this idea is our finding that ida-, a conserved membrane component of the dense core vesicles in which neuropeptides are typically packaged, is an EG []. Finally, UNC- (CAPS), a known facilitator of dense core vesicular release, is enriched []. Plasma membrane fusion of both dense core vesicles and the small, clear vesicles in which classical neurotransmitters are packaged, depend on a common set of calcium-activated components [] most of which are either enriched or present in these cells (see Table and Additional Files , ).In addition to secreting neuropeptides, UNC- motor neurons are also likely to respond to neuropeptide signaling. Transcripts for nine putative neuropeptide receptors are enriched. (Table ). RNAi of two of these, npr- and F59D12., results in locomotory defects that could be indicative of specific functions in DA motor neurons []. npr- is a close relative of npr- (not detected) which has been shown to control social feeding behavior in response to the FLP- (not detected) peptide []. F59D12. is most closely related to melatonin receptors but its in vivo ligand is unknown. Neuropeptides are believed to modulate secretion of classical neurotransmitters []. Neuropeptide specific effects on excitatory motor neuron activity in the Ascaris ventral nerve cord have been reported []. Genetic evidence in C. elegans indicates that acetylcholine release at the neuromuscular junction is enhanced by neuropeptide activity [] and that this pathway depends on the EGL- Gqα protein [] (Table ). These neuropeptides may be released from neurons and also as a retrograde signal from muscle cells [,].Other classes of enriched transcripts are discussed in Additional file (Transcription factors, Cell Adhesion Molecules, Synapse-Associated Proteins, Neurotransmitter Vesicular Release Components, TGF-β Signaling Proteins, Serpentine Receptors, Calcium Channels, Calcium Ion Binding Proteins, Potassium Channels, Innexins, DEG/ENaC Channels, and Stomatins).DiscussionWe have described MAPCeL, a microarray-based strategy for fingerprinting specific C. elegans neurons, and provide evidence that this approach can reveal a comprehensive picture of gene expression in these cells in vivo. unc-::GFP-marked neurons were isolated by FACS from primary cultures of embryonic cells and profiled on the C. elegans Affymetrix gene chip. Because these unc-::GFP neurons differentiate in vitro, it was important to establish that our microarray data provide an accurate representation of gene expression in the intact animal. This conclusion is supported by three observations: () A majority (/) of genes with known expression in unc-::GFP neurons are detected in our microarray data set; () ~% (/) of GFP reporters constructed for transcripts enriched in UNC- motor neurons are expressed in these cells in vivo (Table , Fig. ); () Transcripts known to encode proteins with key roles in unc-::GFP motor neuron differentiation (e.g. axon guidance and outgrowth, synaptogenesis) and function (e.g., neurotransmitter vesicle release, G-protein signaling pathways) are highly represented in our data sets. These findings parallel earlier studies showing that cultured C. elegans neurons and muscle cells adopt apparently normal morphological and physiological characteristics [] and are consistent with evidence favoring a cell autonomous mode of differentiation for C. elegans embryonic cells after an initial phase of inductive signaling events [,]. We have now generated comparable microarray profiles of other motor neuron classes and muscle cells that also show strong congruence with known patterns of gene expression (RMF, SEV, SJB, DMM, unpublished data). We therefore conclude that our approach of profiling GFP marked neurons isolated from primary culture can now be widely applied to fingerprint specific C. elegans embryonic cells. In some cases, however, differentiation of a given neuron is likely to depend on specific intercellular signals that primary cultures will not provide. Thus, in every instance, it will be necessary to confirm microarray data by independent methods as described here.Methods for profiling specific C. elegans cellsPrevious studies have described other methods for cataloging transcripts from specific C. elegans cells. Comparisons of microarray data from mutant animals with either supernumerary or absent sensory neurons in the male tail, have revealed genes that are preferentially expressed in these cells []. However this approach is limited to cell types that can be manipulated by specific mutants. In addition, this method may be insufficiently sensitive to detect changes in smaller subsets of cells due to high background mRNA from cells that are not affected by the mutation (SEV, DMM, unpublished data). This limitation can be overcome by enriching for mRNA from target cells. To this end, Zhang et al. () used an approach similar to the strategy outlined in this paper to identify downstream genes of the MEC- transcription factor in C. elegans touch neurons []. However, this work did not provide a comprehensive cell-specific gene expression profile as we have here perhaps due to the limited enrichment (~%) of GFP-labeled touch neurons. We have now optimized the application of nematode embryonic cell culture and FACS technology to obtain ~% enrichment of GFP-marked neurons and muscle cells (Fig. ) (RMF, SEV, SJB, DMM, unpublished data). These methods have now been successfully applied to profile other classes of C. elegans embryonic cells [,].MAPCeL cannot be used for postembryonic cells because these apparently do not arise in culture []. Microarray profiles of specific larval cells have been obtained, however, by mRNA tagging. In this approach, an epitope-labeled polyA binding protein (FLAG-PAB-) is expressed transgenically under the control of a cell-specific promoter and mRNAs isolated by co-immunoprecipitation with anti-FLAG. This method has been used for microarray analysis of C. elegans body muscle cells and ciliated sensory neurons [,]. We have now successfully used the mRNA tagging strategy to profile specific subsets of motor neurons from C. elegans larvae (SEV, RMF, J. Watson, S. Kim, P. Roy, DMM, unpublished data). Thus, in principle, it should now be possible to obtain an accurate gene expression profile for virtually any C. elegans cell throughout development.UNC- motor neurons are sensitive to a wide range of neurotransmitters and peptidergic signalsAcetylcholine (ACh) release at the DA neuromuscular junction is presumptively triggered by excitatory input from command interneurons. The strength of the DA cholinergic signal, however, may be strongly modulated by other cells that release neurotransmitters from distal locations. For example, dopamine is produced by neurons, none of which are presynaptic to DA motor neurons []. Dopamine, however, is a potent regulator of cholinergic secretory activity in the ventral motor circuit. The dopamine effect is mediated in part by DOP-, a G-protein coupled receptor (GPCR) []. We have confirmed enrichment of the dop- transcript and also detected elevated levels of transcripts encoding GPCRs for acetylcholine and serotonin, additional neurotransmitters known to modulate cholinergic motor neuron activity via G-protein signaling pathways [,]. Enrichment of a GABA metabotropic receptor transcript offers yet another mechanism for exogenous adjustment of neurotransmitter vesicular fusion in DA motor neurons. Indirect evidence indicates that acetylcholine release from ventral cord motor neurons may also be sensitive to neuropeptide signals from other neurons or muscle cells [,]. We have established that unc-::GFP motor neurons express elevated transcript levels for nine different GPCRs with significant homology to insect or mammalian neuropeptide receptors. This signaling complexity is further compounded by the enrichment of transcripts for members of the serpentine GPCR-like family in unc-::GFP neurons (Additional Files , ). Ligands for this outlier group of GPCRs are unknown []. The picture emerging from these data is of a motor neuron festooned with multiple G-protein linked receptors each responding to a different class of neurotransmitter or peptidergic signal (Fig ). In effect, these motor neurons are also functioning as a kind of sensory neuron in which disparate inputs are internally assessed to fine-tune output in concert with temporal requirements for locomotory activity.Figure 8Signaling components detected in unc-::GFP motor neurons.The microarray data also reveal multiple additional classes of receptors and ion channels through which the differentiation and function of unc-::GFP motor neurons could be modulated by extracellular signals (Fig ). Finally, we have detected enrichment of transcripts encoding TGF-β, wingless, and several classes of neuropeptides (Table , Additional Files , ). Thus, in addition to responding to a wide range of stimuli, unc-::GFP motor neurons are also potentially capable of regulating the activities of other cells with a variety of different signals. If an organism as simple as C. elegans builds motor neurons with such sophisticated signaling and response mechanisms, it is tempting to speculate that neurons in other, more advanced species may have evolved even more complex pathways. It is likely, however, that the core signaling systems described here are also conserved. This prediction is underscored by our finding that approximately half of the enriched transcripts (/) and / of EGs (/) detected in unc-::GFP neurons have human homologs (BLAST ≤ e-) (Additional Files , ).Applications of MAPCeLIn addition to confirming expression of genes with known roles in unc-::GFP motor neuron differentiation and function, the microarray data also uncovered strong candidates for new genes governing these events. For example, DA motor axons grow dorsally in response to a ventrally provided repulsive UNC-/netrin guidance cue []. Recent work has shown that the receptor protein tyrosine phosphatase (RPTP), CLR-, positively enhances this response []. As expected, we found that the clr- transcript is enriched in the unc-::GFP motor neuron data set. We also noted enrichment of abl-, the C. elegans homolog of Abelson tyrosine kinase. By analogy to findings in Drosophila in which Abelson tyrosine kinase functions in opposition to RPTP-dependent axon guidance [], we propose that ABL- antagonizes CLR- activity (Fig ). This model predicts that either genetic ablation or RNAi of abl- will suppress the DA motor axon guidance defects of clr- mutants.Another application of this strategy includes the identification of transcription factor target genes. A comparison of expression fingerprints of wildtype cells vs cells that are mutant for a specific transcription factor could reveal downstream genes []. For example, the UNC- transcription factor regulates axon morphology and synaptic strength in embryonically derived unc-::GFP neurons []. UNC- also defines the specificity of synaptic inputs to postembryonically-derived VA motor neurons [,,]. We have now used a combination of MAPCeL and mRNA tagging strategies to identify candidate genes regulated by UNC- in these cells (RMF, SEV, DMM, unpublished data). Gene regulatory motifs to which transcription factors bind may also be revealed as common cis-acting sequences in cohorts of co-regulated genes [].The C. elegans nervous system is composed of exactly neurons. The morphology and connectivity for every neuron has been defined by serial section electron microscopy to generate a detailed wiring diagram for the entire network []. The C. elegans genome is similarly well defined. All chromosomes are completely sequenced and the structure of over , genes described []. Unique combinations of these genes are likely to specify different classes of neurons and their differentiated traits. The problem now is to link the gene map with the wiring diagram. We believe that MAPCeL offers a powerful approach toward achieving this goal.ConclusionWe have described a new method, Micro-Array Profiling of C. elegans cells, or MAPCeL, for generating gene expression fingerprints of subsets of C. elegans neurons. Embryonic motor neurons marked with a reporter gene, unc-::GFP, were isolated by FACS from primary cultures and profiled on the C. elegans Affymetrix array. We confirmed that microarray data generated by this approach reliably identify genes expressed by these motor neurons in vivo. We propose that MAPCeL can now be used to generate a gene expression map for the C. elegans nervous system.MethodsCell CultureEmbryonic cells were obtained using methods previously described []. Briefly, embryos were isolated from gravid adults following lysis in a hypochlorite solution. Intact embryos were separated from debris by flotation on % sucrose. Eggshells were removed by incubation in . ml chitinase (. U/ml in egg buffer) for minutes. Following resuspension in L- medium supplemented with % FBS (L15-) and antibiotics, the embryos were dissociated by passage through a 5μm syringe filter (Durapore). Cells were plated on poly-L-lysine (.%, Sigma) coated single-well chambered coverglasses (Nalge Nunc International) at a density of ~ million cells/ml and maintained in L15- media. Cells were incubated at °C in a humidified chamber. Wildtype (N2) cells were isolated and treated similarly.FACS analysisSorting experiments were performed on a FACStar Plus flow cytometer (Becton Dickinson, San Jose, CA) equipped with a nm argon laser. Emission filters were ± nm for GFP fluorescence and ± nm for PI fluorescence. The machine was flushed with egg buffer [] prior to sorting to enhance viability. μm fluorescent beads were used to calibrate light scattering parameters for the relatively small size of C. elegans embryonic cells. Cells were sorted at a rate of – cells per second through a μm nozzle.Immediately prior to sorting, supernatant from the hour cultures was removed and discarded. ml of egg buffer was added to the chamber coverglass. Cells are loosely adherent to poly-L-lysine and can be easily dislodged with gentle pipetting. ml of egg buffer + cells were drawn into a cc syringe and the suspension filtered with a μm Durapore syringe filter. Propidium Iodide (PI) was added to the cell suspension at a final concentration of μg/ml prior to sorting. Autofluorescence levels were established by flow cytometry of cells isolated from wildtype (i.e. non-GFP) embryos (See Fig 2A). Next, wildtype cells stained with PI were used to define the sorting gate for damaged cells. GFP+ cells containing no PI were sorted to establish the intensity range of GFP fluorescence. Finally, unc-::GFP cells stained with PI were gated (Fig 2B) using the parameters established above. The sorting gate for size and granularity (Fig 2C) was empirically adjusted to exclude cell clumps and debris and to achieve ~% enrichment for GFP-labeled cells. unc-::GFP cells were collected in a ml conical tube containing ml of L15- media. Cells were pelleted using low-speed cenrifugation ( × g) and either plated on peanut lectin-coated slides for visualization [] or used for RNA isolation (see below). Reference cells were obtained from day old cultures of embryonic blastomeres isolated from the non-GFP wildtype strain (N2). In this case, all viable cells (i.e. non-PI stained) were collected by FACS for RNA isolation.RNA isolation, amplification, and hybridizationRNA was prepared from FACS-isolated unc-::GFP cells for comparison to RNA from the wildtype reference strain (N2). Cells were pelleted using low-speed centrifugation ( × g). The supernatant was removed and RNA was extracted with a micro-RNA isolation kit (Stratagene) using the recommended volumes for million cells. Typical yields were pg total RNA/cell. ng of total RNA was subjected to rounds of amplification, as described in the Affymetrix GeneChip Eukaryotic Small Sample Target Labelling Protocol, with the following modifications. ng ( pmol) of T7-dT primer ('-GGCCAGTGAATTGTAATAC GACTCACTATAGGGAGGCGG-(dT)-') was used as opposed to the recommended pmol. RNA cleanup was achieved using the RNeasy mini kit (Qiagen); μl of % ethanol (final concentration = % ethanol) was added to the sample prior to absorption to the column matrix. Eluate was passed through the column × prior to washing to improve yields. The BioArray High Yield RNA Transcript Labeling Kit (Enzo) was used to biotinylate the sample in the second round of amplification. – μg of labeled aRNA (amplified RNA) was fragmented and hybridized to the Affymetrix C. elegans chip according to the Affymetrix Expression Analysis Technical Manual. The Agilent Bioanalyzer was used to assess RNA quality prior to labeling and to confirm fragmentation (< bp) before hybridization.Data analysisThe commercially available C. elegans Affymetrix array was used for all experiments. This chip was designed using the December genome sequence. All probe set information is available at as well as . unc-::GFP neurons were profiled in triplicate; baseline data (all cells) were obtained from four independent experiments with wildtype embryonic cells. Hybridization intensities for each experiment were scaled in comparison to a global average signal from the same array (A complete list of Affy normalized values can be found in Additional file ) []. Expressed transcripts were initially identified on the basis of a "Present" call in a majority of experiments (/ for unc-::GFP and / for wildtype cells) as determined by Affymetrix MAS . (see below) (Additional Files , ). In this approach, a Mismatch (MM) value for each feature is compared to a Perfect Match (PM) value to estimate non-specific binding. This strategy, however, tends to arbitrarily exclude low intensity signals in which PM and MM values may be comparable [,]. To avoid this bias in the detection of transcripts that might be differentially elevated in the unc-::GFP data set, intensity values were normalized using RMA (Robust Multi-Array Analysis) available through GeneTraffic (Iobion) in which the MM values are not considered (Additional file ) [,]. Comparisons of RMA normalized intensities for unc-::GFP vs reference cells were statistically analyzed using Significance Analysis of Microarrays software (SAM, Stanford) [,]. A two-class unpaired analysis of the data was performed to identify genes that differ by ≥ .-fold from the wildtype reference at a False Discovery Rate (FDR) of ≤ %. These genes were considered significantly enriched (Additional file ). This analysis also identified ~ transcripts that are depleted (.×, ≤ % FDR) in unc-::GFP cells vs the wildtype reference (Additional file ). Although of these transcripts are also scored as "present" in the unc-::GFP motor neuron dataset, we attribute their detection to high expression in the small fraction (~%) of non-GFP cells contaminating this preparation (see above). Therefore, we excluded all of these wildtype-enriched transcripts from the list of present calls in the unc-::GFP motor neuron data set (Additional file ). Finally, to compute the overall sum of Expressed Genes (EGs) in the unc-::GFP data set we restored unc-::GFP-enriched genes that were initially excluded from the present list due to high mismatch signals. These considerations produce a final list of , genes that are detected in unc-::GFP motor neurons (Additional file ). (see Logic Tree, Additional file ).Annotation of datasetsA Wormbase mirror was established by downloading code and databases from . Using the acedb perl module, an annotation script was generated that queries the wormbase mirror. Affymetrix IDs have been mapped to specific transcripts in wormbase. Text files containing Affy IDs (one per line) and cosmid names are input into the script which then searches the wormbase mirror and matches Affy ID/cosmid name to a specific transcript. Cosmid names are used for this search when Affy IDs have not been mapped in wormbase. This information is used to acquire other linked annotations (i.e. KOG, common name, RNAi phenotype, Expression data, Kim mountain data and Gene Ontology, etc.).In litero analysisAn extensive literature search was performed using Textpresso . The keywords "DA motor neurons" generated a list of citations, a similar search was conducted using the keywords "I5 pharyngeal neuron" and "SAB neurons" that detected an additional citations. Expression patterns on wormbase were also searched using the "Cell identity" function to identify genes with documented expression in DA, SAB or I5. A list of genes with documented expression in DA motor neurons, the I5 pharyngeal neuron and the SAB neurons was compiled from this information (Additional file ).StrainsNematode strains were maintained at –°C using standard culture methods []. The wildtype strain was N2. Transgenic lines carrying promoter GFP fusions are listed in Additional file .Generating transgenic promoter GFP strainstwk-::GFP ( ng/ul) was microinjected with the myo-::dsRed2 marker ( ng/ul) []. Other transgenics were generated by biolistic transformation with promoter::GFP constructs from the Promoterome project (Additional file ). Primer sequences for "promoterome" constructs can be found at []. Microparticle bombardment was conducted [] in a BioRad Biolistic PDS-/He equipped with the Hepta Adapter. Gold beads ( micron) were coated with DNA at ug/ul. mm NGM plates were seeded with a monolayer of ~, L4/adult unc- (ed3) animals. For each construct, 'shot' was performed using a psi rupture disk at inches of Hg vacuum. After a hr recovery period, animals were washed from the plates with ml M9 buffer and transferred to NGM plates ( ml/plate). Animals were grown at °C for week. To pick transgenic animals, one-half of the plate was 'chunked' and added to a new mm NGM plate; animals with wildtype movement were picked to mm NGM plates and allowed to self. Worms derived from separate plates were considered independent lines; at least lines were obtained for each construct.MicroscopyTransgenic animals and cultured cells were visualized by differential interference contrast (DIC), or epifluorescence microscopy using either a Zeiss Axioplan or Axiovert compound microscopes. Images were recorded with CCD cameras (ORCA I, ORCA ER, Hamamatsu Corporation, Bridgewater, NJ). Some images were recorded on a Zeiss META confocal microscope.Authors' contributionsRMF developed the FACS protocol, RNA amplification methods, performed unc-::GFP microarray experiments, analyzed microarray data, generated/scored GFP reporters, and helped draft the manuscript. SEV helped develop RNA amplification and bombardment protocols, generated/scored GFP reporters and helped draft the manuscript. SJB generated the reference dataset. CS and KO wrote Perl scripts used to annotate datasets, make comparisons, and identify Present genes. JM offered advice on statistical methods. DD and MV provided unpublished GFP reporter constructs as part of the Promoterome project. DMM collected images of GFP lines in the confocal microscope, oversaw all aspects of the project and helped draft the manuscript.Supplementary MaterialAdditional File 1promoter::GFP fusions Strain names, source and references for GFP reporters used to validate microarray data.Click here for fileAdditional File 2Intensity Values Following Global Normalization (Affymetrix MAS .) Raw intensity values were scaled against a global average intensity value calculated for each chip using Affymetrix MAS .. Column A lists Affymetrix IDs for , features on the chip. These are indexed by rows with normalized intensity values for each independent hybridization and its Present/Absent (P/A) call. Columns labeled DMR , , , include intensity values for hybridizations using RNA from all N2 (wildtype) cells. Columns labeled DM39, , include the experimental hybridizations using RNA from FACS-isolated unc-::GFP neurons.Click here for fileAdditional File 3Intensity values after RMA normalization. Raw data from Additional file were normalized with Robust Multi-array Analysis (RMA) (GeneTraffic, version .) (Iobion). Normalized values are listed for the , features on the Affymetrix C. elegans chip. Affy IDs are listed in column A. Columns C-I show RMA normalized values for individual replicates. DMR , , , are the baseline hybridizations using RNA from all N2 (wildtype) cells. DM39, , are the experimental hybridizations using RNA from FACS-isolated unc-::GFP neurons.Click here for fileAdditional File 4N2 (wildtype) Expressed Genes (EGs) genes were called "Present" by MAS . in out of N2 hybridizations and are listed here as EGs. Affymetrix ID, Cosmid Name and common names (Columns A-C respectively) are given for each gene. Column D contains KOG descriptions and Column E indicates the dataset in which an individual gene is present.Click here for fileAdditional File 5unc-::GFP Present Genes genes were called "Present" by MAS . in out of unc-::GFP hybridizations and are listed here.Click here for fileAdditional File 6unc-::GFP Present Genes minus N2 Enriched Genes The list of unc-::GFP "present" calls in Additional file is likely to include transcripts that are highly expressed in the small fraction of contaminating non-GFP cells (~%). These transcripts were removed from the list of unc-::GFP Present calls (Additional file ) by subtracting genes that are highly enriched in N2 cells (Additional file ). The balance of genes is listed here with Affy ID, Cosmid Name, Common Name and KOG description.Click here for fileAdditional File 7unc-::GFP Expressed Genes (EGs) RMA normalization ignores the MAS . P/A calls and therefore does not discard some of the "absent" genes that are excluded from Additional file (unc-::GFP Present Genes). As a result, the unc-::GFP enriched genes (Additional file ) identified by SAM statistics actually includes genes that are not listed in Additional Files and . These additional unc-::GFP enriched genes were therefore restored to the list of unc-::GFP Present Genes in Additional file to provide a more accurate list of genes () that are actually expressed in unc-::GFP cells. These unc-::GFP Expressed Genes (EGs) are listed with Affy ID, Cosmid Name, Common Name and KOG description.Click here for fileAdditional File 8Expressed Genes (EGs) EGs from N2 (Additional file ) and from unc-::GFP (Additional file ) cells were combined to detect a total of , unique EGs.Click here for fileAdditional File 9unc-::GFP Enriched Genes Transcripts elevated . fold above baseline at a False Discovery Rate (FDR) ≤% were considered enriched. transcripts met these criteria and are listed according to statistical rank (Column A). Annotation includes Affy ID, Cosmid Name, Common Name and KOG description (Columns B-E); SAM score, Fold Change and q-value are listed (Columns F-H). We attribute the high ranking () of dpy- (T22B3.) to its use as a coselectable marker in the generation of the unc-::GFP transgenic line [].Click here for fileAdditional File 10N2 Enriched Genes genes are enriched . fold (FDR ≤%) in N2 cells compared to unc-::GFP neurons. These are listed according to statistical rank (Column A). Annotation includes Affy ID, Cosmid Name, Common Name and KOG description (Columns B-E); SAM score, Fold Change and q-value are listed (Columns F-H).Click here for fileAdditional File 11Genes previously known to be expressed in unc-::GFP neurons. Published literature and wormbase were searched to identify genes that are expressed in embryonic neurons that also express unc-::GFP (I5, DA, SAB).Click here for fileAdditional File 12unc-::GFP motor neuron enriched genes with human homologs. Enriched genes (Additional file ) and human genes listed here have BLAST scores ≤ e-.Click here for fileAdditional File 13unc-::GFP EGs with human homologs. Expressed genes (Additional file ) and human genes listed here have BLAST scores ≤ e-.Click here for fileAdditional File 14Summary of unc-::GFP enriched and expressed (EGs) transcripts in selected categories. This file is an expanded version of Table (see text) in which EGs are added to the list of enriched genes. Genes are organized as in Table according to molecular function (KOG description, other description ). Statistical rank is indicated for each enriched transcript and "EG" designates transcripts that are EGs.Click here for fileAdditional File 15Genes Expressed only in unc-::GFP neurons. Affy ID, Cosmid Name, Common Name and KOG description for genes that are expressed in unc-::GFP neurons. These genes are not detected in the reference (N2) dataset.Click here for fileAdditional File 16Gene families represented in unc-::GFP neurons. A comprehensive description of neuronal transcripts organized according to gene family.Click here for fileAdditional File 17Logic tree of data analysis methods.Click here for file
|
PMC2362782.txt
|
TITLE: Influence of metastatic site as an additional predictor for response and outcome in advanced colorectal carcinoma AUTHORS: L Assersohn, A Norman, D Cunningham, T Benepal, P J Ross, J Oates ABSTRACT: Every year, men and women are diagnosed with colorectal carcinoma, and up to % of these will ultimately develop advanced disease. However, there is little information to identify which patients are most likely to benefit from palliative chemotherapy. This analysis is unique in evaluating how the site of metastasis influences response and survival. A database of patients treated within randomized clinical trials using -Fluorouracil (5FU)-based chemotherapy at the Royal Marsden Hospital was analysed. The potential for site of metastasis as a predictive variable for response to chemotherapy and survival was examined, in addition to other clinical parameters. The presence of liver metastases was a better predictor for overall response than either performance status or number of metastatic sites on presentation. Probability of response was significantly decreased by a raised serum carcinoembryonic antigen (CEA) and presence of peritoneal metastases. In liver metastases, a normal serum albumin was as significant a predictor for response as good performance status. The most important predictor for survival was initial performance status. The number of metastatic sites on presentation had no influence on survival. Site of metastasis can predict for response to 5FU-based chemotherapy and patients should be stratified according to the involved site of metastasis in the future. © Cancer Research Campaign BODY: No Body Content
|
PMC3046000.txt
|
TITLE: Elastic Stable Intramedullary Nailing (ESIN), Orthoss® and Gravitational Platelet Separation - System (GPS®): An effective method of treatment for pathologic fractures of bone cysts in children AUTHORS: Marion Rapp, Daniel Svoboda, Lucas M Wessel, Martin M Kaiser ABSTRACT: BackgroundThe different treatment strategies for bone cysts in children are often associated with persistence and high recurrence rates of the lesions. The safety and clinical outcomes of a combined mechanical and biological treatment with elastic intramedullary nailing, artificial bone substitute and autologous platelet rich plasma are evaluated.MethodsFrom / to / we offered all children with bone cysts the treatment combination of elastic intramedullary nailing (ESIN), artificial bone substitute (Orthoss®) and autologous platelet rich plasma, concentrated by the Gravitational Platelet Separation (GPS®) - System. All patients were reviewed radiologically for one year following the removal of the intramedullary nailing, which was possible because of cyst obliteration.ResultsA cohort of children ( girls, boys) was recruited. The mean patient age was . years (range - years). The bone defects (ten humeral, two femoral) included eight juvenile and four aneurysmal bone cysts. Five patients suffered from persistent cysts following earlier unsuccessful treatment of humeral bone cyst after pathologic fracture; the other seven presented with acute pathologic fractures. No peri- or postoperative complications occurred. The radiographic findings showed a total resolution of the cysts in ten cases (Capanna Grade ); in two cases a small residual cyst remained (Capanna Grade ). The intramedullary nails were removed six to twelve months (mean .) after the operation; in one case, a fourteen year old boy (Capanna Grade ), required a further application of GPS® and Orthoss® to reach a total resolution of the cyst. At follow-up (- months, mean . months) all patients showed very good functional results and had returned to sporting activity. No refracture occurred, no further procedure was necessary.ConclusionsThe combination of elastic intramedullary nailing, artificial bone substitute and autologous platelet rich plasma (GPS®) enhances the treatment of bone cysts in children, with no resulting complications. BODY: BackgroundJuvenile bone cysts were first described by Virchow in , but their aetiology still remains unknown []. They can occur in any bone, most often in the long bones and at any age, but mainly in the first two decades []. Despite their benign nature, simple bone cysts interfere with everyday activities. This is because the cysts weaken the cortex, predisposing the bone to pathological fracture. Various treatment options have been reported apart from the principle of 'watch and wait' for spontaneous consolidation. The filling of the cysts with cortisone [-], bone marrow [-] or allogenic bone grafts [-] have been described. Another approach is to stabilize the cyst with elastic stable intramedullary nails, allowing for immediate mobilization. This procedure can also be combined with a bone substitute [,] or the decompression of the cyst with cannulated screws []. However, none of these treatments has been evaluated as being superior to the others in terms of avoiding the persistence of the condition, recurrence or refractures [,].In the case of a fracture, a healing process is initiated with fibrin clot formation, platelet aggregation and degranulation. Platelets contain numerous growth factors such as platelet-derived growth factor, transforming growth factor beta, insulin-like growth factors I and II and epidermal growth factor [-]. Experimental studies have shown that a gravitational platelet-separating system is able to boost the concentration of growth factors. This has the potential to stimulate the prematurely terminated bone-healing processes, especially when combined with autologous bone or bone graft materials [-]. Initial published studies on the use of autologous concentrated platelets in poorly healing dermal wounds [,] and in artificial joint surgery [] have also demonstrated its efficiency.Nowadays in our dynamic and sportive society, children and adolescents want to return to all activities as soon as possible and are afraid of refractures of the cysts arising from minor trauma. Driven by our own mediocre results during the treatment of juvenile bone cysts with prednisolone, cannulated "decompression" screws or ESIN in isolation, we looked for an alternative additional treatment strategy to hasten healing and to minimize the need for repeat operations. This study evaluates the safety and clinical outcome of the treatment with elastic intramedullary nailing (ESIN), artificial bone substitute (Orthoss®) and autologous platelet rich plasma (GPS®) in bone cysts in children.MethodsFrom February to January we offered a combined treatment to all children with bone cysts who had suffered a pathologic fracture or the failure of earlier treatment. The treatment combination consisted of elastic intramedullary nailing (two ascending Titanium Nails, diameter depending on the medullar canal between . an . mm, Fa. Santech Nord, Germany), curettage, artificial bone substitute (Orthoss®, Fa. Geistlich, Germany) and autologous platelet rich plasma (GPS®, Biomet Merck Biomaterials, Berlin, Germany). Orthoss® is an inorganic bone matrix with a macro- and microporous structure derived from bovine material. With the interconnecting pore structure and high inner surface it is an osteoconductive matrix, which is structurally integrated into the surrounding bone and incorporated into the physiological remodeling process. It is indicated for the filling of bone voids following trauma, for reconstruction in orthopedics and in spinal surgery [-]. The material has been in use for more than years and more than million applications are documented []. This new treatment combination was carried out with the informed consent of the parents and the patients themselves.For each subject in the study, the following data are presented: age, gender, location and histology of cyst, earlier treatment history, peri- and postoperative morbidity and further operative procedures. Prior to operation, the distal-proximal, medial-lateral and anterior-posterior extent of every cyst was digitally determined from plain X-ray images.Patients taking medicines known to influence platelet function or patients who had a platelet count < /nl were excluded from the study.The autologous platelet rich plasma was augmented by the commercially available GPS®-System [,]. A blood sample of to ml was taken from each of the patients during anesthesia. The volume taken depended on the size of the cyst and the age of the patient. The preparation of GPS® was performed in the operating theatre during the actual surgical intervention and took minutes. In this procedure the blood was separated into three basic components: red blood cells, platelet poor plasma and - ml platelet-rich plasma (Figure ).Figure 1GPS®-Tube after centrifugation. The GPS® tube after centrifugation: The red blood cells fraction at the bottom of the tube is separated from the buffy coat by a buoy. The platelet-rich plasma is above and is separated from the platelet-poor plasma after the white plunger is manually pushed down.Surgery was always performed under general anesthesia. After reduction of the fracture the elastic intramedullary nailing (ESIN) was performed under fluoroscopic guidance in an ascending matter (-C-configuration). The diameter of the nails (. to . mm) was selected on the basis of the preoperative anterior-posterior radiograph (digitally measured). In the cases of failed earlier treatment, elastic intramedullary nails were removed first. The production of GPS did not require any lengthening of the operation or the time of anesthesia, as it was carried out simultaneously. In all cases the bone cysts were then opened in a minimally invasive manner by an approximately cm long incision. A small specimen for histological investigation was taken and the cyst debrided. Afterwards the to ml GPS® (platelet rich plasma) was mixed with the artificial bone substitute (Orthoss®) and the cyst filled up with this mixture as completely as possible. The treatment protocol involved no further immobilization following intracutaneous wound closure.All patients were reviewed with clinical examination, X-rays and functional evaluation four weeks after the operation, then every three months until complete bone mineralization occurred and removal of the nails was possible. Because long-term results were unknown we arranged one further visit one year after the nails were removed.Treatment results were classified according to the scheme used by Capanna:- Grade = healed - the cyst was completely filled in with bone and the cortical margin thickened- Grade = healed with residual cyst - the cyst was consolidated with bone and the cortical margin thickened but there were still residual cyst parts- Grade = recurrence - the cyst initially consolidated with bone, but large areas of osteolysis and cortical thinning subsequently recurred- Grade = no response - the cyst showed no evidence of response to the treatment [].Grades and were defined as success, whereas grades and represented a failure in treatment. Statistical analysis was descriptive: averages and ranges were determined. Because the expected number of subjects was below twenty, no statistical tests were performed.EthicsThe study confirmed to the Helsinki Declaration and was approved by the local ethics committee of the University of Luebeck [AZ -].ResultsStudy participationsA cohort of twelve children (four girls, eight boys) was recruited. Mean patient age was . years (range - years) at the time of surgery. Histologically the bone defects included eight juvenile and four aneurysmal bone cysts (Table ).Table 1Clinical characteristics of the study populationAgeSexCyst type *Cyst LocationCyst Size (mm)**Acute FracturePrior treatment and ComplicationTime to nail removal (month)Follow-up (month)Outcome Grade (Capanna).4fJBCHumerus proximal47 × × +-.8fABCHumerus proximal39 × × -ESIN + (Cerasorb®) for . years, nail exchange twice, Valgus, Capanna .3mJBCHumerus central75 × × -ESIN alone, . years, Capanna ***.5mABCFemur distal65 × × +-.0mJBCHumerus proximal66 × × -ESIN alone > years, nail exchange three times, Capanna .9mJBCHumerus central57 × × +-.5fJBCHumerus proximal40 × × +-.3fABCHumerus proximal50 × × -ESIN alone, > . years, Capanna .0mABCHumerus proximal62 × × +-.3mJBCFemur proximal36 × × +-.7mJBCHumerus proximal55 × × -ESIN alone > years, nail exchange once, Capanna ..0mJBCHumerus proximal48 × × +-* JBC = Juvenile bone cyst; ABC = Aneurysmal bone cyst** To estimate the size of the defects, every cyst was digitally measured from the plain X-rays in mm prior to operation: distal-proximal, medial-lateral and anterior-posterior.*** One year old boy wished a further GPS® and Orthoss® application at the age of years because of a small residual cyst ( ml) and complicated prior treatment with ESIN alone.Five patients had suffered prior unsuccessful treatment of humeral bone cyst after pathologic fracture with intramedullary nailing and curettage (Figures , , and ) or artificial bone substitution in one case; the other seven presented with acute pathologic fractures (five humeral, two femoral; Figures , , and ). They all received the treatment combination of elastic intramedullary nailing (ESIN), artificial bone substitute (Orthoss®) and autologous platelet rich plasma (GPS®). None satisfied the exclusion criteria.Figure 2Failed Consolidation after treatment with nailing alone. Persistence of cyst after earlier failed treatment (Elastic stable intramedullary Nailing years previously)Figure 3Failed Consolidation after treatment with nailing alone. Elastic stable intramedullary Nailing in combination with Orthoss® and GPS® after earlier failed treatment. Radiograph three months following the initial GPS®/Orthoss® treatment resulting in bone mineralization.Figure 4Complete Consolidation after a long history of failed treatment. Valgus deformity of the humerus after failed earlier treatment of juvenile bone cyst with Elastic stable intramedullary Nails and a different artificial bone substitute. After years of failed treatment the result was classified as Capanna Grade . During removal of the nails and the combined treatment with ESIN-osteosynthesis, Orthoss® and GPS® an additional external fixation was needed due to instability. The Fixateur was removed after weeks, the nails after months.Figure 5Complete Consolidation after a long history of failed treatment. Results months after ESIN and treatment with Orthoss® and GPS®. Clinically the patient achieved an excellent functional result without refracture or deviation of axis.Figure 6Patient presenting with acute pathologic fracture. Consolidation with ESIN, Orthoss® and GPS®. Pathologic fracture of the right humerus after a fall on wet grass.Figure 7Patient presenting with acute pathologic fracture. Consolidation with ESIN, Orthoss® and GPS®. Lateral view X-ray five months after treatment with ESIN, Orthoss® and GPS®. As a consequence, Nails could be removed.Figure 8Patient presenting with acute pathologic fracture. Consolidation with ESIN, Orthoss® and GPS®. Anterior-posterior X-ray five months after treatment with ESIN, Orthoss® and GPS®. As a consequence, Nails could be removed.Figure 9Patient presenting with acute pathologic fracture. Consolidation with ESIN, Orthoss® and GPS®. months after nail removal. Capanna Typ .Two further patients (one boy, one girl) refused to participate. The pathologic fractures of these patients were stabilized with ESIN alone. Another patient excluded from this study presented with a mm × mm × mm sized cyst of the calcaneus. After debridement, the defect was filled with Orthoss® and GPS®. X-rays after and months showed complete mineralization of the former cyst.Peri-and postoperative morbidityNo side effects were obvious before and after the admission of ESIN, Orthoss® and GPS®. No postoperative complications such as infection, deviation of axis, re-operation of intramedullary nailing or refractures occurred.Follow upAt the first outpatient visit, four weeks after the operation, all patients reported complete pain relief from an average time of one week after the surgery. The time taken for the patients to return to full, unrestricted activities was four to six weeks. The radiological findings at four weeks showed the beginning of fracture healing and mineralization of the defect. After three months no deviation of axis was evident and radiological examination revealed that the cysts had begun to heal.The radiological findings at six month showed a total resolution of the cysts in ten cases (Capanna Grade ); in two cases a small residual cyst remained (Capanna Grade ). All fractures healed and complete bone mineralization had occurred. The tiny residual cysts remained on the proximal end of the cyst. The patients with and without a residual cyst had identical treatment strategy. In the two cases with residual cysts, the intraoperative X-rays clearly revealed a technical error: an incomplete filling of these parts of the cysts (Figures , and ). The intramedullary nailing was removed five to twelve months after the operation (mean . months); in one case a fourteen year old boy (Capanna Grade ) wished a further GPS® and Orthoss® application to reach a total resolution. All patients are still being followed up (- months, mean . months), they show good functional results without any movement limitations and no refracture occurred after they returned to sports.Figure 10Incomplete filling of the cyst. Acute pathologic fracture of the right proximal humerus.Figure 11Incomplete filling of the cyst. Intraoperative fluoroscopic control after Orthoss® and GPS® application. Small filling defect in the upper part of the former cyst visible.Figure 12Incomplete filling of the cyst. months after nail removal. Small residual cyst in the area of the former postoperative defect (Capanna Typ ).DiscussionThe question as to whether a conservative or operative treatment is the best treatment strategy in unicameral bone cysts, has been an active one since the Neer's observational series of about cases. Children and adolescents with bone cysts often present first with pathologic fractures. While gross dislocation calls for reduction and stabilization, immobilization can be the treatment of choice for cases with little dislocation or in cysts detected by chance. However, one or more refractures are known to occur in around ten percent of all children []. These complications are the reason why children limit their normal physical activities, which inhibits their social development. In the longer term, all lesions in younger children show high persistence and recurrence rates following conservative treatment. In the proximal humerus and femur - the most frequent localizations - the prognosis following both treatment options was reported to be less satisfactory. As a consequence, NEER supported prompt surgical intervention with curettage and bone grafting as "the best way to rehabilitate these young children" producing successful healing in % to % of cases [].Other treatment strategies such as filling the cysts with cortisone [,,], bone marrow [-,,] or allogenic bone grafts were evaluated [-,]. Long-term studies of percutaneous injection of methylprednisolone acetate have not supported the initial satisfactory results with success rates of only about to % []. Failure rates of about % for cortisone, % for curettage and % of combined procedures with bone marrow have been reported []. In a case series Lokiec reported the consolidation of bone cysts in all ten patients treated with percutaneous autologous marrow grafting allied to multiple perforations of the cysts before injection [], which - in our opinion - will additionally weaken the affected bone. All these described methods may produce consolidation of the cyst but they do not initially enhance the mechanical stability of the weakened bone []. As a consequence patients have to avoid strenuous activities and sports for the duration of the healing process, which might be years [,]. In our experience, this is not seen as a valid therapeutic option by parents and patients.Mechanical stability was achieved by the improvement in pediatric surgery with elastic intramedullary nailing of the pathologic fractures. It is described as a minimal invasive surgery, which supports the healing of the cysts. Stabilization of the bone allows early mobilization of the patients without major complications [,]. The disadvantage is the moderate success rates of about % for complete healing and the prolonged healing period, necessitating the exchange of the elastic stable nails once or twice during therapy []. In two of our patients the removal of the nails was really difficult and one further patient experienced an unsuccessful attempt to remove the nails in another hospital. The difficult changing operations and concomitant prolonged periods of postoperative immobilization render this method unconvincing [,]. In our study earlier treatment with ESIN on its own had failed in four patients, which was also the case in another cyst that was additionally treated with a product containing tricalciumphosphate (Cerasorb®). The duration of their treatment history was four years and two of the children underwent multiple operations.The combined biological and mechanical treatment of simple bone cysts was described for the first time by Kanellopoulos. His treatment employed demineralized bone matrix and autologous bone marrow injection in addition to intramedullary nailing for the stabilization of bone cysts. In seven patients the cysts consolidated completely; two consolidated partially [,]. Although autologous bone marrow collection is considered a relatively simple procedure, it can be associated with numerous complications such as biopsy site bleeding, hematoma or infection [,]. Harvesting enough bone marrow for huge cysts in children is sometimes problematic and causes more pain than filling the cyst and implanting the nails. To avoid these complications, we added GPS® and Orthoss® to our treatment, the rationale being that the application of growth factors from the platelet-rich plasma has been shown to promote tissue repair in many other clinical situations such as cranio-facial surgery [].With the combined mechanical and biological treatment of elastic intramedullary nailing, artificial bone substitute and autologous platelet rich plasma, we describe an option having the possibility of early removal of the intramedullary nailing (six to twelve months, mean . month), thus avoiding changing operations. In our patients it prevented further pathologic fractures and was able to avoid long-lasting limitations in activity. Our approach lead to visible ( - cm) scars on the upper arm or the leg and two further small incisions for the nails, but the children were satisfied with their appearance. No patient showed any changes in their intra- or postoperative vital parameters. The most significant benefits of using the GPS®-System are its autologous nature, the fact that it is endogenously derived and its easy availability. There are no issues of immunogenicity or transmission of infection by using the GPS®-System or the artificial bone substitute Orthoss®. Hass reported the successful use of demineralized human bone matrix (Grafton®) for the treatment of bone cysts in seven children. The disadvantages of this approach are the high costs and the, albeit low, risk of infection [].Another interesting study from published data on patients treated with ChronOS® (Synthes, Switzerland). Treatment with this new synthetic tricalciumphosphate cement resulted in successful healing of different bony lesions in cases; in two others healing with residual cyst. During follow-up from one to twenty months ten defects were observed with partial or subtotal absorption of the injected cement. The volume injected ranged from - ml [].In our series the radiographic evaluation showed complete healing in of patients; only two patients had a small residual cyst in the upper part of the earlier cyst. Apart from one patient, no X-rays were performed later than three years after removal of the implant, because cysts are known not to recur once totally consolidated after treatment []. Rougraff demonstrated, that most cyst recurrence occurred at their proximal or distal ends and suggested that this might be related to incomplete filling of the end of the cysts []. Retrospective analysis of all X-rays from the intraoperative fluoroscopic films confirmed that the filling of the upper part of the cyst was incomplete in our cases with small residual cysts.In all patients the removal of the nails was possible after only six to twelve months (mean . month); such short treatment period has never been previously reported in the literature. Rougraff showed radiographically, that by six to nine months, mature cortical thickening was present in most of his patients, and it was seen in almost every patient by one year. He confirmed that the radiographic features changed very little after one year following treatment []. From our data we can conclude that the combination with Orthoss® and GPS® will fasten the healing of the cysts.Two years after the combined treatment strategy with elastic intramedullary nailing (ESIN), artificial bone substitute (Orthoss®) and autologous platelet rich plasma (GPS®) our preliminary results are promising, but further prospective studies are necessary to validate the efficacy. Although this is a small series that lacks a control group, the results were excellent with no mayor complications such as re-fracture or re-operation (apart from possible early implant removal). The time to healing was short and the children returned early to full daily activities, having suffered restrictions for no more than one month. At the latest examination more than one year after implant removal there was no clinical evidence of cyst recurrence and the functional results were optimal.In our opinion, our promising results as well as the results of the treatment with Grafton® [] and ChronOS® [] justify a prospective multicenter study for further evaluation.ConclusionsThe combination of elastic intramedullary nailing, artificial bone substitute (Orthoss®) and autologous platelet rich plasma (GPS®-System) enhances the treatment of bone cysts in children. It is a safe method without additional perioperative complications and it shortened total treatment time in our series compared to earlier strategies. Secondary procedures such as difficult removal of the elastic stable intramedullary nails followed by a new implantation as well as re-do surgery due to refractures or deviation of axis were avoided. Technically the decisive factor is the debridement of the cyst with the complete filling of the cyst with artificial bone substitute and autologous platelet rich plasma to avoid residual cysts.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsLMW and MMK participated in the planning of the study and did the operations. MR and DS had responsibility for data collection and participated in writing the paper. MMK coordinated the study and had overall responsibility. LMW and MMK revised the manuscript critically for important intellectual content. All authors (MR, DS, LMW and MMK) read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/-///prepub
|
PMC1183205.txt
|
TITLE: Comparing independent microarray studies: the case of human embryonic stem cells AUTHORS: Mayte Suárez-Fariñas, Scott Noggle, Michael Heke, Ali Hemmati-Brivanlou, Marcelo O Magnasco ABSTRACT: BackgroundMicroarray studies of the same phenomenon in different labs often appear at variance because the published lists of regulated transcripts have disproportionately small intersections. We demonstrate that comparing studies by intersecting lists in this manner is methodologically flawed by reanalyzing three studies of the molecular signature of "stemness" in human embryonic stem cells. There are only genes common to all three published lists, suggesting disagreement.ResultsCarefully reanalyzing all three together from the raw data we detect genes upregulated and downregulated in all three studies. The upregulated list was subject to rtRTPCR analysis and % of the genes were confirmed.ConclusionOur findings show that the three studies have a substantial core of common genes, which is missed if only the published lists are examined. Combined analysis of multiple experiments can be a powerful way to distil coherent conclusions. BODY: BackgroundA grave concern in the use of microarray technology is the apparently little agreement between different studies of the same phenomenon carried out in different laboratories, or even in the same lab using different platforms. Particularly evident is the "small intersection" problem: when several competing studies each conclude with a large list of "statistically significant" genes – yet the intersection between the published lists is ridiculously small. This problem undermines confidence in the technology[] or, potentially worse, may misdirect the researchers into suspecting strange biological processes. When the subject matter is itself under contention then controversies may erupt [-]. The repeatability of microarray experiments across labs and platforms has become a hot current topic of discussion in the microarray community, with some studies suggesting that reproducibility across platforms is poor [-] while other studies indicate that biological and laboratory variability are larger sources of discrepancy than platforms or technologies [-], and with other studies pointing out annotations can be a source of problems[].We set out to study whether indeed different studies cannot be reconciled or whether the small intersection problem is artifactual, for one concrete problem of great scientific and extra-scientific importance: human embryonic stem cells (HESC). Isolation of HESC lines has generated the exciting possibility of both access to the basic science of human development as well as the possibility of new hope for cell-based therapy in the clinic[]. The realization of these clinical goals depends on an understanding of the molecular basis of signaling pathways that maintain ESCs in an undifferentiated, pluripotent state, and microarray studies of HESCs should provide a sound basis for discovery and exploration of these pathways.We shall consider microarray studies of HESCs: Battacharya[], Sperger[] and Sato[]. These studies had considerable differences, both in platforms (See Table ) as well as biological. The Bhattacharya and Sperger studies have none or few replicates, hybridizing instead different cell lines to each chip, and contrast to different control samples of pools of differentiated tissues. The Sato study analyzed triplicates of a single cell line compared against "nonlineage induced differentiation" of that same cell line. Maintenance conditions also varied; see Methods for more details.Table 1Summary of platformsBhattacharyaSpergerSatoPlatformcDNA -cDNA – × 4Affy – HGU133aChip design8 × , × × , × × 650ScannerGenePix 4000BGenePix 4000GeneArrayImage Analysis SoftwareGenePix .0Gene Pix .0MAS SuiteSpot/probes (per chip) / 283UniGene1204125 441Empty spots per chip238531-Each study concluded with a list of genes which are up-regulated in stem cells. These lists are important because they should show a good view of all pathways in the cell for which there might be important stem-cell-specific functions, not just developmental signaling pathways, but also potential stem-cell-specific "housekeeping" genes. Most studies have a tendency to focus on developmentally important pathways, such as those that are sufficient for self-renewal; a genome-wide search giving us an unbiased list of up-regulated transcripts is supposed to give us a wider view into the state of stemness.But the three lists of significantly up-regulated genes, as published, are quite different from each other. Their intersection is shown in Figure : seven genes appear in all three studies out of total genes in the union of the three published lists. This is particularly troublesome since all three studies appear to be technically reliable and each study has good reproducibility between replicates (see Methods: "Within platform variations"). The intersection highlights mainly metabolic or housekeeping genes; significantly, few of these have been implicated in any of the nine major developmentally important signaling pathways.Figure 1Common published genes. Intersection (per UniGene) between the published lists of up-regulated genes for each study.The small intersection is a big problem: it is generally taken to mean that the results are not reproducible from laboratory to laboratory and should not be believed. We shall show that this is not the case, that comparing the lists in this manner is methodologically flawed and that the experiments do have a large core of common genes.No experimental scientist would be surprised when results change upon a change of protocol; the analysis of microarray data, from image analysis to the statistical procedure, has as many parameters as a biochemical protocol, and hence there's little surprise that lists obtained by widely different "statistical protocols" are incompatible, so we shall now set about the task of defining and applying a single such statistical protocol to these studies.ResultsIn an experiment addressing differential gene expression, a common criterion to choose a p-value is the largest tolerable amount of false positives; then the power of the test to detect the true positives will depend on the difference between condition and control, the replication variability, the number of replicates and other parameters. In order to be at the intersection between three lists, a gene must have passed such a test three independent times, which it does with a probability equal to the product of the probabilities of each test. Therefore the intersection's p-value is the cube of the lists' p-value. This strongly throws off the balance between sensitivity and specificity, with the result that the intersection becomes insensitive and carries few genes (See section D1 of Additional file – Supplementary Discussion and Methods.). The lists which are good for publishing are inadequate for intersecting; in order to generate the best intersection the lists have to be recomputed.We procured the raw data from all three studies, and re-analyzed it using consistent criteria for spot quality, normalization and summarization, in order to obtain the expression measures for the three studies. Since all three platforms are different (see Methods: "description of studies") they have many genes which are not in common and so could not possibly be at the intersection. We carefully screened UniGene numbers to obtain the set of genes common to all microarray platforms, which is the "universe" of our study. Most of these genes show variations between stem lineage and differentiated control which are no larger than replication errors.Eliminating them increases the power of our analysis and thus we kept for further study only those genes in this set that displayed variations across the samples, since only these could be differentially expressed genes. It should be emphasized that our universe does not include all of the genes possibly involved in stemness. For example some genes, like TDGF1 (Hs385870), Nanog (Hs329296), DNMT3B (Hs252613), FOXD3 (Hs424212), OTX2 (Hs288655), MyosinX (Hs481720), HEY2 (Hs144287), FGF4 (Hs1755), Rex1 (Hs335787) and Nodal (Hs370414), most of which were highlighted as up-regulated in HESCs in various studies, were not included in our "universe" because they are missing from at least one of the three different chip designs.We then analyzed the genes by Integrated Correlation Analysis[] (see Methods), which was introduced to validate the agreement among studies and to select genes that exhibit a coherent behavior across different studies. The idea is that while studying the same system, co-regulated genes should exhibit correlated expression profiles, and these correlations should be maintained across studies. This quality of moving together with other genes we call "coherence". Conversely, when extraneous factors affect a small set of genes in a particular study, the correlations between those genes and the rest shall not be maintained across studies.Figure illustrates how "coherence" of genes between two studies is quantified. For a pair of genes, the correlations between their expression profiles are calculated within each study and the correlation in one study is plotted against the other; points near the identity line represent pairs of genes which maintain their correlation values across studies. The coherence score of a gene g is the correlation coefficient of the points corresponding to all pairs containing g. Using all gene pairs we obtain the integrated correlation score, an overall measurement of agreement between the two studies. We have changed the nomenclature from [], where coherence is called "consistency", because we want to reserve the notion of consistency to mean moving in the same direction from condition to control in all studies (i.e., up-regulated in all studies). Notice that no information about the condition and control is introduced in the definition of coherence; if the genes are both up-regulated in study and both down-regulated in study they are coherent because they behave the same way with respect to each other within each study, even though they are both inconsistent. (See Methods: Coherence and Consistency)Figure 2Coherence Scores. Given the expression profile for four genes A, B, C, and D in two studies and shown in panel (a), we can calculate the correlation of the expression profiles for every pair of genes in each of the two studies. In panel (b) those pairwise correlations are plotted for study against study . For example, pair AB is positively correlated in both studies and pair BD is negatively correlated in study and positively in study . Correlations for pairs AB, AC, AD lie approximately on a line of positive slope, so gene A is called "coherent". The coherence score of A is the correlation coefficient of those points, which is positive. Gene D has a negative correlation coefficient and so is incoherent. Study in panel (c) is study where condition and control have been swapped; all genes are perfectly coherent for studies and , yet each gene which is up-regulated in study is down-regulated in study .Previous use of this analysis applied to classification of cancer [] yielded unimodal histograms of the gene-coherence scores (there called "gene-reproducibility score"); in our study, though, we obtain a strongly bimodal histogram, shown in Figure , indicating that there is a number of genes in strong agreement among the three studies and a number of genes in strong disagreement. (Further details, including bivariate density plots, can be found in Additional file , section D4). There are a number of potential reasons why genes could be strongly incoherent as shown in these histograms discussed later.Figure 3Coherence Scores Distribution. The histograms of the coherence scores are bimodal. (a) Pairwise comparisons, and averaged score over all three comparisons. This implies that in these studies the genes can be divided into two distinct categories, coherent and incoherent. When "erratic" genes are discarded, there is a marked improvement in the agreement between studies. b) Integrated Correlation and Correlation of M-values calculated using the genes in the top percentiles of Coherence Score, red: Bhattacharya-Sperger, blue: Bhattacharya-Sato, black: Sperger-SatoEliminating incoherent genes improves enormously the general agreement between the studies. The improvement in the integrated correlation score and the correlation between log2-fold changes (M-values) is illustrated in Figure 3b. We decided to keep for further analysis the genes in the top % of the gene-coherence score distribution. (Little variation of our final results was observed for percentiles % and % since they only include positive coherence scores; we present the stricter criterion). Not so obviously, the correlation between the M-values between studies also markedly improved from . to ., ., and . respectively. (More detailed results about Integrated Correlations results can be found in Additional file , section D4)Once we restrict ourselves to this set of coherent genes, we study those genes that are up- or down- regulated in stem cells vs. their differentiated controls in each one of the studies. We emphasize that exactly the same statistical tests and criteria were applied to all three studies, with a strict cutoff value selection based both on a p-value and a positive lods []. We used the moderated t-statistics as proposed by [] and the FDR (BH) procedure was used to adjust the p-values for multiple hypothesis []. This is important because the set of coherent genes is enriched in genes that are not statistically independent. The BH procedure [] controls FDR under certain general assumptions (positive regression dependence) and simulations shows its adequacy to control FDR in more general dependence structures [], while the BY procedure [] is assumption-free but is more conservative that BH. However, in our case both procedures led to practically the same results, because the proportion of differentially expressed genes is higher in the coherent set; we report the results from the BH. P-value cutoff was set in ., which implies than the probability of error is - in the pairwise comparison and - when the three studies are considered.The intersection between the lists is now quite larger and statistically significant, as shown in Figure ; the up-regulated and down-regulated genes common to all three studies (vs. expected by chance intersections) are listed in Additional file and Additional file online. Notice that the up-regulated genes in this list are not necessarily the "most" up-regulated for any individual study; yet they are significantly up-regulated for each study.Figure 4Intersection of significant genes. Improved intersection of significant genes. After our screening process, the number of transcripts which are up (a) and down (b) regulated in stem cells for the three studies. Notice that now most genes in each study also appear at least in one other study (% Bhattacharya, % Sperger, % Sato, about % expected by chance), with a very important fraction common to all three (about genes expected by chance at the intersection).We performed real-time RT-PCR analysis of the up-regulated genes to validate our findings. In an initial experiment, H1 HESCs were differentiated for days with RNA samples taken before and after differentiation. We succeeded in analyzing genes out of the intersecting HESC enriched genes in triplicate for expression level changes upon differentiation by real-time RT-PCR (See Additional file for primers used in real-time RT-PCR). After screening for significant differences (within this experiment) between the mean normalized Ct values (Student's t-test, p > ., n = ), between differentiated and undifferentiated samples, genes (%) had differences in expression. Of these, (%) were enriched in undifferentiated HESCs and (%) were enriched in differentiated HESCs. Overall, of the tested genes (%) were enriched in undifferentiated HESCs (p = . × - in the exact binomial test); of them also significant by RT-PCR, for a % of success probability (p = . × -). % were enriched in the differentiated sample and % were either not regulated or not significant in this experiment. (See Additional file for the raw data of real-time RT-PCR results)A comparison between the fold changes found through real-time RT-PCR and those obtained in the three studies is shown in Figure . Our real-time RT-PCR test was carried out under identical cell line, culture and differentiation conditions to the Sato study, so we present comparisons against the mean log-fold-change of the three studies and against Sato's study separately, which has for evident reasons better fit. Positive slopes and correlation coefficients are obtained in all comparisons; correlations are higher when fitting only to genes with real-time RT-PCR log2 fold-changes bigger than ., becoming . when comparing only to Sato's study. These slopes and correlation coefficients are in line with general results in the literature comparing different microarray platforms and real-time RT-PCR results for the same RNA sample.Figure 5RT-PCR results. Comparison between the real-time RT-PCR results and the microarray results. The blue lines are linear fits (without intercept) through all the genes, while the magenta lines fit only the genes with a fold change bigger than . (log2-scale) in RT-PCR analysis (magenta points). (a) log2-fold change of RT-PCR vs. mean of all microarray studies (R2 = ., r = ., p < - for all genes, R2 = ., r = ., for the top genes); (b) log2-fold change RT-PCR vs. Sato study alone, which had identical conditions to our study (R2 = ., r = ., p < - for all genes, R2 = ., r = . for the top genes)DiscussionHow can we reconcile the agreement in Figure with the disagreement in Figure ? Figure displays an incorrect comparison, since intersecting lists in this manner is flawed methodology. The three lists were created from different sets of genes using different statistical criteria for ranking the transcripts; they really are apples and oranges and should not be compared so lightly. To mention but three exemplars of problems with such comparison:- there are more than genes in Sperger's study, yet only are common to all three arrays, so about / of the genes studied by Sperger were not studied in some other study; in fact, more than genes in the published list could not possibly have been at the intersection.- the size of the intersection cannot be larger than the smallest list and is thus controlled by it; in this case the Bhattacharya study had far more stringent criteria and reported the fewest genes; notice that about /3rds of that list were in at least one of the other studies.- different versions of annotation databases were used to create the lists, so some genes have identifiers that changed over time and are missed in an automated comparison; in fact, when a current version of the Affymetrix probeset annotations is used for the Sato study, the intersection increases more than twofold to due to several genes having changed identifiersThese are just exemplars of the general problems of differences in gene universe (a), in gene identity (c), and in statistical criteria (b), which include not only the significance threshold used, but which particular test was employed to assess it (e.g. "moderated t-statistics with fdr corrections for multiple hypothesis"). Other general problems affecting list intersection also include the preprocessing steps, which have been shown to have a substantial impact in the agreement between platforms[,]. So, how should different experiments be compared? Treating them on an equal footing goes most of the way: this means procuring from the authors the raw data, analyzing all studies with exactly the same statistical methods and cutoffs, and working only on the universe of common genes. This still does not make the studies all apples, but it significantly approaches this goal. We further refined this by using the notion of gene coherence; in doing so we discovered that a number of genes are strongly incoherent, behaving erratically across the studies.The incoherent genes may behave so due to environmental interactions, differences in cell culture system or in the differentiated states being compared. Gene expression is the nervous system of cells, imprinted by anything in the environment. A large number of genes are involved in environmental interaction, leading to variability between labs which obscure the common signal we are trying to detect. Among the gene categories that classify the top incoherent genes (see Additional file ), i.e. those that exhibited a significant fold change in respective studies but that were regulated in opposite directions among the studies, the top classifications were mainly involved with metabolic activity, such as genes involved in the response to oxidative stress and reactive oxygen specific genes. The incoherent genes may also reflect differences in the conditions that the labs use to grow cells. The HESC culture system has many poorly understood variables, such as batch to batch variation in the MEFs used to support the HESCs and the use of serum in some culture systems. These are both examples for which there are, as yet, no good predictors for the ability to support "stemness" and can currently only be screened by their ability to support HESC maintenance based on few markers. However the list also includes developmentally important genes, such as MID1(Hs27695), a possible left-right determination factor [], and FGF9(Hs111), implicated in several developmental decisions. Also included were several members of the Wnt pathway, including APC(Hs158932), DIXDC1(Hs116796), and TLE4(Hs444213), so the incoherent genes may also reflect differences in the control samples used in the studies. The Sato group uses HESCs differentiated by withdrawal of CM, which may result in biases in the differentiated cell types produced, while the other groups use pooled panels of RNA from a variety of somatic tissues.The application of microarray technology to the study of HESC biology has the potential to provide a robust foundation for the exploration of molecular networks underlying the state of "stemness" in human embryonic stem cells. In addition to the insight into their utility in clinical applications provided by this exploration, elucidation of these networks in HESCs shall be an important step towards defining the molecular activity driving development in early human embryos.Our list of confirmed transcripts intersects three experiments carried out on different cell lines, maintained in the stemness state by different protocols, and compared against different differentiated states. That we do get a confirmed list under these circumstances bears witness that there is a well-defined set of molecular circuits involved in the state of stemness that can be studied regardless of variations in protocol or cell line. This analysis more than octuples the list of confirmed molecular markers of the stemness state in HESCs [].A balanced global assay of the molecular state of "stemness" should reflect all activities of the cell including those pathways that are necessary as well as those that are sufficient for the functions of an embryonic stem cell. Therefore, we should expect to find activities that might include, for example, those required for the defense of chromosomal or genetic stability as well as those required for self-renewal in an undifferentiated state. Indeed, the two most represented GO Biological Process in a GO/EASE analysis of the enriched genes within our universe are "Traversing start control point of mitotic cell cycle", indicating a prominent role of cell cycle regulators, and "transmembrane receptor protein serine/threonine kinase signaling", indicating a significant role for genes involved in sensing and responding to the developmental environment. Further, "transforming growth factor-beta receptor activity" was the most represented GO Molecular Function classification, indicating a prominent role for TGF-beta signaling in undifferentiated HESCs (See Additional file and Additional file for the complete GO analysis). However, because many of the genes that are important for these and other activities may be excluded from the universe analyzed here, we do not present the results as a complete or unbiased picture of these activities. Instead, this study addresses the validity of using microarray technology for building this foundation by applying consistent analysis to evaluate reproducibility.ConclusionOur study supports concluding that the three studies are compatible and repeatable. The method we've used demonstrates that we can harness the power of several labs to give weight to the intersection. In fact, we may conclude that a list of regulated transcripts from a single lab obtained under a restricted number of maintenance and differentiation protocols should be considered with some reserve, for we only know about coherent behavior of genes when we have several studies to draw from. Our results indicate that publication of the raw data is far more valuable than publication of the analyzed data, and further suggest that the field shall move forth only upon agreement to set standards of backgrounds and statistical methods.Methods(see Additional file , section Methods for a full description)Description of the studiesThe Bhattacharya study has chips. Different HESC lines were hybridized to the red channel (Cy5) of the arrays; of them were lines BG01, BG02, GE01, GE09, TE06, and the sixth sample was a pool of GE01, GE07, and GE09. The control sample, hybridized to the green channel was "total human universal RNA (huURNA) isolated from a collection of adult human tissues to represent a broad range of expressed genes from both male and female donors (BD Biosciences, Palo Alto, CA)". No replicates were performed for individual lines. The Sperger study used a similar design, hybridizing lines H1, H7, H13, H14 and two samples of H9. The control samples were "a common reference pool of mRNA". The Sato study had Affymetryix HGU133A chips, replicates of H1 cells (in Matrigel/Conditioned Medium) and replicates of "nonlineage-directed differentiation" (Matrigel/non-CM). Table summarizes the architecture of chips and the number of spot/genes involved in the three studies.Language and packagesThe statistical analysis was carried out in the R language version . , and packages were from the Bioconductor project . Gene Ontology analysis was carried out using EASE . software [].Raw dataData from the Bhattacharya and Sato studies were obtained directly from the authors. The Sperger data were obtained through the Stanford Microarray Database []. cDNA array data were output files from GenePix .. Affymetrix raw data files were .CEL files.We used the same image analysis criteria in all CDNA arrays to exclude low quality spots (the criteria were different in the published studies).Expression measuresThe marray package from the Bioconductor suite was used for cDNA arrays. Normalization was executed in two steps, first within-print-tip-group location-dependent intensity normalization followed by within-print-tip group scale normalization using median absolute deviation. Single-channel normalization of two-color cDNA was done as proposed by [], using quantile normalization. The GCRMA algorithm was used to summarize Affymetrix data as proposed in []. This algorithm improves the widely used RMA [] by including an extra step to adjust for non-specific binding, and computing the sequence-specific affinities between probes as described [].Within-platform variabilityIn order to assess the quality of the data replications the within-platform variations were analyzed. The within-study reproducibility is overall fairly good in all the studies, even noting that Bhattacharya's and Sperger's design contain different lines of HESC rather than true replicates of a single line. See further details in Additional file .AnnotationsFor both Bhattacharya and Sperger studies, annotations were obtained from SOURCE from the Stanford microarray data homepage[]. For Affymetrix data, annotations packages from Bioconductor were used. The IMAGE clone IDs and the Affymetrix probes were matched using UniGene Cluster Annotation. Genes with no UniGene number were eliminated from the study. Expression values from spots or probesets with duplicated UniGene identifiers were averaged together.Common genesAnnotations were obtained with the raw data from each study. Genes without UniGene identifier were eliminated and duplicated probes/spots were averaged together. After this process there are genes common to all studies as shown in Supplementary Discussion Figure 8a. We filtered for evidence of variation across samples, reducing our set of interesting genes to those showed in Supplementary Discussion Figure 8b. For the cDNA arrays, we select genes where the M-values was bigger than . in at least arrays and in Affymetrix experiment we keep genes whose expression profile had range bigger than ..Coherent genes: the integrated correlation approachIntegrated correlation analysis was introduced in []. For each study s, let us define xg the expression profile for a gene g, and the correlation for the pair of genes p = (g1, g2) in the study. Based on we can asset both overall coherence between studies and gene-specific coherence. The integrated correlation, defined as: quantified the coherence between studies. If this expression is calculated considering only the pairs containing a gene g, then we have a measure of the gene-specific coherence between two studies: , where p = (g, j) When more that two studies are involved, the average over all s and s' is used as a Coherence Score for a gene g, . Confidence Interval for the correlation scores were obtained by bootstrapping.Coherence and consistencyIn [], the coherence score defined above is called reproducibility score and coherent genes (genes with high values of the score) are called "consistent". However we once again stress that this score bears no direct logical relationship to the notion of reproducibility or consistency in the sense of consistent up- or down- regulation in both studies. A simple counterexample makes the point: create a fake Study which is Study with the values of the condition and control swapped; then by the above definitions, all genes are perfectly coherent for studies and , having coherence (reproducibility) scores equal to ; yet each gene which is up-regulated in study is down-regulated by the same amount in study , so all genes are inconsistent. A relationship between the coherence scores and consistent behaviour is predicated on the counter-reciprocal: if a pair of genes is incoherent then both genes cannot be consistent, and hence if a gene has a negative coherence (reproducibility) score it is "likely" (though by no means sure) to be inconsistent by being the "odd one out".Differentially expression criteriaStatistical analysis to determine which genes are differentially expressed was carried out using the package Limma from the Bioconductor project. For assessing differential expression the moderated t-statistics was used as proposed by [] in all the studies. To do so, Limma uses an empirical Bayes method to moderate the standard errors of the estimated log2-fold changes. This results in more stable inference and improved power, especially for experiments with small number of arrays. The p-values of the moderated t-test were adjusted for multiple hypothesis testing, controlling the false discovery rate (fdr) as proposed []. We use a strict cut off criterion for selectivity of the genes based on both the p-values and lods ratio, as proposed in []. The lods (or B-statistic) is the log of the odds that the gene is differentially expressed. For the set of coherent genes we selected here, the cutoff of p = . and positive lods leads to ( up, down) genes while ( up down) is obtained considered only the p-value cut-off. In Supplementary Discussion and Methods the reader can find how the number of selected genes depends on the criteria for different set of coherent genes.Real-time RT-PCR verification of gene expression level in HESCsRelative expression levels of the (%) of genes in the intersection were analyzed in H1 HESCs maintained in CM or differentiated by withdrawal of CM for days by real-time RT-PCR. Raw data were normalized to Ubiquitin-C expression and relative expression levels were determined using PCR efficiency-adjusted ratios [].Authors' contributionsMSF and MOM are the analytical members of the team while SN, MH, AHB are the biological team. Conception of the work came from join discussion between MSF and AHB. MSF and MOM are responsible for analysis and methodology. MH designed the primers for the RTPCR experiment, carried out by SN and AHB; the latter are also responsible for the biological assessment of all analytical results.Supplementary MaterialAdditional File 1Supplementary Discussion and Methods.Click here for fileAdditional File 2Up-regulated genes in the intersection. List of up-regulated genes in the intersection of the studies. In html format, including annotations and links.Click here for fileAdditional File 3Down-regulated genes in the intersection. List of down-regulated genes in the intersection of the studies. In html format, including annotations and links.Click here for fileAdditional File 4RT-PCR primer sequences used. Primer sequences used in real-time RT-PCR.Click here for fileAdditional File 5real RT-PCR results.Click here for fileAdditional File 6Top incoherent genes. List of the top genes with the most negative coherence score.Click here for fileAdditional File 7Up-regulated genes – EASE analysis. EASE analysis for the up-regulated genes in the intersection of the three studiesClick here for fileAdditional File 8Down-regulated genes – EASE analysis. EASE analysis for the up-regulated genes in the intersection of the three studiesClick here for file
|
PMC2629006.txt
|
TITLE: Towards safer, better healthcare: harnessing the natural properties of complex sociotechnical systems AUTHORS: J Braithwaite, W B Runciman, A F Merry ABSTRACT: Objectives:To sustain an argument that harnessing the natural properties of sociotechnical systems is necessary to promote safer, better healthcare.Methods:Triangulated analyses of discrete literature sources, particularly drawing on those from mathematics, sociology, marketing science and psychology.Results:Progress involves the use of natural networks and exploiting features such as their scale-free and small world nature, as well as characteristics of group dynamics like natural appeal (stickiness) and propagation (tipping points). The agenda for change should be set by prioritising problems in natural categories, addressed by groups who self select on the basis of their natural interest in the areas in question, and who set clinical standards and develop tools, the use of which should be monitored by peers. This approach will facilitate the evidence-based practice that most agree is now overdue, but which has not yet been realised by the application of conventional methods.Conclusion:A key to health system transformation may lie under-recognised under our noses, and involves exploiting the naturally-occurring characteristics of complex systems. Current strategies to address healthcare problems are insufficient. Clinicians work best when their expertise is mobilised, and they flourish in groupings of their own interests and preference. Being invited, empowered and nurtured rather than directed, micro-managed and controlled through a hierarchy is preferable. BODY: An important question facing contemporary health systems is how to reduce variability and iatrogenic harm. Many initiatives have been proposed in the belief that this is primarily a structural issue. They have taken the form of reorganisation4 and policy reforms, credentialling and accreditation, and directives to standardise care processes. However, at the heart of healthcare lie interactions between individual patients and clinicians. Improving communication and relationships, enhancing individual decision-making through evidence-based decision support10 and promoting patients’ involvement in and responsibility for their own care are also vital for safer, better care.We propose that stimulating the fundamental transformation needed to improve the safety and quality of healthcare will require harnessing the natural properties which emerge (often spontaneously) at the interface between the socio (human behavioural) and technical components of complex systems. A bottom-up strategy led by clinicians is badly needed to balance the predominantly top-down approaches which frequently result in only modest improvements which are difficult to sustain. Patient safety is what social scientists call a “wicked problem”—one that is messy, persistent and multidimensional. Politicians and bureaucrats seek to shape clinical practice by edict, whereas in reality it is shaped by the behaviours and attitudes of practising clinicians.NATURAL PROPERTIES OF COMPLEX SYSTEMSMany complex systems have similar natural properties and behaviours (table ); these have been identified in fields as diverse as mathematics, sociology, marketing science17 and psychology, and attempts made to apply them to healthcare. Safety is a property which depends on good organisation, tools and infrastructure as well as on the behaviours of individuals (often in teams). The collective values and behaviours of these individuals comprise the culture of the system. Harnessing their undoubted industry, goodwill and energy, and supporting the natural processes by which they interact and cooperate, rather than constantly reorganising them, is the key to changing this culture. The emphasis should be on guiding the natural properties and behaviours of sociotechnical systems (see table ) rather than imposing hierarchical structures and above-down instructions from people who do not actually work at the “coal-face.”Table 1Natural properties and features of complex systemsProperties of complex systemsHealthcare manifestationsNatural networksGroups of clinicians who interact professionally to share information, support, consult, refer and jointly manage patientsNatural hubs and scale-free behaviourOpinion leaders in networks who disproportionately influence policies, events or practicesNatural pathways, connectivity and small worldsCommunication channels facilitating the rapid dissemination of information via “grapevines” and communities of practiceNatural appeal and stickinessMessages and communications that are convincing and are absorbed among clinical cohortsNatural propagation and tipping pointsThe point at which a message, idea or practice whose time has come is readily adopted by a critical mass of cliniciansNatural categories and natural mappingThe identification of clinically relevant problems grouped as accessible data, to facilitate decision-making and solutions to healthcare problemsNatural interest and self-selectionClinicians with common concerns and complementary expertise voluntarily grouped together to collectively resolve coal-face clinical problemsARTICULATING THESE NATURAL PROPERTIESNatural networksThere are two types of network: those that are purpose-designed, funded and imposed by someone in authority (mandated networks), and those that are formed by relationships among clinicians which rest on mutual (often implicit) agreements to participate (natural networks). The former have assigned functions, and are necessary and appropriate for the “hotel,” logistic and infrastructure requirements of healthcare. The latter underpin the health system’s purpose: to enhance health and deliver care.Natural networks respond poorly or not at all to conventional management or control measures. They emerge spontaneously and function with little or no externally imposed structure, but can exert powerful and pervasive influences on how systems actually perform and function, at “street level,” behind the formal organisational charts. They are webs of humans connected personally or via technologies, interacting in multiple ways. With the internet, natural networks of patients and their supporters are also increasing.24Natural networks can be relatively simple or highly complicated,– and individuals may tap into as many networks as there are problems to be solved or ends to be achieved. They occur wherever humans cluster in the pursuit of a common purpose or activity, such as adolescent friendships in schools (see fig ), the sharing of wisdom across boundaries29 or the improvement of patient care.31Figure 1Friendship clusters in a US school. Reprinted with permission from Moody J. Peer influence groups: identifying dense clusters in large networks. Soc Netw ;:–.Medical examples range from geographically dispersed clinicians weakly related by discipline or common interests (for example, in stroke, trauma or ophthalmology), collaborating to improve care in Australia, to teams in America reducing catheter-related infections in ICUs, to action-orientated researchers promoting high levels of cooperation among paediatric clinicians in Finland.33Natural hubs and scale-free behaviourNodes in a network might be thought to be randomly related, each with a similar number of connections. However, most networks are actually “scale-free,” with unequally distributed nodes with varying numbers of connections.– A few nodes with many connections form “hubs” which emerge naturally and become the distributed force field of the network (for example, the Google search engine, King’s Cross station or a prominent, influential clinician). Scale-free networks are less likely to become congested because they concentrate effort efficiently. The internet has such a structure, with a limited number of hubs, an inner layer of strongly peer-connected components and an outer layer with many relatively isolated nodes (fig ). Analogous patterns emerge in healthcare, where hubs (prestigious services or clinicians) can have pervasive influences on practices and attitudes among their well-linked inner layer colleagues, and progressively less on the outer layer of poorly connected isolates.Figure 2Structural typology of the internet. Reprinted with permission from Carmi S, Havlin S, Kirkpatrick S, et al. A model of Internet typology using k-shell decomposition. Proc Natl Acad Sci U S A ;:–.Natural pathways, connectivity and small worldsScholars have sought recently to understand the sociology of different types of networks through an analysis of another natural characteristic, that of “small worlds.” Regardless of the size of any particular network, there are fast routes through them, creating small worlds within large networks. Even among the six billion people on Earth, for example, we can usually map the ties of any two through substantially less than the famous six degrees of separation. Making a small world work using natural pathways does not require personal or corporate knowledge of everyone in a chain or explicit representation of connections, pathways or hierarchies.Natural appeal and stickinessAdministrators are always trying to spread messages or influence people through directives, emails and meetings. However, most communications are ignored, and meetings avoided or attended only by a few. This brings us to another concept—not a natural feature of networks, but a phenomenon related to group behaviour. It comes from marketing, and is known as stickiness. Any message to be both remembered and acted upon needs to be sticky. Stickiness is a function of the intrinsic nature of a message, how it is presented and the effect it has on the recipient. Sticky messages have natural appeal. Awareness of the importance of stickiness challenges communicators to communicate well. Novel or effective communication, smooth transmission modes, embedded cues in the environment and workplace to remind people of the message, forcing functions to facilitate compliance with the message and a critical mass of champions or opinion leaders can all be important in getting a message to stick.Natural propagation and tipping pointsThe stage at which a critical mass for sustained or escalating momentum in any system is reached is what Gladwell calls a “tipping point,” an idea used in classic epidemiology. It is the pivotal juncture at which a concept, social movement or epidemic takes hold. One or more triggers may be needed for a state change. Although there are many potential candidates, several stand out: persuasive, catalytic individuals, memorable, even irresistible messages47 and conducive contexts. A tipping point is not easy to evoke. An example in medicine that has taxed the best minds over the last two decades is “evidence-based medicine,” an undeniably seductive concept, but one which is yet to demonstrate self-propagation and has proved resistant to conventional measures for inducing change. The missing ingredients for the requisite degree of stickiness may be the redesign of the interface between information technology and clinicians at the point of care, work-flow forcing functions making it easier to do the right thing and harder to do the wrong thing, and peer-group self-regulation involving meaningful surveillance between colleagues.50Examples of the power of networks continue to emerge. In the longstanding Framingham Heart Study, clusters of obese individuals were identified in networks of social ties (fig ). The risk of obesity for those with obese friends was increased more than %. The influential factor was social, not geographic, distance. The lesson is a profound reminder that desirable or undesirable behaviours can propagate via networks.Figure 3Framingham Heart Study and obesity networks. Reprinted with permission from Christakis N, Fowler J. The spread of obesity in a large social network over years N Engl J Med ;:–.Natural categories and natural mappingPatients may be placed at avoidable risk by being subjected to a flawed plan, by the flawed execution of a plan or both. To devise and disseminate effective preventive and corrective strategies, it is necessary to identify what is going wrong and then how and why. The major studies on appropriateness of care and on adverse events have not adequately provided this information in a manner that is meaningful for clinicians. Qualitative analysis of the problems identified in these studies is needed. They should be placed into clinically meaningful categories (principal natural categories or groups of problems amenable to a common solution), based on their contextual features and contributing factors. The characteristics of natural categories, with examples, have been described elsewhere. Natural categories will form the basis of the International Classification for Patient Safety, which is currently being developed.56Natural mapping, a concept described by Norman, allows the components and attributes of an entity, schema or problem to be grouped in a representation reflecting their relationships in the real world. In healthcare, this facilitates a hierarchical classification of incidents which allows detailed information to be elicited from reports or documents in a comprehensive, structured but intuitive way, and stored in a database from which meaningful information can readily be extracted and analysed for the generation of solutions to problems with safety and patient care. The process includes seeking consensus from groups of clinicians on what the relevant constructs and concepts are, what terms or descriptions should be used and how they should be represented. Testing these constructs and concepts against real-world information (such as incident reports, complaints or coroners’ recommendations) permits iterative improvement, progressively enhancing their validity. The reliability of these processes is a function of the reproducibility with which incidents can be deconstructed into their components. This in turn depends on how intuitive (or natural) the categories and cascades of questions (maps) are, and how “user-friendly” the overall system is.TOWARDS A BOTTOM-UP SOLUTION TO PATIENT SAFETYPutting natural interest and self selection to workFor every healthcare problem there are networks, hubs and subclusters made up of clinicians with a special interest and expertise in that area. Such people are typically willing to devote time and effort to solving the problems in which they have a natural interest, as part of their professional lives, whether or not adequate funding is available. “Hub” clinicians should be asked to collaborate with others who have a genuine aptitude and passion for addressing a particular problem, to review that problem, devise and implement corrective measures, and ensure sustained surveillance and frequent updating of the solutions (see box ). 60Box A tale of a natural network and propagation of a sticky idea57 59By the late 1980s, important technological advances provided the potential for safer anaesthesia with devices such as pulse oximeters and capnographs. When anaesthetists requested access to these, they were ignored by administrators. Frustrated, a meeting was called in Australia of influential clinicians. The idea had such natural appeal that of people invited attended the meeting in May , the majority of whom represented hubs of natural networks interested in subjects allied to safety and monitoring. All had to pay their way to the meeting which was called at short notice and not under the auspices of any professional college, society or association. Everyone who was asked to present a paper did so and provided copy for the manuscripts which were published months later. These recommended minimum standards which were endorsed by the relevant professional bodies at national, and, later, international levels. Despite opposition from administrators, oximeters were introduced for every anaesthetised patient and capnographs for every intubated and/or ventilated patient. The idea “tipped,” and it became unacceptable to conduct anaesthesia in developed countries without the use of these devices.An analysis of incidents and medicolegal reports over the period – revealed not a single case of hypoxic brain damage or death due to inadequate ventilation or undetected oesophageal intubation, problems which had plagued anaesthesia until the late 1980s. This is an example of quadruple-loop learning (at personal, organisational, national and international levels), and a bottom-up initiative which gained traction by harnessing some of the natural properties of a sociotechnical system.Exploring the propositionWe suggest, in parallel with above-down initiatives to improve the safety and quality of healthcare, that clinical standards be set by expert groups for each of the individual problems which compromise the safety and quality of healthcare by harnessing the natural properties of networks and clinicians’ behaviour. A good start has been made with the development of evidence-based clinical guidelines by organisations such as the National Institute for Health and Clinical Excellence in the UK (NICE) and the Joanna Briggs Institute (another self-propagating network; see box ), but ongoing effort is essential, as new problems will continue to emerge at the leading edge of medical advances. The plethora of new errors associated with the introduction of computer order entry provide an example. Problems to address initially could be those clearly shown to be widespread, costly and amenable to mitigation by proven interventions of reasonable cost– and risk–benefit. 65Box Joanna Briggs Institute for Evidence-Based Healthcare—another self-propagating networkThis organisation was formed in with a single corporate member. It was not under the auspices of any professional body but has alliances with several as well as with hospitals and universities. Within a decade it had centres in countries, with members in countries, and corporate members. The core business is the generation of annually updated evidence-based guidelines based on systematic literature reviews. There are now systematic reviews and packages of guidelines which can be compiled, using software available for members, into customised clinical practice manuals. Software is also available for clinical audit, benchmarking and tracking evidence of practice change. There are now plans to harness the concepts and networks to complement international and national collaborations for translating evidence into practice.Once a problem that needs to be addressed has been identified, a meeting should be arranged to develop a tool to address it. Modest funding should be provided for two or three acknowledged experts to attend and present reviews of the relevant information. A public invitation should be issued to all with an interest in the area. In this way, the natural hubs of existing networks relevant to the problem can be gathered together to strengthen existing links between networks and establish new ones, thereby facilitating the ongoing generation of effective solutions to the particular problem.Ideal tools should: explain the underlying principles; outline the necessary tasks; record that they have been done; facilitate audit; and promote dissemination, stickiness and use of the principles in question. They should be endorsed by the relevant professional organisations as clinical standards, with a requirement that they should be used whenever appropriate. Their design should make compliance easy, and clinicians should be involved in their design to facilitate this. Documented justification of non-compliance should be required. The underlying rationale for the tool, and the process by which it was developed, should be published (ideally in the peer-review literature) and made available on the internet; a plain language version should be developed for patients and lay-carers. The tool should be refined over time on the basis of feedback from those who use it.Natural surveillance and collegial behaviourA transparent process is needed for identifying persistent non-compliance with these tools, with a view to progressively improving standards. It has been shown that formal surveillance may be acceptable to practising clinicians. We recommend a peer-review audit. This is practical because much routine clinical practice involves the following of pathways and algorithms by individual clinicians through habit, so it is possible to identify the routine practices of most clinicians by reviewing as few as or medical records. We propose that peers review each other’s compliance with agreed tools on a regular basis, as part of normal professional practice, and accreditation agencies can then randomly check for evidence of tool use. This would provide a transparent process for identifying persistent non-compliance with standards and an objective basis for opening discussions to understand and address the issues.CONCLUSIONThere is now compelling evidence and widespread acceptance that there are problems in healthcare that must be addressed and that current approaches to addressing them are inadequate. Education, persuasion and attempts to change practice through existing hierarchical structures have largely failed in the face of the entrenched opposing forces of clinical autonomy and the traditional tolerance of individual and regional variations in practice. Evidence-based practice is overdue. It is time to do something different.We recommend harnessing the natural properties of the sociotechnical system that comprises healthcare, and guiding the existing practices which come naturally to clinicians, so that these can be developed and used to promote continuous quality improvement with effective self-regulation. Clinicians, like other professionals, work best when they are allowed to flourish in groupings of their own interests and preference, are empowered rather than directed, and nurtured and influenced by their peers rather than controlled by others. They are likely to become more involved in promoting safer and better care if invited rather than compelled, and should be encouraged to solve naturally occurring problems in voluntary collaborations with their fellow clinicians. To underpin this, we propose facilitating the development and application of clinical standards by means of relevant and effective tools.
|
PMC2362706.txt
|
TITLE: Fatigue in advanced cancer: a prospective controlled cross-sectional study AUTHORS: P Stone, J Hardy, K Broadley, A Kurowska, R A'Hern ABSTRACT: Uncontrolled studies have reported that fatigue is a common symptom among patients with advanced cancer. It is also a frequent complaint among the general population. Simply asking cancer patients whether or not they feel fatigued does not distinguish between the ‘background’ level of this symptom in the community and any ‘excess’ arising as a result of illness. The aim of this study was to determine the prevalence of fatigue among palliative care inpatients in comparison with a control group of age and sex-matched volunteers without cancer. In addition, the correlates of fatigue were investigated. The prevalence of ‘severe subjective fatigue’ (defined as fatigue greater than that experienced by % of the control group) was found to be %. Patients were malnourished, had diminished muscle function and were suffering from a number of physical and mental symptoms. The severity of fatigue was unrelated to age, sex, diagnosis, presence or site of metastases, anaemia, dose of opioid or steroid, any of the haematological or biochemical indices (except urea), nutritional status, voluntary muscle function, or mood. A multivariate analysis found that fatigue severity was significantly associated with pain and dypnoea scores in the patients, and with the symptoms of anxiety and depression in the controls. The authors conclude that subjective fatigue is both prevalent and severe among patients with advanced cancer. The causes of this symptom remain obscure. Further work is required in order to determine if the associations reported between fatigue and pain and between fatigue and dyspnoea are causal or coincidental. © Cancer Research Campaign BODY: No Body Content
|
PMC1468398.txt
|
TITLE: Role of acetylcholine and polyspecific cation transporters in serotonin-induced bronchoconstriction in the mouse AUTHORS: Wolfgang Kummer, Silke Wiegand, Sibel Akinci, Ignatz Wessler, Alfred H Schinkel, Jürgen Wess, Hermann Koepsell, Rainer V Haberberger, Katrin S Lips ABSTRACT: BackgroundIt has been proposed that serotonin (-HT)-mediated constriction of the murine trachea is largely dependent on acetylcholine (ACh) released from the epithelium. We recently demonstrated that ACh can be released from non-neuronal cells by corticosteroid-sensitive polyspecific organic cation transporters (OCTs), which are also expressed by airway epithelial cells. Hence, the hypothesis emerged that -HT evokes bronchoconstriction by inducing release of ACh from epithelial cells via OCTs.MethodsWe tested this hypothesis by analysing bronchoconstriction in precision-cut murine lung slices using OCT and muscarinic ACh receptor knockout mouse strains. Epithelial ACh content was measured by HPLC, and the tissue distribution of OCT isoforms was determined by immunohistochemistry.ResultsEpithelial ACh content was significantly higher in OCT1/ double-knockout mice ( ± % of the content of the epithelium-denuded trachea, n = ) than in wild-type mice (. ± . %, n = ). In wild-type mice, -HT ( μM) caused a bronchoconstriction that slightly exceeded that evoked by muscarine ( μM) in intact bronchi but amounted to only % of the response to muscarine after epithelium removal. -HT-induced bronchoconstriction was undiminished in M2/M3 muscarinic ACh receptor double-knockout mice which were entirely unresponsive to muscarine. Corticosterone ( μM) significantly reduced -HT-induced bronchoconstriction in wild-type and OCT1/ double-knockout mice, but not in OCT3 knockout mice. This effect persisted after removal of the bronchial epithelium. Immunohistochemistry localized OCT3 to the bronchial smooth muscle.ConclusionThe doubling of airway epithelial ACh content in OCT1/-/- mice is consistent with the concept that OCT1 and/or mediate ACh release from the respiratory epithelium. This effect, however, does not contribute to -HT-induced constriction of murine intrapulmonary bronchi. Instead, this activity involves ) a non-cholinergic epithelium-dependent component, and ) direct stimulation of bronchial smooth muscle cells, a response which is partly sensitive to acutely administered corticosterone acting on OCT3. These data provide new insights into the mechanisms involved in -HT-induced bronchoconstriction, including novel information about non-genomic, acute effects of corticosteroids on bronchoconstriction. BODY: BackgroundSerotonin (-hydroxytryptamine, -HT) causes constriction of murine airways that is sensitive to atropine both in vivo and in vitro [,]. This response is markedly reduced after removal of the epithelium in the isolated mouse trachea []. Hence, it has been suggested that stimulation of epithelial -HT2A receptors on mouse tracheal epithelial cells triggers the release of acetylcholine (ACh) from these cells, which then causes airway constriction []. In line with this notion, the presence of ACh, its synthesizing enzyme choline acetyltransferase, and of the high-affinity choline transporter, CHT1, that mediates the rate-limiting step of ACh synthesis, has been demonstrated in the airway epithelium of several mammalian species [-,]. It remains unclear, however, by which molecular mechanism ACh is released from airway epithelial cells. In cholinergic neurons, ACh is synthesized in the cytosol by choline acetyltransferase (ChAT), translocated into synaptic vesicles by the vesicular ACh transporter (VAChT) and then released by exocytosis. VAChT expression has been detected in some airway epithelial cells [,]. However, since -HT-induced constriction of the mouse trachea is insensitive to botulinum toxin A [], it is unlikely that exocytotic ACh release is involved in this activity. Recently, polyspecific organic cation transporters (OCTs) have emerged as alternative mediators for the release of ACh. All known OCT isoforms (OCT1-) are expressed by rat and human airway epithelia []. OCT inhibitors and pre-treatment with OCT-anti-sense-oligonucleotides diminish ACh release from human placental villi []. Recently, we demonstrated that rat and human OCT1 and OCT2 expressed by Xenopus oocytes mediate ACh transport, and that this effect could be blocked by corticosteroids [].Hence, we speculated that corticosteroid-sensitive OCTs may mediate -HT-induced ACh release from airway epithelial cells, thus leading to airway constriction in the mouse. In order to test this hypothesis, -HT-induced bronchoconstriction of small intrapulmonary airways and the sensitivity of this response to corticosterone were studied videomorphometrically in precision-cut lung slices (PCLS) [-] taken from OCT1--deficient mice [,]. PCLS offer the advantage to study smallest bronchi whose bronchoconstrictor response can, otherwise, not directly been visualised. The presence of ACh in murine respiratory epithelium was validated by biochemical techniques and ChAT-immunohistochemistry, and we obtained evidence for a significant role of OCT1 and in the release of ACh from airway surface epithelium. The potential involvement of ACh in -HT-induced bronchoconstriction was tested by using mice deficient in both M2 and M3 muscarinic ACh receptors (M2/3R-/- mice). We demonstrated previously that muscarinic agonists are unable to constrict bronchi taken from M2/3R-/- mice []. Surprisingly, the data obtained with these mutant strains revealed that ACh is not involved in -HT-induced bronchoconstriction. On the other hand, we uncovered a direct involvement of smooth muscular OCT3 in -HT-induced bronchoconstriction which proved to be corticosterone-sensitive.MethodsAnimalsLungs were taken from – wk old male M2/3R-/- mutant mice and M2/3R+/+ wild-type mice of the same genetic background [/J1 ( %) × 129SvEv ( %) × CF1 ( %)], OCT1/-/- mice, OCT3-/- mice, and their corresponding wild-type strain (FVB) (all age- and gender-matched). The generation of the mutant mouse strains used in this study has been described previously []. M2/3R-/- mice and the corresponding wild-type strain were kept under specified pathogen-free conditions, whereas the remaining animals were kept in a standard animal facility.ACh assayFVB and OCT1/-/- mice were killed by isoflurane inhalation. Tracheas were carefully cleaned from adhering tissue and fixed in a Petri dish with the luminal surface facing upwards. A cotton-tipped applicator (Q-tip) was gently rubbed along the luminal surface as described earlier [] and thereafter placed in ml % formic acid in acetone (v/v). Epithelium-intact or denuded tracheas were also placed in ml % formic acid in acetone (v/v) and minced with scissors. After a min incubation on ice, Q-tips were removed and the extraction medium was centrifuged ( min; rpm), and the supernatant was evaporated to dryness by nitrogen. The dried sample was resuspended in μl of the mobile phase of the HPLC system, and μl were injected.ACh was measured by cationic exchange HPLC combined with bioreactors and electrochemical detection as described elsewhere [,]. The BAS microbore system was used (Bioanalytical Systems Inc., West Lafayette, USA). ACh and choline were separated on an analytical SepStik column ( × mm; BAS, Axel Semrau, Sprockhövel, Germany) using a mobile phase of mM phosphate buffer and . mM EDTA (adjusted to pH .). The analytical column was followed by an immobilized enzyme reactor containing acetylcholinesterase to hydrolyze ACh and choline oxidase to produce H2O2 from the breakdown product choline. H2O2 flowing across a platinum electrode is oxidized producing a current which is proportional to the amount of ACh in the sample. Twenty μl samples were injected by an automatic injector. The amount of ACh was calculated by comparison with external standard containing pmol/ μl of both ACh and choline.VideomorphometryPCLS were prepared using a slightly modified version of the protocol described by Martin et al. [], as reported in full detail earlier [,]. Very briefly, mice were killed by cervical dislocation, the pulmonary vasculature was flushed blood-free via the right ventricle, and the airways were filled via the cannulated trachea with low melting point agarose (Sigma, Taufkirchen, Germany). Lungs and heart were dissected in toto, cooled, and PCLS were cut (vibratome VT1000S, Leica, Bensheim, Germany) at a thickness of μm from the left lobe of the lung and incubated in minimal essential medium (MEM; GIBCO, Karlsruhe, Germany) at °C for – h to remove the agarose. Experiments were performed in HEPES-Ringer buffer in a lung slice superfusion chamber (Hugo Sachs Elektronik, March, Germany) mounted on an inverted microscope. Images of bronchi of about μm in diameter were recorded with a CCD camera and analyzed with Optimas . software (Stemmer Imaging, Puchheim, Germany). Only those bronchi were included in the final data analysis which responded to a test stimulus of - M muscarine (or, in case of M2/3R-/- mice, - M U44619, a thromboxane analogue) with a reduction of luminal area of at least %.Epithelia were removed after preparation of PCLS and wash-out of agarose. PCLS were placed in HEPES-Ringer buffer in a Petri dish on a binocular stage and immobilized with a mesh of nylon strings connected to a platinum ring. Under microscopic control, the lumen of selected bronchi was manually rubbed with a fine steel-needle (. mm diameter; Faber, Berlin, Germany) mounted onto a wooded rod, until the epithelium could be seen floating off. The position of treated bronchi within PCLS was recorded to assure subsequent re-identification. PCLS were returned for – h into the equilibrium medium in the incubator before the start of the experiments. After completion of the videomorphometric recordings, PCLS were placed on microscopic slides and cover-slipped. The efficiency of epithelium removal was then assessed microscopically. Only those bronchi were included in the analysis in which at least % of the luminal circumference was found to be devoid of epithelial cells. Epithelium denudation of the entire circumference could not be achieved.Muscarine, atropine, -HT, U44619, and corticosterone were purchased from Sigma, Taufkirchen, Germany. Corticosterone was dissolved in ethanol at - M, and diluted in water to the desired experimental concentration immediately before use.ImmunofluorescenceOCTs. Thoraxes of wild-type FVB mice (n = ) and OCT1/-/- mice (n = ) were dissected, the lungs were filled with Tissue-Tek (Sakura Finetek, Zoeterwoude, Netherlands), and the tissues were shock-frozen in melting isopentane. Cryosections ( μm) were fixed in acetone for min at -°C, preincubated for h in phosphate-buffered saline (PBS) containing % horse serum, and then covered for – h with primary antibodies diluted in PBS. The affinity-purified antibody against OCT1 (dilution :; Alpha Diagnostic, San Antonio, TX, USA) was raised against a amino acid sequence near the C-terminus of rat OCT1, which shares % amino acid identity with mouse OCT1. Two affinity-purified antibodies against OCT2 were used. One was raised against amino acids – (near the C-terminus) of human OCT2 (dilution :; []) that share % amino acid identity with mouse OCT2, and the other one was raised against a amino acid sequence near the C-terminus of rat OCT2 (:; Alpha Diagnostic) sharing % amino acid identity with mouse OCT2. The affinity-purified antibody against OCT3 was raised against amino acids – of human OCT3 (dilution : ; []) that share % identity with mouse-OCT3. Since the OCT3 antibody apparently labelled smooth muscle cells, it was also applied in combination with a mouse monoclonal marker antibody for this cell type, i.e. anti-α-smooth muscle actin antibody directly conjugated to fluorescein-isothiocyanate (clone 1A4; Sigma, Taufkirchen, Germany; dilution :) to ascertain muscular localization. After washing in PBS, the sections were incubated for h at room temperature with Cy3-coupled donkey anti-rabbit IgG (: in PBS diluted; Chemicon, Hofheim, Germany) and cover-slipped with carbonate-buffered glycerol (pH .). The sections were evaluated by epifluorescence microscopy (BX60, Olympus, Hamburg, Germany) or with a confocal laser scanning microscope (TCS SP2; Leica, Mannheim, Germany).We have recently demonstrated the specificity of the primary antibodies in OCT1- overexpressing cell lines []. On the present material, it was further validated by (a) omission of the primary antibody, (b) preabsorption with the corresponding antigen ( μg/ml) for h at room temperature prior to use in immunofluorescence, and (c) evaluation of immunofluorescence in OCT-deficient mice.ChAT. Lungs from FVB mice were prepared as described above. Cryosections ( μm) were dipped in phosphate-buffered % picric acid/ % paraformaldehyde, preincubated for h in PBS containing . % Tween (Sigma) and . % bovine serum albumin (Sigma), and covered overnight with a rabbit antiserum (dilution :) raised against a synthetic peptide corresponding to amino acids – of the predicted rat ChAT protein []. This antiserum specifically recognizes the "common type" of ChAT []. After PBS washes, the sections were incubated for h at room temperature with Cy3-coupled donkey anti-rabbit IgG (:; Chemicon), postfixed for min in % buffered paraformaldehyde, washed, and cover-slipped with carbonate-buffered glycerol (pH .). Micropgraphs were taken by confocal laser scanning microscopy.Control sections were incubated with antiserum that had been preincubated with its corresponding peptide ( μg/ml) for h at room temperature prior to use in immunofluorescence.Statistical analysisData are presented as mean ± standard error of the mean. Non-matched groups were compared by Mann-Whitney U-test. In case of more than two groups, analysis was done first by global Kruskal-Wallis rank sum test, and if significant (p < .) differences were observed, comparison between two groups was made by Mann-Whitney U-test. Throughout, differences were considered as statistically significant when p < ..ResultsACh in murine trachea and respiratory epitheliumWe used an HPLC procedure to determine ACh levels separately in epithelium and underlying tissues in wild-type (FVB strain) and OCT1/-/- mice. Using wet weight of the sample as reference, ACh content of the epithelium-denuded trachea was not significantly different in these strains (FVB: . ± . pmol/mg; n = ; OCT1/-/-: .. ± . pmol/mg, n = ). The relative proportion of epithelial ACh, however, was significantly (p < .) higher in OCT1/-/- mice ( ± % of that in the denuded specimens) than in corresponding wild-type (FVB) mice (. ± . %). In a few additional samples, tracheal specimens with intact epithelium were analysed, yielding . ± . pmol/mg in FVB mice (n = ) and . ± . pmol/mg in OCT1/-/- mice (n = ).Bronchi of about μm in diameter were too small to dissect the respiratory epithelium for biochemical ACh analysis. The ACh synthesizing enzyme, ChAT, was demonstrated in epithelial cells of these bronchi by immunohistochemistry (Fig. ).Figure 1Immunohistochemical localization of ChAT in murine peripheral bronchi. Respiratory epithelial cells are strongly ChAT-immunoreactive in wild-type FVB mice (A). The specificity of this labelling is indicated by its absence after preabsorption of the antiserum with its corresponding antigenic peptide (B). Bar represents μm.Role of the epithelium and of ACh in -HT-induced bronchoconstrictionSmall intrapulmonary bronchi from M2/3R+/+ wild-type mice strongly constricted in response to both muscarine (- M) and to -HT (- M; Fig. ). The magnitude of the -HT-induced bronchoconstriction even surpassed that evoked by muscarine (Fig. ). Mechanical (partial) removal of the epithelium diminished the constriction to muscarine (Fig. ), consistent with the results of a previous study involving the chemical (Triton X-) ablation of the murine tracheal epithelium []. Removal of the airway epithelium also led to a significant reduction in the -HT-induced bronchoconstriction response (Fig. ). However, removal of the epithelium had a more pronounced effect on -HT- than on muscarine-induced bronchoconstriction. Thus, in contrast to intact bronchi, the magnitude of the -HT response was smaller than that evoked by muscarine after epithelium removal.Bronchi from M2/3R-/- mice were entirely unresponsive to muscarine (- M; Fig. ), as reported earlier []. In striking contrast, -HT (- M) induced indistinguishable bronchoconstrictor responses in M2/3R-/- mutant and M2/3R+/+ wild-type mice, both in absolute values and expressed as percent response evoked by the thromboxane analogue, U46610 (- M) (Fig. ).Figure 2Effect of epithelium removal on constriction of peripheral bronchi in PCLS of M2/3R+/+ mice. (A) Reduction of luminal area of intact (control, blue) and epithelium-denuded (denuded, red) peripheral bronchi in response to muscarine (Mus, - M) and -HT (- M). The numbers in parentheses refer to the numbers of bronchi/number of lungs from which they were taken. Panel (B) illustrates the magnitude of the response to -HT (- M) compared to that to muscarine (- M) which was set as %. Control bronchi react slightly stronger to -HT than to muscarine, whereas the -HT response is significantly smaller that the muscarine response after epithelium removal, particularly at min (') after agonist application. The box plots shows percentiles , , (median), , and ; individual data points beyond × S.D. are indicated by * or °. ***p < ., **p < ., *p < . (comparison of corresponding time points by Mann-Whitney U-test). (C) Microscopic appearance of control and epithelium-denuded bronchi. In the left panel, arrowheads indicate thickness of the epithelial layer in a control bronchus. In the right panel, the arrowhead points to a small remnant of epithelium after mechanical denudation of the epithelium.In preparations from M2/3R+/+ wild-type mice, atropine (- M) partially inhibited -HT-induced constriction (Fig. 4A). The same concentration of atropine fully blocked muscarine-induced bronchoconstriction (data not shown, see our previous report []). At a higher concentration (- M), however, atropine reduced -HT-induced bronchoconstriction by approximately % in all strains tested, including M2/3R-/-, OCT1/-/-, OCT3-/-, and corresponding wild-type mice (Fig. 4B, C).Figure 3Changes in luminal area of peripheral bronchi in response to muscarine (Mus, - M), -HT (- M), and U44619 (- M) in wild-type (M2/3R+/+) and M2/3R-/- mice. (A) -HT induces similar responses in both strains. The numbers in parentheses refer to the numbers of bronchi/number of lungs from which they were taken. Panel (B) expresses the -HT-induced constriction in percent of that evoked by U44619 in the first min after agonist application. The box plots shows percentiles , , (median), , and ; * indicates an individual data point beyond × S.D. (C) Original images of a peripheral bronchus of a wild-type and an M2/3R-/- double-knockout animal before and after agonist application. As depicted in (A), there is no constriction in response to muscarine in M2/3R-/- mice. On the other hand, both strains show identical responses to -HT and U44619.Distribution of OCTs in murine bronchiImmunohistochemistry revealed OCT1-immunoreactivity in the apical membrane of ciliated cells (Fig. 5A). This labelling was OCT1-specific since it was absent when the antiserum was preabsorbed with the corresponding antigenic peptide and when tissue from OCT1/-/- mice was used for immunohistochemistry (Fig. 5B, C). No specific OCT2-immunolabelling was observed in the bronchial wall (Fig. 5D–F). Specific OCT3-immunoreactivity was most intense in the bronchial smooth muscle and weaker on epithelial cells (Fig. 5G–G").Figure 4Effect of atropine on -HT-induced bronchoconstriction (reduction of bronchial luminal area) in PCLS. Atropine blocks -HT-induced constriction partially at - M (A), and nearly completely at - M, even in absence of both M2 and M3 muscarinic receptors (B). The numbers in parentheses refer to the numbers of bronchi/number of lungs from which they were taken. (C) Persisting bronchoconstriction in response to -HT (- M) in the presence of - M atropine in different wild-type and knockout strains. The initial -HT-induced bronchoconstriction was set as %.Role of OCTs in -HT-induced bronchoconstrictionSmall intrapulmonary bronchi from OCT1/-/-, OCT3-/-, and OCT1-+/+ wild-type mice reacted with a strong constriction to muscarine (- M) and to -HT (- M) (Fig. 6A, B). The absence of OCT1/ or OCT3 had no significant effect on the -HT bronchoconstrictor response. Corticosterone (- M) significantly reduced the -HT-induced bronchoconstriction both in wild-type and in OCT1/-/- mice but was ineffective in OCT3-/- mice (Fig. 6C, D). The effect of epithelium removal on the inhibitory action of corticosterone on -HT-induced bronchoconstriction was investigated in M2/3R+/+ wild-type mice. In intact bronchi from this strain, ± % (mean ± S.E.M.; PCLS from lungs) of the -HT-induced contraction remained in the presence of corticosterone, so that the corticosterone effect was not as marked as in OCT1-+/+ wild-type (FVB) mice. This small, but significant reduction of -HT-induced contraction by corticosterone in M2/3R+/+ wild-type mice was still present after epithelium removal (remaining contraction: ± %; mean ± S.E.M.; PCLS from lungs).Figure 5Immunohistochemical localization of OCTs in murine bronchi. OCT1-immunolabelling is localized to the apical membrane of ciliated epithelial cells in wild-type FVB mice (arrows in A). The specificity of this labelling is indicated by its absence after preabsorption of the antiserum with its corresponding antigenic peptide (B) and the lack of labelling in OCT1/-/- mice (C). Neither of the two OCT2-antibodies used in this study showed specific labelling of mouse bronchi (D, E). The spotty labelling of epithelial cells observed with the OCT2-antibody raised against the human sequence (E) was also observed in OCT1/-/- mice (F), indicating that this signal is non-specific. Specific OCT3-immunolabelling, documented by its absence in the preabsorption control (inset in G), is observed primarily on the bronchial smooth muscle (sm) and, less intensely, on epithelial cells (G). OCT3-localization in smooth muscle cells is confirmed by double-labelling immunofluorescence with OCT3-antibody and a monoclonal antibody against α-smooth muscle actin (SMA) (G') yielding the yellow signal in the merged image (G'). Bar represents μm in A-F and μm in G-G".DiscussionThe present data clearly demonstrate an epithelium-dependent component of -HT-induced bronchoconstriction in the mouse, consistent with the results of a previous study on the mouse trachea []. It has been suggested that this activity is dependent on the release of ACh from airway epithelial cells []. In the Xenopus oocyte expression system, both OCT1 and , but not OCT3, proved to be able to translocate ACh across the plasma membrane []. In the present study, we found that the airway epithelial ACh content was twice as high in OCT1/-/- than in wild-type mice. This observation supports the concept that OCT1/ may also play a role in the release of ACh from airway epithelia. However, to our surprise, the magnitude of -HT-induced bronchoconstrictor responses was unchanged in PCLS preparations from OCT1/-/- mice, indicating that -HT-induced bronchoconstriction does not require the presence of OCT1 and . Moreover, videomorphometric studies showed that PCLS from M2/3R-/- mice remained fully responsive to -HT. In contrast, PCLS from M2/3R-/- mice do no longer show a bronchoconstrictor response following cholinergic stimulation, as shown in this and in an earlier study []. These data clearly indicate that the release of epithelial ACh is not involved in the -HT-induced bronchoconstrictor response, but that another epithelium-derived constrictory factor contributes to this activity.In previous studies, ACh emerged as a candidate for mediating -HT-induced airway constriction in the mouse because this effect could be inhibited by atropine [-]. In the present study, we found a large reduction of -HT-induced bronchoconstriction only after application of an unusually high concentration of atropine (- M). On the other hand, a much smaller concentration of atropine (- M) was sufficient to fully block muscarine-induced bronchoconstriction. Interestingly, Eum et al. [] also did not observe a significant inhibition of -HT-induced contraction of the isolated mouse trachea at - M atropine. The inhibition of -HT-induced bronchoconstriction by - M atropine persisted in M2/3R-/- mice, clearly indicating that this high concentration of atropine inhibits airway smooth muscle contractility via non-specific effects that are not due to muscarinic receptor blockade. Indeed, atropine has been described as a competitive antagonist at the -HT3-receptor []. Taken together, the present data demonstrate that -HT releases an epithelium-derived bronchoconstrictory factor that is OCT-independent and different from ACh.We made the striking observation that corticosterone exerted an acute inhibitory effect on -HT-induced bronchoconstriction. This acute effect of corticosterone was mediated by OCT3, as demonstrated by its absence in OCT3-/- mice. This finding is of potential clinical relevance since rapid therapeutical effects of a bolus of inhaled glucocorticoids have been reported in asthmatic patients where they reverse airway subsensitivity to β2-agonists [,]. In our model, the inhibitory action of corticosterone on -HT-induced bronchoconstriction is epithelium-independent since it persisted after epithelium removal. In line with this observation, immunohistochemistry demonstrated that OCT3 is located directly on bronchial smooth muscle cells. In principle, all OCT isoforms tested so far are sensitive to corticosteroids that are not substrates for transport by themselves but inhibit transport of other substances []. OCT3, which we identified as being responsible for the acute inhibitory effect of corticosterone on -HT-induced bronchoconstriction, has the highest affinity for corticosteroids []. It also clears monoamines, including catecholamines and -HT, from the extracellular space [], and hence its blockade is expected to increase the extracellular concentrations of these agents. Indeed, acute human bronchial vasoconstriction elicited by corticosteroids has been explained by inhibition of OCT3 with subsequent rise of extracellular noradrenaline and prolonged activation of α1-adrenoreceptors []. However, a separate, specific serotonin transporter (SERT) is highly expressed in the lung [,]. As a result, deficiency or blockade of OCT3 may have little impact on -HT turnover. In agreement with this notion, the magnitude of the bronchoconstrictor response to -HT remained unchanged in bronchi from OCT3-/- mice and the -HT response was reduced rather than augmented by corticosterone. It is therefore unlikely that the observed OCT3-mediated inhibition of -HT-induced bronchoconstriction by acutely administered corticosterone involves direct interference with -HT transport. In view of the electrogenic properties of OCTs [], the acute inhibitory effect of corticosterone on -HT-induced bronchoconstriction might be caused by modulation of membrane potential, but the underlying signal transduction cascade still awaits to be clarified.Conclusion5-HT-induced constriction of murine intrapulmonary bronchi involves two independent pathways. One pathway is dependent on the release of an epithelium-derived constrictory factor that is different from ACh. The second pathway involves the direct stimulation of bronchial smooth muscle cells. This latter pathway is partly sensitive to acutely administered corticosterone acting on OCT3. These data provide new insights into the mechanisms involved in -HT-induced bronchoconstriction, including novel information about non-genomic, acute pulmonary effects of corticosteroids.Competing interestsThe author(s) declare that they have no competing interests.Authors' contributionsWK carried out the epithelium removal, evaluated immunohistochemistry, participated in the design of the study and drafted the manuscript. SW carried out epithelium removal, videomorphometric and statistical analyses. SA carried out videomorphometric and statistical analyses. IW performed the ACh assay and revised the manuscript critically for important intellectual content. AHS provided OCT-deficient mice and revised the manuscript critically for important intellectual content. JW provided M2R/M3R-deficient mice and revised the manuscript critically for important intellectual content. HK provided antibodies, added to the design of the study and revised the manuscript critically for important intellectual content. RVH coordinated the videomorphometric setup and breeding of genetically deficient mice strains, and revised the manuscript critically for important intellectual content. KSL performed and evaluated immunohistochemistry, and participated in the design of the study and drafting of the manuscript. The data presented in this manuscript are part of the doctoral thesis of SA.Figure -HT-induced reduction of bronchial luminal area (bronchoconstriction) in OCT-deficient mice and sensitivity of this response to corticosterone. (A, B) Wild-type FVB mice (OCT1-+/+), OCT1/-/- mice and OCT3-/- mice exhibit no differences in their response to -HT (- M). The numbers in parentheses refer to the numbers of bronchi/number of lungs from which they were taken. (C, D) In wild-type and OCT1/-/- mice, but not in OCT3-/- mice, the bronchoconstriction in response to -HT is significantly reduced by corticosterone (Cs, - M). Panel (D) depicts the bronchoconstrictor response to -HT (- M, min after administration) in the presence of corticosterone (- M), as compared to the response to -HT alone (set as %). **p < ., Mann-Whitney U-test.
|
PMC3068949.txt
|
TITLE: SERPINE2, an inhibitor of plasminogen activators, is highly expressed in the human endometrium during the secretory phase AUTHORS: Robert Kuo-Kuang Lee, Chi-Chen Fan, Yuh-Ming Hwu, Chung-Hao Lu, Ming-Huei Lin, Ying-Jie Chen, Sheng-Hsiang Li ABSTRACT: BackgroundSERPINE2, also known as protease nexin-, belongs to the serine protease inhibitor (SERPIN) superfamily. It is one of the potent SERPINs that modulates the activity of plasminogen activators (PAs). PAs and their SERPIN inhibitors, such as SERPINB2 and SERPINE1, were expressed in the human endometrium and were implicated in implantation. However, expression data about SERPINE2 in the human endometrium is still unknown. Thus, we conducted an investigation to reveal the spatiotemporal and cellular expression of SERPINE2 in the human uterus during the menstrual cycle.MethodsSeven patients who underwent a hysterectomy and samples of archived patients' endometrial curettage or parts of the uterus that were formalin-fixed and embedded in paraffin. Western blotting was performed to evaluate the specificity and sensitivity of the antibody. Immunohistochemistry was conducted to localize the SERPINE2 expression site. Quantitative analysis was conducted to evaluate expression levels of SERPINE2 in various sub-phases of the menstrual cycle.ResultsThe SERPINE2 protein was primarily detected in the uterine fluid during the mid- and late-secretory phases of the menstrual cycle. It was predominantly expressed in the luminal and glandular epithelium, less in the myometrium, and only dispersedly in certain stromal cells throughout the menstrual cycle. A quantitative analysis of expression levels of SERPINE2 in the glandular epithelium revealed that it was highly expressed in the endometrium during the secretory phase compared to the proliferative phase.ConclusionsThe SERPINE2 protein is highly expressed in the endometrium during the secretory phase, indicating that it may participate in tissue remodeling involved in implantation. BODY: BackgroundThe plasminogen activator (PA) system refers to the PA and its cognate inhibitors []. The system includes two forms of PA, tissue-type (PLAT or tPA) and urokinase-type (PLAU or uPA), and four forms of serine protease inhibitors (SERPIN), including SERPINA5 (also called protein C inhibitor, PCI), SERPINB2 (also called plasminogen activator inhibitor , PAI-), SERPINE1 (also called plasminogen activator inhibitor , PAI-), and SERPINE2 (also called protease nexin-, PN-) [].The PA system is associated with many reproductive processes, including ovulation, embryogenesis, and embryo implantation in female reproductive tissues [,]. How SERPIN modulates the proteolytic activities of PLAT/PLAU in reproductive tissue remodeling is of great importance.Tissue remodeling requires a fine-tuned balance between levels of proteases and their cognate inhibitors. The PA is involved in tissue remodeling by converting abundant extracellular plasminogen into plasmin, an active protease, which degrades the extracellular matrix. The classical substrate of plasmin is fibrin, but in fact, many other matrix proteins can be cleaved by this enzyme [].The expression and activity of PLAT and PLAU were detected in the human uterus, including the uterine fluid [-], and the endometrium during cycling [-] and implantation []. SERPINB2 [] and SERPINE1 [,] were demonstrated to be present in the human endometrium. SERPINE1 was even detected in human and mouse uteri during implantation [,], indicating that the PA inhibitor is involved in implantation.SERPINE2 has broad anti-protease activity specific to serine proteases, including trypsin, thrombin, plasmin, PLAU [], and prostasin []. It is widely expressed in various tissues []. In the uterus, it is reported that expression levels of SERPINE2 in the monkey endometrium and placenta during early pregnancy were weak or below the level of detection []. In rats, Serpine2 messenger (m)RNA was exclusively detected in endometrial stromal cells of the uterus, in particular on day . postcoitally, thus suggesting that it may be involved in the implantation process [].Recently, SERPINE2 protein was reported to be expressed in the murine uterus during the estrous cycle, pregnancy, and lactation []. It is predominantly expressed in the luminal and glandular epithelium and weakly in stromal cells and myometrium. It seems that different species have different expression patterns for the SERPINE2 protein. So far, there is no study on this aspect in the human uterus. Herein, we conducted an investigation to reveal the spatiotemporal and cellular expression of SERPINE2 in the human uterus during the menstrual cycle.MethodsSample collectionUterine fluid aspirates were collected under the consent of patients (n = ) who were to undergo a hysterectomy because of a leiomyoma or adenomyosis. After anesthesia and before surgery, uterine fluid in the cavity was directly aspirated using an embryo transfer catheter (Labotect, Goettingen, Germany). Then, the cavity was flushed using mL of saline. The two solutions were mixed for analysis. On day of the operation, blood was aspirated to examine the serum concentration of estradiol and progesterone to evaluate the phase of the menstrual cycle. A sample from endometrial curettage was also formalin-fixed and paraffin-embedded (FFPE) for a histological evaluation. Menstrual cycle can be dated into sub-phases according to the anatomical changes within the endometrial biopsy including, early proliferative (EP) (days -), mid-proliferative (MP) (days -), late proliferative (LP) (days -), early secretory (LS) (days -), mid-secretory (MS) (days -), and late secretory (LS) (days -) phases []. Informed consent for all samples was obtained from all patients.To analyze the SERPINE2 protein expression in different phases of the menstrual cycle, archived FFPE samples of endometrial curettage or parts of the uterus from patients of reproductive age at various sub-phases of the menstrual cycle judged by a pathologist were obtained from the Department of Pathology, Mackay Memorial Hospital (Taipei, Taiwan). Wax blocks, cases of EP, cases of MP, cases of LP, cases of ES, cases of MS, and cases of LS, were prepared from patients who were to undergo a hysterectomy because of a leiomyoma or adenomyosis. All retrieved wax blocks had been treated at the hospital for years. The Institutional Review Board approved this study (reference number: MMH-I-S-).Western blottingTo assess the specificity of the antibody, ng of recombinant glutathione S-transferase (GST)-human SERPINE2 protein (Abnova, Taipei, Taiwan) was resolved by sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) on a % gel slab (. × . × . cm) and was transferred to a nitrocellulose membrane for immunostaining according to a previously described method []. Membranes were blocked with % (w/v) skim milk in phosphate-buffered saline (PBS) (blocking solution) for h, and then incubated with our homemade anti-mouse SERPINE2 antiserum (: ) [,], anti-human SERPINE2 antibody (: , catalog no. AF2980, R&D Systems, Minneapolis, MN, USA), or another anti-human SERPINE2 antibody (: , product no. H00005270-B01, Abnova) in blocking solution for h at °C. After gentle agitation in four changes of PBS for min each, membranes were immunoreacted with horseradish peroxidase (HRP)-conjugated goat anti-rabbit immunoglobulin G (IgG) (: , GE Healthcare Life Sciences, Piscataway, NJ, USA), donkey anti-goat IgG (: , Jackson ImmunoResearch, West Grove, PA, USA), or horse anti-mouse IgG (:, Cell Signaling Technology, Beverly, MA, USA), respectively, in blocking solution for h at °C. After gentle agitation as mentioned above, immunoreactive bands were revealed using an enhanced chemiluminescent substrate according to the manufacturer's instructions (Western Lightning; PerkinElmer, Boston, MA, USA).To evaluate the sensitivity of the antibody, μg of a tissue extract from endometrial curetting was resolved, transferred onto a blot, and detected using anti-mouse SERPINE2 antiserum (: ) or an anti-human SERPINE2 antibody (: ) following the method described above.To examine the expression of SERPINE2 in the uterine fluid, μg of uterine luminal proteins was applied and detected using anti-mouse SERPINE2 antiserum (: ) following the same method.Immunohistochemical staining and signal analysisImmunohistochemical analysis was performed on an automatic staining machine (BenchMark XT, Ventana Medical Systems, Tucson, AZ, USA) using the iVIEW , -diaminobenzidine (DAB) detection kit (Ventana Medical Systems). After tissue sections ( μm) on slides were deparaffinized and hydrated, they were heated to °C for min and then °C for min to induce antigen retrieval using Ventana Cell Conditioner (Ventana Medical Systems). After cooling to room temperature for min, the slides were treated with an iVIEW inhibitor at °C for min to inactivate the endogenous peroxidase activity. Slides were then incubated with anti-SERPINE2 antiserum or antiserum pretreated with SERPINE2 antigen-conjugated beads diluted : in blocking solution at °C for min. After rinsing with PBS, slides were treated with iVIEW biotin-conjugated IgG in blocking solution for min at room temperature. Slides were rinsed again and then incubated with iVIEW streptavidin-conjugated HRP in blocking solution for min at room temperature. Protein signals were developed by iVIEW DAB and hydrogen peroxide for min at °C. Slides were finally incubated with iVIEW copper for min to enhance the signal intensity, counterstained with hematoxylin (Vector Laboratories, Burlingame, CA), and photographed with a Nikon Eclipse E600 microscope (Tokyo, Japan) or Zeiss AxioImager Z1 microscope system (Wetzlar, Germany) equipped with a CCD camera and an automated acquisition system (TissueGnostics, Vienna, Austria). Pictures were acquired using TissueFaxs software (TissueGnostics). The percentage of positively stained cells was determined using HistoQuest software (TissueGnostics). Five endometrial glands on a slide for each individual patient at a given sub-phase of the menstrual cycle were chosen to evaluate the level of staining according to the mean intensity of DAB, refers to the staining signal of SERPINE2, of the stained cells which were counterstained with hematoxylin. The strong intensity was suggested that more than % of glandular epithelial cells had a staining intensity of > . The weak intensity was suggested that more than % of glandular epithelial cells had a staining signal of < . The staining signal other than the two cases was referred to the medium signal. A chi-square test was performed to compare the significance of difference in expression levels of SERPINE2 among patients at different sub-phases of menstrual cycle.ResultsTo assess the specificity of the antibodies, we analyzed two commercially available anti-human SERPINE2 antibodies and homemade anti-mouse SERPINE2 antiserum by Western blotting. As shown in Figure , all antibodies could recognize the recombinant GST-human SERPINE2 protein (Figure 1A), indicating that our antibody has specificity for the human SERPINE2 protein.Figure 1Antibody specificity and presence of the SERPINE2 protein in human uterine fluid. (A) Four hundred nanograms of recombinant human SERPINE2 was resolved on % SDS-PAGE and followed by Western blotting using anti-mouse SERPINE2 antiserum (lane ), an anti-human SERPINE2 antibody (R&D) (lane2), or another anti-human SERPINE2 antibody (Abnova) (lane ). (B) One hundred micrograms of the extract of endometrial curettage was analyzed by anti-mouse SERPINE2 antiserum (lanes and ) and an anti-human SERPINE2 antibody (R&D) (lanes and ). (C) Fifty micrograms of uterine fluid proteins collected from each individual patient (n = ) was Western-blotted using anti-mouse SERPINE2 antiserum (:) (upper panel). EP, MP, and LP indicate early-, mid-, and late-proliferative phases, and ES and MS indicate early- and mid-secretory phases, respectively. The blot was also overexposed to clearly display the staining signal (lower panel).To evaluate the sensitivity of the antibody, a tissue extract from endometrial curettage was resolved on a gel and Western-blotted using the three antibodies. An apparent -kDa protein band corresponding to human SERPINE2 was immunodetected by our anti-mouse SERPINE2 antibody, while many non-specific crossed-reacted protein bands were blotted by the commercial antibodies. In addition, an approximately -kDa protein band corresponding to the complex of SERPINE2 and a certain protease demonstrated in previous studies [,,,] was found. Thus, the sensitivity of our antibody apparently excelled those of the commercial antibodies (Figure 1B). These data indicated that our antibody has high specificity and sensitivity for detecting the human SERPINE2 protein.Using the antibody, SERPINE2 was detected in the endometrial fluid of the human uterus (Figure 1C). It was obvious that there was more SERPINE2 in the secretory phase (lanes and ) than in the proliferative phase (lanes , , , and ). Thus, SERPINE2 is indeed a human uterine secretory protein.An immunolocalization study was conducted to reveal the cellular localization of the SERPINE2 protein in the early secretory phase uterus. As shown in Figure , SERPINE2 was primarily detected on the apical surface of the luminal epithelium (Figure 2A) and glandular epithelium (Figure 2B), but weakly on the myometrium (Figure 2C). In contrast, the protein signal in stromal cells was dispersedly distributed on some unidentified cells, parts of which may be macrophage as judged by cell morphology (Figure 2B). In addition, SERPINE2 was detected on endothelial cells of the vessel (Figure 2C) as demonstrated by a previous study []. When slides were immunostained with control antiserum, no signal was detected (data not shown).Figure 2Localization of SERPINE2 in the human uterus. Longitudinal sections of the early secretory phase uterus (n = ) on the slide were incubated with anti-mouse SERPINE2 antiserum and then treated with biotin-conjugated goat-anti-rabbit IgG and HRP-conjugated streptavidin (brown). For contrast, specimens were further stained with hematoxylin (blue). The representative picture is shown. Magnified pictures of the luminal epithelium (A), glandular epithelium (B), and myometrium (C) are shown. Bar = μm. bv, blood vessel; ge, glandular epithelium; le, luminal epithelium; m, muscle; s, stroma.To further evaluate the expression of the SERPINE2 protein in various sub-phases of the menstrual cycle, we examined endometrial slides prepared from early-, mid-, and late-proliferative as well as secretory phases by immunohistochemistry. The results showed that the signal was relatively weaker during the proliferative phase (Figure 3A-C), while it was strongly detected in the glandular epithelium during the secretory phase (Figure 3D-F), especially in the mid- and late-secretory phase (Figure 3E, F).Figure 3Glandular epithelial expression of the SERPINE2 protein in the endometrium during the menstrual cycle. Sections prepared from endometrial curettage during early-proliferative (EP), mid-proliferative (MP), late-proliferative (LP), early-secretory (ES), mid-secretory (MS), and late-secretory (LS) phases were incubated with anti-mouse SERPINE2 antiserum and then treated as described in Figure . Bar = μm. bv, blood vessel; ge, glandular epithelium; s, stroma.Positively stained cells in the endometrial gland were quantified to analyze expression levels of the SERPINE2 protein. The signal intensity was determined by quantitative software. Scattergrams of the staining intensity (Figure 4B) indicated strongly and weakly stained cells in the glandular epithelium (Figure 4A). The results implied that weakly stained cells in the gland were mostly derived from the proliferative phase, while strongly stained cells were predominantly from the secretory phase (Figure 4C). Thus, SERPINE2 was primarily expressed in secretory-phase endometrial glandular cells from which it was secreted into the lumen of the uterus.Figure 4Quantification of SERPINE2 protein expression levels in endometrial glands. Representative samples were analyzed by automated cell acquisition and quantification software (A). The expression signal of a respective glandular gland was quantified using HistoQuest software and is presented as a scattergram. Each spot on the scattergram stands for the intensity of one cell (B). The relative SERPINE2 protein expression levels in patients' glandular glands at various sub-phases of the menstrual cycle are shown as bar diagrams (C). Differences are significant among patients at various groups (χ = ., p < .).DiscussionIn this study, we demonstrated that the SERPINE2 protein, an inhibitor of PAs, is highly expressed in the human uterus during the secretory phase. However, its levels are low in the uterus during the proliferative phase. It is primarily expressed in the luminal and glandular epithelium, weakly expressed in the myometrium, and dispersedly expressed by certain stromal cells.Proteases are known to be involved in extracellular matrix degradation required for implantation, including cysteine, serine, and matrix metalloproteinases []. PA serine proteases and their cognate inhibitors are involved in implantation [,]. Tissue remodeling is an important biological event for many reproductive processes that occur in the ovary, uterus, and placenta, such as follicle growth, ovulation, the estrous cycle, implantation, and placentation [,]. SERPINE2 was previously demonstrated to primarily be upregulated in dominant follicles during follicle growth in cattle [,] and during ovulation in mice []. It is also highly expressed in the rodent uterus during implantation and pregnancy [].Similar to the rodent data, in this study, we revealed that the SERPINE2 protein was highly expressed in the glandular epithelium of human uterus around the time of the implantation window. These data suggest a role for SERPINE2 in regulating tissue remodeling during implantation.Human SERPINE2 has high amino-acid identities with the rat and mouse SERPINE2, at % and %, respectively. Anti-mouse SERPINE2 antiserum can cross-react with rat SERPINE2 in the uterus [], indicating that it would cross-react with human SERPINE2. In this study, we demonstrated that anti-mouse SERPINE2 antiserum recognized the human SERPINE2 protein. Interestingly, we found that only one SERPINE2 protein form existed in the human uterus, while two isoforms were found in the mouse and rat uterus [].Our anti-SERPINE2 antibody, produced using a highly purified protein as the immunogen [], was very sensitive at detecting tissues in which there was trace expression of SERPINE2 [,], even better than the commercial antibodies. In this study, the commercial antibodies could not specifically detect SERPINE2's expression in endometrial tissues, demonstrating the sensitivity of our antibody. Weak or very low levels of SERPINE2 in the monkey endometrium and placenta in a previous study may have resulted from the fact that they used the peptide or Escherichia coli-expressed protein as the immunogen which is often not in the native conformation of the protein [].Previous studies demonstrated that SERPINE2 can form a complex of approximately kDa with PLAU [,,] or about kDa with prostasin []. A -kDa protein complex was found in the protein extract of murine and rat uteri []. Similarly, the complex was also detected in a human endometrial tissue extract (Figure 1B), indicating that SERPINE2 may act in the human uterus with a certain protease that may be involved in tissue remodeling.Prostasin is highly expressed, but its inhibitor, SERPINE2, is barely expressed in the glandular epithelium of the monkey endometrium during early pregnancy []. However, the SERPINE2 protein is highly expressed in the human endometrium, although prostasin protein expression in the human uterus is still unknown. Whether SERPINE2 regulates the proteolytic activity of prostasin in the human uterus awaits further investigation.PLAT, PLAU, SERPINE1, and SERPINB2 were found to be expressed in the human endometrium [-]. PLAU mRNA is expressed by stromal cells in all phases of the menstrual cycle. However, the PLAU protein is expressed by epithelial and stromal cells in the early-proliferative and late-secretory phases but is nearly undetectable in the mid-secretory phase of the menstrual cycle []. SERPINE1 mRNA and protein were reported to predominantly be expressed by stromal cells in the late secretory phase []. The SERPINB2 protein was found at very low concentrations and was expressed by certain stromal cells in the human endometrium. Expression levels of this protein showed no difference between the proliferative and secretory phases []. Our preliminary results in analyzing the relative mRNA expression levels of PA inhibitors showed that SERPINE2 mRNA was the most highly expressed PA inhibitor in the extract of the human uterus (unpublished observations). Thus, the SERPINE2 protein may be the major PA inhibitor in the human uterus. This was also found in the mouse uterus [].While SERPINE2 was expressed in the mouse and rat uteri [], it was below the level of detection in the monkey uterus []. This is the first report to demonstrate that SERPINE2 is prominently expressed by the human endometrium. We also found that the SERPINE2 protein is present in the uterine fluid. PA activity was also demonstrated to be present in the uterine fluid [-]. Thus, the SERPINE2 protein may exert an inhibitory role of modulating PA activity in the uterine milieu.In conclusion, cellular localization of the SERPINE2 protein in the human uterus suggests that it may play important roles in PA-modulated tissue remodeling. The high expression of the SERPINE2 protein in the secretory phase suggests that it might be associated with embryo implantation.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsRKKL participated in the design of the study and helped draft the manuscript. CCF carried out immunohistochemistry and drafted parts of the manuscript. YMH and MHL participated in sample collection. CHL carried out Western blotting. YJC helped the immunohistochemical signal quantification. SHL conceived of the study, and participated in the project design and coordination. All authors read and approved the final manuscript.
|
PMC1513210.txt
|
TITLE: The time-course of a scrapie outbreak AUTHORS: K Marie McIntyre, Simon Gubbins, Wilfred Goldmann, Emily Stevenson, Matthew Baylis ABSTRACT: BackgroundBecause the incubation period of scrapie has a strong host genetic component and a dose-response relationship, it is possible that changes will occur during an outbreak, especially in the genotypes of cases, age-at-onset of disease and, perhaps, the clinical signs displayed. We investigated these factors for a large outbreak of natural scrapie, which yielded sufficient data to detect temporal trends.ResultsCases occurred mostly in two genotypes, VRQ/VRQ and VRQ/ARQ, with those early in the outbreak more likely to be of the VRQ/VRQ genotype. As the epidemic progressed, the age-at-onset of disease increased, which reflected changes in the genotypes of cases rather than changes in the age-at-onset within genotypes. Clinical signs of cases changed over the course of the outbreak. As the epidemic progressed VRQ/VRQ and VRQ/ARQ sheep were more likely to be reported with behavioural changes, while VRQ/VRQ sheep only were less likely to be reported with loss of condition.ConclusionThis study of one of the largest scrapie outbreaks in the UK allowed investigation of the effect of PrP genotype on other epidemiological parameters. Our analysis indicated that, although age-at-onset and clinical signs changed over time, the observed changes were largely, but not exclusively, driven by the time course of the PrP genotypes of cases. BODY: BackgroundScrapie is an infectious disease, but one that is unusual in that the incubation period leading to clinical disease is governed by a strong host genetic component [] and a strong dose-response relationship []. These factors can interact to yield intriguing epidemiological patterns.During an outbreak, the PrP genotypes of affected animals are expected to change over time. Those encoding PrP genotypes with the shortest incubation periods should be the first to reach clinical onset, followed by their co-occurrence with those encoding PrP genotypes with longer incubation periods, then those with the longest incubation periods. If most sheep are infected at a similar age, this could manifest as a change in the age-at-onset of scrapie over time. The level of infectivity to which susceptible animals are exposed could also increase through the epidemic as the cumulative total of infected animals rises. A higher dose may shorten the incubation period and lead to a reduction in the age-at-onset over time. If PrP genotype, incubation period or age-at-onset change with time during an outbreak, then any linkage between any of these factors and the clinical signs of scrapie could also lead to changes in the clinical manifestation of the disease over time. A time course of clinical signs may also arise from the farmer, who may be able to recognise scrapie on the basis of fewer or different clinical signs as more cases are encountered.These ideas are supported by the reports of farmers running commercial flocks affected by natural scrapie. In a field study undertaken by the Institute for Animal Health (IAH), twenty-four percent of farmers reported a change in the clinical signs being reported through the course of a scrapie epidemic, while thirty-five percent reported a change in the ages of cases.Detection of a time trend, and teasing apart the causal factors behind it, is made difficult by the small number of cases confirmed in most outbreaks and the range of PrP genotypes involved. Furthermore, variability between flocks in the characteristics of outbreaks precludes, in most cases, the combining of datasets.Here, we have investigated in detail the scrapie outbreak of one of very few flocks for which sufficient data are available for time course analysis. Between and there were over confirmed cases of scrapie in a single flock of Welsh Mountain sheep. Almost all cases were in just two genotypes. This outbreak is possibly the largest reported for any flock in Great Britain over the same time period. The scale of the outbreak gave us an opportunity to explore in detail the time course of the scrapie epidemic at the level of PrP genotype and with extensive detail of the clinical presentation of disease.ResultsThe flockTen (out of ) PrP genotypes were found in this flock upon blood sampling in (Table ). A quarter (.%) carried the VRQ allele (associated with the highest risk of scrapie), but only .% were of the most susceptible genotype, VRQ/VRQ. The genotype frequencies correspond to allele frequencies of: ARR, .%; AHQ, .%; ARQ, .%; and VRQ, .%. The ARH allele was not detected in this flock. Animals also had PrP haplotypes ARQ-F141 and ARQ-S241 which occurred at frequencies of .% and .%, respectively. Neither of these haplotypes was associated with the occurrence of scrapie in this flock.The outbreakThe farmer did not know at what point his flock acquired infection. Although he thought that there had been one case of scrapie (unconfirmed) in , he did not recognise any more animals with scrapie-like signs until the start of the current epidemic in September . The affected animals exhibited clinical signs of incoordination (ataxia), nervousness/excitability (nervous) and rubbing/scratching (pruritis); he thought that both the clinical signs and the age of the animals succumbing to disease changed during the course of the epidemic.Between September and February there were confirmed cases in the flock, in animals of genotypes (Table ). The frequency of cases increased to a peak in and subsequently decreased (Figure ). The number of cases in appears unusually low; this may be related to difficulties in reporting scrapie during the UK's foot-and-mouth disease epidemic. More cases were reported in the first than the second half of each year, but seasonal trends were not significant (t13 = ., P = .).Most cases were initially in animals of the VRQ/VRQ genotype, which peaked in , after which cases became more prevalent in animals of the VRQ/ARQ genotype, which peaked in (Figure ). Three cases also occurred in animals of the less-susceptible VRQ/ARR genotype during the decline phase of the epidemic (Figure ). Considering only the time from when the flock was blood sampled, the proportion of animals which subsequently died from scrapie were: .% (/) in VRQ/ARR; .% (/) in VRQ/ARQ; and .% (/) in VRQ/VRQ.The time-courseAs predicted, the genotype of cases changed during the time-course of the outbreak (Figure 2A). Earlier in the epidemic cases were more likely to be of the VRQ/VRQ than the VRQ/ARQ genotype (AOR = .; % CI: .–.). Exploratory analysis suggested that animals of the VRQ/ARR genotype were likely to exhibit clinical signs later in the epidemic than either VRQ/ARQ or VRQ/VRQ animals (Figure 2A), but their small sample size precluded more detailed analyses.As observed by the farmer, cases which occurred later in the epidemic tended to be older than those seen earlier (Figure 2A; F1, = ., P < .). Cases of the VRQ/VRQ genotype developed clinical signs at an earlier age than VRQ/ARQ animals (Figure 2B; log-rank test: χ2 = ., df = , P < .), but age within each genotype did not change significantly during the outbreak (P = .). This implies that the increase in age-at-onset was due to changes in the PrP genotype of cases rather than changes in age within each genotype.Clinical signsAnalysis was undertaken for those clinical signs with sufficient observations (n>): ataxia, change of behaviour, fleece loss, loss of condition, pruritus and trembling. As the epidemic progressed, animals were more likely to be reported showing a change in behaviour, independent of PrP genotype (AOR = .; % CI: .–.). VRQ/VRQ animals were less likely to show loss of condition (AOR = .; % CI: .–.) or fleece loss (AOR = .; % CI: .–.) as the outbreak progressed. This pattern was not seen in VRQ/ARQ cases. Throughout the outbreak, older cases of any genotype were more likely to show ataxia (AOR = .; % CI: .–.) and trembling (AOR = .; % CI: .–.) than younger ones. Finally, animals exhibiting fleece loss were more likely to be of the VRQ/VRQ rather than the VRQ/ARQ genotype (AOR = .; % CI: .–.).DiscussionAlthough there have been several studies describing within-flock outbreaks of scrapie [-] the results of this paper represent the first analysis addressing confounding between the effects of time since the beginning of the epidemic, age at onset and PrP genotype. The major limitations of this study are that we do not know when the flock first became infected with scrapie; and genotypes were unavailable for cases, including the first during the epidemic.An interesting feature of this outbreak is that cases occurred in the VRQ/VRQ, VRQ/ARQ and VRQ/ARR genotypes only, despite the high frequency of sheep of the ARQ/ARQ and VRQ/AHQ genotypes in the flock (Table ). These findings confirm the resistance of the VRQ/AHQ genotype to classical scrapie [,] and the resistance of the ARQ/ARQ genotype to certain scrapie strains in some sheep breeds [,].A key finding of this study is that the cases which occurred later in the outbreak tended to be older. Our analysis has shown that this is not related to changes in the age-at-onset within each genotype during the outbreak. Rather, it is an effect of the PrP genotype of cases, which influences the age at which animals succumb to disease. Our results differ from an earlier study, which reported a decline in the age-at-onset for four outbreaks [], though the decline was significant (P < .) in only one outbreak in a flock of Suffolk sheep []. This Suffolk flock was bred to maximise the incidence of disease; hence, susceptible animals were constantly being introduced to the flock and exposed to an increasing load of infection. By contrast, most, though not all, susceptible animals in our study flock succumbed to scrapie, but their susceptible alleles were not replaced. The resulting pattern of the outbreak means that animals of the same genotype were, on average, exposed to similar loads of infection.Animals of the VRQ/VRQ genotype had a younger age-at-onset than VRQ/ARQ animals (Figure ), in common with other studies [,,]. The age at onset of clinical signs depends on a number of factors, including the infectious load, age-dependent exposure, age-dependent susceptibility and incubation period. Modelling analysis of an outbreak in a Cheviot flock suggested that the incubation period for VRQ/VRQ animals was shorter than that for VRQ/ARQ and that there was evidence for age-dependent susceptibility []. This suggests that the differences in the age-at-onset are likely to reflect differences in incubation period rather than age-dependent exposure. More detailed analyses are needed, however, to tease apart the interactions of various age and genotype effects in the age-at-onset of clinical signs.Our study is the first to report changes in clinical signs during a scrapie outbreak. Animals were more likely to be reported showing a change of behaviour as the epidemic progressed, irrespective of PrP genotype. One hypothesis would be that at the start of the epidemic the farmer required physical signs to identify scrapie, but as the epidemic progressed he became better able to detect disease on the basis of more subtle behavioural changes. Indeed, behavioural changes are reported prior to the onset of other clinical signs in both goats [] and mice [].Temporal changes in clinical signs also differed between PrP genotypes. Early in the epidemic, VRQ/VRQ animals were more likely to show loss of condition and fleece loss/change, while those reported later in the epidemic were less likely to do so. This could be related to the aforementioned greater reliance of the farmer on physical (as opposed to behavioural) signs earlier in the outbreak; however, this seems unlikely, as the pattern did not occur in animals of the VRQ/ARQ genotype. Alternatively the effect may be real, or it may reflect variation in the clinical signs recorded by the seven reporting officers. Exploratory analysis provided some support for this latter hypothesis, but there were too few cases to be able to test this rigorously.ConclusionIn this paper we have described one of the largest scrapie outbreaks in the UK. As the cases were almost exclusively in two genotypes, this has facilitated the investigation of the effect of PrP genotype on other epidemiological parameters. Although age-at-onset and clinical signs changed over time, our analysis indicated that the effect was largely, but not exclusively, driven by the time course in the PrP genotype of cases.MethodsFlock detailsA flock comprising sheep predominantly of the Welsh Mountain breed and numbering about breeding animals was recruited into the aforementioned IAH scrapie field study. The first cases of scrapie within this flock were confirmed in September . As of February , there were confirmed cases of scrapie. The entire breeding flock (n = ) was blood sampled for PrP genotyping in August .Identification of confirmed casesThe data for scrapie cases within the study flock were retrieved from the Scrapie Notification Database (SND) held at the Veterinary Laboratories Agency (VLA). This includes the breed, sex, date of birth, date of death, PrP genotype and clinical signs in animals submitted as suspect scrapie cases. Tissue samples are also obtained from cases, which are subject to routine analysis for evidence of scrapie.Breed of casesThe confirmed cases in the flock were a mixture of pure-bred (n = ) and cross-bred Welsh Mountain (n = ), Welsh Half-bred (n = ) and Beulah Speckled Face (n = ). Because of the small number of cases in most breeds and crossbreeds within the flock, breed was not considered further in the analyses.PrP genotype analysisSusceptibility to scrapie is strongly associated with polymorphisms in the prion protein (PrP) gene [,,], of which five alleles are commonly found (defined by amino acids at codons , and ): VRQ, ARQ, ARH, AHQ and ARR [,]. PrP genotype analysis was undertaken using approximately ml of blood collected into an EDTA-vacutainer from each sheep, as described previously [].Clinical signsThe clinical signs used within analyses were those observed and recorded by the reporting officers who visited the farm when notified by the farmer of a suspect case. Data on clinical signs were available for out of confirmed cases; recorded signs included abnormal head position, ataxia, change of behaviour (or temperament), biting, dribbling (of saliva or cud), fleece loss (or change), loss of condition, nervousness, nibbling reflex, pruritus, teeth grinding, trembling and visual impairment.Statistical methods'Age' in months relates to when clinical signs were reported; 'time' in months was from the first confirmed scrapie case (i.e. the start of the reported epidemic). Analyses were restricted to confirmed clinical cases of scrapie.A binary logistic regression model was used to examine the effects of time and age upon the PrP genotype of cases. Cox proportional hazard models, with age as the survival measure, were used to analyse the effect of time and PrP genotype upon age []. Binary logistic regression models were used to investigate the effect of time, age and PrP genotype on the presence or absence of clinical signs. Statistical significance was determined by a P-value of less than .. Models were derived by step-wise deletion of insignificant main effects, excluding those within significant interaction terms. Within logistic regression models, the adjusted odds ratios (AOR) and % confidence intervals (CI) were calculated for significant independent variables using standard methods.Authors' contributionsKMM carried out the statistical analyses and drafted the manuscript. SG carried out the statistical analyses and critically revised the manuscript. WG and ES undertook the PrP genotyping. MB conceived the study and critically revised the manuscript. All authors read and approved the final manuscript.
|
PMC2547981.txt
|
TITLE: New products AUTHORS: No authors listed ABSTRACT: No Abstract BODY: No Body Content
|
PMC2736169.txt
|
TITLE: TTV and HPV co-infection in cervical smears of patients with cervical lesions AUTHORS: Martina Saláková, Vratislav Němeček, Ruth Tachezy ABSTRACT: BackgroundThe female lower genital tract is a gateway for pathogens entering the host through the mucous membrane. One of the prevalent human viruses is Torque teno virus (TTV). The major reported routes of TTV transmission are fecal-oral and parenteral. Furthermore, other modes of transmission, e.g. sexual contact, are suggested. To investigate the sexual route of TTV transmission, cervical smears of healthy women and those with cervical lesions were screened for the presence of TTV DNA.MethodsTTV DNA was studied in cervical smears of patients with cervical lesions and healthy women. Paired serum samples were available from and women, respectively. All healthy women had normal cytology while patients had histologically confirmed low-grade lesion (LGL) and high-grade lesion (HGL). TTV DNA was detected with primers specific for the non-coding region. In paired cervical smears and serum samples, the phylogenetic group of TTV isolates was determined. The presence of HPV DNA in cervical smears was detected by means of PCR with MY09/ primers.ResultsThe prevalence of TTV DNA in cervical smears of healthy women was .% and was comparable with that in paired serum samples (%). Symptomatic women had significantly higher prevalence of TTV DNA in cervical smears (.%) than healthy controls. The TTV DNA prevalence in patient serum samples was %. The phylogenetic groups of TTV serum isolates were concordant with those of TTV from cervical smears of the same subjects. In cervical smears, a wider variety of TTV isolates was found. The viral loads in cervical smears were to times as high as in sera. The HPV-positive study subjects had significantly higher TTV DNA prevalence than HPV negatives. The prevalence of TTV was not associated with disease severity.ConclusionHigh prevalence of TTV in cervical smears suggests that sexual transmission is another mode of expansion of TTV infection among the population. The higher viral load in cervical smears than in the respective serum samples might indicate active TTV replication in the female genital tract. Nevertheless, cooperation between TTV and HPV needs to be further investigated. BODY: BackgroundThe mucosal surface of the female lower genital tract is the first line of defense against pathogens. The physical barrier created by epithelial cells prevents the penetration of microbes and together with dendritic cells, macrophages, NK cells and chemical elements, they are the main components of the innate immune system. However, many common infectious disease pathogens succeed in entering the host through the mucous membrane after which they spread systematically. One of the most prevalent viral pathogens of cervical epithelial cells is human papillomavirus (HPV). HPV is a very common sexually transmitted infection with prevalence rates ranging from % to % among healthy women around the world []. HPV plays a role in the development of precancerous lesions and squamous cell carcinomas in the anogenital region. However, most individuals eliminate the HPV infection without any evidence of disease and only very few HPV-infected women develop invasive cervical cancer after a long period. The number of sexual partners, smoking, and infections by other sexually transmitted pathogens are factors which increase the risk of cervical cancer in HPV infected individuals [].The human population is frequently infected with Torque teno virus (TTV), a small, non-enveloped DNA virus with a single stranded circular genome, recently classified into a floating family called Anellovirus []. TTV shows a high degree of genetic variability: until now, more than genotypes of TTV divided into major phylogenetic groups have been identified in various clinical specimens from humans, including blood and cervical smears []. To date, no clinical manifestations have been unequivocally associated with TTV, classified into orphan viruses []. Nevertheless, the high genetic diversity of TTV implies the risk that one of TTV genotypes may have pathogenic potential.The major routes of transmission are parenteral and fecal-oral. Since TTV DNA was detected in semen and cervical smears, sexual transmission has also been considered as a possible mode of infection [-]. A study has observed increase in TTV prevalence with increasing number of sexual partners among drug users with liver disease []; on the other hand, several studies have reported that in at-risk groups of prostitutes, the rate of TTV viremia was not significantly different from that in controls [-]. Therefore, the role of sexual transmission remains unclear.To investigate the possibility of sexual transmission of TTV, cervical smears from healthy women and those with cervical lesions were screened for the presence of TTV DNA. For most study subjects, parallel serum samples were also available for comparison and TTV subtyping. Co-infection with TTV and HPV was investigated to support the hypothesis of the sexual route of TTV transmission.MethodsStudy populationCervical smears and sera samples were obtained from the bank of the National Reference Laboratory for Papillomaviruses in Prague. All samples were collected during a prospective study of HPV in patients treated for cervical lesions in – []. Cervical smears from patients (mean age years, age range –) were randomly selected to form two balanced groups of patients, one with low-grade and one with high-grade cervical lesions. The control group of healthy women (mean age years, age range –) was selected as age matched to the patients group. Parallel serum samples were collected from patients and healthy women. For all study subjects, cytology data was available and for patients, histology data at enrollment was obtained. Furthermore, the questionnaire data on medical history, education, socioeconomic status, and number of lifetime sexual partners was accessible. All healthy women had normal cytology with no pathological, clinical or cytological findings while in the patient group, and subjects had histologically confirmed low-grade lesions (LGL) and high-grade lesions (HGL), respectively. No informed consent from patient was needed by course of law in the Czech Republic before . The study in – was approved by the Ethical committee of the institution.DNA extractionCervical brushes were transported in tubes containing phosphate-buffered saline solution (PBS) with mM ethylene diamine tetraacetic acid (EDTA), pH ., and stored at +°C. The tubes were vortexed and cells were collected by centrifugation at , rpm for min. Forty microlitres of each cell pellet were digested in μl of lysis buffer ( mM Tris/HCL, pH ., % Tween-, mM EDTA, pH .) containing μg/ml proteinase K (Fermentas, Hanover, MD) at °C overnight. Proteinase K was inactivated at °C for minutes and the samples were stored at -°C. DNA was extracted from μl of serum by QIAamp Blood kit (QIAGEN Ltd., Crawley, UK) and dissolved in μl of the elution buffer (QIAGEN Ltd., Crawley, UK). The control plasmid with a part of the human beta-globin gene was added to serum sample for the evaluation of DNA quality. Extracted DNA was stored at -°C.The integrity and quality of DNA was assessed by amplification of the human beta-globin gene with primer PC03/PC04. All tested samples were positive for the human beta-globin gene [].HPV DNA detectionThe cervical smears were analyzed for the presence of HPV DNA by means of PCR with MY09/ primers followed by Southern blot hybridization and identification of HPV types on dot blot hybridization with oligonucleotide type specific probes or by sequencing [].TTV DNA detectionTTV DNA detection in cervical smears and serum samples was done by nested PCR directed at the non-coding region of TTV (PCR NCR) with NG133/ and NG132/ [] and with primers specific for the ORF1 N22 region (PCR ORF1) []. The sensitivity of both assays was analyzed by amplification of cloned TTV DNA at , and copies per tube. The PCR NCR system was – times more sensitive, able to detect one copy per reaction (data not shown). Furthermore, ORF system is known to amplify limited number of TTV genotypes while NCR system can detect genotypes from all five so far known major TTV phylogenetic groups. The prevalence with NCR primers in our study was much higher and therefore we used results obtained with this system for the final analysis.Ten microlitres of the PCR product were separated electrophoretically on a % agarose gel (NuSieve :, FMC BioProduct, Rockland, ME).Semiquantitative determination of TTV DNA loadDNA extracted from the sera and cervical smears from patients with cervical abnormalities and healthy women was serially diluted (in -fold steps) in distilled water and the presence of TTV DNA was detected by PCR NCR. The highest dilution (10N) testing positive was used as a relative titer for determining the viral titer per ml of TTV DNA in serum samples and cervical smears.Genotyping determinationIn patients and healthy women positive for TTV DNA in both cervical smears and serum samples, the TTV genotype was determined by a PCR assay based on amplification with TTV group-specific primers derived from the conserved non-coding region [].The statistical analysis was performed using the Fisher exact test. Odds ratios (OR) with % confidence intervals (CI) and two-tailed P values were calculated in × tables using the EPI INFO statistical package ( version) and Graph Pad InStat (version .) (GraphPad Software, San Diego, CA). In all tests, the basic significance level was P = ..ResultsIn total, cervical smears and serum samples were screened by PCR for the presence of TTV. The prevalence rates of TTV in the study groups of healthy women and patients with cervical lesions are shown in Table .Table 1Prevalence of TTV DNA in study groupsMaterialSubjectsCervical smearsSerumOR (%CI), PNTTV positive (%)NTTV positive (%)Patients9571 (.) (). (.–.), P = .004Healthy women5529 (.) ()NSN – number of subjects, NS – not significantThe prevalence rates of TTV DNA were .% in cervical smears of healthy women (/) and % in the parallel serum samples (/). The difference between the TTV DNA prevalence rates from the patient cervical smears and serum samples was statistically significant (.% (/) versus % (/), OR = ., CI .–., P = .). Comparable prevalence rates of TTV DNA were found in the sera from patients and healthy women, while patients showed significantly higher TTV DNA rates in cervical smears than healthy women (OR = ., CI .–., P = .).Overall, .% (/) of cervical smears from patients with cervical abnormalities and .% (/) cervical smears from healthy women were HPV DNA positive. The HPV-positive study subjects showed a significantly higher prevalence of TTV DNA, i.e. .% (/), than the HPV-negatives with .% (/) (OR = ., CI .–., P = .). The patients were divided into two groups according to cervical lesion severity. The TTV prevalence rate was slightly higher in the group with high-grade lesions, i.e. .% (/), than in that with low-grade lesions, i.e. .% (/), but the difference was not statistically significant.The relative viral loads in cervical smears and in the respective serum samples were compared for five patients and five healthy women. All of these subjects had to . times higher TTV DNA titers in cervical smears than in serum samples.TTV genotype was determined in patients and healthy women with both tested specimens positive for TTV. Any of TTV genotypes was recognized in % of patient cervical smears, % of patient serum samples, % of healthy women's cervical smears and % of the healthy women's serum samples. The most prevalent phylogenetic group was TTV (%), followed by TTV (%), the phylogenetic group TTV was found rarely (%) and TTV only in the group of patients (%). Concordant phylogenetic groups of TTV isolates were found in serum samples and the parallel cervical smears from the same subject; however, in cervical smears more diverse TTV genotypes were present (Table ).Table 2TTV phylogenetic group distributionIDTTV phylogenetic groups in cervical smearTTV phylogenetic groups in serum13321,,,,,,,,,,,,,,,,,,,,,,31233TTV phylogenetic group distribution in parallel cervical smears and serum samples in which any TTV genotype was detected, ID is sample number, ID1- are samples from patients with cervical lesions, ID10- are samples from healthy women.The number of sexual partners self-reported by the study subjects did not exceed four. No difference in the prevalence of TTV DNA associated with the number of sexual partners was found; the TTV DNA positivity rates were % (/) in subjects with less than three sexual partners and .% (/) in those with three or four sexual partners.DiscussionTTV infects the majority of the general population worldwide but is not associated with any clinical evidence of disease. The high prevalence of TTV in the general population might be explained by the existence of multiple transmission routes. Several studies have reported the parenteral and fecal-oral routes of TTV transmission and other modes of transmission have also been considered. The presence of TTV DNA in cervical smears and semen suggests that TTV may be spread by sexual contact [].In our study, in agreement with Calcaterra et al., Chan et al., and Fornai et al [-], we observed a high prevalence of TTV in cervical smears. The prevalence of TTV DNA in cervical smears of healthy women was similar to that previously reported for the Czech general population []. In healthy women, no difference was found in the prevalence of TTV between the cervical smears and parallel serum samples. Only one study has so far compared the prevalence of TTV DNA in these two types of specimens. Fornai et al. have reported a higher prevalence of TTV DNA in serum than in cervical smears. Nevertheless, the differences between the study populations can explain this discrepancy. The subjects tested in the Fornai study were HIV-infected individuals who have a higher TTV prevalence and viral load in the serum than the general population [,].The cervical smears from patients with cervical lesions were more frequently infected by TTV in comparison with those from healthy women. The co-presence of HPV and TTV in the patient cervical smears suggests that both viruses have a common transmission route. The co-infection with TTV and other viruses such as HBV, HCV, and HIV l has been reported by others [,,]. The high TTV/HPV co-infection rate, similar to those of HIV, HBV, HCV, supports the hypothesis that these viruses share the same mode of transmission and that sexual transmission is another important mode of TTV infection. HPV is an etiological factor of cervical carcinoma and precancerous cervical lesions. The mode of transmission is mostly sexual. It has been documented that patients with cervical disorders are more likely to start sexual life sooner and to have more sexual partners []. Therefore, it is likely that if TTV is transmitted sexually it will be more frequently present in women infected with HPV. Women with cervical lesions are also more likely to be HPV positive and form a risk group in comparison to healthy women in terms of the mode of transmission. Nevertheless, we did not find the number of sexual partners to be a risk factor for TTV infection, most likely due to the relatively low number of lifetime sexual partners of our study subjects. Many studies have reported that the risk of HPV infection and cervical cancer increases with more than six lifetime partners []. The similarity in the TTV phylogenetic groups in the isolates from cervical smears and serum samples implies the same source of TTV infection. The higher TTV load in cervical smears may indicate that the vagina is the entry site of TTV infection. The higher genotypic variability of TTV in cervical smears may reflect that the vagina is the first place of TTV infection. On the other hand, it might also explain the inability to detect the lower TTV load in the serum.Originally, liver tissue had been thought to be a site of TTV replication since the viral loads in hepatocytes were to times higher than those in the respective plasma samples. This hypothesis was confirmed by the detection of a double-stranded variant of TTV DNA in the liver and hepatocytes by in situ hybridization. The viral titer was to . times higher in saliva than in serum. The assumption that TTV might replicate in oropharyngeal tissues and/or salivary glands was supported by the detection of TTV in the gingival tissue []. In our study, we found to . higher viral loads in cervical smears than in the parallel serum samples. This observation suggests that TTV might also replicate in cells of the cervicovaginal area.The reported higher viral loads of TTV in HIV infected patients might be associated with a poorer prognosis in AIDS. A higher viral load can be explained by the immune stimulation due to HIV infection. CIN biopsies in HIV-positive women have shown higher lymphocyte and macrophage counts in the squamous epithelium of the cervix and an increased intralesional concentration of proinflammatory cytokines modulating HIV replication, with consequently increased HIV genital shedding []. Similarly, higher TTV prevalence in HPV-positive patients can be attributed to the stimulation of the immune system by HPV infection and additional increase in TTV load. Nevertheless, more data is needed to better understand the phenomena. Additional risk to patients with cervical lesions from co-infection with TTV was also investigated. As in our study, TTV prevalence did not differ between patients with low-grade and high-grade cervical lesions, co-infection with TTV and HPV seemed unlikely to promote the progression of precancerous cervical lesions.ConclusionTo conclude, high TTV prevalence in cervical smears suggests that sexual transmission is another mode of expansion of TTV infection among the population. The higher viral loads in cervical smears than in the respective serum samples might indicate active TTV replication in the female genital tract or might be the result of plasma secretion. The TTV shedding is possibly mediated by the immune upregulation in the cervix and vagina due to HPV infection. Nevertheless, natural cooperation between TTV and HPV needs to be further investigated.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMS projected the study, carried out the experimental work, did all statistical analyses and evaluation of the results. MS wrote the first draft of the paper and other coauthors contributed to the final draft. RT and VN participated in the study design and data interpretation. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/-///prepub
|
PMC2564802.txt
|
TITLE: Efficacy and safety of abatacept or infliximab vs placebo in ATTEST: a phase III, multi-centre, randomised, double-blind, placebo-controlled study in patients with rheumatoid arthritis and an inadequate response to methotrexate AUTHORS: M Schiff, M Keiserman, C Codding, S Songcharoen, A Berman, S Nayiager, C Saldate, T Li, R Aranda, J-C Becker, C Lin, P L N Cornet, M Dougados ABSTRACT: Objectives:This double-blind trial evaluated the efficacy and safety of abatacept or infliximab vs placebo. The primary objective of this study was to evaluate the mean change from baseline in Disease Activity Score (based on erythrocyte sedimentation rates; DAS28 (ESR)) for the abatacept vs placebo groups at day .Methods:Patients with rheumatoid arthritis (RA) and an inadequate response to methotrexate (MTX) were randomised :: to abatacept (∼ mg/kg every weeks, n = ), infliximab ( mg/kg every weeks, n = ), or placebo (every weeks, n = ) and background MTX. Safety and efficacy were assessed throughout the study.Results:Similar patient demographics and clinical characteristics were present at baseline between groups, with mean scores of ∼. for HAQ-DI and . for DAS28 (ESR). At months, mean changes in DAS28 (ESR) were significantly greater for abatacept vs placebo (–. vs –., p<.) and infliximab vs placebo (–. vs –., p<.). For abatacept vs infliximab treatment at day , reductions in the DAS28 (ESR) were –. vs –.. At day , the following response rates were observed for abatacept and infliximab, respectively: American College of Rheumatology (ACR) , . and .%; ACR , . and .%; ACR , . and .%; low disease activity score (LDAS), . and .%; DAS28-defined remission, . and .%; good European League Against Rheumatism (EULAR) responses, . and .%; and Health Assessment Questionnaire Disability Index (HAQ-DI), . and .%. Mean changes in physical component summary (PCS) were . and ., and mental component summary (MCS) were . and ., for abatacept and infliximab, respectively. Over year, adverse events (AEs) (. vs .%), serious AEs (SAEs) (. vs .%), serious infections (. vs .%) and discontinuations due to AEs (. vs .%) and SAEs (. vs .%) were lower with abatacept than infliximab.Conclusions:In this study, abatacept and infliximab ( mg/kg every weeks) demonstrated similar efficacy. Overall, abatacept had a relatively more acceptable safety and tolerability profile, with fewer SAEs, serious infections, acute infusional events and discontinuations due to AEs than the infliximab group. Trial registration number:NCT00095147. BODY: Abatacept is a selective T-cell co-stimulation modulator, that modulates the CD80/CD86:CD28 co-stimulatory signal required for full T-cell activation. The mechanism of action of abatacept is fundamentally different to that of other biological disease-modifying antirheumatic drugs (DMARDs) for the treatment of rheumatoid arthritis (RA). The efficacy of abatacept has previously been demonstrated in patients with RA and an inadequate response to methotrexate (MTX) and anti-tumour necrosis factor (TNF) agents, respectively.The ATTEST (for “Abatacept or infliximab vs placebo, a Trial for Tolerability, Efficacy and Safety in Treating rheumatoid arthritis”) trial was designed to obtain data on the magnitude of the treatment effect in RA of abatacept or infliximab (an established inhibitor of TNF for RA) vs placebo, and to obtain relative efficacy and safety data on these two biological treatments in a single study. The study utilised a double-blind, randomised, placebo-controlled design for the first months to validate efficacy responses, and the study duration allowed for the opportunity to directly compare the safety profile of the active biologic treatment groups over year.METHODSPatientsEligible patients met the American College of Rheumatology (ACR) criteria for RA, were at least years of age, had RA for at least year, and had an inadequate response to MTX, as demonstrated by ongoing active disease (at randomisation ⩾ swollen joints, ⩾ tender joints, and C-reactive protein (CRP) levels ⩾ mg/dl using a high sensitivity assay (upper limit of the normal range, .)). All patients had received MTX ⩾ mg/week for ⩾ months prior to randomisation (stable for at least days) and washed out all DMARDs (⩾ days prior) except for MTX. No prior experience of abatacept or anti-TNF therapy was permitted. All patients were screened for tuberculosis (TB) by purified protein derivative (PPD) testing and chest x ray. The protocol used for TB screening was the same as that employed in the “Anti-TNF Trial in rheumatoid arthritis with Concomitant Therapy” (ATTRACT) trial.5Concomitant medications were permitted between days –: oral corticosteroids (⩽ mg of prednisone or equivalent daily (stable for ⩾ out of days prior to randomisation)), and/or stable non-steroidal anti-inflammatory drugs (NSAIDs) including acetyl salicylic acid, and analgesics not containing aspirin or NSAIDs. No MTX dose adjustments were permitted except in the occurrence of adverse events (AEs). Between days –, dose modification was permitted for MTX (⩽ mg weekly) and oral corticosteroids (⩽ mg prednisone or equivalent daily); hydroxychloroquine, sulfasalazine, gold, or azathioprine were also permitted. Premedication prior to infusions of study drug was left at the discretion of the investigator (not required by protocol).Study designATTEST was a randomised, double-blind, double-dummy, placebo- and active (infliximab)-controlled, -month global trial. Adult patients with active RA and an inadequate response to MTX were randomised by centre in a :: ratio to months of abatacept (approximating mg/kg), infliximab ( mg/kg), or placebo treatment by intravenous (IV) infusion, on a background of MTX. Assessors, physicians and patients were blinded to the treatment group assignment for year.The study was approved by the institutional review boards and independent ethics committees at participating sites, and was carried out in accordance with the ethical principles of the Declaration of Helsinki. All patients provided written informed consent before randomisation.Treatment with placebo was limited to days – to provide internal validity to the trial design and the clinical response rates of the two active treatment groups. On day , placebo-treated patients were reallocated to abatacept (with blinding maintained). Patients initially randomised to abatacept or infliximab continued their treatment. Clinical efficacy and safety assessments are presented for the abatacept, placebo and infliximab groups up to day , and for the abatacept and infliximab groups up to day (excluding patients in the placebo group who were reallocated to abatacept at day ).Study drugsAbatacept was dosed according to weight: patients weighing less than kg, – kg, or more than kg received mg, mg, or mg of abatacept, respectively. Infliximab was dosed at mg/kg for all patients. Abatacept was administered by IV infusion on days , and , and every days thereafter, up to and including day (with normal saline received on day ). Infliximab was administered on days , , and , and every days thereafter (normal saline was received at the remaining visit days). Two IV bags were infused simultaneously to ensure blinding to treatment group assignment, one over min (abatacept or placebo) and one over h (infliximab or placebo).ObjectivesThe primary endpoint was to evaluate a reduction in disease activity, measured by Disease Activity Score (based on erythrocyte sedimentation rate levels; DAS28 (ESR)) with abatacept vs placebo at months. Secondary endpoints included mean reduction in DAS28 (ESR) with infliximab vs placebo at months. Additional secondary endpoints at months and year included: mean reduction in DAS28 (ESR) with abatacept vs infliximab; DAS28 (ESR) European League Against Rheumatism (EULAR) responses; low disease activity score (LDAS; DAS28 (ESR) ⩽.); DAS28 (ESR)-defined remission (DAS28 (ESR) <.); ACR , and responses; Health Assessment Questionnaire Disability Index (HAQ-DI) response rates (⩾. improvement from baseline); and mean changes in the physical and mental component summary (PCS and MCS, respectively) scores, and eight subscales of the Short Form- (SF-). Tertiary endpoints included comparative safety at year between the abatacept and infliximab groups.Clinical assessmentsDisease activity was assessed using the validated DAS28 (ESR). The EULAR criteria were used to assess good responses (endpoint DAS28 (ESR) of ⩽. and an improvement from baseline of ⩾.). Signs and symptoms of RA were evaluated using ACR , and response rates based on the ACR criteria.7Physical function was assessed using the HAQ-DI. Health-related quality of life (HRQoL) was assessed using the SF-.10SafetyAll patients who received at least one dose of study drug were evaluated for safety, including all reported AEs, serious AEs (SAEs), and clinically significant changes in vital signs, physical exams, and clinical laboratory test abnormalities.Sample size calculationThe sample size and power were calculated to detect a treatment difference in the primary analysis of a mean change from baseline in DAS28 (ESR) for the abatacept vs placebo groups at day . Prospectively, this study was not powered with a superiority or non-inferiority design to compare the two active arms.Statistical analysesAll patients who received at least one dose of study medication were assessed for efficacy and safety (intent-to-treat population). At day , the abatacept or infliximab groups were compared with the placebo group by analyses of covariance (ANCOVA) for mean changes from baseline in DAS28 (ESR) and in the SF- (PCS and MCS). The model included the change as the dependent variable, with treatment group as a main effect and the baseline score as an additional covariate. The proportion of patients with ACR , and responses, LDAS, DAS28-defined remission, a good EULAR response and a clinical meaningful HAQ-DI response was calculated. The χ2 test was performed to evaluate the differences (and % CIs) between the abatacept or infliximab groups and placebo. At day , the reference group was changed to infliximab.Patients who discontinued the study prematurely were considered as non-responders subsequent to the time of discontinuation for ACR , and responses, good EULAR responses and clinically meaningful HAQ-DI responses. For all continuous measurements (mean changes in DAS28, SF- and the HAQ-DI score), LDAS and DAS28-defined remission the last observations prior to the discontinuation were carried forward (LOCF). To access the effect of antirheumatic medications on the abatacept and infliximab treatment groups, a predefined sensitivity analysis was conducted on the data with the last DAS28 (ESR) score just prior to the initiation of the additional DMARD, or any increase in MTX or corticosteroid use during treatment days – carried forward.Safety was assessed at each visit. Summary statistics were tabulated by treatment group at days and .RESULTSBaseline demographics and clinical characteristicsA total of patients were randomised and treated with abatacept (n = ), placebo (n = ) or infliximab (n = ) (fig ). Baseline demographics and clinical characteristics were similar between groups (table ). At randomisation, patients had active disease despite background MTX (mean DAS28 (ESR) of .–., tender joint counts .–., swollen joint counts .–. and mean HAQ-DI of .–.). The mean dose of MTX was .–. mg and mean treatment duration was .–. months.Figure 1Patient disposition over year. The ATTEST trial was a -month global trial conducted at sites in the US ( sites), Europe ( sites ( in Poland, in Spain, in Sweden, in Russia, in Denmark and in Switzerland)), Canada ( sites), Australia ( sites), Mexico ( sites), Argentina ( sites), Brazil ( sites), Peru ( sites) and South Africa ( sites). Patients were randomised in a :: ratio to months of abatacept (approximating mg/kg), infliximab ( mg/kg), or placebo treatment. During days –, efficacy and safety data are not presented for the placebo group following reallocation to abatacept.Table 1Baseline demographics and clinical characteristicsDemographic/characteristicAbatacept + MTX (n = )Placebo + MTX (n = )Infliximab + MTX (n = )Age, years (SD). (.). (.). (.)Gender, % female83...4Race, % Caucasian80...6Geographic origin: North America, n (%) (.) (.) (.) South America, n (%) (.) (.) (.) Europe, n (%) (.) (.) (.) Rest of the world, n (%) (.) (.) (.)Disease duration, years (SD). (.). (.). (.)Tender joints, n (SD). (.). (.). (.)Swollen joints, n (SD). (.). (.). (.)Erythrocyte sedimentation rate, mm/h (SD). (.). (.). (.)C-reactive protein levels, mg/dl (SD). (.). (.). (.)DAS28 (ESR), n (SD). (.). (.). (.)HAQ-DI, – (SD). (.). (.). (.)Rheumatoid factor positive, n (%) (.) (.) (.)Concomitant medicationsTotal patients on concomitant medications, n (%) () () ()MTX, n (%) () () (.) Dose, mg/week (SD). (.). (.). (.) Duration, months (SD). (.). (.). (.)Corticosteroids, n (%) (.) (.) (.)NSAIDs, n (%) (.) (.) (.)MTX, methotrexate; DAS28 (ESR), Disease Activity Score (based on erythrocyte sedimentation rate levels); HAQ-DI, Health Assessment Questionnaire Disability Index; NSAID, non-steroidal anti-inflammatory drug.Use of additional medicationsConcomitant medications and NSAIDs were used by a similar proportion of patients across treatment groups at randomisation (table ). Between days –, when the protocol permitted adjustments to background medications, .% and .% of abatacept and infliximab-treated patients, respectively, added a DMARD, or increased their dose of MTX/corticosteroids from baseline.DiscontinuationsDuring the first months, discontinuations occurred in ., . and .% of the abatacept, placebo and infliximab groups, respectively. Between days –, . and .%, of the abatacept and infliximab groups, respectively, discontinued. Discontinuation due to AEs and SAEs were highest in the infliximab group in both periods (fig ). Similar numbers of patients from the abatacept and infliximab groups completed year of treatment (fig ).Clinical efficacyDAS28At day , the reduction in DAS28 (ESR), was significantly greater with abatacept vs placebo (–. vs –., p<.; fig 2A). A greater proportion of patients also experienced a good EULAR response (. vs .%), LDAS (. vs .%) and were in DAS28 (ESR)-defined remission (. vs .%), for abatacept vs placebo, respectively (fig 2B). Reductions in DAS28 (ESR) were also significantly greater in the infliximab vs placebo groups at day (–. vs –., p<.; fig 2A), with a higher proportion of patients experiencing a good EULAR response (. vs .%), LDAS (. vs .%) and were in DAS28 (ESR)-defined remission (. vs .%), respectively (fig 2B). At day , the relative efficacy of abatacept and infliximab as assessed by the DAS28 (ESR) was similar.Figure 2Disease Activity Score (DAS28) based on erythrocyte sedimentation rates (ESR). A. DAS28 (ESR) mean changes from baseline at days and . Error bars represent standard error of the mean. B. European League Against Rheumatism (EULAR) good responses, low DAS (LDAS; DAS28 ⩽.) and DAS28 (ESR)-defined remission at day and at day . Data are presented for the intent-to-treat population with a last-observation carried forward analysis for mean changes in DAS28 (ESR), LDAS and DAS28 (ESR)-defined remission. Good EULAR responses were presented for the intent-to-treat population with patients who discontinued the study prematurely considered as non-responders subsequent to the time of discontinuation. Error bars show standard error of the mean. *Adjustment based on covariance with treatment as factor and baseline as covariant.At day , a greater reduction in DAS28 (ESR) was observed with abatacept than with infliximab (fig 2A; –. vs –.; estimate of difference (% CI) = –. (–., –.)). Also, the proportion of patients achieving a good EULAR response (. vs .%, estimate of difference (% CI) = .% (., .)), LDAS (. vs .%, estimate of difference (% CI) = . (., .)), and DAS28 (ESR)-defined remission (. vs .%, estimate of difference (% CI) = . (–., .)) were higher with abatacept compared with infliximab (fig 2B).When disease activity was assessed using a sensitivity analysis (LOCF prior to increase in antirheumatic medications was performed), consistent improvements were demonstrated with abatacept: DAS28 (ESR) mean changes were significantly greater with abatacept vs infliximab at day (–. vs –., respectively (difference from infliximab of –., % CI of –., –.)).American College of Rheumatology response ratesAt day , ACR , and responses were significantly greater with abatacept vs placebo (ACR : . vs .%, p<.; ACR : . vs .%, p<.; and ACR : . vs .%, p = .). ACR , and responses were also significantly higher in the infliximab group vs placebo (ACR : . vs .%, p = .; ACR : . vs .%, p = .; and ACR : . vs .%, p = .). The onset of response, as assessed by ACR response rates, was generally more rapid for infliximab compared with abatacept (fig ), however, by day , responses were similar (fig ). Abatacept and infliximab demonstrated similar responses at day . During the second months of the trial, the responses associated with abatacept were maintained, while those observed with infliximab were not. At day , ACR responses were higher with abatacept than with infliximab (ACR : . vs .%, difference of ., % CI = ., .). In addition, the percentages of ACR and responders were numerically higher with abatacept vs infliximab treatment (with overlapping % CIs for the estimate of difference for ACR : . vs .%, estimate of difference (% CI) = . (–., .); ACR : . vs .%, estimate of difference (% CI) = . (–., .), respectively).Figure 3American College of Rheumatology (ACR) responses over year. Proportion of patients with ACR , and responses was assessed on each visit day. Data are presented for the intent-to-treat population with a last-observation carried forward analysis. *Infliximab was administered on days , , , and then every days thereafter. Abatacept dosing occured at each visit day presentation following the assesment of efficacy.Physical functionAt months, significantly more patients in the abatacept group than in the placebo group demonstrated a clinically meaningful improvement in physical function (HAQ-DI responses: . vs .%, respectively, p = .). Similarly, significantly more patients in the infliximab group vs the placebo group achieved a HAQ-DI response (. vs .%, p = .). At day , HAQ-DI responses were maintained in the abatacept and infliximab groups (. and .%, respectively, estimate of difference (% CI) = . (–., .)).Health-related quality of lifePatients in the abatacept group experienced statistically significantly greater improvements from baseline in the PCS (p<.) and MCS (p = .) of the SF-, and in each of the eight individual subscales compared with placebo, following months of treatment (fig 4A). Patients in the infliximab group also experienced significantly greater improvements from baseline in the PCS (p = .) and MCS (p = .), and all eight subscales of the SF- at day compared with patients in the placebo group. At day , greater improvements from baseline in the PCS were observed with abatacept vs infliximab (difference of ., % CI = ., .). Improvements in the MCS (difference of ., % CI = –., .) and in all eight subscales were also numerically higher with abatacept vs infliximab (fig 4B).Figure 4Health-related quality of life. A. Mean change in physical and mental component summary (PCS and MCS, respectively) scores and the individual subscales of the Short Form- (SF-) from day to day . B. Mean change in PCS and MCS scores and the individual subscales of the SF- from day to day ; Data are presented for the intent-to-to treat population with a last observation carried forward analysis.SafetyA summary of safety for all patients who received at least one dose of study medication is presented in table (excluding the original placebo group between days –).Table 2Summary of safety to day and day 365Days –197Days –365Abatacept + MTX (n = )Placebo + MTX (n = )Infliximab + MTX (n = )Abatacept + MTX (n = )Infliximab + MTX (n = )n (%)*n (%)*n (%)*n (%)*n (%)*Deaths1 (.) (.) (.) (.)SAEs8 (.) (.) (.) (.) (.)Related SAEs3 (.) (.) (.) (.) (.)Discontinuations due to SAEs2 (.) (.) (.) (.)AEs129 (.) (.) (.) (.) (.)Related AEs64 (.) (.) (.) (.) (.)Discontinuations due to AEs3 (.) (.) (.) (.) (.)Serious infections2 (.) (.) (.) (.) (.)Autoimmune symptoms and disorders1 (.) (.) (.) (.) (.)Malignant neoplasms1 (.) (.) (.) (.) (.)*More than one AE or SAE could be reported in each patient; percentages represent the proportion of patients who reported ⩾ event.AE, adverse event; MTX, methotrexate; SAE, serious adverse event.Summary of safety for days –197During days –, AEs were reported by a similar proportion of patients in the abatacept (.%), placebo (.%) and infliximab (.%) groups. Two deaths were reported: one patient in the abatacept group due to a cerebrovascular accident, and one patient in the infliximab group due to fibrosarcoma. Overall, the frequency of AEs, related AEs and discontinuations due to AEs or SAEs was similar for the abatacept and placebo groups. Between days –, SAEs were lower with abatacept vs placebo (. vs .%), the difference was largely attributed to a higher frequency of gastrointestinal disorders, bacterial infections and musculoskeletal disorders in the placebo group. For the infliximab vs placebo groups during the same period, the frequency of AEs (. vs .%), related AEs (. vs .%) and SAEs (. vs .%) were similar; however, a higher proportion of patients in the infliximab group compared with the placebo group reported related SAEs (. vs .%), discontinued due to AEs (. vs .%), and discontinued due to SAEs (. vs %). The higher frequency of SAEs in the infliximab vs placebo groups was largely due to an increase in serious infections (. vs .%, respectively).The most frequently reported AEs were primarily infections and infestations, and most were of mild to moderate intensity. Acute infusional AEs (within h of the start of dosing) were reported in ., . and .% of the abatacept, placebo and infliximab groups, respectively (table ). Infections and infestations were reported in ., . and .% of the abatacept, placebo and infliximab groups, respectively.Table 3Frequently occurring acute infusional events (⩾.% of patients in any group) to day and day 365Acute infusional eventsDays –197Days –365Abatacept + MTX (n = )Placebo + MTX (n = )Infliximab + MTX (n = )Abatacept + MTX (n = )Infliximab + MTX (n = )n (%)n (%)n (%)n (%)n (%)Total patients with AEs8 (.) (.) (.) (.) (.)Hypotension007 (.) (.)Headache2 (.) (.) (.) (.) (.)Nausea2 (.) (.) (.) (.) (.)Flushing1 (.) (.) (.) (.)Dyspnea004 (.) (.)Urticaria01 (.) (.) (.)Pruritus002 (.) (.)Dizziness002 (.) (.) (.)AE, adverse event; MTX, methotrexate.Serious infections were lower in the abatacept group (.%) than in the placebo (.%) and infliximab (.%) groups at day (table ), with patients originating from Latin America (abatacept, .%; placebo, .%; infliximab, .%) and Europe (abatacept, .%; placebo, .%; infliximab, .%). Between days –, two opportunistic infections occurred (a pseudomonal lung infection and a Pneumocysitis jiroveci pneumonia) in the infliximab group.Table 4Serious infectious events to day and day 365Days –197Days –365Abatacept + MTX (n = )Placebo + MTX (n = )Inflixima + MTX (n = )Abatacept + MTX (n = )Infliximab + MTX (n = )n (%)n (%)n (%)n (%)n (%)Total patients with SAEs8 (.) (.) (.) (.) (.)Infections and infestations2 (.) (.) (.) (.) (.)Pneumonia2 (.) (.) (.) (.)Sinusitis1 (.) (.)0Postoperative wound infection01 (.) (.)Soft tissue abscess01 (.)000Infective bursitis01 (.)000Bronchitis001 (.) (.)Cellulitis001 (.) (.)Gastroenteritis001 (.) (.)Herpes zoster001 (.) (.)Lung infection pseudomonal001 (.) (.)Pneumocystis jiroveci pneumonia001 (.) (.)Infection skin ulcer0001 (.)0Encephalitis herpetic00001 (.)Erysipelas00001 (.)Lobar pneumonia00001 (.)Peritoneal tuberculosis00001 (.)Pulmonary tuberculosis00001 (.)Septic shock00001 (.)MTX, methotrexate; SAE, serious adverse event.Autoimmune symptoms or disorders were uncommon (<%), occurring at a similar frequency across groups (table ). Malignancies were reported for four patients, including one abatacept-treated patient (bladder cancer), one placebo-treated patient (non-melanomatous skin cancer) and two infliximab-treated patients (malignant anorectal neoplasm and fibrosarcoma in one patient for each).Summary of safety for days – for abatacept and infliximab groupsDuring the entire -month, double-blind period (days –), SAEs, related SAEs, and discontinuations due to SAEs were lower with abatacept than infliximab (SAEs: . vs .%; related SAEs: . vs .%; and discontinuations due to SAEs: . vs .%, respectively (table )). Overall, AEs, related AEs and discontinuations due to AEs were also lower with abatacept than with infliximab (AEs: . vs .%; related AEs: . vs .%: and discontinuations due to AEs: . vs .%, respectively). An additional infliximab-treated patient with peritoneal TB died during the second months of the trial due to septic shock following surgery. One patient who was randomised to the placebo group died while receiving abatacept between days – from pneumonia and sepsis. The investigator assessment deemed the death possibly related to study treatment.Acute infusional events (. vs .%; table ) were lower with abatacept vs infliximab, respectively.Up to day , infections and infestations were reported in . and .% of the abatacept and infliximab groups, respectively. Serious infections were reported in .% of abatacept-treated patients and .% of infliximab-treated patients (table ). A total of five serious opportunistic infections were reported, all occurring in the infliximab group in one patient for each (encephalitis herpetic, lung infection pseudomonal, peritoneal TB, P jiroveci pneumonia and pulmonary TB). Both cases of TB were in patients who were PPD test negative and chest x ray negative at study entry. Between days –, autoimmune symptoms or disorders were uncommon (<%) and occurred at a similar frequency in the abatacept and infliximab groups (table ). No additional malignant neoplasms were reported in either group between days –.DISCUSSIONThe multi-centre, double-blind, placebo-controlled ATTEST trial, examining the relative efficacy and safety of abatacept or infliximab vs placebo, confirms that both biologics are effective for the treatment of RA. Over the -month study, a relative difference in safety was observed, with fewer SAEs, serious infections, acute infusional events and discontinuations due to AEs in the abatacept group than the infliximab group.Following months of treatment, abatacept and infliximab on a background of MTX significantly reduced the signs and symptoms of disease (DAS28 (ESR); ACR , and ), and improved physical function (HAQ-DI) and quality of life (SF-) compared with placebo, in patients with an inadequate response to MTX. The relative efficacy of abatacept and infliximab at day was similar, as the % CIs for the treatment difference for DAS28 (ESR) scores, ACR response rates, HAQ-DI responses and HRQoL improvements overlapped. However, by day , the % CIs for the treatment difference between abatacept and infliximab for the reduction in DAS28 (ESR), good EULAR responses, LDAS, the ACR response rate and the HRQoL physical summary measure did not include zero, suggesting a difference favouring abatacept for these specific endpoints.While the onset of response (as assessed by ACR responses), initially appeared more rapid with infliximab, similar response rates were noted with abatacept and infliximab by day . Between months and year, abatacept responses were maintained, while those with infliximab were not. The sustained efficacy observed at year compared with months in abatacept-treated patients is consistent with results from other abatacept trials. Similarly, the finding that patients treated with infliximab did not sustain response rates over year is also consistent with previous trials. When the impact of adding additional therapies was assessed using a sensitivity analysis according to the last score prior to addition, reductions in disease activity tended to be consistently higher with abatacept than with infliximab.Overall, abatacept had a relatively more acceptable safety and tolerability profile than infliximab. Over year, fewer SAEs, AEs, infections, and discontinuations due to AEs or SAEs were observed with abatacept relative to infliximab. Generally, opportunistic infections appeared to be relatively uncommon in this study and were mainly observed in the infliximab group, including two events of TB, an event of P jiroveci pneumonia, an event of pseudomonal lung infection and an event of herpes encephalitis. No opportunistic infections were reported in the abatacept group. Autoimmune symptoms or disorders were uncommon (<%) with abatacept and infliximab. All of the malignancies were reported during the first months, including one in the abatacept group, one in the placebo group and two in the infliximab group.The data presented here should be interpreted within the context of several limitations. Although collectively the data support a relatively more acceptable risk-to-benefit profile of abatacept compared with infliximab, the design of this study utilised the infliximab dose of mg/kg, without dose adjustment. At the time of the study, the recommended dose of infliximab (approved labelled starting dose for RA) was mg/kg; however, today regulatory agencies recognise the use of higher doses of infliximab, and physicians use them in a proportion of patients to maintain a durable response. In fact, the dose may be increased in up to % of patients. Although higher doses have been associated with an increased risk of AEs in clinical trials. Finally, this study was not designed to evaluate the effect of abatacept or infliximab vs placebo on radiographic progression.In conclusion, these two biologic therapies with two distinct mechanisms of action have different safety and efficacy profiles; however, abatacept and infliximab both offer clinical improvements to patients with an inadequate response to MTX. Over year, abatacept exhibited a durable response, and had a relatively more acceptable safety and tolerability profile than infliximab, as evidenced by fewer SAEs, serious infections, acute infusional events and discontinuations due to AEs.
|
PMC2746549.txt
|
TITLE: Computer-based quality of life questionnaires may contribute to doctor–patient interactions in oncology AUTHORS: G Velikova, J M Brown, A B Smith, P J Selby ABSTRACT: It is well recognized that oncologists should consider patients' quality of life and functioning when planning and delivering anticancer treatment, but a comprehensive assessment of how a patient feels requires a thorough inquiry. A standardized measurement of patients' quality of life may support clinicians in identifying important problems for discussion during the limited time of the medical consultations. The aim of this study was to assess the feasibility of computer-administered individual quality of life measurements in oncology clinics with immediate feedback of results to clinicians and to examine the impact of the information on consultations. The study employed a prospective non-randomized design with pre-test post-test within subjects comparisons and involved three medical oncologists and cancer patients receiving chemotherapy. The intervention consisted of completion of quality of life questionnaires before the consultations and informing clinicians of the results. The main outcome measures were patients' perceptions of the content of baseline and intervention consultations and satisfaction with communication. A qualitative analysis of clinicians' interviews was performed. When clinicians had the quality of life results they enquired more often about daily activities (Z=−., P=.), emotional problems (Z=−., P=.) and work related issues (Z=−., P=.). There was an increase in the number of issues discussed during the intervention consultation (Z=−., P=.). Patients were highly satisfied with both consultations. The computer measurement was well accepted by patients who felt that the questionnaires were a useful tool to tell the doctors about their problems. The clinicians perceived that the quality of life data broadened the range of the clinical inquiry and helped them identify issues for discussion. Having symptoms and functional problems expressed quantitatively on a scale was useful for detection of change over time.British Journal of Cancer () , –. DOI: ./sj/bjc/ www.bjcancer.com© The Cancer Research Campaign BODY: Transfer of information between healthcare professionals and their patients, in both directions will always remain a critically important element in diagnosis, management and patient support. For cancer patients the issues surrounding their quality of life (QL) are recognized to be central to good patient care. Comprehensive assessment of how a patient feels and functions requires a thorough inquiry. The traditional medical history and physical examination are often insufficient for assessing the full range of health-related problems in cancer patients and in chronically ill patients in general (Calkins et al, ; Passik et al, ). A standardized measurement of patients' symptoms and functioning offers an alternative structured way of collecting subjective information.Advances in health services research over the past decades have provided tools to measure the effects of illness on the daily lives of patients, in a standardized, reliable and valid way (Fitzpatrick et al, ; Ware, ). Multidimensional, self-reported questionnaires (usually called quality of life or health status questionnaires) are currently widely used in clinical research (Fayers et al, ). Several studies have investigated the value of paper-based questionnaires in clinical practice. These looked at patients with chronic diseases causing significant functional impairment and suggested that the health status reports provided accurate information, facilitated doctor-patient communication, but in general did not have a detectable impact on patients' functional status (Kazis et al, ; Calkins et al, ; Wagner et al, ). However, when resource and management suggestions were incorporated into the physicians' feedback, there was an improvement in emotional well-being and social functioning of the patients (Rubenstein et al, ). Two published studies in oncology addressed the impact of individual QL assessments on physicians' behaviour and doctor-patient communication (Detmar and Aaronson, ; Taenzer et al, ). The QL results appeared to stimulate physicians to initiate discussions on specific aspects of patients' health and well-being.The implementation of standardized QL measurement in clinical practice is proving to be difficult due to practical, methodological and conceptual barriers (Deyo and Patrick, ; McHorney and Tarlov, ). The logistical problem of gathering real-time questionnaire data for immediate use in medical practice can be overcome by computer-based administration, scoring and presentation of QL results (Sigle and Porzsolt, ; Taenzer et al, ; Velikova et al, ). However, most clinicians are not trained in the evaluation of questionnaire data and may be uncertain how to interpret and respond to the results. How to present to clinicians individual patient's QL data in a way that helps them to interpret and use it efficiently, remains an important area of research. There is a need to study the steps that would follow the feedback of QL data to clinicians: how clinicians incorporate the information within the structure of the medical interview, does the additional information change their behaviour, do they find it clinically useful, or do they find any unfavourable effects. The evaluation of any new clinical intervention should also include an assessment of patients' attitude and perceptions of the impact on their care.This project was therefore undertaken to assess the feasibility of using computer-administered individual QL measurement in oncology clinics with immediate feedback of results to clinicians and to examine the impact of the QL information on the content of the medical consultations and on patient satisfaction with communication. We will also describe patients' and physicians' acceptance and attitude to the process.METHODSStudy designThis was a prospective non-randomized study using pre-test post-test within subjects comparisons. Cancer outpatients completed two standard questionnaires on a touch-screen computer on two occasions (baseline and intervention visits). Both times they were reviewed by the same physician. Graphic and numeric summaries of the results were given to the clinician only at the second intervention visit. After the visits patients were asked about the content of the consultations, their satisfaction with the visits and their attitude to the use of QL data. Semi-structured interviews were conducted with the clinicians after each intervention consultation and at the end of the study.SampleConsecutive cancer patients receiving chemotherapy or biological therapy at the Medical Oncology Outpatient Clinics at St James's University Hospital, Leeds between October and March were considered for participation. Patients were eligible for the study if they were able to read and understand English, were willing to give informed consent and were expected to attend the clinic at least once after the baseline visit. Three clinicians (two consultants and one specialist registrar) were selected at random from all medical oncologists at St James's Hospital. We aimed at a convenience sample of patients per doctor which would allow us to detect a moderate effect size (.) (Cohen, ) of the intervention on patient satisfaction with % power and % significance level.The project was approved by the Local Ethical Committee at St James's Hospital, Leeds. Written informed consent was obtained from all participating patients and clinicians.Intervention questionnairesThe European Organisation for Research and Treatment of Cancer – Core Quality of Life Questionnaire version . (EORTC QLQ-C30) and the Hospital Anxiety and Depression Scale (HADS) were used. The EORTC QLQ-C30 is a item cancer-specific questionnaire including functional scales (physical, emotional, cognitive, social and role), a global health/QL scale, symptom scales (fatigue, pain, nausea/vomiting), and six single items assessing symptoms and financial impact of disease (Aaronson et al, ). The raw scores for the scales and items were linearly transformed to give standard scores in the range –. Higher scores on the functioning and global health scales indicate better functioning, higher scores on the symptom scales represent a higher level of symptoms.HADS is a -item instrument with two separate subscales for anxiety and depression. Scores range from – on each scale with higher scores indicating more distress. Scores above suggest probable cases of anxiety or depressive illness and scores between and – borderline cases (Zigmond and Snaith, ).Both questionnaires were administered in electronic format using computers with touch-screen monitors (the software was developed by AB Smith and can be obtained by writing to the author). The responses were scored immediately and the results were printed in two different formats – numerical (only present results) and graphical (incorporating present and previous results) (Figure 1Figure 1Example of the graphic print-out of the quality of life results.). The aim of having two different presentations of QL results was to assess clinicians' preferences.Training clinicians in interpretation of QL resultsA -h training session was conducted before the study was commenced. The training session focused on the content of EORTC QLQ-C30 and HADS, the interpretation of the scores and the general population reference data (Hjermstad et al, ; Fayers et al, ). Examples of individual patients profiles were discussed in conjunction with their clinical data.Outcome measuresPatientsPatient perceptions of the content of the consultation were assessed using a study specific checklist. It included seven possible discussion topics: () overall condition: () usual daily activities; () limitations in doing work or leisure activities; () how they feel emotionally; () symptoms of illness; () side effects of treatment; () impact of illness on relationships with family and friends. These topics were derived from the content of EORTC QLQ-C30. Patients were asked to indicate whether or not the doctor discussed any of these topics during the encounter.Patient satisfaction with the clinic visit was assessed with a -item “Cancer Research Campaign Patient Satisfaction with Communication Questionnaire” (LJ Fallowfield, personal communication). This instrument was initially based on the item Patient Satisfaction Questionnaire developed by Ware et al (quoted in Wilkin et al, ). It was extensively adapted and refined for use in UK in two pilot studies involving and oncology patients. The resulting -item questionnaire was psychometrically tested on a further sample of cancer patients. The instrument includes direct statements of opinion about the consultation. Eight items are positively worded and nine negatively worded. Each item is scored from (strongly agree) to (strongly disagree). A high score indicates a high level of satisfaction. The range of scores is from to . Principal components factor analysis revealed three dimensions: () satisfaction with rapport (six items – the doctor answered all questions, seemed to know what she/he was doing, handled the consultation well, did her/his best to keep me from worrying, seemed sympathetic, told me what I wanted to know); () dissatisfaction with doctor's manner (six items – the doctor could be irritated, could have been more respectful, too businesslike and impersonal, lacked experience with my medical problems, made me feel awkward, more attention to my privacy); () understanding (four items – the doctor used medical terms, the doctor told me all there was to know, satisfied with the medical care today, unclear about some things the doctor told me).At the end of the study patient attitude to the intervention was assessed with a short questionnaire covering practical issues of completing computer questionnaires in clinics, questionnaires' content, patient beliefs whether their functioning should be considered by the doctors and the overall usefulness of the intervention for their care (Appendix ).CliniciansThe clinical usefulness of the QL information was discussed with the physicians after each intervention consultation using semi-structured interviews based on a questionnaire used by Wagner et al (). The interviews covered the following issues: () quality of the information gathered with the standard questionnaires –whether the QL scores provided any new information, information confirming doctor's knowledge, information conflicting with the clinical assessment, accurate information and clinically relevant of QL information; () usefulness of information during the consultation – general question on usefulness and for which part of the consultation, usefulness for communication and usefulness for the management of the patient; () perceived prolongation of the intervention consultations and by how many minutes; and () preferences for format of presentation of QL data (numerical or graphic). The questions on the quality of QL information and clinical usefulness had a suggested -points response format – not at all, a little, somewhat, quite a bit and very much, but clinicians were encouraged to provide further comments.An end-of-study meeting was conducted with the three doctors together to discuss their experiences during the study. The following topics were covered: opinion on QL information, clinical usefulness, integration of QL data into the consultations, length of consultations, presentation of QL results and training of clinicians. The discussion was taped and transcribed.Statistical analysisWilcoxon signed rank test was used to compare: () the number of baseline and intervention visits when each of the seven possible discussion topics were included; () the overall number of topics discussed during the baseline and the intervention consultations; and () patient satisfaction with the baseline and the intervention consultations. The data from doctors' interviews and patients' attitude to QL questionnaires were analyzed descriptively. The end-of-study group discussion with clinicians was subjected to qualitative thematic analysis. The transcript was carefully reviewed independently by two researchers (G Velikova and AB Smith) for key phrases. Phrases were organized in clusters, compared with each other and a set of core themes was derived (Miles and Huberman, ).RESULTSPatients' characteristicsForty-two patients were asked to take part in the study. Eight patients (%) refused to participate (two stated that they did not like answering questions, one was taking part in a drug trial involving completion of QL questionnaires and five did not give a reason for refusal). Two patients completed the baseline assessment but refused the 2nd assessment (one felt that the questions were not relevant and the other was too ill to continue) and four did not attend for another clinic appointment during the study period due to changes in treatment plans. The analysis is based on patients completing both parts of the study, females and six males with median age . years (range – years). Eighteen patients had ovarian cancer and patients had malignant melanoma. They were receiving either chemotherapy ( patients) or biological therapy (four patients). Twenty-two patients were married/cohabiting, four were divorced/widowed and two did not endorse this item. Thirteen patients had basic school education, nine studied in college, three had higher university education and three missed this question. Fourteen patients were retired, six continued to work full or part time, four were homemakers, two checked other (education) and two responses were missing. The age and gender of the refusing patients and of the patients who did not complete the study was not significantly different from those of the participating patients (data not shown).All patients completed the computer questionnaires during the clinic waiting time, usually after they had routine blood samples taken and were waiting to see the doctor. For the intervention visit the print-out of the results was attached to the front of the medical notes. Summary statistics of the EORTC QLQ-C30 and HADS results are presented in Tables 1Table 1EORTC QLQ-C30 results for baseline and intervention consultations and 2Table 2HADS results for baseline and intervention consultations.Patient perceptions of the content of the consultationsTable 3Table 3Patient perceptions about the topics discussed during the baseline and intervention consultations (total number of consultations=) presents patient opinion on what topics were discussed during their consultations. Patients felt that if clinicians had the QL results they enquired more often about usual daily activities ( of the intervention vs of the baseline consultations, Z=−., P=.) and emotional problems ( of the intervention vs of the baseline consultations, Z=−., P=.). Physicians also discussed more often limitations in doing work or leisure activities ( intervention vs baseline consultations), but the difference was of borderline significance (Z=−., P=.).There was a small increase in the overall number of issues discussed during the intervention consultation (Z=−., P=.). Figure 2Figure 2Number of areas from the checklist discussed during baseline and intervention consultations (based on patients as one patient did not return the post intervention questionnaire and was excluded from this analysis). shows the distribution of the number of topics covered in baseline and intervention consultations. During the intervention visit there was an increase of the consultations when all seven topics were included and a decrease of the encounters when only two topics were covered.Patient satisfaction with communicationWe observed high patient satisfaction with both consultations and no difference between the two consultations (median and range for baseline and intervention consultations were ., – and ., – respectively). The range of possible total scores on the instrument is from to . No difference was found either when comparing the three different dimensions of satisfaction with communication. The scores for baseline and intervention consultations on each of the three factors were respectively: Satisfaction with rapport (possible range –) – median , range – and median , range –; Dissatisfaction with doctor's manner (possible range –) – median , range – and median , range –; Understanding (possible range –) – median , range – and median , range –.Patient acceptance and attitude to QL interventionTwenty-six patients completed the end-of-study questionnaires (Appendix ). Two patients did not return the questionnaires. For presentation of results the response categories Definitely and Probably Yes/No were combined. All patients who returned the questionnaires were happy to do the computer questionnaires and their visit was not prolonged or made more difficult by the study. Only one patient felt that the standard QL questionnaires were not asking the right questions. The majority of the patients felt that the doctors considered their usual daily activities (n=), how they feel emotionally (n=) and their overall quality of life (n=) when advising them. They wanted the doctors to ask them about their usual daily activities (n=), feelings (n=) and overall quality of life (n=). Twenty-three patients felt the touch-screen questionnaires were useful to tell the doctor how they felt physically and emotionally and were willing to complete them at each hospital visit as part of their usual care. Twenty-three patients thought this personal information should be kept in their medical notes, but only half (n=) wanted to see the print-out of their questionnaire results.Clinical usefulness of the QL data from physician's point of viewInterviews after each intervention consultation (n=)The clinicians discussed the QL results in and did not discuss them in three consultations. On those three occasions the doctors either knew the patient well or had seen him/her very recently. The quantitative results from the semi-structured interviews with the physicians after each intervention consultation are presented in Table 4Table 4Clinicians' evaluation of the usefulness of the QL data after each intervention consultation. Although the doctors knew the majority of the study patients well – out of (%), the QL data still provided new information in half of the cases. The information was accurate and consistent with the medical assessment. There were four occasions when the responses to some questions were somewhat different from the clinical impression. In the first case the physician identified significant insomnia and shortness of breath which were not detected by the questionnaire. This patient needed a lot of help while completing the questionnaire and obviously had problems understanding some of the questions. The second case was a clinically anxious patient who had ‘normal’ scores. In this case the patient's spouse had dictated the responses. In the other two cases the QL results suggested symptoms (shortness of breath and pain) which were not considered to be serious by the physician.The doctors felt that the QL data enhanced communication with the patients and contributed to some of the management decisions. The changes suggested by QL data consisted of stopping chemotherapy (n=), readjusting symptomatic drugs (n=), blood transfusion (n=), counselling about life style (n=), reassurance (n=), discussion of depression (n=).The physicians felt that discussion of QL results may have lengthened nine of all consultations. They estimated the prolongation to be between and min (median min) and considered this acceptable.End of study discussionThe qualitative content analysis of the discussion with the three participating clinicians identified a number of core themes. The doctors felt that the available information broadened the range of the inquiry and also helped them to focus their questions on relevant problems. Using the QL data in the encounters helped in building rapport with the patients and improved communication by giving patients chance to talk and by adding to the probing about less physical aspects. Clinicians were less clear how much the data influenced their decisions on patient management. All doctors commented that there would be times when the information may trigger a referral to other services (psychologists, social workers).The clinicians felt that having symptoms and functional problems expressed quantitatively on a scale could be very useful especially for detection of change over time. There was a strong preference for the graphic format of presentation as it allowed to see changes at a single glance.The physicians expressed concerns that the use of this rather broad, additional data may increase their workload.DISCUSSIONOur study confirms that computer-based individual QL assessment in oncology clinics with immediate feedback of results to clinicians is possible and feasible. The measurement procedure was integrated into the usual clinic routine and all patients completed the task within their waiting time.We found that the QL data may have a positive effect on doctor-patient interactions by highlighting additional areas for discussion during the consultation. Even in these small numbers we showed a significant increase in the enquiries about daily activities and emotional functioning. It is important to emphasize that this is an increase from the patients' point of view as we asked the patients to report what issues were discussed with them. The design of the study allowed us to show that patients can feel a significant difference in the content of their consultations if their doctors had structured QL information. Taenzer et al () published similar findings from a randomized study of lung cancer patients. In structured interviews after consultations, the patients reported that more issues were addressed when the doctors had QL results. Using direct observation of oncology consultations Detmar and Aaronson () found that the QL information stimulated physicians to initiate discussions on wider aspects of functioning, but there was no increase in the patients' rating of physicians' awareness of their problems. Several large studies in chronic diseases also suggested that feedback of health status data may facilitate communication between patients and clinicians and enhance patients' care (Kazis et al, ; Rubenstein et al, ; Wagner et al, ).Our initial hypothesis was that if the QL results stimulate doctors to discuss a broader range of issues during the consultation, this may result in higher patients satisfaction with doctor–patient communication and overall satisfaction with the visit. Therefore we attempted to measure this possible effect by a satisfaction questionnaire. We were aware that patient satisfaction with care questionnaires tended to yield highly positive responses (Fitzpatrick, ; Hall and Dorman, ). Therefore, we carefully chose a questionnaire which was specifically modified and tested in cancer patients and which focused more narrowly on satisfaction with communication during the last clinic visit. Unfortunately even with this questionnaire, we observed that out of patients (one missing questionnaire) had very high satisfaction scores, above (from a possible range –), which represents a significant ‘ceiling’ effect. This makes the detection of improvement difficult. Taenzer et al () reported very similar findings of high satisfaction in their group of lung cancer patients. Indeed, it may be that the concept of satisfaction is not an appropriate outcome measure in patients with advanced cancer with relatively short prognosis. People with serious diseases often give more favourable evaluation of medical care (Ben-Sira, ). Our patients attended a cancer centre with a concentration of experience and expertise in their problems. The patients understood that the options for treatment at this stage are limited and they were grateful for every help and support that was offered. We feel that high satisfaction with care among ill cancer patients is a psychological phenomenon that can not be avoided and should be taken into account when planning future research into patient care.The computer measurement was well accepted and the patients who took part in the study felt that the questionnaires were a useful tool to tell the doctors about their feelings and were happy to complete them as part of their care. These positive results should be interpreted with caution as they reflect the opinion only of patients who completed the end of study questionnaire. We approached patients and eight of them refused to take part. Two of those eight said that they did not like answering questions. One patient refused the 2nd assessment because the questions were not applicable to his situation. Two patients completed the 2nd intervention questionnaires in clinic, but did not fill in the end of study questionnaire asking for their opinion about the procedure. Therefore, we should assume that for those non-compliant patients the procedure might have been burdensome and not worthwhile. Overall, we feel that the proportion of patients who refused to participate (eight patients initially and two patients later i.e. %) and the proportion of patients completing both parts of the study ( out of approached i.e. %) is similar to therapeutic studies using QL measures (Osoba, ). However, our experience from this and other QL studies suggests that there will always be a number of patients (approximately % of all clinic attendants) for whom this approach of collecting QL information may not be acceptable or suitable. These are likely to be people who are physically very ill, emotionally distressed or simply not interested. In addition, the clinicians chose not to use the QL information in three out of patients (%), because they either knew them very well or had seen them recently.We believe that one of the major reasons for not applying QL data in clinical practice is the lack of research data on its clinical usefulness in individual patients and the lack of guidelines how to integrate the data into the decision-making framework of a medical consultation (Till et al, ). Therefore, in this research project we placed an emphasis on detailed exploration of clinicians' assessment of the value of QL information, how they used the data during medical encounters and what were their needs for effective use of the data. The three doctors who took part in the study felt that the QL information was accurate, consistent with their medical assessment and clinically relevant. This practical observation is quite important and suggests that health surveys can be used reliably in clinical practice to gather information about individuals despite the methodological concern that they may not be sufficiently precise (McHorney and Tarlov, ).The main contribution of the QL data was in bringing additional information and broadening the range of inquiry. The main impact on the consultation was in helping building rapport and better communication with the patients. The QL results were felt to be useful for exchanging information with other members of the team and for referrals to other services. The opportunity to have problems expressed on a scale and over time was rated as particularly valuable.The doctors expressed two major concerns. They were apprehensive that the broad additional information may increase their workload. They also recognized how difficult it is to change their set pattern of questioning and behaviour during a medical consultation and include a new intervention. The doctors tended to follow the usual structure of a medical encounter and included the QL data towards the end to cover additional areas and give the patients the opportunity to talk.One of the limitations of this study is the lack of direct observation of the consultations. We felt that any method of observation (i.e. tape or video recording or direct observation) may have an impact on the consultations and patients' and doctors' responses to the study questions. Therefore we decided to focus on patients' and doctors' perception and opinion about what happened during the medical encounters. However we recognize that an in-depth study of the effect of QL data on patient care would require formal analysis of the content of the consultations and we are using this approach in other studies.Our results are encouraging but should be interpreted with caution at this stage as they are based on a small sample of patients and doctors. However they do suggest a potential for influencing doctor-patient interactions when using computer-administered standard QL questionnaires. It is unclear yet whether this broadening of the clinical enquiry will improve the process of care and whether it will bring benefits for the patients like better detection of morbidity, better control of symptoms or better emotional adjustment to cancer. Based on our results and on the findings of other researchers, we believe that this approach deserves further investigation and we are currently conducting a large randomized prospective clinical trial assessing the impact of QL information on the process and outcomes of medical care.
|
PMC2752808.txt
|
TITLE: An ATP and Oxalate Generating Variant Tricarboxylic Acid Cycle Counters Aluminum Toxicity in Pseudomonas fluorescens AUTHORS: Ranji Singh, Joseph Lemire, Ryan J. Mailloux, Daniel Chénier, Robert Hamel, Vasu D. Appanna ABSTRACT: Although the tricarboxylic acid (TCA) cycle is essential in almost all aerobic organisms, its precise modulation and integration in global cellular metabolism is not fully understood. Here, we report on an alternative TCA cycle uniquely aimed at generating ATP and oxalate, two metabolites critical for the survival of Pseudomonas fluorescens. The upregulation of isocitrate lyase (ICL) and acylating glyoxylate dehydrogenase (AGODH) led to the enhanced synthesis of oxalate, a dicarboxylic acid involved in the immobilization of aluminum (Al). The increased activity of succinyl-CoA synthetase (SCS) and oxalate CoA-transferase (OCT) in the Al-stressed cells afforded an effective route to ATP synthesis from oxalyl-CoA via substrate level phosphorylation. This modified TCA cycle with diminished efficacy in NADH production and decreased CO2-evolving capacity, orchestrates the synthesis of oxalate, NADPH, and ATP, ingredients pivotal to the survival of P. fluorescens in an Al environment. The channeling of succinyl-CoA towards ATP formation may be an important function of the TCA cycle during anaerobiosis, Fe starvation and O2-limited conditions. BODY: IntroductionThe TCA cycle is a universal metabolic network in all aerobic organisms. In its catabolic role, this metabolic engine essentially oxidizes acetyl-CoA into CO2 with the concomitant production of NADH, FADH2 and nucleoside triphosphates by substrate-level phosphorylation. The NADH and FADH2 provide the reductive power to generate a proton motive force that is eventually converted into ATP via oxidative phosphorylation []. The eight enzymes that constitute this metabolic pathway work in a synergistic fashion to oxidize acetyl-CoA. In its anabolic function, the TCA cycle helps produce α-ketoglutarate, oxaloacetate, succinyl-CoA, and succinate, metabolites that are vital for the production of non-essential amino acids, fatty acids and heme respectively [], [].However, numerous organisms have also evolved to utilize various modified versions of the TCA cycle in order to survive extreme conditions or produce unique metabolites. Some autotrophic organisms for example invoke a reductive TCA cycle to fix CO2 into acetyl-CoA []. This is in essence a reverse oxidative TCA cycle. In this instance, two molecules of CO2 are converted into one molecule of acetyl-CoA. The key enzymes ATP-citrate lyase, fumarate reductase, and α-ketoglutarate ferredoxin oxidoreductase differentiate this metabolic module from its oxidative counterpart. An alternative TCA cycle has been shown to promote the survival of organisms proliferating in dicarboxylic acids even in the absence of α-ketoglutarate dehydrogenase (KGDH). In this situation, α-ketoglutarate decarboxylase coupled with succinate semialdehyde dehydrogenase enables the organism to generate reducing equivalents []. Modified TCA cycles also appear to be critical in the adaptation to a toxic environment. We have recently identified a modified TCA cycle with decreases in NAD-dependent isocitrate dehydrogenase (NAD-ICDH) and KGDH activities and an increased NADP-dependent isocitrate dehydrogenase (NADP-ICDH) activity []. This alternative TCA cycle plays a critical role in the adaptation of the cellular systems to oxidative stress owing to its ability to produce increased amounts of the antioxidant NADPH and decreased amounts of the pro-oxidant NADH. Furthermore, α-ketoglutarate (KG) also contributes to the detoxification of ROS []. The tailoring of the TCA cycle towards an antioxidant defense network appears to be orchestrated by NAD kinase and NADP phosphatase [].Although many aspects of this metabolic machinery have been delineated, its regulation, its interaction with other metabolic pathways, the significance of its non-cyclic attributes and its integration into the global molecular networks have yet to be fully elucidated. In this report, we have identified an alternative TCA cycle that enables the soil microbe P.fluorescens exposed to Al to generate ATP by a substrate-level phosphorylation module as oxidative phosphorylation, a process reliant on iron (Fe) is severely impeded. Al promotes dysfunctional Fe metabolism []. This novel ATP-producing module works in tandem with AGODH, SCS, OCT, and ICL to generate oxalate, an Al-sequester. The diversion of succinyl-CoA toward ATP synthesis during Al-toxicity, Fe-deprivation, and anaerobiosis is also discussed.Materials and MethodsMicrobial growth conditions and cellular fractionation Pseudomonas fluorescens (ATCC ) was obtained from the American Type Culture Collection. The microbe was maintained and grown in a mineral medium consisting of Na2HPO4 ( g), KH2PO4 ( g), NH4Cl (. g), MgSO4.7H2O (. g), and mM citrate as the sole carbon source. Trace elements were present in concentrations as described previously []. In the Al-stressed medium, the citric acid was complexed to AlCl3▪6H2O ( mM). Control cells were grown in citrate alone. The pH was adjusted to . with 1N NaOH and then autoclaved. The media was dispensed in mL aliquots in mL Erlenmeyer flasks and inoculated with mL of stationary phase cells grown in a citrate (control) medium. The cultures were aerated on a gyratory water bath shaker (model G76, New Brunswick Scientific) at °C and rpm. Bacterial cells were isolated by centrifugation at similar growth phases ( h for control and h for Al-stressed) and then re-suspended in a cell storage buffer (CSB) consisting of mM Tris-HCl, mM MgCl2, mM PMSF (pH .). Membrane and soluble fractions were isolated as described previously []. Purity of each fraction was tested by measuring the activities of glucose--phosphate dehydrogenase (soluble enzyme) and succinate dehydrogenase (membrane-bound enzyme). These cell free extract (CFE) fractions were kept at °C for up to days and the various enzymatic activities were monitored. The concentration of protein in each CFE fraction was determined using the Bradford assay []. Bovine serum albumin served as the standard.Enzymatic assaysThe activities of several membrane-bound and soluble enzymes (.–. µg protein equivalent/mL) were monitored using an Ultraspec Pro spectrophotometer. The activities of aconitase (ACN) and fumarase (FUM) were detected by monitoring the formation of the cis-aconitate and the disappearance of fumarate at nm, respectively []. α-Ketoglutarate dehydrogenase (KGDH), NAD-dependent isocitrate dehydrogenase (NAD-ICDH), malate synthase (MS), and isocitrate lyase (ICL) activities were assessed at nm using the ,-dinitrophenylhydrazine assay (DNPH) as described in []. While the formation of α-ketoglutarate (KG) was monitored in the case of NAD-ICDH and ICL, the utilization of the substrates, glyoxylate and KG were assessed for MS and KGDH respectively. The activity of succinate dehydrogenase (SDH) was probed using ,-dichloroindophenol (DCIP) as the chromophore []. The disappearance of the blue colour due to the donation of electrons from succinate to DCIP was monitored at nm. OCT was monitored by coupling the activity of this enzyme to DCIP and exogenous SDH. Control membrane fraction (. mg protein equivalent/mL) served as the exogenous source of SDH. The control membrane CFE served as an exogenous source of SDH. SCS activity was monitored using the ,′-dinitrothiobenzoic acid (DTNB) assay []. The interaction of DTNB with CoA was assessed at nm. While citrate synthase (CS) was measured with the DTNB assay, malate dehydrogenase (MDH) was monitored via the production of NADH in the presence of malate [].Heme measurementThe heme assay was performed as described in []. A . mg protein equivalent of membrane fraction was treated with µL of % (v/v) Triton X- (in pure methanol) and the volume was adjusted to mL with .1N NaOH. Following a min incubation at room temperature, the precipitate was removed by centrifugation and the supernatant was subjected to spectrophotometric analysis. Soluble fraction was treated in a similar fashion except a . mg/mL equivalent of protein was used and the centrifugation step was not necessary. Heme was monitored at nm. The assay was standardized using bovine hemoglobin.Blue Native Polyacrylamide Gel Electrophoresis (BN PAGE) and two-dimensional (2D) SDS-PAGEBlue Native polyacrylamide gel electrophoresis (BN PAGE) and in-gel activity stains were performed as described previously []. Membrane proteins were prepared in blue native (BN) buffer ( mM Bis-Tris, mM -aminohexanoic acid, pH ., °C) and % β-dodecyl-D-maltoside. The in-gel visualization of enzyme activity was ascertained by coupling the formation of NADH/NADPH to . mg/mL of phenazine methosulfate (PMS) and . mg/mL of iodonitrotetrazolium (INT). AGODH was visualized using mM glyoxylate, . mM CoA, . mM NADP, INT, PMS and reaction buffer. The in-gel activity of Complex I was probed using mM NADH, INT, and reaction buffer containing mM KCN. Complex IV was tested using cytochrome C and diaminobenzidene []. SCS was detected by adding mM succinyl-CoA, and . mM ADP. The formation of ATP was coupled to mM glucose, units of hexokinase, units of glucose--phosphate, units of glucose--phosphate dehydrogenase, . mM NADP, INT, PMS, and reaction buffer []. OCT activity was ascertained by adding mM succinyl-CoA, mM oxalate, µg of protein equivalent from P.fluorescens membrane fraction (which contains SDH), INT, PMS, and reaction buffer. Reactions were stopped using destaining solution (% methanol, % glacial acetic acid) once the activity bands reached their desired intensity. Appropriate controls involving the omission of substrates and enzymes were utilized. Bands were quantified using SCION imaging software for Windows. The resulting bands were documented and used for 2D SDS-PAGE analysis.2D SDS-PAGE gels were performed in accordance with the modified method described in []. Activity bands from blue native gels were precision cut from the gel and incubated in denaturing buffer (% β-mercaptoethanol, % SDS) for min, and then loaded vertically into the SDS gel. SDS-gels (% isocratic) were favoured for the proper separation of protein. Kaleidoscope standards (Bio-Rad) were used to identify the molecular mass (MM) of the proteins. Proteins were detected by staining the gel with silver staining.Metabolite analysisThe cellular levels of various metabolic intermediates and nucleotides were assessed by HPLC. Briefly, a mg protein equivalent to soluble CFE from control and Al-stressed cells (isolated at h and h, respectively) was boiled for min and then injected into an Alliance HPLC equipped with a C18-reverse phase column (Synergi Hydro-RP; µm; ×. mm, Phenomenex) operating at a flow rate of . mL/min and °C. The proper separation of the metabolites was achieved using a mobile phase consisting of mM K2HPO4 (pH .) [], []. The ability of glyoxylate to produce ATP by substrate level phosphorylation was assessed by incubating mg of protein equivalent to the CFE in a phosphate reaction buffer ( mM NaH2PO4, mM MgCl2, pH .) containing mM succinate, mM glyoxylate, . mM CoA, mM ADP, and . mM NADP for h in the presence of mM KCN. The latter inhibits complex IV. The use of succinyl-CoA for ATP production was monitored by incubating . mg membrane protein in a phosphate reaction buffer containing mM succinyl-CoA and mM ADP for h. To ascertain whether ATP was preferentially being generated by substrate-level phosphorylation, the CFE from the control and stressed cultures was incubated with mM citrate, . mM ADP, . mM NADP, and . mM CoA in the absence and presence of mM KCN. The activity of OCT was tested by monitoring the transfer of CoA from succinyl-CoA to oxalate. CFE ( mg of protein equivalent) in a phosphate reaction buffer containing . mM succinyl-CoA and . mM oxalate was reacted for min then quenched for HPLC as described above. Organic acids (oxalate, glyoxylate, and succinate) and nucleotides (ATP, ADP, oxalyl-CoA, and succinyl-CoA) were monitored at nm and nm, respectively. Negative reactions were performed in the absence of substrates. The retention time of all metabolites was confirmed by injecting known standards at various concentrations. The levels of the respective metabolites were quantified using standard curves and the EMPOWER software. Oxalyl-CoA was also identified by HPLC. Following the min reaction, the CFE (. mg protein/mL) with . mM succinyl-CoA and mM oxalate, an unidentifiable peak at min was collected and then treated with 6N KOH for h []. The hydrolytic digest was then monitored at nm for CoA and nm for oxalate. Oxalate and CoA were identified by injecting known standards. 13C-NMR analyses were performed using a Varian Gemini spectrometer operating at . MHz for 13C. Samples were analyzed with a mm dual probe (° pulse, -s relaxation delay, kilobytes of data, and scans). Chemical shifts were referenced to standard compounds under analogous conditions. The formation of succinyl-CoA was probed by incubating a mg protein equivalent of control or Al-stressed CFE for h in a phosphate reaction buffer containing [,-13C2] succinate, mM glyoxylate, . mM CoA, mM ADP, and . mM NADP. Metabolites were identified by comparing to known standards. Reactions were quenched by boiling and then subjected to spectral analysis.Regulation of AGODH activityThe Al-mediated modulation of AGODH activity was further delineated by performing regulation experiments. Control cells ( mg protein equivalent) grown in citrate for h were transferred to an Al-containing medium. Following an h exposure, the cells were harvested and treated for BN PAGE as described above. Similar experiments were performed with Al-stressed cells (grown for h) transferred into a control medium.Statistical analysisData were expressed as means±standard deviations. Statistical correlations of data were checked for significance using the Student t test (p≤.). All experiments were performed at least twice and in triplicate.ResultsAl toxicity perturbs aerobic respiration in P. fluorescensIn order to further delineate the molecular mechanisms involved in the adaptation of P. fluorescens to Al toxicity, the metabolite profile from the CFE of the control and Al-stressed cells were analyzed. Marked variations in such metabolites as citrate, oxalate, and succinate were observed (Figure , Panel ). While the level of glyoxylate was sharply increased in the Al-stressed cells, there was not a pronounced variation in the total amounts of ATP (Figure , Panel ,). To probe the nature of this disparate metabolic profile further, the specific activities of several TCA cycle enzymes and respiratory complexes were assessed. The activities of ACN, KGDH, NAD-ICDH, and SDH were all severely affected in the Al-stressed cells (Table ). In addition, complex I and IV, two key enzymes in the respiratory chain displayed a significant reduction in activity (Table ). No discernable variation in the activity of CS and MDH was observed. Despite the marked decrease of the key enzymes involved in oxidative phosphorylation, the ATP level in the Al-stressed cells did not undergo a corresponding reduction. Hence, it became evident that the cells challenged with Al were fulfilling their energy needs via an alternative mechanism../journal.pone..g001Figure 1HPLC analysis of targeted metabolites in Pseudomonas fluorescens.Panel : Soluble CFE was utilized A) control (citrate) cultures and B) Al-stressed (Al-citrate) cultures. Panel : Succinate and glyoxylate levels in P. fluorescens exposed to A) control B) Al-stressed conditions. White bars = succinate, Gray bars = glyoxylate. Panel : ATP and ADP levels in P. fluorescens exposed to A) control B) Al-stressed conditions. White bars = ATP, gray bars = ADP. n = , p≤., mean±S.D. * represents statistical significance in comparison to control. Cells were isolated at similar growth intervals to afford a proper comparison (control = h, Al-stress = h). Peaks were quantified using EMPOWER software and identified by injecting known standards../journal.pone..t001Table 1The relative activity of selected TCA cycle and ETC enzymes in the presence of Al.EnzymeAl-stressed activityACNa %NAD-ICDHb %KGDHb %SDHc %FUMa %Complex Id %Complex IVd NegligibleICLb %MSb %All activities were expressed as percent of control. a: double bonds were measured at nm. % corresponds to ± nmol x min− mg protein− for ACN and .±. nmol x min− mg protein− for FUM. b: α-ketoglutarate and glyoxylate levels were measured using DNPH at nm. % corresponds to .±. µmol x min− mg protein− for NAD-ICDH, .±. µmol x min− mg protein− for KGDH, .±. µmol x min− mg protein− for ICL, and .±. µmol x min− mg protein−1for MS. c: activity monitored at nm using DCIP. % corresponds to ±. nmol x min− mg protein−. d: Activity was assessed by quantifying BN-gel activity bands using SCION imaging. % corresponds to arbitrary units for complex I and arbitrary units for complex IV.Acylating glyoxylate dehydrogenase (AGODH) is required for oxalogenesisWe have previously demonstrated that ICL was drastically increased in cells treated with Al compared to control. However, the increased activity of ICL was not coupled to the concomitant increase of MS; this could account for the accumulation of glyoxylate []. This glyoxylate appeared to be diverted towards the production of oxalate, a moiety known to render Al innocuous []. The amounts of oxalate were dependent on the concentration of Al in the stressed medium []. Hence, the two possible enzymes involved in the oxidation of glyoxylate to oxalate, glyoxylate dehydrogenase (GODH) and AGODH, were probed by in-gel activity staining. In contrast to control cells, a more intense activity band corresponding to AGODH was observed in the Al-treated cells (Figure , Panel ). No activity bands were observed for GODH in either control or Al-stressed cells. A role for the NADP-dependent glyoxylate was further supported by the observation that this enzyme did not yield any band with NAD (. mM) (data not shown). 2D SDS-PAGE analysis and Coomassie staining revealed a high amount of protein associated with the AGODH activity bands (Figure , Panel ). To evaluate the involvement of AGODH in the adaptive response of P. fluorescens to Al toxicity further, we exposed the Al-treated cells to a control medium for h. Similar experiments were performed with control (citrate) cells exposed to an Al-enriched medium. Exposure of the Al-challenged cells to a control medium resulted in a dramatic decrease in the activity of AGODH (Figure , Panel ). In contrast, control cells incubated in an Al-enriched medium displayed an intense activity band. Thus, AGODH may be pivotal to the adaptation of P. fluorescens to Al toxicity../journal.pone..g002Figure 2Electrophoretic analysis of enzymatic activities in P.fluorescens.Panel : In-gel activity analysis for GODH and AGODH in Lane : control, Lane : Al-stressed conditions. Panel : 2D SDS-PAGE analysis of the expression of AGODH in P. fluorescens exposed to Lane : control, Lane : Al-stressed conditions. Activity bands from Panel were excised, and electrophoresed under denaturing conditions. Panel : Regulation of AGODH activity. Lane : Al-citrate medium; Lane : Al-stressed cells grown in a citrate medium for h; Lane : citrate medium; Lane : citrate medium cells transferred to an Al-citrate medium for h. Control cells were isolated at h and Al-stressed cells were isolated at h, respectively.Al toxicity and substrate level phosphorylationThe oxidative conversion of glyoxylate to oxalate requires CoA as a cofactor with the concomitant formation of a high energy intermediate []. The transfer of CoA from high-energy acyl-CoAs to succinate by SCS is often coupled to the formation of ATP. In this instance, succinate is amply available due to the increased activity of ICL and the perturbation of SDH in the Al-stressed cells. In an effort to decipher how glyoxylate oxidation leads to ATP formation, the CFE from the control and Al-stressed cells were incubated with glyoxylate, 13C-labelled succinate, CoA, and NADP for h. In the CFE obtained from the Al-stressed cells, signals at , , , and ppm. The latter peak is indicative of the carbon in the thioester group from succinyl-CoA (Figure , Panel ). In contrast, the succinate peak ( ppm) was metabolized rapidly in the control CFE. Incubation of control and Al-stressed CFE in citrate, CoA, NADP, and ADP for h in the presence and absence of KCN clearly revealed that the modified TCA cycle was responsible for the production of ATP via substrate level phosphorylation. While there was a drastic change in the ATP producing profile in the control CFE in the presence of KCN, in the stressed CFE, ATP levels did not change significantly (Figure , Panel ). The accumulation of succinyl-CoA and oxalate when the CFE from the Al-stressed cells incubated with glyoxylate and succinate, clearly pointed to an important role of succinyl-CoA in substrate-level phosphorylation (Figure , Panel )../journal.pone..g003Figure 3Alternative TCA cycle and substrate-level phosphorylation in P.fluorescens.Panel : Proton-decoupled 13C-NMR spectra obtained from the incubation of [,-13C2]-succinate incubated with P. fluorescens CFE. Panel : ATP production and influence of KCN. Control and Al-stressed CFE from P.fluorescens were incubated with mM citrate, . mM ADP, and . mM CoA. ATP was measured in both the presence or absence of ( mM) KCN. Panel : P. fluorescens CFE was incubated with mM succinate, . mM NADP, . mM CoA and mM glyoxylate for h. Succinyl-CoA and oxalate levels were quantified using EMPOWER software. A) control and B) Al-stressed cultures. n = , p≤., mean±S.D. * represents statistical significance in comparison to control. □ (open bar) = oxalate, ▪ (closed bar) = succinyl-CoA.To verify the involvement of succinyl-CoA in ATP production, the enzyme SCS was probed. The activity of this enzyme was significantly higher in the Al-stressed cells (Figure , Panel ). More ATP production was observed. In gel enzyme activity revealed a more prominent band in the CFE from the Al-stressed cells (Figure , Panel ). 2D SDS-PAGE analysis aided in establishing that the level of this enzyme was elevated in the stressed-cells (Figure , Panel ) []. Indeed, HPLC analysis pointed to the enhanced production of ATP by this enzyme in contrast to control cells (Figure , Panel ). To probe the nature of this ATP-producing mechanism further, we tested the activity of OCT, a key enzyme required for the production succinyl-CoA from oxalyl-CoA []. A significant increase in the activity of OCT was observed in the Al-treated cells (Figure , Panel ). The CFE from the Al-exposed cells also produced a major peak at min that was attributed to oxalyl-CoA (Figure , Panel ). This moiety readily yielded oxalate and CoA upon hydrolysis (data not shown). The in-gel activity assay and 2D SDS-PAGE helped confirm the overexpression of this enzyme in the Al-stressed cells (Figure , Panel and ) []. As Al is known to interfere with Fe metabolism, a key effector of heme biosynthesis, it was important to evaluate if the production of this metabolite was affected []. A drastic decrease in this moiety was observed in Al-stressed cells. Indeed, in control cells, the level of heme was . µg/mg of protein ±. in the membrane fraction and .61μg/mg of protein ±. in the cytosol. In contrast, the Al-stressed cells contained .64μg/mg of protein ±. in the membrane fraction and .98μg/mg of protein ±. in the cytosolic fraction../journal.pone..g004Figure 4The monitoring of SCS activity and expression in P.fluorescens.Panel : HPLC analysis of ATP production by SCS. The membrane fractions from A) control and B) Al-stressed cells were incubated for min in a phosphate reaction buffer containing ADP and succinyl-CoA. ATP (white bars) and succinate (gray bars) levels were quantified using EMPOWER software. The integration of control was considered to be % (ATP = × and succinate = × arbitrary units). ATP and succinate were identified by injecting known standards. n = , p≤., mean±S.D. Panel : In-gel activity detection of SCS. Panel : Bands were precision excised from Panel , and analyzed by SDS-PAGE. A) control and B) Al-stressed cells../journal.pone..g005Figure 5The monitoring of OCT activity and expression in P.fluorescens.Panel : HPLC analysis of OCT activity. CFE was incubated in a reaction buffer containing . mM succinyl-CoA and . mM oxalate for minutes and the levels of oxalate (white bars), succinyl-CoA (gray bars), and succinate (crossed-bars) were measured. EMPOWER software was utilized to quantify the metabolites. The integration value at time = of the reaction was considered % (oxalate = , succinyl-CoA = , and succinate = arbitray units, respectively). Succinate, oxalate, and succinyl-CoA were identified by injecting known standards. n = , p≤., mean±S.D. Panel : Assessment of oxalyl-CoA levels. CFE was incubated in a reaction buffer containing . mM succinyl-CoA and mM oxalate for min. Oxalyl-CoA was identified by detecting oxalate and CoA following the treatment of the collected peak at min with concentrated KOH. * represents statistical significance in comparison to control. Panel : In-gel activity of OCT. Panel : Bands were precision excised from Panel , and analyzed by SDS-PAGE. A) control and B) Al-stressed cells.DiscussionThe data in this study point to an alternative TCA cycle that P.fluorescens invokes in an effort to fulfill its fluctuating metabolic obligations in an Al-containing environment. In this instance, two moieties namely oxalate and ATP that are critical to the survival of this microbe were essentially generated via an alternative TCA cycle. Oxalate is involved in the immobilization of Al []. As oxidative phosphorylation was impeded, the ATP budget was augmented by enhanced substrate-level phosphorylation. The latter aspect is very important as under Al-stress, oxidative phosphorylation was sharply reduced due to dysfunctional Fe metabolism. Furthermore, citrate the sole carbon source cannot effectively provide ATP via glycolysis. ATP, the universal energy currency, can be generated by substrate-level phosphorylation and oxidative phosphorylation in aerobic organisms.Glycolysis and the TCA cycle are two metabolic networks that can produce ATP via substrate-level phosphorylation in P. fluorescens. Pyruvate kinase and phosphoglycerate kinase are two glycolytic enzymes that contribute to ATP formation during hypoxia []. In fact, numerous cellular systems rely on glycolysis to fulfill their energy requirement. The TCA cycle can also produce ATP by substrate-level phosphorylation, a process mediated by SCS. The microbe Trypanosoma brucei, which is responsible for human sleeping sickness, does invoke an acetate succinate CoA transferase/succinyl CoA synthetase cycle to generate ATP. The transfer of CoA to succinate from acetyl-CoA helps preserve the energy in the thioester bond of succinyl-CoA. The succinyl-CoA is subsequently utilized to generate ATP via the phosphorylation of ADP, a process that is effected by SCS []. In eukaryotes, it appears that SCS exists in two isoforms. One isoform utilizes ADP as the substrate with the concomitant formation of ATP, while the other generates GTP from GDP []–[]. It has recently been demonstrated that an ATP synthase deficient organism can survive by augmenting its ATP production via the substrate-level phosphorylation associated with the TCA cycle [].The data in the present report point to a pivotal role of SCS in ATP production. Incubation of glyoxylate, succinate, CoA, ADP with the cell-free extract of the Al-stressed cells led to a drastic increase in ATP production. In addition, high levels of oxalate were observed within the Al-exposed cells. Furthermore, an increase in OCT activity was recorded in the organism exposed to Al. Hence, SCS and OCT partner to generate ATP in the Al-stressed P. fluorescens. This is in sharp contrast to the downregulation of KGDH, the mediator of succinyl-CoA formation in a normally functioning TCA cycle []. This metabolic reorganisation will undoubtedly allow the organism to sequester Al, generate the anti-oxidant NADPH and produce ATP in an oxygen independent manner. The latter point is extremely important as Al toxicity is known to interfere with Fe metabolism and to drastically reduce the biogenesis of the respiratory complexes []. And as citrate is the only carbon nutrient available, the variant TCA cycle may be the only alternative route to ATP production. Hence, the upregulation of SCS will provide an evolutionary advantage to survive in an environment with limited access to oxygen. Furthermore, the upregulation of SCS will help mitigate, albeit partially, the drastic diminution of KGDH activity. From an evolutionary viewpoint, it may be unwise to completely eliminate a metabolic pathway due to a sporadic environmental flux. Thus, it is quite conceivable that the upregulation of SCS may help in this regard.The utilization of succinyl-CoA in the generation of ATP via substrate-level phosphorylation may be triggered by the inability of the organism to effectively transport e- via the respiratory complexes. These moieties contain heme. However, Al is known to trigger an Fe-deficient situation. This will render the biosynthesis of heme futile [], []. Indeed, a change in the cellular proteome, in an effort to preserve Fe under Fe-limited conditions has recently been shown []. The lack of heme will have a major impact on the downstream processes involving the biogenesis of heme-rich proteins like complex I, II, III and IV. This will undoubtedly perturb oxidative phosphorylation and hence the stimulation of substrate-level phosphorylation associated with the TCA cycle will help compensate for this shortcoming. Furthermore, heme production in Pseudomonas has been shown to utilize glutamate and NADPH [], []. In Al-stressed P.fluorescens, these metabolites may be diverted in an effort to combat oxidative stress. While NADPH is critical to the proper functioning of numerous anti-oxidative enzymes, glutamate has been shown to generate KG, a potent ROS scavenger [], []. This situation may contribute to the decreased heme synthesis observed in the Al-stressed conditions. Hence, enhanced TCA-cycle mediated substrate-level phosphorylation may help signal a dysfunctional Fe-metabolism and evoke a response reminiscent of anaerobiosis []. Therefore, the conversion of succinyl-CoA into ATP may be an important metabolic adaptation under Fe-limitation and O2-deficient conditions, two situations inherent of Al toxicity. The significance of the TCA cycle in metabolic adaptation and as a signaling network is only now beginning to emerge [], []. As in most organisms, succinyl-CoA is a participant in heme synthesis, it is tempting to postulate that the diversion of this high energy metabolite may be involved in signalling anaerobic situations [].The formation of oxalate via a reconfigured TCA cycle may also help the microbe immobilize the toxic metal. It is tempting to postulate that the downregulation of NAD-dependent ICDH and KGDH allows this organism to create a TCA cycle that instead of evolving 2CO2 from acetyl CoA, transforms this moiety into oxalate. The net effect of this metabolic adaptation is in fact a TCA cycle with limited CO2 release, diminished NADH formation and increased production of oxalate, NADPH and ATP. This is a classical example of an organism utilizing pre-existing biochemical reactions to create a novel metabolic pathway aimed at circumventing an environmental challenge. This adaptation provides a glimpse into how metabolic networks may have evolved. Oxalate sequesters Al and NADPH guards against Al-induced ROS. ATP obtained via substrate-level phosphorylation helps the organism survive an anaerobic condition promoted by Al. It is also important to note that ICL and MS that usually work in partnership as part of the glyoxylate cycle were uncoupled in this study []. In the Al-stressed cells, ICL activity was elevated while the activity of MS was relatively unchanged. This modification will allow the channeling of glyoxylate towards the biogenesis of oxalate, a metabolite utilized in the sequestration of Al [], []. It is noteworthy to mention that upregulation of fumarase C, an Fe-independent enzyme recently uncovered in Al-stressed P. fluorescens may help provide an alternate route to malate than the one mediated by MS [].Since Al has been found to create an intracellular oxidative environment, it is not surprising that NADPH-generating enzymes are markedly increased. We have shown the increment in both activity and expression of such enzymes as glucose--phosphate dehydrogenase, NADP-dependent ICDH and malic enzyme (ME) in P. fluorescens subjected to Al toxicity []. The NADPH-generating systems are known to help guard enzymes involved in the direct disposal of such ROS as O2 − and H2O2. Thus, as the demand for NADPH is accentuated during Al-stress, it is not surprising that upregulation of AGODH may contribute to the NADPH budget and the formation of oxalyl-CoA. The energy stored in this thioester bond is eventually transferred to succinate with the formation of the energy-rich succinyl-CoA. Succinate is readily available during Al stress due to the upregulation of ICL. The increased activity of SCS will help phosphorylate ADP into ATP with the participation of succinyl-CoA Hence, this metabolic arrangement allows the production of three products namely oxalate, NADPH and ATP all important guardians against Al toxicity.In conclusion, the present study shows how this alternative TCA cycle enables P.fluorescens to generate the three critical metabolites (oxalate, ATP, and NADPH) with the concomitant reduction in the synthesis of NADH. This fine metabolic-balancing act is crucial for the survival of the microbe. In this instance, succinyl-CoA is key to the adaptation to Al toxicity as it helps generate ATP. The flux in the concentration of succinyl-CoA may indeed reflect the status of the TCA cycle, cellular O2 gradient and Fe homeostasis. It is also important to note that this variant TCA cycle instead of releasing CO2 from acetyl-CoA fixes it into oxalate. Figure depicts a scheme that illustrates the alternative TCA cycle evoked by Al-toxicity../journal.pone..g006Figure 6An alternative TCA cycle generates oxalate and ATP as a consequence of Al toxicity.(↓) represents a decrease in enzyme activity whereas (↑) represents an increase in activity. AGODH = acylating glyoxylate dehydrogenase, SCS = succinyl-CoA synthetase, ICL = isocitrate lyase, MS = malate synthase, ACN = aconitase, NAD-ICDH = NAD-dependent isocitrate dehydrogenase, SDH = succinate dehydrogenase, FUM = fumarase, OCT = oxalate CoA-transferase, and KGDH = α-ketoglutarate dehydrogenase, CS = citrate synthase, MDH = malate dehydrogenase.
|
PMC3046429.txt
|
TITLE: Circulating microRNAs as blood-based markers for patients with primary and metastatic breast cancer AUTHORS: Carina Roth, Brigitte Rack, Volkmar Müller, Wolfgang Janni, Klaus Pantel, Heidi Schwarzenbach ABSTRACT: IntroductionMicroRNAs (miRs) are interesting new diagnostic targets that may provide important insights into the molecular pathogenesis of breast cancer. Here we evaluated, for the first time, the feasibility and clinical utility of circulating miRs as biomarkers for the detection and staging of breast cancer.MethodsThe relative concentrations of breast cancer-associated miR10b, miR34a, miR141 and miR155 were measured in the blood serum of patients with primary breast cancer (M0, n = ) and metastatic disease (M1, n = ), and healthy women by a TaqMan MicroRNA Assay.ResultsThe relative concentrations of total RNA (P = .) and miR155 (P = .) in serum significantly discriminated M0-patients from healthy women, whereas miR10b (P = .), miR34a (P = .) and miR155 (P = .) discriminated M1-patients from healthy controls. In breast cancer patients, the changes in the levels of total RNA (P = .), miR10b (P = .), miR34a (P = .) and miR155 (P = .) correlated with the presence of overt metastases. Within the M0-cohort, patients at advanced tumor stages (pT3 to ) had significantly more total RNA (P = .) and miR34a (P = .) in their blood than patients at early tumor stages (pT1 to ).ConclusionsThis pilot study provides first evidence that tumor-associated circulating miRs are elevated in the blood of breast cancer patients and associated with tumor progression. BODY: IntroductionThe development of breast cancer is a complex multistep process associated with numerous genetic alterations, downregulation of tumor suppressor genes, upregulation of oncogenes and early hematogeneous dissemination of tumor cells [-]. Accordingly, the elucidation of the molecular mechanisms in breast cancer has been the subject of extensive research over the past decade. The search for sensitive, non-invasive markers that represent tumor-associated changes in the peripheral blood might facilitate early detection of breast cancer as well as monitoring of tumor progression and treatment responses.MicroRNAs (miRs) are small, non-coding RNA molecules and consist of approximately nucleotides. MiRs modulate post-transcriptionally the expression of numerous genes, such as of tumor suppressor genes, by binding sequence specifically to their target mRNA and inhibiting their translation into polypeptides. Moreover, miRs are involved in the regulation of different cellular processes, for example, apoptosis, hematopoietic cell differentiation, metabolism, neural development and metastasis [,]. The importance of miRs in the regulation of these processes has been documented by the results obtained in knockout mice. Knocking out the enzyme Dicer, responsible for the maturation of miRs, caused death of a mouse embryo []. As half of human miRs are located in fragile chromosomal regions harboring DNA amplifications, deletions or translocations, their expression is frequently deregulated during tumor development, which contributes to tumor progression [].Apart from their release of DNA and RNA, apoptotic and necrotic cells of the primary tumor also discharge miRs into the blood circulation. Protected from the degradation by endogenous RNase activity, miRs circulate in a remarkably stable form in human blood []. In the year , extracellular serum miRs were for the first time described for patients with diffuse B cell lymphoma []. Although numerous publications have reported on the elevated levels of circulating nucleic acids in the blood of breast cancer patients [-], there are only a few publications dealing with circulating miRs in peripheral blood of breast cancer patients [,-]. To date, most studies describe the profile of miR expression in breast cancer cell lines and primary tumor tissue [].In primary and metastatic breast carcinomas, high transcript levels of miR10b were detected and associated with tumor progression. Overexpression of miR10b in a non-metastatic breast cancer cell line induced invasion and metastasis []. Knockdown of miR34a by small interfering RNA significantly suppressed proliferation in the breast cancer cell line MCF-, indicating that miR34a overexpression may be an acquired feature during carcinogenesis and support cell proliferation in breast tumors []. Consistent with its role in regulation of epithelial to mesenchymal transition (EMT), an essential early step in metastasis, the expression of the miR200 family, including miR200a, miR200b, miR200c, miR141 and miR429, was found to be lost in invasive breast cancer cell lines with a mesenchymal phenotype. Expression of these miRs was also lost in areas of metaplastic breast cancer specimens lacking the adhesion molecule E-cadherin []. Knockout mice, which do not express miR155, showed that this miR played an important role in the immune system. In Hodgkin's lymphoma, diffuse B cell lymphoma and breast cancer miR155 was highly expressed [,].In summary, miRs are involved in the molecular pathogenesis of malignant tumors including breast cancer, and they might represent novel diagnostic markers. To assess the potential of serum miRs as biomarkers in breast cancer, we here examined the transcript levels of miR10b, miR34a, miR141 and miR155 in blood serum of breast cancer patients.Material and methodsPatient design and healthy controlsPatients have been invited to take part in a multicenter study (SUCCESS) and attending local hospitals in the network of Ludwig Maximilians University of Munich. During September to July , blood serum was taken from patients with primary breast cancer about days after surgery before initiation of adjuvant therapy. The median follow-up time of this patient subgroup was three years (range . to . years). During January to April , blood serum from patients with metastatic breast cancer (n = ) was collected to years after surgery of the primary tumor. All patients analyzed had histologically proven epithelial cancer and no known comorbidities. Metastatic spread in M0 patients was excluded by chest radiology, liver ultrasound scan and bone scan. Additionally, healthy women with no history of cancer and in good health based on self-report were recruited as controls. All patients and healthy controls gave their informed consent. The examination of all samples was approved by the local ethics review boards. Table summarizes the clinical and histopathological factors of the breast cancer patient cohort.Table 1Patients' characteristics at the time of primary diagnosis of breast cancer and correlations of the serum RNA and miR values with these parametersParametersPatients (%)Total RNAmiR10bmiR34amiR141miR155Total89Median ± Standard DeviationAge56 years(range - years)Distant metastasis£M059a . ± .5b . ± .2c . ± .. ± .1d . ± .$M130a . ± .5b . ± .6c . ± .. ± .3d . ± .#Tumor stagepT1- (.)e . ± .. ± .2f . ± .. ± .. ± .9pT3- (.)e . ± .. ± .1f . ± .. ± .. ± .#Lymph node metastasisN021 (.). ± .. ± .. ± .. ± .. ± 1N1- (.). ± .. ± .. ± .. ± .. ± .#GradingII26 (.). ± .. ± .. ± .. ± .. ± 1III33 (.). ± .. ± .. ± .. ± .. ± .#Estrogen receptor statuspositive37 (.). ± .. ± .. ± .. ± .. ± .2negative22 (.). ± .. ± .. ± .. ± .. ± .#Progesterone receptor statuspositive33 (.). ± .. ± .. ± .. ± .. ± .1negative26 (.). ± .. ± .. ± .. ± .. ± .#HER2positive30 (.). ± .. ± .. ± .. ± .. ± .9negative26 (.). ± .. ± .. ± .. ± .. ± .3Number of CTC¥positive47 (.). ± .. ± .. ± .. ± .. ± .9negative10 (.). ± .. ± .. ± .. ± .. ± .7ap = ., bp = ., cp = ., dp = ., ep = ., fp = .£M0, patients with localized breast cancer$M1, patients with metastatic breast cancer#of M0 patients¥CTC, circulating tumor cellsp values as determined by Mann and Whitney-U test.Cell cultureFor investigations of miR expression the breast cancer cell lines MDA-MB- and GI- and the micrometastatic breast cancer cell lines BC-M1 and BC-S1 were used. MDA-MB- and GI- were cultured in DMEM (Invitrogen, Karlsruhe, Germany) supplemented with % FCS (fetal calf serum; PAA Laboratories, Cölbe, Germany), mMol L-glutamin (Invitrogen) and U/mL antibiotic-antimycotic solution (PAA). Micrometastatic cells were cultured in RPMI (Invitrogen) with mMol L-glutamin (Invitrogen), U/mL antibiotic-antimycotic solution (PAA), × Insulin-Transferrin-Selenium-A (Gibco, Eggenstein, Germany), ng/mL human EGF (epidermal growth factor; Macs Miltenyi Biotec, Bergisch Gladbach, Germany) and ng/mL human bFGF (basic fibroblast growth factor; Macs Miltenyi Biotec).Extraction of total RNAFor isolation of total RNA from human blood serum and cultured cell lines, the mirVana PARIS kit (Ambion, Darmstadt, Germany) was used. Four hundred μL of serum samples and × lysed cells were incubated with an equal volume of Denaturation Solution for five minutes on ice. According to the manufacturer´s protocol, the RNA extraction was performed by acid-phenol:chloroform, and the precipitation was carried out by ethanol and a filter cartridge. The extracted RNA was eluted in μL of preheated Elution Solution and measured on a NanoDrop ND- Spectrophotometer (Thermo Scientific, Wilmington, DE, USA). The RNA samples were immediately stored at -°C and within few days converted into cDNA.Conversion of total RNA into cDNAReverse transcription was performed by the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, Darmstadt, Germany). The μL-reverse transcription reaction contained . μL mM dNTPs, . μL MultiScribe Reverse Transcriptase ( U/μL), μL × Reverse Transcription Buffer, . μL RNase Inhibitor ( U/μL), nuclease-free water and . μL or μL RNA derived from human serum or cultured cells, respectively. On a MJ Research PTC- Peltier Thermal Cycler (Global Medical Instrumentation, Ramsey, Minnesota, USA) the reaction was carried out at °C for minutes, °C for minutes and °C for minutes.Preamplification of miR141 and miR16 cDNADue to the low expression of miR141, a preamplification of its cDNA was performed. To effectively normalize the expression data of miR141, cDNA of the reference miR16 was also preamplified. Both cDNAs were preamplified in . μL Taq PCR Mastermix and . μL × TaqMan MiRNA Assay mix by using the Taq PCR Mastermix Kit (Qiagen, Hilden, Germany). The PCR was run on a MJ Research PTC- Peltier Thermal Cycler (Global Medical Instrumentation): cycle at °C for minutes, cycles at °C for s, °C for s and °C for s, and a terminal cycle at °C for minutes.Quantitative real-time PCR of miR10b, miR34a, miR141 and miR155For quantitative real-time PCR, the miR-specific TaqMan MicroRNA Assays (Applied Biosystems) for miR16 (reference miR), miR10b, miR34a, miR141 and miR155 were used. In a μL-reaction, μL cDNA or μL preamplified cDNA were mixed with μL TaqMan Universal PCR Master Mix No AmpErase UNG and . μL miR-specific TaqMan MicroRNA Assay Mix on a twin-tec real-time PCR plate (Eppendorf, Hamburg, Germany). The quantitative real-time PCR reaction was performed at °C for minutes and in cycles at °C for s and °C for s on a Mastercycler Realplex (Eppendorf). Melting curve analyses were performed to verify the specificity and identity of PCR products.The obtained data of the miR expression levels were calculated and evaluated by the ΔCt method as follows: ΔCt = mean value Ct (reference miR16) - mean value Ct (miR of interest). The relative expression of miR of interest corresponded to the ^(ΔCt) value.As recommended by the manufacturer (Applied Biosystems), miR16 has been chosen as reference for normalization of the expression levels of our miR panel. To effectively normalize the expression data of miR10b, miR34a, miR141 and miR155, we analyzed the miR16 expression levels and found that the levels remained relatively constant across the serum samples. We calculated a mean value of ., ., and . with a standard deviation of ., . and . for the subgroups of M0 patients, M1 patients and healthy women, respectively.Statistical analysisThe statistical analyses were performed using the SPSS software package, version . (SPSS Inc. Chicago, IL, USA). The chi square or two-tailed Fischer´s exact test was used to identify potential associations of miR concentrations in blood serum with the clinical and histopathological risk factors of the breast cancer patients. For nonparametric comparisons, univariate analyses of the Mann Whitney-U test of two independent variables and bivariate analyses of the Spearman-Rho test were used. Kaplan-Meier plots were drawn on to estimate overall survival and recurrence, and the Log rank test was applied for statistical analyses. Missing data were handled by pairwise deletion. A P-value ≤. was considered as statistically significant. All P-values are two-sided.ResultsExpression pattern of miR10b, miR34a, miR141 and miR155 in breast cancer cell linesBefore examining the signature of the four miRs (miR10b, miR34a, miR141 and miR155) in human blood, we analyzed their expression patterns in the widely used breast cancer cell lines MDA-MB- and GI- as well as in the micrometastatic cell lines BC-M1 and BC-S1 which were established from disseminated tumor cells present in the bone marrow of breast cancer patients without overt distant metastases []. Interestingly, BC-M1 and BC-S1 cells co-express cytokeratin and vimentin consistent with an EMT-like invasive phenotype as well as other cancer stem cell characteristics [,].We quantified the relative expression of these miRs by determining the low cycle threshold (Ct) values. As shown in Figure 1A-D, the relative values of the miRs in MDA-MB- and GI- differed from those in the micrometastatic cell lines. Compared with the low values in MDA-MB- and GI- cells, the transcript levels of miR10b (Figure 1A), miR34a (Figure 1B) and miR155 (Figure 1D) were significantly upregulated in the micrometastatic breast cancer cells BC-M1 and BC-S1. MiR141 expression was clearly upregulated in MBD-MB-, and GI- cells, whereas it was nearly abolished in the micrometastatic cells (Figure 1C).Figure 1MiR expression in breast cancer cell lines as determined by quantitative real-time PCR. Basal expression levels of miR10b (A), miR34a (B), miR141 (C) and miR155 (D) in breast cancer cell lines MDA-MB- and GI-, and micrometastatic breast cancer cell lines BC-M1 and BC-S1. The relative transcript levels of the miRs were determined by the low cycle threshold (Ct) values.Profiling of total RNA, miR10b, miR34a, miR141 and miR155 in blood serum of breast cancer patients and healthy womenIn the box plot of Figure , the relative quantification of circulating total RNA, miR10b, miR34a, miR141 and miR155 in blood serum of healthy individuals and patients with breast cancer (M0, n = and M1, n = ) are depicted. To determine the differences in the relative expression profiles, we performed univariate analyses of the Mann Whitney-U test.Figure 2Levels of total RNA and miRs in blood of breast cancer patients and healthy controls. The box plot and the additionally integrated box plot of miR141 show the different, relative amounts of total RNA, miR10b, miR34a, miR141 and miR155 which circulate in blood of healthy individuals (n = ), M0 patients (n = ) and M1 patients (n = ). The relative transcript levels of miRs were determined by the low cycle threshold (Ct) values. As determined by Mann and Whitney-U test, the significant P-values of the statistical evaluations of serum RNA and miR levels are indicated above the blots.As shown in Figure , M0 patients had significantly higher levels of circulating total RNA in their blood (normalized Ct values), which was taken approximately four weeks after surgery, than healthy individuals (P = .). Counter to expectation, M1 breast cancer patients had similar RNA concentrations to healthy individuals. M0 patients had on average .-fold higher median values of RNA than M1 patients and healthy controls (P = .). In addition, we also analyzed paired serum samples taken pre- and post-operatively from breast cancer patients. Our data showed similar RNA levels before and four weeks after surgery (P = ., data not shown).In respect to the miRs, the median values of miR155 of both subgroups M0 and M1 were .-fold (P = .) and .-fold (P = .) increased in comparison to healthy women. In contrast, the values of miR10b (-fold, P = .) and miR34a (.-fold, P = .) could only discriminate M1 patients, but not M0 patients, from healthy individuals. In M0 and M1 patients, their relative yields of miR141 did not differ significantly between healthy women and women with breast cancer (Figure ).Correlation of circulating total RNA and miRs in serum of breast cancer patients with clinical and histopathological factorsWe compared the relative concentrations of circulating total RNA, miR10b, miR34a, miR141 and miR155 in blood serum of the breast cancer patients with their clinical and histopathological data. For these statistical evaluations, the Mann and Whitney-U test of the non-parametric comparison of two independent variables and Log rank test were used (Table ). As shown in Table and Figure , the transcript levels of miR10b (P = .) and miR34a (P = .) were significantly higher in serum of M1 patients than M0 patients, whereas the levels of miR155 (P = .) and total RNA (.) in M1 patients were lower than in M0 patients. Moreover, high values of total serum RNA (P = .) and miR34a (P = .) correlated with advanced tumor stages within the subgroup of M0 patients (Table ).In addition, we correlated the values of total RNA and miRs with the number of mononuclear blood cells of M0 patients, because leukocytes could also be a source for the release of these RNAs. Unfortunately, no data on the number of leukocytes were available for M1 patients. Surprisingly, M0 patients at early tumor stages had significantly more leukocytes in their blood than patients at advanced tumors (P = .). No significant relationship of the number of leukocytes to the RNA and miR values could be detected.The statistical assessments between the concentrations of miRs and the other clinical and histopathological data as well as the number of circulating tumor cells (CTC) did not reach any statistical significance (Table ). The number of CTC was only determined in blood of M0 patients. Within this subgroup, of patients were analyzable for CTC but only (%) patients had CTC in their blood. This low number of CTC-positive patients impeded an accurate statistical analysis.Kaplan-Meier and Log-rank models were used to assess the prognostic potential of miR levels in serum of breast cancer patients. The median follow-up time was three years (range . to . years). Median expression levels were used for grouping the serum samples according to low and high expression. The serum levels of miR141 could not discriminate breast cancer patients from healthy women but the association with the clinical outcome reached borderline significance (P = . for progression-free survival in M1 patients and P = . for recurrence-free survival in M0 patients). The measurements of the other serum miRs were not associated with either diagnosis or progression of breast cancer.DiscussionEmerging evidence indicates that the deregulation of miRs might play a crucial role in breast carcinogenesis []. However, their precise role in promoting breast cancer progression and metastasis remains under investigation. Previous studies scrutinizing the signature of miRs in tumor patients have mainly been carried out by using tumor tissues. To date, only a handful of publications have dealt with circulating miRs in blood of breast cancer patients [,-]. Based on their regulation of relevant target genes [,,,], we assembled a set of four miRs (miR10b, miR34a, miR141 and miR155) and examined whether their expression profile in blood serum was associated with diagnosis and progression of breast cancer. To our best knowledge, only miRNA10b, miR141 and miR155 have been measured in blood serum so far by other groups [,-] including miRNA10b and miR155 in breast cancer [,,].In the present study, we demonstrated that cell-free serum total RNA and miR155 in M0 patients as well as miR10b, miR34a and miR155 in M1 patients may discriminate breast cancer patients from healthy individuals. Whereas M0 patients at advanced tumor stages had dramatically more RNA in their blood than patients at early tumor stages, M1 patients with overt distant metastases surprisingly displayed lower levels of RNA. This observation cannot be explained by a higher cell turnover in the advanced tumors of M0 patients because these tumors were resected four weeks before taking the blood samples for miR analysis. Notably, because we collected the serum samples postoperatively, we can not exclude that serum RNA may also stem from other tissue sources, such as mononuclear blood cells, CTC or occult micrometastases. However, the statistical evaluations of the number of leukocytes showed no association with the serum RNA concentrations in M0 patients, and an inverse correlation with increasing tumor stages. Since only M0 patients had CTC in their blood, a meaningful statistical comparison between the CTC and RNA data was not possible. However, we did not even observe a tendency of higher RNA and miR levels in serum of CTC-positive M0 patients. These findings indicate that the bulk of circulating RNA may probably not stem from mononuclear blood cells, and the contribution of CTC remains questionable.The kinetics and metabolism of circulating RNA has not yet been clearly elucidated, and a possible caveat of our study results might be that our blood serum samples were collected approximately four weeks after surgery (that is, before initiation of adjuvant therapy), which might limit their use for diagnostic purposes. To address this important concern, we analyzed paired serum samples taken pre- and postoperatively from breast cancer patients. Our data showed similar RNA levels before and four weeks after surgery suggesting that the serum RNA levels did not decrease significantly after surgery of the primary breast tumor.Considering the particular miRs in blood serum of our breast cancer patient cohort, increasing concentrations of miR10b and miR34a significantly correlated with the occurrence of overt metastasis. In respect to miR10b, early studies detected this miR to be down-regulated in breast tumor tissue compared with normal breast tissue [,]. However, in a recent study using plasma and serum, the expression of circulating miR10b did not differ significantly between a breast cancer cohort and healthy controls, but significantly higher levels were observed in cancer patients with ER (estrogen receptor)-negative disease than in those with ER-positive breast cancer []. Our analyses on serum miR10b support the studies of Ma et al. who reported that miR-10b specifically played a role in the metastatic process but not in primary tumor development. They found this miR to be highly expressed in metastatic breast cancer cells, and its overexpression initiated invasion and metastasis in a combination of mouse and human cell models by indirectly activating the pro-metastatic gene RhoC [].With regard to miR34a, discrepant data on the expression of this miR in diverse tumor entities have been described. It was reported that miR34a was downregulated in non-small cell lung carcinomas, pancreas tumor cell lines, colon carcinomas and primary neuroblastomas [,]. In contrast, Dutta et al. detected a high incidence of miR34a overexpression in various tumor types and undetectable expression in only poorly differentiated gastric adenocarcinomas and renal cell carcinomas. In the majority of carcinomas they perceived the participation of miR34a in cell proliferation []. Consistent with the data of Dutta et al. [], we detected high miR34a levels in the blood of our breast cancer cohort, in particular in patients with advanced tumor stages and metastatic disease.In addition, we detected significantly lower serum levels of miR155 in M1 than in M0 patients, but both subgroups had significantly higher miR155 levels than healthy individuals. Our findings are in contrast to another study showing that transcript levels of circulating miR155 in blood did not differ significantly between breast cancer and healthy control cohorts []. Using primary breast cancer tissues and epithelial cells, Kong et al. observed that miR155 may play an important role in the TGFβ-induced EMT, cell migration and cell invasion by targeting the small G-protein RhoA. They indicated that it might be a potential therapeutic target for breast cancer intervention []. Moreover, circulating miR155 was shown to be differentially expressed in the sera of women with hormone-sensitive and -insensitive breast cancer. Women with progesterone receptor-positive tumors had a higher miR155 expression than patients with tumors that were progesterone receptor-negative []. In our breast cancer cohort, we did not observe different serum miR155 levels between hormone receptor-positive and -negative patients. This discrepancy may mainly be explained by the small number of patients analyzed in that report.MiR141 expression was reported to be associated with cancer progression in breast cancer []. Our present findings showed that although the levels of miR141 did not markedly change in the blood of patients with metastatic disease in comparison to patients with primary breast cancer, the quantification of the serum miR141 concentrations might have a prognostic value in particular in node-negative patients. To sustain this potential role of miR141 as prognostic marker, additional results of future prospective studies with more patients and longer follow-up periods are needed.Finally, we assessed the transcript levels of the four miRs in tumor and micrometastatic cell lines. In tumor cell lines, the concentrations of the four miRs differed strikingly from those in the micrometastatic tumor cell lines []. Comparing the miR profiles of the cell lines with those observed in the blood of breast cancer patients revealed that the profile of the micrometastatic cell lines was nearly congruent with that in the blood of M1 patients, whereas the profiles of the other tumor cell lines were not concordant with those in the patient´s blood. Thus, the micrometastatic cell lines might be good models for future functional analyses.ConclusionThe significant increase of miR10b, miR34a and miR155 concentrations in the peripheral blood of breast cancer patients and the observed associations with tumor progression suggest a potential clinical utility of circulating miRs as a new class of future biomarkers.AbbreviationsCt: cycle threshold; CTC: circulating tumor cells; EMT: epithelial to mesenchymal transition; ER: estrogen receptor; miRs: microRNAs; M0: primary breast cancer; M1: metastatic diseaseCompeting interestsThe authors declare that they have no competing interests.Authors' contributionsCR and HS performed all experiments. HS and CR performed the statistical analysis. HS drafted the manuscript and KP revised the manuscript. BR, VM and WJ prepared the clinical material. CR summarized the clinical parameters. HS, CR and KP were involved in conception and design of the study and participated in the discussion and interpretation of the results.
|
PMC2781298.txt
|
TITLE: Comprehensive Functional Analysis of Mycobacterium tuberculosis Toxin-Antitoxin Systems: Implications for Pathogenesis, Stress Responses, and Evolution AUTHORS: Holly R. Ramage, Lynn E. Connolly, Jeffery S. Cox ABSTRACT: Toxin-antitoxin (TA) systems, stress-responsive genetic elements ubiquitous in microbial genomes, are unusually abundant in the major human pathogen Mycobacterium tuberculosis. Why M. tuberculosis has so many TA systems and what role they play in the unique biology of the pathogen is unknown. To address these questions, we have taken a comprehensive approach to identify and functionally characterize all the TA systems encoded in the M. tuberculosis genome. Here we show that putative TA system candidates are present in M. tuberculosis, considerably more than previously thought. Comparative genomic analysis revealed that the vast majority of these systems are conserved in the M. tuberculosis complex (MTBC), but largely absent from other mycobacteria, including close relatives of M. tuberculosis. We found that many of the M. tuberculosis TA systems are located within discernable genomic islands and were thus likely acquired recently via horizontal gene transfer. We discovered a novel TA system located in the core genome that is conserved across the genus, suggesting that it may fulfill a role common to all mycobacteria. By expressing each of the putative TA systems in M. smegmatis, we demonstrate that encode a functional toxin and its cognate antitoxin. We show that the toxins of the largest family of TA systems, VapBC, act by inhibiting translation via mRNA cleavage. Expression profiling demonstrated that four systems are specifically activated during stresses likely encountered in vivo, including hypoxia and phagocytosis by macrophages. The expansion and maintenance of TA genes in the MTBC, coupled with the finding that a subset is transcriptionally activated by stress, suggests that TA systems are important for M. tuberculosis pathogenesis. BODY: IntroductionToxin-antitoxin (TA) systems are ubiquitous in prokaryotic genomes and have been proposed to play a role in several important cellular functions []. These systems typically consist of a two-gene operon encoding a toxic protein that targets an essential cellular function and an antitoxin that binds to and inhibits the toxin. Regulation of toxin activity is achieved through differential stability of the stable toxin and the unstable antitoxin []. In most cases, the antitoxin also acts as a transcriptional autorepressor of the operon, such that degradation of the antitoxin results in transcriptional induction of the TA genes. Most of what we know about TA systems has come from the pioneering work in E. coli, though their role in bacterial physiology is still controversial. Some of the genome-encoded systems are activated in response to environmental stress, resulting in cell stasis from which these cells can recover under more favorable growth conditions [],[]. In contrast, it has also been reported that the MazEF TA system participates in programmed cell death []–[]. Importantly, the HipBA TA system has been implicated in the formation of persister cells, a subpopulation of bacteria that exhibit antibiotic tolerance in an otherwise susceptible population [],[] and may thus contribute to the long treatment times required to cure some infections. Because TA systems were discovered as plasmid stability elements, it has also been proposed that genomic TA loci may similarly stabilize or help to ensure the maintenance of genes encoded nearby in the genome [],[],[]. Finally, it has also been postulated that these systems are simply selfish genetic elements that function only to maintain their own existence in a genome [],[]. Although these studies have provided a wealth of information regarding the function of TA systems in E. coli, their role in the physiology of other microbes remains largely unexplored.The most common mechanism of TA system toxicity is mediated through mRNA cleavage, resulting in translation inhibition [],[],[]. Two well-characterized TA system families of E. coli, MazEF and RelBE, have been shown to act via this mechanism and cleave specific three-nucleotide sequences [],[]. The toxins of the largest family of TA systems in M. tuberculosis, VapBC, contain PIN domains, a motif thought to be associated with ribonuclease function [] and have been shown to block translation via mRNA cleavage []–[]. Transient activation of these mRNases may allow the bacteria to adapt to stress not only by inhibiting replication, but also by degrading existing transcripts, allowing a rapid change in the metabolic program of the bacteria.Given that a TA system is required for the formation of persisters in E. coli, it has been speculated that the TA systems of M. tuberculosis may govern cell division decisions during infection []. In the majority of individuals infected with M. tuberculosis, the bacteria initially grow and then establish a latent, asymptomatic infection that can persist for decades with the potential to reactivate later in life [],[]. These persistent bacteria are thought to adopt a slowly or non-replicating state in response to environmental stresses encountered in the host [],[], yet the mechanisms by which this non-replicating state is achieved are unknown. Because the majority of current antimicrobials require bacterial growth to exert their killing action, these non-replicating persistent bacteria are thought to comprise an important subpopulation of bacteria that is refractory to antibiotic therapy []. A similar antibiotic-tolerant state is elicited by TA system activation in other bacteria [], suggesting that TA systems may contribute to the long duration of antibiotic therapy required to cure tuberculosis.The most extensively studied culture condition to induce cell stasis in M. tuberculosis is hypoxia []–[]. Gradual oxygen depletion in culture results in significant changes in metabolism and gene expression, leading to a non-replicative persistent (NRP) state []. Because bacteria experience an effectively hypoxic environment in vivo as a result of reduced oxygen availability and exposure to nitric oxide (NO), it is thought that hypoxia-induced NRP is similar to the in vivo state [],[],[],[]. Additionally, these two conditions result in a significant overlap in gene expression as they both induce the dormancy regulon, a set of genes under the control of the transcription factor DosR [],[]. Bacilli grown under hypoxia exhibit a tolerance to antimicrobial therapy []. Given that hypoxia results in a state of cell stasis and the formation of antibiotic-tolerant persisters, TA systems are prime candidates for mediating this transition both in vitro and in vivo.Recent bioinformatics studies revealed that the M. tuberculosis genome encodes numerous TA system homologs and many PIN domain-containing proteins, far more than any other intracellular pathogen [], []–[]. Although these studies suggested that there has been a significant expansion of TA systems in M. tuberculosis, these analyses may have missed distantly related homologs and novel families of TA systems and thus the total number of TA systems in M. tuberculosis may be even greater. Additionally, how the M. tuberculosis genome evolved to acquire and maintain these TA systems during its evolution is unclear. To date, there has not been a comprehensive comparative analysis to determine whether the M. tuberculosis TA systems have been selectively maintained in the pathogens of this genus. TA systems are often associated with mobile genetic elements and are thus commonly acquired by horizontal gene transfer [],[], yet only three of the M. tuberculosis TA systems have been definitively assigned to a known genomic island [],[]. Although the evolutionary history of these genes is uncertain, the vast number of TA systems in M. tuberculosis evokes the question of whether the expansion of TA systems in M. tuberculosis plays an important role in the physiology of the bacteria.A subset of the putative M. tuberculosis TA genes have been partially characterized but our knowledge of the full complement of TA systems thus far is very fragmented [], [], []–[]. Therefore, a comprehensive and systematic analysis is needed to provide a foundation on which to investigate the role of this interesting gene family in M. tuberculosis biology. Although recent bioinformatic analyses have expanded the number of putative TA systems encoded in the M. tuberculosis genome [],[],[], the key questions of how many of these genes encode functional TA systems and which of these systems are important in M. tuberculosis biology have not been addressed.Here we report the results of a comprehensive strategy to identify and examine the putative TA systems encoded in the M. tuberculosis genome. Our approach revealed many more putative TA loci than previously appreciated. Expression of each of these systems in M. smegmatis, a fast-growing relative of M. tuberculosis, revealed a subset that encodes bona fide TA systems. Importantly, we identified three novel systems that were not previously recognized and bear no similarity to known TA genes, and thus may represent new families of TA systems. Additionally, by performing comparative genomic analysis across the mycobacterial genus, we made the striking discovery that the vast majority of these systems are conserved only in the MTBC and are absent from mycobacteria outside this complex, including closely-related pathogenic species. The acquisition and expansion of TA systems likely occurred coincident with or after speciation of the MTBC from the last common ancestor, suggesting an important role for these genes in M. tuberculosis evolution. We demonstrate that toxins with homology to RNases inhibit translation and have RNase activity in vitro, while a novel toxin likely functions via a different mechanism. Finally, we show that subsets of these genes are upregulated during hypoxia or macrophage infection, providing strong evidence that these systems are activated during specific stresses likely encountered in the host.ResultsThe genome of Mycobacterium tuberculosis encodes several TA system homologs, as well as novel TA systemsTo broadly search the M. tuberculosis genome for putative TA systems, we reasoned that a combination of approaches would be more powerful than a single strategy. To this end, we utilized three complementary approaches that took advantage of different characteristics of known TA systems. First, we performed PSI-BLAST searches of the M. tuberculosis genome using the toxin and antitoxin protein sequences from each of eight major TA system families: CcdBA, HigBA, HipBA, MazEF, ParDE, RelBE, VapBC, and Doc/PhD (Table S1). When possible, we used sequences from both distantly-related organisms (Gram-negative) and more closely-related organisms (Gram-positive, high-GC) to search for homologs. Second, we expanded this list to include PIN domain-containing proteins and toxin-antitoxin systems identified in previous analyses []. Finally, we used a sequence-independent approach to identify pairs of adjacent genes encoded in the M. tuberculosis genome that bear no homology to known TA systems but share a similar genomic organization []. This method, similar to an approach used to identify novel TA systems in E. coli, included constraints on size, orientation, and spacing of putative TA pairs. Candidate genes from all three methods were filtered using four criteria for the genomic organization of TA systems: ) the putative toxin and antitoxin genes were adjacent to one another, ) the two putative genes were separated by fewer than bp, likely comprising an operon, ) neither gene encoded a protein larger than amino acids, and ) the upstream gene (putative antitoxin) was smaller than the downstream gene (putative toxin). The last criterion was disregarded for cases in which there were small differences in size provided that both the putative toxin and antitoxin had conserved protein domains associated with their proposed function. Additionally, we made an exception in the case of the lone HigBA homolog, as the orientation of the toxin and antitoxin are reversed in this system [].In total, we generated a list of putative TA systems in M. tuberculosis (Figure 1A and Table S1). Our list includes gene pairs that were identified by homology. In cases for which the putative toxins and antitoxins belonged to different TA families, the TA system homology was assigned based on the homology of the toxin gene (Table S1). Toxins that contain PIN domains are most closely related to the VapBC family and thus we have classified TA systems with toxins containing these domains as part of the VapBC family (Figure , Table S2, and Table S3). We identified an additional putative systems that share no sequence similarity to known TA genes and thus may represent novel systems (Figure 1A and Table S2). Two other groups have incorporated homology-independent methods in addition to traditional homology-dependent searches to aid in the comprehensive identification of TA systems in prokaryotic genomes [],[]. Similar to the methods used here, Sevin and Barloy-Hubler utilized a sequence-independent approach that relies on the characteristic size, gene organization and spacing of known TA systems to develop an automated, web-based tool termed RASTA-Bacteria. Makarova, et al. used a novel approach that identifies pairs of genes that are significantly non-uniformly distributed in prokaryotic genomes. Comparison of our results with those of the above studies revealed that our methods identified a large number (–) of putative new TA systems, largely of the novel class, not identified in either the Makarova study nor deemed significant by RASTA-Bacteria (Figure S1)../journal.pgen..g001Figure 1Identification and testing of putative TA systems.(A) Three approaches were used to identify putative TA systems. These are indicated as BLAST (Non-PIN; homologs found through BLAST analysis that do not contain PIN domains), PIN Domain (PIN domain-containing proteins), and Genome Org. Only (novel and not homologous to known TA systems). (B) M. smegmatis cultures with putative toxins or putative toxin-antitoxin pairs under the control of the inducible acetamidase promoter were serially diluted and plated on solid media with (right panel) or without (left panel) .% acetamide. (C) Summary of toxin and antitoxin testing results: not tested (unable to PCR or clone gene products), not toxic (no toxin activity was detected), toxic only (toxicity was not relieved by the putative antitoxin), or a TA system (toxic activity relieved by antitoxin). (D) Functional TA systems were identified as novel, PIN domain-containing proteins, or homologs found by BLAST (non-PIN domain-containing proteins). Genes found by BLAST were further subdivided by the TA system family to which they belong../journal.pgen..g002Figure 2TA system conservation across the genus Mycobacterium.Phylogenetic tree based on 16S rDNA sequences showing conservation of TA systems. The tree was constructed using Neighbor-joining inference method and nodes supported by bootstrap values>% (, replicates) are shown. Nocardia farcinica (Nfa) was used as the outgroup. TA systems are arranged according to family (vapBC, mazEF, relBE, parDE, higBA, and novel); for details see Table S6. Orange represents orthologs (BLAST best reciprocal hits displaying synteny), yellow: BLAST best reciprocal hits residing in different genomic contexts (homologs), blue: pseudogenes residing in similar genomic context, green: pseudogenes residing in different genomic contexts, and black indicates no hits were detected by BLAST. Abbreviations: Nfa Nocardia farcinica; Mab Mycobacterium abscessus; Mgi Mycobacterium gilvum; Msm Mycobacterium smegmatis; Mva Mycobacterium vanbaalenii; Mjl Mycobacterium sp. JLS; Mmc Mycobacterium sp. MCS; Mkm Mycobacterium sp. KMS; Mle Mycobacterium leprae; Mka Mycobacterium kansasii; Mpa Mycobacterium avium str. k10; Mav Mycobacterium avium ; Mul Mycobacterium ulcerans; Mma Mycobacterium marinum; Mtu Mycobacterium tuberculosis; Mmi Mycobacterium microti; Mbo Mycobacterium bovis; Maf M. africanum; Mca Mycobacterium canetti.TA system expansion is unique to the species of the MTBCTo better define when in evolutionary history TA module expansion occurred, we examined both published and unpublished genome sequences spanning the genus Mycobacterium for orthologs and homologs of the known and newly identified TA systems present in M. tuberculosis. We included the draft genomes of several members of the MTBC, including M. canetti, which is the deepest diverging lineage of this group [], as well as the genomes of M. marinum and M. kansasii, which are the most closely related mycobacterial species that lie outside the MTBC [],[]. Finally, completed genomes of rapidly growing environmental mycobacteria and the slow growing pathogens M. leprae, M. ulcerans and M. avium were included. We used reciprocal best BLAST hit and genomic context analyses [] to identify orthologs and homologs of the toxin portion of of the putative M. tuberculosis TA systems described above. For all top BLAST hits, we then determined whether an associated antitoxin was encoded nearby in the genome. putative toxin genes identified by genomic organization were excluded from this analysis as we found no supporting evidence based on either homology or over-expression experiments (see below) that they encoded functional toxins.Analysis of the MTBC identified orthologs of nearly every TA system in each of the genomes analyzed (Figure and Table S3). All of the TA systems were conserved in M. microti, and only one toxin, Rv2653c, which is encoded on a prophage, was absent in M. bovis and M. africanum. Six toxin genes appeared to be absent in M. canetti, including the prophage-encoded toxin Rv2653c. Analysis of the surrounding genomic sequences, however, revealed that sequences homologous to three of these genes, Rv0624, Rv0627 and Rv1102c, were present in the M. canetti genome but had been disrupted by genomic rearrangements, including the insertion of transposon-like sequences (data not shown). The last two toxins absent from the M. canetti genome, Rv0299 and Rv0301, are encoded on a genomic island in M. tuberculosis that has previously been shown to be absent in several M. canetti strains (Table ). These results suggest that the vast majority of TA systems were present in the progenitor of the MTBC. The finding that a small number of TA genes have been lost in M. canetti is consistent with its assignment as the deepest branching member of the MTBC lineage []../journal.pgen..t001Table 1Inferred genomic islands encoding TA systems and associated genes.Genomic IslandTA systems encodedAssociated genes implicated in virulence or stress adaptation Rv0057-Rv0080 a Rv0064A- (vapBC1) None Rv0298-Rv0303 b Rv0298- None Rv0300- (vapBC2) Rv0595c-Rv0614 b Rv0595c-0596c (vapBC4) None Rv0598c-0599c (vapBC27) Rv0608- (vapBC28) Rv0656c-Rv0666 b Rv0656c-0657c (vapBC6) Rv0666 c Rv0659c-0660c (mazEF2) Rv0661c-0662c (vapBC7) Rv0664- (vapBC8) Rv0739-Rv0750 b Rv0748- (vapBC31) None Rv1397c-Rv1398c b Rv1397c-1398c (vapBC10) None Rv1942-Rv2028 a Rv1942c-1943c (mazEF5) mce3 operond (Rv1988-Rv1991c) b Rv1955- (higBA) otsB1 e Rv1959c-1960c (parDE1) dosT and dormancy regulon member fdxA f Rv1962A-1962c (vapBC35) Rv1982A-Rv1982c (vapBC36) Rv1991A-1991c (mazEF6) Rv2009- (vapBC15) Rv2100- a Rv2103c-2104c (vapBC37) Rv2491-Rv2494 b Rv2493- (vapBC38) Rv2801c-Rv2824c b Rv2801c-2802c (mazEF9) Direct repeat (DR) locus; phage resistanceg (Rv2802-) a Rv2808 c, Rv2813 c Rv2871-Rv2872 b Rv2871- (vapBC43) Rv3320c-Rv3324c a Rv3320c-3321c (vapBC44) Molybdopterin locus IIh a []. b []. c []. d []. e []. f []. g []. h [].TA systems encoded in previously described genomic islands. Genomic islands identified by parametric or combined parametric and phylogenetic methods [],[] were examined for the presence of the TA systems identified in this work.BLASTP analysis identified putative toxins in several mycobacterial genomes outside of the MTBC (Figure and Table S3). In most cases, however, the surrounding genomic regions did not demonstrate conservation of gene content and order with the M. tuberculosis genome, suggesting that these related TA systems were acquired independently in each bacterial species rather than having evolved from a common ancestral acquisition event. Alternatively, because TA systems are often associated with mobile genetic elements, these systems may have been present in a common ancestor and subsequently moved in the genome as each species diverged. Three of the five TA systems (coding and pseudogenes) identified in M. leprae and two of the TA systems found in the M. kansasii genome are in regions with similar gene content and order to that of M. tuberculosis, arguing that they evolved from a common ancestral acquisition event. Intriguingly, the only TA module encoded in all genomes analyzed was the novel TA system Rv0909-Rv0910, identified in the homology-independent search.Strikingly, all but two of the TA systems present in M. tuberculosis are absent in the closely related pathogen M. marinum. Closer analysis of these revealed that the antitoxin gene for one of them, Rv3181c, contains two point mutations resulting a significantly truncated, and likely nonfunctional, protein (data not shown). This lack of conservation of TA systems was surprising given that M. marinum is thought to be the closest genetic relative of M. tuberculosis outside of the MTBC [],[]. These findings strongly support the idea that TA gene expansion occurred after the MTBC and M. marinum diverged from their last common ancestor, and suggests that these systems play an important role in the unique biology of the MTBC. In support of this idea and consistent with the known origins of TA systems in other bacteria, we discovered that many of the M. tuberculosis TA systems are encoded in locations in the genome that were previously identified as regions of horizontal gene transfer [],[]. By cross referencing the list of TA systems identified here with previously defined genomic islands [],[], we discovered that (%) of these systems are located in these regions (Table ). It has been proposed that the acquisition of foreign DNA sequences via horizontal gene transfer was a defining event in the speciation of the MTBC [],[],[]. The large number of TA systems that were acquired during this large influx of heterologous DNA, and subsequently maintained, lends support to the idea that these genes are integral to the biology of the MTBC.Inducible expression in M. smegmatis identifies many functional TA systemsTo begin to understand the functions and biological roles of the numerous TA systems of M. tuberculosis, we first sought to determine the number of putative TA systems that encode functional toxins. We used the inducible acetamidase promoter to conditionally express of the putative toxin genes we identified in M. smegmatis. We were unable to clone and express putative genes, likely due to differences between our strain (Erdman) and the published sequence of H37Rv. Toxicity was assessed by plating -fold dilutions of cultures on solid media in the presence or absence of inducer. Genes encoding a toxic protein product, such as Rv0301 and Rv2829c, inhibited growth of cultures on plates with inducer, but did not affect growth of bacteria on plates without inducer (Figure 1B). This method identified a total of genes that resulted in toxicity when expressed, while the remainder of the genes tested did not inhibit growth under inducing or non-inducing conditions (Table S2). Since many genes can be toxic to cells when over-expressed, it was important to show that expression of the cognate antitoxins of these genes could relieve the toxic activity. Therefore, for each gene that was toxic, we then co-expressed the toxin and antitoxin, under the control of the same inducible promoter, to determine if it allowed cell growth in the presence of inducer (Figure 1B). Toxic proteins that were inactivated by their putative antitoxins, as in the case of Rv0300- and Rv2829c-2830c, were then considered functional TA systems. For two of the RelBE homologs, Rv1246c and Rv2866, we were unable to obtain transformants in M. smegmatis. This result is most likely due to low levels of expression from the acetamidase promoter in the absence of inducer. For these toxins, antitoxin activity was assessed by the ability to obtain transformants with the vector containing both the toxin and the antitoxin.In total, we discovered pairs of genes in the M. tuberculosis genome that function as toxin-antitoxin genes in M. smegmatis (Figure 1C and 1D). The majority of the TA systems we identified came from those found by BLAST and PIN-domain containing proteins. Indeed, nearly half of the putative systems identified by these methods are functional TA systems. However, of the genes found by our homology-independent search, a much lower percentage of genes were subsequently found to be functional TA systems. This is not surprising, given that the majority of these genes are not predicted to have protein domains associated with toxin or antitoxin activity. While this method was less efficient, the three novel TA systems identified (Rv0298-, Rv0909-, and Rv2653c-2654c) are of particular interest. Although Rv0299 appears to be distantly related to MazF, neither Rv0910 nor Rv2653c bear any homology to known TA systems, nor to one another, raising the possibility that they may function by novel mechanisms of toxicity (Figure 1D).Comparative genomic analysis reveals a novel, conserved TA systemStrikingly, Rv0910 was the only toxin present in all genomes analyzed and, in all cases, orthologs of the putative antitoxin Rv0909 were also present nearby (Figure 3A). In contrast to many of the other TA systems identified, this operon is not encoded in a genomic island, and its position in the genome is relatively conserved throughout the genus (Figure 3A and Table ). These findings suggest that the Rv0909-Rv0910 system may play a conserved role in the physiology of this otherwise diverse group of bacteria../journal.pgen..g003Figure 3Conservation of a novel TA system.(A) Genomic regions of Rv0909-Rv0910 TA module orthologs (orange) across diverse mycobacteria showing conservation of toxin and cognate antitoxin genes as well as surrounding genomic context, as indicated by color coding of orthologs across species. (B) M. smegmatis Rv0909- orthologs encode a functional TA system. M. smegmatis cells expressing MSEMG_5634 alone or MSMEG_5634- under the control of the inducible acetamidase promoter were serially diluted and plated on solid media with (right panel) or without (left panel) .% acetamide. (A) adapted from the tree-browser function in MicrobesOnline [].To determine whether this novel, conserved putative TA system acts as a TA system in other mycobacteria, we cloned the Rv0909- orthologs MSMEG_5635- from M. smegmatis. We assessed the ability of MSEMG_5634 to inhibit cell growth as well as the ability of MSEMG_5635 to rescue growth inhibition as described above. Plating cells expressing toxin alone on media containing inducer led to growth inhibition that was ameliorated when the cognate antitoxin was co-expressed with the toxin (Figure 3B). These results show that a second member of this novel TA family functions as a TA pair, supporting the idea that Rv0909- represents the founding member of a new TA family.VapB antitoxins show specificity for their cognate VapC toxinsThe VapBC family comprises, by far, the largest family of TA systems in M. tuberculosis. Given that there are a large number of these related genes, we wanted to determine the potential for cross-talk between VapB antitoxins and VapC toxins. Specifically, we wanted to determine if VapB antitoxins can inactivate non-cognate VapC toxins, resulting in the inhibition of toxicity. To address this question we expressed four heterologous VapB and VapC proteins, under the control of separate inducible promoters, and assessed growth in the presence and absence of both inducers. Remarkably, although these are related proteins, our results show that these antitoxins are only able to inhibit their cognate toxins (Figure ). Although we analyzed only a subset of the VapBC family, this data strongly suggests that VapB antitoxins are highly specific for their associated toxins. Given these results, we conclude that cross-talk is similarly unlikely to occur in vivo../journal.pgen..g004Figure 4VapB Antitoxins are specific for their cognate VapC toxins. M. smegmatis cultures carrying VapC toxins under control of the inducible acetamidase promoter and VapB antitoxins under the control of a tetracycline-inducible promoter were assessed for toxin activity. The VapC toxin being tested is indicated on the left side and the VapB antitoxins are indicated across the top of each set of panels. Each set of panels includes strains tested on solid media without (top) and with (bottom) inducers. The cognate toxin-antitoxin pair for each set is indicated in blue.VapC homologs inhibit translation and have RNase activity in vitro Many toxins of TA systems function as RNases and result in translation inhibition when activated [],[]. In particular, PIN domain-containing proteins, including some VapC homologs, have been shown to have RNase function []–[],[]. We expressed the VapC homolog Rv0301 in M. smegmatis and monitored bulk translation via incorporation of 35S-methionine over a six hour time course. As shown in Figure 5A, Rv0301 expression led to inhibition of translation, an effect that was reversed by co-expression of its antitoxin, Rv0300. The effect on protein synthesis preceded the inhibition of growth caused by this toxin (Figure 5B). Likewise, expression of three other VapC homologs (Rv1561, and Rv2829c, Rv3408) also inhibited translation (Figure 5A). As a control, addition of hygromycin, an antibiotic that targets protein synthesis, inhibited incorporation of 35S-methionine (Figure 5A) and growth (Figure 5B), as early as one hour after addition to the media. The modest difference in the kinetics of translation inhibition between toxin induction and addition of antibiotics is likely due to time required for transcription and synthesis of the toxin../journal.pgen..g005Figure 5VapC homologs have RNase activity and inhibit translation but a novel toxin does not.(A) Cultures of M. smegmatis harboring empty vector (pHR100) or acetamide-inducible toxin constructs were treated with .% acetamide. At the indicated times, cells were labeled with of 35S-methionine at °C for min and incorporation of radioactivity was measured. Incorporation at t = was set as % translation. As controls, cells were treated with . µg/ml ciprofloxacin (cip), or µg/ml hygromycin (hyg). The average of three experiments is shown and error bars represent the standard deviation. (B) Cultures of M. smegmatis were grown and induced as described above. The OD600 of each culture was measured at the indicated times. Results are plotted as fold-increase of OD600 at each timepoint as compared to OD600 at t = . The average of three experiments is shown error bars represent the standard deviation. (C) Purified toxins were incubated with MS2 RNA for h at °C. The RNA was then purified and electrophoresed in a % denaturing agarose gel. Included as controls were RNA alone (RNA), His-MBP (MBP), and E. coli MazF (MazF). (D) Purified toxins MazF and Rv0301 were incubated with their respective GST-tagged antitoxins MazE ( µg) and Rv0300 ( µg) and . µg MS2 RNA for h at °C. Reactions were electrophoresed in a % agarose gel.Given the proposed ribonuclease function associated with PIN domains, we reasoned that a likely mechanism for translation inhibition of the toxins tested was RNA cleavage. Indeed, in vitro RNase activity of an M. tuberculosis VapC homolog was recently demonstrated []. To test the ability of these proteins to hydrolyze RNA, we performed in vitro RNA cleavage assays with purified VapC proteins and the viral MS2 RNA, a substrate that has been effectively used to detect RNase activity []. As shown in Figure 5C, incubation of MS2 RNA with E. coli MazF, or with either of two M. tuberculosis VapC homologs, Rv0301 and Rv1561, resulted in degradation of the RNA, though Rv0301 exhibited less potent RNase activity. The incubation of the toxins MazF and Rv0301 with their cognate antitoxins, MazE and Rv0300, respectively, inhibited this RNase activity (Figure 5D). As controls, we also included MS2 RNA incubated with buffer alone, as well as an MBP-His protein fragment, which both failed to cleave MS2 RNA (Figure 5C). These results show that M. tuberculosis VapC homologs inhibit translation and strongly suggest that these toxins affect translation directly via RNA cleavage.A novel toxin inhibits growth but not translationIn contrast to our results with the VapC homologs, expression of the novel toxin Rv0910 did not result in inhibition of translation throughout the duration of the assay (Figure 5A). However, expression of this toxin did inhibit cell growth, suggesting that Rv0910 targets a cellular process other than translation (Figure 5B). Additionally, incubation of MS2 RNA with purified Rv0910 yielded intact MS2 RNA, consistent with this idea. The specificity of the translation assay was verified by treating cultures with the antibiotic ciprofloxacin, a DNA replication inhibitor. Although the antibiotic inhibited growth (Figure 5B) and caused DNA damage as measured by recA expression (Figure S2), it did not affect translation over the course of the experiment (Figure 5A). This important specificity control shows that disruption of other macromolecular synthesis pathways does not affect translation during this assay, and is similar to the results obtained with expression of Rv0910. Taken together, these results strongly suggest that Rv0909- represents a novel TA system that inhibits cell growth via a mechanism distinct from the VapBC family.Subsets of TA systems are expressed in M. tuberculosis under conditions of stressThe presence of such a large number of TA systems presents an obvious question: What is the benefit of having so many of these genes in one organism? We postulated that subsets of TA systems important for M. tuberculosis biology may respond to different cellular stresses. To test this hypothesis, we determined if any of the functional TA systems in M. tuberculosis were transcriptionally activated under two conditions encountered during infection. In particular, we examined the response of TA systems under hypoxic conditions in culture and during infection of IFN-γ-stimulated murine bone marrow-derived macrophages. To assess TA activation, we took advantage of the fact that most antitoxins also function as transcriptional autorepressors, and thus degradation of an antitoxin results in increased transcription of its operon. Although this increase in transcription leads to increased protein synthesis of both the toxin and antitoxin, the unstable antitoxin is typically selectively targeted for degradation by a protease, further increasing transcription of the operon, while the more stable toxin interacts with its cellular target []. In this manner, we are using the increase in transcription as an indirect read-out of TA system activation via degradation of the antitoxin, thus relieving its transcriptional inhibitory activity.We monitored the expression of each of the functional TA systems by quantitative PCR and obtained detectable signal for TA systems during hypoxia and for systems during macrophage infection (Table S4 and Table S5). Two TA systems, Rv2009- and Rv1955-, were induced during hypoxia (Figure 6A). Transcription of hspX and fdxA, two genes belonging to the dormancy regulon that are very highly induced during hypoxia, was also induced, demonstrating that the bacteria experienced a hypoxic environment (Figure 6A). Of the TA systems monitored during macrophage infection, Rv1560- and Rv0549c-0550c were induced. As controls, transcription of hspX and icl were greatly induced, consistent with previous results of the transcriptional response following macrophage infection [] (Figure 6B). It is interesting that the two TA systems that are activated during hypoxia are not activated during macrophage infection, given the effectively hypoxic environment generated via nitric oxide signaling following IFN-γ stimulation of the macrophages. Indeed, these two conditions both result in induction of the dormancy regulon [],[]. However, the TA genes tested here are not part of the DosR dormancy regulon and are likely activated by independent mechanisms. Certainly, it is possible that other TA systems are also activated under these conditions but this was not detectable by the methods used here. In light of these data, it appears that TA systems are regulated independently from one another, and expression of at least four of these TA systems is modulated in response to different environmental stresses. This supports our hypothesis that specific subsets of TA genes are regulated in response to changes in the environment../journal.pgen..g006Figure 6Subsets of TA systems are activated during stress.(A) Cultures were grown in liter roller bottles with a headspace ratio of . ( ml culture) at °C, with slow stirring to induce NRP. At the indicated days, cells were collected and RNA was isolated and amplified. Gene expression was measured using qPCR. One of three similar experiments is shown and error bars represent the standard deviation within this experiment. (B) IFN-γ stimulated bone marrow-derived macrophages were infected with M. tuberculosis at an MOI of . Bacterial RNA from intracellular bacilli was isolated and amplified. Gene expression was measured by qPCR and normalized to 16S rRNA as an internal control. Gene expression from in vitro grown log-phase cultures (log) is included for comparison. One of three similar experiments is shown and error bars represent the standard deviation within this experiment.DiscussionHere we have shown that of putative M. tuberculosis toxin-antitoxin loci, encode functional TA systems. The numbers of both the putative and functional TA systems are significantly greater than in any other organism studied thus far. Our sequence-based searching of the M. tuberculosis genome, in conjunction with previous bioinformatic approaches, has identified TA systems with homology to known TA systems [],[],[]. Because mycobacterial TA genes may have limited sequence identity with distantly related homologs from other bacteria, there may be additional TA systems that have yet to be discovered. Importantly, incorporation of our sequence-independent method identified an additional loci bearing no homology to known TA genes, allowing us to assemble one of the most comprehensive lists of putative TA systems in M. tuberculosis to date.It is striking that nearly all of the TA systems we identified are well conserved among the MTBC but are largely absent from species outside of this complex, including M. marinum. The paucity of TA systems in M. marinum is particularly notable as the two bacteria are highly related, sharing orthologs with an average amino acid identity of % []. Therefore, the massive expansion of TA systems is a distinguishing feature of the MTBC, and the acquisition and maintenance of these genes was likely instrumental for the evolution of M. tuberculosis. We discovered that of the protein-coding genes located within these regions, are TA genes (Table ). This frequency of TA genes (%) is significantly higher than that of the entire M. tuberculosis genome (%), indicating that there is an enrichment of TA loci in these regions. Our results, combined with data from Becq et al. [] and Stinear et al. [], lends strong support to the idea that many of the TA systems were recently acquired via horizontal gene transfer after the divergence between M. tuberculosis and M. marinum. It is possible, however, that some of these genes were acquired in a common ancestor of the slow-growing mycobacteria and subsequently lost in a subset of extant species. In support of the latter hypothesis, M. leprae and M. kansasii, bacteria that are thought to be more distantly related to the MTBC, contain TA system orthologs that are clearly absent from M. marinum (Figure and Table S3). Alternatively, the assignment of M. marinum as the closest relative of the MTBC may be incorrect, as supported by whole-genome comparisons that place M. kansasii as the nearest neighbor of the MTBC []. Given the evidence that the vast majority of TA systems have no counterparts in mycobacteria outside of the MTBC, the most parsimonious explanation for their expansion is that these genes were acquired after speciation. In addition, further amplification of TA systems may have occurred in the MTBC via gene duplication events.It is curious that so many TA loci are present in the M. tuberculosis genome. One possible reason for this expansion is that genomic TA systems may function to stabilize the M. tuberculosis chromosome, akin to the role of TA systems in stabilizing plasmids []. Since toxins are typically more stable than antitoxins, TA systems inhibit post-segregational plasmid loss because cells that do not inherit the episome rapidly deplete the antitoxin protein, leading to toxin activation and inhibition of growth. Indeed, there is recent evidence that chromosomal TA systems may protect adjacent regions of the chromosome from deletion []. In support of this notion, many of the TA systems identified here are encoded on genomic islands that include genes important for M. tuberculosis virulence or physiology (Table ). Alternatively, TA systems may participate more directly in the physiology of M. tuberculosis by functioning as stress response elements, as has been demonstrated in E. coli [],[],[],[]. There is a growing body of evidence that some TA systems are induced during exposure to adverse environmental conditions, such as exposure to antibiotics, allowing cells to respond and adapt to the assault []. Indeed, a screen to identify mutants in M. tuberculosis with altered growth kinetics during transitions in carbon availability revealed numerous TA systems identified in our analysis likely participate in growth rate decisions []. In addition to responding to stress directly, the HipBA TA system participates in generating slowly or non-replicating persister cell subpopulations within a larger group of growing bacteria []. In this way, TA systems provide resistance to non-favorable environmental conditions as persister cells are better able to survive the onslaughts of severe stress than are replicating bacteria. Given the number and diversity of TA systems in M. tuberculosis, it is possible that some serve to stabilize the genome while others serve to provide stress resistance.Of particular interest are TA systems that participate in the biology of M. tuberculosis, either participating in stress-response or persister formation. Our findings that four TA systems are activated during cellular stress supports the notion that these loci participate directly in M. tuberculosis physiology. For example, both Rv1955- and Rv2009- are induced during the transition to hypoxia (Figure 6A), suggesting they play a role in the adaptation of M. tuberculosis to low oxygen conditions. Curiously, both of the hypoxia-induced TA loci, Rv1955- and Rv2009-, are located within the same genomic island []. Although these genes are not part of the “dormancy regulon”, notable members of this regulon, dosT and fdxA, are also in the same genomic island. It may be that acquisition of this entire region helped promote M. tuberculosis' ability to respond to a hypoxic environment. In addition to the TA genes upregulated by hypoxia, Rv1560-Rv1561 and Rv0549c-0550c are specifically upregulated during macrophage infection (Figure 6B), in agreement with previous studies suggesting that these genes may be important during infection [],[]. Taken together, these data suggest that a subset of TA systems is important for M. tuberculosis in vivo, perhaps as stress-response elements.In an attempt to identify novel classes of TA systems, we incorporated a sequence-independent method in our bioinformatics search. Importantly, this strategy revealed three novel TA systems. One of these, Rv0909-, is conserved among the diverse range of the mycobacterial species analyzed (Figure and Figure 3A), suggesting an ancient history and likely fundamental role in mycobacterial physiology. Our results indicate that Rv0910 lacks RNase activity (Figure 5C) and thus probably functions by an alternate mechanism than the majority of M. tuberculosis TA systems. Interestingly, Rv0910 is most similar to the polyketide cyclase group of the START domain superfamily of proteins []. Although it is not clear how a polyketide cyclase would function to inhibit cell growth, our results demonstrate for the first time that these two genes have toxin and antitoxin activities. It is interesting to note that mycobacteria encode numerous polyketide synthases and have an enormous capacity for lipid synthesis []. Since many TA toxins inhibit macromolecular synthesis (translation, DNA synthesis), it is tempting to speculate that Rv0910 may inhibit lipid biosynthesis.Although our approach yielded a total of functional TA systems, it is likely that there are additional TA systems in the M. tuberculosis genome. Our search for putative systems was biased by using characteristics of known TA systems, including constraints on size and gene organization. Therefore, any TA system with different features would have been excluded from our search. Indeed, two other groups have incorporated homology-independent methods for comprehensive discovery of TA systems in microbial genomes and identify an additional – putative systems in the M. tuberculosis genome, depending on the level of scoring confidence chosen (Figure S1). We also cannot rule out the possibility that some of the toxins that did not inhibit growth were simply not expressed at sufficiently high levels using our system. Additionally, although M. smegmatis and M. tuberculosis are similar, we cannot discount the possibility that there are some species-specific factors that may be required for the proper function of some of the TA systems tested.Our results differ significantly from two recent reports in which M. tuberculosis TA systems were expressed in E. coli [],[]. Of the putative systems we tested in M. smegmatis, only have been evaluated in E. coli. Of the functional TA systems we identified, seven were not functional in E. coli and a further putative systems were not assessed. In contrast, only four systems were uniquely toxic in E. coli. The different results obtained in these studies may be due to the use of a distantly related host, E. coli, rather than the more closely related species, M. smegmatis. For example, specific factors such as transcript GC content, may greatly influence the number of potential targets of an RNase, and these differences are minimized by using another mycobacterial species.Given the huge expansion of VapBC homologs in M. tuberculosis, it seems likely that there are many more toxins that function via RNA cleavage than by other mechanisms. It is perplexing that one genome would have so many genes encoding proteins that perform the same function. One possibility is that most of these TA systems participate solely in genome stability and thus redundant function would not be an issue. However, all four of the stress-responsive TA systems we identified are putative RNAses. Therefore, another possibility is that, although they have the same function, specific TA systems are under different regulatory controls, such that only a subset are activated in response to particular stresses. Having a wide variety of mRNases under diverse regulatory mechanisms would allow the cell to adapt to many different conditions. An alternative explanation is that the different mRNases are actually functionally distinct. Under certain conditions, it may be useful for the cell to express an mRNase with a short, ubiquitous cleavage site that will target most of the existing messages in the cell, and allow for newly transcribed messages to be translated. This provides an efficient way to erase the previous transcriptional profile of the bacterium, allowing the cell to reprogram the proteome and thus rapidly change the metabolic state of the cell during conditions of stress (Figure ). Alternatively, some TA systems may target only a limited number of messages and, therefore, not inhibit bulk translation. For example, one possible mechanism to tailor the response of the cell upon expression of an mRNase is through the recognition site at which mRNA cleavage occurs. Indeed, the cleavage sites for two M. tuberculosis MazF homologs target pentad sequences, longer than the three-residue recognition site of MazF in E. coli []. This may allow the targeting of specific messages, giving rise to more subtle changes in the metabolic state of the cell. It may be the presence of so many mRNases that allows M. tuberculosis to regulate growth, survival and metabolism during a wide range of environmental stresses, including those encountered during infection. This hypothesis is consistent with our findings that the vast majority of TA systems are present only in the virulent mycobacteria of the MTBC../journal.pgen..g007Figure 7Model of stress-induced TA system activation and proteome remodeling.(A) Under normal growth conditions, the antitoxin is bound to its cognate toxin, this complex in turn binds its own promoter, inhibiting transcription. (B) In response to stress, the antitoxin is specifically degraded, releasing the toxin to cleaves existing transcripts. Additionally, degradation of the antitoxin results in increased transcription of the TA system. (C) Upon stabilization of the antitoxin, the TA system is inactivated and toxin-antitoxin complex, resulting in transcriptional inhibition of its operon. This pulse of toxin activity functionally erases the previous transcriptional profile allowing the newly stress induced messages to be preferentially translated.Materials and MethodsStrains and plasmidsAll strains, plasmids and primers used in this study can be found in Table S6.Identifying putative TA systemsTA system homologs in M. tuberculosis were identified using PSI-BLAST (NCBI) with a cut-off E-value of −. Iterations were repeated until we obtained no new hits below the cut-off E-value. The input sequences for this analysis included the toxins and antitoxins of major TA system families []. The origins of the input sequences of each TA system family used for BLAST analysis were as follows: CcdBA (F plasmid from E. coli), MazEF (MazEF from E.coli, PemK from Rhodococcus erythropolis), Doc/Phd (enterobacteria phage P1, Phd from Frankia alni ACN14a), RelBE (RelBE from E. coli, PasBA from plasmid pTCF14 of Acidithiobacillus caldus, YoeB/YefM from E. coli), HipBA (HipBA from E. coli), ParDE (ParDE from plasmid RK2 of E. coli), HigBA (HigBA from plasmid Rts1 of E. coli, HigA of Xylanimonas cellulosilytica DSM ), VapBC (VapBC from Frankia sp., StbBC from plasmid pDC3000B of Pseudomonas syringae). Following identification of toxin genes we determined if an adjacent upstream gene smaller than the putative toxin gene was present. Following identification of antitoxin genes we determined if an adjacent downstream gene larger than the putative toxin was present. In both cases, we required a maximum distance of bp between the putative toxin and antitoxin. Homologs for which we were unable to find an adjacent cognate toxin or antitoxin or in which the adjacent gene did not meet our criteria for either size or distance between genes were excluded. In cases where the putative toxin and antitoxin of an adjacent pair were homologous to different TA system families, we assigned the pair based on the homology of the toxin gene. PIN domain-containing proteins were identified as previously described []. To find novel TA pairs we searched the M. tuberculosis genome for pairs of genes as previously described []. In summary, we identified pairs of open reading frames (ORFs) encoding hypothetical proteins in the M. tuberculosis genome of less than amino acids, were less than bp apart, and in which the upstream ORF was smaller than the downstream ORF.Comparative genomics studyThe fully sequenced and annotated genomes were downloaded from the NCBI database. The latest assemblies of unfinished genomes were downloaded from the Sanger Institute website with permission from Dr. Julian Parkhill. or the NCBI database with permission from Dr. Marcel Behr. We used standard BLASTP or TBLASTX to search protein or nucleic acid databases of each genome for homologs of the M. tuberculosis toxin proteins identified in this study. For each toxin homolog or ortholog identified, we determined whether an adjacent putative antitoxin was present. Orthologs were defined as BLAST reciprocal best hits (E-value<−) displaying conserved synteny with other orthologs []. A detailed description of the phylogenetic analysis can be found in Text S1 and Table S3.Assessing toxin and antitoxin activityAll putative M. tuberculosis toxin genes and toxin-antitoxin gene pairs were inserted downstream of the inducible acetamidase promoter in plasmid pHR100. Toxin and antitoxin activity was assessed by growing M. smegmatis carrying the appropriate vector at °C on 7H10 solid media with .% Tween-, µg/ml kanamycin, and .% acetamide to induce gene expression. Growth was assessed after three days of incubation. Cross-talk between non-cognate VapB and VapC proteins was assessed by co-transforming M. smegmatis with the VapC toxin under the control of the acetamidase promoter and the VapB antitoxins under the control of the tetracycline-inducible promoter in plasmid pUV15tetORm []. Toxicity was assessed by patching colonies onto solid media in the absence (7H10 with .% Tween-, µg/ml kanamycin, µg/ml hygromycin) or presence of both inducers (7H10 with .% Tween-, µg/ml kanamycin, µg/ml hygromycin, .% acetamide, ng/ml ATc).Translation assays M. smegmatis cells carrying the appropriate expression vector were grown to early log phase at °C in 7H9 Middlebrook media with µg/ml kanamycin. At OD600 ., acetamide was added to a final concentration of .% to induce gene expression. Control cultures were treated with either . µg/ml ciprofloxacin or µg/ml hygromycin. Samples of ml were harvested by centrifugation at the timepoints indicated and resuspended in . ml media containing µCi of 35S-methionine. After one minute of incorporation at °C, reactions were stopped by adding ml mM sodium azide and immediately frozen in liquid nitrogen. Proteins were precipitated with % trichloroacetic acid and concentrations were determined using Micro BCA Protein Assay Kit (Pierce). Radioactivity incorporated was assessed via a liquid scintillation counter and normalized to protein concentration in each sample. The radioactivity incorporated at t = for each culture was set as % translation and all subsequent measurements were compared to this value.Purification of proteinsToxin and antitoxin proteins for the RNase assay were expressed in E. coli BL21 (DE3) pLysS cells. The toxins were expressed as N-terminal (His)-MBP-TEV tagged fusions while the antitoxins were expressed as N-terminal GST fusions. Protein expression was induced for h at °C with µM IPTG. Toxin proteins were purified using Talon metal affinity resin (Clontech). The resin was washed using buffer ( mM NaPO4, mM NaCl, pH .) containing imidazole at concentrations of , , and mM and eluted using mM imidazole. Antitoxin protein were purified using glutathione resin, washed with column volumes of the buffer indicated above and eluted using mM reduced glutathione. Proteins were subsequently dialyzed in buffer containing mM NaCl and mM Tris-HCl. Toxin proteins were TEV-digested at a ratio of ∶ (TEV∶protein) in buffer containing mM NaCl, mM Tris-HCl and mM DTT to remove the tags.RNase activity1. µg MS2 RNA (Roche) was incubated for h with µg of each purified protein at °C in mM Tris-HCl (pH ). RNA was purified and samples were heated to °C for min and placed on ice for min before loading in a denaturing agarose gel (% agarose, .% formaldehyde, × MOPS buffer). To assess antitoxin activity, µg of purified toxin protein was incubated with antitoxins Rv0300-GST ( µg) or MazF-GST ( µg). RNA from each reaction was electrophoresed in a % agarose gel under non-denaturing conditions. In vitro hypoxiaNRP was induced essentially as described in [] with a larger culture volume ( ml) in liter roller bottles to achieve a headspace ratio of .. Bacteria were pelleted and lysed at the indicated timepoints by bead beating with µl . mm zirconia beads (Biospec) at maximum speed for s in ml Trizol and total RNA was isolated via chloroform extraction and sodium acetate precipitation as previously described. [] Bacterial RNA was amplified using the MessageAmp II Bacteria Prokaryotic RNA Kit per the manufacturer's instructions (Ambion).Macrophage infectionsBone marrow-derived macrophages were isolated from C57BL/ mice and cultured for d in media containing % L-cell supernatant in the presence of antibiotics. Macrophages were stimulated with recombinant mouse IFN-γ at a final concentration of units/ml for h prior to infection. Macrophages were infected using DMEM containing % horse serum at a multiplicity of infection of , incubated for h, washed and fresh medium was added. At the indicated timepoints, M. tuberculosis RNA from inside macrophages was isolated and amplified as previously described [].Quantitative PCR M. tuberculosis and M. smegmatis from in vitro-grown log-phase cultures were pelleted and lysed by bead beating in Trizol as described above and total RNA was isolated [] and used for quantitative real-time PCR (qPCR) with the oligonucleotides specified (Table S7). The cDNA used for qPCR was generated with µg of total RNA using the Superscript III First Strand Synthesis for RT-PCR kit (Invitrogen). Standard curves were generated by measuring the concentration of each message in a reference sample of RNA pooled from all indicated conditions and timepoints analyzed for each experiment. All values reported are given as relative expression of each gene compared to 16S RNA (gene/16S).Supporting InformationFigure S1Venn diagrams illustrating the relationships between putative TA systems identified by three different algorithms utilizing combined homology-dependent and independent methods for finding TA loci in microbial genomes. Only predictions of complete TA pairs are included in this analysis and genes with multiple predicted partners are included only once. (A) Comparison between putative TA systems identified here, in Makarova, et al. [] and by RASTA-Bacteria [] using a strict RASTA-Bacteria cutoff score of >%. (B) Comparison between putative TA systems identified here, in Makarova, et al. [] and by RASTA-Bacteria [] using a strict RASTA-Bacteria cutoff score of >%. Gene lists and Venn diagram figures were generated using the web-based tools Venn Diagram Generator (http://www.pangloss.com/seidel/Protocols/venn.cgi) and Wybiral's Venn Diagram Generator (http://davy.wybiral.googlepages.com/venn.html), respectively.(. MB TIF)Click here for additional data file.Figure S2 recA is induced after treatment with ciprofloxacin. M. smegmatis harboring pHR100 was grown to early log phase and treated with . µg/ml ciprofloxacin. At , , , and h, ml aliquots were taken and RNA was harvested. The expression of recA was measured by qPCR. Three replicates are shown.(. MB TIF)Click here for additional data file.Table S1BLAST analysis to identify M. tuberculosis TA system homologs. Toxin and antitoxin system homologs used for PSI-BLAST analysis, M. tuberculosis BLAST hits and E-values are shown. For each homolog identified, we determined if an adjacent putative toxin/antitoxin was nearby and conserved domains present in these proteins are indicated. M. tuberculosis homologs that were significantly larger than amino acids were excluded.(. MB XLS)Click here for additional data file.Table S2Results of putative TA system testing. Putative toxin genes, along with the method of identification and toxicity results in M. smegmatis are shown. For genes that inhibited growth, the putative antitoxin was co-expressed. These genes were scored as those that relieved toxicity, and are part of a functional TA system (yes), and those that did not (no). The putative antitoxins that were used for testing are indicated in parentheses.(. MB DOC)Click here for additional data file.Table S3Conservation of TA systems across the genus Mycobacterium. The top BLAST hit in each organism is shown along with conserved antitoxin (if present). Reciprocal indicates whether the BLAST hit was the top reciprocal hit to the M. tuberculosis query toxin and synteny indicates whether the BLAST hit resides in a similar genomic context as the M. tuberculosis query toxin. For all strains other than M. leprae, top BLASTP hit is shown and top TBLASTX hit is shown for M. leprae. Putative antitoxin sequences shaded in grey indicate truncated sequences that are likely nonfunctional. Note that there is a possible duplication of Rv3384c in the M. canetti genome, but because the assembly used in this analysis is preliminary, this apparent duplication event may be due to an error in the draft assembly.(. MB XLS)Click here for additional data file.Table S4Expression results for all genes tested during hypoxia. Results of qPCR for each M. tuberculosis gene tested in two experiments after induction of NRP (hypoxia). Data is expressed as gene/16S and the standard deviation (SD) at each timepoint is shown. Timepoints at which we were unable to detect signal for a given gene are indicated (ND).(. MB DOC)Click here for additional data file.Table S5Expression results for all genes tested during macrophage infection. Results of qPCR for each M. tuberculosis gene tested in two experiments at and h after infection of IFN-γ-stimulated wild-type macrophages. Data is expressed as gene/16S and the standard deviation (SD) at each timepoint is shown.(. MB DOC)Click here for additional data file.Table S6Vectors, plasmids, and primers used in this study.(. MB DOC)Click here for additional data file.Table S7Primers used for qPCR of toxin, antitoxin, and control genes.(. MB DOC)Click here for additional data file.Text S1Supporting methods.(. MB DOC)Click here for additional data file.
|
PMC1959193.txt
|
TITLE: SNPs in Multi-Species Conserved Sequences (MCS) as useful markers in association studies: a practical approach AUTHORS: Jacob L McCauley, Shannon J Kenealy, Elliott H Margulies, Nathalie Schnetz-Boutaud, Simon G Gregory, Stephen L Hauser, Jorge R Oksenberg, Margaret A Pericak-Vance, Jonathan L Haines, Douglas P Mortlock ABSTRACT: BackgroundAlthough genes play a key role in many complex diseases, the specific genes involved in most complex diseases remain largely unidentified. Their discovery will hinge on the identification of key sequence variants that are conclusively associated with disease. While much attention has been focused on variants in protein-coding DNA, variants in noncoding regions may also play many important roles in complex disease by altering gene regulation. Since the vast majority of noncoding genomic sequence is of unknown function, this increases the challenge of identifying "functional" variants that cause disease. However, evolutionary conservation can be used as a guide to indicate regions of noncoding or coding DNA that are likely to have biological function, and thus may be more likely to harbor SNP variants with functional consequences. To help bias marker selection in favor of such variants, we devised a process that prioritizes annotated SNPs for genotyping studies based on their location within Multi-species Conserved Sequences (MCSs) and used this process to select SNPs in a region of linkage to a complex disease. This allowed us to evaluate the utility of the chosen SNPs for further association studies. Previously, a region of chromosome 1q43 was linked to Multiple Sclerosis (MS) in a genome-wide screen. We chose annotated SNPs in the region based on location within MCSs (termed MCS-SNPs). We then obtained genotypes for MCS-SNPs in individuals from MS families.ResultsAnalysis of our MCS-SNP genotypes from the 1q43 region and comparison to HapMap data confirmed that annotated SNPs in MCS regions are frequently polymorphic and show subtle signatures of selective pressure, consistent with previous reports of genome-wide variation in conserved regions. We also present an online tool that allows MCS data to be directly exported to the UCSC genome browser so that MCS-SNPs can be easily identified within genomic regions of interest.ConclusionOur results showed that MCS can easily be used to prioritize markers for follow-up and candidate gene association studies. We believe that this novel approach demonstrates a paradigm for expediting the search for genes contributing to complex diseases. BODY: BackgroundAdding to the challenge of disease gene discovery is the scale of genotyping that is required to conduct association studies in regions demonstrating linkage. Because the follow-up analysis of genomic linkage screen results often entails performing candidate gene analyses within a selected genomic region of interest, the number of necessary genotyping assays to thoroughly test candidate genes within each region quickly becomes very large and often cost prohibitive. New approaches to increase effectiveness of genotyping studies are clearly needed. With this in mind, we provide a novel approach that incorporates evidence from positional and informatic approaches to expedite follow-up of candidate regions. The utility of combining these approaches is evident in several recent studies for complex genetic disorders. For example, the emerging concept of "genomic convergence" suggests that parallel investigations of genetic linkage, association, and expression data will speed disease gene discovery []. Application of this process to identify and prioritize candidate genes on chromosome in Alzheimer disease and Parkinson's disease recently led to the successful identification of two genes significantly associated with these diseases [].An area of current interest in genomics and disease gene discovery concerns information that can be gained from assessing evolutionary sequence conservation between species. Such sequences that have remained similar across the millions of years of evolution are believed to indicate regions of biological function []. Researchers are increasingly integrating knowledge of conserved sequences in the selection of SNPs for genetic studies–whether it is to identify evidence of coding regions, potential splice sites, or regulatory regions [,]. In this capacity, conservation can be used as a putative annotation for genomic regions that may have functional importance, even when the precise nature of their function is lacking. This conserved sequence approach has several potential advantages. For example, since noncoding regulatory elements that control neighboring genes can be dispersed across large areas devoid of coding sequences, conserved elements might help discriminate functional regions within large noncoding areas that might not share linkage disequilibrium with coding markers [-]. Conservation can also indicate coding regions that may lack strong annotation support, such as alternatively spliced exons, RNA genes, "novel" genes with no homology to other gene families, or genes expressed at very low levels.A current challenge to genetic association studies is that it is usually not practical to genotype all SNPs in a region of interest. Linkage disequilibrium (LD) patterns can be exploited to identify subsets of "tag" SNPs that can serve as surrogates for detecting associations driven by functional SNPs that are in LD with tag SNPs; however, the statistical power of this approach depends on the degree of LD between markers, and decreases as LD decreases. Testing "functional" SNPs directly should help maximize the likelihood of detecting significant associations with disease phenotypes. By using conservation to prioritize SNPs, the odds may be increased that SNPs impacting the phenotype in question will actually be genotyped. For example, variation in transcriptional levels for key genes can play a significant role in disease risk [,]. SNPs in noncoding cis-regulatory sequences, such as enhancers, repressors, or chromatin structural regulators, might contribute to the genetic component of this process by modulating transcriptional output. It is reasonable to predict that SNPs within conserved regions may be more likely to have phenotypic effects than SNPs in nonconserved DNA.There are currently several publicly available tools to detect evolutionarily conserved sequences across large genomic regions by performing sequence alignments [,]. However, simple pair-wise sequence comparisons have drawbacks for use as a systematic approach in prioritization of conserved regions. With a relatively large region, sequence alignment between any two mammalian species can provide too much aligning sequence, resulting in the over-identification of sequences that are not actually preserved due to selective processes and are thus less likely to be functional []. Conversely, sequence alignment between more divergent species (e.g. between human and Fugu) can provide too little aligning sequence, resulting in the identification of only highly conserved protein-coding regions, while most noncoding regions are unalignable. Fortunately, new alignment methods that compare sequences from multiple species have been developed to minimize these drawbacks. By comparing sequence from three or more vertebrate species, human sequences that are likely to be functional can be detected with improved sensitivity and specificity [].Recently, an algorithm for detecting multi-species conserved sequences (MCS) has been optimized for scoring multi-species alignment data across large genomic regions []. This allows MCS scores to be assigned across any human genomic region, provided that sequences from multiple species have been aligned. While multiple species are used for comparison, a sequence block can be assigned a high MCS score even if not all the compared species show alignment to the region. This makes the MCS method robust when comparing draft genomes, and for identifying elements that may be conserved only within subgroups of species (e.g. within mammals only). In other words, an MCS may be classified as such even if it is not conserved across all the species in the comparison. MCS analysis incorporates genome sequence alignment data from multiple species in order to assign MCS scores to -base pair windows across the human genome. Using this tool, one can identify all sequences in a region of interest that have MCS scores above a defined threshold (e.g., in the top % of genome-wide MCS scores) []. Substantial portions of multi-species conserved sequences are in noncoding regions. Previous analyses suggest that approximately the top –% of MCS scores are very likely to indicate regions of human DNA undergoing evolutionary selection []. This threshold also detects the vast majority of coding exons, yet still detects many noncoding regions. This highlights the importance of MCS in detecting potentially functional sequences within relatively large genomic regions (e.g. several Mb) without relying solely on gene annotation.We have formulated a systematic approach to prioritize SNP markers to be genotyped for association studies in any target region, with a specific application to the 1q43 linkage region in MS [,]. This approach incorporates MCS analysis to prioritize markers within MCSs, termed MCS-SNPs, for high-throughput genotyping. Here we show that annotated SNPs in MCSs were readily identified and are frequently polymorphic. By comparing our findings with publicly available data from the HapMap Project [] we confirm that MCSs have a slightly reduced SNP density. We also provide an online tool and instructions to extract MCS-SNPs for any region of human DNA, via the UCSC genome browser. Our data is consistent with previous descriptions of human SNP variation in conserved regions [], suggesting the MCS tool is useful for identifying these SNPs. These findings indicate that conservation is a useful guide for selecting SNPs that may reside in biologically functional elements, particularly for those in noncoding regions.ResultsWe analyzed SNPs in a ~ .-Mb region of human 1q43 that showed suggestive linkage to multiple sclerosis (MS) [,]. This region was bounded by SNPs rs10925296 and rs1319790 (encompassing chr1:,,-,, of human Build ). We initially defined MCS in this region based on a "-way" species comparison (see Methods). The majority (/) of SNPs that we subsequently genotyped were chosen from within MCS elements (see Methods). An additional SNPs in nonconserved regions were added to provide additional coverage across the 1q43 region. In all, SNPs were within boundaries of known genes and the remaining markers were in intergenic regions. All SNPs were genotyped in our sample population of families ( individuals). Analyses to test for potential association(s) between these SNPs and multiple sclerosis will be presented elsewhere.In order to screen as many MCS-SNPs as possible, we selected MCS-SNPs for genotyping without regard to prior validation status. We therefore examined the fraction of our chosen MCS-SNPs and non-MCS-SNPs that demonstrated polymorphism in our study population (Table ). Approximately –% of SNPs within MCS regions were monomorphic in our study population. Although the percentage of monomorphic SNPs located within non-MCS regions was much less (–%), this was likely an artifact of our bias in selection of non-MCS-SNPs. Non-MCS-SNPs were chosen only if there was good prior bioinformatics evidence supporting their validation; however, prior validation was not a criteria used for selecting the MCS-SNPs.Table 1Descriptive breakdown of SNPs chosen for follow-up genotyping within the . Mb 1q43 regionOur Dataset4-way comparison (%)-way comparison (%)-way comparison (%)Total SNPs768 (%) (%) (%) MCS-SNPs478 (%) (%) (%) Polymorphic311 (%) (%) (%) Monomorphic167 (%) (%) (%) non-MCS-SNPs290 (%) (%) (%) Polymorphic268 (%) (%) (%) Monomorphic22 (%) (%) (%)During our initial study, more genomic sequence data from additional vertebrate species became available. Adding species can potentially increase the sensitivity and specificity of MCS analysis in detecting functional elements []. Therefore, we retrospectively compared the fraction of the MCS-SNPs defined by the "-way" species comparison that were still within MCS as defined by "-way" and "-way" species comparisons (see Methods, Table ). Approximately % of MCS-SNPs identified by "-way" comparison (N = ) were still within MCSs defined by the "-way" comparison (N = ).We examined the entire ~ .-Mb region in greater detail. Using data from the UCSC genome browser [] (Build ) we found over , distinct annotated SNPs in this region. Of all bases in the .-Mb region, .% were assigned MCS scores that meet the threshold for classification as MCS based on the "-way" comparison, indicating an MCS density similar to the genome-wide average (defined as the top %). However, only ~ .% of the annotated SNPs are within MCS elements. Thus, annotated SNPs are less dense in MCS versus non-MCS sequence. This is consistent with potential selection against variation within regions of high conservation, as observed in a previous whole-genome analysis [].These findings indicated that despite the modest reduction in SNP density within MCS, many polymorphic MCS-SNPs were validated in our sample population. Given the recent finding that in general, SNP minor allele frequencies (MAFs) are reduced in conserved sequence [], we examined HapMap data within the .-Mb interval to determine whether our MCS-SNPs exhibited reduced MAFs []. Rates of observed monomorphism were examined, as this could reflect either an excess of very rare alleles in MCS, or that truly polymorphic SNPs were more rare in MCS. Data from the HapMap CEU population ( trios of northern and western European ancestry collected by the Centre d'Etude du Polymorphisme Humain (CEPH)) was used for comparison, as our sample contained almost exclusively U.S. Caucasian individuals. We counted the total number of HapMap CEU SNPs (N = , SNPs) falling in MCS regions defined by "-way" comparisons (N = , SNPs), and then further subdivided these based on location in or outside exons and whether they were polymorphic or monomorphic in the HapMap CEU population (Table ). Within the exons of this region, MCS-SNPs had an average MAF of . and % monomorphism rate, while non-MCS-SNPs in exons had an average MAF of . and % monomorphism rate. Thus, as expected exonic MCS-SNPs had reduced MAF and increased monomorphism. Since variation is known to be reduced in coding sequence, we examined the non-exonic SNPs. Similar trends were observed, though less pronounced than in exonic SNPs. Non-exonic MCS-SNPs had an average MAF of . and monomorphism rate of %, while non-exonic/non-MCS SNPs had an average MAF of . and a % monomorphism rate. The difference in MAFs was not significant by chi-square test (not shown), although the trend reflects genome-wide observation []. However, the increased rate of monomorphism of MCS-SNPs in non-exonic regions was significant (Table ).Table 2Descriptive breakdown of HapMap CEU SNP data across the 1q43 . Mb regionAll annotated SNPs (Build ) Chr1: -240494277CEU population: genotyped SNPs (-way comparison)Total SNPs2824311076*exonic SNPs20874 MCS-SNPs13047 PolymorphicNA23 (MAF = .) MonomorphicNA24 non-MCS-SNPs7827 PolymorphicNA22 (MAF = .) MonomorphicNA5non-exonic SNPs2803511002 MCS-SNPs776414 PolymorphicNA259 (MAF = .) MonomorphicNA155 non-MCS-SNPs2725910587 PolymorphicNA7241 (MAF = .) MonomorphicNA3346Table 3Comparison of monomorphic vs. polymorphic non-exonic SNPs, in HapMap CEU (-way MCS vs. non-MCS)Polymorphic SNPsMonomorphic SNPsChi-squarep-valueNon-exonicMCS2591556..012non-MCS72413346ExonicMCS23247..006non-MCS225An MCS calculator is now available to provide easier access to genome-wide MCS data []. This resource allows MCS elements for a human genomic region of interest to be rapidly exported to the UCSC genome browser as a "custom" track, such that SNPs within MCS can be quickly identified (Figure ). Lists of MCS-SNPs can then be easily retrieved, or visualized on the browser (Figure ). Detailed instructions, for navigating the WebMCS and UCSC websites in order to extract MCS-SNPs, are provided [see Additional File ].Figure 1Scheme for identifying MCS-SNPs.Figure 2UCSC genome browser image showing position of MCS-SNPs in the vicinity of the EXO1 gene. After obtaining the subset of SNPs within MCS, the Table Browser function allows MCS-SNPs to be downloaded in table format or visualized as a custom track on the browser as shown here. Note several MCS-SNPs are in EXO1 ' flanking region, and in introns (e.g. rs4149896, rs2526698) as well as those in exons.Since noncoding MCS-SNPs could affect gene regulation, it may be of interest to prioritize analysis of SNPs that disrupt putative transcription factor binding sites (TFBS). The UCSC browser "TFBS Conserved" track displays predicted TFBS present in regions of human/mouse/rat conservation []. This track could be used as an independent SNP filtering mechanism; alternatively, as many predicted sites overlap with MCS it could easily be used to further prioritize MCS-SNPs. For example, in the Mb chromosome region described here conserved TFBS sites are predicted, of which (%) are within -way MCS regions. For any region of interest, a file of MCS-SNPs could be extracted as a UCSC-ready custom track using the UCSC Table Browser [see Additional File ], and immediately used to identify the subset of MCS-SNPs that overlap conserved TFBS. In the chromosome region HapMap SNPs were found to overlap TFBS; of these, are also within -way MCS.Actual coverage of MCS-SNPs in genotyping studies depends on the coverage of working genotyping assays. For the latter reason, the coverage of MCS-SNP assays in existing whole-genome SNP genotyping platforms is of interest. For both chromosomal regions, we determined the fraction of MCS-SNPs that are in existing Affymetrix K SNP chips and in the Illumina K assay set (Table ). Interestingly, coverage for these assay sets ranges from roughly –% but was roughly twice as great in the chromosome region, independent of platform. Therefore, the large majority of MS-SNPs are not directly represented in these genotyping platforms. For genotyping genomic regions of interest, the effective association coverage of MS-SNPs with these assay sets will be a function of the MS-SNP representation for the assay set used, the ability of assays in the set to "tag" LD across markers, and the LD structure of the population in question.Table 4Coverage of -way MCS-SNPs in whole-genome assay setsChr1: ,,-,,277Chr8: ,,–,,000Total MCS-SNPs906382within MCS:Affymetrix K Nsp SNPs31 (.%) (.%)Affymetrix K Sty SNPs21 (.%) (.%)Total Affy500 K52 (.%) (.%)Illumina K39 (.%) (.%)Since LD can be used to select tag SNPs that provide surrogate information for groups of SNPs having shared LD, it is reasonable to question whether prioritization of MCS SNPs is useful if LD tagging strategies can "cover" MCS regions effectively. To assess this, we used the Tagger tool [] through the HapMap website [] to obtain HapMap CEU tag SNPs for the Mb region (using MAF > = ., r2 > = .). This analysis was also performed for a randomly chosen Mb region on chromosome roughly centered on the GDF6 gene (build hg17, chr8: ,,–,,). This region contains a higher gene density than the Mb chromosome region ( RefSeq annotated genes; ~ kb/gene) but slightly lower -way MCS coverage (.%). For the chromosome and regions, respectively, and SNPs fell in MCS; of these, (%) and (%) did not share LD with any other SNPs and were thus not captured by other tags. Therefore, while a large majority of HapMap-genotyped MCS-SNPs are captured by tag SNPs obtained by these parameters from the CEU data set it may be of interest to augment tagging strategies by including "uncaptured" MCS-SNPs in genotyping studies.DiscussionHere we show that polymorphic SNPs can be readily identified within conserved noncoding regions, and that SNP selection based on conservation data is an approach that can be readily incorporated into genotyping projects. Genetic association studies have typically placed heavy emphasis on coding SNPs, with the idea that these are more likely to have biological impact than noncoding SNPs. SNP analysis in noncoding DNA has typically been restricted to promoter regions, though conservation in more distant noncoding regions is a widespread observation. Our approach prioritizes SNPs in conserved noncoding and coding regions, since conservation is widely accepted to correlate with function in either type of sequence.Although the density of MCS in the ~ .-Mb region is similar to the genome-wide average, the number of exonic bases and total genes per kb are below genome averages. The region contains genes based on RefSeq annotation [], which is far below the genome average of approximately genes per Mb []. Notably, this region contains several large genes (RYR2, kb; RGS7, kb; PLD5, kb) as well as a "gene desert" of ~ .-Mb. Based on RefSeq annotation, .% of genomic bases in the interval are spanned by transcription units, and .% of bases are within exons, less than the estimated genome-wide fraction of coding bases (.%) []. Even with the conservative assumption that % of exon bases fall within MCS elements, at least % of MCS bases must be in noncoding MCS regions (see Results). Since not all exon bases are in MCS regions, this is a minimal estimate. Thus, relative to the whole genome the .-Mb interval has slightly below average density of genes and exonic bases, but has a fraction of noncoding MCS that is similar to genome-wide estimates of noncoding conservation [].There are two similar hypotheses at play when discussing variation in regions of conservation. The first is that regions of conserved DNA sequence, whether coding or noncoding, are likely to have biological function(s). The extension of this idea is that when examining these regions in suspected disease genes, functional variations will be discovered that predispose to disease risk. However, a second hypothesis is that because conserved regions are indeed likely to have important functions, variants that persist within these regions are actually those that are less likely to have functional effects, as variants that impact function are likely to reduce fitness and be subjected to negative selection pressure. This stems from the concept that most perturbations in conserved sequence will reduce reproductive fitness. These conflicting hypotheses can be somewhat reconciled by the notion that many genetic variants that predispose to complex, common diseases in the present day may have had negligible effects on reproductive fitness in recent human evolution. Regions of high conservation harbor biological function and therefore are likely to be under selective constraint. New variants in these regions are often detrimental and thus less likely to persist in populations, which may explain the increased rate of apparent monomorphism (or rare alleles) for MCS markers. However, some persisting MCS variants may have phenotypic effects that influence disease risk. Variants that alter potential TFBS are particularly interesting and may be worthwhile to prioritize, although the density of such MCS-SNPs was on the order of one per gene in the chromosome region. However, since actual binding sites can deviate from the predicted consensus and not all factor binding motifs are known, not all "functional" SNPs that disrupt true binding sites will be identified with this approach.For certain studies, different combinations of species might be more desirable for classifying MCS. For example, inclusion of additional mammals might be useful for detecting functional elements that are not expected to be present in other vertebrates. In fact, diseases that involve the immune response (such as MS or lupus) might be controlled by genes with a more recent or dynamic evolutionary history. We note that users can define MCS with custom-generated sequence alignments based on species data of their choice, using the WebMCS online resource [].Since many MCS-SNPs are not represented in widely used whole-genome genotyping platforms, the inclusion of MCS-SNPs is particularly for candidate gene regions where dense coverage of potentially functional SNPs is of greater interest. For candidate genomic regions (or populations) that are characterized by low linkage disequilibrium, prioritizing MCS-SNPs may have additional value, as markers are less likely to be effectively captured by tagging strategies.ConclusionOur results are consistent with recent reports that SNP density and derived (new) SNP allele frequencies are slightly reduced in noncoding conserved regions as compared to nonconserved regions []. Our findings also confirm that inferences from genome-wide conservation data can be usefully applied to SNP selection for fine-mapping genetic studies. The MCS-SNP approach represents practical knowledge for choosing SNP markers and can be integrated with current approaches for genotyping studies to refine or follow-up regions or genes believed to be involved in disease.MethodsSubjects and phenotypesThis study involved genotyping SNPs in a dataset of multiple sclerosis families ( individuals genotyped). Families were ascertained at the University of California at San Francisco (UCSF) as previously described []. Of these families, families had previous evidence for positive linkage to 1q43 []. Informed consent was obtained for all subjects. This research was performed under protocol # as approved by Vanderbilt University Internal Review Board.Molecular analysis768 SNPs were chosen for genotyping by selecting SNPs from the UCSC genome browser with significant preference given to markers with an MCS score exceeding the top % threshold (see below). The Illumina BeadArray™ platform was used for SNP genotyping. All genotyping was performed by the Duke Genomics Resource Laboratory Core using the Illumina BeadArray™ platform.For this study we used MCS scores that fall in the top % of genome-wide MCS scores for -base pair sequences [], based on "-way" genome-wide alignment between human, mouse, rat, and chick genomes (see below). Selection of SNPs in conserved regions was based on informativeness (with preference given to SNPs with high minor allele frequencies), validation (preference to SNPs confirmed by multiple lines of evidence), location (preference to SNPs spaced at regular intervals), putative function (preference to SNPs in coding, splice site, and mRNA UTR regions), and assay scores provided by Illumina (preference to SNPs generating scores > .). Illumina scores were determined by an algorithm weighing a series of factors to predict the success of each locus within an OPA. Scores ranged between and , with Illumina recommending selection of SNPs generating scores > ..Several SNPs located in % MCS regions were eliminated from the study due to the nature of the variation precluding genotyping on the Illumina platform (e.g. insertion/deletions or multiple mutation events leading to > alleles) and/or the failure to generate an Illumina score > .. Because elimination of these SNPs resulted in the identification of fewer than SNPs in conserved regions, additional SNPs were selected from nonconserved regions, with similar consideration for suitability to the Illumina platform and to create an average spacing of < kb for the SNPs in the ~ .-Mb region of interest. This density was chosen to make most efficient use of the Illumina platform and may help to exploit linkage disequilibrium for future association studies, based on LD patterns in Caucasians []. Most markers fell within intronic and intergenic areas.Statistical analysisWe initially identified multi-species conserved sequences (MCSs) through a "-way" alignment of mouse, rat, and chick genomic sequence to human chromosome 1q43 sequence (Build , chr1:,,-,,) [,]. MCS regions were defined as those regions with MCS scores falling in the top % of genome-wide scores. This allowed us to identify SNPs from MCS regions that were amenable to genotyping on the Illumina platform as described above. Since the time we initially chose the markers, additional genome sequences for other model organisms have been produced. Therefore we also analyzed our SNPs using new MCS scores derived from both "-way" alignments of genomic sequences (human, mouse, rat, chick, and Fugu) and "-way" species alignments (human, chimp, mouse, rat, dog, chick, Fugu, and zebrafish) (Margulies, unpublished). Currently, genome-wide MCS data based on the "-way" species comparison can be obtained online [].HapMap CEU data from the international HapMap project, were used to examine SNP data in the 1q43 region (HapMap Data Rel#/phaseII Jan06, on NCBI B35 assembly, dbSNP b125; []). Comparisons of numbers of monomorphic and polymorphic markers in each group were made using a Chi-square test for significance (Table ). The UCSC browser TFBS conserved track was implemented with a Z score cutoff of . [].Authors' contributionsJLM analyzed genotype and HapMap data and prepared the draft manuscript. SJK helped conceive the study design, selected SNPs and conducted genotyping. NSB and SGG helped coordinate and direct sample handling and genotyping. EHM created the WebMCS tool. SLH and JRO directed patient ascertainment and oversaw blood sample collection. MPV, JLH, and DPM helped conceive and organize the project and helped draft the manuscript.Supplementary MaterialAdditional file 1Supplementary instructions on how to determine, view, and extract MCS-SNPs for a region of interest. This file provides a step-by-step tutorial for obtaining MCS-SNPs from a given region of the human genome.Click here for file
|
PMC2744653.txt
|
TITLE: Effect of traffic pollution on respiratory and allergic disease in adults: cross-sectional and longitudinal analyses AUTHORS: Mar Pujades-Rodríguez, Tricia McKeever, Sarah Lewis, Duncan Whyatt, John Britton, Andrea Venn ABSTRACT: BackgroundEpidemiological research into the role of traffic pollution on chronic respiratory and allergic disease has focused primarily on children. Studies in adults, in particular those based on objective outcomes such as bronchial hyperresponsiveness, skin sensitisation, and lung function, are limited.MethodsWe have used an existing cohort of adults aged – living in Nottingham, UK, for whom baseline health and demographic data were collected in and computed two markers of exposure to traffic: distance between the home and nearest main road and modelled outdoor nitrogen dioxide (NO2) concentration at the home location. Using multiple regression techniques, we analysed cross-sectional associations with bronchial hyperresponsiveness, FEV1, spirometry-defined COPD, skin test positivity, total IgE and questionnaire-reported wheeze, asthma, eczema and hayfever in subjects, and longitudinal associations with decline in FEV1 in subjects followed-up nine years later in .ResultsThere were no significant cross-sectional associations between home proximity to the roadside or NO2 level on any of the outcomes studied (adjusted OR of bronchial hyperresponsiveness in relation to living ≤ m vs > m from a road = ., % CI . to .). Furthermore, neither exposure was associated with a significantly greater decline in FEV1 over time (adjusted mean difference in ΔFEV1 for living ≤ m vs > m of a road = . ml, % CI, -. to .).ConclusionThis study found no evidence to suggest that living in close proximity to traffic is a major determinant of asthma, allergic disease or COPD in adults. BODY: BackgroundMany epidemiological studies have examined effects of exposure to road vehicle traffic on chronic respiratory and allergic disease in children, but research of the effects in adults is limited. A handful of studies of adults have reported that living in close proximity to busy or major roads is associated with an increased risk of wheeze [-], whilst others have shown no effect on wheeze or asthma [-]. Some of this inconsistency may be due to the use of self-reported markers of asthma which are potentially biased, but use of objective markers such as bronchial hyperresponsiveness (BHR) is rare. Lung function measures such as one second forced expiratory volume (FEV1) have been investigated by a few in relation to traffic indices such as proximity to main roads or modelled traffic-related pollutants, but again findings in adults are inconclusive and evidence of longitudinal effects lacking[]. Spirometry has also been used to define chronic obstructive pulmonary disease (COPD) in one study of women which reported an adverse effect of close residential proximity to a busy road[]. Investigations of allergy and atopy in adults have also tended to rely on self-reported outcomes, and use of objective markers such as skin sensitisation or elevated immunoglobulin E (IgE) rare[,].We have therefore used data from an existing population-based cohort of adults to compute markers of exposure to traffic-related pollution and investigate their relation with objective measures of respiratory and allergic disease, namely bronchial hyperresponsiveness, FEV1, spirometry-defined COPD, skin test positivity and total IgE, as well as questionnaire reported wheeze, asthma, eczema and hayfever. In addition to these cross-sectional investigations, we have also used longitudinal measurements made on the cohort to examine effects of exposure on change in FEV1. As the primary traffic related air pollutants are highest at the roadside and decline exponentially[], we have chosen to use distance between home residence and the nearest major road as an objective proxy of exposure to traffic pollution, and to look particularly at dose-response relations across the first m from the roadside where most of the decline occurs. We have also used an alternative marker of exposure based on modelled traffic-related NO2 at the home location, which may better reflect the level and type of traffic. Data on number of years residence in the current home has also enabled us to look more specifically at long-term effects of exposure by restricting analyses to long-term residents only.MethodsStudy populationOur study population is a cohort of adults aged – living in the Gedling area of Nottingham City, UK, who were recruited in as part of a study of the effect of diet on chronic lung disease. Gedling is an area of square miles with an estimated population of , in which covers the north east suburbs of Nottingham and surrounding rural villages. Full details of the study have been described elsewhere[]. Briefly, a representative sample of adults was drawn from the local electoral register and those of eligible age and residing in the study area were invited to take part in the study (figure ). Information on respiratory and allergic disease symptoms, demographics, smoking, diet, and numerous other lifestyle factors were collected using an interview-led questionnaire. FEV1 and forced vital capacity (FVC) were measured by a study nurse using a calibrated dry bellows spirometer (Vitalograph, Buckingham, UK) taking the best of three technically satisfactory manoeuvres with the subject seated; a methacholine challenge performed to determine BHR using the technique described by Yan et al[]; allergen skin tests to Dermatophagoides pteronyssinus, mixed grass pollen, cat fur, Aspergillus fumigatus, and Cladosporium herbarum (Bencard solutions, Brentford, UK) carried out; and a blood sample taken. In total, , individuals participated in the survey, estimated to be between % and % of those eligible [] (figure ). The exact response rate cannot be computed since it was not known what proportion of the non-responders would have been eligible for inclusion. In all surviving individuals were invited to participate in a follow-up survey (follow-up rate was %) in which these measurements, with the exception of BHR, were repeated[]. The surveys were approved by the Nottingham City Hospital and Nottingham University ethics committees.Figure 1Flow diagram showing study participation.Computation of exposure variablesWe took each subjects home address in and converted it into a grid reference using information from the Local Land & Property Gazetteers database held by Gedling Borough Council. To compute our distance exposure variable, we linked this grid reference to a digitised road map of Great Britain (Meridian database; Ordnance Survey, Southampton, UK), which is a geometrically structured :, scale vector database with a coordinate resolution of m. We then calculated the shortest distance (in metres) between each address location and the nearest major road, defined as a motorway (freeway), or 'A' or 'B' class road (principal road as classified by UK Department for Transport), using Geographical Information System (GIS) software (ArcGIS .).To compute our modelled NO2 variable, we linked each home location grid reference to a high resolution map of modelled traffic-related NO2 using ArcGIS. This map was generated by the dispersion model ADMS Roads (CERC, Cambridge, UK), which has been widely used to assess the impact of road traffic on local air quality and extensively validated against monitored roadside concentrations of traffic pollutants[]. Traffic count and composition data supplied by Nottinghamshire County Council and hourly sequential meteorological data (including wind direction and speed) provided by the UK Meteorological Office were inputted into the model. Background concentration data were averaged from pollutant data recorded at automated sites in Nottingham City Centre (urban) and Sutton Bonnington (rural). Modelled annual mean concentrations of NO2 were verified through comparison with observed concentrations (–) recorded at diffusion tube survey sites (background, intermediate and roadside) across the study area and overall modelled values correlated well with observed (r = .). Closer inspection revealed some underestimation of observed concentrations at roadside sites close to major road intersections where slow moving or standing traffic is likely during rush hours. In the absence of more detailed information on traffic count, speed and composition the model is unable to reproduce elevated concentrations at such localized hot spots. However given the good level of agreement between modelled and observed concentrations elsewhere (r = . for the other sites), and the fact that only a minority of the study population would live close to such major intersections, the model parameterisation was deemed satisfactory to invoke deployment across the entire study area at a spatial resolution of metres.Statistical analysesIn cross-sectional analyses we assessed the effect of each exposure variable on each of the following outcome variables using data from the baseline survey: self-reported wheeze in the past year ('Have you had wheezing or whistling in your chest at any time in the last months?'), diagnosed asthma ('Have you ever had asthma?' and 'Was this confirmed by a doctor?'), eczema ever ('Have you ever had eczema or any kind of skin rash?') and hayfever ever ('Have you ever had hayfever or other nasal allergies?'); bronchial hyperresponsiveness (BHR), defined as a methacholine dose provoking a % fall in FEV1 (PD20) of . mg/ml or less; COPD, defined as stage I or above using the GOLD (Global Initiative for Chronic Obstructive Lung Disease) criteria[] (namely having an FEV1/FVC less than %); allergen skin sensitisation, defined as a response to any of the allergens tested at least mm greater than the saline control response in the presence of a positive histamine control; and high total IgE, defined as a concentration above kU/l. Multiple logistic regression analyses were carried out to assess the effect of distance, initially treated as a binary variable (<= m and > m), on each binary outcome, adjusted for age, sex, smoking status (current, never or ex-smoker) and quintiles of Carstairs deprivation score (postcode based index of deprivation based on unemployment, overcrowding, car ownership and occupation[]). To investigate dose-response effects, analyses were then carried out in the subset of respondents living within m from a road with distance fitted as m bands (the smallest categorisation possible given the study sample size). Modelled NO2 was categorised into quintiles and analysed in a similar way to distance.The cross-sectional association between each exposure and FEV1, defined as the best of three satisfactory measurements, was analysed using multiple linear regression controlling for sex, age, age squared, smoking status, pack-years of cigarettes smoked, height, Carstairs deprivation score and age-height interaction. Multiple linear regression was also used to analyse total IgE as a continuous variable (transformed to achieve a Normal distribution by adding and taking the logarithm), with adjustment for age, sex, smoking and Carstairs score.In longitudinal analyses, we computed changes in adjusted FEV1 residuals between and using the method described by Carey et al[]. Briefly, we modelled predicted FEV1 values for each sex separately in the sub-group of non-smoking, non-asthmatic, non-wheezing individuals, with terms for age, height, age squared, and age-height interaction, and calculated the difference from predicted as the adjusted FEV1 residuals in and . Change in FEV1 residual from to was modelled in relation to the exposure variables using multiple linear regression, controlling for the effect of smoking and number of pack-years of cigarettes in and for Carstairs deprivation score as recorded at baseline.For all analyses we tried adjusting for other potential confounders including exposure to passive smoking, older siblings, indoor heating and cooking appliances, pet ownership, vitamin C and E intake, and body mass index. We also tried controlling for social class based on occupation as an alternative marker of socio-economic status to Carstairs score. Since long-term exposure may be most relevant and inclusion of others could weaken any associations, we repeated all cross-sectional analyses restricting to the subgroup of individuals who had lived in the same address for at least years (long-term residents), and longitudinal analysis restricting to individuals who lived in the same address between and , to assess whether the magnitude of effect estimates were altered.All analyses were performed in STATA . for Windows (Stata Corportation, college Station, Texas). The available sample provided % power to detect an odds ratio (OR) of . for wheeze (based on an outcome prevalence of %) in relation to living within m of a road relative to more than m (based on an exposure prevalence of %).ResultsPopulationOf the , participants in the baseline survey, we excluded with invalid or incomplete address information and one who had not provided full lung function data. Cross-sectional analyses were therefore carried out on , subjects (%), of whom , (.%) were long-term residents (median length of residence years; interquartile range to years). For the longitudinal analyses we excluded without valid address information and without lung function data, leaving , individuals (.% of participants in the follow-up survey) for analysis, of whom (.%) had resided at the same address over the period of follow-up. The baseline characteristics of the study subjects and those lost to follow-up are shown in Table . Those followed up in were generally similar to the baseline group, although slightly less likely to be a smoker or in the youngest age group. Although the characteristics of our original adults sampled from the electoral role are not known, Table shows how the age, sex and social deprivation (Carstairs) distribution of our participants compares with that of all Gedling residents in , our target population, using census data from that year [] (Table ). This shows that those included in our cross-sectional and longitudinal analyses were slightly older than the target population but similar with respect to gender and social deprivation (Table ). Cross-sectional analyses of BHR and IgE were carried out on slightly smaller datasets due to missing data (figure ), but demographics were similar to the complete dataset of subjects (Table ).Table 1Baseline characteristics of Gedling study participants included in cross-sectional and longitudinal traffic pollution analyses and those lost to follow-upBaseline characteristicsCross-sectional analysis(N = ) Number (%)Longitudinal analysis(N = ) Number (%)Lost to follow-up(N = ) Number (%)Age group18– (.) (.) (.)– (.) (.) (.)– (.) (.) (.)– (.) (.) (.)– (.) (.) (.)SexMen1299 (.) (.) (.)Women1300 (.) (.) (.)Carstairs deprivation quintiles1st (least deprived) (.) (.) (.)2nd457 (.) (.) (.)3rd523 (.) (.) (.)4th482 (.) (.) (.)5th (most deprived) (.) (.) (.)Smoking statusNever1289 (.) (.) (.)Ex-smoker722 (.) (.) (.)Current smoker588 (.) (.) (.)Distance to a main road< m164 (.) (.) (.) – m226 (.) (.) (.) – m190 (.) (.) (.)> m2019 (.) (.) (.)Modelled quintiles of NO2 (μg/m3)<. (.) (.) (.). – . (.) (.) (.). – . (.) (.) (.). – . (.) (.) (.)>. (.) (.) (.)Outcomes*Current wheeze624 (.) (.) (.)Diagnosed asthma231 (.) (.) (.)COPD** (.) (.) (.)BHR† (.) (.) (.)Ever hay fever662 (.) (.) (.)Ever eczema714 (.) (.) (.)Allergen skin sensitisation788 (.) (.) (.)High total IgE‡ (.) (.) (.)NO2 nitrogen dioxide; COPD chronic obstructive pulmonary disease; BHR bronchial hyper-responsiveness; IgE immunoglobulin E.* For outcomes, values represent number and % positive where %'s are computed excluding any missing values**COPD: missing values in cross-sectional study (% = /) and in longitudinal (% = /)† BHR: missing values in cross-sectional study (% = /) and in longitudinal (% = /)‡ IgE: missing values in cross-sectional study (% = /) and in longitudinal (% = /)Table 2Comparison of Gedling study participants included in data analyses with Gedling census populationCharacteristic1991 Gedling census popn[]*Main cross-sectional analyses N = 2599BHR analysisN = 2383IgE analysisN = 2467Longitudinal analysis of FEV1N = 1329Age group (%)–.....–.....–.....–.....–.....7Sex (%)Men49.....6Women50.....4Mean (SD) Carstairs deprivation score-. (.)-. (.)-. (.)-. (.)-. (.)BHR bronchial hyper-responsiveness; IgE immunoglobulin E.*Age and sex %'s based on Gedling census population aged – (n = ); mean Carstairs based on total Gedling census population (n = )Census output is Crown copyright and is reproduced with the permission of the Controller of HMSO and the Queen's Printer for Scotland.Just under one quarter of individuals (.%, n = ) lived within m of a main road and the median level of modelled NO2 exposure was . μg/m3 (IQR . to .). The concentration of modelled NO2 was significantly related to residential proximity to roads (chi square p-value < .) such that those living within m of a road had a median level of . μg/m3 (IQR . to .) decreasing to . μg/m3 (IQR . to .) for those living more than m away.Effect of proximity to a major road on respiratory and allergic outcomesAfter adjusting for potential risk factors, respondents living within m of a major road were not more likely to have BHR, COPD, positive skin test or high total IgE, or self-reported wheeze, than those living further away (Table ). For wheeze and allergen sensitisation there was weak evidence of a positive dose-response relation across the first m from the roadside (p for trend = . and . respectively), but not for the other outcomes. There were no significant associations between proximity and questionnaire-reported asthma, eczema and hay fever (adjusted OR (% CI) . (. to .), . (. to .) and . (. to .) respectively).Table 3Association between residential proximity to a main road and respiratory and allergic outcomesOutcomesNumber with outcome (%)Adjusted OR*(% CI)p-value†Wheezing in last year≤ m129 (.). (. – .).> m495 (.)1Bands of distance < m46 (.). (. – .). – m45 (.). (. – .) – m38 (.)1COPD≤ m53 (.). (. – .).> m185 (.)1Bands of distance < m16 (.). (. – .). – m24 (.). (. – .) – m13 (.)1BHR≤ m64 (.). (. – .).> m246 (.)1Bands of distance < m13 (.). (. – .). – m25 (.). (. – .) – m26 (.)1Allergic sensitisation≤ m163 (.). (. – .).> m625 (.)1Bands of distance < m51 (.). (. – .). – m70 (.). (. – .) – m42 (.)1High total IgE≤ m109 (.). (. – .).> m440 (.)1Bands of distance < m31 (.). (. – .). – m43 (.). (. – .) – m35 (.)1OR odds ratio; CI confidence interval; COPD chronic obstructive pulmonary disease; BHR bronchial hyper-responsiveness; IgE immunoglobulin E.* OR adjusted for sex, age group, Carstairs deprivation scores and smoking status† p-value for trend across bands of distanceEffect of modelled NO2 on respiratory and allergic outcomesThere was no evidence that increasing modelled NO2 at the home was related to an increase in the risk of wheeze, COPD, BHR, skin sensitisation or high IgE (Table ). Similarly, questionnaire reported asthma, eczema and hayfever were not significantly related to modelled NO2 (adjusted OR for highest versus lowest quintile . (. to .), . (. to .), and . (. to .), respectively).Table 4Association between modelled NO2 level and respiratory and allergic outcomesQuintiles of modelled NO2 (μg/m3)Number with outcome (%)Adjusted OR* (% CI)p for trendWheezing in last year <. (.). . – . (.). (. – .) . – . (.). (. – .) . – . (.). (. – .) >. (.). (. – .)COPD <. (.). . – . (.). (. – .) . – . (.). (. – .) . – . (.). (. – .) >. (.). (. – .)BHR to methacholine <. (.). . – . (.). (. – .) . – . (.). (. – .) . – . (.). (. – .) >. (.). (. – .)Allergen skin sensitisation <. (.). . – . (.). (. – .) . – . (.). (. – .) . – . (.). (. – .) >. (.). (. – .)High total IgE <. (.). . – . (.). (. – .) . – . (.). (. – .) . – . (.). (. – .) >. (.). (. – .)NO2 nitrogen dioxide; OR odds ratio; CI confidence interval; COPD chronic obstructive pulmonary disease; BHR bronchial hyper-responsiveness; IgE immunoglobulin E.* OR adjusted for sex, age group, Carstairs deprivation score and smoking statusEffect of traffic pollution on lung function measurementsIn cross-sectional analyses, those living within m of a main road were seen to have a similar FEV1 to those living further away, and amongst those living within m of a road, there was no trend of reduced FEV1 with increased proximity (Table ). Similarly, there was no association between measured values of lung function and modelled quintiles of NO2 at home location.Table 5Effect of proximity and modelled NO2 on cross-sectional FEV1 and longitudinal change in FEV1Mean (unadjusted)SDβ*% CIp-value†Cross-sectionalDistance (N = )≤ m3202...-. to ..> m3178..10Bands of distance (N = ) < m3143...-. to .. – m3230..-.-. to . – m3220..90Quintile of NO2 (N = ) <.... . – ...-.-. to . . – ...-.-. to . . – ....-. to . >....-. to .25Longitudinal‡Mean change (unadjusted)SDβ**% CIp-value†Distance (N = )≤ m-...-. to ..> m-..00Bands of distance (N = ) < m-...-. to .. – m-...-. to . – m-..70Quintiles of NO2 (N = ) <.-... . – .-..-.-. to . . – .-..-.-. to . . – .-..-.-. to . >.-..-.-. to .08SD standard deviation; β adjusted regression coefficient; CI confidence interval; NO2 nitrogen dioxide.* β coefficient for cross-sectional FEV1 adjusted for sex, age, age squared, height, age-height interaction, smoking status, cigarette pack-years and Carstairs deprivation score, as recorded at baseline** β coefficient for longitudinal change in FEV1 residuals adjusted for smoking status and cigarette pack years in and Carstairs deprivation score as recorded at baseline† p-value for trend across quintiles of modelled NO2 and across bands of distance effects‡ Change in lung function between and 2000In longitudinal analyses of lung function over the years of follow-up, decline in FEV1 was similar for those living within m from the roadside and those living further away and showed no trend with proximity amongst those living within m (Table ). Similarly, there was no significant association between modelled NO2 and change in FEV1 (Table ).Further analysesFurther control for other potential confounders, or occupation-based social class as an alternative to Carstairs deprivation score, did not materially alter any of the results, and restriction of cross-sectional and longitudinal analyses to the sub-group of long-term residents made little difference to the estimates. When total IgE was analysed as a continuous variable rather than a binary variable, no significant associations were seen with distance (adjusted β = -., % CI -. to . for <= m relative to > m, p = .) or modelled NO2 (p for trend = .) (Table ).Table 6Effect of proximity and modelled NO2 on Total IgEExposure variableNMedian Total IgE kU/l (IQR)Adjusted regression coefficient* (% CI)P-valueDistance<= m55327 (–)-. (-., .).> (–.)0Bands of distance< m15323 (–). (-., .).– m21930 (–). (-., .)Ptrend = .> m18127 (–.)0Quintiles of NO2 (μg/m3)<.. (–)..–. (–.)-. (-., .)Ptrend = ..–. (.–.)-. (-., .).–.. (.–.)-. (-., .)>. (–)-. (-., .)IQR Interquartile range* β coefficient from linear regression analysis of log transformed IgEDiscussionIn this population-based study of Nottingham adults, we found no evidence that living close to a main road or in an area of increased traffic-related pollution was associated with an increased risk of asthma or COPD. This was true for both self-reported markers such as disease symptoms and diagnosis, and objective markers: BHR and lung function. Furthermore, in longitudinal analyses, there was no evidence that increased traffic exposure was associated with decline in lung function. We found some suggestion of an adverse effect of home proximity on allergy with a significant exposure-response across the first m from roadside for allergic sensitisation, but not for other markers such as hayfever, eczema or total IgE.The response rate, both to the original cross-sectional survey and in the year follow-up of these subjects, was only approximately %, which raises the possibility of response bias. In the follow-up study, the characteristics of those who participated were generally similar to the original sample, and in particular, participation rates did not differ according to proximity to a main road. Whilst the factors associated with participation in the original survey are not fully known, proximity to a main road is not likely to be one of them since respondents and interviewers were unaware of the current hypothesis of investigation. We did find evidence that our study participants were slightly older than Gedling residents in the census[], but proximity was not associated with age in our dataset (r = -.); socio-economic status was comparable to the census population and again was not related to proximity in our dataset (r = -.). Therefore, whilst we cannot completely rule out the possibility of response bias, it is unlikely to have had a major impact on our study results.A strength of this study is that our exposure variables were computed using GIS techniques from the participant's exact address rather than postcode used in many previous studies. We estimated that by using the postcode rather than the exact address coordinates in the computation of the binary and the m band distance variables, exposure status would have misclassified % and % of respondents respectively. In addition to the commonly used marker of exposure, residential distance from major roads, we used a more sophisticated marker of exposure based on modelled traffic-related NO2 concentrations. As this incorporated factors such as traffic patterns on the roads and meteorological influences, it is likely to be a more accurate marker of traffic pollution exposure. Whilst we have endeavoured to minimise misclassification by our choice of exposure variables, they still do not allow for exposure away from home and the possibility that they are insufficiently accurate to detect any true adverse effects that exist can not be ruled out. We also addressed the issue of our exposure variables being based on current () home location only, which for those who had moved house may not be the relevant exposure. However our subjects had lived in their home for an average of years and estimates were seen to remain similar when we restricted analyses to the subgroup of long term residents only, suggesting we had not missed any effects because of this issue.Our findings for asthma fit with a number of previous studies of adults that also found no adverse effect of living in close proximity to a major road on self-reported symptoms [-]. One study that did report an adverse effect looked at US male veterans and showed that those living within m of major roads had a increased risk of persistent wheeze, with a significant odds ratio of . for heavily trafficked roads (> = , vehicles a day), but a smaller effect (OR = .) which reached borderline significance when all major roads were considered[]. Two other recent studies have shown increased risks of wheeze of borderline significance in relation to increased residential proximity to surfaced roads in Ethiopia[] and living within m of a main street in Switzerland[]. The Ethiopian study differed from most other studies in that it was conducted in a developing country where background pollution was thought to be very low, and it may be that the likelihood of detecting any real effects of home proximity to the roadside are greater in such settings. In our study, the sampling method and geographical area chosen are likely to have provided a sample broadly representative of the general population, but the fact that the Gedling district is primarily urban means that the majority of the sample live in areas with relatively high background concentrations of pollutants. Insufficient contrast in exposure may therefore explain why we were unable to detect any adverse effects of our localised traffic pollution markers in this study population. Comparison with a recent study in Rome that also modelled NO2 revealed less variation in our values (IQR . to . μg/m3) than those that experienced in Rome (IQR . to . μg/m3), although this study also found no positive associations with asthma either[]. It is also possible that in some settings, effects of distance on asthma are evident across a wider range of distances than considered here, as suggested by Gauderman et al who reported a dose-response effect across the entire range of distance to the nearest freeway amongst children living in southern California [].A number of cross-sectional studies of lung function have, like ours, found no adverse effect of exposure to traffic on FEV1 [-]. However in a large study of US adults, Kan et al did find a negative association between traffic density at the residential location and FEV1, although in women only[], and a similar finding was reported in a study of German woman in relation to living within m of a major road[]. In our study, no differential effects by gender were observed. As with asthma, adverse effects on lung function have been reported in Californian children in relation to much larger cut-points in distance ( m bands), although again this was for freeways only []. Longitudinal studies of decline in FEV1 are more scarce, but in contrast to our finding of no effect, significant effects have been reported in relation to traffic-related pollution in Japanese women[] and Swiss adults[]. The latter used modelled PM10 concentrations, an exposure we were unable to analyse in our study as insufficient data were available for model validation. We also looked at spirometry-defined COPD using the same definition as that used previously by Schikowski et al who unlike us reported a significantly increased odds ratio of . in relation to living within m of a busy roads[]. Studies that looked at symptoms of COPD such as chronic cough and dyspnoea have generally found no significant associations[,].Exposure to traffic pollution could plausibly increase the risk of sensitisation to allergens as traffic-related pollutants have been shown to enhance immunological responses to allergens[]. Our finding of weak evidence of an effect on allergic sensitisation shows some consistency with that of Wyler and colleagues who reported an increased risk of skin sensitisation to pollen in relation to level of traffic at the home location[]. However, they found no such effect on sensitisation to indoor allergens or hay fever[]. Allergic sensitisation in adults was also investigated by Heinrich et al using specific IgE to common allergens and no relation to living near busy roads was seen[]. We found no significant effect on hay fever or eczema in adults, which with one exception[], is in agreement with others[,,,]. The lack of consistency of findings across different markers of allergy suggests caution is needed when interpreting one-off findings of adverse effects on allergic outcomes.ConclusionIn conclusion, we found no evidence to suggest that home proximity to major roads is a major determinant of the risk of asthma, COPD or allergic disease, or progression of obstructive lung disease in adults. However, because of relatively high levels of background pollution in our study area and possible misclassification of exposure, we cannot completely rule out an adverse effect, and further study is needed which incorporate life-time exposure to pollution in populations with wide variation in exposure.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMP carried out all the statistical analyses, including data manipulation, interpreted the results and drafted the manuscript for publication. SL and TM supervised the data analysis, helped interpret the results and critically revised the manuscript. DW carried out the pollution modelling and creation of the NO2 exposure variable, helped interpret the results and critically revised the manuscript. JB was involved in the conception and design and in obtaining funding for the study. He also helped interpret the results and critically revise the manuscript. AV designed and obtained funding for the study, provided overall supervision for the study, including the data analysis and interpretation of results, and critically revised the manuscript for publication. All authors give final approval of the version of the manuscript submitted.Pre-publication historyThe pre-publication history for this paper can be accessed here:
|
PMC2631474.txt
|
TITLE: Phase II trial of Modified Vaccinia Ankara (MVA) virus expressing 5T4 and high dose Interleukin- (IL-) in patients with metastatic renal cell carcinoma AUTHORS: Howard L Kaufman, Bret Taback, William Sherman, Dae Won Kim, William H Shingler, Dorota Moroziewicz, Gail DeRaffele, Josephine Mitcham, Miles W Carroll, Richard Harrop, Stuart Naylor, Seunghee Kim-Schulze ABSTRACT: BackgroundInterleukin- (IL-) induces durable objective responses in a small cohort of patients with metastatic renal cell carcinoma (RCC) but the antigen(s) responsible for tumor rejection are not known. 5T4 is a non-secreted membrane glycoprotein expressed on clear cell and papillary RCCs. A modified vaccinia virus Ankara (MVA) encoding 5T4 was tested in combination with high-dose IL- to determine the safety, objective response rate and effect on humoral and cell-mediated immunity.Methods25 patients with metastatic RCC who qualified for IL- were eligible and received three immunizations every three weeks followed by IL- (, IU/kg) after the second and third vaccinations. Blood was collected for analysis of humoral, effector and regulatory T cell responses.ResultsThere were no serious vaccine-related adverse events. While no objective responses were observed, three patients (%) were rendered disease-free after nephrectomy or resection of residual metastatic disease. Twelve patients (%) had stable disease which was associated with improved median overall survival compared to patients with progressive disease (not reached vs. months, p = .). All patients developed 5T4-specific antibody responses and patients had an increase in 5T4-specific T cell responses. Although the baseline frequency of Tregs was elevated in all patients, those with stable disease showed a trend toward increased effector CD8+ T cells and a decrease in Tregs.ConclusionVaccination with MVA-5T4 did not improve objective response rates of IL- therapy but did result in stable disease associated with an increase in the ratio of 5T4-specific effector to regulatory T cells in selected patients.Trial registration numberISRCTN83977250 BODY: BackgroundRenal cell carcinoma (RCC) is the fifth most common cancer worldwide and five-year survival is % for those with metastatic disease. High-dose bolus interleukin- (IL-) is associated with a consistent and durable objective response in % of patients with metastatic RCC and a –% complete response rate [-]. The relatively low frequency of therapeutic responses and significant treatment-associated toxicities, however, has made IL- difficult to recommend for all patients. The objective response rate to IL- was improved in a melanoma clinical trial when combined with gp100 peptide vaccination resulting in a % objective response rate []. In contrast to melanoma where numerous T cell specific antigens have been defined, relatively few antigens have been described in RCC [].5T4 is a membrane glycoprotein expressed at high levels on placental trophoblast and also on a wide range of human carcinomas including clear cell and papillary RCC but rarely on normal tissue [,]. 5T4 overexpression on tumor cells has also been associated with metastatic spread and poor prognosis in cancer patients [,]. 5T4 is not released from the cell membrane and thus can mediate antibody-dependent cell-mediated cytotoxicity (ADCC). In addition, 5T4-transduced renal carcinoma cell lines can be recognized by human T cells in vitro, suggesting that 5T4 can induce cellular immunity as well. 5T4-transfected tumor cells display altered morphology and increased motility suggesting that 5T4 plays a role in tumor progression and invasion []. A recombinant modified vaccinia virus Ankara (MVA) encoding human 5T4 (MVA-5T4) was tested previously in a phase I clinical trial for patients with stage IV colorectal carcinoma []. Vaccinated patients demonstrated few adverse events and nearly all patients developed 5T4-specific antibody and T cell immune responses, which correlated with time to disease progression []. Thus, the expression of 5T4 in RCC, ability to generate 5T4-specific humoral and cell-mediated immunity and the role of 5T4 in tumor progression suggest this would be an ideal antigen for targeted immunotherapy in RCC. Hence, we sought to determine if vaccination with MVA-5T4 could improve the therapeutic responses observed with standard high-dose IL- in patients with metastatic RCC. In order to take advantage of IL- during the contraction phase of the immune response, we designed an exploratory trial in which an initial vaccination was administered alone and subsequent booster immunizations were supported by the addition of high-dose bolus IL-.MethodsPatientsThis phase II trial was an open label study of MVA-5T4 vaccine in patients with metastatic clear cell or papillary RCC eligible for high-dose IL-. A total of patients were enrolled who met these criteria: Eastern Cooperative Oncology Group (ECOG) performance status of to , life expectancy greater than six months, years of age or older; able to provide written informed consent; able to comply with study procedures, hemoglobin > g/dL, granulocyte count > /mm3, lymphocyte count > /mm3, platelet count > ,/mm3, serum creatinine < . mg/dL, total bilirubin < . × the normal upper limits, and AST, ALT, and alkaline phosphatase < × the normal upper limit, or < × the normal upper limit if due to liver metastases. The clinical protocol was approved by the Institutional Review Board.Vaccine preparation5T4-MVA vaccine was produced by homologous recombination of human 5T4 cDNA into deletion region III of MVA under the control of the modified H5 promoter, as previously described []. Individual vials were stored in a secured, monitored, alarmed refrigerator at -°C. A sterile syringe was used to inject mL of solution subcutaneously in the deltoid region.Study designA dose of × pfu ( ml) MVA-5T4 was established as safe in a Phase I trial []. In this trial, the first dose was given by intramuscular injection alone and booster vaccination was given weeks later, followed immediately by high dose IL- (, IU/kg) given every hours up to a maximum of doses. Three weeks later patients received a third booster and second cycle of IL-. All patients underwent re-staging CT scans two weeks later. Clinical responses were determined by RECIST criteria []. For patients without progression an additional two cycles of vaccine/IL- were given at three week intervals. Patients demonstrating benefit after completing two courses of IL- were allowed to continue vaccination every three months for up to one year. In order to monitor the immune responses prior-, during- and post-vaccinations, heparinized blood was collected and processed by centrifugation through Histopaque columns to isolate peripheral blood mononuclear cells (PBMC).Antibody responsesMVA- and 5T4-specific antibody titers were determined by ELISA as described previously []. All test plasma was compared against a pool of plasma taken from healthy (vaccinia naïve) donors. Antibody titers were defined as the greatest dilution of plasma at which the mean optical density (O.D.) of the test plasma was ≥ fold the mean O.D. of the negative control (normal human plasma) at the same dilution. A positive response was defined as a post-vaccination titer ≥ fold of the baseline titer.T cell responsesThe IFN-γ ELISPOT was used to monitor T cell responses, as previously described []. Briefly, frozen PBMCs were thawed and incubated in medium overnight at °C, % CO2 prior to use. ELISPOT plates (PVDF, Millipore) were coated with an anti-IFN-γ capture antibody (human IFN-γ ELISPOT kit, Mabtech). Following blocking, × PBMCs were added to each well and incubated overnight at °C, % CO2 with the appropriate antigens. For positive control CEF (CMV, EBV and Flu virus) amino acid length peptides were used. Subsequently, spots were enumerated using an automated ELISPOT plate reader. The precursor frequency was calculated as the number of spot-forming units from wells containing PBMC and 5T4 overlapping peptides after subtraction of the background (PBMC alone) relative to the number of PBMC seeded per well. A positive ELISPOT response was reported if the mean spot forming units (SFU) per well in response to antigen was ≥ fold the mean SFU/well in wells containing medium alone and the mean SFU/well in response to antigen was ≥ . A positive response was also required to demonstrate ≥ fold increase after vaccination. Phenotypic characterization was done by four color flow cytometry analysis of PBMC using the following antibodies: CD4, CD8, CD25, CCR7, CD45RA, Foxp3, GITR, PD-, IL-, CD152, CD107a, granzyme B and perforin. Isotype matched controls were always included. The change of frequency for specific subset of cells during the post-vaccination period is calculated by subtracting the basal value of pre-vaccination time point. Flow cytometry was done using a FACSCalibur flow cytometer equipped with CellQuest Pro software. T cell function was tested by mixed lymphocyte proliferation assay, as previously described []. A total of healthy donor PBMC were used as normal controls.Statistical analysisSince this was an exploratory study, no formal power calculations were undertaken. The intention-to-treat population included all subjects enrolled in the study and the per-protocol population met all eligibility criteria and completed at least five vaccinations. All safety and efficacy analyses were carried out using the intention-to-treat (ITT) population and analysis of immune response was carried out in the per-protocol population. Descriptive statistics were analyzed using Student's t-test to assess differences between the different study groups with p < . considered significant. Correlations between variables were assessed with adjustments to other variables via linear models. Overall survival (OS) was calculated by the method of Kaplan-Meier, log rank test. OS was calculated from the first date of treatment to date of death, or last known date alive.Role of funding sourceThis work was supported by grants from Oxford Biomedica. The funding sources had no role in the study design, collection, analysis, or interpretation of the data, or in the writing of the report. They also had no access to the raw data. The corresponding author had full access to all data and the final responsibility to submit for publication.ResultsPatient characteristicsTwenty five patients were enrolled in the trial and included in the ITT population. One patient withdrew from the trial early due to relocation and one patient could not tolerate IL-, leaving patients in the per-protocol analysis. The mean age of the ITT population was . ± years (range – years). patients had clear cell carcinomas and patients had papillary histology. Further characteristics are detailed in Table .Table 1Patient characteristics and treatmentsMean age58. (range –)N = %SexMale1768Female832TNM StageTX4160001410272837284312NX00018721002728M000125100HistologyClear cell2184Papillary416Sites of diseaseLung1664Lymph node936Soft tissue728Bone624Kidney520Liver520Pancreas28Adrenal14Prior TherapyNephrectomy2392Chemotherapy832Immunotherapy1040Radiation therapy28Cryoablation14Laser ablation14Treatment CharacteristicsVaccination≤ –≥ 6832No. of IL- Cycles141629363004936Treatment-related toxicityTable shows all adverse events; there were no serious adverse events related to the vaccine in the ITT population. The most frequent side effect related to vaccine administration was fever in patients. Other toxicities were largely expected high-dose IL- related side effects (see Table ).Table 2Adverse events related to vaccine and IL-2Vaccine-related AEsMaximum GradePatientsSystemAdverse EventsN = %ConstitutionalFever1832Pain at injection site1416Injection site reaction1312Myalgia114Chills114IL--related AEsCardiovascularCardiopulmonary arrest414Elevated troponin414Hypotension4936Ventricular tachycardia314Acidosis414ElectrolyteHyperglycemia3416Hypocalcemia314Hyponatremia31144Hypophosphatemia3312Gastro-intestinalIschemic bowel414HematologicAnemia328Neutropenia314Thrombocytopenia3416HepaticElevated transaminases314Hyperbilirubinemia3312NeurologicConfusion3312Syncope328PulmonaryDyspnea314RenalElevated creatinine32288Oliguria328SystemicFatigue314Humoral immune responsesMVA- and 5T4-specific antibody responses were monitored by ELISA at each sampling time point throughout the trial and expressed as a titer [see Additional file ]. All patients showed an increase in MVA antibody titers following vaccination (range, to ,). One patient (#) had detectable MVA-specific titers prior to the first vaccine and this increased further following vaccination. All patients also demonstrated 5T4-specific antibody titers ranging from to , which were evident after ≥ vaccinations in most patients. Two patients (# and ) had detectable 5T4-specific antibody titers prior to vaccination but showed an increase in post-immunization titers.Effector and regulatory T cell responses5T4-specific CD8+ T cell responses were monitored by IFN-γ ELISPOT assay using overlapping 5T4 peptides and full-length protein. Before immunization only a single patient had a detectable T cell response (frequency :,). Following treatment of tested patients (%) had detectable 5T4-specific CD8+ T cell responses with precursor frequencies ranging from :, to :, (Table ). Only of (%) patients with progressive disease exhibited an increase in T cell response compared to of patients (%) with stable disease (Fig. ). Positive T cell responses to MVA and a control CEF peptide pool were detected in all evaluable patients (Table ). The CEF-specific precursor frequencies were highly consistent throughout the study period. The mean frequency of MVA-specific T cells was decreased slightly from : PBMCs pre-vaccination (.%) to : PBMCs post-vaccination (.%).Figure 15T4-specific T cell responses in patients with (A) progressive disease and (B) stable disease.Table 3Antigen specific T cell responsesPatient NumberPeak 5T4 polyclonal precursor frequenciesORR (month)Time point (week)Peptides aloneTime Point (week)Protein + Peptides1-< /,-< /,000PD2-< /,/,231PD3-< /,-< /,000PD5551/,/,701Surgical CR(+)/,/,277SD ()/,/,113SD ()-< /,-< /,000PD91051//993Surgical CR(+)-< /,-< /,000PD11-< /,/,500SD ()-< /,-< /,000PD13151/,/,765SD ()-< /,-< /,000PD15-< /,/,231SD ()-< /,-< /,000SD ()/,/,132SD ()-< /,/,833SD ()-< /,-< /,000PD21-< /,-< /,000PD2291/,/,821PD23-< /,-< /,000Surgical CR (+)-< /,/,048SD (.)-< /,/,868PDThe peak 5T4 specific responses detected at any time point to 5T4 peptides or 5T4 peptide plus protein. *; no detection of 5T4 responses. Positive responses are indicated as bold type.Table 4T cell responses to CEF and MVA antigens by IFN-γ ELISPOTPatient NumberAntigenPeak Ag Specific T cell Precursor FrequenciesPrePost1CEFND1/,299MVAND1/,3642CEFND1/,929MVAND1/,9263CEFNDNDMVA1/,/,3415CEFND1/,658MVAND1/,9936CEF< /,< /,000MVA1/,/,6297CEFND1/,084MVAND1/,9358CEF1/,/,126MVA< /,/17709CEF1/,/956MVA1/,/,45210CEF< /,< /,000MVA< /,/,44211CEF1/,/,362MVA1/,/,13512CEFND< /,000MVA< /,/4545513CEF< /,< /,000MVA< /,/,63014CEF1//619MVA1//,11215CEF1/,/,445MVA1/,/,90116CEF1/,/,316MVA1/,/,68517CEFND1/,216MVAND1/,40519CEF1/,/,335MVA1/,/,27320CEF< /,< /,000MVA1//98421CEF< /,< /,000MVA1/,/94522CEFND1/,789MVAND1/,98823CEF1//,056MVA1/,/,56124CEF1//,078MVA1/,/,25625CEF1/,/,161MVA1/,/,697Abbreviation: ND, not detectedCD8+ effector T cell response were also characterized by staining for T cell activation markers [,]. The mean frequency of CD8+CD107a+ T cells at baseline was .% ± . and increased to .% ± . after vaccination. Figure 2A shows that patients with stable disease had a significantly greater increase in CD8+CD107a+ T cells compared to those with progressive disease (.% ± . vs. .% ± ., p = .). There was also a higher frequency of CD8+perforin+ T cells in RCC patients compared to normal healthy donors (. vs. .%, p = .) and a trend towards decreasing CD8+perforin+ T cells in patients with progressive disease (Fig. 2B). In addition, there was a significant increase in PD- expressing CD4+ (p = . at weeks and p = . at weeks) and CD8+ T cells (p = . at weeks) in patients with progressive disease compared to stable patients (Fig. 2C).Figure 2Characterization of T cell responses. (A) CD8+CD107a+ effector cells, (B) CD8+perforin+ effector cells, (C) PD-+ T cells, (D) CD4+CD25+FoxP3+ Tregs before and after treatment.CD4+CD25+FoxP3+ Tregs were monitored by flow cytometry throughout the trial and functional suppression determined by co-culture proliferation assay. The mean frequency of Tregs in the per-protocol population at baseline was significantly higher than that detected in healthy donors (.% vs. .%, p = .), although the degrees of suppression in proliferation assays was similar (p = .) (data not shown). In patients with progressive disease, the mean Treg frequency was .% (± .) before treatment and increased to .% (± .) after treatment (Fig. 2D). In contrast, patients with stable disease had a mean Treg frequency of .% (± .) prior to treatment which decreased to .% (± .) by weeks (Fig. 2D). The absolute number of Tregs was decreased by % in stable patients following treatment (p = .). Fig. 3E–G shows the kinetics of effector CD8+ T cell responses and Treg frequency in three representative patients with stable disease. The effector/regulatory T cell ratio decreased in patients with progressive disease, whereas stable patients showed a dramatic increase which was maintained for up to months (Fig. 3H).Figure 3Representative effector CD8+ T cell and Treg responses in patients (A-C). effector/regulatory T cell ratio in all patients (D). SD, stable disease (open square), PD, progressive disease (closed square).Clinical responseThere were no objective responses based on the first re-staging CT scans. Twelve of (%) per-protocol patients, however, had stable disease and went on to a second course of vaccination/IL-. Three patients (%) were rendered disease free through surgical resection; patients had complete regression of all metastatic disease (lungs and bone) at initial follow-up and underwent nephrectomy of primary tumors, patient had two intra-abdominal masses that regressed by < % but were surgically resected (pathology showed tumor with significant necrosis in one mass and no viable tumor in the other). The median progression-free survival of the per-protocol patients was . months and median overall survival has not yet been reached (Fig. 4A) at a median follow-up of months.Figure 4Kaplan-Meier analysis of (A) overall (solid line) and progression-free (dashed line) survival of per-protocol patients treated with MVA-5T4 and IL-. (B) Overall survival of stable (solid line) and progressive (dashed line) disease patients. Numbers of patients at risk at , and months are shown below the graph.Median overall survival of the stable patients has not yet been reached (– months) and was months (– months) for those with progressive disease (Fig. 4B, p = .).DiscussionThis study established the safety and feasibility of combining vaccination with MVA expressing 5T4 and high-dose IL- in patients with metastatic RCC. The trial was initially designed to determine the impact of combination treatment on objective response rate since there is a well-defined, consistent response for IL- alone [,]. We did not, however, observe any objective responses by strict RECIST criteria although three patients were rendered disease free by additional surgery. The reasons for this outcome might relate to the study design in which we evaluated initial tumor responses two weeks after completing the first course of IL-, selected in order to continue booster immunizations in a timely manner. Recent reports suggest that the kinetics of immunotherapy may require more time to mediate tumor regression in patients with established disease and, therefore, detection of tumor regression may be delayed [,]. This possibility is supported by patient #, who continues to have a slow but steady regression of tumor over a month period. Thus, our decision to scan at two weeks might have prevented some patients with stable disease from becoming objective responders. The trial was also biased by the early surgical intervention in three patients who were rendered disease free prior to further follow-up imaging. Two of these patients had complete regression of metastatic disease but had large primary renal tumors in place. Primary tumors are known to be more resistant to immunotherapy and often require nephrectomy before or after treatment to optimize response []. We also included four patients with papillary histology in the trial since these tumors express 5T4, but these tumor are also more resistant to IL-, which may have influenced our results [].MVA-5T4 vaccine and high-dose IL- elicited 5T4-specific humoral and cell-mediated immunity. All patients developed an increase in 5T4 antibody titers after vaccination, consistent with previous clinical trials in patients with metastatic colorectal and hormone-refractory prostate cancer [,]. While the pattern of antibody response in our patients was similar to that observed in previous studies, the magnitude of the response was higher in this trial (mean , maximum titer ) compared to colorectal cancer patients treated with MVA-5T4 and chemotherapy (mean , maximum titer ) []. We also observed the induction of 5T4-specific CD8+ T cell responses in % (/) of vaccinated patients and this compares favorably to previous trials [,]. The induction of humoral and T cell immunity in this trial might relate to the underlying tumor histology, since RCC is known to be more immunogenic than other tumors [,] or could be due to the adjuvant effects of high-dose IL-. We further characterized the effector CD8+ T cells in whole PBMC and found that there was an increase in CD107a, a marker of degranulation and cytotoxic function [,]. These cells remained elevated in patients with stable disease but began to decrease at weeks in patients with progressive disease. We saw a similar trend in CD8+perforin+ T cells although this was only significant at weeks. We also found that PD- expression, a pan T cell co-inhibitory receptor, was significantly elevated in both CD4+ and CD8+ T cells in patients with progressive disease [-]. These data suggest that the loss of effector CD8+ T cells or decreased effector function is associated with tumor progression.Since Tregs may suppress tumor rejection by effector T cells and because IL- can promote Treg activity, we evaluated the frequency and functional activity of Tregs in our patients. We previously reported that Tregs are increased in metastatic RCC patients but decreased to normal levels in those patients responding to IL- therapy []. In the current study, we similarly found that the Treg population was increased in patients compared to normal donors without detectable differences in suppressor activity. Patients who achieved stable disease demonstrated a % reduction in the mean number of Tregs within four weeks of completing the first course of IL- (p = .) and supports the notion that patients destined to respond to immunotherapy exhibit a decreased frequency of Tregs. In murine tumor models, the ratio of effector to regulatory T cells was found to be the critical determinant of tumor regression or progression []. Similarly, we found that patients with stable disease exhibited an increase in the effector to regulatory ratio that persisted for at least months; in contrast, patients with progressive disease showed a low ratio at all time points tested. Although we lacked statistical power in our trial to directly compare these groups, these data would support determining the effector to regulatory ratio in future clinical trials.In summary, this study provides safety and feasibility data supporting the combination of MVA-5T4 vaccine and IL- for patients with metastatic RCC. The treatment regimen was associated with induction of 5T4-specific humoral and cellular immunity. Twelve patients had stable disease, which was associated with increased effector T cells, reduced Tregs and increased effector to regulatory T cell ratios, suggesting a benefit from therapy. Although there was insufficient power to make conclusions regarding clinical response, these data suggest that stable disease by current RECIST criteria might harbor subsets of patients who may benefit from immunotherapy. Future randomized studies will be helpful in better delineating the potential effectiveness of MVA-5T4 and IL- for the treatment of RCC.Competing interestsRichard Harrop, William Shingler and Stuart Naylor are employed by Oxford Biomedica U.K. Ltd.Authors' contributionsH. L. K and M.W.C. did the conception and design of the clinical study; H. L. K., B. T and W. S. treated and evaluated patients; G. D. and J. M. provided study materials; S. K-S, D. W. K, W. H. S, D. M. processed samples and analyzed immune responses; H.L.K, S. K-S, J. N. H, R. H., and S. N. did data analysis and interpretation. H.L.K, J. N. H and S. K-S did statistical analysis and wrote the manuscript. All authors have agreed to all the content in the manuscript, including the data as presented.Supplementary MaterialAdditional file 1MVA- and 5T4- specific antibody responses. (A) MVA-specific antibody titers, (B) 5T4-specific antibody titers. The data provided antibody titers specific for MVA- and 5T4- antibodies.Click here for file
|
PMC2803943.txt
|
TITLE: Guyons canal syndrome due to accessory palmaris longus muscle: aetiological classification: a case report AUTHORS: Ramavath Ashok Lal, Sakamuri Raj ABSTRACT: IntroductionAccessory muscles and anatomic variations are well described at the Guyon's canal. Though this case report is similar to variants published in previous reports, it differs from the rest due to rapidity of worsening of symptoms in few months following use of cane.Case presentationWe report a case of year old man with ulnar nerve compression at Guyon's canal by accessory palmaris longus arose from distal third palmaris longus and from deep fascia of forearm. The hypertrophied muscular portion of accessory palmaris longus crosses over the ulnar nerve and artery at Guyon's canal becomes tendinous before merging with the hypothenar muscle.ConclusionWith co-existing anatomical variants, pressure in Guyon's canal might rapidly increase, might be causative factor and cause compression of deep branch of ulnar nerve following frequent dorsiflexion of wrist like in our case. Following division of accessory palmaris longus symptoms rapidly improved. In this article we discuss the aetiological classification, diagnostic criteria and treatment based on available evidence. BODY: IntroductionAccessory muscles are the most frequently described anatomical variations at Guyon's canal .% in cadaver specimens and bilateral in .% (Dodd's et al., ) []. Compression of deep branch of ulnar nerve is mainly due to intrinsic or extrinsic factors (Inapathy ). The most common anatomical accessory muscle variation were accessory abductor digiti minimi with incidence of .% (Dodds et al., ) and accessory palmaris longus with incidence of .%.This case report differs from other published in the literature in rapidly progression of symptoms following use of cane for to support his new hip replacement. This compression of deep branch of ulnar nerve is confirmed by electro diagnostic studies with predominant compression at Guyon's canal.Case presentationA -year-old British Caucasian man presented with pain along hypothenar aspect and severe paraesthesia of six months duration in the area of ring and little finger of left non dominant wrist. The symptoms are aggravated rapidly with progressive weakness following use of cane. This cane was prescribed to support his new revision hip. On further examination in out patient department, revealed hypothenar atrophy and wasting of interosseous muscles, hypoesthesia and paraesthesia over ring and little finger. The left hand grip was weak. Tinel sign positive at wrist. The symptoms were reproduced by Phalen's test.The electrodiagnostic studies revealed, slowing of ulnar motor and sensory mainly at the wrist. The electro diagnostic study of median nerve was normal. A diagnosis of ulnar nerve entrapment in Guyon's canal made and subsequently patient underwent decompression with good recovery.At operation, following surgical exploration of Guyon's canal, revealed accessory palmaris longus muscle originating from the palmaris longus tendon and few muscle fibres from deep fascia of forearm passing through Guyon's canal over the ulnar nerve and vessels and merging with hypothenar muscles (Figure ). We also noticed tortuous ulnar vessels proximal to Guyon's canal. This hypertrophied muscle was compressing common ulnar nerve trunk and ulnar vessel. The muscle was divided widely and decompressed ulnar nerve and ulnar vessel (Figure ). The ulnar nerve was explored proximally three to four inches in to distal fore arm and did not find any other sites of compression. There was no constriction at pisohammate hiatus. After hemostasis and skin closure, a soft bandage applied and no post operative complications detected. Post operatively patient under went hand physio, and free of compression symptoms following review at OPD in two and six weeks.Figure 1Aberrant palmaris longus muscle compressing ulnar nerve and vessels at Guyon's canal. Note: tortuous ulnar vessels.Figure 2Release of compression over ulnar nerve and vessel at Guyon's canal. Note: excised ends of aberrant palmaris longus muscle.DiscussionFollowing the review of literature, [] based on aetiology, the causes of ulnar tunnel syndrome broadly categorized in to primary and secondary.The primary is idiopathic with symptoms of Guyon's canal syndrome with no identifiable lesion, in a study done by Murata et al found % of cases were idiopathic. The secondary causes can be further divided in to traumatic, inflammatory, structural, vascular (Ozdemir ), neuropathic and occupational. Following fractures to hook of hammate [,] fractures of ring and little finger metacarpal bases there were reported evidence of compression of ulnar nerve at the wrist. Inflammatory conditions like rheumatoid arthritis, rheumatoid synovitis can also cause compression of ulnar nerve [].Structural causes can be further divided in to tumorous and nontumorous. The tumorous conditions causing compression over deep branch of ulnar nerve are ganglion [], lipoma [], and giant cell tumour (Hayes JBJS ).Anomalies at Guyon's canal are common .% and bilateral (.%) in cadaver specimens []. The radiographic study by Ziess et al1992 incidence of accessory muscle is .%. Non tumorous conditions are anomalous muscles like unusual thickening of distal edge of palmaris brevis (Mackinon ), thickening of proximal edge of volar carpal ligament or fibrous origin of flexor digiti minimi muscle [], complete reversal of palmaris longus so that tendon arises from the medial epicondyle and muscle belly attaches to flexor retinaculum at the wrist [], accessory palmaris longus muscle arise from palmaris longus tendon insert distally in to pisiform bone and in to the hypothenar muscle traversing through the Guyon's canal (Thomas JBJS ), thickening of roof of Guyon's canal due to origin of anomalous muscle from the flexor carpi ulnaris inserting in to the volar carpal ligament [], High origin of flexor digiti minimi from palmaris longus, with double origin reported (Jeffrey JBJS1971), duplication of hypothenar muscles simulating a hand tumour described (Lipscomb JBJS ), the most common muscle variant at Guyon's canal is abductor digiti minimi (.%) which passes through Guyon's canal [], in one reported case its anomalous variant arose from transverse carpal ligament instead of pisiform bone [].The vascular causes of Guyon's canal syndrome are ulnar artery aneurysm (Yoshii ) or ulnar artery thrombosis [], pain in hypothenar region followed by continuous tremors of ring and little finger is due to distal ulnar neuropathy relieved by Guyon's canal decompression [].The occupational or social causes are mostly contributory due to over use like in carpenters and in our case, resulting in compression of deep branch of ulnar nerve worsened following use of cane with or with out anatomical variation.The diagnosis of Guyon's canal syndrome made by history, clinical examination and electromyography studies. One can clearly differentiate ulnar nerve compression at wrist or above elbow, by eliciting weakness of flexor carpi ulnaris and flexor digitorum profundus to ring and little finger metacarpal and sensory changes to ulnar side dorsum hand. Detailed electro diagnostic studies help in diagnosis.ConclusionWith above check list in mind and careful clinical examination one can diagnose Guyon's canal syndrome. Though anomalies are frequent at Guyon's canal, over use with frequent dorsi flexion position might have aggravated symptoms, early diagnosis followed by exploration and decompression will relieve symptoms like described in our case.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' contributionsRA - main author, wrote the paper, performed the literature search and discussion.SR - aided in editing and senior author of this article.
|
PMC2633281.txt
|
TITLE: Sex-specific dispersal and evolutionary rescue in metapopulations infected by male killing endosymbionts AUTHORS: Dries Bonte, Thomas Hovestadt, Hans-Joachim Poethke ABSTRACT: BackgroundMale killing endosymbionts manipulate their arthropod host reproduction by only allowing female embryos to develop into infected females and killing all male offspring. Because the resulting change in sex ratio is expected to affect the evolution of sex-specific dispersal, we investigated under which environmental conditions strong sex-biased dispersal would emerge, and how this would affect host and endosymbiont metapopulation persistence.ResultsWe simulated host-endosymbiont metapopulation dynamics in an individual-based model, in which dispersal rates are allowed to evolve independently for the two sexes. Prominent male-biased dispersal emerges under conditions of low environmental stochasticity and high dispersal mortality. By applying a reshuffling algorithm, we show that kin-competition is a major driver of this evolutionary pattern because of the high within-population relatedness of males compared to those of females. Moreover, the evolution of sex-specific dispersal rescues metapopulations from extinction by (i) reducing endosymbiont fixation rates and (ii) by enhancing the extinction of endosymbionts within metapopulations that are characterized by low environmental stochasticity.ConclusionMale killing endosymbionts induce the evolution of sex-specific dispersal, with prominent male-biased dispersal under conditions of low environmental stochasticity and high dispersal mortality. This male-biased dispersal emerges from stronger kin-competition in males compared to females and induces an evolutionary rescue mechanism. BODY: BackgroundParasites can induce host population extinction through negative effects on population growth []. This effect can be amplified by the induction of host behaviour that stimulates parasite spreading [-]. Some bacterial endosymbionts that that are predominantly vertically transmitted from females to their offspring can be regarded as such parasites. They are widespread in arthropods and manipulate reproduction of their host [,]. The induced reproductive manipulations comprise parthenogenesis (i.e. infected virgin females produce daughters), feminization (infected genetic males reproduce as females), cytoplasmatic incompatibility (CI; in its simplest form the mating between infected males and uninfected females leads to the death of embryos), and male killing (i.e. infected male embryos die while infected female embryos develop into infected females). Male-killing imposes substantial costs at the individual level (both on infected females and on the males that mate with them) since the death of male offspring halves the number of viable offspring. Such male-killing endosymbionts are known from butterflies, ladybird beetles and flies, in which they affect sex ratio and related life-history parameters []. If both, endosymbiont transmission and host manipulation occur with near-perfect efficiency, infections can approach (near) fixation and host extinction will occur when all males are eliminated out of the population [,]. However, Randerson and colleagues [] showed that hosts can avoid extinction by adaptive alterations of sexual behaviour, thereby raising their inclusive fitness. A prominent example comprises male-killing bacteria that trigger increasing male fatigue and female promiscuity in infected population of a butterfly [].Male killing endosymbionts clearly impair individual fitness of infected males (and implicitly of the males mating with them). Yet, the invasion of male-killers in a metapopulation has strong and more complex effects on the population structure. First, effective population sizes are strongly reduced due to skewed sex ratios []. This is expected to lead to a reduction in genetic variation [], which subsequently reduces the effectiveness of selection against deleterious mutations [] and the rate of adaptive evolution [,]. Secondly, endosymbionts reduce population density and thus relax intraspecific competition []. This relaxation is mainly among kin when eggs are laid in clutches and female offspring benefit from the death of their brothers by e.g. lowered resource competition in a later life phase [,]. Alternatively, male death enhances the fitness of their infected female siblings by prevention of inbreeding []. However, when competition takes place among all patch inhabitants male-killing endosymbionts reduce competition among offspring, thereby potentially benefiting surviving individuals within the local population indifferently of their infection status. This can lead, under certain conditions, to the spread of male-killing endosymbionts in metapopulations (by beneficial group trait selection) in absence of any explicit fitness compensation [].At any rate, male-killing endosymbionts have the ability to alter the demographic properties of their host population by inducing female-biased sex ratios. In combination with environmental factors related to dispersal mortality and demographic stochasticity these changes in population structure are expected to have a strong influence on the evolution of dispersal in spatially structured populations [-]. In a previous contribution [], we already showed that the invasion of male killers selects for increased dispersal under conditions of low environmental stochasticity and high dispersal mortality. These evolved dispersal rates subsequently provoked extinction-colonization dynamics among patches in the metapopulation, thereby leading to stable infection frequencies (from here-on referred to as infection rates), metapopulation extinction of host and endosymbiont, or endosymbiont extinction only. Yet, because male-killing endosymbionts induce pronounced sex-specific effects on fitness and kin structure, they should also influence the evolution of dispersal differently in males and females. This is in agreement with the general predictions of Leturque & Rousset [] that evolved changes in sex ratio may lead to higher dispersal rates and trigger the evolution of sex-specific dispersal. A theoretical background for the evolution of sex-biased dispersal is, however, poorly developed. Perrin & Mazalov [] showed that mating systems are expected to be an important driver of sex-biased dispersal because sex-specific differences in potential reproductive success affect the balance between local resource mate competition and local resource competition.When populations occupy spatially structured habitat, evolutionary changes in dispersal may rescue populations from (human induced) changes in habitat availability and quality []. These adaptive responses can indeed occur in fairly short time spans, as, for example, shown for wind dispersing arthropods [,] and vascular plants []. Metapopulation curing, i.e. the deterministic extinction of parasites but not the host, under environmental conditions that select against dispersal [] could be another prominent example of such an evolutionary rescue. Because male killing endosymbionts generate strong bias in sex ratio, and increase dispersal rates considerably [], we here explore the evolutionary mechanisms leading to male-biased dispersal in infected populations. Secondly, we show that a male-bias in dispersal is responsible for higher rates of metapopulation curing (host extinction) compared to populations with sex-indifferent dispersal.ResultsWe build an individual based simulation model that allowed the evolution of sex-specific dispersal strategies in a metapopulation consisting of patches with carrying capacity K, inhabited by sexually reproducing, polygynous organisms. Dispersal is accompanied by costs (μ) that relate to patch isolation. Environmental stochasticity is modelled by a standard deviation (σ) around the number of offspring (λ). Population dynamics follow logistic growth, endosymbionts are maternally transmitted from mother to daughters (male offspring die during the embryonic stage when the mother is infected). Dispersal strategies were determined by two sex-specific dispersal alleles.Mean dispersal probabilities reached equilibrium after less than generations. Similarly, sex ratio and the proportion of infected individuals stabilised after this number of generations. In general, our simulation results confirmed the evolution towards higher dispersal probabilities under higher environmental stochasticity (σ) and lower costs of dispersal (μ) (Fig ). This pattern holds for males (Fig 1A) as well as females (Fig 1B), but a considerable male-bias in dispersal rate was observed (Fig 1Aversus 1B; Table ). Similar patterns were found for simulation experiments with uninfected metapopulations, although overall dispersal probabilities as well as male-bias are considerably lower in the latter (Fig 1C, 1D). When simulations were run for sex-indifferent dispersal strategies (i.e., one allele coding for dispersal propensity in males and females) in infected metapopulations (Fig 1E) we noticed the evolution of increased dispersal probabilities, particularly under low environmental stochasticity and high dispersal costs.Figure 1Sex specific dispersal rates. Dispersal probabilities for sex-specific strategies in infected (A males; B females) and uninfected metapopulations (C males; D females) and sex-indifferent strategies in an infected metapopulation (E). The x-axis gives dispersal mortality (μ), the y-axis environmental variability (σ).Table 1Bias in sex-specific dispersalScenarioBiasno infection, no reshuffling0. ± .04no infection, reshuffling0. ± .04infection, no reshuffling0. ± .06infection, reshuffling0. ± .04Bias (male dispersers/total number of dispersers) in sex-specific dispersal for the different investigated scenario (mean and SD over the considered ranges of σ ( ≤ σ ≤ .) and μ(. ≤ μ ≤ .)).In order to eliminate effects of kin competition, but retaining all other characteristics of local populations, we performed a reshuffling experiment. By this, individuals within each patch were replaced by individuals of the same sex and with the same infection status randomly selected from the entire pool of individuals in the metapopulation. The genetic structure of the metapopulation is consequently homogenized and kin-competition eliminated. This leads to a decrease in dispersal probability in all scenarios. The decline was especially pronounced under conditions of high dispersal mortality and low environmental stochasticity (compare Fig 1A, B with Fig 2A, B). More interestingly, reshuffling removed the male bias of dispersal (Table ) and dispersal patterns became similar for both sexes (Fig 2Aversus Fig 2B).Figure 2Sex specific dispersal rates after eliminating kin competition. Dispersal probabilities for sex-specific strategies (A males; B Females) after reshuffling (elimination of kin competition). The x-axis gives dispersal mortality (μ), the y-axis environmental variability (σ).Sex-indifferent dispersal strategies had a minor rescue-effect on the entire metapopulation extinction probability, with a slight shift towards decreasing extinction probabilities under conditions that select for high dispersal (Fig 3A, B). In contrast, allowing the evolution of sex-specific dispersal strategies increases the chance of endosymbiont extinction while the host metapopulation survives. As evident from Fig 3C and Fig 3D, the evolution of male-biased dispersal induced curing (only endosymbiont and not host extinction) especially under high dispersal mortality and low environmental stochasticity but also increased curing under conditions of high environmental stochasticity.Figure 3Metapopulation extinction rates. Metapopulation (upper panels A, B) and endosymbiont extinction (lower panels C, D) probabilities for sex-specific (left panels, A, C) and sex-indifferent strategies (right panels B, D). Note decreased metapopulation extinction and increased curing under scenarios with evolution of sex-specific dispersal. The x-axis gives dispersal mortality (μ), the y-axis environmental variability (σ).DiscussionStrong environmental fluctuations and low dispersal mortality are well-acknowledged factors that support the evolution of high emigration probability [-,,]. Our simulations suggest that the presence of male-killing microbial infections should increase overall dispersal rates, too. When dispersal is assumed to be independent of sex, dispersal probabilities showed a significant increase, particularly for those scenarios otherwise favouring low dispersal (i.e. high dispersal mortality and low environmental stochasticity). However, if allowed for sex-specific dispersal, endosymbionts induced pronounced male-biased dispersal rates. Effects of male-killing endosymbionts on demography and life history in e.g. tropical butterflies [], flies [] and ladybird beetles [] are documented. No attempts have been made, so far, to link infections by male-killing endosymbionts to spatial population structure and dispersal. Our results suggest that this would be a worthwhile endeavour.The higher rates of patch extinction induced by male-killing endosymbionts compared to metapopulations of uninfected hosts [], as well as the benefits of relieved competition in patches that are founded shortly after patch extinction select for higher dispersal rates [,-]. The relationship between extinction rate and dispersal might, however, become hump-shaped with very high extinction rates selecting for reduced dispersal []. Yet in our simulations, as well as in simulations without infection [], local extinction rates so high as to select for reduced dispersal never emerged because the system rapidly went globally extinct under such conditions.Sex-biased dispersal is documented to originate from strategies related to inbreeding avoidance, inbreeding depression, asymmetrical mating systems, social structure, or sex-specific dispersal costs [,,]. More generally, the small effective population size and the tighter kin-structure in infected metapopulations due to the rarity of males explain the evolution towards overall higher dispersal rates, compared to metapopulations without male-killing endosymbionts. This particularly holds under conditions of high dispersal mortality and low environmental stochasticity. However, theory generally predicts that strong kin-competition selects against a sex-bias in dispersal [], while in our simulations we demonstrate that strong kin-competition is responsible for the emergence of male-biased dispersal. The explanation for this apparent contradiction can be found in the unbalanced cost-benefits for dispersal in males and females. In Gandon's system [], males and females played the same game, i.e. they both competed for space and mates and have the same costs of dispersal. Taylor [] showed, however, that deviations from these conditions, i.e. sex-specific costs of dispersal and different relatednesses for the two sexes, induce sex-specific dispersal. The latter assumed that a sex bias in relatedness should only be important in haplodiploid populations, but not in diploid organisms. In our system, however, male-killing endosymbionts strongly affect within-population relatedness, because under high infection rates the (few) males in a patch have higher probabilities to share mothers than females. This is due to the fact that in populations with an infection rate I only a fraction of -I females produce sons, while all females produce daughters. Like for Taylor's haplodiploid system, the most related sex (females in a haplodiploid system, but males in our case) should thus evolve higher dispersal rates than patchmates from the other sex.One could also speculate that dispersing males take a greater risk of mating with infected females as the very existence of males in a patch may indicate reduced infection rates in that patch. However, consequences of such an increase in the cost of dispersal could not be observed. Probabilities of mating with uninfected females were not higher for males in their natal patches, because populations may persist at high infection frequencies []. Moreover, the particularly strong kin-competition for males overrules this potential (and sporadic) increased chance of finding uninfected females.Leturque & Rousset [] showed that sex-specific dispersal rates may evolve when reproductive values vary among genotypes and when relatedness is high. Interestingly, their model also shows that the incorporation of sex-biased dispersal may lead to a biased sex ratio in offspring production in finite populations and in populations experiencing spatial heterogeneity in habitat quality. In our model, changes in the sex ratio are due to male-killing endosymbionts and lead to higher and sex-biased dispersal. Yet, we could expect that the changes in sex ratio might trigger an evolutionary response also in offspring sex ratio. If a female can recognize its own infection status, infected females should clearly shift the sex ratio in favour of female offspring as the male offspring is killed anyways. Uninfected females should in contrast produce more male offspring as they would have more fitness contribution via sons. However, that only holds for infected populations because in uninfected ones the sex ratio is :. So ideally, such females should response to both, their own infection status and the infection status of the population.In well connected metapopulations with high inter-patch dispersal, metapopulation extinction probabilities are hardly affected by dispersal. Under these conditions the high dispersal leads to high recolonization rates of empty patches but also to a rapid spread of infections into uninfected populations []. Consequently, endosymbiont extinction probabilities are only slightly affected by sex-specific dispersal when dispersal is generally high (even in females). By contrast, deterministic curing of metapopulations increases with sex-specific dispersal strategies. The evolution of sex-specific dispersal leads to considerably lower dispersal probabilities of females (compared to simulations with sex indifferent dispersal) under environmental conditions characterized by high dispersal mortality and low environmental stochasticity. This leads to a comparable decline in recolonization rates of patches by infected females and the spread of infections. The evolution towards male-biased dispersal in infected metapopulations – promoted by the benefit of reduced kin-competition in males – can consequently be regarded as an evolutionary rescue as it increases the probability of curing for the entire metapopulation.In our simulations we assume global dispersal as opposed to e.g. nearest neighbour dispersal, reflecting airborne dispersal in e.g. arachnids and insects. Because kin-competition is the dominant driver behind the evolution of sex-specific dispersal, we could expect limited dispersal distance (i.e., nearest neighbour dispersal), which maintains some kin-competition even after dispersal, to have a strong influence on the evolutionary dynamics. However, simulation experiments for uninfected metapopulations [] as well as our own simulations (Bonte D, Hovestadt T, Poethke HJ, unpub. data) show that the choice of dispersal-mode has little influence on the evolution of ES dispersal probabilities. This is due to intrinsically high dispersal rates and the absence of any spatial autocorrelation in environmental stochasticity. This demonstrates that kin-competition is predominantly generated through changes in local population structure, i.e. the (much) reduced effective population size in infected populations and not by limited dispersal distance.Kin-competition is documented to select for increased dispersal rates when dispersal cost is high and/or when spatio-temporal variability of environmental conditions is low [,,-]. Male-killing endosymbionts induce strong kin-competition under these environmental conditions, which lead to strong sex-biased dispersal in infected metapopulations. Paradoxically, their induced negative feedbacks on female dispersal rates eventually decrease their own persistence in a host metapopulation. Adaptive dispersal or genotype-biased dispersal strategies are already known to rescue metapopulations from extinction [-]. Here we show that evolution of sex-specific dispersal could enhance persistence of hosts that experience infection by male-killing endosymbionts in two different ways: (i) by decreasing host extinction probabilities (minor effect) and (ii) by inducing curing of the host (major effect) through the combined action of increased male and decreased female dispersal. In addition to the finding that adaptive dispersal can promote the evolution of parasite resistance [], we here show that an evolutionary response of dispersal strategies to male-killing endosymbiont infection may already as such be an adaptation to escape male-killing endosymbiont infections.ConclusionMale killing endosymbionts induce the evolution of sex-specific dispersal, with prominent male-biased dispersal under conditions of low environmental stochasticity and high dispersal mortality. This male-biased dispersal emerges from kin-competition, which is (much) stronger in males because they are all offspring of the (few) females of their high relatedness. In addition, the evolution of sex-specific dispersal rates induces an evolutionary rescue mechanism by either decreasing endosymbiont fixation probabilities (which subsequently would lead to the crash of the host metapopulation) under conditions of high environmental stochasticity, or increasing endosymbiont extinction (curing) under conditions of low environmental stochasticity.MethodsThe modelThe landscapeFor our simulation experiments we use an extended version of an individual-based model [-] of insect dispersal in patchy landscapes of n (= ) habitat patches with equal carrying capacities K (= ).The individualEach individual is characterized by its sex, its affiliation with a specific patch (i), and by four alleles at two different diploid loci that determine male (dm), respectively female (df) dispersal propensity (see below). At initialization allele values are drawn randomly from a uniform distribution [–]. Further, individuals are characterized by their infection status (infected versus uninfected), which they solely inherit from their mother.Population dynamicsLocal population dynamics are governed by density-dependent reproduction of individuals. After mating with a randomly drawn local male (thus assuming polygyny), a female gives birth to Λ offspring, where Λ is a Poisson-distributed number with a patch- and time-specific mean, Λmean(t, patch). For each generation, the mean value of Λmean(t, patch) is drawn from a lognormal distribution with mean λ and a standard deviation σ ( ≤ σ ≤ ). In our simulations, λ was set to , a value typical for arthropod demography []. σ subsequently determines the degree of environmental fluctuations which are assumed to be uncorrelated in space and time. Offspring are randomly assigned to the male or female sex, but male offspring die immediately after conception if the mother is infected. Remaining offspring develop into mature individuals with a density-dependent survival probability s:() s = ( + a N i ) with with a=λ−1KHere Ni represents the population size in patch i. K is the carrying capacity of each patch in the metapopulation. It is important to recognize that the daughters of an infected mother do not exclusively benefit from the death of their brothers. Yet, in groups with infected females, population growth increases as female offspring are released from competition by males.DispersalIn our model, individuals simultaneously disperse before mating and production of offspring; each individual has only one opportunity to disperse. Dispersing individuals die with a probability μ (dispersal mortality), regardless of patch origin. For each individual its emigration probability d is determined by the mean value of their two sex-specific dispersal alleles, (dm,+dm,)/ respectively (df,+df,)/. We assume global dispersal; that is, a successful disperser reaches a patch in the landscape (except its home patch) with the same probability (-μ)/(n-). Dispersal probability is thus unconditional, i.e. we assume that dispersal decisions are not based on patch condition (e.g. infection rate, density or sex ratio). However, dispersal alleles were allowed to change by mutation, thus allowing for the evolution of sex-specific dispersal strategies To promote greater variability of genotypes in the first generations and to reduce the influence of mutations on the stability of the final result, we let mutation rates exponentially decrease from ~. to <. over the course of the simulation experiments ( generations; see e.g. []). A mutation comprises a change into a new random value from the uniform [–] distribution. For the case of symmetric, i.e sex-indifferent dispersal we assume that dispersal is governed by one locus only determining male as well as female dispersal propensity.Simulation experimentsTo infer how the presence of infections in the metapopulation influenced the evolution of dispersal probability in male and female hosts, we compared results of simulation experiments with and without endosymbiont infection. Experiments with infections were run with sex-specific as well as with symmetric dispersal behaviour. For both sets of experiments we estimated the metapopulation extinction probability (host and endosymbiont go both extinct) and the probability of endosymbiont extinction only ('curing'). Simulations were run with an initial fraction of infected females of I = . randomly distributed over patches. Additional simulations showed that the initial infection rate did not influence trait evolution (Bonte D, unpublished data). Because (i) we only observed an effect of carrying capacity on metapopulation dynamics and trait evolution when K << ; and (ii) typical insect habitats can be expected to rarely have capacities below K = [], we ran simulations only for values of K = .Simulations were run for different combinations of dispersal mortality (μ = ., ......; values) and environmental stochasticity (σ = , ., , . ....; values) resulting in a total of scenarios for each of the simulation experiments described above. All scenarios were replicated times. Global host as well as endosymbiont extinction probability was calculated as the number of simulation runs with metapopulation extinction divided by the total number of replicates for each scenario. Equally, the probability of curing was estimated by dividing the number of cured populations by the number of surviving host populations. In order to test the influence of kin-competition on the observed patterns in dispersal evolution, infection rate, and population dynamics, we applied a reshuffling algorithm by which local kin-structure is destroyed []. Here, individuals within each patch i were replaced by individuals of the same sex and with the same infection status randomly selected from the entire pool of individuals in the metapopulation. This experiment was also performed for the scenarios described above with replicates each.Model restrictionsIn our simulations, we do not allow for the evolution of host resistance. As shown by Dyer & Jaenike [], endosymbionts may have only limited capacity to counter newly evolved host resistance because of their small effective population sizes []. Yet the most influential restrictions are probably the applied mating system and the dispersal strategy. Because endosymbionts infections are predominantly associated with arthropods, we assume polygyny since this is the most common mating system in insects and other arthropods []; our simulation results would certainly be very different for a monogamous mating system. Despite evidence that male-killing bacteria may affect sexual selection [,], we only allowed single mating events in order to provide straight mechanisms of inheritance.Authors' contributionsThe work presented here was carried out in collaboration between all authors. DB conceptualized the research questions, implemented the model, analyzed the data, interpreted the results and wrote the paper. TH and HJP designed an earlier version of the model, discussed analyses, interpretation, and presentation. All authors have contributed to, seen and approved the manuscript.
|
PMC2583055.txt
|
TITLE: GEF-H1 Mediated Control of NOD1 Dependent NF-κB Activation by Shigella Effectors AUTHORS: Atsuko Fukazawa, Carmen Alonso, Kiyotaka Kurachi, Sonal Gupta, Cammie F. Lesser, Beth Ann McCormick, Hans-Christian Reinecker ABSTRACT: Shigella flexneri has evolved the ability to modify host cell function with intracellular active effectors to overcome the intestinal barrier. The detection of these microbial effectors and the initiation of innate immune responses are critical for rapid mucosal defense activation. The guanine nucleotide exchange factor H1 (GEF-H1) mediates RhoA activation required for cell invasion by the enteroinvasive pathogen Shigella flexneri. Surprisingly, GEF-H1 is requisite for NF-κB activation in response to Shigella infection. GEF-H1 interacts with NOD1 and is required for RIP2 dependent NF-κB activation by H-Ala-D-γGlu-DAP (γTriDAP). GEF-H1 is essential for NF-κB activation by the Shigella effectors IpgB2 and OspB, which were found to signal in a NOD1 and RhoA Kinase (ROCK) dependent manner. Our results demonstrate that GEF-H1 is a critical component of cellular defenses forming an intracellular sensing system with NOD1 for the detection of microbial effectors during cell invasion by pathogens. BODY: IntroductionThe tight junctions (TJs) of the intestinal epithelium act as a defensive barrier against microbial invaders and many pathogenic bacteria have developed mechanisms to overcome the tight junctional seal to exploit the intestinal epithelium as a replicative niche or to allow entry and dissemination into the host. TJs are multiprotein complexes which consist of transmembrane components, scaffolding proteins and signaling molecules that have the potential to initiate host immune surveillance upon disruption of the tight junctional seal []–[]. GEF-H1 was originally identified in mice as an oncoprotein member of the DBL family that activates RhoA in hematopoietic cells [],[]. GEF-H1 can associate with microtubules in non-polarized epithelial cells or the actin cytoskeleton in polarized epithelial cells and has been proposed to mediate cross talk between the two filament types []–[]. GEF-H1 associates with cingulin within TJs of epithelial cells and regulates paracellular permeability []–[]. Shigella species are human pathogens capable of colonizing the intestinal epithelium by exploiting epithelial cell functions and circumventing the host innate immune response []. The enteroinvasive pathogen S. flexneri can specifically target TJs to overcome the intestinal barrier and gain access to the basolateral membrane compartments of intestinal epithelial cells [], a prerequisite for the invasion of epithelial cells []. Shigella cell invasion depends on the release of a subset of effectors through the type III secretion system (T3SS) both around the bacterial surface and directly into the host cell [],[]. It is known that the ability of Shigella to invade epithelial cells and subsequently spread from cell to cell is pivotal in establishing intestinal infection. This process is associated with a strong inflammatory response, but the bacterial effectors and host cell response mechanisms leading to defense activation are not well understood. [],[].Signaling through nucleotide binding and oligomerization domain (NOD)-like receptor (NLR) NOD1 provides the intestinal epithelium with a mechanism for activating innate immunity during infection by invasive pathogenic Gram negative bacteria [],[]. Intracellular pattern recognition receptors such as NOD1 can detect the D-Glu-meso-diaminopimelic acid (DAP) dipeptide of S. flexneri in macrophages and activate NF-κB [],[]. S. flexneri employs a diverse array of effectors to alter the host actin cytoskeleton and innate immune responses including regulators for Rho GTPases and NF-κB [], []–[]. However, it has not been established which bacterial and host mediators signal the disruption of the apical junctional complex to initiate cellular defense responses in the intestinal epithelium.In this study, we demonstrate that GEF-H1 is a central component of pathogen recognition by NOD1. GEF-H1 and NOD1 together not only detect the presence of peptidoglycan (PGN)-derived muropeptides but also signal in response to Shigella effectors in the cytoplasm. GEF-H1 is recruited into bacterial invasion sites of S. flexneri and the subsequent RhoA activation is required for cell invasion. In addition, GEF-H1 is requisite for the activation of NF-κB dependent gene expression during Shigella cell invasion. NF-κB activation by GEF-H1 is independent from the detection of bacterial products by Toll-like receptor (TLR) and cytokine receptor signaling. Instead, GEF-H1 interacts with NOD1 and is required for NF-κB activation in response to γTriDAP. Importantly, we find that the Shigella effectors IpgB2 and OspB activate NF-κB by a mechanism that depends on both NOD1 and GEF-H1 and requires ROCK activation. GEF-H1 is a central component in a detection system that directs NF-κB activation in RhoA and RIP2 dependent pathways initiated by the action of bacterial effectors and intracellular pathogen pattern recognition.ResultsGEF-H1 is recruited into membrane ruffles induced by S. flexneri at tight junctions of polarized epithelial cellsInteraction of S. flexneri with the polarized epithelium results in the redistribution of TJ associated proteins which results in the loss of barrier function and allows access to the basolateral membrane compartment to facilitate cell invasion []. We utilized GFP expressing S. flexneri to determine the subcellular redistribution of GEF-H1 and its binding partner cingulin in TJs [] during invasion of polarized Madden Darby canine kidney (MDCK) model epithelial cell monolayers. As demonstrated in Figure 1A , S. flexneri initially was found specifically attached to cell-cell contacts and co-localized with GEF-H1 and cingulin in TJs of polarized MDCK cell monolayers. Subsequently, Shigella gained access to the tight junctional structure and altered its composition. As demonstrated in Figure 1B and C , Shigella induced membrane ruffles that extended above the TJs in MDCK cells. GEF-H1 but not cingulin was found to be recruited into bacterial entry sites ( Figure 1B and C ). When Shigella began to gain access to the cytoplasm, both GEF-H1 and cingulin were removed from the tight junctional area of infected cells ( Figure 1D ). Afterwards, Shigella was found either free in the cytoplasm or in compartments which were GEF-H1 positive and the neighboring epithelial cells started to contract above the infected cell ( Figure 1E ). Of note, cells in close proximity to infected cells demonstrated increased GEF-H1 expression in the cytoplasm. Together, these data demonstrate that GEF-H1 is recruited into membrane ruffles induced upon disruption of TJs and the modifications of the cytoskeleton by Shigella../journal.ppat..g001Figure 1GEF-H1 is recruited into membrane ruffles induced by S. flexneri.XY, XZ and YZ sections as well as three-dimensional reconstructions (volume render) of confocal image series of GEF-H1 recruitment to the invasion sites of S. flexneri. Polarized MDCK monolayers were exposed to green fluorescent protein (GFP)-expressing S. flexneri and then fixed and stained at different time points. Endogenous GEF-H1 was stained with anti-GEF-H1 antibody and Texas red secondary antibody. Cingulin was stained with anti-cingulin antibody and Cy5 secondary antibody. Black arrows locate the cross section level. Bars indicate µm. (A) Initially, S. flexneri was found attached to TJs, co-localizing with GEF-H1 and cingulin. (B and C) Thereafter, Shigella gained access to the TJs and induced membrane ruffles and pedestals that extended above the TJs. White arrows represent area indicated by white frame in (B). GEF-H1 was recruited from tight junctional complexes into bacterial entry sites, while cingulin was not and remained associated with the TJs. (D) Subsequently, Shigella was found within the cytoplasm of MDCK cells and GEF-H1 and cingulin were removed from the tight junctional area of infected cells. (E) Upon cell invasion, Shigella was found inside the epithelial monolayer either free in the cytoplasm or associated with intracellular vesicles containing GEF-H1 (white arrows). The cytoplasm of the infected cells appeared to retract and the neighboring epithelial cells started to close above the infected cells.GEF-H1 function is required for host cell invasion by S. flexneri The recruitment of GEF-H1 to invasion sites raised the possibility that GEF-H1 was targeted by bacterial effectors to activate Rho GTPases at the basolateral membrane compartments as a prerequisite for cell invasion by S. flexneri []. We therefore determined the requirement of GEF-H1 and its contribution to Rho GTPase activation during Shigella cell invasion. Invasion of MDCK monolayers by S. flexneri was associated with an increase of GEF-H1 expression in Triton X-soluble protein fractions. The membrane associated pool of GEF-H1 remained high during the interaction of S. flexneri with the polarized epithelium, suggesting that GEF-H1 was rapidly redistributed from apical junctions to new membrane associated binding partners at bacterial entry sites or basolateral membrane compartments ( Figure 2A ). Cingulin was redistributed from the TJ associated Triton X-insoluble to the -soluble protein fractions during the initial phase of Shigella cell invasion potentially initiating the release of GEF-H1 ( Figure 2A ). However, within minutes, cingulin levels in TJ associated membrane fractions recovered. In contrast, cytosolic GEF-H1 protein levels continued to increase over the minute observation period ( Figure 2A ). The increase in cytosolic GEF-H1 expression observed in soluble protein fractions may not only result from subcellular redistribution, but could also include protein neosynthesis, since GEF-H1 mRNA expression was upregulated .±.-fold within hours of S. flexneri infection in MDCK cells ( Figure 2B )../journal.ppat..g002Figure 2GEF-H1 mediates RhoA activation during S. flexneri cell invasion.(A) Western blot analysis of subcellular distribution of GEF-H1 and cingulin upon S. flexneri invasion of MDCK cell monolayer. Densitometric analysis of GEF-H1 and cingulin in soluble and insoluble protein fractions during Shigella infection. Bars represent mean±SD (* p<. compared to the rest of time points). (B) Real-time PCR assessment of GEF-H1 expression in response to S. flexneri invasion of MDCK cells. Bars represent mean±SD (* p<. compared to non-infected cells). (C) GST pull down assays to assess activation of small GTPases in response to GEF-H1 expression. HEK293 cells were transfected with GEF-H1, RhoA, Rac1 and Cdc42 expression vectors, GEF-H1 (Y395A) mutant and constitutive active mutants for Rac1 (G12V) and Cdc42 (G12V). (D) GST pull down assays to assess activation of RhoA to S. flexneri infection. HEK293 cells were transfected with GEF-H1 mutant (Y395A) defective in nucleotide exchange or the empty plasmid as a control and incubated with S. flexneri for minutes. Expression of GEF-H1 (Y395A) decreased baseline activation and prevented the rapid activation of RhoA during Shigella invasion of HEK293 cells. (E) NF-κB activation in response to S. flexneri invasion of HEK293 cells in the absence or presence of dominant negative GEF-H1 (Y395A) mutant. Bars represent mean±SD (* p<. compared to control). (F) Gentamicin protection assays of S. flexneri infection of HEK293 cells transfected with GEF-H1 (Y395A) or dominant negative RhoA (T19N) expression vectors. Bars represent mean±SD (* p<. compared to control plasmid). (G) Gentamicin protection assays of S. flexneri infection of HEK293 cells in the presence of two different GEF-H1 siRNAs. Bars represent mean±SD (* p<. compared to control siRNA).Increases in expression of GEF-H1 in HEK293 cells led to the activation of RhoA, but not Cdc42 and Rac1, as demonstrated by the induction of Rhotekin binding to activated RhoA but not of PAK1 to Rac1 or Cdc42 in GST pull down assays ( Figure 2C ). GEF-H1 was responsible for RhoA activation during cell invasion, since expression of a GEF-H1 mutant (Y395A) defective in nucleotide exchange prevented the activation of RhoA during Shigella invasion of HEK293 cells ( Figure 2D ). In addition, the expression of the catalytically inactive GEF-H1 reduced the basal levels of RhoA activation ( Figure 2D ). Furthermore, expression of GEF-H1 (Y395A) also reduced NF-κB reporter gene activation by ±% during infection of HEK293 cells by S. flexneri ( Figure 2E ). Expression of GEF-H1 (Y395A) or RhoA (T19N) ( Figure 2F ) or the depletion of GEF-H1 expression by two distinct siRNAs significantly reduced cell invasion of HEK293 by S. flexneri ( Figure 2G ). Together these findings demonstrate a critical function of GEF-H1 in RhoA activation required for cell invasion by S. flexneri.GEF-H1 mediates NF-κB activation during Shigella cell invasionSince RhoA activation has been linked to the activation of NF-κB [], we determined whether enhanced cytoplasmic GEF-H1 expression was linked to NF-κB activation during S. flexneri cell invasion. Expression of GEF-H1 in HEK293 cells induced activation of NF-κB in a dose dependent fashion ( Figure 3A ). The activation of the NF-κB pathway by GEF-H1 was dependent on its RhoA specific GEF activity. Co-expression of the dominant negative allele of RhoA (T19N) prevented GEF-H1 mediated NF-κB activation in HEK293 cells ( Figure 3A ). Furthermore, the RhoA target kinase p160/ROCK was required for GEF-H1 mediated NF-κB activation, since the presence of ROCK inhibitor Y27632 decreased GEF-H1 mediated NF-κB activation ( Figure 3A ). GEF-H1 expression induced IL- promoter activity in HEK293 cells which was prevented in the presence of the RhoA (T19N) mutant ( Figure 3B ). GEF-H1 overexpression resulted in the phosphorylation and subsequent degradation of IκBα ( Figure 3C ). Furthermore, the activation of NF-κB through GEF-H1 was partially dependent on the inhibitor of NF-κB kinase complex subunit β (IKKβ), since either the kinase negative mutant IKKβ (K44M) or the dominant negative mutant IKKβ (SS/AA) inhibited GEF-H1 mediated activation of NF-κB ( Figure 3D )../journal.ppat..g003Figure 3GEF-H1 mediates RhoA dependent NF-κB activation.(A) GEF-H1 mediated NF-κB reporter gene expression in the absence or presence of dominant negative RhoA (T19N) or the ROCK inhibitor Y27632 in HEK293. Bars represent mean±SD (*p<. compared to GEF-H1 alone). (B) GEF-H1 and RhoA mediated IL- reporter gene expression. HEK293 cells were transfected with GEF-H1 in the absence or presence of dominant negative RhoA (T19N) or the constitutive active RhoA mutant (G14V) plasmids. Bars represent mean±SD (*p<., compared to GEF-H1 alone). (C) Western blot analysis of IκBα phosphorylation and protein degradation in response to GEF-H1 expression in HEK293 cells. (D) NF-κB reporter activation by GEF-H1 in the absence or presence of IKKβ kinase negative mutant K44M or dominant negative SS/AA mutant in HEK293 cells. Bars represent mean±SD (*p<. compared to control, **p<. compared to GEF-H1 alone). (E) NF-κB reporter activation by S. flexneri in HEK293 cells after depletion of NOD1 and GEF-H1 with specific siRNAs. Bars represent mean±SD (* p<. compared to Shigella responses in the presence of control siRNAs). (F) Western blot demonstrating that specific and effective knock down of endogenous and exogenous GEF-H1 as well as overexpressed NOD1 in HEK293 cells.NF-κB activation upon S. flexneri cell invasion has been linked to recognition of this intracellular pathogen by NOD1 []. Consistent with these observations, depletion of NOD1 with specific siRNAs but not control siRNAs inhibited S. flexneri mediated NF-κB activation in HEK293 cells by ±% ( Figure 3E ). However, NF-κB activation in response to infection by S. flexneri was also inhibited in a dose dependent manner by the siRNA mediated depletion of GEF-H1 ( Figure 3E ). Inhibition of GEF-H1 and NOD1 expression together decreased NF-κB activation upon S. flexneri infection by ±% ( Figure 3E ). Depletion of either GEF-H1 or NOD1 by the utilized siRNAs was specific ( Figure 3F ). Together, these data demonstrate that GEF-H1 is required for NF-κB activation during cell invasion by S. flexneri and that GEF-H1 induced RhoA activation can contribute to the activation of the canonical NF-κB pathway and induction of IL- transcription.GEF-H1 is required for NOD1 dependent NF-κB activationWe next determined the role of GEF-H1 in the activation of NF-κB by the NOD1 ligand γTriDAP. In these experiments, HEK293 cells transfected with the NF-κB luciferase reporter plus NOD1, GEF-H1 or control siRNAs were exposed to either γTriDAP or control compound αTriDAP (Ala-αGlu-meso-DAP). Addition of γTriDAP to HEK293 cells induced a .±.-fold increase in NF-κB promoter activity ( Figure 4A ). As demonstrated in Figure 4A , depletion of GEF-H1 prevented NF-κB activation through specific NOD1 ligands more efficiently than the inhibition of NOD1 expression (±.% versus ±.% reduction), implicating GEF-H1 as a critical component in the recognition of specific PGN-derived muropeptides. In contrast, depletion of neither GEF-H1 nor NOD1 significantly inhibited NF-κB activation in response to TNF-α receptor signal transduction in HEK293 cells ( Figure 4A ). Depletion of GEF-H1 also prevented NF-κB activation induced by overexpression of NOD1 by ±% ( Figure 4B ). Conversely, overexpression of GEF-H1 together with NOD1 synergistically upregulated NF-κB dependent gene transcription by up to ±-fold, while expression of GEF-H1 or NOD1 alone increased NF-κB dependent transcription by ±.-fold and .±.-fold, respectively ( Figure 4C ). However, RhoA activation was not required for signaling of the NOD1 ligand γTriDAP or TNF-α since expression of GEF-H1 (Y395A) or RhoA (T19N) failed to inhibit NF-κB activation ( Figure 4D and E ). Instead, GEF-H1 was able to interact with NOD1 directly or through intermediates, since both were found to co-immunoprecipitate when antibodies directed against either NOD1 or GEF-H1 but not control antibodies were used ( Figure 4F ). Confocal microscopic analysis revealed that NOD1 was recruited to the basolateral membrane compartment in polarized epithelial cells but co-localized with endogenous or exogenously expressed GEF-H1 in cell-cell contacts in MDCK cells or HEK- cells ( Figure 4G ). We concluded from these experiments that GEF-H1 is requisite for signal transduction by NOD1 independent of its GEF activity../journal.ppat..g004Figure 4GEF-H1 interacts with NOD1 and is required for NOD1 dependent NF-κB activation.(A) NF-κB activation in response to active and inactive NOD1 ligands in the absence or presence of GEF-H1 and NOD1 siRNA in HEK293 cells. Bars represent mean±SD (* p<., compared to responses in the presence of control siRNA and γTriDAP). (B) NF-κB activation in response to overexpression of NOD1 in the presence of control or GEF-H1 siRNA in HEK293 cells. Bars represent mean±SD (* p<. compared to control siRNA). (C) NF-κB activation in response to transfection of indicated amounts of expression vectors encoding GEF-H1 and NOD1 in HEK293 cells. Bars represent mean±SD (*p<. compared to GEF-H1 and NOD1 alone). (D and E) RhoA activation is not required for γTriDAP and TNFα signaling. NF-κB activation in response to active and inactive NOD1 ligands and TNFα in HEK293 cells transfected with GEF-H1 (Y395A), dominant negative RhoA (T19N) or control expression vectors. Bars represent mean±SD (*p = NS, control vector vs GEF-H1 (Y395) or RhoA (T19) in the presence of γTriDAP). (F) Co-immunoprecipitation of GEF-H1 and NOD1 with indicated antibodies in HEK293 cells. (G) Confocal microscopic image analysis of the subcellular localization of NOD1, GEFH1 and E-cadherin. MDCK monolayers and HEK cells were transfected with HA-NOD1 vector alone or in addition to vsv-GEF-H1 and then fixed and stained for confocal microscopic analysis. Endogenous GEF-H1 was stained with anti-GEF-H1 antibody and Texas red secondary antibody. E-cadherin was stained with anti-E-cadherin antibody and Texas red secondary antibody. HA-NOD1 was stained with anti-HA antibody and Cy5, Texas Red or FITC secondary antibody; vsv-GEF-H1 was stained with anti-vsv antibody and Cy5 or FITC secondary antibody. Bars indicate µm.GEF-H1 mediates NF-κB activation initiated by Shigella effectorsDuring the course of infection, S. flexneri modulates Rho GTPase function and NF-κB activation through a number of effectors delivered directly into host cells [],[]. We screened for S. flexneri effectors that are able to target the cytoskeleton or replace NF-κB signaling during infection ( Figure 5A and B ). When GFP tagged Shigella effectors were transfected into MDCK cells, IpgB1 and IpgB2 localized to cellular junctions and intracellular vesicular compartments, while OspB associated only with intracellular compartments ( Figure 5A ). OspG, which has been shown to inhibit NF-κB activation, also associated with a subcellular compartment close to the cell membrane. In contrast, OspF remained cytoplasmic, while IpgD was found to be enriched in cell nuclei. Also, VirA demonstrated a cytoplasmic staining pattern, but did not activate NF-κB, although this effector rapidly induced membrane blebbing and cell death when expressed in MDCK cells ( Figure 5A )../journal.ppat..g005Figure 5GEF-H1 and NOD1 mediate NF-κB activation induced by S. flexneri effectors.(A) Confocal microscopic image analysis of MDCK cells transfected with indicated GFP-tagged S. flexneri effectors. Bars indicate µm. (B) NF-κB activation in response to GFP-tagged S. flexneri effectors in HEK293 cells. (C) Three-dimensional reconstructions of confocal microscopic image series of MDCK cells transfected with indicated GFP-tagged S. flexneri effectors and immunostained for endogenously expressed GEF-H1. Bars indicate µm, arrows indicate co-localization of GEF-H1 and S. flexneri effectors. (D) NF-κB activation in response to IpgB1, IpgB2 and OspB expression in the absence or presence of control or GEF-H1 specific siRNAs (*p<. compared to the expression of GFP control alone, **p<. compared to siRNA control).When expressed as GFP fusion proteins in HEK293 cells, IpgB2, OspB and, to a lesser extent, IpgB1, were able to activate NF-κB promotor activity by .±., .±. and .±.-fold above baseline activity of GFP control plasmid. In contrast, OspG, OspF, IpgD, as well as VirA, failed to significantly upregulate NF-κB promoter activity ( Figure 5B ).Furthermore, the subcellular distribution of IpgB1 and IpgB2 partially overlapped with GEF-H1 in cellular junctions of MDCK cells. In contrast, OspB remained in a cellular compartment which had very little overlap with the GEF-H1 containing membrane compartments in MDCK cells ( Figure 5C ).Surprisingly, in the absence of GEF-H1, all S. flexneri effectors failed to induce significant NF-κB activation ( Figure 5D ). When expressed in MDCK cells, IpgB1, IpgB2 and OspB induced NF-κB activity .±., .±. and .±.-fold, respectively, above background levels found in the presence of GFP alone ( Figure 5D ). SiRNA mediated depletion of GEF-H1 significantly reduced the NF-κB activation induced in the presence of IpgB2 and OspB ( Figure 5D ). Collectively, these experiments demonstrated that Shigella effectors can directly or indirectly activate NF-κB activation independently of muramyl dipeptides by a mechanism that requires the presence of GEF-H1.GEF-H1 and NOD1 are required for NF-κB activation by muropeptides and Shigella effectorsTo further define the mechanism by which OspB and IpgB2 induces NF-κB activation, we depleted either GEF-H1 or NOD1 and expressed OspB and IpgB2 in the presence of γTriDAP or control peptide and determined NF-κB activation in HEK293 cells. Remarkably, in the presence of additional γTriDAP, NF-κB activity in OspB and IpgB2 transfected cells increased up to .±. and .±.-fold, respectively, over baseline levels induced by transfection with the GFP control plasmid ( Figure 6A ). Surprisingly, the induction of NF-κB activity by γTriDAP, OspB and IpgB2 alone or in the presence of γTriDAP was dependent on NOD1, as well as GEF-H1, since depletion of both mediators by specific siRNAs but not control siRNAs reduced NF-κB activity down to baseline levels ( Figure 6A ). Depletion of GEF-H1 decreased NF-κB activity induced by γTriDAP by .±.% by OspB in the presence of γTriDAP by .±.% and the activity resulting from IpgB2 overexpression by .±.% ( Figure 6A ). Depletion of NOD1 reduced the NF-κB activity induced by γTriDAP, OspB and IpgB2 by .±.%, ±.% and .±.%, respectively ( Figure 6A ). The inhibition of NF-κB activation was more pronounced after depletion of GEF-H1 compared to NOD1, possibly due to different knockdown efficiencies of the gene specific siRNAs in these experiments. Specific siRNA mediated depletion of NOD1 or GEF-H1 were target specific and did not affect expression levels of IpgB2 or OspB in our cell system ( Figure 6B ). The synergistic increase in NF-κB activation induced by γTriDAP and Shigella effector together was dependent on the caspase recruitment domain (CARD)-containing serine/threonine kinase RIP2 (also known as Rick, CARDIAK, CCK and Ripk2) ( Figure 6C ). However, NF-κB activation induced by IpgB2 or OspB alone was found to be independent of RIP2 and instead was dependent on ROCK activation ( Figure 6D and E )../journal.ppat..g006Figure 6GEF-H1 and NOD1 are required for NF-κB activation by muropeptides and Shigella effectors.(A) NF-κB activation in response to control or active NOD1 ligand (γTriDAP) in the absence or presence of Shigella effectors and control, GEF-H1 or NOD1 siRNAs (* p<. compared to responses in the presence of control siRNA and γTriDAP, **p<. compared to responses in the presence of control siRNA and OspB, # p<. compared to responses in the presence of control siRNA and IpgB2). (B) Depletion of GEF-H1 or NOD1 did not affect protein expression levels of the indicated GFP-tagged S. flexneri effectors in HEK293 cells. (C) NF-κB activation in response to control or active NOD1 ligand (γTriDAP) in the absence or presence of Shigella effectors and control or RIP2 siRNAs (*p<. compared to responses in the presence of control vector, control siRNA and γTriDAP, **p<. compared to responses in the presence of control siRNA, OspB and γTriDAP, #p<. compared to responses in the presence of control siRNA, IpgB2 and γTriDAP). (D) IpgB2 and OspB signaling is RIP2 independent. NF-κB activation in response to control or Shigella effectors expression vectors in the presence of control or RIP2 siRNAs (* p<. compared to responses in the presence of control siRNA). (E) IpgB2 and OspB signaling is ROCK dependent. NF-κB activation in response to control or Shigella effectors expression vectors in the presence or absence of ROCK inhibitor Y- (*p<. compared to responses in the absence of Y-). (F) Gentamicin protection assay with wild type and indicated Shigella mutants in HEK293 cells (*p<. compared to wild type Shigella). (G) NF-κB promoter activation in HEK293 cells in response to infection by wild type and indicated Shigella mutants (*p<., compared to wild type Shigella).We next carried out gentamicin protection assays with Shigella mutants deficient in IpgB1, IpgB2 or OspB to determine the role of these effectors in pathogen uptake and intracellular survival. As demonstrated in Figure 6F , Shigella deficient in IpgB2 were not impaired in their ability to invade and survive in HEK293 cells, while significantly lower numbers of Shigella lacking either IpgB1 or OspB were recovered from infected cells after minutes compared to wild type Shigella ( Figure 6F ). However, Shigella mutants deficient in the expression of IpgB1 IpgB2 or OspB were all characterized by significantly reduced NF-κB activation during invasion of HEK293 cells compared to wild type S. flexneri ( Figure 6G ). These experiments established that GEF-H1 and NOD1 are central to the detection of intracellular Shigella effectors, in addition to their function in recognizing PGN-derived muropeptides. NF-κB activation during Shigella cell invasion involves both RhoA and RIP2 dependent activation pathways which are dependent on GEF-H1.DiscussionThe epithelial interface is involved in constant cross talk with the intestinal microbiota through molecular mechanisms that integrate intestinal epithelial barrier function with mucosal immune regulation. Failure to accurately monitor the intestinal environment or respond adequately to challenges by pathogens results in a breakdown of the intestinal barrier. Multiple signaling components are localized at epithelial TJs [],[]. Most of these signal mediators have been identified through their function in the regulation of epithelial polarization, differentiation and growth control, but very little information exists about their contribution to mucosal host defense responses. We now show that the disruption of TJs by Shigella effectors is linked to the activation of innate immune responses through GEF-H1 controlling NOD1 mediated NF-κB activation.GEF-H1 was recruited to Shigella induced membrane ruffles similarly to NOD1 which has been shown to be enriched in bacterial entry sites in an actin dependent mechanism also required for signal transduction []. Shigella effectors secreted by the T3SS could initiate NF-κB signaling before the PGN release from Shigella multiplying within epithelial cells that has been shown to activate the NOD pathway [].GEF-H1 is indispensable for NOD1 mediated NF-κB activation by γTriDAP, and by the Shigella effectors OspB and IpgB2. However, the downstream signaling events leading to NF-κB activation were specific for either γTriDAP or the Shigella effectors. S. flexneri infection induces NOD1 oligomerization via the homophilic CARD-CARD interaction allowing transient recruitment of RIP2 and IKK, which phosphorylates IκB leading to prolonged activation of NF-κB [],[]. In our experiments, GEF-H1 dependent activation of NF-κB by γTriDAP and its synergistic effect on Shigella effector signaling was RIP2 dependent. The ability of GEF-H1 to mediate γTriDAP initiated NOD1 signal transduction was independent of the GEF function of GEF-H1 and the activation of RhoA. Instead, in γTriDAP signaling, GEF-H1 directly interacted with NOD1 serving as a signaling adaptor through protein motifs which need to be defined in future experiments. In contrast, OspB and IpgB2 mediated NF-κB activation required RhoA mediated activation of ROCK but not RIP2. These findings demonstrate that GEF-H1 is a central component of RhoA and Rip2-mediated NF-kB activation during Shigella cell invasion.In the early stages of Shigella infection, effectors delivered by the T3SS include IpgB1, IpgB2, IpgD and VirA both around the bacterial surface and directly into the host cell. While both IpgB1 and IpgB2 associated with cellular junctions, only IpgB2 was a potent inducer of the GEF-H1 dependent activation of NF-κB. Both IpgB1 and IpgB2 have been linked to Rho GTPase function. IpgB1 is presumed to have major roles in producing membrane ruffles by activating Rac- through ELMO and DOCK180, a Rac- guanine nucleotide exchange factor []. IpgB1 function has been linked to the activation of RhoG, resulting in downstream activation of Rac []. IpgB2 is an IpgB1 homolog that binds to mDia1 which facilitates actin nucleation and the Rho kinase ROCK through its GTPase binding domains []. Therefore, IpgB2 may mimic the activity of RhoA in our experiments resulting in ROCK dependent NF-κB activation. Invasiveness of IpgB1 or IpgB2 mutant Shigella has been recently assessed []. Consistent with our results, Shigella IpgB2 mutants have been found to exhibit the same invasive capacity as the wild type strain in non-polarized and polarized intestinal epithelial cells. In contrast, IpgB1 mutants were % less invasive in the non-polarized epithelium but slightly more invasive in the polarized epithelium []. Therefore, the reduced NF-κB activation in response to IpgB2 mutants was not due to reduced cell invasion, while IpgB1 and OspB play a role in the cell invasion process itself, which may contribute to the reduced NF-κB activation seen in response to Shigella mutants lacking these effectors.Independent of its role in cell invasion, OspB was able to activate GEF-H1 dependent NF-κB activation. OspB is a MxiE regulated gene and is therefore likely expressed after S. flexneri invades host cells. It was therefore surprising that we recovered a reduced number of OspB deficient Shigella mutants in our gentamicin protection assays. OspB expressed in the cytoplasm could initiate GEF-H1 and NOD1 dependent NF-κB signaling after the escape of S. flexneri from the phagosome. It needs to be determined if NF-κB activation by this effector contributes to anti-apoptotic regulation supporting survival of the pathogen in the host cell.VirA is delivered into the host cell cytoplasm near the site of bacterial entry and induces local microtubule degradation []. Degradation of microtubules by the VirA related effector EspG from enteropathogenic E. coli results in the release of various microtubule associated proteins, including GEF-H1 []. VirA activity is assumed to contribute to ruffle formation during Shigella invasion through cross talk between RhoA and Rac-. In our studies, VirA induced responses resulted in rapid induction of cell death in epithelial cell lines tested and failed to induce NF-κB. The function of VirA might be more closely associated with the ability of Shigella to move within the cell than with direct cell invasion. During multiplication within the epithelium, Shigella secretes additional effectors including OspF and OspG. Consistent with previous findings, OspG and OspF both failed to induce NF-κB. OspG has been demonstrated to interfere with the IκBα degradation resulting in repression of NF-κB activation and downregulated inflammatory response to infection []. OspF has a specific phosphatase activity that dephosphorylates and inactivates MAPK leading to blockage of phosphorylation of histone H3 which is required for transcription of a subset of NF-κB regulated genes [],[].In polarized epithelial cells, GEF-H1 is associated with apical polarization complexes concentrated at TJs from which it is released and redistributed to bacterial invasion sites during Shigella infection. Our experiments demonstrate that, in epithelial cells, GEF-H1 is essential for RhoA activation required for cell invasion by Shigella. Like Cdc42 and Rac, RhoA activity is required for Shigella entry []–[]. RhoA is not critical for Shigella induced actin polymerization, but is required for the recruitment of ezrin to bacterial entry sites [].We demonstrate that RhoA activation through GEF-H1 can contribute to the activation of NF-κB upon cell invasion by Shigella. Invasion by extracellular and intracellular pathogens is sensed by various signaling pathways that converge to activate NF-κB which can initiate inflammatory mediator secretion, but is also a critical cell survival signal []. Members of the Rho family of GTPases are involved specifically in the regulation of NF-κB dependent transcription. RhoA mediated activation of atypical protein kinase C can induce phosphorylation of p65/Rel-A at serine , a site that is crucial for modulating the interaction between Rel-A and CREB binding protein []. A recent report demonstrated that Rac1 can interact and co-localize with NOD2, raising the possibility that regulators of Rho GTPases contribute specifically to intracellular innate immune recognition [].Rho GTPases have been shown to be important factors in TLR2, TLR4 and TLR9 signaling [], []–[]. LPS released by Shigella as they escape from the phagosome into the macrophage cytoplasm activates caspase- which induces production of IL-1β and cell death []. However, TLR mediated recognition was not responsible for the rapid activation of NF-κB in response to S. flexneri, its effectors and GEF-H1, since our experiments were carried out in HEK293 cells, which express low levels of TLR1- mRNA and consequently do not respond to TLR ligands []. This may resemble the situation encountered by Shigella in vivo, where intestinal epithelial cells lack membrane expression of CD14 and remain relatively refractory to activation by extracellular LPS or noninvasive Gram negative bacteria []–[]. However, our results do not exclude the possible involvement of GEF-H1 or other Rho-GEFs in bacterial recognition through TLRs in other cell types or in the recognition of other pathogens.Collectively, our findings support a model in which GEF-H1 is a new essential signal mediator in the activation of host defense initiated by the intracellular recognition of enteropathogens by NOD1 ( Figure ). GEF-H1 has two important functions, one in sensing muramyl dipeptides through NOD1 that is independent of its GEF activity, and another in the activation of NF-κB by Shigella effector proteins, which requires its GEF activity and the activation of RhoA. GEF-H1 and NOD1 are central components in RhoA and RIP2 dependent NF-κB signaling pathways activated upon Shigella cell invasion. Shigella release of LPS and PGN into host cells has been considered to be the main cause of the strong inflammatory response that is induced during Shigella infections []. Our experiments extend this model by demonstrating that NOD1 function is not limited to the detection of PGN fragments but, together with GEF-H1, functions in sensing the intracellular action of S. flexneri peptide effectors. This mechanism may allow intestinal epithelial cells to rapidly detect Shigella effectors and activate the NF-κB pathway to initiate survival and inflammatory signals. Our findings raise the interesting possibility that other NLR family proteins are also involved in the sensing of modifications to cellular functions by microbial effectors in addition to their role as pattern recognition receptors. Recognition of altered cell function by bacterial effectors may be important for the ability to distinguish between pathogenic and commensal microorganisms which cannot be achieved based on extracellular pattern recognition receptors alone, since many of their ligands are commonly expressed by commensal, as well as infectious microbiota../journal.ppat..g007Figure 7Proposed model of GEF-H1 function in the activation of innate immune responses to Shigella cell invasion.Upon cell attachment Shigella releases effectors through the T3SS into the host cell to regulate the actin cytoskeleton and facilitate invasion. This releases GEF-H1 from its binding partners in tight junctions inducing its interaction with NOD1. GEF-H1 has at least two functions in the activation of cellular defenses to microbial effectors. One mediates the sensing of muramyl dipeptides through NOD1 that is independent of its GEF activity, and the other is required for the activation of NF-κB by Shigella effector proteins, which requires its GEF activity mediated by the DBL homolgous domain (DH) and the activation of RhoA. RhoA activation by GEF-H1 is required for Shigella cell invasion through the recruitment of ezrin and regulation of the cytoskeleton potential cooperating with the RhoA mimicry of IpgB2. RhoA activation by GEF-H1 contributes to NF-κB activation through ROCK dependent phosphorylation of NF-κB proteins. Through interaction with GEF-H1, NOD1 can function in sensing the intracellular action of S. flexneri peptide effectors in addition to its role in the detection of peptidoglycan fragments.Materials and MethodsExpression vectorsPlasmid encoding vsv-tagged Canis familialis GEF-H1 has been described previously []. The GEF-H1 (Y395A) mutant was generated by introducing a tyrosine to alanine mutation into residue in the conserved QRITKY sequence of the DH domain that is responsible for GEF activity. Expression plasmids for dominant negative mutants of RhoA (T19N) or constitutive active mutants of Rac1 (G12V) and Cdc42 (G12V) were purchased from UMR cDNA Resource Center (University of Missouri Rolla, Rolla, MO). The pEAK13 vectors expressing the kinase deficient mutant IKKβ (K44M) or the dominant negative mutant IKKβ SS/AA in which the serines at positions and in the activation loop were replaced by alanines [] were kindly provided by Dr. Ramnik Xavier, Massachusetts General Hospital, Boston, MA, USA. The pCI-HA-CARD4 vector has been described previously []. GFP Shigella effector expression vectors for IpgB1, IpgB2, OspB, OspG, OspF, IpgD and VirA were generated by RT-PCR and subcloned into pEGFP-C1 (Clontech Laboratories, Mountain View, CA).Antibodies and reagentsMouse anti-GEF-H1 antibody has been described previously [] and was kindly provided by Dr. Karl Matter, University College London, London, UK. Rabbit anti-HA (Y-), rabbit anti-GFP (FL), mouse anti-RhoA (26C4) and goat anti-actin (I-) antibodies were obtained from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Mouse anti-vsv antibody (P5D4) was purchased from Sigma (St. Louis, MO). Rabbit anti-cingulin was purchased from Zymed (San Francisco, CA). Mouse anti-Rac1, mouse anti-Cdc42 and mouse anti-E-cadherin antibodies were purchased from BD Bioscience (San Jose, CA). Mouse anti-HA antibody was purchased from Roche (Indianapolis, IN). All horseradish peroxidase (HRP)-conjugated antibodies against mouse, rabbit or goat IgG were purchased from GE Healthcare (Piscataway, NJ). FITC conjugated anti-rabbit IgG, Texas red conjugated anti-mouse and anti-rabbit IgG and Cy5 conjugated anti-rabbit IgG were purchased from Jackson ImmunoResearch Laboratory (West Grove, PA). γTriDAP and αTriDAP were purchased from AnaSpec (San Jose, CA). Y27632 was purchased from Sigma.Cell culture and transfection protocolsHEK293 cells, MDCK cells were purchased from American Type Culture Collection (Manassas, VA) and maintained in Dulbecco's modified Eagle's medium (DMEM) with g/l glucose and sodium pyruvate (Cellgro, Herndon, VA) supplemented with % heat inactivated fetal bovine serum and a .% penicillin G/streptomycin mixture. HEK293 cells were plated hours before transfection with FuGene6 (Roche Diagnostics, Basel, Switzerland) or Lipofectamine (Invitrogen, Carlsbad, CA) according to the manufacturer's protocols. MDCK cells were seeded at high density onto . µm-pore Transwell™ filter culture inserts (Corning Incorporated, Corning, NY) and studied – hours thereafter, when a transepithelial resistance of ≥ ohm/cm2 was reached. To induce overexpression of genes in epithelial monolayers, MDCK cells were transfected with Nucleofector kits (Amaxa Biosystems, Gaithersburg, MD) according to the manufacturer's protocol.Immunofluorescence stainingEpithelial monolayers were washed three times with HBSS buffer containing Mg2+ and Ca2+, pH ., (HBSS+) and fixed with acetone for seconds at −°C or with methanol for minutes at −°C. For blocking, fixed monolayers were incubated with % normal donkey serum for minutes at room temperature. Primary antibodies diluted with .% normal donkey serum were incubated for hours at °C. Dilution ratios were ∶ for anti-GEF-H1, ∶ for anti-vsv, ∶ for anti-cingulin and anti-E-cadherin antibodies and ∶ for anti-HA antibody. After washing three times, samples were incubated with secondary antibodies (∶ for each secondary antibody).Protein separation and immunoprecipitationCells prepared as described above were washed with HBSS+ and proteins separated into Triton X--soluble and -insoluble fractions or homogenated in NP- buffer (% NP-, mM Tris-HCl at pH ., mM NaCl, mM EDTA, mM EGTA, mM Na3VO4, mM NaF). Electrophoresis and transfer were performed as previously described []. Membranes were blocked with % non-fat dry milk in Tris-buffered saline (TBS) at room temperature for hour and incubated with primary antibodies diluted in blocking solution to a ratio of ∶ at °C overnight. After washing in TBS with .% Tween- (TBS-T), membranes were incubated with appropriate horseradish peroxidase conjugated secondary antibody diluted in blocking buffer for hour at room temperature. Blots were washed times with TBS-T and hybridized bands were detected by Amersham ECL Western blotting detection reagent (GE Healthcare, Piscataway, NJ). HEK293 cells were transfected in well plates using Fugene6 transfection reagent (Roche) according to the manufacturer's protocol. After hours, proteins were separated as described above. Lysates were incubated with protein G plus agarose (Calbiochem, San Diego, CA) at °C for minutes and pre-cleared. Pre-cleared lysates were incubated with anti-vsv or anti-HA antibodies at °C overnight followed by incubation with agarose beads at °C for hours. Precipitated proteins were collected by centrifugation and washed times in washing buffer (.% NP-, mM Tris-HCl at pH ., mM NaCl, mM EDTA, mM EGTA, mM Na3VO4, mM NaF). After washing, proteins were boiled with SDS-PAGE sample buffer at °C for minutes for elution and detected by Western blotting as described above.GTPase activation assayThe quantification of cellular activated small GTPases was performed by precipitation with a fusion protein consisting of GST and the Rho binding domain of Rhotekin (GST-RBD) or the Rac binding domain of PAK1 (GST-PBD). Briefly, HEK293 cells were lysed in ice-cold cell lysis buffer (% NP-, mM Tris-HCl at pH ., mM NaCl, mM MgCl2, % glycerol and tab of Complete Protease Inhibitor Cocktail Tablets) (Roche) and cleared by centrifugation at g at °C for minutes. Cleared lysates were incubated with GST-RBD or GST-PBD bound beads ( µg) at °C for hour. After washing three times, GTP bound small GTPases were captured onto beads and total small GTPases in cell lysates were detected by Western blotting using monoclonal anti-RhoA, Rac1 or Cdc42 antibodies. GTP bound small GTPase amounts were normalized to the total amount of small GTPases in cell lysates by densitometry analysis.Dual-luciferase assayLuciferase assays were performed hours after transfection of different vectors using Dual-Luciferase Reporter Assay System (Promega, Madison, WI) according to the manufacturer's instructions. Renilla luciferase activity was used as an internal control. HEK293 cells were transfected with . µg of pNF-κB-luciferase Firefly (Clontech), or . µg pGL3b-IL8 reporter construct containing bp of human IL- promoter, and . µg of pRL- vector (Promega) and . µg of different vectors with FuGene6 transfection reagent or Lipofectamine as described above. For control experiments, empty vectors of indicated expression vectors were utilized. All experiments were carried out at least times.Small interfering RNA (siRNA) mediated inhibition of gene expressionBoth NOD1 siRNA, GEF-H1 siRNA and RIP2 siRNA were purchased from Santa Cruz Biotechnology. HEK293 cells were plated hours before transfection in well plates and then transfected with . µg of NF-κB reporter construct and . µg of pRL- vector and, depending on the experiments, also with . µg of GFP Shigella effector expression vectors (IpgB1, IpgB2, OspB, or control vector) in the absence or presence of NOD1 siRNA or GEF-H1 siRNA ( nM). Cells were incubated for hours and stimulated with γTriDAP or control (. µg/mL) for hours.Real Time PCRTotal RNA from MDCK cells was isolated using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. The cDNA templates were synthesized using the iScriptTM cDNA synthesis kit and quantitative PCR reactions were performed with iQTM SYBR green PCR supermix (both from Bio-Rad, Hercules, CA). Quantitative RT-PCR analysis was performed using the Bio-Rad iQ5 System and gene expression levels for each individual sample were normalized to GAPDH. The primers used were the following: canine GEF-H1 sense ′- GACTTTGCAGCCGACTCATGG-′, antisense ′- TCCTGGCGCTCCTCGTCGG-′ and canine GAPDH sense ′- TGTCCCCACCCCCAATGT-′, antisense ′- ACCCGGTTGCTGTAGCCA. S. flexneri growth condition and stimulation of epithelial cell monolayersA wild type strain of S. flexneri serotype 2a (2457T) was grown at °C in trypticase soy broth (TSB). One hundred microliters of a stationary phase culture liquid after overnight culture was used to inoculate ml of TSB and bacteria were grown in a shaking incubator for hours at °C to mid-exponential phase as described previously []. Monolayers on the permeable filter supports were serum starved overnight and gently washed with HBSS+ three times. Bacteria were administered to the monolayers from the apical side in a MOI of . Cells were incubated at °C for the time periods indicated subsequent to minute centrifugation at rpm at room temperature. After extensive washing with HBSS+, cells were fixed for immunolabelling or lysed for protein separation or NF-κB activation assays. The same protocol was used for HEK293 cells plated on -well plates. Generation of GFP-Shigella was based on the strategy described previously [].Gentamicin protection assayGentamicin protection assay was performed as described previously. Briefly, bacterial samples at a MOI of were administered to monolayers and incubated for minutes at °C. After washing three times, cells were treated with µg/ml gentamicin for minutes at °C. Intracellular bacteria were released by lysis with % Triton X- solution after washing times. Cell lysates were sequentially diluted and plated on LB agar plates for counting of colony forming units. The same protocol was used for HEK293 cells plated on -well plates. All experiments were carried out at least times and representative data are included.Statistical analysisAll results are expressed as the mean±SD. Statistical analyses were performed using a statistical software package, Statview . (Abacus Concepts, Berkeley, CA). Differences among samples were assessed using a Student's t-test, and p values of less than . were considered statistically significant.
|
PMC2579917.txt
|
TITLE: Exposure to ambient concentrations of particulate air pollution does not influence vascular function or inflammatory pathways in young healthy individuals AUTHORS: Elvira V Bräuner, Peter Møller, Lars Barregard, Lars O Dragsted, Marianne Glasius, Peter Wåhlin, Peter Vinzents, Ole Raaschou-Nielsen, Steffen Loft ABSTRACT: BackgroundParticulate air pollution is associated with increased risk of cardiovascular events although the involved mechanisms are poorly understood. The objective of the present study was to investigate the effects of controlled exposure to ambient air fine and ultrafine particles on microvascular function and biomarkers related to inflammation, haemostasis and lipid and protein oxidation.MethodsTwenty-nine subjects participated in a randomized, two-factor crossover study with or without biking exercise for minutes and with hour exposure to particle rich (number concentrations, NC: ± per cm3, mass concentrations: . ± . μg/m3 and . ± . μg/m3 for PM10-. and PM2., respectively) or particle filtered (NC: ± per cm3) air collected above a busy street. Microvascular function was assessed non-invasively by measuring digital peripheral artery tone following arm ischemia. Biomarkers included haemoglobin, red blood cells, platelet count, coagulation factors, C-reactive protein, fibrinogen, interleukin-, tumour necrosis factor α, lag time to copper-induced oxidation of plasma lipids and protein oxidation measured as -aminoadipic semialdehyde in plasma.ResultsNo statistically significant differences were observed on microvascular function or the biomarkers after exposure to particle rich or particle filtered air.ConclusionThis study indicates that exposure to air pollution particles at outdoor concentrations is not associated with detectable systemic inflammation, lipid or protein oxidation, altered haemostasis or microvascular function in young healthy participants. BODY: BackgroundEpidemiological studies have consistently identified particulate matter (PM) in ambient air as an important risk factor for morbidity and mortality related to cardiovascular diseases []. The biological mechanisms of action of PM are thought to involve altered cardiac autonomic function, endothelial dysfunction, inflammation, oxidative stress, and altered blood hypercoaguability with small particles being more potent than larger particles due to their higher surface area and reactivity [-]. The ultrafine particle fraction of PM with a diameter of less than nm and the ability to translocate through the epithelium of terminal bronchioles and alveoli is thought to be important in relation to health effects, although the extent of translocation has been debated [,]. Traffic-related PM may be particularly relevant as indicated by source allocation in acute mortality studies [], risk of cardiovascular events shortly after exposure in traffic [] and mortality associated with a residential address close to major roads with dense traffic []. Abnormal endothelial function is a strong predictor of adverse cardiovascular outcomes [] and widely recognized in patients with atherosclerosis and its risk factors []. Inhalation of high concentrations of diesel exhaust particles has been shown to impair two important and complementary aspects of vascular functions in healthy humans: the regulation of vascular tone and endogenous fibrinolysis [,] and we recently showed that filtration of indoor air PM in the residence significantly improved microvascular function (MVF) in an aged population []. Similarly, decreased flow or nitroglycerine mediated vasodilation has been associated with high ambient PM levels [,]. In other studies, ambient or personal exposure to PM was associated with increased levels C-reactive protein (CRP), amyloid A and blood coagulation [,], plasma viscosity [], fibrinogen [,], increased counts of neutrophils and platelets [,], expression of adhesion molecules on leukocyte or in plasma [,,] and the oxidation of proteins and lipids in plasma or excreted in urine [,]. Animal experiments also indicate that pulmonary exposure to PM causes microvascular dysfunction and inflammatory responses and that ultrafine particles elicit particularly strong responses [-]. However, most of the mechanistic evidence comes from experimental human or animal studies with high levels of exposure or from panel studies with associated difficulties in exposure assessment and incomplete control of confounders that hamper the establishment of causality.The primary aim of this study was to use carefully controlled exposure to real-life ambient air particles compared with particle filtered air to address the effects of particle exposure on MVF in young and healthy non-smokers. This exposure was previously shown to cause oxidative stress-induced damage to DNA in peripheral blood mononuclear cells in the same volunteers []. Exercise was included in the study in order to mimic real life exposure because it increases the dose by increased ventilation rate []. Changes in hyperaemia induced peripheral artery tonometry (RH-PAT) were used to assess MVF. Biomarkers included haematological parameters, markers of inflammation responses (serum interleukin- (IL-), tumour necrosis factor α (TNF-α), fibrinogen and CRP) and haemostasis (platelet counts and coagulation factors), as well as resistance to lipid oxidation and protein oxidation in plasma in order to elucidate potential mechanisms of action. These markers are related to the pathogenesis of atherosclerosis and associations with air pollution have been found in some studies [,,,,,].MethodsStudy design and populationStudy design and recruitment methods have previously been described in detail []. Briefly, the project was designed as a single blind two-factor crossover study with randomised exposure to fine and ultrafine particles and/or cycling scenarios. Participants were recruited using a notice in both a local newspaper and at the local campus, University of Copenhagen. To avoid problems due to diurnal variation, participants entered the exposure chamber at the same time of the morning on each -h visit at either . or . am. We simulated two exposure scenarios by pumping either non-filtered or particle filtered air into the chambers from one of Copenhagen's busiest roads which has a traffic density of vehicles per -h. Distribution and number concentration of fine and ultrafine particles (– nm) was monitored using a custom built differential-mobility particle sizer (DMPS) placed in the chamber as well as portable condensation particle counters (TSI ; TSI, St. Paul, MN, USA), which the subjects also carried when leaving the chamber for e.g. lung function test, whereas concentrations of O3, NO, NO2 and CO were continuously measured using monitors from Advanced Pollution Instrumentation, San Diego, CA, USA. Filtration order was randomised and all measurements were performed by subjects unaware of use of high efficiency particle filter (HEPA) filter or not in the period. Outdoor air was pumped directly into the chamber using a KVR- ventilator ( m3/h, P = Pa) giving a continuous air exchange. For the particle filtered air a Camfil FARR HEPA filter (226002A1; Sweden) was inserted in-line downstream of the ventilator. Each exposure period included minutes of exercise on an ergometer bicycle after exposure times of minutes and / hours. The median interval between individual exposures was days. Each volunteer was allowed to visit the bathroom, kitchen and for measurements of lung function reported elsewhere [] and the median period outside the chamber was minutes during -h. Blood was sampled after and hours of exposureThe participants, men, women, aged – yrs (median yrs), had normal lung function (baseline FEV1: . ± . L), no personal history of cardiovascular disease, a mean body mass index (BMI) of (SD: .) and were taking no medications. Each participant was his/her own control, excluding confounding by factors that are stable within an individual over time but vary between participants.The study was approved by the local ethics committee (KF ) and in accordance with the Declaration of Helsinki and all participants gave written informed consent before inclusion.Microvascular function (MVF)MVF was measured immediately before blood sampling and non-invasively using PAT during reactive hyperaemia, as previously described in detail [-]. This method assesses microvascular vasodilatation evoked by the endothelial production of NO. NO production is induced by hyperaemia associated shear stress and altered hydrostatic pressure in digital arteries following the release of the cuff.The technique uses finger-mountable pneumatic sensors (Endo-PAT , Itamar Medical Ltd, Cesaria, Israel) specifically designed to continuously digitally record the arterial pulse wave amplitude. Probe pressure is generated by a computer-controlled mini-compressor, which also contains the necessary pressure transducers, signal filters and amplifiers, data storage, signal processing means, and a screen to display the signals.Participants were instructed to fast and refrain from beverages containing alcohol and caffeine for a minimum of hours prior to testing and all tests were performed in a quiet laboratory environment. A blood pressure cuff was placed above the elbow of the right arm for hyperaemia testing while the left arm served as a control. PAT finger probes were placed on the index fingers of both hands. The test consisted of three stages: baseline recording (min. minutes), brachial arterial occlusion induced by inflating the cuff to mm Hg above systolic pressure (exactly minutes), and a post-occlusion recording of the induced reactive hyperaemia response (min. minutes). Data was digitally stored as pulse wave tracings from both hands. A MVF score describing the extent of response to hyperaemia was computed using an automated algorithm. This algorithm used the average amplitude of the PAT signal during the -minute period beginning seconds after the cuff deflation divided by the average amplitude of the PAT signal during a -minute period prior to the cuff inflation to describe the extent of reactive hyperaemia. To eliminate potential confounding of systemic effects of unilateral arm ischemia, this ratio was normalised to the concurrent signal from the control arm. The resulting value was further corrected for baseline signal amplitude. Determination of the reproducibility of this method has been described elsewhere []. Heart rate was monitored during exercise and blood pressure was measured directly before each MVF measurement.BiomarkersRed blood cells (RBC), haemoglobin, platelet counts, coagulation factors (II+VII+X), CRP and fibrinogen were measured at the Department of Clinical Biochemistry, Copenhagen University Hospital as previously detailed []. IL- and TNF-α were measured with commercially available ELISA kits (R&D Systems, Abingdon, UK) at the Sahlgrenska University Hospital, Gothenburg, Sweden.We assessed protein oxidation by the concentration of -aminoapidic semialdehyde in plasma proteins (PLAAS), as described previously []. Susceptibility to oxidation of lipoproteins in plasma was determined by ex vivo oxidation with ,'-azobis(-amidinopropane) dihydrochloride (AAPH) using a fluorescent probe as previously described [].StatisticsAll statistical analyses were performed using SAS software (version ., SAS Inst. Inc., Cary, NC). Due to four missing points for RH-PAT and missing blood samples we used mixed effect model repeated measures analysis to investigate the effect of exposure to particle rich or particle filtered air on the outcome variables: MVF quantified as MVF score, haematological and inflammation biomarkers (haemoglobin, RBC, platelets, fibrinogen, coagulation factors, CRP, IL- and TNF-α), as well as and oxidation of plasma protein (PLAAS) and resistance of lipids to oxidation (AAPH). Participant nested in gender was included as a random factor variable to account for inter-individual variation (random intercept). Exposure in terms of presence or absence of particle filter, lengths of exposure ( and hours) as well as exercise were included as fixed categorical explanatory variables. Age was included as a continuous linear variable because the quadratic term of age had no association with endpoints. We estimated effects of exposure adjusted for possible confounding by inclusion of O3, CO and NOx as continuous variables. The distributions of MVF scores and the biomarkers were skewed and all statistical analyses were performed on the natural logarithm of these data. Correlation coefficients between all dependent variables were assessed using Spearman correlation. Significance in the differences between number concentrations of fine and ultrafine PM, and gaseous parameters (O3, NO, NOx and CO) according to the two exposure scenarios were determined by the paired t-test. The significance threshold was P < . in all analyses.ResultsExposure characterisationThe chemical composition and mass of PM in the chamber air, as well as number and size distribution of PM and the concentration of gases during the two different exposure scenarios and during the corresponding period at nearby monitoring stations at a busy street and in urban background, have recently been reported[] Exposure chamber PM mass concentrations without filtration was around μg/m3 and μg/m3 for PM10-. and PM2., respectively, whilst the -h total number concentration was per cm3 and per cm3 for particle filtered and non-filtered air, respectively (Table ). The particle concentrations measured by the handheld TSI3007 counter were slightly higher than the DMPS based measurements without filtration whereas the difference was larger during filtration. It is our experience that especially at low particle concentrations the DMPS shows lower levels than the handheld counters, which also recorded exposure during the lung function tests outside the chambers. The filter effectively removed particles from chamber air (P < ., t-test). PM mass concentrations were too low to measure during filtration. NOx and NO concentrations were unaffected by filtration, whereas the O3 concentration was significantly (P < ., t-test) reduced (possibly due to reaction with the filter material) and the CO concentration significantly increased (P = ., t-test). During the non-filtered air chamber scenario the levels of PM and gases resembled the composition of a mix of urban background air with penetration and mixing with busy street air. Particles with a median diameter of nm were the most abundant and also represented the major part of the surface area in both indoor and outdoor (background and urban) air and the PM2. fraction was rich in sulphur consistent with substantial contributions from long-range transport [].Table -hour averages of total number concentrations (NC) by DPMS in the chambers and handheld CPC3007 particle counters and mass of particles in the exposure chamber and at outdoor monitoring stations.Exposure chamberOutdoor monitoring stationsNo filtrationFiltrationUrban backgroundBusy urban streetDMPS NC (#/cm3) ± ± ± ± 15100CPC3007 NC (#/cm3) ± ± --PM10 (μg/m3). ± .-. ± .. ± .7PM10-. (μg/m3). ± .-. ± .. ± .9PM2. (μg/m3). ± .-. ± .. ± .0Values are mean ± SD.Biomarkers and function testsMVF score, haematological measurements, oxidation related to proteins and lipids and cytokines are presented in table according to activity and exposure duration.Table 2Geometric means (% CI) of all dependent variables according to exposure, time and activityTime of exposureAllCyclingRestNon-filteredFilteredNon-filteredFilteredNon-filteredFilteredMVF scorea6 h2.(., .).(., .).(., .).(., .).(., .).(., .) h2.(., .).(., .).(., .).(., .).(., .).(., .)Haemoglobin (mmol/L) h8.(., .).(., .).(., .).(., .).(., .).(., .) h8.(., .).(., .).(., .).(., .).(., .).(., .)Red blood cells (×/L) h4.(., .).(., .).(., .).(., .).(., .).(., .) h4.(., .).(., .).(., .).(., .).(., .).(., .)Fibrinogen (μmol/L) h9.(., .).(., .).(., .).(., .).(., .).(., .) h9.(., .).(., .).(., .).(., .).(., .).(., .)Platelets (×/L) h248(, )(, )(, )(, )(, )(, ) h240(, )(, )(, )(, )(, )(, )CF (II+VII+X)b6 h1.(., .).(., .).(., .).(., .).(., .).(., .) h1.(., .).(., .).(., .).(., .).(., .).(., .)C-reactive protein (mg/L) h1.(., .).(., .).(., .).(., .).(., .).(., .) h1.(., .).(., .).(., .).(., .).(., .).(., .)Interleukin- (ng/L) h0.(., .).(., .).(., .).(., .).(., .).(., .) h0.(., .).(., .).(., .).(., .).(., .).(., .)TNF-α (ng/L)c6 h1.(., .).(., .).(., .).(., .).(., .).(., .) h1.(., .).(., .).(., .).(., .).(., .).(., .)PLAAS (pmol/mg)d6 h41.(., .).(., .).(., .).(., .).(., .).(., .) h41.(., .).(., .).(., .).(., .).(., .).(., .)AAPH (lagtime min)e6 h238(, )(, )(, )(, )(, )(, ) h236(, )(, )(, )(, )(, )(, )aMicrovascular function: a detailed description of the score and its relation to endothelial function is provided in the methods sectionbCoagulation factorcTumour necrosis factor αd2-aminoapidic semialdehyde in plasma proteins (pmol/mg protein)eSusceptibility to lipoprotein oxidation in plasma was measured as lagtime (min) of ex vivo oxidation with AAPH (,'-azobis( aminopropane)dihydrochloride)Microvascular function (MVF)Four pulse wave tracings were not recorded due to instrument failure. There was no significant association between MVF score and exercise or exposure to particle rich or particle filtered air (Table and ). The predictive effect of exposure as a categorical variable on MVF ranged from a .% decrease to a .% increase in MVF score (Table ).Table 3The predictive value of the estimates (% change) with % confidence intervals relative to exposure to non-filtered air in mixed effects models.Outcome variables% change (CI)Microvascular function scoreb1. (-., .)Haemoglobin (mmol/L)-. (-., .)Red blood cells (×/L)-. (-., .)Fibrinogen (μmol/L)-. (-., .)Platelets (×/L)-. (-., .)Coagulation factor (II+VII+X)-. (-., .)C-reactive protein (mg/L)-. (-., .)Interleukin- (ng/L)-. (-., .)Tumour necrosis factor-α (ng/L)-. (-.,.)PLAAS (pmol/mg protein)c-. (-., .)AAPH(lagtime min)d1. (-., .)aMixed model regression with subject nested in gender used as random factor and the natural logarithm of the biomarker in question included as a continuous outcome variable. All models adjusted for age, BMI, activity and time of sampling. Non-filtered air was included in the models as a categorical (yes/no) fixed effects predictor variable. Mutual adjustment with gases did not have significant effect on the significance of the main exposure variable (not shown)bA detailed description of the microvascular function score and its relation to endothelial function is provided in the methods sectionc2-aminoapidic semialdehyde in plasma proteinsdSusceptibility to lipoprotein oxidation in plasma was measured as lagtime (min) of ex vivo oxidation with AAPH (,'-azobis(-aminopropane)dihydrochloride)BiomarkersThere was no significant difference between the exposure to particle rich or particle filtered air for any of the measured biomarkers (Table and ). Haemoglobin and RBC were significantly decreased after exercise in these models.The inclusion of gases (O3, NOx and CO) had no significant effect on these findings.Correlation between biomarkers and function testsMVF score was not correlated with any of the biomarkers included in this study (Table ).Table 4Correlation coefficients for all day correlations among biomarkers and microvascular function scoreaHaemoglobinRBCFibrinogenPlateletsII+VII+XCRPIL-6TNFPLAAScAAPHdBMIeMVF scoreb-.-.-..-.-..-.-...015Haemoglobin0.-.-..-.-..-...317Red blood cells (RBC)-.-.-.-.-..-...190Fibrinogen0.-.....-..286Platelets-..-.-..-..209Coagulation factor II+VII+X-.-.-.-..-.117C-reactive protein (CRP)....-.057Interleukin (IL-)...-.022Tumour necrosis factor (TNF)-αg0.-.-.183PLAASc-..072AAPHd0.033aCorrelations were made on raw data. Numbers in bold depict statistically significant values (p < .)bMVF, microvascular function score was positively correlated with platelets (not significant, NS), IL6 (NS), BMI(NS) and AAPH (borderline significant, P = .) and negatively correlated with PLAAS (significant) and the remaining biomarkers (NS)c2-aminoapidic semialdehyde in plasma proteinsdSusceptibility to lipoprotein oxidation in plasma was measured as lagtime (min) of ex vivo oxidation with AAPH (,'-azobis(-aminopropane)dihydrochloride)eBody mass indexDiscussionWe studied the effects of fine and ultrafine particles from ambient air on MVF and biomarkers related to inflammation, haemostasis and lipid and protein oxidation in healthy young adults in order to give insights in the mechanisms of cardiovascular disease related to air pollution. We used a robust and powerful study design with controlled exposure comparing particle levels corresponding to urban air with virtually no exposure as well as exercise to enhance effective exposure was used. Despite our previous reported finding that the exposure was sufficient to induce oxidative damage to DNA in peripheral blood mononuclear cells within the same participants [], no significant effect on MVF or the present biomarkers was found.Endothelial function constitutes an independent predictor of cardiovascular events [,] and the clinical implications of the association between endothelial cell dysfunction and cardiac events are well established [,]. A recent exposure study of young healthy volunteers demonstrated that inhalation of air heavily polluted with emission from a diesel engine and a PM concentration of μg/m3 impaired vasomotor responses to both endothelium-dependent (acetylcholine and bradykinin) and -independent (sodium nitroprusside) vasodilators []. Moreover, this research group also showed that endothelium-dependent vasodilatation occurring in the presence of mild systemic inflammation was persistent hours after the same exposure in volunteers []. Similarly, a reduction in flow-mediated vasodilation was associated with high ambient levels of PM levels in subjects aged – years and sitting for two hours at bus stops []. We have recently shown improved MVF upon removal of particles from indoor air in the homes of healthy elderly subject []. These results provide a mechanistic and plausible link between traffic air pollution and acute myocardial infarction. In the present study, we used ambient air PM levels of μg/m3 on average and the lack of response in MVF and biomarkers may be due to this relatively low concentration, but may also indicate that the blood vessels of young subjects are less sensitive to detrimental effects of air pollution exposure at realistic concentrations. This is also in accordance with the finding that flow- and nitroglycerine-mediated vasodilation was only negatively associated with urban background levels of PM2. and sulphate and/or black carbon among diabetics of a wide age range, indicating that susceptibility including diabetes and/or age is required to detect impairment of vascular reactivity at ambient levels []. Moreover, a recent animal experiment found decreased endothelial function assessed as decreased acetylcholine-induced and unchanged sodium nitroprusside-induced vasodilation in aortic rings after systemic administration of diesel exhaust particles to hyperlipidemic apoE knockout mice, whereas wild-type mice showed an enhanced response []. It should also be considered that in patients with prior myocardial infarction and overt endothelial dysfunction no further deterioration was detected despite enhanced exercise induced coronary ischemia during high level exposure to diesel exhaust [].We used RH-PAT to assess MVF because this functional measure reflects coronary endothelial function. The method can identify individuals with coronary endothelial dysfunction and MVF score correlates well with flow-mediated dilation and reflects the vascular function of both conduit arteries and the microvasculature [,,,]. The role of NO production has been shown by the blunted response after administration of an NO-synthase inhibitor []. This method has also been used in assessment of endothelial dysfunction during pre-eclampsia [] and obstructive sleep apnoea [] as well as improvement after administration of cocoa flavonols [], external counter-pulsation in patients with coronary artery disorders [] and filtration of indoor air PM in aged volunteers []. In the latter study, the geometric mean (% CI) of MVF score was . (., .) among elderly healthy volunteers and lower than in the present study, reflecting the general consensus that aging decreases MVF []. In a study of patients referred for coronary angiography those showing endothelial dysfunction had an average MVF score of ., whilst patients without coronary endothelial dysfunction had an average MVF score of . []. Endothelium-independent vasodilatation assessed by the PAT signal to nitroglycerine was similar in the two groups in that study []. These data support the application of MVF measured by RH-PAT as a convenient non-invasive measure of coronary endothelial function. The assessment of endothelium independent vasodilatation by measuring the PAT response to administration of nitroglycerine is not likely to have contributed with further elucidation of the mechanism of PM-induced vascular effects in the present study, because there was no change in the response to flow-mediated vasodilation.We attempted to address oxidative stress in the plasma compartment as a potential mechanism of action, and found that the particle exposure in the present study did not alter the lagtime of lipoprotein oxidation or specific protein carbonyls (PLAAS) at lysine residues in plasma. This is interesting considering that the same subjects had increased levels of oxidatively damaged DNA in peripheral blood mononuclear cells []. Recently, we also found increased levels of oxidative stress-induced DNA damage the morning following exposure to traffic during biking in streets as compared to biking in a laboratory environment, further supporting that ambient levels of air pollutants are sufficiently high in Copenhagen to induce systemic effects []. In an earlier study we found that personal -h exposure to PM2. was associated with oxidative damage to both DNA and protein measured as PLAAS like in the present study [,], whereas a similar association with lipid oxidation measured as plasma malondialdehyde cannot be compared with the present lack of effect on copper-induced lipid peroxidation. In the present study we also included exposure with similar levels of exercise to enhance exposure due to ventilation, which had a non-significant enhancing effect on the DNA damage []. In the results reported here, exercise significantly and independently decreased levels of haemoglobin and RBC, but had no effect on the other biomarkers.We did not find any changes in haematological or blood coagulation parameters or markers of systemic inflammation related to the particle exposure. Exposure to diesel exhaust at μg/m3 has been associated with diminished fibrinolytic capacity, whereas the plasma concentrations of von Willebrand factor activity, prothrombin fragments, CRP and fibrinogen were unaltered and finally the effects on plasma concentrations of IL- and TNF-α vary from being unaltered to increased. [,]. In another study including volunteers only fibrinogen was affected by exposure to concentrated ambient air particles at mean concentrations of μg/m3 []. A recent wood-smoke exposure study included healthy subjects exposed to particles at μg/m3. This wood-smoke exposure was significantly associated with the concentrations of serum amyloid A, a cardiovascular risk factor, as well as factor VIII in plasma and the factor VIII/von Willebrand factor ratio, whereas IL-, TNF-α, CRP levels and fibrinogen showed no increase []. In some panel studies, CRP levels have been found to be associated with ambient or personal PM exposure [,]. Accordingly, acute phase reactants such as CRP, fibrinogen and amyloid A in plasma may respond at relatively high levels of particle exposure whereas effects on cytokine levels in plasma seem unclear. The recent finding of association between expression of adhesion molecules on leukocytes or in plasma and ambient levels of PM in observational panel studies suggest that these are promising biomarkers for experimental exposure studies [,,].It will remain challenging to compare all aspects of our results to other studies with experimental exposure of humans to particles because of the different composition including diesel exhaust [,], wood-smoke [], and concentrated ambient PM2. [] as well as large differences in exposure concentrations. In addition, there are significant variations in particle constituents between cities []. We used real-life urban background and traffic-generated particles, which is in contrast to the use of specific emission sources which make up a variable and sometimes small component of most urban air profiles. On the other hand this also caused day-day variation in the actual exposure in chambers which may have limited the possibility for detecting effects on the endpoints. The average -h exposure without air filtration in the chambers was around particles per cm3, which is somewhat higher than the particles per cm3 we have found as average -h exposure by means of the same handheld particle counters (TSI3007) in subjects of similar age living freely in Copenhagen []. Moreover, in that study we also found that more oxidative damage to DNA was induced by traffic generated particles than by the same number of particles encountered at home. In the present study the exposure chambers contained mainly traffic generated particles. Nevertheless, it cannot be excluded that the exposure occurring the days before each chamber scenario influenced our findings and it could also be argued that the main change in exposure was the effect of filtration for h, which was not associated with improvement in the measured endpoints. We included exercise as a factor in order to elucidate potential interactions. Thus, our scenarios are representative of outdoor compositions making it easier to determine their contribution to cardiovascular effects and render our results more comparable with epidemiological studies which are based on exposure assessment for ambient air pollution as well as applicable for risk assessment. In earlier studies, which utilised high concentrations of exposure, extrapolation to lower concentrations is difficult, therefore these results provide important complementary data which may prove useful in risk assessment.ConclusionThe present study of healthy, young non-smokers found no change in MVF or biomarkers of inflammation, haemostasis and oxidative stress to protein and lipids in plasma comparing controlled exposure to particle filtered air and realistic particle concentrations corresponding to ambient urban air at a level inducing oxidative damage to DNA in peripheral blood mononuclear cells. Although populations at higher risk, such as elderly or diabetics, are likely to respond differently, the present results lend no support to the notion of altered vascular function and inflammation in the cardiovascular disease pathway related to ambient air PM in a young healthy population.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsEVB participated in the design of the study, recruited and managed the study subjects, measured the microvascular function, performed the statistical analysis and drafted the first manuscript. PM participated in the design of the study and the data analysis. LB carried out IL- and TNF-α measurement. LOD carried out analysis of lipid and protein oxidation. MG and PW carried out the characterisation of the air pollutants in the exposure chambers. PV developed the exposure chambers and participated in subject management. ORN participated in the design of the study and the data analysis. SL conceived of the study, and participated in its design, coordination and data analysis. All authors read and approved the final manuscript.
|
PMC1867859.txt
|
TITLE: The Arabidopsis thaliana Brassinosteroid Receptor (AtBRI1) Contains a Domain that Functions as a Guanylyl Cyclase In Vitro AUTHORS: Lusisizwe Kwezi, Stuart Meier, Lyndon Mungur, Oziniel Ruzvidzo, Helen Irving, Chris Gehring ABSTRACT: BackgroundGuanylyl cyclases (GCs) catalyze the formation of the second messenger guanosine ′,′-cyclic monophosphate (cGMP) from guanosine ′-triphosphate (GTP). Cyclic GMP has been implicated in an increasing number of plant processes, including responses to abiotic stresses such as dehydration and salt, as well as hormones.Principle FindingsHere we used a rational search strategy based on conserved and functionally assigned residues in the catalytic centre of annotated GCs to identify candidate GCs in Arabidopsis thaliana and show that one of the candidates is the brassinosteroid receptor AtBR1, a leucine rich repeat receptor like kinase. We have cloned and expressed a amino acid recombinant protein (AtBR1-GC) that harbours the putative catalytic domain, and demonstrate that this molecule can convert GTP to cGMP in vitro.ConclusionsOur results suggest that AtBR1 may belong to a novel class of GCs that contains both a cytosolic kinase and GC domain, and thus have a domain organisation that is not dissimilar to that of atrial natriuretic peptide receptors, NPR1 and NPR2. The findings also suggest that cGMP may have a role as a second messenger in brassinosteroid signalling. In addition, it is conceivable that other proteins containing the extended GC search motif may also have catalytic activity, thus implying that a significant number of GCs, both in plants and animals, remain to be discovered, and this is in keeping with the fact that the single cellular green alga Chlamydomonas reinhardtii contains over annotated putative CGs. BODY: IntroductionGuanylyl cyclases (GCs) have been identified in many diverse prokaryotes and eukaryotes where they catalyse the formation of the second messenger guanosine ′,′-cyclic monophosphate (cGMP) from guanosine ′-triphosphate (GTP) []. In higher plants cGMP has been shown to act as a second messenger in a large number of physiological responses [], [] that include cGMP-mediated changes of the transcriptome [], NO-dependent signaling [] as well as gravitropic responses [] and plant hormone-dependent responses []–[]. Furthermore, significant and transient increases in intracellular cGMP levels have also been reported in response to plant natriuretic peptides (PNPs) [], [] as well as NaCl and drought stress []. The first functional GC in higher plants was identified with a search motif based on several functionally assigned amino acids in the catalytic domain of known GCs from lower eukaryotes and animals [].Here we show that a rationally designed search motif of the catalytic domain identifies several members of the family of leucine rich repeat receptor-like kinases (LRR RLKs) including an Arabidopsis thaliana brassinosteroid receptor (AtBRI1). A recombinant domain protein was made which tested positive for GC activity in vitro. The implications of this finding for both the projected number of different classes of GCs and the role of cGMP in brassinosteroid signaling are discussed.ResultsExtending the search GC search motif and identification of AtBRI1The original GC search motif [RKS][YFW][GCTH][VIL][FV]X[DNA]X[VIL]X{}[KR] [] (Figure 1A) yielded seven Arabidopsis candidate proteins including AtGC1 that has been demonstrated to have GC activity in vitro []. Two of seven proteins are annotated kinases and one of the two (At1g79680) belongs to the group of wall associated kinase-like proteins (WAKLs). In a quest to identify further candidate GCs in plants we mutated the position in the original search motif. This position is assigned as having a role in dimerisation that may not be a critical requirement for GC functionality. When [D] in position is substituted by [L] in a amino acid recombinant AtGC1(–), a domain that is encoded by the first exon of AtGC1 (Figure 1B), no significant loss in catalytic GC activity occurs (Fig. 1D). This implies firstly that a amino acid domain is sufficient for activity and secondly, that [D] in position is not essential for AtGC1 GC activity. These observation may hold for a number of GCs. Consequently we added [L] to make position [DNAL] (Fig. 2A). This extended motif retrieves Arabidopsis thaliana proteins including the brassinosteroid receptor AtBRI1 (Arabidopsis thaliana brassinosteroid insensitive (NP_195650., At4g39400.). Furthermore, since the catalytic asparagine [N] in position , while quite conserved in some classes of nucleotide cyclases is substituted by [AILSTYHDE] in many annotated GCs [], we further modified the motif to [RKS][YFW][GCTH][VIL][FV]X{}[VIL]X{}[KR]. All three motifs are specific for GC rather than adenylyl cyclases (ACs) since they contain the residues [GCTH] in position facing the purine and determining substrate specificity for GTP rather than ATP []–[] and GenBank contains no annotated ACs that conform to either of the motifs../journal.pone..g001Figure 1Site-directed mutagenesis and functional testing of AtGC1(–).(A) Original amino acid search motif for GCs []; substitutions are in square brackets, X represents any amino acid and curly brackets define the number of amino acids. (B) AtGC1 (At5g05930): The position of the GC catalytic centre is marked in red; the underlined aspartic acid [D] is the amino acid that has been changed into a leucine [L] by site directed mutagenesis. The open arrow marks the conserved PPi-binding arginine [R] and the C-terminal putative metal binding site is highlighted in aquamarine. The green triangles point to exon borders and the solid arrow shows the border of the fragment that we have tested for GC activity. (C) SDS-PAGE of the purification steps of the recombinant protein AtGC1(–); “M” is the molecular weight marker, “FT” is the protein in the flow-through, “W” is the wash and “E” is the eluted recombinant protein. (D) In vitro GC activity assay. The control (cont.; empty bar) was obtained by omitting protein in the reaction mix and the concentration of GTP was 1mM and that of Mn2+ or Mg2+ was mM. The values for the wild-type protein (N-terminal fragment of amino acids containing a [D] in position of the catalytic centre) and the mutated protein (N-terminal fragment of amino acids containing [L] in position of the catalytic centre) are represented with solid bars. The bar values represent the mean cGMP (+/−SEM) generated in minutes in three samples and the response pattern is representative of independent experiments.The third and most relaxed motif appears in Arabidopsis proteins of which contain [EQCTDRVHY] at position or which is presumed responsible for metal ion binding []. Metal binding is normally associated with aspartic or glutamic acid ([DE]). Within the group of proteins; contain a PPi-binding [R] between and amino acids N-terminal of the core motif.These Arabidopsis proteins (Table ) represent a highly significant (p<1e−) enrichment for the gene ontology GO categories of phosphorus metabolic processes, protein metabolic process, cellular macromolecular metabolic process and biopolymer metabolic process. It is noteworthy that several are annotated as leucine rich repeat receptor like kinases (LRR RLKs) including AtBRI1 (Fig. ). In AtBRI1, the putative catalytic core was identified within the cytosolic kinase domain../journal.pone..g002Figure 2Structural features of the GC catalytic domain and the Arabidopsis thaliana brassinosteroid receptor AtBRI1.(A) The amino acid long original search motif (modified after [] with an inclusion of [L] in position ). Red amino acids are functionally assigned residues of the catalytic centre. The residue in position does the hydrogen bonding with the guanine, the amino acid in position confers substrate specificity and the residues in positions and stabilise the transition (GTP/cGMP). (B) Representation of the domain organisation of AtBRI1 containing a signal peptide (SP), leucine rich repeats (LLRs) including an island, a transmembrane domain (TM), a GC centre embedded in the kinase domain. The position ( or ) outside the catalytic centre is implicated in Mg2+/Mn2+-binding (aquamarine). (C) Amino acid sequence of the intracellular C-terminal region of AtBRI1. The kinase domain is underlined (yellow), the GC domain is boxed in green, putative Mg2+/Mn2+-binding sites are highlighted (aquamarine), the proposed PPi binding is underlined in black, and the recombinant protein (AtBRI1-GC) tested for GC activity in vitro is delineated by solid triangles (σ). (D) Alignment of AtBRI1-like sequences. AtBRI1 (At4g39400), LeBRI1 (tomato|TC185049, Q9LJF3), OsBRI1 (Os06g0691800), VvBRI1 (grape |TC70352, Q9ZWC8), PlBRI1 (poplar|TC57820, Q9ZWC8), PsBR (BAC99050), OsBR (Os08g25380), OsSR160 (BAD34326., AP006156.)../journal.pone..t001Table 1Unique Arabidopsis proteins retrieved with the search pattern: [R]X{,}[RKS][YFW][GCTH][VIL][FV]X{}[VIL]X{}[KR]X{,}[D]I.D.AnnotationAt1g09050Function unknownAt1g14370Protein kinase APK2aAt1g17750Leucine-rich repeat transmembrane protein kinaseAt1g28440Leucine-rich repeat transmembrane protein kinaseAt1g48220Serine/threonine protein kinase, similar to Pto kinase interactorAt1g69270Leucine-rich repeat family protein, protein kinase familyAt1g69910Protein kinase family proteinAt1g73080Nucleotide binding leucine-rich repeat RK, immune responseAt1g76370Protein kinaseAt2g01860Pentatricopeptide (PPR) repeat-containing proteinAt2g02800Protein kinase APK2bAt2g26330Leucine-rich repeat protein kinase, putative ERECTAAt2g32800Kinase family protein with dual protein kinase domainsAt3g02130Leucine-rich repeat transmembrane protein kinaseAt3g02810Protein kinase family proteinAt3g07070Protein kinase family proteinAt3g24790Protein kinase family proteinAt3g46340Leucine-rich repeat protein kinase, similar to light repressible receptor PKAt3g46350Leucine-rich repeat protein kinaseAt3g46400Leucine-rich repeat protein kinase, similar to light repressible receptor PKAt4g20270Leucine-rich repeat transmembrane PK, CLAVATA1 receptor kinaseAt4g39400BRI1 (ATBRI1-GC)At5g05930AtGC1At5g07180Leucine-rich repeat protein kinase, putative ERECTAAt5g10530Lectin protein kinase, similar to receptor lectin kinase 3At5g16500Protein kinase family proteinAt5g62230Leucine-rich repeat protein kinase, putative ERECTAPK, protein kinase; RK, receptor kinase. Note: The listed proteins constitute a significant overrepresentation (p<1e−) in the Fatigo+ (level ) categories of phosphorus metabolic processes, protein metabolic process, cellular macromolecular metabolic process and biopolymer metabolic process.The choice of for further testing AtBRI1 was informed by several factors including the fact that brassinosteroids are physiologically well characterised growth regulators that await further elucidation of their signal transduction networks as well as the availability of a number of AtBRI1 mutants that can support these investigations. Brassinosteroid receptors have been identified in several other species and these also contain the conserved GC motif (Fig. 2D).The recombinant protein that we decided to synthesise and test for in vitro activity contains the predicted GC catalytic centre of AtBRI1 (At4g39400) and additional amino acids on both the N- and the C-terminus (Fig. 2C). This peptide (AtBRI1-GC) is part of the cytoplasmic domain containing the N-terminal part aspartic acid ([D] at − from the catalytic centre) implicated in metal binding [] as well as a metal binding [D] in position relative to the C-terminus of the motif.Testing a recombinant GC domain (AtBRI1-GC) for activityThe capacity of the recombinant putative GC domain of AtBRI1 (AtBRI1-GC) to generate cGMP from GTP was assessed with two independent methods. Firstly, we used an enzyme immunoassay to check if a reconstituted recombinant AtBRI1-GC could function as a GC in vitro. The results indicate that the recombinant protein can cyclase GTP and does so preferably in the presence of Mg2+ (Fig. 3A). In order to verify the result obtained with this anti-body based detection method we also used mass spectrometry. Firstly, we established that the Q-TOF mass chromatogram could detect cGMP at fmol concentrations (Fig. 3D, right inset) much like the enzyme immunoassay. We detected neither cGMP in the solution containing the recombinant protein only (Fig. 3B) nor in the reaction mix in the absence of the protein (Fig. 3C). Our sample generates cGMP in a time dependent way (Fig. 3D). We calculated that after min. incubation in the presence of mM GTP fmoles cGMP/µg protein were generated and after min. > pmoles cGMP/µg protein (Fig. 3D). We noted that values of the amount of cGMP generated obtained with the mass spectroscopy read higher than those obtained with the enzymatic assay and this observation has been made consistently in independent in vitro experiments with recombinant proteins (data not shown). In addition, it is noteworthy that plant GC activities are reportedly low and certainly not at the levels observed for some soluble animal GCs []. The main reason for this is that higher activities might require co-factors (e.g. Ca2+, chaperones or co-proteins) or post-translational modifications that are not present in the recombinant tested in vitro../journal.pone..g003Figure 3Preparation of recombinant AtBRI1-GC and functional testing in vitro. (A) Testing of GC activity with an enzyme immunoassay. The control contains the reaction mixture without the substrate ( µg recombinant protein in mM Tris-HCl (pH .), mM isobutyl methylxanthine (IBMX), mM Mg2+ and mM Mn2+), the other columns represent cGMP generated in the presence of mM GTP and either Mn2+ or Mg2+ after min. The bar values represent the mean (+/−SEM). (B) Extracted mass chromatogram of m/z [M-]− ion of cGMP generated by µg recombinant protein. The inset shows an SDS-PAGE of AtBRI1-GC expressed in E. coli BL21 (pLysS) (DE3) and purified with Ni-NTA agarose under denaturing conditions. Cleared lysate (lane ), flow through (lane ), first wash (lane ), second wash (lane ) and eluted recombinant protein (lane ). ‘M’ is the molecular weight marker. (C) Extracted mass chromatogram of m/z [M-]− ion of the reaction mix without AtBRI1-GC. (D) Two superimposed extracted mass chromatogram of m/z [M-]− ion of cGMP generated by µg recombinant protein after and min. respectively in the presence of mM Mg2+. (Note that the sample was diluted times as compared to the experiment presented in Fig. 2A). The left inset represents the mass of the peak in the chromatogram, the right inset is the calibration curve with ., and fmoles on the column.We also used mass spectrometry to test AtBRI1-GC for adenylyl cyclase activity in the presence of mM ATP as the substrate and could not detect significant amounts of cAMP generated after min. of incubation (result not shown) thus indicating that, at least in vitro, the recombinant protein has the predicted substrate preference for GTP.DiscussionBrassinosteroids (BRs) are polyhydroxylated plant steroid hormones with an essential role in co-regulating many processes including embryogenesis, cell elongation and vascular differentiation [], []. Brassinosteroid Insensitive- (BRI1) was first identified from mutant analysis and then cloned and found to be a leucine rich repeat receptor like kinase [] located in the plasma membrane [], []. Based on the binding of the ligand BR to the leucine rich repeat extracellular domain, BRI1 has been identified as a BR receptor in Arabidopsis [], [] and therefore a critical signal component. BRI1 is ubiquitously expressed in Arabidopsis and potential BRI1 kinase substrates have been identified such as transthyretin-like protein which is phosphorylated in vitro by the kinase domain of BRI1 []. Several models have been developed to describe the signaling events following perception of BR by BRI1 (see []) involving other membrane associated proteins and activation of transcription factors. The observation that AtBRI1 does harbor a functional GC domain within the cytosolic part of the molecule might suggest that cGMP is a second messenger in some BR dependent processes. However, this hypothesis remains to be tested. Several genes that regulate physiological functions are stimulated by BR as well as being influenced by cGMP. An example for this dual dependence is plant cell elongation []. Microarray studies revealed that genes involved in cell wall expansion such as expansins and pectinesterases are up-regulated by both BR [] and membrane permeable cGMP treatments [].Both BR and gibberellin interact to regulate plant growth. Some of these interactions are antagonistic but in other cases, BR can potentiate gibberellin activity []. Gibberellin, itself, stimulates increases in cGMP []. It is conceivable that in some instances the GC domain of BRI1 could stimulate cGMP production and so potentiate gibberellin activity. On a speculative note, there may be key molecules within specific cells that specify decreased cytoplasmic kinase activity and enhance the GC activity of the AtBRI1.There are several recessive alleles of AtBRI1 in the cytoplasmic kinase domain. Of these mutants, bri- is the only mutant in the GC catalytic region ([E] to [L]) and it is insensitive to BR and also has reduced kinase activity when tested in a heterologous system [], []. Interestingly, this mutation should not affect the GC activity as it occurs at position which can be any amino acid. Three other mutants have been found in the region that we show confers GC activity in vitro, being: bri-, [A] to [T], bri1-- [Q] to stop (which would exclude the GC catalytic domain from the truncated protein) and bri1- [G] to [N] []. The domain that we have identified (Fig. ) occurs within the kinase domain []. We demonstrate that the isolated amino acid recombinant peptide (AtBRI1-GC) has GC activity in vitro (Fig. ). The relative importance of the two functions in the action of the receptor remains to be demonstrated bearing in mind that previously work has focused on the kinase domain as the GC domain had not been identified. Interestingly, a number of enzymes have recently been identified as “moonlighting” proteins with dual functions []; the kinase and GC activity of AtBRI1 could be yet another example.On a more general level, the finding implies that functional GC domains may be part of a large variety of different multifunctional signaling molecules and receptors in particular. It is noteworthy that the atrial natriuretic peptide receptors NPR1 and NPR2 both signal through cGMP and have an AtBRI1-like domain organisation with an extracellular ligand-binding domain, a transmembrane domain and an intracellular kinase and GC domain [], [].Finally, the fact that two recombinant proteins (AtGC1(–) and AtBRI1 - GC ) of less than amino acids have GC activity in vitro begs a reexamination of the minimal catalytic requirement for GCs and may suggest that the number of different GC domains is significantly higher than currently assumed. This is in keeping with the fact that the single cellular green alga Chlamydomonas reinhardtii contains a surprisingly large number (>) of annotated putative GCs [] and with the increasing number of biological processes discovered that are modulated by the second messenger cGMP [], [].Materials and MethodsIdentification of GC catalytic domainAnnotated GCs were retrieved from NCBI and their catalytic domains [] were used for Blast [] queries of “The Arabidopsis Information Resource” (TAIR) database and GenBank. The catalytic domains were aligned using Clustal X [] and the alignments at the catalytic centre of the catalytic domain was used to derive the search motifs []. Derived search motifs were tested for accurate and specific detection of nucleotide cyclases by querying the Protein Information Resource (www-nbrf.georgetown.edu) using the Pattern Match option on the PIR-NREF link. Search motifs were also used to query the Arabidopsis genome via the Arabidopsis server (www.arabidopsis.org) using the “Patmatch” function.Site-directed mutagenesisA non-methylated double strand was synthesized using . µM Forward (′ -ATACTGCCTATTCGATCTTCCCTTGGTGAGTGATG- ′) and . µM Reverse (′ – CATCACTCACCAAGGGAAGATCGAATAGGCAGTAT- ′) primers from a clone containing the AtGC1 gene (At5g05930). The site where the mutagenesis occurred is underlined. Phusion High-Fidelity DNA Polymerase (New England Biolabs) was used in accordance with the manufactures instructions to amplify the plasmid. The original methylated template plasmid was digested using DpnI (New England Biolabs) leaving the amplified plasmid which was transformed into E. coli Topo 10F competent cells (Invitrogen). Single colonies were selected and the clones were analyzed by DNA sequencing.Synthesis of recombinant proteinSince no introns are in the putative GC domain of AtBRI1, genomic DNA from Arabidopsis thaliana (Col.) was used as the PCR template. PCR amplifications of AtBRI1 GC domain (At4g39400.: –) (Forward primer with BamH11 site: ′ GCTAGGATCCTGGAAGCTCGGGTTT ′, reverse primer with EcoR1 site: ′ TCCAGAATTCTCAAGCAACTTTTAAATGT ′) were performed on a Mastercycler personal (Eppendorf, Hamburg Germany), in five µL reaction volumes. Each reaction contained . mM MgCl2, µM dNTPs, . µM reverse and forward primer, . ng of genomic DNA, and . units of Taq polymerase (Fermentas GmbH, St. Leon-Rot, Germany). The thermal cycling parameters were: initial denaturation at °C for min., followed by sec. at °C, °C for sec. and °C for min. for cycles, followed by a final extension at °C for min. A fragment of bp was excised from the gel and purified using the GFX purification kit as per manufacturer's instruction (Amersham Biosciences, Little Chalfont, UK). The fragment was cloned into the pCR®T7 TOPO®-NT vector (Invitrogen Ltd., Paisley, UK) and used to transform E. coli BL21 (DE3) pLysS cells (Invitrogen Ltd., Paisley, UK); colonies were selected and inserts verified by sequencing.cGMP measurementsGC activity in vitro was assessed by measuring cGMP generated from GTP in the presence of µg purified protein, mM Tris-HCl (pH .), mM isobutyl methylxanthine (IBMX), mM Mg2+ and/or mM Mn2+ and mM GTP []. Product levels were measured by cGMP enzyme immunoassay Biotrak (EIA) System (Amersham Biosciences, Little Chalfont, UK) with the acetylation protocol as described in the supplier's manual. The anti-cGMP antibody is highly specific for cGMP and has approximately times lower affinity for cAMP.Mass spectroscopic determinations of cGMP were done with a Waters API Q-TOF Ultima in the W-mode. The samples were introduced with a Waters Acquity UPLC (Waters Microsep, Johannesburg, South Africa) at a flow rate of µL/min. and separation was achieved by a Phenomenex Synergi (Torrance, CA) µm Fusion -RP (×. mm) column. A gradient of solvent “A” (.% formic acid) and solvent “B” (% acetonitrile) over min was applied. During the first min. the solvent composition was kept at % “A” followed by a linear gradient of min. to % “B” and re-equilibration to the initial conditions. Electrospray ionisation in the negative mode was used at a cone voltage of V. The running parameters were optimised for sensitivity and specificity.
|
PMC2678267.txt
|
TITLE: Deaminase-Independent Inhibition of Parvoviruses by the APOBEC3A Cytidine Deaminase AUTHORS: Iñigo Narvaiza, Daniel C. Linfesty, Benjamin N. Greener, Yoshiyuki Hakata, David J. Pintel, Eric Logue, Nathaniel R. Landau, Matthew D. Weitzman ABSTRACT: The APOBEC3 proteins form a multigene family of cytidine deaminases with inhibitory activity against viruses and retrotransposons. In contrast to APOBEC3G (A3G), APOBEC3A (A3A) has no effect on lentiviruses but dramatically inhibits replication of the parvovirus adeno-associated virus (AAV). To study the contribution of deaminase activity to the antiviral activity of A3A, we performed a comprehensive mutational analysis of A3A. By mutation of non-conserved residues, we found that regions outside of the catalytic active site contribute to both deaminase and antiviral activities. Using A3A point mutants and A3A/A3G chimeras, we show that deaminase activity is not required for inhibition of recombinant AAV production. We also found that deaminase-deficient A3A mutants block replication of both wild-type AAV and the autonomous parvovirus minute virus of mice (MVM). In addition, we identify specific residues of A3A that confer activity against AAV when substituted into A3G. In summary, our results demonstrate that deaminase activity is not necessary for the antiviral activity of A3A against parvoviruses. BODY: IntroductionEukaryotes have evolved numerous innate immune defenses against invading pathogens. The apolipoprotein B mRNA-editing catalytic polypeptide-like (APOBEC3) proteins comprise a family of seven cytidine deaminases []–[] that may each form distinct intrinsic barriers to endogenous retrotransposons and invading viruses []–[]. The most characterized member of the family is APOBEC3G (A3G), which restricts Vif-deficient human immunodeficiency virus (HIV-) []–[]. In addition to HIV-, the APOBEC3 proteins inhibit a diverse array of viruses including simian immunodeficiency virus (SIV), human T cell leukemia virus (HTLV1), murine leukemia virus (MLV), mouse mammary tumor virus (MMTV), and hepatitis B virus (HBV) [],[],[]–[]. Interestingly, human APOBEC3A (A3A) lacks activity against retroviruses but dramatically inhibits replication of the parvovirus adeno-associated virus (AAV) []. The molecular mechanisms that govern specificity of APOBEC3 antiviral activity are not yet fully understood.The APOBEC3 proteins share structural and functional features with zinc-dependent deaminases and possess cytidine deaminase activity [],[]. The cytidine deaminase domains (CDDs) of APOBEC3 proteins contain an active site with the conserved consensus motif H-X-E-X23–-P-C-X2–-C (where X is any amino acid). It has been proposed that the histidine and the two cysteine residues coordinate a Zn2+ ion, while the glutamic acid residue serves an essential role in catalysis as a proton shuttle []–[]. APOBEC3 proteins contain either a single CDD (A3A, A3C and A3H) or two tandem CDDs (A3B, A3D/E, A3F and A3G). In the case of A3G, both domains contain an intact active site consensus sequence motif but only CDD2 appears to be catalytically active, while CDD1 is responsible for interaction with the HIV nucleocapsid proteins and packaging []–[]. Sequence alignment shows that A3A is highly homologous to the C-terminus of A3B and A3G proteins [].The mechanism of the antiviral activity of APOBEC3 proteins against retroviruses has been studied extensively [],[],[]. APOBEC3 proteins are incorporated into virus particles, and encapsidation is mediated via interactions with Gag, viral RNA and cellular RNAs [],[],[],[]–[]. Encapsidated A3G is delivered into target cells where it deaminates dC to dU on newly synthesized minus strand cDNAs during the process of reverse transcription []–[],[]–[]. Deaminated genomes can be degraded by the action of the cellular base excision repair machinery [], although recent reports suggest that uracil-DNA glycosylase (UNG2) is not required for the A3G antiviral function [],[]. In addition, hypermutated proviruses will contain sequence changes that inactivate the virus by generating alternate splicing, premature translation, and nonfunctional proteins [].Whether cytidine deamination is the principal mechanism for the antiviral activity of APOBEC3 proteins remains controversial []. APOBEC3 chimeric proteins and catalytically inactive mutants demonstrate that inhibition of HIV- can be achieved by A3G that lacks deaminase activity []–[]. A3G can inhibit lentivirus reverse transcriptase (RT) and prevent accumulation of reverse transcripts and viral cDNA in target cells in a deamination-independent manner [],[]–[]. Additionally, subsequent steps of viral integration have been shown to be affected by APOBEC3 proteins [],[]. The idea of deaminase-independent antiviral activity has been challenged by others who have argued that cytosine deamination is required for efficient inhibition of retroelements at low levels of APOBEC3 expression [],[],[]. In the case of HBV, the mechanism for inhibition by APOBEC3 proteins is also controversial but has been suggested to be deaminase-independent due to infrequent editing, and may be caused by blocking reverse transcription and expression of HBV antigens [],[],[].Although it lacks activity against retroviruses, A3A is a potent inhibitor of both parvovirus and the human transposon LINE- [],[],[]. Parvoviruses are small eukaryotic viruses that infect humans and a variety of other animal species []. The parvovirus genome consists of a linear single-stranded DNA (ssDNA) molecule approximately . kb in length, with hairpin structures at both ends that function as origins for viral DNA replication. The genome contains two major ORFs that encode the nonstructural replication proteins (NS or Rep) and the structural capsid proteins (Cap). The family is divided into autonomous parvoviruses and dependoviruses, which require a helper virus for efficient replication and progeny production. The minute virus of mice (MVM) is an autonomous parvovirus while adeno-associated virus (AAV) is a dependovirus that uses adenovirus as a helper virus. In our previous studies we found that despite the absence of detectable AAV editing, mutations in conserved active site residues of A3A abrogated the antiviral activity against AAV [].A3A inhibition of parvovirus provides an attractive system in which to decipher the relative contribution of deamination and deaminase-independent mechanisms to antiviral activity. Unlike the other viruses known to be inhibited by APOBEC3 proteins, the parvoviruses replicate exclusively in the nucleus, do not pass through RNA intermediates, do not have a reverse transcription step in their replication schemes, and use DNA hairpins for priming replication []. Moreover, A3A is a simplified system for probing APOBEC3 functions because it has a single CDD and is not restricted to a specific subcellular compartment [],[].To ascertain whether deaminase activity is required for inhibition of parvovirus replication, and to understand the functional significance of amino acid divergence between A3A and A3G, we analyzed the properties of a panel of A3A point mutants and A3A/A3G chimeras. The proteins were tested for deaminase activity in vitro and for antiviral activity in rAAV production assays. We identify mutants that lack deaminase activity but retain the antiviral effect, supporting the idea of a deamination-independent mechanism. In addition to AAV, we show that A3A inhibits DNA replication of the autonomous parvovirus MVM. Chimeric A3A/A3G proteins generated by exchanging divergent sequences demonstrate loss-of-function for A3A and gain-of-function for A3G. Our mutants also reveal residues in the linker and pseudoactive site domains that are important for deamination, target specificity and antiviral activities of A3A. Together, these studies reveal domains of A3A that are responsible for its distinct antiviral activity and suggest that A3A can inhibit parvoviruses through a mechanism separate from its function as a cytidine deaminase.ResultsInhibition of rAAV Production Is Not Dependent Upon the Deaminase Activity of A3AWe previously showed that A3A antiviral activity was dependent on the integrity of conserved amino acids in the active site that are responsible for proton shuttling and zinc coordination []. These results suggested that the catalytic domain of A3A must be intact for antiviral activity against AAV. To extend these studies, we generated additional point mutations at conserved residues (F75 and F95) previously shown to be required for deaminase activity of APOBEC1 [],[]. We also mutated the 99SPC101 residues to AAA (SPC) in the active site domain of A3A []. The position of these mutants is indicated in Figure 1A../journal.ppat..g001Figure 1Deamination is not required for antiviral activity of A3A.(A) Schematic of A3A and active site mutants. Domains marked are the cytidine deaminase domain (CDD), the linker (LINK), the pseudoactive site (PAS) and the hemagglutinin epitope tag (H). Active site residues conserved among APOBEC3 proteins (H-X-E-X28-PC-X4-C) are indicated in bold. Asterisks indicate specific mutations generated for this study; F75 is indicated with a red asterisk. (B) Immunofluorescence to detect HA-tagged APOBEC3 proteins (red) expressed by transfection in U2OS cells. (C) In vitro assay for cytidine deaminase activity. Proteins were generated by IVT, immuno-precipitated by the HA epitope, and incubated with a radiolabeled substrate (T28TCAT29). The deaminated molecules were cleaved by treatment with UDG followed by high pH, and the products were resolved by PAGE (upper panel). Arrows indicate the substrate and deaminated product. The panel below shows an immunoblot to detect in vitro translated proteins. (D) Inhibition of AAV. Production of rAAVLuc was assessed by transfection of 293T cells with AAV plasmids in the presence of APOBEC3 expression constructs ( µg). Production of rAAV was assessed by transduction of target cells and quantitation of luciferase activity. Presented is the average of three independent experiments normalized to vector only control (mock). The panels below show immunoblots to detect HA-tagged wild type and mutant A3A proteins in transfected 293T cell lysates. Ku86 served as a loading control.First, we determined the cellular localization of the mutant proteins in transfected U2OS and HeLa cells (Figure 1B and data not shown). Wild-type A3A was located throughout the cell, while A3G was predominantly cytoplasmic (Figure 1B). Most of the A3A mutants displayed cellular localization patterns similar to wild-type protein. The double mutant FF7595LL showed a nuclear punctate pattern (also observed in some cells transfected with E72Q and SPC), which might reflect misfolded protein.We next measured the catalytic activity of the proteins using an in vitro deamination assay. We previously demonstrated deaminase activity for A3A packaged into HIV- virions []. To bypass the requirement for packaging, we directly immunoprecipitated A3A from lysates of transfected cells []. We further adapted the assay to control for different expression levels by generating A3A by coupled in vitro transcription/translation (IVT) using wheat germ extract. An advantage of this method is that it allows assessment of proteins that are unstable or difficult to express in cells, such as FF7595LL and A3A truncations (Figure S1). Cytidine deaminase activity of the immunoprecipitates was measured by incubation with a radiolabeled deoxyoligonucleotide containing a single deoxycytidine target site. Wild-type A3A protein generated by IVT had deaminase activity (Figure 1C) and demonstrated similar activity to A3A produced by transfection (Figure S1). These results showed that the IVT generated protein was catalytically active and also suggested that A3A does not require a mammalian cellular co-factor for its catalytic activity. Analysis of the active site mutants in the IVT system showed that, consistent with previous studies, mutants in conserved active site amino acids (H70R, E72Q, C106S, and SPC) had lost deaminase activity [],[]. Although the F75 and F95 residues have both been shown to be required for APOBEC1 deaminase activity, the F95L mutant of A3A retained deaminase activity in our analysis, while the F75L mutant was inactive.The antiviral activity of A3A mutants was compared to wild-type A3A and A3G by transfection of 293T cells in the recombinant AAV (rAAV) production assay (Figure 1D). The APOBEC3 expression vectors were cotransfected with plasmids required for rAAV replication and packaging. The Rep and Cap proteins were supplied in trans to allow replication of an AAV vector and packaging of the ssDNA genome into virus particles. In this study we employed rAAV expressing luciferase (rAAVLuc), which allowed for quantitative assessment of virus production by transduction of target cells. Immunoblotting confirmed that proteins of the expected size were expressed (Figure 1D). Wild-type A3A completely blocked rAAV production, as previously reported []. In contrast, neither A3G nor the active site A3A mutants H70R, E72Q, SPC and C106S inhibited rAAV production. The F95L mutant that retains deaminase activity was active against rAAV. Surprisingly, F75L also inhibited rAAV despite its lack of deaminase activity in the in vitro assay. The F75L protein is therefore a separation-of-function mutant of A3A that facilitates analysis of the relative contribution of the deaminase-dependent and -independent mechanisms to AAV inhibition. The double mutant FF7595LL did not inhibit rAAV production but was poorly expressed and showed altered cellular distribution. Together, these data demonstrate that deaminase activity is not required for the anti-AAV effect of A3A.A3A Inhibits Parvovirus Replication by a Deaminase-Independent MechanismTo exclude differences in AAV inhibition due to disparities in protein expression levels, we compared the mutants F75L and F95L with wild-type A3A over a dose-response (Figure 2A). Immunoblotting revealed that expression levels of the mutants F75L and F95L were approximately ten-fold lower than wild-type A3A when equal amounts of DNA were transfected. However, despite differences in the deamination ability of the mutants, they displayed similar antiviral activity to wild-type A3A when equivalent protein levels were compared. We tested the F75L mutant against a panel of oligonucleotides that contained a cytosine in each of the possible tri-nucleotide configurations, but found no evidence of deamination above background levels on any sequence (Figure S2A and Table S1). Therefore the lack of detectable deaminase activity for F75L in vitro was not caused by an altered target sequence preference. We also tested the deaminase activity of A3A mutants in cell lysates by adapting the quantitative fluorescence resonance energy transfer (FRET) assay recently developed for A3G []. This FRET assay measures cleavage of a target oligonucleotide dual-labeled with fluorophores (Figure S2B). We observed dose-dependent deaminase activity with increasing amounts of cell lysates from 293T cells transfected with the A3A plasmid. Background levels of activity were obtained with the defective mutants E72Q and C106S. F95L showed deaminase activity in this assay, whereas F75L was not above background (Figure S2B). This result supports the observations from the in vitro deaminase assay and the conclusion that AAV is inhibited in the absence of deamination../journal.ppat..g002Figure 2A3A mutants inhibit AAV DNA replication.(A) Titration of wild-type and mutant A3A expression vectors in rAAVLuc production assays. Production of rAAVLuc was assessed by transduction of target cells and quantitation of luciferase activity. Presented is the average of four independent experiments normalized to vector alone control (mock). The panels below show immunoblots to detect HA-tagged wild-type and mutant A3A proteins in transfected 293T cell lysates. (B) Southern blot detection of low molecular weight DNA extracted from 293T cells transfected for rAAVLuc production in the presence of mock ( µg) A3G ( µg), A3A (, ., . and . µg) and mutant A3A expression vectors ( µg). The DNA was digested with Dpn-I, separated by gel electrophoresis, and hybridized with a radiolabeled luciferase probe.In our previous studies of AAV production in the presence of wild-type A3A, we found no detectable evidence of AAV sequence changes but viral replication was inhibited []. To detect effects of A3A on the accumulation of rAAV DNA, we used Southern blotting of low molecular weight DNA extracted from the transfected cells during rAAV production (Figure 2B). As controls, we compared the A3A mutants to wild-type A3A and A3G. Replicated rAAV DNA was detected by hybridization with a luciferase probe. Although F75L was slightly less effective than F95L, both mutants inhibited the accumulation of rAAV DNA (Figure 2B), suggesting that inhibition of AAV replication is not dependent on deamination.To examine the effect of A3A and mutants on replication of wild-type parvovirus genomes, we used two different viral systems. AAV2 depends upon helper virus for replication, while MVM replicates autonomously. First we used immunofluorescence to assess the effect of APOBEC3 proteins in cells infected with AAV2 and adenovirus helper virus. As previously shown [], replication centers detected by staining for the viral Rep protein were present in cells that expressed A3G but were absent in those with A3A (Figure 3A). Advanced stage viral replication centers were detected in cells expressing inactive mutants. In contrast, A3A mutants that inhibited rAAV production (F75L and F95L) also blocked formation of viral replication centers. These data demonstrate that observations made with rAAV production also apply to inhibition of wild-type AAV replication../journal.ppat..g003Figure 3A3A inhibits AAV2 and MVM DNA replication.(A) Inhibition of wild-type AAV replication. U2OS cells were transfected with plasmids for APOBEC3 proteins and then infected with AAV and adenovirus. HA-tagged APOBEC3 (red) and AAV Rep proteins (green) were detected by immunofluorescence using specific antibodies. (B) Inhibition of wild-type MVM replication. Southern blot detection of low molecular weight DNA extracted from A9 cells cotransfected with an infectious MVM clone together with APOBEC3A expression plasmids. The DNA was digested with Dpn-I, separated by gel electrophoresis and hybridized with a radiolabeled MVM probe. Left line is a marker (M). Replicative intermediates of ssDNA (SS), monomer (M), and dimer (D) are indicated to the right. Panel below shows immunoblots for APOBEC3 and the NS1 protein of MVM.Replication of an infectious plasmid clone of MVM was also dramatically inhibited by co-transfection of wild-type A3A, but not the C106S mutant (Figure 3B, lanes –). Interestingly, replication of an MVM genome bearing a large in-frame deletion within the capsid gene, and therefore unable to generate the wild-type capsid proteins necessary to produce single-stranded progeny DNA (MVM-ΔBglII), was also inhibited by A3A (Figure 3B, lanes –). In separate experiments, the expression of the full spectrum of MVM RNA and protein generated from a non-replicating full-length MVM plasmid was not affected by expression of A3A (data not shown), suggesting that A3A directly affects parvovirus genome replication.The Linker and Pseudoactive Domains of A3A Are Required for Antiviral ActivityTo identify further residues required for A3A antiviral activity, we compared the sequence of A3A to the C-terminus of A3B and A3G. Alignment of the amino acid sequences showed that A3A is most closely related to A3B. Two main regions with variable sequences (VS1 and VS2) differ from A3G (Figure ). Using the structure of the C-terminal domain of A3G as a template [], we predicted the secondary structure of A3A (Figure ) and this suggested that the VS1 would be located in the loop between the β2 strand and the α1 helix, and that VS2 partially overlaps with the α4 helix. To define regions of A3A responsible for the antiviral activity against parvoviruses, we tested the contribution of the linker and pseudoactive site subdomains []. We first generated chimeric proteins between A3A and A3G (Figure 5A) joined at the shared PmlI site. The N-terminus of A3A (residues –) was fused to the C-terminal PmlI fragment of A3G (residues –) to form the chimera A3ApmlA3G. The reciprocal chimera A3GpmlA3A was also generated. We also assessed the activity of the C-terminus of A3G (A3G-CT, residues M197 to N384) and a fusion with A3A at the PmlI site (A3G-CTpmlA3A). Cellular localization of the mutants was tested by immunofluorescence in transfected cells (Figure S3A). We found that localization of the chimeric proteins to the cytoplasm was determined by the N-terminal domain of A3G as previously reported []. We also measured the deaminase activity of chimeric proteins immunoprecipitated from transfected cells in the in vitro assay with the T28CCCGT28 deoxyoligonucleotide substrate (Figure 5B, upper panel). Full-length A3A and A3G both produced robust deamination, but all of the chimeras were less active. The mutants were also tested for activity against rAAV in the transfection assay (Figure 5C). Neither A3G nor A3G-CT had any inhibitory effect against rAAV despite being expressed well in transfected cells. Fusion of the linker and pseudoactive site subdomains of A3A onto A3G or A3G-CT was not sufficient to confer antiviral activity (A3GpmlA3A and A3G-CTpmlA3A). The A3ApmlA3G chimera, which possesses the linker and pseudoactive site of A3G fused onto A3A, showed diminished antiviral activity. This reduction was confirmed in a dose-response titration (Figure 5D). Together these data suggest that the linker and pseudoactive site regions of A3A are important for deamination and antiviral activity, but that these domains are not sufficient to confer activity to A3G../journal.ppat..g004Figure 4Alignment of APOBEC3 amino acid sequences for A3A with the C-terminus of A3B and A3G.Residues exchanged in A3A are boxed and the mutant designation is indicated above. The PmlI site and stretches of variable sequence VS1 and VS2 switched in the chimeras are also indicated. The asterisks mark individual amino acid mutants, and the diamond indicates the start of A3G-CT (residues –). Residue numbers are indicated on the right side. Predicted secondary structure of A3A is indicated below the alignment with α-helices in black and stranded β-sheets in grey. A3A secondary structure modeling was generated by Swiss-Model using the crystal structure of the C-terminal fragment of A3G (Protein Data Bank accession number 3E1A) as template []../journal.ppat..g005Figure 5Activity of A3A/A3G PmlI based chimeras.(A) Schematic of A3A (dark grey) and A3G (light grey). Domains marked are the cytidine deaminase domains (CDD), the linker (LINK), the pseudoactive site (PAS), and the hemagglutinin epitope tag (H). Below are the chimeras generated at the PmlI site. (B) In vitro assays for cytidine deaminase activity. Proteins were immunoprecipitated from transfected cells by the HA epitope and incubated with the indicated radiolabeled substrate in UDG-dependent assays. The upper panel uses a substrate oligonucleotide with target sequence CCCG, and the middle panel uses an oligonucleotide with the specific A3A target sequence TCA. The substrate and deaminated products are indicated. The bottom panel shows an immunoblot to detect proteins in immunoprecipitates. A band corresponding to the light-chain IgG used for immunoprecipitation is indicated (*). (C) Production of rAAV in the presence of APOBEC3 proteins. 293T cells were transfected with APOBEC3 constructs ( µg, except for A3A . µg), together with plasmids required for rAAVLuc production. Virus production was assessed by transduction of target cells and quantitation by luciferase assay. Panels below show immunoblots for APOBEC3 proteins (HA) in transfected cells and Ku86 as a loading control. (D) Dose-response for A3A and the A3ApmlA3G chimera in the rAAV production assay. Panels below show immunoblots for APOBEC3 (HA) and Ku86 proteins in transfected cells.Careful inspection of the deaminase assay autoradiograms (Figure 5B, upper panel) revealed that the deaminated products for A3A and A3G have slightly different mobilities, which likely reflects differences in deamination target specificity [],[],[]–[]. The deamination product for A3GpmlA3A migrated similarly to that of A3A, suggesting it had gained the target site specificity of A3A. To test this possibility, we analyzed the deaminase activity on a deoxyoligonucleotide containing the specific A3A target sequence TCA (Table S1). While A3A was highly active on the TCA substrate, A3G was inactive (Figure 5B, middle panel). Although the level of deamination by the A3GpmlA3A chimera was less than that of A3A, this mutant was similarly active on both substrates (Figure 5B, compare upper and middle panels). These observations suggest that the region encompassing the linker and pseudoactive site is involved in target site selection []. In our assay the C-terminal fragment of A3G was inactive, although recent studies reported that a similar fragment of A3G (residues –) showed mutator activity in bacteria [],[] and deaminase activity in vitro []. However, these studies used GST-tagged A3G-CT in the bacterial assays or purified recombinant protein in their in vitro assays, while we have analyzed protein from cell lysates. Furthermore, A3G-CT was shown to be significantly less active than the full-length protein []. Our results may also reflect differences in experimental conditions or that protein produced by transfection may be less active due to RNA inhibition or lack of dimerization.Deaminase Activity Is Dispensable for Activity against rAAVBased on the analysis of chimeric proteins, we generated further mutants in which non-conserved residues of A3A were substituted with those from A3G (see Figure ). Amino acids that differed between the two proteins were changed throughout A3A (Figure 6A). Most of the mutants consisted of A3A residues replaced with the analogous A3G sequence. Two residues lacking in A3G were deleted from A3A (ΔWG), and in another mutant unique residues from A3G were inserted into A3A (EPWVR). The two main variable regions that contain stretches of divergence (VS1 and VS2 in Figure ) were switched in two stages that changed or residues at a time (Figure 6A). All mutants displayed the same pattern of cellular localization as wild-type A3A (Figure S3B). Mutant proteins were synthesized by IVT and evaluated for deaminase activity in the in vitro assay (Figure 6B). Mutants PT, MAK, and SK retained wild-type levels of activity. The EPWVR and ΔWG mutants had diminished activity compared to wild-type A3A. The chimeras in variable stretches VS1 and VS2 lacked detectable deaminase activity. When these mutants were included in the rAAV production assay, all of them retained the ability to inhibit AAV (Figure 6C). The VS1 mutants with diminished antiviral activity (GFLE and PHKHGFLE) were also compared to wild-type A3A over a dose response (Figure S4). When similar levels of protein were compared for their effect on rAAV production, the VS1 mutants GFLE and PHKHGFLE showed less activity than wild-type A3A (∼% inhibition compared to ∼%) (Figure S4). This observation suggests that the VS1 region in A3A contributes to the antiviral activity. Together these data demonstrate that residues outside of the putative enzymatic active site of A3A, in both the N-terminus and C-terminus, are required for efficient deamination but that this does not correlate with antiviral activity against parvovirus replication../journal.ppat..g006Figure 6Mutants of A3A with residues replaced with the corresponding sequences of A3G.(A) Schematic of variable segments VS1 and VS2, and the chimeric mutants generated for A3A in these regions. The VS1 segment corresponds to A3A residues to , and VS2 corresponds to A3A residues to . (B) In vitro deamination assay. Proteins were immunoprecipitated from transfected cells and incubated with radiolabeled oligonucleotide (T28CCCGT28) for h in UDG-dependent assays. Arrows indicate the substrate and deaminated product. The panel below shows an immunoblot to detect proteins in immunoprecipitates. (C) Production of rAAV in the presence of wild-type (. µg) and mutant A3A proteins ( µg). Panels below show immunoblots for APOBEC3 proteins (HA) in transfected cells and Ku86 as a loading control.Activity against AAV Can Be Conferred to A3GAnalysis of the A3A mutants suggested that the two stretches of residues divergent between A3A and A3G (VS1 and VS2) are important for deamination, and may play a role in antiviral function. Therefore, we determined whether the reciprocal switch (where residues in A3G were substituted with the sequences from A3A) would generate a gain-of-function (Figure 7A). Sequences from A3A were incorporated into constructs that express the C-terminal fragment of A3G which can localize in the nucleus (Figure 7B). The mutant proteins were tested for their effect on AAV in the virus production assay (Figure 7C). Proteins of the expected size were expressed at similar levels. The C-terminal fragment that contained the complete sequence from VS1 of A3A (A3G-CT/KNLLCGFY) acquired significant inhibitory activity against AAV. It was less active than wild-type A3A and reduced rAAV production by approximately % when equal levels of protein where compared (Figure S5). When incorporated into full-length A3G, the VS1 region of A3A increased deamination in vitro, whereas the VS2 sequences decreased activity (Figure S6). Incorporation of the A3A sequences into full-length A3G did not confer AAV inhibition, presumably due to cytoplasmic localization or interference by the N-terminus (Figure S6). Thus A3G-CT/KNLLCGFY provides the first gain-of-function mutant for A3G and demonstrates that the VS1 region of A3A (residues to ) contributes to the antiviral activity against parvovirus../journal.ppat..g007Figure 7Identification of a gain-of-function mutant for A3G.(A) Schematic of full-length A3G and the C-terminal fragment A3G-CT. Chimeras of A3G-CT were generated with variable segments VS1 and VS2 replaced with sequences of A3A. (B) Immunofluorescence to detect localization of HA-tagged APOBEC3 and chimeric proteins (red) expressed by transfection in U2OS cells. Cell nuclei were detected by staining with DAPI (blue). (C) Production of rAAV in the presence of APOBEC3 proteins. Virus production was assessed by transduction of target cells and quantitation by luciferase assay. Immunoblots show similar expression levels for APOBEC proteins in transfected cells. Ku86 served as a loading control. The A3G-CT/KNLLCGFY mutant demonstrated activity against AAV. The asterisks indicate that the inhibition with A3A and A3G-CT/KNLLCGFY was statistically significant (p<.) when compared to mock by Student t test.DiscussionIn this report we provide multiple pieces of evidence to show that A3A inhibition of AAV can occur through a deaminase-independent mechanism. Mutation of A3A active site residues that are essential for catalytic activity (H70R, E72Q, SPC99-101AAA and C106S) led to the loss of activity against AAV. However, other mutants (F75L and mutants in VS1 and VS2) separated the deaminase activity from the ability to inhibit AAV. Together these data indicate that the integrity of the active site is important but that deaminase activity is not required for AAV inhibition. In APOBEC1, aromatic residues analogous to F75L and F95L of A3A are required for both deaminase activity and binding to nucleic acids [],[]. In the case of A3A, we found that F95 is not required for deaminase activity. This result probably reflects the differences in structure and nucleic acid specificity between APOBEC3 proteins and APOBEC1, as revealed by recent structural studies [],[]. We found that the F75L mutant was deaminase-defective in our in vitro deaminase assay. The lack of deaminase activity on a panel of target oligonucleotides demonstrated that the F75L A3A mutant is truly deaminase-deficient, and has not simply changed its target site preference and eluded detection in the deaminase assay. Although the in vitro deaminase assay has limitations, we demonstrated that lysates containing F75L also lacked deaminase activity in the FRET assay []. Evidence that deaminase activity is not essential for inhibition of AAV by A3A is consistent with the absence of signs of deamination in AAV DNA in cells expressing A3A [].Deaminase-independent inhibition of ΔVif-HIV and retroelements by A3G has been controversial [],[]. The deaminase-defective A3G mutants in E259 retains anti-viral activity when over-expressed [],[]. However, when equivalent proteins levels are compared in transient transfections or in stable cell lines, the deaminase deficient mutant has significantly less potent antiviral activity than wild-type A3G [],[],[]. In our studies we assessed the dose response of A3A mutants by comparing their antiviral activity against AAV in titration experiments. We identified deaminase-defective mutants (F75L and mutants in VS2) that displayed similar activity against AAV as wild-type A3A when analyzed at comparable protein levels. In addition, the F75L and VS2 mutants displayed the same subcellular localization as the wild-type protein and thus their phenotype cannot be ascribed to protein mislocalization.In addition to providing evidence for deaminase-independent antiviral activity, our study also offers insights into the structural basis of APOBEC3 protein function. The linker and pseudoactive site domains in the N-terminus of A3G are required for HIV- virion incorporation [],[]. We demonstrate that these domains also influence the target site specificity of APOBEC3. A3G prefers the target site (T/C)CC [],[],[],[],[], while A3A is more flexible, showing preference for (T/C)CA []. Replacement of the linker and pseudoactive sub-domains at the end of A3G with those from A3A modified the target site preference towards the A3A-specific consensus target TCA. Thus, A3A residues in the C-terminus contribute to its target specificity, in agreement with data from chimeric and mutant proteins of other APOBEC3 family members []. Residues in the VS2 region have been implicated in target site specificity [], but in our hands the exchange of VS2 residues between A3A and A3G did not affect target specificity.It is unclear why A3A has more potent in vitro deaminase activity than other APOBEC3 proteins. Deaminase activity of A3G resides in the C-terminal CDD [],[],[],[],[], which shares % identity with A3A. An intriguing difference between A3A and A3G is the presence of two additional residues (WG) within the PCX2–4C motif of A3A. Deletion of these amino acids in the A3A/ΔWG mutant slightly reduced in vitro deaminase activity (Figure ). The variable region VS1 is situated immediately upstream of the H-X-E-X23–-P-C-X4-C conserved motif, and replacement of A3A sequences with those from A3G caused a decrease in deaminase activity. Substitution of the VS1 region in A3G with sequences from A3A increased deaminase activity compared to wild-type A3G (Figure S6). This suggests that the VS1 region (residues –) may contribute to the increased enzymatic activity of A3A [],[]. Interestingly, the VS1 region in located in the active center loop of the C-terminal domain of A3G and disruption of this loop results in greatly impaired A3G deaminase activity []. In the variable region VS2, substitution of A3A amino acids with those from A3G also decreased deamination, suggesting that this region is important for catalytic activity. This observation is supported by the decrease in deaminase activity observed for the reciprocal A3G mutants (A3G/YDP and A3G/YDPLYK) (Figure S6). Together these results indicate that regions outside of the active site contribute to catalytic activity of A3A. In support of our observations, a recent mutagenesis study of A3G also suggested that C-terminal residues (residues –) are important for deaminase activity [].Multiple mechanisms have been proposed to explain the deaminase-independent inhibition of retroelements and HBV by APOBEC3G [],[]. In the case of retroviruses, the APOBEC3 proteins have been suggested to inhibit RT, prevent accumulation of reverse transcripts and viral cDNA in target cells, and block integration [],[],[],[]. Biochemical studies have shown that purified recombinant A3G inhibited RT-catalyzed DNA elongation in vitro and this was independent of its deaminase activity []. A3G can inhibit HBV in the absence of extensive editing [], and has been suggested to be due to inhibition of early steps in viral reverse transcription and strand elongation []. AAV inhibition differs from these other systems because it does not involve an RT-mediated step or any RNA substrates. Since A3A also inhibits MVM replication and does not affect production of viral proteins [], we favor the idea that A3A inhibits parvovirus replication through a direct interaction with the viral DNA or the replication machinery. Binding to ssDNA by A3G is proposed to inhibit RT processivity [],[], and binding of A3A to ssDNA in the parvovirus genome could physically block movement of the DNA polymerase along the viral template. Although this inhibition would be independent of catalytic activity, amino acids in the active site may be required for efficient nucleic acid binding, explaining the loss of antiviral activity for mutants such as E72Q or C106S. Preliminary results suggest that the F75L mutant retains its ability to bind nucleic acid (data not shown). Potential binding sites could be the viral ITR or the ssDNA/dsDNA junction, which may reduce Rep binding and inhibit DNA synthesis. A3A is not found in high molecular weight complexes that have been reported to modulate A3G activity [],[], however we cannot exclude the possibility that AAV inhibition might be mediated through A3A interactions with AAV Rep or cellular proteins that are required for AAV replication []. It will be interesting to investigate whether recombinant A3A can block viral DNA replication in an in vitro replication assay where we could test the direct activity of A3A on AAV replication []. This approach, however, is currently limited by the requirement for purified recombinant A3A and its mutants.Although cells are a standard system used to study parvovirus replication, it remains unclear whether endogenous A3A restricts parvovirus infection in vivo. It is unknown which cells represent the primary site of AAV infection and replication in vivo. It has been shown that expression of APOBEC3 proteins is induced in response to interferon-α (IFNα) []. The levels of A3A that achieve inhibition of AAV in transfected cells are within the range of endogenous A3A levels detected in peripheral blood mononuclear cells (PBMCs) and macrophages activated with IFNα (Figure S7). Therefore, it would be interesting to test whether cells refractory to parvovirus replication will allow AAV replication when endogenous A3A levels are reduced. In addition to parvoviruses, A3A is active against LINE1 and other retrotransposons [],[],[],[],[]–[] where hypermutation is not detected. It will be informative to determine whether this occurs by a mechanism similar to parvovirus inhibition. In summary, our study demonstrates that the DNA cytidine deaminase activity of A3A is not required for inhibition of parvovirus replication. The combination of the single-domain cytidine deaminase A3A with the simple model system of parvovirus replication provides a valuable tool to uncover new mechanisms for the antiviral activity of the APOBEC3 proteins.Materials and MethodsCell Lines293T, HeLa, A9 and human osteosarcoma U2OS cell lines were purchased from the American Tissue Culture Collection. Cells were grown as monolayers in Dulbecco's modified Eagle's medium (DMEM) supplemented with % fetal bovine serum and antibiotics at °C in a humid atmosphere containing % CO2.Expression PlasmidsExpression plasmids encoding cDNAs for A3A (NM_145699) and A3G (NM_021822) and the A3A mutants H70R, E72Q, and C106S in the pcDNA3. (+) vector with a hemagglutinin (HA) tag at the C-terminus have been previously described []. New A3A and A3G mutants were generated by site-directed mutagenesis using the QuikChange kit (Stratagene) (Table S1). The truncated form of A3G (A3G-CT, residues -) was generated by PCR amplification of its C-terminus. Plasmids expressing AAV Rep/Cap proteins (pXX2) and the adenovirus helper proteins (pXX6) have been described []. The rAAV vector plasmid (pACLALuc) consists of the luciferase gene amplified from pGL3basic (Promega) cloned into an ITR-flanked expression cassette under the control of the CMV promoter and the BGH polyadenylation signal (pACLA). The complete ITR-flanked expression cassette in pACLALuc is . Kb.Production of Recombinant AAVRecombinant AAV production assays were performed as previously described [],[]. Briefly, 293T cells were seeded at .× cells/well in -well plates and the next day were co-transfected with pXX6 (. µg), pXX2 (. µg), pACLALuc (. µg) and APOBEC3 expression vector ( µg unless otherwise stated) or pcDNA3.(+) control vector ( µg). Dose-response titrations maintained the total amount of effector DNA by addition of pcDNA3.(+). Transfections were performed in duplicate or triplicate using polyethyleneimine (PEI) []. Cells were harvested h post-transfection after two washes in ice-cold PBS, and one third of each sample was removed for immuno-blotting. The other two thirds of the cells were used to generate rAAVLuc virus lysates by freeze/thaw cycles followed by centrifugation. Virus lysates were used to transduce 293T cells in wells in triplicate. Transduced cells were incubated with Steady-Glo luciferase substrate reagent (Promega) h post–transduction and lucifierase activity was quantified in triplicate in well Lumiplates (Greiner Bio-One) in a TopCount NXT scintillation and luminescence counter (PerkinElmer). rAAV production experiments are presented as mean+SEM of the relative value (%) of at least three independent experiments, and compared to mock transfections with pcDNA3.(+).ImmunoblottingImmunoblotting was performed essentially as described []. Cell pellets from rAAVLuc production assays were lysed in lysis buffer ( mM NaCl, . mM KCl, . mM Na2HPO4, . mM KH2PO4, mM NaO3V, mM β-glycerol phosphate, mM NaF, .% NP40, and .% Triton-X ) supplemented with Complete protease inhibitor cocktail (Roche) for min on ice. The lysates were clarified by centrifugation at ,×g for min. Protein concentrations from whole cell lysates were quantified by BCA assay (Bio-Rad), and µg of protein was loaded per well onto polyacrylamide gels. Proteins were separated in –% or % Acrylamide Bis-Tris NuPage gels in MOPS buffer (Invitrogen) and transferred onto Hybond nitrocellulose membranes (Amersham Biosciences). Membranes were probed with anti-HA 16b12 monoclonal antibody (mAb) (Covance) and anti-Ku86 mAb (Santa Cruz). Bound antibody was detected by incubation with goat anti-mouse antibody conjugated to horseradish peroxidase (Jackson ImmunoResearch), and the bands were visualized with enhanced chemiluminiscence reagent (ECL Western Lightning Kit, PerkinElmer) followed by autoradiography.ImmunofluorescenceAPOBEC3 protein localization was determined by indirect immunoflorescence []. U2OS or HeLa cells were grown on glass coverslips in well plates and transfected with . µg APOBEC3 expression vector using Lipofectamine (Invitrogen). After – h, cells were washed with PBS, fixed with % paraformaldehyde for min and extracted with .% Triton X- in PBS for min. Cells were incubated with % BSA for min, followed by incubation with anti-HA mAb 16b12 (∶). A ∶ dilution of goat anti-mouse conjugated Alexa Fluor (Invitrogen) and DAPI (Sigma Aldrich) in % BSA in PBS was added to cells and samples were incubated for h at room temperature. The coverslips were mounted in Fluoromount-G (Southern Biotech) and cells were visualized by fluorescence microscopy (Diaphot inverted microscope, Nikon). For AAV replication U2OS cells were seeded on glass coverslips and transfected with APOBEC3 expression vector, and infected h post-transfection with wild-type AAV and/or adenovirus. After hr, the cells were fixed and stained with anti-HA and anti-Rep antibodies and DAPI as described above. Rep from the input virus is undetectable in this assay, so positive Rep staining is indicative of AAV replication.Southern Blot HybridizationLow molecular weight AAV episomal DNA was analyzed by Southern hybridization with a 32P-labeled luciferase probe as previously described []. Briefly, 293T cells grown in well tissue culture plates to % confluency were co-transfected with plasmids for rAAV production using Lipofectamine following the manufacturer's protocol. The following effector plasmids were included: pcDNA3.(+) ( µg), A3G (1ug), A3A ( µg, . µg, . µg and . µg), F75L ( µg) and F95L ( µg). After h, the cells were collected and washed in PBS. One third of each sample was removed for immunoblotting. DNA was isolated from pellets by a modified HIRT protocol [] and digested with DpnI (New England Biolabs) to remove input plasmid. DNA was processed by gel electrophoresis on a % agarose gel in TAE buffer. The pACLALuc plasmid was digested with SmaI as control. The gel was depurinated in . M HCl, denatured in M NaCl, . M NaOH, and neutralized in . M Tris pH ., . M NaCl. DNA was then transferred to a Hybond XL membrane (Amersham Biosciences) and UV-cross linked. The membrane was hybridized with a 32P labeled luciferase probe generated by PCR using the primers described in Table S1 and labeled with 32P dCTP using Radivue II labeling kit (Amersham), and visualized in a FLA- phosphorimager (Fuji).Southern blot replication assays of wild-type MVM and MVM-ΔBglII were performed as previously described []. In vitro Deaminase Assays293T cells were seeded at × cells/well in -well plates and transfected after one day with µg of APOBEC3 expression vector. Two days post-transfection, the cells were rinsed twice with cold PBS and lysed for min on ice in lysis buffer ( mM Tris, pH ., mM KCl, mM NaCl, mM EDTA, .% Triton X-, mM DTT). Lysates were clarified by centrifugation at ,×g for min and pre-cleared with µl of High flow protein-G-Sepharose (Amersham). The lysate was incubated with anti-HA mAb 3F10 (Roche) for h at °C. The lysate-antibody was then incubated with High flow protein-G-Sepharose for - h at °C. The resin was washed three times with lysis buffer. One-fifth of the resin was removed for immunoblot analysis and the remainder was washed once with deaminase reaction buffer ( mM Tris, pH ., % glycerol, mM KCl, mM NaCl, mM EDTA, and mM DTT). PAGE purified deoxyoligonucleotide (Table S1) was ′-end 32P labeled and added into µl of deaminase reaction buffer. The reaction was incubated at °C for h, stopped by heating to °C for min, cooled on ice, and then centrifuged to collect the resin at the bottom of tube. The supernatant was incubated with uracil DNA glycosylase (New England Biolabs) in buffer containing mM Tris, pH ., mM DTT for h at °C and treated with mM NaOH for h at °C. The samples were incubated at °C for min, °C for min and separated by % TBE/urea-PAGE. The gel was dried, exposed to a phosphorimager screen and analyzed using a FLA- scanner (Fuji).For in vitro synthesis of A3A and mutants we employed the TNT Coupled Wheat Germ Extract System (Promega) using T7 polymerase. Translation reactions were performed with non-labeled amino acids following the manufacturer's protocol in µl final volume that included µg of pcDNA3.(+) plasmid encoding APOBEC3. Reactions were incubated for min at °C. After incubation, µl of TritonX- buffer ( mM Tris, pH ., mM KCl, mM NaCl, mM EDTA, .% Triton X-, mM DTT) were added to the reactions and used to analyze deaminase activity after immunoprecipitation as described above.In order to measure deaminase activity directly from IVT reactions, after min incubation at °C translation reactions were centrifuged at ,×g for min. An aliquot of supernatant ( µl) was removed for immunoblot analysis. The reaction mixture ( µl of supernatant) was incubated with ′-end 32P labeled deoxyoligonucleotide in µl of deaminase reaction buffer and assayed for deaminase activity following the same procedure described above.FRET-Based Deaminase Activity Assay from Cell LysatesTo quantify deaminase activity from cell lysates, we used a modification of the FRET-based protocol described by Thielen et al. []. 293T cells were seeded at × cells/well in cm plates and a day later transfected with µg of APOBEC3 expression vector. After to h, cells were resuspended in lysis buffer ( µl) and lysates were obtained as described above. Cell lysate ( µl) was mixed with µl of FRET deaminase buffer ( mM Tris, pH ., mM KCl, mM NaCl, mM EDTA) containing pmol of dual-labeled probe (Table S1) and . units of UDG (New England Biolabs). Reactions were incubated at °C for min followed by addition of µl of 4N NaOH and incubation at °C for min. Reactions were neutralized with µl of 4N HCl and µl of M Tris-HCl (pH ). 6FAM fluorescence was measured at °C in an Mx30005P (Stratagene). Two-fold serial dilutions of each lysate were analyzed in duplicate. Fluorescence detected in 293T cells transfected with pcDNA3. (mock) was substracted from all samples and deaminase activity is shown as relative fluorescence units (RFU).Detection of Endogenous A3AHuman monocytes were purified from leukocyte enriched blood samples (New York Blood Center) using CD14+ magnetic beads (Miltenyi Biotec) according to manufacturer instructions. The CD14+ monocytes were cultured with ng/ml GM-CSF (Invitrogen) for days in order to differentiate them into macrophages. The monocyte-derived macrophages were plated in well plates at cells per well and cultured with or without 2000U of Universal Type I Interferon (PBL Biomedical Laboratories) for hours. The cells were then collected in lysis buffer. Cell lysates from human peripheral blood mononuclear cells (PBMCs) were obtained from F. Chisari (The Scripps Research Institute) []. PBMCs were treated with 1000U of IFN-α or kept untreated for hrs before cell lysates were generated. Lysates of PBMCs ( µg) and macrophages ( µg) of each sample were run on –% or % Bis-Tris gels and analyzed by immunoblotting as described above. Anti-A3A (raised against an N-terminal peptide) (∶ dilution) or anti-recombinant A3A (∶ dilution) polyclonal rabbit sera were used for A3A detection as described.Supporting InformationFigure S1APOBEC3A proteins synthesized from coupled in-vitro transcription-translation are active in UDG-dependent deaminase assays. Deaminase activity of A3A and mutant proteins generated by cell transfection and in-vitro coupled transcription-translation (IVT) was analyzed in UDG-dependent deaminase assays. (A) 293T cells were transfected with plasmids for A3A and mutants. Cells were harvested at hrs post-transfection, and lysates were subject to immunoprecipitation (IP) with anti-HA antibody (3F10). / of the IP was incubated with a ′-end 32P labeled T28TCAT28 deoxyoligonucleotide and tested in UDG-dependent deaminase assays. Arrows indicate substrate deoxyoligonucleotide and cleaved deaminated product. Bottom panel shows an immunoblot corresponding to / of the IP analyzed with an anti-HA antibody (16B12). Asterisks indicate bands corresponding to IgG light chain. A3A truncations are: TruncA (aa –), TruncB (aa –), and TruncC (aa –). (B) Wild-type and mutant A3A proteins were synthesized by IVT as described in Methods. pcDNA3.(+) was included in IVTs as mock. A3A proteins were immunoprecipitated with 3F10 antibody and tested in the UDG-dependent deaminase assay. Immunoprecipitated A3A from transfected cells was included as a control. Middle panel shows / of the IP protein analyzed by immunoblotting with 16B12 antibody. Bottom panel shows immunoblotting of lysates to demonstrate that equal amounts of protein were generated by IVT. (C) Wild-type and mutant A3A proteins were tested directly from IVT reactions for deaminase activity. Bottom panel shows an immunoblot of / of the IVT-synthesized proteins loaded into the deaminase reactions.(. MB PDF)Click here for additional data file.Figure S2Lack of deaminase activity for F75L is not due to modified target sequence preference and is supported by the lack of deamination measured by FRET. (A) To rule out differences in target sequence preference, A3A and F75L were tested in UDG-dependent deaminase assays against different target sites (Table S1). Target sequence preference of A3A and F75L was determined on a panel of four 32P-labeled deoxyoligonucleotide substrates, each containing four target sites. The - base is shown at the top of each lane and the + base is shown on the left. The structure of the cleaved products is shown on the right. (B) Quantification of deaminase acitivity by FRET in cell lysates obtained from 293T cells transfected with A3A, E72Q, F75L, F95L and C106S. Cell lysates were incubated with a dual-labeled probe containing the CCCG target sequence (Table S1). The TAMRA fluorophore quenches emission by the 6FAM fluorophore. After deamination and treatment with UDG and high pH, 6FAM fluorescence emission can be detected from the cleaved probe. Two-fold serial dilutions of each lysate were analyzed by duplicate and represented as mean±SEM. Results show that deaminase activity is detected in cell lysates obtained from cells transfected with A3A and F95L, while deaminase activity of E72Q, F75L and C106S is not distinguishable from the background level. Bottom panel shows an immunoblot of cell lysates.(. MB PDF)Click here for additional data file.Figure S3Localization of A3A/A3G chimeras. Immunofluorescence to detect localization of HA-tagged APOBEC3 and chimeric proteins (red) expressed by transfection in U2OS cells. Cell nuclei were detected by staining with DAPI (blue). (A) Localization of wild-type APOBEC3 proteins and A3A/A3G chimeras. (B) Localization of A3A mutants with sequences incorporated from A3G.(. MB PDF)Click here for additional data file.Figure S4Inhibition of rAAV production by A3A/A3G chimeras in the VS1 region. Dose-response for A3A and mutant proteins in the rAAV production assay. Virus production was assessed by transduction of target cells and quantitation by luciferase assay. Panels below show immunoblots for APOBEC3 (HA) and Ku86 proteins in transfected cells.(. MB PDF)Click here for additional data file.Figure S5Dose-response antiviral activity of A3G-CT/KNLLCGFY. Comparison of A3A and A3G-CT/KNLLCGFY antiviral activity over a dose-response in rAAV production assays. Comparable levels of A3A and A3G-CT/KNLLCGFY resulted in ∼% and ∼% inhibition respectively. Bottom panels show immunoblots for A3A and A3G-CT/KNLLCGFY (HA), and Ku86 protein as a loading control. The asterisks indicate that the inhibition with A3G-CT/KNLLCGFY was statistically significant (p<.) when compared to mock by Student t test.(. MB PDF)Click here for additional data file.Figure S6Chimeric A3G proteins with sequences replaced with VS1 and VS2 from A3A. (A) Immunofluorescence to detect localization of HA-tagged APOBEC3 and chimeric proteins (red) expressed by transfection in U2OS cells. Cell nuclei were detected by staining with DAPI (blue). (B) In vitro deamination assay. Proteins were immunoprecipitated from transfected cells and incubated with the radiolabeled substrate in the standard assay. Arrows indicate the substrate and deaminated product. The panel below shows an immunoblot to detect proteins in immunoprecipitates. Arrows indicate bands corresponding to APOBEC3 proteins. Asterisk indicates bands corresponding to IgG light chain. (C) Production of rAAV in the presence of APOBEC3 proteins. Virus production was assessed by transduction of target cells and quantitation by luciferase assay. Immunoblots show similar expression levels for APOBEC3 proteins (HA) in transfected cells. Ku86 served as a loading control.(. MB PDF)Click here for additional data file.Figure S7Antiviral activity of A3A is achieved with physiological levels of transfected A3A. (A) Antiviral activity of A3A was analyzed in a rAAVLuc production experiment over a dose response. Expression levels of A3A were analyzed by immunoblotting using an anti-HA antibody (top blot), and compared to endogenous levels of A3A expressed in human PBMCs incubated with IFNα. A3A was detected using a rabbit polyclonal antisera raised against an N-terminal A3A specific peptide (bottom blot). Arrow indicates bands corresponding to A3A. Background band below A3A is indicated with an asterisk. (B) In panel B the expression levels of A3A are compared to endogenous levels of A3A induced in human macrophages in response to IFNα. A3A was detected with polyclonal rabbit serum generated to recombinant A3A protein.(. MB PDF)Click here for additional data file.Table S1Oligonucleotides used in this study. Sequences corresponding to the oligonucleotides used in this study. Name, template and type of experiments are indicated for each oligonucleotide. D.A. (deaminase assays). N/A (not applicable).(. MB DOC)Click here for additional data file.
|
PMC2649506.txt
|
TITLE: Serial Position Learning in Honeybees AUTHORS: Randolf Menzel ABSTRACT: Learning of stimulus sequences is considered as a characteristic feature of episodic memory since it contains not only a particular item but also the experience of preceding and following events. In sensorimotor tasks resembling navigational performance, the serial order of objects is intimately connected with spatial order. Mammals and birds develop episodic(-like) memory in serial spatio-temporal tasks, and the honeybee learns spatio-temporal order when navigating between the nest and a food source. Here I examine the structure of the bees’ memory for a combined spatio-temporal task. I ask whether discrimination and generalization are based solely on simple forms of stimulus-reward learning or whether they require sequential configurations. Animals were trained to fly either left or right in a continuous T-maze. The correct choice was signaled by the sequence of colors (blue, yellow) at four positions in the access arm. If only one of the possible signals is shown (either blue or yellow), the rank order of position salience is , and (numbered from T-junction). No learning is found if the signal appears at position . If two signals are shown, differences at positions and are learned best, those at position at a low level, and those at position not at all. If three or more signals are shown these results are corroborated. This salience rank order again appeared in transfer tests, but additional configural phenomena emerged. Most of the results can be explained with a simple model based on the assumption that the four positions are equipped with different salience scores and that these add up independently. However, deviations from the model are interpreted by assuming stimulus configuration of sequential patterns. It is concluded that, under the conditions chosen, bees rely most strongly on memories developed during simple forms of associative reward learning, but memories of configural serial patterns contribute, too. BODY: IntroductionLearning of stimulus sequences requires memory of the temporal order of occurrences. Under natural conditions temporal sequence is often combined with spatial sequence, e.g. in navigational tasks. Position in space of objects is defined both by the temporal sequence of experience for the navigating animal and by its relation to surrounding cues, giving the position of each item a unique spatial character. Memories developed for sequential spatial positions of items may, therefore, be embedded in a large-scale relational spatial memory (a mental map). The memory formed under these conditions has been recognized as episodic or episodic-like resembling key features of memories that allow humans to mentally experience a previous occasion in space and time [], []. One characteristic feature of episodic-like memory is the configuration of serial patterns into unique episodes []. It is not too far fetched to ask whether an insect like the honeybee is able to create episodic-like memory because bees are known to navigate with reference to a map-like spatial memory [], perform configural forms of compound learning (such as positive and negative patterning in olfactory conditioning, []), master serial conditional discrimination like matching-to-sample and non matching-to-sample tasks [], extract rules from multiple training sets (e.g. symmetrical vs asymmetrical patterns, [] sequences of turn is mazes with multiple choice points, []), and organize their foraging activities according to multiple circadian time windows according to occasion setting conditions (review: [], []). However, none of these experiments allowed rejecting more simple explanations, e.g the familiarity of signals, the recency of experience, differences of the strengths of memory traces and other characteristics of associative learning. In such a scenario one would expect the learning of the temporal-spatial sequences to be defined predominantly with respect to the evaluating conditions for the choice, the reward following the correct choice, and the fact that positions closer to the evaluating signal may have a greater impact on memory.The role of order in a purely temporal sequences has been studied intensively after Ebbinghaus’ground braking discovery of the primacy and recency effect [], []. Many examples are known meanwhile in which the last and the first items are better remembered than the middle items (bow-shaped memory function [], []. According to the kind of errors made in recognizing the serial order of items (e.g. words, letters, numbers), several models were developed after Lashley's [] account of creating a theoretical concept of serial order learning []. Such concepts range from assuming rather simple associative phenomena to specific coding of sequences of items, and it is generally agreed that a dominance of the recency effect does not require the assumption of a memory for the whole sequence.Here we ask whether an insect, the honeybee, learns sequences of two colors in a context that attempts to simulate a navigational task. The bees fly in a T-maze which due to its narrow channels stimulates their neural distance measuring device (odometer) so strongly that they appear to experience multiples of the actual flight length []. They learn to turn right or left at the T-intersection according to the sequential color pattern they have experience during their flight in the access arm. In such a task temporal and spatial components of serial order are tightly connected and are associated with the outcome (reward) after a decision has been made based on serial discrimination. From a learning-theoretical point of view the task examined here belongs to those in which the animal has to discriminate patterns of compound stimuli (color, position, sequence) to form conditional discriminations (e.g. serial feature positive or negative tasks, []). If bees were to solve the task by some episodic-like or configural color sequence memory we would expect rather equal salience of the four sequential color positions and unique patterning effects reflecting memory of the whole or at least part of the sequential pattern. If, however, simple associative phenomena dominate their choice behavior one would expect a deviation from the bow-shaped function of the serial stimulus salience and a salience rank order that reflects the distance from the evaluating conditions (reward). It is known that sensory memory for visual and olfactory stimuli in reward learning in bees allows for an interval between the cue and the reward of several seconds [], []. Since the four serial color signals in our experiments are experienced within a few seconds prior to the reward, both a configuration into a unique sequential pattern and a dominance of simple associative phenomena are possible. There is a rich literature on bees learning to associate particular signals to motor routines like moving right or left [], []. In some of these experiments bees were also exposed to sequential signal/turn relations, and it was found that they learn multiple associations under particular training conditions [], []. The sequences of signals tested in matching to sample (or matching to non-sample) paradigms [], [] need to be experienced within up to sec before the match. None of these experiments have yet addressed the question of how bees evaluate positions of sequences of visual signals that are experienced within short intervals as guiding signals for alternative turns.I find that bees learn to discriminate a series of two colors in four positions. Stimulus salience follows a rank order according to the distance of stimulus position from the choice point with highest salience to the closest position indicative of a recency effect. Discrimination is predicted to a large extent by positional salience, but divergence from this rule and data from the transfer tests also indicate configural phenomena.MethodsHoneybees (Apis mellifera carnica) were trained from the hive to a T-maze (at a distance of meters) during two summer periods. The T-maze was located between trees which allowed the bees a view of the canopy and the sky. The T-maze consisted of a . m long entrance tunnel (× cm), two . m long tunnels to the right and left of the T intersection, and two . m long connecting tunnels that led the bees back to the entrance area after they had been feeding (continuous T-maze, Fig. ). The bottom and the walls were covered with a random black-and-white pattern the structures of which appeared at a visual angle of approximately ° when the bee flew in the middle of the tunnel. The top of the tunnel was made of UV-transmitting plexiglass. The color signals (either blue, B or yellow, Y) were arranged inside the entrance tunnel in such a way that the bees had to fly between two identical cm color stripes (called here: signals). These color signals appeared at different positions numbered , , and and were placed at distances of cm. No. was closest to the T-intersection ( cm away from the T intersection). Bees were trained to fly in such a continuous T-maze. They entered it via the central tunnel (entrance tunnel), turned to the right or to the left at the T intersection, depending on the pattern of color signals at positions , , and , and were rewarded at one of the two feeders (F). After feeding to completion, they flew out through one of the connecting tunnels. During the training session the flight path of the bee was guided by revolving doors (a–h, Fig. )../journal.pone..g001Figure 1Bees approached the continuous T-maze from the hive at a distance of m, entered it via the entrance tunnel, and proceeded through the entrance tunnel toward the T-junction after passing the positions (–) of the color signals (either blue, B or yellow, Y).During training revolving doors (a–h) first guided the bees to enter the arm which provided food reward (F) at the end of the respective side arm, and, after sucking to completion, back out of the maze via the respective connecting tunnel. The bottom and walls of the tunnels were covered with a random black-and-white pattern, the structures of which appeared at a visual angle of approximately ° when the bee flew in the middle of the tunnel. The top of the tunnel was made of UV-transmitting plexiglass. The color signals were arranged inside the entrance tunnel in such a way that the bees had to fly between two identical cm color stripes.A group of – bees shuttled regularly between the hive and the T-maze. They recruited newcomers, which became the experimental bees of the day. Two to four experimental bees were trained and tested only on the day of training. The group of recruiting bees was caged during the training and testing of the experimental bees. Sucrose concentration during training of the experimental bees was adjusted such that no further bees were recruited. The training of the experimental bees consisted of two phases: an initial phase lasting – minutes in which the bees learned to use only the entrance tunnel to access reward and to fly fluidly through the tunnel. No tests were performed during the initial training phase. In a second phase lasting – hours bees continued learning the particular arrangement of color signals at the four positions – which they had already experienced in the first training phase. Tests began when bees reached asymptotic performance after about hours of training in the second phase. All bees mastered the task, and all bees trained were included in the tests. The flight time in the entrance tunnel was .+/−. s. Each experimental bee was trained to only one signal pattern, and all tests were performed on the day of training. In average each bee arrived times during the initial training phase and times during the second training/test phase.Two kinds of tests were performed, within-training tests and between-training tests. In the first case food was available at the correct position, but both doors at the choice point were open. If the bee chose the correct arm of the maze, it was rewarded, and the choice was recorded as correct; if it chose the wrong arm it did not receive a reward, flew out through the connecting tunnel by closing the respective doors, and the choice was recorded as wrong. When it entered the entrance tunnel again it was guided to the feeder by closing the door into the wrong tunnel at the T intersection (no choice record). Thus any one bee made only one decision during a within-training test. During the between-training tests all doors were open, no feeder was available and the two feeder areas were covered with new paper of the same black-and-white random pattern. Bees were allowed to fly through the tunnel in any direction, but they usually flew into the tunnel via the entrance tunnel and exited via the respective connecting tunnel. Decisions were counted when a bee approached the T intersection via the entrance tunnel and flew at least half the length into one of the two arms of the T-maze. Each experimental bee made – choice flights during the between-training tests which lasted for minutes. No difference was found between the results of the within-training and the between-training tests. Therefore, the data were pooled. Tests for the same signal patterns were repeated during the – hours of the test phase, and the different test patterns (including the transfer tests) followed each other in a pseudorandom fashion. Training for both patterns was continued during the test phase, and test patterns were always different from the last training pattern.The serial position tasks involved two colors (B,Y) at four positions (, , , and ). Training patterns differed with respect to numbers and positions. The bee learned two patterns in sequential approach flights, one that was associated with a right turn in the T maze, and one associated with a left turn in the T maze. The notation of the training pattern will show the respective color (B, Y) at the respective position and the training side. Ø represents a position without a color signal. For example, [Ø B B Ø]r vs. [Ø Y Y Ø]l means bees learned to fly into the right arm when a blue signal appeared at positions and and no signal appeared at positions and , and during the same training session the same bees learned to fly into the left arm when yellow signals appeared at positions and and none at positions and . Table and summarize all experimental conditions, and tables S1 and S2 in the supplementary material give the choice data for all experiments../journal.pone..t001Table 1Summary of all discrimination experiments listing the training conditions, the patterns trained (for notation see Methods), the consecutive number of the experiment, the number n of choices, and the respective figure.ColumnExperimentTraining patternsNo. of experimentTotal number of choicesFigure1Differences in number and positionX Ø Ø Ø vs. X X Ø Ø47a, Fig. “X Ø Ø Ø vs. X X X Ø48, Fig. “X Ø Ø Ø vs. X X X X47, Fig. “Ø X Ø Ø vs. X X X X2, Fig. 5One signalB Ø Ø Ø vs. Y Ø Ø Ø1555 Fig. “Ø B Ø Ø vs. Ø Y Ø Ø16, Fig. “Ø Ø B Ø vs. Ø Ø Y Ø1730 Fig. “Ø Ø Ø B vs. Ø Ø Ø Y18, Fig. Two signals9DS at different positions12 different positions25, , 27b, , , , , Fig. 4A 10SD at different positions12 different positions24, , , , 19a, , Fig. 4B 11DD at different positions12 different positions27a, , , , 19b, , Fig. 4C 12Three signalsD D D36, Fig. 5A “D D S40, , Fig. 5A “D S D36a, Fig. 5A “D S S41, Fig. 5A “S D D14, 33a119 Fig. 5B “S D S37, 42a120 Fig. 5B “S S D12, Fig. 5B The consecutive experiment number indicates the sequence of experiments, assigning a new number to experiments performed on a different day. A number combined with a letter indicates that an additional experiment was performed on the same day. The first column (column Nr.) helps to coordinate this table with the table S1 and S2 in the supplementary material giving the choice values for all tests../journal.pone..t002Table 2Summary of all transfer experiments listing he training conditions, the patterns trained (for notation see Methods), the transfer patterns, the consecutive number of the experiment, the number n of choices, and the respective figure.ColumnExperimentTraining patternsTransfer testsNo. of experimentTotal number of choicesFigure1Color transfersØ B Ø Ø vs. B B B BØ Y Ø Ø vs. Y Y Y Y232 Fig. 6A “Ø Y Ø Ø vs. Y Y Y YØ B Ø Ø vs. B B B B619 Fig. 6B “Y B Ø Ø vs. B Y Ø ØB B Ø Ø vs. Y Y Ø Ø2156 Fig. 6C “B Y B Ø vs. Y B B ØB B B Ø vs. Y Y B Ø970 Fig. 6D 5Color and position transferØ B Ø Ø vs. B B B BØ G G Ø G Ø Ø Ø638 Fig. 6Position transferB Ø Ø Ø vs. Y Ø Ø ØFour different patterns1546 Fig. 8A “Ø Y B Ø vs. Ø B Y ØFour different patterns2237 Fig. 8B “B B Ø Ø vs. Y B Ø ØFour different patterns2042 Fig. 8C “Y Y Ø Ø vs. B Y Ø ØFour different patterns2677 Fig. 8D “Ø B Y Ø vs. Ø Y Y ØFour different patterns30126 Fig. 8E “Ø B B Ø vs. Ø B Y ØFour different patterns31110 Fig. 8F “B Ø B Ø Y Ø B ØTwo different patterns3348 Fig. 8G 13Position transferB Ø Ø B vs. Y Ø ØBTwo different patterns3417 Fig. 8H “B Y B Ø vs. Y B B ØTwo different patterns934 Fig. 8I “B Y B Ø vs. Y B Y ØTwo different patterns3655 Fig. 8J 16Position and number transferØ X Ø Ø vs. X X X XSeven different patterns2, Fig. The consecutive experiment number indicates the sequence of experiments, assigning a new number to experiments performed on a different day. A number combined with a letter indicates that an additional experiment was performed on the same day. The first column (column Nr.) helps to coordinate this table with the table S1 and S2 in the supplementary material giving the choice values for all tests.Experimental design and statistics: Each test for a given pattern resulted in about to decisions made by each experimental bee, leading to up to decisions per test. Tests for the same pattern were repeated several times for the same group of experimental bees on the same day, and many experiments were repeated with different experimental bees (see tables S1 and S2 in the supplementary material). The permutations of the three variables (color: B, Y; number of signals: –; positions: –) shall be presented in a systematic fashion but were carried out during two summer periods in an unsystematic way.Statistic: To test pattern discrimination within a single experiment we used Fisher's exact tests. To analyze differences in performances across different experiments we used the G-test. We also used a paired t-test to compare the flight times for correct and incorrect choices, and Pearson correlation and linear regression to analyze the predictions of a model [].ResultsThe experimental design includes three variables, colors (B, Y), numbers (–), and positions (–). The role of these variables in guiding the bee in their choices at the T-intersection will be studied by systematically varying them independently and in combination. The measure of performance will be the probability of correct choices in discriminating the two patterns associated with the left and right turn (Chapter A: Discrimination tests, Fig. – ). In a first series of experiments (section A (), Fig. ) the color will be the same for both patterns but the numbers of signals and their positions will be different. Thus we first ask whether bees can use the numbers of signals to learn the turn in the T-maze. Then we shall vary the numbers of differently colored signals and ask whether the different positions provide the same or different salience for learning to orient in the T-maze. In addition, it will be interesting to search for any pattern effect possibly indicative for configural phenomena. I shall present the discrimination values first for one signal (section A (), Fig. ), then for two (section A (), Fig. ), and then for three signals (section A (), Fig. )../journal.pone..g002Figure 2Percentage of correct choices for tasks in which bees were trained to discriminate between patterns offering signals of the same color but differing in number and position.X indicates either a blue or a yellow signal, but it was always the same color in the two alternatives (see text). Asterisks indicate significant discrimination between the training patterns: Fisher's exact test, P XØØØ vs. XXØØ = ., P XØØØ vs. XXXØ = ., P XØØØ vs. XXXX<., P ØXØØ vs. XXXX<.. Numbers inside the bars indicate the number of choices analyzed../journal.pone..g003Figure 3Percentage of correct choices for tasks in which bees were trained to discriminate between single signals of two different colors, B or Y.Asterisks indicate significant discrimination between the training patterns: Fisher's exact test; P BØØØ vs. YØØØ<., P ØBØØ vs. ØYØØ<., P ØØBØ vs. ØØYØ = ., P ØØØB vs. ØØØY = .. Numbers inside the bars indicate the number of choices analyzed../journal.pone..g004Figure 4Percentage of correct choices for tasks in which bees were trained to discriminate between patterns offering two signals in the same position, this signal providing either the same (S) or different (D) colors.A, B and C show the six DS, SD and DD combinations respectively. Asterisks indicate significant discrimination between the training patterns: Fisher's exact test; A) P DSØØ<., P ØDSØ<., P ØØDS = ., P DØSØ<., P DØØS<., P ØDØS<.. B) P SDØØ = ., P ØSDØ = ., P ØØSD = ., P SØDØ = , P SØØD = , P ØSØD = . C) P DDØØ<., P ØDDØ = ., P ØØDD = ., P DØDØ = ., P DØØD = ., P ØDØD = .. Numbers inside the bars indicate the number of choices analyzed../journal.pone..g005Figure 5Percentage of correct choices for tasks in which bees were trained to discriminate between patterns offering three signals in the same position.These signals presented either the same (S) or different (D) colors. A Four combinations that presented a different signal in position . B Three combinations that presented a similar signal in position . Asterisks indicate significant discrimination between the training patterns: Fisher's exact test; A) P DDD<., P DDS<., P DSD<., P DSS<.. B) P SDD = ., P SDS = ., P SSD = .‥ Numbers inside the bars indicate the number of choices analyzed.Configural phenomena of combinations of color, number and position may become apparent in generalization tests which ask whether the bee transfers a learned sequential pattern more strongly to one or another test pattern that was not experienced during training. A large number of such transfer tests will be presented in Chapter B (Transfer Tests, Fig. – ). Again the trained patterns will be systematically varied according to numbers and positions of these sequential patterns../journal.pone..g006Figure 6Percentage of choices toward the left arm of the maze for color transfer tests.Bees were trained to discriminate between patterns offering signals of the same color but differing in number and position (A, B), between patterns offering a DDØØ combination (C), or between patterns offering a DDSØ combination (D), and were then tested with transfer patterns which differed from the training patterns only in color. The trained patterns are given in the lower line and the transfer patterns in the line above. In the figure the choice values for the trained patterns are indicated by horizontal lines (…‥ or ‥_‥). The choice values for the transfer patterns are shown by columns. Asterisks indicate a significant transfer (i.e., the percentage of choices for a given transfer pattern significantly differ from only one of the training patterns and is similar to the other one): Fisher's exact test; A) P ØGØØ vs ØBØØ = ., P ØGØØ vs BBBB<., P GGGG vs ØBØØ<., P GGGG vs BBBB = .; B) P ØBØØ vs ØYØØ = , P ØBØØ vs YYYY = ., P BBBB vs ØYØØ = ., P BBBB vs YYYY = ; C) P BBØØ vs YBØØ<., P BBØØ vs BYØØ = ., P YYØØ vs YBØØ = ., P YYØØ vs BYØØ<.. D) P BBBØ vs BYBØ = ., P BBBØ vs YBBØ<., P YYBØ vs YBBØ = ., P YYBØ vs BYBØ<.. Numbers inside the bars indicate the number of choices analyzed../journal.pone..g007Figure 7Percentage of choices toward the left arm of the maze for color, number and position transfer tests.The trained pattern is given in the lower line, and the transfer patterns in the line above. In the figure the choice values for the trained patterns are indicated by horizontal lines (…‥ or ‥_‥). The choice values for the transfer patterns are shown by columns. Asterisks indicate a significant transfer (i.e., the percentage of choices for a given transfer pattern differ significantly from only one of the training patterns and is similar to the other one): Fisher's exact test, P ØGGØ vs ØBØØ = ., P ØGGØ vs BBBB = ., P GØØØ vs BBBB = ., P GØØØ vs ØBØØ<.. Numbers inside the bars indicate the number of choices analyzed../journal.pone..g008Figure 8Percentage of choices toward the left arm of the maze for position transfer tests.A) Bees were trained to discriminate between patterns offering a single signal, either yellow or blue, in position and tested with a pattern offering the single signal yellow or blue in positions or . B, E, F) Bees were trained to discriminate between patterns offering a two signal yellow and blue in position , and tested with a pattern offering the same signals in positions , or , . C, D) Bees were trained to discriminate between patterns offering a two signal yellow and blue in position , and tested with a pattern offering the same signals in positions , or , . G) Bees were trained to discriminate between patterns offering a two signal yellow and blue in position , and tested with a pattern offering the same signals in positions , . H) Bees were trained to discriminate between patterns offering a two signal yellow and blue in position , and tested with a pattern offering the same signals in positions , . I, J) Bees were trained to discriminate between patterns offering a three signal yellow and blue in position , , and tested with a pattern offering the same signals in positions , , . Numbers inside the bars indicate the number of choices analyzed. The trained pattern is given in the lower line, and the transfer patterns in the line above. In the figure the choice values for the trained patterns are indicated by horizontal lines (…‥ or ‥_‥). The choice values for the transfer patterns are shown by columns. A) P ØBØØ vs BØØØ = ., P ØBØØ vs YØØØ = ., P ØYØØ vs BØØØ = ., P ØYØØ vs YØØØ = , P ØØØB vs BØØØ = ., P ØØØB vs YØØØ = ., P ØØØY vs BØØØ = ., P ØØØY vs YØØØ = .. B) P ØØYB vs ØYBØ = ., P ØØYB vs ØBYØ = ., P ØØBY vs ØYBØ = ., P ØØBY vs ØBYØ = ., P YBØØ vs ØYBØ = ., P YBØØ vs ØBYØ = ., P BYØØ vs ØBYØ = , P BYØØ vs ØYBØ = .../journal.pone..g009Figure 9Percentage of choices toward the left arm of the maze for position and number transfer tests.Bees were trained to discriminate between patterns offering signals of the same color but differing in number and tested with pattern offering one or two signals of the same color in different positions. The trained pattern is given in the lower line, and the transfer patterns in the line above. In the figure the choice values for the trained patterns are indicated by horizontal lines (…‥ or ‥_‥). The choice values for the transfer patterns are shown by columns. Asterisks indicate a significant transfer (i.e., the percentage of choices for a given transfer pattern significantly differ from only one of the training patterns and is similar to the other one): Fisher's exact test; P XØØØ vs ØXØØ<., P XØØØ vs XXXX = ., P ØØXØ vs ØXØØ = ., P ØØXØ vs XXXX = ., P ØØØX vs ØXØØ = ., P ØØØX vs XXXX<., P XXØØ vs ØXØØ<., P XXØØ vs XXXX = ., P ØØXX vs ØXØØ = ., P ØØXX vs XXXX<., P XØØX vs ØXØØ<., P XØØX vs XXXX = ., P ØXXØ vs ØXØØ = ., P ØXXØ vs XXXX<.. Numbers inside the bars indicate the number of choices analyzed.The large number of possible combinations of the three parameters color, number and position can be reduced due to the fact that the two colors (B,Y) provide the same salience, and that the two directions of turns in the T-maze (right, left) are fully symmetrical.A. Discrimination tests() Differences in numbers and positionsI first asked whether bees discriminate between two alternatives associated with right or left turns that differed in the numbers and positions of the signals, but not in their color. Fig. gives the results for permutations in which one alternative was always one signal (either B or Y indicated by X) at either position or , and the other alternative two, three or four signals. Different numbers are well discriminated even for a difference only by one signal in position , and discrimination improves if the alternatives differ in more than one signal. There is no difference between the choice values in experiments with two or three signals (each versus a single signal at position one) indicating that either position contributes less to discrimination than position and , or that the position effects are not cumulative. Taking the results of experiments with two or three signals together (each versus a single signal at position ), and comparing them with those having four signals (again versus a signal at position one), a significant difference is found (P = .) indicating that the fourth position in the context of all four positions contributes to discrimination. In experiments with the single signal at either position or the same discrimination values indicate a similar salience of positions and .() One signal: differences in position and colorIf a single signal differs in color for the two alternatives, bees learn the task particularly well when the signals are at position , equally well when at positions and , but do not discriminate between the two alternatives when the signals are at position (Fig. ). Choice values are significantly higher for color difference at position than those of the first and second bar in Fig. , indicating that positions and are most salient for discrimination. Thus large differences in numbers of signals (Fig. ) have a lower salience than color differences if they appear at position (Fig. ). Color differences appearing at position do not lead to significant discrimination (Fig. ). Thus position may not contribute to discrimination if animals are trained to use the color of a single signal.() Two signals: differences in position and colorIn the case of two signals at different positions, one can reduce the multitude of permutations to patterns in which the same position in the two alternatives presents either the same (S) or different (D) colors. Thus for positions and patterns of [B B Ø Ø] vs. [Y B Ø Ø], or [Y B Ø Ø] vs. [B B Ø Ø], or [Y Y Ø Ø] vs. [B Y Ø Ø], or [B Y Ø Ø] vs. [Y Y Ø Ø] belong to the same group of [D S Ø Ø] patterns. Since it was found that the two colors B and Y have the same salience, and no preference existed for one of the two sides, there was no need to test all possible permutations. Thus our notation for the three types of patterns DS, SD, DD at permutations of positions out of four is: DS at positions and , and , and , and , and , and (Fig. 4A); SD at these same positions (Fig. 4B), as well as DD at these same positions (Fig. 4C).Bees discriminate two signal patterns whenever differences appear in either position or or in both of these positions (Fig. ). If the difference appears in position then they may or may not discriminate the patterns (compare Fig. 4A, third column with Fig. B, second and fourth column, and Fig. 4C, third column). The only case in which different signals in position are discriminated (Fig. 4B second column) comes from experiments with a rather large number of decisions (n = ). In all cases in which only the fourth position presented different signals, no significant discrimination between the two patterns is found (Fig. 4B last two columns). A possible indication of the rank order of salience in positions from to can also be seen when results from experiments are compared in which the signals are the same or different in one position, while other positions provided different signals. Consider the examples in which position presents the difference and position either a similar signal or no signal at all (Fig. 4A, C: [D S Ø Ø], [D Ø S Ø], [D Ø Ø S], [D Ø D Ø]. The highest scores are found for [D Ø S Ø] and [D Ø Ø S]. Although the respective discrimination scores are not significantly different, the discrimination scores follow [D Ø S Ø]>[D Ø Ø S]>[D Ø D Ø]) indicating that the similar signal at position may reduce discrimination. This effect is also seen if position presented a similar signal and position presented a different one (Fig. 4B: [S D Ø Ø], as compared to Fig. 4A: [D S Ø Ø]: [D S Ø Ø]>[S D Ø Ø]). A similar tendency is seen for position (Fig. B: [Ø S D Ø], as compared to Fig. 4A: [Ø D S Ø]; [Ø D S Ø]>[Ø S D Ø], P = .), and for position if, for example, position provides a different signal but position a similar signal (Fig. 4B: [Ø Ø S D], as compared to Fig. 4A: [Ø Ø D S]; [Ø Ø D S]>[Ø Ø S D]). No such effect is seen for position (Fig. 4C: [Ø Ø D D] as compared to Fig. 4A: [Ø Ø D S], P = .).The combined effect of differences and similarities in these two signal patterns on the rank order of positions – can be seen in a comparison of several other cases presented in Fig. . For example, comparing the discrimination scores in Fig. 4A with those in Fig. 4B indicates that the respective scores in Fig. 4A are, in most cases, significantly higher than those in Fig. 4B (Fisher's exact test: PDSØØ vs SDØØ = .; PØDSØ vs ØSDØ = .; PØØDS vs ØØSD = .; PDØSØ vs SØDØ<.; PDØØS vs SØØD = .; PØDØS vs ØSØD = .). One would expect that different signals at both positions may add up and make it easier for the animal to discriminate the two alternatives. This is not the case, as a comparison between Fig. 4A and Fig. 4C shows. The scores given in Fig. 4C should be higher if the effect of differences were to add up, in fact, the corresponding patterns for D S versus D D are not significantly different (Fisher's exact test: PDSØØ vs DDØØ = .; PØDSØ vs ØDDØ = .; PØØDS vs ØØDD = .; PDØSØ vs DØDØ = .; PDØØS vs DØØD = .; PØDØS vs ØDØD = .). This effect may indicate that the single signals at the positions are not learned independently, but rather as a sequential pattern, which allows better discrimination if the two signal sequences differ when compared to similar sequential signals. This question will be further addressed with transfer tests (see below).() Three signals: differences in position and colorSeven triple patterns are examined for the three positions , and ([D D D], [D S D], [D S S], [S D D], [S D S], [D D S] and [S S D]). Position is excluded from these experiments because it was found earlier that signals at position may not contribute at all or only marginally to discrimination. In each of the seven triple patterns, two experiments are performed which, because of the earlier findings showing that both colors and direction in the T-maze are interchangeable, are pooled in Fig. . The particular color/position assignments for these experiments are given in the legend for Fig. . Again, higher discrimination scores are found for different signals at position . All scores in Fig. 5A are significantly higher than those in Fig. 5B, where the difference between those two groups of triple patterns is that those in Fig. 5A have patterns with different signals at position and those in Fig. 5B have the same signals in position (Fisher's exact test: PD** vs S**<.). Differences only in position are discriminated, but those in position are not discriminated. The latter case is particularly interesting because the number of choices in these tests is rather high (n = ).A rank order of salience for positions and can be seen when comparing those triplets which differ only in position or . Animals trained to a [DDD] situation show a significantly higher percentage of correct choices than the animals trained to a [SDD] situation (Fig 5A, G-test: PDDD vs DSD = .). The same tendency is observed when position provides the same signal and the other two positions the differences (Fig , G-test: PDDD vs DSD = ., after α level correction, differences should be taken as significant only if P<.). These results show that position have the higher salience, followed by position . Position does not contribute to discrimination in these triple patterns indicating that similar signals at positions and/or can overshadow the small contribution of position as seen in the dual patterns (Fig. 4B).B. Transfer TestsHidden pattern effects may be uncovered by transfer experiments in which the color at a particular position (Fig. ), or its position or the number of signals and their positions (Fig. – ) are changed in the test. Using such tests, we ask how well the animals are prepared to generalize from the learned to the transfer test stimulus conditions. Such transfer tests are always in between-training tests, because the animals are not rewarded under any of the test conditions. Since one cannot predict which choice (right or left) would be the correct choice, Figs. – plot the choice probability as a percentage of one direction (left).() Color transfer testsColor transfer tests are carried out after training to either equal or different numbers of signals in the two alternatives (Fig. ). Bees transfer easily from blue to gray signals (Fig. 6A, Fig. ; G indicates gray), or from yellow to blue signals (Fig. 6B–D). If the difference between the two signals during training lies in the positions and of the color, then bees decide according to the match of position (Fig. 6C).() Color, number and position transferThe findings of the color transfer experiments are corroborated by results of transfer tests in which both color and position are changed (Fig. ). Bees choose according to the match with the color of position , and appeared to ignore positions –. If position does not provide a color signal during the transfer test (Fig. 8A, C, D), but position shows the color signal which the animal learned at position , the choice does not differ from the trained conditions, indicating that similar signals at position in training and transfer patterns overshadow those in other positions. However, if position does not provide any information in the transfer test the learned signal at position is transferred to position .() Position transferSystematically changing stimulus positions by keeping the number and the colors of the stimuli constant indicates several additional rules about the role of stimulus position. Well-discriminated patterns (Fig. 8B: [Ø Y B Ø] vs. [Ø B Y Ø], Fig. 8E: [Ø B Y Ø] vs. [Ø Y Y Ø]; Fig. 8G: [B Ø B Ø] vs. [Y Ø B Ø], Fig. 8H: [B Ø Ø B] vs. [Y Ø Ø B]) are less well discriminated if the respective patterns are moved backwards from the T-choice point, to higher position numbers (Fig. 8B to [Ø Ø Y B] and [Ø Ø B Y]; Fig. 8E to [Ø Ø B Y], Fig. 9G: [Ø B Ø B] and [Ø Y Ø B]; Fig. 8H to [Ø Y Ø B] and [Ø Y Ø B]. Position is relevant for this position transfer because the training pattern [Ø B Y Ø ] is transferred to [Ø Ø B B ] and the training pattern [Ø Y Y Ø ] to [Ø Ø Y Y] (with a significant difference between the choices in the transfer patterns, P = ., Fisher exact test, see Fig. 8E).() Position and number transferCombined transfer tests for positions and numbers of signals are run after three training pairs [Ø B Ø Ø] vs. [B B B B], [Ø Y Ø Ø] vs. [Y Y Y Y], and [B Ø Ø Ø] vs. [B B B Ø] (Fig. ). The respective patterns for the two colors are pooled because no significant difference was found between the respective transfer tests, and these patterns are expressed in Fig. as [ØX Ø Ø] vs [XXXX] for the training patterns, and [X Ø Ø Ø[ , [Ø Ø X Ø], [Ø Ø Ø X], [X X Ø Ø], [Ø Ø X X], [X Ø Ø X], [Ø X X Ø] for the transfer pattern). Position again turned out to be the most important stimulus position (e.g. [X X Ø Ø] or [X Ø Ø X] or [X Ø Ø Ø] are chosen as the trained pattern [X X X X] and [Ø X X Ø] as the alternative trained pattern [Ø X Ø Ø]). This indicates that the animals do not interpret the signal in the trained pattern [Ø X Ø Ø] as being presented at position , but learned it as the signal for position . Patterns [X X Ø Ø] and [X Ø Ø X] are transferred to trained pattern [X X X X], and patterns [Ø Ø X Ø] and [Ø Ø Ø X] are transferred to [Ø X Ø Ø], indicating that either the animals always turn according to trained pattern [X X X X] if there is a signal at position and according to trained pattern [Ø X Ø Ø] if there is no signal at position , or that the number of signals also play a role, irrespective of position. The significantly lower transfer to pattern [X X Ø Ø] than to pattern [X Ø Ø X] ( P = ., Fisher's exact test) shows that bees learn to relate position to the alternative trained pattern. Transfer of pattern [X Ø Ø X] to [X X X X] results in significantly higher choice values than the choice of the trained pattern [X X X X] (P = ., Fisher's exact test), indicating that similar signals at position in the trained patterns reduce the choice for the trained signals, and that position contributes to the choice.Stimulus conditions which are not learned ([Ø Ø B Y] vs [Ø Ø Y B]; [B B B Ø] vs [B Ø Ø Ø ]) do not lead to significant transfer (transfer pattern tested: [Ø Ø Y Y], [Ø Ø B B], [Y B Ø Ø ], [B Y Ø Ø ], data not shown). In another case of a non-significant training effect ([B B B B] vs [B Ø Ø Ø]) the transfer to [B Ø B Ø] (n = ) is significantly different from those to [Ø B B Ø] (n = ) and [Ø B B B] (n = ) (P = ., and P = . respectively, Fisher's exact test, data not shown) indicating that a transfer test can uncover a weak learning effect with respect to differences in signals at positions –. Fig. 8B, and Fig. 8F–J show cases in which either no significant transfer is found or the transfer is strongly reduced although the training patterns are well discriminated: Fig. 8B: [Ø Ø Y B] and [Ø Ø B Y] after training [Ø Y B Ø ] vs [Ø B Y Ø ]; Fig. 8F: [B Y Ø Ø], [Ø Ø B Y] and [Ø Ø B B] after training to [Ø B B Ø] vs (Ø B Y Ø]; Fig. 8G: [Ø B Ø B], [Ø Y Ø B] after training to [B Ø B Ø] vs [Ø V B Ø]; and Fig. 8H: [Ø B Ø B] after training to [B Ø Ø B] vs [Y Ø Ø B]. In all these cases the transfer patterns provide on the one side the same patterns as the trained pattern, but since it is shifted to a different position a mismatch results between the learned signals at the respective position. This indicates that the sequence in the pattern and the signal positions both play a role in transfer.Taking all these transfer results together, we can derive the following rules: () Generalization to different colors is readily performed. In this case, decisions are made with reference to position and number of signals. () Signals at position provide the highest salience, and a shift to position , or lead to a gradual reduction of salience. This gradient resembles the one found in the training experiments, which showed that position has the highest salience, position a somewhat lower one, position a much lower one and position a very low salience. () Transfer is reduced or lost if position provides the signal for discrimination during training, but is empty in the transfer patterns (Fig. 8G, H, I, J). () If discrimination learning is based on position , as in the case of the experiment shown in Fig. , the signals at position override other criteria such as the number of signals. () Number matching can also guide choice behavior (Fig. ). () Serial patterns appear to play a role, too, but the effect is small (See Fig. 8B, C, E, F).C. Model Calculation() Modeling the positional salience scores (PSS)The results from both the discrimination and the transfer experiments clearly show that serial position is the main parameter. The question I shall address next is how well the data can be understood by assuming a particular rank order of positional salience, and whether additional factors may guide the bee's decisions.Position provides the most salient signal and position the least, if any. To address the question of whether position is the only relevant parameter I modeled the effect of positional salience on discrimination by assuming two relations between position and a positional salience score (PSS). The linear model relates positions , , , and to the PSS of ., ., ., and ., and the stepwise model assumes an equal PSS of . for positions and , . for position , and zero PSS for position . Since the animals learn the difference between the two signal patterns the model creates a cumulative PSS that takes into account the differences at the respective positions weighted with their respective salience scores. The PSS are calculated for the discrimination tests, transfer tests are not considered.It is not obvious how the animals might have related the differences at each position. Therefore, two calculations of the cumulative differences in PSS are run. In calculation (all signals were learned independently) it is assumed that animals learn each signal in both alternatives, weight them according to the PSS, and the cumulative differences in PSS result from the sum of the differences. No PSS is given to a position without a signal. For example (numbers in brackets are the respective PSS according to the linear model): [B (.) B (.) B (.) B (.)] vs [B (.) Ø () Ø () Ø ()] lead to differences in cumulative PSS: +.+.+. = .. In calculation (signals are learned with respect to each other) it is assumed that only those signals are learned that are different. The respective numbers for the same examples are (again according to the linear model): [B (.) B (.) B (.) B (.)] vs [B (.) Ø () Ø () Ø ()] lead to differences in cumulative PSS: +.+.+. = .. The same two calculations are performed for the stepwise model. Since choice performance is expressed in % for each of the two alternatives and the model requires the calculation of differences between the choice performances, the % values are first linearized by calculating the corresponding probit values and then subtracting these probits. This gives a probit value for each pair of trained patterns.The four models give the following results (Pearson correlation): linear model, calculation : P<., r2 = .; linear model, calculation : P<., r2 = .; step model, calculation : P<., r2 = .; step model, calculation : P<., r2 = .. The correlation coefficients for the four models are only marginally different. The best correlation is found for the linear model and calculation (all signals are learned independently). This result is shown in Fig. . Although the model calculations show a correlation between the salience scores and discrimination, the rather low correlation coefficients do not allow distinguishing between the different assumptions behind these models../journal.pone..g010Figure 10Linear regression between discrimination scores (ordinate) and the sum of position salience scores (PSS, abscissa) as calculated with the linear model applying calculation (all signals are learned independently, see text).The discrimination scores (probit) result from converting the percentage of correct choices for each of the two alternative patterns into probit values and subtracting these probit values. The PSS for position is set to ., that of position to ., that of position to ., and that of position to . (calculation , see text). Pearson correlation: P<., r2 = .. The data points marked with the numbers of the respective experiments (see table ) indicate outliers (selected by eye); numbers , and for experiments in which discrimination appears to be better than expected from the model calculation, numbers , 19b, , 33a, , for experiments in which the discrimination scores appear to be lower than expected from the model calculation (see text). The thin lines show % confidence range.Nine data points in Fig. are marked as particularly clear deviations from close correlation. Three data points result from experiments in which performance appears to be better than expected from the model (exp. , , ), and data points in which performance is less than expected from the model (exp. , 19b, , 33a, 36a, ). Better performance than expected is found if only one signal is provided (exp. : [Ø Ø B Ø] vs [Ø Ø Y Ø]) or if one of the two signals shows a difference in position and the second signal is the same at either position or (exp. : [B Ø B Ø] vs [Y Ø B Ø]; exp. : [B Ø ØB] vs [Y Ø ØB]). Lower performance is found for serial signals which resemble components of a regular pattern in the two alternatives as, for example, reversed order (exp. : [Ø Y B Ø] vs [Ø B Y Ø]; exp. : [B B Y Ø] vs [Y B B Ø]; exp. : [Y B B Ø] vs [B Y B Ø]; exp. : [Y Y B Ø] vs [B B Y Ø]) or two equal signals in a series of three (exp. 33a: [Y Y B Ø] vs [Y B B Ø]). These results could indicate that bees may evaluate the serial positions not only independently, but also to some degree as a pattern or a configural unit.DiscussionThe task honeybees had to solve in these experiments was to fly either to the right or to the left in a T-maze to obtain food reward depending on two different patterns of up to sequential blue or yellow signals. The sequential patterns were presented in a flight tunnel with black and white patterns on the floor and walls. The aim was to simulate a navigational task during a foraging episode, with the parameters involved in sequential landmark experience during approach flights more precisely controlled than it would be possible in a natural foraging range. Bees are known to interpret the length of their flights through a narrow tunnel as up to five times longer than their flights in the open due to their distance estimation via visual flow field [], [], [], and they are also known to learn the sequence of landmarks on their foraging trips [], [], []. Therefore, it might well be that bees apply a form of observational or latent learning in such a simulated navigational task, and thus the sequential signals may be learned in spatial-temporal relation to each other, rather than in their temporal contiguity with respect to the reward. If the sequential signals would be learned in relation to each other with the potential to be grouped together to form unique configured experiences we expect rather equal salience of the various positions, and the appearance of phenomena indicative of stimulus configuration. If, however, associative learning dominates learning strategy the most recently experienced signal before reaching the reward might be weighted highest, and thus the salience of the sequentially experienced positions should be different. This is what I found. Position closest to the intersection and the reward has the strongest impact on discrimination and generalization, position provides close to equal salience as position , position has a lower salience than position , and position contributes only weakly or not at all to discrimination and generalization. Thus the temporal contiguity between the sequential signals and the reward appears to be the most important parameter. Configural phenomena are not absent but have a low impact (see below).These results can be interpreted in several ways. I first consider the possibility that training in a narrow tunnel may not simulate a natural navigational situation. Several reasons may account for this effect. a) Similar context conditions: Although bees may experience longer flight distances in the tunnel the same location in the navigational space of the bee may not allow the bee to associate different serial positions of “landmarks” to different flight directions. This possibility is important because similar training conditions with bees flying over short distances in boxes or tunnels were used in the past to study questions related to navigation in bees (review: []). My results call into question whether such tests conditions allow generalization to the navigational context. This argument of caution may be particularly relevant for a recent study in which bees were trained to visual signals presented in a narrow tunnel, where these signals were denoted as “landmarks” []. b) Temporal order during learning: Bees as other animals [] localize important items in time. They match their choices to the average of reward as experienced in multiple visits [], [], they monitor the gradient of reward over multiple visits [], they activate time linked memories in diurnal rhythms [], [] and they learn to visit sequential feeding places at long distances during one foraging bout (unpublished observation). However, in all these cases temporal sequences appeared at much longer intervals (minutes to hours instead of seconds). Therefore, the temporal weighting of information may follow different rules of integration and memory formation. c) Temporal order during memory retrieval: The weight of recent experience is often higher in memory retrieval shortly after learning, whereas longer retrieval intervals favor may more balanced weights for sequential items [] (see also below). Furthermore, the last learned item overshadows earlier learned items (bees: [], []). These recency effects has been studied in the context of navigation for multiple phases of learning, and thus may not apply for the training conditions used here, but it will be necessary to ask in future experiments whether the recency effect seen here depends on the interval between learning and retention test. Taken together these arguments favor the conclusion that sequential signal learning in the T-maze does not mimic a navigational task.It thus appears that the training conditions in the T-maze favor associative learning of multiple sequential stimuli experienced within the time span of sensory/working memory as tested in trace conditioning paradigms. Two parameters may be instrumental for the rank order of positional salience, limited time span and limited capacity of sensory/working memory. A multitude of associative learning phenomena favor temporal recency to the evaluating (reinforcing) stimulus (e.g. Pavlovian and instrumental conditioning, forgetting, recovery from extinction, delayed matching-to-sample, overshadowing, and others; review with respect to the bee: []). In bees the optimal CS-US interval is in the range of a few seconds both in instrumental color learning of freely flying animals and in odor conditioning of the proboscis extension response. Sensory memory can be extended to about sec by an autoshaping procedure in which free-flying bees were trained in a dual forced-choice tests to expect delayed reward []. Since the bees experienced the signals in the access arm of the T–maze within a few seconds (.+/−. s) before arriving at the reward site I conclude that the time span for trace conditioning is not the limiting factor. It is thus more likely that limited capacity of the sensory store or cue competition between successively experienced stimuli is the critical factor.Evidence for a limited capacity of the sensory store comes from matching-to-sample experiments which were carried out with two sequential visual stimuli under similar conditions as applied here [], []. In such experiments it was found that bees learned to use signals but not for a later match, indicating that the storage capacity for item numbers may be very much limited. In experiments aiming to elucidate the question whether bees have some competence of counting (see below) the number of visual signals referred to reached to . Thus the rank order of salience may reflect a sensory/working memory store limited to items in bees.One way of conceptualizing cue competition within sensory memory is to assume a form of overshadowing between stimuli. Overshadowing of stimuli equipped with different salience has been found in instrumentally trained [] and in classical conditioned bees []. If sequential stimuli are equipped with different salience according to position, the first one experienced providing least salience, the last one experienced highest salience, the last ones will overshadow the first ones and association will be reduced accordingly. Another way of looking into cue competition would be to understand sensory memory as a shift register: whenever something new comes in a previously stored item has to leave (the first-in-first-out rule, see for example []). In any case it appears that the stimulus traces initiated sequentially during the approach flight may act on each other competitively rather than supporting each other as assumed, e.g. in the associative chaining or positional coding hypotheses [].Sequentially experienced signals have numerical attributes. Thus the question arises whether bees extract such attributes from the pattern of signals as tested here. The number of sequentially experienced landmarks was found to be a guiding factor in bee navigation [], and it was concluded that bees judge distance flown not only on the basis of their visual odometer but also on the basis of the sequence and “number” of landmarks passed by. This capacity might reflect a basic form of precounting [] which has been assigned to many animals including insects [], [], [], [], [].Recently the capacity of sequential numerosity has been demonstrated for bees flying in a tunnel, and passing by up to four signals at varying spatial separation []. This study showed that bees appear to learn up to three sequential signals, and since they transferred the trained “number” to novel signals it was concluded that they might perform some form of “exact counting”. Irrespective of the validity of this claim it is important in our context that the estimated upper limit of sequential signal numbers in a tunnel is three - a number that coincides with what I have found in the experiments presented here, and what is known form many studies in animals and humans (e.g. monkey: [], human babies: []) . With respect to the question of whether my results are indicative of exact counting one might be skeptical because of the strong rank order of position salience. Although bees discriminate sequences of similar signals (Fig. ) the effect is small, and the dominance of the signal in position reduces the number effects in most other test conditions. The results of transfer tests after training [X X X X] versus [Ø X Ø Ø] (Fig. ) could indicate that the numbers of signals are of importance, and that even position may contribute to this effect. However, the transfer results could also be explained if one assumes that bees behave according to what they have learned about position (with a signal in pattern [X X X X], no signal in pattern [Ø X Ø Ø]). Approximate counting, the discrimination of vs , vs , vs is supported by my data, but since the number effect is limited to (or ) it is not possible to test a characteristic property of approximate counting namely that it follows Weber's law (equal ratios are discriminated equally, []). Clues to numeric ability could be derived from the observation that sequential pattern effects are less strong in patterns with signals as compared to those with two signals, indicating that configuration of sequential patterns may be limited to two signals. Thus the numerosity effect in associative learning as studied here may be limited to two, and is overshadowed by the strong position salience of position .Sequences can be learned only if the animal keeps the temporal order in its sensory/working memory. It is well known from studies in humans and animals that sensory memory (also called primary, short-term or working memory with different emphasis on experimental paradigms) is limited in time and capacity. Serial memory tests have been used since James [], Ebbinghaus [] and Müller [] to characterize the organization and temporal dynamics of this initial memory. Human subjects learning word lists and animals learning a series of visual features or landmarks show a serial position function (position meaning temporal position in most test conditions) with high retention scores at the beginning and the end of the series (primacy and recency effect, []. However, such a U-shaped retention function may not be the general case, since different stimuli (visual or auditory), different training-test intervals and different animal species may give rather divergent results. Under certain conditions the recency effect is favored, under other conditions the primacy effect, and sometimes both effects are detectable (U-shape function). For example Wright [] [] found that the recency effect dominates for short retention intervals ( – sec), and the primacy effect dominates for longer retention intervals ( s, s). Different animal species and different stimulus modalities influence the shape of the positional retention function (review []). The data reported here document the recency effect. This is in line with the observation that short retention test intervals favor the recency effect.Next I want to ask whether there is any evidence for configural phenomena in the T-maze serial signal learning task. Since the signals are experienced well within the time span of working memory one might expect that these signals might be grouped together to form a unique compound. Although the serial salience effect is strong there is evidence that bees indeed learn the sequential patterns also as stimulus sequence patterns. () A comparison of the results in Fig. 4A and Fig. 4C shows that the discrimination scores for signals in positions and (Fig. 4C) do not add up and, in some cases, the scores are even lower than the respective ones for patterns with a difference only in position (Fig. 4A). Discrimination is reduced if the two sequences have common features, whereas signals are better discriminated if they differ in pattern features. Thus signals at the positions , and (to some extent) are not only learned independently, but also as a sequential pattern indicating that these sequences are also learned as configured patterns. () Transfer experiments show that patterns are learned for their respective relative positions, and signals are not only weighted according to the positional salience rank order. For example, the transfer pattern [Ø X X Ø] is chosen more strongly than the trained pattern [X X Ø Ø ] (Fig. 8C), and the transfer pattern [X X Ø Ø] is less strongly chosen than the trained pattern [ØXX Ø] (Fig. 8E). () Although the model calculation on the basis of positional salience of the isolated signals captures most of the data quite well, I selected a few examples in which better performance or worse discrimination than expected was found (Fig. ). Sequential patterns were better discriminated if only one signal provided the difference (exp. : [Ø Ø B Ø] vs [Ø Ø Y Ø]) possibly indicating a lack of interference from other signals. Better discrimination was also seen if one of the two signals showed a difference in position and the second signal was the same at either position or (exp. : [B Ø B Ø] vs [Y Ø B Ø]; exp. : [B Ø ØB] vs [Y Ø ØB]). This effect may hint at a priming effect of the first signal encountered: the discriminative signal might receive additional value. Lower performance was found for serial signals which resembled components of a regular pattern in the two alternatives as, for example, reversed order (exp. : [Ø Y B Ø] vs [Ø B Y Ø]; exp. : [B B Y Ø] vs [Y B B Ø]; exp. : [Y B B Ø] vs [B Y B Ø]; exp. : [Y Y B Ø] vs [B B Y Ø]) or two equal signals in a series of three (exp. 33a: [Y Y B Ø] vs [Y B B Ø]). Taken together these results indicate that bees learn these sequential patterns to some extent as configural units in addition to their isolated functions. Training against the dominance of the positional rank order may allow isolating these configural components. Configuration of sequential patterns in mammals has been interpreted as indicating relational representations in time and space [], []. The structure of such representations resemble key components of episodic-like memory, e.g. the configuration of sequentially experienced odors as a unique episode. It will be a question for future experiments to test whether a comparable memory structure exists in an insect, the honeybee.Supporting InformationTable S1Discrimination tests. The table gives the choice data of all experiments for discrimination tests. Rows (column a) are numbered as in Table (test). Column b gives the number of the experiment as in Table , and the number of tests performed in the particular experiment. Column c shows the training pattern, first for the training to the left side of the T-maze, and then for the right site. Column d gives the number of animals trained and tested in the respective experiment. Column e gives the choices summed up for all tests first for the choice of the left arm and then the choice of the right arm of the T-maze. Column f gives the % of correct choices.(. MB DOC)Click here for additional data file.Table S2Transfer tests. Rows and columns a, b, c and d are the same as in table S1. Column e gives the patterns presented during the transfer tests together with the choices for these patterns. As in table S1 the choices of the left arm are shown first and then the ones for the right arm of the T-maze. Column f gives the % of choices for the left arm.(. MB DOC)Click here for additional data file.
|
PMC1635055.txt
|
TITLE: Severe malnutrition with and without HIV- infection in hospitalised children in Kampala, Uganda: differences in clinical features, haematological findings and CD4+ cell counts AUTHORS: Hanifa Bachou, Thorkild Tylleskär, Robert Downing, James K Tumwine ABSTRACT: BackgroundThe aim of this study was to describe the clinical features, haematological findings and CD4+ and CD8+ cell counts of severely malnourished children in relation to human immunodeficiency virus (HIV) infection.MethodsThe study was conducted in the paediatric wards of Mulago hospital, which is Uganda's national referral and teaching hospital. We studied severely malnourished children (presence of oedema and/or weight-for-height: z-score < -) and have presented our findings. At admission, the CD4+ and CD8+ cells were measured by the flow cytometry and HIV serology was confirmed by Enzyme linked Immunoassay for children > months of age, and RNA PCR was performed for those ≤ months. Complete blood count, including differential counts, was determined using a Beckman Coulter counter.ResultsAmong the children, (%) were female; the median age of these children was months (Interquartile range – months), and no difference was observed in the HIV status with regard to gender or age. The children showed a high prevalence of infections: pneumonia (%), diarrhoea (%), urinary tract infection (%) and bacteraemia (%), with no significant difference with regard to the HIV status (HIV-positive versus HIV-negative children). However, the HIV-positive children were more likely to have persistent diarrhoea than the HIV-uninfected severely malnourished children (odds ratio (OR) ., % confidence interval (CI) .–.). When compared with the HIV-negative children, the HIV-positive children showed a significantly lower median white blood cell count ( versus ) and lymphocyte count ( versus ). The CD4+ cell percentages were more likely to be lower in children with non-oedematous malnutrition than in those with oedematous malnutrition even after controlling for the HIV infection.The novel observation of this study is that the CD4+ percentages in both HIV-positive and HIV-negative children without oedema were lower that those in children with oedema. These observations appear to imply that the development of oedema requires a certain degree of immunocompetence, which is an interesting clue to the pathophysiology of oedema in severe malnutrition. BODY: BackgroundSevere malnutrition has been associated with acquired immunodeficiency (AID) among children worldwide, and it is referred to as Nutritionally Acquired Immunodeficiency Syndrome or NAIDS [,]. With the advent of the human immunodeficiency virus (HIV) pandemic, there has been a tendency to overlook the role of malnutrition in immunodeficiency, and indeed, only a handful of studies have investigated the CD4+ and CD8+ lymphocyte subsets in severely malnourished children [,].There is little information on the effect of the added burden of HIV infection on the clinical features [-] and cellular immunity of severely malnourished children. The objective of this study was to report the clinical features, haematological findings and CD4+ and CD8+ lymphocyte subsets of severely malnourished children with regard to their HIV status.Subjects and methodsAll severely malnourished children consecutively admitted to the paediatric wards of Mulago hospital, which is Uganda's national referral and teaching hospital, during the two peak seasons of malnutrition, namely, September-November and September-December were followed up from the time of admission to outcome (death or discharge). In this study, we included a total of severely malnourished children (presence of oedema and/or weight-for-height: z-score < -) after obtaining the informed consent of their parents or caregivers (Figure ); the age of these children was below months. The risk factors for death in the first peak () have been stated in a previous report that describes the methodology in greater detail.Figure 1Study profile showing the enrolment process of the children in the study.In this paper, we report the complete results of the HIV tests as well as the CD4+ and CD8+ cell counts and percentages of the children. In the case of children, complete laboratory data could not be obtained; this was due to the lack of reagents in the case of children; inadequate blood volume in , haemolysis in and absence of blood sample in . The basic characteristics of the children with complete results were compared with those of the children with incomplete results.The following parameters were recorded for all the children: demographic characteristics (age and sex), clinical features (weight, height/length and presence of oedema and diarrhoea), haematological tests (haemoglobin concentration, white blood cell (WBC) count and differentials and presence of malarial parasites), HIV tests (ELISA and RNA PCR), microbial tests (blood and urine culture and sensitivity), immunologic tests (CD4+ and CD8+ cell counts and percentages) and chest x-ray reports. We used the CD4+ cell percentage to categorize children with or without the HIV infection. The clinical definition of malnourishment classified all our patients into category C because all the children were severely malnourished. The haemoglobin concentrations were evaluated according to the WHO criteria: < g/dL and < mg/dL are referred to as severe anaemia and very severe anaemia, respectively.Laboratory methodsBlood was collected in -ml EDTA vacutainer tubes (Becton Dickinson, Franklin Lakes, NJ, USA) in the mornings between – am by venipuncture and transported within h to Uganda Virus Research Institute (UVRI) laboratory, Entebbe for serological testing. HIV testing was performed using the standard HIV algorithm of two enzyme-linked immunoassays (EIA) in parallel. Western blot and real-time polymerase chain reaction (RT-PCR) were performed to confirm a positive EIA test for children below months of age and those with indeterminate results on EIA.TriTEST reagents (CD3, FITC/CD4, PE/CD45, PerCP and CD3, FITC/CD8, PE/CD45, Per CP) were used to stain PBMC for CD4+/CD8+ cell counting according to the manufacturer's instructions. FACScan instrument and MultiSET software were used to perform flow cytometry and report the absolute CD4+ and CD8+ cell counts of each sample by using the dual-platform approach (Becton Dickinson, Franklin Lakes, NJ, USA). Complete blood count, including differential counts, was assessed using a Beckman Coulter counter []. Blood was stained within h of collection, and the observations were analysed within h.Severe malnutrition was defined according to the WHO classification and the presence of severe wasting (weight-for-height < SD of the NCHS/WHO reference values with no oedema) and/or oedematous malnutrition (presence of symmetrical oedema involving at least the feet) []. The children were divided in two groups; HIV-positive and HIV-negative groups.The study protocol was approved by the Regional Committee for Medical Ethics, Bergen, Norway (REK Vest), Makerere University Faculty of Medicine Ethics and Research Committee, Mulago Hospital Ethics Committee and the Uganda National Council for Science and Technology.Statistical analysis was performed using SPSS version . Medians were used to calculate the central tendency and interquartile range (IQR) for the spread of haemoglobin concentration, WBC, total lymphocyte and CD4+ and CD8+ cell counts. Children were grouped by their gender (male or female), age in months (≤ months and > months), presence or absence of oedematous malnutrition and HIV infection and CD4+ levels (CD4+ cell percentage < % and < %). Chi square and Wilcoxon-Mann-Whitney tests and multivariate analysis were used to determine differences with regard to the HIV status, gender and type of severe malnutrition (oedematous versus non-oedematous). A -tailed p value of < . was considered significant. Binary logistic regression models were constructed using the HIV status as the outcome variable. The appropriate important baseline data of clinical significance was included in a regression model and used for adjustment. The chi-square test was used to select variables according to their statistical significance (p < .). Dummy variables were created for the categorical variables used. The chosen dependent variables were tested for interactions, and the very significant variables were stratified to assess for the possibility of effect modification. Positive interactions remained in the final model. Independent variables that showed a persistently non-significant relationship with the dependant variable during modelling were excluded from the final model.ResultsOf the children, (%) were female, and the median age of these children was . months (IQR –). The age of half the children was between – months, and that of a few children (%) was below months (Table ). The age distribution was not affected by their HIV status. Almost half the children (/) had oedematous malnutrition (kwashiorkor and marasmic-kwashiorkor). These characteristics (sex, age and type of malnutrition) were comparable to those of the children with incomplete laboratory data.Table 1Characteristics of children aged below months with severe malnutrition during peak malnutrition periods.Age groupHIV-positive children n = 123HIV-uninfected children n = 192TotalMonthsMaleFemaleMaleFemale0–.–.–.–.–.–.9255416Total754812171315HIV infection was detected in / children (approximately %). The HIV-infected children were less likely to present with oedema (OR ., % CI .–.). Of all the severely malnourished children, only (%) had no identifiable infection on admission, (%) had only one type of infection and the majority, that is, (%) had more than one type of infection on admission. The infections included pneumonia (%), diarrhoea (%), urinary tract infection (%), bacteraemia (%), malaria (%) and oral thrush (%) (Table ). Overall, there was no significant difference in the prevalence of infection with regard to the HIV status. However, the HIV--infected children were more likely to have persistent diarrhoea and oral thrush (Table ).Table 2Characteristics, co-existing medical conditions and diagnosis of children aged < months with severe malnutrition.HIV-positive childrenn (%)HIV-negative childrenn (%)Odds ratio(% CI)Symptoms and signs(n = )(n = )Diarrhoea (all) () (). (.–.)Persistent diarrhoea (> weeks) () (). (.–.)*Oral thrush20 () (). (.–.)*Bilateral oedema (nutritional) () (). (.–.)*Severe dehydration7 () (). (.–.)Chest x-ray findings(n = )(n = )Bronchopneumonia26 () (). (.–.)Interstitial pneumonia40 () (). (.–.)Suspected tuberculosis14 () (). (.–.)Blood tests(n = )(n = )Malarial parasites10 () (). (.–.)Severe anaemia (Hb < g/dL) () (). (.–.)Bacteraemia24 () (). (.–.)Urine tests(n = )(n = )Bacteruria33 () (). (.–.)*Statistically significantThe median haemoglobin concentration of these children was below g/dL. There was no significant difference in the haemoglobin concentration with regard to the type of severe malnutrition or HIV status (Table ). The total WBC count was significantly lower in the HIV-positive children (. × ; IQR .–.) than in the HIV-negative children (. × ; .-.) (p = .). Among the HIV-infected children, the total WBC count was lower in the non-oedematous children than in the oedematous children. However, this difference was not observed among the HIV-uninfected children (Table ).Table 3Laboratory data of all severely malnourished children aged < months grouped by their HIV status.HIV-positive median (IQR)n = 123HIV-negative median (IQR)n = 192Haemoglobin7. (.–.). (.–.)Total WBC (/L). (.–.). (.–.)**Neutrophils (/L). (.–.). (.–.)Neutrophils (%) (–) (–)Monocytes (/L). (.–.). (.–.)Monocytes (%) (.–.) (.–)Total lymphocytes (/L). (.–.). (.–.)**Lymphocytes (%) (–) (–)CD4+ cell count (/L) (–) (–)***CD4+ cells (%) (–) (–)***CD8+ cell count (/L) (–) (–)***CD8+ cells (%) (–) (–)***CD4+/CD8+ ratio0. (.–.). (.–.)****p < ., ** p < ., and *** p < .001Table 4Laboratory data of <-month-old severely malnourished children categorised by their HIV status and malnutrition type.HIV-positive childrenHIV-negative childrenOedema n = median (IQR)No oedema n = median (IQR)Oedema n = median (IQR)No oedema n = median (IQR)Haemoglobin (g/dL). (.–.). (–.). (.–.). (.–.)White blood cells (/L). (.–). (.–). (.–). (.–)Neutrophils (/L). (.–.)*. (.–.). (.–.). (.–.)Neutrophils (%). (–). (–). (–). (–)Monocytes (/L). (–)*. (–). (–). (–)Monocytes (%). (–)*. (.–.). (–). (–.)Total lymphocytes (/L). (.–.). (.–.)*. (.–.). (.–.)Lymphocytes (%). (–). (–). (–). (–)CD4+ cell counts630. (–)***. (–) (–)*** (–)CD4+ cells (%). (–)***. (–). (–)***. (–)CD8+ cell counts1046 (–). (–). (–)** (–)CD8+ cells (%). (–)*. (–). (–). (–)CD4+/CD8+ cell ratio0. (.–.)***. (.–.). (.–.). (.–.)*p value < ., **p value < ., and ***p value < ., comparing each HIV status group with the type of severe malnutrition.The total lymphocyte counts were . × (IQR .–.) in the HIV-positive children and . × (IQR .–.) in the HIV-uninfected children (p = .). The absolute lymphocyte counts were . × (IQR .–.) in the HIV-infected children and . × (.–.) in the HIV-uninfected children (p < .). The total lymphocyte, monocyte and neutrophil counts were lower in the HIV-positive children with no oedema than in those with oedema; this was not observed among the HIV-negative children.Regardless of their HIV status, children with severe non-oedematous malnutrition (marasmus) had significantly lower CD4+ count, CD4+and CD8+ percentages and CD4+/CD8+ ratios than those with oedematous malnutrition (kwashiorkor and marasmic-kwashiorkor) (Table and Figure )Figure 2Box and whisker plot showing the median and the interquartile range of the percentages of CD4+ cells in severely malnourished children who were grouped based on their HIV status and type of malnutrition.The CD4+ cell percentage was below % in one-third of the severely malnourished children. Among these one-third children, the CD4+ cell percentage was between %–% in % (/) children, indicating moderate cellular immunosuppression, while it was below % in % children, indicating severe cellular immunosuppression. Both these categories of cellular immunosuppression were present in HIV-infected and HIV-uninfected groups (Table ). The CD4+ cell percentages were more likely to be below % (OR ., CI .–.) and between %–% (OR ., CI . – .) in children with non-oedematous malnutrition than in those with oedema. This difference persisted even after controlling for their HIV status (Figure ).Table 5Distribution of all severely malnourished children by malnutrition type, cellular immunological category and HIV status.HIV-positive children n (%)HIV-uninfected children n (%)TotalOedeman = 53n = 119n = 172CD4+ ≥ %* () () ()CD4+ %–%** () () ()CD4+ < %*** () () ()No oedeman = 70n = 73n = 143CD4+ ≥ %* () () ()CD4+ %–%** () () ()CD4+ < %*** () () ()*No evidence of suppression, **Evidence of moderate suppression, ***Severe suppressionDiscussionIn this study, we have reviewed an old problem – severe malnutrition in children – by using modern techniques for assessing the immunocompetence that has developed over the last decades in response to the HIV/AIDS pandemic. We have used the up-to-date laboratory techniques for the assessment of lymphocyte subsets in recognised laboratories. Since some of the severely malnourished children are also HIV positive, it is now possible to describe the clinical and laboratory features of the two groups of patients: severely malnourished children with HIV infection and those without HIV infection. We have noticed the well-known fact that the clinical features of severe malnutrition and HIV/AIDS overlap in young children. This affects the possibility of an accurate clinical diagnosis of HIV infection in settings with poor resources and often with inadequate HIV-testing facilities, particularly in cases wherein the two conditions, namely, malnutrition and HIV infection co-exist []. Therefore, a question that arises is whether there are any clinical differences that may raise a suspicion of HIV infection in cases of severe malnutrition?In our study in Uganda, both the groups showed a high prevalence of multiple infections, including pneumonia, diarrhoea, bacteraemia, malaria, urinary tract infection and oral thrush. For respiratory, blood stream or urinary tract infections, no significant difference was observed with regard to the HIV status. The only two conditions that were over-represented among the HIV-positive children were persistent diarrhoea and oral thrush. In view of the remarkable difficulties encountered while clinically differentiating between malnutrition and HIV infection, we strongly support the establishment of routine counselling and testing for HIV- infection among paediatric patients with severe malnutrition in settings where HIV infection is an existing problem. In our study, we observed that there was a high acceptability of counselling and testing for HIV- infection; another reason for the absence of hesitation in organising routine counselling and testing for HIV- infection is that the study subjects were in the paediatric age group.The drop-out cases in this study were mostly due to random factors, and we believe that these did not affect the selection of the study subjects in any systematic manner. In addition, the basic characteristics of the of the severely malnourished children analysed in this study were the same as those of the drop-out cases.The median CD4+ cell counts and percentages observed in this study were compared to the recently reported median CD4+ counts and percentages of healthy Ugandan children younger than years []. Among the HIV-uninfected children without oedema, as many as one-third had signs of immunosuppression with a CD4+ percentage below %; this number was in in the HIV-uninfected children with oedema. Almost % of the HIV-positive children without oedema had signs of immunosuppression that was revealed as a CD4+ cell percentage below %. Approximately half the HIV-positive children with oedema had CD4+ counts below %. Very low CD4+ cell percentages consistent with a laboratory diagnosis of AIDS have rarely been described in HIV-uninfected children with or without mixed infections. Reports on the proportions of T cell and CD4+ cell percentages in severely malnourished children have shown inconsistent findings [,,,]. The difference in the results may be influenced by the differences in study designs and sample size.Alterations in the haematological functions in malnutrition have been documented []. A recent study has reported cases of malnourished children with mixed infection in whom the monocyte counts were higher than those in malnourished children with only respiratory infection although their HIV status was not reported []. Therefore, both granulocyte and lymphocyte suppression observed in this study is an indication of reduced haemopoietic function, and the additional burden of HIV- infection appears to further reduce this function.The CD4+ cell percentages in this study were lower in children who presented with non-oedematous severe malnutrition, and this finding was consistent in both HIV-infected and uninfected groups. Earlier studies reported that oedematous malnutrition had lower T cell counts [,], while others found no difference in the T cell count with regard to the type of malnutrition []. The reason for these controversies is unclear. The only known fact is that severe malnutrition alters the immunological competence through a number of mechanisms, including apoptosis of the thymus gland [, ] and micronutrient deficiencies []. Likewise, the rapid destruction of the CD4+ T lymphocytes by the HIV- virus has been well established. However, mechanisms leading to cellular immunological alterations in cases wherein severe malnutrition and HIV- virus infection co-exist are yet unclear.It is interesting to notice that the HIV-positive children less often present with oedema, that is, oedema was present in just over % of the HIV-positive children, while it was seen in over % of the HIV-negative children. Severe wasting in the absence of oedema is a common feature observed in severe malnutrition with concurrent HIV infection [,,-]. The novel observation of this study was that the CD4+ percentages were lower in both HIV-positive and HIV-negative children without oedema than in children with oedema. Both the above-mentioned observations appear to imply that the development of oedema requires a certain degree of immunocompetence, which is an interesting clue to the pathophysiology of oedema in severe malnutrition.ConclusionSevere protein energy malnutrition is associated with the depletion of the haematological and lymphocyte subsets, and this depletion is exacerbated by the presence of HIV- infection. Cell-mediated immunosuppression is more marked in non-oedematous severe malnutrition, regardless of the HIV status.Competing interestsThe author(s) declare that they have no competing interests.Authors' contributionsAll authors have participated in the design of the study, interpretation of the results, statistical analysis and writing of the manuscript. HB supervised patient recruitment, follow-up and data collection. All authors have read and approved the final manuscript.
|
PMC1440853.txt
|
TITLE: Comparison of gene coverage of mouse oligonucleotide microarray platforms AUTHORS: Ricardo A Verdugo, Juan F Medrano ABSTRACT: BackgroundThe increasing use of DNA microarrays for genetical genomics studies generates a need for platforms with complete coverage of the genome. We have compared the effective gene coverage in the mouse genome of different commercial and noncommercial oligonucleotide microarray platforms by performing an in-house gene annotation of probes. We only used information about probes that is available from vendors and followed a process that any researcher may take to find the gene targeted by a given probe. In order to make consistent comparisons between platforms, probes in each microarray were annotated with an Entrez Gene id and the chromosomal position for each gene was obtained from the UCSC Genome Browser Database. Gene coverage was estimated as the percentage of Entrez Genes with a unique position in the UCSC Genome database that is tested by a given microarray platform.ResultsA MySQL relational database was created to store the mapping information for , mouse genes and for the probes in five microarray platforms (gene coverage level in parenthesis): Affymetrix430 . (.%), ABI Genome Survey (.%), Agilent (.%), Codelink (.%), Sentrix (.%); and four array-ready oligosets: Sigma (.%), Operon v. (.%), Operon v. (.%), and MEEBO (.%). The differences in coverage between platforms were highly conserved across chromosomes. Differences in the number of redundant and unspecific probes were also found among arrays. The database can be queried to compare specific genomic regions using a web interface. The software used to create, update and query the database is freely available as a toolbox named ArrayGene.ConclusionThe software developed here allows researchers to create updated custom databases by using public or proprietary information on genes for any organisms. ArrayGene allows easy comparisons of gene coverage between microarray platforms for any region of the genome. The comparison presented here reveals that the commercial microarray Sentrix, which is based on the MEEBO public oligoset, showed the best mouse genome coverage currently available. We also suggest the creation of guidelines to standardize the minimum set of information that vendors should provide to allow researchers to accurately evaluate the advantages and disadvantages of using a given platform. BODY: BackgroundThe wide use of DNA microarrays to query expression of genes has created the need for updated, consistent and meaningful annotations on the probes included in the microarrays. We refer to gene annotation as a recognizable label or gene id identifying the gene that is targeted by a given probe. Gene ids should be stable, widely used and allow reliable associations among genomic databases. Several microarray annotation systems are available for investigators, aiming to address specific user demands. For instance, the KARMA [] web server provides periodically updated gene annotations of Keck arrays [] and Affymetrix® GeneChips® [], and can also annotate user-provided lists of accession numbers for pair-wise comparisons, even for different species. However, providing a gene list is not always a straight forward process given the large heterogeneity in the format that vendors provide sequence identifiers for probes. For instance, one platform can include identifiers in a Genbank header format such as GB|AY073000.|AAL60663. and others may include different types of ids separated by commas or some other character within a single column. In addition, different sequence identifiers in several columns may be provided by vendors and choosing only one of them may not be the best solution. The Resourcerer database [] tackles this problem by pre-computing gene annotations on a more exhaustive list of microarrays and oligosets for a number of species []. This database is centered on 'tentative consensus' (TC) sequences which are used as gene definitions. TCs group EST sequences that can be aligned and clustered in distinct groups, and these are periodically updated as new ESTs from GenBank become available. Functional annotations on these TCs generate the Gene Indices resource available from the TGI website []. TCs allow for cross species comparisons through the Tentative Orthologue Groups (TOGs) database. However, Gene Indices are not stable and cross referencing to other genomic databases is not easy. A different approach has been taken by Mattes[] who created a set of Perl scripts that use UniGene and LocusLink as gene identifiers, providing a more universal gene definition that can be cross referenced with other databases. Unfortunately, the recent shift of NCBI from LocusLink to the Entrez Gene database format [] has limited the functionality of these scripts and rendered them obsolete. The DRAGON [] database [,], and the DAVID [] software [] provide web based services of gene annotation with similar objectives. None of them, however, allows for chromosome or genomic-region specific comparisons of gene coverage.The objective of this study was to compare gene coverage from currently available whole mouse genome microarrays for any region of the genome. We only used the information about probes provided to researchers by vendors before the purchase of a microarray for the purpose of choosing the platform that best fits their needs. We have developed a platform for microarray annotation that not only provides gene annotations for probes but also genomic positions for tested genes in the mouse genome. Coverage comparisons can be obtained for any genomic region of the last available mouse assembly build. The level of coverage of five whole mouse genome microarrays and four oligosets was compared in the present study. Microarrays and oligosets will be referenced here by the short name provided in Table . The results were stored in a relational database that can readily be queried for coverage comparisons based on genome position. Figure shows a flowchart diagram for the databases and methods used for the annotation system and for querying gene coverage comparisons.ResultsNumber of genes in the genomeThe total number of genes in the genome was defined as the number of Entrez Genes with a unique genomic position at the UCSC Genome Browser Database [] (see methods). A total of , mapped sequences from the Known Gene, RefGene, and mRNA tracks were associated with Entrez Genes. A total of genes could not be used because they are located in unordered scaffolds in Build . of the mouse genome assembly. Multiple sequence alignments in the genomes were found for a total of , sequences. For example, the M10062 cDNA aligns with chromosomes , , , , , , , , , , , X, and Un_random (not chromosome assigned contigs) chromosomes. This cDNA is identified as the Iap gene in the Entrez Gene database. This is a retrotransposon that can be found in several chromosomes and does not have a unique position. Genes like this, and other not so extreme cases, cannot be considered in region-specific coverage comparisons and were therefore discarded. Table shows the number of genes that could be assigned to specific position in the genome in each of the source files. The file Mm.gb_cid_lid in Table was used to incorporate associations between sequences and Unigene ids in the genexref table. Although we do not use Unigene annotations of probes to identify the targeted genes (see discussion section for explanation) this allowed us to associate Entrez Gene ids with EST accession numbers, which are commonly used in microarray genes lists. A total of , genes could be found having a unique position in the August mouse genome assembly (Build .). The distribution of these genes in the genome is shown in Figure .Microarrays probe annotationsThe Resourcerer database provides gene annotations for most of the available microarrays and oligosets for mouse and other organisms []. However, database updates are done at four-month intervals, and in our experience, gene annotations change periodically and many Entrez Gene ids in the Resourcerer annotations are obsolete. Microarray vendors also provide gene annotations on their probes, but they vary greatly in the number and quality of the annotations. For instance, ABI is the platform with the largest number of probes annotated with a gene id by the vendor (Table ). However, these are Celera Genomics® gene identifications and cannot be directly matched to public domain genes. Therefore, we opted for performing our own gene annotation of probes using the most updated information at hand (see methods). Our genexref table, in the ArrayGene database, stored cross reference information between , sequence identifiers and , Entrez Genes. The different kinds of sequence identifiers included in the database are shown in Figure . Performing in-house annotations allowed us to discard any probe that could be associated with more than one gene. Probe annotation files, referred to as Genelists, were obtained directly from vendors' websites and they identify the sequences from which oligonucleotide probes where designed. The format and amount of information provided in these lists varied greatly, Affymetrix being the most comprehensive in the number of different annotations. The efficiency of the gene annotation process varied between platforms depending on the amount of sequence identifiers provided by vendors and the level of specificity of the information provided (Figure ). Specificity is defined here as the number of associations that can be inferred between a probe and Entrez Genes from all the probe annotations provided by the vendor. The gene annotations from the ArrayGene system do not include any probe that could be associated with more than one gene nor genes with an uncertain position in the genome. This approach created a more conservative set of gene annotations than those included in the Resourcerer database. In most of the platforms we could not match the number of probes annotated with gene ids by the vendor (Entrez Gene ids, gene symbols, UniGene ids, etc) given our stringent criteria to select a unique gene identifier ("Unknown seq id" in Figure ). The level of redundancy, i.e. the number of probes hybridizing to the same gene, also varied between platforms (Figure ), the least redundant being the Sigma platform (. probes per gene on average), though it has the least number of probes. However, the newer ABI platform has a comparable level of redundancy (. probes per gene). The most redundant platforms is the Affy array with an average . probe sets per gene.Gene coverage from mouse whole genome microarrays and oligonucleotide setsGene coverage was estimated as the proportion of Entrez Genes with a unique position in the UCSC Genome database that is tested by a given microarray platform. The genome wide coverage varied from different platforms, ranging from .% to .% (Table and Figure ). The lowest coverage was observed, as expected, for the oldest platform, the Sigma oligoset, with a total of , probes testing , genes. Agilent and Codelink showed very similar coverage levels (.% and .%, respectively). Sentrix is the ready-to-use mouse microarray with the highest gene coverage, with .% of the publicly available genes tested. This was followed by the public oligonucleotide set, MEEBO (.%), that Sentrix was based on. The Operon AROS Arrays Oligosets showed clear improvement in gene coverage levels as new releases of their oligo data set have become available. The latest release (Operon4) shows .% coverage, higher than Operon3 which only covered .% and even Agilent and Affy (.% and75.%, respectively).Gene coverage by chromosomeThe differences between platforms in terms of gene coverage are well conserved across chromosomes (Figure ). Some changes, though, can be observed in specific cases. For instance, the Affymetrix platform has a particularly low coverage for mouse chromosome and (.% and .%), being outperformed even by the older Operon3 oligoset (.% and .%, respectively). For the complete list of gene coverage by chromosome per platform see [Additional file ] of the supplementary material. The database can be easily queried for gene coverage comparison on any region of any chromosome. Figure shows an example output report for a . MB region of mouse chromosome .DiscussionThe use of microarrays has recently been extended to the study of natural genetic variation affecting the expression of genes at the transcript level. Such techniques, originally coined as Genetical Genomics [], treats gene expression as a quantitative trait suitable for QTL analysis. Using this approach, several groups have been able to detect loci affecting the expression of thousands of genes, both in cis and trans, in yeast [,], mouse [], and humans [,]. Furthermore, Schadt et al. [] were able to identify cis-acting QTL for gene expression (eQTL) causing obesity in mouse. However, the success of this approach depends heavily on the level of gene coverage from the microarray platform being used. Previous QTL mapping knowledge can provide candidates regions for cis-eQTL that researchers would like to exhaustively test with a microarray platform. This is of prime importance when microarrays are used to compare congenic lines with background strains and the only difference between individuals is a small chromosomal region [-]. A priori knowledge of the level of gene coverage in the congenic region is essential to assess the significance of the results from such studies, creating the need for chromosome and region specific gene coverage comparisons between microarray platforms. Consistent gene identifiers are needed that can be mapped to specific locations in the genome. We have created a system to perform in-house gene annotations on sequences from a number of different sources. This system is centered on our ArrayGene database which maintains an extensive set of cross references between sequence identifiers and Entrez Gene ids. This system was used to annotate the sequences mapped by the UCSC Genome Browser Database with gene ids and created a gene-centered database called Aligndb. These databases, and a set of Perl scripts and modules, provide a platform for annotating and maintaining updated gene annotations of microarrays. We have annotated and compared gene coverage from five mouse oligonucleotide microarrays and four oligosets used for microarray spotting. Only genes that could be uniquely mapped to a single position in the genome with the UCSC Genome Browser database were included in the comparison. Genome coverage was then estimated as the proportion of such genes that is tested by a given platform. None of the platforms tested provided % coverage of the mouse genome, and their level of coverage depended greatly on the date of release. Newer sets have better coverage, most likely due to a better state of genome assembly and annotation at the time of design. It should be noted that nonetheless the highest coverage was found in the newest commercial microarray available, Sentrix by Illumina, it was followed by a non-commercial oligoset that was the basis of the Sentrix platform, called MEEBO (Exonic Evidence Based Oligonucleotide). This oligoset was developed by a collaborative effort between researchers at UCSF, Stanford, Rockefeller, Basel, and the Stowers Institute and it was based on an early draft of the genome, NCBI Build (Jan ), which is previous to the genomic assemblies used by most of the arrays tested here (Table S4 for details).Significant differences were also found in the level of redundancy (Figure ) and the amount of information that vendors provide for their platforms. The latter is a critical point at the time of data analysis if biological inferences are to be made from expression data. For instance, not all probes in the MEEBO array are annotated. The authors only provide an accession number for .% of non-control probes. They do provide a gene symbol for % of them, but the user has no means of finding the gene from sequence-specific annotations alone for .% of the non-control probes. However, the case is different for Operon4 array where although % of probes are annotated with at least one sequence identifier, many of them (,) could not be associated with any Entrez Gene, representing .% of the not annotated probes by ArrayGene for this platform. This is also the case for , probes in the Codelink array with no Entrez Gene association (Unknown seq id probes in Figure ). Whether this is a consequence of using obsolete accession numbers that are not included in current development of Entrez Gene or Unigene, or if it reflects big omissions in the curation of Entrez Gene, this should not affect the comparison in the present study. We expect that by imposing the same restrictions and conditions to all platforms our comparison can reflect real differences in gene coverage level. However we are aware that older platforms may be penalized because of the use of old sequences or ESTs that are not included in the Entrez Gene database and RefSeqs that are obsolete. It also must be noted that by using gene annotations such as gene symbols or Unigene identifications provided by vendors, we assume that the annotation properly associates the probe with the source sequence. This not only can add errors to the gene annotation process, but can increase the percentage of unspecific-probe calls since gene symbols sometimes can be associated with more than one gene. However, in some platforms, gene symbols were the only probe annotation available and not using them would restrict the inclusion of those platforms in the present study. Therefore, we opted for using gene symbols as probe annotations but not Unigene ids since these are automated sequence clusters with no human curation, are not stable and could lead to errors, especially overestimation of unspecific probes if the UniGene build used at the time of generation of the gene list is different from that used to build the genexref table of ArrayGene.Although we believe that our experimental design allowed for a fair and comprehensive comparison of gene coverage between platforms, this work also led us to identifying obstacles that researchers will encounter when trying to answer the questions of how many genes are being tested in a given platform and what is its level of coverage for a given region. The problems that we have encountered to perform optimal gene annotations of probes are associated with the lack of complete gene lists with essential information for the probes. Furthermore, accession numbers may become obsolete and gene names can change with time. The only data that is completely uniform and stable is the sequence of the probe. Sequences would allow any person to perform an alignment with the genome and identify the position being tested to determine what are the genes been tested. The accession number of the original sequence from which the probe was designed can also provide valuable information for the user. Unfortunately, these are not always readily available and for this reason we could not use them in this study. For proprietary reasons vendors very often provide only a consensus or representative sequence identifier which does not necessarily correspond to the original sequences used for probe design. Other probe annotations such as RefSeq accession, gene symbol, gene ontology classification, pathway, etc, can also be very useful for any potential user of the platform. However, from this work we see the need for standardization of a minimal set of information that vendors should provide along with the microarray itself. This should include: () accession number or other appropriate identifier for the original sequence used to design the probe; () database source of the sequence; and () type of probe (test vs. control). The position of the probe within the source sequence or the probe sequence is also invaluable information that, for example, would allow the user to assess possible hybridization artifacts due to polymorphisms between the target RNA and probe. Additional information such as RefSeqs id, gene name, gene symbol, and genomic position of the probe can also be very useful, helping the user to avoid performing genome-wide error prone annotations. These higher order annotations can become obsolete with the release of new genomic information, and constant updates would be needed for vendors to remain current. Some microarray vendors have already taken this approach and we believe that this will provide them with a marketing advantage, even if the features and quality of the microarrays themselves are comparably similar.ConclusionThe results of this study provide researchers with a comparison of the level of gene coverage of different microarrays platforms for the mouse genome, which is particularly useful for genetical genomic studies. The Sentrix microarray by Illumina and the MEEBO public oligo set that this array is based on provide the highest gene coverage of the mouse genome. The bioinformatics tools generated here can prove useful for microarray users studying other organisms with a sequenced genome. The annotation tools included in the ArrayGene software are flexible enough to be easily adapted to use input files from any database that associates genes with sequence ids. We also provide comments on the limitations from current microarray annotations to encourage the scientific community to develop minimum annotation guidelines that designers and vendors of microarrays can follow to better fulfill research needs.MethodsGene definitionThe Entrez Gene database at NCBI was used as the reference to assign consistent gene annotations across platforms. The Entrez Gene database groups GenBank and RefSeq accession numbers and gene symbols by unique, non redundant and traceable gene ids []. These ids could also be cross-referenced to ENSEMBL transcripts and genes, and to any id that can be related to a GenBank accession number. Tables associating gene ids to GenBank and RefSeq sequences were obtained from the NCBI ftp server (files downloaded on December 15th ). A third table associating UCSC Known Genes to ENSEMBL transcripts was obtained from the UCSC ftp server (used files and source URLs in Table ). A single table called genexref was created in the ArrayGene MySQL database, with three columns: gene id (from Entrez), sequence id, and type of sequence (i.e. Ensembl, genomic, mRNA, protein, RefSeq, symbol, synonym, NIA transcript, or Unigene). This table is the core of the gene annotation process of probes in microarrays. Only current Entrez Gene ids were used. The file gene_history from NCBI was used to delete obsolete genes gene ids from the database.Genomic location for genesThe physical location of genes was defined as the position where any sequence associated with the gene could be aligned with confidence by sequence comparison. Genomic alignment results were obtained from the UCSC Genome Browser Database [] for the August mouse genome assembly (Build .). Text files from this server contain the results from aligning all mouse mRNA sequences deposited at GenBank against the mouse genome sequence using BLAT. All sequence alignments in the UCSC database have at least % identity. The track of "Known Genes" in the UCSC Genome Browser provides genomic coordinates only for mRNA that could be associated with a protein in SWISS-PROT, TrEMBL, or TrEMBL-NEW. Similarly, the track called RefSeq Gene contains codon and intron positions for RefSeq sequences. UCSC tracks were used in a hierarchical order: ) Known Genes, ) RefSeq Genes, and ) mRNA sequences. Genes mapping to unordered scaffolds or to multiple positions from BLAT alignments were not considered. A table labeled genemap was created in MySQL to store the best, if any, mapping information for every gene following the above criteria. The table stores the genomic coordinates from a single sequence per gene. As a result of importing these tracks in hierarchical order, coordinates from known Genes are preferred over RefSeqs, and these are over mRNA ones. The genemap table is updated with every new release of the mouse alignment from NCBI. The table is stored in an alignment-specific database herein called Aligndb. The actual name of this database is provided by the user when a new genome alignment is available, and the current database is called mm7, in reference to the name provided by the UCSC genome browser to the annotated Build . assembly. The proportional contribution from each UCSC track to the genemap table is shown in Table . This table shows the number of genes in each track that could not be used because they mapped to multiple positions in the genome or to unordered scaffolds. The last column of the table shows the number of genes for each track that could be mapped to a unique known position in the genome and that were not already present in the database (i.e. were not included in any track previously imported).Database architecture and maintenanceTwo MySQL databases were created and are maintained separately. The ArrayGene database contains the genexref table which provides cross references between sequence identifiers and Entrez Genes and the arrays_table table storing all the information about microarray annotations (Figure ). A second database is created every time a new mouse genome build is released. This database contains the genemap table providing mapping positions for genes probed in microarrays. The ArrayGene package was created to build and maintain both databases using command line and web forms and is included as [Additional file ] in this report. Future new versions of the software will be available from the authors' website []. This system is a generic tool for the comparison of gene coverage providing a user interface to generate reports and can be used for any species with a sequenced genome and a database of genes associated with sequences, which are used to produce probes in microarrays. The input of files for both gene information and gene mapping are format independent and the programs can be customized through command line options to use any data file that is in tabular form. The software is distributed along with a user manual under the General Public License V. [] and is freely available for the research community. No genomic or microarray data is included in the package and the user must follow the instructions included with the software to install and populate the database. The software depends on MySQL v. .. or later versions, Perl (only tested on v. ..) and on some Perl modules described in installation instructions of the User Guide included as [Additional file ].Gene annotation of microarray probesThe import_array Perl script is designed to extract the probe annotations provided by vendors for their platforms in files commonly called gene lists. The script can parse a text file, extract probe and sequence ids (even from fasta-style description lines), connect to the ArrayGene database and find the Entrez Gene id for each probe. Finally, import_array can either write an output file with genomic annotations for probes or create a table in the database with this information. When multiple sequences are associated with a given probe in the gene list the Perl script checks if they all match the same gene. The program can look for sequence ids in columns maximum, and it can detect inconsistencies between them if they point to different Entrez Gene ids in the database. It also detects single sequences that are associated with more than one gene in the Entrez Gene database. In any of these two cases, since a unique gene cannot be associated with a probe it is annotated as a cross hybridizing probe. The ids of all genes associated with that probe are stored but are not used in genome coverage comparisons. Reports about gene coverage in annotated microarrays are done by a series of CGI scripts providing an intuitive web interface to the database.Microarray platforms and oligonucleotide setsThis study compared the gene coverage of mouse whole genome microarray platforms that are currently available to investigators. One older oligoset was also included to evaluate coverage improvements with time of release ([Additional file ] in supplementary material provides time of release for some platforms, URL and filename for the Genelists used here). We compared four commercial one-color arrays, Affymetrix [] Mouse Genome . Array, Amersham [] Codelink Mouse Whole Genome, Sentrix Mouse- Expression BeadChip from Illumina [], and Applied Biosystems [] Mouse Genome Survey. We also compared the two-color Agilent [] Mouse Oligo Microarray Kit; the commercial oligoset Operon [] Array-Ready Oligo Set V. and one previous version (Operon V. ); the Sigma-Genosys Mouse Oligonucleotide library (available through Lab on Web []), and the Mouse Exonic Evidence Based Oligonucleotide (MEEBO) [] produced by a group of investigators at UCSF, Stanford, Rockefeller, Basel, and the Stowers Institute. Table lists all these platforms, indicating their full and short name used throughout this paper. A special mention should be made about the ABI microarray which contains almost , proprietary sequences, which are not in the public domain. The probes in this platform were designed based on the Celera Mouse Genome Alignment (Celera, Rockville, MD), which contains gene annotations based in proprietary methods. The present study compared gene coverage by using sequence annotations equivalent to public accession numbers available for , probes. However, these probes may target genes without an exact counterpart in the public domain given large methodological differences that exist for defining a gene between the Celera and the public domain approaches.Gene lists for every platform included in this study, except for ABI, were downloaded from the vendors' websites. ABI's gene-list was obtained directly from the vendor.Authors' contributionsRAV collected information from public databases and created the ArrayGene software and database and drafted the manuscript. JFM supervised the design and development of the database and guided the gene coverage comparison between microarrays platforms and helped draft the manuscript. Both authors read and approved the final manuscript.Supplementary MaterialAdditional File 1Microarray platforms compared in this study. Table provides details of the Genelists used, including filename, URL, date of release, date updated, and date obtained. The column called annot update lists the date that the Genelist available for the platform was last updated by the vendorClick here for fileAdditional File 2Comparative gene coverage from whole mouse genome microarrays and oligo set. Table shows the number of Entrez Genes with a single genomic position in the genome by the UCSC Genome Browser Database, and the number of genes that are tested by each platform as absolute counts and as percentage from the number of genes in the chromosome.Click here for fileAdditional File 4ArrayGene software with an install script for UNIX platforms and a README file for installation and usage instructionsClick here for fileAdditional File 3Reference manual for the installation and use of the ArrayGene software. The manual includes a list of system requirements and the full list of options for all the tools included in the package.Click here for file
|
PMC2008431.txt
|
TITLE: A Study of the Immunosuppressive Activity of Methylene Dimethane Sulphonate (MDMS) in Relation to its Effectiveness as an Anti-Tumour Agent AUTHORS: B. W. Fox, Connie J. Gregory ABSTRACT: The anti-tumour action of methylene dimethane sulphonate (MDMS) has been further investigated in relation to its immunosuppressive properties. Following a dose of mg/kg, the proportion of permanent regressions of Yoshida lymphosarcoma transplants is lower in animals treated during the first days of tumour growth. Re-implants on day to those animals in which regression of the tumour had occurred indicated that the immune response to the tumour increases during the first days of tumour growth.Studies of the effect of MDMS on the primary antibody-forming cell response of mice to sheep red cell antigens showed this drug to be an immunosuppressant comparable in strength to x-radiation. MDMS given to rats prior to tumour transplantation also acted as an immunosuppressant in this system resulting in an increased rate of tumour growth. For both responses the maximum immunosuppressive effect was obtained when the interval between drug administration and antigenic challenge was minimal. BODY:
|
PMC3038535.txt
|
TITLE: Dementia, a special supplement AUTHORS: T. S. Sathyanarayana Rao, K. S. Shaji ABSTRACT: No Abstract BODY: Dementia is fast emerging as a huge public health challenge in the developing world. We need to prepare ourselves to face this crisis. We must know more about various aspects of dementia and its management. Better understanding of dementia in developing countries is also likely to provide valuable new information on risk factors, genetics, and potential new treatments. That is why we chose “Dementia” as the theme.We are pleased to present a supplement of Indian Journal of Psychiatry (IJP) on Dementia. Dementia being a neglected area in our country - both concerning the research and service, there are many invited articles from abroad. We hope this will act as an impetus to further the cause of dementia. A special issue on a specific topic and on this scale is first in the history of IJP. We have tried to get articles from many people who have contributed to the knowledge base. Most of them responded wholeheartedly and well in time. We are delighted to acknowledge their help and support. We remain deeply indebted to them.We look forward to your feed back. Please send in your comments, criticisms and suggestions.Yours sincerelyT. S. S. Rao, Editor, Indian Journal of PsychiatryK. S. Shaji, Guest Editor, Indian Journal of Psychiatry supplement on Dementia
|
PMC1574296.txt
|
TITLE: Complications after endovascular stent-grafting of thoracic aortic diseases AUTHORS: Gabriele Piffaretti, Matteo Tozzi, Chiara Lomazzi, Nicola Rivolta, Roberto Caronno, Patrizio Castelli ABSTRACT: BackgroundTo update our experience with thoracic aortic stent-graft treatment over a -year period, with special consideration for the occurrence and management of complications.MethodsFrom December to June , patients with thoracic aortic pathologies underwent endovascular repair; there were males (%) and females, mean age ± years (range –). Fourteen patients (%) were treated for degenerative thoracic aortic aneurysm, patients (%) for penetrating aortic ulcer, patients (%) for blunt traumatic injury, patients (%) for acute type B dissection, patients (%) for a type B dissecting aneurysm; patients (%) with thoraco-abdominal aortic aneurysms were excluded from the analyses. Fifteen patients (%) underwent emergency treatment. Overall, mean EuroSCORE was ± (median , range –). All procedures were performed in the theatre under general anesthesia. All complications occurring during hospitalisation were recorded. Follow-up protocol featured CT-A, and chest X-rays , and months after intervention, and annually thereafter.ResultsPrimary technical success was achieved in all patients; procedures never aborted because of access difficulty. Conversion to standard open repair was never required. Mean duration of the procedure was ± minutes (median , range –). Mean blood loss was mL (range – mL). The mean length of the aorta covered by the SGs was ± mm (range –). The LSA was over-stented in cases (/, %). Overall -day operative mortality was .% (/). Major complications included pneumonia (n = ), cerebrovascular accidents (n = ), arrhythmia (n = ), acute renal failure (n = ), and colic ischemia (n = ). Overall, endoleak rate was %.ConclusionAlthough this report is a retrospective and not comparative analysis of thoracic aortic repair, the combined minor and major morbidity rate was lower than previous reported to results of either electively and emergency performed conventional repair. BODY: BackgroundEndovascular treatment for aortic disease has emerged as an alternative mode of treatment for thoracic aortic diseases (TADs) that is particularly attractive for patients with severe co-morbidities who would not be ideal candidates for open surgery [-]. Because initial results have been encouraging, this new therapy has been used more liberally, and short-term results are well documented [,,] However, despite early promising clinical results, effect on the long-term durability of the aortic repair have not been yet established, they are associated with complications, as are all surgical methods [-].Here, we present our experience with TAA stent-graft (SG) treatment, with special consideration for the occurrence and management of complications.Materials and methodsFrom December to June , patients with thoracic aortic pathologies underwent endovascular repair; there were males (%) and females, mean age ± years (range –). Fourteen patients (%) were treated for degenerative thoracic aortic aneurysm, patients (%) for penetrating aortic ulcer, patients (%) for blunt traumatic injury, patients (%) for acute type B dissection, patients (%) for a type B dissecting aneurysm. Five additional patients who received hybrid treatment (visceral revascularization and endovascular thoracic SG) for thoraco-abdominal aortic aneurysms were excluded from this series. Hence, patients wree considered for the retrospective analyses. Patient data including procedure details, postoperative complications, secondary procedures, and mortality were collected. Demographics and co-morbidities are listed in Table .Table 1Demographic data and co-morbiditiesdiseasenr%hypertension3472obesity (BMI > )2145COPD1940IHD1328CRF817previous aortic surgery715hypercholesterolemia510previous cardiovascular surgery510CVD48diabetes mellitus48BMI = body mass index; COPD = chronic obstructive pulmonary disease; IHD = ischemic heart disease CRF = chronic renal failure; CVD = cerebrovascular diseaseFifteen patients (%) underwent emergency treatment because of blunt traumatic injury (n = ), complicated dissection (n = ), penetrating ulcer (n = ), or ruptured degenerative aneurysm. Eight patients (%) had infrarenal aortic aneurysm repair in their medical history.The Ishimaru classification were used to define the location of the proximal neck of the lesion: "zone " involves the innominate artery, "zone " the origin of the left common carotid artery, "zone " the origin of the left subclavian artery, "zone " extends into the descending thoracic aorta distally to the origin of the left subclavian artery, "zone " involved the middle to the lower third of the descending aorta: according to the Ishimaru classification, the aortic pathology was located in the aortic arch (zone –) in patients, in proximal/mid-descending aorta (zone ) in patients, and in the distal descending aorta (zone ) in patients. Mean thoracic aortic aneurysm diameter was . ± cm (range .–), mean dissecting aneurysm diameter was . ± cm (range –), and mean ulcer diameter was . ± . cm (range .–.).Spiral CT or CT-angiography (CT-A) of the entire thoracic-abdominal aorta was performed preoperatively in all patients to determine the anatomical feasibility for the endovascular procedure, aneurysm size, location, extension, and relationship between the aneurysm and significant aortic branches such as the left subclavian artery (LSA) and the visceral vessels (e.g celiac trunk, superior mesenteric artery, renal arteries) (Figure ). When elective surgery was scheduled, all patients underwent spirometry, echocardiography; a cardiologist was consulted to evaluate eventual either pharmacologic stress testing or coronary angiography. CT-A of the epiaortic trunks and intracranial vessels were carried out in order to detect any abnormalities (e.g. agenesia), stenosis, patency of the anterior spinal artery, or previous ischemic areas.Figure 1Preoperative CT-A shows an -cm saccular aneurysm of the distal arch (A, B); MIP reconstructions shows the involvement of the LSA (C).All patients (or their relatives) were informed in detail about the endovascular operation and gave written consent. All procedures were performed in the theatre equipped with a movable radiolucent surgical table combined with a digital angiographic system using a ceiling-suspended C-arm (Isocentric mobile C-arm-Siemens®; Munich-Germany), under general anesthesia by a cardiovascular surgeon in close collaboration with an interventional radiologist.All patients received short-term prophylactic antibiotic therapy (generally, vancomycin gr b.i.d.) and heparinization intraoperatively. All patients were prepared and draped also for potential conversion to thoracotomy if it became necessary. The patients were initially positioned on their backs with the left shoulder slightly raised. Cardiopulmonary bypass (CPB) and a cell-saver (Compact-Dideco®-Modena, Italy) were eventually available. Prophylactic use of cerebrospinal fluid (CSF) drainage to prevent spinal cord ischemia was not used. Surgical exposure included the common femoral artery in cases [monolaterally (n = ) and bilaterally (n = )], an extra-peritoneal approach for the external iliac artery (n = ), a -mm Dacron conduit end-to-side onto the common iliac artery (n = ), and transperitoneal approach (n = ) during an explorative laparptomy to repair associated multiple liver fractures in a polytraumatized patients who sustained traumatic aortic injury. Once we gained the remote access, administration of IU heparin (Epsoclar®-Biologici Italia Laboratories; Novate Milanese-IT) was followed by insertion of the SG system. Calibrated angiographic pigtail catheter (Cook Inc.®-Bloomington; IN-USA) was placed in the ascending aorta, followed by a digital subtraction angiography (DSA) to confirm the beginning of the aneurysm, to mark the LSA (e.g. a 5F catheter was preliminary advanced through a percutaneous left brachial artery access), and to determine the final landing zone. Devices were then positioned with confirmatory DSA before deployment, to enable precise positioning. Dimensions of SG were determined on the basis of contrast-enhanced spiral CT images and angiographic images, and were oversized by % to %, depending on the type of pathology; if multiple SGs wee needed, implantation began with the distal device first and then continued cephalad to the proximal end of the thoracic lesion. A complete overlap of at least cm was used, with more overlap for severely angulated segments. Just before the device was released, systolic arterial blood pressure was lowered by nitro-prusside or B-blocker. Once deployed, SG were assessed with a combination of DSA and TEE to identify the presence of endoleak, evaluate the degree of device apposition, and LSA revascularization from the ipsilateral vertebral artery (Figure ). An inflatable balloon was eventually used to model the SG to the aortic wall.Figure 2Intraoperative preliminary DSA (A) confirm the presence of a "lusoria artery" (arrow). Final DSA (B) confirmed the complete exclusion of the aneurysm and the patency of the left common carotid artery (black arrow) and the re-filling of the LSA via the ipsilateral LVA (white arrow).Several types of SG were used: Excluder/TAG [W.L. Gore & ass.®-Flagstaff, AZ-USA (n = )], TX-/TX- [Cook Inc®-Bloomington, IN-USA (n = )], or Talent/Valiant [Medtronic Inc®-Minneapolis, MN-USA (n = )] straight tube devices. Simultaneous endovascular procedures included AAA exclusion (n = ), axillary artery stenting (n = ): in the former patients, both the thoracic and abdominal lesions were successfully excluded. Abdominal aortic SGs used were as following: Zenith [Cook Inc®-Bloomington, IN-USA (n = )] or Excluder [W.L. Gore & ass.®-Flagstaff, AZ-USA (n = )].Patients were not routinely admitted to the surgical intensive care unit postoperatively. A complete CT scan was performed before the discharge in all patients to confirm complete sealing of the aneurysm and identify any endoleaks. A clinical neurological investigation was carried out in case of neurological complications, with additional diagnostics imaging if required.Technical success rate was defined as successful deployment of the SG without perigraft endoleak. Endoleak was classified as primary (diagnosed within days of endovascular repair) or secondary (diagnosed more than days after intervention) (Figure ). All complications occurring while the patient remained in the hospital or later, were recorded. A >% rise in renal parameters was considered indication of renal complications. Mortality rates were calculated at three intervals: during the in-hospital stay (perioperative), -days after the operation (early), and in follow-up (late).Figure 3Follow-up CT-A showed late type Ib endoleak (arrow) after SG treatment of a type B dissecting aneurysm (A). Preliminary DSA (B) confirmed the re-perfusion of the false channel through a distal re-entry tear (arrow): additional SG was used to seal the aneurysm as confirmed by final DSA (C) with TEE control.The follow-up protocol featured spiral-CT or CT-A, plus conventional -view chest X-rays to detect changes in SG configuration at , and months after intervention, and then annually. All films were reviewed by both an attending radiologist and an attending surgeon. During the follow-up, DSA control was performed in patients with persistent endoleak, stent migration, or aneurysm sac enlargement.ResultsIn all cases, the SG was placed successfully; neither procedures aborted because of access difficulty, nor conversion to standard open repair was necessary. The median delay of the emergency procedure for traumatic injury from the original accident was ± minutes (range –), whereas the overall median delay for the urgent procedures was ± minutes (range –). Overall, the mean duration of the procedure was ± minutes (range –). Median blood loss was mL (range – mL). The average total amount of contrast was ± mL (range –). A total of SGs were implanted: patients received a single SG, patients two SGs, and patients three SGs. The mean length of the aorta covered by the SGs was ± mm (range –), whereas mean SG diameter was ± mm (range –). The LSA was over-stented in cases.Mortalityoverall -day operative mortality was .%, .% in those undergoing elective treatment and .% in those undergoing emergency treatment. In the elective group a cause of death was an hemorrhagic shock due to a secondary rupture owing to an undetected endoleak, days after the procedure. In the emergency patients, we experienced hemorrhagic shock (n = ) due to rupture of a complicated type B dissection, and a sudden irreversible cardiac failure (n = ). During follow-up (mean months, range –), two further patients died. The causes of death were acute myocardial infarction (n = ), and acute respiratory failure (n = ); mean interval was .-months (range -days--months). Overall, at , and years, survival for elective and emergency cases were .%, .% and %, respectively.Neurological sequelaethere was no paraparesis or paraplegia due to ischemia of the spinal cord. Four patients (.%) developed cerebrovascular events after the operation: we observed cases of transient ischemic events (TIA); and strokes that occurred days after the intervention in a -years old male patient with a zone type B dissecting aneurysm and was caused by recurrent peaks of hypertension that did not respond to pharmacological therapy, and days after the intervention in a -tears male patient with a zone aortic arch aneurysm. Patients with TIA did not developed definitive neurological deficits.Endoleaksoverall, endoleak rate was %. Conversion to open surgery was never required. Early endoleaks were detected before discharge in patients: one type 1a endoleak and type endoleak from the LSA. The last one was successfully excluded by coils embolization at the origin of the LSA. One type 1a endoleak developed after caudal migration of the SG, leaded to secondary rupture and finally death. Late endoleaks occurred in patients. We observed type 1b endoleak that was treated by additional SG on postoperative month , and two type leaks from the LSA (n = ) that was sealed by coils embolization, and from an intercostal artery (n = ) that is still under survey with no evidence of sac growth. New onset of endoleaks, at later follow up, were not observed.Overstented supra-aortic arteriesthe LSA was fully covered in patients (%). None of the patients developed an upper arm ischemia. In cases in which the origin of the left common carotid artery (LCA) was covered with a SG, a carotid-carotid PTFE bypass was performed before stent-grafting. Preventive bypass procedure to the LSA was performed in one case only in a -years old male patient with a zone arch aneurysm; the preoperative CT-A revealed the occlusion of the right vertebral artery, hence being at high risk of perioperative cerebrovascular accident we carried out an extra-anatomical bypass to the LSA in addition to the carotid-carotid bypass, before the SG procedure. Unfortunately, the patient developed a major stroke (left part plus posterior) two days after the intervention. In the remainder cases, left upper arm ischemia did not develop, and retrograde flow through the ipsilateral vertebral artery (LVA) was demonstrated by final DSA and postoperative duplex examination. Retrograde reflux leaks from the subclavian artery into the aorta occurred in cases and were sealed by coils.Other complicationsoverall, the most common complication was pneumonia (n = ), followed by arrhythmia [atrial fibrillation (n = ) and atrial flutter (n = ]). Six patients (%) developed renal impairment [(early (n = ), late (n = )]. Temporary hemofiltration or dialysis was necessary in patients. Two patients demonstrated new, clinically silent splenic infarctions on the postoperative CT but did not develop abscess or subsequent infectious complications; both lesions resolvedcompletely. Colic ischemia developed in a -year male patient on day twenty-seventh, postoperatively and required an Hartman intervention: he did well and was discharged without further complication. Among minor complications, the most frequent was post-implant syndrome (PIS) that developed in patients (%). No patients had retrograde acute ascending thoracic dissections. Thoracic lesion shrinkage, defined as a sac reduction of mm at least after -months CT-A follow-up control, was observed in patients (%).DiscussionMortality is the most important criterion in appraising the results of TAA treatment. Conventional intervention carries a mortality rate that ranges between % and %, up to % for emergency repair [,]. In their largest series, Stanford equipe reported a perioperative mortality of % []. The % thirty-day mortality we attained with endovascular repair, is similar to previous reports of procedure-related mortality in patients undergoing endovascular treatment of TAA and dissections and the benefits of the intraluminal approach are apparent [,,,]. Moreover, the % thirty-day mortality in our emergency cases are quite encouraging: considering that all the procedures were performed in high-risk patients, we believe that this early mortality rate is acceptable. It is interesting to note that patients with TAD who required emergent treatment had a procedure related mortality lower than reported, although most of the patients had active bleeding at presentation. This result suggest that, even among patients with the highest risk, endovascular repair can be undertaken with acceptable mortality. Given the high mortality in patients who undergo emergent open repair, these patients would seem to benefit most from endovascular treatment. Indeed, all the late deaths in our cohort were not related to the SG procedure; in order to reduce mortality and morbidity rates, careful selection criteria must be followed when treating patients endovascularly: using objective criteria could help in selecting patients for endovascular treatment [].Surgeons have often focused on death, bleeding, and paraplegia as the major adverse outcomes of thoracic aortic surgery [,,-,]. Stroke may not be adequately appreciated as a common sequel of descending aortic operations. Most clinical reports on aortic surgical procedures simply state an overall incidence of stroke without specific analysis of this complication [,]. In our series, stroke was found to be disturbingly frequent after aortic surgery. In multivariable analysis, a number of risk factors were found to be predictive of stroke for open surgery: these included a history of diabetes mellitus (possibly caused by chronic damage to the cerebral microvasculature), and emergency surgery; in our experience, all but one cerebrovascular accidents were embolic-technically related, and appeared to be caused by the manipulation of wires and catheters in an atherosclerotic aortic arch []. Thus, meticulous technique must be used during device deployment to avoid atheroembolic cerebrovascular events.Spinal cord ischemia and postoperative neurologic deficit with surgical repair of degenerative aneurysms and thoraco-abdominal aneurysms are believed to be multifactorial in etiology. Their occurrence is influenced by the duration and severity of the ischemic insult, the neuronal metabolic rate during this ischemic period, the postoperative hypotension, and a subsequent reperfusion injury [,,,]. In-addition, the extent of the aneurysm involvement with the thoracic aorta has been shown to influence the expected rate of neurologic deficit after open thoracic aortic aneurysm and thoraco-abdominal repair []. Despite many improvements in the technique and adjuncts, the rates of paralysis in conventional open surgery range from .% to % in high-volume centers [,,-]. However, since a large number of intercostal arteries are acutely occluded even in endovascular repair, especially when long SG are implanted (a maximum of mm in our patients), paralysis can still occur. In our own patients, there has been no case of paralysis up to now, despite the average length of the "stented" thoracic aorta was mm, which means that to pairs of intercostal arteries were occluded at least. The aortic segment from T7 to T11, regarded as particularly sensitive, was covered with SG in of our patients.Because of the absence of aortic cross-clamping and reperfusion, endovascular SG repair of the descending thoracic aorta may lower the incidence of spinal cord ischemia. Those cases with TAA stent-grafting with concomitant open AAA repair also undergo a period of aortic cross-clamping, potentially causing an even more severe initial spinal cord ischemia as well as adding an element of reperfusion injury to the ischemic insult, being associated with an increased risk of paraplegia, since the absence of the normal lumbar or ilio-lumbar arteries sacrificed by AAA surgery may place these candidates at an increased risk for spinal cord ischemia [,-]. We noted no paraplegia or paraparesis, although the distal thoracic aorta was stented in eight patients with previous abdominal aortic surgery, and four patients had concomitant endovascular aortic multilevel disease.Although in the initial steps of previous published studies, endoleaks in the thorax seemed markedly less significant than of aneurysms of the infrarenal aorta, it soon became obvious that endoleaks after thoracic endovascular therapy are a serious complication [,,,,]. Degenerative disease and atherosclerotic aneurysm formation of the thoracic aorta are diffuse and progressive in character, often involving the entire thoracic aorta, but changes within the proximal and distal neck caused by disease progression have been reported causing type endoleak []. Neck dilatation after endovascular aneurysm repair has previously been recognized in patients by a group at Stanford University. In our series patients displayed neck dilatation []. Our policy in treating type Ib endoleaks was to implant a cuff at the endoleak site. These late complications show the necessity of mandatory life-long follow-up. Finally, disease progression proximally or distally, especially in patients with diffuse degenerative aneurysms, may result in slow enlargement of the landing zones, loss of fixation, and sac pressurization [,]. Isolated cases of secondary rupture after SG have been described in many reports. In all of these cases, technical errors of untreated primary or secondary endoleaks could be incriminated. In this series post-procedural rupture was observed in one patient. Multiple factors may have contributed to these late endoleaks. The association between action of vector forces in SG and the often severe tortuosity in the thoracic aorta might induce even stronger vector forces after aneuryms repair.Renal failure (RF) has been reported as one of the leading cause of postopretaive death following conventional descending aneurysm and thoraco-abdominal repair; also endovascular repair involved new potential risks specific to the procedure, which may be associated with a substantial rate of postoperative RF []: the incidence of long-term RF following endovascular aortic aneurysm exclusion is poorly documented in the literature; the incidence of postoperative RF ranged between .% and .% in previous reported series []. In the present study, transient perioperative RF occurred in % of patients; moreover, it has been noted a trend toward this being a more common event in those patients with pre-existing RF [,,]. We identified a % prevalence of preoperative RF in our own population undergoing endovascular repair. Our incidence of renal complications is similar than rates reported by others []. Most likely the etiology is multifactorial, resulting from a combination of mechanical, atheroembolic and contrast related contributors. In particular, the relationship between contrast type and volume and renal function has been established []. Contrast induced renal insufficiency is reversible, yet in our patients, the noted creatinine rise was permanent in only case of those patients who had deterioration of renal function: this suggests factors other than contrast exposure alone as contributory, suggesting atheroembolic sequelae. Strategies may be adopted to minimize the likelihood of adverse renal effects associated with aortic SG at all stages of evaluation and intervention. While our routine practice involves the use of non-ionic contrast agents in association with vigorous hydration and use of osmotic diuretics, we have greatly reduced our reliance upon angiography employing iodinated contrast agents in our patients with RF.ConclusionAlthough this report is a retrospective and not comparative analysis of thoracic aortic repair, the combined minor and major morbidity rate was lower than previous reported to results of either electively and emergency performed conventional repair.Despite the favorability of these findings, complications unique to endovascular repair of descending thoracic aortic lesions (eg, RF, stroke, leakage) did occur. These complications in part reflect the results of an evolving technology that involves a requisite learning curve and developmental issues. A reduction in these method-related complications can in all probability be achieved only by improved SG and optimal patient selection and matching.In addition to the usual concerns of death, bleeding, and paraplegia, surgeons must consider stroke heavily when deciding to operate on the thoracic aorta. Meticulous technique must be used during device deployment to avoid atheroembolic cerebrovascular events.Renal insufficiency is present in a large percentage of patients being considered for endovascular aneurysm repair. It should not be considered a contraindication to the evaluation or management of these patients provided that care is taken to pursue a strategy that minimizes the risk of renal injury.Competing interestsThe author(s) declare that they have no competing interests.Authors' contributionsGP: study design; data collection; performed the statistical analysis.MT: critical analysisCL: data collectionNR: critical analysisRC: critical analysisPC: critical and statistical analysesAll authors read and approved the final manuscript.
|
PMC2726965.txt
|
TITLE: Perioperative management in distal pancreatectomy: results of a survey in European participating centres of the DISPACT trial and a review of literature AUTHORS: Helge Bruns, Nuh N Rahbari, Thorsten Löffler, Markus K Diener, Christoph M Seiler, Matthias Glanemann, Giovanni Butturini, Christoph Schuhmacher, Inga Rossion, Markus W Büchler, Tido Junghans ABSTRACT: BackgroundConcomitant treatment in addition to intervention may influence the primary outcome, especially in complex interventions such as surgical trials. Evidence-based standards for perioperative care after distal pancreatectomy, however, have been rarely defined. This study's objective was therefore to identify and analyse the current basis of evidence for perioperative management in distal pancreatectomy.MethodsA standardised questionnaire was sent to European centres recruiting patients for a randomized controlled trial (RCT) on open distal pancreatectomy that would compare suture versus stapler closure of the pancreatic remnant (DISPACT trial, ISRCTN ). Perioperative strategies (e.g., bowel preparation, pain management, administration of antibiotics, abdominal incision, drainages, nasogastric tubes, somatostatin, mobilisation and feeding regimens) were assessed. Moreover, a systematic literature search in the Medline database was performed and retrieved meta-analyses and RCTs were reviewed.ResultsAll centres returned the questionnaire. Consensus for thoracic epidural catheters (TECs), pain treatment and transverse incisions was found, as well as strong consensus for the placement of intra-abdominal drainages and perioperative single-shot antibiotics. Also, there was consensus that bowel preparation, somatostatin application, postoperative nasogastric tubes and intravenous feeding might not be beneficial. The literature search identified meta-analyses and RCTs demonstrating that bowel preparation, somatostatin therapy and nasogastric tubes can be omitted. Early mobilisation, feeding and TECs seem to be beneficial for patients. The value of drainages remains unclear.ConclusionMost perioperative standards within the centres participating in the DISPACT trial are in accordance with current available evidence. The need for drainages requires further investigation.Trial registrationClinical trial registration: ISRCTN BODY: BackgroundShort-term outcomes after major abdominal surgery are influenced by indication, surgical intervention, the surgeons' expertise, hospital volume and perioperative management [-]. In order to qualitatively assess the impact of the various factors on outcome, a rigorous means of evaluation such as the randomized controlled trial (RCT) design is necessary. Mortality and morbidity in pancreatic surgery has decreased in high-volume centres in recent years [,] mainly by virtue of standardised surgical and optimised perioperative management []. In high volume centres, mortality from distal pancreatectomy is as low as % while morbidity remains as high as % [,-]. Morbidity is determined by perioperative factors such as bowel preparation, incision type, analgesia, mobilisation, and feeding regimens, amongst other factors []. In addition, long-term survival after pancreatectomy as cancer treatment strongly depends on a centre's operative volume [].A recent study of consecutive pancreatic resections demonstrated local complications to be more frequent than systemic complications []. Thus, the formation of pancreatic fistulas can be considered the most important complication after a pancreatic left resection. Currently, the Study Centre of the German Surgical Society is conducting a large RCT (DISPACT trial, ISRCTN ) to compare the occurrence rates of pancreatic fistula after stapler versus hand-sewn closure of the pancreas after left resections []. Both expert opinions and evidence-based standards were taken into account in the development of the study protocol. The consensus-assisted development of study protocols has proven to be beneficial and may indeed be helpful in identifying when consensus is present but not justified by evidence []. All centres contributing to DISPACT are centres with a substantial caseload of pancreatic left resection (Table ). The study protocol of DISPACT has been approved by the ethics committees of all participating centres and has been recently published []. Since the DISPACT trial aims to evaluate the effect of two surgical procedures (stapler and hand-sewn closure) on the development of pancreatic fistula, surgical and perioperative standardisation is of special interest. The present study was launched during an investigator meeting to identify the current practices concerning perioperative management amongst the various participating centres. This study aims to identify the level of consensus of relevant factors in perioperative management. In addition, an extensive literature search was conducted to compare findings with current best available evidence.Table 1Total pancreatectomies and distal pancreatic resections performed in at centres taking part in DISPACT TrialCentreNumber of pancreatic resectionsNumber of distal pancreatectomiesNumber of patients randomized in 2008Amsterdam, Netherlands76128Berlin Charité Mitte, Germany20124Berlin Charité Virchow, Germany1974820Berlin Lichtenberg, Germany38104Bochum St. Josef, Germany214467Dresden-Friedrichstadt, Germany5252Freiburg, Germany83152Gent, Belgium6863Heidelberg, Germany4238745Homburg, Germany6110-Cologne-Merheim, Germany2510-Liverpool, Great Britain128131Ljubljana, Slovenia902510Mannheim, Germany71104Marburg, Germany48131Munich-LMU, Germany911913Munich-TU, Germany872722Regensburg, Germany70196Verona, Italy2696617Würzburg, Germany2072Wuppertal, Germany2431Total2155463172Median (range) (–) (–) (–)MethodsSurvey of current practiceTwenty-three European centres participating in the DISPACT trial were evaluated using a standardized questionnaire to evaluate perioperative management [see Additional file ]. Standards in bowel preparation, pain and feeding management, antibiotics prophylaxis, incision type, drainages, necessity and duration of intensive care, mobilisation and perioperative somatostatin therapy were assessed. These relevant items were identified in a previously published meta-analysis and included in the questionnaire []. The survey was designed by T.L., N.N.R. and C.M.S and mailed to all centres taking part in the DISPACT trial. The survey was completed and returned by all local surgical sub-investigators. Consensus levels were classified for each outcome as published elsewhere []. Briefly, a discrete scale was used and the results were stratified along consensus levels – , which in turn corresponded to being standard in <% (no consensus), –% (overall agreement), –% (consensus), and >% (strong consensus) of participating centres, respectively.For time-critical factors such as the removal of intra-abdominal drainages, thoracic epidural catheters (TECs) and gastric tubes; the discharge from intensive care unit (ICU) or intermediate care (IMC) wards; and the duration of intravenous feeding and somatostatin therapy, median time and interquartile ranges (IQRs) have been calculated.Literature search of best practiceOnly RCTs or meta-analyses of RCTs assessing perioperative care were eligible for inclusion. Since specific evidence of perioperative care for distal pancreatectomy and pancreatic surgery was expected to be low, articles on general abdominal surgery including hepatic and colorectal surgery were included as well and evaluated in hierarchical order (distal pancreatectomy > pancreatic surgery > abdominal surgery > hepatic and colorectal surgery). Studies of paediatric, laparoscopic, and transplantation surgery were excluded. A search algorithm was developed and an extensive systematic Medline search using Boolean operator functions and wildcards (i.e. the asterisk symbol) was performed independently by three authors of this study (H.B., N.N.R. and T.L.). The search algorithm used single relevant keywords, medical subject headings (MESH), and their combinations (Appendix ). No other literature databases than Medline were searched and a subsequent analysis of the identified literature was performed. All published articles were evaluated.Reference lists of the retrieved literature were manually cross-searched for additional publications independently by three researchers (T.L., N.N.R and H.B.) under the supervision of an experienced researcher (C.M.S.). Moreover, captured citations were filtered for study design in order to identify all RCTs and meta-analyses. Titles, abstracts, and full text articles were screened for the selection of relevant studies.ResultsSurvey of current practiceAll centres participating in the DISPACT trial returned completely filled out questionnaires either per fax, mail or email. The annual caseload of pancreatic resections for each centre within the year is given in Table . All centres performed at least pancreatic resections, which was the requirement for participation in the DISPACT trial. Pancreatic left resections (median ; range –) are performed less commonly than other resections. Only two centres have not randomized patients for DISPACT within the last year. Median patients (range –) were randomized. The random ratio (randomized left resections/total left resection * ) was %.Bowel preparationMechanical bowel preparation (MBP) as a standard was performed at eight of the participating centres (Table ). Enemas were made use of at seven centres and one centre performed orthograde lavage; hospitals did not use any kind of MBP before left pancreatectomy (Table ).Table 2Perioperative standards in the European centres for pancreatic surgeryn%consensus levelmedian duration (IQR)Bowel preparation none1565overall agreement- enema730- orthograde lavage14-Type of incision Midline+- Transverse+1878consensus- Other24-Intraoperative single-shot antibiotic prophylaxis No14- Yes2296strong consensus-Intra-abdominal drainages No00 Yes23100strong consensus4 (–)Postoperative care IMC** (–.) ICU** (-) Nursery ward522- Not specified14-Pain management (thoracic epidural catheter) No313- Yes*2087consensus4 (–)Post-operative gastric tube No1357overall agreement- Yes10431 (–)Intravenous feeding No1252overall agreement- Yes11482 (.–)Somatostatin therapy No1565overall agreement- Yes8356 (.–)+ One centre reported using both midline and transversal incisions frequently.*Local anaesthetics and opioids were used at % of the centres (n = ), % (n = ) used local anaesthetics without opioids, and % (n = ) did not specify administered drugs.** Patients were transferred from the ICU to the IMC at three centres. While there is no consensus whether patients should be transferred to the ICU or IMC, there is an overall agreement in % of centres (n = ) that patients should not be returned to nursery wards immediately after distal pancreatectomy.Type of incision, intraoperative antibiotics and placement of abdominal drainagesThe transverse incision was exclusively used by hospitals and the midline by four hospitals. Two centres used other types of incisions and one centre used both incisions, depending on the individual surgeon's choice (Table ). Intraoperative single-shot antibiotics were performed in all but one centre (Table ).All centres placed intra-abdominal drainages. Drains were left for median of four days (IQR –), with a maximum of eight days at one centre (Table ).Postoperative carePostoperatively, centres transferred their patients to ICU or IMC wards, whereas five centres transferred patients directly to nursery wards. Patients were discharged from ICU and IMC after a median of one day (IQR -) and two days (IQR –.), respectively (Table ).Pain managementThoracic epidural catheters (TECs) were used in hospitals and kept in place for a median duration of four days (IQR –) (Appendix ). Fifteen centres used a combination of local anaesthetics and opioids for the medication of TECs and three centres used exclusively local anaesthetics (Appendix ). Two centres did not specify the medication.Nasogastric tube and feeding regimensPostoperatively, a nasogastric tube was placed for a median of one day (IQR –) in hospitals (Table ). Eleven centres used parenteral feeding for a median of two days (IQR .–) (Table ). Oral feeding with fluids started on day (IQR (–) and solid food on day (IQR –) (Table ). One centre reported keeping patients on parenteral feeding till day and starting oral feeding with fluids and solids on day .Table 3Mobilisation and oral feedingBegin of...postoperative day (median (IQR))...mobilisation1 (–)...oral feeding Fluids1 (–) Solid food2 (–)MobilisationMobilisation of patients started at postoperative day (IQR –) (Table ). Eight centres started mobilisation on the operative day, while centres started on first postoperative day. One centre reported starting mobilisation on the second postoperative day.Somatostatin therapyFifteen centres did not use Somatostatin as a routine after pancreatic surgery. In eight centres, somatostatin was used as a standard after left pancreatectomy for a median of six days (IQR .–) (Appendix ).Literature search of best practiceAll of the identified studies focus on large bowel surgery. Besides pancreatic surgery, other procedures have been included based on the similarity in concepts of perioperative management and determinants of complications.A total of meta-analyses of RCTs and RCTs were identified and reviewed [see Additional file ].Bowel preparationWe identified two meta-analyses [,] which included RCTs and six RCTs [-] with a total number of patients and patients, respectively [see Additional file ]. All of these studies failed to identify a benefit from bowel preparation; both meta-analyses [,] and two of the six RCTs [,] found beneficial effects of no bowel preparation in terms of decreased anastomotic leakage and/or wound infections. Slim et al. [] and Guenaga et al. [] identified a rate of .% vs. .% (p = .) and .% vs. .% (p = .) for bowel preparation versus no bowel preparation, respectively [see Additional file ].Type of incisionWhile complication rates were not different in the meta-analysis by Brown et al. [] in transversal versus midline incisions, a recent RCT by Fassiadis et al. [] identified a decreased risk of incisional hernia after transverse incision for abdominal aortic aneurysm repair only (p = .; [see Additional file ]. Moreover, Brown et al. [] suggested that transverse incisions may be less painful (odds ratio [OR] for less analgesic use (mg morphine equivalent) for total hospital stay, -.; % CI [-.; -.]) and may affect pulmonary function less than midline access (OR for percentage change in forced expiratory volume in one second on the last day of measurement, .; % CI [.; .]).Perioperative antibiotic prophylaxisRegarding antibiotic prophylaxis, the meta-analysis by Song et al. [], which included RCTs, showed single-shot antibiotics to be as effective as long-term antibiotics in the prevention of surgical wound infections [see Additional file ]. The recent study by Fujita et al. [] identified a dose-dependent correlation between surgical site infections and antibiotics (p = .; [see Additional file ].Abdominal drainagesGurusamy et al. [] included five trials and patients in their meta-analysis [see Additional file ]. No significant influence of drainages on complication rates after liver resection was found [see Additional file ] []. Petrowsky et al. [] were able to identify surgical procedures after which abdominal drainages significantly decrease complications (i.e. oesophageal resection and total gastrectomy) and in which abdominal drainages can be omitted (i.e. hepatic, rectal or colonic resection with primary anastomosis, and appendectomy) [see Additional file ]. In a RCT by Conlon et al. [], no correlation between the placement of abdominal drainages and the need for interventional drainages and surgical exploration after septic complications in patients undergoing pancreatectomy was detected [see Additional file ]. For hepatectomy, Sun et al. [] were able to identify a correlation of abdominal drainages and postoperative wound complications in a RCT (% vs. %, p < .; [see Additional file ].Pain managementIn their meta-analysis of RCTs, Marret et al. [] did not identify any differences in length of hospitalisation between patients receiving epidural or parenteral analgesia after colorectal surgery [see Additional file ]. Werawatganon et al. [], who included patients from nine RCTs in their analysis, and Wu et al. [] identified better pain relief with epidural analgesia compared to intravenous analgesia [see Additional file ]. In a meta-analysis by Jørgensen et al. [], a decreased rate of gastrointestinal paralysis with comparable effect on pain was found for epidural local anaesthetics compared to opioid-based medication [see Additional file ]. In RCTs, Flisberg et al. [] and Mann et al. [] were able to identify better pain relief with epidural analgesia and less frequent side effects compared to i.v. analgesia during rest (p < .) and mobilisation (p < .) [] and rest (p < .) and coughing (p < .) [], respectively [see Additional file ].Nasogastric tubesTwo meta-analyses by Nelson et al. [,] did not identify any advantage of routinely placed gastric tubes after abdominal surgery [see Additional file ]. Additionally, an earlier return of bowel function was found in patients without gastric tube in both studies (p < .) [,] [see Additional file ]. Pesseaux et al. [] included patients after hepatic resection in a RCT and did not find an advantage for nasogastric tubes. They demonstrated an increased risk of pulmonary complications (i.e. pneumonia and atelectasis, p = . and p = ., respectively) within the nasogastric tube group [see Additional file ] [].Feeding regimensAndersen et al. [] included RCTs with patients in their meta-analysis. No advantage for late return to oral diet was found [see Additional file ] []. Han-Geurts et al. [], who included patients in a RCT on early (i.e. within the first two postoperative days) versus late (i.e. more than two days after procedure) postoperative return to oral diet, also failed to identify a significant difference in or beneficial effect of keeping patients starved in terms of postoperative ileus and recovery [see Additional file ].MobilisationThree RCTs [-] on postoperative mobilisation and physiotherapy were available [see Additional file ]. While Browning et al. [] concluded that early mobilisation may decrease hospitalisation length, Mackay et al. [] did not find a benefit of chest physiotherapy in high risk patients after open abdominal surgery as far as pulmonary complications are concerned [see Additional file ]. In the larger study by Fagevik Olsén et al. [], chest physiotherapy decreased postoperative pulmonary complications and improved the mobilisation of patients (% in physiotherapy group vs. % in control group; p < .; [see Additional file ].Somatostatin therapyWhile Connor et al. [] and Alghamdi et al. [] identified a reduced complication rate in their meta-analyses on somatostatin versus no somatostatin after pancreatic surgery (p = . [] and p = . []), Zeng et al. [] did not find any significant effect on morbidity [see Additional file ]. All included meta-analyses [-] failed to identify any effect of somatostatin on mortality rates [see Additional file ].No beneficial effect of somatostatin was found in all included RCTs [,] on somatostatin after pancreatic surgery [see Additional file ]. Hesse et al. [] did not identify any reduced rate of pancreatic fistula. Even worse, Shan et al. [] in their study found delayed gastric emptying to be more frequent in the somatostatin group.DiscussionSurgical and perioperative management remains a matter of debate. Besides surgical expertise and technique [,], perioperative procedures significantly influence morbidity [,]. This survey of centres participating in a large multicenter trial on pancreatic left resection revealed that general perioperative standards can be detected in pancreatic surgery. In most items, at least overall agreement was present. Strong consensus was found in two items (i.e. placement of abdominal drainages and intraoperative antibiotic prophylaxis) and consensus was present for the usage of TECs for pain medication and transversal incisions as the abdominal cavity opening (Table ). The DISPACT centres agreed overall that no bowel preparation, gastric tube, somatostatin therapy or intravenous feeding was necessary after left pancreatectomy (Table ).While most of the strategies and standards in the DISPACT group are in accordance with current evidence, this survey has detected some differences in the perioperative management after pancreatic left resection despite current available evidence from meta-analyses and RCTs.Bowel preparationAs early as , Hughes [] has questioned the beneficial effect of bowel preparation in large-bowel surgery. However, this tradition has remained a standard in abdominal surgery for decades. Within the DISPACT group, there was overall agreement that no bowel preparation is necessary before pancreatic left resection (Appendix ). Based on the results from our literature research, this standard is in accordance with current evidence. Recent meta-analyses [,] and RCTs [-] did not identify any benefit from bowel preparation before abdominal surgery (Table 3A). In both meta-analyses by Guenaga et al. and by Slim et al. [,], an increased rate of anastomotic leakage was identified. Moreover, bowel preparation may even be disadvantageous in terms of wound infections [-].In eight (%) of DISPACT centres, bowel preparation was used as a standard prior to left pancreatectomy. This is in contrast to the results of studies conducted in colorectal surgery demonstrating no significant differences for anastomotic leakage rates and overall morbidity [see Additional file ] [-]. While there is no evidence that enema or orthograde lavage may be harmful in pancreatic surgery, it seems likely that there is no benefit from bowel preparation. Therefore, bowel preparation can no longer be considered as a necessary standard in pancreatic surgery. Most DISPACT centres have therefore already stopped any bowel preparation.Type of incisionA recent meta-analysis did not show any significant difference between midline and transverse incision as far as complication rates, incisional hernias and recovery times [see Additional file ] [,]. Amongst the DISPACT centres, there was consensus (%) that a transverse incision may be more beneficial for distal pancreatectomy (Table ). While the meta-analysis by Brown et al. [] failed to identify any difference between groups with midline and transverse abdominal incisions, Fassiadis et al. [] found increased rates of incisional hernia in patients undergoing surgery for aortic aneurysms after midline versus transversal incision ( of patients versus of patients, respectively). This finding is confounded by the underlying connective tissue disorders which is a causative agent for both diseases and therefore may not be relevant for benign and malignant pancreatic disorders [-]. Therefore, the practise of the DISPACT centres is in accordance with the evidence.Antibiotic prophylaxisAntibiotic prophylaxis significantly reduces surgical site infections as long as local tissue concentrations reach an effective level [see Additional file ] [,]. Both single-shot and long-term antibiotics are equally effective. As a result, antibiotic prophylaxis is a common procedure in surgery.All but one centre agreed on using single-shot antibiotics before left pancreatectomy (Appendix ). Given the ubiquitous use of antibiotic prophylaxis in surgery, it can be suspected that one centre chose a different regime and may have used a long-term antibiotic prophylaxis.Abdominal drainagesPlacement of drainages has a long tradition in abdominal surgery and can be dated back to the 19th century. While many surgical procedures originating from these times have been omitted, Lawson Taits advice of "when in doubt, drain" is still considered standard in many surgical centres []. Not surprisingly, all centres taking part in the DISPACT trial agreed on using intra-abdominal drainages as a standard after distal pancreatectomy (Table ). There were considerable differences in time concerning the removal of drainages, though; the median day of removal was the fourth postoperative day (IQR –). Some centres kept drainages as a standard for up to eight days.The detection of postoperative pancreatic fistulas according to the consensus definition requires a drainage for at least three days []. Since the occurrence of pancreatic fistula is the primary endpoint of the DISPACT trial, placement of abdominal drainages had to be included as a standard in the study protocol. Unfortunately, there is evidence that intra-abdominal drainages are not beneficial in all cases and sometimes may even be harmful [see Additional file ] [-]. For a number of procedures (i.e. liver resection, appendectomy and colonic or rectal resection with primary anastomosis), there is no need for abdominal drainages while for other procedures (i.e. oesophageal resection or total gastrectomy) there are proven benefits []. Moreover, wound complications may even be more common when drainages are placed [].A recent RCT by Conlon et al. [] failed to show a decreased mortality or rate of complications by intraperitoneal drainages after pancreatectomy [see Additional file ]. A total number of patients were included. Forty of them underwent distal pancreatectomy; data was consistent among the different procedures in pancreatic surgery. No reduced need for interventional drainages and surgical exploration for septic complications was found in the drainage group [].In summary, the value of drainages in pancreatic surgery remains unclear. Further studies are necessary to resolve the currents uncertainties regarding drainages after pancreatic surgery.Postoperative careMost centres (%) agree that patients undergoing distal pancreatectomy should be transferred to ICU or IMC wards (Table ). In % of the DISPACT centres, patients are transferred directly to regular nursery wards. No meta-analyses or RCTs were identified in our Medline research regarding this topic.Pain managementPain management is generally considered a key factor of fast track concepts in surgery and a determinant of hospitalisation length [,]. In (%) of the DISPACT centres, a thoracic epidural catheter was used for pain management. Fifteen centres (%) applied regimens consisting of local anaesthetics and opioids (Table ). While overall time of hospitalisation is not significantly different between patients receiving epidural and parenteral opioids, epidural administration significantly decreases side effects [see Additional file ] [-,,]. In RCTs published by Filsberg et al. [] and Mann et al. [], epidural versus intravenous application of pain medication was compared [see Additional file ]. Opioid-related gastrointestinal side effects especially are less frequent in epidural analgesia groups [,]. In the study by Jørgensen et al. [], no difference in pain relief could be detected between opioid-based analgesia and analgesia using epidural application of local anaesthetics only. Within the DISPACT group, thoracic epidural catheters (TEC) were clearly preferred for pain management and placed for median of four days (IQR -, Table ). The preference for opioids in this context seems to be questionable.Nasogastric tubesMost DISPACT centres do not apply post-operative gastric tubes as a standard after distal pancreatectomy (Table ). Ten centres (%) keep nasogastric tubes in place for a median of one day (IQR –) (Table ). Both meta-analyses and RCTs with a total number of included patients could not detect benefits from nasogastric tubes after abdominal surgery [see Additional file ] [-]. Contrary to traditional belief, gastric tubes as a postoperative standard seem to be associated with a higher risk of pulmonary complications and delayed return of bowel function [see Additional file ] [-]. Postoperative removal of nasogastric tubes is a standard based on evidence.Feeding regimensAdministration of oral fluids was performed on median postoperative day (IQR –) and of solid food on day (IQR –) (Table ). Most centres did not perform intravenous feeding (Table ). Both the RCT by Han-Geurts et al. [] and the meta-analyses by Andersen et al. [] did not find any benefit in keeping patients starved after abdominal surgery [see Additional file ]. Since an early (i.e. second postoperative day) return to an oral diet is not associated with increased postoperative complications, there is no rationale in keeping patients starved [see Additional file ] [,].MobilisationEarly postoperative patient mobilisation (median postoperative day (IQR –) is performed in most of the centres (Table )). While early mobilisation actually may reduce length of hospitalisation, the value of postoperative chest physiotherapy remains unclear [see Additional file ] [-]. Mackay et al. [] could not detect any significant decrease in pulmonary complications with chest physiotherapy in high risk patients. Contrary to this, Fagevik et al. [] reported a significant decrease in postoperative pulmonary complications and an improved mobilisation with chest physiotherapy [see Additional file ].Somatostatin therapyFifteen (%) DISPACT centres do not use somatostatin as a postoperative standard (Appendix ). All reviewed studies agree that prophylactic somatostatin does not decrease mortality [see Additional file ] [-,]. As far as meta-analyses are concerned, the effect of somatostatin on the occurrence of pancreatic fistula remains unclear [-]. All reviewed RCTs not only agree that somatostatin does not have any influence on occurrence of pancreatic fistula, but Shan et al. [] also found increased rates of delayed gastric emptying after prophylactic somatostatin administration and provided decreased plasma motilin levels as a possible mechanism [see Additional file ] [,]. Therefore, somatostatin as a postoperative standard should be avoided.ConclusionEvidence-based perioperative management is an important cornerstone for a successful outcome after complex surgical procedures. The results of this survey and the comparison with the current available evidence detected some inconsistencies in perioperative management of patients with pancreatic left resection. At least overall agreement was present in most items and was – with the exception of abdominal drainages – in accordance with strategies that were identified as beneficial for the patients from our literature research. Evidence-based perioperative treatment in pancreatic surgery requires the further conduct of randomized controlled trials.AbbreviationsDISPACT: DIStal PAnCreaTectomy; ICU: Intensive Care Unit; IMC: InterMediate Care; IQR: Inter Quartile Range; ISRCTN: International Standard Randomized Controlled Trial Number; MBP: Mechanical Bowel Preparation; MESH: Medical Subject Headings; RCT: Randomized Controlled Trial; TEC: Thoracic Epidural Catheter.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsHB performed the analysis, literature research and drafted the manuscript. NRR and TL performed the survey and literature search. MKD and TJ designed the study. CMS designed the study and reviewed the manuscript. MG, GB, CS, IR and MWB reviewed and contributed substantially to the manuscript. All authors read and approved the final manuscript.Appendix 1Items and corresponding search algorithms used for Medline search of literature-based standardsBowel preparation(Pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw] OR hepatic surgery [tw] OR hepatic resection [tw] OR liver surgery [tw] OR hepatectomy [tw] OR colorectal surgery [tw] OR colon resection [tw] OR colon surgery [tw] OR abdominal surgery [tw] OR gastrointestinal surgery [tw]) AND (Mechanical bowel preparation [tw] OR bowel preparation [tw] OR bowel cleansing [tw] OR colon preparation [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH] NOT child [MESH] NOT infant [MESH])Thoracic epidural catheter((Pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw] OR hepatic surgery [tw] OR hepatic resection [tw] OR liver surgery [tw] OR hepatectomy [tw] OR colorectal surgery [tw] OR colon resection [tw] OR colon surgery [tw] OR abdominal surgery [tw] OR gastrointestinal surgery [tw]) AND (Thoracic epidural catheter [tw] OR epidural catheter* [tw] OR epidural analgesia [MESH] OR epidural opioid [tw] OR epidural local anaesthetic [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH])Somatostatin(Gastrointestinal surgery [tw] OR pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw]) AND (somatostatin [tw] OR octreotide [tw] OR vapreotide [tw] OR lanreotide [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH])Incision(Surgery [MESH]) AND (midline incision [tw] OR transverse incision [tw] OR abdominal incision [tw] OR midline laparotomy [tw] OR transverse laparotomy [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH])Antibiotic prophylaxis(Pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw] OR hepatic surgery [tw] OR hepatic resection [tw] OR liver surgery [tw] OR hepatectomy [tw] OR colorectal surgery [tw] OR colon resection [tw] OR colon surgery [tw] OR abdominal surgery [tw] OR gastrointestinal surgery [tw]) AND (Antibiotic prophylaxis [tw] OR single shot [tw] OR antimicrobial prophylaxis [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH] NOT child [MESH] NOT infant [MESH])Nasogastric tube(Pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw] OR hepatic surgery [tw] OR hepatic resection [tw] OR liver surgery [tw] OR hepatectomy [tw] OR colorectal surgery [tw] OR colon resection [tw] OR colon surgery [tw] OR abdominal surgery [tw] OR gastrointestinal surgery [tw]) AND (Nasogastric decompression [tw] OR nasogastric tube [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH] NOT child [MESH] NOT infant [MESH])Abdominal drainage(Surgery [MESH]) AND (abdominal drain* [tw] OR intraperitoneal drain* [tw] OR intraabdominal drain* [tw] OR prophylactic drain* [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH] NOT child [MESH] NOT infant [MESH])Postoperative feeding(Pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw] OR hepatic surgery [tw] OR hepatic resection [tw] OR liver surgery [tw] OR hepatectomy [tw] OR colorectal surgery [tw] OR colon resection [tw] OR colon surgery [tw] OR abdominal surgery [tw] OR gastrointestinal surgery [tw]) AND (feeding [tw] OR postoperative feeding [tw] OR enteral feeding [tw] OR nasogastric feeding [tw] OR nasojejunal feeding [tw] OR parenteral nutrition [tw] OR enteral nutrition [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH] NOT child [MESH] NOT infant [MESH])Patient mobilisation(Pancreas surgery [tw] OR pancreas resection [tw] OR pancreatic surgery [tw] OR pancreatic resection [tw] OR distal pancreatectomy [tw] OR pancreatic left resection [tw] OR pancreaticoduodenectomy [tw] OR hepatic surgery [tw] OR hepatic resection [tw] OR liver surgery [tw] OR hepatectomy [tw] OR colorectal surgery [tw] OR colon resection [tw] OR colon surgery [tw] OR abdominal surgery [tw] OR gastrointestinal surgery [tw]) AND (mobilisation [tw] OR mobilization [tw] OR postoperative mobilisation [tw] OR patient mobilisation [tw] OR postoperative mobilization [tw] OR patient mobilization [tw]) AND (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [MESH] OR random allocation [MESH] OR double-blind method [MESH] OR single-blind method [MESH] OR clinical trial [pt] OR clinical trials [MESH] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR placebos [MESH] OR placebo* [tw] OR random* [tw] OR research design [mh:noexp] OR Meta-Analysis [ptyp] OR systematic review [tw] NOT (animal [MESH] NOT human [MESH]) NOT case reports [pt] NOT Comment [PT] NOT letter [PT] NOT child [MESH] NOT infant [MESH] NOT child [MESH] NOT infant [MESH])Supplementary MaterialAdditional file 1Survey on peri-operative standards in distal pancreatectomy. A copy of the research instrument used for the survey.Click here for fileAdditional file 2Meta-analyses and randomized controlled trials on perioperative management. Appendix summarizes relevant data from the meta-analyses and randomized controlled trials found by the literature search of best practice.Click here for file
|
PMC2447266.txt
|
TITLE: StressDB: A Locally Installable Web-Based Relational Microarray Database Designed for Small User Communities AUTHORS: Madhusmita Mitra, Nigam Shah, Lukas Mueller, Scuth Pin, Nina Fedoroff ABSTRACT: We have built a microarray database, StressDB, for management of microarray data from our studies on stress-modulated genes in Arabidopsis. StressDB provides small user groups with a locally installable web-based relational microarray database. It has a simple and intuitive architecture and has been designed for cDNA microarray technology users. StressDB uses Windows™ as the centralized database server with Oracle™ 8i as the relational database management system. It allows users to manage microarray data and data-related biological information over the Internet using a web browser. The source-code is currently available on request from the authors and will soon be made freely available for downloading from our website athttp://arastressdb.cac.psu.edu. BODY: No Body Content
|
PMC3003677.txt
|
TITLE: Unusual presentation of peritonitis with persistent clear aspirate: a case report AUTHORS: Ebru Asicioglu, Arzu Kahveci, Elif Ari Bakir, Atilla Bulur, Hakki Arikan, Mehmet Koc, Serhan Tuglular, Cetin Ozener ABSTRACT: IntroductionPeritonitis is the most frequent complication of peritoneal dialysis. Diagnosis of peritonitis includes symptoms and signs of peritonitis with a cloudy aspirate of more than WBC/ml, as well as positive cultures. Although sterile peritonitis has been reported in the literature, to the best of our knowledge this is the first report of an unusual presentation of peritonitis without any white blood cells in the peritoneal aspirate despite multiple positive peritoneal cultures.Case presentationAn -year-old Caucasian man who had been on continuous cycling peritoneal dialysis for years was admitted to our hospital with general malaise, loss of appetite, weight loss and somnolence. He did not describe abdominal pain or fever. Even though his peritoneal fluid was consistently negative for leukocytes and clear, he had peritonitis with different organisms consecutively.ConclusionsOur case report shows that any patient on peritoneal dialysis presenting with evidence of infection (fever, peripheral leukocytosis) without an obvious cause should have aspirate cultures done even if the aspirate is clear and abdominal pain is absent. Our case report may change the initial work-up and management of these patients. We believe this report is of interest to general medicine and emergency room physicians as well as nephrologists. BODY: IntroductionPeritonitis is the most frequent complication of peritoneal dialysis (PD). Diagnosis of peritonitis includes symptoms and signs of peritonitis with a cloudy aspirate of more than WBC/ml, as well as positive cultures. Although sterile peritonitis has been reported in the literature [], to the best of our knowledge this is the first report of an unusual presentation of peritonitis without any white blood cells in the peritoneal aspirate despite multiple positive peritoneal cultures.Case presentationAn -year-old Caucasian man who had been on continuous cycling peritoneal dialysis (CCPD) for years was admitted to our hospital with general malaise, loss of appetite, weight loss and somnolence. He did not describe abdominal pain or fever. His medical history included renal failure due to nephrolithiasis and he had not experienced peritonitis before. His physical examination was unremarkable and he was experiencing no abdominal tenderness at the time of his admission. His white blood cell count was /ml with percent neutrophils. His C-reactive protein (CRP) level was mg/L (Normal range: - mg/L). His blood cultures were negative. He was started on ampicillin and sulbactam and ciprofloxacin treatment empirically. Even though his peritoneal fluid was clear and there were not any leukocytes present, peritoneal fluid cultures taken on admission revealed Pseudomonas aeruginosa. The ampicillin and sulbactam was switched to piperacillin and tazobactam and ciprofloxacin treatment was continued since the isolate was sensitive to both. Two weeks after his admission, he developed a cough with sputum production. There was a new infiltration on his chest X-ray consistent with pneumonia and his CRP level increased from mg/L to mg/L. He was switched to imipenem and vancomycin therapy. He still did not describe abdominal pain or fever and there was no abdominal tenderness on physical examination. Even though the peritoneal fluid count did not show any leukocytes, a control peritoneal fluid culture revealed Stenotrophomonas maltophilia which was sensitive to trimethoprim-sulfamethoxazole. Trimethoprim-sulfamethoxazole was added to treatment. Despite treatment with imipenem, vancomycin and trimethoprim-sulfamethoxazole for two weeks, his condition deteriorated rapidly and he developed septic shock. While analysis of his peritoneal fluid was still negative for white blood cells, peritoneal cultures obtained during the septic shock revealed vancomycin-resistant Enterococcus faecium (VRE). His blood cultures were negative, however stool cultures also revealed VRE. He was started on linezolide therapy however his condition deteriorated rapidly and he died six weeks after hospitalization.DiscussionPeritonitis causes significant morbidity and mortality in PD patients. It can result in death either from sepsis or from ensuing complications. Gram-positive organisms, especially Staphylococcus epidermidis and Staphylococcus aureus, are the most common bacteriologic causes of peritonitis. Recently, the incidence of Gram-positive infections has decreased due to newer techniques and improved exit-site care, while the incidence of Gram-negative organisms has remained stable []. Gram-negative organisms now account for to percent of all peritonitis episodes []. Patients with Gram-negative peritonitis have a worse outcome, are more likely to be hospitalized and more likely to die within six months after peritonitis []. In our case report, peritoneal fluid cultures yielded three different organisms consecutively. The first organism was P. aeruginosa which is one of the most important causes of Gram-negative peritonitis and is identified in . percent of culture positive cases []. The major risk factors for infection are a history of antibiotic therapy within days and concomitant exit-site infection []. Our patient did not have any risk factors for this infection. The second microorganism was Stenotrophomonas maltophilia which is an environmental saprophyte that has been rarely reported to cause peritonitis. The annual rate of isolation for this organism is not known since it has only been reported as case reports. It is mainly a nosocomial pathogen and is resistant to multiple antibiotics []. It can be isolated from water, soil, hospital equipment and humans. The major risk factors include prior broad-spectrum antibiotic use, neutropenia, underlying illness and prolonged hospital stay []. The patient described in this case report had risk factors for this organism, such as prolonged antibiotic use and hospital stay. The third microorganism was VRE, which was first described in the s and has since become increasingly prevalent, causing infections that include PD-associated peritonitis []. However, the annual rate of isolation is not known. Risk factors for VRE include severe underlying disease, immunosuppression, prior use of vancomycin and other antibiotics, abdominal or cardiothoracic surgery and indwelling urinary or central venous catheters []. Patient had multiple risk factors for VRE infection, such as vancomycin use and prolonged hospital stay. The persistent growth of organisms in his peritoneal culture and his lack of response to antibiotic therapy warranted early removal of the catheter. Failure to respond to antibiotic therapy necessitating early catheter removal is common in patients with S. maltophilia and/or VRE peritonitis. Our patient was repeatedly advised immediate surgical removal of his PD catheter due to the risk of septic shock and death; however, he refused to undergo the procedure despite regular counselingThe majority of patients with peritonitis have cloudy fluid and abdominal pain. The diagnosis is confirmed by a dialysate aspirate leukocyte count exceeding WBC/ml with at least percent polymorphonuclear leukocytes. A small percentage of patients do not have white blood cells in the peritoneal fluid at presentation with cell counts increasing in time; this is called an impaired initial cell reaction []. This delay may be due to slower cytokine response to infection []. Peritonitis episodes with cell counts less than WBC/ml have been reported in the literature [,]. However, peritonitis without abdominal pain and/or white blood cells in the peritoneal fluid is an extremely rare entity and to the best of our knowledge, this is the first report of a case of peritonitis with a persistently clear aspirate without any white blood cells. The peritoneal fluid cultures of our case report yielded different organisms consecutively, however the peritoneal cell counts were consistently nil. It can be argued that the positive cultures in the absence of increased white blood cell count in the peritoneal fluid could be indicative of colonization or improper culture collection. The method used for peritoneal culture at our center is as follows: sediment obtained after centrifugation of ml of fluid is inoculated into aerobic and anaerobic Bactec bottles, on aerobic and anaerobic blood, chocolate and MacConkey agar plates. Gram and Giemsa stains are also performed. It is highly unlikely that the only three improper cultures collected at our center so far have been in the same patient. A low white blood cell count in peritoneal fluid in the presence of antibiotic use has been reported. However, since the first organism was isolated before the use of any antibiotic therapy and the consecutive organisms were both resistant to the antibiotic therapy administered at that time, it is not possible to explain the absence of white blood cells by prior therapy. Immunosuppression is another possible mechanism for the lack of white blood cells in our case report; however, he did not have any disease, condition or any other laboratory values indicating that he was indeed immunosuppressed.We believe our case report demonstrates that clear aspirate and the absence of white blood cells in the peritoneal fluid may not rule out the presence of peritoneal infection in patients on PD. Therefore, any patient on PD presenting with evidence of infection (fever, peripheral leukocytosis) without an obvious cause should have aspirate cultures done even if the aspirate is clear and abdominal pain is absent.ConclusionsPatients on PD presenting with evidence of infection (fever, peripheral leukocytosis) without an obvious cause should have aspirate cultures done even if the aspirate is clear and abdominal pain is absent.ConsentWritten informed consent was obtained from the patient's next of kin for publication of this case report and any 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' contributionsEA, AK, EAB, AB and HA made substantial contributions to the design, and the acquisition and interpretation of data. MK, ST and CO revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript.
|
PMC2709932.txt
|
TITLE: Signaling Transduction Network Mediated by Tumor Suppressor/Susceptibility Genes in NPC AUTHORS: Minghua Wu, Xiayu Li, Xiaoling Li, Guiyuan Li ABSTRACT: Nasopharyngeal carcinoma (NPC) is a polygenetic disease. SPLUNC1, UBAP1, BRD7, NAG7, NOR1, NGX6 and LTF genes were found to be tumor suppressor/susceptibility genes in different stages of NPC. SPLUNC1, an early warning molecular diagnosis marker, inhibits the bacteria clone formation, and is an innated immune molecule. SPLUNC1 can negatively regulate the ERK/MAPK signaling transduction pathway to inhibit NPC cell proliferation and induce apoptosis. BRD7, a transcript regulation factor, interacts with BRD2, and promotes apoptosis induced by BRD2. Its promoter is regulated by c-Myc and SP1. BRD7 inhibits NPC cell cycle progression, preventing passage through G0/G1 by suppressing ras/MEK/ERK, Rb/E2F and Wnt signaling pathways. Abnormal activation of BRD7 is crucial to cell cycle turbulence in NPC. NGX6, a metastasis-associated protein, can negative-regulate the EGF/Ras/MAPK signaling transduction pathway, and interacts with ezrin protein to inhibit NPC cell invasion and metastasis. LTF, also a metastasis-associated protein, can negatively regulate MAPK signal transduction pathways, such as JNK2 and ERK, to inhibit NPC cell proliferation and growth. Taken together, it was found that these tumor suppressor/susceptibility genes can regulate key molecules involved in cell signal pathways such as ras/MEK/ERK, Rb/E2F and EGFR ras/MEK/MAPK, and can regulate the expression of some adhesion molecules such as ezrin, nm23 and α-catenin. According to functional genomics and signaling transduction pathways, we have described a signaling cross-talk network between the tumor suppressor/susceptibility genes involved in NPC. These tumor suppressor/susceptibility genes may be potential treatment targets for NPC in the future. BODY: INTRODUCTIONNasopharyngeal carcinoma (NPC) is a multifactorial disease that presents a challenge to clinicians and biologists in various fields including epidemiology, genetics, virology and immunology []. NPC has a striking geographical distribution and Southern China is a major nasopharyngeal carcinoma-endemic region. Chinese emigrants continue to have a high incidence of the disease, but the rate of NPC among ethnic Chinese born in North America is considerably lower than those born in China []. This epidemiologic evidence implies that both environmental factors and genetic susceptibility play roles in the development of NPC. The possible existence of several susceptibility genes (including two of them on the 4q [] and 3p [] chromosomes) pointed to the importance of epidemiology and genetics in NPC. Cellular gene alterations also contribute to NPC development, especially inactivation of tumor suppressor genes, such as SPLUNC1, UBAP1, BRD7, NOR1, NGX6 and LTF, which are involved in different stages of NPC initiation and progression (Fig. ) []. SNP of BRD7 (C450T and A737C), NGX6 (rs879284), UBAP1 (rs1049557) and NOR1 (ss2220003 and ss3211583) have been demonstrated to be important genetic susceptibility risk factors for NPC [-]. Xiao et al. found that the expression of UBAP1 decreases during NPC development and progression from normal epithelium of nasopharynx to hyperplastic epithelium of nasopharynx, then, to atypical hyperplastic epithelium of the nasopharynx and finally, to nasopharygeal carcinoma [].Although combination radiotherapy/chemotherapy is sensitive for stage I and II NPC and improves survival rates, unfortunately the majority of NPC is diagnosed in advanced stages because of nonspecific presenting symptoms (cervical nodalenlargement, headache, nasal and aural dysfunction), delay in seeking treatment after the onset of symptoms, and the difficulty of a thorough nasopharyngeal exam []. In light of this, more precise molecular diagnosis and targeted treatments of NPC need to be developed. In this review, the function of tumor suppressor/susceptibility genes in different stages of NPC and signal transduction mediated by tumor suppressor/susceptibility genes will be discussed. Specifically, we will provide evidence for molecular diagnosis, prognosis and targeted treatments in NPC.SPLUNC1, AN EARLY WARNING MOLECULAR DIAGNOSIS MARKER, IS INVOLVED IN ERK/MAPK AND NF-κB SIGNALING PATHWAYS, WHICH INDUCE CELL APOPTOSIS IN NPCSPLUNC1 (short palate, lung, and nasal epithelium clone ) is isolated from human nasopharyngeal epithelia by suppression subtractive hybridization (SSH) and cDNA microarray analysis []. Its expression is downregulated in % of nasopharyngeal carcinoma (NPC) biopsies but upregulated in % of lung cancer biopsies. SPLUNC1 is a gene transcript of the PLUNC gene family and also is a member of the BPI (bactericidal permeability increasing protein)/LBP (lipopolysaccharide-binding protein) family with putative bactericidal/bacteriostatic functions. All BPI and LBP, including SPLUNC1 are bactericidal proteins and have the function of neutralizing endotoxin [-]. SPLUNC1 is downregulated in slight atypical hyperplasia of nasopharynx endepidermis, and its expression gradually decreases with progression of malignant nasopharynx endepidermis from slight atypical hyperplasia to nasopharyngeal carcinoma []. It is found to scavenge bacterial endotoxin [], EB virus [] and it also can combine with nanobacteria [] and ameliorate its deleterious effects. Furthermore, it is involved in the earliest stage of NPC carcinogensis, where it plays a role in innate immunity. Thus, SPLUNC1 is an early warning molecular diagnosis marker and is used to carry out early diagnosis and crowd risk forecasting for NPC.The MAPK (mitogen-activated protein kinase) cascade relays the presence of extracellular stimuli such as growth hormones to the nucleus and controls the expression of hundreds of genes. MAPKs control major cell fate decisions such as proliferation, differentiation and apoptosis, mainly by inducing alterations in gene expression []. The MAPK cascade includes four main signaling pathways, ERK, JNK, P38 and ERK5. SPLUNC1 enhances apoptosis of NPC HNE1 cells and EBV-transformed human B-lymphocytes []. SPLUNC1 evidently inhibits the expression of phosphorylated ERK (p-Tyr-) and its BPI domain contributes to this inhibitory effect. SPLUNC1 inhibits the expression JNK2, not JNK1, and has no effect on total P38 protien or phosphorylated P38 (p-Tyr-). At the same time, SPLUNC1 can promote expression of NF-κB (Nuclear factor NF-κB) and I-κα protein, which is an inhibitory effect dependent upon its BPI domain, but the inhibitory effect of phosphorylated I-κα(p-Ser-) has nothing to do with its BPI domain. Therefore, SPLUNC1 can regulate ERK/MAPK and NF- kB signaling pathways to induce cell apoptosis in NPC [].DOWNREGULATED EXPRESSION OF RASSF1A IS INVOLVED IN NF-κB SIGNALING PATHWAYRas-association domain family of proteins (RASSF) are characterized by the Ras-association (RA) domain at the C-terminal. RASSF1A is a target tumor suppressor gene on 3p21. in NPC. The aberrant hypermethylation of RASSF1A and high EBV load might be important events in NPC pathogenesis and they may be useful molecular diagnostic markers for this cancer [-]. Downregulation of RASSF1A expression is dependent on the activation of intracellular signaling of NF-κB involving the C-terminal activating regions (CTARs) of LMP1 [].BRD7, A NUCLEAR TRANSCRIPTION FACTOR, REGULATES CELL CYCLE PROGRESSION BY MEK/ERK/MAPK, Rb/E2F AND Wnt/β-CATENIN PATHWAYSCell cycle progression is a strictly controlled process and it needs precise interplay between many molecules. The regulatory pathway controlling G1–S is the Rb/E2F pathway: in the presence of extracellular growth-stimulatory signals, cyclinD1 and its kinase partner cdk4 form an activated complex and phosphorylate Rb and Rb family members, thus inactivating Rb and allowing E2F/DP transcription factors to exert their transactivation activity. This in turn leads to transcription of a number of genes essential for DNA replication and entry into S phase, thus the cell enters into S phase and completes the irreversible cell cycle progression []. BRD7 is a nuclear transcription regulation factor containing a bromodomain that is found in many chromatin-associated proteins and in nearly all known nuclear histone acetyltransferases (HATs) and has been found to play an important role in chromatin remodeling and transcriptional activation []. The transcriptional regulation of BRD7 is achieved by binding to acetylated histone H3 at Lys14. The bromodomain is essential for this role. Chromatin remodeling, not chromatin modification, is the major mechanism of BRD7 mediated gene transcription []. The downexpression of the BRD7 gene has been shown to be critical to the pathogenesis of NPC []. Promoter methylation inhibits BRD7 expression in NPC [] and its promoter is regulated by c-Myc and Sp1 []. BRD7 interacts with BRD2 and stimulates apoptosis induced by BRD2. BRD7 protein inhibited cell growth and cell cycle progression from G1 to S phase by transcriptionally regulating some cell cycle associated genes such as Rb and E2F3 in NPC cells []. Zhou et al. [] and Peng et al. [] found that potential targets of BRD7 were mainly involved in MEK/ERK/MAPK, Rb/E2F and Wnt/β-catenin signaling pathway.Extra-cellular signals can be transmitted into the nucleus cell cycle machinery through diverse signaling pathways, among which MEK/ERK/MAPK is one intensively studied pathways []. The over-expression of BRD7 in NPC cells results in the downregulation of the c-jun, p-MEK and p-ERK1/ expression and Ap- promoter inactivity, therefore, BRD7 plays a negative role in the MEK/ERK/MAPK pathway []. E2Fs and DPs protein are both transcription factor families that can form heterodimers that play central roles in the expression of genes essential for S phase entry []. The expression of DP2 and E2F3 were both decreased by the induction of BRD7, peaking .- and .-fold, respectively [] and the transcription targets of E2F/DP, such us replication factor A (RFA), replication factor C37 and subunit (RFC37 and RFC38), were all downregulated as a result of the decrease of E2F3/DP2. RFA and RFC are important components for DNA replication initiation by contributing to origin unwinding, primering and firing during DNA replication []. BRD7 interacts with Ceap- (centrosome associated protein-, also termed BLOS2) by C-terminus of BRD7 and the central region of Ceap-. Through this binding, Ceap- can translocate from cytoplasm to the nucleus where it selectively inhibits the transcriptional suppression activity of BRD7 towards certain target genes including E2F3 and cyclin A []. In addition, Kim et al. [] reported that BP75, the most homologous gene of BRD7 (bromodomain containing protein ), was found to bind dishevelled- and enhance Wnt signaling by inactivating GSK-3β (glycogen synthase kinase- beta). In NPC cells, the induction of BRD7 increased the expression of α-catenin which “hold” β-catenin in the complex and inhibit β-catenin accumulation in the nucleus to induce the downregulation of cyclinD1, E2F3 []. As a nuclear transcription regulation factor, the nuclear localization of BRD7 is critical for the expression of cell cycle related molecules and cell biological function. NLS (nuclear localization signal) is an essential motif affecting BRD7 nuclear distribution, and NLS-deleted BRD7 shifts the nuclear localization mostly to the cytoplasm, and failed or reduced to negatively regulate the expression of cell cycle related molecules, cyclin D1 and E2F3, and cell cycle progression from G1 to S phase [].NGX6, A METASTASIS-ASSOCIATED PROTEIN, CAN NEGATIVE-REGULATED EGF/Ras/MAPK SIGNALING PATHWAY AND INTERACT WITH EZRIN PROTEIN TO INHIBIT INVASION AND METASTASIS OF NPC CELLSCancer metastasis is a very complicated biological process involving many sequential steps. Cell-cell and cell-extracellular matrix (fibronectin, laminin, collagen, etc.) interactions are involved in the metastatic process []. NGX6 (nasopharyngeal carcinoma (NPC)-associated gene ) is a metastasis-associated gene located on human chromosome 9p21–. Loss of NGX6 expression is associated with lymph node or distance metastasis in colorectal carcinomas [] and NPC []. The NGX6 protein includes two transmembrane (TM) regions. The extracellular region contains one epidermal growth factor (EGF)-like domain and three potential N-glycosylation sites, and the short cytoplasm (CYTO) contains a tyrosine residue that is a potential tyrosine kinasephosphorylation site. The EGF-like domain is a sequence of aa residues long, which has a significant homology to EGF. The EGF-like domain, which contains a unanimous sequence of CX7CX –3GXCX10– CXCX3YXGXRCX –n, is involved in the interaction between receptors and ligands in cell adhesion and signal transduction []. Molecules containing EGF-like domains are mostly involved in cell adhesion, matrix formation, injury, recovery and chemotaxis []. N-Glycosylation sites are signatures of some cell adhesion molecules []. NGX6 modulates the adhesion and invasion process via both the EGF-like domain and the CYTO region [], and could delay cell cycle G0-G1 progression and thus inhibit cell proliferation by negatively regulating the EGFR Ras/Mek/MAPK pathway in NPC cells. As an EGF-like domain gene [], NGX6 influences the expression of some important cell adhesion-related molecules. In NPC cells, the expression of RhoA and vitronectin are both decreased following the induction of NGX6, and the induction of NGX6 up-regulates the expression of KRT1, integrin a2, integrin b7, PSCD2L, CD9, ezrin, nm23-H1, VE-cadherin and catenin a2. RhoA is implicated in the invasion of human microvascular endothelial cells (HMEC-). Ectopic expression of active-RhoA GTPase induces the expression of the MMP- metalloproteinase []. Vitronectin can promote cell adhesion and spreading. In normal tissue, vitronectin has a homogeneous periductal occurrence, with local accumulations much lower than in carcinoma tissues []. KRT1 is a member of the keratin gene family associated with differentiation []. Integrins are a family of transmembrane glycoproteins that participate in a wide range of cellular events including cell adhesion, proliferation, apoptosis, differentiation and cell-surface-mediated signaling []. PSCD2L (Cytohesin-) can specifically interact with CD18 (integrin b2) and can promote cell adhesion to ICAM1 []. CD9 is a cell-surface glycoprotein, that can modulate cell adhesion and migration and that can also trigger platelet activation and aggregation []. Nm23-H1 is a tumor metastasis inhibitor in many tumors []. VE-cadherin is a calcium-dependent cell--cell adhesion glycoprotein []. The up or down-regulation of those adhesion molecules reflects a possible role of NGX6 in tumor invasion and metastasis.Ezrin has been suggested to be a mediator of cell motility, is essential for the maintenance of cell-cell adhesion, and is inhibitory towards cell matrix adhesion in human colonic epithelial cells. Ezrin, a linker between membrane protein and cytoskeleton, plays an important role in cell morphology, cytoskeleton reorganization, adhesion, invasion and metastasis. NGX6 can interact with ezrin via its cytoplasmic region [, ]. NGX6 and ezrin expression is negatively correlation in tissue from NPC biopsies, and the positive ratio of ezrin expression may be associated with the clinicopathological phase []. Ezrin is an important promoting factor in the development and metastasis of NPC. The positive expression ratio of ezrin in NPC patients is greater than that in non-NPC patients, and the positive ratio of ezrin in NPC patients with lymph nodes metastasis was much greater than that in NPC patients without metastasis. Ezrin can also bind to cell adhesion molecules such as CD44, CD43, CD46, ICAM-, ICAM- and ICAM-, and be coprecipitated with E-cadhein and β-catenin, all of which are implicated in cell migration and metastasis []. The interaction of NGX6 with ezrin suggests that NGX6 participates in cell migration and metastasis modulation. NGX6 plays an inhibitory role in the migration and invasion of NPC cells by interacting with ezrin and down-regulating the expression of ezrin and ezrin-related signaling molecules.NAG7, AN ESTROGEN RECEPTOR REPRESSOR, STIMULATES THE INVASIVE POTENTIAL OF HUMAN NPC CELLS BY REGULATING OF ERA EXPRESSION AND THE H-ras/p-c-RAF AND JNK/AP-/MMP1 SIGNALING PATHWAYSNAG7 (NPC-associated gene-) located on 3p25., also known as ERR- (estrogen receptor repressor-), is down-regulated in many NPC biopsy samples and in the NPC cell line HNE1 []. As an estrogen receptor repressor, NAG7 can interact with the estrogen receptor α (ERα), and reduce 17β-estradiol (E2) induced activation of ERα transcriptional activity in transient transfection assays of mammalian cells []. ERα is a nuclear transcription factor that regulates gene expression by binding to specific estrogen-responsive elements (ERE) on target gene promoters. ERα promotes breast cancer cell growth, but paradoxically, it also inhibits cancer cell invasion in vitro and metastasis in vivo [-]. ERα plays opposing roles in promoting cancer cell growth and inhibiting its invasion and metastasis. NAG7 is a negative regulator of ER, and has a double effect on proliferation and invasion of NPC cell lines. In NPC HNE1 cells, overexpression of NAG7 inhibits the proliferation by arresting cell cycle progression from G1 to S phase and inducing apoptosis [-]; but it increases the adhesion, motility and invasion of cells, both in vitro and in vivo, by down-regulating ERα expression in a E2-independent manner. NAG7 suppresses ER expression and stimulates cell invasion via the H-ras/p-c-Raf and JNK/AP-/MMP1 pathways []. The Ras proto-oncogene, a small GTP/GDP-binding protein, is a central component of mitogenic signaling and is a critical player in cancer progression. It affects the progression of human malignancies by increasing proliferation, enhancing invasion and altering cytoskeletal organization []. Activation of a linear pathway Ras–Raf–MAPK kinase (MEK) results in phosphorylation and activation of MAPK []. NAG7 overexpression in HNE1 cells increases the expression of H-Ras, which leads to the activation of p-c-Raf, while the expression of K-Ras, N-Ras and total c-Raf is unchanged []. NAG7 gene influences the H-ras/p-c-Raf pathway, leading to MAPK signal activation. MAPKs consist of ERK, c-Jun NH2-terminal kinase (JNK), and p38 cascades. JNK activity is essential for cancer proliferation, transformation, invasion and metastasis. Moreover, a positive correlation exists between JNK and lymph node metastases in breast cancer patients [-]. It has been reported that JNK may induce changes in the expression of c-Jun in hepatocellular carcinoma cells, resulting in increased expression and enzymatic activity of MMP1 []. Degradation of collagen is one of the most critical steps in the invasive and metastatic processes of cancer progression, and up-regulation of MMP- is associated with increased cellular invasion and poor prognosis in patients with colorectal, esophageal and advanced gastric cancers [, ]. MMP- may also promote cell adhesion and motility by influencing cytoskeletal arrangements through its association with different cell adhesion molecules such as CD44 and integrins []. NAG7 significantly increases the expression of JNK2 and c-Jun and down-regulated c-Fos, while leaving ERK and p38 unaffected. It also causes an increase in the transcriptional activation of AP-, which up-regulates the production of MMP- []. E2 has been reported to up-regulate MAPK phosphatases in breast cancer cells [], but NAG7 stimulates the JNK2/AP-/MMP1 pathway in an E2-independent manner. This suggests that NPC tumorigenesis does not involve estrogen stimulation [].LTF, A METASTASIS-ASSOCIATED PROTEIN, CAN NEGATIVE-REGULATE THE MAPK SIGNALING TRANSDUCTION PATHWAYLTF (Lactotransferrin, also referred to as lactoferrin, LF) is a secreted iron-binding glycoprotein that defends against microbial pathogens in innate immunity []. Recently, LTF has been found to have antitumor activity thus regulating tumorigenesis [-]. LTF is identified from NPC susceptibility locus in chromosome 3p21.-., which is a site of frequent loss of heterozygosity in NPC [, ]. The expression of LTF is downregulated in NPC biopsy samples [], and is negatively associated with the progression and metastasis of NPC. LTF is expressed at a significantly lower frequency at the T3/T4 stage or in NPCs with local lymph node metastasis compared to that of NPC at early stages or without metastasis []. LTF may act as a tumor suppressor, downregulating the progression and metastasis of NPC. Induction of LTF gene expression in NPC or modulation of cyclin D1, p21 and p27 expression, as well as Rb phosphorylation and signaling through the MAPK pathway may inhibit the development and progression of human NPC []. LTF treatment inhibits cyclin D1 expression and Rb phosphorylation, and increases levels of p21 and p27 expression. Upregulation of p21 and p27 may downregulate the expression of cyclin D1 and Rb phosphorylation contributing to NPC growth inhibition []. LTF treatment for or hr dramatically reduces expression levels of JNK2 and c-Jun in addition to ERK1/ phosporylation or c-fos expression, respectively. Extending treatment for longer periods resulted in a continued decrease in the levels of p-ERK1/ and c-fos expression in –8F NPC cells in vitro. However, LTF treatment does not modulate expression levels of total ERK1/, STAT3 and p53 and STAT3 phosphorylation []. Inhibition of NPC cell proliferation by LTF is partially mediated through modulation of the MAPK signal transduction pathway, but is independent of p53 and STAT3 signaling.THE INCREASED ACTIVITY OF p16 AND p27 BY SUPPRESSOR/SUSCEPTIBILITY GENES IS RELATED TO MULTIPLE SIGNALING CASCADESBoth p16 and p27 are cyclin-dependent kinase inhibitory proteins (CKI) that suppress activity and are frequently inactivated in NPC. p16 negatively regulates cyclin D1 activity to suppress CDK4, subsequently to control the G1/S checkpoint; and phosphorylated p27 allows the cdk2/cyclin E complex to remain activated allowing for cell cycle progression []. A majority of NPCs exhibit low level expression of p16 and p27 in dysplasia epithelium of nasopharynx. Suppressor/susceptibility genes increase activity of p16 and p27 by suppressing multi-signaling cascades, such as ERK/ MAPK, JNK/c-Jun, Rb/E2F.In conclusion, the development and progression of NPC involves in the alteration in expression of numerous oncogenes or tumor suppressor/susceptibility genes and the aberrations of a large variety of signaling pathways. According to functional genomics and signaling transduction pathways, we have described a signaling cross-talk network between the tumor suppressor/susceptibility genes involved in NPC (Fig. ). These tumor suppressor/susceptibility genes may be potential treatment targets for NPC in the future.
|
PMC2596794.txt
|
TITLE: Pharmaceutical induction of ApoE secretion by multipotent mesenchymal stromal cells (MSCs) AUTHORS: Suzanne Zeitouni, Brian S Ford, Sean M Harris, Mandolin J Whitney, Carl A Gregory, Darwin J Prockop ABSTRACT: BackgroundApolipoprotein E (ApoE) is a molecular scavenger in the blood and brain. Aberrant function of the molecule causes formation of protein and lipid deposits or "plaques" that characterize Alzheimer's disease (AD) and atherosclerosis. There are three human isoforms of ApoE designated ε2, ε3, and ε4. Each isoform differentially affects the structure and function of the protein and thus the development of disease. Homozygosity for ApoE ε4 is associated with atherosclerosis and Alzheimer's disease whereas ApoE ε2 and ε3 tend to be protective. Furthermore, the ε2 form may cause forms of hyperlipoproteinemia. Therefore, introduction of ApoE ε3 may be beneficial to patients that are susceptible to or suffering from these diseases. Mesenchymal stem cells or multipotent mesenchymal stromal cells (MSCs) are adult progenitor cells found in numerous tissues. They are easily expanded in culture and engraft into host tissues when administered appropriately. Furthermore, MSCs are immunosuppressive and have been reported to engraft as allogeneic transplants. In our previous study, mouse MSCs (mMSCs) were implanted into the brains of ApoE null mice, resulting in production of small amounts of ApoE in the brain and attenuation of cognitive deficits. Therefore human MSCs (hMSCs) are a promising vector for the administration of ApoE ε3 in humans.ResultsUnlike mMSCs, hMSCs were found not to express ApoE in culture; therefore a molecular screen was performed for compounds that induce expression. PPARγ agonists, neural stem cell conditioned medium, osteo-inductive media, dexamethasone, and adipo-inductive media (AIM) were tested. Of the conditions tested, only AIM or dexamethasone induced sustained secretion of ApoE in MSCs and the duration of secretion was only limited by the length of time MSCs could be sustained in culture. Upon withdrawal of the inductive stimuli, the ApoE secretion persisted for a further days.ConclusionThe data demonstrated that pre-treatment and perhaps co-administration of MSCs homozygous for ApoE ε3 and dexamethasone may represent a novel therapy for severe instances of AD, atherosclerosis and other ApoE-related diseases. BODY: BackgroundApolipoprotein E is a kDa secreted lipoprotein, first discovered by Shore and Shore in , see reference . Like other apolipoproteins, its primary function is to scavenge cholesterol, associated lipids and proteins for transport to the liver for processing. In healthy individuals, a plasma concentration of approximately mg/dL is sufficient to maintain systemic lipid homeostasis []. Although primarily expressed in liver [], ApoE contrasts with the other apolipoproteins in that it is also expressed in the spleen, lungs, kidneys, myoblasts, and macrophages [,]. ApoE is also highly expressed in the brain, by microglia and astrocytes [,] and there have also been reports of ApoE production by neurons []. Although poorly understood, the role of ApoE in the central nervous system is thought to be comparable to its role in plasma, sequestering cholesterol, lipids and other macromolecular debris from neural tissue [].The effectiveness of ApoE as a lipid scavenger is closely related to the ApoE isoform. ApoE polymorphism is thought to be uniquely exhibited by humans [] in that it consists of three isoforms, ApoE ε2, ε3, and ε4. Although the isoforms vary only at two amino acid residues, the changes are sufficient to profoundly alter the tertiary structure and affect lipid binding properties of the protein [,]. ApoE polymorphism occurs at residues and , ApoEε2 has two cysteines, ApoEε3 had a cysteine and an arginine and ApoEε4 has two arginines at these positions. ApoE ε3 is the most common isoform with an incidence of approximately % in the population [,,], followed by ε4 in approximately %, and then ε2 at approximately %.Due to the profound effects on the functional and structural properties of the protein, different isoforms of ApoE are associated with different diseases such as stroke, multiple sclerosis, Parkinson's disease, alcoholic cirrhosis of the liver and type diabetes [,-]. The role that ApoE isoforms play in disease development seems to be closely associated with different affinities of the various isoforms for the low density lipoprotein-receptor and lipids themselves. For instance in the case of Alzheimer's disease (AD), individuals that are homozygous or heterozygous for ApoE ε4, have a significantly increased risk of developing the disease with an earlier onset [] compared with individuals with the most common ApoE ε3/ε3 genotype [,,,-]. Likewise, individuals homozygous for Apo ε2 or with an ApoE ε2/ε3 genotype also have normal susceptibility for AD. Although ApoE4 is not an absolute determinant of the disease –% of AD patients have at least one copy of the allele. However, the role of ApoE isoforms is complex with additional reports also implicating ApoE ε2 with hyperlipidaemia and ApoE ε4 with hypercholesterolaemia. [,]. In consideration of the putative association of ApoE ε2 and ε4 with disease it is reasonable to assume that safe administration of ApoE ε3 could provide beneficial effects for some individuals.Multipotent mesenchymal stromal cells (MSC), also known as marrow stromal cells or mesenchymal stem cells, are a heterogeneous population of non-hematopoietic cells representing .–.% of the total bone marrow []. The in vivo function of MSCs remains controversial, but they are generally regarded as hematopoietic support cells and also a source of progenitors for structural tissues [].The existence of such cells was first suggested by Cohneheim in the 1870s [] but Friedenstein and his colleagues were the first to isolate and culture MSCs in the 1970s [-], followed by Caplan and colleagues [,] who were the first to propose the phrase "mesenchymal stem cells". These cells are adherent to plastic and are known to differentiate to osteoblasts, chondrocytes and adipocytes in vitro and in vivo [,,,]. However, more recent findings suggest that MSCs may differentiate to additional cell types []. Another remarkable characteristic of MSCs are their immunosuppressive qualities, suggesting that they may survive and engraft in allogeneic recipients []. Because MSCs are easily expanded in culture, readily engraft into host tissues, and may locally modulate the immune response, they represent ideal candidates for cytotherapy. Specifically, MSCs represent an extremely useful tool for the treatment of deficits in ApoE function if harbouring the ApoE ε3 isoform.In a previous study we demonstrated that murine MSCs (mMSCs) injected into the lateral ventricles of day old ApoE null mice secreted small amounts of ApoE into the tissue []. As ApoE null mice mature, the animals develop cognitive impairment, but upon behavioural analysis, the presence of mMSCs resulted in improved cognitive behavioural testing compared to the control groups []. Therefore, human MSCs (hMSCs) seemed a promising vector for the administration of ApoE ε3 in human diseases. In contrast to mMSCs, unmanipulated human MSCs (hMSCs) do not synthesize ApoE mRNA in vitro, but expression could be observed after days of adipocyte differentiation []. Pharmaceutical induction of ApoE secretion has been demonstrated in macrophages, hepatocytes, adipocytes and other cell types by a variety of agents. In two studies employing rat hepatocytes, ApoE secretion was stimulated using dexamathasone, insulin, or a combination of both [,]. In macrophages, secretion of ApoE was induced by treatment with TGFβ and dexamethasone []. Expression of ApoE mRNA by adipocytes was increased by pioglitazone and ciglitazone, neither of which affected the ApoE secretion of macrophages [].This study examined pharmaceutical induction of endogenous ApoE expression by hMSCs. We demonstrate here that dexamethasone alone or adipo-inductive media (AIM) containing dexamethasone, indomethacin and isobutylmethylxanthine resulted in the expression of high levels of ApoE by hMSCs in vitro. Maximal expression could be attained after approximately – days of dexamethasone or AIM treatment depending on the donor source of the cells. We found no correlation between the ApoE expression levels and the degree of cell expansion prior to the assay, nor did we find any correlation between sex or age and ApoE secretion kinetics. The maximal rate of secretion ranged between .–. ng cell- day-. After withdrawal of the stimulus, ApoE expression remained approximately days, but sometimes much longer, depending on the donor. These in vitro results demonstrate that ApoE expression by hMSCs is entirely possible through pharmaceutical induction without the necessity for genetic manipulation. Therefore, MSCs may represent a safe and feasible strategy for treatment of diseases that result from a functional deficit of ApoE.ResultsCulture and characterization of human MSCsWe employed MSCs from a total of eight donors for the study. Human MSCs were recovered from the mononuclear fraction of whole bone marrow and cultured as described in the Materials and Methods. The adherent component of the cultures, containing hMSCs is relatively pure, but frequently contains traces of contaminating osteoblasts, fibroblasts and senescent cells. To assay for enrichment of multipotent hMSCs, we performed differentiation assays to osteoblasts, adipocytes and chondrocytes. The hMSCs readily differentiated into all three lineages when subjected to the appropriate conditions (Figure ). MSCs formed mineralized osteoblasts as detected by the calcium binding dye, alizarin red S; they formed adipocytes with oil red O stainable lipid droplets; and formed proteoglycan-filled cartilage pellets that stained purple with toluidine blue when subjected to pellet culture in the presence of bone morphogenic protein and tumor necrosis factor β.Figure 1Differentiation of human MSCs. Characterization of MSCs by differentiation assays. One donor is presented. Row , high (right) and low power (left) micrographs of osteogenic MSCs stained with Alizarin Red S for calcium. Row , high (right) and low power (left) micrographs of adipogenic MSCs stained with oil red O for lipid. Row , high (right) and low power (center and left) micrographs of toluidine blue (purple) stained sections of MSC pellets induced to form chondrocytes and cartilage. All donors were assayed for adipogenesis and osteogenesis with the exception of chondrogenesis, where donors were assayed.Induction of ApoE productionUnmanipulated hMSCs did not secrete detectable levels of ApoE when the conditioned culture media was assayed by ELISA as shown in the controls in figures , , and . Since previous studies have reported successful induction of ApoE secretion from other cell types, we decided to screen some candidates for ApoE inducing activity. A total of nine conditions were tested; . μM dexamethasone (Dex), neural stem cell conditioned media (NSC-CM), AIM, osteogenic media (ost), μM ciglitazone (cig ), μM ciglitazone (cig ), μM troglitazone (trog ), μM troglitazone (trog ). Confluent MSC monolayers from donors were cultured in the experimental conditions for days. At day increments, conditioned medium was recovered for ApoE ELISA. Of the conditions tested, five induced ApoE expression after days of treatment; these were Dex, NSC-CM, AIM, ost and cig (Figure 2a). Cig transiently induced ApoE secretion by day of culture but these levels were absent at subsequent time points. Ost maintained ApoE expression to day , the end point of the experiment, but the levels remained very low. Since mMSCs were shown to secrete ApoE in brains of mice [], we tested conditioned media from murine neural stem cells. Transient expression of ApoE was induced after days of culture, but these levels were undetectable at later time points. Since the ELISA detects only human ApoE, the levels were derived from the hMSCs rather than trace levels in the conditioned media itself. In contrast, Dex and AIM induced high and robust expression of ApoE, which persisted until the end of the experiment at days.Figure 2a – Assays of ApoE production. Some conditions known to induce ApoE production in other cells and their effect upon MSCs. NSC conditioned media AIM and osteogenic conditions are also included. Experiment was repeated with donors and data from one of the donors is presented, n = , average ± standard deviation is shown. b – Detection of ApoE in MSCs by western blotting. Western blot demonstrating ApoE ( kD) production (upper) in early (passage , left column) and late (passage , right column) passage MSCs. Lanes were normalized to GAPDH expression (lower). Experiment was repeated donors. c – Dex-mediated expression of ApoE in MSCs is not accompanied by adipocyte differentiation. ApoE expression by MSCs was induced by exposure to Dex or AIM. After days, the monolayers were stained with oil red O to visualize lipid droplets (left). Lipid droplets were evident in AIM treated, but not Dex treated MSCs. However, ApoE expression was high in both cases when media was assayed by ELISA (right).Figure 3ApoE production kinetics. ApoE production during days of treatment with . μM dexamethasone (panel a) or adipoinductive media (panel b). Experiment was repeated for donors, data from one donor is presented, n = , error bars are standard deviations.Figure 4ApoE production following withdrawal of inductive media. MSCs were treated with the inductive media (. μM dexamethasone (Dex) or AIM (Adipo)) through for days, then with CCM for a further weeks. ApoE levels drop abruptly after day . Experiment was repeated donors, data from one donor is presented, n = , error bars are standard deviations.Figure 5Dose responses. Dexamethasone dose response (high dose, panel a, and low dose, panel b). Experiments were repeated with donors, data from one of the donors is presented, n = , error bars are standard deviations.Figure 6ApoE production per cell. Rate of ApoE production normalized to cell number. Cells were treated with inductive media (. μM dexamethasone (Dex) or AIM (Adipo)) for – days. Experiment was performed on donor, n = , error bars are standard deviations.To confirm ELISA data, ApoE was also detected in protein extracts from MSCs treated with Dex, AIM and control media, containing an appropriate level of vehicle DMSO, by western blotting. A single band ( kDa) corresponding to ApoE was present only in the Dex and AIM treated MSCs (Figure 2b). The position of this band was later confirmed by comparison to immunoblotted recombinant human ApoE (data not shown). To confirm that Dex was solely responsible for the induction of ApoE secretion, we tested whether the two additional components of AIM induced ApoE expression. Isobutyl-methyl xanthine, a phosphodiesterase inhibitor, and indomethacin, a ligand of the adipogenesis master-regulator peroxisome proliferator-activated receptor, both failed to induce ApoE expression alone or in combination, nor did they induce terminal adipogenesis (data not shown). Furthermore, Dex did not appear to induce ApoE as a by-product of terminal adipogenesis, since lipid droplets could not be detected in long term treated cultures (Figure 2c). Thereafter, Dex and AIM were the selected conditions for subsequent experiments.Kinetics and Maintenance of ApoE productionTo examine the kinetics, longevity, and donor-dependency of the inducible ApoE secretion, a series of time course experiments were performed with hMSCs from eight donors. In the first series of experiments, we examined the rate of response to Dex and AIM in terms of attaining maximal ApoE secretion. Confluent monolayers of hMSCs were established from passage (approximately doublings per cell) cultures. Cells were treated with Dex or AIM and for controls, the cells were treated with either vehicle (DMSO) or not at all. At – day intervals, media were recovered and ApoE levels were measured by ELISA (Figure ). For all MSC preparations tested the control cultures did not secrete ApoE over the entire experimental period (data not shown). For the AIM and Dex conditions, maximal ApoE expression was attained after days and was sustained for up to days. After long term culture (about days), the ApoE yield dropped due to apoptosis of the MSCs that frequently occurs after sustained culture at confluency. The ApoE yields were comparable between all of the MSC preparations at about – ng day- for the Dex (Figure 3a) and – ng day- for the AIM conditions (Figure 3b). There were no correlations between the rate of response to stimulus and donor age or sex. When stimulated by Dex only, there was a linear increase in the rate of ApoE secretion from hMSCs. However, there was an exponential increase in the rate of ApoE secretion when treated with AIM and this was accompanied by differentiation into adipocytes. Since adipocytes are a source of ApoE in vivo, this probably accounts for the higher overall ApoE output, and the accelerated rate of secretion.The next study was designed to address the longevity of ApoE secretion after withdrawal of stimulus. Confluent passage MSC cultures were incubated in Dex and AIM media for days to attain maximal expression of ApoE. The cells were then cultured in the absence of stimulus for an additional weeks (Figure ). After weeks, ApoE expression dropped to about % of the maximum for AIM treated cells and the Dex treated cells expressed barely detectable levels. This suggested that for sustained ApoE expression, induction is required at least weekly.In consideration of the prospect of clinical application, we examined the dose dependency of ApoE induction on MSCs. Although AIM media was most effective in inducing ApoE secretion by MSCs, the compounds have not been safely co-administered in vivo. On the other hand, dexamethasone is a standard therapeutic and would be acceptable for administration to humans at low doses. We performed dose response curves ranging from . μM to μM and . μM to . μM found that the dose of Dex could be reduced from . μM to . μM with insignificant reduction in ApoE secretion (Figure 5a). The lower dose response series demonstrated that approximately half maximal ApoE expression could be attained (Figure 5b).ApoE production and cell numberTo provide insights on cell viability and proliferative potential during induction, we examined ApoE secretion on a per cell basis. MSCs were allowed to reach % confluency before beginning treatment with AIM or Dex. Media and cells were collected at day , and . At each time point, the amount of ApoE secreted per day was normalized to the number of cells in the monolayer. Overall cell recovery was slightly reduced when compared to the untreated controls suggesting that DMSO primarily affected MSC proliferation (data not shown). However, the presence of Dex and AIM did not affect long term viability, since the cultures slowly expanded over the day experimental period. In the case of AIM, normalized ApoE levels reflected the unmodified measurements (Figure 3b) with a continuous increase in ApoE levels over time suggesting that as the cells divide in culture, they retain their ApoE secretion potential. In the case of the Dex conditions, normalized ApoE levels dropped slightly, suggesting expansion of some cells with limited or no potential for ApoE secretion (Figure ).We then tested the ability of the ApoE to bind to VLDLs. To reduce background, ApoE expression was induced for days by Dex treatment in serum free media. MSC viability was not significantly affected during the brief exposure to serum free conditions and ApoE expression occurred at levels comparable to expression in serum containing media (Figure 7b). The conditioned media was then incubated with biotinylated VLDLs and then subjected to a series of biotin-mediated depletions using streptavidin coated microtiter plates (Figure 7a). Upon ELISA of the depleted media, ApoE levels were significantly reduced when compared to controls that were treated identically, but lacked VLDL (Figure 7b). This suggests that the MSC-derived ApoE meets at least one of its in vivo functions. ApoE also binds to the LDL receptor and facilitates internalization by hepatocytes, macrophages and astrocytes. To examine the potential of MSC-derived ApoE to accelerate lipid uptake, we generated synthetic micelles containing fluorescently labelled cholesterol ester and free cholesterol. When such lipid micelles were added to serum free media conditioned by the ApoE (Figure 8a), it bound to them and could be enriched from the media by centrifugation (Figure 8b). When compared to untreated controls, the ApoE conditioned media catalysed the formation of larger micelle aggregates, many of which could be identified by the naked eye (data not shown). Furthermore, ApoE conditioned media accelerated the uptake of the fluorescent lipid aggregates by huh- hepatocytes (Figure 8c).Figure 7Binding of ApoE to VLDL. ApoE was secreted into serum free media by MSCs. The conditioned media was then added to biotinylated VLDLs (panel a). After biotin depletion, the remaining ApoE in the conditioned medium was measured by ELISA (panel b). When compared with buffer alone, biotinylated VLDLs depleted the ApoE from the conditioned medium. Data from one of two donors is presented, n = , error bars are standard deviations.Figure 8Binding of ApoE to synthetic micelles and uptake by hepatocyte cells. ApoE was secreted into serum free media by MSCs. Fluorescent cholesterol-ester and cholesterol was then added to the media (panel a). Upon recovery of the micelles by centrifugation, ApoE could be detected when analyzed by western blotting (panel b). When ApoE conditioned media was mixed with lipids and incubated with huh- hepatocarcinoma cells, it accelerated uptake when compared with control media. Phase contrast/fluorescent merge indicating hepatocytes taking up fluorescent cholesterol (green) from the medium (panel c).DiscussionThere are a number of reports where ApoE has been successfully administered via direct gene delivery or virally in a variety of disease models including Alzheimer's disease [], hypercholesterolemia [-], hyperlipidemia [], atherosclerosis [-], experimental stroke [] and hematopoietic diseases []. Direct gene introduction or direct viral therapy can be efficient and sustainable in some cases, but it is associated with safety concerns. To circumvent some of these problems, ApoE has been administered through hematopoietic stem cells (HSCs) in rodent models [-]. The HSCs are administered to radioablated animal models and long term hematopoietic stem cells reconstitute the bone marrow compartment. In some studies, ApoE is introduced to the HSCs virally, and in others, the inherent ApoE is utilized mostly through expression by resultant macrophages. These studies suggest that bone marrow transplantation is a promising vector for cytotherapeutic ApoE administration but the process of radioablation in humans is a dangerous procedure, and graft versus host disease is a common problem with allogeneic recipients. Viral integration, especially with long term repopulating HSCs is also a source of concern for these strategies.In consideration of all of these caveats, a safer cytotherapeutic vector would be beneficial. MSCs are safely recovered from humans via a simple iliac crest aspirate and can be expanded by the million [,]. Their characteristics suggest that they would be an excellent delivery tool for ApoE; for instance, when administered into animal models, they survive for long periods in the brain (up to months) and subcutaneously [,-]. When in the brain, hMSCs secrete trophic factors that stimulate neural stem cell proliferation and probably their own survival []. Upon subcutaneous implantation, the hMSCs distribute themselves close to capillary beds suggesting that they have the capacity to secrete proteins into the blood []. MSCs are also immunomodulatory, with the capacity to suppress a variety of rejection mechanisms [], reducing the possibility of graft versus host complications, and improving the probability of allogeneic transplant. Finally, transplanted hMSCs do not rapidly proliferate in adult tissues in the same way repopulating HSCs do, suggesting that they may also be safer for accommodating viral transgenes.In our previous study [], mMSCs were administered to the lateral ventricles of ApoE null mice. As they age, untreated ApoE null mice develop cognitive deficits that resemble Alzheimer's disease in addition to profound atherosclerotic pathogenesis. Remarkably, upon administration of the murine MSCs, the mice exhibited signs of improvement in some of the behavioural tests. When the brains of the experimental animals were examined, a small amount of murine ApoE could be detected. Although the presence of other mMSC derived factors could have contributed to functional recovery, the presence of ApoE months after MSC administration strongly suggested that the functional deficit of ApoE in the null mice had been attenuated. We therefore hypothesized that hMSCs may provide a promising vector for administration of the ApoE3 isoform for the treatment of Alzheimer's disease and possibly atherosclerosis.We could not detect ApoE expression in vitro, when cultured under standard conditions, nor could we detect expression when cultured in murine neural conditioned medium. This was surprising, since administration of murine MSCs in vivo, stimulated expression. We hypothesize that the ApoE null environment may have been sufficient to stimulate expression or more likely, the hMSCs required contact with the neural cells to initiate expression. Due to the limited availability of human neural tissue, and in view of potential inter-species incompatibility, we decided to attempt induction of expression using commonly available pharmaceuticals. This approach has been utilized with a variety of cell lines and a variety of reagents [-] and also in MSCs when differentiated completely into adipocytes []. Administration of adipocytes directly into the brain cannot be performed, therefore we tested related conditions in an attempt to induce ApoE expression without lipid droplet accumulation that occurs during adipogenic differentiation. In contrast with previous literature where adipocytes were employed [], PPARγ agonists gave very limited success (Figure 2a) demonstrating that the compounds function in a cell type-dependent manner. Another series of studies demonstrated efficacy with the hydrocortisone analog, dexamethasone on murine macrophages and rat hepatocytes [-]. We therefore tested dexamethasone at the equivalent concentration employed for adipogenic differentiation and we achieved high levels of ApoE expression without complete differentiation into adipocytes. Indomethacin and Dex also induced ApoE secretion, but we also observed high levels of unwanted adipogenesis (data not shown). We acquired the genotype of of the donors and found them to be Apo ε3/ and Apo ε2/. These donors had comparable ApoE expression levels when compared with the other uncharacterized MSC preparations. This confirmed that MSCs with preferable Apo ε3/ genotype would be compatible with the Dex induction strategy.Permanent induction of ApoE would be beneficial, dismissing the need for constant drug administration. However, withdrawal of both Dex and AIM caused a reduction of ApoE output over weeks. Although it is unclear whether this is due to outgrowth of a non-expressing component of MSCs or due to attenuation of expression, the extent of proliferation required to completely ablate expression is not expected in these confluent cultures. Of interest, is the reduced expression of ApoE in the AIM treated cultures, which, by day have mostly differentiated into lipid-filled adipocytes. It appears that ApoE is not a prerequisite for maintenance of the terminal adipocyte phenotype. We did not observe a qualitative reduction in lipid filled cells suggesting that dedifferentiation had not occurred, but this phenomenon cannot be completely discounted.We found that doses as low as . μM were sufficient to maintain maximal expression of ApoE, and doses as low as . μM could sustain half maximal levels of ApoE expression. These levels are comparable with published therapeutic plasma levels of dexamethasone, which are in the region of . – . μM for immunomodulation []. These results suggest that relatively safe plasma levels that can be attained clinically would be sufficient for ApoE induction by implanted MSCs.Nevertheless, chronic Dex administration is not without serious side effects [] such as immunological inhibition, diabetes/hyperglycemia, osteoporotic symptoms, gastric irritation, weight loss or gain, glaucoma, muscle pain/weakness and exacerbation of psychosis. The side effects increase with duration and dose. Since expression of ApoE was maintained for at least days by hMSCs after withdrawal of stimulus, it is possible that less frequent administration of Dex (every days, for instance) may maintain ApoE levels while reducing the probability of harmful effects. Furthermore, corticosteroids with faster clearance and shorter effect durations, such as prednisone or hydrocortisone may improve the risk/benefit ratio [].In terms of efficacy, we found that maximal ApoE expression per cell was in the region of . ng per cell in hr when treated with Dex. Since the plasma concentration is about mg L- [], and the average blood volume in an adult human is . L, it would take an implant containing million cells about days to attain % of normal systemic ApoE levels, a dose that is protective against atherosclerotic plaque formation []. However, systemically infused MSCs have the capacity to migrate to sites of injury and inflammation, suggesting that the local dose of ApoE might be much higher at the lesions [,,]. Direct injection of a lower dose of MSCs into the brain may provide long term relief of Alzheimer's disease while stimulating the local population of neural stem cells to repair existing damage []. In healthy humans, the blood brain barrier (BBB) is somewhat resistant to Dex penetration due to multidrug resistance receptor action [] and this is a potential problem for treatment of Alzheimer's disease. There are, however, alternative corticosteroids, such as prednisolone, which are predicted to have similar effects and have been reported to have increased permeability into the cerebrospinal fluid [,-]. Furthermore, there is extensive evidence that the BBB is compromised in individuals with Alzheimer's disease suggesting that Dex may have access to implanted MSCs in severe cases [-].It can be argued that MSCs may not require pharmaceutical ApoE induction at all since there have been reports that they express various neuronal markers in vitro [] and in vivo [,,] and thus may spontaneously differentiate into astrocytes in the brain. Indeed, murine MSCs were shown to express nestin, β-III tubulin, neurofilament marker and GFAP (an astrocyte marker) in vitro [], and GFAP, β-III tubulin, and neurofilament in vivo after transplantation into mice [,]. Human MSCs were also shown to express GFAP []in vivo. However, it is noteworthy that in every instance where MSCs were implanted into the murine brain, only a small proportion expressed the astrocyte marker, GFAP. Therefore, although some hMSCs may differentiate into ApoE-secreting astrcocytes, it remains important to pre-treat the hMSCs with Dex to ensure that the majority of the cells secrete ApoE immediately upon implantation.Since functional ApoE has been expressed in E. coli [] and in insect cells [], it is likely that the MSC-derived ApoE is also functional. Nevertheless, we confirmed that the protein could associate with cholesterol esters and could also bind to VLDLs (Figure and ). MSC-derived ApoE also accelerated uptake of lipid by hepatocytes (Figure ). Since the physiological role of ApoE is dependent binding of lipid and the LDL-receptor, these data suggest that the ApoE would satisfy its role in vivo. Of particular note, is the observation that MSCs produce high levels of ApoE in the absence of serum (Figure 7a). This raises the possibility of generating ApoE preparations for therapeutic use without the necessity for MSC administration. Since MSC conditioned media inherently contains neuroprotective cytokines [], dialysed conditioned media containing ApoE may represent an efficacious neuroprotective cocktail for clinical use.ConclusionIn this study we have shown that expression of potentially therapeutic levels of functional ApoE by hMSCs can be induced with dexamethasone. These data demonstrate that co-administration of hMSCs genotyped for homozygous expression of ApoE ε3 and chronic cortisol treatment may represent a novel therapy for severe instances of ApoE related diseases.MethodsCulture of human multipotent stromal cellsMutlipotent stromal cells were acquired from the Tulane University adult stem cell distribution facility. In accordance with institutional review board approved protocols, the cells were prepared from mL posterior iliac crest bone marrow aspirates derived from males and female between and years of age. After a brief interval of monolayer culture to exclude non-adherent hematopoietic cells, hMSCs were expanded to % confluency prior to passage or use in experiments. Cells were cultured according to standard MSC culture conditions in complete culture medium (CCM), consisting of alpha minimal essential medium (GIBCO, Invitrogen, Carlsbad, CA) containing % (v/v) FBS (Hyclone, Logan, UT and Altanta Biologicals, Norcross, GA), mM L-glutamine and units ml- penicillin and μg ml- streptomycin (GIBCO, Invitrogen) [,].Differentiation of human multipotent stromal cellsTo confirm that the cell preparations from the mononuclear layer were multipotent, a panel of differentiation assays were performed.Osteogenic differentiation and Alizarin Red S stainingFor osteogenic differentiation, confluent monolayers of hMSCs were incubated in CCM supplemented with - M dexamethasone, μg mL- ascorbic acid and mM β-glycerol phosphate (Sigma, Poole, UK) for days with changes of medium every days. The monolayers were then stained with mM Alizarin Red S pH . (Sigma) for min and washed times with distilled water. Micrographs were taken using an inverted microscope (Nikon Eclipse, TE200).Adipogenic differentiation and Oil Red O stainingAll reagents were purchased from Sigma. Confluent monolayers of hMSCs in well plates ( cm2 per well) were incubated in adipoinductive medium (AIM) consisting of CCM containing . μM dexamethasone, × - M isobutylmethylxanthine, and × - M indomethacin (Sigma). Media was changed every – days. After days, the adipogenic cultures are fixed in % formalin for min and stained with fresh Oil Red-O solution . (w/v) (Sigma) in % (v/v) isopropanol in phosphate buffered saline (PBS) for min. The dishes were washed times with excess PBS and visualized using an inverted microscope (Nikon Eclipse, TE200).Chondrogenic differentiation and processingMicromass pellet chondrogenic differentiation was carried out in accordance with the protocol of Sekiya et al. on , pelleted cells []. After days of differentiation, the chondroid pellets were washed in PBS and fixed in % paraformaldehyde. The pellets were then embedded in paraffin, sectioned, then stained with toluidine blue to visualize sulphated proteoglycans and chondrocyte lacunae [].ELISA detection of ApoESample preparation for ELISAhMSCs were plated in each well of a six well plate at cells per cm2 and grown to % confluency. The media was then changed to the appropriate conditions and maintained for the appropriate duration with changes every – days. All experiments were performed in triplicate. Media samples were taken at intervals defined in the Results and stored for ELISAs at -°C. Dexamethasone and dimethyl sulphoxide (vehicle) were purchased from Sigma and diluted into CCM from a × stock solution. AIM and osteogenic media were prepared as described above and the individual components were added to CCM from stocks. For production of NSC conditioned medium, a frozen vial of murine neural stem cells (NSCs, a gift from Dr Jeffrey Spees, Department of Medicine, Cardiovascular Research Institute, University of Vermont) was employed. NSC conditioned medium was produced by incubation of the NSCs in NSC media [neurobasal alpha medium containing B27 supplement, L-glutamine, penicillin, streptomycin (Invitrogen), ng mL- epidermal growth factor, ng mL-, fibroblast growth factor (Sigma)] for days. NSC conditioned medium was mixed : with CCM prior to addition to the hMSCs. Ciglitazone was purchased from Tocris Biochemical (Bristol, UK) and troglitazone was purchased from Cayman Chemical (Ann Arbor, MI). The drugs were added to CCM from a stock dissolved in DMSO. All assays were conducted in parallel with vehicle, or unconditioned media controls.Human ApoE sandwich ELISAA polyclonal goat-anti-human ApoE antibody (Academy Biomedical, Houston, TX) was diluted in PBS at :. Each well of a high binding microtiter plate (Fisher Lifesciences) was coated with μL of the antibody solution for hrs at °C. The coating solution was then removed followed × min washes with μL PBS. Wells were then blocked by addition of PBS containing .% (v/v) Tween20 (Fisher Lifesciences) and % (w/v) bovine serum albumin (Sigma) for hrs at room temperature. Block solution removed and the wells were washed for × min in PBS containing .% (v/v) Tween20 (PBST). Media samples ( μL) were added to the wells in triplicate. Standard solutions of recombinant human ApoE (Calbiochem, Gibbstown, NJ) were diluted in PBST. Samples were incubated in the microtiter plates for hrs at °C. Plates were then washed for × for mins in PBST, and the detection antibody ( μL) was added consisting of a : dilution of HRP conjugated goat anti human ApoE IgG (Academy Biomedical) in PBST at room temperature. After hrs, the wells were washed in PBST and developed by addition of μL TMB (Pierce, Rockford IL).After approximately mins, the reactions were stopped with the addition of μL N sulphuric acid (Fisher Lifesciences) and the resultant yellow substrate was measured by absorbance ( nm) on a well plate reader (Fluostar Optima, BMG).Cell number evaluationCells were counted using a DNA intercalating dye that fluoresces upon incorporation (CyQuant dye, Invitrogen). Briefly, monolayers were recovered by trypsinization, washed in PBS, then resuspended in lysis buffer (PBS containing .% Triton ×, mM MgCl2) containing a fold dilution of the CyQuant dye. At various dilutions, the labelling suspensions were aliquoted into a microtiter plate and fluorescence was measured at / nm on a well plate reader (Fluostar Optima) using opaque black plates (Fisher Scientific). Fluorescence was proportional to DNA content, and thus cell number, when compared to known cell standards.ApoE detection by western blottingSample preparationMSCs were plated on cm plates at cells per cm2 and grown to % confluency. The cells were then maintained under experimental conditions for days and recovered by scraping. The media was collected for ELISAs. Cells were washed in PBS, flash frozen in liquid nitrogen and then stored at -°C for protein extraction. The proteins were solubilized using an extraction solution consisting of PBS containing mM MgCl2, % (w/v) SDS (Sigma), .% Triton X- (Fisher Lifesciences) and fold protease inhibitor cocktail (Roche Diagnostics, Nutley, NJ). Protein yield and quality was evaluated by gel electrophoresis (Novex electrophoresis system, Invitrogen) followed by silver staining (Invitrogen).Western blot assayApproximately μg of protein were added to the appropriate volume of × LDS-PAGE sample buffer (Invitrogen) containing mM β-mercaptoethanol and M urea (Sigma Aldrich). The samples were heated at °C for mins and electrophoresed on a % NuPage bis-Tris gel using the MOPS buffering system followed by transferral to PVDF for min (iBlot, Invitrogen). Filters were blocked in PBS-T containing % (w/v) powdered milk (Santa Cruz Biotechnology, Santa Cruz, CA) for hrs. For detection of ApoE the blots were incubated overnight in an HRP conjugated goat anti human ApoE (Academy Biomedical) at : in block buffer. For detection of GAPDH, blots were probed with a monoclonal antibody at a dilution of : (clone 6C5, Chemicon, Temecula, CA). For secondary detection, a rabbit anti-mouse IgG antibody coupled to horse radish peroxidase was used at a dilution of : (Sigma). The blots were developed in peroxidase substrate ( mM Tris pH . containing μM paracoumaric acid, μM luminol and .% (v/v) hydrogen peroxide, Sigma) for mins prior to exposure to photographic film (Pierce).VLDL binding assayMSCs were transferred to serum free complete culture medium containing Dex or vehicle. After days, the conditioned medium was filtered through a . μm membrane. One mL of the conditioned medium was added to μg of biotinylated human VLDL solution (Intracel, Fredrick, MD). The mixture was incubated for hr with rotation, and large VLDL aggregates were removed by centrifugation at , g for min. The mixture was then depleted of biotinylated components by sequential min incubations in wells of a streptavidin coated microtiter plate (Streptawell High Bind, Roche Diagnostics). The supernatant was then subjected to ELISA assay.Production and characterization of lipid micelles containing labelled cholesterol ester, cholesterol and MSC derived ApoEConditioned serum free medium containing – ng mL- ApoE was incubated in μg mL- ,-difluoro--(-pyrrolyl)--bora-3a,4a-diaza-s-indacene--undecanoate (cholesteryl BODIPY / C11, Invitrogen), and μg mL- -[N-[(-nitro--,-benzoxadiazol--yl)methyl]amino]--norcholesterol (NBD cholesterol). Cholesteryl BODIPY / C11 is a blue cholesteryl ester and NBD cholesterol consists of a cholesterol moiety covalently conjugated to a green fluorophore /. Lipids were allowed to self-assemble at °C with slow mixing. In some cases, the micelles were recovered by centrifugation, washed in ice cold PBS and assayed for ApoE content by western blotting. For cell binding studies, Huh- hepatocytes (a gift from Srikanta Dash, Tulane Health Sciences Center) were expanded in well plates containing standard media at °C with % (v/v) CO2. Upon initiation of the experiment, mL of micelle-containing serum free medium was added to the cultures and at hourly intervals, cultures were washed in warm PBS, then the medium was replaced by serum free medium without lipid. Cells were visualized by epifluorescent and phase microscopy. Kinetics of the uptake of NBD cholesterol was compared between ApoE containing and vehicle containing control media.Statistical AnalysesMeasurements were performed in triplicate for each media sample taken and measurements were considered acceptable when the variation was less than % of the value. For each condition tested, – replicate cultures were prepared and assayed. Data were presented as the mean of the measurements from replicate cultures with standard deviations. Representative data are presented from one donor in figures, but multiple donors were assayed yielding the same results.Authors' contributionsSZ Conceptualized experimental strategy, carried out experiments, interpreted data, co-wrote manuscript. BSF Carried out experiments, interpreted data. SMH Carried out experiments. MJW Prepared some of the MSCs, carried out experiments. CAG Conceptualized experimental strategy, interpreted data, co-wrote manuscript. DJP Conceptualized experimental strategy, interpreted data, co-wrote manuscript. All authors read and approved the final manuscript.
|
PMC3024224.txt
|
TITLE: Buckle up safely: a cluster randomised trial to evaluate the effectiveness of a pre-school based program to increase appropriate use of child restraints AUTHORS: Rebecca Q Ivers, Lisa Keay, Julie Brown, Lynne E Bilston, Kate Hunter, Judy M Simpson, Mark Stevenson ABSTRACT: BackgroundRoad traffic crashes for car occupants are a leading cause of death and serious injury in children from high and middle income countries globally. Correct use of appropriate child restraints can significantly reduce death and serious injury but there is a need for well powered trials to examine effectiveness of programs to increase optimal child restraint practices. The aim of this trial is to examine the effectiveness of a comprehensive intervention to increase the use of appropriate child restraints, and decrease incorrect use of child restraints in pre-school aged children traveling in cars.Methods and DesignA cluster randomised controlled trial will be conducted, involving pre-school or childcare centres in low income areas of Sydney, Australia, over one calendar year. The intervention is an educational program involving an in-service for centre staff, distribution of educational materials to parents, a parent workshop demonstrating restraint use, subsidised restraints for parents in need, and vouchers for a free restraint checking service. Blinded assessors will observe restraint use at all centres at the end of the calendar year. Data will be analysed on an intention-to-treat basis; the primary analysis will compare the proportion of each of the two outcome measures (use of appropriate restraints, and incorrect use of restraints) at each centre between intervention and control groups. Detailed process evaluation will be conducted, including examination of implementation and utilisation of various elements of the program by both centres and families.DiscussionThis assessor blinded cluster randomised trial is powered to provide credible evidence about the efficacy of an education and distribution program in a pre-school setting to increase appropriate use, and decrease incorrect use of child restraints.Trial registrationAustralian New Zealand Clinical Trials Registry ACTRN12609000612213. BODY: BackgroundRoad traffic crashes are a leading cause of death and serious injury in children in high income countries worldwide, accounting for % of all child injury deaths []. Road traffic crashes for car occupants are a leading cause of death and serious injury in Australian children. There were passengers under years killed in traffic crashes between and in Australia[], and about children aged - are seriously injured each year as car occupants[], with about a third of these aged - [].The effectiveness of child restraint systems in preventing injury for children involved in traffic crashes is well documented. Data from the United States showed a % reduction in risk of death for children aged - years using child restraints []. For toddlers, Zaloshnja recently reported that the adjusted odds of injury were % lower for toddlers in child seats than for toddlers restrained in an adult seatbelt[]. Children aged - years using booster seats had a % reduction in risk of injury compared to those in adult seatbelts[]. An in-depth crash investigation study found that for children less than years of age, dedicated child restraints performed best in terms of safety, followed by lap-sash belts, then lap-only belts[]. A recent case-control study of children aged - years found no serious or fatal injuries among optimally restrained children, while % of sub-optimally restrained children sustained serious or fatal injuries[]. There is therefore clear evidence that children are at substantially lower risk of serious injury and death in the event of a car crash if they are restrained in an appropriate child restraint rather than an adult seatbelt.However, for child restraints to be effective they need to be appropriate for the age (or height and weight) of the child. Inappropriate restraint use is defined as the use of a restraint by a child who is not suitable for that restraint on the basis of their age and/or size. It is recommended that children graduate to an adult seatbelt without a booster seat when a good fit is achieved, usually once children are - cm tall (about years old)[,]. However, many children move to an adult seatbelt too early. A recent New Zealand study found that only % of children requiring a booster seat were using one, and in the --years age group, % required a booster seat but only % were using one []. Other studies have shown that inappropriate restraint is highest among children aged - years [-]. Recent Australian research showed that, of children meeting the booster seat height-weight criteria, % had been prematurely graduated into an adult seat-belt[].Factors that have been associated with premature graduation from one restraint type to another in recent Australian studies include low parental education [-], poor parental knowledge of appropriate restraint transition points [], larger families [,], older age of child [], travelling in other people's vehicles and parenting style []. Studies conducted in the US have identified perceived restraint comfort, lower income status and educational attainment as factors associated with a higher risk of early graduation to a seatbelt-only restraint [,].Although children may be restrained in an age-appropriate child restraint, incorrect use of restraints can also degrade the level of protection provided in a crash. Incorrect use is when the restraint system is not used as intended and can occur in both child restraints and adult seatbelts with or without booster seats. Examples include incorrect positioning and/or tension of the belt and incorrect attachment of the restraint to the vehicle. Incorrect restraint use substantially increases injury risk. In its most serious form, an incorrectly used restraint may fail to control the child's motion in a crash, leading to impact with the vehicle interior and other occupants. An Australian retrospective case review of - year olds in crashes found that children incorrectly using restraints were . times more likely to be seriously injured than those correctly using their restraints, irrespective of restraint type []. More recent work has shown that % of children in the state of New South Wales (NSW) are incorrectly restrained, with the prevalence of incorrect use highest among those using dedicated child restraints (% of those using forward facing restraints and % of those using booster seats, compared to % of those using adult seatbelts). In addition, the data have shown that the forms of incorrect use were more serious for children restrained in seatbelt, booster seats and harnesses, all typically used by children in the - year age group []. Risk factors for incorrect use of child restraints include lower educational levels, families who remove the seat frequently from their vehicle, drivers who are not the parent of the child, and young, small children []. Incorrect use of booster seats is more likely to occur in children from low socioeconomic and non-English speaking background (NESB) families [].Previous research has predicted that the greatest gains are to be made by targeting both the use of appropriate restraints and the correct use of restraints []. Studies have also demonstrated the effectiveness of various interventions to increase child restraint use. A recent systematic review [] found that incentives or the provision of a free child restraint combined with education increased the use of restraints - but concluded that, although the interventions seemed promising, there was a need for high quality well-controlled studies to evaluate effectiveness of programs aimed at increasing use of child restraints. A recent US trial found that a pre-school based intervention to increase restraint use was not effective but also reported that few families received the intervention as planned []. Intervention studies aimed at decreasing incorrect use of restraints are less common, with only one published study, which showed an education program was effective in increasing correct use of restraints []. There have been no Australian trials examining effectiveness of interventions to increase the use of appropriate child restraints or decrease incorrect use of child restraints.There is a clear need for a well powered trial of interventions aimed at improving the quality of restraint use. Given the lack of effectiveness of previous trials, there is also a need to understand clearly the process aspects of the trial.The aim of this trial is therefore to examine the effectiveness of a comprehensive intervention to ) increase the use of appropriate child restraints and ) decrease incorrect use of child restraints in pre-school aged children travelling in cars.MethodsThe study is a cluster randomised trial, designed in line with the CONSORT statement [], with randomisation of pre-school or childcare centres. Eligible centres will be identified from early childhood education services in the Sydney metropolitan area via the Kids and Traffic program at Macquarie University who provide road safety education services to all pre-schools and childcare centres state-wide. Directors of eligible centres will be approached by mail and phone by research staff to participate in the trial. Once the directors have consented to join the study and to be randomised, randomisation will occur into treatment and control groups.Eligible centres will include community pre-schools or childcare centres with a minimum of families with a child in the eligible age range (- years). Children with their 3rd birthday in and older will be included as inappropriate use begins to increase dramatically from this age and year old children are widely enrolled in pre-school programs. Because interventions are most needed in families from low educational backgrounds, eligible centres will be those based in local government areas (LGA) of Sydney metropolitan area ranked in the lowest % of LGAs in the Sydney Metropolitan Area according to the Socioeconomic Indices For Areas (SEIFA): Index of Education and Occupation scores (SEIFA scores <). LGAs to be included are Fairfield, Campbelltown, Liverpool, Blacktown, Canterbury, and Bankstown. English must be used as the instructional language in the centre and physical layout of the centre and parking must allow for safe observation of vehicles upon arrival, without major disruption to traffic flow. Finally, centres must have no active engagement or specific policies on child restraints, or programs where they actively engage with families on this issue. A complete list of eligible centres will be created, stratified by service type (pre-school or long daycare), and each will be allocated a unique identifier. The centres will be approached in random order until enough centres of each service type have been recruited. The number of each service type will be chosen to match the distribution of service types in these areas, with an even number of each type and totalling . An authorised representative of the centre will sign a record of informed consent.A baseline assessment will be conducted before randomisation to determine whether the control and intervention groups have similar baseline restraint use patterns. While direct observation of children in situ would be the best way to measure baseline levels of appropriate restraint use, such observation and necessary contact with research assistants may in itself act as an intervention. To avoid such potential contamination of the control group, baseline measures will therefore involve a brief self-report questionnaire to be completed by parents/guardians in both intervention and control centres when they sign children in or out of the centre. Information will be collected about number of children in the family, age of children, type and frequency of restraint use and seating position in car. Although this technique will be less accurate than the direct observations made post intervention, previous work has shown that parent interview tools may be used to accurately describe child restraint use [], and it has been determined that the risk of measurement error is less important than any possible contamination of the control group which may occur by more detailed measures. This process will also remove any ethical dilemmas that may arise if children in the control group were observed to be restrained inappropriately. Any self-report bias should be similar between groups.Restricted randomisation will be conducted using the unique identifiers and summary data from the baseline survey to ensure adequate balance of important factors. The five balance criteria will be (i) service type exactly balanced between groups; (ii) self-reported restraint choice for the - year-old children attending the service (proportion appropriately restrained): mean proportions in each group differ by no more than % in relative terms (i.e. ratio of means between . and .); (iii) mean proportion of families with annual household income below AU$, in each group differ by no more than % relatively; (iv) mean proportion of families who speak a language other than English at home in each group differ by no more than % relatively; and (v) mean size (number of children aged - years enrolled) of centres in each group differ by no more than % relatively. The randomisation process will be fully concealed as the statistician will allocate the centres at a site remote to the recruitment site and will only have access to a list of centres recruited. Of the centres in the study, centres will be randomly allocated to each of the intervention and control groups.InterventionThe intervention will be a comprehensive education and restraint distribution program, to be delivered in pre-school settings under the Kids and Traffic program. The Kids and Traffic program offers a road safety orientated education service to all child care services in NSW. Funded by the Roads and Traffic Authority (RTA) of NSW, the licensed early childhood services registered with the program are sent road safety information, programming advice, posters and information to share with families at least once each year. It has developed a range of strategies, based on the principles of early childhood learning and best practice in early childhood education which support road safety education in pre-school/daycare settings. These include professional development workshops primarily for staff and others designed for both staff and families. The program will be enhanced as part of this trial such that specific material will be developed around increasing use of appropriate child restraints and decreasing incorrect use of child restraints. In line with interventions tested in smaller trials in other settings [,], this material will include an educational component (including staff and parent workshops), and will be further enhanced by distribution of child restraints at reduced cost and ongoing distribution and reinforcement of educational materials at the centre level.The program to be tested in the trial includes the following components:. Centres will receive the enhanced workshop from the Kids and Traffic program, with the focus on restraint use. This will require in-service training for staff, and resources for incorporating messages about appropriate restraint use into programming for the centre. Centre staff will also receive monthly policy support from the Kids and Traffic Program staff, and will be provided with extra printed information that can be incorporated into parent newsletters to reinforce messages.. Educational material will be provided to intervention centres, consisting of pamphlets and posters about optimal child restraint practices, and information about frequently asked questions, with photographs of appropriate use and incorrect use. Written material will be translated into a range of community languages. These materials will contain specific information about what types of restraints should be used by children in the target age group, and common forms of incorrect use and the impact these have on crash safety. Centres will be asked to display restraint information provided in prominent positions. Approximately weeks following the staff inservice, a parent information session will be held at each centre. This session will include a presentation by the Kids and Traffic personnel, and will include viewing of a DVD created for the trial detailing the benefits of appropriate and correct restraint use and reinforcing recommendations for correct and appropriate use provided in print material, as well as hands-on demonstrations of child restraints. All parents enrolled at each intervention centre will receive material on restraints including the DVD, brochures on appropriate restraints, and written and pictorial information on restraint use. All families will be given a free voucher for a visit to an authorised child restraint fitting station for personalised advice on correct restraint use and will receive information on location of fitting stations in their area.. Subsidised child restraints will be made available at a minimal cost ($AUD50, approximately % of recommended retail price) for those who require them.The control group will be offered the standard Kids and Traffic program rather than the enhanced program described above. The current program offers minimal information about use of child restraints so, even if implemented, it is unlikely to have a significant impact on appropriate and correct restraint use. Control sites will be given the option of receiving the enhanced Kids in Traffic workshop at the end of the trial.Outcome measures will be both use of appropriate restraints for the age of the child, and incorrect use of restraints, and will be collected months after the intervention has been implemented using a proven observational method []. Masked observers will attend each site over a - hour period during morning drop-off times. Potential participants will be approached as they pull into a parking spot outside the institution, and informed consent will be gained from all participating parents. Initial observations will be made with the child in situ. The driver of the vehicle will then be invited to participate in a brief interview. All refusals will be recorded and brief observational and demographic data collected to allow evaluation of potential bias between participants and non-participants. A brief structured interview will be conducted with those consenting to record parent/guardian age, gender, education, and information about age and gender of child, usual child restraint use, number and age of children in family. A detailed examination of the restraint installation by trained observers will be conducted while the interview is taking place.Although it would be ideal to measure outcome after a longer time period, it is difficult in practice to follow-up children over two calendar years. Most of the children aged - at the beginning of the intervention year will move to primary (elementary) school the following year. It would not be practical to collect follow-up data from children at this time as they will move to a range of different schools and direct observation will not be feasible. While month follow-up data could be collected by phone, likely differential reporting bias would mean that such outcome data would be severely compromised.In order to assess compliance and uptake of the program, records will be kept of which families attend the education sessions, and receive individualised fitting services or subsidised child restraints. Information about alternative road safety or restraint programs administered by either intervention or the control sites will be collected. Importantly, as understanding practical issues about how to implement such programs is vital to further roll-out of successful programs, process measures will also be evaluated by asking the director and staff of the intervention group centres to complete a short questionnaire about the implementation of the program.Research assistants conducting the pre-and post-trial observations will be blinded to the intervention/control status of each centre. Furthermore, research staff will be trained appropriately and outcome measurements will be based on standardised objective criteria, minimizing opportunities for measurement error. As pre-trial observations are collected by self-administered questionnaires, research staff will not observe children until after the intervention has been implemented. For both intervention and control centres, research staff conducting post-trial observations will present parents or guardians of children with written information about appropriate use of child restraints and offer advice if any child is observed in inappropriate restraints or using restraints incorrectly. Appropriate bilingual staff will be employed to conduct baseline interviews, implement the program and conduct post-trial observations, to meet the language needs of participating families.On March new national child restraint laws were enacted in NSW, mandating appropriate restraint use by children up to age . Enforcement of these laws was delayed until June to allow parents and carers time to fully understand and comply with the new laws. The new laws will impact on all pre-school centres and parents in the trial. The impact of the new laws is that the baseline rate of appropriate child restraint use may increase in all centres, whether control or intervention. It is unlikely that the rate of incorrect use will change as the social marketing and education campaigns behind the new laws are focused more on use of appropriate restraints rather than on reducing incorrect use. Baseline interviews will be conducted before the enforcement of the new law at all centres, and for most centres before the March introduction.Sample size and statistical analysisThis study has % power to detect a % increase in use of appropriate child restraints from % to %, as significant at the % level. Previous research has indicated that educational and incentive based interventions may increase use of restraints by about % []. In NSW the use of appropriate child restraints in centres in Sydney varied from -% []. However, among centres from local government areas with low educational attainment, the average use of appropriate child restraint use is -%. If the effect of the implemented RTA general marketing program raises the base rate in controls to %, the study will be adequately powered to find an effect of a % increase. However, it will also have power to detect a % effect if the base rate of restraint use remains between -%. Increasing the rate to % is feasible: the base rate of child restraint use in centres in areas with high education and high socioeconomic status is around %, so this increase is possible. Previous work has allowed estimation of the intra-class correlation as ., based on proportion using appropriate restraint at centres []. Allowing for a minimum of eligible children at each centre (with an average number of slightly more than anticipated) would require centres. Further allowing for % centre drop-outs through the course of the study requires intervention centres and control sites. This sample size will also allow % power to establish whether incorrect use of restraint systems can be reduced from % of children using restraints to %. Direct observation of children in pre-school settings in Sydney found that % of children using forward facing child restraints and % of children in booster seats were incorrectly using the restraints [].Data will be analysed on an intention-to-treat basis. The primary analysis will compare the proportion of each of the two outcome measures (appropriateness and correctness of restraint use) at each centre between the intervention and control groups using Donner's adjusted chi-square test[]. Logistic regression modelling will be used to explore predictors of each outcome measure, using generalised estimating equations (GEE) to allow for clustering of data by centre.The study protocol was approved by the Human Research Ethics Committees of the University of Sydney and the University of NSW.DiscussionThis assessor blinded cluster randomised trial is powered to provide credible evidence about the efficacy of an education and distribution program in a pre-school setting to increase appropriate use, and decrease incorrect use of child restraints. The intervention is an evidence based program that could be implemented on a larger scale if found to be effective, via the existing, government funded, Kids and Traffic program. The study results will also highlight whether such programs can increase appropriate restraint use over and above the increases found as part of new enforcement, education and legislation introduced simultaneously, and further, whether it is possible to offset the higher rates of incorrect use associated with increased use of appropriate restraints. The results of this study will therefore have major relevance to policy makers and child safety advocates as well as researchers in both Australia and internationally. If such interventions can be shown to be efficacious, they show great promise for implementation. The results will also be of great relevance to researchers and policy makers in the US and other high income countries.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsRI, LB, JB and LK led the methodological design of the study, supported by JS, KH and MS. RI and LK drafted the paper and all authors contributed to revisions. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/-///prepub
|
PMC2717082.txt
|
TITLE: Leisure-time physical activity, cardiorespiratory fitness and feelings of hopelessness in men AUTHORS: Maarit Valtonen, David E Laaksonen, Jari Laukkanen, Tommi Tolmunen, Rainer Rauramaa, Heimo Viinamäki, Jussi Kauhanen, Timo Lakka, Leo Niskanen ABSTRACT: BackgroundLeisure-time physical activity (LTPA) and cardiorespiratory fitness contribute to mental health. Hopelessness has been linked to impaired mental health, cardiovascular events and mortality. Previous studies have focused on physical exercise and depression. We examined the associations of LTPA and cardiorespiratory fitness with feelings of hopelessness.MethodsIn this cross-sectional study leisure-time physical activity, maximal oxygen uptake (VO2max), hopelessness and cardiovascular risk factors were assessed in a population-based cohort of men aged – years old at baseline.ResultsMen feeling more hopeless about their future and reaching goals were less physically active, less fit and had a higher prevalence of many cardiovascular risk factors than men with lower levels of hopelessness. In a logistic regression model adjusted for age, smoking, alcohol consumption, cardiovascular disease and socioeconomic status, men engaging in less than min/week of moderate-to-vigorous LTPA were % (% CI – %) more likely to feel hopeless than those engaging in at least . h/wk of LTPA. After further adjusting for elevated depressive symptoms the association of LTPA and hopelessness remained significant. VO2max was also associated with hopelessness, but not after adjustment for depressive symptoms.ConclusionModerate and vigorous LTPA and cardiorespiratory fitness were inversely associated with hopelessness in these middle-aged men. These findings suggest that physical inactivity and poor cardiorespiratory fitness is an important associate of hopelessness, a distinct element of low subjective well-being. BODY: BackgroundHope is an important component of physical and psychological well-being []. Hopelessness is associated with dissatisfaction with life, depression and suicidality []. Previous studies have shown that hopelessness is a predictor of cardiovascular morbidity [-] and mortality [], independently of depression and other confounding factors. Hopelessness is also associated with the incidence of hypertension, myocardial infarction, and cardiovascular mortality and with accelerated progression of carotid atherosclerosis in middle-aged men [-]. Even though hopelessness is a rather prevalent condition in general population [] and is inversely associated with health [-], the mechanisms underlying this phenomenon remain unclear.Leisure-time physical activity (LTPA) and cardiorespiratory fitness seem to protect against chronic diseases such as the metabolic syndrome [,], type diabetes [] and cardiovascular disease (CVD) [,]. Moreover, physically active lifestyle may improve mental health []. Previous studies have mainly focused on exercise and depression. Both cross-sectional [-] and prospective studies [-] have shown conflicting results on the association between physical activity and depression. For example, Tolmunen and colleagues [] reported that cardiorespiratory fitness is inversely related to elevated depressive symptoms in middle-aged Eastern Finnish men. However, physical activity, rather than cardiorespiratory fitness, has been associated with better mental health and mood []. Mechanisms behind these associations are unclear. Confounding factors related to depression such as an unhealthy lifestyle and several chronic diseases must also be taken into consideration in assessing the relationship of physical exercise and psychological well-being.The associations of physical activity and cardiorespiratory fitness with hopelessness have not been reported earlier. We therefore investigated the associations of LTPA and cardiorespiratory fitness with hopelessness in middle-aged Finnish men.MethodsThe Kuopio Ischemic Heart Disease Risk Factor Study (KIHD) is a population-based cohort study in a sample of middle-aged men in Eastern Finland. Baseline data were collected between and from male participants aged – years old.The Research Ethics Committee of the University of Kuopio approved the study. All study subjects gave their written informed consent. This study includes men who had complete data on physical activity, VO2maxand depressive symptoms.Assessment of LTPAThe validated KIHD -month LTPA Questionnaire was used as described previously [,]. This is a detailed quantitative questionnaire assessing the duration, frequency and mean intensity of the most common lifestyle and structured LTPA of middle-aged Finnish men as recalled over the previous months. Low-intensity LTPA was defined as <. metabolic equivalents (METs). One MET is defined as metabolic expenditure at rest, corresponding to an oxygen uptake of . ml O2/kg. Cut-off of ≥ . METs for at least moderate LTPA included brisk walking, skiing, jogging, bicycling, ball games and forestry. Vigorous LTPA was defined as ≥ . METs. The durations of LTPA were calculated in min/week.Assessment of cardiorespiratory fitnessA graded symptom-limited exercise test was performed on an electrically braked cycle ergometer. VO2max was measure directly with breath-by-breath respiratory gas-exchange analysis as previously described [].Assessment of hopelessness and depressive symptomsThe psychological questionnaires included two items that measured hopelessness [-]. These items were "the future seems to be hopeless, and I cannot believe that things are changing for the better" and "I feel that it is impossible to reach the goal I would like to strive for". Participant responded using -point scale ( = absolutely agree, = somewhat agree, = cannot say, = somewhat disagree, or = absolutely disagree). Hopelessness score, with a range of to , was created by reverse-coding and summing the items.We used the Human Population Laboratory Depression Scale (HPL Scale) to assess depressive symptoms. The HPL Scale is a self-administered -item depression score that was specifically developed for screening general population samples [,]. A cut-off ≥ was used previously to classify men with elevated depressive symptoms [].Assessment of features related to the metabolic syndrome, diabetes and cardiovascular diseaseBody mass index (BMI), waist circumference and blood pressure were measured as previously described [].Fasting blood glucose was measured using a glucose dehydrogenase method. Diabetes was defined as fasting blood glucose concentration ≥ . mmol/L or a clinical diagnosis of diabetes []. Serum insulin was determined with a Novo Biolabs radioimmunoassay kit. The cholesterol contents of lipoprotein fractions and serum triglycerides were measured enzymatically. Fibrinogen was measured based on the clotting of diluted plasma with excess thrombin. Serum C-reactive protein (hCRP) was measured with a high-sensitivity immunometric assay [].Definition of the metabolic syndromeThe National Cholesterol Education Program (NCEP) criteria were used (for men, three or more of the following: fasting blood glucose levels ≥ . mmol/l, triglycerides ≥ . mmol/l, HDL cholesterol <. mmol/l, blood pressure ≥ / mmHg, waist girth > cm) [,].Other assessmentsMedical history, socioeconomic status, smoking habits and alcohol consumption were assessed with questionnaires [,].Statistical AnalysisParticipants were categorized by tertiles based on their scores for hopelessness. Differences between men in the highest third of hopelessness and men in the lower thirds were assessed with one-way ANOVA, and where indicated, the chi-squared test. LTPA and VO2max were categorized by tertiles for logistic regression analyses. The associations of LTPA and VO2max with hopelessness were estimated using logistic regression models adjusting for covariates. Durations of LTPA (in min/week) and triglyceride, insulin and hCRP concentrations are presented as medians (interquartile ranges); other data are presented as means ± or simple percentages. Triglyceride and insulin concentrations were corrected for skewing using log transformation for statistical analysis but are presented using untransformed values. Statistical significance was considered to be P < .. All statistical analyses were performed with SPSS . for Windows (Chicago, IL).ResultsBaseline clinical characteristicsMen with increased hopelessness had more pronounced features of the metabolic syndrome and inflammation (Table ). Moreover, men in highest third of hopelessness were less physically active and had a lower VO2max.Table 1Characteristics of the middle-aged men according to tertiles of hopelessness.LowMiddleHighP value*N757803868Age (years). (.). (.). (.)< .001Body mass index (kg/m2). (.). (.). (.).011Waist to hip ratio0. (.). (.). (.)< .001Waist girth (cm). (.). (.). (.)< .001Systolic blood pressure (mmHg). (.). (.). (.).014Diastolic blood pressure (mmHg). (.). (.). (.).201Fasting blood glucose (mmol/l). (.). (.). (.).004HDL cholesterol (mmol/l). (.). (.). (.).864Serum triglyserides (mmol/l). (., .). (., .). (., .).090Fasting serum insulin (mU/l). (., .). (., .). (., ,).048C-reactive protein (mg/l). (., .). (., .). (., .).001Fibrinogen (g/l). (.). (.). (.)< .001Total LTPA (MET hours/year) (, ) (, ) (, ).041Total LTPA (minutes/week) (, ) (, ) (, ).704LTPA < . METs (minutes/week) (, ) (, ) (, ).728LTPA ≥ . METs (minutes/week) (, ) (, ) (, ).030LTPA ≥ . METs (minutes/week) (, ) (, ) (, )< .001VO2max (ml × kg- × min-). (.). (.). (.)< .001Alcohol consumption (g/week) (, ) (, ) (, ).096Smoker (%)...< .001Adult socioeconomic status7.(.). (.). (.)< .001Metabolic syndrome (NCEP†) (%)...< .001Cardiovascular disease (%)...< .001Diabetes (%)....140Elevated depressive symptoms(points on the HPL scale). (, .). (, .). (., .)< .001Data are displayed as means ± SD, medians (interquartile ranges), or percentages.*ANOVA for continuous variables; chi-square for categorical variables. †NCEP, National Cholesterol Education Program.Association of LTPA with hopelessnessAfter adjustment for age, men with at least moderate LTPA more than . h/week had a % lower risk of being hopeless than men with ≤ min moderate LTPA/week (Table , Model , P < .). After further adjustment for potential confounding or mediating factors (Models and ), the association remained significant.Table 2Odds ratios (% confidence intervals) for having feelings of hopelessness according to categories of physical activity and cardiorespiratory fitness in middle-aged men.HOPELESSNESS (upper vs. lower tertiles)Model 1Model 2Model 3Model 4Total LTPA (min/week)< min/week1111270– min/week0. (.–.). (.–.). (.–.). (.–.)> min/week0. (.–.). (.–.). (.–.). (.–.)p0....022Model 1Model 2Model 3Model 4Low-intensity LTPA (<. METs, min/week)< min/week1111111– min/week0. (.–.). (.–.). (.–.). (.–.)≥ min/week0. (.–.). (.–.). (.–.). (.–.)p0....393Model 1Model 2Model 3Model 4Moderate-to-vigorous LTPA (≥ . METs, min/week)≤ min/week111161– min/week0. (.–.). (.–.). (.–.). (.–.)≥ min/week0. (.–.). (.–.). (.–.). (.–.)p< ....002Model 1Model 2Model 3Model 4Vigorous LTPA (≥ . METs, min/week)< min/week111110– min/week0. (.–.). (.–.). (.–.). (.–.)≥ min/week0. (.–.). (.–.). (.–.). (.–.)p< .< .< .< .001Model 1Model 2Model 3Model 4VO2max(ml•kg-•-1min)≤..–.. (.–.). (.–.). (.–.). (.–.)≥.. (.–.). (.–.). (.–.). (.–.)p< ....242Data are OR (% Cl).Model , adjusted for age category (logistic regression).Model , adjusted for age category, smoking, alcohol consumption (abstainers and low, moderate and high intake according to g/wk consumption), cardiovascular disease and adulthood socioeconomic status.Model adjusted for Model and body mass index, serum C-reactive protein and plasma fibrinogen.Model adjusted for Model and elevated depressive symptoms (Human Population Laboratory Depression Scale).The association of vigorous LTPA with hopelessness was even stronger (Table , Model , P < .). Further adjustment for potential confounding factors (Models and ) did not alter the association. Total LTPA was similarly associated with hopelessness, but low-intensity exercise was not (Table ).In additional analyses with separate adjustment of variables in Model by VO2max, diabetes and the metabolic syndrome, none of these potentially mediating factors weakened the association (data not shown). Adjusting for the separate components of the metabolic syndrome (fasting serum HDL, triglycerides or insulin, fasting blood glucose, waist girth and blood pressure) did not alter the association of moderate-to vigorous or vigorous LTPA with hopelessness either. Furthermore, in the association with hopelessness there was no interaction between moderate-to-vigorous LTPA and diabetes (P = .), CVD (P = .), or the metabolic syndrome (P = . for the interaction). There also was no interaction between high-intensity LTPA and these variables for the association with hopelessness.We categorized the men by VO2max tertiles to study whether cardiorespiratory fitness modifies the association between moderate LTPA and hopelessness (Figure ). The association seemed to be stronger in the most fit group, but the interaction term for VO2max and LTPA with respect to hopelessness was not significant (P = .). Physically unfit and sedentary men were twice as likely to feel hopeless than fit and physically active men.Figure 1Tertiles of hopelessness score in relation to physical fitness and at least moderate LTPA. LTPA categories were adjusted for age, presence of CVD, adult socioeconomic status, smoking and alcohol consumption. Unfit and physically inactive men had the highest risk for hopelessness. Those men who exercised most and were fit had the lowest risk (versus the unfit and least active category, OR ., % CI .–., P for the trend < .). The interaction term for VO2max and LTPA with respect to hopelessness was not significant (P = .).Association of cardiorespiratory fitness with hopelessnessMen with VO2max over . ml • kg- • min- were % less likely to express feelings of hopeless than those with VO2max below .. ml • kg- • min- after adjusting for age. After further adjustment for potentially confounding variables (models and ), cardiorespiratory fitness was still associated with hopelessness. After further adjustment separately for potential mediating factors, such as LTPA, diabetes, CVD and the metabolic syndrome, the association remained significant (data not shown). Of the components of the metabolic syndrome, only waist girth attenuated the association between VO2max and hopelessness significantly (data not shown). Moreover, there were no interaction between VO2max and diabetes (P = .), CVD (P = .) or the metabolic syndrome (P = . for the interaction) in the association with hopelessness.LTPA, hopelessness and depressive symptomsHPL depression and hopelessness scores correlated moderately (r = ., P < .). The average HPL depression score in men in the highest third of hopelessness categories was higher than in the lower tertiles (Table ). However, in the logistic regression analysis elevated depressive symptoms did not decrease the associations of moderate or vigorous LTPA and hopelessness (Table , Model , P = . and P < .). Moreover, there was no interaction between LTPA and depressive symptoms in the association with hopelessness (moderate LTPA P = ., vigorous LTPA P = . for the interaction). In separate analyses, depressive symptoms and LTPA were not associated.Cardiorespiratory fitness, hopelessness and depressive symptomsUnlike with physical activity, after adjusting for depressive symptoms the association of cardiorespiratory fitness with hopelessness was no longer significant (Table , Model ). There was no interaction between VO2max and depression in the association of VO2max with hopelessness (P = .).DiscussionThis is the first study to show that men who were physically active during their leisure-time were less likely to feel hopeless about their future and reaching goals than sedentary men. Cardiorespiratory fitness seemed to equally be accompanied by hopelessness. Moreover, these associations persisted even after adjusting for BMI, inflammatory markers and other confounding factors.All adults are recommended to engage in at least min of moderate-intensity exercise per day []. In our study those men exercising for at least . h per week were % less likely to express feelings of hopelessness than sedentary men (< min/week) even after adjustment for potential confounding factors, BMI and inflammatory markers. The association of vigorous LTPA with hopelessness was even stronger, reducing the likelihood of hopelessness by % if practiced at least one hour per week. Of importance, potential mediating factors such as VO2max, diabetes, the metabolic syndrome and or its components and other potential confounding factors did not attenuate these associations.The effect of depressive symptoms on the associations of LTPA and VO2max with hopelessness was distinctly different. Odds ratios for total, moderate and vigorous LTPA for hopelessness were not attenuated in logistic models adjusting further for depressive symptoms, whereas the association of VO2max and hopelessness was no longer significant. Cardiorespiratory fitness was strongly associated with elevated depressive symptoms even after adjustment for various potential confounding and mediating factors. The relationship between LTPA and depression, however, was not found in this cohort, as reported previously []. This suggests that hopelessness and depression are overlapping, but distinct entities. The findings also suggest that moderate or vigorous LTPA may ameliorate or protect against feelings of hopelessness even if VO2max does not improve.Although hopelessness is considered to be a part of depression, it is also quite common in the absence of depression. According to Haatainen and colleagues [] the prevalence of hopelessness measured with the Beck Hopelessness scale, was .% in a homogenous sample of Finnish adults. After excluding those with any self-reported mental disorder diagnosed or treated by a physician during the preceding year the prevalence of hopelessness was still as high as .% []. Studies from this [-] and other cohorts[] suggest that the correlation between hopelessness and depression scales is only moderate (r = . – .). Moreover, these studies [-] suggest that the effects of feelings of hopelessness and depression can be disentangled, and that hopelessness may be a more powerful predictor of adverse cardiovascular outcome than depressive symptoms. The present findings suggest that LTPA helps to maintain an optimistic perspective on future and personal capabilities.The distinction between hopelessness and depression could explain the difference in the associations of hopelessness with LTPA and VO2max. It is important to note, however, that cardiorespiratory fitness is only partly a reflection of physical activity. Although moderate and vigorous LTPA obviously has an effect on VO2max, it is also determined partly by genetic factors []. How this may relate to feelings of hopelessness and depression is still unclear.The strength of this cross-sectional study is its large population-based design and detailed assessments of different psychosocial factors and features related to cardiovascular disease. Furthermore, VO2max was measured directly by a maximal symptom-limited cycle ergometer exercise test with analysis of respiratory gas exchange, an accurate and highly reproducible measure of cardioresipiratory fitness []. Physical activity was not assessed using physical activity monitors, but the questionnaire used to measure physical activity is valid and repeatable []. Hopelessness and depressive symptoms were measured using self-reported questionnaires instead of structured interviews. Although these questionnaires are relatively simple and are designed for epidemiological use rather than clinical diagnosis, they have been well validated [,]. Both of these scales have been widely used in the prediction of outcomes and as outcomes [-,,,]. The HPL depression scale is significantly correlated with the Beck Depression Inventory in an outpatient population (r = .) [], and it is also similar to other symptom checklists, such as the Center for Epidemiologic Studies Depression scale []. However, we lack information of correlation between the HPL hopelessness questionnaire and other hopelessness scales such as Beck Hopelessness scale. The findings of this study cannot be generalized to women and men in different age and ethnic groups. Because of the cross-sectional design, we cannot draw conclusions about the directionality of the associations of physical activity and VO2max with hopelessness.ConclusionPhysically active middle-aged men are less likely to feel hopeless about their future and reaching goals than sedentary men. Cardiorespiratory fitness was also related to reduced feelings of hopelessness, but this association was partially explained by depressive symptoms. Considering that hopelessness is an important determinant of mortality, cardiovascular morbidity and low subjective well-being, this study provides additional evidence for the health benefits of physical activity and fitness. Physically active lifestyle not only helps to live a physically healthier life, but it may also improve happiness by helping to maintain a positive attitude and optimistic perspective on the future and oneself.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMV participated in the design of the study, performed the statistical analysis and drafted the manuscript. DEL participated in the design of the study and helped to draft the manuscript. JL, TT, RR, HV, JK and TK have been involved in drafting the manuscript or revising it critically for important intellectual content. LN conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:
|
PMC2629569.txt
|
TITLE: Towards Predictive Computational Models of Oncolytic Virus Therapy: Basis for Experimental Validation and Model Selection AUTHORS: Dominik Wodarz, Natalia Komarova ABSTRACT: Oncolytic viruses are viruses that specifically infect cancer cells and kill them, while leaving healthy cells largely intact. Their ability to spread through the tumor makes them an attractive therapy approach. While promising results have been observed in clinical trials, solid success remains elusive since we lack understanding of the basic principles that govern the dynamical interactions between the virus and the cancer. In this respect, computational models can help experimental research at optimizing treatment regimes. Although preliminary mathematical work has been performed, this suffers from the fact that individual models are largely arbitrary and based on biologically uncertain assumptions. Here, we present a general framework to study the dynamics of oncolytic viruses that is independent of uncertain and arbitrary mathematical formulations. We find two categories of dynamics, depending on the assumptions about spatial constraints that govern that spread of the virus from cell to cell. If infected cells are mixed among uninfected cells, there exists a viral replication rate threshold beyond which tumor control is the only outcome. On the other hand, if infected cells are clustered together (e.g. in a solid tumor), then we observe more complicated dynamics in which the outcome of therapy might go either way, depending on the initial number of cells and viruses. We fit our models to previously published experimental data and discuss aspects of model validation, selection, and experimental design. This framework can be used as a basis for model selection and validation in the context of future, more detailed experimental studies. It can further serve as the basis for future, more complex models that take into account other clinically relevant factors such as immune responses. BODY: IntroductionOncolytic viruses are live replicating viruses that selectively infect cancer cells and kill them []–[]. Healthy cells are largely spared. The idea is to inoculate the virus into a cancer patient, and let the virus spread throughout the tumor, thereby driving it into remission. Selectivity for cancer cells occurs because cancer cells tend to lack important genes that normally shut down the replication cycle of the virus. For example, the adenovirus ONYX- has been engineered such that it only replicates in p53−/− cells, a characteristic of many cancers []. Certain animal viruses by chance have the ability to replicate in human cancer cells, while healthy human cells are not permissive. An example is Newcastle disease virus, which can replicate in tumor cells that lack interferons [], []. In general, a wide array of viruses is being explored as potential oncolytic viruses.Oncolytic viruses have shown promising results in clinical trials []. Cancers have been found to respond to treatment, leading to tumor remission in some cases. Consistent and sustained eradication or control of cancers has, however, been very difficult to achieve. This is caused in part by our lack of understanding regarding the dynamics that underlie the spread of oncolytic viruses through tumors. Without such a rigorous understanding, much of the work is based on trial and error. In such scenarios, mathematical models can be very useful to complement empirical work. Mathematical analysis allows us to see the whole spectrum of possible outcomes, and provides a means to logically suggest ways to optimize treatment. Limited mathematical analysis of oncolytic virus therapy has been performed in the past []–[]. This work is largely qualitative in nature, examining how variation in viral and host parameters influences the outcome of treatment. For example, it has been suggested that maximizing the virus-induced rate of tumor cell killing is not going to lead to the best treatment outcomes. Instead, an intermediate and optimal rate of virus-induced cell death optimizes treatment success [], []. This work was based on the analysis of the equilibrium properties of the model. That is, the lower the total number of cancer cells that remain as the dynamics converge to steady state, the better the predicted outcome of therapy.While such steady state analysis can provide some valuable qualitative insights, it has limitations. The main problem is that in such infection dynamics models, the population of cells and viruses can show extensive oscillations before converging to a steady state. During these oscillations, the populations of cells and viruses can potentially go extinct, and the system might never reach equilibrium. Therefore, it is important to understand these oscillatory dynamics, and how they relate to the chances that the cancer cell population is driven extinct.This paper aims to analyze these dynamics in detail in an attempt to provide a more realistic description of oncolytic virus dynamics. This is a difficult task because these infection dynamics, and in particular the occurrence of population oscillations, can be dependent on particular details of the models that are of a biologically uncertain nature. To address this issue, we avoid concentrating on a particular model, but take a more general approach. Through specific restrictions about biological assumptions, we analyze a class of mathematical models that aim to describe viral spread through a tumor in different settings. We seek to determine conditions under which the virus is successful at eliminating the tumor, and the conditions when virus therapy fails. In order to underline the insights that we gain from this general framework, we also consider specific models that are examples of the general framework. This modeling framework provides the basis for experimental validation and testing procedures, which will allow us to accurately predict the time course of cells and viruses at least in relatively simple scenarios, such as in vitro experiments or simple in vivo scenarios. In this context, we fit the models to previously published experimental data and discuss implications for model testing, model selection, and experimental work. A predictive model of a complex in vivo situation (e.g. including immune responses) will obviously be more difficult to attain, but can arise from a thorough understanding of the simpler in vitro scenario that we examine here.ResultsThe modeling frameworkWe will model the dynamics of oncolytic virus replication by ordinary differential equations that describe the development of the average population sizes of cells and viruses over time. This approach is based on very well established mathematical models that describe the general dynamics of virus spread both in vivo and on an epidemiological level [], []. Instead of considering a specific model, however, we will take a generalized approach and consider a class of models. The general modeling framework used in our study is as follows. We take into account two populations: uninfected tumor cells, x; and infected tumor cells, y. The population of free viruses is not modeled explicitly. Because the turnover of free viruses is much faster than that of infected cells, we simply assume that the free virus population is in a quasi-steady state and proportional to the number of infected cells. The basic model is given as follows:() () The function F describes the growth properties of the uninfected tumor cells, x, and the function G describes the rate at which tumor cells become infected by the virus. These functions are unknown and can potentially take a variety of forms, which will be discussed below. The coefficient β in front of the infection term represents the infectivity of the virus. Finally, virus-infected cells die with a rate ay. We will not include immune responses in our considerations. While immune responses will certainly be an important factor for oncolytic virus dynamics in vivo, our goal is to first understand those dynamics in a simpler setting without the presence of immune responses. These models would be suitable to describe the growth of oncolytic viruses in relatively simple in vitro or in vivo settings. Once an understanding of such simple systems has been achieved, additional biological complexities (such as the presence of immune responses) can be added to the model.This class of models is characterized by the existence of equilibria, the number and nature of which depends on the tumor growth term F and the infection term G. In the most general sense, the equilibria of the system are defined by the following two equations:() ()We will explore the equilibria and their properties depending on the tumor growth term, F, and the infection term, G.The term F reflects the growth properties of an uninfected tumor. It comprises both division and death rates. The simplest assumption that can be made about the term F is that growth is exponential (or, more precisely, the division and death happen according to an exponential law, and the division rate is higher than the death rate). While this can be true during early stages of tumor growth, tumor growth certainly deviates from an exponential pattern at larger sizes for a variety of reasons, for example space or nutrient limitations. Therefore, more complicated tumor growth terms involving some form of saturation must be considered []. In this respect, we can distinguish between two basic scenarios: First, while the rate of tumor growth saturates and slows down at higher tumor sizes, the tumor has the potential to keep growing towards infinity. Growth would stop once the tumor has reached a lethal size. Second, it can be assumed that growth not only slows down, but comes to a halt as the tumor size reaches a critical level, which can be called the carrying capacity of the tumor. This could happen when the division rate equals the death rate of the cells.Regarding the infection term, the assumption used most often in mathematical models is that it is directly proportional to the number of infected and uninfected cells [], []. This, however, assumes mass action or perfect mixing of populations, which is unrealistic, especially in the context of tumors. Instead, virus spread is likely to be slower, limited by spatial constraints. Since the virus released from one infected cell cannot reach all susceptible tumor cells in the population, the infection rate must be a saturating function of the number of susceptible tumor cells. Similarly, not all infected cells present in the population will be able to contribute to the generation of newly infected cells, for example if they are spatially separated from susceptible cells.In the following section, we will define different classes of infection terms that have biologically reasonable characteristics, and investigate how they influence the properties of the model. These are based in part on mathematical work done in the context of infectious disease epidemiology [], []. Subsequently, we will examine how changing the tumor growth term influences the model predictions.Different classes of infection terms and their propertiesLet us consider two different classes of viral growth, see figure (a,b). Tumor-virus systems belonging to class I are characterized by the following property: if the number of uninfected tumor cells is high relative to the number of infected cells, virus growth does not slow down as the number of infected cells rises. Virus growth is exponential. Biologically, this can be interpreted as virus replication in a non-solid tumor where cells mix relatively freely. In other words, infected cells are not clustered together in a mass but are interspersed among uninfected cells. This is shown schematically at the top of figure (a) (the white circles represent uninfected cells, and the black circles - infected cells). In this case, if the number of uninfected cells is relatively large, then every infected cell is likely to be surrounded by uninfected cells to which the virus can be passed on. Alternatively, a similar picture can be achieved by a very high motility of the virus. In either case, all infected cells contribute to viral spread and growth is exponential. We call this “fast virus spread”. On the other hand, with tumor-virus systems that belong to class II, the virus growth rate decreases as the number of infected cells rises, even if the number of uninfected cells is very large. The biological interpretation is that infected tumor cells are clustered together, figure (b). This can occur in solid tumors, which typically show a high degree of spatial arrangement. In this case, as the number of infected tumor cells increases, most infected cells will be surrounded by other infected cells and not by uninfected cells. Hence, they cannot pass on the virus and cannot contribute to virus spread. Only cells at the periphery of the infected cell mass have uninfected cells in the neighborhood and can contribute to new infection events. We refer to this model of infection as “slow virus spread”../journal.pone..g001Figure 1Two classes of virus growth captured by the mathematical models.(a) According to class I or fast virus growth, virus growth is exponential as long as the number of uninfected cells is significantly larger than the number of infected cells. This can correspond to a high degree of mixing between infected and uninfected cells. As the virus population grows, the number of cells that contribute to virus spread remains constant because most infected cells will have an uninfected cell in their vicinity. (b) According to class II or slow virus growth, virus growth slows down and saturates as the virus population increase in size, even if the number of uninfected cells is relatively large. This can correspond to spatial clustering of the infected cells. Only infected cells at the surface have uninfected cells in their neighborhood and can thus contribute to virus transmission. As the number of infected cells rises, the number of “active” cells that can contribute to virus transmission declines.Next let us connect this classification with the mathematical model, and in particular, with the infection term, βyG(x,y). The function G(x,y) is related to the proportion of the total population of the infected cells which participates in the infection process. It is plotted in figure as a function of the number of tumor cells, x, and we examine the shape of these plots. Let us take a closer look at the schematic at the top of figure (b). Because of the geometrical arrangement of the cells in this case, only the infected cells on the surface of the black core will be able to infect other cells (it is out of the cells in the smaller colony presented). Now, let us increase the system size, such that the number of infected and uninfected cells grows in the same proportion. Again, only the infected cells close to the surface of the infected core will participate in the infection process. However, now the proportion of the surface cells is much smaller ( out of cells). As the size of the system increases, the proportion of such “active” cells (that is, cells capable of infecting other cells) decreases. This is what is depicted in the graph in figure (b), where the function G(x,y) declines following the peak. (For very small system sizes, the proportion of cells participating in infection is formally zero because of the lack of uninfected cells, therefore the graph of the function G starts at zero, reaches a peak, and then declines for high values of x). Next, we take a look at the cell arrangement at the top of figure (a). Here, the populations are well-mixed, and as the system grows, a constant fraction of infected cells will be able to infect new cells. This is reflected in the corresponding graph of G(x,x/a), which reaches an asymptote and does not decline. In Table we list several examples of fast and slow growth laws../journal.pone..t001Table 1Examples of different virus spread terms, G(x,y). G(x,y) Law of virus spread Fast (frequency dependent) Fast Slow Slow Slow SlowIn general, we can prove that the two scenarios above are the only possible outcomes, given the biological requirements imposed on the function G. As x increases, this function increases, and can either approach zero or a nonzero level. If it approaches a non-zero level, this does not necessarily need to occur via a monotonic approach to the asymptote. It is possible that the function G first increases, peaks, and then converges to a non-zero asymptote. For intermediate values of x the function G may have a more complicated structure than that shown in figure , but in the absence of any biological evidence of that it is a safe bet to assume the simplest shape with a minimal number of local extrema.How does the shape of G help us draw meaningful conclusions about the behavior of the biological system? It turns out that the function G is essential in determining the number and the stability properties of the equilibria of the system, and thus it will help us reason about long-term predictions on the treatment outcome.Equations (–) can be combined in a single equation,()where the function y(x) is a relationship between the number of infected and uninfected cells at equilibrium as the total system size grows; it is obtained from equation () and depends on the exact rate of cancer growth, F. If the cancer growth is exponential (F = ), we have y(x) = x/a, that is, at equilibrium, the infected cells comprise a fixed fraction of uninfected cells. Thus the function G(x,x/a) depicted in figure is just the left hand side of the equation for the equilibria, equation (). The right hand side is represented by horizontal dashed lines, whose level decreases with the viral replication rate β. The number of intersections corresponds to the number of equilibria in the system.We can see that the two graphs in figure exhibit different numbers of equilibria. First we consider figure (a), fast virus spread. In this case, the model always contains a parameter region in which exactly one equilibrium exists. If the viral replication rate, β, lies below a threshold (β<βc) then no equilibrium exists. If the viral replication rate lies above that threshold, the following is observed. As shown in figure (a) exactly one equilibrium is found. In other cases, it is possible that there are two or more equilibria for intermediate viral replication rates. (For example, if the function G(x,x/a) first rises and achieves a maximum before descending to its horizontal asymptote, or if it goes through a number of local extrema before approaching a horizontal asymptote.) The most important universal feature in all fast growth scenarios is that for sufficiently high values of β, there is exactly one equilibrium. Next, consider Figure (b), slow virus spread. Again, for an equilibrium to exist, the viral replication rate needs to lie above the threshold β>βc. If this is the case, the system is always characterized by the presence of not one, but two equilibria. Again, in some cases, it is possible that the intermediate values of β correspond to more than two equilibria.The biological interpretation of this analysis is as follows. We saw that for both modes of infection, if the values of the viral replication rate β are small, no equilibria exist. This translates into an uncontrolled cancer growth. This is an intuitive result: for low viral replication rates, treatment is impossible. A less intuitive result is connected with the number of equilibria once β is above its threshold value.The cancer-virus system displays a fundamentally different behavior depending on whether it is characterized by one or two equilibria. If there is only one equilibrium, then the dynamics will be governed by the properties of this equilibrium only. Because the number of tumor cells is relatively low at this equilibrium, this outcome corresponds to containment of the tumor by the virus. For convenience, we call this internal equilibrium EI. On the other hand, the situation is more complicated if the system is characterized by two equilibria. The first equilibrium, at which the number of tumor cells is lower, is again the internal equilibrium, EI, and can be interpreted as containment of the tumor by the virus. The second equilibrium can be shown to be an unstable saddle node equilibrium, call it ES. The presence of the saddle equilibrium means that the dynamics are qualitatively different depending on the initial conditions. If the initial number of tumor cells is relatively low and close to the internal equilibrium, then the dynamics are governed by this internal equilibrium, EI, leading to a degree of tumor control. If the initial number of tumor cells is higher and around or above the saddle node equilibrium ES, then the number of tumor cells increases in an uncontrolled fashion. Hence, in this regime, uncontrolled cancer growth is always a possible outcome.We conclude that our biologically defined modes of virus spread correspond to very different mathematical properties. Models of class I (fast virus spread) contain a parameter region (of high enough β) in which only a single equilibrium is observed. In this case, the model contains a parameter region in which uncontrolled cancer growth is impossible. Models of class II (slow spread) never have only one equilibrium and the saddle node equilibrium ES is present whenever the internal equilibrium EI exists. In this class of models, no matter how high β is, uncontrolled cancer growth is always a possibility.Effect of the tumor growth termFor the purposes of classification of the virus spread terms, we looked at the changes in G as the number of infected and uninfected cells grew in the same proportion. This led to a direct evaluation of the number of equilibria for exponential cancer growth (F = ). While mathematically the simplest scenario, exponential growth is an unrealistic assumption, because the growth of cells is bound to saturate as the tumor grows. Our methods allow to study any reasonable cancer growth law in a very natural way.Let us model a slow-down of the tumor growth rate as the number of tumor cells increases. This can be done in two different ways. On the one hand, we can assume that while tumor growth slows down, it never stops, such that the tumor can grow towards infinity over time. That is, there is no upper limit to the number of tumor cells; in practical terms growth will stop when the organism dies. An example is what we call “surface growth”, where only the cells around the surface of the tumor can give rise to viable daughter cells and can contribute to tumor spread. This can apply to solid tumors that have a high degree of spatial structure. Surface growth in 2D and 3D are listed in Table . The parameter η determines the tumor size at which saturation comes into play. Another possibility that falls into this category is that the rate of tumor growth becomes linear as the number of tumor cells increases. In this case, tumor growth is even slower; we refer to it as “linear growth”../journal.pone..t002Table 2Examples of different tumor growth terms, F(x+y). F(x+y) Growth Law 1Exponential Linear Surface growth in 2D Surface growth in 3D Logistic GompertzianOn the other hand, it is possible that there is a natural limit or carrying capacity, W, that limits tumor growth []. Thus, we will assume that growth slows down and eventually stops as the number of tumor cells increases. This can occur in a variety of ways. Tumor growth can be exponential until the number of cells approaches carrying capacity and the rate of tumor cell growth becomes zero. For example, this can be described by the logistic growth term (see Table ) []. Alternatively, we can assume that tumor growth first saturates according to the surface growth or linear growth patterns described above, and only reaches the carrying capacity once the tumor has grown to a significantly larger number of cells. Another example of a growth with a carrying capacity is a Gompertzian type growth [], Table .As mentioned before, the term F reflects implicitly both division and death properties of uninfected tumor cells. For example, an exponential growth is characterized by a net expansion rate resulting from exponential division and death processes. The logistic growth is a consequence of saturation of the division rate while the (exponential) death rate remains constant. In fact, any process with a sub-exponential division rate and an exponential death will be characterized by a finite carrying capacity. On the other hand, an unlimited (but saturated) growth (such as surface growth) implicitly includes death which happens slower than exponentially. If we were to add an exponential death term to a surface growth, it would lead to a limited growth with a carrying capacity. Our framework includes all these and any other reasonable functional forms of cellular growth.In the following, we will examine the effect of different types of tumor growth terms on the properties of the model. We will do this first in the context of the faster virus infection terms that belong to class I, and then in the context of the slower infection terms that belong to class II. Note that our analysis is quite general and the particular growth laws listed in Table are merely an illustration; the results are not restricted to these particular growth laws.Effect on fast virus growthWith this class of virus infection term, we found that in the context of exponential tumor growth, G(x,y(x)) with y(x) = x/a approaches a nonzero asymptote for large values of x (note that it can either rise monotonically to the asymptote, or first go through one or more local maxima before declining towards the asymptote). In either case, for any equilibrium to exist, the viral replication rate needs to lie above a threshold β>β c, and there exists a parameter region (characterized by values of β greater than a threshold) in which only the internal equilibrium EI is present. In this parameter region, tumor control is the only outcome.Introducing saturated tumor growth (or changing the function F in any way) will lead to a different functional form of y(x) in equation (). A universal feature is that any tumor growth slower than straight exponential growth will lead to smaller values of y(x) and thus to higher values of G. Therefore, as a result of tumor growth saturation, the asymptote becomes higher for slower tumor growth terms. This means that only the internal equilibrium EI can exist, as with exponential growth. The only difference lies in the viral replication rate threshold beyond which this equilibrium can exist and beyond which tumor control is possible. The slower the tumor growth, the lower the viral replication rate threshold required for virus-mediated control.If we assume saturated but limited tumor growth (i.e. growth stops at carrying capacity W), then the picture is similar for the most part, with one difference. After the term G(x,y(x)) has approached the asymptote, the curve G takes an upward turn in the vicinity of x = W, i.e. when the number of cells approaches carrying capacity. This means that the model acquires an additional equilibrium, which corresponds to the cancer growing to its carrying capacity W. In the systems with unrestrictive growth, this was equivalent to unlimited growth of the cell population to infinitely large sizes. This is illustrated with the dotted line in Figure (a). We can see that for x≪W, the curves for limited and unlimited growth laws look identical, and near the carrying capacity W they deviate../journal.pone..g002Figure 2The effect of a carrying capacity.The function G(x,y(x)) is plotted for two particular choices of the virus spread law and three different laws of cancer growth: exponential, surface growth and linear growth. (a) Fast virus spread, G(x,y) = x/(x+y+) and (b) slow virus spread, G = x/(x+)/(y+). The solid lines correspond to the unlimited cancer growth; the dotted lines - to a growth up to a given size, W. The parameters are: a = , η = and W = .So far, we have concentrated on the case where G(x,y(x)) increases monotonically towards an asymptote. Alternatively, the term G(x,y(x)) can rise to a peak and then decline toward a non-zero asymptote. In this case, including saturation into the tumor growth term F(x,y) leads to similar consequences. However, the hump in the function can disappear, eliminating any parameter region in which both equilibria can exist. In other words, with slower tumor growth, there is no parameter region anymore in which the tumor can escape the effect of the virus and grow out of control. Whether this occurs or not depends on the relative size of the two spatial scales involved. The first scale is defined by the tumor size at which the virus infection function G saturates and peaks in the context of exponential growth; this is entirely dependent on the properties of the viral growth term. Let us call this scale sv, where the subscript refers to “viral”. The second scale is given by the colony size at which the tumor growth law starts to deviate from exponential; we will call this scale st (where the subscript refers to “tumor”). When xt≤sv, the asymptotic value of G becomes sufficiently large such that the hump disappears. The disappearance of the hump makes treatment easier, and this occurs if tumor growth slows down before virus growth does.Effect on slow virus growthHere, we assume slower virus growth terms that belong to class II. In the context of exponential growth, the function G(x,y(x)) first increases, and then declines towards zero. This means that if equilibria exist, both the internal equilibrium EI and the saddle node equilibrium ES are aways present. Consequently, the possibility always exists that the tumor can out-run the virus infection and grow uncontrolled. Taking into account saturated tumor growth has the following effect (Figure (b)). (i) The function G can remain qualitatively the same; that is, it rises to a peak and then declines towards zero. (ii) Alternatively, the picture can change such that it does not decline towards zero, but towards a non-zero asymptote, while remaining a one-humped function. (iii) Finally, the picture can change further such that the function G increases monotonically towards an asymptote. Which outcome is observed depends on the exact nature of the functions F and G and also the relative size of the two spatial scales involved: the tumor size at which the virus infection term G saturates and peaks (sv), and the size st at which the pattern of tumor growth starts to deviate from exponential. Lowering the value of st relative to sv shifts the outcome from scenario (i) to (iii). As the value of st becomes similar to the value of sv, the model contains parameter regions in which only the internal equilibrium EI exists and in which uncontrolled tumor growth is impossible. If st≪sv, then the hump in the function G disappears, and the saddle node equilibrium ES is never present. In this case, virus-induced tumor control is the only outcome, and uncontrolled tumor growth cannot be observed. In biological terms, saturation of tumor growth at lower sizes promotes successful virus therapy. These arguments apply to all saturated tumor growth scenarios. With saturated but unlimited tumor growth, the function G approaches an asymptote for large tumor sizes x. For tumor growth that is limited by a carrying capacity W, the function G eventually deviates from the asymptote and rises again, indicating the presence of an equilibrium that describes tumor growth towards carrying capacity rather than towards infinity. Lowering the carrying capacity W has the same effect as lowering the parameter st that determines the tumor size at which growth starts to saturate: it shifts the outcome from scenario (i) to (iii).Summary of model propertiesIn summary this analysis has provided the following insights. We examined two types of infection terms and found that they strongly influence the dynamics of oncolytic virus spread. In the first class of models, virus spread was fast because infected cells are mixed among uninfected cells. In this case, tumor control is always observed if the viral replication rate lies above a threshold. In these parameter regions, loss of tumor control is not observed. In the second class of models, virus spread was assumed to be slow, because infected cells are clustered together in space. In this situation, the model can be characterized by bistability. If the initial number of tumor cells lies below a threshold, tumor control is observed. If the initial number of tumor cells lies above this threshold, uncontrolled tumor growth is observed. If tumor growth only saturates at high numbers of tumor cells or not at all, then uncontrolled tumor growth is always possible in parameter regions in which tumor control is possible. If tumor growth saturates at lower levels, there are parameter regions in which only the tumor control outcome is observed and in which uncontrolled tumor growth is not possible. If tumor growth saturates at even lower levels, then the bistability and the dependence on initial conditions vanishes completely.Properties of the internal equilibriumThe above analysis concentrated on the equilibria. By examining which equilibria exist under different conditions, we can obtain information about the ability of the virus to control the cancer, and about the possibility that the cancer grows despite the presence of the virus. If the dynamics are governed by the internal equilibrium EI, then the virus keeps the tumor cell population at relatively low levels and prevents uncontrolled tumor growth. We have discussed the conditions under which this can be achieved and interpreted these conditions from a biological angle. If the virus does control the tumor, however, additional questions arise. The virus can either control a persisting tumor at low levels, or the virus can drive the tumor cell population extinct.Because we are considering ordinary differential equations that describe the average behavior of the cell and virus populations, true extinction cannot occur in this model. The number of cells can, however, drop to very low levels. If the average number of cells is below one, we can assume that tumor extinction is a likely event. Therefore, if the number of tumor cells at equilibrium lies below one, we can say that the virus is likely to drive the tumor extinct. However, even if the equilibrium number of cells lies above one, the tumor cell population can still go extinct during oscillatory dynamics that can occur before the dynamics reach equilibrium. Therefore, we need to understand the properties of the internal equilibrium in more detail. We will examine this in the context of both fast and slow virus growth. We will only assume saturated tumor growth and not consider straight exponential tumor growth.Fast virus growthOne of the most important parameters that influence the properties of the internal equilibrium is the replication rate of the virus β. In general, the faster the replication rate of the virus, the lower is the equilibrium number of tumor cells. Further, it can be shown that if the viral replication rate β crosses a threshold, the behavior near the equilibrium becomes oscillatory. Both promote the eradication of the cancer. In general, the internal equilibrium can either be stable or unstable, depending on the particular model under consideration as well as parameter values.Let us first consider the case where the equilibrium is stable. Then we can distinguish between two parameter regions. Denote the size at which tumor growth slows down and deviates from exponential by st. In the first parameter region, the value of st is large compared to a value related to the virus scale, sv (for the exact definition see the Supporting information S1). In this parameter region, we observe a viral replication rate threshold, at which the equilibrium number of tumor cells drops sharply from relatively high values to values of the order (figure 3a). This replication rate threshold can be defined for individual models that belong to this class and defines the condition for cancer eradication. If the tumor size at equilibrium drops to small values (of the order of cell), stochastic effects are very likely to lead to extinction. This is further supported by changes in the oscillatory approach to the equilibrium, which we have investigated in the context of individual models (figure b). At this viral replication rate threshold, the amplitude of the initial oscillations can increase sharply, as can the time it takes for the dynamics to approach the stable equilibrium (the real part of the eigenvalues of the Jacobian matrix rapidly approaches zero). Since pronounced oscillations reduce the number of tumor cells well below one, tumor eradication is the likely outcome. Note, however, that this drastic change in the oscillatory pattern is not observed in all models that belong to this class. The sharp drop in the equilibrium value is, however, a universal feature of models that belong to this class../journal.pone..g003Figure (a) The equilibrium number of uninfected cancer cells as a function of the viral replication rate β for fast virus growth. There is a threshold viral replication rate at which the number of cancer cells drops sharply from relatively high values to values of the order of one. This can be considered a tumor extinction threshold. (b) Dynamics of the uninfected cancer cells if the viral replication rate lies below (left) and above (right) this threshold. If the viral replication rate lies below the threshold, limited oscillations are observed that dampen out quickly. If the viral replication rate lies above the threshold, extensive oscillations are observed that reduce the cancer cell population to very low levels, and that dampen out very slowly (dampening not observed on time scale shown here). These plots were made by using a specific model from the fast virus growth category, that is G = (ε+) x/(x+y+ε). Note that the transition in oscillations is not a universal feature of all models in this class. Parameters were chosen as follows: r1 = ; a = .; ε = ; η = ; x0 = ; y0 = ; For (b), β = . and β = ., respectively.Now assume the other parameter region in which the scale st is small. In this case, no such viral replication rate threshold exists. Instead, the equilibrium number of tumor cells declines proportional to the viral replication rate β. Numerical simulation of individual models, however, indicates that the minimum number of tumor cells can decline exponentially with an increase in the viral replication rate, although this could not be proved in general. Taken together, these findings indicate that in the parameter regions where virus replication is fast enough such that there is an oscillatory approach to the equilibrium, tumor eradication is the likely outcome.As mentioned above, it is also possible that the internal equilibrium EI is unstable. In this case, we observe oscillations that diverge away from the equilibrium if the viral replication rate β is sufficiently fast. That is the amplitude of the oscillations increases over time. This is likely to correlate with extinction of the tumor, especially if the number of tumor cells at equilibrium is relatively low. This is because the oscillations will reduce the number of tumor cells well below the equilibrium value over time. Thus we conclude that for sufficiently large values of β, the cancer will be driven extinct by the virus through (convergent or divergent) oscillations.Slow virus growthIn this case, the tumor size at the internal equilibrium is again negatively correlated with the viral replication rate β. Similarly to fast virus growth, the internal equilibrium can be stable or unstable depending on the individual model and on the parameter values. The dynamics will be discussed for both stable and unstable equilibria EI.If the equilibrium is stable, the approach is again oscillatory if the viral replication rate is sufficiently large (Figure 4a and 5a). Numerical simulations of individual models indicate that the minimum tumor size during these oscillations can decline exponentially with the viral replication rate β, although again this could not be proved in a general setting. These results indicate, however, that if oscillations are observed it is likely that the cancer is eradicated by the virus (Figure ). Note that this assumes that the initial number of tumor cells is sufficiently small such that the population is in the region of attraction of the internal equilibrium. If this is not the case, the virus fails and unlimited virus growth occurs because the long-term outcome depends on the initial conditions as discussed above (Figure a). In addition to these dynamics, the following can occur (Figure (a)). Assume that the tumor cell population is reduced to low levels during the initial oscillations, but not to extinction. As the tumor cell population rises again, it can actually cross over to values larger than the saddle equilibrium ES. Consequently, the cancer will grow uncontrolled and virus therapy will fail../journal.pone..g004Figure 4The phase portrait for a system with a slow virus propagation term.(a) The intermediate equilibrium, EI, is stable (the basin of attraction is shaded), (b) EI is unstable../journal.pone..g005Figure 5Dynamics in fast virus growth models assuming that the internal equilibrium EI is (a) stable and (b) unstable.(a) If the internal equilibrium is stable, then the dynamics can converge to this equilibrium via damped oscillation if the initial number of cancer cells is relatively low. On the other hand, if the initial number of cancer cells is relatively high, then uncontrolled cancer growth is observed. (b) If the internal equilibrium is unstable, then diverging oscillations are observed. Eventually, these diverging oscillations take the populations beyond the saddle node equilibrium, leading to unlimited cancer growth. Before that occurs, however, it is most likely that the cancer has been driven extinct in a stochastic setting because the diverging oscillations drive the tumor size to ever decreasing values. These plots were obtained from a specific model that belongs to the slow virus growth class, i.e. . Parameters were chosen as follows: (a) r = ; β = .; a = .; ε1 = ; ε2 = ; η = ; x0 = and ,, respectively; y0 = . For (b) r = ; β = ; a = .; ε1 = ; ε2 = $; η = ; x0 = ; y0 = .On the other hand, if the internal equilibrium is unstable, then the following is observed (Figure 4b and 5b). If the viral replication rate is fast enough, the populations show diverging oscillations away from the equilibrium (Figure 4b), i.e. the amplitude of the oscillations increases over time. During these diverging oscillations, the minimum number of tumor cells declines over time. Hence, the tumor is likely to hit extinction. Again, there is the possibility that during the oscillations the tumor cell population crosses over to values larger than the saddle equilibrium ES. In this case, the tumor cell population would grow uncontrolled to ever increasing levels.To summarize, for slow virus growth oscillations around the internal equilibrium have the potential to drive the tumor cell population extinct. However, the bi-stability of this system causes problems since there is always the possibility that the populations can escape to large numbers, leading to uncontrolled tumor growth.Application of models to experimental dataHere we fit our models to previously published experimental data and discuss implications for model validation, model selection, and further experimental work. We examined data published by []. This study considered A549 human lung cancer nude mouse xenografts, and infected them with the wild-type adenovirus Ad309 and a mutant virus Ad337 (characterized by a deletion in the E1b-19kD gene). The resulting dynamics were investigated under two conditions. (i) Under the first condition, the cancer cells were used to establish subcutaneous tumors in the mice. When the tumors reached a certain size, the virus was injected into the tumor. (ii) In a second scenario, infected cells were first mixed with uninfected cells, and the mixture was injected into the mice. The first scenario corresponds to spatially more restricted virus growth, while in the second scenario there is a higher degree of mixing between infected and uninfected tumor cells due to the experimental protocol. For both scenarios, we fitted models that differ in the infection term G and the tumor growth term F. We performed non-linear least squares regression, using standard software. The exact models that were used are provided in Figures and . The parameter estimates obtained for all fits are tabulated in the Supporting Information S1../journal.pone..g006Figure (a) Data on the growth of A549 human lung cancer nude mouse xenografts in the absence of the virus []. Different tumor growth models were fitted, see Table . The parameter values and the root mean square values are summarized in the Supporting Information S1. The graph on the right plots the predicted long-term growth curves. (b) Growth dynamics in the presence of the wild-type virus Ad309, which was injected into an established tumor. Both a slow model and a fast model were fitted. For the slow model, G = x/(x y1/+ε). For the fast model, G = x/(x+y+ε). Tumor growth was assumed to be logistic, F = −(x+y)/W. For the slow model, different parameter combinations are shown that fit the data to a similar degree (slow1, slow2). The graph on the right shows the predicted long term dynamics. Parameter values and root mean square values are given in the Supporting Information S1../journal.pone..g007Figure (a) Data on the growth of A549 human lung cancer nude mouse xenografts in the presence of the wild-type virus Ad309, assuming that infected and uninfected cells were mixed before the tumor cells were injected into the mouse []. A slow and a fast model were fitted. For each model, different parameter combinations were found that fit the model comparably (slow1, slow2, fast1, fast2).The graph on the right side shows the predicted long term dynamics for the different models and parameter combinations. (b) Infection with the mutant As337 virus, where again the infected and uninfected cells were mixed before the tumor was injected into the mice. Again, a slow and a fast model were fitted, and with each model different parameter combinations were found that provided a comparable fit to the data. As before, the graph on the right hand side shows the predicted long-term dynamics. For the slow model, G = x/(x y1/+ε). For the fast model, G = x/(x+y+ε). Tumor growth was assumed to be logistic, F = −(x+y)/W. Parameter values and root mean square values are given in the Supporting Information S1.We first fitted the control tumor growth in the absence of the virus (Figure (a)). Both exponential growth and saturated growth models (logistic, gompertzian, surface and linear, see table ) were applied. The saturated growth models fit the data better than exponential growth. All saturated growth models fit the data well, the logistic growth yielding the lowest (by a small margin) root mean square (RMS) error. For convenience, we chose logistic tumor growth as the basis for analyzing the effect of virus infection.First, consider experimental condition (i), where the tumor was allowed to grow in the mice before the virus was inoculated. Only the wild-type virus Ad309 is considered. In this experiment, tumor growth was significantly reduced by the virus. However, tumor size reached a plateau by day , despite the persistence of the virus, leading to the conclusion that the virus failed to eradicate the tumor cell population. For fitting purposes we considered one fast and one slow virus spread term, the first and the third in table . Figure (b) shows that both a fast and a slow virus growth model can fit the data. However, extrapolating beyond the experimental time frame, very different long term outcomes are observed. The fast model predicts that the tumor remains at relatively low levels, controlled by the persisting virus infection. With the slow model, we show two parameter combinations which both fit the data well, but which are characterized by different long term outcomes. For one parameter combination (slow ), damped oscillations are observed that lead to persistence of both the tumor and the virus at relatively low levels. For the second parameter combination (slow ), the tumor cell population escapes control and grows to high levels. The virus population persists at low and ineffective levels (not shown). Therefore, not only do different models predict different long term dynamics; within one model, different parameter combinations that describe the data equally well can give rise to different predictions regarding the long-term dynamics. Our discussion of this and other results is postponed until the end of this section.Next consider experimental condition (ii), in which infected cells were mixed with uninfected cells at a ratio of ∶ before the tumor was injected into the mice. In this case, the viruses were generally more effective. Tumor growth was prevented, and the number of tumor cells declined to low levels. Figure fits a slow and a fast model to data that document infection with the wild type virus Ad309 (Figure 7a) and the mutant virus Ad337 (Figure 7b). Consider the wild type As309 virus first (Figure 7a). All models fit the data well. Again, the predictions about the long-term dynamics vary, not only between models, but between different parameter combinations of the same model. Two qualitatively different outcomes are depicted in Figure 7a. On the one hand, the cancer can grow out of control following the initial reduction in the number of cancer cells. On the other hand, the virus maintains control of the cancer, which persists but is suppressed to relatively low levels. Thus, the encouraging but limited trend shown by the data cannot be used to conclude efficient virus-mediated tumor control. Longer experimental studies are needed in order gain insights into the eventual outcome of treatment, and to differentiate between the various model predictions. Figure 7b shows the same analysis for the mutant virus Ad337. As before, the slow and fast model can both fit the data well, and within one model, different parameter combinations are possible. The long-term dynamics show different outcomes, depending on the model and the parameter combinations. They include long term virus-mediated cancer control, as well as uncontrolled cancer growth. In the context of the experimental data, however, these long-term dynamics will not be observed, as the cancers regressed completely in the experiments. During the initial decline of the cancer cell population in the model, the number of cells drops to such low levels that extinction is actually the likely outcome in practical terms. However, what this tells us is that if by chance the cancer cell population does not hit extinction in the experiments, it is entirely possible that the cancer cell population rebounds and grows to high levels, depending on the model and its parameters.As mentioned above, the experiments include both a highly spatial setting where the virus was inoculated into an already established tumor, and a mixed setting where infected and uninfected cells were mixed before the tumor cells were placed into the mouse. Therefore, it can be tempting to examine whether the relative goodness of fit for the fast and slow models is different in these two situations. As explained, however, each model can fit the data with several alternative parameter combinations. There are many more solutions to the least squares regression than shown here. Therefore, it does not make sense to compare the goodness of fit for slow and fast models. For instance if the fits obtained for the slow model are slightly better than those obtained for the fast model, it is quite possible that there exists another parameter combination in the fast model that is better yet, and that has not been encountered so far. This brings us to the fundamental problem of nonlinear data fitting and model validation, which is an interesting issue in itself and will be discussed briefly here.As with many (and perhaps most) other nonlinear models, the parameter space where the minimization of the RMS error is performed, is multidimensional and is characterized by many shallow local minima. Most standard fitting routines get “stuck” at local minima, and even more sophisticated algorithms aimed at finding the global minimum are not very useful, because the difference between the global minimum and many runner-ups is usually insignificant and cannot serve as an indicator of the “right” or “best” fit. Therefore, to attack the fitting problem, one is required to repeat the minimization procedure multiple times, either by performing an exhaustive span of the space of the initial guesses, or by implementing a Monte-Carlo method. The statistics of the outcomes are then analyzed in the hope to find clusters of good fits, which are then assumed to be indicative of the solution of interest.The unfortunate part is that most of the times, these sophisticated statistical techniques are not very useful because the data sets that the fitting is applied to are simply too sparse, and they probably do not contain enough information to distinguish between models. Some of the deficiencies of experimental data sets are (i) an insufficient number of time-points, (ii) a very large experimental error at each time point, due to the experimental difficulties as well as a small sample size, and (iii) the long-term dynamics is often not captured due to the time-constraints of the experiment. In other words, no sophisticated statistical data manipulation can help distinguish between models if the data set is too sparse, short and contains large scatter.What can we conclude from these considerations and our own attempts to validate the models based on published experimental data? The good news is that at least some of the models contain parameter combinations which describe the existing data reasonably well. The bad news is that model validation/rejection was not possible in the particular system that we used. If data were collected over longer periods of time, and with a larger sample size, then the number of parameter combinations that can fit the data would be significantly reduced, and allow for more meaningful model comparison.DiscussionIn this paper we presented the first modeling approach that tries to analyze the dynamics of oncolytic viruses in a general setting, going beyond particular models in which results can easily depend on mathematical terms chosen. Previous approaches to modeling oncolytic virus dynamics, and virus dynamics in general, have been based on particular models that include uncertain and unrealistic assumptions. The most striking is the assumption about the infection term, which usually assumes perfect mixing of populations, and which is certainly violated in any biologically realistic setting.Our method can be considered a hybrid between such space-free, mass-action approaches, and much more complex methods involving spatial network ideas, e.g. []–[]. The former approach fails to capture spatial and geometric constraints which play an important role in infection spread. The latter approach is only analytically tractable to a certain degree; also, it usually relies on a particular, given, set of rules that govern the infection spread. Our investigation aims to capture general trends that arise from different assumptions on the infection mechanism. It combines the analytical tractability of simple dynamical systems with a more realistic modeling of infection spread.We found that based on the infection term, we can divide models into two categories with fundamentally different behavior. In one group, virus growth is relatively fast because the infected cells are dispersed among the uninfected cells rather than being clustered together. In this case most infected cells contribute to virus spread. In these models, there is a clear viral replication rate threshold beyond which the number of cancer cells drops to levels of the order of one or less, corresponding to extinction in practical terms. Under this parameter region, this is the only outcome in this class of model. In the other category, infected cells are assumed to be clustered together to some degree in a mass, which might be realistic for solid tumors. In this case, only the infected cells located at the surface of the cluster contribute to virus spread because they are in the vicinity of uninfected cells. The infected cells located in the center of the cluster are surrounded only by other infected cells and therefore do not contribute to virus replication. The larger the number of infected cells, the smaller the proportion of cells that can pass on the virus. In this scenario, virus therapy is more difficult. If tumor growth saturates only at relatively large sizes or does not saturate, then even in the parameter regions where the dynamics can converge to tumor control or eradication, there can be the possibility that the cancer can outrun the virus if the number of cancer cells lies above a threshold at the start of virus therapy. This is because of the existence of the saddle node equilibrium which ensures dependence of the outcome on initial conditions. This might be problematic in clinical settings, because there is only a relatively small window between the size at which the tumor becomes detectable (about cells) and the size at which it can induce mortality (around cells).Tumor growth saturation at lower levels introduces a parameter region in which only the tumor control outcome is possible. A further reduction in the number of tumor cells at which growth saturation occurs can abolish the existence of the saddle node equilibrium altogether. In this case, the only outcome is tumor control. This result makes intuitive sense: earlier saturation of tumor growth slows down the cancer, and makes it easier for the virus to gain the upper hand. It also means that if the tumor is found early, it might be possible to slow down tumor growth by means of more conventional drug therapy, enabling the virus to control the cancer and to prevent runaway growth. There is indication in clinical data that a combination of chemotherapy and oncolytic virus therapy leads to better results than either approach alone [].Another important finding of our study is that the basic results regarding the outcome of oncolytic virus therapy do not depend on the particular tumor growth terms used in the model. The exact kinetics of tumor growth are still poorly understood and a source of uncertainty. We examined straight exponential growth, as well as a number of more realistic options, including saturated but continued growth at high numbers of cancer cells, as well as cessation of growth as the number of tumor cells approaches an upper limit. While there are minor differences (such as the existence of a stable equilibrium at large tumor sizes vs continues slow growth), the properties of the tumor control equilibrium are largely independent from the exact way in which tumor growth is modeled.Throughout this paper we discussed the ability of the virus to eradicate the tumor in the context of our mathematical model that aims to describe oncolytic virus growth in relatively simple settings. It is important to point out that even simple scenarios could be characterized by complicating conditions which are not captured in the model and which make actual tumor extinction difficult to achieve. For example, tumor cells might become resistant to the virus by for example down-regulating the receptor required for viral entry []. Related to this, cells could temporarily become resistant to virus-induced effects depending on the stage of the cell cycle []. Such effects can be easily incorporated into our framework, if data suggest that they play a role in determining the dynamics of oncolytic virus growth.The framework presented here aims to bring us closer towards predictive computational models of oncolytic virus replication in vitro. Both additional computational and experimental work will be necessary to advance this framework. On the theoretical side, it will be important to also explore spatially explicit and stochastic models. The ordinary differential equations are desirable because they can be applied to experimental data in a relatively straightforward way. At the same time, however, spatial aspects of population growth can only be captured in a phenomenological way. Hence, it will be important to consider a spatially explicit model and to compare its properties to the results obtained here. On the empirical side, it will be important to run experiments that document the growth of specific oncolytic viruses e.g. in a culture of specific tumor cells. These data can be fitted to the various models explored here to determine which model describes the data best and which models can be rejected. This can be done in a variety of setting: a culture where cells and viruses can mix well; a 2D tissue culture which imposes a degree of spatial constraints; and a 3D tissue culture which can impose further spatial constraints. Different models will apply to these different scenarios. This will allow us to test the theoretical notions presented here, and to obtain a set of models that are predictive for the relevant scenarios.Of course, for clinical relevance, oncolytic virus replication needs to be considered in the context of more complex settings. Most importantly, the virus is immunogenic, and immune responses can inhibit the spread of the virus and can even drive it extinct. Such components will have to be incorporated into a mathematical model that describes the replication of an oncolytic virus in vivo. However, before we have obtained a solid understanding of the principles that govern the dynamics of oncolytic viruses in simpler settings, it is unlikely that modeling can contribute much to understanding the more complicated in vivo scenario. The modeling framework discussed here provides a basis to incorporate increasing amounts of biological complexity in the future, and thus to gradually improve our understanding of the key factors that determine the outcome of oncolytic virus therapy.Materials and MethodsThe results described in this paper are based on the analysis of ordinary differential equations. Extensive mathematical details are provided in the Supporting Information S1.Supporting InformationSupporting Information S1(. MB PDF)Click here for additional data file.
|
PMC1513589.txt
|
TITLE: A new extract of the plant calendula officinalis produces a dual in vitro effect: cytotoxic anti-tumor activity and lymphocyte activation AUTHORS: Eva Jiménez-Medina, Angel Garcia-Lora, Laura Paco, Ignacio Algarra, Antonia Collado, Federico Garrido ABSTRACT: BackgroundPhytopharmacological studies of different Calendula extracts have shown anti-inflamatory, anti-viral and anti-genotoxic properties of therapeutic interest. In this study, we evaluated the in vitro cytotoxic anti-tumor and immunomodulatory activities and in vivo anti-tumor effect of Laser Activated Calendula Extract (LACE), a novel extract of the plant Calendula Officinalis (Asteraceae).MethodsAn aqueous extract of Calendula Officinalis was obtained by a novel extraction method in order to measure its anti-tumor and immunomodulatory activities in vitro. Tumor cell lines derived from leukemias, melanomas, fibrosarcomas and cancers of breast, prostate, cervix, lung, pancreas and colorectal were used and tumor cell proliferation in vitro was measured by BrdU incorporation and viable cell count. Effect of LACE on human peripheral blood lymphocyte (PBL) proliferation in vitro was also analyzed. Studies of cell cycle and apoptosis were performed in LACE-treated cells. In vivo anti-tumor activity was evaluated in nude mice bearing subcutaneously human Ando- melanoma cells.ResultsThe LACE extract showed a potent in vitro inhibition of tumor cell proliferation when tested on a wide variety of human and murine tumor cell lines. The inhibition ranged from to %. Mechanisms of inhibition were identified as cell cycle arrest in G0/G1 phase and Caspase--induced apoptosis. Interestingly, the same extract showed an opposite effect when tested on PBLs and NKL cell line, in which in vitro induction of proliferation and activation of these cells was observed. The intraperitoneal injection or oral administration of LACE extract in nude mice inhibits in vivo tumor growth of Ando- melanoma cells and prolongs the survival day of the mice.ConclusionThese results indicate that LACE aqueous extract has two complementary activities in vitro with potential anti-tumor therapeutic effect: cytotoxic tumor cell activity and lymphocyte activation. The LACE extract presented in vivo anti-tumoral activity in nude mice against tumor growth of Ando- melanoma cells. BODY: BackgroundPlants have a long history of use in the treatment of cancer. Active principles of Angelica Gigas, Catharanthus roseus, Podophyllum peltatum, Podophyllum emodii, Taxus brevifolia, Ocrosia elliptica, and Campototheca acuminata have been used in the treatment of advanced stages of various malignancies in the clinical setting [,]. Furthermore, many phytochemicals with different pharmacological properties have shown responses for the prevention or treatment of different tumors, e.g., flavones, flavanols, isoflavones, catechins, and taxanes [-]. Numerous drugs are used in cancer chemotherapy but most exhibit cell toxicity and can induce genotoxic, carcinogenic and teratogenic effects in non-tumor cells [,]. These side effects limit the use of chemotherapeutic agents despite their high efficacy in treating target malignant cells. Therefore, the search for alternative drugs that are both effective and non-toxic in the treatment of cancers is an important research line []. In fact, increased efforts are being made to isolate bioactive products from medicinal plants for their possible utility in cancer treatment [].Flowers of the plant Calendula officinalis, commonly known as "Marigold", are used in the West and in Asia for their anti-inflammatory properties [,].Phytopharmacological studies of different calendula extracts have shown anti-viral activity, anti-HIV properties of therapeutic interest [], and anti-genotoxic properties []. In clinical studies, Calendula was highly efficacious in the prevention of acute dermatitis in cancer patients undergoing postoperative irradiation []. Its cytotoxic effect on tumor cell lines in vitro and its anticancer efficacy in vivo was briefly outlined yrs ago []. Chemical constituents of C. Officinalis include some triterpenes, triterpene oligoglycosides, and flavonol glycosides [,]. The aim of the present study was to evaluate the in vitro cytotoxic anti-tumor and immunomodulatory activities of a novel extract of the plant Calendula Officinalis (LACE: Laser Activated Calendula Extract). We show that LACE demonstrated a potent in vitro growth inhibition of several tumor cell lines, whereas it induced proliferation and activation of peripheral blood lymphocytes (PBLs). The mechanisms of this inhibition were identified as cell cycle arrest and apoptosis induction. Furthermore, the LACE extract presented anti-tumor activity in vivo in nude mice.MethodsPreparation of the Calendula Extract (LACE)The aqueous extract of Calendula officinalis (LACE) was obtained by subjecting the flower of this plant to the following patented process (patent n° EP B1). First, the flowers were washed with water and comminuted by conventional methods. The flowers were treated with laser radiation at a wavelength of nm for min, followed by the suspension of to g of laser-treated plant in L of water. The suspension was placed on a rocking platform at °C for – days and periodically treated with laser during this period. Then, the liquid phase was separated from the solid, an ochre-colored aqueous extract was obtained and stored in a freezer at -°C. A second aqueous extract of Calendula officinalis was obtained following an identical process but without laser treatment (calendula extract, CE).For the assays, the extract was concentrated to dryness in a lyophilizer, the dried preparation was homogenized in culture medium at concentration of mg/ml (stock solution), and filtered through a . μM Millipore.Cell lines and cell cultureThe following tumor cell lines were used: B16 murine melanoma, B9 murine MCA-induced fibrosarcoma, ANDO- human melanoma, MDA MB231 human breast cancer, AGS human gastric cancer, DU- human prostate cancer, A- human lung cancer, IMIN PC- human pancreatic cancer, DLD1 colon carcinoma, HeLa human cervical adenocarcinoma, U937 monocytic and Jurkat T lymphoma leukemias, and NKL, which was established from PBLs of a patient with LGL leukemia []. All cell lines were obtained from the American Type Culture Collection (Manassas, USA) except for the B9 cell line, which was generated at our laboratory, and the Ando , IMIM PC-, and NKL cell lines, kindly provided by P. Coulie (Unite de Genetique Cellulaire, Louvain University, Brussels, Belgium), F. X. Real (Instituto Municipal de Investigaciones Medicas, Barcelona, Spain), and Dr. M. Lopet-Botet (Universidad Pompeu-Fabra, Barcelona, Spain), respectively. Cell lines derived from solid tumors were grown at °C in a humidified atmosphere of % CO2 in DMEM culture medium (Gibco, Paisley UK) supplemented with % heat-inactivated fetal bovine serum (Life Technologics, Milan Italy), antibiotics, and glutamine. U937 and JURKAT leukemia cells were cultured in RPMI. The NKL cell line was cultured in RPMI with % heat-inactivated human AB serum (Sigma Chemical, St Louis, MO; USA) and human recombinant IL- ( U/ml; purity>%, specific activity, × U/mg) (Roche, Nutley, NJ; USA).Lymphocyte proliferation assayHuman lymphocytes were isolated from venous blood by Ficoll-Hystopaque separation method. Proliferation of PBLs was analyzed in vitro using -bromo-'-deoxyuridine (BrdU) labeling of DNA-synthesizing cells with the Brdu colorimetric ELISA Cell Proliferation kit, (Roche Diagnostics). PBLs were seeded in well-microculture plates at a cell density of × per well. A dose/response curve was performed using different concentrations of LACE or CE ( mg/ml to μg/ml). Concanavalin A ( μg/ml, Sigma) and IL- were used as positive control. After h of culture in presence or absence of LACE, BrdU labeling reagent (final concentration μM) was added and the cells were cultured during h, then cells were fixed for min and incubated with anti-BrdU for h at °C. μl of tetramethyl-benzidine (TMB) was used as substrate. Optical densities were determined at nm using an ELISA microplate reader (Biotek, Power-Wave XS). Controls were: culture medium, cells cultured only in medium and cells incubated with anti-BrdU in absence of BrdU. All experiments were plated in triplicate wells and were performed at least three times.In vitro cytotoxicity assaysThe effect of LACE or CE on tumor cell proliferation was assessed by measuring BrdU incorporation with the kit described above. Cells were plated in -well microculture plates ( × cells/well). After h of culture at °C, a dose/response curve was performed as described above. Every h, the culture medium was replaced and LACE or CE was added. After – h BrdU labeling reagent was added and cultured for a further – h. Assays were also performed by counting viable cells using Trypan Blue exclusion. Briefly, cancer cell lines were seeded into culture tissue-flask (.– × /culture tissue-flask) or -well plates (NKL cells) and incubated for h at °C in a humidified atmosphere of % CO2. Cells were then treated with μg/ml of LACE in the culture medium, which was replaced every h. After – days, the cells were collected by centrifugation and a small sample of the cell suspension was diluted in .% Trypan Blue and cells were counted in a hemocytometer chamber. Each cell sample was counted in this way at least three times and each assay was repeated at least three times.Cell cycle distribution analysisBriefly, cells were plated in six well plates ( × per well) or in culture tissue-flask ( × ) and exposed continuously for days to μg/ml of LACE. The rate of DNA synthesis was examined by a BrdU incorporation method using FITC BrdU Flow Kit (BD Pharmingen) according to manufacturer's instructions. BrdU was then detected by using a method with DNasa cell treatment and FITC-conjugated anti-BrdU antibody. Cells were washed with ml of × BD Perm/Wash Buffer and μl of -amino-actinomycin D was added. Analysis was performed with cells using Cell Quest Software and FACScan flow cytometer (Becton-Dickinson).Analysis of expression of cyclins and cyclin-dependent kinases (CDKs)The cell lines AGS and JURKAT were exposed to μg/ml LACE during – hours. The cells were washed with PBS and permeabilized with Citofix/Citoperm min to °C, incubated with anti-cyclin D1 and E antibodies (BD Pharmigen) and finally propidium iodide was added. The cells were analyzed by inmunofluorescence as mentioned above.Western blot analysis was performed for the analysis de CDK (CDK1/Cdc2, CDK2, CDK4, CDK6) and cyclins (A, B y D3). The Kit TransFactor Extraction (BD Biosciences) was utilized to extract whole proteins. Protein samples were separated by SDS/PAGE (%) and then electroblotted onto a polyvinylidene difluoride membrane (Kit, Mini Trans-Blot Electrophoretic Transfer Cell; Bio-Rad). The membranes were saturated with % nonfat dry milk in TBST ( mM Tris HCl, mM NaCl and .% Tween ) and then incubated with anti-CDK1/Cdc2, CDK2, CDK4, CDK6, cyclin A, cyclin B and cyclin D3 antibodies. The goat anti-Mouse IgG-HRP (BD Biosciences Pharmingen) was used as the second antibody. The membranes were washed thoroughly with TBST and incubated for to min with colorimetric solution, Opti-4CN (Bio-Rad, Madrid, Spain).Annexin V binding assay to detect apoptotic cellsAfter treatment of cancer cells with LACE for four days, cells were detached from the culture tissue-flask with PBS containing mM EDTA. These cells were then collected together with floating cells, washed twice with cold PBS, and resuspended in binding buffer at a concentration of × 106cells per ml. μl of solution was incubated for min at °C with μl of Annexin V-PE antibody (BD Biosciences), and μl of -amino-actynomycin D was then added. Cells were incubated for min in darkness, and μl of staining buffer was added before flow cytometry analysis.Assay for active caspase- expressionFITC conjugated monoclonal anti-active-caspase- antibody (BD Biosciences) was used to determine whether the protease Caspase- was involved in apoptosis induced by LACE. After treatment with LACE for days, the cancer cells were washed twice with cold PBS and fixed and permeabilized in Cytofix/Cytoperm buffer. Then, cells were incubated with FITC-conjugated monoclonal rabbit anti-active human-caspase- antibody for min. Cells were washed twice and μl of × Perm Wash Buffer was added before analysis by flow cytometry.In vivo toxicity assaysFor in vivo studies, immunocompetent - to -week-old Balb/c, C57/BL6 and CBA mice and -month-old Wistar rats were obtained from the Animal Centre of our institution. Mean weight of mice and rats was g and g, respectively. Animals were housed in wire-topped plastic boxes kept on a hour light/ hour dark cycle under pathogen-free conditions. All studies of the animals were performed according to guidelines approved by our institution. Animals were randomly divided into groups of mice or rats. LACE extract (. ml) was administered orally by cannula daily for weeks and the control group was similarly treated with . ml water. LACE concentrations used were: mg/kg/day, mg/kg/day, mg/kg/day, and mg/kg/day. After weeks of treatment, the animals were maintained for weeks under standard conditions and assessed daily for systemic (listlessness, weight loss) or local (alopecia, skin reaction and leg motility) toxicity.Effect of LACE on solid tumor growth in nude miceAthymic nude mice weeks were obtained from Charles River (CRL, Barcelona) and all studies of the animals were performed according to guidelines approved by our institution. The human melanoma cell line ANDO- ( × cells) was injected subcutaneously at the back foot pad of mice. When tumors became visible about one week after injection, the animals were randomized into six groups: LACE-oral treated ( mg/Kg of weight), LACE-intraperitoneal treated ( mg/Kg of weight), taxol-intraperitoneal treated ( mg/Kg of weight), control, controls treated with saline solution by oral or intraperitoneal route. LACE was administered orally three times a week during weeks, LACE intraperitoneal route twice a week during weeks and taxol was administered intraperitoneally twice a week during three weeks. All products were firstly administered in day after cells injection. Tumor growth (or the long tumor diameter) was monitored with calipers three times a week measuring long tumor diameter. Results are expressed as mean size of tumors from each group ± SD. Each group was composed of mice and the assays were performed at least three times. A comparison of the survival percentage between treated and control group was performed.ResultsLACE extract exerts mitogenic activity on PBLsA complete dose-response study was performed to determine the action of LACE or CE on human PBLs, plating × cells in a -well tissue plate for – h with one of concentrations of extract as follows: concentration n° = mg/ml, concentration n° = half of concentration n° , and so on successively to a concentration n° of . μg/ml, with concentration n° representing cells cultured in medium alone. Fig. 1a depicts the absorbances when PBLs were incubated with LACE extract, showing a significant increase in proliferation of PBLs at extract concentrations of – μg/ml. The proliferation induced by LACE was to -fold greater than in the control, reaching % of the level of proliferation induced by Concanavalin A and % of the proliferation induced by IL- (Fig. 1a). If the results of PBLs proliferation are expressed as Stimulation index (SI: absorbance of treated lymphocytes divided by absorbance control or unstimulated lymphocytes): SILACE = ., SIIL2 = ., SIconA = ..Interestingly, LACE induced proliferation without previous stimulation of PBLs. When these assays were performed using the non laser-treated calendula extract, CE, the absorbance results were practically identical, indicating that both extracts produce a similar increase in PBL proliferation (data not shown). The principal chemical components of the aqueous extracts of Calendula Officinalis are: polysaccharides, proteins, fatty acids, carotenoids, flavonoids, triterpenoids and saponins (data not shown).LACE extract inhibits in vitro tumor cell proliferationExponentially growing cell lines (.– × cells) were exposed to increasing concentrations of LACE or CE for – h in -well tissue plates as described above, and a dose/response curve was performed. Figure 1b depicts absorbance results when B9 murine fibrosarcoma cells were cultured with LACE extract. These results showed a significant decrease in absorbance, mainly in the – μg/ml concentration range, with an IC50 concentration of μg/ml. Similar results were found for HELA and Ando cell lines (data not shown). When these assays were performed with the non laser-treated CE extract, the decrease in absorbance was notably smaller (Fig, 1b). These results indicate that CE extract produced significantly less inhibition of tumor cell proliferation in comparison with LACE extract.Following these screening assays, LACE extract was used in further experiments at a concentration of μg/ml. Similarly results were obtained when other studied tumor cell lines were cultured with μg/ml of LACE (a decrease in absorbance was observed with respect to untreated cells) (Fig. ). Paclitaxel (Taxol) was used as control in these experiments, yielding similar results to those obtained with LACE (Fig. ). The NKL leukemia cell line was the exception, since treatment with LACE extract induced an increase in proliferation. It should be taken into account that NKL cells are dependent on IL- for their growth and LACE and IL induce similar levels of proliferation.The correlation between optical densities and number of viable cells was analyzed, culturing – × tumor cells in tissue-flask in medium alone (control) or with μg/ml of LACE for – days. As shown in Table , there was a significant decrease in the final number of viable cells, with a growth inhibition of –% versus control cells, except for the treated leukemia cell line U937, which showed a growth inhibition of %. Under an inverted phase contrast microscope, LACE-treated tumor cells showed morphological changes: rounded and granulated morphology, some vacuoles coming from cytoplasm, cell shrinkage and detached from culture plates of a high number of cells. These morphological changes were not observed in lymphocytes treated with LACE.Cell cycle phase distribution analysis of treated cellsCells were found accumulated in G2/M phase (.%) when PBLs were cultured in medium alone. However, when PBLs were cultured with addition of μg/ml LACE, cells re-entered cell cycle, appearing in S phase (Fig. ). In contrast, when a tumor cell line, e.g., AGS, was cultured in medium alone, cells were in cell cycle (.% G0/G1, .% S, and .% G2/M), and when μg/ml LACE extract was added, cells showed a significant accumulation in G0/G1 phase (.% of cells) at the expense of the S phase population (Fig. ). Similar results were found in Jurkat cells, where the percentage of LACE-treated cells in G0/G1 phase was .% (Fig. ). However, in other tumor cell lines, e.g., HELA cells, a slowing rather than an arrest of cell cycle was observed, with a partial accumulation of the cells in G0 /G1 phase (Fig. ). Fractions of cells in the studied tumor cell lines are listed in the Table .LACE inhibits expression levels of CDKs- and cyclins-associated to G1 phaseWe have found G1 cell cycle arrest in cells treated with LACE. The cyclins D1 and E are important regulator of progression through of G1 phase and entry into S phase. The AGS and JURKAT cells were treated with LACE during – hours. After this period, a decrease in the expression of these proteins was detected by immunofluorescence (Fig. 4a). The JURKAT cells cultured in alone medium present an expression of cyclin D1 and E of .% and .%, respectively; when these cells were cultured with LACE the cyclin D1 expression decrease to .% and the cyclin E expression to .% (Fig. 4a). Similar results were found in AGS tumor cell line, so the cyclin D1 expression decreases of .% to .% and the cyclin E expression decreases of .% to .% (Fig. 4a).The western blot analyses of total cell lysates showed a decrease of protein levels of CDK1/Cdc2, CDK2, CDK4, CDK6, cyclin A and cyclin D3, in LACE-treated cells (Fig. 4b). The cyclin B expression was no altered in response to LACE (Fig. 4b).LACE induces apoptosis in tumor cellsTreatment of cancer cells with LACE led to the generation of a sub-G1 population suggestive of apoptosis. To quantitatively examine the ability of LACE to induce apoptosis, cancer cells were treated with and μg/ml for days. Cancer cells that were untreated or treated for days were incubated with Annexin V-PE antibody in a buffer containing -amino-actinomycin D and analyzed by flow cytometry. Apoptosis of B9 cells increased from .% in cells cultured in medium alone to .% in those cultured with μg/ml of LACE, and to .% in those cultured with mg/ml of LACE (Fig. ). Similar results were found in the other tumor cell lines (Table ). In contrast, LACE did not produce apoptosis in PBLs (Fig. ).Expression of active human caspase-3Since caspases are the main enzymes involved in the apoptotic pathway, it was investigated whether active caspase- was involved in LACE-induced apoptosis. Thus, the cancer cell lines were treated with LACE for days, and cells were then permeabilized, fixed, and stained for active human caspase- and analyzed by flow cytometry. In the case of AGS cell line, the results clearly illustrated that untreated cells were primarily negative for presence of active-caspase-, whereas around % of LACE-treated cells showed detectable active caspase- (Fig. ). Expression of active caspase- was also detected in other tumor cell lines, whereas some lines, e.g., HELA, showed no caspase- expression after LACE treatment (Fig. ). These results indicate that apoptosis in HELA cells was not induced by caspase-. In PBLs, as expected, LACE treatment did not induce caspase- expression (Fig. ).PBL subpopulations activated by LACEWe also analyzed the PBL subpopulations that proliferated after LACE treatment. Figure depicts representative results of PBL subpopulations treated with μg/ml LACE for – h. In control PBLs, the subpopulations were .% CD4+, .% CD19+ and .% CD56+/CD3+. After LACE treatment, these populations significantly increased to .% CD4+, .% CD19+ and .% CD56+/CD3+ (Fig. ). Therefore, cells in proliferation were mainly B lymphocytes, CD4+ T lymphocytes, and NKT lymphocytes. Activated CD markers of lymphocytes (e.g., HLA-DR and CD69) were expressed by the lymphocyte subpopulations that proliferated with LACE treatment (Fig. ), indicating that these lymphocytes were activated.Toxicity of LACE extract in mice and ratsAfter the above in vitro activities of LACE were established, investigation of its in vivo toxicity was the next step. In a preliminary study, Balb/c, C57/BL6, and CBA mice were orally administered with or mg/kg body weight of LACE daily for consecutive days and no deaths were observed during this period. No differences in systemic or local toxicity were observed between LACE-treated groups and controls, although a dose increase to mg/kg body weight resulted in a % reduction in the -day survival of these mice (LD50) (data not shown). At autopsy, the animals exhibited hepatic toxicity. LACE was also orally administered to rats at a dose of , and mg/kg body weight daily for consecutive days, and no deaths or increases in systemic or local toxicity were observed. However, when the daily dose was mg/kg body weight, % of rats died by day (LD50) (data not shown). After days of treatment, surviving mice and rats were maintained under standard conditions and their mortality was recorded at days post treatment. During this period, no toxic side effects, e.g., debility, loss of body weight or death, were observed.In vivo anti-tumor effect of LACEWe have established in our laboratory a solid tumor model with nude mice bearing ANDO- human melanoma cell line. We next investigated the effect of LACE on this tumor model. The mice administered with LACE, oral ( mg/kg of weight) or intraperitoneal ( mg/kg of weight) route, showed similar tumor growth inhibition reached to about % (Fig. 8a). Similar results were found with Taxol (Fig. 8a). There was not differences between the different control groups. The untreated mice showed a fast progressive increase in tumor volume on the day while in groups treated with LACE o Taxol occurred later on day (Fig. 8a).On the other hand, the survival of mice in LACE-treated groups was monitored and compared to control groups. At the end of assays, the survival was: % in control groups, % in LACE-oral group, % in LACE-intraperitoneal group and % in taxol group (Fig. 8b).DiscussionThe main finding of the present study is the in vitro anticancer efficacy of a novel extract obtained from Calendula Officinalis against various cancer cell lines derived from human or murine solid tumors of different etiologies. In some cases, this extract, LACE, achieved % inhibition of growth (Fig. , Table ). Importantly, this inhibition effect is exerted on tumor cells from different solid tumors and is not specific to a single tumor cell tissue. The in vitro growth inhibition of LACE extract was similar to that reported for Taxol in tumor cell lines [], as shown in Figure .The principal chemical components of the aqueous extracts of Calendula Officinalis are: polysaccharides, proteins, fatty acids, carotenoids, flavonoids, triterpenoids and saponins. The carotenoids components may be excited by visible radiation. Laser treatment of the calendula extract is necessary to detect this biological activity. A similar calendula extract without laser treatment, CE, produced only a slight inhibition of tumor cell growth (Fig. 1b). The increase in this biological activity is due to treatment with laser radiation, which may induce conformational changes, excitation or degradation of different molecules of the CE extract.Studies performed to assess the mechanisms involved in the above in vitro effects showed that LACE induced cell cycle arrest in G0/G1 (Fig. and Table ). Furthermore, mechanistic investigation showed that LACE-induced G1 arrest mainly mediated via a down-regulation of cyclins D1, D3, E y A, and CDK1-Cdc2, CDK2, CDK4 and CDK6 (Fig. ). These novel data may have clinical relevance, since most human malignancies exhibit aberrations in cell cycle regulation []. In fact, most anticancer agents derived from plants exert their effect via apoptosis induction in cancer cells [-]. The present data demonstrated that LACE induces apoptotic death and that the rate of apoptosis increases with higher LACE concentrations (Fig. , Table ). The induction of apoptosis involved caspase--dependent mechanisms in some but not all tumor cell lines (Fig. ), suggesting differential molecular determinants of apoptosis induction in different tumor cell lines.In leukemia cell lines, LACE treatment inhibited % of growth in Jurkat cells and only % of growth in U937 cells, which may be explained by the different origin of the cells (lymphoma and monocytic, respectively) and/or the much higher growth rate of U937 cells, which double in number in less than h. Interestingly, LACE induced proliferation of NKL leukemia cells (Fig. ). The NKL cell line is dependent on IL- for its growth in vitro [], indicating that it is in arrest phase of the cell cycle. Treatment with IL- or LACE produces cell cycle re-entry and similar proliferation values in NKL cells (Fig. ). Likewise, a compound isolated from the fungus Coriolus Versicolor, Protein-bound polysaccharide K (PSK), also induces proliferation of NKL cells in absence of IL- []. PSK has shown anticancer activity in vitro in experimental models and in human clinical trials [,]. These anti-tumor activities can be largely attributed to activation of NK cells [,].The second novel finding of this study is that LACE treatment induces proliferation and activation of human PBLs (Fig. and Fig. ). The CE extract, without laser treatment, also produces a similar increase in PBL proliferation. PBLs, which do not proliferate in vitro and are found in G2/M arrest, re-entered cell cycle with LACE treatment (Fig. ).PBL subpopulations in LACE-induced proliferation were CD4+, CD19+, and mainly CD3+/CD16/+ (Fig. ). The latter correspond to NKT cells, which have been shown to recruit and promote a response by downstream effectors in an IFN-γ-dependent manner, activating both NK and CTL anti-tumor activity [,]. These data, considered alongside the cytotoxic activity of LACE extract in tumor cell lines, indicate that it might have anticancer properties in vivo. In fact, preliminary results from our laboratory indicate that LACE can inhibit the growth of mouse tumor cells in vivo (unpublished results). Results obtained with PBLs and NKL cells indicate that cells in cell cycle arrest re-enter cell cycle with LACE treatment. In contrast, LACE treatment produces phase G1 cell cycle arrest in tumor cells in cell cycle (Fig. ). Further research is required to determine whether a single active principle is responsible for this dual in vitro activity. Purification of LACE extract is in progress to identify its active principles. A further finding of this study was the low in vivo toxicity of this extract.The LD50 for orally administered LACE was mg/kg body weight in mice and mg/kg body weight in rats. Notably, the antiproliferative activity of LACE was not accompanied by systemic toxicity in mice or rats at a dose of mg/kg body weight. The LACE extract showed in vivo anti-tumor activity in nude mice, the growth of Ando- tumor was reduced in a % (Fig ). The anti-tumor efficiency of LACE was similar to obtained with other commonly used chemotherapy drug, paclitaxel. Furthermore, LACE produced higher prolongation of lifespan in tumor-bearing mice (Fig. ).The results of the present study are encouraging because LACE has shown significant inhibition of tumor growth in vitro and in vivo, therefore LACE or some components might be a promising chemotherapy candidate in treating cancers in clinic. Further experiments will focus on purification studies and the in vivo efficacy of LACE in other experimental mouse cancer models.ConclusionLACE extract have demonstrated in vitro growth inhibition of various tumor cell lines. This is due to induction of cell cycle arrest and apoptosis. In contrast, it induced proliferation and activation of PBL cells. In addition, LACE present anti-tumor activity in vivo in nude mice.AbbreviationsLACE, Laser Activated Calendula Extract; CE, Calendula extract; PBLs, Peripheral Blood Lymphocytes; LGL, Large Granular Lymphocyte; TMB, Tetramethyl-benzidine.Competing interestsThe authors declare that they have no competing interest. Materials for these studies were partially supported by a grant from Bomsund S.L. Malaga, Spain. The extraction process is patented by Bomsund S.L. The authors have no financial or non-financial relationship with the company Bomsund S.L. The authors have no interest, arrangement or affiliation with any product or organization that could be perceived as a real or apparent conflict of interest in the context of this manuscript.Authors' contributionsEJM and AGL performed the assays. IA and AC helped in some experiments. AGL and FG designed the study and drafted the manuscript. All authors have read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:
|
PMC2556669.txt
|
TITLE: Predictors of HIV infection and prevalence for syphilis infection among injection drug users in China: Community-based surveys along major drug trafficking routes AUTHORS: Yujiang Jia, Fan Lu, Gang Zeng, Xinhua Sun, Yan Xiao, Lin Lu, Wei Liu, Mingjian Ni, Shuquan Qu, Chunmei Li, Jianbo Liu, Pingsheng Wu, Sten H Vermund ABSTRACT: ObjectiveTo assess the predictors and prevalence of HIV infection among injection drug users in highly endemic regions along major drug trafficking routes in three Chinese provinces.MethodsWe enrolled participants using community outreach and peer referrals. uestionnaire-based interviews provided demographic, drug use, and sexual behavior information. HIV was tested via ELISA and syphilis by RPR.ResultsOf the participants, .% were HIV-infected, with persons living in Guangxi having significantly lower prevalence (.%) than those from Xinjiang and Yunnan (.% and .%, respectively). Syphilis seropositivity was noted in .%. Longer duration of IDU, greater awareness of HIV transmission routes, and living in Xinjiang or Yunnan were associated with HIV seropositivity on multivariable analysis. Independent risk factors differed between sites. In Guangxi, being male and having a longer duration of IDU were independent risk factors for HIV infection; in Xinjiang, older age and sharing needles and/or syringes were independent factors; in Yunnan, more frequent drug injection, greater awareness of HIV transmission routes, and higher income were independent predictors of HIV seropositivity.ConclusionPrevalence rates of HIV among IDUs in China are more than two out of three in some venues. Risk factors include longer duration of IDU and needle sharing. Also associated with HIV were factors that may indicate some success in education in higher risk persons, such as higher knowledge. A systemic community-level intervention with respect to evidenced-based, population-level interventions to stem the spread of HIV from IDU in China should include needle exchange, opiate agonist-based drug treatment, condom distribution along with promotion, and advocacy for community-based VCT with bridges to HIV preventive services and care. BODY: BackgroundInjection drug use (IDU) represents the largest single cause of HIV transmission in China, accounting for nearly half of the infections at the end of []. Ministry of Public Security data suggest that the number of registered drug users has risen steadily at a rate of about % per year, from , in to . million in [-]. The total number, including unregistered drug users, is thought to be much higher, perhaps . million []. China has the second largest estimate (midpoint: . million) of IDUs worldwide, following only Russia []. The first large outbreak of HIV in China was identified in among injection drug users (IDUs) in Dehong Prefecture of Yunnan Province on the Myanmar (Burma) border in southwest China []. The specific HIV subtypes first seen in Dehong spread along drug trafficking routes to IDUs in nearby cities in Yunnan [,]. Since then, serious epidemics among IDUs have been identified in Xinjiang (), Guangxi and Sichuan (), Guangdong (), Gansu (), and Jiangxi () []. The HIV epidemic routes coincided with the major drug trafficking roads from the "Golden Triangle" into China. Molecular epidemiology suggests that the major spread of the initial drug-related epidemic in China started in Yunnan and took two major routes: northbound to Sichuan, Guizhou, Gansu, Ningxia and Xinjiang, and eastbound to Guangxi, Guangdong and Guizhou [,,-]. Before , the HIV-infected cases in China were reported mainly from Yunnan [].Xinjiang and Sichuan first reported HIV infections among drug users in ; the HIV epidemic was first detected among drug users in Guangxi in . In subsequent years, HIV spread rapidly among IDUs in Yunnan, Xinjiang, and Guangxi and by the end of , all provinces, municipalities and autonomous regions in mainland China, as well as Hong Kong, Macao, and Taiwan, had reported cases of HIV/AIDS among drug users from to . Yunnan reported the highest number of annual HIV/AIDS cases in mainland China [].Yunnan's proximity to one of the world's largest illicit drug (especially heroin) production and distribution centers, the "Golden Triangle", contributes to drug trafficking and the availability of heroin [,,]. Only a small portion of heroin/opium is trafficked into Xinjiang from the "Golden Crescent" []. Currently, Yunnan, Xinjiang and Guangxi have remained the top three of the hardest-hit regions fueled by IDU in China [,,,,-]. However, no systematic community-based interventions have been undertaken in these regions. Only a small fraction of IDUs receive counseling and testing services and even fewer have participated in methadone maintenance treatment and needle exchange programs that were started in . Several studies have described the different HIV transmission risk factors among IDUs based in detoxification and detention centers in China [,]. However, there are few studies that used community-based recruitment of IDUs from multiple provinces []. A behavioral survey among drug users in Yunnan, Xinjiang, Hubei, and Beijing found that most of the drug users reported behaviors associated with high rates of HIV/AIDS acquisition, such as unsafe sexual practices and using drugs intravenously (.%) []. Of those who used drugs intravenously, .% reported sharing needles. The general knowledge about HIV/AIDS among this group was relatively poor. In order to understand the threat of HIV epidemic expansion and guide appropriate HIV prevention among IDUs in three highly endemic regions along drug trafficking routes in China, we conducted this community-based survey to assess the prevalence of HIV and syphilis and predictors for HIV infections.MethodsStudy sitesThis study was conducted in three sites along major drug (heroin) trafficking routes in Nanning City, Guangxi Zhuang Autonomous Region; Yili Prefecture, Xinjiang Uygar Autonomous Region; and Honghe Prefecture, Yunnan Province (Fig. ). We chose these three drug trafficking routes/provinces because HIV epidemics in these areas shared certain characteristics. All three regions were hardest hit by HIV, IDU has been the predominant route of transmission for HIV, and non-Han minority ethnic groups account for a large portion of the IDUs. Most of these IDUs live in relatively poor socioeconomic conditions. Guangxi, located along the major drug trafficking trade route bordering Yunnan on the west and Vietnam on the southwest, hosts million people. Nanning is Guangxi's capital city and has a population of almost million, % of whom belong to Zhuang ethnic and other non-Han minority ethnic groups. Xinjiang covers a very large area, with million people in far northwestern China, and has the longest boundary in China. From the northeast to the southwest, Xinjiang borders eight countries: Mongolia, Russia, Kazakhstan, Kirghizstan, Tajikistan, Afghanistan, Pakistan, and India. Yili Prefecture, located in the northwest of Xinjiang, hosts million people: .% Han, .% Kazak, .% Uygar, and .% belong to other minorities. Yunnan is located in southwestern China and borders Myanmar, Laos, and Vietnam. Ethnic minorities account for .% of Yunnan's population of million. Honghe Prefecture is located in the south of Yunnan Province. The population of Honghe is about . million and .% belong to Hani and Yi ethnic groups, while .% belong to other non-Han minorities.Figure 1Location of study sites. This study was conducted in three sites along major drug (heroin) trafficking routes in Yili Prefecture, Xinjiang Uygar Autonomous Region (A); Honghe Autonomous Prefecture, Yunnan Province (B), and Nanning City, Guangxi Zhuang Autonomous Region (C).Study design and study populationCommunity-based surveys were completed from November to January . The size of the IDU population was estimated in each community and geographic mapping was conducted for each site in the study's targeted communities. The participants were primarily enrolled by the trained staff using community outreach and peer referral "snowball" techniques. The peer referrals were limited to a maximum of five participants in order to enroll a relatively representative sample in the IDU community. Eligibility criteria required that participants be ≥ years old and have injected drugs at least one time in the last three months. Blood was collected for HIV and syphilis testing. All eligible participants were provided with risk reduction and coping counseling, both pre- and post-test. Written informed consent was received for all participants. Survey information was collected anonymously and remained confidential. The surveys also served as part of ongoing comprehensive IDU-focused surveillance activity, combining behavioral and biological information []. The study was approved by the Institutional Review Board (IRB) of the National Center for AIDS/STD Control and Prevention of the China Centers for Disease Control and Prevention, as well as the IRB of Vanderbilt University.MeasuresParticipants were recruited and completed all study procedures in either Chinese and/or the local language. All interviews were conducted by trained staff in both Chinese and the local languages to provide information including (Table and ): () demographic characteristics, e.g., sex, age, marital status, residency, ethnicity, years of education, monthly income, and study site; () drug use behaviors, e.g., duration of drug use, frequency of injecting drugs in the last week, ever shared needle and/or syringe during injection, the number of people shared needle and/or syringe with in the last injection, frequency of shared injection needle and/or syringe in the last six months, always carried a needle and syringe when out, and how many times a needle and syringe was used before trashing it; and () sexual behaviors, e.g., living with regular sex partners in the last year, ever had sex with regular sex partner in the last year, condom use with regulars sex partners in the last sex act, frequency of condom use with regular sex partners in the last year, regular sex partners ever used drugs, regular sex partners knew you used drugs, shared needle and/or syringe with regular sex partners, ever had sex with non-regular sex partners in the last year, the number of non-regular sex partners in the last year, condom use with non-regular sex partners in the last sex act, frequency of condom use with non-regular sex partners in the last year, ever paid money or provided drugs for sex in the last year, the number of sex partners paid or provided drugs for sex in the last year, condom use during paid or provided drugs for sex in the last sex act, frequency of condom use during paid or provided drugs for sex in the last year, ever provided sex for money or drugs for sex in the last year, the number of sex partners who had sex for money or drugs in the last year, condom use during sex for money or drugs in the last sex act, and frequency of condom use during sex for money or drugs in the last year. Knowledge about risk of HIV transmission routes was assessed by correctly answering five questions that were related to modes of HIV transmission (blood, sex, and mother to infant). The participants were further asked whether they had ever received voluntary HIV counseling and testing (VCT). All of the above questions in the questionnaire were selected by a panel of consultants of the national behavioral and biological sentinel surveillance in China [,].Table 1Demographic factors associated with HIV infection among injection drug users in three highly endemic regions of ChinaFactorsN*% (HIV+)†OR (% CI)PSex Female12234.().<. Male56055.().(.–.)Age < years234462().. ≥ years35350.().(.–.)Marital status Married19551.().. Single39549.().(.–.). Separated8063.().(.–.).05Residency Local65852.().. Other province1711.().(.–.)Ethnicity Han39048.().. Other28557.().(.–.)Years of education > years48752.().. ≤ years18652.().(.–.)Monthly income ≤ Yuan22050.().. > Yuan26058.().(.–.)District Nanning, Guangxi20716.().<. Yili, Xinjiang20566.().(.–.)<. Honghe, Yunnan27767.().(.–.)<.*: Total N may not equal , because of missing data; N: the number of participants being tested; †: %: the prevalence of HIV infection; HIV+: the number of HIV positive.Table 2Factors associated with HIV infection among injection drug users in three highly endemic regions of ChinaFactorsN*% (HIV+)†OR (% CI)PKnowledge of three major transmission routes for HIV Yes54853. ().. No12940. (). (.–.)Received voluntary counseling and testing No62851. ().. Yes3470. (). (.–.)Drug use behaviorsDuration of drug use (injection plus non-injection) < years11437. ().. ≥ years43253. (). (.–.)Duration of injection drug use < years18634. ().. ≥ years38156. (). (.–.)Frequency of drug injection in the last week < times47053. ().. ≥ times17651. (). (.–.)Ever shared needle and/or syringe during injection No32849. ().. Yes34154. (). (.–.)Ever shared needle and/or syringe in the last injection No58651. ().. Yes7860. (). (.–.)No. of people shared needle and/or syringe in the last injection = .().. >.().(.–.)Frequency of shared injection needle and/or syringe in the last months Never51652.().. Sometimes13751.().(.–.). Always683.().(.–.).2Always carried a needle and syringe with you when you were out Yes17153.().. No40957.().(.–.)How many times a needle and syringe was used before trashing it times25259.(). times19456.().(.–.). > times10548.().(.–.).08Sexual behaviorLiving with regular sex partners in the last year Yes28251. ().. No38052. ().(.–.)Ever had sex with regular sex partners in the last year Yes25751.().. No2352.(). (.–.)Condom use with regular partner in the last sex act Yes7652.().. No18651.().(.–.)Frequency of condom use with regular sex partner in the last year Always8145. ().. Sometimes14654. ().(.–.). Never3253. ().(.–.).5Regular sex partners ever used drugs No19949. ().. Yes9552. (). (.–.)Regular sex partner knew you used drugs No8148.().. Yes21451.().(.–.)Ever shared needle and/or syringe with a regular sex partner No4753.().. Yes4654.().(.–.)Ever had sex with non-regular partners in the last year No52853. ().. Yes13248. (). (.–.)No. of non-regular sex partners in the last year .().. >.().(.–.)Condom use with non-regular sex partner in the last sex act Yes4254.().. No9245.().(.–.)Frequency of condom use with non-regular sex partners in the last year Always7339. ().. Sometimes3658. ().(.–.). Never2458. ().(.–.).2Ever paid or provided drugs for sex in the last year No50756.(). Yes4143.().(.–.)No. of sex partners ever paid or provided drugs for sex in the last year .().. >.().(.–.)Condom use during paid or provided drugs for sex in the last sex act Yes1353.(). No2839.().(.–.)Frequency of condom use during paid or provided drugs for sex in the last year Always560.(). Sometimes1266.().(.–.) Never2330.().(.–.).3Ever provided sex for money or drugs in the last year No8332.(). Yes3438.().(.–.)No. of sex partners for money or drugs in the last year - >.()-Condom use during sex for money or drugs in the sex act Yes1330.(). No1643.().(.–.)Frequency of condom use during sex for money or drugs in the last year Always00- Sometimes560.()- Never714.()-Syphilis sero-status Negative61450.().. Positive3348.().(.–.)*: Total N may not equal , because of missing data; N: the number of participants being tested for HIV; †: %: the prevalence of HIV infection; HIV+: the number of HIV positive.All collected serospecimens were stored at the Prefecture-level CDC laboratories and transported to Provincial-level CDC for HIV testing. Two Enzyme-Linked ImmunoSorbent Assays (ELISA, Vironostika HIV Uni-form II plus O™, BioMérieux, Marcy L'Etoile, France; Beijing Wantai Biologic Medicine Co., China) were performed. Both samples testing positive were considered HIV-positive; both samples testing negative were considered HIV-negative. A repeat second ELISA was used as a tiebreaker for discordant results. Western blot confirmation of cases was possible in one province consistently, one province intermittently, but was not used in the third province. Syphilis serostatus was determined by screening for the antibody to Treponema pallidum antigen (p15, p17, and p47) and by a positive rapid-plasma reagin (RPR) test (Macro-Vue RPR™ Card Test, Becton-Dickinson, USA).Statistical analysisData were entered with EpiData. After corrections, data were then converted and analyzed using the Statistical Package for the Social Sciences (SPSS for Windows™; SPSS Inc., Chicago, Il, USA). The data were analyzed using unadjusted odds ratios with % confidence intervals for the odds ratio point estimates. Tests for significance of categorical data used a Chi-square test or Fisher's exact test. A multivariable logistic regression model was constructed with all variables in the univariate model whose p value was less than .. Thus, we report independent risk factors for HIV infection, controlling for confounding and interaction from other putative risk factors.ResultsSocio-demographic characteristicsWe included eligible participants (.%) for the analyses; persons were excluded because of refusing to participate or not meeting eligibility criteria. Of the participants, .% were males; .% were of the majority Han ethnicity; .% had < years of education; and .% were single, .% married, and .% separated (Table ). Their average age was . years old (S.D. ± .) and .% were under years old; .% were local residents; and .% had ≤ Yuan monthly incomes (Table ).HIV knowledge and VCTOf the participants, .% were aware of all three transmission routes (blood, sex, and mother-to-child); only .% of the participants had ever received VCT (Table ).Drug use and sexual behaviorsOf the participants, .% had used illicit drugs > years; .% injected drugs for ≥ years; and .% reported a history of sharing needles and/or syringes. To judge current users, we determined that .% had injected drugs more than twice in the prior week. Of the .% participants who reported using a shared needle and/or syringe in the last injection, three-quarters of them shared with more than one person. Of the participants, .% reported never carrying a needle and syringe when they were out. .% of the participants reported used a needle and syringe more than once before trashing it. One-fifth of participants reportedly had sex with non-regular partners in the last year. One-third of subjects reported always using condoms when having sex with their regular partner in the last year, while .% reported always using condoms when having sex with non-regular sex partners in the last year. Over the last year, .% had paid money or provided drugs for sex and only .% of them reported using condoms consistently. .% provided sex for money or drugs and none of them reported using condoms consistently (Table ).Prevalence of syphilis seropositivity and predictors for HIV seropositivityOf the participants, .% were RPR reactive for syphilis. .% were HIV-seropositive, with persons living in Guangxi having significantly lower prevalence (.%) than those from Xinjiang and Yunnan (.% and .%, respectively). In univariate analyses, risk factors associated with HIV sero-positive status included male sex, "separated" marital status, local residency, minority (i.e., non-Han) ethnicity, study site (Yili, Xinjiang and Honghe, Yunnan), awareness of HIV transmission routes, having received VCT, longer duration of drug use, and longer duration of IDU (Table and ). Sexually-related factors, age, years of education, and syphilis seropositivity were not associated significantly with HIV seropositive status.Multivariable logistic regression analyses suggested that a longer duration of IDU (Adjusted OR = .; %CI: .–.), greater awareness of HIV transmission routes (AOR = .; %CI: .–.), and living in Yili, Xijiang (AOR = .; %CI: .–.) and Honghe, Yunnan (AOR = .; %CI: .–.) versus in Nanning, Guangxi, were independent risk factors for HIV sero-positivity for the three sites (Table ). In Nanning, Guangxi, being male (AOR = .; %CI: .–.) and having a longer duration of IDU (AOR = .; %CI: .–.) were independent risk factors for HIV sero-positive. In Yili, Xinjiang, older age (AOR = .; %CI: .–.) and ever sharing of needles and/or syringes (AOR = .; % CI: .–.) were independent risk factors for HIV sero-positive. In Honghe, Yunnan, higher frequency of drug injection (AOR = .; %CI: .–.), greater awareness of HIV transmission routes (AOR = .; %CI: .–.), and higher income (AOR = .; % CI: .–.) were independent risk factors for HIV sero-positive status.Table 3Factors associated with HIV infection among injection drug users in three highly endemic regions of China, as predicted by a multivariable logistic regression modelFactorsAOR (% CI)PThree sites Duration of injection drug use: ≥ years versus < years3. (.–.)<. Awareness of HIV transmission: awareness vs. unawareness2. (.–.)<. Yili Prefecture, Xinjiang: vs. Nanning, Guangxi7. (.–.)<. Honghe Prefecture, Yunnan: vs. Nanning, Guangxi15. (.–.)<.001Site , Nanning, Guangxi Duration of injection drug use: ≥ years vs. < years4. (.–.). Sex: Male vs. female3. (.–.).04Site , Yili, Xinjiang Shared injection needle and/or syringe: Yes vs. No5. (.–.). Age: ≥ years old vs. < years old3. (.–.).03Site , Honghe, Yunnan Frequent drug injection: – time/week vs. ≥ times/week3. (.–.)<. Awareness of HIV transmission: awareness vs. unawareness2. (.–.). Monthly income: ≥ Yuan vs. < Yuan1. (.–.)<.05AOR = Adjusted Odds RatioDiscussionWe assessed the prevalence and predictors of HIV sero-positive among IDUs with serious illicit drug problems in China using community-based cross sectional surveys with consistent sampling procedures in all three provinces (or autonomous regions). HIV prevalence was very high (.%), but was lower in persons living in Guangxi (.%) compared to Xinjiang and Yunnan (.%). The HIV prevalence rates were remarkably similar to those from the same sites among IDUs from detoxification or detention centers [], and were significantly higher than estimates from community-based surveys in other regions in China [].Lower rates are reported in other provinces. For example, in January , HIV prevalence rates of % to .% were reported in six community-based surveys of , IDUs in Guangxi and Yunnan's adjacent provinces of Sichuan (.%), Guangdong (.%), and Guizhou (%), with even lower prevalence noted in sites in Fujian (.%), Henan (%), and Hubei (%), provinces located farther from Guangxi and Yunnan []. Higher HIV prevalence rates among IDUs in – surveys are seen in those regions of Guangxi, Xinjiang, and Yunnan where rapid spread of the virus among drug users occurred earliest; HIV was first reported in Yunnan in [,]. Overall prevalence was noted to be .% among IDUs from detoxification centers in Honghe and Wenshan Prefectures of Yunnan Province in , having declined subsequently. One may speculate that rates have dropped due to deaths and/or prevention successes []. Five out of prefectures in Yunnan have reported high HIV prevalence rates among IDUs, ranging from .% to .% [,,]. Biological sentinel surveillance data show that HIV prevalence rates have reached .% in certain sites of Xinjiang and .% in certain sites of Guangxi in []. The majority of the participants in sentinel surveillance were recruited from detoxification or detention centers and they are likely to be higher risk injectors than IDUs in community settings. These differences could also reflect the availability of proactive testing in the detoxification or detention centers rather than a proven difference between the sub-group and a wider population of IDUs.High HIV prevalence among IDUs, prevalent needle sharing and high frequency of injecting practices suggest an urgent need to improve drug addiction treatment and risk reduction measures in China. We found that .% of the participants had shared needles and/or syringes and .% had injected drugs more than twice in the last week. An HIV epidemic becomes self-perpetuating (endemic) and even a modest level of risk behavior can lead to a substantial rate of infection in the face of efficient needle/blood transmission [,]. Because they live along major drug trafficking routes, many of the HIV-infected IDUs in our survey will continue to serve as a major source for continued transmission and further spread unless drug abuse treatment, antiretroviral therapy, and risk reduction are implemented, as indicated[].While longer duration of IDU, shared injection needle and/or syringe, and higher frequency of injection were the independent risk factors for HIV infection [,,,], greater awareness of HIV was associated (unexpectedly) with higher HIV prevalence. This may suggest some successes in educating IDUs. Higher income was also a risk factor. We speculate that drug users with higher incomes may use drugs more often; they may also have a greater awareness of HIV issues. There was some diversity in associated risk factors among the IDU subgroups in the three regions where HIV prevalence was especially high. Although a high portion of participants know HIV transmission routes in all three sites, the needle sharing rates and unprotected sexual behaviors were still high among IDUs. Most astonishingly, a very small portion (overall .%) of participants reported ever receiving VCT, a gateway for the prevention programs. This indicated that a large proportion of IDUs who have been infected with HIV don't know their status and could continue to spread the virus [,]. China has scaled up HIV control efforts since []; however, low HIV testing rates (≈% nationwide) remain an impediment to prevention and care. Lack of affordable accessibility to sterile needles and syringes was the major reason for high risk sharing of "works" in this study. Other data suggested social norms that foster stigma, discrimination associated with drug use and HIV/AIDS, fear of arrest due to illegal practice, knowing a positive result, a lack of coping skills, and knowledge of HIV risks are the other reasons for the low rate of HIV testing among IDUs [,]. This suggested that risk reduction education alone cannot help drug users and their sex partners make lasting behavioral changes. The community-based needle exchange programs and elimination of any barriers to accessing clean needles and syringes could reduce the prevalence of needle sharing among IDUs[,]. In addition to providing accurate and up-to-date information on risky behaviors, effective community-based prevention programs not only make clean needles and condoms available and accessible, but also focus on enhancing individuals' motivation to change their behavioral patterns, teaching concrete strategies, and behavioral skills to reduce risk, providing tools for risk reduction, and reinforcing positive behavior change.We found that there were significant differences between sex, age, marital status, residency, ethnicity, education level, and monthly income among the participants in the three study sites. A larger portion of participants who were single and belong to the Han ethnic group, with > years of education and higher income, were recruited in Honghe, Yunnan than in the other two sites. Yili, Xijiang's participants were more likely to be younger, belong to non-Han ethnic groups (.% Wei ethnic group in Yili, Xijiang; .% Hani and Yi ethnic groups in Honghe, Yunnan and .% Zhuang ethnic group in Nanning, Guangxi), and receive lower levels of education. Nanning, Guangxi's participants were more likely to have less monthly income (.% with ≤ Yuan RMB monthly income). We found that higher income in Honghe, being male in Nanning, and old age in Yili were independently associated with HIV infection. There could be other factors beyond this study, besides gender, age and the sharing of needles, such as the actual availability of syringe distribution and exchange programs, condom distribution and promotion, and other social determinants of health that account for the differences for the HIV prevalence rates in the three study sites. China's central government has scaled up HIV/AIDS control efforts since [], including setting up national policy framework for responding to HIV/AIDS, increasing funding inputs, and expanding collaborations with international organizations. However, responses to drug use and the HIV/AIDS epidemic vary significantly at provincial and lower administrative levels. A literature review indicated that Yunnan and Guangxi provinces have done far more than other provinces in supporting, implementing, and advocating for harm reduction interventions for IDUs []. Some local governments are not fully motivated to confront drug abuse and HIV/AIDS problems [].Among IDUs in other studies from China, risky sexual behaviors have been reported as a risk factor for HIV infection [,,], although we did not find this association in our three populations. Most of our participants that lived in remote rural areas of Honghe, Yunnan and Yili, Xinjiang were less likely to receive health education and services. Furthermore, due to relatively poor economic status and lower levels of education, they may be more likely to be involved in drug smuggling and abuse, and unprotected sexual behavior. Risk reduction programs should give high priority to these poorer, more isolated IDUs who are also more likely to be of minority ethnic origin. Because of the high prevalence of HIV and often risky sexual behavior among IDUs, there is a great potential for IDUs serving as a bridge population to transmit HIV to the general population. The overlapping of risk behaviors among at-risk persons facilitates the rapid HIV spread from IDUs to other risk groups, e.g., from female sex workers and their clients to their clients' regular partners. We found that low condom use rates and the high proportion of female drug users who had reported engaging in commercial sex underscore the importance of behavioral surveillance in IDUs to provide early warnings and more effective interventions. This highlighted the need for condom distribution and promotion. As noted in this study, most of the target IDUs interviewed already knew the causes of HIV; the problem is not knowledge translation, it is more basic social determinants of health. They don't have access to free condoms. Free condoms should be provided widely to sex trade workers and IDUs.The prevalence of syphilis by RPR in our high risk IDUs was .% (/), similar to estimates in sentinel surveillance sites using RPR screening in , IDUs in the same three provinces (average: .%, range from . to .%) []. Syphilis seropositivity did not predict HIV, suggesting that most infections were due to injection-related behaviors. Other studies have reported an association between HIV infection and other STDs among IDUs [-]. Syphilis should be considered one indicator of high sexual risk behavior among IDUs []. Previous studies of syphilis among IDUs have suggested that while a high prevalence of syphilis and low HIV prevalence may be found in clinical or community settings, the reverse pattern of high HIV prevalence and low prevalence of syphilis may be more common in detoxification centers where IDUs, who are heavier drug users, are overrepresented [,]. The patterns of STD co-morbidity among IDUs vary significantly by venue and high risk group [,].Strengths of this study include its substantial sample size, the geographic diversity of our venues, and community-based outreach and peer referral using "snowball" and mapping strategies. There are also limitations. First, IDUs recruited into the study may have been higher risk such that their HIV prevalence may not exactly reflect the true background rate among IDUs in the study community. Second, recall bias and social desirability bias are possible, since the drug use and sexual behavioral information was collected based on self-reporting. Most information about drug use and sexual behaviors in the last year were used in the data collection, instead of collecting the behaviors in more recent period, in the last three or six months. Third, our cross-sectional study cannot ascertain a causal association between predictors and HIV infections. Fourth, we do not include a complete list of factors in this study. Other factors beyond this study may also account for the differences.China has initiated harm reduction projects, including needle exchange programs, methadone treatment, condom promotion, and VCT programs among drug users [,,,,,]. China Center for Disease Control and Prevention provincial authorities have been organizing the needle exchange and methadone treatments since early [,,]. China plans to scale up harm reduction projects, including needle exchange programs and methadone treatments, since only a small portion of IDUs have been covered by these programs so far. Our data suggest the urgent need for expanded community-level needle exchange programs, opiate agonist-based drug treatment, and advocacy for community-based VCT with bridges to HIV preventive services and care. Condom distribution along with condom promotion should also be highlighted. In vulnerable target populations where condom use is directly related to availability, condom distribution and promotion is crucial to helping curb the spread of HIV and other STDs. These prevention and treatment efforts are likely to require an infrastructure that not only provides operational and financial support, but also creates an environment in which IDUs feel comfortable and safe in seeking help without any barriers. Implementation research programs can critically assess these programs and provide insight as to where they might be improved.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsYJ participated in the development of the manuscript, coordinated the analysis, and drafted the manuscript. FL, ZG, and XS were responsible for securing funding, supervising data collection, and preparation of the manuscript. YX provided data analysis, and drafted and reviewed the manuscript. CL and PW served as the statisticians for the manuscript. LW, LL, MN, and SQ oversaw all recruitment efforts in the field, supervised HIV and syphilis tests, and were an active part of the preparation of the manuscript. SHV provided input with guidance on the data analysis and interpretation, and co-wrote the manuscript. All authors read and approved the final manuscript.
|
PMC3023765.txt
|
TITLE: Double Dissociation of Amygdala and Hippocampal Contributions to Trace and Delay Fear Conditioning AUTHORS: Jonathan D. Raybuck, K. Matthew Lattal ABSTRACT: A key finding in studies of the neurobiology of learning memory is that the amygdala is critically involved in Pavlovian fear conditioning. This is well established in delay-cued and contextual fear conditioning; however, surprisingly little is known of the role of the amygdala in trace conditioning. Trace fear conditioning, in which the CS and US are separated in time by a trace interval, requires the hippocampus and prefrontal cortex. It is possible that recruitment of cortical structures by trace conditioning alters the role of the amygdala compared to delay fear conditioning, where the CS and US overlap. To investigate this, we inactivated the amygdala of male C57BL/ mice with GABA A agonist muscimol prior to -pairing trace or delay fear conditioning. Amygdala inactivation produced deficits in contextual and delay conditioning, but had no effect on trace conditioning. As controls, we demonstrate that dorsal hippocampal inactivation produced deficits in trace and contextual, but not delay fear conditioning. Further, pre- and post-training amygdala inactivation disrupted the contextual but the not cued component of trace conditioning, as did muscimol infusion prior to - or -pairing trace conditioning. These findings demonstrate that insertion of a temporal gap between the CS and US can generate amygdala-independent fear conditioning. We discuss the implications of this surprising finding for current models of the neural circuitry involved in fear conditioning. BODY: IntroductionA key finding from studies of Pavlovian fear conditioning is that amygdala function is required to associate neutral conditioned stimuli (CSs) with aversive events (unconditioned stimuli; USs). This has been demonstrated in many fear conditioning situations where the neutral stimulus is a discrete stimulus or a physical context. Although there are different theories for the role of the amygdala in this conditioning process (e.g. []–[]), there is general agreement that the amygdala is involved in some aspect of fear learning. These models have revealed a great deal about the biochemical, genetic, and epigenetic mechanisms involved in the circuitry that underlies fear learning and memory (e.g. []–[]). However, within this circuitry, there are a number of caveats about the contribution of various structures to different components of this task.When a CS is paired with a US, typically the US occurs coincident with or immediately after CS offset. This delay-cued fear conditioning results in a strong CS-US association that depends on plasticity largely thought to occur in the amygdala []. Insertion of an interval (the trace interval) between the CS and US recruits the dorsal hippocampus (DH), prelimbic cortex (PrL), anterior cingulate cortex, and entorhinal and perirhinal cortices to this trace-cued fear conditioning []–[]. Surprisingly little is known about the contributions of the amygdala to trace fear conditioning []. Given the well-documented importance of the amygdala in fear learning, one should expect it to also be critical for trace fear conditioning []. However, because of the additional circuitry recruited by trace fear conditioning, it is possible that the amygdala is less critical for the acquisition or consolidation of the trace fear memory. Although the nature of the interaction between the hippocampus and amygdala in contextual fear conditioning is well studied, it is not yet understood how these regions interact to support trace conditioning. One possibility is that the hippocampus maintains a representation of the CS during the trace interval and interacts with cortical regions to assign salience and predictive value to that representation. However, it is not yet clear whether amygdala activity is a critical component of this task. It may be that unlike delay and contextual conditioning, activity in the hippocampus and cortex are sufficient to support acquisition of trace fear conditioning.To investigate the involvement of the dorsal hippocampus and amygdala in trace fear conditioning, we temporarily inactivated these regions via intra-cranial infusion of muscimol, a GABAA receptor agonist, prior to trace or delay fear conditioning. To further examine amygdala contributions to trace and contextual conditioning, muscimol was infused before or after conditioning and before different trace fear conditioning protocols.MethodsSubjectsThese studies used , – week old, male C57Bl6/J mice, individually housed in standard colony cages, maintained on a / hour light/dark cycle. All animals received ad lib food and water, and all studies were authorized by the Oregon Health & Science University Institutional Animal Care and Use Committee (Protocol ID B11039) and conducted in accordance with the ethical guidelines of the Society for Neuroscience and the National Institutes of Health.DrugsMuscimol (Sigma-Aldrich, St. Louis, MO) in . ul PBS, was infused at , ., or . ug/side at a rate of . µl/min through stainless steel cannulae into the DH ( g) or amygdala ( g). The infusion cannulae were attached with PE50 to Hamilton syringes (Hamilton, Reno, NV) controlled by a microinfusion pump. Injection cannulae were left in place for s after infusion.ApparatusTraining and testing for contextual conditioning was conducted in . cm circular plexiglas chambers described in []. An dB CS was administered through a sound generator (Coulbourn Instruments, Whitehall, PA), and a . mA footshock US was administered through the rod floor with a shock scrambler/generator (Coulbourn). To provide a distinct olfactory cue the apparatus was wiped down with .% acetic acid prior to conditioning or testing sessions. Training and context testing sessions were controlled by an IBM-PC running Graphic State software (Coulbourn). Testing for cued conditioning was conducted in rectangular conditioning chambers (Med-Associates, St. Albans, VT) with white Plexiglas floors, situated in sound attenuating cubicles. These chambers were located in a room different from the conditioning room. CS was generated with ANL- (Med) and administered through speakers mounted on the left wall of the chambers, controlled by an IBM-PC running MED-PC (Med). The altered context was cleaned with % ethanol. In both contexts ventilation fans provided 65dB background noise.Surgical ProceduresMice were anesthetized with isoflurane (%–%), mounted in a stereotaxic apparatus (David Kopf Instruments, Tujunga, CA), and the scalp was scrubbed and excised to expose the skull. Holes were drilled for anterior and lateral coordinates; guide cannulae were lowered into place with the stereotaxic, to match D/V coordinates; and permanently secured with dental cement []. DH coordinates were A/P -., M/L ., D/V -. mm and amygdala coordinates were A/P -., M/L ., D/V -. mm from bregma []. Stainless steel stylets were inserted into the cannulae to maintain patency during the day post-surgical recovery period.Trace fear conditioningSeparate experiments examined the effects of , , or trace conditioning trials. Each session began with activation of a house light, minutes after which a s CS was activated, followed s later by a s US presentation []. In the two- and four-trial experiments, each CS-US presentation was separated by a variable s inter-trial interval (ITI). After the final trial, mice remained in the chambers for an additional s, after which the houselight was inactivated and mice were returned to their home cages. The session lengths were ., ., and . min for the one-, two-, and four-trial experiments, respectively.Delay fear conditioningDelay fear conditioning was similar to trace fear conditioning, except that the US was presented during the final s of the CS. Mice received two CS-US pairings in a . min session.TestingTo test for contextual learning, hr following training, mice were placed in the training apparatus, the house light was activated and freezing was assessed for min. To test for cued learning, hr following training mice were placed in the cued-testing apparatus and assessed for generalized freezing for min, followed by two min-long CS presentations separated by a min ITI, followed with a min post-CS period. Freezing was assessed across the entire session, and for analysis the minute, Pre-CS, ITI, and Post-CS periods were combined and reported as “Altered” and the CS presentations were combined and reported as “CS”. Freezing was defined as the absence of all movement except respiration [], assessed for one second at ten second intervals by an observer unaware of treatment assignments and reported as percent freezing [].HistologyAll brains were post-fixed in % paraformaldehyde in PBS for at least hours, sliced on a cryostat, mounted and nissl stained with cresyl-violet. Infusion sites were confirmed by observing gliosis along the infusion cannula tracts. All dorsal hippocampal infusions were within target regions and all but three amygdala infusions were within or adjacent to the amygdala, Figure . Missed placements were excluded from analysis../journal.pone..g001Figure 1Schematic representation of DH and amygdala cannulae placements.Placements determined to be outside of the target regions are marked with crosses, placements within the target region are marked with solid circles.AnalysisAnalysis was conducted with RExcel .. []. Data were analyzed with one-way ANOVA followed with Tukey's post-hoc tests. All data are reported as mean ± the SEM, significant differences from vehicle were defined as p<..ResultsDorsal Hippocampal Inactivation Produces Deficits in Contextual But Not Delay Fear Conditioning (Figure 2A)Infusion into the DH prior to -pairing delay conditioning disrupted contextual fear conditioning [F(,) = ., p<.] but not Altered or CS freezing. Post-hoc analysis of contextual freezing demonstrated that vehicle-treated mice froze significantly less than sham-treated controls, and muscimol-treated mice froze less to context exposure than vehicle-treated mice or sham controls (Figure 2A). These results demonstrate that perturbation of the DH by vehicle infusion produces small deficits in contextual learning during delay fear conditioning, and that inactivation of this brain region with muscimol produces large deficits in contextual learning, confirming that the DH is critical to contextual but not delay fear conditioning../journal.pone..g002Figure 2Effects of DH or amygdala inactivation on contextual and cued components of delay and trace fear conditioning.Infusion of muscimol into the DH disrupts acquisition/consolidation of trace (B) and contextual (A & B) but not delay conditioning (B), while muscimol infusion into the amygdala disrupts delay (C) and contextual (C & D), but not trace conditioning (D). These results suggest that the amygdala may be differently involved in trace conditioning than in contextual and delay conditioning. Additionally, deficits in vehicle compared to sham in contextual (A, B, C & D) and CS freezing (B & C) show that disruption of DH or amygdala can interfere with conditioning. Subjects per group from top to bottom by panel, were: (A) , , ; (B) , , ; (C) , , ; (D) , , , . Data represented as Mean ± SEM, * denotes significant difference from vehicle control group.Dorsal Hippocampal Inactivation Produces Deficits in Contextual and Trace Fear Conditioning (Figure 2B)Infusion into the DH prior to -pairing trace conditioning affected contextual [F(,) = ., p<.] and CS [F(,) = ., p<.], but not Altered freezing (Figure 2B). Post-hoc analysis revealed that in both contextual and CS freezing muscimol infused mice froze significantly less than vehicle controls, and that vehicle infusion significantly impaired contextual learning compared to sham controls, while a trend toward a deficit was observed in CS freezing. These results demonstrate that vehicle infusion into the DH disrupts contextual learning, and that inactivation of the DH with muscimol disrupts both contextual and trace-cued fear learning compared to vehicle controls.Amygdala Inactivation Produces Deficits in Contextual and Delay Fear Conditioning (Figure 2C)Infusion into the amygdala prior to delay fear conditioning affected contextual [F(,) = ., p<.] and CS [F(,) = ., p<.], but not Altered freezing (Figure 2C). Post-hoc analysis showed that inactivation of the amygdala produced deficits in contextual and CS freezing compared to vehicle or sham controls, and that vehicle infusion produced deficits in CS freezing compared to sham controls, confirming that the amygdala is critically involved in acquisition of contextual and delay fear learning.Amygdala Inactivation has No Effect on Trace Fear Conditioning (Figure 2D)Infusion into the amygdala before -pairing trace fear conditioning affected contextual [F(,) = ., p<.], but not CS or Altered freezing (Figure 2D). Post hoc analysis revealed that muscimol (. or . ug/side) produced deficits in contextual freezing compared to vehicle controls, suggesting that amygdala activity is not required for trace fear conditioning.Amygdala Inactivation Effects in Trace Conditioning Are Independent of Training Protocol (Figure 3A, B)To determine if the lack of effect of amygdala inactivation on trace fear conditioning was particular to a -pairing trace fear conditioning protocol, mice received -pairing (Figure 3A) or -pairing (Figure 3B) trace fear conditioning. There were significant effects of infusion on contextual conditioning in both -pairing [F(,) = ., p<.] and -pairing [F(,) = ., p<.] but no effects on CS or Altered freezing in either task. Post-hoc analysis showed that in both -pairing and -pairing experiments muscimol infusion prior to training produced deficits in contextual freezing compared to vehicle or sham treated groups, and that vehicle infusion produced deficits compared to sham controls. These results suggest that the effects of amygdala inactivation on fear conditioning generalize across multiple protocols../journal.pone..g003Figure 3Parametric manipulations demonstrate that amygdala inactivation has similar effects on -pairing and -pairing trace fear conditioning, and that amygdala inactivation prior to or immediately after training of -pairing trace conditioning has similar effects.(A) Inactivation of the amygdala prior to -pairing or (B) -pairing trace fear conditioning produces deficits in the contextual component but does not affect the cued component or freezing to an altered context. Additionally, comparison of sham and vehicle groups demonstrates that infusion of vehicle into the amygdala produces deficits in contextual but not trace conditioning and does not affect freezing to an altered context. (C) Inactivation of the amygdala -minutes pre-training or immediately post-training produced deficits in the contextual but not the cued component of -pairing trace fear conditioning, and did not affect freezing to an altered context. Subjects per group from top to bottom by panel, were: A , , ; B , , ; C , , . Data represented as Mean ± SEM, * denotes significant difference from vehicle control group.Amygdala Inactivation Pre- or Post-Training Produces Deficits in Contextual But Not Trace Fear Conditioning (Figure 3C)To determine if the effects of muscimol infusion into the amygdala on contextual learning were due to interference with processes related to acquisition, such as locomotor activity during training, muscimol or vehicle was infused into the amygdala min pre- or immediately post-training (Figure 3C). In this study, all groups received muscimol or vehicle infusion both pre- and post-training, to equate handling conditioning between groups []. Thus the “vehicle” group received two vehicle infusions, the “pre-training” group received muscimol pre- and vehicle post-training, and the “post-training” group received vehicle pre- and muscimol post-training. Muscimol infusion pre- or post-training produced deficits in contextual learning [F(,) = ., p<.], but had no effect on CS or Altered freezing. Post-hoc tests showed that mice receiving pre- or post-training muscimol froze significantly less to the context than vehicle controls.DiscussionThe key finding from our results is that inactivation of the amygdala did not affect acquisition or consolidation of trace fear conditioning, even though delay and contextual fear conditioning were impaired. Further, inactivation of the DH impaired trace and contextual conditioning, but did not affect delay conditioning. Together, these findings suggest that insertion a temporal gap between the CS and the US allows amygdala-independent fear conditioning to occur. The dissociation between hippocampal and amygdala involvement in trace and delay fear conditioning in our experiments is evident in Figure , which summarizes these key findings../journal.pone..g004Figure 4Double dissociation of effects of muscimol infusion on -pairing trace and delay cued freezing.Inactivation of the DH with muscimol produces deficits in trace but not delay conditioned CS freezing, while inactivation of the amygdala produces deficits in delay but not trace conditioned CS freezing. These findings suggest that amygdala independent circuitry can support trace fear conditioning. Data are Muscimol (. ug/side) freezing to CS from experiments – as percent vehicle.The absence of an effect of amygdala inactivation on trace fear conditioning in the present findings is surprising, but it is unlikely to be explained by insensitive experimental manipulations. Animals that did not show trace fear conditioning deficits showed deficits in contextual fear conditioning and the same muscimol injections resulted in robust deficits in delay fear conditioning. Further, parametric control experiments demonstrate that the lack of effect of amygdala inactivation on trace fear conditioning is not due to under- or over-training as amygdala inactivation had no effect with one, two, or four CS-US pairings. Finally, the finding that muscimol injections into the DH disrupted trace fear conditioning demonstrates that the behavioral parameters in this paradigm were sensitive to disruption by neurobiological manipulations. Thus, the absence of an effect of amygdala inactivation on trace fear conditioning is strengthened by the presence of effects of amygdala inactivation on contextual and delayed fear conditioning, as well as by the effects of hippocampal inactivation on trace fear conditioning.Our amygdala infusion cannulae were targeted toward the basolateral nucleus (BLA) of the amygdala, which is critically involved in contextual fear conditioning []–[], whereas the lateral amygdala (LA) supports auditory processing in fear conditioning but not necessarily contextual learning []–[]. Thus, one possible explanation for our finding that amygdala inactivation disrupted contextual but not trace fear conditioning could be that our infusions inactivated the BLA but not the LA, and the LA supports trace conditioning. However, we demonstrate that inactivation of the amygdala with muscimol produces deficits in delay as well as contextual fear conditioning. Thus, our infusions are sufficient to affect delay fear conditioning, either through LA inactivation or because the BLA plays a critical role in delay conditioning. This makes it unlikely that our results can be explained by insufficient amygdala inactivation by muscimol infusion. Additionally estimates suggest that muscimol diffusion with this procedure is approximately mm3 []–[], suggesting that our infusions were sufficient to inactivate the entire amygdala. While such diffusion precludes conclusions about which amygdala nuclei were inactivated in these studies, the pattern of results clearly suggests that the amygdala plays a different role in trace than in delay and contextual conditioning.In these studies the GABAA receptor agonist muscimol was used to inactivate the DH and amygdala. As muscimol is selective to GABAA receptors it is possible that our results demonstrate distinct roles for GABAA receptor expressing neurons in the amygdala in trace and delay fear conditioning. Additionally, it should be noted that muscimol temporarily inhibits neural activity, and it is possible that plasticity supporting learning could occur independent of local neural activity. Whether these findings are because of distinct GABA receptor involvement in these forms of learning or distinct regional involvement in these tasks will require further study.Previous studies have shown that amygdala-independent contextual fear conditioning can occur with intensive training []–[], suggesting that it is possible under unique training conditions for amygdala-independent fear conditioning to occur. One view of amygdala function in fear conditioning is that rather than acting as the seat of plasticity, it plays a modulatory role increasing the strength of thalamo-cortical associations between CS and US []–[]. It may be that extensive training establishes CS-US associations through pathways independent of the amygdala, and that trace conditioning facilitates learning through these circuits. However, it is also possible that the amygdala is engaged by trace conditioning [], [], but amygdala activity is not necessary for associative learning. Future work will need to determine the precise role that the amygdala plays in the trace circuit.Alternate Pathways to Trace ConditioningBrain regions involved in trace fear conditioning are highly interconnected. Notably, the ventral hippocampus, thought to act as a conduit between the DH and amygdala [], has neurons that extend projections to both the amygdala and PrL []. Additionally, the ventral hippocampus mediates connectivity between the entorhinal/perirhinal cortices and the amygdala []. Thus, it could be that activity in the hippocampus and cortex can support acquisition of trace conditioning independent of the amygdala, and generate responses to the CS by hijacking amygdala output through ventral hippocampal projections to the amygdala. However, the PrL is also important form trace conditioning []–[], [] and its connectivity suggests that it could play a more direct role.Stimulation of the PrL increases freezing in rats [], suggesting that the PrL positively modulates fear responses. Additionally, neuronal activity in the PrL during the CS increases during trace fear conditioning [], thus it may be critically involved in generating fear responses to the trace-conditioned CS. However, the pathways through which these responses occur are not yet known. PrL connectivity suggests multiple routes through which it could directly initiate fear responses. PrL projections to the periaquaductal gray could bypass the amygdala entirely allowing the PrL to directly evoke fear responses []–[]. An additional route through which the PrL could generate fear responses is through projections to the intercalated-cell masses of the amygdala. These cells modulates inhibition of the CeA [], which is involved in fear response. Thus, PrL projections to the ITCm could allow the PrL to hijack amygdala output, controlling fear expression and subjugating the amygdala to an output structure rather than a center for CS-US association.Clearly there are multiple mechanisms that may support trace fear conditioning. Our findings demonstrate that inserting a temporal gap between the CS and US may allow fear memories to form independent of amygdala activity, although it is not yet clear whether plasticity in the amygdala supports this task. These findings add a level of complexity to current thinking about the circuitry underlying fear learning and add to a body of literature that shows that trace fear conditioning is supported by circuitry distinct from that of delay and contextual conditioning.
|
PMC2335120.txt
|
TITLE: Can subtle changes in gene expression be consistently detected with different microarray platforms? AUTHORS: Paola Pedotti, Peter AC 't Hoen, Erno Vreugdenhil, Geert J Schenk, Rolf HAM Vossen, Yavuz Ariyurek, Mattias de Hollander, Rowan Kuiper, Gertjan JB van Ommen, Johan T den Dunnen, Judith M Boer, Renée X de Menezes ABSTRACT: BackgroundThe comparability of gene expression data generated with different microarray platforms is still a matter of concern. Here we address the performance and the overlap in the detection of differentially expressed genes for five different microarray platforms in a challenging biological context where differences in gene expression are few and subtle.ResultsGene expression profiles in the hippocampus of five wild-type and five transgenic δC-doublecortin-like kinase mice were evaluated with five microarray platforms: Applied Biosystems, Affymetrix, Agilent, Illumina, LGTC home-spotted arrays. Using a fixed false discovery rate of % we detected surprising differences between the number of differentially expressed genes per platform. Four genes were selected by ABI, by Affymetrix, , by Agilent, by Illumina, and by LGTC. Two genes were found significantly differentially expressed by all platforms and the four genes identified by the ABI platform were found by at least three other platforms. Quantitative RT-PCR analysis confirmed out of of the genes detected by two or more platforms and out of of the genes detected by Agilent only. We observed improved correlations between platforms when ranking the genes based on the significance level than with a fixed statistical cut-off. We demonstrate significant overlap in the affected gene sets identified by the different platforms, although biological processes were represented by only partially overlapping sets of genes. Aberrances in GABA-ergic signalling in the transgenic mice were consistently found by all platforms.ConclusionThe different microarray platforms give partially complementary views on biological processes affected. Our data indicate that when analyzing samples with only subtle differences in gene expression the use of two different platforms might be more attractive than increasing the number of replicates. Commercial two-color platforms seem to have higher power for finding differentially expressed genes between groups with small differences in expression. BODY: BackgroundMicroarray technologies are now commonly used for genome-wide surveying of gene expression. With the availability of an increasing amount of data from different studies, there is a growing need for comparison and combination of datasets. This would be helpful to increase statistical power and to compare biological processes. Comparisons across different studies are, however, complicated by the use of different platforms. Over the past years, many microarray platforms, based on different technologies, have been developed by commercial and academic institutions. How reliable and consistent the results from different platforms are is still a matter of debate [-]. Initially, platforms comparison studies were mainly focused on comparison between commercial chips (mainly Affymetrix) and in-house spotted microarrays [-]. In recent years, more comprehensive studies were done, some of them reporting agreement between platforms [-] and some of them not [-]. The largest comparison was performed within an FDA-initiated program for evaluation of the reproducibility, quality and consistency of microarray platforms (MicroArray Quality Control, MAQC). In general, a high agreement between platforms was reported [-]. Our study is an extension to previously published studies in several aspects: we investigated the capabilities of five microarray platforms with high technological diversity to identify differences in gene expression in a challenging and highly controlled biological condition, where the expected level of transcriptional regulation was low, the number of differentially expressed genes small, and the number of biological replicates small, but realistic.The biological question addressed was the finding of differential gene expression in the hippocampus between transgenic mice overexpressing a splice-variant of the doublecortin-like kinase- gene, δC-doublecortin-like kinase (DCLK)-short, which makes the kinase constitutively active []. The DCLK gene has recently been implicated in crucial aspects of embryonic cortical development by controlling neurogenesis, neuronal migration and neuronal vesicle transport [-]. DCLK-short is not expressed during embryogenesis, is abundantly expressed in adult limbic brain structures, particularly in the hippocampus [], and has mild kinase activity in vitro [,]. The biological function of DCLK-short expression in the adult hippocampus is largely unknown and the transgenic mice have subtle phenotypes with no obvious differences in basal outcomes (Schenk et al, in preparation). Microarray-based expression profiling of the hippocampus tissues from δC-DCLK-short and controls should reveal the biological processes in which the gene is involved.The main aim of this paper is to compare the performance of different microarray platforms to detect differences in gene expression in biologically related samples. The performance of and the consistency between the microarray platforms on the level of affected genes and gene sets are reported here. The biological findings will be discussed in more detail elsewhere (Schenk, in preparation).ResultsExperimental set-upGene expression in the hippocampus of five wild-type mice and five transgenic mice was evaluated with five microarray platforms (Table ): Applied Biosystems (ABI), Affymetrix (AFF), Agilent (AGL), Illumina (ILL), and home-spotted oligonucleotide arrays (LGTC). Ten chips were used for each platform. For the two-color arrays, a wild-type sample was always co-hybridized with a transgenic sample and the design was balanced with respect to dye. Platform-specific processing of the signal was kept to a minimum as to not introduce processing artefacts. After careful performance evaluation, different normalization methods were chosen for one and two-color, but within the groups of one- and two-color platforms the method was kept constant as not to introduce differences due to the normalizaton algorithm. Differential gene expression was evaluated with an empirical Bayes linear regression model (EBLRM) from the R package limma []. Raw and normalized data are available from Gene Expression Omnibus (GEO) under series GSE8349.Table 1Description of the platforms under study.PlatformName#Probes#Unique UniGene IDs#Ensembl transcripts *one/two color# DEGs (% FDR)# DEGs UniGene dataset (N = ,)# DEGs Ensembl dataset (N = ,)ABIApplied Biosystem- AB170035,,,858one445AFFAffymetrix – Mouse Genome v2. Array45,,,553one13072112AGLAgilent- WMG G4122A41,,,845two3,,,003ILLIllumina- Sentrix Mouse- Expression BeadChip46,,,629one541924LGTCHome-spotted -mer oligonucleotide arrays (Sigma- Compugen collection),,,104two132235All Platforms10,,-***** More probes could correspond to the Ensembl transcript.** The number of differential genes obtained comes from the integrated analysis of data from all platforms within one statistical model which takes the platform into account (cf. Discussion section)Detected transcriptsThere was a large difference between the platforms in the number of probes which generated a signal above background. AGL had the highest number of present calls, LGTC the lowest. To make a fair comparison across platforms, we re-annotated all probe sequences and mapped them to the Ensembl transcript database. In addition to providing the most up-to-date annotation, alternatively spliced transcripts are considered separately so that possible inconsistencies between platforms due to measuring different splice variants would be excluded. The number of detectable Ensembl transcripts was high on AGL (,), intermediate on AFF, ILL, and ABI (around ,) and low on LGTC (,) (Table ). The low number of detectable transcripts on the LGTC platform is mainly due to background problems, causing negative control spots to occasionally give high signals. The overlap between detectable transcripts is highest between AFF and AGL (%) and lowest for all LGTC combinations.Table 2Overlap in detectable Ensembl transcripts across platforms. The number of detectable transcripts is presented on the diagonal (bold), with the total number of interrogated transcripts for each platform between parentheses. The overlap in the number of detectable transcripts for each pair of platforms is presented in the right side of the table, with the total number of interrogated transcripts shared between each pair of platforms between parentheses. The pair-wise overlap in detectable transcripts as a percentage of the overlapping set of interrogated Ensembl transcripts is presented in the left side of the table.Detectable transcriptsABIAFFAGLILLLGTCABI13331 () () () () ()AFF55.% () () ()()AGL56.%.% () () ()ILL44.%.%.% () ()LGTC10.%.%.%.% ()Differentially expressed genes identified on each platformThe number of significantly differentially expressed genes (DEGs) detected with a fixed False Discovery Rate (FDR) of % greatly varied across platforms (Table ): probes were selected by ABI, by AFF, , by AGL, by ILL, and by LGTC. As expected, the observed degree of differential gene expression was small. The absolute expression differences for the DEGs were in the following range: . – .-fold (ABI), . – .-fold (AFF), . – .-fold (AGL), . – .-fold (ILL), and . – .-fold (LGTC). The only two DEGs with a more than two-fold change in expression (as found with multiple microarray platforms and confirmed by qPCR) were: Plac9 (up) and Gabra2 (down).We further investigated the surprisingly high number of DEGs detected by AGL. When intensities instead of ratios were taken into the statistical analysis, no differential genes were detected at a FDR of % unless dye and array effect were included in the model. With the latter model (model in the Methods section), , genes were selected, among which all the , genes selected by the log ratios-based analysis. This and the more elaborate evaluation presented in Additional file suggest three major explanations for the good performance of the AGL platform: co-hybridization of samples from the two different biological groups to the same array, doubling of the number of observations with the same number of arrays used for the one-color systems, and low noise levels. These conclusions are in accordance with observations from earlier studies [,].The low number of DEGs on the ABI platform may be partly attributable to the use of different batches of arrays, but including the batch effect in the statistical model did not result in more DEGs.Analysis of overlapping DEGs across platformsTo be able to compare results across platforms, we created two data subsets with genes or transcripts interrogated by all platforms. For the first subset all GenBank accessions that were used by the array suppliers for their probe design were mapped to Unigene (UG) database, while averaging signal intensities from probes that mapped to the same UG entry. For , UG IDs data was available for all platforms. For the second subset, we mapped all probes to the Ensembl transcript database. There were , Ensembl transcripts that were interrogated by all platforms.Results for the subset of genes with overlapping UG identifiers are reported in Table and show the same trend already observed in the complete datasets. In Table the overlaps in DEGs selected by each pair of platforms are reported. Two genes were selected by all platforms (Plac9, 9230117N10Rik). The genes identified by ABI were selected on at least three other platforms. Overall, correspondence between platforms appears to be low. This is likely due to the use of a fixed statistical threshold. A higher correlation was found when evaluating the ranks of genes based on significance score. In Figure the ranks for each gene are plotted for each pair of platforms. A scattersmooth function [] is used for better visualization of the data cloud. As can be seen, in the area of the highly ranked genes (roughly from rank – rank ) there is a higher correlation between platforms than in the area of lower ranked genes. This is expected because only genes with significantly differential expression should be correlated while no correlation and complete scattering is expected for unchanged genes. We also considered the moderate t-statistics from the EBLRM which takes into account the direction of changes in the gene expression. The Pearson correlation coefficients (cP) of the t statistics within pair of platforms ranged between .–. (Table ). Correlations between pairs of platforms belonging to the same type (one- or two-color) where higher than between those of different types, with cP = . between AFF – ILL and between AGL – LGTC. Given the fact that the correlations are calculated based on all genes of which the biggest majority does not change in expression, higher correlations are not to be expected.Table 3Overlap in DEGs in the UG subset (normal face) and Pearson correlation coefficients (bold face) between t statistics in each pair of platforms.C(Pearson)\ # DEGsAFFABIILLAGLLGTCAFF72412254ABI0.264343ILL0..2219103AGL0...17159419LGTC0....4722Figure 1Scattersmooth plots of the correlation between the ranks (according to p values) of genes in the UG dataset of the platforms. Red corresponds to denser areas, while yellow corresponds to non dense areas. The scattersmooth uses an algorithm for smoothing of two dimensional histograms with smoothed densities (). This graph is more meaningful than a traditional scatter plot of the p values or of the -log p values, where the smallnumber of DEGs in our datasets originates graph blurred with thousands of overlapping dots and empty areas. Since the different signal to noise ratio is varying in the platforms and affects the statistics differently, plots of the ranks are more meaningful than plot of p values and statistics.The results of the analysis of the Ensembl transcript-mapped overlapping probes were highly similar in terms of overlap (Table ), and correlations of ranks and t-statistics (data not shown).ValidationQuantitative reverse transcription PCR (qRT-PCR) was used to validate the results of the different microarray platforms [see Additional file ]. As expected the two genes found as DEGs by all five microarray platforms were confirmed to display differential expression. The fold-changes found by qRT-PCR were slightly higher than those found by any of the microarray platforms, confirming previous observations that ratios tend to be compressed in microarray experiments [,,]. For out of tested genes that were significant (FDR<.) on at least two platforms, qRT-PCR experiments confirmed differential expression (Student's t-test: p < .). Lgals1, that was found by AFF and ILL only, did not reach significance in the qRT-PCR experiment due to large variability in the wild-type group. We selected genes (ranked from to ) that were found by AGL only covering the range from highly to lowly expressed genes, to ascertain whether the high number of genes selected by AGL was due to false positives. Eight out of these genes were confirmed by qRT-PCR (p < .), including Spp1 and Camkk1. These two genes were ranked among the top- genes on all platforms, except for Camkk1 on ABI. Pip5k2a, Ttc3, and Acsl1 were confirmed by qRT-PCR, but had an average ranking on the other platforms, and thus are truly found by AGL only. Of the genes that were found by AGL only but could not be confirmed by qRT-PCR, Gnb1l and Sgip1 were border-line significant in the qRT-PCR experiment (p = .). Interestingly, Taf12, although significant on AGL only, displayed very consistent fold-changes on the five microarray platforms (-. to -.). Probably its fold-change was so low that it was hard to confirm by qRT-PCR.Gene set analysisAnalysis at the level of gene sets (as annotated in the Gene Ontology -GO- [] and Kyoto Encyclopedia of Genes and Genomes -KEGG- [] libraries) may reveal greater similarities between platforms than analysis at the level of individual genes, since different but functionally-related genes could give hints to aberrations in the same biological processes []. The Global Test was used to evaluate the differential regulation of gene sets []. This method is based on a model for predicting a response variable from the gene expression measurements of a set of genes. Unlike commonly overrepresentation test or Gene Set Enrichtment Analysis, it has optimal power in small sample size experiments and is able to identify gene sets where many genes display a small but consistent effect []. Furthermore, the test enables the control for array and dye effects, and produces easily interpretable p-values that can be compared across experiments.We ranked the gene sets based on their Global Test significance and compared each pair of platforms (Figure ). Like for the analysis of individual genes, the highly ranked gene sets showed good agreement across platforms. Again, the best correlations were observed between pairs of platforms of the same type: AFF-ILL (both one-color) and LGTC-AGL (both two-color) with Spearman correlation coefficients of . and . respectively. In agreement with the lower number of DEGs found by ABI, the results from ABI did not correlate well with those of the other platforms. Similar results were observed using the gene sets from KEGG (data not shown).Figure 2Scattersmooth plots between the ranks of the GO gene sets (according to Global Test p values).The list of gene sets that were consistently identified by at least three platforms is dominated by genes involved in GABAergic signaling (Table ). Gabra2, found down-regulated on all platforms and confirmed by qRT-PCR [see Additional file ], is the most influential gene in these gene sets. Different genes on different platforms contribute to the significance of these gene sets as a whole: e.g. Chrna4 (AFF, AGL, LGTC), Chrna3 (AGL), Glra3 (LGTC), Glra4 (ILL) for gene set GO:. In general, this was due to near-background signals of these genes on most platforms.Table 4Gene sets highly ranked across platforms. For each platform, gene sets were ranked by their association with the phenotype under study, using the p-value from the global test. Displayed are those gene sets that rank highly in the majority of platforms; for GO sets, the sum of the highest three (out of five) ranks had to be below ; for KEGG sets, the sum of the highest four (out of five) ranks had to be below . Columns and display GO/KEGG IDs and names of the gene sets (in parentheses: GO classification: BP = biological process; MF = molecular function). Columns to display the ranks of the gene sets for each of the platforms.GENE SET IDNameABIAFFAGLILLLGTCGOGO:0007214gamma-aminobutyric acid signaling pathway (BP)23246232633GO:0004890GABA-A receptor activity (MF)24842111348GO:0016917GABA receptor activity (MF)28340141252GO:0030594neurotransmitter receptor activity (MF)4332322166GO:0042165neurotransmitter binding (MF)4742211174GO:0006821chloride transport (BP)8531732221GO:0015698inorganic anion transport (BP)10227213376GO:0005230extracellular ligand-gated ion channel activity (MF)167229101791GO:0006820anion transport (BP)18018333498GO:0015276ligand-gated ion channel activity (MF)190020411175GO:0050900leukocyte migration (BP)1389812691726KEGG4080Neuroactive ligand-receptor interaction25132104512ECM-receptor interaction1892116332010ABC transporters – General273154534660T cell receptor signaling pathway231657831760Nicotinate and nicotinamide metabolism5884161203030DNA polymerase91112241384514Cell adhesion molecules (CAMs)39132317115900Terpenoid biosynthesis13125419624640Hematopoietic cell lineage64422976DiscussionThe aims of the present study were to compare the ability of different microarray platforms to detect differences in gene expression, when levels of regulation and numbers of regulated genes are low, and to investigate the influence of the platform in the biological interpretation of the results.We show that even when gene expression differences between groups are small, several microarray platforms are able to consistently detect them. This is an important point, since in most previously published microarray platform comparisons, including the toxicogenomics MAQC study where biological replicates were analyzed, differences between samples analyzed where much larger than in our study [,,-]. The MAQC papers conclude that the cross-platform correlation is higher for fold-changes than for t-statistics. This is not true for our study. This apparent contradiction is because high fold-changes, which we simply do not have in our study, are more likely to be measured consistently, and contribute most to the Pearson correlation coefficient. Cross platform consistency in our study may compare favorably to another platform comparison study within a biological setting: Tan et al. reported a low agreement between platforms (Affymetrix, Agilent, Amersham) in the analysis of the effect of serum withdrawal []. In their case, the amount of interrogated genes shared by all platforms was low. In our study, the number of common probes is bigger (N~ ,) and allows for more reliable comparisons since a bigger and possibly more representative set of probes is taken into consideration.In contrast to other papers, we did not apply any filter to our data. In the reanalysis of the Tan dataset by Shi and collaborators [] the authors claimed that the use of the unfiltered dataset gave a poor agreement between platforms, while restricting the analyses to a small filtered subset gives highly reproducible results. Even if several filters are commonly used, strict investigation on the possible bias introduced in the data because of the exclusion of genes has not been done. Since filters of the data may affect individual datasets differently, we have avoided using them in order to reflect the true unbiased gene expression signatures. The drawback is that the correlation measures are more affected by biological and technical noise.The choice of the type of cut-off is still a matter of debate, and several authors suggested using a mixed cut-off of p-values and Fold Changes (FCs) [,]. However, even if a FC cut-off makes DEGs determination easier and from the technical point of view is more direct, it can eliminate the possibility of finding small differences in the data that are biologically interesting, as demonstrated in the current study (where only two genes showed a FC > ). Furthermore, the FC statistics do not have the probabilistic characteristics guaranteed by theoretical conditions that allow to be sure about what the method does [,].The degree of overlap between DEGs can be influenced by the overlap in interrogated and detectable transcripts as well as the method for matching of the probes. The overlap in interrogated transcripts was >%, as expected for these whole genome microarray platforms. The overlap in probes with signal above background was also in the same range. However, by adding the two effects, one can explain as much as % of the difference between two platforms and this can be even more for home-spotted arrays were the numbers of detectable transcripts are often reduced due to local background problems. The overlap may be further reduced due to the interrogation of different splice variants that are mapped to the same UG identifier. The Ensembl transcript mapping accounts for alternatively spliced transcripts. However, the correlation between platforms in the Ensembl transcript-mapped dataset was, in our case, not higher than in the UG dataset. This could be due to complications in the mapping process: AFF probe sets sometimes cover more than one transcript, and for ABI oligonucleotide sequences were not provided but only bp regions in which the probes were designed. Furthermore, there is considerable redundancy in the Ensembl transcript dataset due to multiple splice variants from the same gene being detected by all platforms, which may introduce biases in the downstream analyses. In this respect, the use of the recently released whole genome exon arrays for gene expression probably provides an attractive alternative, coping with such a problem.AGL selected a ten-fold higher number of DEGs and significant gene sets than all other platforms. This is partly attributable to the high signal to noise level of this platform, as evident from the number of probes with signal higher than background. Still, this huge difference was unexpected and we investigated the behavior of the AGL data in more detail, and compared this with AFF and LGTC data using different approaches [see Additional file ]. Briefly, the AGL log ratios show a bigger variability than AFF log intensities, measured by the a posteriori standard deviations. This difference remains after multiplying the variance of AFF intensities by the square root in order to calculate the variance in the ratio between two samples. To check whether the doubled number of observations on the AGL were the cause for finding many more differentially expressed genes, we left AGL arrays out one by one and repeated the EBLRM analysis. The number of DEGs decreased steadily from , ( arrays, samples) to ( arrays, samples). This is on the same order of magnitude as the number of DEGs of AFF ( arrays, samples, DEGs), but still five times larger.This suggests that the direct comparison of the wild-type and transgenic mouse samples on the same array drives the better performance, which is accordance with previous observations [,]. It argues against using either a common reference design or one-color protocols when comparing two groups of samples []. However, this does not explain the differences in performance between AGL and LGTC arrays. We found that AGL's technical replicates were much more reproducible than those of LGTC: Pearson correlation coefficients were .–. for AGL and .–. for LGTC, illustrating the differences in quality between commercial and home-spotted arrays. Overall, our study suggests that the differences in amount of DEGs found by the different platforms were mainly caused by differences in signal to noise ratios, and the numbers of observations between one and two-color platforms, when using the same number of arrays. Our qRT-PCR experiments validated differential gene expression in most cases, also for genes found by AGL only, indicating that these are not just false positives.Our results illustrates once more that typical sample sizes used in microarray experiments, three samples per group, can be too small to enable reliable detection of subtle effects such as in this study. Even though using samples per group still does not yield enough power for some platforms, it is possible to use our data as basis for estimation of sample size for the platforms considered. We are undergoing this work and the detailed analysis, beyond the scope of this paper, shall appear elsewhere. Our preliminary results confirm that AGL and AFF have comparable power, so the different outcomes observed by us are for the largest part due to the larger effective sample size involved in two-colour platforms design.We investigated whether the power of the analysis could be enhanced by merging data from all five platforms in one statistical model. We applied an EBLRM on the UG subset and included samples, platforms and dye (only for the two-color arrays) as confounders. At an FDR of ., genes were selected (Table ). Among these, most had been selected as DEGs by the individual platforms with the exception of genes. However, we could not validate the differential expression of the top of those genes by qRT-PCR, mainly due to large biological variation within groups. These genes seem to have been selected in the merged analysis due to the technical consistency on the microarray platforms allied to the larger pooled sample size.This study also aims to elucidate the biological function of delta-DCLK-short expression in the hippocampus. Recent loss and gain of function studies strongly suggest involvement of the DCLK gene in neurogenesis, neuronal migration, vesicle transport, microtubule-directed retrograde transport, neurotransmission and apoptosis [-,-]. Thus, DEGs identified in this study may be involved in these processes. The present study focuses on comparison of different array platforms and therefore the results of the biological function will be discussed more extensively elsewhere (Schenk in preparation). However, it is interesting to note that the DEGs and the significant gene sets revealed by the different microarrays are biologically meaningful. For example, numerous gene sets related to GABA-ergic neurotransmission emerged as highly significant in out of platforms. Intriguingly, similarly as the DCLK gene, excitatory GABA signalling has been shown to control neurogenesis, neuronal migration and differentiation of neuroblasts [,]. DCLK-short expression starts postnatally around day , a timepoint that is characterized by a switch in excitatory GABAergic responses to inhibitory responses [,]. The added value of the use of different microarray platforms lies in the prioritization of the pathways for follow-up experiments. When analyzing data from a single platform, many spurious gene sets apparently not related to the biological process under study (e.g. chemotaxis) ranked highly, probably due to the relatively small expression differences observed. By comparing platforms, a biologically meaningful consensus could be distilled.ConclusionThe present study suggests that the choice of a platform can be mainly governed by practical and cost considerations. However, our data demonstrate that, given the much higher number of identified DEGs, commercial two-color platforms may be preferred when two groups with small differences in expression are to be compared. In these situations, a direct-comparison design helps to maximize signal-to-noise ratios in the ratios between the two groups through minimization of the array effect and the possibility for more replicates with the same number of arrays. Since we performed this study with a clear underlying biological question, we could demonstrate that there was agreement across platforms in the perturbed biological processes identified. Consistency between platforms helped to prioritize biological processes relevant for the biological question under study. The relevant gene sets were detected with an only partly overlapping set of genes. Our data indicate that when analyzing samples with only subtle differences in gene expression the use of two different platforms might be more rewarding than increasing the number of replicates on the same platform.MethodsMICE5 Wild-type male C57/BL6j and transgenic male mice over-expressing DCLK-short with a C57/BL6j background were individually housed days prior to the start of the experiment. Animals were housed under standard conditions, 12h/12h light/dark cycle and had access to food and water ad libitum. Wild-type (N = ) and transgenic (N = ) tissue samples were collected by taking the brain from the skull and quickly dissecting out both hippocampi. Dissection was performed at °C to prevent degradation of RNA. Hippocampi were put directly in pre-chilled tubes containing Trizol reagent (Invitrogen Life Technologies, Carlsbad, CA, USA). All animal treatments were approved by the Leiden University Animal Care and Use Committee (UDEC# ).RNA extractionAfter transfer to ice-cold Trizol, hippocampi were homogenized using a tissue homogenizer (Salm&Kipp, Breukelen, The Netherlands) and total RNA was isolated according to the manufacturer's protocol. After precipitation, RNA was purified with Qiagen's RNeasy kit with on-column DNase digestion. The quality of the RNA was assessed with the RNA Labchip kit in combination with the Agilent Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA), using the Eukaryote Total RNA Nano assay according to the manufacturer's instructions. Total RNA was amplified using Ambion's MessageAmp kit, with incorporation of modified nucleotides (biotin--UTP (AFF, ILL), aminoallyl-UTP (AGL, LGTC), DIG-UTP (ABI)). For AGL and LGTC, aminoallyl-cRNA was coupled to Cy3 or Cy5 monoreactive dyes (GE Healthcare).Experimental designLabelled cRNAs of individual wild-type and transgenic mice were hybridized on different microarray platforms (Table ): Applied Biosystem (ABI), Affymetrix (AFF), Agilent (AGL), Illumina (ILL), and home-spotted glass microarrays containing the 22K mouse Sigma-Compugen collection generated at the Leiden Genome Technology Center (LGTC). Ten microarrays were used for each platform. For the one-color platforms (ABI, AFF, ILL), each individual RNA was hybridized to one microarray. A direct design was used for hybridization of the two-color arrays (AGL, LGTC), i.e. each microarray was hybridized with two RNA samples from different groups. All samples were hybridized once in Cy3 and once in Cy5. Dye-swapped hybridizations were done with non-identical sample pairs [see Additional file ].Quantitative RT-PCRQuantitative RT-PCRs were done on the Lightcycler480 (Roche), using the universal probe library (UPL, Roche) or SYBR-Green (when amplification efficiencies with UPL were below %). The RNA samples used for validation were the same as in the microarray experiments. Each cDNA was analyzed in quadruplicate, after which the average threshold cycle (Ct) was calculated per sample. Differential expression was evaluated with a Student's t-test, considering the biological replicates in each group.MappingTwo approaches were used to obtain an overlapping gene set that was measured on each platform. The first is based on the annotation provided by the manufacturer, while the second is an in-house performed probe sequence-based annotation.. GenBank accession numbers that were used for the design of the microarray probes were used for querying the Mus musculus Unigene (UG) database build #. All UG IDs that occurred at least once on each platform were included in the UG set (N = ,). With this UG set, a UG dataset was created for each platform by extracting the expression values for the relevant probes. When multiple probes were present for the same UG ID, the average of the signal of the probes was used as expression value.. For AGL, LGTC, and ILL, probe sequences provided by the manufacturer were directly used for annotation. For AFF, the probe sequences in a probe set were concatenated, after removal of potential overlap. ABI did not reveal the exact probe sequences but a bp region, in which the probes were located. Gmap [] was used for alignment of the sequences to the Ensembl mouse genome sequence (build NCBIM34). Hits with a match score higher than . (matches – gaps/query size []) were considered genuine matches. Chromosomal start and end positions of the hits were compared to the exon positions in the Ensembl database (version .34e). Subsequently, the Ensembl Transcript database was queried with only the exons that matched (part of) the probe sequence. Only transcripts with a match score >. on all platforms were included the Ensembl transcript set (N = ,). When multiple probes were present for the same transcript, the average of the signal of the probes was used as expression value.Based on each of the above overlapping gene sets, a dataset was created for each platform, which was analyzed separately [for UG: see Additional file ]. For completeness, the complete datasets (including also the non overlapping probes) were also analyzed in parallel.Preprocessing proceduresThe quality of the arrays was assessed by visual inspection of the raw images and pairwise MA-plots. No arrays were excluded from the analysis since the variance on the log-ratios was comparable between arrays. For the ABI platform, we observed differences in the signal distribution between two batches of arrays hybridized on two different days, for the other platform no quality problems were observed. Each dataset was loaded into the R environment directly as a raw data matrix (for ABI and ILL) or using the limma package (AFF, AGL, and LGTC). No background correction was applied to the two-color microarray platforms since the background correction increased noise levels in the low intensity range considerably. For AFF analysis, only perfect match probes were taken into account and probesets were summarized with the "median polish" method. The data from the one-color platforms were normalized with variance stabilization and normalization function implemented in the vsn package []. From all the normalization methods tested, vsn was most robust, whereas the performance of alternative normalization algorithms was more platform-dependent. Two-color arrays were normalized with loess [] since vsn normalization did not correct all the intensity-dependent non-linear behaviour in the data. Raw and normalized data are available in GEO under series GSE8349.Present callsFor Affymetrix chips, probes were said to be present when the MAS5. present call algorithm called the probe "P" (present) on all arrays. For the other platforms, probes were said to be present when their signal intensity was above the signal from the lowest % of platform-specific negative control probes on all arrays. For the two colour platforms, this requirement was imposed on the intensities of both the green and the red channel. Lists of present probes for each platform were then mapped to the ENSEMBL transcript database to generate a list of unique ENSEMBL transcript IDs with detectable expression.Statistical analysesDetermination of Differentially Expressed Genes (DEGs)Each dataset was analyzed for determination of DEGs using an Empirical Bayes Linear Regression Model (EBLRM). The following models were used for this purpose:) one-color datasetsyi = αi + βi group + εi2) two-color datasets- log ratioswi = αi + εi3) two-color datasets-intensitiesyi = αi + βi group + γi dye + δi array + εiwhere i is the ith item of the datasets, yi is the intensity signal, wi is the log ratios of the signal in Cy3 dye vs the Cy5 dye; αi, βi, γi, δi, εi were the coefficients of the intercept, group (transgenic vs. wild type), dye (Cy5 vs. Cy3 – only for two-color arrays), array (only for two-color arrays), and error terms, respectively. All the effects were considered to be random. DEGs were defined as the probes for which the βi were significantly different from , since βi is the estimate for the group (wild-type or transgenic) effect. Analysis were performed with the limma package [], using the lmFit function. P values were adjusted for multiple testing using the False Discovery Rates (FDR) method suggested by Benjamini and Hochberg []. FDR not greater than % was considered as statistically significant. Numbers and percentages of overlapping items in the list of DEGs among the platforms were calculated.Genes of UG and Ensembl were ordered by their p values obtained from the EBLRM and their Spearman correlation coefficients (cS) were calculated for pairs of platforms. Pearson correlation coefficients (cP) were calculated to quantify the correlation between the statistics produced by the EBLRM in the overlapping datasets.Gene set analysisThe association between the multiple functionally-related genes belonging to the same gene sets (according to the GO [] and KEGG [] libraries) and the group was assessed using the Global Test []. A logistic model with a gamma p-value estimating method was used for all platforms. For the two-color arrays, intensities were extracted and a model including array and dye effects as confounders was used. Gene sets were ordered by their p values obtained from the global test and Spearman correlation coefficients were calculated for pairs of platforms. Multiple testing was corrected via the FDR method []. FDR not greater than % was considered as statistically significant.Two-color platforms data analysisAnalyses of the two-color platforms data were done using log ratios per array, whenever possible. However, for the gene set analysis and for the analysis of the merged datasets separate channel intensities were needed. These were then extracted from the raw data, normalized using vsn and, to account for technical variability, the analysis model also included array and dye as confounders.SoftwareAll the analyses were performed using R software environment [] version .. and BioConductor [] packages vsn [], loess [], multtest [], Affy [], globaltest [], limma [], AnnBuilder [] and the function scattersmooth [].When metadata packages were available at the BioConductor website (in our case, for AGL and AFF platforms), we used them for the annotation. Otherwise (for ABI, ILL, and LGTC) the annotation packages were produced using the AnnBuilder package in R [].AbbreviationsABI: Applied Biosystems; AFF: Affymetrix; AGL: Agilent; DCLK: δC-doublecortin-like kinase; DEG: Differentially expressed genes; EBLRM: Empirical Bayes Linear Regression Model; FC: Fold Change; FDA: Food and Drug Administration; FDR: False discovery rate; GABA: Gamma amino butyric acid; GO: Gene Ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes; ILL: Illumina; LGTC: Home-spotted LGTC arrays; MAQC: MicroArray Quality Control Consortium; qRT-PCR: Quantitative reverse transcription PCR; UG: Unigene.Authors' contributionsPP conducted all the statistical analyses and drafted the manuscript, PAtH contributed to the study design, the data analysis and the drafting of the manuscript, EV worked on the biological interpretation of the results, GJS isolated the RNA and contributed to the biological interpretation, RHAMV performed the validation via qRT-PCR, YA was responsible for the microarray experiments, MdH and RK reannotated the microarray probe sequences, GJBvO corrected the manuscript, JdD and JB helped in the setting up of the experiments and contributed to the interpretation of the results, RM was mainly involved in the supervision of the statistical analyses and statistical interpretation of results.Supplementary MaterialAdditional file 1suppl.material.pedotti. contains a more detailed comparison of the performance of AGL and AFF microarrays.Click here for fileAdditional file 2table.S1.pedotti. contains a list of the genes selected for the validation with qRT-PCR and its results.Click here for fileAdditional file 3table.S2.pedotti. contains the hybridization design for the two color arrays (AGL and LGTC).Click here for fileAdditional file 4all.data.UG. contains the expression data of all the platforms for the subset of genes with overlapping UG identifiers.Click here for file
|
PMC2360727.txt
|
TITLE: Hepatitis C virus and non-Hodgkin's lymphoma: findings from the Swiss HIV Cohort Study AUTHORS: S Franceschi, J Polesel, M Rickenbach, L Dal Maso, N M Probst-Hensch, C Fux, M Cavassini, B Hasse, A Kofler, B Ledergerber, P Erb, G M Clifford ABSTRACT: Infections with hepatitis C virus (HCV) and, possibly, hepatitis B virus (HBV) are associated with an increased risk of non-Hodgkin's lymphoma (NHL) in the general population, but little information is available on the relationship between hepatitis viruses and NHL among people with HIV (PHIV). We conducted a matched case–control study nested in the Swiss HIV Cohort Study (SHCS). Two hundred and ninety-eight NHL cases and control subjects were matched by SHCS centre, gender, age group, CD4+ count at enrolment, and length of follow-up. Odds ratios (OR) and corresponding % confidence intervals (CI) were computed using logistic regression to evaluate the association between NHL and seropositivity for antibodies against HCV (anti-HCV) and hepatitis B core antigen (anti-HBc), and for hepatitis B surface antigen (HBsAg). Anti-HCV was not associated with increased NHL risk overall (OR=.; % CI: .–.), or in different strata of CD4+ count, age or gender. Only among men having sex with men was an association with anti-HCV found (OR=.; % CI: .–.). No relationships between NHL risk and anti-HBc or HBsAg emerged. Coinfection with HIV and HCV or HBV did not increase NHL risk compared to HIV alone in the SHCS. BODY: People infected with HIV (PHIV) have an approximately -fold increased risk of non-Hodgkin's lymphoma (NHL), mainly high-grade B-cell NHL, compared to the general population (Dal Maso and Franceschi, ). The role of HIV seems to be indirect and related to the effect of the virus on immunoregulation. Conversely, the causal role of oncogenic viruses such as Epstein–Barr virus (EBV) and, in much rarer instances, Kaposi's sarcoma herpes virus in the onset of NHL in PHIV is well established (Jaffe et al, ).An association between hepatitis C virus (HCV) infection and NHL in the general population has emerged in the last decade (Gasparotto et al, ; Negri et al, ) and a meta-analysis of studies from eight countries has shown a pooled relative risk of . (% confidence interval (CI): .–.) (Dal Maso and Franceschi, in press).Owing to shared routes of transmission, HCV and HIV coinfection is common and is associated with higher HCV RNA levels and increased risk of progression of HCV-related liver disease than in the presence of HCV infection alone (Sulkowski et al, ). Some investigators also attempted to evaluate whether coinfection with HCV increased the risk of NHL among PHIV, but did not find any association (Besson et al, ; Levine et al, ; Engels et al, ; de Sanjosé et al, ; Waters et al, ). None of these studies, however, included more than cases of NHL positive for both HIV and HCV (Besson et al, ; Levine et al, ; de Sanjosé et al, ; Waters et al, ).We have therefore, carried out a much larger case–control study nested in the Swiss HIV Cohort Study (SHCS) to assess the relationship between HCV infection and NHL in PHIV. On account of some reports of increased NHL risk in HIV-negative individuals seropositive for hepatitis B surface antigen (HBsAg) (Kim et al, ; Talamini et al, ), we also evaluated markers of HBV infection.MATERIALS AND METHODSSubjectsThe SHCS is an ongoing study that has been enrolling PHIV since , with some retrospective enrolment going back to , from seven large hospitals in Switzerland (www.shcs.ch). Follow-up visits take place every months and clinical events, such as opportunistic diseases, selected cancers (i.e. Kaposi's sarcoma, NHL, cervical cancer and Hodgkin's lymphoma), and death have always been recorded.The database used for the present study included information recorded for the SHCS up to October , and has been supplemented with independent information on cancer occurrence from a record-linkage study between the SHCS and eight Swiss Cantonal Cancer Registries (Clifford et al, ). Cases included in the present nested case–control study were PHIV who developed NHL, and for whom results on antibodies against HCV (anti-HCV), or, in their absence, a blood sample taken before NHL diagnosis (serum or plasma) to test for anti-HCV, was available. Out of a total of NHL cases identified, were excluded because information on some matching criteria (see below) was not available, and because HCV serology status was impossible to obtain. The present study finally included NHL cases (Table ). Primary brain lymphomas (PBL) were distinguished from non-PBLs, and non-PBLs were classified, when possible, by major histological subtype (i.e. diffuse large B-cell lymphomas, Burkitt lymphomas or other) using the World Health Organisation classification of lymphoid neoplasms (Jaffe et al, ).For each NHL case three control subjects were chosen at random among SHCS participants who had at least the same length of follow-up as the matched-case, and for whom HCV serology, or a blood sample to perform the test, was available. Matching criteria were: () centre of enrolment in the SHCS; () gender; () age group (in -year groups from to to ⩾ years); and () CD4+ count (<, –, –, ⩾ cells μl−) at enrolment. For five control subjects, the retrieved blood sample was inadequate for anti-HCV testing, leaving control subjects available for the present study (Table ). Follow-up was censored at the reference date (i.e. the date of NHL diagnosis for cases, and the date occurring after a similar length of follow-up for matched control subjects).As incidence of NHL was much greater in the early part of the SHCS and we did not match for year of enrolment, median year of enrolment was earlier for NHL cases (; interquartile range: –) than for control subjects (; –). Median year at cancer diagnosis for NHL cases was (interquartile range: –).HIV-transmission category was classified into three groups: () intravenous drug users (IDUs), () men having sex with men (MSM), and () heterosexual and others. ‘Others’ (i.e. transmission through blood and blood products and unknown transmission category) included only NHL cases and control subjects.This study was approved by the local ethical committees of the clinics collaborating with the SHCS and of the International Agency for Research on Cancer. Written informed consent was obtained from all SHCS participants.SEROLOGYAll hepatitis virus markers were measured in blood samples taken before NHL diagnosis or the corresponding reference date for controls. As systematic anti-HCV testing in the SHCS started in April , anti-HCV results could not be obtained from medical records for (.%) NHL cases and (.%) control subjects. New HCV testing on stored serum aliquots was, therefore, performed at the Institute for Medical Microbiology, University of Basel, Switzerland using third–generation ELISA (AxSYM anti-HCV, Abbott Diagnostic Division, Weisbaden, Germany). Reactive results were confirmed by immunoblot. Among PHIV for whom anti-HCV results were available from both medical records and new HCV testing, anti-HCV status agreed in .Information on antibodies against hepatitis B core antigen (anti-HBc) came from SHCS records for NHL cases and control subjects. New testing was performed on stored serum aliquots from NHL cases and control subjects, using microparticle enzyme immunoassay (AxSYM core TM, Abbott Diagnostic Division). Anti-HBc status could not be evaluated in one NHL case and three controls whose anti-HCV status was known.For HBsAg, information came from SHCS records for NHL cases and control subjects. New testing was performed on stored serum aliquots from NHL cases and control subjects, using microparticle immunoenzyme assay (AxSYM HBsAg version ., Abbott Diagnostic Division). Hepatitis B surface antigen status could not be evaluated in six NHL cases and controls whose anti-HCV status was known.Testing for antibodies against HBsAg and against HBe antigen was not systematically carried out in the SHCS, and testing for HCV RNA started only in February . Thus, these three additional viral markers were not considered in the present study. The CD4+ count was measured by flow cytometry.Statistical analysisConditional logistic regression was used to calculate odds ratios (ORs) and corresponding % CIs. All analyses were conditioned on all matching variables (see Subjects) and adjusted for HIV-transmission category and year of enrolment (before , –, after ). Additional adjustment for highly active antiretroviral therapy (HAART) use did not substantially change the ORs reported. Unconditional logistic regression was used for the analyses stratified by HIV transmission category, adjusted for all matching variables and year of enrolment.Heterogeneity of the ORs between strata of selected variables was tested by comparing the overall maximum-likelihood estimate for anti-HCV seropositivity to stratum-specific maximum-likelihood estimates. The test statistic was compared to the χ2 distribution with degrees of freedom equal to the number of strata minus one (Rothman and Greenland, ).RESULTSTable shows the distribution of NHL cases and control subjects by matching variables. Among the seven SHCS centres, Zurich alone contributed % of study participants. The majority of study participants were male (%) and % were between and years of age. Forty per cent had a CD4+ count at enrolment below cells μl−, and in % follow-up before the reference date was months or longer (Table ).Figure shows seropositivity for the three hepatitis virus markers by HIV transmission category among control subjects. Anti-HCV was detected in .% of IDUs, .% of MSM and .% of heterosexuals and other. Seropositivity for anti-HBc was nearly as frequent as seropositivity for anti-HCV among IDUs, but substantially higher than anti-HCV among MSM (.%) and heterosexuals and other (.%). Seropositivity for HBsAg was found in .% of IDUs, .% of MSM and .% of heterosexuals and other.None of the three hepatitis virus markers considered showed an association with NHL risk (Table ). ORs were . (% CI: .–.) for anti-HCV, . (% CI: .–.) for anti-HBc and . (% CI: .–.) for HBsAg seropositivity.Table shows the influence of anti-HCV seropositivity on NHL risk in separate strata of CD4+ count at enrolment, age, gender and HIV transmission category. Anti-HCV+ PHIV did not show an increased NHL risk compared to anti-HCV− PHIV in any separate stratum except for HIV transmission category, where a significant association between anti-HCV and NHL risk emerged among MSM (OR=.; % CI: .–.). In no instance, however, was the effect of anti-HCV seropositivity on NHL risk significantly heterogeneous across the strata of the variables considered (Table ).Seropositivity for anti-HBc and HBsAg was not associated with NHL risk in any stratum of CD4+ count at enrolment, age, gender or HIV-transmission category (data not shown).Figure shows the percent distribution of CD4+ count at lymphoma diagnosis and NHL subtype separately among anti-HCV+ and anti-HCV− NHL cases. Although % CIs always overlapped, a slightly lower proportion of anti-HCV+ than anti-HCV− NHL cases had less than CD4+ cells μl− at cancer diagnosis (. vs .%, respectively), or were diagnosed with PBL (. vs .%, respectively).DISCUSSIONSeropositivity for HBV and HCV did not increase NHL risk among PHIV in the SHCS. Our findings on HCV, the most studied hepatitis virus in respect to NHL risk (Dal Maso and Franceschi, in press), are consistent with previous studies of PHIV that also did not show an association (Besson et al, ; Levine et al, ; Waters et al, ). Compared to earlier reports on the topic, our study was, however, much larger, allowing stratification by, and more accurate allowance for, other correlates of NHL risk.As expected, levels of seropositivity for hepatitis viruses were high among PHIV, most notably among IDUs. The higher anti-HBc/anti-HCV seropositivity ratio among MSM (.) and heterosexuals and other (.) compared to IDUs (.) confirms the much greater efficiency of the sexual-route of transmission for HBV than HCV (IARC, ).We were not able to review the slides of NHL cases, and histological type was not specified in % of the cases, thus limiting the chance of detecting qualitative differences between anti-HCV+ and anti-HCV− lymphomas. We did detect, however, a slight under-representation of anti-HCV+ compared to anti-HCV− cases among NHL occurring at CD4+ counts below cells μl− and among PBL, which are the two lymphoma subtypes where EBV is known to be most strongly implicated (Jaffe et al, ). Among transplant patients, for instance, the majority of NHL is associated with EBV, but NHL has also been reported in EBV− patients and tends to occur longer after organ transplant, when immunosuppression is milder, than in EBV+ patients (Leblond et al, ).As among transplant patients, the strong role of immunodeficiency and EBV in NHL (Jaffe et al, ) makes the evaluation of possible weak risk factors such as HCV infection much more difficult among PHIV than in the general population (Negri et al, ). A hint of an association between NHL risk and HCV infection emerged among MSM, who had lower levels of anti-HCV seropositivity than heterosexuals and IDUs. Indeed, nearly all IDUs were anti-HCV+, thus preventing any meaningful evaluation of the effect of HCV infection on NHL risk. Furthermore, MSM (median age ) were older than subjects in the other HIV-transmission categories (median age and for IDUs and heterosexuals and other, respectively). A two-fold increased NHL risk was also found among anti-HCV+ PHIV aged years or older, although it did not reach statistical significance. Older age may be a correlate of longer exposure to HCV, and, hence, higher risk of HCV-related complications including NHL. Indeed, the relative risk for cirrhosis in anti-HCV+ vs anti-HCV− PHIV increased between the pre-HAART and HAART era (Giordano et al, ) and the excess of hepatocellular carcinoma currently seen in PHIV (Clifford et al, ) had not clearly emerged earlier in the epidemic (Beral and Newton, ).An important strength of the SHCS is the fact that it is very representative of PHIV in Switzerland. It has been estimated that, since the beginning of the HIV epidemic, % of PHIV, and % of people diagnosed with AIDS in Switzerland have been enrolled in the SHCS (www.shcs.ch). Weaknesses of the SHCS include incomplete information on time of HIV seroconversion, which prevented us from using years of follow-up as an exact proxy for duration of HIV infection. In addition, no information was available on EBV infection and on the activity of HCV infection, that is the presence of serum HCV RNA. The vast majority of persons acutely infected with HCV develop chronic persistent viraemia and, especially among PHIV, anti-HCV seropositivity is likely to correspond to chronic HCV infection (Sulkowski et al, ).In conclusion, coinfection with HIV and HCV or HBV did not increase NHL risk compared to HIV alone in the SHCS. However, as the life expectancy of PHIV increases, the influence of HCV infection on cancer risk deserves further study as the high prevalence of HCV could result in huge attributable risks, even in the presence of weak relative risks.
|
PMC1382264.txt
|
TITLE: Recovery and evolutionary analysis of complete integron gene cassette arrays from Vibrio AUTHORS: Yan Boucher, Camilla L Nesbø, Michael J Joss, Andrew Robinson, Bridget C Mabbutt, Michael R Gillings, W Ford Doolittle, HW Stokes ABSTRACT: BackgroundIntegrons are genetic elements capable of the acquisition, rearrangement and expression of genes contained in gene cassettes. Gene cassettes generally consist of a promoterless gene associated with a recombination site known as a -base element (-be). Multiple insertion events can lead to the assembly of large integron-associated cassette arrays. The most striking examples are found in Vibrio, where such cassette arrays are widespread and can range from kb to kb. Besides those found in completely sequenced genomes, no such array has yet been recovered in its entirety. We describe an approach to systematically isolate, sequence and annotate large integron gene cassette arrays from bacterial strains.ResultsThe complete Vibrio sp. DAT722 integron cassette array was determined through the streamlined approach described here. To place it in an evolutionary context, we compare the DAT722 array to known vibrio arrays and performed phylogenetic analyses for all of its components (integrase, -be sites, gene cassette encoded genes). It differs extensively in terms of genomic context as well as gene cassette content and organization. The phylogenetic tree of the -be sites collectively found in the Vibrio gene cassette pool suggests frequent transfer of cassettes within and between Vibrio species, with slower transfer rates between more phylogenetically distant relatives. We also identify multiple cases where non-integron chromosomal genes seem to have been assembled into gene cassettes and others where cassettes have been inserted into chromosomal locations outside integrons.ConclusionOur systematic approach greatly facilitates the isolation and annotation of large integrons gene cassette arrays. Comparative analysis of the Vibrio sp. DAT722 integron obtained through this approach to those found in other vibrios confirms the role of this genetic element in promoting lateral gene transfer and suggests a high rate of gene gain/loss relative to most other loci on vibrio chromosomes. We identify a relationship between the phylogenetic distance separating two species and the rate at which they exchange gene cassettes, interactions between the non-mobile portion of bacterial genomes and the vibrio gene cassette pool as well as intragenomic translocation events of integrons in vibrios. BODY: BackgroundThe integron/gene cassette system is now known to be widely distributed amongst the Bacteria and is a major contributor to the dispersal of genes by Lateral Gene Transfer (LGT). The system can facilitate LGT by virtue of the fact that the integron encodes a site-specific recombination (SSR) system [,]. The targets for SSR are gene cassettes []. These are independently mobilizable units of DNA, usually composed of a single gene bound by a recombination site, designated a -base element (-be) or alternatively attC. The DNA integrase catalysing SSR, IntI, is encoded by the integron. This integrase recognises two families of recombination sites. The first of these is the cassette associated -be [] and the second is a site contained within the integron designated attI. IntI can catalyze a reversible reaction by which gene cassettes can be inserted into, or excised from, an integron via recombination between the attI site and a -be or between two -be sites [,]. Multiple recombination events are common. As a consequence, integrons are normally associated with arrays comprising multiple cassettes, which in some species of Vibrio can number well in excess of a hundred []. Gene cassettes do not normally include a promoter. Instead, where it has been examined, transcription of cassette genes is driven by a promoter (Pc) located in the integron itself []. Thus, the integron and gene cassettes comprise both a gene capture and gene expression system.Integrons were first identified in the context of multi-drug resistance. In this context, pathogenic bacteria are often resistant to many antibiotics as a consequence of possessing an integron that has captured several gene cassettes containing resistance genes []. These integrons recovered from clinical environments are often embedded in other types of mobile elements such as plasmids and transposons []. It is now clear however that integrons are also a common feature of the chromosomes of various bacteria, being found in the gamma-proteobacteria (vibrios, xanthomonads, pseudomonads, Shewanella), beta-proteobacteria (Nitrosomonas europea) and spirochaetes (Treponema denticola) (Figure ). Another notable feature of chromosomal integrons is the wide diversity of functions encoded by the genes found in their gene cassettes []. This, coupled with the observation that cassette arrays can be quite large, clearly hints at a system that can greatly impact on the adaptive potential of bacteria []. Given that cassette associated genes are part of the gene pool that is LGT-associated and mobilizable, they also represent a community resource that can be shared between individuals [].Figure 1Best maximum likelihood phylogeny of representative integron integrases (IntI). For integrons carried on a mobile element such as a plasmid or transposon, the taxon name represents this genetic element and the host name is in parentheses. Some integron integrases can be attributed to a specific integron class, as indicated by the name of the clade to which they belong. Support values displayed above the nodes represent the consensus of the best maximum likelihood distance trees of bootstrap pseudo-replicates of the dataset (only values >% are displayed). The sequences of the intI genes from Vibrio sp. DAT865, DAT868 and DAT722 strains were obtained in this study.Bacteria of the Vibrio genus are rapidly becoming a model system for the study of the chromosomal integron/gene cassette system [-]. As noted above, their cassette arrays are characteristically large, encompassing several percent of genomic coding capacity in many cases. Representatives of this genus display such a high level of variation in terms genome size and content that "species" units can rarely be identified by the sequencing of one or a few genetic markers []. Many complete integron arrays have now been assembled from various species of Vibrio as a result of whole genome sequencing initiatives. These include V. cholerae, V. parahaemolyticus, V. vulnificus (two strains) and V. fischeri [-]. These collective sequencing efforts have helped to highlight the enormous amount of novel genetic diversity contained within integron arrays. However, it is also becoming clear that whole genome sequencing is a relatively blunt instrument for assessing gene cassette diversity. This is because, although individual strains of Vibrio can contain large arrays, the total number of cassettes harboured by any single individual is very small compared to the overall size of this community resource []. As an alternative approach to recovering gene cassettes, the cassette PCR technique has been developed []. This method selectively amplifies gene cassettes to the exclusion of other genomic sequences and can be applied to both metagenomic (environmental) DNA and to the DNA of defined strains. However, cassette PCR only recovers cassettes in isolation (i.e. outside their genetic context) and cannot aid in the assembly of contiguous arrays. Also, although gene cassette PCR is selective, there is always the possibility for some false positives (non-cassette DNA fragments being amplified) []. When a gene cassette is found within an array, there is no doubt about its nature.To more rapidly access and analyse the mobile gene cassette pool, we adopted a genomics approach that allows us to specifically isolate DNA fragments containing large integron gene cassette arrays and exclude the vast majority of the genome that is common to most members of a species. No such large cassette array has previously been isolated and fully sequenced without completely sequencing the genome of its host. Using our streamlined approach, we isolate and sequence the integron from a close relative of the widespread marine bacterium Vibrio harveyi, a well-known member of the core vibrio group to which V. parahaemolyticus also belongs (Figure ). To place it in an evolutionary context, we compare it to known vibrio integron arrays and perform a phylogenetic analysis of all of its components (integrase, -be sites, gene cassette encoded genes). This reveals strong interaction between vibrio integrons through LGT, high variability of array contents and wide functional diversity of gene cassettes. These analyses also yield insights on the processes of gene cassette recruitment from non-mobile genes and the possibility of cassette de-recruitment. Differences in the genomic context of integrons from various vibrios suggest several events of intragenomic translocation for these genetic elements.Figure 2Best maximum likelihood phylogeny of Vibrio species based on the RNA polymerase B gene (rpoB). The strain which had its entire integron gene cassette array sequenced in this study is boxed. Organisms that had their complete genome sequenced are in bold (the sequence of the rpoB genes of all other strains were generated in this study). The gene immediately upstream of the integron, when known, is indicated next to the taxon name: white dot (hypothetical protein VC1310), black dot (ribosomal protein RplT) or grey dot (hypothetical protein VCA0034). Support values displayed above the nodes represent the consensus of the best maximum likelihood trees of bootstrap pseudo-replicates of the dataset (only values >% are displayed).Results and discussionCharacterization of the Vibrio sp. DAT722 cassette arrayThe method used to isolate a complete vibrio gene cassette array takes advantage of the fact that cassettes are linked, having been assembled at a single locus, the integron attI site. By making genomic libraries and screening individual clones by cassette PCR or with PCR primers targeting the intI gene, we can recover large inserts that comprise exclusively, or in large part, of contiguous arrays of mobile gene cassettes. Here we demonstrate this principle on an environmental strain of a Vibrio sp. from an aquaculture facility in Darwin, Australia. We show that using a fosmid vector (insert size between – kb) to construct a genomic library ( clones), we can recover the whole kb integron array in four PCR screening rounds, the first of which identifies a clone bearing the intI. A structured query language (SQL) script was developed to facilitate annotation of integron arrays, made difficult by the presence of non-coding elements such as -be sites. The script performs a search for short DNA motifs on a DNA segment and detects intervening ORFs, outputting annotation in GenBank format.The gene cassette array of Vibrio sp. isolate DAT722 has an average G+C content of .%, compared to .% for the flanking DNA regions. A G+C content lower than the genomic average of the host is typical of vibrio integrons, and differences ranging from . to .% can be found in completely sequenced vibrio genomes harbouring an integron (data not shown). The DAT722 array contains gene cassettes (Figure ), falling between the shortest vibrio array of cassettes (V. fischeri) and the longest array of cassettes (V. vulnificus CMCP6) (Table ). However, DAT722 contains a high diversity of cassettes, comprising different cassette types out of a total of cassettes. A cassette type represents either a non-paralogous cassette or a group of paralogous cassettes, the latter being defined as a group of cassette that have reciprocal BLASTP or BLASTN matches with an e-value of 1E- or less). This number of different cassette types is only equalled or surpassed by much larger arrays (V. cholerae with cassette types of a total cassettes and V. vulnificus CMCP6 with cassette types of a total cassettes). Consistent with this cassette diversity, the DAT722 array also encodes a wide variety of functions including DNA modification and primary metabolic enzymes (Table ). Also typical of gene cassette arrays, there is a high proportion of cassettes either with no detectable homolog ( cassettes) or with sequence matches only to hypothetical proteins ( cassettes). Only a minority of cassettes encodes a protein to which a putative function can be assigned: cassettes can be assigned a specific function with reasonable certainty and another a broad function (for example matching a general super-family of proteins such as acetyltransferases).Figure 3Diagram of the Vibrio sp. DAT722 integron gene cassette arrayTable 1Characteristics of Vibrio integronsVvuCMaVvuYJVchVpaVfiVspDATlength (kb)% of host genome2.%.%.%.%.%n.d.bno. cassettes2191871736936116 non-coding10757118725 coding112130162612991no. cassette types1418895593194 paralogous groups182844729 singlets1236051522985a VvuCM (Vibrio vulnificus CMCP6); VvuYJ (Vibrio vulnificus YJ016); Vch (Vibrio cholerae N16961); Vpa (Vibrio parahaemolyticus RIMD2210633); Vfi (Vibrio fischeri ES114); VspDAT (Vibrio sp. DAT722) bn.d. (not determined)Table 2Identifiable functions of the Vibrio sp. DAT722 gene cassette encoded proteinsCassette no.Putative functionSource of best hitaE-valueIdentity (%)DNA modification11DNA topoisomerase IShewanella amazonensis1E-966857NUDIX hydrolaseDeinococcus radiodurans1E-294659DNA topology modulation proteinBacillus cereus5E-4551Primary metabolism54-methyl-(B-hydroxyethyl)-thiazole monophosphate biosynthesisBacillus anthracis3E-152931beta-phosphoglucomutaseVibrio vulnificus4E-867075selenocysteine lyaseAnabaena variabilis4E-445578maltose O-acetyltransferaseMethanosarcina acetivorans9E-4147Other2lysogenic conversion proteinBacteriophage P2-EC55E-413320antibiotic biosynthesis monooxygenaseRalstonia eutropha3E-346424acetyltransferaseStreptomyces avermitilis6E-273339acetyltransferaseVibrio parahaemolyticus9E-869954acetyltransferaseLegionella pneumophila1E-173184acetyltransferaseParacoccus denitrificans3E- & 91acetyltransferaseVibrio vulnificus3E-6766103acetyltransferaseVibrio fischeri4E-617669histone acetyltransferaseVibrio vulnificus1E-163174haemagglutinin associated proteinVibrio cholerae1E-1168897cold shock domainVibrio vulnificus5E-7177102toxin-antitoxin plasmid stability systemVibrio vulnificus4E-4894aBLASTP hits from non-redundant database as of July 2005Interestingly, Vibrio gene cassette arrays vary enormously in the proportions of their contents which is composed of non-coding cassettes. Cassettes are defined as "non-coding" when they have no BLASTX matches and they do not contain ORFs longer than bp. Given this definition, the proportion of the array made-up of non-coding cassettes can vary from % (V. cholerae) to % (V. vulnificus CMCP6) and does not appear to be correlated with array size. There seems to be some correlation, however, between array size and content in paralogous (whether coding or non-coding) cassettes, as all large arrays (the two V. vulnificus strains and V. cholerae) have in excess of % of cassettes repeated at least twice in the array while smaller arrays all have around % of paralogs (Table ).One remarkable feature of non-coding cassettes is that they either have very close homolog(s) in other vibrios (>% nucleotide identity) or no homologs at all. In the Vibrio sp. DAT722 array, of non-coding cassettes, have no significant BLASTN hits and are closely related to non-coding cassettes from other vibrios. None of these cassettes seem to represent pseudo-genes, as they do not display any matches to public databases using amino acid translations of all possible reading frames as query (BLASTX). Furthermore, all cassettes defined as non-coding that do have BLASTN matches in other vibrios are similar to these homologs across their whole length. Non-coding cassettes also tend to be paralogous in all vibrio integrons. Although there are only copies of the most prevalent non-coding cassette in the Vibrio sp. DAT722 array, there is a highly paralogous family of non-coding cassettes which has representatives in V. vulnificus CMCP6 and in V. vulnificus YJ016.Coding cassettes are also frequently paralogous (i.e. have reciprocal BLASTP or BLASTN matches with an e-value of 1E- or less). For example, they account for out of the paralogous cassettes in the DAT722 array. The V. cholerae array displays the most paralogy among coding cassettes, boasting proteins with some degree of paralogy. However, only two of its paralogous groups have more than members, even when a generous BLASTP similarity cut-off is used to determine paralogy (e-value 1E-). Coding cassettes differ most significantly from non-coding cassettes by having homologs both inside and outside of the vibrios. Out of coding cassettes in the Vibrio sp. DAT722 array, have no BLASTP or BLASTX hit (e-value cutoff 1E-), have a best match to other vibrios and have a best match outside vibrios.Variability of gene cassette array compositionComparison of cassette content between vibrio strains/species arrays reveals important variability. The number of cassettes shared between the arrays of two different species of Vibrio, counting repeated paralogous cassettes only once, varies between and (arrays contain anywhere from to cassette types) (Table ). The two V. vulnificus strains, respectively containing and cassette types, only have cassettes in common. This represents a –% shared content (based on the two different array sizes), which is much lower than the –% shared gene content of their whole genomes when compared with the same homology criteria (BLAST cut-off of 1E-). The difference in shared gene content between arrays and genomes is even more pronounced when comparing different species. For example, V. vulnificus CMCP6 shares –% of its genomic genes with V. cholerae but only –% of its array genes.Table 3Gene cassettesa shared between the integrons of Vibrio speciesVvuCMbVvuYJVchVpaVfiVspDATVvuCM14134148212VvuYJ-881110110Vch--951046Vpa---5929Vfi----314VspDAT-----94aParalogous cassettes are only counted once, the middle diagonal displays the number of different cassette types found in each strainbVvuCM (Vibrio vulnificus CMCP6); VvuYJ (Vibrio vulnificus YJ016); Vch (Vibrio cholerae N16961); Vpa (Vibrio parahaemolyticus RIMD2210633); Vfi (Vibrio fischeri ES114); VspDAT (Vibrio sp. DAT722)Origins and fates of gene cassettesThere are multiple examples of cassette-encoded vibrio genes only found in a single vibrio strain but showing strong similarity to chromosomal genes from other types of bacteria. Two striking examples are the proteins coded by cassettes and from the DAT722 array. Cassette encodes a putative DNA topoisomerase that has % amino acid identity to a similarly annotated protein from Shewanella amazonensis (Table ). The cassette protein also shares % identity with proteins coded by uncultured marine alpha-proteobacteria from the SAR116 cluster. The protein coded by cassette shares % identity (e-value 4E-) with selenocysteine lyase from the cyanobacteria Anabaena variabilis yet does not display homology to any other protein in current databases. Another notable case is the hypothetical protein encoded by cassette , whose only homologs are found in two Salmonella and two Bordetella species, with which it shares ~% amino acid identity. It is not possible to perform meaningful phylogenetic analyses of these genes to get a more complete picture of their evolutionary history, as there are not enough known homologs. However, the high degree of similarity between these cassettes and their non-cassette homologs, combined with the lack of any vibrio homologs, suggest that they have been recruited from chromosomal genes of other types of bacteria.Vibrio cassette-encoded ORFs are also found to have non-cassette ORFs homologs in other vibrio strains. The three most striking cases from the DAT722 array are cassettes , and . Cassette codes for a putative beta-phosphoglucomutase that has % amino acid identity with a non-cassette ORF occuring in both V. vulnificus genomes, as well as more distant homologs in a variety of other bacteria (Figure 4A). Cassette (Figure 4B) codes for a haemagglutinin associated protein and shares % amino acid identity with a V. cholerae cassette-encoded homolog. This cassette product also has % identity with a non-cassette protein found in V. parahaemolyticus and Photobacterium profundum. Cassette (Figure 4C) encodes a putative acetyltransferase that shares % identity with a non-cassette ORF in V. fischeri, and is additionally found in many other bacteria. No traces of -be site can be found next to the non-cassette homologs, and it is unclear wether these vibrio non-cassette ORFs closely related to cassette-encoded ORFs represent cassette generation events or non-specific cassette insertion events. The latter scenario of cassette insertion outside an integron context has been previously documented for gene cassettes generally associated with class integrons []. The nature of this secondary site insertion appears to generate a non-functional -be thereby preventing subsequent mobilization [].Figure 4Best maximum likelihood phylogenies of some Vibrio sp. DAT722 gene cassette encoded proteins. The letter in brackets following the taxon name indicates the genetic context of the gene coding for the protein: C (found in a gene cassette), N (host organism has an integron, but this gene is not found in a gene cassette), X (host organism does not have an integron). Support values displayed above the nodes represent the consensus of the best maximum likelihood distance trees of bootstrap pseudo-replicates of the dataset (only values >% are displayed).There is one clear-cut example of cassette insertion outside an integron emerging from the analysis of the DAT722 array. Cassette encodes two short proteins together forming a toxin-antitoxin stability system widespread in vibrio integrons [] (a phylogenetic tree of the antoxin protein, which is highly similar to the tree of the toxin protein, is displayed in Figure 4D). The amino acid sequence of the antitoxin protein from DAT722 is % identical to its non-cassette homolog in Shewanella oneidensis and but only % identical to its relative encoded by a cassette in the sequenced V. fischeri array. This DAT722 cassette-encoded protein also shares significant similarity to a Xanthomonas non-cassette homolog (% identity). It seems likely in this case that a vibrio gene cassette has been inserted non-specifically in the S. oneidensis genome, since the Shewanella homolog clusters strongly inside a clade of vibrio cassette-encoded homologs. Although Shewanella oneidensis is known to possess a functional integron, it only harbours cassettes and the vibrio-like antitoxin gene is found at a different chromosomal location (Figure ).Non-coding cassettes differ from those that are coding in terms of distribution, as no homologs outside the vibrios can be found for these cassettes. Instead, closely related homologs in other vibrio species and frequent paralogs in the same array are found for most non-coding cassettes. For example, non-coding cassette is % and % identical to paralogous cassettes and at the nucleotide level and shares % identity with homologous non-coding V. parahaemolyticus and V. vulnificus cassettes. This high level of sequence conservation from one species to another suggests that non-coding cassettes are under purifying selection. The degree of sequence conservation of paralogous non-coding cassettes could also be enhanced by a process of gene conversion, where recombination between paralogs leads to a more homogeneous set of gene sequences []. This sequence conservation, linked with their limited distribution and frequent paralogy leads to two possibilities: ) They could be selfish elements associated with vibrio integrons; ) They have a function beneficial to the integron (contain promoters that regulate neighboring gene cassette encoded gene expression or can be transcribed as RNA molecules that play a role in integron regulation). In either case, they most likely have originated within the vibrios and spread to most representatives of this group harbouring integrons.Translocation of integron gene cassette arraysThe genomic context of an integron and its associated cassette array can provide important information on its history. The context of the Vibrio sp. DAT722 cassette array differs significantly from the context of its closest sequenced relative, the V. parahaemolyticus array. Although the same genes (in an identical order) are found ' of the integrase gene in both arrays, the context suddenly changes past this conserved block. None of the next genes upstream of this block in DAT722 has a homolog in a nearby location in V. parahaemolyticus, some of them are found . mb away on the large V. parahaemolyticus chromosome and two of them match genes found some kb after the end of the array. Similarly, genes found after the end of these two arrays are completely different, with the exception of homologous transposase genes (IS3 family) being found near the last cassette of both arrays. As suggested by Rowe-Magnus et al. [], the fact that the core vibrio group (which includes V. parahaemolyticus and Vibrio sp. DAT722) integron integrases are flanked by a different gene than their V. cholerae and V. vulnificus homologs most likely represents an integron translocation event in the ancestor of the former group (Figure ). It therefore seems that an integron was inserted at a given location (next to the VC1310 gene) in the ancestor of the V. harveyi / V. parahaemolyticus core vibrio group, and that the array might have then been moved by a transposition or genome rearrangement event. However, since we do not know the degree of synteny (conservation of gene order) between the V. parahaemolyticus and Vibrio sp. DAT genomes, the possibility that the genome of either of these strains was rearranged on both side of its integron cannot be excluded.Translocation events like the potential integron movement in the V. parahaemolyticus and/or Vibrio sp. DAT722 lineage(s) seem likely, as the V. cholerae integron has certainly been moved from its original position on the large chromosome, being the only Vibrio integron found on the smaller of the two chromosomes typically found in this genus []. As is the case for the arrays of Vibrio sp. DAT722 and V. parahaemolyticus, the block of genes flanking the integron integrase gene is conserved between V. cholerae and V. vulnificus (limited to genes rather than in this case, including the ribosomal protein gene directly flanking the integrase gene, rplT), with entirely different contexts beyond this block. This block of genes would have accompanied the integron when it was translocated from one V. cholerae chromosome to the other. Transposases could have been involved, as the V. cholerae integron (including the genes block linked to it) is flanked by genes coding for such proteins on both sides.Those translocation events do not seem to occur solely within and between chromosomes. Vibrio salmonicida carries a large MDa plasmid (pRSV1) harbouring an integron []. The integrase of the pRSV1 integron is clearly a vibrio-type integrase, not one of the class or SXT constin integron integrases previously found on vibrio plasmids. This is shown by its strongly supported clustering with vibrio-type integrases in phylogenetic analyses (Figure ). One of the gene cassettes of this integron encodes a transposase and another a dihydrofolate reductase type I usually found in class integrons. It therefore seems that the V. salmonicida integron originates from a translocation (possibly transposition) event of a chromosomal vibrio-type integron to a plasmid and later acquired cassettes from class integrons.The frequent association of vibrio integrons with transposases, along with the variability in their genetic context and the DNA molecule carrying them (plasmid, large or small chromosome), suggest that some of them could be mobilized. As discussed above, there is strong evidence for intragenomic and intracellular translocation of integrons, but what remains to be established is whether lateral transfer of whole gene cassette arrays between strains and species can occur.Gene cassette transfers within and between Vibrio speciesTo examine the movement of gene cassettes between strains and species, we extracted and aligned all -be sites from the vibrio cassette arrays of Vibrio sp. DAT722, V. parahaemolyticus RIMD2210633, V. cholerae N16961, V. vulnificus YJ016, V. vulnificus CMCP6 and V. fischeri ES114(a total of sequences). Figure presents the phylogenetic analysis of -be based on this alignment. Three major clades emerge in the resulting tree: ) All V. fischeri -be sites cluster together to the exclusion of all others with strong statistical support (% bootstrap value); ) most (/ or %) V. cholerae -be sites fall in a single clade along with four -be sites from other species with robust support (% bootstrap value); ) a large proportion (/ or %) of -be sites from either of the two V. vulnificus strains form a cluster with a single -be site from V. parahaemolyticus (weakly supported, % bootstrap value). The vast majority of clusters beside these three are formed of -be sites from a variety of species.Figure 5Phylogenetic analysis of the -be sites found in Vibrio chromosomal integrons. Each terminal branch is colour-coded according to the organism hosting the particular -be site it represents (see legend). The tree displayed in the best distance neighbor-joining tree obtained using PAUP*. Percent bootstrap support values represent the consensus of distance neighbor-joining trees obtained from pseudo-replicates of the dataset. They are only displayed for important backbone nodes when the value exceeds %. internal nodes are supported with >% bootstrap value, of which represent mixed species clades.The major clades observed for V. fischeri, V. cholerae and V. vulnificus contrast with the absence of such phylogenetic structuring for V. parahaemolyticus or Vibrio sp. DAT722. The presence/absence of species-specific clades in the -be sites tree seems to be correlated with the phylogenetic distance of a species from others represented the dataset. In the most phylogenetically distant species of the dataset, V. fischeri, no exchange of -be sites with other species (indicated by mixed clades) is observed. V. vulnificus and V. cholerae also lack close relatives in the dataset, both being represented by relatively long branches in the rpoB tree (Figure ). They both display one major species-specific clade along with some exchange of -be sites with other species. Vibrio sp. DAT722 and Vibrio parahaemolyticus are relatively close species (Figure ) and neither of them displays a species-specific clade in the -be site tree.Of nodes with a bootstrap support value >%, contain only -be sites from the same species and contain -be sites from multiple species, indicating transfer of gene cassettes between species. All species of vibrio are involved in such transfers except for the more phylogenetically distant V. fischeri ( events involve V. parahaemolyticus as either donor or recipient, V. cholerae, Vibrio sp. DAT722 and V. vulnificus). Among the well-supported nodes containing -be from a single species only, nodes contain -be sites solely from V. vulnificus. Of these , contain -be sites from a single strain and contain -be sites from both CMCP6 and YJ016. Transfer of gene cassettes can therefore be assumed to happen at both the subspecies and species levels among vibrios. The frequency of such transfer events seems to be broadly correlated with phylogenetic distance between the proponents (with a higher frequency between close relatives). Comparison of closely related vibrios, such as Vibrio sp. DAT722 and V. parahaemolyticus or individual strains of V. vulnificus, shows evidence of frequent transfer events that disrupt vertical inheritance patterns, ultimately leading to the absence of a major monophyletic clade for their -be sites. At the other extreme, transfer of cassettes between V. fischeri and other vibrios must be very rare or absent, as illustrated by the monophyly of all of its -be sites.A phenomenon that could contribute to enhancing the degree of homogeneity between the -be sites found in a given species is gene conversion. This non-reciprocal recombination process occurs between relatively small tracts of DNA and is known to lead to the sequence homogenization of multiple paralogous genes found in a given organism, such as rRNA genes []. This could also be the case with -be sites, and linked with the process of nucleotide substitutions due to replication or repair errors, could lead to the creation and maintenance of distinct pools of -be sites in divergent organisms.Evolutionary rate of gene cassette arraysMixed phylogenetic clades composed of -be sites from different vibrio species, observed with V. cholerae and V. metschnikovii by Rowe-Magnus et al. [] and here with a more substantial dataset, suggest frequent transfer of gene cassettes between vibrios. However, the phylogenetic structure present in the -be site tree, where -be sites from a given species cluster more frequently with others sharing a similar host, also suggest a correlation between phylogenetic distance and frequency of cassette exchange. This could be due to a lower efficiency of integrons at integrating gene cassettes from divergent species, as their -be sites did not co-evolve with the integrase and attI site of the host species. This hypothesis is supported by the recent findings of Biskri et al. [] that the range of -be sites efficiently recombined by the V. cholerae integrase is narrower than the range of -be sites efficiently recombined by class integrases. Performing similar assays of recombination efficiency using combinations of -be sites, attI sites and integrases from different vibrio species could allow direct testing of a correlation between phylogenetic distance and frequency of cassette exchange in vibrios.Looking at the difference between two gene cassette arrays relative to their genomic context can yield insight on the rate of evolution of those arrays versus the rate at which the rest of the genome evolves. The two most closely related strains of Vibrio for which we have complete array sequences, V. vulnificus YJ016 and V. vulnificus CMCP6, share –% of their gene cassettes ( cassettes on an array size of and non-repeated cassettes, respectively). This represents a minimum of gene cassette gain/loss events between the arrays ( of YJ016 cassettes are not found in CMCP6 and the latter has additional cassettes). This divergence is striking, as these strains are very closely related, displaying differences of only bp in the bp of their intI genes and bp in a bp fragment of their rpoB gene. Furthermore, the shared content of the array is also misleading for the estimation of cassette gain/loss events, as shared cassettes do not necessarily represent an evolutionarily conserved trait, but can have different histories, as is suggested by the lack of conservation in gene order between the arrays as opposed to the relatively conserved organization of the surrounding genomic genes []. Given that the impact of acquiring/losing a gene on organismal fitness would generally be greater than that of a nucleotide substitution, integrons are of immense evolutionary value for their host. Comparison of the cassette content of integrons from a large number of closely related strains, combined with a multi-locus sequence analysis to estimate the nucleotide substitution rate in the genome of these strains, could allow for a quantitative estimate of the relative evolutionary rate of integrons.ConclusionThe entire kb integron gene cassette array of Vibrio sp. DAT722 was isolated and sequenced using a streamlined approach that could be used with most, if not all, integron-containing bacteria (Figure ). The ease of preparation and high efficiency of fosmid libraries combined with the possibility to screen for integrons by PCR (targeting either of their main components, the integron integrase gene or -be sites), makes this approach widely applicable. Non-coding genetic elements such as -be sites are not readily identified by automated genome annotation tools, which focus mostly on identifying ORFs. Even for the identification of ORFs encoded within gene cassette arrays, these programs perform poorly, as they often identify the -be sites as being part of gene cassette ORFs or being ORFs themselves. This is in part due to the frequent use of alternative start codons in gene cassette encoded ORFs []. The approach implemented here, a simple motif search for the conserved ends of -be sites identifies gene cassettes with almost % efficiency, provided that sequence information is available from gene cassettes of an integron belonging to the same class as the target. By searching for the -be sites first and subsequently identifying ORFs that fall in-between these elements (allowing for minimal overlap), the accuracy of the annotation is improved significantly.Although the mechanistic aspects of integrons are now relatively well understood (recombinase activity, preferential integration of cassettes at the attI site, etc.), little is known on their general biology, such as the size range of their gene cassette arrays, the composition and diversity of their genetic pool and their rate of evolution. Isolating and comparing gene cassette arrays from multiple closely related organisms is necessary to understand their role in natural environments. Because of the almost ubiquitous presence of large integron gene cassette arrays in vibrios and the wide distribution and ease of isolation of these bacteria, they represent an ideal model system to study this genetic element and its impact on natural populations. The comparative analyses performed here confirm previous studies as to the functional diversity contained in vibrio integrons as well as the high variability of their gene cassette content [,]. Additionally, evolutionary analyses of the gene cassettes of the Vibrio sp. DAT722 array revealed that these mobile elements can be recruited from genomic genes of various origins and can also be de-recruited, becoming a 'sedentary' gene after a non-specific integration event and the loss of its associated -be sites. This suggests a two-way interaction between the gene pool made up of gene cassettes and the larger gene pool composed of non-mobile bacterial genes. Our phylogenetic analysis of -be sites suggests that the gene cassette pool itself is fragmented along phylogenetic lines. The -be sites linked to gene cassettes are not homogeneous across all vibrios but tend to be more similar if found in closely related strains or species, suggesting more frequent transfers of gene cassettes among close relatives. Accumulation of sequence data on complete gene cassette arrays will allow for more precise estimates of the rates of gain, loss and transfer of gene cassettes and perhaps even on the degree of gene order conservation, provided that the arrays of very closely related organisms can be isolated.MethodsDNA extraction and strain characterizationThe Vibrio harveyi ATCC14126 reference strain was obtained from the American Type and Culture Collection. All other strains used in this study were isolated from marine environments in Australia: strains labeled DAT were obtained from aquaculture tanks used to raise mud-crab larvae in Darwin, Northern Territory (Vibrio sp. DAT722, Vibrio sp. DAT1706, Vibrio sp. DAT865, Vibrio sp. DAT868, Vibrio sp. DAT447); strains labelled SH originate from Sydney Harbour (New South Wales) seawater samples (Vibrio sp. SH32 and Vibrio fischeri SH36). All vibrios were isolated by plating water samples directly on thiosulfate citrate bile sucrose (TCBS) agar plates. Pure cultures of the strains were grown in rich medium ( g tryptone, g NaCl, . g MgCl2·6H2O, . g KCl, make up to L with distilled H2O, adjust pH to . with M NaOH). Genomic DNA was extracted as described in Wilson []. A bp DNA fragment of the RNA polymerase B gene (rpoB) (from nucleotide position to in reference to the E. coli rpoB gene), was amplified by PCR from the genomic DNA of all vibrio strains using general vibrio primers (rpoB1315F '-GGCGAAGTGGACGATATCGACC-' and rpoB2442R '-GATCGAGTCTTCGAAGTTGTA-'). A phylogenetic analysis of these rpoB fragments along with their homologs from completely sequenced vibrio genomes was performed to identify relationships between the isolates.PCR amplification and sequencing of PCR productsPCR amplifications were carried out in a final volume of μl containing – ng of template DNA, . mM of each primer, and . μl of PCR Master Mix (PROMEGA). The reactions were performed with an initial denaturation at °C for min., cycles with a denaturation at °C for sec., primer annealing at °C for sec., and primer extension at °C for min. PCR products were gel purified with the MinElute kit (QIAGEN) and cloned in TopoTA (INVITROGEN). Two clones were sequenced from both directions for each PCR product using an ABI377 automated DNA sequencer and BigDye v3. chemistry.Screening strains for the presence of integronsGenomic DNA isolated from a pure culture of each strain was screened for the presence of an integron using one of two sets of primers: one set targeted part of intI (VintI_R '-CTGATATWMGWACMGTACAAGARC-') and most of the gene known to be upstream of it in strains from the vibrio core group such as V. parahaemolyticus, VC1310 (VC1310_F '-CTGAATGTCTTATTTGCCTTTGG-'); the second set targeted Vibrio gene cassettes -be sites (V59be_F '-CACACCTTARSNSRGCGTTA-' and V59be_R '-TCCCTCTTGARCNSYTTGTTA-'). Strains SH32 and SH36 were positive only for gene cassettes. Strains DAT722, DAT868, DAT865 were positive for both intI flanked by VC1310 and gene cassettes. Once the PCR products were confirmed by sequencing, these three strains were selected for genomic fosmid library construction.Construction and screening of fosmid librariesFosmid libraries were constructed from the genomic DNA of strains DAT722, DAT865 and DAT868 using the EPIFOS kit (EPICENTRE). Purified genomic DNA was run on a low-melt % agarose gel (AMRESCO) in a pulse-field gel electrophoresis apparatus (BIORAD) and the DNA of ~ kb in size was cut out, purified from the agarose, ligated to the fosmid vector, packaged in phage capsids and used to infect E. coli as described in the EPIFOS kit manual. colonies from each of the resulting libraries were picked from agar plates and used to inoculate -well blocks containing ml of LB broth with . μg/ml of chloramphenicol (to select for the presence of fosmids) in each well. These cultures were grown overnight and glycerol stocks were made by mixing μl of culture from each well with μl of % glycerol in a -well plate. 10ul was sampled from each well and pooled by row in eppendorf tubes ( μl per tube). The latter tubes were centrifuged to remove the LB broth and cells were resuspended in μl of pH . Tris-HCl buffer. μl of resuspended cells was added as template to PCR reaction tubes for screening. PCRs were required to screen each library ( reaction per wells row, rows per -well block, -well blocks per library). The PCR screening was done using primers targeting the intI gene and its upstream neighbor VC1310. For all positive rows of the blocks, μl was taken out of each well from that row in the corresponding glycerol stock to inoculate a new PCR screening reaction. Clones testing positive in this second round of screening were re-streaked on LB agar plates containing . μg/ml of chloramphenicol, which were used to inoculate a liquid culture for extraction of pure fosmid DNA (as described in the EPIFOS kit manual). From to positive clones were obtained for each clones library, with an average of or ~ % of positives.Sequencing reactions were performed on one intI / VC1310 positive fosmid for each of the DAT722, DAT865 and DAT868 libraries using primer VintI_F ('-CTTGTACGGTACGAATATCAGC-') to obtain the complete sequence of their intI genes (completing the partial intI sequence on the intI / VC1310 fragment). An intI-bearing fosmid from the library with the most positive clones for intI (the DAT722 library, positive clones / total clones) was then selected for whole shotgun sequencing.Assembly of the complete Vibrio sp. DAT722 integronThe intI-containing DNA insert from the DAT722 library was cleaved from its fosmid vector using EcoRI and separated from it by PFGE. The insert DNA recovered from the low-melt agarose gel (EPIFOS kit manual) was then sheared in kb fragments by nebullization and cloned in the TopoTA vector (INVITROGEN). Sub-clones were sequenced to obtain × coverage of the complete insert. Fragments were assembled using PHRED-PHRAP, which was also used to assess the quality of the sequencing. Low-quality regions were targeted by PCR giving a final average coverage of . (average for all the fosmid clones sequenced here). This yielded a kb contig that included the intI gene, kb of flanking DNA on one side and three gene cassettes integrated at attI on the other. To extend the array in the direction of the gene cassettes, a set of PCR primers targeting the last cassette of the contig was designed and used to re-screen the library. Both ends of library clones positive using this set of primers were sequenced to determine overlap with the original kb contig. A fosmid clone with minimal overlap (– kb) was selected for sequencing. This "walking" procedure was repeated until the end of the array was reached (i.e. gene cassettes were absent at one end of the last fosmid sequenced). This yielded a total of four fosmid clones of kb, kb, kb and kb in size, covering the entire kb DAT722 integron gene cassette array as well as kb of DNA on one side and kb on the other (total length of kb).The integron gene cassette array DNA sequence determined in this study has been submitted to the GenBank nucleotide sequence database and assigned accession number [GenBank:DQ139261].Detection and annotation of gene cassettes in DNA sequence data59-be sites were located in the DNA sequence using a Transact SQL procedure run in MSSQL . The procedure searched for the conserved regions identified in the -be sites of previously analyzed vibrio strains (': TAACAA, ': GCGTTA). The results were then filtered by ensuring that the distance between the ' conserved region and the ' conserved region fell in a length range of to bp, as previously observed for other vibrio -be sites. These possible -be sites were then ranked based on how well they conformed to the consensus motif found in other vibrio -be sites on a scale of to and potential -be sites with scores below (likely false positives) were discarded.Once the -be sites had been identified, ORFs were located using another T-SQL procedure and then filtered by size (>50aa) and proximity to existing -be sites. ORFs that intruded too far ( bp or more) into a -be site were disregarded or shortened. Overlapping ORFs were then examined manually and the most likely of them was retained (usually the ORF with a BLASTP match in the NCBI database). Once the -be sites, gene cassettes and ORFs had been found, they were automatically annotated in GenBank format using an ASP.net script. Cassettes in which no ORF was found according to the criteria defined above were used as a BLASTX query against the GenBank database to identify potential pseudogenes. Also, to confirm the "non-coding" nature of these cassettes with no apparent ORFs, all homologs of such cassettes found using BLASTN were aligned (>% of potential non-coding cassettes had homologs). If all homologous cassettes were conserved across their whole length and no mutation pattern reminiscent of the variability of synonymous codon positions could be observed, the cassette was confirmed as non-coding. The SQL script used to annotate gene cassette arrays can be obtained from the authors upon request.Comparison of Vibrio gene cassette arraysThe DNA sequences of all integron gene cassette arrays found in completely sequenced vibrio genomes (along with the ORFs they encode) were retrieved from the KEGG database []. A PERL script was written to perform reciprocal BLASTP and BLASTN of the gene cassette contents of two arrays against each other (to detect homologs) and of an array against itself (to detect paralogs). The script can be obtained from the authors upon request. A cut-off e-value of 1E- was established to determine homology.Sequence alignments and phylogenetic analysisThe -be sites present in all gene cassette arrays from sequenced vibrio genomes as well as the DAT722 array were identified and retrieved using the Transact SQL procedure described above. They were then aligned using CLUSTALW []. The alignment was subsequently edited manually to remove ambiguous characters, leaving unambiguous positions for phylogenetic analysis. The latter was performed in PAUP* .04b [] and applying the Neighbor-Joining tree reconstruction method using the GTR (general time-reversible) nucleotide substitution model. Support values represent the consensus of Neighbor-Joining trees constructed from bootstrap pseudo-replicates of the original dataset.Phylogenetic analyses were also carried at the DNA level for the rpoB gene. Third codon positions did not display mutational saturation and were therefore included in the analyses. The rpoB trees were constructed with PAUP* .04b, applying the heuristic-search option and using the TBR branch-swapping algorithm. Maximum likelihood was used as the tree reconstruction method, with the nucleotide substitution model (GTR), gamma rates parameter α, proportion of invariable sites and nucleotide frequencies determined using MODELTEST []. The confidence of each node was determined by building a consensus tree of maximum likelihood trees from bootstrap pseudo-replicates of the original dataset.Amino acid alignments of the Vibrio sp. DAT722 gene cassette encoded ORFs and their homologs as well as the IntI protein family were constructed using CLUSTALW and edited manually to remove ambiguous characters. Maximum likelihood phylogenetic analyses were applied to these datasets using PROML with the JTT amino acid substitution matrix, a rate heterogeneity model with gamma-distributed rates over four categories with the α parameter estimated using TREE-PUZZLE, global rearrangements and randomized input order of sequences ( jumbles). Bootstrap support values represent a consensus (obtained using CONSENSE) of Fitch-Margoliash distance trees (obtained using PUZZLEBOOT and FITCH) from pseudo-replicates (obtained using SEQBOOT) of the original alignment. The settings of PUZZLEBOOT were the same as those used for PROML, except that no global rearrangements and randomized input order of sequences are available in this program. PROML, CONSENSE, FITCH and SEQBOOT are from the PHYLIP package version .6a []. TREE-PUZZLE and PUZZLEBLOOT can be obtained from the programs website [].Authors' contributionsYB drafted the manuscript, conceived the study and performed the evolutionary analyses. CLN carried out the DNA sequencing and assembly of the fosmid clones and helped draft the manuscript. MJJ wrote the script for the annotation of gene cassette arrays. AR and BCM performed the functional annotation of gene cassettes and helped draft the manuscript. MRG participated in the conception and design of the approach, WFD took part in drafting the manuscript and helped coordinate the study. Finally, HWS participated in the experimental design and drafting of the manuscript.
|
PMC1977604.txt
|
TITLE: An in vivo assessment of adriamycin-loaded albumin microspheres. AUTHORS: J. A. Goldberg, N. Willmott, D. J. Kerr, C. Sutherland, C. S. McArdle ABSTRACT: ImagesFigure 1Figure BODY:
|
PMC1664620.txt
|
TITLE: Spinal Delivery of p38: TNF-alpha Inhibitors AUTHORS: Edward Tobinick ABSTRACT: No Abstract BODY: The new study by Boyle and colleagues provides important data on basic science mechanisms involved in pain and inflammation []. Their data, along with that from previous studies, provides further basic scientific evidence documenting p38–TNF-alpha interactions, and suggests that spinal p38 or spinal TNF-alpha blockade may have clinical relevance [,]. The present study documents that p38 activation may be occurring predominantly in microglia. The present study, therefore, joins other recent work which suggests the importance of p38-glial-TNF-alpha interactions in neuroinflammation and synaptic signaling [–]. This increasing evidence may have clinical relevance not only to arthritis pain, but also to the pathogenesis of various forms of neuropathic pain and Alzheimer disease [–].Because the present study suggests that spinal delivery may be more effective than systemic delivery when attempting to intervene in spinally-mediated inflammatory mechanisms, the authors note the potential importance of developing compounds that may bypass the blood-brain barrier. The present author speculates that the rapid and significant clinical effects noted following perispinal administration of etanercept in small pilot studies suggest that perispinal administration of p38 inhibitors may also allow these compounds to reach the spinal cord and dorsal root ganglia in therapeutically effective amounts [,]. It is hypothesized that this may be possible via passage through the vertebral portion of the cerebrospinal venous system, and this may explain the efficacy of perispinal etanercept in the above cited studies [–]. Previous studies have documented that epidural administration of large molecules may result in delivery to the endoneurial space []. This evidence, along with the basic scientific evidence provided by the present study of the potential clinical importance of spinal delivery, supports consideration of investigation of this novel route of administration.
|
PMC1239908.txt
|
TITLE: A method for finding single-nucleotide polymorphisms with allele frequencies in sequences of deep coverage AUTHORS: Jianmin Wang, Xiaoqiu Huang ABSTRACT: BackgroundThe allele frequencies of single-nucleotide polymorphisms (SNPs) are needed to select an optimal subset of common SNPs for use in association studies. Sequence-based methods for finding SNPs with allele frequencies may need to handle thousands of sequences from the same genome location (sequences of deep coverage).ResultsWe describe a computational method for finding common SNPs with allele frequencies in single-pass sequences of deep coverage. The method enhances a widely used program named PolyBayes in several aspects. We present results from our method and PolyBayes on eighteen data sets of human expressed sequence tags (ESTs) with deep coverage. The results indicate that our method used almost all single-pass sequences in computation of the allele frequencies of SNPs.ConclusionThe new method is able to handle single-pass sequences of deep coverage efficiently. Our work shows that it is possible to analyze sequences of deep coverage by using pairwise alignments of the sequences with the finished genome sequence, instead of multiple sequence alignments. BODY: BackgroundInformation concerning the allele frequencies of single-nucleotide polymorphisms (SNPs) is needed to select an optimal subset of common SNPs for use in association studies []. One approach to finding common SNPs with allele frequencies is to generate DNA sequences from a sufficient number of samples in a population. This approach requires that computational methods have an ability to handle thousands of sequences from the same genome location (sequences of deep coverage). In this paper, we describe a computational method for finding common SNPs with allele frequencies in sequences of deep coverage. We present results from the method on human expressed sequence tags (ESTs) of deep coverage, which are currently a major source of DNA sequences of deep coverage. The method is also expected to be useful for finding common mutations in sequences of deep coverage produced in a cancer genome project [].The PolyBayes program is widely used to find SNPs in redundant DNA sequences [,]. It first constructs a multiple sequence alignment based on pairwise alignments of each sequence with a high-quality genomic sequence called an anchor. Then it identifies and removes paralogous sequences that have a high number of observed differences with the anchor sequence. Next it computes an SNP probability score for each column of the multiple sequence alignment based on a rigorous Bayesian formula. The formula uses the prior probabilities of all the nucleotide permutations for the column, which are estimated from the quality scores of the bases on the column.We enhance the PolyBayes program in several aspects to handle single-pass sequences (query sequences) of deep coverage. First, all the paralogous regions of the finished human genome sequence are included as anchor sequences. Each query sequence is assigned to the corresponding anchor sequence that is different from each of the remaining anchor sequences at some positions but is identical to the query sequence at most of the positions. This approach separates paralogous sequences by making use of the positions where paralogous sequences differ but sequences from the same genome location agree.Second, pairwise alignments of corresponding query and anchor sequences are used to construct profiles, one per anchor sequence. At each position of an anchor sequence, its profile contains the numbers and types of high-quality query bases that are aligned to the position of the anchor sequence. Candidate SNPs are produced based on the profiles, instead of multiple sequence alignments for the following reason. As the number of single-pass sequences in a multiple sequence alignment increases, the number of gap columns in the alignment increases but the number of identity columns in the alignment does not increase. Thus, it is difficult to construct an accurate multiple sequence alignment for single-pass sequences of deep coverage.Third, because the pairwise alignment of corresponding query and anchor sequences may contain regions of low similarity due to sequencing errors or contaminants, the highly similar regions of the alignment are found by a dynamic programming algorithm. Only the highly similar regions are used in generation of the profile.Our computer program named PolyFreq was compared with PolyBayes on eighteen data sets of human EST sequences of deep coverage. Results from PolyFreq and PolyBayes indicate that PolyFreq ran to completion and used almost all input sequences in computation of the allele frequencies of SNPs for every data set.ResultsThe method for finding SNPs with allele frequencies was implemented as a computer program. The source code of the program is freely available for academic use [, see Additional file ]. The program takes as input a set of high-quality anchor sequences and a set of query sequences with quality scores. The set of anchor sequences includes all the paralogous regions of the genome for the set of query sequences. The anchor and query sequences are from the same species.The program reports candidate SNPs in the anchor sequences. For each candidate SNP, the program reports its position in the anchor sequence, its local context in the anchor sequence, and base types with a frequency greater than a cutoff. The frequency of a base type is also given in a rational form with the number of query bases of the type as the enumerator and the total number of query bases as the denominator.To evaluate PolyFreq, eighteen data sets of human EST sequences of deep coverage were constructed as follows. Eighteen clusters of human EST sequences, each containing at least , EST sequences with trace data, were randomly selected from the April, release of the UniGene database []. The eighteen UniGene clusters also contain EST sequences without trace data. For each of the eighteen UniGene clusters, an EST data set was obtained by selecting all EST sequences with trace data from the cluster. The set of quality score sequences for each of the eighteen data sets was produced from the trace data with Phred []. The quality score q of a base is obtained by the formula q = - log p, where p is the estimated error probability of the base []. For example, a quality score of corresponds to an error probability of .. The EST sequences in each of the eighteen sets were produced from to cDNA libraries derived from diverse human tissues and cell lines []. Each of the eighteen data sets of full-length EST sequences without any masking was used as a query set.For each query set, its set of anchor sequences was obtained by comparing the query sequences with the finished genome sequence at the BLAT web server []. By using stringent settings for BLAT, a set of two human anchor sequences was produced for each of three query data sets, and a set of one human anchor sequence was produced for each of the remaining query data sets. Each set of anchor sequences was screened for repeats with RepeatMasker [].The PolyFreq program was run on each pair of query and anchor sets. The PolyFreq program ran successfully to completion for each of the eighteen data sets. The following values were used for the parameters of the program: , minimum depth of coverage for each candidate SNP; .%, minimum minor allele frequency; bp, minimum perfect block length; , minimum base quality score in the perfect block; %, minimum percent identity for query-anchor alignments; %, minimum percent identity for the highly similar regions of query-anchor alignments.Although PolyBayes was not designed to deal with data sets of deep coverage, we tested PolyBayes on the eighteen data sets of deep coverage to see how PolyBayes would behave on the data sets. Because PolyBayes takes only one anchor sequence, the corresponding anchor sequence was selected and given to PolyBayes for each data set. On eight of the eighteen data sets, PolyBayes ran successfully to completion. On the remaining data sets, PolyBayes terminated abnormally without producing any output file after running for a few hours. The default values for all the parameters but the SNP probability output cutoff of PolyBayes were used; PolyBayes terminated abnormally more frequently under other parameter values. A value of . for the SNP probability output cutoff was used to produce a lower number of false positives than the default value of ..The abnormal termination of PolyBayes might be related to the deep coverage of the data set and full-length EST sequences with low-quality ends or contaminants. Thus, for each set of full-length EST sequences, a set of trimmed EST sequences was produced by removing the ends of every sequence that are not highly similar to the corresponding anchor sequence. For each data set, the number of trimmed sequences was close to the number of full-length sequences. The PolyBayes program was also run on each set of trimmed sequences. It ran to completion for thirteen out of the eighteen data sets.All the tests were performed on a Dell workstation with two .-Ghz processors and Gb of main memory. PolyFreq took less than one hour on every data set, whereas PolyBayes was two to ten times slower than PolyFreq on every data set. The memory requirements of PolyFreq and PolyBayes on the data sets were similar and were about to times the input size.The PolyFreq and PolyBayes programs were compared on every data set for which PolyBayes ran successfully to completion on either the set of full-length sequences or the set of trimmed sequences. For each data set, results produced by PolyFreq on the set of full-length sequences were included in the comparison, whereas results produced by PolyBayes on both sets of full-length and trimmed sequences were included. The SNPs from the dbSNP database [] that were mapped by the following method to the anchor sequences were used as true SNPs for the comparison. Each SNP from dbSNP is specified by a local sequence context. For each data set of EST sequences with a RefSeq sequence [], each SNP from dbSNP that occurs in the RefSeq sequence was determined by finding the exact occurrence of its sequence context in the RefSeq sequence. Each SNP from dbSNP in the RefSeq sequence was mapped to a corresponding anchor sequence by using a spliced alignment of the RefSeq and anchor sequences. Because four data sets had no RefSeq sequence, no SNPs from dbSNP were mapped to the anchor sequences for the data sets.For each program on every data set with a RefSeq sequence, the number of true positives, the number of false positives, and the number of false negatives were computed. The number of true negatives was not collected because of its large value. Also reported were the number of sequences in the data set and the number of sequences that were used by the program to compute candidate SNPs. The comparison results are shown in Table .Table 1Results by PolyFreq and PolyBayes on eighteen data sets of EST sequencesData setSizePolyBayes (trimmed)PolyFreq (full length)PolyBayes (full length)TPFPFNNSUTPFPFNNSUTPFPFNNSUHs.11958944031217050149152457439112152501531Hs.1296731665748111457511131662T/AHs.148340160333691563498158336791560Hs.17062215141371242931810150727311365Hs.1785511685562121632614111676T/AHs.180909101735069833206101231646996Hs.1871992041T/AN/A1997T/AHs.198281315691102230771554163149T/AHs.35092710175421197692171015613910995Hs.356331144125593181141014362859239Hs.35657228220462253401722821T/AHs.4395527163T/AN/A6873T/AHs.4444674033N/A805N/A4028N/A679Hs.4466281490432121338512111486T/AHs.5206404120T/A952274099T/AHs.5224638294T/AN/A8280T/AHs.5243904462T/A948244454T/AHs.54457775371460431716101747751718175391556The mark N/A means that no SNP from dbSNP was mapped to the anchor sequence because of lack of a RefSeq sequence. The mark T/A means that PolyBayes terminated abnormally without producing any output file. A candidate SNP from the program is considered as true positive (TP) if it is in dbSNP or false positive (FP) otherwise. A SNP from dbSNP that occurs in the data set is considered as false negative (FN) if it is not reported as a candidate SNP from the program. The number of sequences used (NSU) by the program in generation of candidate SNPs is reported.The results in Table indicate that PolyFreq could handle the data sets of full-length reads with problem regions and with very deep coverage. PolyFreq used , to , sequences on the five data sets for which PolyBayes terminated abnormally. On the data sets for which PolyBayes ran to completion, PolyFreq was similar to PolyBayes in the number of true positives and the number of false negatives, and is better than PolyBayes in the number of false positives. PolyBayes used significantly fewer sequences than PolyFreq on some of the data sets. Note that the ability to use as many sequences as possible is necessary for accurate computation of the allele frequencies of SNPs.DiscussionWe originally developed a method for assembling sequences of deep coverage. The method constructs multiple sequence alignments of large width for contigs. The method has to deal with a large number of gap columns in the multiple sequence alignment. We later agreed with one of the reviewers that it is not necessary to construct multiple sequence alignments for analysis of sequences of deep coverage. The reviewer also suggested that we focus on SNP analysis in sequences of deep coverage. Those suggestions motivated us to develop the method reported in this paper.The PolyFreq program keeps PolyBayes' feature of performing comparisons between query and anchor sequences, instead of performing comparisons among query sequences. In addition, PolyFreq constructs profiles by using the highly similar regions of pairwise alignments of corresponding query and anchor sequences, instead of multiple alignments of query and anchor sequences. Thus, the efficiency and accuracy of PolyFreq are not significantly affected by query sequences of deep coverage. On the contrary, PolyFreq can compute the allele frequencies of SNPs more accurately in query sequences of deep coverage.As sequencing costs are significantly reduced in the future, single-pass sequences from hundreds to thousands of individuals will be produced. Those sequences will be of deep coverage. Our current work suggests that it is possible to analyze sequences of deep coverage by using pairwise alignments of the sequences with the finished genome sequence, instead of multiple sequence alignments.MethodsWe first present the major steps of our method for finding common SNPs with allele frequencies in a set of query sequences and a set of anchor sequences. Then we describe each step in detail. The method consists of the following major steps:. Compute an alignment of anchor sequences for each pair of anchor sequences.. Compute an alignment of query and anchor sequences for each pair of similar query and anchor sequences.. For each query sequence, find the corresponding anchor sequence that is different from each of the remaining anchor sequences at some positions but is identical to the query sequence at most of the positions.. Find the highly similar regions of their alignment for each pair of corresponding query and anchor sequences.. For each anchor sequence, use the highly similar regions of every alignment involving the anchor sequence to construct a profile for the anchor sequence. At each position of the anchor sequence, its profile contains the numbers and types of high-quality query bases that are aligned to the position of the anchor sequence.. Report each candidate SNP with major and minor allele frequencies if its minor allele frequency is greater than a cutoff.In step , for each pair of anchor sequences, an alignment of the sequences in given orientation as well as an alignment of the sequences in opposite orientation is constructed with GAP3, a global alignment program specially designed for genomic sequences with long different regions between similar regions []. One of the two alignments with a larger score is saved for the pair of sequences. The alignments saved in this step are to be used in step for finding the corresponding anchor sequence for each query sequence.In step , pairs of similar query and anchor sequences are found with DDS2, which produces a high-scoring chain of segment pairs (ungapped alignment fragments) between the two sequences in the pair []. For each pair of similar query and anchor sequences, an alignment of the sequences in the pair is constructed with GAP22, an improved version of the GAP2 program [] for quickly computing an alignment in a small area of the dynamic programming matrix, which is determined based on the chain of segment pairs. If the percent identity of the alignment is greater than a cutoff, then the alignment is saved for the pair of sequences.In step , for each query sequence that is highly similar to two or more anchor sequences, the corresponding anchor sequence for the query sequence is selected among the anchor sequences through pairwise comparisons. Initially, one anchor sequence is taken as the current leader. Then the rest are compared with the current leader one at a time. Consider the comparison between the current leader and the current challenger. The winner between the two anchor sequences is produced by using the alignment of the two anchor sequences and their alignments with the query sequence. A common match occurs at a position of the query sequence, a position of the current leader, and a position of the current challenger if the three positions are pairwise aligned on each of the three alignments and contain the same base. The winner between the two anchor sequences is the one with a larger number of uncommon matches in its alignment with the query sequence. The winner becomes the current leader. After all the pairwise comparisons, the final leader is the corresponding anchor sequence for the query sequence.In step , for each pair of corresponding query and anchor sequences, the highly similar regions of the alignment of the two sequences are identified in linear time with LCP, a program for finding regions of a sequence that meet a content requirement []. Each of the highly similar regions found by LCP has a percent identity greater than or equal to a cutoff p and is strictly optimal. The score of a region of the alignment is the sum of scores of every base match and every base difference in the region, where the score of every base match is - p and the score of every base difference is - p. A region is optimal if its score is not less than the score of any other region that overlaps with it. An optimal region is strictly optimal if it is not completely contained in any optimal region other than itself.In step , only substitutions in the highly similar regions of every alignment of corresponding query and anchor sequences are used to construct a profile for the anchor sequence because the remaining regions of the alignment have a high rate of difference, which is likely due to sequencing errors or contaminants in the query sequence. Additional requirements are introduced below because a long highly similar region may still contain a packet of sequencing errors in the middle. A sufficiently long section in a highly similar region of an alignment is a perfect block if the section consists only of exact base matches and the quality score of each query base in the section is greater than or equal to a cutoff []. A substitution in a highly similar region of an alignment is acceptable if it is immediately flanked on each side by a perfect block. Acceptable and unacceptable substitutions are illustrated in Figure 1A.Figure 1Acceptable and unacceptable substitutions in a pairwise alignment and a candidate SNP from a real data set. (A) The line shows an alignment of query and anchor sequences with thick parts indicating highly similar regions. The large rectangular box gives a detailed view of the small box. On the left is an unacceptable substitution that is flanked by a block of low-quality bases, and on the right is an acceptable substitution that is flanked by a perfect block on each side. The quality value of each query base in the large box is shown next to the base. (B) Shown is a candidate SNP with allele frequencies from PolyFreq on a real data set (Hs. ) in Table .For each anchor sequence, its profile contains four counts at each position: one count for each query base type. For example, the base A count at the anchor position is the number of acceptable substitutions at the anchor position and at a query sequence position with base A, in a highly similar region of an alignment of the anchor sequence with the query sequence. A count for a query base type at the anchor position is if there is no acceptable substitution at the anchor position and at any query sequence position with the query base type. The frequency of each of the four counts is the count divided by the sum of the four counts if the sum is positive.In step , each profile is scanned for candidate SNPs. A candidate SNP occurs at an anchor sequence position if the sum of the four counts for the position is greater than or equal to a cutoff and at least two of the four counts have a frequency greater than a cutoff. All candidate SNPs with allele frequencies are reported along with a local anchor sequence region for each candidate SNP. A candidate SNP with allele frequencies from one of the examples in Table is shown in Figure 1B.Author's contributionsXH designed the strategy for solving the problem and provided guidance to JW. JW worked out the details of the strategy, developed the program, and produced results on data sets with the program. XH wrote the paper and JW formatted it in Word. All authors read and approved the final manuscript.Supplementary MaterialAdditional File 1A file named PolyFreq.tar is included. The file contains the source code of all programs in the package. The file is unpacked by using the Unix command "tar xvf PolyFreq.tar" on a Unix or Linux computer.Click here for file
|
PMC2572155.txt
|
TITLE: Comparable contributions of structural-functional constraints and expression level to the rate of protein sequence evolution AUTHORS: Maxim Y Wolf, Yuri I Wolf, Eugene V Koonin ABSTRACT: BackgroundProteins show a broad range of evolutionary rates. Understanding the factors that are responsible for the characteristic rate of evolution of a given protein arguably is one of the major goals of evolutionary biology. A long-standing general assumption used to be that the evolution rate is, primarily, determined by the specific functional constraints that affect the given protein. These constrains were traditionally thought to depend both on the specific features of the protein's structure and its biological role. The advent of systems biology brought about new types of data, such as expression level and protein-protein interactions, and unexpectedly, a variety of correlations between protein evolution rate and these variables have been observed. The strongest connections by far were repeatedly seen between protein sequence evolution rate and the expression level of the respective gene. It has been hypothesized that this link is due to the selection for the robustness of the protein structure to mistranslation-induced misfolding that is particularly important for highly expressed proteins and is the dominant determinant of the sequence evolution rate.ResultsThis work is an attempt to assess the relative contributions of protein domain structure and function, on the one hand, and expression level on the other hand, to the rate of sequence evolution. To this end, we performed a genome-wide analysis of the effect of the fusion of a pair of domains in multidomain proteins on the difference in the domain-specific evolutionary rates. The mistranslation-induced misfolding hypothesis would predict that, within multidomain proteins, fused domains, on average, should evolve at substantially closer rates than the same domains in different proteins because, within a mutlidomain protein, all domains are translated at the same rate. We performed a comprehensive comparison of the evolutionary rates of mammalian and plant protein domains that are either joined in multidomain proteins or contained in distinct proteins. Substantial homogenization of evolutionary rates in multidomain proteins was, indeed, observed in both animals and plants, although highly significant differences between domain-specific rates remained. The contributions of the translation rate, as determined by the effect of the fusion of a pair of domains within a multidomain protein, and intrinsic, domain-specific structural-functional constraints appear to be comparable in magnitude.ConclusionFusion of domains in a multidomain protein results in substantial homogenization of the domain-specific evolutionary rates but significant differences between domain-specific evolution rates remain. Thus, the rate of translation and intrinsic structural-functional constraints both exert sizable and comparable effects on sequence evolution.ReviewersThis article was reviewed by Sergei Maslov, Dennis Vitkup, Claus Wilke (nominated by Orly Alter), and Allan Drummond (nominated by Joel Bader). For the full reviews, please go to the Reviewers' Reports section. BODY: BackgroundThe first grand generalization of molecular evolution is that proteins evolve at widely different rates but each particular protein has a characteristic rate that remains relatively constant over long evolutionary spans []. In other words, there seems to exist a molecular clock that ticks at widely different paces for different protein-coding genes. What determines this characteristic rate is, arguably, one of the central questions of evolutionary biology. To our knowledge, the first explicit hypothesis on the interplay of factors that determine the rate of a protein's evolution belongs to Wilson et al. who proposed, in their classic review on molecular evolution, that the sequence evolution rate should be a function of, firstly, the intrinsic functional constraints that affect the given protein and, secondly, on the biological function of the protein in the organism: Ri = f(Pi)f(Qi) where Ri is the sequence evolution rate, f(Pi) is the functional-constraint factor, and f(Qj) is the dispensability factor []. Testing this hypothesis at the time was hardly feasible, so given that the functions and structures of proteins are, indeed, widely different and so are the rates of sequence evolution, it was (more or less) tacitly assumed that the first term in Wilson's equation was the decisive one.Things changed with the advent of functional genomics and systems biology in the beginning of the 21st century when it became possible to measure the correlations between many "genomic" variables [-]. Quite surprisingly, it turned out that there was little if any correlation between the essentiality of genes for the reproduction of organisms and their rates of evolution: at best, non-essential genes evolve slightly faster than essential genes [-]. Equally unexpectedly, highly significant negative correlation has been shown to exist between a gene's expression level and its evolutionary rate, that is, highly expressed genes evolve significantly slower than lowly expressed ones [,-]. Many other correlations between genomic variables have been examined including but not limited to the number of protein-protein and genetic interactions, position in various kinds of networks, and the propensity to be lost during evolution [,-]. Generally, the evolutionary variables, namely, the sequence evolution rate and the propensity for gene loss, are positively correlated with each other and negatively correlated with "phenomic" variables such as expression level, number of interactions and others [,].Most of the observed correlations between genomic variables are relatively weak. The link between expression level and sequence evolution rate appears to be by far the strongest and most consistent across a range of diverse organisms, a finding that led to the striking hypothesis that expression level or, more precisely, the rate of translational events is indeed the dominant determinant of the sequence evolution rate [,,]. It has been further hypothesized that the underlying cause of the covariation between the sequence evolution rate and expression level is the selection for robustness to protein misfolding that is increasingly important for highly expressed proteins owing to the toxic effects of misfolded proteins [,]. Misfolded forms of highly expressed proteins are thought to be strongly deleterious for a cell, so there could be a strong selection to avoid their accumulation. Recent detailed computer simulations of protein evolution suggest that the toxic effect of protein misfolding, indeed, could suffice to explain the observed covariation of expression level and sequence evolution rate [].The hypothesis on the crucial role of the selection for robustness to misfolding, dependent on the translation rate, in protein sequence evolution is a drastic departure from the more traditional thinking that links the evolution rate, primarily, to intrinsic structural-functional constraints that are thought to substantially differ for different proteins [,]. We were interested in directly assessing the relative contributions of effects mediated by the translation rate and intrinsic structural-functional constraints to the evolution rates of protein sequences. The results of the analysis described here indicate that both contributions are substantial and comparable in magnitude.ResultsRationale and ApproachThis analysis relies on the high abundance and diversity of multidomain proteins in eukaryotes [,]. The simple underlying idea is that different domains of the same protein are translated at the exact same rate. Accordingly, under the mistranslation-induced misfolding (hereinafter MIM) hypothesis, distinct domains within the same multidomain protein, on average, would be expected to evolve at substantially closer rates than the same domains when contained in different proteins. In the extreme, that is, under the obviously over-simplifying assumption that the intrinsic structural-functional constraints that affect different domains are the same, the constituent domains of any multidomain protein would evolve at the same rate (within sampling error). Conversely, the remaining difference in the evolution rates of domains in multidomain proteins is attributable to the differences in the intrinsic constraints. So, at least, in principle, by measuring the extent of "homogenization" of evolutionary rates of domains in multidomain proteins, it should be possible to disentangle and compare the contributions of translation rate-mediated factors and translation-independent, intrinsic ones.The approach is schematically illustrated in Figure . In addition to the potential for directly assessing the contributions of translation rate and domain identity (structural-functional constraints), this approach has the major advantage of not relying on experimental gene expression data that are, inevitably, noisy. Furthermore, the interpretation of experimental data on gene expression is ambiguous because the quantity that is actually measured in most experiments is the transcript level rather than the number of translation events per se, which is the decisive factor under the MIM hypothesis [,].Figure 1Scheme of the comparison of evolution rates of domains located in multidomain proteins and in separate proteins. rix – evolutionary rates of all instantiations of domain X in human (Arabidopsis) proteins; rjy – evolutionary rates of all instantiations of domain Y in human (Arabidopsis) proteins; rax – evolutionary rate of domain X in multidomain protein A; ray – evolutionary rate of domain Y in multidomain protein A; r'ax – evolutionary rate of the randomized sequence segment within the boundaries of domain X in multidomain protein A; r'ay – evolutionary rate of the randomized sequence segment within the boundaries of domain Y in multidomain protein A.The only substantial caveat we are aware of is the possibility that, because of alternative splicing, different domains of some multidomain proteins are actually expressed at different rates. However, it seems unlikely that alternative splicing would have a major effect on the results considering that there is limited correspondence at best between alternative splice form structure and domain architectures of proteins []. Furthermore, as described below, we performed the comparisons of fused and separated domains in both mammals where alternative splicing is most abundant and in plants where it is thought to be much less common [].Comparison of the evolutionary rates of fused and separated protein domainsIn order to implement the scheme shown in Figure , we mapped structurally distinct domains from the SCOP/ASTRAL database onto alignments of orthologous proteins from human and mouse, and Arabidopsis and cottonwood. Evolutionary rates were calculated for the complete sets of domains, domains fused within multidomain proteins, and domains contained in different proteins (see Materials and Methods for details). All the analyses described below were performed on protein superfamilies, the intermediate level of the structural classification of proteins implemented in the SCOP database []; however, very similar results were obtained with the higher (fold) and lower (family) levels of proteins classification (data not shown).Examination of the distributions of evolutionary rates among different domains is interesting in itself: mean evolutionary rates of abundant domains show a span of almost orders of magnitude (Figure ), in a qualitative agreement with previous observations made by comparison within protein families [] or functional categories of genes []. The distributions typically are broad but follow a general bell-shaped form (some are bimodal), in support of the notion of an intrinsic rate of domain evolution for which the mean (or median) of the respective distribution can be taken as a reasonable proxy. Domains present within multidomain proteins do not show systematic differences in evolutionary rates compared to solo domains as illustrated by two examples in Figure .Figure 2Evolution rate distributions of highly abundant domains. The rates were estimated from comparisons of human-mouse orthologous protein sequences. X-axis: log10 of domain evolution rate, Y-axis: probability density function..Figure 3Evolutionary rate distributions for two superfamilies of highly abundant domains: comparison of fused and solo domains. A – P-loop-containing NTP hydrolases; B – PH-like domains. The rates were estimated from comparisons of human-mouse orthologous protein sequences. X-axis: log10 of domain evolution rate, Y-axis: probability density function.We then addressed the issue of homogenization of the rates of evolution of domains that is predicted by the MIM hypothesis to result from the fusion of domains within a multidomain protein. Figure shows anecdotal evidence for two proteins, each consisting of distinct domains. For one of these proteins, homogenization is obvious (Figure 4A) whereas the other one shows no obvious sign of homogenization (Figure 4B). These examples are characteristic of the diversity of the evolutionary regimes of domains, so that homogenization is seen in many but by no means all multidomain proteins, and some actually display the opposite trend (Additional Files and , and see below). This striking variability notwithstanding, the results of the analysis of the complete sets of domains unequivocally reveal substantial homogenization as illustrated by the comparison of the probability density functions for the difference (ratio) of the evolutionary rates for all domain combinations and for domain pairs fused within multidomain proteins. The difference in evolutionary rates between a pair of domains within a multidomain protein tends to be substantially less than the difference between rates for the same pair of domains found in different proteins in both human (Figure 5A) and Arabidopsis (Figure 5B).Figure 4Examples of domain evolutionary rates in multidomain proteins. A – complement component precursor (C2; NP_000054). B – protein regulating synaptic membrane exocytosis (RIMS2; NP_001093587). The curves indicate the rate distributions for the constituent domains of multidomain proteins (as in Figure dots indicate the rates for the corresponding domains (color-coded) in the given protein.Figure 5Differences of evolution rates for domains contained in the same multidomain protein and in different proteins, compared to randomized "domain" sequences from multidomain proteins. X-axis: log10 of the ratio of evolution rates of a pair of domains. Y-axis: probability density function. Red, domain pairs within multidomain proteins; black, all possible pairwise combinations of the same domains (from both multidomain proteins and proteins in which the respective domains occur separately); green, randomized domains within multidomain proteins (control for sampling error). The randomized domains were obtained by randomly shuffling the columns in a multidomain protein alignment, and evolution rates were calculated for regions within the original domain boundaries. Va, Vm, and Vr are the normalized variances of the distributions for all domain pairs, domain pairs within multidomain proteins, and randomized domains, respectively. A – Human proteins: Va/Vm/Vr = ././.. B – Arabidopsis proteins: Va/Vm/Vr = ././..However, homogenization is far from being complete as shown by comparing the distributions of rate differences between domains in multidomain proteins to control distributions obtained with sequence segments within the original domain boundaries in randomized alignments of multidomain proteins. The variance of the distribution of the rate differences between the domains of multidomain proteins was ~. times greater than the variance for the control distribution for human proteins (Figure 5A) and ~. times greater in the case of Arabidopsis (Figure 5B). Taken together, these findings indicate that fusion of domains in multidomain proteins leads to substantial homogenization of their evolutionary rates although highly significant rate differences between the domains remain.We then examined the correlations between the mean rate differences of domain pairs that are conjoined in multidomain proteins and the same domain pairs found in different proteins. If fusion within a multidomain proteins, on average, has no effect on the evolutionary rates of the domains involved, the geometric mean of the ratio of the rates for the given pair of domains should be equal to that for all combinations of these domains (up to the sampling error), so the slope of the regression line is expected to be equal to . Conversely, if the evolution rates of the constituent domains in multidomain proteins are completely homogenized, all rate differences for domain pairs in multidomain proteins should be close to (again, subject to a sampling error), and the slope of the regression line would be equal to as well. The results show that both in human (Figure 6A) and in Arabidopsis (Figure 6C), there is a limited but statistically highly significant, positive correlation between the rate differences of domains in the two classes of domain pairs. This correlation was in a sharp contrast with the results of similar comparisons that were performed with sequences of multidomain proteins that were randomized over the entire lengths and the rates of evolution were then compared for regions within the original domain boundaries (compare Figure 6A with 6B, and Figure 6C with 6D). The slope of the linear trendline in the log-log scale was ~. for the domain pairs in human genome and ~. for the domain pairs in Arabidopsis genome. For instance, if the mean evolution rates of two domains in human proteins differ by a factor of , then an ~. fold difference in rates can be expected when these domains are fused within a single multidomain protein; similarly, in the case of Arabidopsis, an ~. fold difference can be expected. These results indicate that the contributions of translation-rate related factors and intrinsic structural-functional constraints to the rate of protein sequence evolution are comparable.Figure 6Correlations between the evolutionary rate differences of domains within multidomain proteins and the same domain pairs from separate proteins. Each data point corresponds to a pair of SCOP superfamilies (S1, S2) that is observed at least once in the same multidomain protein in a genome. X-axis: mean of log10 of the ratios of evolution rates of all combination of domains (Di1, Dj2) where Di1 ∈ S1 and Dj2 ∈ S2. Y-axis: mean of log10 of the ratios of evolution rates of all combination of domains (Di1, Dj2) where Di1 ∈ S1, Dj2 ∈ S2 and (Di1, Dj2) belong to the same multidomain protein. A – Human proteins. B – Randomized human protein sequences. C – Arabidopsis proteins. D – Randomized Arabidopsis protein sequences.DiscussionThe comparative analysis of the differences between domain evolution rates within multidomain proteins and in separate proteins reveals comparable contributions of factors that we consider to be translation rate-related (and hence uniformly affecting all domains of a multidomain protein) and intrinsic factors that differ between domains, regardless of whether or not they belong to the same multidomain protein. Continuing the line of thought under which selection for tolerance to amino acid misincorporation (caused by mistranslation, transcription error or mutation), or misfolding robustness, is the dominant factor of evolution [,], it seems possible to interpret these results in terms of a generalized MIM hypothesis. Thus, the rate of protein sequence evolution would be considered to depend on two factors:(i) Intrinsic misfolding robustness that depends on the structure of the given domain, and in particular, on its characteristic stability and designability []. It has been shown that proteins (or domains within multidomain proteins) with a greater amino acid residue contact density (more designable ones.) evolve faster than less designable proteins (domains) although the effect was relatively small and differed greatly among organisms [,]. Misfolding of a protein molecule is assumed to incur a specific fitness cost which might be unrelated to the protein's function, such as through a cytotoxic effect of the misfolded protein [].(ii) Translation rate that serves as an amplifier of the fitness cost of misfolding (roughly, the total cost is at least proportional to the number of translation events, considering that the error rate of translation is orders of magnitude greater than those of replication and transcription []) and, accordingly, of the strength of selection for the robustness to amino acid misincorporation.The analysis of evolutionary rates of individual domain in multidomain proteins allows one to disentangle the two factors by removing the effect of translation rate (the amplifier). The results show that the remaining difference in evolutionary rates of domains is much less than the total difference, that is, the amplifying effect of translation is substantial.The MIM hypothesis is a highly attractive concept not only because it introduces a single, dominant determinant of protein evolutionary rate but also because the key role of misfolding robustness is compatible with fundamental biological features of all cells. Indeed, all cells encode numerous chaperones that prevent misfolding and enormously elaborate molecular machines, such as proteasomes, that to a large extent are dedicated to the selective degradation of misfolded proteins [].Nevertheless, for a more complete explanation of the observed variation of protein (domain) evolution rates, it might be desirable to take into account other extrinsic variables, in addition to the expression level, such as the number of physical and genetic interactions, functional importance (fitness cost of knockout), and perhaps, others. An example of such a pluralistic explanatory framework is the concept of a gene's "status" according to which different extrinsic variables independently but concordantly contribute to sequence evolution [].We showed here that the contributions of intrinsic and extrinsic factors to the rate of protein sequence evolution are of comparable magnitude. Definitive, concrete interpretation of these terms and decomposition of each of them into more specific contributions are crucial goals for further experimental, computational and theoretical studies that ultimately might allow us to reach an adequate understanding of protein evolution.Materials and methodsA non-redundant set of human protein sequences (one per locus) and a set of Arabidopsis protein sequences (both from NCBI RefSeq release ) were used as queries in a BLASTP search [] with the e-value threshold of . against a database consisting of ASTRAL . domain sequences []. A non-overlapping set of best ASTRAL hits was used to map SCOP . [] domains to the human and Arabidopsis proteins. "Double-solo" domains, that is, domains belonging to a single-domain protein that was the only representative of the respective SCOP family, were discarded. This procedure yielded domains from different SCOP families that mapped to human proteins and domains belonging to different SCOP families that mapped to Arabidopsis proteins.In parallel, a set of orthologs for human and Arabidopsis proteins was obtained using the bi-directional best hit scheme [] against a set of mouse (NCBI RefSeq release ) and cottonwood (DOE Joint Genome Institute Populus genome release .) proteins, respectively. This procedure yielded , human-mouse and , Arabidopsis-poplar orthologous protein pairs. Pairwise alignments of the orthologous protein sequences were generated using the MUSCLE program [].Segments of human-mouse (Arabidopsis-cottonwood) ortholog alignments that corresponded to the SCOP domains were extracted, and the evolutionary distances between human and mouse (Arabidopsis and cottonwood) domains were calculated using the PROTDIST program of the PHYLIP package (JTT evolutionary model; gamma-distributed site rates with shape parameter .) []. To control for sampling error, full-length alignments were randomized by permuting the alignment columns after which the domains were extracted using the same coordinates as in the original alignment, and the distances were calculated as indicated above.Smoothing of the individual data points to produce the distribution curves shown in Figures , , , was performed using the Gaussian-kernel method [].Competing interestsThe authors declare that they have no competing interests.Authors' contributionsMYW and YIW performed the data collection and analysis. YIW and EVK designed the analysis and interpreted the results. EVK conceived of the study and wrote the manuscript. All authors read and approved the final version.Reviewers' commentsReviewer's report : Sergei Maslov, Brookhaven National LaboratoryThe manuscript reports a simple yet conclusive test of whether the variability of rates of protein evolution can be explained exclusively by the difference in their expression levels. The unequivocal answer is: "No".This conclusion comes from comparison of the evolutionary rates of different domains within the same multi-domain protein, which obviously have identical expression levels. Authors find that while the intraprotein domain-to-domain variability of evolutionary rates (black curve in Figure ) is dramatically reduced compared to that of the same domains in different proteins (red curve in Figure ), it still remains considerably broader than in the null-model (green curve in Figure ). In this null-model authors simulate the variability due to the sampling error by first reshuffling each of the alignments over the whole aligned region encompassing both domains and then recalculating the variability of the resulting homogenized domains. The fact that the null-model variability is smaller than that of the actual alignment indicates that some systematic difference in evolutionary rates of different domains still remains even when their expression levels are identical. This residual variability is tentatively attributed to domain-specific structural-functional constraints.The null-model itself is very important and (in my view) it is not adequately explained in the text of the manuscript. Indeed, the fact that the domain-to-domain variability of evolutionary rates is considerably broader than in the null-model is one of the central observations of the manuscript. Even though the Methods section contains a more detailed description of the null-model, I feel that authors should outline it in a few sentences right where it is first mentioned in the manuscript. This description should make it clear that the alignments are reshuffled over the entire aligned region encompassing all of the domains. Such reshuffling effectively homogenizes the evolutionary rates of individual domains.Authors' response: such a description was added and, hopefully, clarifies the presentation of the results.My main comment on the manuscript is regarding the claim made in the title, the abstract, and many times in the main text. Authors claim that the relative contributions of the expression level and the domain-specific structural-functional constraints to variability of evolutionary rates are *closely comparable*. If closely comparable means that, when considered separately, they reduce the variability by approximately equal factors, then it is not at all obvious from the manuscript. From Figure one gets the impression that the reduction in variability when the expression level differences are taken into account is much stronger than the residual variability due to structural-functional constraints. Could authors perhaps make a more quantitative comparison of relative contributions of these two factors?Authors' response: Actually, the comparison of the contributions of translation rate and structural functional constraints is quantitative as illustrated in Figure . To quote: "For instance, if the mean evolution rates of two domains in human proteins differ by a factor of , then an ~. fold difference in rates can be expected when these domains are fused within a single multidomain protein; similarly, in the case of Arabidopsis, an ~. fold difference can be expected." We believe that this comparison justifies the conclusion that these contributions are indeed comparable. However, we removed the claim that they are "closely" comparable. Furthermore, in the revised manuscript, we provide a quantitative comparison in Figure 5as well, in terms of the normalized variance of the distributions of evolutionary rate ratios. The results obtained with these two approaches are compatible, at least, qualitatively.My second comment or rather a question to the authors: does taking the expression level into account explains all of the variability of *silent* substitution rates? That is what one expects if all of the residual variability visible in Figure is due to structural-functional constraints.Authors' response: This certainly deserves to be tested. However, the prediction is not as clear-cut as it might immediately seem because the selection pressure that affects silent positions (that is, codon usage) could actually depend on the intrinsic tolerance of domain to amino acid misincorporation. So for the moment we decided that this analysis was not an integral part of the study.Reviewer's report :Dennis Vitkup, Columbia UniversityThe paper by Maxim Wolf et al. investigates one of the main puzzles of molecular evolution – the existence of the protein-specific molecular clock. While the rate of evolution for different proteins (genes) varies by several orders of magnitude, it seems to be roughly constant for orthologous protein in different species. This observation, first made in the sixties, begs for an explanation. What factors set the rate of accepted mutations for each protein? I think it was first demonstrated by Laurence Hurst et al. in that a major determinant of the clock rate is gene expression. It accounts for about % of the rate variance among different proteins. Other functional determinants (protein length, number of protein-protein interactions, metabolic flux, etc) were shown by other studies to account for smaller fractions of the rate variance (usually between %–%).Recently, Drummond et al. ( PNAS, Cell) proposed an interesting hypothesis that the clock rate depends strongly on gene expression primarily due to the protein misfolding. Specifically, the evolution is significantly constrained by the optimality of codons used in each protein. Non-optimized codons will generate a significant fraction of toxic (due to protein aggregation) misfolded proteins. The amount of misfolded proteins is proportional to the number of translational events and consequently depends on gene expression.One comment I have about the paper is that it primarily tests the independent contribution of the gene expression versus other structural-functional constraints. It does not test the validity of the misfolding hypothesis that expression affects the evolutionary rate primarily through misfolding. The authors need to make this very clear. There may be many other explanations for the observed expression-rate correlation, for example, energy conservation. As far as I know, currently there are no experimental confirmations of the proposed misfolding theory.I think the method used to investigate the homogenization of rates is quite elegant. I think that in addition to the fact that, as authors state, "the interpretation of experimental data on gene expression is ambiguous because the quantity that is measured in most experiments is the transcript level rather than the number of translation events", the analysis of the multi-domain proteins escapes the difficulty that expression of any protein is very heterogeneous in many different conditions. There is no single expression level for a protein or a domain. Analysis of domain fused into a single ORF insures that they are present in exactly the same amount (or require similar number of translation events) under all possible environmental, developmental, and regulatory conditions. Maybe authors need to add a comment about this.I think another important caveat is that the fusion of domains into a single protein usually ensures, at least, partial homogenization of protein function. This is important as authors do not check that the fused domains have the same molecular functions. For example, in metabolic network fused domain are almost always direct neighbors in a metabolic pathway. In contrast, single domain with the same structure (from the same SCOP family) may be responsible for very different metabolic functions. The same is certainly true for other (non-metabolic) proteins. Consequently, the homogenization of the evolutionary rate specifically due to expression will be estimated from above. The real expression-based homogenization will be lower.It is interesting to note that in the Figure the majority of protein domains demonstrate multimodal behavior. I think all distributions are clearly multi-modal, although with different peak weights. It may be more logical to use the lower SCOP hierarchy for the analysis in which most distributions are single peaked. But the authors state that similar results were obtained with a lower level of the SCOP classification.In future studies it may be also interesting to investigate the rate homogenization due to tight (non-transient) protein complexes. Proteins in such complexes are usually expressed at very similar rates and should, by analogy to fusions, demonstrate a similar decrease in the evolutionary rate difference.I do not understand completely the logic behind the quantitative estimation of the expression-based homogenization. To make this estimation the authors consider the correlation between the rates in single domains and the corresponding rate in multi-domain proteins (Figure ). It seems to me that to estimate the relative contribution one needs also to estimate the variation of the rates for the same orthologous domains. For example, difference in the homogenization obtained by the analysis of human-to-rat versus human-to-mouse orthologs.Authors' response: The analysis presented here already includes an internal control for rate variation in many evolutionary lineages, namely, those of paralogous domains. Therefore, we believe that the control for variation in orthologous lineages suggested by the reviewer might not be necessary in this case.Also the estimation of relative contributions were made by several studies before (I think also by the Koonin et al.) using independent contribution, i.e. partial correlation, analysis. It may be appropriate to comment on these studies.Authors' response: Somewhat unclear: independent contributions of what? Different evolutionary and functional variables? This is briefly discussed in the text (see Ref. ).Interestingly, the significantly different homogenization in observed in the case of Arabidopsis compared to the case of human proteins. What can be the reasons for this difference?Authors' response: Yes, a very curious difference but so far we have no clue as to what it might be owing to.Reviewer's report : Claus Wilke (University of Texas-Austin, nominated by Orly Alter) and Alan Drummond (Harvard University, nominated by Joel Bader)This study quantifies the extent to which protein evolutionary rates are determined by intrinsic structural or functional properties of a protein rather than by external factors, in particular expression level. The study uses a clever method to achieve this goal: it compares the evolutionary rates of individual domains when they are fused into multi-domain proteins to the evolutionary rates of the same domains when they evolve independently. The main result is that both intrinsic and external factors affect domain evolutionary rates to a comparable extent.This is a wonderful, insightful and important study. Both the method and the results will have an immediate impact on studies of evolutionary rates. We expect that numerous groups (hopefully including the authors) will rapidly extend the present results to other organisms, as the results warrant. In particular, this study emphasizes the role of mutational tolerance in protein domains as a separate issue from the multiplicative role of expression level, a perspective which should be paradigmatic for future investigations.The authors note that their results fit well with the hypothesis of mistranslation-induced protein misfolding (MIM) which we have recently introduced (Drummond and Wilke, Cell ). Under the MIM hypothesis, the selection pressure arises from the fitness cost of misfolded proteins. There is no reason to assume that different domains would have the same propensity to misfold under translation errors, nor that they would impose the same fitness cost when misfolded. Therefore, we expect expression level to homogenize evolutionary rates only to the extent to which two domains have similar misfolding propensities.The analysis invites a few questions. From the authors' finding that both intrinsic and extrinsic factors contribute roughly equally to the evolutionary rate of domains, it may be tempting to infer that the same conclusion holds for entire genes. Is this inference valid? Perhaps. A way to answer this question in the affirmative would be to show that the evolutionary rate of a multidomain protein is in general well-estimated by the weighted average of the rates for each constituent domain, or that domains cover nearly all of the sequence (in other words, that variation in domain evolutionary rates explains nearly all the variation in gene evolutionary rates). If the evolutionary rates of genes are strongly influenced by features outside of domains, such as the number/length of linker regions, then the inference would be invalid. The authors may at least wish to clarify exactly what their analyses show and what is left to inference.Authors' response: Our findings for domains would might not quantitatively hold for the whole proteins, in particular, because polypeptide chains outside the domain boundaries have a tendency to fold into disordered or, at least, non-globular structures. However, as long as the regions outside of domains do not dramatically affect the evolutionary rates of the domains themselves, we believe that the present findings can be safely extrapolated to the entire repertoire of the combinations of globular domains.Along the same lines, what do we know about the domains being analyzed? Do they appear more often in fast-evolving or slow-evolving genes? Highly expressed or weakly expressed genes? The unstated assumption of the analysis is that the domains show no bias in their origins. Perhaps the assumption could be stated, if not tested, so that we may understand the caveats (if any) in generalizing the present results to whole proteomes.Authors' response: Again, no assumption is made that the currently detectable structural domains are unbiased "in their origins". For all we know, we might expect the proteins that contain identifiable domains to be biased toward higher "status" (higher expression levels, lower evolutionary rates etc). However, because we compare pairs of domains in multidomain proteins against random pairings of the same domains, there is no need for this domain set to be unbiased relative to other proteins (protein segments) for the conclusions to be valid.While the authors have done an admirable job of putting the MIM hypothesis in perspective, at times the language is still even more exuberant than that of the hypothesis' original authors. For example, MIM is referred to as "the dominant factor of evolution" (p.), whereas in previous works it is referred to as "a dominant constraint on coding-sequence evolution". The difference is nontrivial. The MIM hypothesis was advanced to explain variation in evolutionary rates between coding sequences. The difference in evolutionary rates between an essential protein-coding gene and a random snippet of junk DNA is not mostly due to MIM costs; it is primarily due to selection for a folded, functional polypeptide. The premise of the MIM hypothesis is that, between two coding sequences which are both under pressure to produce functional proteins, that pressure may be more or less constant between them, so that the differences in their evolutionary rates arise mostly from other factors such as MIM.Authors' response: The readers will, beyond doubt, benefit from these clarifications from the authors of the MIM hypothesis. The differences in evolutionary rates between junk DNA (e.g., pseudogenes) and protein-coding genes are indeed substantial, indicating that selection for (any kind of) folded, functional polypeptide is a major factor in evolution compared to which differences between different protein folds might be typically small.In their Discussion, the authors nicely divide the mistranslation-induced misfolding (MIM) hypothesis into two parts: a domain's robustness to translation errors to avoid misfolding into a costly molecule, and the cost-multiplier of translation frequency. The description, however, should be modified. For example, "i) Intrinsic misfolding robustness that depends on the structural features of the given domain that are captured by the concept of designability." We believe that designability is a relatively minor contributor to robustness, whereas thermodynamic stability is likely to be a major contributor. Designability, the number of sequences which fold into a particular structure, is a property of a structure, whereas thermodynamic stability is a joint property of the structure and the specific amino-acid sequence. Consider instead: "i) Intrinsic misfolding robustness that depends on the structure and sequence of the given domain. It has been shown."Authors' response: we agree that stability is likely to be more important than designability; the text was modified essentially as suggested.Similarly, factor (i) of the generalized MIM hypothesis states that " [m] is folding of a protein molecule is assumed to incur a specific fitness cost, primarily, through the poisoning effect of misfolded proteins". First, we argue in favor of both direct toxicity (i.e., "poisoning effect") and indirect toxicity (i.e., overwhelming the chaperone machinery so that another insult, such as heat shock, becomes lethal). But more importantly, the nature of the fitness cost is not an assumption of the MIM hypothesis, which only posits that misfolding is costly. The basis of that cost remains an open question, as we (Cell ) are at pains to make clear. We do argue (rather than assume!) that the evidence speaks against certain costs, such as loss of function and biosynthetic cost. To avoid confusion, it may be best to say that "misfolding of a protein molecule is assumed to impose a specific fitness cost which may be unrelated to the protein's function, such as through a cytotoxic effect."Authors' response: The text was modified essentially as suggested.We had extensive comments on a previous version of the manuscript; all these comments have been addressed satisfactorily in the revision. We are truly grateful to the authors for their insistence on a careful representation of our previous work, and believe the manuscript is unusually balanced in its presentation.Supplementary MaterialAdditional file 1Comparison of the domain evolution rate differences in all human proteins and in multidomain proteins (text).Click here for fileAdditional file 2Comparison of the domain evolution rate differences in all Arabidopsis proteins and in multidomain proteins (text).Click here for file
|
PMC2367030.txt
|
TITLE: A novel Mutein of TNFα Containing the Arg-Gly-Asp Sequence Shows Reduced Toxicity in Intestine AUTHORS: H. Shikama, K. Miyata, N. Sakae, K. Kuroda, K. Nishimura, S. Yotsuya, M. Kato ABSTRACT: The effects of human tumour necrosis factor-α (TNFα), or its mutein (F4168) having the cell adhesive Arg-Gly-Asp sequence at the N-terminus, on intestinal injury, were examined. Histopathological examination revealed that an intravenous injection of TNFα resulted in marked haemorrhage or oedema in the caecum of rats, whereas F4168 showed no such effects even at the same therapeutic dose. Moreover, the number of neutrophils that adhered to endothelial cells or infiltrated the mucosal tissue was much higher after TNFα injection compared with F4168 in vivo. The enhanced adhesion of neutrophils on to human umbilical vein endothelial cells also occurred when the latter were pre-stimulated with TNFα but not with F4168 in vitro. The expression of the cell adhesion molecules including endothelial leukocyte adhesion molecule- or intercellular adhesion molecule- on F4168- stimulated human umbilical vein endothelial ceils was significantly lower than that stimulated with TNFα. These results suggest that the Arg-Gly-Asp sequence introduced into the TNFα molecule abrogates the side effect of this cytokine such as tissue injury or shock, and that F4168 could be useful for systemic therapy. BODY: No Body Content
|
PMC2787512.txt
|
TITLE: Hepatitis A virus (HAV) packaging size limit AUTHORS: Krishnamurthy Konduru, Siham M Nakamura, Gerardo G Kaplan ABSTRACT: BackgroundHepatitis A virus (HAV), an atypical Picornaviridae that causes acute hepatitis in humans, grows poorly in cell culture and in general does not cause cytopathic effect. Foreign sequences have been inserted into different parts of the HAV genome. However, the packaging size limit of HAV has not been determined. The purpose of the present study is to investigate the maximum size of additional sequences that the HAV genome can tolerate without loosing infectivity.ResultsIn vitro T7 polymerase transcripts of HAV constructs containing a -nt fragment coding for a blasticidin (Bsd) resistance gene, a ,-nt fragment coding for the same gene fused to GFP (GFP-Bsd), or a ,-nt fragment containing a hygromycin (Hyg) resistance gene cloned into the 2A-2B junction of the HAV genome were transfected into fetal Rhesus monkey kidney (FRhK4) cells. After antibiotic selection, cells transfected with the HAV construct containing the resistance gene for Bsd but not the GFP-Bsd or Hyg survived and formed colonies. To determine whether this size limitation was due to the position of the insertion, a bp fragment coding for the Encephalomyocarditis virus (EMCV) internal ribosome entry site (IRES) sequence was cloned into the ' nontranslated (NTR) region of HAV. The resulting HAV-IRES retained the EMCV IRES insertion for - passages. HAV constructs containing both the EMCV IRES at the ' NTR and the Bsd-resistance gene at the 2A-2B junction could not be rescued in the presence of Bsd but, in the absence of antibiotic, the rescued viruses contained deletions in both inserted sequences.ConclusionHAV constructs containing insertions of approximately - nt but not , nt produced viable viruses, which indicated that the HAV particles can successfully package approximately nt of additional sequences and maintain infectivity. BODY: BackgroundHepatitis A virus (HAV), a member of the Picornaviridae family, causes acute hepatitis in humans. The - nm non-enveloped HAV icosahedral particles encapsidate a . kb single-stranded positive-sense RNA genome [], which contains a long open reading frame (ORF) flanked by ' and ' end non-translated regions (NTR). The long ' NTR of approximately nucleotides (nt) has a complex structure and contains an internal ribosome entry site (IRES) required for viral translation. The ' NTR is short and ends in a poly(A) tail []. The HAV long ORF encodes a polyprotein of approximately kDa that undergoes co- and post-translational processing into smaller structural (VP0, VP3, and VP1-2A) and non-structural (2B, 2C, 3A, 3B, 3C, and 3D) proteins [,]. HAV 3C is a cysteine proteinase (3Cpro) responsible for most of the polyprotein cleavages and is the only protease coded in the HAV genome [-]. The 2A-2B junction is the primary cleavage site of the HAV polyprotein processed by 3Cpro [,]. The VP0 undergoes structural cleavage, and an unknown host cellular protease cleaves the VP1-2A junction [].HAV is a hepatotropic virus transmitted through the fecal-oral route. Pathogenesis of HAV is poorly understood, and it is unclear whether the virus needs to replicate in extra-hepatic sites before reaching the liver. After binding to its cellular receptor HAVCR1 [,], the HAV genome is delivered to the cytoplasm by an unknown mechanism. Once in the cytoplasm, the HAV genome is translated, transcribed, and encapsidated without in general causing cytopathic effect. The virus is eliminated by the immune system and does not establish chronic infection. Inactivated HAV vaccines are safe and effective, and are currently used in most of the world to prevent and treat HAV infection [,,].Considerable interest has been devoted to develop HAV as an expression vector for combination vaccines, expression of proteins in the liver, and basic research on this poorly understood human pathogen. We have previously shown that replication-competent HAV constructs containing inserts of - nt coding for malaria and FLAG-tag epitopes at the N-terminus of the HAV polyprotein were stable for at least passages []. An HAV recombinant containing -nt insertion at the 2A-2B junction was stable for up to five passages []. HAV constructs carrying a seven amino acid human immunodeficiency virus gp41 epitope at the surface of the HAV particles elicited an immune response against gp41 in infected animals [,]. Recently, we showed that a -nt fragment coding for a blasticidin (Bsd) resistant gene inserted at the 2A-2B junction of wild type HAV was stable for passages [], conferred Bsd resistance to infected cells, and was used to develop an antibiotic resistance titration assay to evaluate anti-HAV antibodies in preparations of human immunoglobulins []. Although foreign sequences have been successfully inserted into different parts of the HAV, the packaging size limit of HAV has not been determined. To study the maximum size of foreign sequences that HAV could tolerate, we cloned exogenous sequences into the 2A-2B junction and/or N-terminus of the polyprotein. A -nt fragment coding for a Bsd resistance gene was engineered into the 2A-2B junction of HAV and maintained for passages with or without antibiotic selection. A recombinant HAV containing the Encephalomyocarditis (EMCV) internal ribosome entry site (IRES) of nt cloned at the 'NTR between the HAV IRES and the initiation codon of the HAV polyprotein only maintained the insert for a few passages. However, recombinant HAV constucts containing approximately , nt at the 2A-2B junction or both the EMCV IRES at the 'NTR and the Bsd resistance gene at the 2A-2B junction could not be rescued from transfected cells. Our results indicated that HAV has a restricted packaging size limit that can accommodate approximately nt of additional sequences.ResultsInsertions at the 2A-2B junctionTo determine the packaging size limit of the HAV genome, we inserted foreign sequences into a -nt polylinker engineered between 3Cpro cleavage sites at the 2A-2B junction of the HAV cDNA in pHAVvec9 [] (Figure ). A -nt fragment containing a Bsd resistance gene coding for a Bsd deaminase, a -nt fragment coding for the same resistance gene fused to the GFP protein (GFP-Bsd), or a -nt fragment containing a hygromycin (Hyg) resistance gene coding for a hygromycin-phosphotransferase were inserted into the polylinker of pHAVvec9, and termed pHAVvec9-Bsd, pHAVvec9-GFP-Bsd, and pHAVvec9-Hyg, respectively (Figure ). All the inserts lacked translation initiation and termination codons to allow expression of the coded proteins as part of the HAV polyprotein. To rescue viruses, T7 polymerase in vitro transcripts of the HAV constructs were transfected into FRhK4 cells. A day after transfection, cells were split : and grown in medium that contained Bsd .- μg/ml or hygromycin - μg/ml. As expected, some cells transfected with transcripts from pHAVvec9-Bsd survived selection with μg/ml Bsd and developed colonies (Figure 2A). However, cells transfected with the T7 in vitro transcripts from pHAVvec9-GFP-Bsd did not survive antibiotic selection, show GFP fluorescence, or produce infectious virus even without Bsd selection. Cells transfected with in vitro transcripts from pHAVvec9-Hyg also did not survive antibiotic selection. A virus stock termed HAVvec9-Bsd was prepared from these Bsd-resistant cells and used to infect naïve FRhK4 cells in the presence of μg/ml Bsd. Cells infected with HAVvec9-Bsd but not mock-infected cells survived selection with Bsd, indicating that HAVvec9-Bsd was infectious. However, HAVvec9-Bsd produced -. log less virus than HAVvec9 and parental HAV/ []. The stability of Bsd gene in HAVvec9-Bsd virus was assessed during serial passages in the absence of Bsd. RT-PCR (Figure 2B) and nucleotide sequence analysis (data not shown) revealed that the Bsd gene insert was stable during the serial passages in FRhK4 cells. These data suggested that HAV could tolerated insertions of approximately -, nt at the 2A-2B junctionFigure 1Schematic representation of HAV constructs containing insertions in the 2A-2B junction. Antibiotic resistance genes were cloned the 2A-2B junction of the HAV genome in pHAVvec9 []. This plasmid contains a polylinker coding for SalI, SnaBI, and KpnI restriction sites flanked by Gly hinges (grey color letters) and HAV 3Cpro cleavage site (arrows) engineered into 2A-2B junction of pT7HAV. The blasticidin resistance gene (Bsd) coding for blasticidin deaminase, green fluorescent protein (GFP)-Bsd fusion protein, and hygromycin resistance gene (Hyg) coding for hygromycin phosphotransferase were cloned into the SalI and KpnI sites in the polylinker of pHAVvec9, and the resulting plasmids were termed pHAVvec9-Bsd, pHAVvec9-GFP-Bsd, and pHAVvec9-Hyg, respectively.Figure 2Rescue and stability of the HAV constructs containing antibiotic resistance genes in the 2A-2B junction. (A) Subconfluent FRhK4 cells were transfected with in vitro T7 polymerase transcripts of pHAVvec9, pHAVvec9-Bsd, or pHAVvec9-GFP-Bsd, or pHAVvec9-Hyg or mock-transfected. Cells were and incubated for -weeks in selection medium containing μg/ml Bsd for all transfections except for cells transfected with pHAVvec9-Hyg transcripts, which were grown in the presence of μg/ml Hyg. Phase contrast micrographs were taken with a Zeiss Axiovert microscope at × magnification. (B) Stability of the recombinant HAV. Viral RNA was extracted and fragments were amplified by RT-PCR using HAV primers forward VP1 coding for nts - and reverse 2B primer coding for nts -. RT-PCR fragments amplified from RNA extracted from HAV/ (lane ), HAVvec9 (lane ), HAVvec9-Bsd passage (lane ), and HAVvec9-Bsd passage -times in the presence (lane ) or absence (lane ) of Bsd were analyzed by TAE-%-agarose gel electrophoresis. The RT-PCR fragments from HAVvec9-Bsd and HAVvec9 are indicated by arrowheads and their sizes given in base pairs (bp). The size of the DNA molecular weight markers (lane M) is indicated in bp.Insertions at the ' NTRTo assess whether the packaging size limitation of -, nt was site-specific, we inserted a nt fragment into the ' NTR of the HAV genome between the HAV IRES and the initiation codon of the HAV polyprotein. We chose the EMCV IRES because it is a strong ribosome entry site compared to the weak HAV IRES [,] and has been widely used in other viral systems [-]. We hypothesized that a virus with both IRESes would have a translation advantage compared to the HAV IRES alone. The EMCV IRES -nt fragment was cloned adjacent to the polyprotein initiation codon to assure that it would drive translation of the HAV polyprotein (Figure ). FRhK4 cells were transfected with T7 transcripts of the double IRES HAV construct. After two weeks of incubation, IF analysis revealed that % of the cells had the characteristic cytoplasmic granular fluorescence of HAV-infected cells compared to % of the cells transfected with pHAVvec9 or HAVvec9-Bsd transcripts (Figure 4A). A virus stock was prepared and termed HAV-IRES, and the presence of the EMCV IRES was assessed by RT-PCR analysis (Figure 4B). To do so, viral particles were purified by sedimentation through a % sucrose cushion, and viral RNA was extracted from the pellet. As a negative control, the same amount of transfected T7 RNA transcripts ( μl) was purified in parallel by sedimentation through a % sucrose cushion, and RNA was extracted from the pellet. As a positive RT-PCR control, we used RNA extracted from purified HAV/ particles. RT-PCR analysis showed that the HAV-IRES but not HAV/ particles contained the expected nt EMCV IRES fragment, which was verified by automated nucleotide sequencing. RT-PCR fragments were not amplified from the negative control sample. These data indicated that the -nt EMCV IRES insert was packaged into the HAV-IRES particles. To further assess the stability of the inserted EMCV-IRES, we performed serial passages of HAV-IRES in FRhK4 cells at weekly intervals. At each cell passage, a virus stock was produced and analyzed by RT-PCR as described above. After passages, most of the EMCV IRES was deleted leaving only a small insert of nt (data not shown). These data indicated that HAV was capable of packaging the -nt insert coding for the EMCV IRES but that HAV-IRES was an unstable recombinant virus.Figure 3Schematic representation of the HAV construct containing the EMCV IRES at the 'NTR of the HAV genome. A -nt cDNA fragment containing the EMCV IRES was inserted into the HAV 'NTR between the HAV IRES and the initiation codon (ATG) of the HAV polyprotein. The resulting construct was termed pHAV-IRES, and contained a short synthetic polylinker containing KpnI, SnaBI, SacII, EcoRI, and AciI restriction sites followed by nts - of the EMCV IRES (dashed rectangle) inserted into the XbaI site in VP4 in such a way as to recreate the native initiation codon of the HAV polyprotein. The mature HAV proteins and the viral poly (A) tail (AAA) are indicated in the schematic representation of the HAV genome organization.Figure 4Rescue and stability of the HAV constructs containing the EMCV IRES at the 'NTR. (A) IF analysis of FRhK4 cells transfected with in vitro RNA transcripts from pHAVvec9, pHAVvec9-Bsd, or pHAV-IRES or mock-transfected. Cells were fixed with acetone two weeks post-transfection, and stained with anti-HAV neutralizing monoclonal antibodies K2-4F2 and K3-4C8 and FITC-conjugated goat anti-mouse antibodies. Micrographs were taken with a Zeiss microscope at × magnification. (B) Analysis of the stability of HAV recombinants containing the EMCV IRES. RT-PCR analysis of genomic RNA extracted form HAV/, HAV-IRES, HAVvec9-Bsd virions and amplified using primers corresponding to nts - and - of HAV. As negative control, T7 polymerase in vitro transcripts from pHAV-IRES were spiked into media, layered on top of a % sucrose cushion, and sedimented by ultracentrifugation. RNA extracted from the pellet was used for the RT-PCR analysis (Naked-RNA control). The size of the DNA molecular weight markers (lane M) is indicated in bp.Insertions at both the 2A-2B junction and the 'NTRTo rule out that constrains at the specific insertion sites unrelated to packaging limited the viability of the HAV constructs, we cloned both the Bsd resistance gene into the 2A-2B junction and the EMCV IRES into the ' NTR (Figure ). T7 polymerase in vitro transcripts of the resulting construct, pHAVvec9-Bsd-IRES, were transfected into FRhK4 cells. Transfected cells did not survive selection μg/ml Bsd indicating that HAV did not tolerate the double insertion. IF analysis of transfected cells grown in the absence of Bsd selection showed that less than % of cells contained HAV antigens (data not shown). RT-PCR analysis of RNA extracted from particles purified by sedimentation through a % sucrose cushion revealed that the EMCV IRES and the Bsd resistance gene contained deletions (Figure ). Nucleotide sequence analysis of PCR fragments showed the presence of only nt from the EMCV IRES at the 'NTR and nt from the Bsd resistance gene at 2A-2B junction. Interestingly, the polylinker, 3Cpro cleavage sites, and Gly hinge flanking sequences were preserved. These data indicated that HAV cannot tolerate insertions larger than nucleotides, and that this limitation in size is not due to site-specific restrictions of the inserted sequences.Figure 5Schematic representation of the HAV construct containing the EMCV IRES at the 'NTR and the Bsd resistance gene at the 2A-2B junction. The EMCV IRES was inserted into the HAV 'NTR in pHAVvec9-Bsd. The construct containing both the EMCV IRES at the 'NTR and the Bsd resistance gene at the 2A-2B junction was termed pHAVvec9-Bsd-IRES.Figure 6Rescue and stability of the HAV constructs containing the EMCV IRES and Bsd resistance gene. Viral RNA was extracted from purified virions and analyzed by RT-PCR as described in Figures 2B and 4B.DiscussionIn this paper we studied the packaging size limit of HAV. The virus tolerated the insertion of approximately nt at the 2A-2B junction and nt at the 'NTR. The insertion of -nt fragment coding for a Bsd resistance gene flanked by 3Cpro cleavage sites was remarkably stable for passages even in the absence of antibiotic. However, HAV could not accommodate larger inserts in the same site, and viruses containing GFP-Bsd and Hyg selectable markers could not be rescued from transfected cells. It should be pointed out FRhk4 cells transfected with eukaryotic expression vectors containing the GFP-Bsd and Hyg resistance genes survived selection with Bsd and Hyg, respectively. Moreover, FRhK4 cells transfected with an eukaryotic expression vector containing the GFP-Bsd fusion protein fluoresced similarly to cells transfected with a construct containing only GFP (data not shown). Since HAVvec9-Bsd was very stable, it is unclear why we could not rescue an HAV construct containing GFP-Bsd. One simple explanation is that we exceeded the packaging size limit of HAV. However, cells transfected with pHAVvec9-GFP-Bsd in vitro transcripts did not survive selection with Bsd suggesting that packaging alone was not the only factor affecting infectivity. Therefore, we hypothesized that site-specific or sequence-specific constraints at the 2A-2B junction site prevented HAV replication. To test this hypothesis, we introduced a completely different sequence at another site of the HAV genome. To that effect, we cloned a -nt fragment coding for the EMCV IRES between the HAV IRES and the initiation codon of the HAV polyprotein. Because HAV containing the EMCV IRES was unstable and tended to delete the insert in a few passages, we limited our study to viruses produced at the initial passage. The failure to rescue virus from the construct containing the Bsd resistance gene at the 2A-2B junction and the EMCV IRES at the 'NTR indicated that the maximum size of foreign sequences that can be incorporated into the HAV genome of a viable virus is approximately nt. Our study showed that HAV can package genomes of approximately , nt composed of the full-length genome and an insertion of approximately nt, which can accommodate an additional amino acid protein into the HAV polyprotein. Although the size of the extra sequences is somehow limited, HAV could be used as an expression vector for the development of combination vaccines and delivery of genes to the liver.ConclusionIn this study we showed that viable HAV can package full-length viral genomes containing insertions of approximately nt.MethodsCells and virusesFetal rhesus monkey kidney (FRhK4) cells, a gift of S. Emerson, National Institutes of Health (NIH), were grown in Dulbecco's modified Eagle's medium supplemented with % fetal bovine serum. The cell culture-adapted human HM- strain of HAV was derived from an infectious cDNA clone [], termed HAV/ [], and grown in FRhK- cells. Nucleotide positions of the HAV genome are according to the published HM175 HAV cDNA sequence [].Plasmid constructionsStandard molecular biology methods [] were used to construct the HAV recombinants. PCR-based DNA fragments were amplified using expand high fidelity PCR kit (Roche) in cycles consisting of °C for sec, °C for min, and °C for - min. For overlap PCR, DNA fragments were denatured at °C and annealed at °C for min in each step. The blasticidin (Bsd), fusion of green fluorescent protein with Bsd (GFP-Bsd), and hygromycin (Hyg) resistance genes were generated by PCR with eliminating the translation initiation (ATG) and termination codons and introducing ' SalI and ' KpnI restriction sites for cloning into the polylinker at the 2A-2B junction of the HAV polyprotein in pHAVvec9 []. To clone inserts in-frame at the 2A-2B junction, PCR products and pHAVvec9 were digested with SalI and KpnI, gel purified, and ligated. The oligonucleotides used to generate the PCR fragments are described in Table . The sequences inserted into the HAV geome are described in Table . Constructs were verified by automated nucleotide sequence analysis. The following plasmids were used in this work:Table 1PCR PrimersName*PCR FragmentOligonucleotide Sequence&Bsd forward (SalI)Bsd5'-GTCGACGTCGACCAGGCCAAGCCTTTGTCTCAAGAA-'Bsd reverse (KpnI)Bsd5'-CGGTTAGGTACCGCCCTCCCACACATAACCAGAGGG-'GFP forward (SalI)GFP-Bsd5'-GTCGACGTCGACGCCTCCAAAGGAGAAGAACTTTTC-'Hyg forward (SalI)Hyg5'-GTCGACGTCGACAAAAAGCCTGAACTCACCGCGACG-'Hyg reverse (KpnI)Hyg5'-CGGTTAGGTACCGTTAGCCTCCCCCATCTCCCGATC-'EMVC IRES forward (XbaI)EMCV IRES5'-TTAGTCTAGATGGTACCTACGTACCGCGGAATTCCGCCCCTCTCCCTAACGTTACTGGCCGAA-'EMVC IRES reverse (XbaI)EMCV IRES5'-TTTCTAGACATGTTCATATTATCATCGTGTTTTTCAA-'*Restriction enzymes in brackets cut the PCR fragment amplified with the oligonucleotides.&The restriction site in each oligonucleotide is underlined.Table 2Insertions in the HAV 2A-2B junction or 'NTRConstructInsertSiteSize (nt) *Infectivity&pHAVvec9polylinker2A-2B66+pHAVvec9-BsdBsd2A-2B396+pHAVvec9-GFP-BsdGFP-Bsd2A-2B1098-pHAVvec9-HygHygromycin2A-2B1032-pHAV-IRESEMCV IRES5'NTR606+pHAVvec9-Bsd-IRESBsd2A-2B396-EMCV IRES5'NTR606* pHAVvec9-Bsd, pHAVvec9-GFP-Bsd, and pHAVvec9-Hyg, and pHAVvec9-Bsd-IRES contain additional -nt corresponding to the polylinker inserted at the 2A-2B junction.&Infectivity (+) or lack of infectivity (-) of the T7 polymerase in vitro synthesized transcripts from the plasmid constructs was assessed in FRhK4 cells by IF analysis.pT7HAV contains the infectious cDNA cell culture-adapted HM- strain of HAV under the control of a T7 RNA polymerase promoter in pGEM1, and the in vitro transcripts were infectious in FRhK4 cells [].pHAVvec9 was derived from pT7HAV as described previously []. pHAVvec9 contains a polylinker with unique SalI, SnaBI, and KpnI restriction sites.pHAVvec9-Bsd contains Bsd gene cloned into the polylinker of the HAV cDNA in pHAVvec9 [].pHAVvec9-GFP-Bsd contains a GFP-Bsd fusion cloned into the HAV cDNA polylinker in pHAVvec9. The GFP-Bsd fusion was amplified by PCR from pTracer-CMV/Bsd (Invitrogen). To create unique SalI site in GFP-Bsd insert, a silent substitution mutation (from GTCGAC to GTAGAC) was introduced to eliminate SalI site in the GFP gene sequence using overlap PCR.pHAVvec9-Hyg contains a Hyg resistance gene cloned into the HAV cDNA polylinker in pHAVvec9. The Hyg resistance gene was amplified by PCR from pCFB-EGSH (Stratagene).pHAV-IRES contains the EMCV IRES (corresponding to - nt of the EMCV genome) cloned between the HAV IRES and the HAV polyprotein initiation codons in pT7HAV (Figure ). To construct pHAV-IRES, a -nt cDNA fragment containing nt - of EMCV flanked at the ' end by a XbaI site and at the 'end by the two tandem initiation codons of the HAV polyprotein and an XbaI restriction site was amplified from pCITE- (Novagen), a plasmid containing the EMCV IRES. This amplified cDNA fragment was cut with XbaI and cloned into the unique XbaI site adjacent to the HAV initiation codons in pT7HAV.pHAVvec9-Bsd-IRES was constructed by cloning the EMCV IRES from pHAV-IRES into pHAVvec9-Bsd. To do so, the ,-nt BspEI and BstEII fragment of pHAV-IRES containing the EMCV IRES was cloned into pHAVvec9-Bsd cut with BspEI and BstEII at nts and of the HAV cDNA, respectively. The resulting construct contained both the EMCV IRES and the 'NTR and the Bsd resistance gene at the 2A-2B junction of the HAV cDNA.In vitro RNA synthesis and transfectionSynthesis of full-length HAV RNA transcripts was performed using T7 RNA polymerase []. To do so, plasmids were linearized with HaeII, which cuts immediately downstream of the poly(A) of the HAV cDNA [,]. Subconfluent FRhK4 cells in -cm2 flasks were transfected with in vitro synthesized RNA transcripts using DEAE-dextran as facilitator [] or Lipofectamine (Invitrogen). After min incubation at room temperature, fresh media was added to replace transfection solution and incubated at °C. Cells were split weekly into new flasks, and an aliquot was passed to -well chamber slides for immunofluorescence (IF) analysis. To prepare viral stocks, monolayers with approximately % of the cells expressing HAV antigens as assessed by IF analysis were subjected to three freeze-thaw cycles and clarified by low-speed centrifugation. Supernatants containing the virus stocks were stored at -°C.Immunofluorescence (IF) analysisFRhK4 cells were grown in -well Permanox chamber slides (Nunc, Inc.) at °C in a CO2 incubator for - h. Cell culture media was aspirated and the cells were fixed with cold acetone for min at -°C. Fixed cell monolayers were air-dried, blocked with PBS-% FBS at room temperature, and stained with a mix of anti-HAV neutralizing murine monoclonal antibodies K2-4F2 and K3-2F2 [] and fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse antibody (KPL, Inc). Stained slides were air-dried, coverslips were mounted with PermaFluor aqueous medium (Shandon Immunon, PA), and fluorescent micrographs were taken with a Zeiss Axioscope microscope at × magnification.RT-PCR and nucleotide sequence analysisRNA was extracted from HAV particles using QIAamp viral RNA mini kit (Qiagen). RT-PCR was carried out using the Superscript-II enzyme kit (Invitrogen). The cDNA was synthesized with primers corresponding to HAV nt position - or -. PCR fragments were amplified using primers corresponding to HAV nts - and - or - and -. Amplified DNA fragments were gel purified, and sequenced using the ABI Prism BigDye terminator cycle-sequencing ready reaction kit (Applied Biosystems) and with ABI Prism (model ) analyzer (Applied Biosystems).Competing interestsThe authors declare that they have no competing interests.The findings and conclusions in this article have not been formally disseminated by the Food and Drug Administration and should not be construed to represent any Agency determination or policy.Authors' contributionsKK and GGK conceived the study. KK and GGK designed the experiments and drafted the manuscript. KK and SMN performed the experiments.
|
PMC3041663.txt
|
TITLE: Utility of electronic patient records in primary care for stroke secondary prevention trials AUTHORS: Alex Dregan, Michael A Toschke, Charles D Wolfe, Anthony Rudd, Mark Ashworth, Martin C Gulliford ABSTRACT: BackgroundThis study aimed to inform the design of a pragmatic trial of stroke prevention in primary care by evaluating data recorded in electronic patient records (EPRs) as potential outcome measures. The study also evaluated achievement of recommended standards of care; variation between family practices; and changes in risk factor values from before to after stroke.MethodsData from the UK General Practice Research Database (GPRD) were analysed for , participants with an index first stroke between and from family practices. For each subject, the EPR was evaluated for the months before and after stroke. Measures relevant to stroke secondary prevention were analysed including blood pressure (BP), cholesterol, smoking, alcohol use, body mass index (BMI), atrial fibrillation, utilisation of antihypertensive, antiplatelet and cholesterol lowering drugs. Intraclass correlation coefficients (ICC) were estimated by family practice. Random effects models were fitted to evaluate changes in risk factor values over time.ResultsIn the months following stroke, BP was recorded for %, cholesterol for % and body mass index (BMI) for %. ICCs by family practice ranged from . for BP and BMI to . for LDL and HDL cholesterol. For subjects with records available both before and after stroke, the mean reductions from before to after stroke were: mean systolic BP, . mm Hg; diastolic BP, . mm Hg; total cholesterol, . mmol/l; BMI, . Kg/m2. There was an absolute reduction in smokers of % and heavy drinkers of %. The proportion of stroke patients within the recommended guidelines varied from less than a third (%) for systolic BP, just over half for BMI (%), and over % (%) on alcohol consumption.ConclusionsElectronic patient records have potential for evaluation of outcomes in pragmatic trials of stroke secondary prevention. Stroke prevention interventions in primary care remain suboptimal but important reductions in vascular risk factor values were observed following stroke. Better recording of lifestyle factors in the GPRD has the potential to expand the scope of the GPRD for health care research and practice. BODY: BackgroundIndividuals who survive stroke are at increased risk of recurrent stroke []. The risk of recurrence is highest during the first year following a stroke but remains elevated for at least years [,]. Clinical practice recommendations for secondary prevention of stroke, including the UK Intercollegiate Stroke Working Party (ICSWP) guidelines [] and the American Heart Association (AHA) guidelines, [] emphasise the requirement to intervene to reduce the risk of further strokes, and other cardiovascular events, by targeting major modifiable vascular risk factors including elevated blood pressure (BP), and high cholesterol levels, as well as implementing antiplatelet therapy and lifestyle change when appropriate. These recommendations draw on evidence from meta-analyses of randomised controlled trials and other studies that provide evidence of the protective role of antiplatelet, cholesterol lowering and antihypertensive therapy against the risk of recurrent strokes and other vascular events [-].Few nationally representative, population-based studies of stroke secondary prevention, and changes in vascular risk factor values following stroke, have been reported. The present study therefore aimed to evaluate the implementation of stroke secondary prevention in primary care by using electronic patient records (EPR) from a large population. The analyses had the additional purpose of informing the design of a cluster randomised trial of computerised decision support. The development of the intervention for the study has been reported elsewhere []. The reported analyses therefore explored the feasibility of using data from EPRs to evaluate outcome measures for intervention trials of stroke prevention.MethodsParticipantsData for this study were derived from the UK General Practice Research Database (GPRD). The GPRD is a large database that contains electronic patient records from over family practices in the UK. The geographic distribution and demographic profile of the GPRD are representative of the UK population []. The validity of medical diagnoses and prescribing information in GPRD has been confirmed in several studies []. In the GPRD, data are designated as 'up-to-standard' (UTS) when they are judged to be of high enough quality to be used for research.Data for the present analyses were extracted from the records of UK general practices with a registered population of approximately . million in . The initial sample comprised , participants. The last data were collected in August . Stroke patients were identified using the READ and OXMIS codes for stroke that were described previously [,]. Transient ischemic attacks were not included. A first stroke was defined as a first diagnosis of stroke during the period //-// inclusive with a minimum of months of up-to-standard record free from stroke codes preceding the index event. This ensured that all stroke patients represent new diagnoses. Incidence and case-fatality rates for stroke in GPRD appear to be comparable to estimates obtained from other epidemiological studies [].For the present analyses, only those individuals for whom a first stroke event was recorded between 1st January and 31st December were retained because these represented recent years' data with improved quality data recording and greater relevance to contemporary clinical practice. Data for individuals with stroke diagnoses first recorded after the death date were excluded. The final study sample comprised , participants with a first stroke event between 1st January and 31st December . The mean age of the participants was years and % were women.Data for analysisThe electronic patient record was evaluated for each subject for the months preceding, and for the months after, the stroke index date. The stroke index date was defined as the first date on which a stroke medical code was recorded. Data were analysed for six continuous variables: systolic blood pressure (BP; mm Hg); diastolic BP (mm Hg); total cholesterol (mmol/l); low density lipoprotein (LDL) cholesterol (mmol/L); high density lipoprotein (HDL) cholesterol (mmol/L); and body mass index (BMI; Kg/m2). For each of the continuous measures, the mean of all values recorded in each participant in the months before, or after, stroke was estimated.Electronic patient records were also evaluated for a further ten dichotomous outcomes including whether: the blood pressure was recorded; antihypertensive drugs were prescribed; cholesterol was measured (including total, LDL or HDL cholesterol); statins were prescribed; atrial fibrillation was recorded; antiplatelet drugs were prescribed; whether anticoagulants were prescribed; whether the body mass index was recorded; whether smoking or alcohol habits were recorded. These outcomes were extracted from the clinical, referral and therapy files using READ medical codes or BNF formulary codes.Recording of blood pressure, cholesterol, atrial fibrillation, and body mass index (BMI) were characterised by whether or not one or more measurements were recorded in either the year before or the year after stroke. Drug prescriptions were analysed in the same way. Antihypertensive drugs included ACE inhibitors, angiotensin-II receptors, calcium-channel blockers, centrally acting antihypertensive drugs, diuretics, beta-blockers, and adrenergic neurone blocking drugs. Antiplatelet drugs included aspirin and dipyridamole. Anticoagulants included oral (e.g. warfarin) anticoagulants and just a small number of parenteral (e.g. heparin) anticoagulant prescriptions. Information concerning smoking included whether the participant's smoking habit was recorded, as well as whether the individual was recorded as a current smoker in the months. Information concerning alcohol included whether the alcohol intake was recorded, and whether alcohol intake was higher than recommended limits. Recommended upper limits were units per week for men and units per week for women.Statistical analysisFor binary outcomes, because of right censoring following stroke arising from early mortality, a series of time-to-event analyses were implemented to estimate the proportion of stroke participants with the outcomes of interest during the months following the stroke index date. For continuous measures, the study investigated the mean values for systolic and diastolic blood pressure, cholesterol levels (total, LDL, and HDL), and BMI values in the months before and after the stroke index date. For each measure, the participant-specific mean was estimated by stroke index year. For lifestyle measures, the study estimated the proportion of stroke participants that, at any time during the months before or after the stroke index date, were recorded as current smokers or consuming over the recommended units of alcohol per week. For participants with outcome data available both before and after the stroke index date, a random-effects model FOR REPEATED MEASURES, with participant as the random effect, was used to estimate gender and age adjusted differences (and associated confidence intervals). For the binary measures (smoking and alcohol) a two-sample proportion test was used to examine differences (and associated confidence intervals) in the proportions before and after stroke.Additional analyses were carried out to explore the variation in the distribution of outcome measures at different family practices. The intraclass correlation coefficient (ICC) by family practice was calculated using one-way ANOVA for the continuous outcome measures: systolic and diastolic blood pressure, cholesterol, and BMI index. All analyses were carried out using STATA version .EthicsThe study was implemented utilising an anonymised dataset from the General Practice Research Database. The study protocol was approved by the Independent Scientific Advisory Committee (ISAC) of the Medicines and Healthcare products Regulatory Agency (MHRA) (ISAC Protocol No. 07_027R).ResultsIn the -year study period from 1st January to 31st December , a total of , first-ever stroke diagnoses were recorded in the GPRD. The frequency of records of blood pressure, cholesterol, BMI, and lifestyle factors are presented in Table . Risk factor recording was more frequent after stroke than before. In the months following stroke, BP was recorded for %, cholesterol for %, BMI for % and smoking and alcohol habits for % and % respectively. Risk factor management was greater after stroke than before, and generally increased over time, with between two thirds and three quarters of patients being treated with antihypertensive drugs or statins. A smaller proportion (%) of patients had a record of anticoagulant prescription in the months before the stroke index date which increased to12%in the months after the stroke index date. Low anticoagulant prescriptions appeared to be consistent with the small proportion of patients recorded as having atrial fibrillation in the period. A third (%) of patients with atrial fibrillation were treated with anticoagulants in the months before the stroke index date compared to % in the months after the stroke index date.Table 1Primary care management in the months before and after stroke by year of stroke.2003200420052006Total(N = ,)(N = ,)(N = ,)(N = ,)(N = ,)BP recordBefore stroke3, (%), (%), (%), (%), (%)After strokea3, (%), (%), (%), (%), (%)Antihypertensive drugs prescribedBefore stroke2, (%), (%), (%), (%), (%)After stroke a3, (%), (%), (%), (%), (%)Antiplatelet drugsBefore stroke1, (%), (%), (%), (%), (%)prescribedAfter stroke a2, (%), (%), (%), (%), (%)AnticoagulantsBefore stroke270 (%) (%) (%) (%), (%)prescribedAfter stroke a444 (%) (%) (%) (%), (%)Atrial fibrillationBefore stroke139(%)(%)(%)(%)(%)After stroke a238(%)(%)(%)(%)(%)Cholesterol recording (Total, LDL or HDL)Before stroke1, (%), (%), (%), (%), (%)After stroke a2, (%), (%), (%), (%), (%)Statin prescriptionBefore stroke1, (%), (%), (%), (%), (%)After stroke a2, (%), (%), (%), (%), (%)BMI recordBefore stroke1, (%), (%), (%), (%), (%)After stroke a2, (%), (%), (%), (%), (%)Smoking recordBefore stroke1, (%), (%), (%), (%), (%)After stroke a2, (%), (%), (%), (%), (%)Alcohol recordBefore stroke972 (%), (%), (%), (%), (%)After stroke a1, (%), (%), (%), (%), (%)Figures are frequencies (column percents).a 'After stroke' figures are estimated from time-to- event analysesTable describes substantial variation in adherence to stroke secondary prevention recommendations among family practices. In general the proportion of patients with measures recorded ranged from to % at different practices. Mean values for continuous measures and intraclass correlation coefficients for the main outcome measures by family practice are presented in Table . The intraclass correlation coefficient for the systolic BP in the months before stroke was . and . in the months after stroke. For diastolic BP, the intraclass correlation coefficient was ., and for both BMI and total cholesterol ., both before and after stroke. A higher intraclass correlation coefficient of . was observed for both LDL and HDL cholesterol both before and after stroke.Table 2Centiles for family practice-specific proportions (%) of stroke secondary prevention measuresLowest performingfamily practiceLower quartileMedianUpper quartileHighest performingfamily practiceBP recordedBefore stroke33717783100After stroke0768287100Antihypertensive treatmentBefore stroke0677379100After stroke0778186100Antiplatelet treatmentBefore stroke0344350100After stroke0535965100Anticoagulant treatmentBefore stroke036925After stroke07101450Atrial fibrillationBefore stroke003525After stroke025743Cholesterol recordedBefore stroke0384656100After stroke053616988Statin measureBefore stroke0233038100After stroke0536068100BMI recordedBefore stroke0253344100After stroke026375283Smoking recordedBefore stroke0415059100After stroke0556372100Alcohol recordedBefore stroke0142234100After stroke012254480Table 3Descriptive statistics and intraclass correlation coefficient (ICC) by family practice for continuous measuresNMean (SD)ICC (SE)SBP measure(mmHg)Before stroke16,.(.). (.)After stroke17,.(.). (.)DBP measure(mmHG)Before stroke16,.(.). (.)After stroke17,. (.). (.)Total cholesterol(mmol/l)Before stroke10,.(.). (.)After stroke13,.(.). (.)LDL cholesterol(mmol/l)Before stroke57442.(.). (.)After stroke7,.(.). (.)HDL cholesterol(mmol/l)Before stroke7,.(.). (.)After stroke9,.(.). (.)BMI measure(kg/m2)Before stroke6,.(.). (.)After stroke8,.(.). (.)Table presents the results based on patients who were selected because their records provided information both before and after the stroke event. There were quantitatively important reductions from before- to after-stroke in systolic and diastolic blood pressure as well as total and LDL cholesterol, BMI, current smoking and excessive drinking. Systolic BP was 6mm Hg lower after stroke than before, total cholesterol was . mmol/l lower after stroke than before, LDL cholesterol was . mmol/l lower, BMI was . kg/m2 lower, and there was a reduction in the proportion of active smokers of % and excessive drinkers of %. There was also evidence of a declining secular trend in values for systolic and diastolic blood pressure, as well as for total and LDL cholesterol. Tests for trend gave P values of . or smaller for each of these measures either before or after stroke.Table 4Mean values for stroke prevention measures for participants with values recorded both before and after stroke.2003200420052006OverallSystolic BP (N = ,)Before stroke148.....54After stroke141.....48Difference( % CI)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)Diastolic BP (N = ,)Before stroke82.....64After stroke78.....83Difference( % CI)**-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)Total cholesterol (N = ,)Before stroke5.....18After stroke4.....49Difference( % CI)**-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)LDL cholesterol (N = ,)Before stroke3.....99After stroke2.....38Difference( % CI)**-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)HDL cholesterol (N = ,)Before stroke1.....43After stroke1.....42Difference( % CI)-.(-.;.)-.(-.;.)-.(-.;.)-.(-.;-.)-.(-.;-.)BMI values (N = ,)Before stroke27.....78After stroke27.....44Difference( % CI)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)-.(-.;-.)Current smokers (N = ,)Before stroke31%%%%%After stroke23%%%%%Difference( % CI)-%(-%;-%)-%(-%;-%)-%(-%;-%)-%(-%;-%)-%(-%;-%)Excessive drinking (n = ,)Before stroke13%%%%%After stroke8%%%%%Difference( % CI)-%(-%%)-%(-%;-%)-%(-%%)-%(-%;%)-%(-%;-%)Figures are means except where indicated.Table presents the proportion of patients who achieved recommended target values for the measures analysed. For each measure, the denominator was the total with the item recorded. Between one quarter to a third of patients achieved systolic BP target, but the proportion was almost two-thirds for diastolic BP. Just under a half of patients achieved total cholesterol targets while just over half of stroke patients achieved LDL cholesterol targets. A similar finding emerged with respect to BMI values. More encouraging results emerged with respect to HDL cholesterol, smoking, and alcohol units per week where less than a tenth to around a fifth of the patients with values recorded were outside the recommended guidelines.Table 5Proportion of stroke patients with values recorded who were within the range recommended by the Intercollegiate Stroke Working Party guidelines in the months following a stroke index dateN2003200420052006OverallBlood pressureSystolic (< mmHg),%%%%%Diastolic (< mmHg),%%%%%CholesterolTotal (< mmol/l or -%),%%%%%LDL (< mmol/l or -%),%%%%%HDL (> . mmol/l),%%%%%Body Mass IndexWeight loss (kg/m2),%%%%%SmokingNon-smokers7,%%%%%Alcohol units per weekMen = < ;women < ,%%%%%DiscussionMain findingsThe present results show that electronic patient records from primary care provide a valuable tool for evaluating vascular risk factor values including blood pressure, cholesterol, BMI and smoking. Data on drug prescriptions provide the opportunity to assess adherence with guidelines in secondary stroke prevention. The data presented may be used to inform the design of stroke prevention studies in primary care.This study represents one of the first primary care-based stroke studies to compare risk factor values from before to after stroke. Among participants with values available both before and after stroke, there were clinically important overall reductions in blood pressure, cholesterol, BMI, cigarette smoking and excessive drinking from before to after stroke. These differences appeared to diminish slightly over time, possibly as a result of a declining secular trend in pre-stroke risk factor values. Reductions in BMI may be favourable but might sometimes be a sign of impaired nutrition after stroke. Increased prescribing of antihypertensive drugs, statins after stroke may have contributed to observed reductions in blood pressure and cholesterol values. These might potentially translate into better outcomes for stroke patients, including survival and reduced recurrence rates [,,]. The substantial variation in drug prescriptions at a practice level reinforce previous findings [] and extend these findings to GP consultations. Several explanations may be offered for the observed between-practice variation including practitioner-level factors, including the acceptability of guidelines, and practice-level factors including practice management and the practice population. These suggestions could not be tested here but need considering when designing cluster-level interventions to improve adherence to recommended guidelines.Comparison with earlier studiesThe present study findings are consistent with previous population-based studies that found recent declines in the proportion of smokers, as well as in mean total cholesterol levels, and mean systolic and diastolic blood pressure in stroke patients [,]. The study findings on the proportion of stroke participants treated with anticoagulants and antiplatelet drugs following stroke is consistent with previous studies [,-]. Grau et al. [] reported that antiplatelet drugs were administered to % of their sample. Rudd et al. [] found that % of stroke participants had a cholesterol measure available following stroke with about % prescribed a statin. Results for the proportion of participants on anti-hypertensive treatment are similar to those reported previously [,,,]. Given sampling difference as well as variation in the follow-up period length between studies may explain modest variations in figures.What this study addsCan data from primary care electronic records inform the evaluation of stroke secondary prevention in primary care? The findings of this study suggest that GPRD data represent potentially a valuable resource to evaluate whether adherence to recommended stroke prevention guidelines could improve stroke prognosis. However, this potential value must be qualified in several respects. Several aspects of stroke management are specific to stroke subtypes. However, most strokes are recorded in GPRD using codes that do not distinguish between haemorrhage and infarction. More detailed classification of stroke types using information from CAT or MRI scans is not usually feasible. A further concern is the low level of recording of lifestyle factors recording, in particular obesity, smoking and alcohol consumption. As a consequence we cannot be confident that our evaluation of the extent to which the recommended guidelines regarding these factors is generalisable to other stroke patients. Moreover, the lack of additional socio-demographic data places constraints on the ability to verify that missing data on different outcome measures is ignorable.This study provides new evidence on the magnitude of risk factor reductions during the months following index stroke. Against a background of declining secular trends in values for blood pressure and cholesterol, this study showed that there were important reductions in blood pressure, cholesterol, body mass index, cigarette smoking and excessive alcohol drinking from before to after stroke ( data is incomplete because a -month delay in data collection from practices by the GPRD). Epidemiologic studies suggest that even modest reductions in systolic blood pressure of - mmHg, or diastolic of -3mm Hg, may be associated with a % to % reduced risk of stroke [,]. The reduction in cholesterol observed is also quantitatively important in the context of evidence from clinical trials of cholesterol lowering therapy []. Reductions in BMI may be regarded favourably but may also be indicative of nutritional problems following stroke. Little evidence is available concerning trends in smoking and alcohol use in stroke populations, but there appear to be important reductions following stroke. However, the present results emphasise the need for improved data recording with respect to these risk factors for stroke recurrence. This is important considering the potential value of EPRs for health care research and policy development. The study also documents important practice level variations in the proportion of stroke participants with records or treatment for several risk factors. The factors underlying practice level variation could further help the implementation of the guidelines for secondary stroke prevention, an outcome that will be addressed through our research.Limitations of this studyThis study had the strength of a large sample drawn from a large number of family practices from across the UK. However, several limitations of the data must be acknowledged. It was not generally possible to distinguish sub-types of stroke. Only % of strokes could be classified as either haemorrhages or infarctions. The stroke type has relevance for certain aspects of secondary prevention including the prescriptions of statins, aspirin and antiplatelet drugs, and anticoagulants. However, lowering BP has been shown to reduce the risk of both ischemic and hemorrhagic stroke, while smoking and alcohol consumption are also associated with risk of stroke [,,]. Also, while all stroke diagnoses were recorded prospectively by practices, stroke patients may be admitted directly to hospital and their first contact with the practice may be for review at some later date. The stroke onset date may be therefore imprecisely recorded in primary care, perhaps often being recorded after the true stroke date. However, the validity of clinical data included in the GPRD has repeatedly been shown to be high, and data are rigorously checked and regularly audited. Another limitation of this research is the lack of a reference method for stroke diagnosis as discussed elsewhere []. It is also important to acknowledge that, for an appreciable number of participants, information about their outcome measures was not available. The study findings are comparable, however, to those from previous studies [,,]. Stroke secondary prevention was evaluated with respect to the most recent UK guidelines which represent a update of the edition. Trends showed some lack of consistency between and . This is probably explained by the shorter duration of follow-up for some participants with a stroke index date compared to those with a stroke during previous years.ConclusionsThe present study has documented, in a large nationally representative sample, that electronic based patient records represents potentially a valuable source for evaluating guidelines implementation with respect to secondary stroke prevention. The broad nature of the GPRD data allowed us to identify and operationalise several outcome measures stipulated in the ICSWP as important targets for secondary stroke prevention. The study findings indicate a consistent increase in the screening and treatment of stroke patients over time, but also the need for further improvement both in screening and drug prescriptions. The overall message of the study is that GPRD data is suitable for the implementation of cluster randomised trials to evaluate the success or failure of health interventions on stroke patients' outcomes.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsAD and MG conceptualised the paper worked on the data set, conducted the analyses, and wrote of the paper. MG, AMT, CW, AR, and MA acquired the permits and data set and provided feedback on the paper. All authors read and approved the final manuscript.Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/-///prepub
|
PMC2628281.txt
|
TITLE: DETORQUEO, QUIRKY, and ZERZAUST Represent Novel Components Involved in Organ Development Mediated by the Receptor-Like Kinase STRUBBELIG in Arabidopsis thaliana AUTHORS: Lynette Fulton, Martine Batoux, Prasad Vaddepalli, Ram Kishor Yadav, Wolfgang Busch, Stig U. Andersen, Sangho Jeong, Jan U. Lohmann, Kay Schneitz ABSTRACT: Intercellular signaling plays an important role in controlling cellular behavior in apical meristems and developing organs in plants. One prominent example in Arabidopsis is the regulation of floral organ shape, ovule integument morphogenesis, the cell division plane, and root hair patterning by the leucine-rich repeat receptor-like kinase STRUBBELIG (SUB). Interestingly, kinase activity of SUB is not essential for its in vivo function, indicating that SUB may be an atypical or inactive receptor-like kinase. Since little is known about signaling by atypical receptor-like kinases, we used forward genetics to identify genes that potentially function in SUB-dependent processes and found recessive mutations in three genes that result in a sub-like phenotype. Plants with a defect in DETORQEO (DOQ), QUIRKY (QKY), and ZERZAUST (ZET) show corresponding defects in outer integument development, floral organ shape, and stem twisting. The mutants also show sub-like cellular defects in the floral meristem and in root hair patterning. Thus, SUB, DOQ, QKY, and ZET define the STRUBBELIG-LIKE MUTANT (SLM) class of genes. Molecular cloning of QKY identified a putative transmembrane protein carrying four C2 domains, suggesting that QKY may function in membrane trafficking in a Ca2+-dependent fashion. Morphological analysis of single and all pair-wise double-mutant combinations indicated that SLM genes have overlapping, but also distinct, functions in plant organogenesis. This notion was supported by a systematic comparison of whole-genome transcript profiles during floral development, which molecularly defined common and distinct sets of affected processes in slm mutants. Further analysis indicated that many SLM-responsive genes have functions in cell wall biology, hormone signaling, and various stress responses. Taken together, our data suggest that DOQ, QKY, and ZET contribute to SUB-dependent organogenesis and shed light on the mechanisms, which are dependent on signaling through the atypical receptor-like kinase SUB. BODY: IntroductionHow intercellular communication mechanisms coordinate the activities of cells during organogenesis is an important topic in biology. In higher plants shoot apical meristems and floral meristems are the ultimate source of above-ground lateral organs, such as leaves, flowers, and floral organs []. Meristems are organised into three distinct meristematic or histogenic layers, called L1, L2, and L3 [], and cells of all histogenic layers contribute to organogenesis [],[]. The L1 layer gives rise to the epidermis while the L2 and L3 layers contribute to internal tissues. In Arabidopsis ovules, for example, the integuments that eventually develop into the seed coat are entirely made up of L1-derived cells, while L2 cells generate the inner tissue [].Classic studies have demonstrated that meristematic layers communicate [],[], but it is only recently that the biological relevance and the molecular mechanisms are being elucidated []–[]. For example, work on the receptor-like kinase (RLK) BRASSINOSTEROID INSENSITIVE (BRI1) has provided evidence that the epidermis both promotes and restricts organ growth []. Furthermore, microsurgical experiments indicated that the epidermis also maintains cell division patterns in subtending layers []. These are but two examples that highlight the importance of the epidermis and inwards-oriented signaling in this inter-cell-layer cross-talk required for correct organ size and shape. At the same time, radial outward-oriented signaling also takes place during organogenesis. Known scenarios include transcription factors or small proteins that are synthesized in inner layers and move outwards into overlaying cell layers in a controlled fashion []–[]. The so far best-characterised case of such a movement underlies radial patterning of the root [],[]. In addition, the epidermally-expressed RLKs CRINKLY4 (CR4) from corn or its Arabidopsis homolog ACR4 are necessary for epidermis development and may receive signals from underlying cell layers []–[].Inter-cellular communication during floral morphogenesis in Arabidopsis also depends on signaling mediated by the leucine-rich repeat transmembrane receptor-like kinase (LRR-RLK) STRUBBELIG (SUB) []. Analysis of sub mutants indicated that SUB is required for proper shaping of floral organs such as carpels, petals and ovules. At the cellular level SUB participates in the control of cell shape and/or the orientation of the cell division plane in floral meristems and ovules. In addition, SUB, also known as SCRAMBLED (SCM), affects specification of hair cells in the root epidermis [],[]. Recent evidence suggests that the SUB protein is confined to interior tissues in floral meristems, developing ovules and young roots although SUB mRNA is monitored throughout those organs []. In particular functional SUB:EGFP fusion protein is absent from cells that show a mutant phenotype in sub mutants, but can either be found in adjacent cells, as in floral meristems and ovules, or in cells that are separated from mutant cells by two cell diameters, like in the root. The non-cell-autonomous effects of SUB were corroborated by an analysis of sub- plants expressing a functional SUB:EGFP transgene under the control of different tissue-specific promoters. Thus the data indicate that SUB undergoes posttranscriptional regulation, acts in a non-cell-autonomous fashion and mediates cell morphogenesis and cell fate across clonally distinct cell layers in an inside-out fashion []. The SUB protein is a member of the LRRV/STRUBBELIG-RECEPTOR FAMILY (SRF) family of receptor-like kinases [],[]. It is predicted to carry an extracellular domain with six leucine-rich repeats, a transmembrane domain, and a cytoplasmic intracellular domain with the juxtramembrane and kinase domains. Interestingly, phosphotransfer activity of the SUB kinase domain is not essential for its function in vivo [] and thus SUB seems to belong to the family of atypical or “dead” receptor kinases [],[].Very little is known regarding signaling through atypical receptor-like kinases in plants []. In addition, it remains to be understood how cellular morphogenesis is coordinated across cell layers []–[]. It is therefore of great interest to investigate the molecular basis of SUB signaling and function. Here we present the identification and analysis of three genetic factors that may relate to SUB signaling. Our results show that mutations in QUIRKY (QKY), ZERZAUST (ZET), and DETORQUEO (DOQ) result in a sub-like phenotype. Molecular cloning of QKY revealed that the predicted QKY protein is likely a transmembrane protein with four C2 domains indicating a role for QKY in Ca2+-dependent signaling. Global gene expression profiling of the mutants corroborates the morphological analysis but also suggests additional and distinct roles for each gene. Furthermore, the data indicate that SUB signaling plays previously unknown roles in cell wall and stress biology.ResultsIsolation of sub-Like MutantsWe applied a forward genetic approach to isolate additional factors of the SUB signaling pathway, based on the hypothesis that mutations in some of the genes that are part of the SUB pathway should result in sub-like (slm) mutant phenotypes. We thus screened M2 families of an ethylmethane sulfonate-mutagenized Ler population for slm mutants (see Materials and Methods). In this experiment we identified two new sub alleles [] as well as several novel mutants with sub-like phenotypes (Figures – ). These fell into three different complementation groups, which map to distinct positions on chromosome (Table ). We termed two of the genes DETORQUEO (DOQ) and ZERZAUST (ZET), respectively. DETORQUEO refers to a latin term that means “to twist out of shape”. ZERZAUST is a German term for “disheveled”. We also isolated three mutant alleles of QUIRKY (QKY), which plays a role in fruit dehiscence (L.F., unpublished results; S.J. and Martin F. Yanofsky, unpublished observations) and the numbering of qky alleles was coordinated. Thus, our genetic approach resulted in the identification of three loci, DOQ, QKY, and ZET, mutations in which result in a sub-like phenotype and that, together with SUB, define the STRUBBELIG-LIKE MUTANT (SLM) class of genes../journal.pgen..g001Figure 1Comparison of the overall above-ground morphology of sub-, doq-, qky-, and zet- mutants.The given stages correspond across equivalent panels. (A–E) Wild-type Ler. (F–K) sub-. (L–Q) doq-. (R–W) qky-. (X-Ac) zet-. (A, F, L, R, X) An open stage flower from a -day old plant. Note the misorientation of petals due to twisting in the basal end of the petal structure (arrows). (F, R, X) Petals can also show small notches. (B, G, M, S, Y) Top view of a -day inflorescence. (G, M, S, Y) Flower phyllotaxis is irregular. Arrows mark prematurely opened flower buds. (C, H, N, T, Z) Top view of a -day rosette. (H, N) Leaf petioles can be twisted (arrows). (N) doq- leaves have longer petioles and narrow blades (arrow heads). (D, I, O, U, Aa) Morphology of mature siliques. Typical twisting observed in (I) sub-, (U) qky- and (Aa) zet- mutants. doq- exhibits more subtle twisting in both (O) siliques and (P) stems. (E, J, P, V, Ab) A lateral view of a section of stems from a -day plant. (V) qky- and (Ab) zet- mutant show twisting in stems equivalent to (J) sub- plants. (K, Q, W, Ac) Plant height of mutants (left) in comparison to Ler (right). (K) sub- and (W) qky- mutants show a clearly reduced plant height. (Q) The doq- and (Ac) zet- mutants show only a slight reduction. Scale bars: (A, D, E, F, I, J, L, O, P, R, U, V, X, Aa, Ab) . mm, (B, C, G, H, M, N, S, T, Y, Z) mm, (K, Q, W, Ac) cm../journal.pgen..g002Figure 2Comparison of ovule morphology in sub-, doq-, qky-, and zet- mutants.(A) Wild-type Ler. The arrow marks one of the elongated cells of the distal outer integument. (B) sub-. A mild phenotype is shown. Note the irregular size and shape of cells at the distal outer integument (arrow heads, compare to (A)). (C, D) sub-. Strong phenotypes are depicted. Note the half-formed outer integument. (D) shows an example where the outer integument shows several gaps. (E) doq-. This mutant shows a mild ovule phenotype comparable to the one depicted in (B). The arrow heads highlight the disruption of the regular cell files of the distal outer integument. (F) qky-. Note the variability of the phenotype. The specimen to the left shows only a mild disorganisation of the cell shape (arrow head). The one to the right shows a strong phenotype (compare to (C)). (G) zet-. This mutant shows a variable phenotype. The ovule to the left shows gaps in the outer integument while the one to the right exhibits only mild alterations in cell shape and size (arrow head). Abbreviations: fu, funiculus; ii, inner integument; mp, micropyle; oi, outer integument. Scale bars: µm../journal.pgen..g003Figure 3Analysis of cellular defects in -day old main roots and stage floral meristems of sub-, doq-, qky and zet- mutants.(A–E) Expression of the GL2::GUS reporter in whole-mount main roots. (A) Wild-type Ler root. The reporter is detected in regular files of non-hair cells. (B–E) GL2::GUS reporter expression is patchy. (F–K) Mid-optical longitudinal sections through stage floral meristems obtained by confocal microscopy of propidium-iodide-stained specimen. (F) Wild-type Ler. Note the regular arrangement of cells in the L1 and L2. (G) sub-. The arrow marks a periclinal cell division event. (H) doq-. The region marked by the square (i) is shown at higher magnification in (Hi). The arrows highlight aberrant oblique and periclinal cell divisions in the L1 and L2, respectively. (I) qky-. The region marked by the square (i) is shown at higher magnification in (Ii). The arrow labels a periclinal cell division in the L2. (J–K) zet-. (J) The region marked by the square (i) is shown at higher magnification in (Ji). The arrows highlight periclinal cell divisions. A cell undergoing cell separation is indicated. (K) Disintegrating cells are marked. Abbreviations: cd, cell disintegration; cs, cell separation; L1, L1 cell layer; L2, L2 cell layer. Scale bars: (A–E) µm, (F–K) µm../journal.pgen..t001Table 1Summary of sub-like mutants identified.MutantGene Symbol slm identifierChromosomeBoundary markersMap interval (AGI)* detorqueo- DOQ slm8 1F10O3(48ID) . Mb–. MbNF21M122 zerzaust- ZET slm26 1CER453151(58ID) . Mb–. MbF13011(164ID) zerzaust- slm72 quirky- QKY slm3 .(RsaI) . Mb–. MbF25A4(BglII) quirky- slm17 quirky- slm38 *: Values represent approximate AGI positions of molecular markers listed.1Developed within this study.2Source: TAIR database and attributions therein.Comparison of sub-, doq-, zet-, and qky- PhenotypesWe identified one mutant allele of DOQ and two and three independent alleles of ZET and QKY, respectively (Table ). All sub, doq, zet and qky alleles were recessive and behaved in a Mendelian fashion (not shown). The various zet and qky mutants did not noticeably differ in their respective phenotypes and the three qky alleles are likely to be nulls (see below). Thus, zet- and qky- or qky- were used as reference alleles for further analysis. In addition, we used the well-characterised sub- mutant for comparison []. This mutation likely represents a null-allele since it results in a stop codon and a predicted shorter SUB protein lacking the transmembrane and intracellular kinase domains. Thus, it is expected that SUB-dependent signaling across the plasma membrane is blocked in sub- mutants.At the macroscopic level sub- mutants are known to be affected in several above-ground organs [] (Figures – ) (Tables , ). Inflorescences are characterised by reduced height, an irregularly twisted stem, and an aberrant phyllotaxis of flowers. Flowers open prematurely and show a large percentage of twisted and often notched petals. Furthermore, all flowers exhibit twisted carpels and about percent of sub- ovules showed aberrant initiation of the outer integument (Figure ) (Table ). This results in outer integuments with gaps that often resemble “multifingered clamps” or “scoops”. Also ovules with a fully developed outer integument show defects. In particular the distal or micropylar cells of the outer integument can show aberrant size and shape. In addition, about percent of sub- plants show at least one leaf per rosette with twisted petioles../journal.pgen..g004Figure 4Comparison of above-ground morphology of sub-, doq-, qky-, and zet- double mutants.The given stages are identical across panels. (A–F) sub- doq-. (G–L) sub- qky-. (M–R) sub- zet-. (S–X) doq- qky-. (Y-Ad) doq- zet-. (Ae–Aj) qky- zet-. (A, G, M, S, Y, Ae) An open stage flower from a -day old plant. Note the petal misorientation due to twisting along the longitudinal axis of petals. (M) sub- zet-, (S) doq- qky- and (Y) doq- qky- flowers show particularly aberrant perianth morphology, with thin petals and heavy notching evident. (B, H, N, T, Z, Af) Top view of a -day inflorescence. (B) sub- doq-, (H) sub- qky- and (N) sub- zet- inflorescences show aberrant phyllotaxy. (B, H, N, Z) Premature opening of flower buds is apparent. (C, I, O, U, Aa, Ag) A lateral view of a stem segment from a -day plant. (D, J, P, V, Ab, Ah) Morphology of a mature silique. Stem and silique twisting is exaggerated as compared to single mutant parentals, with the exception of the qky- zet- double mutant (Ag, Ah). (D, P, V, Ab) Note floral organs have not abscised. (E, K, Q, W, Ac, Ai) Top view of a -day rosette. Arrows indicate exaggerated petiole twisting in leaves. (Ai) qky- zet- leaf petioles are not twisted. (F, L, R, X, Ad, Aj) Plant height comparisons between Ler (left), single mutant parental (middle) and double mutant (right) -day plants. (F) sub- doq- and (L) sub- qky- mutants are significantly dwarfed, whereas (R) sub- qky- and (X) doq- qky- plants exhibit modest reductions in height. (Ad, Aj) doq- zet- and qky- zet- plant height is not significantly different from parental mutants. Scale bars: (A, C, D, G, I, J, M, O, P, S, U, V, Y, Aa, Ab, Ae, Ag, Ah) . mm, (B, H, N, T, Z, Af) mm, (E, K, Q, W, Ac, Ai) cm, (F, L, R, X, Ad, Aj) cm../journal.pgen..t002Table 2Ovule defects observed in sub and slm single and double mutants.GenotypeOvule DefectSamples (N)Enclosed structure<% oi absent≥% oi absentSevere malformationDefective ii sub- .%.%.%%% doq- .%.%%%% qky- .%.%.%%% zet- .%.%.%%<% sub- doq- .%.%.%%.% sub- qky- .%.%.%.%.% sub- zet- .%.%.%.%.% doq- qky- .%.%.%%.% doq- zet- .%.%.%.%.% qky- zet- .%.%.%%.%216oi, outer integument; ii, inner integument../journal.pgen..t003Table 3Cellular defects in sub and sub-like mutant floral meristems.Periclinal cell division defectsOther cellular defectsMutantL1L2CW sep. Cell dis. Samples (N)Ler %.%%% sub- .%.%%% doq- .%.%%% qky- .%.%%% zet- %.%.%.%641Cell wall separation.2Cell disintegration.At the cellular level sub- exhibits aberrant cell shape and robustly scorable numbers of periclinal, rather than anticlinal cell division planes in cells of the L2 layer of stage floral meristems [] (Figure ) (Table ). Interestingly, in this analysis we could also observe a previously unnoticed low number of periclinal divisions in the L1 of sub- floral meristems. Furthermore, sub/scm mutants develop root hairs at discordant positions in the epidermis [],[]. This defect in root hair patterning can be followed by expressing the bacterial β-glucuronidase gene under the control of the Arabidopsis GLABRA2 (GL2) promoter (GL2::GUS) [],[] (Figure ). This reporter conveniently labels the regular files of non-hair cells in the epidermis of wild-type roots and exhibits an irregular expression pattern in sub/scm mutants.Upon examination, many aspects of the phenotypes of sub- and the other slm mutants were comparable. The doq-, qky- and zet- mutants showed aberrant floral phyllotaxis and flowers with twisted petals (Figure ). In addition, we observed premature floral bud opening in doq- and zet-. Petals of zet- mutants showed notches similar to sub- petals but zet- floral organs were generally more misshapen. In rosettes, nearly all plants showed examples of leaf petiole twisting. In particular, qky- mutants showed leaf twisting comparable to sub-, whereas doq- rosette leaves showed in addition elongated petioles and narrow blades. In contrast, zet- rosette leaves did not show major defects at the gross morphology level. Irregular twisting of siliques and stems was apparent in doq-, qky- and zet- mutants, although the twisting in doq- siliques and stems was more subtle. Plant height was most affected in qky- while doq- and zet- showed only a slight reduction. Mature ovules of doq- mutants were mildly but consistently affected (Figure ) (Table ). The outer integument did not fully extend to the funiculus (reduced campylotropy). In addition, its distal cells were shorter and of irregular shape. Mature ovules of qky- and zet- closely resembled ovules of sub- mutants with respect to gaps and altered cell shape in the outer integument. Interestingly, zet- showed a slightly higher percentage of malformed outer integuments. As was the case for sub [] the inner integument in all other slm single mutants appeared to be unaffected.Further, doq-, qky- and zet- mutants were investigated for cellular defects in the root epidermis and in L1/L2 cells of stage to floral meristems (Figure ) (Table ). In day old main roots, all three mutants showed misregulation of GL2::GUS reporter expression comparable to sub-, indicating that root hair patterning was similarly affected. The three mutants also exhibited cell shape and periclinal cell division plane defects in the L1 and L2 cells of floral meristems. Floral meristems of doq- showed a higher percentage of those defects while in zet-, cell separation and disintegration could also be observed. The latter finding indicates that ZET is also required for cell viability.Taken together the analysis of the slm single mutants revealed that there is a large functional overlap of the corresponding genes and that individual SLM genes participate in subsets of SUB-dependent processes, such as ovule development, cellular behavior of L2 cells of stage floral meristems, and root hair patterning. The results also suggest, however, that SLM genes have additional functions unrelated to each other. For example, ZET has a particular function in cell survival in floral meristems as has DOQ in the regulation of leaf shape.Double Mutant AnalysisTo investigate further the genetic relationship between SLM genes we generated all possible double-mutant combinations and analysed the respective mutant phenotypes. The results are summarized in Figures and and Table ../journal.pgen..g005Figure 5Comparison of ovule morphology in different slm double mutants.SEM micrographs of stage--V ovules are depicted. (A–C) sub- doq-. (D–F) sub- qky-. (G–I) sub- zet-. (J–L) doq- qky-. (M–O) doq- zet-. (P,Q) qky- zet-. Each double mutant shows a range of ovule phenotypes. (A, D, J, M and P) Mild phenotypes are depicted. Note the irregular size and shape of cells at the distal outer integument (arrow heads). (B, C, D, G K, N Q) Stronger ovule phenotypes are observed whereby specimens have half formed outer integuments. (C, F, I, L, O) In extreme cases, sub- doq-, sub- qky-, sub- zet-, doq- qky- and doq- zet- ovules can show malformation of the inner integument. (C) This specimen is an example where the inner integument shows several gaps. (F, I, L, O) Inner integuments can also appear as finger-like protrusions. The qky- zet- double mutant is an exception, whereby inner integument defects are rarely observed. (E, H) Severely malformed ovules can also appear as integument-like and nucellus-like structures emerging from placental tissue. (R) Fertility was assayed as the average seed number per silique (n = ). Abbreviations: fu, funiculus; ii, inner integument; mp, micropyle; oi, outer integument. Scale bars: µm. sub- doq- Petals of sub- doq- double mutants mostly resembled doq- petals (Figure 4A), however, stem twisting was more similar to sub-. Other aspects of the sub- doq- phenotype, such as silique twisting, plant dwarfism and rosette leaf petiole twisting were more exaggerated when compared to either single mutant. Ovules of sub- doq- plants showed an increase in outer integument defects, although not as strong as in some other double mutant combinations. In addition, about % of ovules of sub- doq- plants showed defects in inner integument morphology (Figure 5C) (Table ), with gaps of variable sizes and finger-like protrusions. The fertility of sub- doq- plants was severely reduced (Figure 5R). sub- qky- Petal and carpel twisting in sub-l qky- double mutants was similar to that shown by each single mutant (Figure 4G). In contrast, twisting of siliques, stems, and leaf petioles was more pronounced as was the reduction in plant height (Figure 4I–L). In addition, ovule development was more heavily affected compared to the single mutants (Figure 5D–F) (Table ), with % of ovules showing inner integument defects and a corresponding reduction in fertility (Figure 5R). About % of ovules were hardly recognizable as such, but rather resembled a mass of cells with integument-like outgrowths (Figure 5E). sub- zet- Overall, the sub- zet- double mutants showed the most disturbed morphology (Figure 4M–R). Perianth organs were twisted, narrower and notched, while carpels were heavily twisted. As a result the overall structure of the flower was irregular. The stem phenotype was knotted rather than twisted. Leaf morphology was misshapen with more pronounced leaf petiole twisting and plants showed prominent dwarfism. Consistent with this exaggerated phenotype ovules of sub- zet- plants showed severe defects with % of ovules resembling a mass of cells with integument-like outgrowths (Figure 5H) (Table ). About % of sub- zet- ovules exhibited short gapped outer integuments and similarly malformed inner integuments (Figure 5I). Again, fertility was reduced in sub- zet- plants (Figure 5R). doq- qky- Perianth organs and carpels in doq- qky- mutants were more twisted as compared to the parental lines (Figure 4S). Twisting of leaf petioles, stems, and siliques was also more pronounced as was dwarfism. Ovules of doq- qky- mutants showed a less exaggerated phenotype compared to some of the other double mutant combinations. Still, a large proportion of doq- qky- ovules showed gaps in the outer integument (Figure 5J–L) (Table ) and aberrant inner integument morphology was seen in % of ovules (Figure 5L). Fertility was strongly reduced in doq- qky- double mutants (Figure 5R). doq- zet- Perianth morphology in doq- zet- mutants was about equivalent to that observed in sub- zet- and doq- qky- flowers (Figure 4Y). Twisting of stems appeared slightly more pronounced compared to each single mutant while siliques were drastically more twisted. Plant height was comparable to zet- single mutants. Overall a higher percentage of doq- zet- ovules showed gaps in the outer and inner integuments (Figure 5M–O), with % of ovules exhibiting severe malformations comparable to sub- zet- mutants (Table ). The reduction in fertility was comparable to that of doq- qky- double mutants (Figure 5R). qky- zet- The qky- zet- double mutants largely phenocopied zet- single mutants in above-ground morphology (Figure 4Ae–Aj) including ovules (Figure 5P,Q) (Table ). One exception to this was leaf petiole twisting as this aspect was most similar to the phenotype in qky- single mutants.In summary, the pleiotropic phenotypes of slm single and double mutants complicated the double mutant analysis. Although an exaggerated phenotype was often observed in a double mutant combination it was usually difficult to decide whether a double mutant displayed an additive or synergistic phenotype, or whether a particular mutation was epistatic to another. Overall, the double mutant analysis reinforced the notion that SLM genes likely do not act in a single linear pathway but have both overlapping and separate functions.Phenotypic Analysis of doq, qky, and zet Plants Carrying a 35S::SUB TransgeneTo further explore the genetic relationship between SLM genes we first tested whether SUB expression was responsive to other genes of this group. We investigated SUB expression in inflorescence apices and stage – flowers (see below) from several slm mutants by quantitative real time PCR (qRT-PCR). As can be seen in Figure 6A only very moderate changes in SUB expression were detected with a slight but consistent reduction of SUB expression in flowers of doq-, qky- and zet- mutants. In contrast, SUB transcript levels were unaltered in those tissues and mutants when assessed by a transcriptome analysis (see below). Given that SUB is expressed at very low levels to begin with [] we reasoned that the observed mild effects may not be the result of direct effects but rather be a consequence of indirect influences due to tissue sampling or the altered morphology of the mutants. To test the relevance of the effects we asked whether rendering SUB independent of its normal transcriptional regulation, by ectopically expressing SUB using the cauliflower mosaic virus 35S promoter, could result in phenotypic rescue of doq, qky and zet mutants. We analysed the phenotypes of 35S::SUB doq-, 35S::SUB qky- and 35S::SUB zet- plants (Figure 6B). In particular we scored stem twisting, flower morphology and silique twisting of individual transgenic T1 lines. Transgene functionality was demonstrated by the wild-type appearance of 35S::SUB sub- plants. Despite demonstrating comparable transgene expression (Figure 6C), the other tested combinations did not show phenotypic rescue (35S::SUB doq-: T1 lines scored, 35S::SUB qky-: >, and 35S::SUB zet-: >), indicating that SUB does not act as a downstream target gene of DOQ, QKY or ZET. Taken together our data suggest that SUB is not directly regulated at the transcriptional level by the other SLM genes../journal.pgen..g006Figure 6Analysis of SUB expression in the slm plant lines.(A) Quantitative expression analysis of SUB in the slm mutants. Steady-state mRNA levels were measured in apex and flower tissues by quantitative real-time PCR. UBC21 expression was used for normalization. (B) Phenotypes of 35S:SUB slm transgenic lines. A 35S::SUB:myc construct was transformed into each slm mutant background. T1 plants were analysed for rescue of flower, silique and stem mutant phenotypes. Note only sub- mutants exhibit rescue by SUB overexpression. (C) Transgenic 35S:SUB:myc T1 plants represented in (A) were assayed for SUB transgene expression using semi-quantitative RT-PCR (top panels). GAPC [] served as internal control (bottom panels). Scale bars: . mm.Genome-Wide Transcriptome Analysis and Global Expression ProfilingTo shed light on the molecular nature of SUB-dependent processes and to systematically investigate the similarities between the SLM genes at the mechanistic level, we applied a transcriptome analysis. Since mutants with defects in genes of overlapping functions should also exhibit overlapping alterations in transcript profiles we compared wild type (Ler) with plants of two sub alleles , as well as doq-, zet- and qky- mutants using the Affymetrix ATH1 GeneChip platform (see Materials and Methods). The analysis of two independent sub alleles was aimed at obtaining a robust set of SUB-responsive genes for the different pairwise comparisons and the list of SUB-responsive genes represents the overlap of misexpressed genes in sub- and sub- mutants []. Since we had observed tissue-specific differences in the phenotypes of slm mutants, which suggested tissue-specific sub-functions for the corresponding genes, we tried to capture these differences also in the sampling for transcriptome analysis. To this end, we sampled inflorescence apices plus flowers up to stage (apex data set), representing mostly proliferative tissues, while stage – flowers were collected to represent maturation stages (flower data set). The samples thus covered SUB-related aspects such as the control of cell shape/division plane in L2 cells of floral stage meristems, and the regulation of petal, carpel and ovule development.Since traditional analysis tools to identify differential gene expression in whole genome transcriptome data are not well suited for comparisons of multiple samples we developed a meta analysis tool based on Z-score statistics (see Materials and Methods). It offers a variety of benefits for the simultaneous analysis of multiple pairwise comparisons, such as normalization for the overall biological effect observed in the individual experiments, increased sensitivity, and the possibility of querying data using Bolean logic. Using this tool we identified and significantly misexpressed genes, respectively, in the apex and flower samples of sub plants (Table ). The other mutants were characterised by higher numbers of misexpressed genes (Table ) indicating that DOQ, QKY, and ZET affect more processes than SUB. More importantly, we systematically analyzed the overlap of misexpressed genes from individual slm mutants compared to wild type. The results of all pairwise comparisons are given in Table . For example, a pairwise comparison between sub and doq- in the apex sample revealed a gene overlap. This corresponded to % of genes misexpressed in sub and about % of genes aberrantly expressed in doq-. Comparable values were observed for sub versus qky- and sub vs zet- comparisons. The overlap with SUB-dependent genes was even more pronounced in the flower data set and in the sub-doq- comparison we found common genes, representing % and % of the genes misexpressed in sub and doq-, respectively. The sub qky- and sub zet- comparisons revealed higher values with % and % of SUB-responsive genes being found to overlap. Thus, a significant number of SUB-responsive genes are also sensitive to DOQ, QKY and ZET function. Similar pairwise comparisons were made between doq-, qky- and zet- (Table ). Again there was considerable overlap of misexpressed genes between different mutants. Most strikingly, in stage – flowers % and % of the QKY-responsive genes are also misregulated in doq- and zet-, respectively. This finding implies that the activity of more than two-thirds of QKY-responsive genes also depends on DOQ and ZET function in stage – flowers. Taken together the microarray data corroborate the morphological analysis of the single mutants and are compatible with the notion that the different SLM genes share overlapping functions../journal.pgen..t004Table 4Genes misexpressed in sub and slm mutants.Mutant profiles (Apex)Genes1 Mutant profiles (Flower)Genes2 sub sub doq- doq- qky- qky- zet- zet- 4971Genes misexpressed in apex tissues sampled (up to flower stage ).2Genes misexpressed in flowers (stages –)../journal.pgen..t005Table 5Comparison of genes co-misexpressed in sub and slm mutants.Apex Overlap profiles1 Genes% sub % slm Flower Overlap profiles1 Genes% sub % slm3 sub_ doq- .%.% sub_ doq- .%.% sub_ qky- .%.% sub_ qky- .%% sub_ zet- .%.% sub_ zet- .%.%1Denotes a comparison (overlap profile) made between two mutant transcriptome profiles.2Percentage of sub misexpressed genes.3Percentage of slm misexpressed genes represented in listed comparisons.Percentages are calculated based on mutant gene list totals (Table )../journal.pgen..t006Table 6Comparisons of genes co-misexpressed in slm mutants.Apex Overlap profiles1 Genes% slm % slm Flower Overlap profiles1 Genes% slm % slm doq-1_ qky- .%.% doq-1_ qky- .%.% doq-1_ zet- .%.% doq-1_ zet- .%.% qky-8_ zet- .%.% qky-8_ zet- .%.%1Denotes a comparison (overlap profile) made between two mutant transcriptome profiles.,3Percentage of slm misexpressed genes (2first appearing, 3second appearing) in listed comparisons.Percentages are calculated based on mutant gene list totals (Table ).To further study the complexity of the functional relationship among SLM genes we carried out additional pairwise comparisons, examined the identity of genes present in the overlap lists, and identified transcripts which are misregulated in all four mutants (i.e., a sub, doq-, qky- and zet- “quadruple” overlay) (Figures ,; Tables ,; Datasets S1, S2). While doing so we uncovered striking trends in the transcriptome data. For example, we found that more than % of the misregulated genes found in any sub/slm overlap identified in the apex samples are indeed shared by all mutants (Figure 7A). More significantly, of the transcripts identified as common SLM responsive in the apex were consistently elevated in expression in all mutants, suggesting that a core function of the SLM genes is to repress this set of genes in proliferating tissue (Figure 7B). Strikingly, this trend was reversed in differentiating tissue represented by the floral samples. Here the misregulated genes shared by all mutants made up more than % of transcripts identified in pairwise comparisons (Figure 7A) and of the genes were reduced in expression (Figure 7B). In addition to supporting the notion of functional overlap between the SLM genes, this analysis also demonstrated that diversification of SLM function observed in different tissues is based on a dramatic switch in the underlying molecular mechanism. While signaling through SLM proteins in proliferating tissue mainly causes repression of target gene transcription, SLM mediated signals lead to a transcriptional activation of downstream genes in differentiating cells, suggesting that SUB and other SLM proteins interact with a radically different regulatory environment. This notion is supported by another striking observation: while we found strong similarities between different genotypes within one tissue we observed very little overlap in misregulated genes between the same genotype in the two tissues analysed (Figure ). For example, only four misregulated transcripts were shared in apex and flower samples of sub mutants. Similar observations were made for the other single and pairwise comparisons. In addition, the core target gene sets of apex and flower, derived from quadruple comparisons, did not share any transcripts../journal.pgen..g007Figure 7Comparison of gene expression changes between sub and slm mutants.(A) Venn diagrams showing the overlap of genes detected as significantly misexpressed (z-score<− or >) in sub, doq-, qky- and zet- mutants. Analyses were performed using data generated from apex (left) and late-stage flower (right) probe sets. Gene lists that compared sub to each independent slm mutant were initially generated (numbers in boxes) and then overlapped for a final comparison (quadruple overlap set). (B) Line graphs depicting the expression profiles of genes present in the quadruple overlap set across sub and slm mutants in apex (left) and flower (right) tissues. Note how the up- or downregulation of individual genes coincides across all mutant genotypes../journal.pgen..g008Figure 8Comparison of genes misexpressed in apex and flower overlap sets.(A) Venn diagram depicting the overlap of genes misexpressed in apex and flower quadruple overlap sets. (B) Box diagram detailing the overlap of genes misregulated in apex and flower gene sets for individual sub and slm comparisons.To investigate the nature of SLM-dependent processes we assessed the known or predicted functions of the apex and flower core sets of SLM-responsive genes by searching The Arabidopsis Information Resource (TAIR), the literature and making use of the AtGenExpress expression atlas [] with the help of the AtGenExpress Visualisation Tool (AVT, http://jsp.weigelworld.org/expviz/) (Tables , ). In both core sets we observed a strong representation of genes that are involved in processes, such as cell wall biosynthesis and function, hormone signaling and abiotic and biotic stress responses. For example, two genes in the apex core set encode the inositol oxygenase family enzymes MIOX2 and MIOX4 required for biosynthesis of uridine-diphospho-glucuronic acid (UDP-GlcA), a precursor for various cell-wall matrix polysaccharides []. In addition, GERMIN-LIKE PROTEIN (GLP6) is a member of a gene family encoding putative extracellular GERMIN-LIKE proteins [] some of which play a role in biotic stress responses and cell wall biology. The flower core set of SLM-responsive genes also included several genes coding for cell wall proteins, such as ECS1 [], At2g05540 (Glycine-rich protein), At5g03350 and At5g18470 (lectin family proteins), and At2g43570 and At4g01700 (chitinases). In another example that relates to hormone signaling, five of the upregulated apex core set genes were inducible by jasmonates (JA) while the flower core set was characterized by a large group of genes that relate to different aspects of signaling mediated by salicylic acid (SA) (see Discussion). SLM-responsive genes whose expression is sensitive to JA encoded for example JAZ1, a central negative regulator of JA signaling [],[], ERD5 a mitochondrial proline dehydrogenase (ProDH) [], AtLEA5/SAG21 [],[], and AtTTPG, a class III trehalose--phosphate phosphatase involved in the biosynthesis of trehalose [],[]. All of these genes are believed to contribute to the plant's response to dehydration, cold and oxidative stress [], [], []–[]. We also observed SLM-dependency of genes involved in the homeostasis of the auxin indole--acetic acid (IAA) and the production of the Arabidopsis phytoalexin camalexin and indole glucosinolates. Glucosinolate metabolites play important roles in plant nutrition, growth regulation and the interactions of plants with pathogens and insect herbivores [],[]. In the apex core set we identified IAA-LEUCINE RESISTANT-LIKE GENE (ILL6), which belongs to a gene family encoding IAA amino acid conjugate hydrolases involved in the release of the active auxin from corresponding inactive amino acid conjugates [],[]. CYP83B1 and CYP79B2 encode cytochrome P450 monooxygenases that regulate the pool of indole--acetaldoxime (IAOx) []–[] a key branching point of primary and secondary metabolic pathways and a precursor of indole glucosinolates, camalexin [], and IAA (reviewed in [],[]). In addition, the flower core set contained two downregulated cytochrome P450 genes, CYP71A13 and CYP71B15, which promote distinct steps in camalexin biosynthesis []–[]../journal.pgen..t007Table 7List of genes co-misexpressed in slm mutant apex tissues.Gene/KeywordGene IdentifierDescription1 Genes upregulated MIOX2 At2g19800Myo-inositol oxygenase family gene / biosynthesis of nucleotide sugar precursors for cell-wall matrix MIOX4 At4g26260Myo-inositol oxygenase family gene / biosynthesis of nucleotide sugar precursors for cell-wall matrix GLP6 At5g39100Germin-like family protein / cell wall biology, pathogen response, nutrient reservoir JAZ1 At1g19180JASMONATE-ZIM-DOMAIN PROTEIN / nuclear-localized protein negatively regulating jasmonate signaling, inducible by JA ERD5 At3g30775EARLY RESPONSIVE TO DEHYDRATION . Proline dehydrogenase / osmotic stress responsive, inducible by JA ATSIP2 At3g57520ARABIDOPSIS THALIANA SEED IMBIBITION / glycosyl hydrolase, catabolism of raffinose family oligosaccharides, inducible by JA AtLEA5 At4g02380LATE embryogenesis abundant-like protein / response to cold, oxidative stress, reactive oxygen species, inducible by JA AtTTPG At4g22590Trehalose--phosphate phosphatase / trehalose biosynthetic process, sugar-based signaling, inducible by JA ILL6 At1g44350AA-LEUCINE RESISTANT (ILR)-LIKE GENE / IAA-amino acid hydrolase CYP79B2 At4g39950Cytochrome P450 family/tryptophan metabolism / converts Trp to indo--acetaldoxime (IAOx), a precursor to IAA, camalexin and indole glucosinolates CYP83B1 At4g31500Cytochrome P450 monooxygenase / biosynthetic pathway of indole glucosinolatesNodulin MtN21-likeAt2g39510Nodulin MtN21 family protein/function unknown.UnknownAt1g74055Unknown protein Genes downregulated AtDof5. At5g66940Dof-type zinc finger domain-containing protein / putative transcription factor, possibly involved in vascular developmentGene keywords and gene identifiers are shown.1Gene descriptions were derived from TAIR and attributions therein../journal.pgen..t008Table 8List of genes co-misexpressed in slm mutant flowers.Gene/Keyword1 Gene IdentifierDescription2 Genes upregulated ATHB40 At4g36740Homeodomain leucine zipper class I (HD-Zip I) protein/ transcription factor GLTP3 At3g21260GLYCOLIPID TRANSFER PROTEIN 3Glycine-richAt2g05540Glycine-rich proteinLactoylglutathione lyaseAt1g15380Lactoylglutathione lyase family protein SUS3 At4g02280sucrose synthase /putative sucrose synthase activity Genes downregulated AGP5 At1g35230Arabinogalactan-protein AIG1 At1g33960AVRRPT2-INDUCED GENE / exhibits RPS2- and avrRpt2-dependent induction early after bacterial infection ANAC036 At2g17040Arabidopsis NAC domain containing protein / transcription factor ATARD3 At2g26400acireductone dioxygenase (ARD/ARD')family protein , putative ATGH9B17 At4g39000ARABIDOPSIS THALIANA GLYCOSYL HYDROLASE 9B17/ O-glycosyl hydrolase activity ATNUDT6 At2g04450ARABIDOPSIS THALIANA NUDIX HYDROLASE HOMOLOG / DNA metabolism, putative pathogen response protein COBL5 At5g60950COBRA-LIKE PROTEIN PRECURSOR/ phytochelatin synthetase, putative CYP71A13 At2g30770CYTOCHROME P450 MONOOXYGENASE 71A13/IAA and camalexin biosynthetic process, pathogen defence response CYP71B15 At3g26830CYTOCHROME P450 FAMILY, 71B15/camalexin biosynthesis CYP96A2 At4g32170CYTOCHROME P450 FAMILY, 96A2 ECS1 At1g31580ECOTYPE SPECIFIC / plant cell wall-associated protein. EDS1 At3g48090ENHANCED DISEASE SUSCEPTIBILITY / SA-mediated signaling, defense against pathogens EDS5/SID1 At4g39030ENHANCED DISEASE SUSCEPTIBILITY / MATE-transporter family protein / SA biosynthesis, defense against pathogens EXP3 At2g18660Expansin family proteinABAAt5g13200GRAM domain-containing protein/ABA-responsive protein-related, function unknownAnkyrinAt5g54610Ankyrin repeat family proteinAuxin-responseAt5g35735Auxin-responsive family proteinCalcium-ion bindingAt3g47480Calcium-binding EF hand family proteinCalcium-ion bindingAt5g39670Calcium-binding EF hand family proteinCalcium-ion bindingAt3g50770Calmodulin-like protein , predicted calcium binding activityChitinaseAt2g43570Chitinase protein, putativeChitinaseAt4g01700Chitinase protein, putativeDEDOL-PPAt5g58782Dehydrodolichyl diphosphate synthase (DEDOL-PP), putativeDELTA-VPEAt3g20210Delta vacuolar processing enzyme/seed coat developmentDisease resistanceAt1g57650Disease resistance protein RPP1-WsA, putativeDisease resistanceAt1g72900Disease resistance protein (TIR-NBS class), putativeDisease resistanceAt1g72920Disease resistance protein (TIR-NBS class), putativeGlucosidaseAt3g03640Beta-glucosidase GRX480 At1g28480Glutaredoxin family protein / regulates protein redox state, JA-mediated signaling pathway, SA-mediated signaling pathwayHIN1-relatedAt1g65690Harpin-induced protein-related/HIN1-related protein ICS1/SID2/EDS16 At1g74710isochorismate synthase / salicylic acid biosynthetic processLectin familyAt5g03350Legume lectin family protein, putativeLectin familyAt5g18470Curculin-like (mannose-binding) lectin family proteinLipaseAt5g55050GDSL-motif lipase/hydrolase family proteinLipid transfer proteinAt3g22600Protease inhibitor/seed storage/lipid transfer protein (LTP) family proteinLipid transfer proteinAt5g55450Protease inhibitor/seed storage/lipid transfer protein (LTP) family proteinLRR proteinAt2g24160Leucine rich repeat protein family, putative disease resistance proteinMethionine metabolismAt4g21850Methionine sulfoxide reductase domain-containing proteinMIF proteinAt3g51660Macrophage migration inhibitory factor (MIF) family protein NIMIN1 At1g02450Transcriptional regulator/ negative regulator of SA-dependent signal transduction, modulates PR gene expression NRAMP5 At4g18790NRAMP METAL ION TRANSPORTER 5OxidoreductaseAt5g245302OG-Fe(II) oxygenase family protein PBP1 At5g54490PINOID-BINDING PROTEIN / responsive to auxin stimulus PBS3/GDG1 At5g13320AVRPPHB SUSCEPTIBLE / Required for accumulation of salicylic acid, activation of defense responses and resistance PCC1 At3g22231PATHOGEN AND CIRCADIAN CONTROLLED /regulated by the circadian clock, responsive to pathogensPlastocyanin domainAt4g30590Plastocyanin-like domain-containing protein PR1 At2g14610PATHOGENESIS-RELATED GENE / pathogen defense response, SAR, response to vitamin B1 PR2 At3g57260PATHOGENESIS-RELATED GENE / glycosyl hydrolase family protein, cold responsive, SAR PR4 At3g04720PATHOGENESIS-RELATED GENE / hevein-like protein precursor, responsive to virus and ethylene stimulus, SAR PR5 At1g75040PATHOGENESIS-RELATED GENE , thaumatin-like protein/ SAR, response to UV-B, regulation of anthocyanin biosynthetic processProteaseAt3g51330Aspartyl protease family proteinProtein bindingAt1g10340Ankyrin repeat family proteinProtein bindingAt4g14365Zinc finger (C3HC4-type RING finger) family protein/ankyrin repeat family proteinProtein kinaseAt4g11890Protein kinase family proteinProtein kinaseAt4g23150Protein kinase family proteinReceptorAt2g37710Receptor lectin kinase/ responsive to SA stimulus RLK1 At5g60900RECEPTOR-LIKE PROTEIN KINASE RLK6 At4g23130RECEPTOR-LIKE PROTEIN KINASE SAL2 At5g640003′(′),′-bisphosphate nucleotidaseSelf-incompatibilityAt1g26795Self-incompatibility protein-related SIB1 At3g56710SIGMA FACTOR BINDING PROTEIN TAT3 At2g24850TYROSINE AMINOTRANSFERASE / responsive to wounding and JA treatmentTransporterAt3g10780emp24/gp25L/p24 family protein / putative transporter protein UBC31 At1g36340UBIQUITIN-CONJUGATING ENZYME WAK1 At1g21250WALL ASSOCIATED KINASE / response to salicylic acid stimulus WAKL8 At1g16260WALL ASSOCIATED KINASE-LIKE WRKY33 At2g38470WRKY-type DNA binding protein/ defence response to bacterium and fungus WRKY46 At2g46400WRKY-type DNA binding protein WRKY70 At3g56400WRKY-type DNA binding protein / transcriptional repressor, JA mediated signaling pathway, SA mediated signaling pathway, SAR WRKY53 At4g23810WRKY-type DNA binding protein/ positive regulation of transcription, regulation of defense response WRKY38 At5g22570WRKY-type DNA binding protein / defence response, SAR ZAT10 At1g27730Salt tolerant zinc finger / transcriptional repressor activity, cold, water deprivation, salt stress, wounding, ABA and chitin responsiveUnknownAt1g19020Expressed proteinUnknownAt3g60420Expressed proteinUnknownAt5g52975Expressed proteinUnknownAt1g13470Hypothetical proteinUnknownAt1g52970Hypothetical proteinUnknownAt1g56060Hypothetical proteinUnknownAt3g13950Hypothetical proteinUnknownAt3g14850Hypothetical proteinUnknownAt3g48640Hypothetical proteinUnknownAt3g48650Hypothetical protein, predicted pseudogeneUnknownAt4g23610Hypothetical proteinUnknownAt5g19240Hypothetical proteinUnknownAt2g18680Unknown proteinUnknownAt2g14560Unknown proteinUnknownAt1g65500Unknown protein1Gene keywords and gene identifiers are shown.2Gene descriptions were derived from TAIR and attributions therein.ABA, abscisic acid; DR, disease resistance; JA, jasmonic acid; PR, pathogenesis-related; SA, salicylic acid; SAR, systemic acquired resistance.Molecular Identification of QKY As a first step to molecularly define the SLM pathway, the QKY gene was identified by map-based cloning (Figure and Materials and Methods). In all three EMS-induced qky alleles we could identify single point mutations in At1g74720 (Figure 9A). Further, two distinct T-DNA insertions in this locus result in qky mutants phenocopies (not shown). Combined this demonstrates that At1g74720 is QKY. Sequence analysis predicted that QKY contains no introns and encodes a transmembrane protein of amino acids with a calculated molecular mass of . kDa harboring four C2 domains (Figure 9B). In addition, two transmembrane domains are embedded in a plant-specific phosphoribosyltransferase C-terminal region (PRT_C, PFAM identifier PF08372), which according to the PFAM database occurs characteristically at the carboxy-terminus of phosphoribosyltransferases and phosphoribosyltransferase-like proteins []. Interestingly, this domain often appears together with C2-domains, as in QKY. A related domain topology, albeit with only three C2 domains, is found in several proteins present in humans, Drosophila melanogaster and Caenorhabditis elegans (C. elegans), collectively referred to as multiple C2 domain and transmembrane region proteins (MCTPs) []. The role of animal MCTPs is not well defined despite the fact that in C. elegans a single MCTP gene is present and essential for embryo viability []. QKY represents the first described plant MCTP-class gene. In accordance with proposed transmembrane topography for MCTPs the QKY C2 domains most likely have an intracellular localisation as QKY lacks a distinguishable N-terminal signal peptide and all known C2 domains are cytoplasmic. C2 domains are autonomously folding modules and form Ca2+-dependent phospholipid complexes, although some exceptions are known that do not bind Ca2+ or phospholipids or both []–[]. Other animal proteins with multiple C2 domains, but only one transmembrane region, include the synaptotagmins [], the extended synaptotagmins [], and the ferlins []. Synaptotagmins and ferlins are known to play a role in membrane trafficking. Although there is general resemblance in domain topology, very little primary sequence conservation is observed between QKY and the animal MCTPs, synaptotagmins, and ferlins../journal.pgen..g009Figure 9Molecular characterization of QKY.(A) Molecular organization of the QKY locus on chromosome . The horizontal black line highlights genomic DNA. The red arrow indicates the orientation and extent of the QKY open reading frame. The small black boxes mark the ′ and ′ untranslated regions, respectively. The orange arrow and box denote the end of the preceding and the start of the next annotated open reading frames, respectively. The positions of the various point mutations or T-DNA insertion sites are marked. (B) Predicted domain topology of the QKY protein. The sequence and numbering of the C2 domains, and the altered C-termini of the mutant protein variants, are given. The italic capital letters represent the de novo residues present in the predicted mutant QKY proteins from the two T-DNA lines. Stars mark the premature stops in the predicted mutant proteins. The green box denotes the plant phosphoribosyltransferase C-terminal region (PRT_C). Abbreviations: aa, amino acids; N, amino terminus; C, carboxy terminus; TM, transmembrane domain.The qky-, qky-, and qky- alleles likely cause a complete loss of QKY function. All three mutations are predicted to introduce stop codons leading to shorter proteins with variable numbers of C2 domains but always lacking the two transmembrane domains (Figure ). Thus, the three mutants likely contain truncated QKY variants that are not properly located to the membrane. Membrane localisation, however, seems essential for QKY function as all three mutants show identical phenotypes, regardless of the number of C2 domains still present in the predicted truncated proteins. In summary, the results suggest that QKY is a membrane-bound Ca2+-signaling factor.Discussion DOQ, QKY, and ZET Contribute to SUB-Dependent Organ DevelopmentIn this work we set out to genetically identify additional components of the SUB-dependent mechanism regulating ovule development and floral morphogenesis. To this end we identified second-site mutations that resulted in a sub-like phenotype, reasoning that such mutations should reside in genes that either act directly in the SUB signaling pathway, or, that SUB and these genes affect parallel pathways that converge on a common molecular or developmental target. A defined set of criteria was applied in the identification of candidate mutants. At the developmental level, candidates were chosen that showed a sub-like ovule phenotype and twisting of petals, carpels, siliques and the stem. In addition, they were analysed for two types of sub-like cellular defects: an increased frequency of periclinal cell divisions in the L1/L2 layers of stage floral meristems and a defect in root hair patterning. Mutations in three genes, DOQ, QKY, and ZET, satisfied all criteria and were considered as sub-like mutants.Although slm mutants were phenotypically very similar there were some noteworthy distinctions. For example, doq- was generally characterised by a weak ovule phenotype and mild silique/stem twisting. At the same time, doq- showed the highest frequency of periclinal cell divisions in the L1/L2 layers of stage floral meristems. Furthermore, rosette leaves of doq- plants had elongated petioles and narrow leaf blades, features not observed in other slm mutants. These findings indicate that DOQ is less important for ovule, silique and stem development compared to other SLM genes, but plays a more prominent role in the regulation of cell shape/division plane in the young floral meristem and leaf development. In contrast ZET does not seem to be involved in leaf development as zet mutants exhibited apparently normal leaves. This clearly distinguishes ZET from the other SLM genes, which all seem to be important for leaf development. Plants with a defect in ZET activity also showed more defects in perianth development when compared to other slm mutants. Cell separation and disintegration observed in stage floral meristems of zet mutants indicated that ZET may function in cell adhesion and viability, roles that seem to be unique among SLM genes.Since the phenotypes of sub and the other slm mutants overlap we assessed the genetic relationship among SLM genes by testing whether the expression of SUB is sensitive to SLM activity. This was apparently not the case, at least not in a relevant manner, since SUB expression was unaltered in the respective other slm mutants in our transcriptome analysis and showed only minor downregulation in slm flowers when assayed by qRT-PCR. In addition, a functional 35S::SUB transgene was unable to rescue the phenotype of doq, qky and zet mutants, including the floral defects. To further investigate the genetic interactions between SLM genes we performed a comprehensive double mutant analysis. The pleiotropic and variable phenotypes of slm single and double mutants complicated this task. Although an exaggerated phenotype was often observed in double mutants it was sometimes difficult to decide whether this represented an additive or synergistic phenotype. In addition, the various effects were often tissue dependent and thus precluded conclusions at the whole-plant level. One exception was qky- zet- where zet- appears to be largely epistatic to qky-. This result indicates that ZET either acts upstream of QKY in a genetic pathway, or that ZET generally acts prior to QKY. Focusing on ovule development simplified our analysis, since a stringent set of morphological criteria could be applied. In sub- zet-, sub- qky-, and doq- zet- double mutants a mass of cells with integument-like structures often developed in place of true ovules. Furthermore, all double mutants, with the exception of qky- zet-, often showed aberrant inner integuments, in contrast to slm single mutants that only displayed defects in the outer integument. Taken together, these synergistic effects indicate that the SLM genes contribute to similar aspects of ovule development and show that QKY and ZET function is still present in ovules of sub plants. In addition, the analysis supports the notion that DOQ may play a more peripheral role in ovule development in comparison to SUB, QKY, and ZET. The exact genetic relationship between SLM genes during this process remains to be elucidated.The phenotypic analysis of slm single mutants suggested that SLM genes variably contribute to several common functions but it also showed that they have distinct roles, both of which depend on tissue context. To characterise the molecular mechanisms that are under control of SLM genes and to narrow down the overlap of SLM gene function, we used transcript profiling of mRNA isolated from two different tissues. Our Z-score based meta-analysis revealed and largely non-overlapping SUB-responsive genes in the apex and flower data sets, respectively (Table ). That we consistently observed larger numbers of affected transcripts in the flower data set might reflect the relatively higher importance of SLM genes for later stages of flower development, but it could also be explained by differences in the nature of the samples. While the flower sample only contained tissue from two well-defined floral stages, the apex sample contained the shoot apical meristem as well as flowers from stage to stage and thus was much more diverse. Consequently, the sensitivity of the array to detect changes that are limited to only a few developmental stages was reduced. When comparing the number of misregulated transcripts across the various genotypes, we found that fewer genes were affected in sub mutants than in the other slm mutants. This observation suggests that DOQ, QKY and ZET play broader roles than SUB. The result might have been expected for ZET, as zet mutants show the most severely affected flowers of all slm mutants, while sub, qky and doq mutant flowers, however, are more alike and thus their morphology is less congruent with a more dramatic change at the molecular level. Meta-analysis revealed a consistent overlap in misregulated genes between sub and the other slm mutants within a given tissue and even allowed us to identify a set of core targets shared by all four SLM genes. Interestingly, these transcripts were consistently up- or downregulated in all genotypes, while the direction of the change was dependent on tissue context. These results demonstrate not only that SLM genes affect common processes with a high tissue specificity, but also that the molecular mechanisms employed by SLM genes in the different tissues are distinct. In addition, the identification of common targets that show consistent behaviour across all genotypes underlines the sensitivity and specificity of our meta-analysis. QKY Encodes a Putative Transmembrane Protein Involved in Ca2+-Mediated SignalingSequence analysis of QKY suggests that QKY is the first described plant representative of the previously described MCTP family [] and that it might act as a membrane-bound protein involved in a process regulated by Ca2+ and phospholipids. While the function of animal MCTPs is unknown, human MCTP2 is a membrane protein located to intracellular vesicular structures []. Interestingly, the C2 domains of human MCTPs were found to bind Ca2+ with high affinity but lacked any phospholipid binding capacity []. Other membrane-bound proteins with multiple C2 domains are the synaptotagmins and ferlins [],[]. Synaptotagmins contain two C2 domains and an amino-terminal transmembrane domain while most ferlins carry between four and six C2 domains and a carboxy-terminal transmembrane domain. Members of these two protein families function during regulated exocytosis, a process in which specific vesicles are signaled to fuse with the plasma membrane. Processes that rely on regulated exocytosis include neurotransmitter release at synapses and plasma membrane repair, a basic cellular process that mends physical injuries inflicted upon the plasma membrane []–[]. During the repair process lysosomes [] or specialised vesicles, such as the enlargosomes [],[], fuse with the plasma membrane providing new membrane material and facilitating resealing.The synaptotagmins Syts and are required for Ca2+-regulated synaptic vesicle exocytosis during neurotransmitter release into the synaptic cleft [],[],[] while Syt VII promotes lysosomal exocytosis during plasma membrane repair in fibroblasts []. The C. elegans ferlin FER- is localized to the membranes of membranous organelles (MOs) and promotes the fusion of MOs with the plasma membrane during the development of crawling spermatozoa [],[]. Mutations in dysferlin result in progressive muscular dystrophies []–[] and dysferlin appears to be required for Ca2+-dependent sarcolemma resealing during membrane repair in skeletal muscle fibres [],[].Given the predicted domain topology of QKY one could speculate that QKY, and possibly other SLM genes, could participate in the control of vesicle trafficking. In principle, the predicted membrane localisation of QKY also allows for the possibility that the SUB and QKY proteins interact directly. A role of QKY in vesicle trafficking would help to explain the non-cell-autonomy of SUB function in inter-cell-layer signaling []. Thus, in one possible scenario SUB could directly or indirectly regulate QKY which in turn might affect vesicular transport of factors mediating the non-cell-autonomy of SUB-dependent signaling. Further work needs to address this exciting hypothesis.Many SLM-Responsive Genes Relate to Cell Wall Biology, Hormone, and Stress ResponsesTo obtain additional insights concerning the molecular processes influenced by SLM function we investigated the known and predicted functions of SLM-responsive genes. The core sets of direct or indirect SLM targets included many genes encoding proteins relating to cell wall biology and a notable enrichment of genes that are inducible by hormones, such as JA and SA, and presumed to play a role in the plant's response to various stresses. For example, a group of genes misregulated in slm mutants encode proteins with a known or predicted role in the biosynthesis and/or function of the cell wall, such as MIOX2/, GLP6, and putative glycine-rich proteins, lectin family proteins, and chitinases. Defects in SLM-activity also resulted in a deregulation of the basal expression of JA-inducible genes with a presumed role in abiotic stress responses to cold or dehydration. The altered expression of genes responsible for camalexin and glucosinolate metabolism, secondary metabolites involved in the defense against pathogens, indicates that SLM-activity appears to play a role in biotic stress responses as well.The notion that SLM activity somehow relates to stress is reinforced by the interesting finding that a major group of SLM-responsive genes in flowers relate to various aspects of salicylic acid (SA)-dependent signaling. SA is synthesized upon exposure of plants to abiotic stresses, such as ozone or UV-C light, and plays an important role in pathogen defense (for reviews see for example []–[]). Three of the SLM-responsive genes are known to be required for the pathogen-induced production of SA: ICS1/SID2/EDS16 encodes an isochorismate synthase that synthesizes SA [],[], EDS5/SID1 encodes a MATE family transporter potentially involved in the transport of SA intermediates [], and PBS3/GDG1 encodes an adenylate-forming enzyme required for SA accumulation [],[]. Other SLM-dependent genes are involved in further aspects of SA signaling. EDS1 encodes a protein with a lipase-like domain [], contributes to rapid SA accumulation in response to many SA-inducing stimuli, and is part of a central regulatory node of SA signaling. NPR1/NIM1 constitutes another, though SLM-independent, central regulator of SA signaling. At low SA levels NPR1 is present in the cytoplasm as a homo-multimeric complex. Increased SA levels result in the dissociation of NPR1 oligomers into monomers, which enter the nucleus where they interact with specific TGA transcription factors. These NPR1-TGA complexes regulate expression of defense genes, such as PATHOGENESIS-RELATED (PR) genes and WRKY-type transcription factor genes. Interestingly, we found NIMIN1 and GRX480, both of which act in the NPR1 pathway, to be downregulated in slm mutants. NIMIN1 encodes an interactor of NPR1 that counteracts NPR1 activity []. GRX480 codes for a glutaredoxin that interacts with TGA2 [] and is possibly involved in cross-talk between SA and JA signaling. Another important class of regulators implied in plant immunity, the regulation of PR genes and SA-mediated signaling are WRKY transcription factors. Interestingly, we found five WRKY genes to be downregulated in slm mutants and three of them are known direct targets of NPR1: WRKY38, WRKY53, and WRKY70 []. WRKY53 is a positive modulator of systemic acquired resistance (SAR) [] while WRKY70 is involved in the cross-talk between SA and JA responses by acting as a negative regulator of JA-inducible genes [],[]. In addition, WRKY70 is required for resistance against several bacteria, fungi and an oomycete [],[],[],[]. Another SLM-responsive WRKY gene, WRKY33, is a positive regulator of resistance against the necrotrophic fungi Botrytis cinerea and Alternaria brassicicola []. Fitting in the overall picture was the observation that expression of five PR genes was reduced in slm mutants, including PR1, PR2, and PR5 whose activities often undergo coordinated changes in response to various stimuli including SA []. In addition, PR1 is likely a direct target of a NPR1/TGA complex [],[].Many of the SLM-responsive genes involved in SA-signaling and pathogen defense are normally transcribed at basal levels while their expression is induced several fold in the case of pathogen attack and increased levels of SA. Our transcriptome analysis of slm mutants suggests that SLM function is required for basal steady-state expression of SA-responsive genes. Thus, SLM function could be involved in priming components of the cellular defense machinery, thereby enabling floral cells to respond faster to invading pathogens. In this scenario SLM genes would contribute to basal resistance.Another scenario is based on the hypothesis that QKY may play a role in membrane traffic. Since membrane repair or cell wall synthesis require extensive vesicle trafficking [],[] SLM-dependent processes could be involved in the regulation of the composition of the plasma membrane or the cell wall and alterations in the two cellular compartments of slm mutants could be interpreted as wounding stress. For example, it was suggested that the SLM- and SA-responsive WALL-ASSOCIATED RECEPTOR KINASE1 (WAK1) gene transmits changes in the cell wall to the interior of the cell. In addition, WAK1 protects plant cells against negative aspects of the pathogen response []. The RLK THESEUS1 (THE1) has been proposed to act as a sensor of cell wall integrity []. It was shown that in the absence of proper cellulose synthesis THE1 modulates the activity of genes affecting pathogen defense, cell wall cross-linking and stress responses. Although THE1 and the SLM genes do not share common transcriptional targets, they affect a functionally related spectrum of genes. Thus, it is conceivable that one aspect of SLM function relates to the surveillance of cell wall integrity or the repair of plasma membranes.The two proposed models are not mutually exclusive and it will be an exciting challenge to further dissect the biological function of SLM genes in development, cell biology, and stress response.Materials and MethodsPlant Work and Genetics Arabidopsis thaliana (L.) Heynh. var. Columbia (Col-) and var. Landsberg (erecta mutant) (Ler) were used as wild-type strains. The sub- mutant was described previously []. Plants were grown in a greenhouse under Philips SON-T Plus Watt fluorescent bulbs on a long day cycle ( hrs light). Dry seeds were sown on soil (Patzer Einheitserde, extra-gesiebt, Typ T, Patzer GmbH & Co. KG, Sinntal-Jossa, Germany) overlying perlite, stratified for days at °C and then placed in the greenhouse. Plant trays were covered for – days to increase humidity and support equal germination. Ler seeds mutagenized with ethylmethane sulfonate (EMS) were obtained from Lehle Seeds (Round Rock, TX, USA). ' M2 plants, corresponding to about ' M1 plants, were screened for plants exhibiting a sub-like phenotype. All sub-like mutants described in this paper were outcrossed three times to Ler prior to further analysis. Two qky T-DNA insertion mutants (line SALK_140123 and SALK_043901) [] were obtained from the ABRC (http://www.arabidopsis.org). The GL2::GUS line [] in Ler was crossed into slm mutants for analysis of root hair specification.Recombinant DNA WorkFor DNA and RNA work standard molecular biology techniques were used []. PCR-fragments used for cloning were obtained using either PfuUltra high-fidelity DNA polymerase (Stratagene) or TaKaRa PrimeSTAR HS DNA polymerase (Lonza, Basel, Switzerland). All PCR-based constructs were sequenced. Information regarding all primers used in this study is given in supporting Table S1.Generation of the 35S::SUB ConstructIn order to generate a translational fusion between SUB and the small c-myc tag SUB cDNA was amplified by PCR from plasmid H2H6T7 [] using primers SUB-CmycF and SUB-Cmyc-R. A 3xmyc-tag was amplified from a pFastBac-HT A plasmid (Invitrogen) containing a 2xmyc tag using primers cmyc-F and cmyc-R. Both PCR fragments were gel purified and an overlap PCR was set up using primers SUB-Cmyc-F and cmyc-R. The overlap PCR product was cloned into vector PCRII TOPO (Invitrogen) and was designated PCRII TOPO SUB:3xmyc. SUB:3xmyc was then cloned into pCAMBIA2300 (www.cambia.org). To this end pCAMBIA2300 was digested with BamHI and PmlI to remove the 35S::GUS cassette. PCRII TOPO SUB:3xmyc was digested with BamHI and BsrBI and the resulting SUB:3xmyc fragment was cloned into BamHI/PmlI-digested pCAMBIA2300 resulting in plasmid SUB:3xmyc pCAMBIA2300. To obtain the 35S promoter [] the vector pART- [] was digested with NotI, blunt-ended by T4 DNA polymerase treatment and the 35S:NOS cassette was isolated by gel purification and digested with XbaI, resulting in a 35S promoter fragment with a ′ blunt and ′ sticky end. SUB:3xmyc pCAMBIA2300 was blunt-ended after SpeI digestion at the ′ end and redigested with BlnI to create an XbaI-compatible ′ end. Ligation resulted in the final 35S::SUB:myc construct (abbreviated 35S::SUB). The plasmid was verified by sequencing.Generation of 35S::SUB slm PlantsTransformation of sub-, doq-, qky- and zet- mutants with the 35S::SUB construct was performed using the floral dip method [] and Agrobacterium tumefaciens strain GV3101 []. Transgenic T1 plants were selected on µg/ml Kanamycin plates and subsequently transferred to soil for phenotypic inspection. About percent of the independent 35S::SUB sub- T1 plants scored showed a wild-type phenotype (/ scored). This indicates that the c-myc tag at the carboxy-terminus of SUB may result in a reduction of SUB functionality. To assay transgene expression RNA was isolated from inflorescences using the NucleoSpin RNA II kit (Macherey-Nagel, Düren, Germany). First-strand cDNA was synthesized from µg of total RNA using Moloney Murine Leukemia Virus (M-MuLV) reverse transcriptase (New England Biolabs, Frankfurt, Germany). Semi-quantitative PCR was performed using Taq DNA polymerase (New England Biolabs, Frankfurt, Germany) and a transgene-specific primer pair (sRT-SUBcmyc_F, sRT-SUBcmyc_R). Between and thermal cycles were tested. The GAPC gene was used as positive control [].Map-Based Cloning of QKY To map the QKY locus at high resolution, an F2-mapping population was generated. F1 plants from qky/qky (Ler) and QKY/QKY (Col) crosses were allowed to self-pollinate, and the F2 progeny were screened for qky individuals based on twisted inflorescence morphology. DNA was isolated and used for PCR-based amplification of molecular markers. Indel polymorphism data was derived from the Monsanto Ler sequence database at TAIR []. Primer sequences for indel and CAPS markers are shown in supporting Table S1. Marker amplifications from mutant individuals restricted the qky map interval to kb, as defined by markers F25A4(BglII) and .(RsaI). Candidate genes were analysed via T-DNA insertion mutant analysis and/or sequence determination revealing that At1g74720 carried mutations in various qky alleles. To confirm qky allelic mutations, nucleotide sequences were obtained from both strands of PCR-amplified fragments. In all three EMS-induced qky alleles transitions result in stop codons. The qky- allele carries a G to A transition at position (genomic DNA, relative to the start ATG triplet (+)) resulting in a shorter predicted protein (W743*). The qky- allele carries another G to A transition at position (W902*), and in qky- a C to T transition was found at position (Q217*). We also determined the genomic integration sites of two T-DNA insertions in At1g74720: SALK_140123 is located at position and SALK_043901 at position . Both lines exhibit a qky phenotype. SALK_140123 is predicted to carry a shorter QKY protein of residues with the last residues (RSHKGSHVMTPADDAGQAVLRLELTEPQR*) being encoded by the T-DNA. SALK_043901 results in a predicted shorter protein of residues with residues – (LFVV*) being encoded by the T-DNA.Near full-length QKY cDNA sequence was assembled from sequences derived from four publicly available RIKEN RAFL cDNA clones []. The cDNA clones partially overlapped (RAFL16--G10, RAFL22--B15, RAFL22--H20, RAFL22--F19) with one clone (RAFL22--F19) containing the ′ poly(A) stretch. Additional ′ RACE experiments [] did not result in more extended ′ cDNA sequences and comparisons of the available QKY genomic and cDNA sequences did not reveal introns.Bioinformatic AnalysisProtein domain searches were conducted using the PFAM database []. Transmembrane topology was predicted using the TMHMM webserver [].Quantitative Real-Time PCRTissue was harvested as described for the whole-genome transcriptome analysis. RNA was extracted as outlined above. First-strand cDNA was synthesized from . µg of total RNA via reverse transcription, using the Transcriptor High Fidelity cDNA Synthesis Kit (Roche Diagnostics GmbH, Mannheim, Germany). Quantitative real-time PCR was performed on a Roche LightCycler using the SYBR Green I detection kit according to the manufacturer's recommendations (Roche). Amplification of UBC21/At5g25760 served as a normalization control []. Primer sequences are summarized in supporting Table S1. Using the comparative Ct method, all gene expression levels were calculated relative to UBC21.Microscopy and Art WorkPreparation and analysis of propidium iodide-stained samples for confocal laser scanning microscopy, scanning electron microscopy, and histochemical localisation of ß-glucuronidase (GUS) activity in whole-mount tissue was done essentially as described [], [], []–[]. Confocal laser scanning microscopy was performed with an Olympus FV1000 setup using an inverted IX81 stand and FluoView software (FV10-ASW version ...) (Olympus Europa GmbH, Hamburg, Germany). After excitation at nm with a multi-line argon laser, propidium iodide fluorescence (– nm slit width) and autofluorescence (– nm slit width) was detected. One-way scan images (scan rate . s/pixel, × pixels, Kahlman frame, average of four scans) were obtained using an Olympus × objective (UApo/ ×/. Oil Iris). Confocal Z-stack imaging of floral meristems was performed using . m sections. Plants or various plant organs were analysed under an Olympus SZX12 stereo microscope. Images were obtained using a ColorView III digital camera (Olympus) and Cell∧P software (version . build , Olympus), saved as TIFF files, and adjusted for color and contrast using Adobe Photoshop CS3 (Adobe, San Jose, CA, USA) software.Whole-Genome Transcriptome Analysis and Global Expression ProfilingTissue was harvested from -day plants grown in continuous light at °C. Wild type and mutants were in the Ler background. The inflorescence apex plus flowers up to stage were kept separate from later stage – flowers (the oldest – closed flower buds). The tissue was immediately frozen in liquid nitrogen and stored at −°C. RNA was extracted with the Plant RNeasy Mini kit (Qiagen). µg of total RNA was used for probe synthesis with the MessageAmp II-Biotin Enhanced kit (Ambion) according to the manufacturer's instructions. Affymetrix ATH1 GeneChips were hybridised, washed, and stained as described []. The gcRMA implementation in the Bioconductor and R software packages was used for background correction, quantile normalization and expression estimate computation [],[] (www.bioconductor.org) []. Expression estimates were transformed to linear scale and the fold change (FC) was calculated for each gene in each mutant-wildtype comparison by dividing the expression estimate of the mutant condition by the value of an according control. Comparisons were done on a single replicate basis for all possible pair wise comparisons. A Z-score for each fold change was calculated by dividing the difference of log2 transformed FC and the mean of the FC population by the standard deviation of the FC population (). Genes were considered as differentially expressed if they displayed a Z-score> in a single comparison and an average Z-score> over all comparisons or, if they displayed a Z-score<− in a single comparison and an average Z-score over all comparisons <−.Accession NumbersThe QKY cDNA sequence was deposited at GenBank under the accession number FJ209045. The expression profile data have been deposited to the EMBL-EBI ArrayExpress repository under the accession E-MEXP-.Supporting InformationDataset S1This list details the genes that show significant changes in expression level in all Ler versus sub-like mutant (slm) comparisons. This list thus comprises the apex core target gene list.(. MB XLS)Click here for additional data file.Dataset S2This list details the genes that show significant changes in expression level in all Ler versus sub-like mutant (slm) comparisons. This list thus comprises the flower core target gene list.(. MB XLS)Click here for additional data file.Table S1List of primers(. MB DOC)Click here for additional data file.
|
PMC3018617.txt
|
TITLE: Peripheral T-Cell Lymphoma: Review and Updates of Current Management Strategies AUTHORS: Tiffany Tang, Kevin Tay, Richard Quek, Miriam Tao, Soo Yong Tan, Leonard Tan, Soon Thye Lim ABSTRACT: The classification of T-cell and natural-killer- (NK-) cell lymphomas has been updated in the 4th edition of the World Health Organization (WHO) classification of tumors of the haematopoietic and lymphoid tissue published in . Based on recent epidemiological studies, NK-cell lymphomas occur almost exclusively in Asia and South America, although T-cell lymphomas appear to occur in the East as commonly as in the West. Due to the low prevalence of this disease, diagnosis and optimal treatment of patients have not been studied prospectively in large randomized trials. Nevertheless, there has been development in the understanding of T-cell lymphomas and how they should be managed; FDG-PET emerges as an increasingly important tool in diagnosis, gene-expression signatures may aid with prognostication in the future, and novel therapies are currently being studied to improve outcomes in T-cell lymphomas. More work, however, needs to be done, and international collaboration will be pertinent to deriving meaningful results from future clinical studies. BODY: . IntroductionT-cell lymphoma and natural-killer- (NK-) cell lymphoma represent the smaller subsets of non-Hodgkin's lymphoma (NHL) that appear to have a geographical predilection to Asia. In Europe and North America, T-cell and NK-cell lymphoma account for –% of all cases of NHL whilst in Asia, this percentage is as high as % []. T-cell lymphomas, as a group, carry a poorer prognosis compared to their B cell counterpart []. In the subgroup of patients with a low international prognostic index (IPI) score of -, -year overall survival (OS) was % in those with T-cell lymphomas and % in those with B-cell lymphomas, and this difference in survival was also reflected in patients with higher IPI scores []. T-cell lymphomas, however, represent a heterogeneous group of diseases with variations in clinical characteristics, prognosis and response to treatment. This paper reviews the developments in the understanding of mature peripheral T-cell lymphoma (PTCL) and its management. .. WHO Classification of T-Cell LymphomaThe 4th edition of the World Health Organization (WHO) classification of tumours of the haematopoietic and lymphoid tissue published in builds upon the previous edition published in , and reclassifies clinicopathological entities based on better understanding from research in the last few years []. The WHO classification now contains different T-cell lymphomas, more than the previous classification. These can be subclassified according to whether they are predominantly nodal, extranodal, cutaneous, or leukemic (Table ). Two new rare EBV-related T-cell diseases affecting children in Asia or Central/South America are recognized in this classification. These are systemic EBV T-cell lymphoproliferative diseases of childhood and hydroa vacciniforme-like lymphoma. Furthermore, anaplastic large-cell lymphoma (ALCL) ALK−positive (ALK+) is now recognized as a distinct entity from the ALCL ALK-negative (ALK−) lymphomas, which occurs in an older population, and which bears a far worse prognosis []. To eliminate redundancy, precursor T-lymphoblastic lymphoma/leukaemia has lost its “precursor” prefix as the term “lymphoblastic” implies this. Finally, three new variants of cutaneous T-cell lymphoma (CTCL) were introduced, and these are primary cutaneous aggressive epidermotropic CD8+ cytotoxic T-cell lymphoma, primary cutaneous gamma-delta T cell lymphoma and primary cutaneous small/medium CD4+ T-cell lymphoma... EpidemiologyPTCL was thought to occur more frequently in Asia. Rüdiger et al. reported frequencies of PTCL in Vancouver to be .% of all NHL compared to .% in Hong Kong []. The international PTCL and NKL project reported PTCL and NKL rates of –% in Western countries compared to –% in Asian countries []. It the study, it was also interesting that the incidence of ALCL (both ALK+ and ALK−) in North America (%) was almost four-times higher than in Asia (%). Data from Au et al. and our own institution's suggest that the rate of PTCL in the East may actually be similar to the West []. The perceived higher rates of PTCL in the East could perhaps be due to the higher incidence of NKL in the East and the differences in diagnostic evaluation []. In their institution, they evaluated consecutive cases of TCL and NKL and found that these accounted for .% and .% of all NHL, respectively. We evaluated a total of patients with malignant lymphomas from to and found that extranodal NKL and PTCL comprised .% (/) and .% (/) of all cases []... Clinical Characteristics of TCLPeripheral T-cell lymphoma, not otherwise specified (PTCL-NOS), Angioimmunoblastic T-cell lymphoma (AITL), and ALCL are the most common subtypes and account for up to % of all T-cell lymphomas []. AITL tends to occur in elderly patients, and patients often present with disseminated lymphadenopathy, hepatosplenomegaly, and autoimmune phenomena []. ALCL ALK+ typically occurs in young men and presentation can be nodal or extranodal, involving the skin, bone, soft tissues, lung, and liver []. On the other hand, ALK‒ ALCL tends to occur in an older population of patients, the presentation is usually nodal, and the disease runs a more aggressive clinical course []. PTCL-NOS, represents a heterogeneous category of nodal and extranodal T cell lymphomas that cannot be grouped into defined entities. Most present with peripheral lymph-node enlargement, B symptoms and at diagnosis, the disease is advanced []. Other uncommon T-cell lymphomas include enteropathy-associated T-cell lymphoma (EATL) [], adult T-cell leukaemia/lymphoma (ATLL) [], hepatosplenic T-cell lymphoma (HSTL) [], and subcutaneous panniculitis-like T-cell lymphoma (SPTCL) []. It is now recognized that EATL lymphoma consists of two distinct forms: classical or type I EATL and type II EATL. Type I EATL is associated with coeliac disease while type II EATL occurs in Asia and is not associated with coeliac disease []. ATLL is caused by infection with the human T-cell-lymphotropic virus type (HTLV-), and is rare outside HTLV-I-endemic areas such as the Caribbean, Japan, and parts of central Africa []. HSTL is rare; patients present with hepatosplenomegaly and systemic symptoms, and in % it occurs in the context of chronic immune suppression []. SPTCL occurs more commonly in women, and is associated with autoimmune conditions such as systemic lupus erythematosus []... Diagnostic Workup of PTCLThe histopathological diagnosis of PTCL can prove to be a challenge even for the experienced pathologist. In addition to standard histological analysis and immunophenotyping, clinical information is important in accurately classifying the TCL. To highlight this point, in the same international T-cell lymphoma study, a change in diagnosis of PTCL-NOS to ATLL (Adult T-cell lymphoma/leukaemia) occurred in .% of cases after pathologists were informed of the patient's HTLV- (human T-cell leukaemia virus) status [].The standard panel of immunostains used in the workup of a T-cell lymphoma includes CD20, CD2, CD3, CD4, CD5, CD8, CD30, CD56, TCR-β, TIA- (T-cell intracellular antigen ), Ki67, and in situ hybridization assay for EBV (Epstein-Barr virus) encoded RNA (ribonucleic acid). PTCL-NOS describes a heterogeneous population of lymphomas of the T-cell lineage, but which does not fulfill criteria to be classified into the other categories of TCL. Most cases are positive for CD4, TCRαβ is expressed and CD5 and are frequently downregulated []. AITL is characterized by a polymorphous infiltration of the lymph node with prominent proliferation of high-endothelial venules and follicular dendritic cells. CD3, CD4, CD10, BCL6, PD1, and CXCL13 are typically positive, and EBV-positive B cells are nearly always present []. AITL is closely related to follicular helper T cells and the follicular variant of PTCL, although they differ genetically []. ALCL, ALK+ and ALCL, ALK− appear similar histologically. There are morphological variations within ALCL but all have hallmark cells, which are cells with eccentric, horseshoe, or kidney-shaped nuclei, often with eosinophilic region near the nucleus. In both subtypes, CD30, CD2, and CD4 and cytotoxic molecules including that of perforin, granzyme B, and TIA- are positive. In ALCL, ALK+, the epithelial membrane antigen (EMA) is characteristically present, and CD3 is negative in % of cases. In ALCL, ALK−, EMA is positive only in a small proportion of patients, and CD3 is positive [, ] (Table ).As mentioned above, Type II EATL is distinct to classical EATL. In classical or type I EATL, the tumour ulcerates through the wall of the mucosa, and the intestinal mucosa surrounding the tumour shows enteropathy comprising villous atrophy, crypt hyperplasia, and increased lamina propria lymphocytes and plasma cells. There are large cytotoxic T cells that often express CD103 and lose both CD4 and CD8 markers, and the intraepithelial lymphocytes adjacent may have an identical immunophenotype. In type II EATL, neoplastic cells are more monotonous with a distinctive CD3+, CD8+, and CD56+ phenotype. There are increased intraepithelial lymphocytes, but not the villous atrophy or crypt hyperplasia typically seen in celiac disease or classical EATL []... Initial Assessment and Staging of T-Cell LymphomasStaging of PTCL is similar to that of other NHL and involves a complete physical examination and routine laboratory tests such as a complete blood count, renal-function test, liver-function test, and serum lactate dehydrogenase level, unilateral or bilateral bone marrow examination (aspirate, trephine, flow cytometry), and radiological imaging with either computerised tomography (CT) scanning or FDG PET/CT (fluorodeoxyglucose positron emission tomography) assessment. . FDG PET/CTPET scans have an established role in the staging and assessment of disease response in B-cell lymphomas, with the ability to differentiate between fibrosis and active disease in a residual mass as well as in prognostication []. In contrast, the role of PET in T-cell lymphomas continues to evolve. A recent review of patients with T-cell lymphomas who underwent PET examination at diagnosis or response assessment reported variations in PET-positivity and maximum standardized uptake value (SUV) across the different T-cell-lymphoma subtypes. PET positivity ranged from % in cutaneous ALCL to % in AITL to % in ATLL. Of note, % of patients had diseases that were not picked up on diagnostic CT of the neck, chest, abdomen, and pelvis. The sites that were missed included cutaneous, subcutaneous, muscular masses of the scalp, upper and lower extremities, and lymphadenopathy in the epitrochlear and popliteal regions, indicating perhaps a role for PET imaging from vertex to feet for patients with TCL []. .. Prognostic Indicators in T-Cell LymphomaThe different histological subtypes of PTCL confer different prognoses. Systemic ALCL portends a better prognosis than non-ALCL PTCL [–], and within the ALCL subpopulation, those that are ALK- positive have a better prognosis than those that are ALK- negative (-year OS, % versus %) []. The international prognostic index (IPI) initially described by Shipp et al. [] was able to separate patients with PTCL-NOS and systemic ALCL into distinct prognostic risk groups; however, even with low IPI scores, patient with PTCL had poor outcomes []. A separate prognostic model for patients with PTCL-NOS called the prognostic index for PTCL-U (unspecified); PIT, was proposed []. In this retrospective analysis of patients with PTCL-U, multivariate analysis showed that age, performance status, LDH level, and bone marrow involvement were predictive of survival, and these variables were used to develop a new prognostic model that could separate the patients into distinct prognostic groups.Curiously, neither a high IPI nor PIT score predict for a poorer outcome in AITL []. It is postulated that the clinical effects of AITL is due to the dysregulation of the immune system rather than the direct effects of tumour growth, given that hypergammaglobulinemia and eosinophilia accompany this condition, and since IPI and PIT scores generally reflect tumour burden, it was unable to prognosticate in AITL. Furthermore, recent work in gene-expression profiling reveals that AITL is closely associated with follicular helper cells, and that more than % of genes in the AITL gene signature were contributed by nonneoplastic cells. The AITL gene signature included genes involved in humoral immune response, recruitment of inflammatory cells, and modulation of vasculogenesis and the extracellular matrix []. The CTCL have the best -year OS; primary cutaneous ALCL: % and subcutaneous panniculitis-like PTCL: %. EATL and hepatosplenic PTCL are associated with particularly dismal prognoses, with reported -year survival rates of only % and %, respectively... Gene-Expression Profiling to Better Classify and Risk Stratify PTCLEven with the revised WHO classification of tumors and the development of several prognostic models, PTCL-NOS remains heterogeneous in terms of biology and outcome. Using microarrays to analyse the gene expression profiles of nodal TCL, Ballester and colleagues have managed to distinguish between AITL, ALCL, and T-lymphoblastic lymphoma, well-characterised entities in the WHO classification []. With the same technique, they managed to identify molecular subgroups amongst the PTCL-NOS: PTCL-U1, PTCL-U2, and PTCL-U3. The genes overexpressed in PTCL-U1 included genes associated with a poorer outcome in other tumors such as CCND2 and interestingly, although not significant, the survival of this group was worse than the other groups. PTCL-U2 was associated with genes involved in T-cell activation and apoptosis; NFKB1 and BCL2, and PTCL-U3, with genes involved in the IFN/JAK/STAT pathway, possibly shedding light on the pathogenesis of these tumours. This work, however, is preliminary, and confirmatory studies are needed... TreatmentNo randomized phase III trial has been conducted in PTCL. Thus, by default, Cyclophosphamide, Doxorubicin, Vincristine, and Prednisolone (CHOP)-like chemotherapy, is the most common regimen used despite its lack of efficacy. The international T cell Project confirms the poor prognosis of patients with PTCL treated with CHOP-like regimen []. To improve the outcomes of these patients, several strategies have been investigated.. Increased Dose-Intensity Chemotherapy RegimensAt least two large studies have shown the utility of a dose-intense regimen in improving outcomes in aggressive large-cell lymphomas, including PTCL. The first study compared the ACVBP (doxorubicin, cyclophosphamide, vincristine, bleomycin, and prednisolone) regimen with CHOP in patients aged between to years of age and included (%) patients with T-cell lymphomas. The dose-intense regimen was shown to significantly improve -year event-free survival (EFS) and overall survival (OS) (% versus %, P = . and % versus %, P = ., resp.) []. The second study was the NHL-B1 trial of the German high-grade non-Hodgkin's lymphoma study group (DHNSL) and it had a -by- factorial design and compared -weekly versus a -weekly CHOP with or without Etoposide (CHOEP) in patients under years, of whom % had T-cell lymphoma. In pooled analysis, the addition of Etoposide to CHOP chemotherapy in this group of patients significantly improved complete response (CR) rates as well as -year EFS (.% versus .%, P = . and .% versus .%, P = ., resp.) although there was no difference in OS []. These results were further reinforced in a recent review of all patients with PTCL and NK-cell lymphoma treated prospectively in phase II or III protocols of the DHNSL []. Nonetheless, these studies were not designed specifically to address the optimal regimen in patients with PTCL, but suggest that a dose-dense approach is feasible. Other attempts at increasing dose density in the setting of PTCL include the use EPOCH (Infusional etoposide, vincristine, and doxorubicin with bolus cyclophosphamide and oral prednisolone) [] and hyper-CVIDD (hyperfractionated cyclophosphamide, vincristine, dexamethasone, and liposomal doxorubicin) alternated with methotrexate and cytarabine [] regimens. Peng et al. reported the use of EPOCH in patients with PTCL, the largest study of EPOCH use in PTCL patients. These patients included seven patients with PTCL-U, seven with NKT lymphoma, five with ALCL, one with mycosis fungoides/Sezary syndrome, and another with SPTCL. Of these, seven were previously treated. The overall response rate was % (/) with % (/) achieving CR. Grade III/IV neutropenia rate was .% with no treatment-related mortality in this study []. In a review of patients with HSTL attending the M.D Anderson Cancer Centre from to , patients received hyper-CVIDD alternated with methotrexate and cytarabine. Of these, achieved CR and subsequently proceeded with stem-cell transplantation (SCT). Survival was reported as , , and months for these patients. One patient achieved PR and passed away months later [].. High-Dose Chemotherapy and Autologous Stem Cell RescueUncontrolled prospective and retrospective studies suggest that upfront high dose chemotherapy with autologous stem cell rescue (HDC+ASCR) as consolidation in patients who attained complete remission following first-line therapy may result in superior disease control and improved OS rates (Table ) [–]. In a retrospective review of patients with PTCL, the GEL-TAMO group showed that a -year OS rate of % was achieved in the patients who were transplanted in first CR compared to % in those treated in a salvage setting []. In another retrospective study of patients with AITL undergoing autologous HSCT, patients transplanted at first remission had significantly better -year and -year progression-free survival (PFS) compared to those transplanted following chemotherapy-sensitive relapse (-year PFS, % versus %; year PFS, % versus %). The estimated OS rates were also better for those who received a transplant during first CR compared to those transplanted following chemotherapy-sensitive relapse (-year OS was % versus %; -year OS was % versus %) []. Reimer et al. conducted the largest prospective trial on upfront HDC+ASCR for patients with PTCL []. Of the patients enrolled, underwent HDC+ASCR. The estimated -year OS rate was % for patients who underwent HDC+ASCR compared to only % for those who did not. However, this was a single-arm study and about a third of the patients did not undergo HDC+ASCR because of a variety of reasons, including disease progression in (%). This study highlights two important issues. Firstly, HDC+ASCR following first remission shows promise in improving survival in patients with an estimated -year survival rate of %. This compares favorably with published data. However, this strategy is limited by the fact that about a third of patients suffered early disease progression. Thus, the challenge is in identifying an effective regimen that can induce a high remission rate so that more patients can undergo HDC+ASCR. (Table ).. Graft Versus Lymphoma Effect of Allogeneic Stem Cell TransplantationHistorically, allogeneic SCT is associated with a lower relapse rate than HDC+ASCR in intermediate to high grade lymphomas. The groups that have subsequently attempted to study the role of allogeneic SCT in patients with PTCL have generally found that it is feasible, however, the treatment-associated toxicities are significant. A prospective phase II trial was conducted on patients with relapsed or refractory PTCL, and they were subjected to salvage chemotherapy and reduced-intensity allogeneic SCT []. At a median followup of months, patients had CR, had partial response (PR), and had stable disease (SD) status. -year OS and PFS were % and %, respectively. patients (%) had CMV reactivation, (%) developed graft versus host disease, and one died of Enterobacter sepsis. In a retrospective review of patients with PTCL who underwent allogeneic SCT, the -year OS and event-free survival rates were % and % respectively []. Notably, the -year OS for chemoresistant patients was %, a result that would not be achieved with conventional approaches alone. Treatment-related mortality, however, was %.. Novel Agents6.. Anti-Metabolites (Gemcitabine and Pralatrexate)The addition of anti-metabolites to the treatment of PTCL has been studied. Gemcitabine, a pyrimidine analogue, has been added to a CHOP plus etoposide regimen in a pilot study of patients with newly diagnosed PTCL. The overall response rate was % ( patients), with (%) achieving complete remission or unconfirmed complete remission (uCR). patients (%) developed grade IV neutropenia, but there were no treatment-related deaths []. Pralatrexate is another antimetabolite, which has garnered much interest in the treatment of PTCL. Pralatrexate is an antifolate agent that has high affinity for the reduced folate carrier- (RFC-), whose expression is induced by oncogenes H-ras and c-myc, resulting in selective accumulation of the drug in tumour cells. In a phase I-II study of patients with both pretreated B- and T-cell lymphoma, Pralatrexate was seen to induce a higher overall response rate in the T-cell sub group compared to the B-cell subgroup (% versus %) []. The PROPEL (Pralatrexate in patients with relapsed or refractory peripheral T-cell lymphoma) study soon followed as a prospective, phase II, and single-armed study in patients with PTCL who had a median of cycles of chemotherapy. Despite having been heavily pretreated, the overall response rates were a remarkable %, with % achieving CR or uCR []. However, only % of patients had a sustained response beyond weeks... Immunotherapy (Alemtuzumab and Denileukin Diftitox)Alemtuzumab is a humanized monoclonal antibody against CD52, which is highly expressed on malignant T cells, but also in normal B- and T- lymphocytes []. It has been studied in combination with CHOP for the treatment of PTCL in two phase II studies. The first, reported by Gallamini et al., enrolled patients whilst the second by Kim et al., had patients [, ]. Both showed high CR rates of % and %, respectively; however, the both studies documented serious infective complications. In the first study, a quarter of their patients developed life-threatening infections, including invasive aspergillosis, JC viral encephalitis, pneumocystis carinii pneumonia, and suspected tuberculosis []. In the second study, febrile neutropenia occurred in % of patients while patients, (%) had CMV reactivation, patients developed CMV disease (pneumonitis or retinitis), and died of treatment-related complications, resulting in the premature closure of the trial []. Denileukin diftitox (Dd) is a recombinant fusion protein containing the IL- (Interleukin-) protein and a truncated portion of the diphtheria toxin which targets cells expressing the IL- receptor. It was approved by the FDA in for the treatment of CTCL [] and is now being actively studied for its utility in the treatment of PTCL. In a phase II trial, patients with relapsed/refractory PTCL (CTCL excluded) were treated with single-agent Dd times weekly with tumour staging repeated after every cycles []. Six patients (%) achieved CR and (%) showed PR. The median PFS was months after a median of . prior therapies. There were no grade / toxicities reported. Dd was then combined with CHOP in patients with a % CR rate and a median PFS of months []. Again, Dd was well tolerated, and multicenter randomized Phase III trials comparing CHOP to Dd-CHOP are in progress... Histone Deacetylase Inhibitors (HDACis)Romidepsin was first isolated from the bacteria Chromobacterium violaceum from a soil sample obtained from the Yamagata prefecture in Japan. It had minimal antibacterial activity but showed antitumor effect in human cell lines []. patients with relapsed PTCL, treated with a median of previous lines of chemotherapy, were given Romidepsin. The OR rate was % ( patients) and the median duration of response was . months []. Currently, a phase 2b study of Romidepsin in progressive or relapsed PTCL is ongoing at several international centers.As a class, HDACis are thought to exert their antitumor effect by modulating gene expression thereby influencing cell differentiation or apoptosis. Both Vorinostat, an oral HDACi as well as Romidepsin, have been approved by the FDA for the treatment of CTCL [, ]... ConclusionThere is still substantial work to be done in the classification, prognostication, and management of mature T-cell lymphomas. With the advent of high-throughput sequencing, technology is available to quickly identify putative genes involved with each subgroup of TCL, which may allow refinement of the classification and prognostication process. In light of the low incidence of this tumour, international collaboration will be important in accruing patients for larger randomised phase III trials to evaluate the various novel treatment agents.
|
PMC3007540.txt
|
TITLE: Di-μ-chlorido-bis{chlorido[(R)/(S)-,-diphenyl--(-pyridyl-κN)--pyrazoline-κN ]zinc(II)} AUTHORS: Miquel Barceló-Oliver, Angel Terrón, Angel García-Raso, Iztok Turel, Marta Morell ABSTRACT: In the centrosymmetric binuclear title compound, [Zn2Cl4(C20H17N3)], the coordination geometry of the ZnII ion can be described as a distorted ZnN2Cl3 trigonal bipyramid (τ = .), arising from the N,N′-bidentate ligand, a terminal chloride ion and two bridging chloride ions. The N atoms occupy one axial and one equatorial site and the terminal chloride ion occupies an equatorial site. The dihedral angle between the pyridine and pyrazole rings is . ()°. In the crystal, aromatic π–π stacking [centroid–centroid separations = . () and . () Å] and C—H⋯Cl and C—H⋯π interactions help to establish the packing. BODY: Related literatureFor background to the biochemistry of zinc, see: Casas et al. ( ▶). For the synthesis of the ligand, see: Barceló-Oliver et al. ( ▶). For the fluorescent properties of related ligands, see: Bissell et al. ( ▶); Silva et al. ( ▶); Wang et al. (2001a ▶,b ▶). For stability constants, see: Yamasaki & Yasuda ( ▶). For geometrical analysis, see: Addison et al. ( ▶). For a description of the Cambridge Structural Database, see: Allen ( ▶). ExperimentalCrystal data [Zn2Cl4(C20H17N3)] M r = .31Monoclinic, a = . () Å b = . () Å c = . () Åβ = . ()° V = . () Å3 Z = 2Mo Kα radiationμ = . mm− T = K0. × . × . mm Data collection Enraf–Nonius KappaCCD diffractometerAbsorption correction: multi-scan (SADABS; Bruker, ▶) T min = ., T max = . measured reflections3467 independent reflections2206 reflections with I > 2σ(I) R int = . Refinement R[F > 2σ(F )] = . wR(F ) = . S = . reflections235 parametersH-atom parameters constrainedΔρmax = . e Å− Δρmin = −. e Å− Data collection: COLLECT (Nonius, ▶); cell refinement: DENZO and SCALEPACK (Otwinowski & Minor, ▶); data reduction: DENZO and SCALEPACK; program(s) used to solve structure: SHELXS97 (Sheldrick, ▶); program(s) used to refine structure: SHELXL97 (Sheldrick, ▶); molecular graphics: ORTEP- for Windows (Farrugia, ▶); software used to prepare material for publication: SHELXL97 and PLATON (Spek, ▶).Supplementary MaterialCrystal structure: contains datablocks I, global. DOI: ./S1600536810026127/hb5500sup1.cif Structure factors: contains datablocks I. DOI: ./S1600536810026127/hb5500Isup2.hkl Additional supplementary materials: crystallographic information; 3D view; checkCIF report
|
PMC1955407.txt
|
TITLE: Do Public Schools Provide Optimal Support for Children With Diabetes? AUTHORS: Dustin Melton, Jessica Henderson ABSTRACT: No Abstract BODY: To the Editor:The obesity epidemic among children has received considerable attention in the media, partly in response to two important reports by the Institute of Medicine on the topic (,). The parallel rise in diabetes mellitus among children is an equally important topic but is less discussed, even though diabetes has become one of the most common chronic diseases among youth. The estimated prevalence of type or type diabetes mellitus among U.S. children or youth aged years or younger was . per , (which means that there were about , children or youth aged < years with diabetes in the United States) in – (). The public schools are important locations for secondary prevention interventions to help these children and youth minimize their risk for complications associated with diabetes. We investigated the diabetes knowledge and preparedness among members of the faculty and staff of schools in Oregon, as well as barriers to school support for students with diabetes.We asked administrators and staff listed in the Public School Directory on the Oregon Department of Education Web site in to complete the School Diabetes Survey developed by Lewis et al (). The -item survey examined in-school support for students with diabetes, including the diabetes-related knowledge and training of teachers and staff and the schools' policies and procedures regarding diabetes.The schools that responded to the survey represented % of the K– public schools in the state with an e-mail address in the directory of the Oregon Department of Education, and the student population of the schools represented % of the total student population in Oregon. Of the schools that indicated their school level, % were elementary schools only, % were middle schools only, % were high schools only, and the rest were mixed levels. The population of the schools ranged from to students.The number of students identified with diabetes in each school ranged from none to . One of every four schools reported that they had no students with diabetes. About % of the schools indicated that teachers or school staff received training about diabetes. Likewise, % stated that they had a school diabetes policy in place. This finding means at least one in five schools either did not have a policy regarding the management of students with diabetes, did not have a person trained to understand the needs of students with diabetes, or both (Table ).Almost all schools (%) responded that they had a policy that allowed students to measure their own blood glucose at school; only two (one high school and one elementary school) responded that they did not. Most schools (%) reported believing that the responsibility for providing glucose-containing food or beverages or both for students with diabetes resides with the parents. Three of every four schools had an accessible refrigerator to house glucose-containing foods. Only two thirds of schools had a full-time staff person who was trained to administer insulin injections. Nearly half of the schools perceived that the main barrier to providing optimum support for students with diabetes was the lack of a nurse on staff each day (Table ). Lack of teacher and staff training and lack of funding were also cited as being the main barrier.The results from our study were similar to those from a study of Lewis et al of schools in the East, which showed that % of schools did not have a staff member with training about diabetes and that % did not have a diabetes management policy (). Absence of a nurse on site daily was a common finding in that study also, indicating that these issues regarding the capacity of schools to assist students with diabetes most likely exist in other states.The National Association of School Nurses recommends that schools should have one school nurse for every students (). Unfortunately, the national ratio is just one for every children. This means that many children receive care from a secretary, teacher, or counselor. Under these conditions, it is critical that schools redouble their efforts to make sure they have a safe and secure process for providing optimum support for students with diabetes.One educator in this study described the lack of support provided to students with diabetes this way: "even though diabetes is serious, it does not receive the same publicity and criticism as does state testing, school report cards, etc. That's what makes the news and gets attention. Diabetes, along with all the other ailments, is serious, but get us (especially small rural schools) medical help." The findings of this study indicate that at least one in five Oregon schools is lacking what it needs to provide a healthy environment for students with diabetes: either a nurse on staff during all school hours, diabetes training for teachers and staff, or both.
|
PMC2639598.txt
|
TITLE: Inhibitory effects of rat bone marrow-derived dendritic cells on naïve and alloantigen-specific CD4+ T cells: a comparison between dendritic cells generated with GM-CSF plus IL- and dendritic cells generated with GM-CSF plus IL- AUTHORS: George Tiurbe, Anja Matuschek, Ulrike Kämmerer, Manuela Schneider, Arnulf Thiede, Karin Ulrichs, Christoph Otto ABSTRACT: BackgroundUnlike mouse immature bone marrow (BM)-derived dendritic cells (DC), rat immature BMDC have not been thoroughly characterised in vitro for the mechanisms underlying their suppressive effect. To better characterise these mechanisms we therefore analysed the phenotypes and immune inhibitory properties of rat BMDC generated with GM-CSF plus IL- (= IL- DC) and with GM-CSF plus IL- (= IL- DC).ResultsBoth IL- DC and IL- DC exhibited lower surface expression of MHC class II and costimulatory molecules than mature splenic DC. They had a strong inhibitory effect on responsive T cells in vitro and despite their weak function as antigen-presenting cells they induced anergic T cells. However, only anergic T cells induced by IL- DC had a suppressive effect on responsive T cells. Induction of suppressive/tolerogenic T cells by IL- DC required direct contact between antigen-specific T cells and IL- DC. In addition, IL- DC and IL- DC prolonged allograft survival in an antigen-specific manner.ConclusionA unique phenotype of immature BMDC was isolated from the cultures. The mechanisms underlying the suppressive effect may be caused by their inability to deliver adequate costimulatory signals for T-cell activation. In addition, IL- DC but not IL- DC induce anergic T cells with suppressive function. This indicates that IL- DC and IL- DC may differ in the quality of their costimulation although no differences in the surface expression of costimulatory molecules were found. BODY: BackgroundIn recent years it has become clear that dendritic cells (DC) are not only potent inducers of adaptive immune responses, but also essential mediators in the induction and maintenance of T-cell tolerance []. The biological properties of DC depend on their phenotypically distinct states of development []. Their delaying effect on allograft rejection has been demonstrated in several rodent models [reviewed in []].Mouse and human DC have both been studied thoroughly [reviewed in []]. Rat DC have been investigated particularly by groups interested in transplantation research [-]. They were not studied thoroughly, although established culture methods exist for the generation of bone marrow-derived rat DC (BMDC) [,]. The maturation of BMDC varies from species to species despite comparable culture conditions. In mice, for example, low doses of granulocyte macrophage colony stimulating factor (GM-CSF) combined with interleukin (IL)- induce the formation of mature BMDC [], whereas in rats the same combination produces immature BMDC []. The effect of GM-CSF and IL- on the generation of rat BMDC is not clearly known.In the present study we examined the ability of IL- DC and IL- DC to inhibit both the activation of naïve T cells and the restimulation of antigen-specific T cells in vitro. We also analysed their in vivo potential to prolong allograft survival.MethodsGeneration of rat BMDCFemur and tibia bones of young (– weeks) Lewis rats were extracted and disinfected in % ethanol. Both ends of the bones were cut and the bone marrow (BM) cells were flushed with ml phosphate-buffered saline (PBS). The BM cells were cultured at a cell density of – × cells/ml in culture dishes (Falcon, Becton Dickinson Biosciences). The RPMI culture medium [] was supplemented with ng/ml recombinant rat GM-CSF (R&D Systems, Heidelberg, Germany) and ng/ml recombinant rat IL- (Miltenyi Biotech GmbH, Germany) or ng/ml rat IL- (Miltenyi Biotech GmbH). On day , non-adherent cells and cells growing in clusters were collected.Cell isolationSplenic DC (S-DC), naïve and antigen-specific T cells were isolated from Lewis rats as described previously []. Allopeptide P1-specific T cells were induced by immunization with P1 [].Activation of naïve T cells and restimulation of antigen-specific T cellsNaïve T cells ( cells/well) were incubated with Gy irradiated IL- DC, IL- DC, or mature S-DC ( cells/well) for days at °C in a % humidified CO2 atmosphere. Allopeptide P1-specific T cells ( cells/well) were incubated with irradiated ( Gy) IL- DC, IL- DC, or S-DC ( cells/well) loaded with P1 (. μg/well) for days at °C in a % humidified CO2 atmosphere. T-cell proliferation in -well, round-bottom plates was measured after 3H-thymidine (. μCi/well) incubation for the last h before harvesting. Radioactivity was determined as previously described []. Results (mean ± SD) were expressed in counts per minutes (cpm).Transwell experimentsSome of the assays were performed in -well transwell plates (Corning Life Sciences, The Netherlands). The upper compartment contained immature rat BMDC ( × ), the lower compartment antigen-specific T cells ( × ). After days of culture, the transwells were removed and P1-loaded S-DC (/well) as well as antigen-specific T cells ( × /well) were added to the lower compartments. The cultures were then incubated for another days and pulsed with . μCi/well [3H]-thymidine for the last h of culture. The incorporation of [3H]-thymidine was measured as described [].Heterotopic heart transplantation with and without antigen-loaded BMDCThe animal experiments were conducted in accordance with European, national and institutional animal care policies. Ten million BMDC were incubated with μg P1 for min in μl PBS, washed times with PBS and transferred intravenously via the penile vein into Lewis rats under full anesthesia with isoflurane day before transplantation. Fully vascularised, heterotopic heart transplantation was performed, and graft survival was monitored as described [].ResultsIL- DC and IL- DC exhibited an identical phenotypeOn approximately day of culture low adherent cell clusters were obvious and both the number and size of the clusters increased during culture (Fig. 1A). The BMDC isolated from these clusters on day were positive for Ox62 (a marker for rat DC), Ox6 (anti-MHC class II), and the macrophage marker ED1 (CD68) (Fig. 1B–D). More than % of the cells isolated from the clusters were double positive for Ox62 and Ox6 without further purification. The flow cytometric analysis revealed that IL- DC and IL- DC expressed MHC II and the costimulatory molecules CD40, CD80 and CD86 at nearly identical levels on their surface (Fig. ).Figure 1Morphology and immunostaining of IL- DC and IL- DC. Rat BMDC were isolated from cell clusters on day of culture (A). Cells prepared on cytospin slides stained positive for monoclonal antibodies Ox62 (rat DC marker) (B), Ox6 (MHC class II) (C), and CD68 (D). Shown are representative IL- DC results, which are similar to those for IL- DC. Magnification: × (A, C, D) and × (B).Figure 2IL- DC and IL- DC exhibit no obvious differences in their phenotype. IL- DC and IL- DC and mature splenic DC (sDC) coexpressed Ox62 and CD68, whereas macrophages generated in the presence of M-CSF ( μg/ml) were only positive for CD68. Broken lines indicate background staining obtained using an irrelevant isotype control. The first number represents the percentage of cells staining positive for the indicated marker and the second number represents the mean fluorescence intensity. The results shown are representative for independent flow cytometric analyses. The antibodies were purchased from Serotec, Ltd, Oxford, United Kingdom, with the exception of HM40- and 24F (BD Biosciences, Heidelberg, Germany).IL- DC and IL- DC neither activated naïve T cells nor restimulated antigen-specific T cellsNaïve T cells did not proliferate in the presence of IL- DC and IL- DC, whereas mature S-DC induced a strong T-cell proliferation (Fig. 3A). IL- DC and IL- DC loaded with the allogeneic peptide P1 [] were not able to restimulate P1-specific T cells (Fig. 3B). Different numbers (, and ) of P1-loaded IL- DC or IL- DC suppressed the proliferation of antigen-specific T cells induced by P1-pulsed S-DC (Fig. 3C, D). For both BMDC types, IL--specific mRNA was not detectable by RT-PCR (Additional File , Table ). This may indicate that they revealed properties of immature DC.Figure 3IL- DC and IL- DC inhibit the proliferation of responsive T cells. IL- DC and IL- DC were unable to activate either naïve T cells in the mixed leukocytes culture (A) or antigen-specific T cells in the T-cell proliferation assay (B). The effect of mature S-DC on the proliferation of naïve T cells is shown in the inset of Fig. A. The antigen-specific restimulation of P1-specific T cells by P1-loaded S-DC (P1) is shown in B (inset). Different numbers of IL- DC and IL- DC influenced the S-DC-induced proliferation of antigen-specific T cells (C-D). The addition of P1-loaded IL- DC and IL- DC to the proliferation assay on day of the -day culture prevented the marked increase in T-cell proliferation that had been induced by P1-loaded S-DC between days and (E-F). The results (mean ± standard deviation) shown are representative for different experiments. Control: Naïve T cells (A) and antigen-specific T cells (B) cultured alone; ∅: P1-unloaded S-DC.Table 1Primers and PCR conditions.PrimerSequence ('- > ')PCR productAnnealing temperatureGAPDHFor: GGT CGG TGT GAA CGG ATT TGRev: GTG AGC CCC AGC CTT CTC CAT319 bp62°CMHC IIFor: CAG GAT CTG GAA GGT CCARev: AGC TGT GGT TGT GCT GA517 bp55°CCD40For: CGC TAT GGG GCT GCT TGT TGA CAG;Rev: GAC GGT ATC AGT GGT CTC AGT GGC401 bp58°CCD80For: TGG TGA AAC ACC TGA CCARev: GTT TCT CTG CTT GCC TCA517 bp50°CCD86For: TGG GAA ACA GAG CTC TCARev: AGG TTG ATC GAC TCG TCA518 bp53°CPrimers specific for rat IL-12p35 was purchased from Invitrogen/Biosource. Total RNA was isolated from cells with ml of the ready-to-use reagent Trizol Reagent (Invitrogen GmbH, Germany) according to the manufacturer's recommendations. Reverse transcription of μl RNA was performed using the GeneAmp RNA-PCR-Kit (Applied Biosystems GmbH, Germany). Five μl cDNA was amplified using Gold AmpliTaq DNA-Polymerase (. U/μl) and the specific primers ( μmol/L each) mixed in nuclease free water (Promega GmbH, Germany) to an end volume of μl per sample.In the BMDC-uninfluenced T-cell proliferation assay, the strongest increase in proliferation of antigen-specific T cells occurred between days and of the -day culture. The addition of P1-pulsed IL- DC or IL- DC to the cultures on day halted this strong increase in T-cell proliferation within h (Fig. 3E, F).IL- DC-T and IL- DC-T showed anergic propertiesP1-specific T cells were named IL- DC-T when incubated with IL- DC and IL- DC-T when incubated with IL- DC. Following purification, the DC-T were transferred to P1-pulsed mature S-DC and their proliferation rate after h was low (Fig. ). The anergic-like effect of IL- DC-T and IL- DC-T could be negated by adding exogenous IL- ( ng/ml) to the cultures (Fig. ).Figure 4IL- DC-T and IL- DC-T do not proliferate and demonstrate anergy. IL- DC-T and IL- DC-T, purified and transferred to second cultures, did not proliferate in response to subsequent stimulation with P1-loaded splenic DC (S-DC), but required the addition of IL- (B). In contrast, antigen-specific T cells not precultured with IL- DC or IL- DC had proliferation rates of , ± , cpm (not shown). The results (mean ± standard deviation) are representative for different experiments.IL- DC-T had an inhibitory effect on antigen-specific T cellsDifferent numbers (, and ) of IL- DC-T were added to assays containing P1-loaded S-DC and freshly isolated antigen-specific T cells. Measurement of proliferation revealed that IL- DC-T had a dose-dependent inhibitory effect on T-cell proliferation (Fig. 5A). Antigen-specific T cells cultured with IL- DC-T in transwell plates did not reduce the control T-cell proliferation (Fig. 5B). Antigen-specific T cells must have direct contact with IL- DC in precultures to have an inhibitory effect on T-cell proliferation in second cultures. The IL- DC-T, in contrast, had no influence on the proliferation of antigen-specific T cells (not shown).Figure 5IL- DC-T inhibit the proliferation of antigen-specific T cells in a dose-dependent manner. Different numbers of IL- DC-T which were in contact with IL- DC during the first culture (without transwell) were added to the proliferation assay containing freshly isolated P1-specific T cells and P1-loaded S-DC. The IL- DC-T inhibited their restimulation in a cell number dependent manner (A). Antigen-specific T cells which were not in contact with IL- DC during the first culture (transwell experiment) had no inhibitory effect (B). The results (mean ± standard deviation) are representative for different experiments.P1-loaded IL- DC and IL- DC prolonged antigen-specific cardiac allograft survivalTen million of P1-loaded BMDC administered intravenously to each Lewis rat day before they received heart allografts let them survive days longer than those animals transfused with unpulsed BMDC, which had no effect on allograft survival (Table and Fig. ). The survival of allografts from third party donors (Brown-Norway rats) was not affected by the transfer of P1-pulsed IL- DC and IL- DC (Table ).Figure 6The in vivo effect of P1-loaded IL- DC and IL- DC. P1-loaded IL- DC and IL- DC prolonged the survival of allogeneic hearts grafts from Wistar-Furth in Lewis recipients. In contrast, unpulsed IL- DC and IL- DC did not influence allograft survival. The allogeneic control group I demonstrates the natural rejection of Wistar-Furth allografts by Lewis recipients (see also Table ).Table 2The effects of allopeptide pulsed and unpulsed IL- DC and IL- DC in vivo.GroupOrigin of DCDonor of heart allograftsSurvival time (d) of heart allografts in LEW recipientsMST ± SD (d))Median (d)n1 Syngeneic Ctr---)LEW> (×)> Syngeneic Ctr + P1---LEW> (×)> Allogeneic Ctr I---WF7(×), (×). ± .. Allogeneic Ctr I + P13)---WF5(×), (×). ± .. × IL- DCLEW )WF6, (×). ± ,. × IL- DC + P1LEWWF9(×), . ± .. × IL- DCLEWWF7(×), . ± .. × IL- DC + P1LEWWF8, (×), (×). ± . Allogeneic Ctr II---BN7(×).. Allogeneic Ctr II + P1---BN7(×),. ± .. × IL- DCLEWBN7(×).. × IL- DC + P1LEWBN7(×), . ± .. × IL- DCLEWBN7(×).. × IL- DC + P1LEWBN7(×)..) Mean survival time ± standard deviation in days (d).) No DC treatment performed.) Animals were immunised with P1 days prior to transplantation.) LEW: Lewis rats, WF: Wistar Furth rats, BN: Brown Norway rats.DiscussionIn the present study, we compared the effects of types of rat BMDC on the proliferation of naïve and antigen-specific T cells in vitro and on the survival of allogeneic heart allografts. Both DC types displayed lower surface expression of MHC class II and costimulatory molecules compared to mature splenic DC. IL- DC and IL- DC had a strong inhibitory effect on responsive T cells. This suppressive effect was detectable within h after the BMDC were added to cultures with antigen-specific T cells and mature splenic DC (Fig. 3E, F). To our knowledge this is the first description of the time course BMDC needed to suppress T-cell proliferation. We found that antigen-specific T cells became anergic after incubation with IL- DC and IL- DC. The same effect has been described for human IL- modified DC on CD4+ T cells []. Anergic T cells isolated from cultures with P1-loaded IL- DC suppressed the DC-mediated activation of responsive T cells in a cell count-dependent manner. They share this suppressive effect with regulatory T cells []. We hypothesise that IL- DC and IL- DC may differ in the quality of their costimulation, with IL- DC inducing suppressive IL- DC-T whereas IL- DC do not. It should be emphasised, however, that we found no differences in the surface expression of costimulatory molecules between the two BMDC types.The results of transwell experiments showed that the contact between IL- DC and the antigen-specific T cells is a prerequisite for inducing suppressive IL- DC-T. However, we made no attempt in this study to determine whether the anergic IL- DC-T mediate their suppressive effect via cell-cell contact or by soluble factors. Vendetti et al., for example, reported that inhibition mediated by anergic murine T cells is dependent on cell-cell contact []. They also described an inhibitory effect of anergic T cells on the antigen-presenting function of mature DC.IL- DC and IL- DC, loaded with allopeptide P1, prolonged allograft survival (Fig. ). In addition, survival time can be improved by increasing the number of transferred cells. The application of million P1-pulsed IL- DC, for example, prolonged survival time to a median of . ± . days (not shown) from the . ± . days achieved with million P1-pulsed IL- DC. Allograft survival was prolonged only when the BMDC were pulsed with the immunodominant allogeneic peptide P1 involved in allograft rejection []. Our results accord with those of Chowdhury et al. [], who showed that the presentation of allogeneic peptides by tolerogenic thymic rat DC greatly prolongs allograft survival. Compared to the results of Chowdhury et al., our – day prolongation of allograft survival may seem meager, but considering the strength of the allogeneic immune response induced by alloreactive T cells and the fact that no immunosuppressive drugs were used, our findings appear very promising.ConclusionThe data suggest that rat IL- DC and IL- DC have suppressive/regulatory properties comparable to those described for immature mice BMDC. They demonstrate a strong inhibitory effect on responsive T cells, probably the consequence of a reduced surface expression of costimulatory molecules paired with the inability to deliver adequate costimulatory signals to T cells. IL- DC and IL- DC are identical in phenotype and in some of their effects, but they are different in their capacity to induce suppressive T cells. IL- DC induce T cells with suppressive/regulatory function whereas IL- DC do not. This may indicate that IL- DC and IL- DC differ in the quality of their costimulation. Further studies are necessary to test this hypothesis.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsGCT designed the study, set up the experiments, collected the data, co-drafted the manuscript and provided images and figures. AM set up the flow cytometric experiments, participated in data collection, analysed and interpreted the results, and provided images and figures. UK revised the article for scientific content. MS performed the transplantation experiments and carried out the histological analysis. AT participated in editorial support and research funding. KU assisted in the study design, experimental concept, and data interpretation. CO drafted the manuscript, designed the study, analysed and interpreted the results and provided images and figures. All authors read and approved the final manuscript.Supplementary MaterialAdditional file 1IL- DC and IL- DC do not express IL-. Shown are the results of reverse transcriptase-polymerase chain reaction (RT-PCR). The used primers are listed in Table .Click here for file
|
PMC2777393.txt
|
TITLE: Evaluating eHealth: How to Make Evaluation More Methodologically Robust AUTHORS: Richard James Lilford, Jo Foster, Mike Pringle ABSTRACT: In the third in a series of articles on evaluating eHealth, Richard Lilford and colleagues consider the evaluation of health IT systems as they are rolled out following preimplementation testing. BODY: This is the third in a series of three articles on evaluation of eHealth. Summary PointsEvaluation of information technology (IT) systems often requires a mixed methods approach.External evaluations have many advantages, especially in terms of standardisation, independence, and the possibility of using controlled before and after designs.Difficulties arise when commissioners ask external evaluations to also provide formative assessments designed to assist in the implementation itself. Under these circumstances the summative results, which encapsulate the overall benefits and harms of a system, may be rendered less generalisable.We think researchers and commissioners should resist the current fashion of asking external academic teams to combine formative with summative assessments.eHealth—the organisation and delivery of health services and information using information technology (IT) systems—is playing an increasingly important role in shaping health care systems. However, as Catwell and Sheikh described in the first article in this series [], IT systems can introduce harms as well as benefits. Catwell and Sheikh argued for a general scheme of evaluation starting with careful specification of need and pre-implementation testing in their article. This philosophy of pre-implementation testing resonates strongly with the UK Medical Research Council (MRC) framework for evaluation of complex interventions [], the tenets of safety science (which endorses the use of analytic procedures to predict the failure rate of a system still in the design phase), and established principles in the IT field where “alpha testing” is routine. But how should IT systems be evaluated as they are rolled out following pre-implementation testing? This is the aspect of eHealth we will consider in this essay.eHealth in the UKOur approach to the evaluation of eHealth has been strongly influenced by the roles we have played in commissioning evaluation research on behalf of the National Programme for Information Technology (NPFIT), an initiative by the Department of Health in England to move the English National Health Service (NHS) towards a single, centrally mandated electronic care record for patients and to connect general practitioners to hospitals. The NHS is investing several billion pounds sterling each year in IT. Investment in a programme of evaluation alongside the NPFIT programmes has provided an excellent opportunity to identify newly installed IT systems and to commission prospective studies to assess these programs as they are rolled out. Such evaluations fulfil a real need, since a recent systematic review showed that “most of the high quality literature regarding multifunctional health information technology systems comes from benchmark research institutions” and that “little evidence is available on the effect of multifunctional commercially developed systems” [] such as those that are increasingly implemented in the NHS.Because these evaluations would inevitably be commissioned under intense political and media attention, NPFIT took the principled decision to contract the University of Birmingham to commission the research independently under Department of Health procurement rules. NPFIT could thus influence what was commissioned (i.e., research topics) but not the results obtained.All research commissioners have to take some responsibility for determining the form that research takes. This is particularly so when, as in the case of the evaluation of NPFIT, it is the research commissioner, rather than the researcher, who puts the ball in play. It is the research commissioner who specifies what is to be researched, over what time scale, and with what level of resource. During the course of commissioning evaluation studies for NPFIT, we identified four tricky issues that we think both commissioners of eHealth research and eHealth researchers need to consider, namely: () which research methods are suitable for the evaluation of highly complex interventions with diffuse effects, such as IT systems; () whether it is necessary to make observations at both the patient and the system level; () whether to conduct research that strengthens or improves the intervention being evaluated (formative research) and/or research that examines the benefits or outcomes of that intervention (summative research); and () whether to evaluate research both externally and internally.The Case for Multiple Methods ResearchThere is a consensus about the evaluation of clinical treatments, such as drugs in which randomised control trials are state of the art. No such consensus exists yet for the evaluation of highly complex service interventions such as computer systems. However, we believe that the best way to evaluate eHealth is through “methodological pluralism” []–[]. That is, research commissioners and research teams need to recognise the importance of undertaking combined quantitative and qualitative work when evaluating IT systems. Quantitative research can provide important numerical information about how IT systems are performing and is important in theory building, which is necessary to understand how interventions work (not just whether they worked in a particular set of instances) and hence to inform judgements about the generalisability of results from one context to another. Qualitative research can provide information on topics such as ease of use, which will ultimately affect whether the IT system is successful. So, for example, quantitative data may show that computer decision support has little impact on clinical error, while qualitative work explains why—for instance, clinicians may experience alert fatigue. More controversially qualitative research can also contribute to parameter estimation, particularly under a Bayesian framework []. More detailed accounts of methodological pluralism can be found elsewhere []–[],[]–[].Observations at Patient and System LevelAlthough IT systems can sometimes be studied at the level of individual patients (e.g., computerised decision support) []–[], they often need to be studied at the organisational level []. In some cases, this is because IT systems simply cannot be restricted to certain individuals in a group (for example, a computerised theatre booking system). In other cases, “contamination” (where an intervention “leaks” from a person in an intervention group to one in a control group) may dilute any positive effects at the individual level []. The primary unit of analysis in evaluation of IT systems is therefore likely to be at the organisational/workgroup level (e.g., wards, hospitals, practices). In this respect, IT assessments frequently have much in common with other forms of Service Delivery and Organisation research [].Like other system-based interventions, IT may impact at many levels in the organisation and may have many effects (good or bad) at these different levels []. We have conceptualised these levels in the form of a causal chain (Figure ). IT interventions, like other patient safety and service delivery interventions, frequently need to be studied at all of these levels [] for several reasons. First, data of different types can be collated from all points on the causal chain to provide information not just on what happened (e.g., to what extent did prescribing improve) but on why it happened (e.g., did clinicians over-ride computer generated suggestions, and if so, why?). Computer systems may have negative influences on clinical consultations [],[] and time efficiency [], and may encounter cultural barriers [], or be affected by broader political forces. Feedback from decision support systems is frequently ignored. All these problems need to be tackled, and collection of information at many levels is necessary to generate theories about possible explanations and remedies. For example, in-depth studies have shown that the psychological effects of clinical computer systems are more positive when both patient and clinician can view the computer screen []../journal.pmed..g001Figure 1Causal chain showing levels where IT may impact.The potential impact of IT at different levels in a health care organisation. These boxes show endpoints that can be measured at different stages of the causal pathway. These endpoints include system effects (operational effects), effects on mediating variables, and endpoints at the patient level such as clinical errors and their sequelae.Second, the robustness of research findings is increased if a service delivery implementation, such as an IT installation, has impacted positively at many levels in the causal chain; consistent change in endpoints acts as a form of “triangulation.” That said, it should be emphasised that the final purpose of an IT system is to impact positively on the patient. The various endpoints at the service level may be necessary, but not necessarily sufficient, conditions for a positive impact at the patient level. Third, multiple measurements across the chain, including time-efficiency, throughput, and patient satisfaction are all necessary to model cost-effectiveness/cost-benefit. Finally, base-line observations across the chain provide evidence on context. For example, failure to find improvement in clinical processes is likely if practice is already very good. Likewise, prevailing attitudes in an organisation can undermine a computer system as seen at the Cedars Sinai Hospital in California [].Impact at the patient level (Figure ) can be divided into the patient experience (including satisfaction), mortality, and morbidity. An effect on mortality and morbidity may not be observed, even if the implementations are effective (i.e., there is a high chance of a false negative result). This high risk of a false negative results arises because the signal-to-noise ratio is high when mortality and morbidity are used as measures of service quality []. For this reason, it is important to collect error rates/clinical process data wherever possible. Such errors are typically much more prevalent than the outcomes they portend. The topic of measurement of clinical processes/errors has been explored in detail elsewhere []–[],[],[]. One particular problem with IT, however, is that the intervention and measuring systems are not necessarily independent. For example, replacing clinical notes with an IT system substitutes both the platform for delivery of care and the platform for measuring the quality of that care.Formative and Summative AssessmentBy formative assessment, we mean studies that provide timely feedback to those who are responsible for implementing an IT system [],[]. By summative assessment, we mean the provision of generalisable knowledge that will inform decision makers elsewhere and into the future, well beyond the life cycle of a particular application. Clearly these are not watertight distinctions. Well-conducted formative studies are seldom totally devoid of generalisable lessons, even those in which feedback is specific and immediate []. Summative research, on the other hand, tends to produce results over a longer time scale and its purpose is often directed at future IT implementations rather than at the projects that were the subject of study.Can researchers and managers have their cake and eat it by combining summative and formative research into single research projects? Our answer is only a cautious “yes.” Some of the results of assessments typically become available earlier than others. For example, effects on intermediate variables such as staff acceptability tend to be available earlier than the results of time series analyses of error rates. However, if the results of formative research are fed back into implementations, this may influence summative results. In circumstances where formative research will not be part and parcel of future implementations, summative research with formative research nested within it may yield greater estimates of benefit than future implementations shorn of a formative component.Because external evaluation will not necessarily be included in future implementations, we would prefer a world in which external assessors carry out summative rather than formative assessments. However, the funders of interventions may expect or demand feedback of interim data to inform implementations. Clearly, there is tension here—in our opinion a tension that is too often wished away under the banner of subjectivist/interpretivist research.A pragmatic policy for research commissioners under these circumstances is to commission both formative and summative assessments from the research team, but to try to identify any effects of the former on the latter. To do this it is necessary to track developments over time at intervention and control sites, noting whenever formative results are fed back to the implementation teams. The hypothesis to be tested is that improvements are incremental and then become stable over time, suggesting that the lessons of earlier applications have been incorporated in successive applications. The step-wedge design []—a design in which later adopters of an IT system act as controls for early adopters—seems to have particular promise in the evaluation of healthcare IT systems in that it has logistic and certain scientific advantages over standard parallel designs. The largest project commissioned under the NPFIT evaluation programmes (an evaluation of comprehensive hospital-based IT systems as they are rolled out) follows a step-wedge design, although as a “natural experiment” rather than randomised trial.External and Internal AssessmentsAn element of formative assessment is a tenet of good implementation policy and thus some formative internal evaluation should be intrinsic to IT implementations. Such an evaluation lies outside an externally sponsored research programme. Any formative internal evaluation carried out by the implementation team should be described, along with other features of the IT system and its context, as with any complex intervention. However, external assessments add value to the evaluations carried out in-house by the implementation teams.External assessment can provide expertise in the measurement of endpoints (for example, error rates/quality) where special expertise is needed. In quality measurement many methodological traps lie in wait for nonexperts [],[],[]. For instance, internal assessors may be (subconsciously) biased in measuring outcomes and/or they may have different levels of performance over time—learning or fatigue effects. External assessors can be masked with respect to time, place, and the hypothesis, and error rates from different epochs can be reviewed in parallel so that any learning effect can be allowed for. External assessment can also bring in contemporaneous control observations, thereby helping to distinguish between temporal trends and causal associations []. Wherever possible before and after controlled studies should be used to reduce bias []. Finally, external assessment is independent from the implementation teams and hence credible to a wider audience. Figure shows an idealised scheme depicting the development and deployment of IT systems in healthcare. Here, we illustrate the concept of formative versus summative assessment on the one hand and internal versus external assessments on the other. These types of assessments sometimes have to be combined, but we would prefer to separate them into two different paradigms: () formative assessments that are carried out by internal teams (or through collaborations between internal teams and “consultants”); and () summative assessments that are independent, operate over larger time frames, are conducted by disinterested “academic” teams, and incorporate both before and after measurements and external controls. Finally, Figure also draws attention to our belief that more attention should be given in guidelines for reporting evaluation studies in health informatics to the potential effect of feedback during the course of a study []../journal.pmed..g002Figure 2Diagrammatic representation of development and deployment of IT systems.Diagrammatic representation of the development and deployment of IT systems to demonstrate the ideas of internal versus external assessment on the one hand and formative versus summative evaluation on the other.ConclusionsIn this essay we have articulated a set of scientific principles for evaluating highly complex service interventions such as IT systems. In doing so, we have stuck closely to accepted scientific principles and have attempted to show that, with appropriate care, these principles can be incorporated into the evaluation of major service delivery interventions in the real world.The programme we have put in train to evaluate NPFIT broadly speaking embodies these principles and has managed to track interventions prospectively, thus fulfilling its main aim. The topics of investigation included in our evaluation programme are (quite properly in our view) heavily influenced by NPFIT. However, the findings of the programme are independent of NPFIT in the sense that NPFIT cannot suppress or alter them. Most, but not all, of the commissioned research was of a purely summative nature. More generally we discern a growing pressure across all service delivery evaluation projects to ask researchers to be “all things to all people” and provide both formative research for service managers and scientific data with international significance, all in the same study. We argue against this tendency. However, since managers understandably want short-term results to help with implementations rather than longer-term results to provide new knowledge for the world as a whole, they may decide, to take their cheque book elsewhere. This is a price the authors of this essay would willingly pay to ensure that the evaluation of eHealth and other complex service interventions is as robust as possible.
|
PMC2447399.txt
|
TITLE: Turning Informal Thesauri Into Formal Ontologies: A Feasibility Study on Biomedical Knowledge re-Use AUTHORS: Udo Hahn ABSTRACT: This paper reports a large-scale knowledge conversion and curation experiment. Biomedical domain knowledge from a semantically weak and shallow terminological resource, the UMLS, is transformed into a rigorous description logics format. This way, the broad coverage of the UMLS is combined with inference mechanisms for consistency and cycle checking. They are the key to proper cleansing of the knowledge directly imported from the UMLS, as well as subsequent updating, maintenance and refinement of large knowledge repositories. The emerging biomedical knowledge base currently comprises more than conceptual entities and hence constitutes one of the largest formal knowledge repositories ever built. BODY: No Body Content
|
PMC3006942.txt
|
TITLE: (E)-N′-(-Bromo--hydroxybenzylidene)--hydroxy--methoxybenzohydrazide methanol solvate AUTHORS: Shi-Yong Liu, Zhonglu You ABSTRACT: In the title compound, C15H13BrN2O4·CH4O, the two benzene rings form a dihedral angle of . ()°. An intramolecular O—H⋯N hydrogen bond is observed. In the crystal structure, molecules are linked through O—H⋯O and N—H⋯O hydrogen bonds, forming a two-dimensional network parallel to (). BODY: Related literatureFor the medicinal applications of hydrazone compounds, see: Hillmer et al. ( ▶); Zhu et al. ( ▶); Jimenez-Pulido et al. ( ▶); Raj et al. ( ▶); Zhong et al. ( ▶). For crystal structures of hydrazone compounds, see: Khaledi et al. ( ▶); Warad et al. ( ▶); Back et al. ( ▶); Vijayakumar et al. ( ▶). For related structures, see: Cao ( ▶); Xu et al. ( ▶); Shafiq et al. ( ▶). ExperimentalCrystal data C15H13BrN2O4·CH4O M r = .23Monoclinic, a = . () Å b = . () Å c = . () Åβ = . ()° V = . () Å3 Z = 4Mo Kα radiationμ = . mm− T = K0. × . × . mm Data collection Bruker SMART CCD area-detector diffractometerAbsorption correction: multi-scan (SADABS; Bruker, ▶) T min = ., T max = . measured reflections3368 independent reflections2230 reflections with I > 2σ(I) R int = . Refinement R[F > 2σ(F )] = . wR(F ) = . S = . reflections225 parameters1 restraintH atoms treated by a mixture of independent and constrained refinementΔρmax = . e Å− Δρmin = −. e Å− Data collection: SMART (Bruker, ▶); cell refinement: SAINT (Bruker, ▶); data reduction: SAINT; program(s) used to solve structure: SHELXTL (Sheldrick, ▶); program(s) used to refine structure: SHELXTL; molecular graphics: SHELXTL; software used to prepare material for publication: SHELXTL.Supplementary MaterialCrystal structure: contains datablocks global, I. DOI: ./S1600536810022063/ci5098sup1.cif Structure factors: contains datablocks I. DOI: ./S1600536810022063/ci5098Isup2.hkl Additional supplementary materials: crystallographic information; 3D view; checkCIF report
|
PMC2213643.txt
|
TITLE: Observed communication skills: how do they relate to the consultation content? A nation-wide study of graduate medical students seeing a standardized patient for a first-time consultation in a general practice setting AUTHORS: Tore Gude, Per Vaglum, Tor Anvik, Anders Baerheim, Hilde Eide, Ole B Fasmer, Peter Graugaard, Hilde Grimstad, Per Hjortdahl, Are Holen, Tone Nordoy, Helge Skirbekk, Arnstein Finset ABSTRACT: BackgroundIn this study, we wanted to investigate the relationship between background variables, communication skills, and the bio-psychosocial content of a medical consultation in a general practice setting with a standardized patient.MethodsFinal-year medical school students (N = ) carried out a consultation with an actor playing the role of a patient with a specific somatic complaint, psychosocial stressors, and concerns about cancer. Based on videotapes, communication skills and consultation content were scored separately.ResultsThe mean level of overall communication skills had a significant impact upon the counts of psychosocial issues, the patient's concerns about cancer, and the information and planning parts of the consultation content being addressed. Gender and age had no influence upon the relationship between communication skills and consultation content.ConclusionCommunication skills seem to be important for final-year students' competence in addressing sensitive psychosocial issues and patients' concerns as well as informing and planning with patients being representative for a fairly complex case in general practice. This result should be considered in the design and incorporation of communication skills training as part of the curriculum of medical schools. BODY: BackgroundEffective patient care relies on the physician's understanding of the patient's biological, psychosocial, and cultural background. In the initial consultation with a patient in a general practice setting, the optimal content of a consultation constitutes the basis for the following: reliably diagnosing biological as well as relevant psychosocial factors, exploring the patient's ideas, concerns, and expectations, and informing the patient about his or her condition including a plan for further treatment [,].The training in specific communication skills during medical school is based on the notion that independent of the physician's medical knowledge, the practice of communication skills will have a significant impact upon the character and quality of the consultation. The reason for consulting the doctor may be concerns about more than symptom-related distress []. In some studies, successful levels of communication skills have been linked to students' diagnostic efficiency, patient stress, patient compliance, medical errors, general outcome, outcome in chronic disorders, and patient and physician satisfaction [-]. However, the intermediate variable between communication skills and such outcome measures, the content of a consultation in a general practice setting, has still not been sufficiently explored.Our hypothesis has been that there is a relatively strong association between communication skills and the content of the consultation. We are well aware of the claims that focus on training in communication skills may be at the expense of basic doctoring skills. Further, if adequate somatic knowledge exists, communication skills may have only a marginal additional impact upon the content of the consultation [].One way to explore these issues is to observe students or physicians in their interaction with patients in a clinical setting. In this regard, one basic methodological problem should be addressed. Should one employ a standardized patient (i.e. an actor in a constructed but typical patient setting) or many different patients? This study design has adopted the first option so all interviewers attending the study could be compared in their skills, but the generalization of the findings may be limited to the specific case being played by the actor, even if the case in its relative complexity is fairly representative of patients in general practice. Findings indicate that a significant case-by-student interaction between communication skills and decision making exists and that individuals' communication skills vary systematically with specific cases []. The length of the test being observed may also influence the results. A -minute consultation, ordinary in a general practice setting and part of the present study design, has been found to be too short to assess generalizable interviewing skills []. Concerning background variables, we would presume that female and older students will develop communication skills promoting psychosocial issues more easily and more effectively than males and younger students.We have conducted an observational study of a consultation with a standardized and common primary care patient performed by graduate medical students from all universities in Norway. Our research questions were:. Will overall level of communication skills be related to the content part like the somatic, the psychosocial, the cancer concern, and the informing/planning indices addressed in the consultation when controlled for gender and age?. Will certain characteristics like gender, age, and level of communication skills differ between students who obtain high scores on the somatic, the psychosocial, and the information/planning indices of the consultation from those obtaining low counts?MethodsSubjectsFinal-year medical students (N = ) attending all four medical schools in Norway in the spring of were invited to participate in a test examination about two months before their graduation. The aim was to recruit students from each site. At two schools, this number of students responded to the first call by mail. At the third school, students initially responded, and five more responded after the first reminder, raising the number to . At the fourth school, only students initially responded; after two reminders, the number increased to . Thus, the sample to be investigated consisted of students, % women (N = ), aged ± (range – ).The sample has to be viewed as an "availability" sample, and we cannot know for sure how representative the attendees were of the national student population. The data from our survey sample showed that the gender proportion among the final-year students in spring (N = ) was % females (N = , ratio ., CI % . – .); in the observational sample (N = of the same ), the gender proportion of females was % (N = , ratio ., CI % . – .). Mean age among the was , (CI % , – ,) years, among the , (CI % , – ,). When we compared the scores on a self-report questionnaire assessing their self evaluation of communication skills (Oslo Inventory of Self-reported Communication Skills – OSISCS), the participants in the survey study yielded approximately the same mean score (, +. on a – scale) one year earlier (when they were in their 5th year of medical school; medical school in Norway lasts for six years) as the attendees (, +.) with minor variance between genders []. With overlapping confidence intervals regarding gender and age and close to similar scores on self-reported skills, none of these differences should be viewed as significant. Thus we presume that we have achieved a fairly representative sample.Medical schoolsThe four medical schools in Norway use identical governmental admission criteria, have the same curriculum length (six years), but have some differences in design. One of the schools offers a "traditional" curriculum divided into a pre-clinical and a clinical part. The remaining three schools do not have this sub-division, running "integrated and/or problem-based" curricula, where students see patients from the onset. Since the differences between schools are not extensive and given the limited sample size (N = ), we have run the analyses without differentiating between curricula.The standardized patientA common primary patient's clinical story with multiple problems was constructed through discussions within the research team comprising nine experienced clinicians (four GP's, one oncologist and four psychiatrists). The patient was a -year-old woman visiting this "physician" for the first time, complaining of irregular menstrual bleedings (see the case story in the Appendix). Her story also contained psychosocial distress related to recent of divorce and relocation, a stressful job and a fear of uterine cancer, of which her mother may have died years earlier. During the consultation, she was to appear as an avoiding and non-complaining person, not disclosing her concerns easily.Four professional actors, one from each of the four university cities, were instructed to simulate this patient. A professional instructor was hired to train them together in order to standardise the role as much as possible.The consultationThe consultations were carried out at each of the four schools and all interviews were videotaped. Prior to the consultation, the students completed the OSISCS and received the same instructions: They were told that they were in their obligate part of the postgraduate period working as an assistant with a general practitioner and had minutes available to see a female patient, NN, years old, for her first consultation due to irregular menstrual bleeding. Their task was to carry out a complete first-time consultation in a general practice setting. They might propose examinations/laboratory tests to be undertaken during the actual visit, getting the results then and there. Further examinations had to be planned as part of the follow-up procedure (see Appendix).Assessment of the consultation content – outcome variablesA rating instrument, Consultation Content, with dichotomous items related to consultation content for this particular patient, was constructed after thorough discussions within the research group (Table ). Two independent raters scored the videotapes according to this list of consultation content items as zero (not present) or (present) according to whether the student-physician had addressed the specific issue or not in the consultation. Thus, total content counts could range between zero and .Table 1Items used in evaluating the consultation content.Has the student performed/addressed:A. Biomedical diagnoses:. Myomas (as cause of the irregular bleedings)□ yes□ no2. Low hemoglobin□ yes□ noB. Subjective complaints3. Exhaustion□ yes□ no4. Headache□ yes□ no5. Vertigo□ yes□ no6. Sleep problems□ yes□ no7. Reduced appetite□ yes□ noC. Concern aspects8. Mother dead from uterine cancer□ yes□ no9. Fear of inherited cancer□ yes□ noD. Psychosocial aspects10. Three children□ yes□ no11. All responsibility for them□ yes□ no12. Divorced one year ago□ yes□ no13. Moved to a new place during the last year□ yes□ no14. Exploring the job situation□ yes□ no15. Problems with divorce□ yes□ noE. Informing/planning:Treatment plan (MAAS – )□ yes□ noInformation (MAAS – )□ yes□ noThe two raters obtained very similar counts and a high correlation (r = .) between them.Since the construction of the content scale was not based upon psychometric properties, we found it most reasonable to look at the face validity. The consultation content items (Table ) were divided into three Content indices: The Somatic Index comprising the two diagnostic items (myomas and anemia) together with the five illness items (numbers – ). Due to very low internal consistency (Cronbach's α = .), the two diagnostic items were excluded and with the five illness items left constituting the Illness index (), alpha increased to .. The Concern Index () comprised two items related to the patient's concern about possible cancer (numbers – , α = .), and the Psychosocial Index () comprised the six psychosocial items (numbers – , α = .).We also constructed a fourth index called The Informing/planning Index () by combining two items (item – treatment planning and item – information, α = .) from the Maastricht History and Advice Checklist (MAAS) []. This index covered the action taken by the student at the end of the consultation. MAAS items were scored on a – scale, yielding a median of ,. A cut off > was used to demonstrate presence of the item during the consultation. Inter-rater reliability between the MAAS raters was (ICC(.)) = ..We then divided the students into groups with high or low counts on the Content indices by merging the Psychosocial and Concern indices into a Psychosocial/concern index in order to simplify the classification. Together with the Illness and the Informing/planning indices, these three indices were dichotomized at their medians. By combining high and low scores on the indices, eight sub-groups emerged with different combinations (from all high to all low). These groups would then be tested against level of communication skills in order to check for correlation linearity.Assessment of communication skillsThe Arizona Communication Interview Rating Scale (ACIR) consists of items (Table ) all with a scale from (least) to (best) []. Three trained raters scored the videotapes and one separate rater scored / of them in order to check the inter-rater reliability (not the same as above). The Intra-class correlation coefficient between raters was similar to that between the MAAS raters ICC(.) = .. The overall ACIR mean was , (.). The psychometric properties of ACIR have been found satisfactory in an earlier study []. A Principal Component Analysis with a Varimax rotation yielded a typical one-factor solution, so the overall mean was used in all analyses. The internal consistency of ACIR in our sample was . (Cronbach's α). As one way of validating ACIR, we correlated ACIR-mean against the MAAS-mean (without the two items constituting the Informing/planning index being used as a content index) giving a value of Pearson's r = ..Table 2Item-list for Arizona Communication Interview Rating Scale (ACIR)12345Items(lowest score)(highest score). Organization□□□□□. Timeline□□□□□. Transitional utterances□□□□□. Open questioning□□□□□. Smooth progress□□□□□. Avoiding repetition□□□□□. Summarizing□□□□□. Understandable information□□□□□. Documentation□□□□□. Eye contact□□□□□. Attentiveness□□□□□. Response to concerns□□□□□. Feed-back□□□□□. Additional questions□□□□□The question of tautology may be raised as the rating of the ACIR items could be influenced by how the content items were rated and vice versa, even if separate raters were used. We, therefore, wished to control for this possibility using a special procedure. When carefully scrutinizing the ACIR items one by one, we categorized them as content-related and non-content-related. Items number , , , , , and were considered non-content-related. This categorization, which of course may be disputed, is highly consonant with the separation of ACIR items performed in a previous study by Aspegren et al. []. This group of items (numbers , , , , and , Cronbach's α = .) was viewed as more typical for civil social conversation, i.e. not strictly related to a specific medical context. The mean of these non-content related items could be used as a substitute for the ordinary overall ACIR-mean in the multiple analyses to test for possible differences in association with our outcome variables (Content indices).StatisticsMeans, frequencies, Anova, Linear Regression, Principal Component Analysis with Varimax rotation, and Reliability testing were conducted in the SPSS . version.ResultsThe frequency of every single content item being addressed in the consultation is shown in Table , varying from % of the consultations concerning exhaustion (Somatic) to % concerning divorce problems (Psychosocial). The content of Myomas and anemia was addressed in % and % respectively. Descriptive data concerning overall ACIR mean and total counts of Content indices present in the consultation by gender and age groups are shown in Table . No significant differences were detected between female and male students or age groups.Table 3Frequency of the consultation content items addressed during the interview.Consultation content items (numbers from Table )How frequently addressed %Exhaustion (),5Three children (),4Mother dead of ca. (),9Headache (),4Divorce (),6Concern for own ca. (),7Myomas (),8Anemia (),2Insomnia (),3Job situation (),2Moved (),4Vertigo (),7Responsibility (),9Appetite (),0Divorce problems (),9Table 4Levels of ACIR overall mean and consultation content total count by gender and age groups.ACIR overall meanAnova F-valueConsultation content countAnova F-valueWhole sample3, (.), (,)Female3, (.). n.s., (,). n.s.Male3, (.), (,)Age < yrs., (.). n.s., (,). n.s.Age – yrs., (.), (,)Age > yrs3, (.), (,)In four linear regression analyses, the four consultation Content indices (Illness, Psychosocial, Concern, and Informing/planning) were used as dependent variables and the overall ACIR-mean as the independent variable with gender and age as control variables. The level of communication skills had a significant impact upon the Psychosocial, the Concern and the Informing/planning indices, controlled for gender and age, yielding standardized Beta-values of . for the concern index (explaining % of the variance), . for the Informing/planning index (%), and . for the psychosocial index (%). The Illness index had no significant relationship to level of communication skills (Table ). When substituting the overall ACIR-mean with the ACIR non-content-related mean, no deviating associations were detected (Beta-values deviating at a maximum of .).Table 5Standardized Beta-values from Linear Regression Analyses for Arizona overall mean as predictor for the four consultation content indices controlled for gender and age.Standardized Beta-valuesIllnessPsycho-socialConcernInform/planGender0.-...01Age-....09ACIR-..**.***.**Expl.var. R20....** p < ., *** p < ., no significance = n.s.When exploring characteristics for the eight sub-groups derived from combinations of the dichotomized Illness, Psychological/concerns, and the Informing/planning part-counts, overall ACIR-mean was significantly higher in those with high score (above median) on all three indices (Illness, Psychosocial/concern, and Informing/planning) compared with those having low score on all three or high on the Illness but low on the two others (F = ,, p = .). Gender and age did not vary between the eight sub-groups.DiscussionOne of the main findings in this study is that the level of communication skills was significantly related to the total content index, especially to the Psychosocial, the Concerns, and the Informing/planning indices, but not to the Illness index, controlled for gender and age. As shown in Table , the strongest relationship was with the Concern index, while the Informing/planning and Psychosocial indices yielded somewhat weaker relationships. This implies that the higher the level of communication skills, the more adequately addressed was the content of the consultation related to the patient's possible stressors; i.e. her fear of having the cancer from which her mother may have died years earlier, responsibility for her children, the divorce she had been through, and the burden of moving to a new place, a.o. The students with the higher level of communication skills were also more able to perform adequate informing/planning procedures at the end of the consultation.These results were supported by the fact that students scoring high on all three dichotomized indices (the Illness, the Psychosocial/Concerns, and Informing/planning) had significantly higher levels of communication skills than those scoring low on all three indices, as well as those with high score on the Illness, but low scores on the other two indices. No variation in gender and age characterized these sub groups, showing that it was the level of communication skills that mattered. The risk of tautology when exploring associations between phenomena with instruments possibly covering some of the same issues has been considered and attended to by testing any possible difference when using the overall mean in contrast to the no-content-related mean of ACIR. When no such difference existed, it is reasonable to presume that the relationship we have found between communication skills and the content indices in this specific consultation was real and not confounded by lack of construct validity for the two instruments.The finding that there was no significant relationship between a high level of communication skills and a high count on the Illness index is important. This may be due to a low variance in the Illness index, but this does not coincide with Table . Another interpretation of this finding, therefore, may be that the somatic content of this particular consultation (the diagnoses of possible myomas and anemia as well as the subjective complaints) depends more on biomedical knowledge than the communication skills themselves and that the presence of these symptoms was not related to psychosocial aspects. Along the same line, we found that those who scored high on the Illness index but low on the Psychosocial and Informing/planning indices had a significantly lower level of communication skills compared with those with high scores on all three indices. This supports the view that graduate students who are very competent in evaluating somatic complaints nevertheless may fail to identify important psychosocial stressors and concerns, in this case due to the lack of communication skills. We know that these issues are important both for patient compliance and satisfaction as well as for the further course and outcome of many somatic disorders [,-]. In order to obtain valuable information on the emotional burdens of patients who may be less willing than the "patient" in this study to disclose such issues, the need for specific communication skills is obvious. In a study by Pfeiffer et al. [], a decrease in students' communication skills was found through the last part of medical school, after an initial increase, especially their skills in obtaining an adequate social history. One interpretation offered was that this can be due to a medical culture which de-emphasizes communication skills in favor of somatic knowledge. This can also be a possible explanation of our findings. Therefore it is important to identify factors that may enhance as well as hamper the development of communication skills among medical students.This study has strengths and limitations. The strength is the nation-wide sample, while a limitation can be the representativeness of our sample (N = ) compared with the available cohort of last year's students (N = ). We cannot overlook the possibility that the students who participated in our study were more motivated than those who did not attend. Such motivation could be based upon an understanding of the need to focus upon communication skills or it could be the wish to test their level of skills shortly before their final exams. With no deviance in gender proportion and mean age, nor in the level of self-assessment of communication skills that the students reported in their fifth year compared to what the out of the same students reported one year later, we may presume that the representativeness may be acceptable.Another limitation is of course the single patient case used in this study. Even if we may argue that our case story is representative for a patient consulting a general practitioner, we should be cautious in generalizing our conclusions regarding the students' communication skills and the relation to consultation content too far. As mentioned above, previous studies have shown the need for more comprehensive studies with observation of communication skills based on more variation in cases than we have used in this study [,].Another limitation could be the variation in the actors' performances of the patient role. Even if they were instructed and trained together preceding the interviews, some variation in their withdrawal behaviour could be observed in the videotapes.According to the instructions the students were given before interviewing, max. minutes should be used for interviewing. A few of them used up to minutes, but as the counts did not improve with more time, this factor was not considered important.ConclusionThis study has shown that the level of communication skills and the content of the consultation with regard to psychosocial issues, patient concerns and the informing and planning procedures (with a representative patient in a general practice setting) among graduate medical students are significantly correlated. On the other hand, no such relationship between communication skills and the somatic/illness content of the consultation was detected. Even if the students attending this study had good biomedical knowledge on this specific case, they may have overlooked important psychosocial stressors as well as the patient's concerns, and may have lacked the necessary skills to conclude this single patient consultation in a fruitful way. In this light, medical schools and researchers, utilizing a broader spectrum of patient stories, should continue to identify factors that promote as well as inhibit medical students' learning of communication skills.Competing interestsThe author(s) declare that they have no competing interests.Authors' contributionsTG participated in the design of the study, organized and took part in the data collection, scored the interviews, performed the data analyses and were the main writer, PV participated in the design of the study and gave major contributions to the writing, TA participated in the design of the study, took part in the data collection, and contributed to the writing, AB participated in the design of the study, took part in the data collection, and contributed to the writing, OBF participated in the design of the study, took part in the data collection, and contributed to the writing, HE scored the interviews and contributed to the writing, PG scored the interviews and contributed to the writing, HG participated in the design of the study, took part in the data collection, and contributed to the writing, PH participated in the design of the study, took part in the data collection, and contributed to the writing, AH participated in the design of the study, took part in the data collection, and contributed to the writing, TN participated in the design of the study, took part in the data collection, and contributed to the writing, HS scored the interviews and contributed to the writing, and AF scored the interviews and contributed to the writing.AppendixInstructions to the actor- You are going to play the role of Eva, a year old woman, divorced and working at the local library. You have an avoidant personality structure and you look pale and exhausted when you enter the office, sitting down in the chair you are offered with a heavy sigh as it is good to be seated.The most typical characteristics of avoidance are:- A pervasive pattern of social inhibition- Feelings of inadequacy- Hypersensitivity to negative evaluations- Avoiding working situations involving cooperation with others due to fear of critics- Restraint within intimate situations due to fear of being shamed or ridiculed- Views self as socially inept, personally unappealing, or inferior to othersComplaints – IllnessTo the doctor's initial question, you tell that the reason for coming is irregular menstrual bleedings. When the doctor asks you to tell more about it, you start with telling about a duration of half a year with irregular intervals and stronger, more lasting bleedings than before. On this background you asked for a consultation, and got one in two weeks.If the doctor asks you about your own opinion of the condition, you may disclose your feeling of exhaustion and thoughts about an approaching menopause. But you have been increasingly worried about something serious turning up. Up till the last year, you have had regular menstrual bleedings with days between each starting point.But the last half of a year, bleedings have appeared with form to days intervals, lasted for eight to nine days (earlier five to six days). Now you have to use double menstrual pads to avoid bleeding through, but yesterday it happened and last night you had to rise from bed in order to change pads due to fresh red blood with cloths appearing.Your first menstrual bleeding occurred when you were years old. You have had four pregnancies, given birth to three children. The fourth pregnancy terminated in a spontaneous abortion two years after your second child was born. years ago. All labours were uncomplicated. After the divorce one year ago, you have not established any new intimate relationship due to fear of one more failure. Therefore, you have stopped using prevention.You smoke cigarettes a day and use no medication except some occasional Aspirin due to headache with some relief.You have up to now had a good health with no need for consulting a doctor after you moved to this new place, where you now live, one year ago.If and when the doctor asks whether you need a sick-leave, you respond that it could be good for you, but you are not sure if you are ill enough and therefore have to give a self-report instead.If the doctor asks more about your life situation, you may tell (not necessarily disclose all of it immediately) that you have moved from your earlier living place into a new flat as the only provider for your three children because of the divorce from your husband one year ago. Your economy is strict as you have to pay high interest on your loan, but you are in balance.Your three children are: Nils years (high school), Trine years (junior high school) and Peter years (day-care centre). Your ex-husband, working on an oil-drilling platform met another woman and moved together with her far away.You have no contact with him, he calls once in a while the children. They have adapted pretty well to the new surroundings. You have not been able to establish a social network in your new living place, are afraid of taking initiative as others may think you are silly and uninteresting to join with.You are educated as a librarian and got the job with only a min. walk from home.In the beginning you liked it, but he last months you have dreaded going to work feeling insufficient. Your new boss demands everyone to be creative and starting new projects thus, favouring your peers doing this, while the more withdrawn, like yourself, become scapegoats.You try to avoid conflicts with your boss, but she is more and more critical towards you. If the doctor asks you explicitly about your real situation, you disclose your feeling of increasing bad temper, exhaustion, and problems with insomnia leading to tiredness in the morning. Your appetite is reduced, but you have not lost weight.If depressions in the family, especially mother, are explored by the doctor, you answer "not as far as I know".The last couple of months you have had some headache, had vertigo and had problems with reading as the letters have merged together.You have less energy to mobilize and have thought a lot about your mother who died only years old from "cancer in her belly" some years ago.With hesitation, you disclose fear of the same faith hitting you after you have got these menstrual problems. Uterine cervix smear two years ago was normal.If you have the opportunity you may disclose your concern about reduced ability to care for your children. Your two teen-agers are in opposition to and criticize you. You feel it as a burden when they quarrel a lot and it is difficult for you to intervene in their expanding activities. It is a lot more drugs and alcohol abuse among young people here than where you came from and you feel uncomfortable with Trine's dating a year older truck driver.Results from the investigation (to be informed about on the student's request):At the gynaecological investigation, abundant blood with some cloths occurs. The uterine body is enlarged as with eight weeks pregnancy. The consistency is flexible with a bulky surface. Both ovarian are normal. Blood pressure /, hearth rate , regular, Hb... Additional clinical investigations are normal.Pre-publication historyThe pre-publication history for this paper can be accessed here:
|
PMC2779992.txt
|
TITLE: Efficacy and safety of deferasirox doses of > mg/kg per d in patients with transfusion-dependent anaemia and iron overload AUTHORS: Ali Taher, Maria D Cappellini, Elliott Vichinsky, Renzo Galanello, Antonio Piga, Tomasz Lawniczek, Joan Clark, Dany Habr, John B Porter ABSTRACT: The highest approved dose of deferasirox is currently mg/kg per d in many countries; however, some patients require escalation above mg/kg per d to achieve their therapeutic goals. This retrospective analysis investigated the efficacy (based on change in serum ferritin levels) and safety of deferasirox > mg/kg per d in adult and paediatric patients with transfusion-dependent anaemias, including β-thalassaemia, sickle cell disease and the myelodysplastic syndromes. In total, patients pooled from four clinical trials received doses of > mg/kg per d; median exposure to deferasirox > mg/kg per d was weeks. In the overall population there was a statistically significant median decrease in serum ferritin of μg/l (P< ·) from pre-dose-escalation to the time-of-analysis; significant decreases were also observed in adult and paediatric patients, as well as β-thalassaemia patients. The adverse event profile in patients who received deferasirox doses of > mg/kg per d was consistent with previously published data. There was no worsening of renal or liver function following dose escalation. Deferasirox > mg/kg per d effectively reduced iron burden to levels lower than those achieved prior to dose escalation in patients with transfusion-dependent anaemias. This has important implications for patients who are heavily transfused and may require higher doses to reduce body iron burden. BODY: Deferasirox (Exjade®) is a once-daily oral iron chelator approved for the treatment of transfusional iron overload in patients with transfusion-dependent anaemia. During a comprehensive series of -year core (Cappellini et al, ; Galanello et al, ; Piga et al, ; Vichinsky et al, ; Porter et al, ; Taher et al, ) and ongoing (Porter et al, ) extension trials, deferasirox was evaluated at starting doses ranging from to mg/kg per d, with maintenance doses of up to mg/kg per d in the extension trials. The efficacy of deferasirox was primarily evaluated based on the assessment of serum ferritin, which remains the most widely used method for evaluating body iron burden. The measurement of serum ferritin is convenient and inexpensive, and serial measurements provide a relatively robust marker of iron burden; significant correlations between changes in serum ferritin and liver iron concentration (LIC) have been identified across various underlying anaemias (Cappellini et al, ; Porter et al, ). Current guidelines for assessing iron burden and monitoring the efficacy of chelation therapy recommend the use of serum ferritin (Thalassaemia International Federation, ).Pharmacological studies have shown that the pharmacokinetics of orally administered deferasirox at doses of , and mg/kg per d is linear and drug exposure is dose proportional in patients at steady state (Nisbet-Brown et al, ). In addition, across the clinical trial programme, response to deferasirox was shown to be dependent on dose. A subsequent analysis based on data from a single study highlighted the additional impact of transfusional iron intake on response (Cohen et al, ). In this analysis %, % and % of patients in the high (>· mg/kg per d), intermediate (·–· mg/kg per d) and low (<· mg/kg per d) iron intake categories respectively, achieved a reduction in LIC with mg/kg per d, while the equivalent proportions at a dose of mg/kg per d were %, % and % (Cohen et al, ). Therefore it is clear that some patients require dose escalation to > mg/kg per d in order to achieve their therapeutic goals, i.e. to have a sufficiently rapid decrease of iron burden in heavily iron-overloaded patients, or to successfully prevent iron accumulation in patients with acceptable iron burden levels.However, when deferasirox was first licensed in , there was a comparatively small clinical database and a limited number of patients had been exposed to doses of > mg/kg per d. As such, doses above mg/kg per d were not recommended in the prescribing information for deferasirox when the drug was first approved. As a larger number of patients have now been exposed to deferasirox doses of > mg/kg per d for a prolonged period of time, a retrospective analysis has been performed to investigate the efficacy and safety of such doses. Data for this analysis have been pooled from four prospective clinical trials in patients with various transfusion-dependent anaemias, including β-thalassaemia, sickle cell disease (SCD) and the myelodysplastic syndromes (MDS).MethodsEligibility criteriaMale or female patients aged ≥ years with transfusional iron overload, as defined by a LIC of ≥ mg Fe/g dry weight (dw) at the start of deferasirox treatment, were eligible for inclusion in this analysis. All patients were required to have baseline serum creatinine levels below the upper limit of normal (ULN) and were excluded if alanine aminotransferase (ALT) levels were > U/l in the previous year. More detailed enrolment criteria for the four studies have been published elsewhere (Cappellini et al, ; Vichinsky et al, ; Porter et al, ; Taher et al, ).Institutional Review Board or Ethics Committee approval was obtained at each participating institution. All patients (or parents/guardians) provided written informed consent. All studies were conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki and were closely monitored by Novartis personnel or a contract organization for compliance to the protocols and the procedures described in them.Study design and dosingThis pooled analysis was performed using data from the core and ongoing extension phases of four prospective multicentre, open-label trials in patients who received deferasirox > mg/kg per d (Cappellini et al, ; Vichinsky et al, ; Porter et al, ; Taher et al, ); a summary of these trials is provided in Table I. All four trials used comparable efficacy and safety monitoring and assessment methods and are therefore suitable for pooled analysis.Table ISummary of trials included in this analysis.Novartis trial numberDesignPatient populationPatients exposed to deferasiroxPatients contributing to this analysis, n (%)CICL670A0107E (Study ) Cappellini et al ()Randomized, comparative deferasirox versus DFO ( year), followed by deferasirox only ( years)Adult and paediatric patients with β-thalassaemia major55599 (·)CICL670A0108E (Study ) Porter et al ()Single arm of deferasirox, non-comparativeAdult and paediatric patients with MDS, DBA and other anaemias of diverse aetiologies18434 (·)CICL670A0109E (Study ) Vichinsky et al ()Randomized, comparative deferasirox versus DFO ( year), followed by deferasirox only ( years)Adult and paediatric patients with SCD18532 (·)CICL670A02402E (Study ) Taher et al ()Single arm of deferasirox, non-comparativeAdult and paediatric patients with β-thalassaemia major237* (·)DFO, deferoxamine; MDS, myelodysplastic syndromes; DBA, Diamond-Blackfan anaemia; SCD, sickle cell disease.*Fifteen patients from a single site were excluded from the final analysis because routine monitoring of study documents at the site could not confirm the accuracy of the data reported (Taher et al, ).Initial deferasirox doses in studies – were – mg/kg per d, depending on baseline LIC; doses above mg/kg per d were not permitted (Cappellini et al, ; Vichinsky et al, ; Porter et al, ). In the extension studies, investigators were allowed to increase the doses above mg/kg per d up to mg/kg per d in patients in whom the control of body iron was judged inadequate after review/approval by the Safety Monitoring Committee. Dose increases in steps of – mg/kg per d were recommended after a minimum of months’ treatment for patients with baseline serum ferritin of > μg/l who had an increasing serum ferritin trend, or for those with baseline serum ferritin of > μg/l who did not have a serum ferritin decrease.During the -year core phase of study , all patients started on deferasirox mg/kg per d except three who started on mg/kg per d (Taher et al, ). Doses of > mg/kg per d were not initially recommended in the extension study, although this was subsequently changed following a protocol amendment. Dose increases of – mg/kg per d were recommended for patients who had confirmed rises in serum ferritin of ≥ μg/l above baseline, or for patients whose serum ferritin remained at > μg/l without showing a downward trend.AssessmentsEfficacy was assessed based on the monthly serum ferritin observations before and after escalation to doses of > mg/kg per d.Safety was evaluated throughout treatment after escalation to doses of > mg/kg per d, based on the incidence and type of adverse events (AEs) and serious AEs, and routine laboratory testing including serum creatinine and liver enzyme levels.Patient populations and statistical methodsTo evaluate patients who were intended to receive deferasirox > mg/kg per d and to avoid variability based on tablet strength (as some patients may have unintentionally received doses of > mg/kg per d due to the fact that the only tablet strengths available are , or mg), only patients with a prescribed dose of > mg/kg per d and an actual daily dose of ≥· mg/kg per d were analysed. The term ‘doses of > mg/kg per d’ used in the subsequent text is defined based on these two criteria.All efficacy analyses are based on patients treated with doses of > mg/kg per d, as defined previously, and who had serum ferritin assessments at study baseline and at least once after escalation to doses of > mg/kg per d. All safety analyses are based on patients treated with doses of > mg/kg per d who had at least one safety assessment performed after escalation to doses of > mg/kg per d. Transfusional iron intake was calculated according to the method previously described by Cohen et al ().Laboratory changes over time were summarized by averaging the observed values over -month periods. The reference point prior to dose escalation was calculated by averaging the observed values over the previous months. Data were mainly summarized descriptively. Graphical representations were also used to follow change over time; the boxes in Figs and display the 10th and the 90th percentiles and the medians are connected. The last observed serum ferritin value was compared to pre-dose escalation values using paired Wilcoxon tests when the number of patients with both a pre-dose escalation and post-dose escalation value was >.Fig 1Relative change (%) in serum ferritin levels from pre-dose escalation (efficacy population).Note: The boxes represent the 25th and 75th percentiles, while the whiskers correspond to the 10th and 90th percentiles. The medians are connected.ResultsPatient characteristicsOf patients who received deferasirox across the four studies, (·%) received doses of > mg/kg per d. Most of these patients had their doses escalated due to inadequate control of serum ferritin levels. This population was composed primarily of patients with β-thalassaemia (Table II). Median serum ferritin levels at pre-dose escalation in paediatric and adult patients were and μg/l, respectively.Table IIPatient characteristics prior to dose escalation.Variable> mg/kg per d (n = )Mean age ± SD, years19· ± ·4Median (range)· (–)Age, n (%)< years113 (·)≥ years151 (·)Male:female129:135Race, n (%)Caucasian138 (·)Black28 (·)Oriental56 (·)Other42 (·)Underlying anaemia, n (%)β-thalassaemia225 (·)MDS–DBA/rare anaemias7 (·)SCD32 (·)Median serum ferritin (range), μg/lStart of deferasirox treatment3937 (– )Pre-dose escalation3880 (– )MDS, myelodysplastic syndromes; DBA, Diamond-Blackfan anaemia; SCD, sickle cell disease.Transfusional iron intakeFor the patients who had their doses escalated to > mg/kg per d, overall median transfusional iron intake was · mg/kg per d for the months preceding dose escalation, which remained constant once doses had been escalated.Deferasirox dosingAfter dose escalation, patients (·%) received an average deferasirox dose of <· mg/kg per d, patients (·%) an average dose of ≥· to <· mg/kg per d, patients (·%) an average dose of ≥· to <· mg/kg per d and three patients (·%) an average dose of ≥· mg/kg per d. Overall median exposure to deferasirox was weeks, while median exposure from the first to last administration of deferasirox > mg/kg per d was weeks. The median of the last dose prior to escalation was · mg/kg per d, while the median dose after escalation was · mg/kg per d. Twenty-six patients (·%) had dose reductions, primarily decreases according to the study protocol (e.g. body weight change), while patients (·%) had drug interruptions, mostly because of missed doses. The median duration of drug interruption was d (range –).Change in serum ferritin levelsOverall, patients (·%) met the requirements for the efficacy analysis. Pre-escalation serum ferritin levels ranged from c. to μg/l across all underlying anaemias (Table III). In this efficacy population, the median relative change in serum ferritin at the time-of-analysis was −·% compared with the level preceding dose escalation to > mg/kg per d (Fig ; Table III). In these patients there was a statistically significant median decrease in serum ferritin of μg/l (P< ·; Table III) from pre-dose-escalation to the time-of-analysis.Table IIIMedian of relative and absolute change in serum ferritin at last observed assessment following dose escalation to > mg/kg per d, by subgroup (efficacy population).Populationn*Pre-escalation serum ferritin (range), μg/lRelative change from pre-escalation, %Absolute change from pre-escalation, μg/l [% CI]P-value†All patients2613880 (– )−·− [−, −]<·0001β-thalassaemia2223794 (– )−·− [−, −]<·0001SCD324516 (– )−·− [−, ]nsDBA/rare anaemias73469 (– )−·− [−, ]–Adults1503930 (– )−·− [−, −]·001Paediatrics1113843 (–)−·− [−, −]<·0001DBA, Diamond-Blackfan anaemia; SCD, sickle cell disease; CI, confidence intervals; ns, non-significant.*Patients with available pre- and post-escalation serum ferritin levels.†P-value based on paired Wilcoxon test, absolute serum ferritin change versus pre-escalation.In patients with β-thalassaemia, deferasirox doses of > mg/kg per d significantly reduced serum ferritin at the last observation time to below the levels prior to dose escalation by μg/l (P< ·; Table III). In patients with SCD, median serum ferritin levels decreased by μg/l. In paediatric patients (aged < years), dose escalation to > mg/kg per d significantly reduced serum ferritin levels by μg/l compared to the levels prior to dose escalation (P< ·, n= ; Table III). Similarly, doses of > mg/kg per d also provided significant reductions in serum ferritin levels by μg/l in adult patients (P= ·, n = ). There was no significant difference between paediatric and adult patients with respect to the relative median change in serum ferritin from pre-escalation to last observed assessment (−· vs. −·; P= ·).Safety and tolerabilityForty-nine (·%) patients had AEs leading to either dose reduction or drug interruption ( patients had drug interruption, three patients had dose reduction and one patient had both). Median dose prior to interruption was · mg/kg per d, while median dose prior to reduction was mg/kg per d. Thirty-two of these patients continued treatment with deferasirox until the completion of the study. Of the remaining patients, nine patients still were treated with deferasirox at the cut-off date of this pooled analysis, three patients discontinued the study due to unsatisfactory therapeutic effect as assessed by the investigator, two patients discontinued due to AEs (see section below), two patients withdrew consent, and one patient was lost to follow-up.Study discontinuationsAt the time of analysis patients (·%) were still on study, while (·%) had completed the extension studies. Forty-seven dose-escalated patients (·%) permanently discontinued due to: insufficient change in serum ferritin as assessed by the investigator (n = ), AEs (n = ), consent withdrawal (n= ), lost to follow-up (n= ), abnormal laboratory values (n= ) and administrative problems (n= ). The AEs leading to discontinuation were increased ALT, lenticular opacities (classed as a serious AE), increased serum ferritin, cardiac failure in two patients (serious AEs), abdominal discomfort, hypotension (serious AE), dyspnoea (serious AE), gastrointestinal haemorrhage (serious AE) and marrow transplantation due to β-thalassaemia (serious AE). Four of the AEs resulting in discontinuation (lenticular opacities, increased serum ferritin, abdominal discomfort, gastrointestinal haemorrhage) were suspected by the investigators to be drug-related. Median dose prior to discontinuation was · mg/kg per d; the median time between study discontinuation and the last dose was d (range − to ).Adverse eventsOverall, (·%) and (·%) patients experienced an AE before and after escalation to > mg/kg per d, respectively. Most AEs were mild to moderate in nature. Ten patients (·%) only had AEs after dose escalation, three of whom received doses > mg/kg per d at baseline; nine had baseline serum ferritin levels > μg/l.Forty-three patients (·%) had a drug-related (investigator-assessed) AE after dose escalation to > mg/kg per d. The most common drug-related AEs before and after escalation are shown in Table IV. Of the eight patients who had an ALT increase after dose escalation, two had the same AE before dose escalation and four had elevated ALT before receiving deferasirox. None of the three patients who had a serum creatinine increase after dose escalation had this AE before escalation.Table IVNumber (%) of patients with drug-related adverse events as assessed by investigators (observed in > patient after dose escalation to > mg/kg per d).Frequency, n (%)AEBefore dose escalationAfter dose escalationTotal exposure (patient years)··4ALT increase* (·) (·)Vomiting19 (·) (·)Nausea26 (·) (·)Abdominal pain17 (·) (·)Upper abdominal pain4 (·) (·)Blood creatinine increase* (·) (·)Abdominal discomfort2 (·) (·)Diarrhoea16 (·) (·)AST increase* (·) (·)Transaminase increase* (·) (·)ALT, alanine aminotransferase; AST, aspartate aminotransferase.*Reported as an AE by the investigator.Forty-two patients (·%) had a serious AE following escalation to > mg/kg per d; those assessed as possibly drug-related were lenticular opacities, increased transaminases, gastrointestinal haemorrhage, abdominal pain and abnormal liver function tests (n= for each). After escalation there were no cases of pancreatitis or obstructive jaundice, one (·%) case of gastritis, one (·%) case of cholecystitis and five (·%) cases of cholelithiasis; no cases of gastritis, cholecystitis or cholelithiasis were considered to be drug-related. There were no deaths among patients who received doses of > mg/kg per d.Renal parametersAfter escalation to > mg/kg per d, the median of relative change in serum creatinine remained unchanged at the end-of-study compared with pre-escalation (Fig ). Two patients (·%) had serum creatinine above the ULN, at two consecutive assessments more than d apart after dose escalation. One of them had normal serum creatinine levels prior to dose escalation. One patient had a urine protein:creatinine (UPUC) ratio >· mg/mg at two consecutive visits after escalation to > mg/kg per d. Eight patients (·%) had a UPUC ratio >· mg/mg at one visit, seven of whom had a normal UPUC ratio prior to dose escalation. Twenty-seven patients (·%) had one episode of abnormal UPUC ratio <· mg/mg, of whom had a normal ratio prior to dose escalation.Fig 2Serum creatinine levels before and after dose escalation to doses > mg/kg per d (safety population).Note: The boxes represent the 25th and 75th percentiles, while the whiskers correspond to the 10th and 90th percentiles. The medians are connected.Hepatic safetyNine patients (·%) had ALT levels of > × ULN but < × ULN at two consecutive assessments at least d apart, all of whom had ALT >ULN prior to escalation to doses > mg/kg per d. Three patients, two of whom who had ALT < ULN prior to dose escalation, had ALT values of > × ULN at two consecutive visits after escalation.Auditory and ophthalmic assessmentsOne patient, who permanently discontinued treatment, had a lenticular opacity assessed as drug-related after escalation to doses of > mg/kg per d ( mg/kg per d); this patient had a normal eye examination at baseline. Six additional patients had ocular abnormalities suggestive of impaired vision (four patients with blurred vision and one each with reduced visual acuity and papilloedema) reported after dose escalation; none were assessed as drug related by the investigator. In / patients who had an ocular AE of interest, ophthalmological examination revealed normal (n = ) or clinically insignificant (n = ) results at the onset of the AE. The lenticular opacity experienced by one patient was detected by ophthalmologic examination. In two of the remaining four patients, the AEs were preceded and followed by normal ophthalmological examinations. Ophthalmological examinations were not available in the remaining two patients. Four patients reported some hearing loss after dose escalation to > mg/kg per d, although none discontinued and one already had hearing loss on lower deferasirox doses. One of these episodes was assessed as drug related.DiscussionPrevious studies have indicated that some patients require deferasirox doses of > mg/kg per d in order to achieve their therapeutic goals. Here we report for the first time that such doses are effective with a comparable safety profile to doses of < mg/kg per d. The patients who were escalated to doses of > mg/kg per d presented with very high median serum ferritin levels prior to escalation ( μg/l overall); patients (·%) had levels > μg/l, which has been associated with an increased risk of iron overload-related complications (Olivieri et al, ; Olivieri & Brittenham, ) and shown to require intensification of chelation therapy (Thalassaemia International Federation, ).Despite limited exposure (median weeks), this analysis demonstrates that escalation of deferasirox to doses of > mg/kg per d effectively decreased serum ferritin compared with levels prior to dose escalation in patients whose serum ferritin levels were not adequately controlled before escalation. The serum ferritin decreases were significant in the overall population, in adult and paediatric patients, as well as patients with β-thalassaemia. Although patients with SCD and other anaemias were included in this analysis, the patient numbers were too low to provide any meaningful information. This may suggest that iron overload in these populations was sufficiently well managed on deferasirox doses of ≤ mg/kg per d and they generally did not require escalation to > mg/kg per d. However, more than half of the SCD patients (n = , %) had drug interruptions (≥ in % of patients) after dose escalation, which may have impacted on the serum ferritin response. It is also worth noting that the measurement of serum ferritin levels is a less reliable marker of iron overload in SCD than in other anaemias. This is because ferritin is an acute phase reactant therefore levels can be greatly influenced by inflammation (Brownell et al, ), which is an important factor in the pathophysiology of SCD.Patients who required escalation to doses of > mg/kg per d were heavily iron overloaded, which was reflected by high serum ferritin levels at baseline. These patients also received an intermediate amount of transfusional iron intake (– ml/kg per month) after dose escalation. This emphasizes the need to dose deferasirox based on goal of therapy, iron burden and transfusional iron intake, and highlights that dose escalation may be required in patients with severe iron overload who do not respond sufficiently to doses up to mg/kg per d.The safety profile observed in patients who received deferasirox doses of > mg/kg per d was consistent with that observed in the -year core trials (Cappellini et al, ; Galanello et al, ; Piga et al, ; Vichinsky et al, ; Porter et al, ). Importantly, there were no AEs observed following escalation to > mg/kg per d that were not observed at lower doses before escalation. Serum creatinine levels remained unchanged at the end of study period compared with the pre-escalation level, which suggests that the risk of impaired renal function is low while administering doses of > mg/kg per d in heavily iron-overloaded patients. Forty-two patients experienced a serious AE following dose escalation to > mg/kg per d, which was consistent with those observed prior to dose escalation and in the -year core trials (Cappellini et al, ; Galanello et al, ; Piga et al, ; Vichinsky et al, ; Porter et al, ; Taher et al, ) and typical of those expected in the target population. The safety of doses > mg/kg per d was not assessed in patients with serum ferritin levels ≤ μg/l as only one patient had such low levels at dose escalation. It is therefore not possible to comment on the safety of doses > mg/kg per d in patients with low levels of iron loading. However, in general, high doses should be reserved for patients with very high levels of iron loading who are not responding to doses < mg/kg per d.Limitations of this study include the fact that it was a retrospective analysis and patients who were included were likely to be those who tolerated doses < mg/kg per d, which is important to consider when reviewing the safety data.In conclusion, these findings indicate that escalation of deferasirox doses to > mg/kg per d effectively reduces iron burden to lower levels than those achieved prior to dose escalation, without an increase in the incidence of AEs or a worsening of renal or liver function. Deferasirox dose should be titrated to meet individual patient needs as dictated by transfusional iron intake and the goal of therapy (reduction or maintenance of body iron stores). Before considering dose escalation to > mg/kg per d to reduce body iron burden in patients with severe iron overload, it is important to assess both the response to therapy (serum ferritin monitoring) and compliance at doses ≤ mg/kg per d. The dose-escalated patients should be regularly monitored and, once body iron stores fall into an acceptable range, deferasirox dose should be decreased to an appropriate maintenance level.
|
PMC3004823.txt
|
TITLE: -oxo-,-dihydro-'-deoxyguanosine as a biomarker of oxidative damage in oesophageal cancer patients: lack of association with antioxidant vitamins and polymorphism of hOGG1 and GST AUTHORS: Stéphanie Lagadu, Mathilde Lechevrel, François Sichel, Jean Breton, Didier Pottier, Rémy Couderc, Fathi Moussa, Virginie Prevost ABSTRACT: BackgroundThe present report was designed to investigate the origins of elevated oxidative stress measured in cancer patients in our previous work related to a case-control study ( cases, controls) on oesophageal cancers. The aim was to characterize the relationship between the levels of -oxo-,-dihydro-'-deoxyguanosine (-oxodG), antioxidant vitamins and genetic susceptibility.Methods8-oxodG was analysed in peripheral blood mononuclear cells (PBMCs) by High Performance Liquid Chromatography with Electrochemical Detection (HPLC-ED). Analysis of gene polymorphisms in GSTM1 and GSTT1 was performed by multiplex PCR and in GSTP1 and hOGG1 by a PCR-RFLP method. Reversed-phase HPLC with UV detection at nm was used to measure vitamins A and E in serum from the same blood samples.ResultsWe observed that in our combined population (cases and control, n = ), there was no statistically significant correlation between the levels of -oxodG and (i) the serum concentration of antioxidant vitamins, vitamin A (P = .) or vitamin E (P = .), or (ii) the incidence of the Ser326Cys polymorphic variant (P = .) of the hOGG1 gene. Also, the levels of -oxodG were not significantly associated with polymorphisms in metabolite-detoxifying genes, such as GSTs, except for the positive correlation with Val/Val GST P1 allele (P < .).ConclusionsThe weakness of our cohort size notwithstanding, vitamins levels in serum and genetic polymorphisms in the hOGG1 or GST genes do not appear to be important modulators of -oxodG levels. BODY: BackgroundOesophageal cancer remains an important public health concern worldwide with an estimated burden of , new cases in []. The two major histological types of oesophageal cancers, squamous cell carcinoma (SCC) and adenocarcinoma (ADC) differ substantially in their underlying patterns of incidence and key etiologic factors. Alcoholism and smoking are the major established risk factors for SCC, whereas Barrett's oesophagus or gastro-oesophageal reflux disease (GORD) are consistently associated with an increased risk of ADC.Oxidative stress and reactive oxygen species (ROS) are thought to play a role in oesophageal carcinogenesis. ROS may result from external factors such as smoking, and alcohol metabolism, or may be produced endogenously via inflammatory conditions such as oesophagitis or GORD or may also be due to precancerous lesions (Barrett's oesophagus), as has been shown experimentally in rats [,]. Diet influences incidence of oesophageal cancers. An adequate diet of fruits and vegetables is associated with a decreased incidence [], presumably due to a better supply of antioxidants.Among the various markers of oxidative stress, -oxo-,-dihydro-'-deoxyguanosine (-oxodG) is particularly popular. It is generated by the oxidation of DNA under physiopathological conditions or environmental stress, but is also a by-product of normal cellular metabolism. It is a premutagenic oxidized-DNA lesion as it is able to mispair with adenine, thus generating G:C to T:A transversion mutations, unless the lesion is repaired prior to DNA replication []. Moreover, affordable analytical methods are available for its quantification. In population studies, -oxodG is often determined in DNA extracted from white blood cells, peripheral blood mononuclear cells (PBMC) or lymphocytes, although PBMCs being mostly lymphocytes, the distinction is rarely made. These cells are considered to be representative of the whole organism in terms of the level of exposure of to oxidative stress. However, it has been suggested that the apparent high levels of -oxodG could be due to artefactual oxidation of DNA during the treatment of the samples. The European Standards Committee on Oxidative DNA Damage (ESCODD) has now been set up within the European laboratory network to improve and harmonise -oxodG measurement methods [-].In a previous study [], we have described the optimisation of an analytical procedure to measure -oxodG in PBMCs by using HPLC coupled with electrochemical detection (HPLC-ED). In that study [], the protocol was applied to the analysis of -oxodG in PBMCs of subjects (n = ) from a case-control study that included both, SCC and ADC cases. Control samples (n = ) exhibited . ± . molecules of -oxodG per unaltered guanosines, levels which correspond to the median values reported by the latest ESCODD trial for HPLC measurement in lymphocytes from healthy young men []. In comparison, oesophageal cancer patients (n = ) showed higher oxidative DNA damage as indicated by the -oxodG levels of . ± . per , '-dG (Student's t-test, P < .). This difference remained significant even after technical (storage, sampling period, '-dG levels) and individual (age, sex, smoking, alcohol) confounding factors were taken into account (P < ., generalized linear regression model). Moreover, data on smoking habits and alcohol consumption of the volunteers were available, and could be correlated with the observed levels of oxidatively-damaged DNA.The aim of the present study was to characterize the relationship between the levels of oxidative stress, antioxidant vitamins and genetic constitution in oesophageal cancers. An elevated level of oxidative DNA lesions could be related to exogenous or endogenous parameters. Therefore, factors that may influence the extent of oxidative DNA damage such as the nutritional status and genetic polymorphisms were included in this study.Antioxidant vitamins, such as vitamin A and vitamin E are effective free radical scavengers and can also be useful markers of antioxidant status. Presumably, a higher production of ROS due to severe oxidative stress, characteristic of oesophageal cancers, could lead to a higher metabolic consumption of the antioxidant vitamins, and this would be reflected in their lower serum levels. This "antioxidant hypothesis" was examined in the subjects included in our study by determining the serum concentrations of vitamins A and E.Oxidatively damaged bases in DNA are preferentially repaired by base excision enzymes. The hOGG1 gene encodes the human -oxo-guanine DNA glycosylase that cleaves the -oxo-guanine base from damaged DNA. The single-nucleotide polymorphism at codon (Ser326, rs ) is the most well-studied polymorphism of hOGG1. Homozygous carriers of the variant form of the hOGG1 Ser326Cys gene appear to have a reduced capacity to repair oxidised DNA lesions [-], although this is controversial []. A genetic susceptibility toward SCC of the oesophagus linked to the Ser326Cys polymorphism in the hOGG1 gene has been described []. We have measured this polymorphism in our population and correlated it with the corresponding -oxodG level.Polymorphism also exists in genes encoding enzymes involved in the metabolism of xenobiotics that act as an indirect source of free radicals. Genetic polymorphisms in genes involved in detoxification such as glutathione-S-transferases (GST), GSTM1, GSTT1 and GSTP1 could potentially affect the susceptibility of an individual to the adverse effects of environmental risk factors involved in oesophageal cancer. The above three genes, are expressed in the oesophageal mucosa. We have earlier reported a significant increase in the risk of oesophageal cancers correlated with the null genotypes of GSTM1 and GSTT1 but not with the GSTP1 Ile/Val polymorphism []. The polymorphisms in the GST genes were analysed according to their histological status, among controls and cases of oesophageal cancers. These polymorphisms were revisited in the present study to investigate their correlation with the levels of -oxodG.MethodsPatients and controlsFollowing an approval from the ethical committee (Comité Consultatif pour la Protection des Personnes en Recherche Biomédicale, Basse-Normandie), consenting patients and control subjects were recruited between and within the context of a case-control study aimed at identification of various biomarkers suitable for molecular epidemiology of oesophageal cancers []. The control group (n = ) included healthy donors, who had no clinical history of chronic diseases or cancer and were living in the Lower Normandy, France. Seventeen oesophageal cancer patients from the University Hospital of Caen, France, were selected based on the availability of biological samples. Diagnosis was performed at the Department of Hepato-gastroenterology, University Hospital of Caen, France, and the Department of Anatomopathology, François Baclesse Center, Caen, France. Out of the patients, presented with SCC, with ADC and with leiomyoma, a rare histological subtype. All cases were newly diagnosed and previously untreated. Individual data related to age, sex, alcohol consumption and smoking habits of the subjects have been published earlier [] and are summarized in Table . Twenty ml of venous blood samples were collected before performing any procedure such as surgery, radio- or chemotherapy. The PBMCs were separated and used for quantification of -oxodG and genetic polymorphisms from blood samples of all individuals (n = ), while the serum was used for quantification of the vitamins A and E from all except three samples (n = ), for which the volumes were insufficient.Table 1Description of individual dataParameterControls ()Patients ()Age (years) ± ± 9Sex men/women27//3Smoking (cigarettes per week) ± 79a113 ± 97bAlcohol (grams/week) ± 140a289 ± 258bNumbers represent values for mean ± standard deviation; a: n = ; b: n = .PBMC collection, DNA isolation and hydrolysisCare was taken to avoid artefactual oxidation of DNA during its extraction and hydrolysis. PBMCs were isolated from ml out of the ml blood samples using Unisep Maxi tubes (Novamed). These were stored in liquid nitrogen until being used for DNA isolation. Latter was performed using the "protocol G" described by Ravanat et al. [] with modifications aimed at optimisation of the analytical procedure with minimum delays []. Other modifications included addition of desferrioxamine to extraction and digestion buffers.-oxodG HPLC-ED analysisAn optimised method for the quantification of -oxodG in PBMCs has been described previously []. Briefly, the DNA hydrolysate was analysed by HPLC with an electrochemical detector (Coulochem II; ESA Inc., Chelmsford, MA) using a Supelcosil reversed-phase C18 HPLC column ( × mm, μm -Supelco) equipped with a C18 guard column. The eluant was mM potassium dihydrogen phosphate, pH ., containing .% methanol, at a flow rate of . ml/min. The potentials applied to the analytical cell (ESA ) were + mV and + mV for E1 and E2, respectively. 'dG was measured in the same run of corresponding -oxodG with a UV detector (Pharmacia LKB VWM ) at nm situated after the ED cell. Acquisition and quantitative analyses of chromatograms were carried out using Eurochrom software (Knauer). The amount of -oxodG in DNA was calculated as the number of -oxodG molecules/ unmodified 'dG.HPLC determination of serum vitamin A and EConcentrations of vitamins A and E were measured in the sera obtained from the blood samples of all subjects, except for ( control, patients). The serum fraction was obtained after the isolation of PBMCs from blood by centrifugation at × g for min. Samples from control and cancer subjects were stored in the same conditions, at -°C for several years until analysis.Simultaneous determination of vitamin A and E was performed by HPLC as previously described [], with the following modifications. The HPLC system consisted of a Summit Dual Gradient System including a diode array detector from Dionex (Voisin le Bretonneux, France). The stationary phase consisted of a LiChroCART® - LiChrospher® RP-, μm protected by a guard column filled with the same stationary phase both from Merck Chemicals, France. The mobile phase consisted of methanol and the flow rate was . ml/min. Separations were carried out at °C. Vitamin A and E peaks were integrated at nm and the specificity of the detection was based on retention factors and comparison of UV-Visible spectra with those collected from the standard samples. For the calibration of the method and for quality control we used lyophilised standard and quality control serums from RECIPE (Munich, Germany). μL of samples of either serum or a standard solution or quality control sample, were added to μL of a solution of ethanol containing tocopheryl acetate ( μM) that was used as an internal standard. After stirring the mixture for seconds, the vitamins were extracted with μL of hexane ( min of stirring). The organic phase was evaporated under nitrogen and the residues dissolved in μL of methanol and μL were injected into the chromatograph. All procedures were performed in a room with glass windows that prevented penetration of direct sunlight.GSTM1, GSTP1, GSTT1 and hOGG1 genotyping analysisDNA was extracted by the phenol-chloroform method using an aliquot out of the ml venous blood samples of the subjects. Determination of GSTM1, GSTP1 and GSTT1 polymorphisms in the subjects was performed as previously described []. Analysis of deletion polymorphism in GSTM1 and GSTT1 was performed by multiplex PCR and that of single nucleotide polymorphism in GSTP1by a PCR-RFLP method as previously described []. In addition to these polymorphisms, subjects were also genotyped for the presence of either the serine or cysteine codon at position (rs ) of the hOGG1 gene by PCR-RFLP, using primers and conditions as previously described []. Briefly, the PCR amplification of the bp fragment consisted of a -min denaturation at °C followed by cycles of °C for min, °C for min and °C for min. A final extension step of °C for min was included. We used a simple RFLP method to identify the Ser326Cys by virtue of an Fnu4HI restriction site. The hOGG1 PCR product was digested with Fnu4HI overnight at °C. Recovery of two digested fragments (/124bp and /170bp) indicated presence of the Cys326 allele, while an undigested amplicon indicated the Ser326 allele.Statistical analysisAll statistics and graphics have been performed with the SAS System release (SAS Institute Inc., Cary, NC, USA). Distributions of -oxodG were normalised by logarithmic transformations. Mean values were compared by Student's t-test or ANOVA and correlations between -oxodG and antioxidants were evaluated by Pearson correlation test. All statistical analyses were two-sided.ResultsBlood levels of -oxodG and vitamins A and EThe mean serum concentrations of vitamin A were . μM and . μM, while those for vitamin E were . μM and . μM, in patients and controls respectively (Table ).Table 2Biochemical parameters of the study groupParameterPatients (mean ± s.d.)Controls (mean ± s.d.)P-valuebpatient vs. control8-oxodG/ 'dGc7. ± . (n = ). ± . (n = )P < .001Vitamin A (μM). ± . (n = )a2. ± . (n = )aP = .895Vitamin E (μM). ± . (n = )a38. ± . (n = )aP = .204PBMCs were collected and processed for measuring -oxodG. Vitamins were extracted from the serum samples for estimation. a Numbers in parentheses indicate the number of samples analysed as some samples had to be omitted due to insufficient volume of serum collected. A total of out of the samples were analysed for vitamins. b Student's t-test was applied. c Values taken from our data published earlier [].There was a high intersample variability in the levels of vitamins across subjects, as indicated by the wide range of values. The mean values in the subjects were in the range of values reported recently by others for these vitamins [-]. There were no significant differences in the levels of vitamins A and E between the control and cases. Further, there was no significant correlation found between the levels of -oxodG and those of vitamin A (R = .; P = .) or vitamin E (R = .; P = .) when cases and controls were combined (Pearson correlation test, two-sided). However, a positive correlation between the levels of -oxodG and vitamin A (R = .; P = .) and vitamin E (R = .; P = .) was observed when only cases (n = ) were taken into account (Figure ).Figure 1Correlation between -oxodG levels and vitamin A (a) and vitamin E (b) in cancer patients group (n = ). -oxodG level is expressed as the number of molecules of -oxodG per 'dG; R = . and P = . for correlation between -oxodG and vitamin A; and R = . and P = . for correlation between -oxodG and vitamin E; Log of -oxodG (Y-axis) is plotted against vitamin A and E concentrations as indicated; circles, values for individual data; full line, linear regression; dotted line, % confidence limit.Levels of -oxodG and hOGG1 polymorphismThe potential relationship between -oxodG and the Ser326Cys polymorphism in the hOGG1 gene was examined in the pooled population of cases and controls. Comparisons of means of -oxodG between genotypes were done with ANOVA after logarithmic transformation. As shown in Figure , there was no statistically significant association between levels of -oxodG in DNA and hOGG1 Ser326Cys polymorphism (P = .). The prevalence of the Cys allele, hOGG1326Cys, was . in the controls and . in the cases (Table ).Figure 2Levels of -oxodG according to hOGG1 genotypes. Data from patients and controls were combined (n = ) and analyzed by ANOVA (P = .). -oxodG level is expressed as the number of molecules of -oxodG per 'dG and Log of -oxodG (Y-axis) is plotted against frequencies of hOGG1 genotypes as indicated. circles, values of individual sample.Table 3Genotype frequency of hOGG1 Ser326Cys in patients with oesophageal cancerGenotypeControls (n = )(%)Patients (n = )(%)Total (n = )(%)Ser/Ser22 () () ()Ser/Cys19 () () ()Cys/Cys2 () ()Cys allele frequency0...22Numbers in parentheses represent the relative percentage in the group.Levels of -oxodG and frequency of GST allelesThe level of -oxodG according to GST polymorphisms was examined in the pooled population after logarithmic transformation. There was no difference with the null genotypes of the GSTM1 (Student t test; P = .), and GSTT1 (Student t test; P = .), whereas there was a strong difference between GSTP1 variants (ANOVA, P < .) (Figure ).Figure 3Levels of -oxodG according to genotypes of GSTM1, GSTP1 and GSTT1. Data from patients and controls were combined (n = ). -oxodG level is expressed as the number of molecules of -oxodG per 'dG and Log of -oxodG (Y-axis) is plotted against frequencies of the various genotypes as indicated, GSTM1 (P = .), GSTP1 (P < . for Val/Val vs Ile/Ile and Ile/Val) and GSTT1 (P = .); circles: values for individual data.DiscussionOxidative damage to DNA is considered to be an important risk factor for carcinogenesis. -oxodG is a key biomarker in this process because it is one of the most frequently encountered product of oxidatively-damaged DNA and also one that can be easily detected in samples of tissues or urine [-]. We have previously reported a significantly higher level of -oxodG in circulating blood cells from oesophageal cancer patients compared to control subjects []. Similar observations have been made for colorectal carcinoma [], lung cancer [,,] and leukaemia [,]. In our study, none of the individual variables such as smoking, alcohol, sex or age, was shown to influence -oxodG concentrations. The aim of the present study was to identify other factors that could modulate -oxodG levels. We have attempted to characterize the relationship between oxidative stress, evaluated in terms of levels of -oxodG in PBMCs, and the levels of antioxidant vitamins and the genetic constitution, in a population consisting of healthy volunteers and oesophageal cancer patients.Vitamin C, vitamin E, carotenoids, and other antioxidants present in fruits and vegetables could contribute to cancer prevention by protecting DNA from oxidative damage, according to the "antioxidant hypothesis". By inference, the endogenous levels of these antioxidant vitamins in the serum of oesophageal cancer patients are expected to be low. Likewise, under conditions of severe oxidative stress also, their serum levels may be low as these would be consumed in redox reactions involving ROS.Many recent epidemiological studies have confirmed that a high intake of fruits and vegetables is associated with a decreased risk of upper aero-digestive tract cancers [,-]. One of the possible mechanisms of this protective effect is the antioxidant activity of vitamins A, C and E. These vitamins are effective antioxidants in vitro, and might be expected to protect against cancer. Calişkan-Can et al. [] found lower levels of β-carotene and vitamins A, C and E in lung cancer patients compared to healthy controls. Foksinski et al. [] observed that the mean levels of all the measured antioxidant vitamins were significantly lower in smokers in comparison with non-smokers. Gackowski et al. [] reported that vitamin E levels were significantly reduced in the plasma of lung cancer patients (smokers and ex-smokers) compared to healthy smokers and that the levels of vitamins A and E in plasma of colorectal carcinoma patients were lower than in the control group []. In contrast, other studies found no significant differences between healthy smokers and non-smokers for either serum vitamin A or vitamin E [,,]. Also, several large-scale antioxidant supplementation trials have failed to show any clear evidence for a decrease in cancer risk [,]. In our study, we found that the endogenous serum levels of vitamins A and E were similar in oesophageal cancer patients and in controls. Notably, despite the fact that our study subjects come from the same geographical area, there was substantial intersample variability, especially for the cancer cases. These differences could reflect the balance between absorption and tissue secretion, and may also be genetically determined. A recording of dietary habits (fruit and vegetable consumption) could have added a complementary and an interesting feature to our study. Determination of the two major water-soluble antioxidants, ascorbate and glutathione would also have brought complementary information. However, as particular conditions are required for sample collection, processing and storage to prevent their oxidation and degradation, these could not be analysed in this retrospective study.Correlation between the levels of vitamins and -oxodG has been reported. In their analyses of cross-sectional studies, Moller & Loft [] identified studies showing an inverse correlation between oxidatively damaged DNA and antioxidant levels, reporting no correlation and two, a positive correlation. A lack of a correlation between -oxodG and antioxidant vitamins has also been reported by others [,,]. In a recent paper, Sram et al. [] found a negative correlation between -oxodG and β-carotene and vitamin E but a weak positive association with vitamin A. Similar positive correlations were reported for vitamin A in chemical workers exposed to vinyl chloride monomer [], carotenoids and vitamin E []. We did not find any correlation between the levels of -oxodG and vitamins in our study group (cases and controls combined).Interpretation of these correlative data must be made with extreme caution because the precise effects of antioxidants on mutagenesis and carcinogenesis remain unclear. An antioxidant, including a vitamin antioxidant, is essentially a redox (reduction-oxidation) agent that provides protection against free radicals, but may promote free radical generation under certain circumstances or may exert pro-oxidant effects. Conversely, recent meta-analysis on supplementation trials indicates increased risk of mortality [], suggesting a pro-oxidant activity at high doses or in cancer-risk subjects (smokers and workers exposed to asbestos). Further, besides vitamins, several food-derived phytochemicals or bioreactive substances in natural products likely contribute to the protection of DNA from damage and thus influence the generation of -oxodG in vivo. The number of such antioxidants exceeds that of antioxidant vitamins. The availability of these unidentified antioxidants in individual diet could thus affect the correlation between levels of -oxodG and antioxidant vitamins. Some dietary components also could up-regulate DNA repair without having any recognised antioxidant function. Interestingly, a positive association was observed in our study between the levels of -oxodG and those of the two vitamins, but only in the cases and not in the controls. However, this observation should be interpreted with caution, in the light of the foregoing discussion. Moreover, to arrive at a more convincing conclusion, our data would have to be expanded and adjusted for possible confounders such as age which can become the predominant, independent determinant of oxidative damage as has been discussed recently [].In view of the conflicting reports in the literature and the results of the present study, the "antioxidant hypothesis" seems open to criticism. Is there indeed a relationship between antioxidant vitamins and oxidatively-damaged DNA? Secondly, are the concentrations of antioxidants and -oxodG in the blood representative measures of the situation in the target tissue of the carcinogenesis and a true reflection of overall cellular DNA damage? Thirdly, do we have reliable tools to examine this correlation? The choice and reliability of biomarkers such as -oxodG has also been debated [,,]. The reliability of -oxodG is influenced by its method of detection since its artefactual production is a serious concern. Notably, the values of -oxodG reported in this study are low and reach the background level of -oxodG recommended by ESCODD for HPLC-ED measurement, indicating that these were not an artefact.It is known that individuals have different responses to oxidative damage and that the risk for oxidative stress-related cancer varies according to both, the environmental exposure and the genetic background. The human -oxoguanine DNA glycosylase1 (hOGG1) is one of the major enzymes involved in DNA base excision repair (BER). A positive relationship between hOGG1 mRNA expression and -oxodG suggests that the expression level of hOGG1 may be interpreted as a biomarker of exposure to oxidative DNA damage [,]. On the other hand, some studies indicated that there was no interaction between these parameters [,,], which could be explained by the fact that hOGG1 is weakly expressed in certain tissues such as the aerodigestive tract tissue [].The activity of hOGG1 can be impaired by a polymorphic mutation at codon , the hOGG1 Ser326Cys polymorphism. However, the phenotypic impact of hOGG1 Ser326Cys polymorphism is unclear. Recent data have shown that the hOGG1 Ser326Cys polymorphism is associated with a reduced DNA repair capacity for oxidatively-damaged DNA [], whereas according to others, the converse is true [,,]. This led to the conclusion that both, wild type and the hOGG1Cys326 variant-encoded proteins should be functional and probably do not exhibit significant differences in repair activities and hence the polymorphism at codon would probably be neutral [,].Many epidemiological studies have investigated the association of the Ser326Cys polymorphism in the hOGG1 gene indicating an increased risk for head and neck cancers but the reports are conflicting [,,]. Studies on the prevalence of this polymorphism in susceptibility to oesophageal cancer also show conflicting results. Xing et al. [] reported a positive association between the Cys326 variant and oesophageal cancer risk in Asians population whereas Tse et al. [] reported no association in Caucasians. In the present study, the small number of samples did not allow us to make a comparison of the genotype distribution between cases and controls in order to determine whether the hOGG1326Cys allele contributed to the risk of oesophageal cancer. However, the distribution of hOGG1 Ser326Cys genotype in our controls (.) is in agreement with the frequencies previously described in Caucasian population. This frequency is classically lower than that in Asians [,,], suggesting that this allele may be differently distributed among ethnic groups and may not confer a particular susceptibility to oesophageal cancer in Caucasian population. The allelic distribution of this polymorphism in our combined population followed Hardy-Weinberg equilibrium.Besides DNA repair activity, enzymes involved in the detoxification of xenobiotics such as glutathione S-transferases may influence the extent of oxidative damage in humans. We genotyped our study population for the GSTM1, GSTT1 and GSTP1 genes. Our results indicate no association between GSTM1 and GSTT1 null polymorphisms and -oxodG levels in DNA from PBMCs. On the other hand, we found a statistically significant association between GSTP1 Val/Val homozygote carriers and a high level of -oxodG (Figure ). However, as no obvious relationship was found between the frequency of the Val allele (Val/Val and Ile/Val combined) and the level of -oxodG, we consider this result questionable. Indeed, correlation of GST polymorphisms with -oxodG levels in WBCs or lymphocytes varies with the context of exposure: polycyclic aromatic hydrocarbons [,], benzene [], fine particulate matters [] and hyperbaric oxygen [].ConclusionsIn conclusion, although the power of our study is limited, it seems likely that vitamin levels in serum and polymorphisms in the hOGG1 or GST genes are not important modulators of -oxodG levels. Further studies in a larger cohort of patients using other biomarkers of oxidative damage and antioxidant status are required to better understand the possible involvement of oxidative stress in oesophageal cancer.List of abbreviationsADC: adenocarcinoma (ADC); ESCODD: European Standards Committee on Oxidative DNA Damage; GST: glutathione-S-transferase; HPLC-ED: HPLC coupled with electrochemical detection; PBMC: peripheral blood mononuclear cells; SCC: squamous cell carcinoma; -oxodG: -oxo-,-dihydro-'-deoxyguanosine.Competing interestsThe authors declare that they have no competing interests.Authors' contributionsSL: genotyping analysis of polymorphisms, data analysis; ML: interpretation of data concerning polymorphisms, critical revision for important intellectual content; FS: conception and design of the study, interpretation of data, final approval of the version to be published; JB: analysis of -oxodG, interpretation of data, critical reading of the manuscript; DP: statistical analysis of the data; RC: interpretation of data concerning vitamins and critical reading of the manuscript; FM: analysis of vitamins and interpretation of these data; VP: coordination of project, interpretation of data and writing of manuscript. All authors have read and approved the final manuscript.
|
PMC3032976.txt
|
TITLE: High levels of genetic variability and differentiation in hilsa shad, Tenualosa ilisha (Clupeidae, Clupeiformes) populations revealed by PCR-RFLP analysis of the mitochondrial DNA D-loop region AUTHORS: Sabuj Kanti Mazumder, Md. Samsul Alam ABSTRACT: The hilsa shad, Tenualosa ilisha (Clupeidae, Clupeiformes) is an important anadromous clupeid species from the Western division of the Indo-Pacific region. It constitutes the largest single fishable species in Bangladesh. Information on genetic variability and population structure is very important for both management and conservation purposes. Past reports on the population structure of T. ilisha involving morphometric, allozyme and RAPD analyses are contradictory. We examined genetic variability and divergence in two riverine (the Jamuna and the Meghna), two estuarine (Kuakata and Sundarbans) and one marine (Cox's Bazar) populations of T. ilisha by applying PCR-RFLP analysis of the mtDNA D-loop region. The amplified PCR products were restricted with four restriction enzymes namely, XbaI, EcoRI, EcoRV, and HaeIII. High levels of haplotype and gene diversity within and significant differentiations among, populations of T. ilisha were observed in this study. Significant FST values indicated differentiation among the river, estuary and marine populations. The UPGMA dendrogram based on genetic distance resulted in two major clusters, although, these were subsequently divided into three, corresponding to the riverine, estuarine and marine populations. The study underlines the usefulness of RFLP of mtDNA D-loop region as molecular markers, and detected at least two differentiated populations of T. ilisha in Bangladesh waters. BODY: IntroductionThe hilsa shad, Tenualosa ilisha, belonging to the sub-family Alosinae of the family Clupeidae (Clupeiformes, Pisces), occurs in foreshore areas, estuaries, brackish-water lakes and freshwater rivers of the western division of the Indo-Pacific faunistic region. Its marine distribution extends from Iran and Iraq in the Persian Gulf to the west coast of India in the Arabian Sea and the Bay of Bengal (Pillay and Rosa Jr, ).Hilsa shad is the largest single fishable species in Bangladesh, present in almost all the major river systems, estuaries and marine environments (Bay of Bengal), and at present contributing to approximately % of the total fish production and % of fishery-capture (inland and marine) with a biomass of , metric tons from inland fisheries and , metric tons from marine (DoF, ). Hilsa is termed the national fish of Bangladesh, due to its popularity and economic importance. However, the production of hilsa in Bangladesh has waned when compared to earlier estimates. Until , hilsa shad fishery was prosperous upstream in the rivers of Bangladesh, especially in the Padma and Meghna. Fishery has entered into a severe decline upstream and is nowadays mainly concentrated downstream, as well as in estuaries, coastal areas and the sea (Nurul Amin et al., ). Due to the low water discharge from the upstream Ganges at the Farakka barrage (in West Bengal, India) with the consequent heavy siltation, the indiscriminate exploitation of juveniles (Jatka), disruption of migration routes, loss of spawning, feeding and nursing grounds, increased fishing pressure, etc., have all contributed to this decline.The hilsa shad is largely an anadromous species, but two other ecotypes - a fluvial potamodromous type and a marine type - have been recognised. The potamodromous stocks appear to remain in the middle reaches of the rivers throughout the year and breed therein. The anadromous stocks, whose normal habitat is the lower region of the estuaries and the foreshore areas, ascend the rivers during the breeding season and return to the original habitat after spawning (Raja, ). The upstream migration during the main breeding season depends largely on the commencement of the south-west monsoon and consequent flooding of the major rivers of Bangladesh, Burma and India. However, the exact spawning season for the species is still controversial, as spawning varies from a few months to all the year round. However, it is not known whether migratory populations mix during migration or whether they pass each other spatially and temporally. Therefore, the exact stocks are still in dispute.Reports on the stock structure of the valuable tropical shad T. ilisha, based on samples collected from Bangladesh and India and through morphometric, allozyme and RAPD analyses, are contradictory. Dahle et al.() claimed to have discriminated three populations of hilsa shad in Bangladesh waters, such as Chandpur (Meghna river), Barguna (Brackish water, estuarine) and Cox's Bazar (sea water) by using RAPD markers. On the basis of a single polymorphic locus, PGM, Rahman and Naevdal () claimed that there existed two separate gene pools of hilsa shad in Bangladesh waters. Milton and Chenery (), on the basis of otolith chemistry and morphometry, and Salini et al. (), on the basis of allozyme and morphometric analyses, inferred, however, that there was a single stock of hilsa in Bangladesh waters, this including the Bay of Bengal. Brahmane et al. () identified two groups of T. ilisha in India, one comprising the Ganges and Yamuna rivers and the other the Hooghly and Narmada rivers, by using RAPD markers.Among the different markers available for population genetic analysis, differences in mitochondrial DNA are probably the most widely used, since they follow maternal inheritance, do not undergo rearrangements or recombination and present higher mutation rates than those of nuclear genes (Avise, ). RFLP analysis of the mtDNA control region is a simple technique for revealing genetic variation in the mitochondrial genome of an organism.In this study, we analyzed haplotype and gene diversity in five samples of T. ilisha, using PCR-RFLPs of the mtDNA D-loop region as genetic markers. The aim of the present study was to compare mtDNA genetic diversity in T. ilisha samples collected from rivers, estuaries and the sea, and to delineate genetic differentiation among populations of the three major aquatic environments. We also report here whether there are any genetic differences between distant river populations of T. ilisha, as well as between estuarine populations. The purpose of the present study is to examine the usefulness of mitochondrial D-loop region diversity to supplement allozyme and RAPD data on the population genetic structure of T. ilisha.Materials and MethodsFish samples and extraction of DNANinety hilsa (T. ilisha), with an average body weight of about g, were collected from five geographic locales: Balashi (Gaibandha, upstream of the Jamuna River)), Chandpur (middle of the Meghna River), Kuakata (estuaries), Sundarbans (estuaries) and Cox's Bazar (Bay of Bengal) with the help of artisanal fishers (Figure ). Immediately after landing the fish on board, a piece of dorsal fin from each was carefully collected and preserved in % ethanol. The samples were stored at - °C until further use for the extraction of DNA. Total DNA (nuclear + mitochondrial DNA) of each individual fish was extracted from approximately mg of fin tissue by standard proteinase-K digestion, phenol:chloroform:isoamyl alcohol extraction and the alcohol precipitation method as described by Islam and Alam (). The quality of the DNA samples was checked by electrophoresis on % agarose gel and the quantity determined by using a spectrophotometer, prior to PCR amplification of the mtDNA D-loop region.Amplification of mitochondrial DNAThe . kb D-loop region of the mtDNA, including a part of the cytochrome b and 12SrRNA genes, was amplified from each of the samples by using the primer-pairs L- (' → ', CAT ATT AAA CCC GAA TGA TAT TT) and H- (' → ', ATA ATA GGG TAT CTA ATC CTA GTT T) (Martin et al., ). The mtDNA fragment was amplified in a reaction volume of μL containing ng template DNA, μL of 10X buffer ( mM Tris-HCl, pH = ., mM MgCl2, mM KCl), μL of . mM dNTPs (dATP, dCTP, dGTP, dTTP), μM of each primer and units of Taq DNA polymerase (GENEI Bangalore, India). An oil-free thermal cycler (Master Cycler Gradient Eppendorf, Germany) was used for PCR amplification with the following cycle parameters: initial denaturation at °C for . min, followed by cycles of denaturation at °C for s, annealing at °C for s and extension at °C for . min. A final extension step at °C for min followed the last cycle. The PCR products were confirmed by electrophoresis on agarose gel and subjected to digestion with restriction endonucleases.RFLP analysisFour restriction enzymes (XbaI, EcoRI, EcoRV and HaeIII) were used for digesting the amplified fragments. Ten microlitres of the PCR product were used for each digestion following manufacturer's (Promega) recommended conditions. The restricted PCR products were electrophoresed on .% agarose gel containing ethidium bromide and photographed under UV light using a Geldoc camera (UVP Inc). Two molecular weight markers, λ-DNA-HindIII digest and bp DNA ladder were run on each gel along with the digested PCR products. The sizes of mtDNA restriction fragments were measured by using the software, DNAfrag (Nash, ).Genetic data analysis and dendrogram constructionRestriction patterns generated from each restriction endonuclease were given letter designations in the order of their frequencies. Haplotype A refers to the most common digestion pattern in the analyzed samples. The remaining alphabetical profile names (B, C, etc.) indicate variant digestion patterns reflecting their frequencies in order. Composite haplotypes were constructed from all the enzymes used and arranged from the enzyme generating the fewest polymorphic patterns to that generating the most (XbaI, EcoRI, EcoRV and HaeIII in this order). The presence or absence of restriction sites in the control region was inferred for each of the four enzymes from a series of restriction fragment patterns. The site codes across the amplified region for a restriction enzyme were concatenated and each fish was assigned a code of four letters that described its composite, multi-enzyme haplotype.A single data matrix comprising the composite haplotypes and their frequencies in the five populations was constructed. Haplotype and gene diversity, both calculated as per Nei (), and a hierarchical analysis of population subdivision performed using the analysis of molecular variance with simulated samples (AMOVA, Excoffier et al. ), were implemented in ARLEQUIN v. . (Excoffier et al., ). Both the Tajima () D-test and the Fu () Fs-test were applied to test deviations from neutral molecular evolution. Significance was assessed in both cases by generating random samples (number of simulated samples: ) under the hypothesis of selective neutrality and population equilibrium, as implemented in ARLEQUIN. Significance levels of pair-wise FST, under the hypothesis of no differentiation between populations, were determined by means of permutations of haplotypes between populations. Genetic distance values between population-pairs were calculated from the FST values of the respective population-pair by using the formula: D (genetic distance) = -log( - FST) (Reynolds et al., ). Distance data were used to draw up an unweighted pair-group method of averages (UPGMA) dendrogram by using MEGA version (Tamura et al., ) software.ResultsPCR amplification of the D-loop region resulted in a product of approximately . kb with no detectable size differences between T. ilisha samples. Using this sequence, four restriction enzymes (XbaI, EcoRI, EcoRV and HaeIII) were selected to obtain RFLP markers. All enzymes produced polymorphic banding patterns and restriction sites. Restriction of the PCR fragment with the four restriction endonucleases resulted in a total of restriction profiles in the samples collected from five locations in Bangladesh. The cleavage patterns and estimated lengths of the restriction fragments for different restriction enzymes are shown in Table . The cleavage patterns produced due to variations in restriction sites were three for XbaI and EcoRI, five for EcoRV and seven for HaeIII. In some cases, however, the sum of the fragment sizes did not exactly equal the total size of the amplified region, probably due to small fragments being lost or bands of similar size co-migrating. The different haplotypes (composite genotypes) detected in individuals of the five populations, along with their numerical frequencies, are presented in Table .Out of the haplotypes, only two, VI and XI, were distributed randomly among the five populations, the others being either specific for a particular population or shared by two or three (mostly two) populations (Table ). Haplotypes II, III, IV, V, VII, IX and X were specific to the Balashi population, XII and XIII to the Chandpur, XVII, XVIII, XIX, XXI, XXIII, XXIV, XXV and XXVI to the Cox's Bazar, XXVII, XXVIII, and XXIX to the Sundarbans and XXXV to the Kuakata. Haplotype I was shared by the Balashi, Chandpur and Sundarbans populations, VIII by the Balashi and Chandpur, XVI by the Chandpur and Sundarbans and XX by the Cox's Bazar and Kuakata. Haplotype I was dominant in the Balashi and Chandpur population, haplotype VI in the Chandpur, haplotypes XVII and XX in the Cox's Bazar and XXVIII in the Sundarbans. None of the haplotypes were found to be dominant in the Kuakata population.Polymorphic haplotypes were observed in all the five populations. The rates of haplotypes (number of haplotypes observed divided by the sample size) ranged from . (Chandpur) to . (Sundarbans) (Table ). Haplotype diversity was high in all the five populations, ranging from . (Chandpur) to . (Sundarbans). The average gene diversity across the loci was highest in the Sundarbans population (.) and lowest in the Kuakata (.). The overall Fixation Index (FST) across the populations was . and the population specific FST indices ranged from . (Sundarbans) to . (Chandpur). Tajima's D test (Tajima, ) and Fu's F test (Fu, ) were not significant (p > .) in all the populations.Variance components and F-statistics analogs (Phi-st) were calculated with AMOVA. No significant differences among samples were found. Hierarchical analyses indicated that .% variation was contained within local populations while .% was distributed among the populations. However, significant differentiation (FST) values were found among the five populations of T. ilisha. The FST value between the Chandpur and Sundarbans populations was the highest and that between the Chandpur and Balashi the lowest. The FST values between the Balashi-Chandpur, Cox's Bazar-Kuakata, and Sundarbans-Kuakata population pairs were insignificant, while these values between the remaining seven were found to be significant (Table ). Genetic distance values calculated from the corresponding FST values between the population-pairs are shown in Table . The highest genetic distance was observed between the Chandpur and Sundarbans population, while the lowest was observed between the Chandpur and Balashi population. The UPGMA dendrogram based on genetic distance resulted in two clusters. The first cluster contained four populations and the second contained only the Cox's Bazar population. The first cluster was subsequently divided into two sub-groups. The two river populations, Balashi and Chandpur, formed one group and the two estuarine populations, Kuakata and Sundarbans, the second (Figure ).DiscussionMitochondrial DNA-RFLP has been proved to be an effective technique for population discrimination. The present study was an attempt to reveal genetic variability and stock discrimination of hilsa populations in Bangladesh waters, including freshwater, estuarine and marine environments, by PCR mtDNA RFLP analysis.The detection of different haplotypes in only individuals of five T. ilisha samples underlines the usefulness of RFLPs of the D-loop region as molecular markers for investigating the geographic structure of the species. High diversity indices were obtained within each sample (Table ). The average haplotypic diversity in T. ilisha observed in the present study, fell within the upper part of the range (.-.) for some other fishes as reported by Avise et al. (), and reached as high as that (.) obtained in red seabream collected from four locations of western Japan through PCR-RFLP analysis of the mtDNA D-loop region by Tabata and Mizuta (). However, since the estimation of haplotype diversity is based on haplotype frequencies alone, it is dependent on the number of restriction enzymes used (Graves and McDowell, ). These results suggest that genetic variability in T. ilisha is quite high. High levels of genetic diversity appear to be commonly observed in migratory fishes with large panmictic populations (Santos et al., ). This is because large effective population size and high migration rates minimize the effect of genetic drift as a force that lowers intra-population genetic variability. However, the presence of private alleles indicates that these populations are not completely panmictic. The non-significant values (p > .) obtained from Tajima's D and Fu's Fs statistical tests indicate that the sampled populations of T. ilisha are in genetic equilibrium, which means that apparently there is no pressure of selection on the population with regard to mitochondrial DNA haplotypes.The test for population differentiation gave significant p-values (p < .), indicating that composite haplotypes were not distributed randomly with respect to locality. F-statistics (Weir and Cockerham, ) yielded an overall FST value of . which indicates that genetic exchange occurring among the populations was not sufficient to prevent either genetic differentiation or structuring into genetically differentiated subpopulations in T. ilisha in Bangladesh. Geographic differentiation of the hilsa shad T. ilisha has also been reported previously. Brahmane et al. () identified two groups of hilsa in India, one comprising the Ganges and Yamuna rivers and the other the Hooghly and Narmada rivers. Dahle et al. () reported the existence of three stocks of hilsa shad in Bangladesh waters, namely the Chandpur (Meghna river), Barguna (Brackish water, near Kuakata)) and Cox's Bazar (salt water), when using RAPD markers. Our results suggest there are at least two, one marine, and the other riverine-estuarine, or possibly three (riverine, estuarine and marine), stocks of hilsa in Bangladesh. The second possibility complies with the hypothesis of three groups of hilsa shad, anadromous, potamodromous and marine, as reported by Pillay and Rosa Jr (). Now, the question is whether the T. ilisha feeding in the Chandpur area migrate further upstream to spawn in the Padma and the Jamuna rivers or not. No significant difference in FST value between the samples from Balashi (the Jamuna) and Chandpur (the Meghna) suggest a single panmictic riverine population of T. ilisha in Bangladesh. However, this conclusion may be contradicted through the presence of some private haplotypes in the Balashi and Chandpur populations. These results also run counter to those of Shifat et al. (), who reported that the two river populations of T. ilisha, the Meghna and the Padma were different stocks, though they observed certain alleles shared by some individuals of these two river samples. The presence of private haplotypes indicates the extent of mixing between populations. Due to their locations (Figure ) and the mode of migration of the species, the possibility of mixing in the Chandpur (the mid Meghna river), Kuakata and Sundarbans populations is higher, reflected by the lower number of private haplotypes in these populations. Higher numbers of private haplotypes have been obtained in the two most distant populations, Balashi and the Cox's Bazar (Table ), which indicates that the low number of samples is not the main cause of the high number of private haplotypes in these populations. Therefore, we would like to opine that certain differences exist among the river populations, perhaps due to preferential movements of a fraction of the migratory groups, although the difference is not high enough to distinguish the populations as separate stocks (sub-populations).We observed significant differentiation between the riverine and marine (Cox's Bazar) populations, but not between the marine and one of the estuarine. There is no evidence of spawning in the sea, but there is evidence of T. ilisha estuarine spawning in Bangladesh (BOBP, ). Thus, fishes from the Cox's Bazar region must migrate to any of the rivers or at least up to the estuaries. Our results suggest that fishes from the Cox's Bazar region may go up to the Kuakata region, but not to the Meghna and Jamuna rivers, as significant differentiations between the Cox's Bazar and the two river populations have been observed, but not between the Cox's Bazar and Kuakata (Table ).The five samples were grouped into two clusters on the basis of genetic distance among population-pairs. Nevertheless, the two clusters were finally divided into three groups corresponding to the three different environments, riverine, estuarine and marine (sea). The slight distance between the Balashi and Chandpur population indicates their belonging to the same stock. However, the sharing of two haplotypes by all the populations has led us to postulate the presence of a certain degree of gene flow among those studied.Hilsa shad (T. ilisha) is the national fish of Bangladesh. The delicious taste and contribution of this fish has lead to its position as one of the most economically important fish in this country. Furthermore, hilsa fishery constitutes the largest single species fishery in the riverine, estuarine and marine ecosystems of the country. It is essential to recognize that geographically and genetically different populations as different stocks should be managed separately (Carvalho and Hauser, ). Our studies on mtDNA PCR-RFLP in T. ilisha indicate that population sub-division does indeed exist in this species. On the basis of the AMOVA, we can conclude that there are three stocks of hilsa with a substantial level of inter-population genetic divergence among river, estuarine and marine populations. The high level of haplotype variability found in only five populations underlines the usefulness of RFLPs of the D-loop region as molecular markers to investigate the geographic structure of T. ilisha. We analyzed only individuals from each population, and the electrophoretic analysis of restriction fragments is also not so perfect as sequencing the PCR product. Therefore, to reach a more definite conclusion, larger samples from throughout the distribution range of the species in the country should be analyzed by sequencing the mtDNA D-loop region and/or with faster evolving molecular markers, such as micro-satellite loci.Figure Map of Bangladesh showing the five sites (•) from where the samples of T. ilisha were collected. The three big rivers, the Padma, Jamuna and Meghna, are also shown. These three rivers constitute the major riverine fishery of T. ilisha in Bangladesh. R: river.Figure UPGMA dendrogram summarising the genetic distance between pairs from the five different populations of T. ilisha.
|
PMC2033638.txt
|
TITLE: Treatment with para-chlorophenylalanine antagonises the emetic response and the serotonin-releasing actions of cisplatin in cancer patients. AUTHORS: A. B. Alfieri, L. X. Cubeddu ABSTRACT: To test the role of serotonin in chemotherapy-induced nausea and emesis, ten cancer patients were pretreated with the serotonin synthesis inhibitor para-chlorophenylalanine (PCPA). PCPA ( g hourly for or days prior to cisplatin) reduced the spontaneous urinary excretion of -hydroxyindoleacetic acid (-HIAA), inhibited the increase in urinary -HIAA induced by cisplatin and markedly attenuated the acute period of nausea and vomiting associated with the cytotoxic drug. These results indicate that gastrointestinal serotonin mediates cisplatin-induced emesis and that the amount of serotonin released by cisplatin is a major factor in determining the severity of the acute period of emesis experienced by the patient. BODY:
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.