Improved delivery of the OVA-CD4 peptide to T helper cells by polymeric surface display on Salmonella

Background Autotransporter proteins represent a treasure trove for molecular engineers who modify Gram-negative bacteria for the export or secretion of foreign proteins across two membrane barriers. A particularly promising direction is the development of autotransporters as antigen display or secretion systems. Immunologists have been using ovalbumin as a reporter antigen for years and have developed sophisticated tools to detect specific T cells that respond to ovalbumin. Although ovalbumin-expressing bacteria are being used to trace T cell responses to colonizing or invading pathogens, current constructs for ovalbumin presentation have not been optimized. Results The activation of T helper cells in response to ovalbumin was improved by displaying the OVA-CD4 reporter epitope as a multimer on the surface of Salmonella and fused to the autotransporter MisL. Expression was optimized by including tandem in vivo promoters and two post-segregational killing systems for plasmid stabilization. Conclusions The use of an autotransporter protein to present relevant epitope repeats on the surface of bacteria, combined with additional techniques favoring stable and efficient in vivo transcription, optimizes antigen presentation to T cells. The technique of multimeric epitope surface display should also benefit the development of new Salmonella or other enterobacterial vaccines.


Background
Escherichia coli is the prototype bacterium used for the production of desired foreign proteins in vitro. A particularly interesting approach has been the export of foreign proteins on the bacterial surface of E. coli by taking advantage of the specific inherent properties of autotransporter proteins. These proteins, which represent a category of type V secretion systems of Gram-negative bacteria, are exported to the periplasm by the sec machinery and assembled into the outer membrane by the Bam and Tam proteins [1][2][3]. The carboxy-terminal end of an autotransporter protein forms a beta-barrel structure with a central pore originally thought to channel the amino-terminal side of the protein or passenger domain to the bacterial surface. Newer models propose that Bam participates in making the pore, implying that the term "autotransporter" is a misnomer [3]. For some autotransporter proteins, the translocated segment is cleaved and released from the bacteria, which can be useful for direct purification procedures from spent medium. Some cleaved translocators remain surface associated by non-covalent bond. Whether secreted or on the bacterial surface, translocators can act as adhesins, mediators of biofilm formation, enzymes for intercellular spreading, cytotoxins or modulators of immune responses [3]. Early studies recognized the use of autotransporters as export machineries for foreign proteins and antigens. Translocator domains of a variety of autotransporter proteins have been modified with in-frame fusions of recombinant proteins for display on the bacterial surface or for delivery to the spent medium [4,5]. Among these autotransporters, the Salmonella MisL autotransporter adhesin [6] has previously been engineered to express four copies of the Plasmodium falciparum immunodominant epitope (NANP) on the surface of Salmonella enterica [7]. MisL was able to express even more copies of NANP on E. coli and to release them in the medium after the addition of a cleavage site for the surface protease OmpT [8]. The epitopes were shown to induce immune responses. Similarly, viral and parasitic protein epitopes were successfully expressed by recombinant MisL on Salmonella vaccine strains and induced epitope-specific immune responses with protective properties [9,10].
Here, we took advantage of the transport property of MisL to export a fused antigenic model peptide on the surface of Salmonella as an activation signal for specific cellular immune responses. The generation of a T helper cell 1 (Th1) immune response is crucial for successful control of the facultative intracellular pathogen Salmonella and several studies have highlighted the importance of the CD4+ T cell response during infection [11]. The ability to track antigen specific T cells is important for understanding the initiation and maintenance of T cell responses during various infections and in response to vaccines [12]. Since endogenous T cells specific for any given antigen are present in small numbers, it makes the tracking of such cells difficult during the early phases of an immune response before clonal expansion has occurred. Adoptive transfer models using TCR-transgenic T cells specific for model antigens such as ovalbumin have thus provided a vital tool for tracking antigen specific T cell responses [13][14][15]. A key aspect of such studies is to obtain efficient and stable expression of a foreign antigen by a genetically engineered pathogen. The expression systems in Salmonella using full-length ovalbumin constructs available thus far have resulted in suboptimal responses in vivo [16][17][18]. The current investigation highlights the use of a novel construct based on the polymeric surface display of an ovalbumin reporter epitope to amplify the signal for the improved activation and detection of cognate CD4+ T cells.

Results
Construction, expression and export of the MisL-OVA-CD4 fusion proteins The Salmonella autotransporter protein MisL was used to display a specific MHC class II ovalbumin peptide (OVA-CD4) on the surface of a Salmonella vaccine vector. The OVA-CD4 peptide was fused through genetic engineering to the N-terminus of the translocator domain of MisL. The construct was based on the pnirLTB-MisL plasmid [7] that encodes a truncated MisL protein with the signal sequence of the heat-labile enterotoxin of E. coli in frame. This truncated MisL lacks the aminoterminal 423 amino acids of the mature MisL. The expressed MisL protein included an additional 210 amino acids of its alpha domain to avoid potential protein misfolding in the periplasm and ensure correct conformation in the outer membrane for transit of the passenger domain to the bacterial surface and display of its N-terminus [7,8]. The plasmid used the nirB promoter that is activated under the anaerobic conditions of the intestinal environment and induces improved systemic immune responses towards foreign antigen expressed by orally administered Salmonella vaccine strains [19][20][21]. This vector is not a high copy number plasmid as antibody responses against antigens delivered by Salmonella are improved when expressed from low to medium copy number plasmids that minimize metabolic stress upon the carrier strain [22,23]. Previous addition of the spiC promoter, which is activated in antigen-presenting cells, was used successfully to express the MisL protein or the 987P fimbriae of E. coli as viral epitope carriers on the surface of a Salmonella vaccine vector. Oral immunization experiments in mice also suggested improved in vivo expression by showing better humoral immune responses against the foreign antigens [10,24]. Thus, we prepared a similar promoter tandem construct by inserting the spiC promoter between the nirB promoter and the Shine Delgarno site for MisL transcription ( Figure 1). Expression of the OVA-CD4-MisL fusion protein was evaluated with Salmonella strain SL7207 carrying either pZS1202 (nirBp) or pZS1204 (nirBp and spiCp). Using an antibody specific for the OVA-CD4 epitope, the fusion protein of expected size (approximately 60 kDa) was detected by western blot analysis for both constructs with approximately three times more protein being expressed when both promoters were used, as determined by densitometry ( Figure 2A). The OVA-CD4 specific antibody was confirmed to react with ovalbumin but not with Salmonella-expressing MisL alone. The detection of the fusion protein in Salmonella grown under aerobic conditions suggests some promoter leakiness, although the high affinity of the anti-OVA-CD4 antibody used in these assays might render detection very sensitive. Previous data indicated that magnesium activated fimbrial subunit gene expression from the spiC promoter in a phoP mutant [24]. However, using the same mutant strain, a significant effect due to magnesium was not detectable for MisL, possibly due to the fact that MisL is a much larger protein than the previously studied fimbrial subunit. In order to demonstrate that the OVA-CD4-MisL fusion protein was crossing the cytoplasmic membrane and behaved like the native MisL protein, outer membranes prepared from strain SL7207 carrying the plasmid with the misL fusion gene were analyzed by western blotting. The OVA-CD4-MisL fusion protein was detected in the outer membrane preparation using the ovalbumin-peptide specific antibody ( Figure 2B), showing that the fusion protein was not affected in its export and localization properties. Moreover, fused OVA-CD4 peptide was susceptible to proteinase K degradation, characterizing its accessibility on either one or both outer membrane surfaces.

Stabilized plasmids
To ensure that attenuated Salmonella strains would keep the OVA-CD4-MisL plasmid construct in the absence of antibiotic selection after administration to mammalian hosts, two post-segregational killing systems from the F plasmid of E. coli, the fimAB and ccdAB genes were engineered into the plasmid (Figure 1). Efficiency of plasmid maintenance was investigated by growing Salmonella SL7207 with or without the stabilized plasmid in antibiotic-free media for approximately 530 generations. The plasmid-bearing rate was calculated by dividing the number of bacteria recovered from  antibiotic plates by the number of bacteria recovered from antibiotic-free plates. As shown in Figure 3, plasmid pZS1204 with no stabilization system was gradually lost during after 20 daily passages, however plasmid pZS1205 was kept in essentially all the bacteria. A derivative of the latter plasmid, pZS1205-4, which expresses four copies of the OVA-CD4 was 100% maintained for one week and in nearly half of the bacteria for 26 days. In contrast, the high copy number plasmid pUC18-OVA that constitutively expressed the full-length ovalbumin protein was lost in five days. These results confirmed that high copy number plasmid constructs have the undesirable effect of being unstable in the absence of antibiotic selection. More importantly, the data demonstrated that the fimAB and ccdAB genes have an efficient plasmid stabilization property that prolongs the life span of foreign antigenexpressing Salmonella.

Surface exposure of the OVA-CD4 epitope peptide on Salmonella
To determine whether the OVA-CD4 portion of the fusion protein was located on the periplasmic or extracellular face of the outer membrane, intact Salmonella expressing the fusion protein were subjected to treatments with various concentrations of proteinase K. Presence of the OVA-CD4 peptide was then determined by western blot analysis with the OVA-CD4 specific antibodies ( Figure 4A). Parallel western blots were probed for the periplasmic β-lactamase with an anti-ampicillinase antibody to evaluate concentrations of proteinase K that would be effective on the bacterial surface without permeating into the periplasmic compartment ( Figure 4B). Proteinase K at a concentration of 11 μg/ml completely degraded the OVA-CD4 peptide without affecting the periplasmic β-lactamase, indicating that the latter peptide was exposed on the bacterial surface. A comparable result was obtained with higher concentrations of proteinase K, albeit a partial proteolytic effect on the β-lactamase suggested some permeation of the protease.
Construction and display of oligomeric OVA-CD4 epitopes on the surface of Salmonella Having engineered OVA-CD4 encoded DNA to be flanked by restriction sites with matching 5′ extensions (XbaI and NheI) on plasmid pZS1205, it was possible to construct new homocomposite plasmids that add OVA-CD4 peptides in tandem to the amino-terminal end of MisL. We prepared plasmids capable of expressing MisL fusion proteins carrying dimeric and tetrameric OVA-CD4 epitopes and compared their levels of expression and surface display. Western blot analysis of these constructs showed that each addition of OVA-CD4 epitope copies to MisL significantly increased expression of the fusion protein ( Figure 4C). Densitometry of the steadystate detectable proteins indicated that the (OVA-CD4) 2 -MisL dimer and (OVA-CD4) 4 -MisL tetramer constructs had approximately 2.7 and 4.8 times more OVA-CD4 epitopes, respectively, than the (OVA-CD4) 1 -MisL monomeric epitope fusion protein. We speculated that the lower molecular bands seen with the OVA tetramer were due to degradation of surface epitopes by the outer membrane protease PgtE [24], therefore we repeated the experiment with an insertion mutant of Salmonella SL7207 deficient for PgtE expression. Densitometry revealed greatly reduced degradation products for the tetramer construct and 1.7 and 4.4 times more OVA-CD4 epitopes were expressed with the dimer and tetramer MisL fusion proteins compared to the monomer construct ( Figure 4D). All the ovalbumin epitopes expressed by Salmonella were accessible to the proteinase K proteolytic activity, consistent with their surfaceexposure. The increased numbers of smaller products for the tetramer construct in the absence of proteinase K pointed towards bacterial auto-degradation processes. Surface display of the OVA-CD4 epitope was visually confirmed by fluorescence microscopy. Using the OVA-CD4 specific antibody with a FITC-labeled secondary antibody to label the surface of Salmonella SL7207 expressing one, two or four copies of OVA-CD4 fused to MisL, the intensity of the fluorescence increased accordingly, as determined by the number of additional epitope Figure 3 Plasmid segregational stability. Salmonella SL7207 with the stabilized pZS1205 plasmid (☐) or the nonstabilized pZS1204 plasmid (￭), each expressing (OVA-CD4)-MisL, with the stabilized pZS1205-4 (△) that expresses four copies of OVA-CD4 fused to MisL, or pUC18-OVA (▲) that expresses full-length ovalbumin were grown in LB broth at 37°C without antibiotics. Serial passages were conducted daily by diluting the overnight cultures into fresh media (10 −3 dilutions for the first week and 10 −5 dilutions for later). Plasmid stability was estimated by calculating the numbers of ampicillin-resistant CFUs divided by the total number of CFUs and multiplying the number by 100 to obtain a percentage of bacteria containing the plasmid. copies ( Figure 5). Fluorescence intensities per bacterial cells using the NIH ImageJ software suggested that the OVA-CD4 dimer and tetramer fusion constructs displayed respectively 2.1 and 3.4 times more epitope as compared to the monomer construct. Fluorescence intensities with the PgtE mutant containing the same plasmids showed similar increases in epitope expression, with 2.4 and 5 times more epitope displayed by the dimer and tetramer constructs respectively (data not shown). The incremental levels of fluorescence and western blot bands confirmed higher expression and surface exposure of the OVA-CD4 peptide with increased numbers of tandem repeats in the constructs.

-MisL
The relative efficiency of OVA-CD4 epitope expression harboring pZS1205-4, which encodes the peptide tetramer fused to MisL, or pUC18-OVA, which encodes the fulllength protein [25], was compared by western blot analysis with the anti-OVA-CD4 antibody ( Figure 6). E. coli was used as host to compare ovalbumin expression under  uninduced and induced conditions. The signal of the MisL fusion construct was 2.6 or 5.6 times stronger than the one from the full-length protein with or without IPTGinduction, respectively.

In vitro antigen presentation
The efficacy of the Salmonella SL7207pZS1205-4 [(OVA-CD4) 4 -MisL] constructs in stimulating ovalbumin specific CD4 T cell responses was tested using in vitro antigen presentation assays with OTII TCR transgenic system. The responses obtained with the current construct were compared to a known full-length ovalbumin expressing system (pUC18-OVA) that has been used extensively in the field [25,26], using the same Salmonella strain. The OTII T cells were labeled with CFSE in these in vitro assays and the dilution of CFSE was used to measure T cell proliferation. When T cells were co-cultured with APCs pulsed with parental strains of Salmonella that do not express ovalbumin epitopes, they did not undergo proliferation as evident from the lack of CFSE dilution ( Figure 7A). APCs pulsed with Salmonella expressing the full length ovalbumin or (OVA-CD4) 4 -MisL resulted in activation of T cells and their proliferation. The OTII T cell expansion was significantly higher in response to (OVA-CD4) 4 -MisL as compared to full-length ovalbumin, when each was expressed by the same Salmonella strain ( Figure 7A and B).

In vivo antigen presentation
To test the efficacy of the various bacterial constructs in vivo, an adoptive transfer model was used, as described in the Methods section. The OTII T cells expressed the congenic marker CD45.1 and could thus be tracked in the recipient mice (CD45.2). Seven days after oral infection with the different bacterial strains, the mice were euthanized and lymphocytes were isolated from the spleen and mesenteric lymph nodes and analyzed by flow cytometry. These studies showed that there was an increase in the frequency of the OTII T cells during infection with the ovalbumin-expressing Salmonella SL7207 compared to the controls that received PBS or Salmonella alone ( Figure 8A). The frequency of OTII T cells (CD45.1 CD4+) in the mice that were administered the strain expressing (OVA-CD4) 4 was higher as compared to the mice infected with the strain that expressed full-length ovalbumin. No deletion of OVA-CD4 repeats was detected during the time of the in vivo experiment (data not shown). Based on the analysis of surface marker expression for activation, such as up regulation of CD44 and CD25, and down regulation of CD62L, there was a higher frequency of activated OTII T cells in the mice infected with the (OVA-CD4) 4 -expressing Salmonella ( Figure 8B). These differences were also  apparent in the total numbers of OTII T cells in the mesenteric lymph nodes ( Figure 8C) and spleens (data not shown), with significantly higher numbers of OTII T cells present in the mice infected with the (OVA-CD4) 4 -expressing Salmonella compared to the mice infected with the full-length ovalbumin-expressing Salmonella. Together these data showed that the newly developed construct permits robust expression of the model ovalbumin epitope both in vitro and in vivo and can be used as reliable model to track antigen specific T cell responses.

Discussion
Our studies demonstrated that the MisL autotransporter protein was capable of displaying the OVA-CD4 epitope on the surface of Salmonella when the epitope was grafted onto the amino-terminal side of MisL. Moreover, MisL fusion proteins engineered to carry the epitope in tandem copies increased the efficiency of surface display. Based on our previous work on foreign antigen expression by Salmonella, we designed a prototype plasmid to optimize in vivo antigen presentation. For this, a tandem in vivo inducible promoters was inserted upstream of the OVA-CD4-MisL open reading frame in a low to medium copy number plasmid and two sets of stabilization systems were introduced into the construct. When compared to a traditional Salmonella construct expressing full-length ovalbumin in the cytoplasm from a multicopy number plasmid with a strong constitutive promoter, our prototype construct expressing four OVA-CD4 epitopes was significantly more efficient at presentation for activating CD4 T cells, both in vitro and in vivo.
A variety of microbes [14,27,28], including wild type or attenuated entero-invasive Salmonella [16,25,29], have been genetically modified to express ovalbumin with its well-characterized CD4 and CD8 T cell epitopes as a reporter system to study cellular immune responses in hosts. Many of these studies did not attempt to optimize the expression level and bacterial delivery mechanism for the most efficient immune recognition of the relevant epitopes. For this study, we designed and tested an improved ovalbumin reporter construct based on previous studies with Salmonella vaccine vectors that have established general principles for the improvement of immune responses. First, surface display of a foreign epitope was reported to induce a better immune response than when the same epitope remained in the bacterial cytoplasm [20]. A recent study suggested that location on the bacterial surface was even more important for immunity than high abundance or immunodominance since Salmonella mainly survive in antigen-presenting cells, rendering surface antigens more accessible to the cell for processing and presentation [30]. Second, surface-exposed foreign epitopes presented as multimers induced better antibody responses than the same epitopes presented as monomers on a Salmonella vaccine vector [10]. Because it is capable of expressing antigens on a Salmonella surface, the MisL protein has the advantage of tolerating large fragments of proteins in a fusion construct. This property was used to incrementally add OVA-CD4 epitopes to MisL and demonstrated a corresponding stepwise increase of epitope expression on the surface of Salmonella. The addition of tandem epitope copies did not affect bacterial growth. Although surface expression of larger amounts of foreign epitopes can induce some degradation by the PgtE protease of Salmonella, as shown previously [24] and here with four copies of the OVA-CD4 epitope, PgtE did not significantly affect steady-state expression of the surface exposed epitope. A third generic rule to improve antigen delivery for better immune responses is the avoidance of high copy number plasmids that express large foreign antigen from constitutive promoters [23,31]. Such vector constructs are metabolic stressors for the bacteria and are rapidly eliminated or mutated, resulting in reduced levels of immune response [22]. In general, metabolic stress can be minimized by both the use of a low copy number plasmid and the expression of the smallest protein fragment needed for an experimental task, such as a major immunogenic epitope fused to a Salmonella protein. Here, we demonstrated that the multicopy number plasmid pUC18-OVA that expresses the full-length ovalbumin was lost in vitro after only a few passages in the absence of antibiotic selection, suggesting that it is not an optimal construct. In contrast, the low copy number pZS1204 was far more stable than pUC18-OVA. Safe remedies to stability issues for the expression of foreign proteins have been the use of chromosomal integration or single-copy number plasmids together with strong constitutive or in vivo inducible promoters [32][33][34]. As an alternative to stabilize protein expression, one can borrow plasmid post-segregational killing genes. We used the latter approach by inserting a cassette comprising two of these set of genes that we have previously used with success [10] to obtain pZS1205 and confirmed significantly improved plasmid stability over 500 generations. In addition, approaches shown to augment the expression of ovalbumin or other foreign antigens in animal models were based on the engineering of in vivo-activated promoters [35][36][37]. For this study, we took advantage of a tandem promoter system that we had previously shown to express antigen efficiently both in vitro and in vivo [10,24].
The improved efficiency of antigen presentation with the engineered Salmonella displaying stably several copies of the OVA-CD4 epitope on its surface was demonstrated here by the enhanced CD4+ T cell multiplication both in vitro and in vivo, using a classical bacterial fulllength ovalbumin construct in the same bacterial background for comparison. Although we did not evaluate the individual effect of each construction step on this immune response, we used a systematic approach based on an extensive body of literature and our previous studies to engineer an efficient ovalbumin reporter system. Among the many autotransporter proteins that are capable of presenting heterologous surface antigens or secreted proteins, MisL has already demonstrated its adaptability as an antigen delivery system in the context of experimental Salmonella vaccines [7][8][9][10]38]. Some of these studies showed that MisL was tolerant to short epitope multimers. Here, we confirmed multimer display with a longer epitope fused to MisL and demonstrated efficient activation of CD4+ T cells. The identification of an increasingly larger panoply of autotransporter proteins capable of exporting foreign proteins might help to partially bypass display or secretion limitations due to the nature, size and number of epitopes added [4,39]. Although the construct described here was developed to create an improved ovalbumin reporter system for the study of CD4+ T cell responses to Salmonella, the concept of epitope presentation in the form of tandemly repeated homopolymers might be applicable for other purposes, including their cytoplasmic delivery by a bacterial injection apparatus or invasive mechanism for a CD8+ T response [10,40,41]. Polymerization of protective epitopes should also benefit vaccine development.

Preparation of outer membrane proteins
Outer membrane fractions were prepared from spheroplasts, as described previously [46]. Bacteria from 10 ml overnight cultures (approximately 10 10 CFU) were pelleted by centrifugation and suspended in 90 μl of 30 mM Tris-HCl (pH 8.0) with 20% sucrose. 10 μl of lysozyme (1 mg/ml in 0.1 M EDTA) was added and the bacteria were incubated for 30 min on ice. The obtained spheroplasts were stabilized with MgCl 2 (20 μM final concentration) and centrifuged (16,000 × g, 2 min). The periplasmic proteins were removed with the supernatant and the spheroplasts were resuspended in 100 μl of a 10 mM Tris-HCl (pH8.0), 100 mM NaCl, 10 mM MgCl 2 solution containing 1 μg of DNase per ml, and lysed by sonication (two times 1 min with a Cup Horn accessory at amplitude output 10, using a model XL2020 sonicator; Heat System, Farmingdale, N.Y.). Cytoplasmic membrane proteins were solubilized by incubating the membranes for 20 min at room temperature with N-

Indirect immunofluorescence assay (IFA)
For IFA, the bacterial strains were spread on glass slides, fixed with cold methanol and incubated with rabbit antiovalbumin antibody. The slides were washed three times with PBS-T, incubated with FITC conjugated goat antirabbit-IgG and washed again with PBS-T. Dried slides were mounted with 10 μl antifade reagent with DAPI, overlaid with glass coverslip, and images were acquired on a Leica TCS SP5 inverted confocal microscope with a 63× (1.2 NA) water immersion lens. Antibody staining was visualized with 488 nm excitation from an Argon laser and DAPI was excited with a 405 nm pulsed diode laser. Relative fluorescence intensities were calculated from measurements with the NIH ImageJ software.

Plasmid stability
To determine the persistence of a plasmid in dividing cells, bacteria were passaged in LB broth without antibiotics for 26 days (approximately 530 generations) using daily 10 −3 dilutions of overnight cultures for the first week and 10 −5 dilutions for the remainder. Plasmid stability was estimated by determining at several passage points the numbers of antibiotic-resistant live bacteria containing the plasmid (colony-forming units or CFUs on ampicillin-containing LB agar plates) divided by the total number of live bacteria (CFUs on LB agar plates). This number was multiplied by 100 to obtain a percentage of bacteria containing the plasmid [10].

Mice
OTII TCR transgenic mice on a CD45.1 background were maintained in a specific pathogen-free (SPF) facility in the Department of Pathobiology at the University of Pennsylvania in conformance with institutional guidelines for animal care. C57BL/6 mice were purchased from Taconic farms and housed within the SPF facility. All animal studies were carried out in compliance with the guidelines of the Institutional Animal Care and Use Committee of the University of Pennsylvania.
OTII T cell isolation CD4+ T cells were enriched from lymphocytes isolated from the spleen and pooled peripheral lymph nodes of OTII TCR transgenic mice (Ovalbumin specific CD4+ TCR transgenic mice) using CD4+ MACs beads (Miltenyii Biotech). The cells were labeled with the cytoplasmic dye CFSE (carboxyfluorescein diacetate succinimidyl ester, Molecular Probes, Eugene, OR), to track proliferation and used for the in vitro assays or transferred into mice for testing in vivo efficacy of the bacterial constructs.

In vitro antigen presentation assay
Antigen-presenting cells were isolated from spleens of WT mice; CD11C + cells were enriched from RBC lysed splenocytes using CD11c + MACs beads (Miltenyii Biotech). The purified APCs were resuspended in complete RPMI (10% FBS) and plated in 24 well plates (1×10 5 cells/well). The APCs were pulsed with the various bacterial constructs at a multiplicity of infection (MOI) of 10 for 2 hours, following which the cells were washed to remove the extracellular bacteria, treated with gentamicin and used for the co-culture studies. The ability of the pulsed APCs to process and present the ovalbumin derived from the bacteria on MHC II molecules for T cell activation, was determined by co-culturing them with CFSE labeled OTII T cells (5×10 5 cells/well) for a period of 4 days, as described previously [14]. At the end of the culture period, the cells were counted and stained for flow cytometric analysis.
In vivo antigen presentation assay 2×10 6 purified OT1I cells (CD45.1) were adoptively transferred (intravenously) into B6 (CD45.2) recipient mice and the mice were infected orally 24 hours later. For infections, log phase cultures of the various bacterial strains were washed and re-suspended in PBS and administered by oral gavage (10 9 bacteria/mouse). 10 minutes before infection mice were given 0.1 ml of 5% NaHCO 3 to neutralize the stomach acidity. The mice were sacrificed on day 7 post infection and the spleens and mesenteric lymph nodes were harvested for flow cytometric analysis.

Flow-cytometry
Lymphocytes were isolated from spleens and cervical lymph nodes by mechanical homogenization followed by lysis of RBCs (for spleens) using lysis buffer (0.846% NH 4 Cl). Freshly isolated cells were stained with the antibodies purchased from eBioscience (San Diego, CA) or BD Biosciences (San Jose, CA). The stained samples were run on a FACSCanto (BD, San Jose, CA) and results were analyzed using FlowJo software (TreeStar Inc., Ashland, OR).