|Home | About | Journals | Submit | Contact Us | Français|
Typhoid fever and non-typhoidal bacteremia caused by Salmonella remain critical human health problems. B cells are required for protective immunity to Salmonella but the mechanism of protection remains unclear. Here, we immunized WildType, B cell deficient, antibody-deficient and class-switched antibody-deficient mice with attenuated Salmonella and examined protection against secondary infection. As expected, WildType mice were protected and B cell deficient mice succumbed to secondary infection. Interestingly, mice with B cells but lacking secreted antibody or class-switched antibody had little deficiency in resistance to Salmonella infection. The susceptibility of B cell deficient mice correlated with marked reductions in CD4 T cell IFN-γ production after secondary infection. Together, these data suggest that the primary role of B cells in acquired immunity to Salmonella is via the development of protective T cell immunity.
Typhoid fever is caused by infection with Salmonella typhi and is a serious health concern worldwide, causing an estimated 21 million cases and 216,000 deaths per year (1). Non-typhoidal salmonellosis (NTS) is caused by other Salmonella serovars and is a growing problem among HIV-infected adults and HIV-negative children in Africa and Asia (2–5). Currently, there are two vaccines for typhoid fever that each provide limited protection but are not widely used in endemic areas (6, 7). There is no available vaccine for NTS, although numerous target antigens have recently been defined (8). The development of novel, effective vaccines for typhoid and NTS requires greater understanding of Salmonella-specific T and B cell responses (9).
Immunity to Salmonella is studied using a well-established murine model of typhoid, in which Salmonella typhimurium causes fatal disseminated disease in susceptible, Nramps mice (10, 11). After oral infection, Salmonella can gain access to the mammalian host by invading M cells in the Peyer’s Patches of the small intestine (10). Salmonella subsequently disseminate via the lymphatic system and replicate within phagocytic cells of the spleen, liver and bone marrow. Salmonella actively inhibit phagolysosomal fusion and infected macrophages require activation via IFN-γ to kill bacteria (12). Salmonella-specific Th1 cells that produce IFN-γ are essential for controlling bacterial growth, and mice lacking αβ CD4 T cells, Th1 cells, or IFN-γ eventually succumb to primary infection with attenuated bacteria (13, 14). Patients with primary genetic deficiencies in IL-12 or IFN-γ receptor signaling suffer from repeated disseminated Salmonella infections (15, 16). Thus, Th1 cells play an important role in mediating protective immunity in both human and murine Salmonellosis.
The resolution of primary Salmonella infection confers robust protective immunity against secondary challenge. CD4 T cells are essential for this acquired resistance and depletion of CD4 T cells eliminates the protective effect of vaccination with attenuated Salmonella (17). More surprisingly for an intra-macrophage infection, B cells are also essential for acquired immunity to Salmonella, and immunized B cell-deficient mice display enhanced susceptibility to secondary infection (18–20). However, the protective role of antibody in secondary immunity is somewhat controversial. Passive transfer of antibody is reported to be protective in some studies, while others have observed no protective effect (18, 19, 21). Furthermore, neither IgA, nor mucosal immunoglobulins are required for protective immunity to Salmonella (8, 22). B cells can contribute to protective immunity via antigen presentation to Salmonella-specific Th1 cells (18, 23), or as an important source of inflammatory cytokines during infection (24, 25). However, it remains unclear whether the contribution of B cells to protective immunity is largely mediated by antibody-dependent or antibody-independent mechanisms.
Here, we examined the role of B cells in protection against infection with virulent Salmonella using transgenic mouse strains that lack B cells, class-switched antibody, or antibody secretion and demonstrate that antibody production is largely dispensable for protection against secondary Salmonella infection. In contrast, B cells are required for optimal priming of Salmonella-specific Th1 cells that mediate bacterial clearance.
Balb/c (WildType) and JhD/Balb/c (B cell Deficient) mice (National Cancer Institute, Frederick, MD) were used at 6–12 weeks of age. Transgenic m+s IgM and mIgM use the B1–8 heavy chain, have a restricted BCR repertoire, were maintained on a JhD/Balb/c background (26) and were provided by Dr. Shlomchik (Yale University, New Haven, CT). Transgenic mice were intercrossed with JhD/Balb/c mice and used at 6–12 weeks of age. Homozygosity at the JHD locus was maintained by interbreeding with JhD mice and PCR screening of mIgM heavy chain was done using the following primers (Vh186.2 5′ 216 CTACTGGATGCACTGGGTGA and Vh186.2 3′ 459 TTGGCCCCAGTAGTCAAAGTA). All mice were housed in specific pathogen free conditions for breeding and experimentation.
Attenuated S. typhimurium BRD509 (ΔaroA/ΔaroD) and parental virulent strain SL1344 were grown overnight in Luria-Bertani broth and diluted in PBS after estimating bacterial counts by spectrophotometry. Mice were immunized IV with 5×105 BRD509 and challenged orally with 5×107 SL1344 after oral administration of 100μl 5% NaHCO3. Infection doses were confirmed by plating serial dilutions onto MacConkey agar plates. Any moribund infected mice were euthanized as stipulated in our IACUC protocol. Bacterial growth in vivo was calculated by plating serial dilutions of organ homogenates onto MacConkey agar and bacterial counts were determined after overnight incubation at 37°C.
Salmonella-specific CD4 and CD8 T cell responses were visualized, as previously described (27). Immunized mice were injected IV with 1×108 Heat-Killed Salmonella typhimurium (HKST) and spleens harvested three or five hours later. A single cell suspension was surface stained using FITC-, PE-, PE-Cy5-, PE-Cy7-, APC-, eF450-, AF700- and APCeF780-conjugated antibodies to CD3, CD4, CD8, Gr-1, CD11c, CD11b, F4/80, B220, and CD44 in Fc block (spent 24G2 supernatant, 2% rat serum, 2% mouse serum). Cells were fixed, permeabilized, and stained intracellularly using PE conjugated anti-IFN-γ. All staining reagents were purchased from BD Biosciences (San Jose, CA) or eBioscience (San Diego, CA). Samples were analyzed by flow cytometry using a FACSCanto and data analyzed using FlowJo Software (Tree Star).
Blood was collected by retro-orbital bleeding and sera prepared and stored at −20°C. Salmonella-specific IgM and IgG antibodies were measured by ELISA, as previously described (27).
Statistical analysis was performed using unpaired t tests (Prism 4, GraphPad Software, Inc., La Jolla, CA). Survival data was compared using a Log-rank (Mantel–Cox) test (Prism 4). Statistical differences between groups are highlighted with *, P < 0.05; **, P < 0.01; or ***, P < 0.001.
Defining protective immune responses to Salmonella infection is a prerequisite for development of new effective vaccines against typhoid and NTS (10). Although CD4 T cells are critical for protective immunity to Salmonella, the contribution of B cells has not been clearly defined. Salmonella-specific antibody production, inflammatory cytokine production, and direct antigen presentation to T cells have each been proposed as mechanisms to explain the protective role of B cells during secondary infection (18, 19, 23, 24, 28). We sought to investigate whether B cells provide secondary protective immunity against Salmonella primarily in an antibody dependent or independent manner. Given previous data showing that serum transfer can protect (19), but that neither IgA nor mucosal immunoglobulin is required (8), we hypothesized that systemic IgG is essential for secondary clearance of bacteria. To test this hypothesis, we examined immunity in B cell deficient mice (JhD), transgenic mice with B cells that cannot class switch or secrete antibody (membrane IgM- mIgM), and mice with B cells that cannot class switch but are able to secrete IgM, (membrane + secretory IgM- m+s IgM) (26). Although the mIgM and m+s IgM transgenic mice have a restricted BCR repertoire, they do not have significant deviations in naive B cell and T cell subsets (Supplemental Fig. 1A–1D and (29)). All four strains (WildType, B cell deficient, mIgM, and m+s IgM mice) survived vaccination with attenuated S. typhimurium and had largely cleared bacteria from the spleen 44 days later (Supplemental Fig. 1E). This confirmed previous reports that resolution of primary infection with attenuated Salmonella does not require B cells (18, 19).
To examine acquired immunity to secondary Salmonella infection, naive and immunized mice from all four strains were challenged orally with virulent S. typhimurium (Fig. 1A). Regardless of the B cell compartment, all naïve mice succumbed to primary infection with virulent Salmonella at a similar rate (Fig. 1A). In contrast, immunized WildType mice resisted secondary infection with virulent Salmonella, while B cell deficient mice succumbed to secondary challenge (Fig. 1A). Surprisingly, m+s IgM mice that lack class-switched antibody also survived secondary infection with Salmonella, demonstrating a similar degree of protective immunity to wild-type mice (Fig. 1A). Furthermore, most mIgM mice that lack all secreted antibodies were resistant to secondary Salmonella infection. However, approximately 25% of these mice eventually died from infection, and this was statistically different from the survival of WildType and B cell deficient mice (Fig. 1A). Together, these data confirm that B cells are essential for resistance to secondary infection with virulent Salmonella, and surprisingly demonstrate that production of class-switched antibodies is not required for protective immunity. Additionally, although secreted IgM antibodies may contribute to secondary protection, the mechanism of B cell-mediated protection against secondary Salmonella infection is largely antibody-independent in this vaccination and rechallenge model.
Given these findings, it was important to confirm the absence of circulating Salmonella-specific antibody in each B cell deficient strain examined above. Serum was collected nine days after secondary infection, and Salmonella-specific antibody responses were examined. Nine days after secondary infection, both WildType mice and IgM Ab Only (m+s IgM) mice had modest levels of circulating Salmonella-specific IgM (Fig. 1B), but only WildType mice developed Salmonella-specific IgG (Fig. 1B). These results confirm that only WildType mice produced a class-switched antibody response to Salmonella, but that IgM Ab Only mice developed low Salmonella-specific IgM responses during secondary infection.
Given the fact that mice lacking all antibodies had a 25% death rate following virulent challenge, it seemed likely that bacterial clearance was hindered at late time points in these mice, perhaps because IgM is required for clearance from a particularly persistent anatomical site such as the mesenteric lymph nodes (30). Thus, we examined the rate of bacterial clearance in immunized mice lacking B cells, class-switched antibody or all antibody. Three days after secondary infection, WildType mice had lower bacterial loads in the spleen than B cell deficient mice (Fig. 2A), demonstrating that B cells are required for rapid secondary clearance of bacteria. At this early time point, no significant differences were apparent between antibody deficient strains and B cell deficient mice, but antibody deficient mice had a trend towards lower CFUs in the spleen (Fig. 2A). No significant differences were detected in liver CFUs at this same early time point (Fig. 2B). Nine days after secondary infection, mice lacking B cells had much higher bacterial loads in both the spleen and liver compared to WildType mice (Fig. 2A and B). In marked contrast, mIgM and m+s IgM mice had lower CFUs in both spleen and liver (Fig. 2A and B). Together these data demonstrate that the rate of bacterial clearance during secondary infection is largely unaffected by the absence of antibody, despite a requirement for B cells. This finding contrasts with prior studies that showed a protective effect of serum transfer (19, 21). However, these studies were not designed to test an antibody-independent role of B cells and both described protection against low dose challenge. Our finding has broad implications since the measurement of circulating immunoglobulin is often used as an indicator of vaccine efficacy.
It is clear from previous work that secretion of IFN-γ by Th1 cells is critical for the resolution of Salmonella infections (14, 31). We confirmed this by depleting CD4 and CD8 T cells in immunized WildType mice and challenging with a virulent strain of Salmonella. T cell depletion caused a significant increase in bacterial loads during secondary infection (Supplemental Fig. 2A). It has been suggested that antibody can enhance T cell responses to Salmonella by allowing bacterial uptake via Fc receptors on dendritic cells (32). B cells also can present antigen and secrete cytokines that shape the development of protective T cell responses. Thus, we examined the effect of B cell or antibody deficiency on the generation of Salmonella-specific Th1 cells.
WildType, B cell deficient, m+s IgM and mIgM mice were immunized with attenuated Salmonella and Salmonella-specific CD4 T cell responses examined 42 days later. As previously reported (33, 34), immunized WildType mice had a large population of CD4 T cells that produced IFNγ in response to HKST stimulation (Fig. 3A and Supplemental Fig. 2B). In marked contrast, immunized B cell deficient mice had lower numbers of IFNγ-producing Th1 cells in response to HKST (Fig. 3A and Supplemental Fig. 2B). This difference was antibody-independent since immunized m+s IgM and mIgM mice had similar levels of IFNγ-producing CD4 T cells as WildType mice (Fig. 3A and Supplemental Fig. 2B). In fact, mIgM mice, which lack all secreted antibodies, had a larger population of Salmonella-specific IFNγ-producing CD4 T cells. Interestingly, IFNγ-producing CD8 T cells were also slightly reduced in immunized B cell deficient mice, but this was not statistically significant (Fig. 3B and Supplemental Fig. 2B). Taken together, these data indicate that B cells, but not antibody, are required for shaping the development of protective CD4 Th1 responses to Salmonella. A similar role for B cells has been reported in other infection models such as LCMV and Pneumocystis (35, 36). Although B cells may directly present antigen and drive Salmonella-specific Th1 responses, a recent study demonstrated that B cell production of IL-6 is important for maximal Th17 responses, and B cell production of IFNγ contributed to Th1 development (24). A recent study has also shown that B cells can affect secondary responses to Salmonella infection via a MyD88 and IL-10-dependent mechanism (37). Thus, B cells likely contribute to protective CD4 responses, via antigen presentation and production of specific cytokines that drive effector lineage commitment during primary responses. It is not yet clear whether these required B cells are necessarily Salmonella-specific, however the limited B cell repertoire in IgM Only and No Antibody mice did not affect protective immunity. We also attempted to address this issue using in vitro re-stimulation and B cell tetramer pull-down experiments in previously infected mice, but did not detect an elevated frequency of Salmonella-specific B cells using either of these approaches. However, it remains possible that expanded Salmonella-specific B cells contribute to immunity to secondary infection.
Together, our data demonstrate that antibody production plays only a minor role in Salmonella immunity while B cells are required for the development of protective T cell immunity. These findings will be important for the development of new effective vaccines against typhoid and NTS.
1This work was supported by Grants AI091298 (to MRN), AI087830 (to SSW), AI043603 (to MJS), AI055743 and AI073672 (to SJM), and T32 GM008244 (to UMN MSTP) from the National Institutes of Health.