UBC Faculty Research and Publications

Proteomic analysis of the secretome of Leishmania donovani Silverman, J M; Chan, Simon K; Robinson, Dale P; Dwyer, Dennis M; Nandan, Devki; Foster, Leonard J; Reiner, Neil E Feb 18, 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-13059_2007_Article_1897.pdf [ 1.23MB ]
JSON: 52383-1.0228388.json
JSON-LD: 52383-1.0228388-ld.json
RDF/XML (Pretty): 52383-1.0228388-rdf.xml
RDF/JSON: 52383-1.0228388-rdf.json
Turtle: 52383-1.0228388-turtle.txt
N-Triples: 52383-1.0228388-rdf-ntriples.txt
Original Record: 52383-1.0228388-source.json
Full Text

Full Text

Open Access2008Silvermanet al.Volume 9, Issue 2, Article R35ResearchProteomic analysis of the secretome of Leishmania donovaniJ Maxwell Silverman*†‡, Simon K Chan§¶¥, Dale P Robinson†#, Dennis M Dwyer¥, Devki Nandan*†, Leonard J Foster†# and Neil E Reiner*†‡Addresses: *Department of Medicine (Division of Infectious Diseases), University of British Columbia, Faculty of Medicine, 2733 Heather St, Vancouver, British Columbia, V5Z 3J5, Canada. †Vancouver Coastal Health Research Institute, 2647 Willow St. Vancouver, British Columbia, V5Z 3P1, Canada. ‡Department Microbiology and Immunology, University of British Columbia, Faculty of Science, 2350 Health Sciences Mall, Vancouver, British Columbia, V6T 1Z3, Canada. §Canada's Michael Smith Genome Sciences Centre, 570 West 7th Ave - Suite 100, Vancouver, British Columbia, V5Z 4S6, Canada. ¶Bioinformatics Graduate Program, University of British Columbia, 100-570 West 7th Avenue, Vancouver, British Columbia, V5Z 4S6 Canada. ¥Laboratory of Parasitic Diseases, Division of Intramural Research, NIAID, National Institutes of Health, 4 Center Drive, Bethesda, Maryland, 20892, USA. #Department of Biochemistry and Molecular Biology, University of British Columbia, Faculty of Science, 2350 Health Sciences Mall, Vancouver, British Columbia, V6T 1Z3, Canada. Correspondence: Neil E Reiner. Email: ethan@interchange.ubc.ca© 2008 Silverman et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.The Leishmania donovani secretome <p>Analysis of Leishm ia-conditioned medium resulted in the identification of 151 proteins apparently secreted by the parasitic proto-zoan Leishmania donovani and suggested a vesicle-based secretion system.</p>AbstractBackground: Leishmania and other intracellular pathogens have evolved strategies that supportinvasion and persistence within host target cells. In some cases the underlying mechanisms involvethe export of virulence factors into the host cell cytosol. Previous work from our laboratoryidentified one such candidate leishmania effector, namely elongation factor-1α, to be present inconditioned medium of infectious leishmania as well as within macrophage cytosol after infection.To investigate secretion of potential effectors more broadly, we used quantitative massspectrometry to analyze the protein content of conditioned medium collected from cultures ofstationary-phase promastigotes of Leishmania donovani, an agent of visceral leishmaniasis.Results: Analysis of leishmania conditioned medium resulted in the identification of 151 proteinsapparently secreted by L. donovani. Ratios reflecting the relative amounts of each leishmania proteinsecreted, as compared to that remaining cell associated, revealed a hierarchy of protein secretion,with some proteins secreted to a greater extent than others. Comparison with an in silico approachdefining proteins potentially exported along the classic eukaryotic secretion pathway suggested thatfew leishmania proteins are targeted for export using a classic eukaryotic amino-terminal secretionsignal peptide. Unexpectedly, a large majority of known eukaryotic exosomal proteins wasdetected in leishmania conditioned medium, suggesting a vesicle-based secretion system.Conclusion: This analysis shows that protein secretion by L. donovani is a heterogeneous processthat is unlikely to be determined by a classical amino-terminal secretion signal. As an alternative, L.donovani appears to use multiple nonclassical secretion pathways, including the release of exosome-like microvesicles.Published: 18 February 2008Genome Biology 2008, 9:R35 (doi:10.1186/gb-2008-9-2-r35)Received: 10 November 2007Revised: 22 January 2008Accepted: 18 February 2008The electronic version of this article is the complete one and can be found online at http://genomebiology.com/2008/9/2/R35Genome Biology 2008, 9:R3535.2http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RBackgroundLeishmania spp. are the causative agents of a group of tropi-cal and subtropical infectious diseases termed the leishmani-ases. These infections disproportionately affect poorerpeoples in developing areas of the world. Because of the debil-itating and disfiguring results of infection, these diseases area great barrier to socioeconomic progress in endemic areas.As of 2001, it was estimated that 12 million people worldwidehave been infected with leishmania, and 2 million new casesare believed to occur each year [1]. Recent environmentalchanges such as urbanization, deforestation, and new irriga-tion schemes have expanded endemic regions and have led tosharp increases in the number of reported cases [2-4]. Inaddition, visceral leishmaniasis is establishing itself in previ-ously unaffected areas by piggy-backing on the spread of theHIV epidemic [5]. Leishmania co-infection with HIV hasbecome a serious global health threat. The two infections areinvolved in a deadly synergy, because leishmania infectionexacerbates the immunocompromised state of infected indi-viduals, thereby promoting HIV replication and resulting inearlier onset of AIDS [6]. The combination of HIV co-infec-tion, expansion of endemic regions, and evolving drug resist-ance [7] has created great need for more effective anti-leishmanial drugs and other control measures. Progress incontrolling the leishmaniases requires improved appreciationof the biology of the parasite to allow novel treatment strate-gies to be designed.Members of the genus Leishmania are digenetic protozoans.The organisms exist either as flagellated, motile promastig-otes within the alimentary canal of their phlebotomine sand-fly vector or as nonmotile amastigotes that reside withinphagolysosomes of mammalian mononuclear phagocytes.Promastigote surface coat constituents have been the focus ofconsiderable interest [8-10], and many of these - includingglycoproteins, proteoglycans, and glycolipids - have beenshown to play protective roles [8,11,12]. Surface-associatedmolecules are considered to make up the vast majority ofleishmania secreted material [9]. Through these studies, ithas become evident that there are a number of unusual fea-tures that typify exocytosis by this group of trypanosomatids.For example, in these highly polarized cells, regulated secre-tion is thought to occur solely at the flagellar pocket, a deepinvagination of the plasma membrane from which the singleflagellum of leishmania emerges [9,13]. Leishmania areknown to synthesize and traffic most surface molecules, suchas lipophosphoglycan and leishmanolysin GP63, along theclassical endoplasmic reticulum-Golgi apparatus-plasmamembrane pathway [9]. As mentioned, these surface mole-cules are ultimately delivered to the flagellar pocket, and it isthought that the pocket retains its role as the primary if notsole site of secretion in nonflagellated amastigotes [9]. Thusfar, no leishmania candidate virulence factors have beenfactors, and little attention has been paid to their intracellularor extracellular trafficking pathways.Whether leishmania use a classical amino-terminal signalsequence peptide to direct the export of most secreted pro-teins through the flagellar pocket or a different mechanism isunclear. Two leishmania surface glycoproteins, a proteophos-phoglycan and GP63, are initially synthesized with a cleava-ble amino-terminal signal sequence [9]. However, the vastmajority of characterized leishmania secreted proteins haveno identifiable secretion signal sequence, with the exceptionof those that are initially membrane bound [9,14,15]. The lackof a clear amino-terminal secretion signal sequence amongthe majority of characterized leishmania secreted proteinssuggests the existence of important nonclassical pathways ofsecretion.Despite the potential importance of protein secretion byleishmania, only a small number of leishmania proteins havebeen examined in detail from this perspective [14,16-18]. Ide-ally, one would like to know the identities of all of the compo-nents of any complex system in order to fully comprehendfunctionality. Consequently, we set out to identify all, or asmany as possible, of the proteins secreted by leishmania. Tothis end, we designed a quantitative proteomic approachbased on SILAC (stable isotopic labeling of amino acids inculture) [19-21]. SILAC involves culturing cells with eithernormal isotopic abundance amino acids or with stable iso-tope-enriched amino acids (for instance, L-arginine versus13C6-L-arginine) until essentially all proteins of the cell arelabeled. The two populations or samples to be compared arethen mixed and analyzed by nanoflow liquid chromatogra-phy-tandem mass spectrometry (LC-MS/MS). We used thisapproach to analyze the extent to which any given leishmaniaprotein was secreted into promastigote conditioned medium(Cm) by relating it to the level of the same protein thatremained cell associated (CA). In this report, we identified358 proteins in combined Cm/CA mixtures from Leishmaniadonovani and, based on a quantitative analysis, we concludethat 151 were actively secreted. The general properties of theidentified secreted proteins allowed us to postulate potentialmechanisms of secretion as well as functional roles within thecontext of infection.ResultsLeishmania conditioned medium contains a multiplicity of enriched proteinsThe main objective of this study was to characterize as com-prehensively as possible the proteins actively secreted by pro-mastigotes of L. donovani into culture medium. Beforeproceeding with the SILAC and LC-MS/MS analysis, wesought to develop a system in which we were confident thatGenome Biology 2008, 9:R35shown to traffic through the flagellar pocket. This is not sur-prising, however, given that no ultrastructural work hasaccompanied descriptions of leishmania candidate virulencethe proteins we were detecting in Cm were not artifacts andwere in fact present due to bona fide secretion. Previousinvestigations of protein secretion by leishmania were35.3http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rhampered by the presence of degradation products and by therequirement of the cells for serum [14,22]. In light of thesecomplexities, we included a nontoxic protease inhibitor, soybean trypsin inhibitor, in the promastigote culture mediumduring collection and isolation of Cm to minimize degrada-tion of secreted proteins by proteases. Secondly, we reducedCm collection time to 6 hours or less in order to allow cultureof promastigotes under serum-free conditions. Pulse-chaselabeling of leishmania with 35S-methionine followed by isola-tion of serum-free Cm showed clearly that leishmaniasecreted numerous proteins (Figure 1a). Here, an equalnumber of trichloroacetic acid-precipitated counts/minute ofCm and whole cell lysate (WCL) were analyzed, allowing us tocompare directly the intensities of protein bands from Cmand WCL. The results show that some of the leishmania-secreted proteins (arrows in Figure 1a) were clearly enrichedin the Cm. It is also important to note that the clearly distinctprotein separation patterns of leishmania Cm and WCL indi-cate that the proteins detected in Cm were unlikely to be arti-facts present due to lysis of cells during culture or processingTo control further for the possibility of false positive proteindetection in Cm caused by inadvertent lysis of promastigoteseither spontaneously (due to programmed cell death) or dur-ing isolation of Cm, using an enzymatic assay we measuredthe amount of cytosolic marker glucose 6-phosphate dehy-drogenase (G6PD) [23] present in Cm. The total amount ofG6PD activity detected in Cm was compared with activitiesfound to be associated with serial dilutions of the total massof promastigotes that was used to generate the Cm. As shownin Figure 1b, the amount of G6PD detected in Cm neverexceeded the total enzyme activity that was associated with5% of the promastigotes used to generate the Cm. Notably,there was also no difference in the amount of G6PD detectedin Cm collected from promastigotes that had been grown ineither stable isotope or normal isotopic abundance culturemedium (data not shown) during the SILAC analysisdescribed below.Quantitative mass spectrometry identifies a wide array of leishmania-secreted proteinsLeishmania Cm contains enriched proteins and is minimally contaminated by incidental cell lysisFigure 1Leishmania Cm contains enriched proteins and is minimally contaminated by incidental cell lysis. (a) Leishmania promastigotes were metabolically labeled, as described in Materials and methods. Conditioned medium (Cm) from labeled cells and the cells themselves were collected in parallel, and the proteins present in the Cm and corresponding whole cell lysate (WCL) of promastigotes were precipitated in 10% trichloroacetic acid (TCA). Equal numbers of TCA-precipitated counts/minute of Cm and WCL were fractionated on a 5% to 20% gradient polyacrylamide gel. Arrows indicate proteins specifically enriched in leishmania Cm. The autoradiograph shown is representative of three independent experiments. (HMW) High molecular weight marker. (b) To control for inadvertent lysis of organisms during collection of Cm, glucose-6-phosphate dehydrogenase (G6PD) activity in Cm collected from isotopically labeled and nonlabeled cells was measured as described in Materials and methods and compared with the activity associated with deliberately lysed promastigotes. The data shown are the means of measurements from three independent experiments. 0.01 units of G6PD were assayed as a control in each experiment. The asterisks shown indicate a significance difference when compared with Cm (P < 0.001), calculated by one-way analysi of variance followed by Bonferroni's correction for multiple comparisons (GraphPad Prism 4.0).Cm  WCL(kDa)HMW203045669722014(a) (b)50% 25% 10% 5% 1% Cm G6PD0. **Percent deliberate lysisUnits of G6PDGenome Biology 2008, 9:R35(Figure 1a). Serum-free leishmania Cm collected from stationary phasepromastigotes was fractionated either by one-dimensionSDS-PAGE or by in-solution isoelectric focusing and analyzed35.4http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rby LC-MS/MS using a linear trapping quadrupole-Fouriertransform hybrid mass spectrometer (see Materials andmethods, below). We set three criteria that had to be met forany protein detected by mass spectrometry to be included inthe leishmania 'secretome'. First, we only considered proteinsto be identified if at least two unique tryptic peptidesequences from that protein were detected (see Materials andmethods for peptide criteria limits). Second, we required aparticular protein to be observed in at least three out of fourindependent experiments. This resulted in the identificationof 358 proteins (listed in Additional data file 1) in the pooledCm and CA samples, with an estimated false discovery rate ofless than one protein in 200. Interestingly, by these criteriawe did not detect G6PD in any of the LC-MS/MS analyses,probably because the amount of G6PD was below the detec-tion limit of the mass spectrometer.of standard methods for measuring total protein concentra-tion (see Materials and methods, below), so we estimated theprotein content of Cm samples from an initial LC-MS/MSanalysis and mixed these with an equal amount of oppositelylabeled CA protein. Because this method of equalization isimprecise, we normalized all Cm/CA ratios within an experi-ment to histone H2B (GeneDB:LmjF19.0050). H2B was con-sistently detected in Cm, most likely as a result of both generalcell lysis and apoptosis [24,25]. After normalization, the val-ues were loge transformed (Additional data file 2) and Cm/CAratios for all identified proteins were calculated as the meanCm/CA ratio for all peptides from that protein across allexperiments (Additional data file 2) [26,27]. These SILACratios reflected the degree of enrichment of individual proteinspecies in leishmania Cm, and a frequency distribution isshown in Figure 2. Across all experiments the overall mean ±Quantitation of leishmania secreted proteins in CmFigure 2Quantitation of leishmania secreted proteins in Cm. Conditioned medium (Cm)/cell associated (CA) ratios from each of four independent analyses were normalized to the ratio for that of histone H2B followed by log normal (Ln) transformation (see Additional data file 2). Mean values across the four experiments for each protein identity were calculated as described in Materials and Methods. GraphPad Prism 4.0 was then used to calculate the mean and standard deviation (SD) of all Cm/CA values for all proteins found in leishmania Cm (mean = 1.39, SD = 0.85). The same program generated a frequency distribution of the mean Cm/CA values based on the relative frequency of each value within the dataset. Proteins with mean Cm/CA values greater than 2 SDs (1.7; solid black horizontal line) above the value for histone H2B were considered to be members of the secretome. Mean Ln transformed ratios of protein abundance (Cm/CA) where identities lying above the black line are actively secreted, and values falling below the dotted line are likely present due to apoptosis or lysis. X-axis numbers correspond to each protein identity, with 358 in total. The mean and SD from at least three measurements for each protein are shown.Hist . H2Bcutoff50 100 150 200 250 300 350-3-2-1123456Mean Ln Transformed ratio valuesGenome Biology 2008, 9:R35The method of preparation of Cm for LC-MS/MS analysis didnot provide sufficient amounts of protein to allow reliable usestandard deviation Cm/CA value for the 358 proteins was 1.35± 0.85 (Figure 2).35.5http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RWe used the Cm/CA ratio of histone H2B to define the thirdcriterion for inclusion in the secretome. We considered leish-mania proteins with a mean Cm/CA peptide ratio at least twostandard deviations (1.7) above the ratio for histone H2B(after transformation = 0) to be actively secreted by leishma-significant margin of error around it, the proteins (number-ing 151 in total) with Cm/CA ratios of 1.7 or greater were likelyto be bona fide secreted proteins. This conservative approachprovided a high level of specificity for 'secretion' at theexpense of sensitivity.We used Western blotting to examine a select group of pro-teins in paired Cm and CA samples to determine the extent towhich this orthogonal method of detection would correlatewith the SILAC/mass spectrometry analysis. Here we exam-ined four proteins: heat shock protein (HSP)70, with a Cm/CA value of 1.86, above the cut-off of histone H2B plus twostandard deviations (or +1.70); HSP83/HSP90, with a Cm/CA ratio of 1.50 falling just below the cut-off; elongation fac-tor-1α (EF-1α) with a ratio of 0.69; and secreted acid phos-phatase (SacP), which was not detected by LC-MS/MS. Asshown in Figure 3, SAcP was detected as a dispersed band inCm, but it was completely absent from the aliquots of WCLanalyzed (lanes 1 and 2). Both HSP70 and HSP90 were alsoclearly enriched in Cm, with HSP70 to a greater extent thanHSP90 (compare Cm with WCL lane 2). On the other hand,the bulk of EF-1α was retained intracellularly (Figure 3). Thisqualitative analysis indicated that the SILAC/LC-MS/MSresults correlated closely with conventional protein detectionby Western blotting in respect of providing a semiquantita-tive estimate of protein secretion by leishmania. Additionally,these findings indicated that the arbitrary third criterion forinclusion in the secretome was both valid and in fact highlyrigorous, because HSP90 - a protein falling just below thesecretome cut-off (Cm/CA of 1.7; Figure 2) - was clearly foundto be enriched in Cm by Western blotting (again compare Cmwith WCL lane 2).The results for SAcP both by mass spectrometry and Westernblotting were of particular interest and appeared to be a spe-cial case. Whereas this ecto-enzyme, which was previouslyreported to have an amino-terminal secretion signal [28], washighly enriched in Cm (Figure 3), its absence from the LC-MS/MS analysis suggested that its absolute abundance mustbe quite low. This is addressed further under Discussion(below).The results of the Western blotting also indicated that therewas minimal contamination of Cm by incidental lysis. Figure3 shows the protein profile of 5% of the cells (selected basedon the maximum amount of lysis that may have occurredaccording to the results of the G6PD analysis; Figure 1b) to bemarkedly distinct from that of the leishmania Cm (compareCm with WCL lane 1). The distinct profiles of Cm and WCLobserved in the metabolic labeling experiment (Figure 1a)also indicated that contamination of Cm through lysis wasnegligible.Leishmania HSPs are enriched in CmFigure 3Leishmania HSPs are enriched in Cm. Leishmania conditioned medium (Cm) was collected from a culture containing about 2 × 109 promastigotes. The proteins contained therein were precipitated and then solubilized directly in Laemmli sample buffer. One half of this volume (containing a known amount of protein) was loaded into the lane labeled 'Cm'. From the 2 × 109 promastigotes recovered from the culture, 5% (1 × 108) were removed and processed in parallel with the remaining 95%. The 5% figure was chosen based upon the estimated (see Figure 1) maximum number of organisms that may have undergone incidental lysis during the incubation period and the collection centrifugation. Proteins contained in the whole cell lysate (WCL) prepared from the 1 × 108 cells were precipitated and resolubilized in a volume equal to that used to resolubilize the proteins precipitated from Cm collection. To allow for direct comparison to be made, half of this volume was then loaded into WCL lane #1. Protein was also precipitated from the WCL prepared from the remaining 95% of the cells and after solubilization in sample buffer an amount of protein was loaded into WCL lane #2 equal to that loaded into the lane labeled Cm. After transfer to nitrocellulose membrane, blots were probed, stripped, and reprobed with the indicated antibodies. The data shown are representative of results obtained in at least three identical experiments. EF1alpha, elongation factor-1α; Hsp, heat shock protein; SAcP, secreted acid phosphatase.Cm   1               2WCLSAcPHsp90Hsp70EF1alphaGenome Biology 2008, 9:R35nia (Figure 2, solid line). In choosing this rather conservativeyet arbitrary cut-off, we reasoned that if H2B was representa-tive of proteins externalized by apoptosis then, by allowing aGene Ontology analysis of the leishmania secretomeTo develop an understanding of how protein secretion byleishmania might be related to specialized functions or proc-35.6http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Resses, we used the Leishmania Genome [29] and the GeneOntology (GO) [30] databases in conjunction with theBlast2GO analysis tool [31] to determine whether any classesof proteins were more likely to be found in among the leish-mania secreted proteins. This analysis resulted in 85% of theproteins detected in leishmania Cm having one or more GOterm assignments (Additional data file 3).After tallying the number of leishmania secreted proteinsassigned to each GO term, it was clear that many of thesecreted proteins (Figure 4a) were involved in turnover andsynthesis of protein and nonprotein macromolecules. In fact,27 out of the 151 secreted proteins (18%) were predicted to beinvolved in protein translation (GO: 0006412), which wasmore than in any other discrete biologic process (Figure 4a).Beyond this, as shown in Figure 4a, the leishmania secretedproteins identified by LC-MS/MS were found to be involvedin a wide array of processes, including proteolysis(GO:0006508), protein folding (GO:0006457), and biologicregulation (GO:0065007).Consistent with the biological process GO analysis, a full 50%of leishmania secreted proteins were involved in proteinbinding interactions, for example binding to ATP(GO:0005524), ions (GO:0043167), or other proteins(GO:0005515; Figure 4b). Other highly represented functionsincluded pyrophosphatase activity, hydrolase activity, andoxidoreductase activity (GO:0016462, GO:0016787, andGO:0016491). It is noteworthy that nearly 20 proteins thatfell below the secretion cut-off were annotated as havingtransporter activity (GO:0005215), whereas no such activitywas found for the secreted proteins (Additional data file 3).Of interest, there appeared to be a trend toward concentra-tion of a distinct set of processes and functions in the group of151 leishmania proteins making up the leishmania secretome.As shown in Figure 5a, when compared with the total group of358 proteins consistently identified in Cm, there appeared tobe enrichment of proteins involved in processes related togrowth (GO:0040007), RNA metabolism (GO:0016070), andbiopolymer modification (GO:0043412), including proteinamino acid phosphorylation (GO:0006468). Consistent withthese biologic process assignments, molecular functions suchas kinase activity, peptidase activity, and translation factoractivity (GO:0016301, GO:0008233, and GO:0003746,respectively) appeared to be more prevalent among the groupof 151 leishmania secreted proteins than among the totalgroup of 358 proteins consistently identified in Cm (Figure5b). We used the GOSSIP [32] statistical framework to deter-mine whether any GO terms were significantly enriched in thesecreted proteins when compared to other Cm proteins. Manyof the processes and functions discussed and depicted in Fig-ure 5 had significant (P < 0.05) single test P values. However,there was no a priori basis for an association between thesecreted proteins and any GO term) [32], no terms werefound to be significantly enriched in the group of 151 secretedproteins. This may be due to our small sample size of individ-ual GO terms associated with at most 358 proteins. In con-trast, these statistical tests are regularly carried out on samplesizes in the tens of thousands [32] of genes or proteins. Inaddition, statistical significance may not have been achievedbecause we were comparing two datasets with a high proba-bility of overlap, because we looked for enrichment of GOterms associated with the group of 151 proteins in leishmaniasecretome compared with GO terms associated with the totalgroup of 358 Cm proteins. In fact, some of the Cm proteinsbelow the cut-off may be actively secreted and certainly werefound to be exported by some mechanism, including celldeath. For these reasons, we consider that the apparent con-centration of GO associations shown in Figure 5 may in factbe meaningful.In addition to members of the secretome having pleiotropicfunctions, they were also predicted to have a variety of subcel-lular localizations. Nearly one-third of leishmania secretedproteins were predicted to be cytoplasmic (GO:0005737) byGO, and these had associations with both membrane bound(GO:0043227) and nonbound intracellular organelles(GO:0043228), including ribosomal proteins, nuclear pro-teins, mitochondrial proteins, and glycosomal proteins(Additional data file 3). Only five secreted proteins were pre-dicted to be integral membrane proteins, and none of thesecretome proteins were predicted to be associated with theendoplasmic reticulum.Bioinformatics analysis of secreted proteins in the leishmania genomeWe screened the leishmania genome database for proteinscontaining a classical amino-terminal secretion signal pep-tide, in order to generate a putative list of classically secretedproteins for comparison with the proteins identified by LC-MS/MS. We modified a bioinformatics approach previouslyused to identify proteins secreted by Mycobacterium tuber-culosis [33] and applied it to the genome of Leishmaniamajor [34]. Proteins were considered highly likely to besecreted if the sequence included a classical amino-terminalsecretion signal peptide and lacked additional transmem-brane (TM) domains. Additional TM domains would havesuggested that the protein was membrane bound and there-fore unlikely to be released from the cell. The majority ofleishmania surface expressed proteins are associated with theplasma membrane via a glycophosphotidylinositol (GPI) lipidattachment [9], and some of these GPI-attached surfaceproteins, such as GP63, are known to disassociate from themembrane and can be detected in Cm [35]. In light of this, asa final step we screened the proteins positive for a signalGenome Biology 2008, 9:R35after correcting for multiple testing using both a false discov-ery rate (the most common correction method) and a familywise error rate (which is more correct in this context becausesequence and negative for multiple TM domains for GPI-link-age attachment sites and considered positive proteins to besecreted (Additional data file 4). Using these parameters, we35.7http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RHigh prevalence GO assignments in the leishmania secretomeFi ure 4High prevalence GO assignments in the leishmania secretome. The secretome sequences were categorized according to (a) biological process and (b) (a)(b)protein metabolic proces sbiosynthetic processbiopolymer metabolic processmacromolecule biosynthetic processtranslationnucleic aci d metabolic processRNA metabolic processproteolysi scellular component organization and biogenesisamino aci d and derivative meta bolic processbiologi cal regulationdevelopmental processcatabolic processprotein folding010203040Percent of the Secreted Proteinsbindingcatalytic activit ynucleo tide bindingpurine nucleo tide bindingadenyl nucleotide bind ingATP bindinghydrolase activi tynucleic acid bindingprotein bindingtransferase activ itypeptidase activi tymetal ion bi ndingoxidor eductase activ itynucleoside-triphosphatase activityion bindingpyrophosphatase activi tyhydrolase activi tyligase activi tyhydrolase activi ty0510152025303540455055Percent of Secreted ProteinsGenome Biology 2008, 9:R35molecular function. Nonredundant processes and functions assigned to at least ten leishmania-secreted protein sequences are displayed. Bars indicate the number of protein sequences found under each Gene Ontology (GO) term expressed as a percentage of the total 151 actively secreted proteins.35.8http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RGO assignments concentrated in the leishmania secretomeFigure 5GO assignments concentrated in the leishmania secretome. (a) Biological process and (b) molecular function Gene Ontology (GO) terms with an equivalent or greater number of assignments in the secreted proteins (black bars) compared with all of the conditioned medium (Cm) proteins (white bars) are shown. Bars indicate the number of actively secreted protein sequences found under each GO term expressed as a percentage of the 151 actively secreted proteins (black bars) compared with the total number of Cm sequences found under that GO term expressed as a percentage of the total 358 proteins identified in the Cm (white bars).(a)(b)biop olymer metabolic processRNA metabo lic pr ocessbiop olymer modificationprotein modification post-trans lati onal protein modificationprotein amino acid phosphorylationtransl ationtransl ational init iationtransc ription terminationproteolysi sdevelopmentregulation  of growthgrowth0510152025translation factor acti vi tytransl ation regulator activi tytransl ation init iation f actor activit yser ine-type pept idase act ivityprotein-tyrosine kinase activityprotein serine/threonine kinase activityprotein kinase activitytRNA ligase activit yisomera se act ivit yoxidoreductase activitytransf erase activit yexopept idase activit ypeptidase act ivit ymetallopepti dase act ivit yendopeptidase act ivit yDNA bindingzinc ion bindingvitami n bind ingtransi tion metal ion bindingpyridoxal phosphate binding012345678Genome Biology 2008, 9:R3535.9http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rfound that the leishmania genome encodes 217 proteins thatcontain a classical secretion signal peptide, of which 141 areannotated as hypothetical proteins (Additional data file 4). Ofthe remaining 76 proteins, approximately one-third appear tobe gene duplications, leaving 50 unique leishmania proteinswith a known or putative classical eukaryotic secretion signalpeptide. It is of interest that only one of the proteins we pre-dicted to be secreted via an amino-terminal secretion signal -LmjF16.0790, a chitinase - has previously been demonstratedto be secreted by leishmania promastigotes [16,36], althoughwe did not detect this protein in our LC-MS/MS analysis.Our analysis also suggests that SAcP does not contain a clas-sical secretion signal, contrary to a previous report [37].Based upon the Von Heijne algorithm [34], the latter studypredicted the presence of a 23-amino-acid amino-terminal'signal peptide'. Subsequently, this leader peptide was shownto be sufficient for secretion of a green fluorescent proteinfusion construct expressed in L. donovani [27]. The SignalPalgorithm we used is the updated version of the 1985 Von Hei-jne algorithm. The lack of concordance in these predictionshighlights the limitations of bioinformatics, while reinforcingthe well known fact that signal sequences are highly variable.Our bioinformatics analysis also confirmed the annotation inthe Leishmania Genome database [29] that none of thehistidine secretory acid phosphatases found in the genomesof L. major or L. donovani infantum have classical amino-terminal secretion signals. Interestingly, only the membranebound acid phosphatases of L. major are annotated ascontaining classical secretion signal peptides, whereas thesame is not true of the orthologs in L. donovani infantum,and these membrane bound proteins would have beenexcluded by our TM domain screen. Only 14 of the proteinspredicted to be secreted through a classical signal sequence-dependent mechanism were detected in leishmania Cm byMS, and only two of these, GeneDB:LmjF04.0310 andLmjF36.3880, had sufficiently high SILAC ratios to beincluded in the secretome (Additional data file 4). Althoughthere are several possible explanations for failing to detect aprotein by LC-MS/MS, the lack of correlation between themeasured and the in silico predicted secretomes suggests thatleishmania utilize nonclassical secretion signals and path-ways to regulate the export of the majority of secretedproteins.Evidence that proteins released by leishmania may originate in exosome-like vesicles, apoptotic vesicles, and glycosomesSomewhat unexpected was the finding that leishmania Cmcontained all of the proteins identified previously to be asso-ciated with exosomes isolated from both B lymphocytes anddendritic cells, with the exception of those for which the leish-mania genome does not contain an ortholog (Additional datafile 5). In fact, more than 10% of the proteins found in theleishmania secretome were previously detected in exosome-like microvesicles released from other eukaryotic cells (Table1), including B lymphocytes [38], dendritic cells [24], and adi-pocytes [39]. Recently, mammalian adipocytes were shown tosecrete microvesicles, which were referred to as adiposomes[39]. These adiposomes contained 98 proteins, 13 of which weconcluded to be actively secreted (Table 1). At least 25 addi-tional adiposome proteins were detected in leishmania Cmwith relative abundances lower than the secretome cut-off(Additional data file 5). The concordance of the proteomicdata between these higher eukaryotic secreted microvesiclesand the leishmania secretome is remarkable. These findingssuggested that leishmania secrete exosome/adiposome-likemicrovesicles carrying proteomic cargo that is similar in com-position to host microvesicles. In support of this, using scan-ning electron microscopy, we observed 50 nm microvesiclesspecifically located at the mouth of the leishmania promastig-ote flagellar pocket (Figure 6a,b), as well as evenly distributedacross the cell surface of cells with the apparent morphologyof amastigotes undergoing differentiation axenically (Figure6c).Microvesicles budding from the flagellar pocket and plasma membrane of leishmaniaFigure 6Microvesicles budding from the flagellar pocket and plasma membrane of leishmania. Stationary phase leishmania promastigotes were fixed and coated for scanning electron microscopy as described in Materials and methods. (a) A leishmania promastigote, (b) 10× magnification of the exposed flagellar pocket region of panel a (square) after stage rotation, and (c) a promastigote in the process of differentiating into an amastigote. Arrowheads point to (a) (c)(b)5um 500nm 5umGenome Biology 2008, 9:R35microvesicles.35.10http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RSurprisingly, DNA-binding histone proteins were reliablydetected by LC-MS/MS in Cm of stationary phase promastig-otes (Additional data files 1 and 5). Histone proteins havebeen detected in dendritic cell exosomal preparations andwere shown to enrich in these preparations after the cellswere treated with an apoptosis-inducing agent [24]. The den-dritic cell vesicles containing histone proteins were moreelectron dense and migrated to a slightly higher sucrose den-sity than the exosomes [24]. This led the authors to concludethat the histone-containing vesicles were indeed a distinctpopulation of vesicles, termed apoptotic vesicles or blebs[24]. The detection of histones in Cm of stationary phaseleishmania (Additional data files 1 and 5), along with the sig-nificant number of apoptotic leishmania known to be presentin a stationary phase population (approximately 43 ± 5%)[25], suggests that promastigotes may have been releasingapoptotic vesicles as well as exosomes.many of the major glycolytic enzymes that normally reside inglycosomes of kinetoplastid organisms [40] (Table 1 andAdditional data file 5). Relevant to these findings, leishmaniahave been shown to utilize peroxisomal targeting signals(PTSs; PTS1 and PTS2) to direct proteins to the glycosome[41], and a screen of the leishmania genome identifiedapproximately 100 proteins with either a PTS1 or a PTS2 tar-geting signal [42]. Remarkably, our MS analysis of leishma-nia Cm detected nearly half of these predicted glycosomalproteins, with ten being detected at high enough relativeabundance to be considered bona fide secreted proteins(Table 1). These findings suggest that leishmania releaseeither whole glycosomes or glycososomal cargo into the extra-cellular environment.DiscussionTable 1Leishmania-secreted proteins associated with exosome-like and glycosomal vesiclesGeneDB accession number Protein identificationa Mean Cm/CA ratio Microvesicle associationLmjF35.3340 6-Phosphogluconate dehydrogenase, decarboxylating, putative 3.01 GLYLmjF29.0510 Cofilin-like protein 2.80 DCLmjF36.6910 Chaperonin, putative, T-complex protein 1 (theta subunit), putative 2.61 APLmjF01.0770 Eukaryotic initiation factor 4a, putative 2.60 DCLmjF28.2860 Cytosolic malate dehydrogenase, putative 2.20 APLmjF24.2060 Transketolase, putative 2.19 GLYLmjF33.2550 Isocitrate dehydrogenase, putative 2.16 APLmjF28.2770 Heat-shock protein hsp70, putative 2.14 BC, DC, APLmjF35.3860 T-complex protein 1, eta subunit, putative 2.14 APLmjF12.0250 Cysteinyl-tRNA synthetase, putative 2.07 GLYLmjF14.1160 Enolase 2.05 BC, DC, APLmjF36.2030 Chaperonin Hsp60, mitochondrial precursor 2.03 APLmjF23.1220 T-complex protein 1, gamma subunit, putative 2.00 APLmjF05.0350 Trypanothione reductase 1.99 GLYLmjF36.2020 Chaperonin Hsp60, mitochondrial precursor 1.98 APLmjF36.1630 Clathrin heavy chain, putative 1.98 BC, APLmjF16.0540 Aspartate carbamoyltransferase, putative 1.95 GLYLmjF27.2000 Hypothetical protein, conserved 1.90 GLYLmjF31.1070 Biotin/lipoate protein ligase-like protein 1.89 APLmjF26.1240 Heat shock protein 70-related protein 1.86 BC, DC, APLmjF04.0960 Adenylate kinase, putative 1.77 GLYLmjF27.1260 T-complex protein 1, beta subunit, putative 1.76 APLmjF30.3240 Glutamyl-tRNA synthetase, putative 1.74 GLYLmjF21.0810 Methionyl-tRNA synthetase, putative 1.74 GLYLmjF36.3210 14-3-3 Protein-like protein 1.74 DC, APLmjF33.2540 Carboxypeptidase, putative, metallo-peptidase 1.73 GLYa The Mascot algorithm {2008 609/id} was used to identify the protein names and the GeneDB accession numbers [29]. Microvesicles: BC, B-cell lymphocyte exosome [38,62]; DC, dendritic cell exosome [24]; AP, adipocyte exosome (adiposome) [39]; and GLY, leishmania glycosome [42]. CA, cell associated; Cm, conditioned medium; Hsp, heat shock protein.Genome Biology 2008, 9:R35In addition to exosomal and apoptotic vesicle-associated pro-teins, we also found that the leishmania secretome includedTo our knowledge, this report is the first proteomic analysis ofprotein secretion by Leishmania, or, for that matter, anyother kinetoplast. This quantitative proteomic analysis35.11http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rshowed that L. donovani released a wide array of proteinswhen in the stationary phase of growth (Additional data file1). Based on previous studies concerned with the pathogene-sis of leishmania as well as other intracellular pathogens[17,43], we anticipated that leishmania may secrete virulenceeffectors into their extracellular environment, including thecytosolic compartment of infected host cells. By examiningthe composition of the leishmania secretome and generatingquantitative information concerning the relative enrichmentof secreted proteins, we expected to identify candidate leish-mania effector proteins that may be involved in virulence. Asexpected, protein export was found to be heterogeneous, withsome proteins exported to a higher degree than they wereretained by the cell, whereas for others the opposite was true(Figure 2). It was our assumption that proteins with higherCm/CA ratios were more likely to be actively secreted thanthey were to be externalized as a result of either incidentallysis or apoptosis. In light of this, we used the relative abun-dance data and a rigorous statistical cut-off (Cm/CA valuesgreater than the ratio for H2B by at least two standard devia-tions) to define proteins actively secreted by leishmania.Based on this analysis, we consider 151 proteins in this data-set to be bona fide members of leishmania secretome. On theother hand, we recognize that in implementing this rigorouscut-off we probably sacrificed some sensitivity. Thus, it isprobable that at least some proteins with ratios falling belowthe cut-off are actively secreted as well.Next, we inspected the leishmania secretome for potentialvirulence factors. Candidate virulence factors were dividedinto four categories: proteins putatively involved in intracel-lular survival; proteins with known immunosuppressivefunctions; proteins involved in signal transduction; and pro-teins involved with transport processes (Table 2). By far thelargest class of candidate virulence factors was comprised ofproteins that may be required for intracellular survival. Theleishmania secretome showed a remarkable abundance ofproteasome subunits (such as GeneDB:LmjF35.4850,LmjF36.1600, LmjF21.1700, LmjF21.1830, LmjF27.0190,LmjF36.1650, and LmjF34.0650) and proteases such as theoligopeptidases (GeneDB:LmjF09.0770; Tables 2 and 3, andAdditional data file 1), of which many had high Cm/CA val-ues. In addition, proteolysis was one of the most common GOterms assigned to the leishmania secreted proteins. Althoughthe frequency of this term did not reach statistical signifi-cance (see Gene Ontology analysis of the leishmania secre-tome, under Results, above), this term appeared to besomewhat over-represented among the proteins in the upperhalf of the ratio distribution (Figures 4a and 5a). It seemslikely that the secretion of at least some of these proteins maybe part of a stress response. On the other hand, some of theseproteins may be involved in pathogenesis. One potentialmechanism is the direction of their proteolytic activitiesmajor histocompatibility complex class I and II molecules,thereby preventing antigen loading and reduced efficiency ofantigen presentation, as has been described for leishmania-infected cells [44]. These findings suggest that secreted leish-mania proteins with proteolytic activities may contribute topathogenesis, and further investigation of this is warranted.Also likely to be involved in intracellular survival are secretedantioxidants, and more generally proteins with oxidoreduct-ase activity, such as iron superoxide dismutase(GeneDB:LmjF32.1820). Other examples of these were foundin the leishmania secretome (Figure 5b and Additional datafile 1), and these may provide protection from intracellularfree radical attack. In addition, some members of the secre-tome, such as the putative 14-3-3 protein, are known to havepowerful antiapoptotic properties in other systems [45]. Thatleishmania infection inhibits host cell apoptosis is well known[46,47], and these antiapoptotic secreted proteins may beactive in prolonging the lifespan of infected host cells.An important inclusion to the category of proteins with func-tional roles in intracellular survival were nucleases, such asGeneDB:LmjF23.0200, an endoribonuclease, which wasfound to be the second most highly secreted protein (Table 3).This endoribonuclease belongs to a class of proteins that acton single-stranded mRNA and are thought to be inhibitors ofprotein synthesis [48]. These nucleases may aid in purinesalvage, which is obligatory for leishmania because they areincapable of de novo purine synthesis [49].Myo-inositol-1-phosphate synthase (GeneDB:LmjF14.1360),the protein with the highest relative abundance ratio (Table3) and therefore the most enriched in the Cm, may also play arole in intracellular survival (Table 2). Leishmania myo-inosi-tol-1-phosphate synthase has been shown to be essential forgrowth and survival in myo-inositol limited environments[50]. Leishmania myo-inositol-1-phosphate synthase knock-outs were found to be completely avirulent [50] in mice, sug-gesting that the phagolysosomal lumen may be a myo-inositollimited environment. Myo-inositol-1-phosphate synthase isrequired for de novo biosynthesis of myo-inositol, a precursorof vital inositol phospholipids such as those found in the GPImembrane anchors of nearly all leishmania surface proteinsand other glycoconjugates such as GP63 and lipophosphogly-can. The massive export of this essential enzyme into Cm isintriguing and warrants further study.The leishmania secreted protein kinetoplastid membraneprotein-11 (GeneDB:LmjF35.2210), identified in the SILAC/mass spectrometry analysis, was previously characterized ashaving immunomodulatory effects on host cells duringleishmania infection [51]. Furthermore, we found that theleishmania secretome contains an ortholog of the mamma-Genome Biology 2008, 9:R35toward degradative enzymes resident in phagolysosomes topromote intracellular survival. A second possibility mightinvolve direction of their proteolytic activities to degradelian macrophage migration inhibitory factor(GeneDB:LmjF33.1750), a protein with known immunosup-pressive and immunomodulatory properties [52] in humans.35.12http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RTable 2Leishmania candidate virulence factors enriched in CmGeneDB accession numbers Protein identificationa Mean Cm/CAratiobPutative functioncSignal transduction proteinsLmjF35.2420 Phosphoinositide-binding protein, putative 3.51 KinasedLmjF33.1380 Mitogen-activated protein kinase 11, putative, MAP kinase, putative 2.42 KinaseLmjF09.0770 Oligopeptidase b, serine peptidase, clan SC, family S9A-like protein 2.31 Cell-cell signalingLmjF34.2820 Regulatory subunit of protein kinase a-like protein 2.30 KinaseLmjF28.2740 Activated protein kinase c receptor (LACK) 2.28 Kinase receptorLmjF25.0750 Protein phosphatase, putative 2.27 PhosphataseLmjF35.1010 Casein kinase, putative 2.08 KinaseLmjF10.0490 Mitogen-activated protein kinase 3, putative, MAP kinase 3, putative 1.73 Kinase, signal transductionLmjF31.2790 ADP-ribosylation factor, putative 1.67 Small GTPase mediated signal transductionImmunosupressive proteinsLmjF35.2210 Kinetoplastid membrane protein-11 2.33 Immunosuppressive [51]LmjF33.1750 Macrophage migration inhibitory factor-like protein 2.21 Immunosuppressive [52]LmjF25.0910 Cyclophilin a 1.73 Immunosuppressive [103]Proteins Involved in Intracellular survivalLmjF14.1360 Myo-inositol-1-phosphate synthase 3.93 Inositol biosynthesis [50]LmjF23.0200 Endoribonuclease L-PSP (pb5), putative 3.91 mRNA salvage, inhibition protein synthesisLmjF11.0630 Aminopeptidase, putative, metallo-peptidase, clan MF, family M17 3.05 ProteolysisLmjF28.1730 Proteasome regulatory non-ATP-ase subunit 2, putative 2.40 ProteolysisLmjF34.1040 Uracil phosphoribosyltransferase, putative 2.34 Pyrimidine salvageLmjF09.0770 Oligopeptidase b, serine peptidase, clan SC, family S9A-like protein 2.31 Invasion, proteolysisLmjF32.1820 Iron superoxide dismutase, putative 2.29 AntioxidantLmjF13.0090 Carboxypeptidase, putative, metallo-peptidase, clan MA(E), family 32 2.14 ProteolysisLmjF28.2770 Heat-shock protein hsp70, putative 2.14 Protein stabilityLmjF26.1570 Thimet oligopeptidase, putative, metallo-peptidase, clan MA(E), family M32.10 ProteolysisLmjF05.0960 Dipeptidyl-peptidase III, putative, metallo-peptidase, clan M-, family M49 2.08 ProteolysisLmjF14.1160 Enolase 2.05 Plasminogen binding [104], InvasionLmjF19.0160 Aminopeptidase, putative, metallo-peptidase, clan MG, family M24 2.04 ProteolysisLmjF21.1830 Proteasome alpha 5 subunit, putative 2.03 ProteolysisLmjF05.0350 Trypanothione reductase 1.99 AntioxidantLmjF21.0760 Proteasome regulatory non-ATP-ase subunit 5, putative,19S proteasome regulatory subunit1.97 ProteolysisLmjF23.0270 Pteridine reductase 1 1.92 AntioxidantLmjF26.0810 Glutathione peroxidase-like protein, putative 1.89 AntioxidantLmjF27.0190 Proteasome alpha 7 subunit, putative 1.89 ProteolysisLmjF26.2280 Nitrilase, putative 1.86 Carbon-nitrogen hydrolaseLmjF26.1240 Heat shock protein 70-related protein 1.86 Protein stabilityLmjF31.1890 Peptidase m20/m25/m40 family-like protein 1.85 ProteolysisLmjF06.0140 Proteasome beta 6 subunit, putative,20S proteasome beta 6 subunit, putative1.85 ProteolysisGenome Biology 2008, 9:R35LmjF29.0120 Proteasome regulatory non-ATPase subunit, putative 1.81 ProteolysisLmjF34.0650 Proteasome regulatory non-ATP-ase subunit 11, putative,19S proteasome regulatory subunit, metallo-peptidase, Clan MP, Family M671.79 Proteolysis35.13http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RIt is possible that this leishmania ortholog could share thesefunctions and affect host immune responses during leishma-nia infection.Manipulation of host cell function via interference with sign-aling pathways is a well known virulence tactic of intracellularpathogens [53-57]. After internalization, leishmania infectedmacrophages exhibit defective signaling in response to vari-ous stimuli [54,55,57]. Based on our analysis, we estimatethat at least ten secreted leishmania proteins are predicted tobe involved in some manner in signal transduction (Table 2and Additional data file 3). In this regard, we found thatkinase activity was concentrated in the upper half of thesecretome ratio distribution (Figure 5b). Secreted leishmaniasignaling intermediates such as the mitogen-activated pro-tein kinases 3 and 11 (GeneDB:LmjF33.1380 andLmjF10.0490) and the protein tyrosine phosphatase-likeprotein (GeneDB:LmjF16.0230) have the potential to affectmacrophage cell signaling after internalization [56]. Anotherinteresting signaling related protein, the putative phosphoi-nositide-binding protein (GeneDB:LmjF35.2420), was one ofmost highly secreted proteins (Table 3). This protein mightinfluence macrophage cell signaling through its potentialbinding of inositol containing signaling intermediates thatare products of phosphatidyloinositol 3 kinase. Notably, GOinvolved in coordinating intracellular vesicle trafficking proc-esses, including both endocytosis and exocytosis [58]. Assuch, this putative sorting nexin may be considered to be aleishmania candidate virulence factor for its potential tomodulate vesicle trafficking in infected cells (Table 2).Somewhat unexpected was the finding of proteins in leishma-nia Cm known to be involved in vesicular transport (Tables 2and 1), such as the phosphoinositide binding proteindiscussed above, the small GTP-binding protein Rab1(GeneDB:LmjF27.0760) and a putative ADP-ribosylation fac-tor (GeneDB:LmjF31.2790). We have classified these pro-teins as candidate virulence factors because, although thesetransport vesicle regulatory proteins may normally regulatevesicle trafficking in leishmania, ectopically following secre-tion, they may have the potential to affect vesicle trafficking ininfected cells. For example, it is tempting to speculate thatthese leishmania secreted proteins could directly affectphagosome maturation through modulating transport to andfusion with host multivesicular bodies, endosomes, andlysosomes.Another interesting and unexpected aspect of the leishmaniasecretome was the presence of numerous proteins related totranslational machinery (Figure 4a and Additional data fileLmjF35.0750 Proteasome activator protein pa26, putative 1.78 ProteolysisLmjF35.2350 Aminopeptidase P, putative, metallo-peptidase, clan MG, family M24 1.77 ProteolysisLmjF02.0370 Proteasome regulatory non-ATPase subunit 6, putative 1.77 ProteolysisLmjF36.3210 14-3-3 protein-like protein 1.74 Anti-apoptoticLmjF21.1700 Proteasome alpha 2 subunit, putative 1.66 ProteolysisLmjF36.1600 Proteasome alpha 1 subunit, putative 1.64 ProteolysisLmjF35.4850 Proteasome alpha 1 subunit, putative 1.60 ProteolysisProteins involved in vesicular transport processesLmjF35.2420 Phosphoinositide-binding protein, putative (sorting nexin 4) 3.51 Transport of proteins and other substancesLmjF27.0760 Small GTP-binding protein Rab1, putative 2.27 Endosomes/Golgi traffickingLmjF32.1730 Coatomer epsilon subunit, putative 2.09 Intracellular protein transportLmjF36.1630 Clathrin heavy chain, putative 1.98 Endocytosis, trans-Golgi to lysosome traffickingLmjF18.0700 Hypothetical protein, conserved 1.85 HEAT repeat, intracellular protein transportLmjF31.2790 ADP-ribosylation factor, putative 1.67 Intracellular protein transportaThe Mascot algorithm {2008 609/id} was used to identify the protein names and the GeneDB accession numbers [29]. bMean of normalized, log normal (Ln) transformed conditioned medium (Cm)/cell associated (CA) peptide ratios for at least three of four experiments. cPutative functions and locations are derived from the GeneDB database {2007 238/id} unless otherwise noted. dProtein sequence contains an amino-terminal secretion signal peptide according to SignalP.Table 2 (Continued)Leishmania candidate virulence factors enriched in CmGenome Biology 2008, 9:R35analysis identified this putative phosphoinositide-bindingprotein (GeneDB:LmjF35.2420) as a sorting nexin 4-like pro-tein (Additional data file 3). Sorting nexins are known to be3). The functional basis for this is unclear at this time. Per-haps the turnover of these proteins is extremely high andexcess machinery is disposed of via secretion in addition to35.14http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rthe reported processes of ubiquitination and proteasomemediated degradation. Interestingly, clathrin-coated vesiclesisolated from rat liver [59] were found to contain more than30 of the same translation related proteins we found in leish-mania Cm, including a putative leishmania eukaryotic trans-lation initiation factor 1A (GeneDB:LmjF16.0140), theprotein with the fifth highest enrichment ratio (Table 3).Appreciation of the multifunctional nature of proteins isincreasing, and the possibility exists that these proteins, per-haps purposely packaged in leishmania secretory vesicles,may play ancillary roles in pathogenesis or pathogen survival,similar to what appears to be the case for EF-1α [17].The protein secretion pathways utilized by leishmania are notwell understood. According to our analysis, only two of the151 proteins in the leishmania secretome contain a classicalamino-terminal secretion signal (Additional data files 2 and4). The fact that more than 98% of the secretome lacks a tar-geting signal indicates that nonclassical secretion pathwaysare probably the dominant means by which leishmania pro-teins are secreted. In support of this argument, the leishma-nia secretome included a large number of proteins previouslyidentified as components of exosomes secreted from varioushigher eukaryotic cell types (Table 1). Leishmania Cm alsocontained many proteins shown to be cargo of clathrin-coatedvesicles. Rat liver clathrin-coated vesicles were found to con-tain a total of 346 proteins, and in addition to the 30 transla-tion-related proteins mentioned above, an further 30 of theseproteins were detected in leishmania Cm, including clathrinof these clathrin-coated vesicles and that of mammalian exo-somes were strikingly similar [24,59-62]. Furthermore, leish-mania have been shown to form clathrin-coated vesicles [63],and clathrin-directed trafficking in leishmania was shown tobe essential for survival in macrophages [64]. Taken together,these findings suggest that leishmania may use clathrin-coated vesicles as a transport mechanism to direct vesicletrafficking at least, if not exocytosis of proteins from endo-somal compartments to the extracellular milieu. Based onthese findings, we propose that leishmania protein secretionprobably involves the release of exosome-like vesicles, whichmay or may not be clathrin-coated. Moreover, we suggest thatat least three distinct vesicular secretion processes contributeto the secretome, including exosomes, apoptotic vesicles, andglycosomes (Table 1).Exosomes are small vesicles, 50 to 100 nm in diameter, whichare released by fusion of either multivesicular endosomes orsecretory lysosomes with the plasma membrane of eukaryoticcells [65-67]. Exosomes were initially described in reticulo-cytes as a mechanism for shedding organellar proteins andexcess transferrin receptor during differentiation into maturenuclei free red blood cells [68]. Somewhat later, the pro-teomes of B lymphocyte and dendritic cell exosomes weredescribed [24,38]. Dendritic cell exosomes have attracted asignificant amount of attention because of their immunostim-ulatory properties as cell-free, peptide-based vaccines [69-72]. The striking correspondence between the leishmaniasecretome and these exosomes strongly suggests that proteinTable 3Highly enriched leishmania secreted proteinsGeneDB accession number Protein identificationa Mean Cm/CAratiobFunctionc Predicted locationcLmjF14.1360 Myo-inositol-1-phosphate synthase 3.93 Inositol biosynthesis [50] CytosolLmjF23.0200 Endoribonuclease L-PSP (pb5), putative 3.91 Nuclease, mRNA cleavage CytosolLmjF15.1203 60S acidic ribosomal protein P2 3.73 Translation [102] RibosomedLmjF35.2420 Phosphoinositide-binding protein, putative 3.51 Phosphoinositol binding, signal transductionCytosoldLmjF16.0140 Eukaryotic translation initiation factor 1A, putative3.26 Translation Cytosol, exosomes [24,38,39]LmjF32.2180 Hypothetical protein, conserved 3.08 Translation initiation NucleusLmjF11.0630 Aminopeptidase, putative, metallo-peptidase, Clan MF, Family M173.05 Proteolysis CytosolLmjF35.3340 6-Phosphogluconate dehydrogenase, decarboxylating, putative3.01 Glucose cataboloism CytosolLmjF04.0310 Beta-fructofuranosidase, putative 2.93 Carbohydrate metabolism CytosoldLmjF36.3840 Glycyl tRNA synthetase, putative 2.87 Translation CytosolaThe Mascot algorithm {2008 609/id} was used to identify the protein names and the GeneDB accession numberes [29]. bMean of normalized, log normal (Ln) transformed conditioned medium (Cm)/cell associated (CA) peptide ratios for at least three of four experiments. cPutative functions and locations are derived from the GeneDB database {2007 238/id}unless otherwise noted. *Protein sequence contains an amino-terminal secretion signal peptide according to SignalP.Genome Biology 2008, 9:R35(GeneDB:LmjF36.1630) and HSP70. Significantly, bothclathrin and HSP70 have been found in exosomes releasedfrom various human cells [24,38,39]. In fact, the proteomessecretion by leishmania involves the release of intraluminalvesicles originating from either the tubular lysosome [73] ormultivesicular endosomes, or both. It is tempting to speculate35.15http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rthat leishmania exosomes, like dendritic cell exosomes[69,70], may be capable of modulating the host immuneresponse, although it likely that their properties may be quitedistinct.The formation of membrane blebs at the plasma membraneof apoptotic mammalian cells and their subsequent releaseare phenomena that have attracted significant attention[24,66,74]. As mentioned above, these apoptotic vesicles havebeen found to contain histone proteins and cytochrome c oxi-dase subunits. That leishmania undergo apoptosis is wellestablished [25], and our finding that they release cyto-chrome c oxidase subunits and histones into Cm (Additionaldata files 1 and 5) suggests that they release apoptotic vesi-cles. Moreover, it has been shown that cultures of stationaryphase leishmania promastigotes contain up to 43% apoptoticcells, and when the latter are removed by sorting the remain-ing nonapoptotic population is incapable of establishing andmaintaining an infection [25]. These findings, taken togetherwith our detection of apoptotic vesicle marker proteins, his-tones 1 through 4, in leishmania Cm (Additional data files 1and 5), strongly suggest the possibility that leishmania apop-totic vesicles may be involved in pathogenesis. This could takethe form of immune evasion, wherein (similar to activation ofthe 'silent phagocytosis' pathway used to internalize and clearvery early apoptotic cells by mammalian macrophages [75])these apoptotic vesicles would promote inhibition of macro-phage activation before invasion by viable leishmaniapromastigotes.Somewhat more difficult to explain from our findings is thesuggestion for whole glycosome release, based upon both thecharacterized and the putative glycosomal proteins wedetected in leishmania Cm (Table 1). Notably, many of theleishmania Cm proteins that were bioinformatically predictedto be glycosomal by the presence of PTS1 or PTS2 have beenidentified in purified glycosomes of the closely related kineto-plast Trypanosoma brucei brucei [76]. Our identification ofthe two most prevalent leishmania glycosomal membraneproteins in promastigote Cm (Additional data file 5) suggeststhat intact glycosomes were being exported from the cell. Thisis as opposed to a model in which these organelles were fusingwith the flagellar pocket to release their cargo, in which casewe would not have expected to have detected glycosomalmembrane proteins per se. As we suggested above to poten-tially explain the secretion of translation machinery proteins,release of glycosomal proteins may be related to a stressresponse, but the targeted release of glycosomes with a morespecialized function remains a possibility.As previously stated, it is our hypothesis that the proteinswith higher relative abundance in leishmania Cm are morelikely to play an active role in pathogenesis than those pro-be less likely to contribute to pathogenesis. Although this is areasonable working model, it is not absolute and does notmean that proteins secreted in lesser abundance may not beof interest. In fact, EF-1α, a candidate virulence factor thathas been shown to inhibit macrophage activation [17], had aCm/CA peptide ratio in the lowest 20% of the ratio distribu-tion (Figure 2, and Additional data files 2 and 5), and wellbelow the cut-off for active secretion used to define the secre-tome. These data, especially when combined with the findingsthat apoptotic leishmania are required for leishmania diseasedevelopment [25], support the interpretation that many ofthe proteins found in leishmania Cm are potential candidatesfor unique and essential roles in leishmania virulence, andfurther analysis will be required to prioritize those thatshould receive additional attention.It should be mentioned that three leishmania proteins previ-ously described to be secreted, namely SacP [14], chitinase[36], and silent information regulator (SIR)2 [18], were notidentified in this LC-MS/MS analysis of leishmania Cm. Onepossible explanation for why these identifications were notmade is that they have extremely low intracellular concentra-tions, with nearly all of the synthesized protein beingsecreted. Under these conditions other proteins present in thecell at a higher concentration could mask the CA peptide sig-nals in the mass spectrometry. Importantly, the SILAC/massspectrometry analysis was designed to compute ratios ofsimultaneously detected spectra from mixed Cm and CA sam-ples. The absence of a CA signal in the mass spectrometrywould have provided a denominator of zero, thereby notallowing for the computation of a meaningful Cm/CA ratioand exclusion from the analysis. Thus, no matter how abun-dant these peptides might be in Cm, without a comparable CAsignal these proteins would not be included in the leishmaniasecretome, as defined by this study. It is possible that thisexplanation may also account for why chitinase and SIR2were not identified in the secretome, especially consideringthat neither have characterized intracellular functions.Finally, we conducted these experiments using L. donovanidonovani. Sequencing of the L. donovani genome is currentlyunderway. As such we used the completed L. major genometo assign protein identities to the mass spectra gathered fromleishmania Cm. Although the genomes for these two speciesare thought to be very similar, as their similar life cycles, biol-ogy, and expression profiles would indicate [77], it is possiblethat genomic difference between species prevented identifi-cation of some Cm proteins.Examination of the secretome led to several additional find-ings worth noting. First, the number of proteins known to beassociated with small vesicles outstripped by far the numberof proteins identified that had classical secretion signals. Thisfinding suggests that the main secretory route for leishmaniaGenome Biology 2008, 9:R35teins secreted to a lesser extent. Following this logic, export ofproteins with lower Cm abundance may be related to eitherroutine waste disposal or apoptotic blebbing, and these mayinvolves the release of small vesicles. Second, for the majorityof candidate virulence factors that were identified, it seemsmost likely that they may function to influence the survival of35.16http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rleishmania within the phagolysosome, although this remainsto be formally tested. As the collection time for Cm was lim-ited because of the need to culture organisms in the absenceof serum, proteins in the secretome that may be involved inpathogenesis are likely to act during early stages of infection.During this early stage, they may contribute to the observeddelay of phagosome maturation [78]. It has been proposedthat delayed phagosome maturation represents a window ofopportunity during which internalized promastigotes can dif-ferentiate into the more acid-tolerant amastigotes [79,80].Whether the amastigote secretome is similar to or distinctfrom that of stationary phase promastigotes is not known atthis time. However, given the relatively low stage-specific dif-ferences in gene expression that have been described [81], wedo not regard significant differences to be likely. Third, tar-geting of virulence factors into host cell cytosol has beenshown to be an effective strategy used by intracellular patho-gens to remodel the environment and to influence host cellfunction [17,82-85]. After invading their macrophage hosts,leishmania have been shown to block cell activation, to inhibitmicrobicidal activity [86-88], and to attenuate antigen-pre-senting cell function [57,89,90]. A broad picture of the pro-teins secreted by leishmania in cell free culture provides abasis for investigation of effector proteins that may be activein host cells either within the phagolysosome or within hostcytosol.ConclusionThis quantitative proteomic analysis identified a large anddiverse pool of proteins in leishmania Cm and allowed us todefine the leishmania secretome based on measurements ofrelative protein abundance in Cm that could only beexplained by active secretion. The identities of proteinswithin the secretome revealed many candidates for furtherstudies concerned with potential contributions to virulenceand pathogenesis as well as to investigate mechanisms ofsecretion. Moreover, the data also indicate clearly that leish-mania use predominantly nonclassical targeting mechanismsto direct protein export. This leads us to propose a model inwhich protein export occurs largely through the release ofmicrovesicles, perhaps including exosome-like vesicles,apoptotic vesicles, and glycosomes.Materials and methodsCell cultureL. donovani Sudan strain 2S promastigotes were cultured inmedium M199 supplemented with 10% fetal bovine serum(FBS; Gibco Cell Culture, Div. of Invitrogen Life Technolo-gies, Gaithersburg, MD, USA), 1% penicillin and streptomy-cin, 20 mmol/l HEPES (Stem Cell Technologies, Vancouver,British Columbia, Canada), 6 μg/ml hemin, 2 mmol/l L-were split 1:10 into fresh medium in 25 or 75 cm3 cell cultureflasks. For SILAC analysis, promastigotes were transferred tocustom made M199 without L-arginine and L-lysine (CaissonLaboratories, North Logan, UT, USA) supplemented with10% partially dialyzed FBS (Gibco), 1% penicillin and strepto-mycin, 20 mmol/l HEPES (Stem Cell Technologies), 6 μg/mlhemin, 2 mmol/l L-glutamine, 10 μg/ml folic acid, 100 μmol/l adenosine, and one of two SILAC media formulations (nor-mal isotopic abundance arginine [42 mg/l] and lysine [73mg/l]; and 13C6-arginine [43.5 mg/l] and 2H4-lysine [75 mg/l]) at 26°C. Organisms were cultured in this medium for atleast 14 days and split 1:10 every third day, in order to achieve100% labeling of cellular proteins before analysis. Stable iso-tope-labeled amino acids were purchased from CambridgeIsotope Laboratories (Andover, MA, USA). Except where oth-erwise noted, reagents were obtained from the Sigma-AldrichChemical Company (St. Louis, MO, USA).Isolation of promastigote CmStationary phase promastigotes that had been grown either inmedium containing normal isotopic abundance arginine andlysine or in medium containing 13C6-arginine 2H4-lysine Lwere collected by centrifugation at 300 × g for 10 minutes ina Beckman GS-6R centrifuge (Beckman-Coulter, Fullerton,CA, USA) and washed in Hanks balanced salt solution.Organisms were then concentrated tenfold by re-suspensionin medium M199 without FBS and supplemented with 2mmol/l L-glutamine, 10 mmol/l HEPES, 10 μg/ml soya beantrypsin inhibitor (Sigma-Aldrich), and either normal isotopicarginine and lysine or 13C6-Arg and 2H4-Lys in the concentra-tions given above for 4 to 6 hours at 26°C. Cm was isolatedfrom cells by centrifugation at 300 × g for 10 minutes in aBeckman GS-6R. Supernatant was then subjected to centrifu-gation once more to ensure that no cells remained in suspen-sion. Cm and cell pellets were either used immediately forenzymatic analysis or stored at -20°C for mass spectrometryanalysis. A minimum of 5 × 108 promastigotes in culture wasrequired to generate Cm with signals of adequate strength formass spectrometry analysis.Four times as many stationary phase organisms wererequired to generate sufficient Cm for detection of proteins byeither metabolic labeling and autoradiography or by Westernblotting. Two billion organisms were cultured in M199 con-taining normal isotopic arginine and lysine (Sigma-Aldrich).For autoradiography, cells were collected and washed asabove, and then starved of methionine by resuspension inRPMI-1640 medium without methionine and cysteine(Sigma-Aldrich) with 1% FBS. After 1 hour 50 μCi/ml of 35Smethionine (Sigma-Aldrich) was added and cells were cul-tured for a further 2 hours to allow labeling to occur. Afterwashing to remove serum, cells were incubated for 4 hours inserum-free RPMI-1640 medium without methionine andGenome Biology 2008, 9:R35glutamine, 10 μg/ml folic acid, and 100 μmol/l adenosine at26°C in a EchoTherm Chilling Incubator (Torrey Pines Scien-tific, San Marcos, CA, USA). Every third day the organismscysteine, containing 10 mmol/l L-glutamine, 1 mmol/lHEPES, and 10 μg/μl Soya bean trypsin inhibitor, at whichpoint the cells were separated from the Cm by low speed cen-35.17http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rtrifugation to avoid mechanical lyses of cells. Pelleted cellswere lysed on ice in lysis buffer (50 mmol/l Tris [pH 7.4], 1%Triton X-100, 0.15 mol/l NaCl, 1 mmol/l EGTA, 1 mmol/lphenylmethylsulfonyl fluoride, 10 μg aprotinin/ml, and 10 μgleupeptin/ml). Cell lysates were clarified by centrifugation ina microcentrifuge at maximum speed for 20 min at 4°C. Theresulting WCL supernatants and the Cm were precipitatedwith trichloroacetic acid at 10% final concentration. The pre-cipitates were solubilized in Laemmli sample buffer and equalcounts/minute of Cm and WCL were separated by SDS-PAGE(5% to 20% gradient) followed by autoradiography.For Western blotting, Cm was collected as above, but organ-isms were concentrated in normal isotopic M199. After sepa-rating Cm from the cells, WCLs were generated by sonicatingthe cell pellets to mimic lysis that may have occurred inad-vertently during culture or centrifugation. Briefly, cell pelletswere solubilized in 0.5 mmol/l Tris Laemmli sample bufferwithout SDS, bromophenol blue, or β-mercaptoethanol, butincluding protease inhibitors leupeptin and aprotinin both at1 μg/ml and 10 μg/ml phenylmethylsulphonyl fluoride. Thesolution was sonicated three times at a power setting of 3 for10 seconds. The lysate was cleared of insoluble material bycentrifugation for 5 minutes at 10,000 × g. Following clarifi-cation the supernatant proteins were precipitated followingthe procedure bellow. The pellet was resuspended in Laemmlisample buffer without β-mercaptoethanol or bromophenolblue.Protein precipitationFor Western blotting and metabolic labeling analysis, pro-teins present within promastigote Cm were precipitated usingpyrogallol red, as described previously [91]. Briefly, sodiumdeoxycholate was added to Cm to a final concentration of0.02% and the solution was mixed for 30 minutes at 4°C tofacilitate precipitation. Cm was then mixed with an equal vol-ume of pyrogallol red solution (containing 0.05 mmol/l pyro-gallol red, 0.16 mmol/l sodium molybdate, 1.0 mmol/lsodium oxalate, 50 mmol/l succinic acid, and 20% methanol[vol/vol]) and the pH adjusted to 2.0 with 2N HCl. The result-ing solution was incubated at room temperature for 1 to 2hours followed by 12 to 24 hours at 4°C. The Cm protein pre-cipitates were harvested by centrifugation at 11,000 × g for 60minutes at 4°C followed by two washes with ice cold acetone.The pellets were allowed to air dry before solubilization inLaemmli sample buffer without β-mercaptoethanol orbromophenol blue at 95°C for 30 minutes. Protein concentra-tions of the Cm and WCLs were measured using the BioRadDC Protein Assay (BioRad Laboratories Inc., Hercules, CA,USA).G6PD assayPromastigote cell pellets were lysed by sonication to generate10 μg/ml soya bean trypsin inhibitor, protease inhibitors leu-peptin and aprotinin both at 1 μg/ml, and 10 μg/ml phenyl-methylsulphonyl fluoride. After clearance by centrifugationat 11,000 × g, serial twofold dilutions of the lysate were madein medium M199 supplemented as above to yield final con-centrations of 50%, 25%, 10%, 5%, and 1% (vol/vol). The con-centrations of G6PD in 100 μl of Cm and in serial dilutions ofWCL were assayed in 55 mmol/l Tris-HCl and 3.3 mmol/lMgCl2 buffer at pH 7.8, containing 3.3 mmol/l glucose-6-phosphate and 2 mmol/l NADP. Enzyme was obtained fromthe Sigma Chemical Company for a positive control. To gen-erate a reference, 0.01 units of G6PD were stabilized in 5.0mmol/l glycine with 0.01% bovine serum albumin (pH 8.0)and assayed along with sample and WCL dilutions. Enzymereactions were carried out at 30°C and the change in absorb-ance, caused by changing NADP concentration, over 5 min-utes was measured at 340 nm.LC-MS/MS of promastigote conditioned medium and data analysisTo identify proteins specifically secreted by leishmania intoculture medium, direct quantitative comparisons of proteinabundance in Cm versus CA were made on a protein-by-pro-tein basis. The Cm was collected from leishmania grown inmedium containing heavy isotopes of arginine and lysine, andcompared with cell-associated material prepared from pro-mastigotes grown in medium containing normal isotopicabundance amino acids. In some cases the reciprocal analysiswas also carried out as well with identical results.Approximately equal amounts of labeled and unlabeled pro-tein (estimated from a preliminary LC-MS/MS analysis) fromCm and CA were mixed together and analyzed either by gel-enhanced LC-MS/MS exactly as described previously [92] orby peptide-level isoelectric focusing (IEF) combined with LC-MS/MS. For IEF, the protein mixture was solubilized indigestion buffer (50 mM NH4OH, 1% sodium deoxycholate,pH 8.0), denatured by heating to 99°C for 5 minutes, reducedby incubation with 1 μg dithiothreitol for 30 minutes at 37°C,alkylated with 5 μg iodoacetamide for 30 minutes at 37°C andfinally digested by the addition of 1 μg porcine trypsin(Promega, Madison, WI, USA) overnight at 37°C. After diges-tion, the sample was acidified by addition of an equal volumeof sample buffer (3% acetonitrile, 1% trifluoroacetic acid, and0.5% acetic acid) and the deoxycholate that fell out of solutionwas pelleted at 16,100 × g for 5 minutes. Peptide mixtureswere then desalted on STop-And-Go Extraction (STAGE) tips[93] before being resolved into 24 fractions from pH 3 to 10on an OFFGEL IEF system (Agilent Technologies, SantaClara, CA, USA), in accordance with the manufacturer'sinstructions. Fractions from the IEF were diluted with anequal volume of sample buffer, and each was desalted againon a STAGE tip. Each gel or OFFGEL fraction was analyzedGenome Biology 2008, 9:R35a WCL in 1 ml medium M199 with the appropriate concentra-tions of either normal isotope or nonradioactive isotopearginine and lysine, 1 mmol/l L-glutamine, 1 mmol/l HEPES,on a linear trapping quadrupole-Fourier transform tandemmass spectrometer, as described previously [19]. Fragmentspectra were extracted with ExtractMSN.exe (v3.2) using the35.18http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rdefault parameters (ThermoFisher Scientific, Ottawa, ON,CA); monoisotopic peak assignments were corrected withDTASuperCharge (default parameters [94]); and the result-ing peak list was searched against the protein database for L.major plus the sequences of all human keratins and porcinetrypsin (5 November 2006 version, 8,324 sequences) usingMascot (v2.1 [95]).MSQuant [94] was used to parse Mascot result files, to recal-ibrate mass measurements, and to extract quantitative ratios.The final nonredundant list of proteins was generated usingfinaList.pl, an in-house script available on our website [96].The false discovery rate for protein identifications based ontwo or more peptides with a measured mass accuracy under 3ppm (the overall average was 0.61 ppm), a Mascot score of 25or greater, and length 8 residues or more was estimated to beless than 0.5%, using reversed database searching. All identi-fied peptides with their associated parameters can be found inAdditional data file 1. SILAC ratios were extracted exactly asdescribed previously [19]. The mean loge transformed ratiosfrom four independent analyses and the relative standarddeviations can be found in Additional data file 2.Western blottingFollowing isolation of Cm, lysis of the corresponding cell pel-let, and precipitation of proteins in both fractions, equivalentamounts of protein from the Cm and WCL were fractionatedby SDS-PAGE. Proteins were transferred to nitrocelluloseand probed with anti-EF-1α (Upstate Biotechnologies Inc.,Lake Placid, NY, USA) following the manufacturer'sinstructions, as well as leishmania-specific antibodies to his-tidine secreted acid phosphatase [97] and against HSP70 andHSP90 [98] (a kind gift from Dr Joachim Clos).Scanning electron microscopyStationary phase promastigotes were washed in phosphate-buffered saline and fixed in 2.5% gluteraldehyde in 0.1 mol/lsodium cacodylate buffer (pH 7.2) containing 0.146 mol/lsucrose and 5 mmol/l CaCl2 at 22°C under vacuum in a micro-wave: 2 minutes at 100 W, 2 minutes without microwaves, 2minutes with 100 W, and then repeated. Subsequently, fixedorganisms were rinsed in the same buffer in the microwavefor 40 seconds at 100 W two times and post-fixed in 1% OsO4in 0.1 mol/l sodium cacodylate containing 2 mmol/l CaCl2and 0.8% potassium ferricyanide (Polysciences, Warrington,PA, USA) at 22°C under vacuum in a microwave following thesame steps used in the gluteraldehyde fixation. Cells werewashed in distilled water at room temperature and allowed toadhere to poly-L-lysine (Sigma) coated coverslips. Subse-quently the coverslips were dehydrated through an ascendingethanol series from 50% to 100%, each for 40 seconds at 100W in a microwave. The fixed cells were critically point driedwith liquid CO2 in a Balzars 020 Critical Point Dryer (BalzarsHitachi S-2600 VPSEM (Hitachi High Technologies, Fin-champstead, Wokingham, Berkshire, UK) at the University ofBritish Columbia Bioimaging Facility.Bioinformatics screen of the genome of Leishmania major to identify candidate secreted proteinsThe genome of L. major was accessed at the GeneDB L. majordatabase [29]. Predictions of signal peptides and signal pepti-dase cleavage sites were made by SignalP [99]. Once thesewere provisionally identified, a filter was applied to removethose that contained more than one TM region predicted byTMpd [100]. Proteins with just one TM region were againscreened to filter out those whose single TM domain did notoverlap with the signal peptide coordinates. Finally, theseputative classically secreted, non-TM proteins were screenedfor GPI attachment sites at the carboxyl-terminus using theGPI prediction program GPI-SOM [101].Gene OntologyGO [30] annotations were performed using Blast2GO [31]. Anonredundant database was used as reference for Blastpsearches with an expectation value minimum of 1 × e-3 and ahigh scoring segment pair cut-off of 33. Annotations weremade with default parameters. Briefly, the pre-eValue-Hit-Filter was 1 × e-6, the Annotation cut-off was 55, and the GOWeight was 5. The statistical framework GOSSIP [32] wasused to identify statistically enriched GO terms associatedwith leishmania secreted proteins when compared to the GOterms associated with all of the proteins identified in leshma-nia Cm. GOSSIP generates 2 × 2 contingency tables for eachGO term in the test group and uses a Fisher's exact test to cal-culate P values for each term. The P values are then adjustedfor multiple testing by calculation of the false discovery rateand the family wise error rate [32].Statistical analysisStatistical analyses of Cm/CA ratios and G6PD concentra-tions were performed using GraphPad Prism version 4.00 forWindows (GraphPad Software, San Diego, CA, USA).AbbreviationsCA, cell associated; Cm, conditioned medium; EF-1α, elonga-tion factor-1α; FBS, fetal bovine serum; G6PD, glucose 6-phosphate dehydrogenase; GO, Gene Ontology; GPI, glyco-phosphotidylinositol; HSP, heat shock protein; IEF, isoelec-tric focusing; LC-MS/MS, liquid chromatography-tandemmass spectrometry; PTS, peroxisomal targeting signal; SacP,secreted acid phosphatase; SILAC, stable isotopic labeling ofamino acids in culture; SIR, silent information regulator; TM,transmembrane; WCL, whole cell lysate.Genome Biology 2008, 9:R35Union Ltd, Lichtenstein) and coated with gold palladiumusing a Nanotech SEMPrep II sputter coater (Nanotech Ltd.,Prestwick, U.K.). Samples were observed and imaged using aAuthors' contributionsJMS was involved in study design, data collection and analy-sis, interpretation of results and manuscript preparation.35.19http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. RSKC helped in design and implementation of the bioinformat-ics screen. DPR was involved in LC-MS/MS data collectionand analysis. DD provided leishmania-specific antibodies tosecreted acid phosphatase. DN helped with design ofbiochemical analyses. LJF was involved in study design, datacollection and analysis, and manuscript preparation. NERwas involved in study design, interpretation of results, andmanuscript preparation.Additional data filesThe following additional data files are available with theonline version of this paper. Additional data file 1 is a tablelisting all the proteins, and the peptides contributing to theiridentification, detected in leishmania Cm. Additional data file2 is a table showing a complete list of the SILAC ratios calcu-lated for each Cm protein in each experiment, including themeans of the four experiments. Additional data file 3 is a tablelisting all the GO terms associated with the leishmania Cmproteins. Additional data file 4 is a table listing the proteinspredicted by bioinformatics to be secreted under the controlof an amino-terminal secretion signal peptide; also shownhere are the proteins with predicted GPI attachment sites andthose proteins determined to be present in leishmania Cm bythe SILAC LC-MS/MS analysis. Additional data file 5 is atable listing the leishmania Cm proteins, their mean SILACratios, and any documented microvesicle associations forthese proteins.Additional data file 1The pr teome of leishmania Cm358 ins h d at least two nonoverlapping peptides that were detected quan ified i  three or more individual analyses of leishma ia Cm protein . The p ptid s corresponding to e ch id n-tifica i  re shown. Prote n id tities we  det rmined as cribed in Mater als and meth ds and for Tables 1 to 3.Cl k her  for file 2m/CA peptide rat os of l ish an a Cm oteinsd r in ng w ich pr t ins w e to be c sider for analy-s s ( s desc ibed  Material  and m thods and fo  Addi ional Data Fil 1), th sured Cm/CA rati s w n rmalize   the m s-ured valu f to H2B ch nd pe dent exp im n . Th  normalized val s re he log orm l (L ) tran fo d ( n Ln transfor d Cm/CA ati , xper ents [Exp ] 1 to 4) t reduceh  s r ad e data. Th f t e L  r form ratios foreac prot i  denti y w e  alculat d (m  L r n formedv lue ). The la ive t ard dev ati ns f th  p tid r ti s for alysis ar c u ed.3GO n ys s of l hm n  Cm pr teinst t f t  pro ns ted in l ishma  Cm. *Pro-n w a in - rmi al s cr ti n sig al p p ides, an †p otei sow  o b ti ic GO IDs li  h GO i e t c i  u b ras o  w t ch p , a d GO T rm l s s th  ter  s c -t  i  eac GO ID C, c llul  c p r ; F, m lecula func; P, b l gic p o s4B inf rm tics nal is f cla sic ly c et lei ia rishm ni  pr t i i te  to b c sically secr t byg o w e r fo pro i s i i g an m n -te m nal r s gn ep d . MS, rot i  t t  in h  SILAC/ sp tr m try n y ; § rot i s c by m s sp c met yw r tio ov ec u -off; PI, rot s f und t  oa GPI t hm nt it ; *  r in v ously r po b  by le s .5M ov s oci ti s of l sh  C  pr nP t h m C /CA p  rati th w tandd de t on b v  ha n H2B w c n rh . *P t i s w t min - e i l ec l p pt s,† h wn  be tig . M c v s o ip ays th ve c s oci d w t he pr ID. AP dip y  i; BC, B-c l ly p cyt x o ; DC, d c e lx ; G , gly .AcknowledgementsThe authors would like to thank Dr Joachim Clos for the kind gift of leish-mania-specific antibodies, and the University of British Columbia Bioimagingfacility for their excellent technical assistance with electron microscopy.This work was supported by Canadian Institutes of Health grants MOP-8633 and MOP-83063 (NER), and MOP-77688 (LJF). LJF is the CanadaResearch Chair in Organelle Proteomics, a Michael Smith FoundationScholar, and a Peter Wall Institute Early Career Scholar. Additional supportwas provided by an award from the CIHR/MSFHR Strategic Training Pro-gram in Bioinformatics (SKC) and by an infrastructure grant from theMichael Smith Foundation (RUA042021).References1. Chappuis F, Sundar S, Hailu A, Ghalib H, Rijal S, Peeling RW, Alvar J,Boelaert M: Visceral leishmaniasis: what are the needs fordiagnosis, treatment and control?  Nat Rev Microbiol 2007,5:873-882.2. Anis E, Leventhal A, Elkana Y, Wilamowski A, Pener H: Cutaneousleishmaniasis in Israel in the era of changing environment.Public Health Rev 2001, 29:37-47.3. Costa CH, Werneck GL, Rodrigues L Jr, Santos MV, Araujo IB, MouraLS, Moreira S, Gomes RB, Lima SS: Household structure andurban services: neglected targets in the control of visceralleishmaniasis.  Ann Trop Med Parasitol 2005, 99:229-236.4. Croft SL, Vivas L, Brooker S: Recent advances in research andcontrol of malaria, leishmaniasis, trypanosomiasis andschistosomiasis.  East Mediterr Health J 2003, 9:518-533.5. Sinha PK, Pandey K, Bhattacharya SK: Diagnosis & managementof leishmania/HIV co-infection.  Indian J Med Res 2005,121:407-414.past, present and future.  Curr Med Chem 2007, 14:1153-1169.8. Spath GF, Garraway LA, Turco SJ, Beverley SM: The role(s) of lipo-phosphoglycan (LPG) in the establishment of Leishmaniamajor infections in mammalian hosts.  Proc Natl Acad Sci USA2003, 100:9536-9541.9. McConville MJ, Mullin KA, Ilgoutz SC, Teasdale RD: Secretorypathway of trypanosomatid parasites.  Microbiol Mol Biol Rev2002, 66:122-154.10. Vergnes B, Sereno D, Madjidian-Sereno N, Lemesre JL, Ouaissi A:Cytoplasmic SIR2 homologue overexpression promotes sur-vival of Leishmania parasites by preventing programmed celldeath.  Gene 2002, 296:139-150.11. Descoteaux A, Avila HA, Zhang K, Turco SJ, Beverley SM: Leishma-nia LPG3 encodes a GRP94 homolog required for phos-phoglycan synthesis implicated in parasite virulence but notviability.  EMBO J 2002, 21:4458-4469.12. Ma D, Russell DG, Beverley SM, Turco SJ: Golgi GDP-mannoseUptake Requires Leishmania LPG2. A member of a eukary-otic family of putative nucleotide-sugar transporters.  Journalof Biological Chemistry 1997, 272:3799-3805.13. Overath P, Stierhof YD, Wiese M: Endocytosis and secretion intrypanosomatid parasites - tumultuous traffic in a pocket.Trends in Cell Biology 1997, 7:27-33.14. Bates PA, Dwyer DM: Biosynthesis and secretion of acid phos-phatase by Leishmania donovani promastigotes.  Mol BiochemParasitol 1987, 26:289-296.15. Rogers ME, Ilg T, Nikolaev AV, Ferguson MA, Bates PA: Transmis-sion of cutaneous leishmaniasis by sand flies is enhanced byregurgitation of fPPG.  Nature 2004, 430:463-467.16. Shakarian AM, Dwyer DM: The Ld Cht1 gene encodes the secre-tory chitinase of the human pathogen Leishmania donovani.Gene 1998, 208:315-322.17. Nandan D, Yi T, Lopez M, Lai C, Reiner NE: Leishmania EF-1alphaactivates the Src homology 2 domain containing tyrosinephosphatase SHP-1 leading to macrophage deactivation.  JBiol Chem 2002, 277:50190-50197.18. Sereno D, Vanhille L, Vergnes B, Monte-Allegre A, Ouaissi A: Exper-imental study of the function of the excreted/secreted Leish-mania LmSIR2 protein by heterologous expression ineukaryotic cell line.  Kinetoplastid Biol Dis 2005, 4:1.19. Foster LJ, Rudich A, Talior I, Patel N, Huang X, Furtado LM, Bilan PJ,Mann M, Klip A: Insulin-dependent Interactions of Proteinswith GLUT4 Revealed through Stable Isotope Labeling byAmino Acids in Cell Culture (SILAC)*.  J Proteome Res 2006,5:64-75.20. Ong SE, Mann M: A practical recipe for stable isotope labelingby amino acids in cell culture (SILAC).  Nat Protoc 2006,1:2650-2660.21. Ong SE, Blagoev B, Kratchmarova I, Kristensen DB, Steen H, PandeyA, Mann M: Stable isotope labeling by amino acids in cell cul-ture, SILAC, as a simple and accurate approach to expres-sion proteomics.  Mol Cell Proteomics 2002, 1:376-386.22. Bates PA, Gottlieb M, Dwyer DM: Leishmania donovani: identifi-cation of glycoproteins released by promastigotes duringgrowth in vitro.  Exp Parasitol 1988, 67:199-209.23. Ralton JE, Naderer T, Piraino HL, Bashtannyk TA, Callaghan JM,McConville MJ: Evidence that intracellular beta1-2 mannan isa virulence factor in Leishmania parasites.  J Biol Chem 2003,278:40757-40763.24. Thery C, Boussac M, Veron P, Ricciardi-Castagnoli P, Raposo G,Garin J, Amigorena S: Proteomic analysis of dendritic cell-derived exosomes: a secreted subcellular compartment dis-tinct from apoptotic vesicles.  J Immunol 2001, 166:7309-7318.25. van Zandbergen G, Bollinger A, Wenzel A, Kamhawi S, Voll R, KlingerM, Muller A, Holscher C, Herrmann M, Sacks D, Solbach W, LaskayT: Leishmania disease development depends on the presenceof apoptotic promastigotes in the virulent inoculum.  PNAS2006, 103:13837-13842.26. Foster LJ, Zeemann PA, Li C, Mann M, Jensen ON, Kassem M: Differ-ential expression profiling of membrane proteins by quanti-tative proteomics in a human mesenchymal stem cell lineundergoing osteoblast differentiation.  Stem Cells 2005,23:1367-1377.27. Foster LJ, de Hoog CL, Zhang Y, Zhang Y, Xie X, Mootha VK, MannGenome Biology 2008, 9:R356. Wolday D, Berhe N, Akuffo H, Desjeux P, Britton S: EmergingLeishmania/HIV co-infection in Africa.  Med Microbiol Immunol2001, 190:65-67.7. Mishra J, Saxena A, Singh S: Chemotherapy of leishmaniasis:M: A mammalian organelle map by protein correlationprofiling.  Cell 2006, 125:187-199.28. Shakarian AM, Joshi MB, Ghedin E, Dwyer DM: Molecular dissec-tion of the functional domains of a unique, tartrate-resistant,35.20http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rsurface membrane acid phosphatase in the primitive humanpathogen Leishmania donovani.  J Biol Chem 2002,277:17994-18001.29. The Leishmania major Genome Database.  2007 [http://www.genedb.org/genedb/leish/].30. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM,Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M,Rubin GM, Sherlock G: Gene ontology: tool for the unificationof biology. The Gene Ontology Consortium.  Nat Genet 2000,25:25-29.31. Conesa A, Gotz S, Garcia-Gomez JM, Terol J, Talon M, Robles M:Blast2GO: a universal tool for annotation, visualization andanalysis in functional genomics research.  Bioinformatics 2005,21:3674-3676.32. Bluthgen N, Brand K, Cajavec B, Swat M, Herzel H, Beule D: Biolog-ical profiling of gene groups utilizing Gene Ontology.  GenomeInform 2005, 16:106-115.33. Gomez M, Johnson S, Gennaro ML: Identification of secretedproteins of Mycobacterium tuberculosis by a bioinformaticapproach.  Infect Immun 2000, 68:2323-2327.34. Ivens AC, Peacock CS, Worthey EA, Murphy L, Aggarwal G, BerrimanM, Sisk E, Rajandream MA, Adlem E, Aert R, Anupama A, ApostolouZ, Attipoe P, Bason N, Bauser C, Beck A, Beverley SM, Bianchettin G,Borzym K, Bothe G, Bruschi CV, Collins M, Cadag E, Ciarloni L, Clay-ton C, Coulson RM, Cronin A, Cruz AK, Davies RM, De GJ, et al.:The genome of the kinetoplastid parasite, Leishmaniamajor.  Science 2005, 309:436-442.35. Chen DQ, Kolli BK, Yadava N, Lu HG, Gilman-Sachs A, Peterson DA,Chang KP: Episomal expression of specific sense and antisensemRNAs in Leishmania amazonensis: Modulation of gp63level in promastigotes and their infection of macrophages invitro.  Infect Immun 2000, 68:80-86.36. Joshi MB, Rogers ME, Shakarian AM, Yamage M, Al-Harthi SA, BatesPA, Dwyer DM: Molecular characterization, expression, and invivo analysis of LmexCht1: the chitinase of the human path-ogen, Leishmania mexicana.  J Biol Chem 2005, 280:3847-3861.37. Shakarian AM, Ellis SL, Mallinson DJ, Olafson RW, Dwyer DM: Twotandemly arrayed genes encode the (histidine) secretoryacid phosphatases of Leishmania donovani.  Gene 1997,196:127-137.38. Raposo G, Nijman HW, Stoorvogel W, Liejendekker R, Harding CV,Melief CJ, Geuze HJ: B lymphocytes secrete antigen-presentingvesicles.  J Exp Med 1996, 183:1161-1172.39. Aoki N, Jin-no S, Nakagawa Y, Asai N, Arakawa E, Tamura N, TamuraT, Matsuda T: Identification and characterization of microves-icles secreted by 3T3-L1 adipocytes: redox- and hormone-dependent induction of milk fat globule-epidermal growthfactor 8-associated microvesicles.  Endocrinology 2007,148:3850-3862.40. Aoki N, Jin-no S, Nakagawa Y, Asai N, Arakawa E, Tamura N, TamuraT, Matsuda T: Identification and characterization of microves-icles secreted by 3T3-L1 adipocytes: redox- and hormone-dependent induction of milk fat globule-epidermal growthfactor 8-associated microvesicles.  Endocrinology 2007,148:3850-3862.41. Michels PA, Bringaud F, Herman M, Hannaert V: Metabolic func-tions of glycosomes in trypanosomatids.  Biochim Biophys Acta2006, 1763:1463-1477.42. Moyersoen J, Choe J, Fan E, Hol WG, Michels PA: Biogenesis ofperoxisomes and glycosomes: trypanosomatid glycosomeassembly is a promising new drug target.  FEMS Microbiol Rev2004, 28:603-643.43. Opperdoes FR, Szikora JP: In silico prediction of the glycosomalenzymes of Leishmania major and trypanosomes.  Mol Bio-chem Parasitol 2006, 147:193-206.44. Beatty WL, Russell DG: Identification of mycobacterial surfaceproteins released into subcellular compartments of infectedmacrophages.  Infect Immun 2000, 68:6997-7002.45. Chakraborty D, Banerjee S, Sen A, Banerjee KK, Das P, Roy S: Leish-mania donovani Affects Antigen Presentation of Macro-phage by Disrupting Lipid Rafts.  J Immunol 2005, 175:3214-3224.46. Johnson BJ, Le TTT, Dobbin CA, Banovic T, Howard CB, FloresFdML, Vanags D, Naylor DJ, Hill GR, Suhrbier A: Heat Shock Pro-W, Laskay T: Inhibition of the spontaneous apoptosis of neu-trophil granulocytes by the intracellular parasite Leishmaniamajor.  J Immunol 2002, 169:898-905.48. Ruhland A, Leal N, Kima PE: Leishmania promastigotes activatePI3K/Akt signalling to confer host cell resistance toapoptosis.  Cell Microbiol 2007, 9:84-96.49. Rappu P, Shin BS, Zalkin H, Mantsala P: A Role for a Highly Con-served Protein of Unknown Function in Regulation of Bacil-lus subtilis purA by the Purine Repressor.  J Bacteriol 1999,181:3810-3815.50. Al-Salabi MI, de Koning HP: Purine Nucleobase Transport inAmastigotes of Leishmania mexicana: Involvement in Allop-urinol Uptake.  Antimicrob Agents Chemother 2005, 49:3682-3689.51. Ilg T: Generation of myo-inositol-auxotrophic Leishmaniamexicana mutants by targeted replacement of the myo-inositol-1-phosphate synthase gene.  Mol Biochem Parasitol 2002,120:151-156.52. Carvalho LP, Passos S, Dutra WO, Soto M, Alonso C, Gollob KJ, Car-valho EM, Ribeiro de Jesus A: Effect of LACK and KMP11 on IFN-gamma Production by Peripheral Blood Mononuclear Cellsfrom Cutaneous and Mucosal Leishmaniasis Patients.  Scandi-navian Journal of Immunology 2005, 61:337-342.53. Flaster H, Bernhagen J, Calandra T, Bucala R: The MacrophageMigration Inhibitory Factor-Glucocorticoid Dyad: Regula-tion of Inflammation and Immunity.  Mol Endocrinol 2007,21:1267-1280.54. Descoteaux A, Turco SJ, Sacks DL, Matlashewski G: Leishmaniadonovani lipophosphoglycan selectively inhibits signal trans-duction in macrophages.  J Immunol 1991, 146:2747-2753.55. Nandan D, Lo R, Reiner NE: Activation of phosphotyrosinephosphatase activity attenuates mitogen-activated proteinkinase signaling and inhibits c-FOS and nitric oxide synthaseexpression in macrophages infected with Leishmaniadonovani.  Infect Immun 1999, 67:4055-4063.56. Nandan D, Reiner NE: Attenuation of gamma interferon-induced tyrosine phosphorylation in mononuclear phago-cytes infected with Leishmania donovani: selective inhibitionof signaling through Janus kinases and Stat1.  Infect Immun1995, 63:4495-4500.57. Nandan D, Knutson KL, Lo R, Reiner NE: Exploitation of host cellsignaling machinery: activation of macrophage phosphotyro-sine phosphatases as a novel mechanism of molecular micro-bial pathogenesis.  J Leukoc Biol 2000, 67:464-470.58. Olivier M, Gregory DJ, Forget G: Subversion mechanisms bywhich Leishmania parasites can escape the host immuneresponse: a signaling point of view.  Clin Microbiol Rev 2005,18:293-305.59. Krsmanovic T, Pawelec A, Sydor T, Kolling R: Control of Ste6Recycling by Ubiquitination in the Early Endocytic Pathwayin Yeast.  Mol Biol Cell 2005, 16:2809-2821.60. Girard M, Allaire PD, McPherson PS, Blondeau F: Non-stoichiomet-ric Relationship between Clathrin Heavy and Light ChainsRevealed by Quantitative Comparative Proteomics of Clath-rin-coated Vesicles from Brain and Liver.  Mol Cell Proteomics2005, 4:1145-1154.61. Aoki N, Jin-no S, Nakagawa Y, Asai N, Arakawa E, Tamura N, TamuraT, Matsuda T: Identification and characterization of microves-icles secreted by 3T3-L1 adipocytes: redox- and hormone-dependent induction of milk fat globule-epidermal growthfactor 8-associated microvesicles.  Endocrinology 2007,148:3850-3862.62. Clayton A, Turkes A, Navabi H, Mason MD, Tabi Z: Induction ofheat shock proteins in B-cell exosomes.  J Cell Sci 2005,118:3631-3638.63. Faure J, Lachenal G, Court M, Hirrlinger J, Chatellard-Causse C, BlotB, Grange J, Schoehn G, Goldberg Y, Boyer V, Kirchhoff F, Raposo G,Garin J, Sadoul R: Exosomes are released by cultured corticalneurones.  Mol Cell Neurosci 2006, 31:642-648.64. Wubbolts R, Leckie RS, Veenhuizen PT, Schwarzmann G, Mobius W,Hoernschemeyer J, Slot JW, Geuze HJ, Stoorvogel W: Proteomicand biochemical analyses of human B cell-derived exosomes.Potential implications for their function and multivesicularbody formation.  J Biol Chem 2003, 278:10963-10972.65. Denny PW, Morgan GW, Field MC, Smith DF: Leishmania major:Genome Biology 2008, 9:R35tein 10 Inhibits Lipopolysaccharide-induced InflammatoryMediator Production.  Journal of Biological Chemistry 2005,280:4037-4047.47. Aga E, Katschinski DM, van ZG, Laufs H, Hansen B, Muller K, Solbachclathrin and adaptin complexes of an intra-cellular parasite.Experimental Parasitology 2005, 109:33-37.66. Gokool S: {sigma}1- and {micro}1-Adaptin Homologues ofLeishmania mexicana Are Required for Parasite Survival in35.21http://genomebiology.com/2008/9/2/R35 Genome Biology 2008,     Volume 9, Issue 2, Article R35       Silverman et al. Rthe Infected Host.  Journal of Biological Chemistry 2003,278:29400-29409.67. Fevrier B, Raposo G: Exosomes: endosomal-derived vesiclesshipping extracellular messages.  Curr Opin Cell Biol 2004,16:415-421.68. Van NG, Porto-Carreiro I, Simoes S, Raposo G: Exosomes: a com-mon pathway for a specialized function.  J Biochem 2006,140:13-21.69. Blott EJ, Griffiths GM: Secretory lysosomes.  Nat Rev Mol Cell Biol2002, 3:122-131.70. Blanc L, de Gassart A, Geminard C, Bette-Bobillo P, Vidal M: Exo-some release by reticulocytes - An integral part of the redblood cell differentiation system.  Blood Cells Mol Dis 2005,35:21-26.71. Andre F, Chaput N, Schartz NE, Flament C, Aubert N, Bernard J,Lemonnier F, Raposo G, Escudier B, Hsu DH, Tursz T, Amigorena S,Angevin E, Zitvogel L: Exosomes as potent cell-free peptide-based vaccine. I. Dendritic cell-derived exosomes transferfunctional MHC class I/peptide complexes to dendritic cells.J Immunol 2004, 172:2126-2136.72. Andre F, Andersen M, Wolfers J, Lozier A, Raposo G, Serra V, RueggC, Flament C, Angevin E, Amigorena S, Zitvogel L: Exosomes incancer immunotherapy: preclinical data.  Adv Exp Med Biol2001, 495:349-354.73. Zitvogel L, Fernandez N, Lozier A, Wolfers J, Regnault A, Raposo G,Amigorena S: Dendritic cells or their exosomes are effectivebiotherapies of cancer.  Eur J Cancer 1999, 35(Suppl 3):S36-S38.74. Zitvogel L: Eradication of established murine tumors using anovel cell-free vaccine: dendritic cell-derived exosomes.Nature Med 1998, 4:594-600.75. Mullin KA, Foth BJ, Ilgoutz SC, Callaghan JM, Zawadzki JL, McFaddenGI, McConville MJ: Regulated Degradation of an EndoplasmicReticulum Membrane Protein in a Tubular Lysosome inLeishmania mexicana.  Mol Biol Cell 2001, 12:2364-2377.76. Caby MP, Lankar D, Vincendeau-Scherrer C, Raposo G, Bonnerot C:Exosomal-like vesicles are present in human blood plasma.Int Immunol 2005, 17:879-887.77. Kurosaka K, Takahashi M, Watanabe N, Kobayashi Y: SilentCleanup of Very Early Apoptotic Cells by Macrophages.  JImmunol 2003, 171:4672-4679.78. Colasante C, Ellis M, Ruppert T, Voncken F: Comparativeproteomics of glycosomes from bloodstream form and pro-cyclic culture form Trypanosoma brucei brucei.  Proteomics2006, 6:3275-3293.79. Zhang WW, Matlashewski G: In vivo selection for Leishmaniadonovani miniexon genes that increase virulence in Leish-mania major.  Mol Microbiol 2004, 54:1051-1062.80. Lodge R, Descoteaux A: Modulation of phagolysosomebiogenesis by the lipophosphoglycan of Leishmania.  ClinImmunol 2005, 114:256-265.81. Lerm M, Holm A, Seiron A, Sarndahl E, Magnusson KE, Rasmusson B:Leishmania donovani requires functional Cdc42 and Rac1 toprevent phagosomal maturation.  Infect Immun 2006,74:2613-2618.82. Scianimanico S, Desrosiers M, Dermine JF, Meresse S, Descoteaux A,Desjardins M: Impaired recruitment of the small GTPase rab7correlates with the inhibition of phagosome maturation byLeishmania donovani promastigotes.  Cell Microbiol 1999,1:19-32.83. Srividya G, Duncan R, Sharma P, Raju BV, Nakhasi HL, Salotra P:Transcriptome analysis during the process of in vitro differ-entiation of Leishmania donovani using genomicmicroarrays.  Parasitology 2007, 134:1527-1539.84. Marti M, Good RT, Rug M, Knuepfer E, Cowman AF: Targetingmalaria virulence and remodeling proteins to the hosterythrocyte.  Science 2004, 306:1930-1933.85. Fossati G, Izzo G, Rizzi E, Gancia E, Modena D, Moras ML, NiccolaiN, Giannozzi E, Spiga O, Bono L, Marone P, Leone E, Mangili F, Hard-ing S, Errington N, Walters C, Henderson B, Roberts MM, CoatesARM, Casetta B, Mascagni P: Mycobacterium tuberculosis Chap-eronin 10 Is Secreted in the Macrophage Phagosome: IsSecretion Due to Dissociation and Adoption of a PartiallyHelical Structure at the Membrane?  J Bacteriol 2003,185:4256-4267.2006, 6:136-147.88. Nandan D, Reiner NE: Leishmania donovani engages in regula-tory interference by targeting macrophage protein tyrosinephosphatase SHP-1.  Clin Immunol 2005, 114:266-277.89. Blanchette J, Racette N, Faure R, Siminovitch KA, Olivier M: Leish-mania-induced increases in activation of macrophage SHP-1tyrosine phosphatase are associated with impaired IFN-gamma-triggered JAK2 activation.  Eur J Immunol 1999,29:3737-3744.90. Dasgupta B, Roychoudhury K, Ganguly S, Akbar MA, Das P, Roy S:Infection of human mononuclear phagocytes and macro-phage-like THP1 cells with Leishmania donovani results inmodulation of expression of a subset of chemokines and achemokine receptor.  Scand J Immunol 2003, 57:366-374.91. Overath P, Aebischer T: Antigen Presentation by MacrophagesHarboring Intravesicular Pathogens.  Parasitology Today 1999,15:325-332.92. Solbach W, Laskay T: The host response to Leishmaniainfection.  Adv Immunol 2000, 74:275-317.93. Caldwell RB, Lattemann CT: Simple and Reliable Method ToPrecipitate Proteins from Bacterial Culture Supernatant.Appl Environ Microbiol 2004, 70:610-612.94. Chan QWT, Howes CG, Foster LJ: Quantitative Comparison ofCaste Differences in Honeybee Hemolymph.  Mol CellProteomics 2006, 5:2252-2262.95. Rappsilber J, Ishihama Y, Mann M: Stop and go extraction tips formatrix-assisted laser desorption/ionization, nanoelectro-spray, and LC/MS sample pretreatment in proteomics.  AnalChem 2003, 75:663-670.96. MSquant.   [http://MSquant.sourceforge.net]. 4 Feb 200897. The Matrix Sicence Website.  Electrophoresis  [http://www.matrixscience.com/]. 11 Feb 200898. The Foster Lab.   [http://www.proteomics.ubc.ca/foster/software].11 Feb 200899. Bates PA, Kurtz MK, Gottlieb M, Dwyer DM: Leishmania dono-vani: generation of monospecific antibody reagents to solu-ble acid phosphatase.  Exp Parasitol 1987, 64:157-164.100. Hubel A, Brandau S, Dresel A, Clos J: A member of the ClpB fam-ily of stress proteins is expressed during heat shock in Leish-mania spp.  Mol Biochem Parasitol 1995, 70:107-118.101. Emanuelsson O, Brunak S, von HG, Nielsen H: Locating proteinsin the cell using TargetP, SignalP and related tools.  Nat Protoc2007, 2:953-971.102. The Tmpred Website.  2008 [http://www.ch.embnet.org/software/TMPRED_form.html].103. Fankhauser N, Maser P: Identification of GPI anchor attach-ment signals by a Kohonen self-organizing map.  Bioinformatics2005, 21:1846-1852.104. Aoki N, Jin-no S, Nakagawa Y, Asai N, Arakawa E, Tamura N, TamuraT, Matsuda T: Identification and characterization of microves-icles secreted by 3T3-L1 adipocytes: redox- and hormone-dependent induction of milk fat globule-epidermal growthfactor 8-associated microvesicles.  Endocrinology 2007,148:3850-3862.105. Soto M, Requena JM, Garcia M, Gomez LC, Navarrete I, Alonso C:Genomic organization and expression of two independentgene arrays coding for two antigenic acidic ribosomal pro-teins of Leishmania.  Journal of Biological Chemistry 1993,268:21835-21843.106. Aoki N, Jin-no S, Nakagawa Y, Asai N, Arakawa E, Tamura N, TamuraT, Matsuda T: Identification and characterization of microves-icles secreted by 3T3-L1 adipocytes: redox- and hormone-dependent induction of milk fat globule-epidermal growthfactor 8-associated microvesicles.  Endocrinology 2007,148:3850-3862.107. Colgan J, Asmal M, Yu B, Luban J: Cyclophilin A-Deficient MiceAre Resistant to Immunosuppression by Cyclosporine.  JImmunol 2005, 174:6030-6038.108. Vanegas G, Quinones W, Carrasco-Lopez C, Concepcion JL, Alberi-cio F, Avilan L: Enolase as a plasminogen binding protein inLeishmania mexicana.  Parasitol Res 2007, 101:1511-1516.Genome Biology 2008, 9:R3586. Reiner NE: Targeting cell regulation promotes pathogen sur-vival in macrophages.  Clin Immunol 2005, 114:213-215.87. Roy CR, Salcedo SP, Gorvel JP: Pathogen-endoplasmic-reticu-lum interactions: in through the out door.  Nat Rev Immunol


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items