The virulence of the L. interrogans serovar Pomona strain LPF was maintained by iterative passages in Golden Syrian hamsters. Recently weaned hamsters were intraperitonially infected with 500 µL of approximately 1 × 10⁴ virulent leptospires. The animals were sacrificed after the appearance of symptoms, such as loss of weight and mobility (approximately 5 days post-infection). Kidney and liver were removed and macerated. The leptospires recovered from these organs were cultured at 28°C in semi-solid EMJH-modified medium supplemented with 10% rabbit sera, followed by sub-culture in liquid EMJH (Difco^®- USA) supplemented with 10% rabbit sera, until the density of approximately 10⁸ cells/mL was reached.

The culture was harvested by centrifugation at 12800× g at 4°C for 10 min. Pellet was washed (×5) by resuspension in 35 mL phosphate-buffered saline (PBS) containing 5 mM MgCl₂, followed by centrifugation under similar conditions described earlier. The pellet was weighed and resuspended in DeStreak Rehydration Solution (GE Healthcare, USA) at the proportion of 5 µL/mg of the bacteria. The cells were lysed by vigorous vortexing followed by mechanical pressure through a syringe. The cellular debris was separated by centrifugation at 20800× g for 10 min at room temperature and the supernatant was collected. Total protein content was determined according to the Bradford method (Pierce Biotechnology, USA), following the manufacture’s protocol. Bovine serum albumin (BSA) was used to generate a standard reference curve (0-10 µg). The protein solution was mixed at the appropriate proportions with the Bradford reagent dye, and after 30 min of incubation at room temperature, the readings were taken at 595 nm. The protein concentration from each sample was calculated based on BSA standard absorbance curve. Samples of 700 µg of protein were adjusted to 340 µL with DeStreak Rehydration Solution (GE Healthcare, USA), along with 0.8% (v/v) IPG buffer, with a pH range of 3-10 (GE Healthcare). The protein extracts were obtained in triplicate from bacteria cultured from kidney and liver of infected animals.

First-dimension isoelectric focusing was performed using the IPGphor-System (GE Healthcare, USA), and the second dimension was conducted on the Ettan DALTsix system (GE Healthcare). The IPG gel strips (18 cm) with a linear separation of immobilized pH ranging from 3 to 10 were rehydrated directly with the solubilized samples. The strips were covered with mineral oil to prevent dehydration and oxidation, and the process was carried out overnight at room temperature in the Immobiline Dry-Strip Reswelling Tray (GE Healthcare). The focusing protocol was 30 V for 180 Vh, 150 V for 300 Vh, 350 V for 350 Vh, 500 V for 500 Vh, 1000 V for 1000 Vh, 3000 V for 3000 Vh, and 5000 V for 65000 Vh, with a 50 µA/strip maximum-setting at 20°C. The strips were equilibrated twice (reduced and alkylated) for 15 min in 15 mL equilibration solution (0.05 M Tris-HCl, pH 8.8, 6.0 M urea, 30% [v/v] glycerol, and 2% [w/v] SDS), first with the addition of 1% DTT, and finally with 2.5% iodoacetamide. After equilibration, the strips and the molecular-mass marker proteins were attached to the 12% SDS-PAGE, 1 mm thickness, using 1% agarose. The electrophoretic conditions were as follows: 5 W/gel for 30 min and 17 W/gel until the end of the running. The gels were stained for 24-48 h using Coomassie Blue R350 (PhastGel Blue R - GE Healthcare) under gentle agitation, and then were rinsed with ultrapure water (MilliQ grade). The destaining of the spots was carried out in the solution of 40% (v/v) ethanol and 10% (v/v) acetic acid. The gels were digitalized in transparency mode, and analyzed with Image Master-2D Platinum version 6.0 software (GE Healthcare, USA). The 2-DE gels were performed in triplicates for the protein extracts of leptospires cultured from kidney and liver of infected hamsters.

The samples were analyzed by MALDI-TOF (Matrix Assisted Laser Desorption Ionization-Time of Flight) MS, using α-cyano-4-hydroxycinnamic acid as the matrix on an Ettan MALDI-TOF/Pro instrument (Amersham Biosciences, USA). Briefly, the gel spots were isolated and dehydrated with acetonitrile (ACN). Then, the treated spots were rehydrated with 40 ng/μL trypsin solution in 50 mM ammonium acetate, and the reaction was carried out overnight at 37°C. The tryptic peptides were extracted from the gel spots using an H₂O/ACN/TFA (1:1:0.05) solution and this procedure was repeated 3X. The peptide solutions were concentrated and desalted with ZipTip C-18 pipette microcolumns (Millipore) prior to analysis. The samples were pre-mixed with the matrix solution (1:1 [v/v]) and approximately 0.5 μL was dropped on the sample loader and let to dry over the bench. The MS was performed on the reflectron mode and internal calibration (trypsin autolysis peptide). External calibration was also carried out in case the trypsin peptides were not present or clear.

The MALDI-TOF obtained mass lists were analyzed by peptide mass fingerprinting (PMF) algorithms on the server MASCOT (http://www.matrixscience.com/cgi/search_form.pl?FORMVER=2&SEARCH=PMF) for matches with known protein sequences deposited on the public or proprietary databases. The searches were performed against MSDB non redundant and L. interrogans serovar Copenhageni databases without any MW or pI restrictions. One missed cleavage site per peptide was allowed, and an error tolerance of ±0.2 Da on the mass measurement was used for the search and modifications such as, carbamidomethylation of cysteines and random oxidation of methionines were taken into account. The positively identified protein matches were considered to be those that were statistically significant by the probability-based MOUSE scoring of the server MASCOT (score>42), which takes into account a value of p<0.05 for the random matches.

The indentified L. interrogans proteins were further analyzed by the bioinformatics tools. The web-based tool P-CLASSIFIER [21] that considers five primary localization sites of Gram-negative bacteria (cytoplasm, extracellular space, inner membrane, outer membrane, and periplasm) was used to predict the subcellular localization of the identified proteins (http://protein.bii.a-star.edu.sg/localization/gram-negative/), taking into account the greatest probability. The LipoP server [22] (http://www.cbs.dtu.dk/services/LipoP/) was used to predict the lipoproteins and discriminate between lipoprotein signal peptides (SpII), other signal peptides (SpI), and n-terminal membrane helices (TMH) that might be present in the proteins. This program is specific for Gram-negative bacteria and is based on the lipobox consensus sequence. For the purpose of analysis of functional category of the identified proteins, the genomic database of L. interrogans serovar Copenhageni (http://aeg.lbi.ic.unicamp.br/world/lic/), (http://cmr.jcvi.org/cgi-bin/CMR/GenomePage.cgi?org=ntli03) and the Entrez Gene of National Center for Biotechnology Information (NCBI) (http://www.ncbi.nlm.nih.gov/sites/entrez?db=gene) were used.

All animal studies were approved by the Ethics Committee of the Faculdade de Medicina Veterinaria, Universidade de São Paulo, Sao Paulo, Brazil.

Protein extracts were fractioned by 2-DE gels, focalized in the pH range of 3-10 and molecular mass separated in 12% acrylamide gels. Representative Coomassie Blue R stained gels, from three independent experiments, of protein extracts of Leptospira cultured from kidney and liver are shown in Fig. (1), where some identified proteins are illustrated. The protein spots were observed to be mainly distributed between de pIs 5-7 and were almost absent in the basic or acid extreme regions of the strips (Fig. 1). Our data are in agreement with the published studies that have shown leptospiral-protein focalization between pIs 4 and 7 [19], including proteins of whole Leptospira [2].

Fig. (1). Representative 2-DE gels of L. interrogans serovar Pomona protein samples. Whole-cell protein extracts of virulent Leptospira cultured from kidney (A) and liver (B) were separated by 2-DE, with pH 3-10 and molecular mass of 10-70 kDa. Protein spots were excised for identification by MALDI-TOF MS. Examples of known leptospiral proteins, including some characterized lipoproteins (**) and some novel hypothetical proteins (*) identified in this work are depicted.

Fig. (2). Functional classification of the proteins of L. interrogans serovar Pomona. Distribution of proteins identified by 2-DE/MALDI-TOF MS of virulent leptospires, cultured from kidney and liver of infected animals. Categorization was based on the genome annotation of L. interrogans serovar Copenhageni genome (http://aeg.lbi.ic.unicamp.br/world/lic/).

All the isolated and visible major protein spots were excised from each gel and were separately processed for MS. After MALDI-TOF MS, the obtained PMF mass lists were searched and compared against the L. interrogans serovar Copenhageni genomic database for the identification of the respective leptospiral proteins, and the NCBI for the general protein match. From the 6 2-DE gels obtained, a total 895 spots were analyzed, resulting in positive identification of 286 spots (31.9 %) as leptospiral proteins. The 108 recognized proteins are representative of all functional categories used in genome annotation, and their respective percentages are depicted in Fig. (2). The proteins were assigned in: (a) biosynthesis of small molecules, (b) cellular processes, (c) central intermediary metabolism, (d) degradation, (e) DNA/RNA metabolism, (f) energy metabolism and carbon, (g) hypothetical, (h) membrane components, (i) other cell structure components, (j) pathogenicity, virulence, and adaptation, (k) protein metabolism, (l) regulatory functions, and (m) undefined category. The distribution in the functional categories of the 108 proteins identified in L. interrogans showed that the allocation of the proteins is similar to all categories, with the highest number of proteins being presented in the hypothetical category (Fig. 2). This data is very important, because it validates the expression of these leptospiral proteins with unknown function. A complete list with general analysis of identified proteins of L. interrogans serovar Pomona cultured from kidney and liver of infected animals is presented in Table 1 (supporting information). Sequences of matched peptides of each identified protein are shown in Table 2, supporting information.

The identification of the surface proteins is an important theme of the current leptospiral research. Owing to their location, the leptospiral surface proteins are expected to be relevant in the host-cells interactions in the context of pathogenesis and their potential ability to promote heterologous immunity. Predicted membrane proteins, according to P-CLASSIFIER program, with or without signal peptidase I or II cleavage sites, as predicted by LipoP program, are compiled in Table 1 (supporting information). The analysis of the subcellular localization of the proteins, as predicted by P-CLASSIFIER [21] revealed that the identified proteins are allocated in all the bacterial cell compartments: cytoplasm (78), periplasm (8), inner membrane (4), outer membrane (14), and extracellular (3) (Table 1, supporting information).

The 108 identified proteins in L. interrogans serovar Pomona were also observed in L. interrogans serovars Lai and Copenhageni genomes [7, 8, 23], 99 of them were conserved in both the strains of L. borgpetersenii serovar Hardjo and 80 were present in L. biflexa serovar Patoc sequenced strains [9, 10]. Protein conservation among these later strains is depicted in Table 3 (supporting information) that shows (in gray) that 8 of the hypothetical proteins validated in this work are absent in the saprophytic non-pathogenic L. biflexa.