Surface proteomics and label-free quantification of Leptospira interrogans serovar Pomona

Leptospirosis is a re-emerging zoonosis with a global distribution. Surface-exposed outer membrane proteins (SE-OMPs) are crucial for bacterial–host interactions. SE-OMPs locate and expose their epitope on cell surface where is easily accessed by host molecules. This study aimed to screen for surface-exposed proteins and their abundance profile of pathogenic Leptospira interrogans serovar Pomona. Two complementary approaches, surface biotinylation and surface proteolytic shaving, followed by liquid chromatography tandem-mass spectrometry (LC-MS/MS) were employed to identify SE-OMPs of intact leptospires. For quantitative comparison, in-depth label-free analysis of SE-OMPs obtained from each method was performed using MaxQuant. The total number of proteins identified was 1,001 and 238 for surface biotinylation and proteinase K shaving, respectively. Among these, 39 were previously known SE-OMPs and 68 were predicted to be localized on the leptospiral surface. Based on MaxQuant analysis for relative quantification, six known SE-OMPs including EF- Tu, LipL21, LipL41, LipL46, Loa22, and OmpL36, and one predicted SE-OMPs, LipL71 were found in the 20 most abundant proteins, in which LipL41 was the highest abundant SE-OMP. Moreover, uncharacterized LIC14011 protein (LIP3228 ortholog in serovar Pomona) was identified as a novel predicted surface βb-OMP. High-abundance leptospiral SE-OMPs identified in this study may play roles in virulence and infection and are potential targets for development of vaccine or diagnostic tests for leptospirosis.


Introduction
Leptospirosis is a neglected zoonosis with high global prevalence, mainly in tropical and subtropical regions including Southeast Asia, Oceania, the Indian subcontinent, Caribbean, and Latin America [1,2]. Leptospirosis causes approximately 1.03 million cases and 58,900 deaths annually worldwide [3]. The severity of the disease ranges from asymptomatic infection to severe manifestations with multi-organ dysfunction, such as renal and hepatic failure, pulmonary hemorrhage, and myocarditis [2]. The major etiologic agent of the illness is pathogenic Leptospira spp., but some cases are associated with intermediate species such as L. inadai and L. wolffii [4]. Most wild and domestic animals can be reservoir hosts harboring the pathogens in their kidneys. Humans are considered as accidental hosts by direct contact with leptospires shed in urine of infected animals or indirect exposure to water or soil contaminated with the urine [2].
Bacterial outer membrane proteins (OMPs) often play key roles in bacterial pathogenesis such as adhesins, porins, targets for antibodies, and receptors for various host molecules. OMPs are likely to be crucial for adaptation and response to host conditions and surrounding environments [5]. Surface-exposed (SE) domains of OMPs on pathogenic leptospires are important because of their location on a tip of cell surface where facilitates bacterial-host interactions [6]. Identification of leptospiral SE-OMPs should be helpful not only for a better understanding of Leptospira pathogenicity but also new targets for the development of vaccines and diagnostic tests [7]. A variety of surface protein assessment methods have been established for isolation of bacterial surface-associated proteins [8,9]. Surface biotinylation method using hydrophilic biotins with membrane impermeable properties can label surface proteins on intact cells [10][11][12]. As a complementary approach, proteolytic shaving of surface proteins under conditions that preserve cell integrity can harvest SE-OMPs [12][13][14]. Proteinase (proK) has broad cleavage sites resulting in coverage of most surface proteins [9]. Pinne and Haake previously used surface biotinylation and surface shaving with proK to identify new 4 surface-exposed leptospiral OMPs [12]. The quantitative profile on peptides or proteins from different biological samples and conditions become important for advanced medical research. MaxQuant is one of the most widely used computational software to quantify protein from mass spectral peak intensity to reveal protein abundance profiling [15]. This software is easy-to-use, and available for label-free quantification (LFQ) and analysis of data obtained from most MS platforms.
In this study, a combination of surface biotinylation and surface shaving methods followed by liquid chromatography tandem-mass spectrometry (LC-MS/MS) were used for high-throughput identification of SE-OMPs on L. interrogans serovar Pomona. The results were subsequently analyzed by MaxQuant software for quantitative profiling of leptospiral SE-OMPs. Our findings revealed known and putative SE-OMPs and the abundance profile of pathogenic Leptospira.

Leptospira and culture
Leptospira interrogans serovar Pomona, kindly provided by Ben Adler, was originally from Lee Smythe, World Health Organization/Food and Agricultural Organization/Office International des Epizooties Collaborating Centre for Reference and Research on Leptospirosis, Queensland Health Scientific Services, Australia. All experiments used low-passage leptospires, which was directly isolated from infected hamsters [16] followed by less than five in vitro passages. Leptospira was cultivated at 30˚C in Ellinghausen-McCullough-Johnson-Harris (EMJH) medium (Difco, MD, USA) containing 10% bovine serum albumin (BSA) supplement solution [17] until the log phase (~5×10 8 cells/ml) was reached.

Preparation of leptospiral surface proteins
The surface biotinylation and proteinase K shaving assays were performed following previously described protocols with some modifications [12]. Leptospires were harvested by lowspeed centrifugation at 2,000×g for 7 min at room temperature (RT). The leptospiral cell pellets were gently washed twice with BSA-free EMJH base medium and approximately 1×10 10 cells were prepared for each method as follows.
Cell surface biotinylation. The surface biotinylation was carried out using Pierce Cell Surface Protein Biotinylation and Isolation Kit (Thermo Scientific) according to the manufacturer's instruction. The cell pellets were resuspended with BupH Phosphate Buffered Saline (BupH-PBS) containing Sulfo-NHS-SS-Biotin at various concentrations up to 1 mg/ml. After incubation at RT for 30 min, the reaction was stopped with BupH-PBS containing 100mM glycine and the inactivated biotin was removed by washing twice with BupH-PBS. The labeled pellets were resuspended with BupH-PBS containing cOmplete protease inhibitors cocktail (Roche) before sonication on ice for 30 min using Ultrasonic Processor (GE Healthcare, Buckinghamshire, UK) (pulse on for 15 sec, pulse off for 45 sec, 35% amplitude). The cell lysates were centrifuged at 15,000×g for 5 min at 4˚C and biotin-labeled proteins in the supernatant were purified using NeutrAvidin Agarose column according to the manufacturer's instruction. The eluted proteins were collected for LC-MS/MS. Cell surface shaving with proteinase K. The cell pellets were resuspended with Proteinase K Solution (Promega, Madison, WI) at various concentrations up to 2 μg/ml. After incubation at 37˚C for 30 min, the reactions were stopped with cOmplete protease inhibitors cocktail (Roche) solution. The cell lysates were centrifuged at 15,000×g at 4˚C for 5 min and the supernatant containing cleaved surface proteins was collected for LC-MS/MS analysis. The remaining cell pellets were subjected to SDS-PAGE and Western blotting.

Live/Dead fluorescence viability staining
Cell membrane integrity was determined using LIVE/DEAD BacLight Bacterial Viability Kit (Thermo Scientific) according to the manufacturer's instruction. Equal volumes of SYTO9 and propidium iodide (PI) were mixed. The biotinylated cells and the remaining cell pellets after proK shaving were incubated with dye mixture at a ratio of 1:1,000 in the dark at RT for 15 min. The stained cells were observed under a fluorescence microscope (Olympus) at 400× magnification. Leptospiral cells treated with cold absolute methanol for 5 min on ice were used as control cells with compromised membrane.

Sodium dodecyl sulphate-polyacrylamide gel electrophoresis
Protein samples were prepared following previously described protocols [18] except biotinlabeled samples were mixed with 1× sample buffer without reducing agent. The samples were characterized by 15% polyacrylamide gel and stained with Coomassie Brilliant Blue R-250 (Bio-Rad, Germany).

Western blotting
Proteins on polyacrylamide gel were transferred onto nitrocellulose membrane. After incubating with 1% BSA in PBS plus 0.05% Tween-20 (PBST) at RT for 1 h, the membrane was incubated with polyclonal antibodies (kindly provided by David A. Haake, UCLA) against OmpL1 (1: 2,000) and FlaA1 (1: 2,000) at RT for 1 h. Afterwards, the membrane was incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (1: 5,000, KPL, MD, USA) at RT for 1 h. For biotinylated proteins, the membrane was incubated with HRP -conjugated streptavidin (1:5000; BD Pharmingen). After each incubation, the membrane was washed three times for 5 min with PBST. The protein bands were detected with ECL chemiluminescent substrate (Amersham ECL Prime, GE Healthcare) under ChemiDoc XRS+ System (Bio-Rad).

In-gel digestion
Protein samples were prepared for mass spectrometry using previously described protocols [18,19] with some modifications. The entire sample lane on Coomassie Brilliant Blue R-250 stained polyacrylamide gel was diced into approximately 1 mm 3 pieces and destained three times for 10 min with 25 mM ammonium bicarbonate (Ambic) in 50% acetonitrile (ACN). The gel pieces were dehydrated with 100% ACN for 5 min and completely dried by a speed vacuum device (Thermo scientific). The samples were reduced with 10 mM DTT in 25 mM Ambic at 56˚C for 45 min and alkylated with 55 mM iodoacetamide in 25 mM Ambic at RT for 30 min in the dark. The gel pieces were dehydrated and completely dried before incubating with Sequencing Grade Modified Trypsin (Promega) on ice for 60 min. The excess solution was replaced with 25 mM Ambic and the protein samples were incubated at 37˚C overnight. The digested peptides were extracted from the gel pieces with 50% ACN in 0.1% formic acid (FA) and dried in vacuo. The dried samples were acidified with 0.1% FA and desalted with C-18 spin columns (Thermo Scientific) according to the manufacturer's instruction. The dried peptides were finally resuspended in 0.1% FA before applying to LC-MS/MS.

Liquid chromatography tandem-mass spectrometry
Peptide analysis by LC-MS/MS was performed using an EASY-nLC1000 system (Thermo Scientific) coupled to a Q-Exactive Orbitrap Plus mass spectrometer (Thermo Scientific) equipped with a nano-electrospray ion source (Thermo Scientific). The peptides were eluted in 5-20% acetonitrile in 0.1% FA for 40 min followed by 20-40% acetonitrile in 0.1% FA for 10 min and 40-98% acetonitrile in 0.1% FA for 10 min at a flow rate of 300 nl/min. The MS methods included a full MS scan at a resolution of 70,000 followed by 10 data-dependent MS2 scans at a resolution of 17,500. The normalized collision energy of HCD fragmentation was set at 32%. An MS scan range of 350 to 1400 m/z was selected and precursor ions with unassigned charge states, a charge state of +1, or a charge state of greater than +8 were excluded. A dynamic exclusion of 30 s was used.

Protein identification and label-free quantification by MaxQuant
Protein identification and label-free quantitative analysis were performed using MaxQuant software suite version 1.6.1.0 and its built-in Andromeda search engine. The mass spectra data (raw file format) derived from LC-MS/MS was searched against the whole L. interrogans serovar Copenhageni Fiocruz L1-130 protein database available in Uniprot database (www. uniprot.org, taxonomy = 267671). The data (three independent experiments) from each surface protein assessment method were analyzed in separate MaxQuant run for protein identification. In addition, three biological replicates from each method were assigned to different experiments and analyzed in the same software run for quantitative proteomics. The default parameters for LFQ were used with some additional settings as follows: Group specific parameters; Instrument = Orbitrap; Digestion = Specific with Trypsin/P; Label free quantification = LFQ with LFQ min. ratio count = 2, Fast LFQ was selected, LFQ min. number of neighbors = 3, LFQ average number of neighbors = 6. The MaxQuant output file (pro-teinGroups.txt format) was further analyzed using Microsoft Excel. Any contaminants and reverse identified proteins were removed from the total proteins list. The proteins repeatedly identified by at least 2 out of 3 replicates were considered proteins that were truly yielded by each method. The results of known leptospiral SE-OMPs retrieved from PubMed and Scopus are shown in S1 Table.

Optimization of cell surface biotinylation with Sulfo-NHS-SS-Biotin
Initially, the concentration of Sulfo-NHS-SS-Biotin was optimized for surface protein labeling. SDS-PAGE showed approximately equal amounts of total protein used for each concentration of biotin ( Fig 1A). Band intensities of biotinylated proteins on the Western blot reached the highest signal when the biotin was used at the final concentration of 0.4 mg/ml (Figs 1B and S1), therefore subsequent labeling was performed using 0.4 mg/ml biotin. Only minimal signals were seen in the unlabeled control (0 mg/ml of biotin). The labeling process might disrupt the surface layer and potentially release proteins from interior compartments. The Live/Dead fluorescence viability staining was performed to determine membrane integrity before and after surface biotinylation. Most leptospires were stained with SYTO9 (green), which indicated intact cells, before ( Fig  2A) and after ( Fig 2B) biotin labeling suggesting that the membrane integrity was mainly preserved after the labeling process. In contrast, all leptospires treated with methanol used as control cells with damaged membrane were stained red with PI as expected ( Fig 2C).
After surface labeling and purification through the avidin agarose column, the purity of surface proteins in the eluted fraction was determined by immunoblotting using antisera against OmpL1 and FlaA1, known leptospiral SE-OMP and periplasmic proteins, respectively (Fig 3). OmpL1 but not FlaA1 was detected in the eluted fraction (S2 Fig), indicating that biotinylation optimally occurred on the cell surface, and the purified proteins mainly contained OMPs. The surface-exposed proteins were further analyzed by LC-MS/MS.

Optimization of cell surface shaving with proteinase K treatment
The optimal concentration of proK that cleaved only surface proteins and still maintained cell membrane integrity was determined. After proK treatment, the presence of OmpL1 and FlaA1 in the cell pellets containing shaved cells was determined by immunoblotting. When a comparable amount of each sample was used (Fig 4A), the band intensity of OmpL1 but not FlaA1 was reduced at 1 μg/ml proK treatment (Fig 4B) indicating that surface proteins were cleaved with the least effect on periplasmic proteins. The cell viability staining with SYTO9/PI showed that lytic cells increased as a dose-dependent manner (Fig 5). Most leptospiral cells remained intact after treatment with 1 μg/ml proK. Therefore, proK treatment at a final concentration of 1 μg/ml was used to cleave SE-OMPs from remaining intact cells. The supernatant portion obtained from treated cells was used to identify SE-OMPs by LC-MS/MS.

Identification of proteins from surface biotinylation and proteinase K shaving by LC-MS/MS
Surface biotinylation and proK shaving of intact L. interrogans serovar Pomona were subsequently performed according to the optimized protocols. Due to unavailable protein database of serovar Pomona, the mass spectrum data were searched against the Uniprot database of L. interrogans serovar Copenhageni Fiocruz L1-130 containing a total of 3,655 proteins. After MaxQuant analysis, the biological replicates were normalized and averaged the data (mean of two or three data) to obtain a single data set of LFQ intensity from each method (S1 Appendix). The analysis demonstrated that surface biotinylation yielded a total of 1,001 proteins, proK shaving yielded a total of 238 proteins, and both methods identified a total of 1,019 proteins (Fig 6A). The biotinylation and proK shaving identified unique 781 and 18 proteins, respectively, while 220 proteins overlapped between the two methods. Previously reported 39 SE-OMPs were identified by at least one method (S1 Table). The number of known SE-OMPs identified by biotinylation and proK surface shaving were 38 and 12, respectively. Of these, 11 known SE-OMPs were shared by both sample groups (Fig 6B).
According to the previous studies on reverse vaccinology, two bioinformatics approaches predicted a total of 272 leptospiral SE-OMPs [34,35], of which 23 proteins were known SE-OMPs but 2 proteins were currently missing from the database. Of 247 remaining proteins, there were 68 predicted SE-OMPs identified by at least one method (S2 Table). The predicted 64 and 22 SE-OMPs were detected in biotinylated and proK-shaved samples, respectively ( Fig  6C). The number of predicted SE-OMPs overlapped in both sample groups was 18. Based on this information, total known and predicted SE-OMPs obtained from both techniques accounted for 10.50% (107/1,019) of all identified proteins. Of these, 29 known and predicted SE-OMPs represented 13.19% (29/220) of total identified proteins shared by both methods (Fig 6D).

Quantitative surface proteomics of pathogenic Leptospira
The quantitative profiling of protein abundance was presented as the abundance ranking of LFQ intensity of all identified proteins obtained from each method. The quantitative abundance profiling of known SE-OMPs and predicted SE-OMPs are shown in S1 and S2 Tables, respectively, while the 20 most abundant proteins of each method are shown in Tables 1 and 2.  The highest abundant SE-OMP was LipL41, whatever the method used. Known SE-OMPs including EF-Tu, LipL21, LipL46, Loa22, and OmpL36, and predicted SE-OMP, LipL71 were shown in the top 20 high abundance, in which EF-Tu, LipL46, and OmpL36 were overlapping proteins. Moreover, there were 9 known SE-OMPs and 5 predicted SE-OMPs ranked in the top 50 high abundance (S3 Table). Of these, 8 proteins including EF-Tu, LipL21, LipL41, LipL46, LipL71, McpA, OmpL1, and OmpL36 were common in both sample groups while Loa22, OmpL32, SdhA, SppA, LIC10314, and LIC12615 were specific in each group. However, outer membrane proteins (LipL31, LipL32, and LipL45), periplasmic proteins such as flagellin (FlaA-1 and FlaA-2), cytoplasmic proteins such as global regulator (Rpo family and elongation factor), metabolic enzymes, chaperone proteins (GroEL and DnaK), and proteins with unknown functions were also presented in high quantity.

Subcellular localization prediction of high-abundance unknown proteins
The uncharacterized proteins; LIC10175, LIC10176, LIC10314, LIC10411, LIC10483, LIC11182, LIC11848, LIC12621, LIC13432, were ranked in the top 50 high abundance proteins (S3 Table). They were further predicted to be SE-OMPs using web-based bioinformatics tools following previously described criteria [34]. The amino acid sequences of those uncharacterized proteins in serovar Pomona were used for predictions. The prediction demonstrated that LIP3228 (an orthologous protein of LIC10411) contained a signal peptide and a βb structure but lacked a transmembrane α-helix (S4 Table). Therefore, LIP3228 was finally predicted to be a βb-OMP. However, the remaining 8 uncharacterized proteins were not predicted as SE-OMPs based on those criteria.

Discussion
The location on the exterior of cell surface enables SE-OMPs of pathogens to promptly interact with host molecules and surrounding environments. SE-OMPs of pathogenic leptospires play roles in virulence, including adherence, invasion, colonization, and interaction with the host environment, including the immune system [6].
Several methods were previously used to identify leptospiral SE-OMPs, such as cell surface labeling [10][11][12], surface proteolysis shaving [12][13][14], surface immunofluorescence [12,36], surface immunoprecipitation [37], and in silico analysis [12,34,35]. Of these, surface labeling with biotin and surface shaving with proteinase K can be used as high-throughput screening to identify SE-OMPs. Both methods are practical and reliable for identification of surface proteins of leptospires [10][11][12][13][14] and other bacterial pathogens [38][39][40][41]. In this study, we used two complementary methods, surface biotinylation, and proK shaving previously used to characterize novel leptospiral SE-OMPs [12], to identify proteins localized on the surface of L. interrogans serovar Pomona, thereby providing stronger evidence for surface localization and enhancing the efficacy of total protein coverage compared to the use of only a single method.
In addition, to reduce contamination of cytoplasmic and periplasmic proteins, we optimized the concentration of biotin and proK to minimize leptospiral membrane disruption during biotin labeling and proteolytic shaving process. The Sulfo-NHS-SS-Biotin used in this study is a hydrophilic and cell membrane-impermeable reagent containing sulfonate group on the N-hydroxysuccinimide ring that reacts with primary amines (-NH2) on surface proteins of intact cells. After labeling and purification, Western blot revealed that surface proteins were mainly obtained in the biotinylated protein samples (Fig 3). In parallel, the proK shaving was performed to digest the surface-exposed portion of OMPs at the optimal concentration of proK to prevent cell lysis. The proK was used because it is potent, active at a wide pH range, and low peptide bond specificity adjacent to the carboxyl group of aliphatic and aromatic amino acids resulting in a broad range of surface proteins in the cleaved fraction [9]. The immunoblotting revealed that the surface proteins were predominantly released into the supernatant fraction (Fig 4). The outer membrane of leptospires are fragile and easily disrupted [42], therefore leptospiral cells were handled as gently as possible during protein preparations to minimize membrane degradation. However, a certain degree of cell lysis was observed by fluorescence viability staining after the experimental process (Figs 3 and 5).
The conventional proteomics approach using two-dimensional gel electrophoresis (2DE) coupled with matrix assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) [10,43], mainly identifies abundant proteins but inefficiently identifies highly hydrophobic or membrane proteins. In more recent studies, LC-MS/MS, a highthroughput and high-resolution method, was used to identify leptospiral membrane proteins from the samples of subcellular fractionation using Triton X-114 [44,45]. Moreover, LC-MS/ MS-based studies and isotope labeling couple revealed protein abundance of leptospires [46]. In this study, LC-MS/MS-based surface proteomics and label-free quantification with Max-Quant was used to identify SE-OMPs and calculate their abundance profiles. Surface biotinylation and proK shaving identified a total of 1,001 proteins and 238 proteins, respectively (Fig 6). The difference of results obtained from these two methods was explained by the fact that proK could not cleave all leptospiral surface proteins, especially those lacking proK cleavage sites [12] and biotin might fail to label the proteins that their lysine residues are not exposed [10]. Therefore, the number of SE-OMPs was likely an approximate value.
The previous surface proteome of Leptospira revealed that LipL21, LipL32, and LipL41 were abundant on the cell surface [10]. All three proteins were also identified and ranked in the top 20 abundance list in this study (Tables 1 and 2). However LipL32 was previously confirmed as a subsurface lipoprotein [13], therefore LipL41 was the highest abundant SE-OMP followed by EF-Tu, LipL21, LipL46, LipL71, Loa22, OmpL36. Moreover, we also identified McpA, OmpL1, OmpL32, SdhA, SppA, LIC10314, and LIC12615 as high-abundance SE-OMPs since their abundances were ranked in the top 50 abundance list (S3 Table). These proteins are conserved among pathogenic leptospires, expressed during mammalian infection, and involved in leptospiral pathogenesis [47][48][49][50][51][52]. For example, the leptospires mutant lacking Loa22 expression was attenuated in animal models [50]. LipL46 and OmpL32 were detected in Leptospira residing in tissues of infected animals [51,52]. OmpL36, LipL46, and EF-Tu were able to interact with several host components [47][48][49]. Previous studies demonstrated the potential vaccine candidates and target proteins for diagnostic tests of high abundance SE-OMPs. For example, LipL21, LipL41, and OmpL1 have been tested as vaccine candidates and certain combinations conferred synergistic effect on protection [53,54], supporting the rationale to use high-abundance SE-OMPs in multi-subunit or chimeric vaccines. Recombinant LipL21, LipL41, and Loa22 specifically reacted with sera from leptospirosis patients and specific antibodies against these abundance proteins could recognize their native forms on leptospiral cells [55,56], implying the beneficial application of high-abundance SE-OMPs in diagnosis of leptospirosis.
Reverse vaccinology was previously used for screening leptospiral OMPs and SE-OMPs as new vaccine candidates [34,35]. For example, reverse and three-dimensional structural vaccinology predicting conserved βb transmembrane proteins and OM lipoproteins was employed to select the surface-related vaccine candidates [34]. βb-OMPs and OM lipoproteins are major types of membrane proteins that contain SE-region on the cell surface of Gram-negative bacteria [57,58]. Of all uncharacterized proteins listed in the top 50 abundance, only LIC10411 (LIP3228 ortholog in serovar Pomona) contained the predicted property of βb-OMP and shared overlapping in both sample groups. Moreover, it is a conserved and abundance protein in pathogenic serovars [59]. Therefore, it is a promising novel vaccine candidate. However, other characteristics of good vaccine candidates such as in vivo expression, immunogenicity, and role in pathogenesis should be further evaluated.
Although low-passage leptospires were used to avoid the loss of virulence or the change of their protein expression after high-passage in vitro culture as previously reported [60,61], we might not detect SE-OMPs that were not expressed or were extremely low abundant under in vitro culture conditions. In addition, only the whole L. interrogans serovar Copenhageni Fiocruz L1-130 protein database was available for searching matched mass spectra of proteins, unique proteins in our L. interrogans serovar Pomona might not be identified. Despite the precautions taken to avoid membrane disruption, a small degree of cell lysis was inevitable and known cytoplasmic proteins and known periplasmic proteins were detected in the surface protein-enriched fractions. In addition, some proteins were named by annotation that might mislead as cytoplasmic proteins, for example, EF-Tu was already confirmed as a surface-exposed outer membrane protein of Leptospira [49]. Due to the high sensitivity of LC-MS/MS, even trace amount of proteins contaminated from undesired compartments might be detected, as shown by detection of FlaA1 in the LC-MS/MS results but not on the Western blot. Moreover, some proteins might have multiple subcellular localizations, for example, chaperone proteins including GroEL, DnaK and ClpB [62].
In conclusion, our results demonstrated that the complementary strategy of surface biotinylation and proteinase K shaving followed by LC-MS/MS with label-free quantification was useful to expand the repertoire of surface proteins and the abundance profile of virulent L. interrogans serovar Pomona. We identified several high-abundance SE-OMPs including EF-Tu, LipL21, LipL41, LipL46, LipL71, Loa22, McpA, OmpL1, OmpL32, OmpL36, SdhA, SppA, LIC10314, and LIC12615. Moreover, we reported the in silico-based characterization of LIC10411(LIP3228 ortholog) to be a putative SE-OMP. However, its subcellular localization should be confirmed. Leptospiral surface proteome obtained from this study is useful for further investigation of novel virulence factors of pathogenic leptospires and serves as new targets for vaccine development as well as diagnostic tests for leptospirosis.  Table. Known surface-exposed outer membrane proteins obtained by surface biotinylation and surface shaving. (DOCX) S2 Table. Predicted surface-exposed outer membrane proteins obtained by surface biotinylation and surface shaving.