B cell activating factor (BAFF) from neutrophils and dendritic cells is required for protective B cell responses against Salmonella typhimurium infection

Mice lacking B cells are more susceptible to S. typhimurium infection. How B cells contribute to protective immunity against Salmonella and what signals drive their activation are still unclear. Neutrophils (Nphs), monocytes (MOs), and dendritic cells (DCs) are involved in early immune responses to control the initial replication of S. typhimurium. These cells can produce B cell activating factor (BAFF) required for mature B cell survival and may help regulate B cell responses during Salmonella infection. Using BAFF reporter mice (BAFF-RFP+/-), we discovered that an i.p. infection with a virulent strain of S. typhimurium increased BAFF expression in splenic conventional DCs (cDC) and inflammatory Ly6Chi MOs/DCs four days post-infection. S. typhimurium infection induced the release of BAFF from Nphs, a decrease of BAFF-RFP expression and expansion of BAFF-RFP+ Nphs in the spleen and peritoneal cavity. After S. typhimurium infection, serum BAFF levels and immature and mature B cell subsets and plasma cells increased substantially. Conditional knockout (cKO) mice lacking BAFF in either Nphs or cDCs compared to control Bafffl/fl mice had reduced up-regulation of systemic BAFF levels and reduced expansion of mature and germinal center B cell subsets after infection. Importantly, the cKO mice lacking BAFF from either Nphs or cDCs had impaired induction of Salmonella-specific IgM Abs, and were more susceptible to S. typhimurium infection. Thus, Nphs and cDCs are major cellular sources of BAFF driving B cell responses, required for mounting optimal protective immunity against lethal Salmonella infection.


Introduction
B cell activating factor (BAFF, also known as BLyS or Tnfsf13b) is an essential cytokine required for B-cell survival, maturation and homeostasis. It binds to three different receptors: BAFFR (BAFF receptor), TACI (Transmembrane activator and calcium-modulating cyclophilin ligand interactor) and BCMA (B-cell maturation antigen), which are predominantly expressed on mature B cells [1,2]. Mice missing either BAFF or BAFFR lack peripheral mature B cells, underscoring the crucial function of BAFF in B cell development, response to antigens (Ags) and survival [3][4][5]. Increased production of BAFF has been implicated in the pathogenesis of several autoimmune diseases, including systemic lupus erythematosus, rheumatoid arthritis, and Sjögren's syndrome [6][7][8]. Although less explored, BAFF also plays an important role in host responses against pathogens. BAFF is essential from protection against viral infections such as West Nile virus (WNV) and promotes WNV-specific Ab responses [9]. BAFF production in response to blood-borne T cell-independent (TI) Ags results in robust classswitched Ab responses [10][11][12]. BAFF also contributes to control of bacterial infections such as B. burgdorferi, A. pleuropneumoniae and M. tuberculosis by boosting humoral immune responses [13]. Increased BAFF levels due to genetic variants in the Baff gene results in greater protection against malaria infection [14]. Furthermore, a subset of common variable immunodeficiency patients with a mutation in the Baffr gene are associated with recurrent bacterial sinopulmonary infections [15], underscoring the importance of BAFF in certain bacterial infections.
We and others have shown that myeloid cells like neutrophils (Nphs), monocytes (MOs) and conventional DCs (cDCs) are major producers of BAFF during infection and inflammation [9,12,16,17]. Following intestinal infection, S. typhimurium colonizes mesenteric lymph nodes and the spleen, where myeloid cells are recruited from the bone marrow (BM) to control the initial bacterial replication [18][19][20][21]. Nphs are rapidly recruited in the infected tissue upon Salmonella infection and actively phagocytose the bacteria to prevent their spread [19,22]. Inflammatory MOs are also recruited to infected tissues, and in addition to being very active in taking up Salmonella, they are the main producers of inflammatory cytokines and inducible nitric oxide synthase (iNOS) that promotes bacterial killing [20,21]. While DCs can produce inflammatory cytokines, they are thought to primarily play a role in Ag presentation and initiating adaptive T cell immunity against Salmonella [20,23]. Myeloid cells may also provide activating and survival signals to B cells supporting adaptive immune responses [24][25][26][27]. Whether BAFF produced by myeloid cells plays a role in B cell regulation and protective immune responses to S. typhimurium has not been investigated.
Mice lacking B cells are more susceptible to virulent Salmonella infection, suggesting that B cells play a role during primary S. typhimurium infection [28]. While passive transfer of immune serum was protective in some studies [29,30], another study found that the protective role of B cells during secondary infection with virulent Salmonella was independent of Ab secretion [31]. A number of studies suggest that B cells contribute to Salmonella immunity mostly during secondary infections [30][31][32]. However, systemic infection studies with attenuated Salmonella strains have shown that extrafollicular (EF) Ab responses help limit the spread of infection after primary challenge [33,34]. How B cells contribute to immunity during an acute systemic infection of virulent Salmonella is less well understood. In the present study, we identified specific BAFF-producing myeloid cell populations affected by systemic Salmonella infection and investigated their requirement for evoking protective B cell responses. We utilized BAFF reporter mice and Baff fl/fl mice we developed [9] and infected i.p. with a virulent Salmonella strain to investigate primary systemic immune responses to acute infection, as previous studies have done with attenuated Salmonella strains [18,33]. Infection of BAFF reporter mice by S. typhimurium led to significant changes in BAFF-expressing Nph, MO, and cDC subsets. The selective removal of BAFF from Nphs or cDCs reduced the expansion of splenic B cell subsets, BAFF serum levels, and Salmonella-specific IgM Ab responses, revealing that BAFF produced from Nphs and cDCs is required for mounting B cell responses to S. typhimurium infection. Importantly, the cKO mice lacking BAFF from either Nphs or cDCs were more susceptible to S. typhimurium infection. Thus, BAFF produced by both Nphs and cDCs play a key role in driving protective B cell immunity against lethal S. typhimurium infection.

Characterization of responses of BAFF-RFP +/reporter mice to S. typhimurium infection
In order to analyze which BAFF-producing cells drive B cell activation during primary Salmonella infection, we used our BAFF reporter mice (BAFF-RFP), where one Baff allele expresses the RFP protein as a measurement of BAFF expression and the other allele expresses wild type (WT) Baff [9]. Although BAFF-RFP +/mice, as expected, express reduced levels of BAFF compared to WT mice, they are still capable of producing mature B cells. We first examined whether the immune responses to S. typhimurium infection of BAFF-RFP +/mice were similar to WT mice. Mice were infected i.p. with 500 CFU of virulent S. typhimurium and monitored daily for survival, clinical scores, and body weight. In agreement with previous studies, B6 mice succumbed to S. typhimurium infection within a week [18,30,35]. WT and BAFF-RFP +/mice showed similar signs of illness and decrease in body weights during the first 3 to 4 days post infection (p.i.) (Fig 1A), and during this period, BAFF-RFP +/infected mice showed significantly enlarged spleens and increased splenic bacterial burdens comparable to WT mice ( Fig 1B).
Primary infection with virulent S. typhimurium via the oral route leads to major changes in splenic myeloid cell populations early in the course of infection [19,21,23]. We observed similar changes in BM and splenic myeloid cell populations in BAFF-RFP +/mice, as well as WT mice, 3-4 days after systemic Salmonella infection (Fig 1C and 1D and S1 Fig in S1 File). BM Nphs, cDC precursors, and inflammatory Ly6C hi MOs decreased in BAFF-RFP +/mice and WT mice with disease progression (Fig 1C and S1A Fig in S1 File). In contrast, splenic Nph numbers from BAFF-RFP +/and WT infected mice substantially expanded by day 3 and remained elevated at day 4 p.i. (Fig 1D and S1B Fig in S1 File). Splenic cDCs were significantly decreased by day 4 p.i. (Fig 1D and S1B Fig in S1 File), whereas inflammatory Ly6C hi MO numbers increased in the infected spleens at day 3 and later decreased as the disease progressed ( Fig 1D and S1B in S1 File). As previous studies have shown, the expansion of Nphs in the spleen is most likely due to their recruitment from the BM [18][19][20][21]. Thus, BAFF-RFP +/mice showed a similar progression of disease and early recruitment of Nphs from the BM to the spleen as WT mice (Fig 1A and 1B and S1 Fig in S1 File). In addition, these data show that i.p. infection with a virulent S. typhimurium strain is a good model for examining early systemic immune responses as it leads to disease progression and changes in innate cell populations similar to those described in oral infection studies [19].

PLOS ONE
CD8 -cDC subsets, as well as inflammatory Ly6C hi MO and Ly6C hi DC subsets, upregulated the BAFF-RFP signal upon infection. In contrast, BAFF-RFP expression per cell did not change in DC and MO subsets from the BMs of infected BAFF-RFP +/mice (S2A Fig in S1 File). We conclude that S. typhimurium infection upregulates BAFF expression specifically in cDCs and MOs in the spleen.

S. typhimurium infection induces BAFF release from Nphs, expansion of splenic BAFF-RFP + Nphs and increased systemic levels of BAFF
Nphs express the highest BAFF levels compared to other myeloid cells [9,16,37]. BM and splenic Nphs from naïve BAFF-RFP +/mice consistently had constitutively high levels of BAFF-RFP expression (Fig 3A) [9]. By day 3 p.i. BAFF-RFP levels in BM Nphs were downregulated and decreased further at day 4 p.i (Fig 3A and 3B, upper panels). A decrease in BAFF-RFP expression in splenic Nphs was only evident at day 4 p.i. (Fig 3A and 3B, lower panels). These data suggest that Salmonella directly induces activation of Nphs and the downregulation of BAFF expression. To further test this possibility, we treated purified BAFF-RFP +/-BM Nphs in vitro with bacterial lysates or live S. typhimurium. A combination of G-CSF plus GM-CSF was used as positive control, as it is known to enhance BAFF release from Nphs [37]. 24 hrs after stimulation of Nphs with S. typhimurium the BAFF-RFP signal was decreased in

PLOS ONE
Nphs (Fig 3C and 3D). As expected, since the BAFF-RFP signal is a measurement of BAFF expression [9] (see also Material and methods), we also detected significantly lower Baff mRNA in purified Nphs after S. typhimurium treatment compared to untreated Nphs ( Fig 3E). Baff mRNA also decreased in WT Nphs 24 hr after stimulation with S. typhimurium (S2B Fig in S1 File). S. typhimurium downregulated Baff mRNA expression in WT and BAFF-RFP +/-Nphs as early as 6 hrs after treatment (S2C Fig in S1 File).
We next tested whether Salmonella directly induced BAFF protein release from Nphs obtained from either WT or BAFF-RFP +/mice. Stimulation with either bacterial lysates or with live S. typhimurium enhanced BAFF release in both WT and BAFF-RFP +/-Nphs ( Fig 4A). As expected, BAFF levels in cell supernatants from purified Nphs from BAFF-RFP +/mice were about half the levels detected in WT mice ( Fig 4A). S. typhimurium triggered BAFF release from Nphs within 6 hrs (S2D Fig in S1 File). The substantial increase in BAFF release from Nphs and recruitment of Nphs from the BM to the spleen suggested that Salmonella might induce increased systemic levels of BAFF. Indeed, WT and BAFF-RFP +/mice infected with S. typhimurium have substantial increases in serum BAFF levels by day 4 p.i. (Fig 4B). In addition, BAFF-producing Nph numbers declined in the BM ( Fig 4C) and expanded in the spleen and peritoneal cavity ( Fig 4D and S2E Fig in S1 File). BAFF-RFP + DC precursors decreased in the BM (Fig 4C). In contrast to Nphs, BAFF-RFP + cDC did not change significantly in the spleen (Fig 4D).
We conclude that S. typhimurium infection induces: 1) the recruitment of BAFF-RFP + Nphs to the spleen and peritoneal cavity; 2) BAFF release from Nphs; 3) increases in BAFF expression in splenic cDCs and MOs; and 4) at the same time, increases in systemic levels of BAFF.

S. typhimurium infection induces the expansion of T1 B cells, FO B cells and plasma cells
As BAFF is a key B cell survival factor [1,38,39], we examined whether the increase in serum BAFF levels and expansion of BAFF-producing cells after S. typhimurium infection affects the expansion of splenic B cell subsets in WT and BAFF-RFP +/mice. Transitional 1 (T1) and follicular (FO) B cell numbers were significantly increased at day 4 p.i compared to uninfected WT and BAFF-RFP +/mice (Fig 5, S3A and S3B Fig in S1 File). An increase in splenic MZ B cells and a trend towards increasing T2 cells were observed in WT mice (Fig 5). In contrast, the precursors of MZ B cells (MZP) were significantly decreased in both mice strains. Moreover, splenic B220 lo CD138 + PC numbers were significantly up-regulated in WT and BAFF-RFP +/mice ( Fig 5 and S3B Fig in S1 File). Thus, S. typhimurium infection, together with the expansion of myeloid BAFF-producing cells, induced changes in the splenic B cells, particularly, an expansion of T1, FO B cell and PCs.

Salmonella-specific IgM responses and protection against S. typhimurium infection
We next addressed whether BAFF plays a protective role during systemic Salmonella infection. Previous studies showed that mice lacking B cells were more susceptible to virulent S. typhimurium primary oral infection [28]. Similarly, B cell deficient μMT mice succumbed to virulent S. typhimurium i.p. infection significantly earlier than WT mice (S3C Fig in S1 File). Therefore, B-cell mediated immunity plays a protective role during S. typhimurium infection.
Using myeloid cell BAFF cKO mice [9], we determined which myeloid BAFF sources contribute to increases in systemic BAFF levels, expansion of splenic B cells and protective immunity after S. typhimurium infection. We first assessed the role of Nph-derived BAFF by using Baff fl/fl Mrp8 Cre (BAFF Nph cKO) mice, where Baff is selectively deleted in Nphs (14). BAFF Nph cKO mice infected with S. typhimurium showed some increases in BAFF serum levels 4 days after infection, but the serum BAFF increases overall were significantly lower than in infected Baff fl/fl control mice (Fig 6A). In addition, the numbers of splenic transitional T1 and T2 B cells, FO B cells and PCs were increased to a lesser extent in infected BAFF Nph cKO mice compared to control mice ( Fig 6B). Furthermore, while CD38 -GL7 + germinal center (GC) B cell numbers were up-regulated in Baff fl/fl mice, they did not expand in BAFF Nph cKO mice after S. typhimurium infection (Fig 6C). However, there was no significant difference between GC B cells from infected Baff fl/fl mice and infected BAFF Nph cKO mice, suggesting that Nphs do not play a major role in GC B cell expansion. To address potential changes in B cell activation in response to S. typhimurium infection, we examined levels of CD69, an early B cell activation marker [40,41], and MHC class II, an indicator of the Ag presenting ability of B cells [42,43]. T1, T2 and FO B cells, but not MZ had significantly reduced up-regulation of CD69 in infected BAFF Nph cKO mice compared to Baff fl/fl mice (S4A Fig in S1 File, upper panels). In contrast, the changes induced by S. typhimurium infection in MHC class II expression on B cell subsets were comparable in BAFF Nph cKO mice and Baff fl/fl mice (S4B Fig in S1 File, upper panels). Consistent with decreased expansion of PCs, BAFF Nph cKO mice did not up-regulate Salmonella-specific serum IgM, that was induced in Baff fl/fl mice after S. typhimurium infection (Fig 6D).
We next tested if BAFF-producing Nphs play a protective role during S. typhimurium infection. BAFF Nph cKO mice infected i.p. with 25 CFU S. typhimurium had significantly increased mortality compared to Baff fl/fl control mice (Fig 6E). Mrp8 Cre mice were used as  Fig in S1 File). The infected BAFF Nph cKO mice also showed more advanced signs of illness and higher systemic bacterial titers than Baff fl/fl mice (Fig 6F). We conclude that during Salmonella infection BAFF produced by Nphs contributes to the increased systemic BAFF levels, expansion and activation of splenic B cells and PCs, Salmonella-specific IgM responses. and plays a protective role against S. typhimurium infection.

BAFF produced by cDCs is required for B cell expansion, Salmonellaspecific IgM responses and protection against S. typhimurium infection
DCs play a significant role in the activation of adaptive immunity during the course of Salmonella infection [20,23]. Splenic cDCs up-regulate BAFF expression during systemic S. typhimurium infection (Fig 2). Thus, we assessed if BAFF production by cDCs plays a role in B cell responses and protective immunity after infection. Uninfected Baff fl/fl zDC Cre (BAFF cDC cKO) mice [9], where BAFF is selectively depleted in cDCs, showed lower serum BAFF levels compared to controls (Fig 7A). These data indicate that in contrast to BAFF from Nphs, BAFF from cDC contributes to the homeostasis of systemic BAFF. After S. typhimurium infection BAFF cDC cKO mice increased serum BAFF levels to a lesser extent than Baff fl/fl control mice ( Fig 7A). However, in contrast to Nph BAFF-depleted mice, the lack of BAFF from cDCs completely prevented the increases in FO B cells and PC numbers observed in control mice ( Fig 7B). Also, after infection, MZ B cells were significantly lower in BAFF cDC cKO mice compared to Baff fl/fl mice (Fig 7B). Moreover, the expansion of splenic GC B cells observed in Baff fl/fl mice after S. typhimurium infection, was not detected in the BAFF cDC cKO mice ( Fig 7C).
Interestingly, in contrast to BAFF depletion in Nphs, BAFF depletion in cDCs did not affect CD69 up-regulation triggered by S. typhimurium infection (S4A Fig in S1 File, lower panels). However, the MHC II up-regulation occurring in MZ and MZP B cells from Baff fl/fl control mice upon S. typhimurium infection was completely ablated in BAFF cDC cKO mice (S4B Fig  in S1 File, lower panels). As in the BAFF Nph cKO mice, the increase in Salmonella-specific serum IgM Ab that occurred in control mice after infection, did not occur in BAFF cDC cKO mice ( Fig 7D). Finally, similar to the BAFF Nph cKO mice, BAFF cDC cKO mice also had increased mortality (Fig 7E), higher clinical scores and bacterial burden (Fig 7F) compared to the Baff fl/fl control mice, suggesting that BAFF from cDCs also plays an important protective role in S. typhimurium infection. Survival rates and systemic bacterial burdens were not significantly different in control zDC Cre mice and Baff fl/fl mice (S5B Fig in S1 File). In contrast to BAFF Nph cKO mice and BAFF cDC cKO mice, the mortality of BAFF MO cKO mice was comparable to Baff fl/fl control mice (S5C Fig in S1 File), suggesting that BAFF from MOs does not contribute to protection against Salmonella infection.
We conclude, that BAFF produced by both Nphs and cDCs is required for optimal specific humoral responses and protective immunity against Salmonella infection. Interestingly, BAFF from Nphs and cDCs regulate B cell responses differently. Nph-derived BAFF, but not BAFF from cDCs, regulate immature B cells. In contrast, unlike Nph-derived BAFF, cDC-derived BAFF is indispensable for mature B cell expansion in response to Salmonella infection.

Discussion
Innate immune cells such as Nphs, cDCs and MOs contribute to immunity against a number of infections, including Salmonella [20,[44][45][46][47]. The role of innate immune cells goes far beyond pathogen killing and involves production of cytokines that critically link the innate and adaptive immune system [13,[24][25][26]. B cells are major contributors to protective acquired immunity to Salmonella infection [48]. Here we showed that BAFF produced from two different innate myeloid cell sources-Nphs and DCs-regulate B cell responses and are required for optimal protective immunity against S. typhimurium infection.
To our knowledge Nph-derived BAFF has not been previously shown to contribute to protective immune responses against infectious diseases. In our previous study, BAFF from Nphs was not required for protective immunity to WNV [9]. BAFF secretion from CD11b + cells in response to the parasite Trypanosoma cruzi was reported to contribute to protection by expansion of total mature B cells [49]; however, which CD11b + cell subsets were the essential sources of BAFF during T. cruzi infection was not investigated. Deniset et al. reported that splenic Nphs interact with B cells in the red pulp and marginal zones, promote T cell-independent (TI) Ab responses and contribute to systemic clearance of Streptococcus pneumoniae [50]. However, whether Nphs helped B cell responses to S. pneumoniae via BAFF was not addressed.
Here we definitively demonstrated that Nphs-derived BAFF plays a protective role during Salmonella infection. In addition, we demonstrated that Nph-derived BAFF significantly contributes to increased systemic BAFF levels, is required for the induction of Salmonellaspecific IgM responses, and plays a major role in sustaining B cell expansion during S. typhimurium infection (Fig 6). In agreement with our results, Coquery et al. found that in an autoimmune hyper inflammatory environment, activated Nphs may contribute to excessive serum BAFF levels and promote B cell responses [51]. BAFF produced by Nphs is required not only for Salmonella-driven expansion of mature B cells and PCs, but also for the optimal expansion of newly formed/transitional B cell (NF-B cell) subsets (T1 and T2 B cells). These data are particularly interesting, given the co-localization of NF-B cells and Nphs in the splenic red pulp [52]. A growing number of studies support a role for NF-B cells in protection against infections [52].
During infection, when splenic Nphs are exposed to microbial and inflammatory signals, they acquire "B helper" functions and work together with B cells to promote adaptive immune responses [27,53,54]. In addition to the red pulp, Nphs colonize the peri-MZ areas of the spleen where, via BAFF and other factors, they help generate TI antimicrobial Ab responses by promoting MZ B cell differentiation into PCs [24,55]. Nph B cell helpers via production of Pentraxin 3, a complement activator, activate MZ B cells and elicit class switching and plasmablast expansion, promoting TI and T-cell dependent responses to S. pneumoniae infection [56]. Whether Nphs directly interact with mature B cells like FO and GC B cell remains controversial [54]. Recent studies have implicated recruitment of Nphs in splenic B-cell follicles and GCs in the presence of appropriate inflammatory signals [24]. Upon Salmonella infection, optimal expansion of NF-B cells, PCs as well as mature B cell subsets are reduced in mice lacking BAFF in Nphs. An attractive model is that during Salmonella infection, some Nph-BAFF in the EF area and peri-MZ zone support NF-B cells and PCs expansion, while other Nphs infiltrate the B cell follicle and via BAFF support the expansion of FO B cells. Together our data highlights BAFF-production and regulation of B cells as new important Nph function essential for host survival to Salmonella infection [20].
In addition to changes in BAFF-producing Nphs, splenic cDCs and inflammatory MOs upregulate BAFF expression upon Salmonella infection. The lack of BAFF from MOs did not affect mouse survival, suggesting that MO-derived BAFF does not play a role in protective immunity to Salmonella. However, mice lacking BAFF in cDCs were more susceptible than control mice to S. typhimurium infection (Fig 7). We have previously shown that cDC-BAFF, in contrast to Nph-BAFF, plays an important role in protective B cell responses against viral infection [9]. Another study suggested DCs by upregulating BAFF expression contribute to protective humoral immune responses against A. pleuropneumoniae infection [57]. Interestingly, BAFF from cDCs plays a more prominent role than Nph-BAFF in contributing to systemic BAFF levels and B cell responses against Salmonella infection: whilst FO B cell and PC expansion was reduced in BAFF Nph cKO mice, it was completely abrogated in BAFF cDC cKO infected mice. Previous studies have suggested that BAFF produced by DCs may drive PC responses [12,58]. LPS from Peyer's patch resident commensal bacterium Alcaligenes enhanced BAFF production by DCs, promoting antigen-specific IgA production [59]. Our results demonstrate that DC-BAFF is required for PC expansion in vivo during bacterial infection. In contrast to Nphs, DCs are key players in adaptive immunity to Salmonella [20,60,61]. cDCs are required to activate Salmonella-specific T cells and during acute Salmonella infection cDCs subsets are rapidly activated and reorganized within the spleen [20,23]. Upon S. typhimurium infection a subset of CD8 -cDCs is redistributed from the MZ to the B cell follicles, while other CD8 -cDCs and CD8 + cDCs are increased in the red pulp, in addition to their normal presence in the T-cell areas [23]. As cDC-BAFF is required for FO and GC B cells expansion, it is possible that cDC redistribution in the B cell areas is key to support FO and GC B cell responses. Further studies are needed to address cross talk and localization of Nph and cDC with B cells in the spleen during Salmonella infection.
Although BAFF from both Nphs and cDCs plays a major role in regulating B cells, these BAFF sources affect B cell responses to Salmonella infection differently. First, Nph-BAFF regulates T1 B cells while DC-BAFF does not. Second, BAFF from Nphs, but not from cDCs, contributes to Salmonella-induced CD69 up-regulation in several B cell subsets. Third, BAFF from cDCs, but not Nphs, is required for MHC class II activation in MZ and MZP B cells. Either location in the spleen or specific crosstalk with B cells may explain the diverse effects on B cell activation by Nph-derived or cDC-derived BAFF. Our results suggest that while Nph-BAFF regulates early B cell activation [41], cDC-BAFF may affect the Ag presenting ability of B cells (33,76,77). MZ and MZP B have the highest MHC class II expression among B cells [43,53]. In addition, the location of cDCs in the splenic MZ makes them good candidates to support MZ B cell Ag presenting capability via BAFF [62,63]. Earlier studies have suggested that during primary infection, B cell Ag presentation, but not Ab secretion, may help the development of protective T cell immunity against secondary Salmonella infection [31,32,64].
Infection with attenuated Salmonella leads to IgG production during EF responses and a delayed GC response [33,34]. In contrast, primary infection with virulent Salmonella induced an unexpected early expansion of GC B cells that required BAFF produced by cDCs. At these early timepoints after infection we did not detect any change in Salmonella-specific serum IgG and IgA Abs in either control Baff fl/fl mice, BAFF cDC cKO mice or BAFF Nph cKO mice. However, the surprising early increase in GC B cells suggests that systemic primary infection with virulent Salmonella may induce a rapid adaptive immune response. Consistent with this possibility, McSorley et al. reported that Salmonella-specific T cell responses in mesenteric lymph nodes and Peyer's patches occur within hours after oral infection with virulent Salmonella [65]. Since systemic Salmonella infection rapidly targets the spleen, it is possible that Salmonella-specific T cells develop at early time points in this model. Further studies are needed to determine whether after acute systemic infection, BAFF from splenic cDCs or Nphs regulates the ability of B cells to present Ags, induces localized changes in class-switched Abs and/ or activate protective Salmonella-specific CD4 + T cells [48].
BAFF from Nphs and cDCs is also required for Salmonella-specific IgM production after infection. Salmonella-specific IgM Abs produced early after infection opsonize extracellular bacteria controlling tissue colonization and rapid increase in systemic bacterial burdens [33]. The fact that mice lacking BAFF in either Nphs or cDCs have higher bacterial levels in blood than control mice, thus, could be due to a deficiency in specific IgM Abs. Our findings are consistent with previous studies showing that Abs are required for vaccines to protect against infection with virulent Salmonella [30,66]. B1 B cells rapidly produce IgM and protective Ab responses against S. typhimurium [67]. Our data do not rule out a role for other B cell subsets like B1 B cells in contributing locally to IgM production.
A limitation of our study is that infected mice die within a week after systemic infection with virulent Salmonella. However, our results clearly indicate that: 1) BAFF plays a protective role during primary Salmonella infection, and 2) this is most likely due to the regulation of B cells, given that BAFF-producing DCs and Nphs are required for the early up-regulation of specific B cell subsets and the Salmonella-specific IgM Abs induced by systemic acute infection. Further studies are needed to determine whether BAFF and B cell protective role occurs via IgM response, antigen presentation or cytokine production.
Stimulation with bacterial Ags induces BAFF release from DCs, MO and macrophages in vitro [13,57]. Here we report that Salmonella directly and rapidly induces BAFF release from Nphs. Nphs constitutively express high levels of BAFF during homeostatic conditions [9,37]. Nphs are the main infiltrating cell population during acute inflammatory responses to Salmonella [22,68], and as we report here, are the predominant BAFF source expanded in the spleen and peritoneal cavity after systemic Salmonella infection. Cytokines produced by innate cells in response to inflammation can activate Nphs and induce BAFF release [37,69,70]. Virulent Salmonella induced BAFF release from Nphs in vitro to the same extent as a combination of G-CSF and GM-CSF, a potent inducer of BAFF release from Nphs [37]. Interestingly, S. typhimurium while inducing BAFF release from Nphs, also downregulated BAFF expression in vivo and in vitro. TNFα treatment also induced BAFF secretion and down-regulation of BAFF expression in human Nphs, similar to what we report here in mouse Nphs treated with S. typhimurium [71]. BAFF secretion and down-regulation of BAFF expression were induced in vitro by Salmonella within 6 hrs, timing comparable to cytokine-induced BAFF release from mouse Nphs [37]. These effects were not only rapid, but also driven by a Salmonella lysate as well as by live bacteria, suggesting that either bacterial -PAMPS or -proteins can directly affect BAFF in Nphs. However, it is also possible that inflammatory cytokines rapidly induced by S. typhimurium infection play a role [72].
Our study highlights the importance of the relationship between innate cells and adaptive B cell responses during Salmonella infection. Although earlier studies have implicated B cells in protection during secondary infections [30][31][32], our work emphasize that B cells are also relevant for primary acute Salmonella infection. Taken together our data demonstrate that BAFF from Nphs and cDCs shapes early mature B cell responses and strengthens protective host responses against primary S. typhimurium infection. An improved understanding of niches of BAFF producing cells regulating B cell responses against Salmonella, as well as other harmful bacteria, may help designing novel strategies for vaccine and therapeutics development.

Mice
C57BL/6J (B6), mice were purchased from Jackson Labs (Bar Harbor, ME) and housed or bred in our facilities at the University of Washington. BAFF-RFP and Baff fl/fl mice (on the B6 background) were developed at the University of California at Davis Mouse Biology Program (UC Davis, CA). Generation of these mice strains are fully described elsewhere [9]. Briefly, BAFF-RFP reporter mice were generated by replacing a Tnfsf13b/Baff allele with a targeting construct expressing IRES-RFP. The Tnfsf13b/Baff is functionally knocked-out where the endogenous reporter expresses the RFP protein translated under the control on an IRES site. Thus, the BAFF-RFP signal is a measurement of BAFF expression. These heterozygous BAFF-RFP +/mice express the RFP protein on one Baff allele and the WT Baff on the other [9]. Baff fl/fl mice were used as controls as these mice express wild type Baff until Cre-mediated deletion. Baff fl/fl mice were crossed with Mrp8 Cre mice [73], zDC Cre mice [74] and Cx3cr1 Cre mice [73] to generate conditional knockout (cKO) mice where Baff is selectively deleted in either Nphs (Baff fl/fl Mrp8 Cre ), cDCs (Baff fl/fl zDC Cre ) and MOs (Baff fl/fl Cx3cr1 Cre ), respectively. All mouse strains were verified by sorting of appropriate cell subsets and assessing relative expression of Baff mRNA in sorted cells by qPCR [9].
Animals were housed under standard barrier conditions in groups of up to five animals in individually ventilated cages within a specific pathogen free environment. Mice were age-and sex-matched and used at 8-14 weeks of age.

Ethics statement
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All procedures were approved and conducted according to regulations of the University of Washington Institutional Animal Care and Use Committee (2242-08 and 4391-01). All research staff completed the required training for animal care or handling. Humane end points were defined prior to the experiments and all possible measures were taken to minimize animal suffering.

Bacterial strains and mouse infection
The experiments were performed using Salmonella enterica serovar Typhimurium strain SL1344 kindly provided by Kelly Smith (University of Washington, Seattle, WA). The inoculum containing S. typhimurium was prepared as described [75]. Briefly, a single bacterial colony was grown overnight in Luria-Bertani (LB) broth containing 50 μg/ml of streptomycin at 37˚C with shaking and back-diluted 1:50 in fresh media. After 4 hours, the OD (600 nm) of bacteria was measured and bacteria appropriately diluted in cold PBS. Mice were infected i.p. in a total volume of 200 μl with 7 or 25 CFU live bacteria for survival studies and with 500 CFU live bacteria for all other experiments. CFUs of the inoculi were verified by plating on LB agar plates containing 50 μg/ml streptomycin (LB/strep plates). Mice were monitored daily for 10 days or twice a day when clinical symptoms were expected to worsen. Survival, clinical scores and mice weights were recorded daily. Mouse survival is given as percentage of live animals per time point. Clinical scores and humane end points in response to systemic S. typhimurium infection were defined as described previously [76]. Clinical evaluations of mice were performed based on mice activity, feeding, appearance of fur, hunched position, ataxia and tremor. Humane endpoints were determined when mice showed severe clinical symptoms due to S. typhimurium infection and were immediately euthanized. Some mice during survival studies succumbed to infection before meeting criteria for euthanasia.

Tissue harvest and cell isolations
Mice were sacrificed by CO 2 asphyxiation at the indicated time points, and spleen and bone marrow (BM) were harvested for cell subset analyses. Mice and spleens were weighed, and weights were used for determining splenomegaly (mean of spleen weight/body weight (%)). A small part of each spleen was homogenized by 1.4 mm-bead disruption in a FastPrep-24 instrument (MP Biomedicals) in PBS with 0.025% Triton X-100. Bacterial burdens were assessed by plating serial dilutions of the spleen homogenates on LB/strep plates. Blood bacterial levels were assessed by immediately plating 100 ul of blood from retro-orbitol bleeds or serial dilutions of blood on LB/strep plates before the blood clotted. Preparation of splenic cell suspensions and isolation of BM cells and peritoneal exudate cells (PEC) from uninfected or S. typhimurium-infected mice were performed as described previously [9,10,77].

Isolation of Nphs and stimulations assays
BMs from Naïve WT and BAFF-RFP +/mice were harvested; Nphs were enriched from BM cell suspensions using a magnetic-activated cell-sorting (MACS) neutrophil isolation kit (BM >95% purity, Miltenyi Biotec). Enriched Nphs were counted and seeded in 48 or 24 well plates at 2 or 4 x 10 6 cells/well respectively in RPMI 1640 supplemented with 10% FBS at 37˚C in 5% CO 2 . Nphs were stimulated with S. typhimurium whole cell lysates (200 μg/ml), live S. typhimurium (1:2 and 1:10 MOI) or G-CSF+GM-CSF (100 ng/ml) for 6 or 24 hrs. Unstimulated Nphs were used as a negative control. The S. typhimurium whole cell lysate was prepared by homogenizing an overnight bacterial culture by 0.1 mm-bead disruption in a FastPrep-24 instrument (MP Biomedicals) in PBS. The protein content of the lysate was measured using Pierce BCA protein assay kit as per manufactures instructions. After incubation the cell supernatants were collected and stored at -80˚C. Half of the Nph cells were processed for flow cytometry to assess the BAFF-RFP signal; the other half of the cells were lysed and RNA extracted for BAFF qRT-PCR analyses as described previously [9].

Salmonella specific Ab and BAFF protein detection
Sera were isolated from blood, collected via the retro-orbital route, and stored at -80˚C until use. Salmonella-specific serum IgM Ab were detected by ELISA. Maxisorp plates (Nunc) were coated with 50 μg/ml Salmonella whole cell lysate and the amount of specific Ab were determined as described previously using a standard curve [77]. Briefly, 2 μg/ml anti-mouse IgM (Jackson ImmunoResearch), were coated, and serial dilutions of known concentration of recombinant IgM (Southern Biotech, Birmingham, AL) were used for the standard curve. Serum samples and recombinant IgM were diluted in 4% milk casein in PBS containing 0.05% Tween-20, and primary mouse Abs were detected using isotype specific goat anti-mouse antibodies conjugated to horseradish peroxidase (Southern Biotech, Birmingham, AL). Serum levels of BAFF from in vivo experiments and in supernatants from stimulated Nph cultures were determined using the Mouse BAFF/BLyS/TNFSF13B DuoSet ELISA kit (R&D systems, Inc., Minneapolis, MN) according to the manufacturer's instructions.

Statistical analyses
All statistical analyses were performed with Prism software (GraphPad Software). A p value of < 0.05 was considered significant. Survival data were analyzed by Mantel-Cox log-rank test. Statistical analyses of flow cytometry experiments and BAFF protein detected by ELISA comparing more than two groups were performed with One-Way ANOVA corrected for multiple comparisons using the Holm-Sidak method. Analyses between two groups were performed using unpaired Student t test. Flow cytometry and BAFF protein experiments involving BAFF cKO mice, timeline data of clinical scores, mouse weights, and IgM Ab assays of various groups were analyzed using a Two-way ANOVA with Holm-Sidak's multiple comparison test.