Murine and Bovine γδ T Cells Enhance Innate Immunity against Brucella abortus Infections

γδ T cells have been postulated to act as a first line of defense against infectious agents, particularly intracellular pathogens, representing an important link between the innate and adaptive immune responses. Human γδ T cells expand in the blood of brucellosis patients and are active against Brucella in vitro. However, the role of γδ T cells in vivo during experimental brucellosis has not been studied. Here we report TCRδ−/− mice are more susceptible to B. abortus infection than C57BL/6 mice at one week post-infection as measured by splenic colonization and splenomegaly. An increase in TCRγδ cells was observed in the spleens of B. abortus-infected C57BL/6 mice, which peaked at two weeks post-infection and occurred concomitantly with diminished brucellae. γδ T cells were the major source of IL-17 following infection and also produced IFN-γ. Depletion of γδ T cells from C57BL/6, IL-17Rα−/−, and GMCSF−/− mice enhanced susceptibility to B. abortus infection although this susceptibility was unaltered in the mutant mice; however, when γδ T cells were depleted from IFN-γ−/− mice, enhanced susceptibility was observed. Neutralization of γδ T cells in the absence of TNF-α did not further impair immunity. In the absence of TNF-α or γδ T cells, B. abortus-infected mice showed enhanced IFN-γ, suggesting that they augmented production to compensate for the loss of γδ T cells and/or TNF-α. While the protective role of γδ T cells was TNF-α-dependent, γδ T cells were not the major source of TNF-α and activation of γδ T cells following B. abortus infection was TNF-α-independent. Additionally, bovine TCRγδ cells were found to respond rapidly to B. abortus infection upon co-culture with autologous macrophages and could impair the intramacrophage replication of B. abortus via IFN-γ. Collectively, these results demonstrate γδ T cells are important for early protection to B. abortus infections.


Introduction
Brucellosis is a widespread and economically important agricultural and human disease in many regions of the world. The etiologic agent of brucellosis in cattle is Brucella abortus, a Gram-negative, facultative, intracellular pathogen [1]. The pathological manifestations of brucellosis are diverse, depending upon host, and in humans include arthritis, endocarditis, and meningitis, while animal brucellosis is characterized by spontaneous abortions [2]. Although use of the B. abortus vaccine strains S19 and RB51 reduces disease incidence and prevents B. abortus-induced abortions, these vaccines are less than ideal because of their limited efficacy and potential to cause disease in humans [3,4]. Consequently, the need exists for improved animal protective measures that will combine safety and efficacy to all species at risk, including domestic herds and wildlife [5]. While much emphasis has been placed on identifying new vaccine candidates, relatively little attention has been paid to the effective coordination of the innate immune response by the host. Moreover, the need exists to determine protective innate immune responses against brucellosis to lessen the progression of infection and to allow a protective, adaptive immune response to develop [6].
cd T cells constitute a small percentage of circulating T cells in adult humans and in mice, but comprise a majority of the intraepithelial lymphocyte population in the gut and in other epithelial mucosa [7]. Moreover, cd T cells are a major subset in adult ruminants and make up a majority (up to 70%) of circulating lymphocytes in neonatal calves [8]. cd T cells are the first to develop and are recruited and expand in response to various infections in humans, rodents, and other animals [7].
Studies conducted with TCRd 2/2 mice demonstrate cd T cells can play a crucial role in protection from bacterial infection [9][10][11][12], and it has been proposed that mice lacking cd T cells have an altered or impaired response to Gram-negative bacteremia [10]. Different mechanisms have been proposed for the protection conferred by cd T cells in murine models. Moore et al. [10] postulates decreased levels of IFN-c and TNF-a observed in Klebsiella-infected TCRd 2/2 mice results in enhanced susceptibility to infection. However, in studies with Nocardia-infected mice, it is shown that TCRd 2/2 mice are unable to effectively recruit inflammatory cells, leading to increased susceptibility to infection (and death) [9]. cd T cells have also been shown to be a major producer of IL-17 in response to infection with intracellular bacteria [13,14].
In humans, Vc9d2 T cells represent the major subtype of cd T cells in blood, but make up only 1-5% of all circulating peripheral T cells [15]. However, the number of Vc9d2 T cells can dramatically increase in early response to infection by a number of intracellular pathogens, including Brucella [16]. Also, B. suis is shown to produce soluble factors that can directly activate Vc9d2 T cells to produce IFN-c and TNF-a [17]. In vitro, Vc9d2 T cells exhibit strong cytolytic activity against Brucella-infected cells and are able to impair intracellular growth of B. suis in autologous macrophages [17]. Inhibition of bacterial multiplication is partially attributed to IFN-c and TNF-a production by Vc9d2 T cells, while granule exocytosis is a mechanism used by Vc9d2 T cells to reduce the intracellular numbers of B. suis by lysing infected macrophages [15]. In addition, Vc9d2 T cells are shown to play a part in innate immunity to Brucella by releasing LL-37 as an antimicrobial defense mechanism [18]. While the effects of human Vc9d2 T cells against Brucella have been studied in vitro, this subset of cd T cells is absent in nonhuman primates [15]. In addition, the role of cd T cells in response to Brucella has not been studied in vivo in an animal model [6]. Here we demonstrate cd T cells can be protective against B. abortus infection in both murine and bovine models.

Results
TCRd 2/2 mice have impaired innate immunity to B. abortus infection To assess the role of cd T cells in protection against brucellosis, TCRd 2/2 and wild-type (wt) C57BL/6 mice were infected with B. abortus, and at select time points (1,3,7,14,21, and 28 days postinfection), the extent of splenic colonization was determined ( Figure 1A). At one week post-infection, TCRd 2/2 spleens contained ,ten-fold more Brucella than spleens from wt mice, and by two weeks post-infection, no significant differences in colonization were observed between TCRd 2/2 and wt mice. To determine if the protective role of cd T cells was limited to innate immunity, wt and TCRd 2/2 mice were immunized with RB51 and eight weeks later challenged with B. abortus 2308. Four weeks after challenge, vaccination efficacy was assessed. RB51 was found to be equally protective in wt and TCRd 2/2 mice ( Figure 1B), indicating the protective function of cd T cells against B. abortus must be limited to early infection and most likely to innate immunity. In an effort to determine if the enhanced susceptibility of TCRd 2/2 mice to infection was dose-dependent, both wt and TCRd 2/2 mice were infected with varying doses, 5610 3 , 5610 4 , or 5610 5 CFUs of B. abortus, and splenic colonization was assessed one week postinfection ( Figure 1C). TCRd 2/2 mice infected with either 5610 4 or 5610 5 CFUs of B. abortus were more susceptible to infection than wt mice ( Figure 1C). The kinetics of infection and susceptibility following infection with 5610 4 CFUs are similar ( Figure 1D) to that found with 5610 5 CFUs of B. abortus ( Figure 1A).

Splenic cd T cells proliferate in response to B. abortus infection
To determine if cd T cells are induced subsequent B. abortus infection, C57BL/6 mice were infected with 5610 4 CFUs of B. abortus, and the relative proportions of TCRcd + , TCRab + , CD4 + , Figure 1. cd T cells are required for innate, but not acquired, immunity to B. abortus infection. A. TCRd 2/2 and C57BL/6 mice were infected with 5610 5 CFUs of B. abortus 2308 i.p and splenic colonization was determined 1, 3, 7, 14, 21, and 28 days post-infection. The mean 6 SEM of three mice/time point/group, except day 7, which contains 10 mice/group, is shown. B. TCRd 2/2 and C57BL/6 mice were dosed i.p. with sterile PBS or 1610 8 CFUs of B. abortus RB51 vaccine. Mice were challenged 8 wks post-immunization with 5610 5 CFUs of B. abortus 2308, and splenic colonization was determined four wks post-infection. The mean 6 SD of 3-4 mice/group is shown; * P,0.05 versus unvaccinated mice of the same genotype. C. C57BL/6 and TCRd 2/2 mice were infected with 5610 3 , 5610 4 , or 5610 5 CFUs of B. abortus 2308 i.p., and splenic CFUs were determined at seven days post-infection. The mean 6 SD of 4-5 mice/group is shown; * P,0.05 versus wild-type mice infected with the same dose of B. abortus. D. TCRd 2/2 and C57BL/6 mice were infected with 5610 4 CFUs of B. abortus 2308, and splenic colonization was determined 1, 7, and 14 days postinfection. The mean 6 SEM of 4 mice/group except day 7 contains 13 mice/group is shown; * P,0.05 versus wild-type mice infected with B. abortus. doi:10.1371/journal.pone.0021978.g001 and CD8 + T cells were determined at varying time points postinfection when compared to splenic T cells of naïve mice. Approximately a 3-fold increase in the proportion of cd T cells was observed in the spleens of B. abortus-infected C57BL/6 mice, which peaked at two weeks post-infection and occurred concomitantly with diminished brucellae (Table 1; Figure 1A and D). As the total number of splenic mononuclear cells increased dramatically during infection, the increase in the proportion of cd T cells observed was due to an increase in the absolute number of cd T cells, rather than a reduction in the number of other lymphocyte subsets. The proportion of CD4 + T cells also appeared to decrease by day 28 following infection, which could be due to an increase in the proportion of B cells, a phenomenon that has been observed in other murine models in which mice are infected with intracellular bacteria, such as Mycobacterium [19].
cd T cells from B. abortus infected mice produce IL-17 and IFN-c Since cd T cells are enhanced by B. abortus infection, we queried which cytokines are produced. cd T cells (.95% purity) were purified from C57BL/6 mice infected 14 days earlier. This time point was selected since it was the peak of splenic cd T cell induction. For comparison, total T cells (neutralized of cd T cells) were also assayed for cytokine production in parallel with the purified cd T cells. Both T cell subsets were stimulated with ionomycin and PMA for three days ( Figure 2). cd T cells were found to be the major producer of IL-17 during infection and were also found to produce IFN-c. IL-17 production by purified cd T cells from B. abortus-infected mice was also found to be induced by TCRcd stimulation (data not shown). Minimal IL-17 was produced by the enriched TCRab cell fraction; however, this fraction was found to contain elevated levels of IFN-c and IL-6 relative to cd T cells. IL-4, IL-10, and TNF-a were not detected in any cell culture supernatants.
cd T cells do not require IL-17Ra, IFN-c, or GM-CSF to confer protection against systemic B. abortus infection Since cd T cells were found to be the main source of IL-17 during infection, we sought to determine if the protection conferred by cd T cells was IL-17-dependent. Wt and IL-17 receptor deficient (IL-17Ra 2/2 ) mice were neutralized of cd T cells via neutralizing mAb (control mice received normal hamster IgG) and were then infected with B. abortus. Neutralization of cd T cells enhanced the susceptibility of both wt and IL-17Ra 2/2 mice ( Figure 3A and B), and IL-17 receptor deficiency did not impact splenomegaly ( Figure 3A), nor tissue colonization by B. abortus ( Figure 3B) at one week post-infection relative to infected wt mice. Thus, these data show that the protective effect by cd T cells during infection is independent of IL-17. Intracellular cytokine staining revealed that while cd T cells were the main source of IL-17 during infection, the proportion of cells producing IL-17 was actually diminished by B. abortus infection ( Figure S1A), which may explain while IL-17Ra was dispensable for protection.
IFN-c is required for immunity to experimental brucellosis [20][21][22] and also is produced by cd T cells from B. abortus-infected mice. Therefore, wt and IFN-c 2/2 mice (B6 background) were neutralized of cd T cells via neutralizing mAb (control mice received normal hamster IgG) and were then infected with B. abortus ( Figure 3C and D). Deficiency of either IFN-c or cd T cells enhanced susceptibility to colonization by ,10-fold at one-week post-infection, while mice deficient in both IFN-c and cd T cells possessed .300-fold more viable B. abortus in their spleens than did wt mice ( Figure 3D). IFN-c 2/2 mice neutralized of cd T cells prior to infection with B. abortus displayed ruffled fur and a hunched posture (data not shown), while no clinical symptoms of disease were observed in any other treatment group. Thus, the protection conferred by cd T cells was independent of IFN-c. In addition, cd T cell deficiency in wt mice resulted in enhanced splenomegaly and brucellar colonization, while splenomegaly in IFN-c 2/2 mice was similar to that observed wt mice ( Figure 3C and D). GM-CSF has been found by others to be required for protection to intracellular bacterial pathogens [23,24]; however, GM-CSF was not important for protection against B. abortus and was dispensable for cd T cell-mediated protection against colonization and splenomegaly at one week post-infection ( Figure 3E and F).
cd T cell-mediated protection against B. abortus infection is dependent on TNF-a While cd T cells were found not to be a major producer of TNF-a ( Figure 2), others have shown TNF-a is required for protection to B. abortus [25]. Thus, mice were neutralized of cd T cells, TNF-a, or both, prior to infection with B. abortus. Neutralization of either TNF-a or cd T cells from wt mice enhanced susceptibility to B. abortus colonization one week after infection ( Figure 4A and B). However, in vivo neutralization of TNF-a did not exacerbate susceptibility to colonization or splenomegaly in mice neutralized of cd T cells ( Figure 4B), indicating TNF-a is required for cd T cell-mediated protection against B. abortus. Depletion of cd T cells did not appear to affect TNF-a production by splenocytes from B. abortus-infected mice at 4 or 7 days post-infection ( Figure 4C), and stimulation of splenocytes with plate-bound anti-cd TCR mAb did not enhance TNF-a production (data not shown). Also, TNF-a production by splenocytes from naïve and B. abortus-infected mice was similar ( Figure 4C), and intracellular TNF-a levels in splenocytes from both B. abortus-infected and naïve mice were similar as measured by flow cytometry (data not shown). However, increased IFN-c production by splenocytes from B. abortus-infected mice was observed in mice depleted of TNF-a and/or cd T cells ( Figure 4D), suggesting possible compensation by IFN-c. Others have shown that TNF-a preferentially activates cd T cells as measured by upregulation of surface CD69 expression [26]. While B. abortus infection did lead to activation of cd T cells (and NK cells to a lesser extent), this effect was independent of TNF-a at 4 and 7 days post-infection, as mice neutralized of TNF-a in vivo still displayed enhanced CD69 expression by their cd T cells and, in fact, more so than mice simply subjected to B. abortus infection ( Figure 4E and data not shown). In addition, it has been shown that macrophages from TCRd 2/2 mice display impaired TNF-a production when stimulated in vitro with LPS [27]. However, no differences in the ability of wt or TCRd 2/2 peritoneal macrophages to produce TNF-a or control brucellae infection was evident ( Figure 4F and G).

Bovine cd T cells can impair the intramacrophage growth of B. abortus in autologous macrophages
In order to corroborate our findings with murine cd T cells in a bovine model, a co-culture system was used in which bovine macrophages were infected with B. abortus, and then fresh media containing either media only, or media including autologous cd T cells were added to the macrophage-containing wells. Intracellular bacterial burden was then determined at several time points postinfection. We found that at 5 days post-infection, bovine cd T cells could augment the clearance of B. abortus in autologous macrophages. This effect varied depending on the donor of leukocytes, as cd T cells from Calf #1 were protective ( Figure 5A); however, cd T cells from Calf #2 ( Figure 5C) and Calf #3 were not as effective ( Figure 5E). Protection appeared to correlate with IFN-c production, and the addition of cd T cells to macrophage containing wells from all donors resulted in an increase in IFN-c concentration ( Figure 5B, D, and F); however, the addition cd T cells from Calf #1 resulted in the greatest increase in IFN-c production ( Figure 5B). Neutralization of IFN-c in vitro also abrogated the protective effect cd T cells from Calf #1, indicating an IFN-c-dependent mechanism. To assess whether cd T cells could protect against B. abortus in an in vivo adoptive transfer model, Rag-1 2/2 mice were depleted of NK cells and reconstituted with bovine macrophages, macrophages plus autologous cd T cells, or macrophages plus autologous CD4 + T cells prior to infection with B. abortus (leukocytes for adoptive transfer were all derived from Calf #1). When mice were sacrificed seven days later, adoptive transfer of macrophages with cd T cells, but not transfer of macrophages only, or macrophages with CD4 + T cells, resulted in a significant reduction in splenic colonization by brucellae ( Figure 5G).

Bovine cd T cells respond rapidly to B. abortus infection and alter the transcriptional profile of autologous macrophages
To determine the transcriptional responses of protective bovine cd T cells during infection at several time points, non-adherent cd T cells were aspirated from macrophage-containing wells, and RNA was extracted. Following aspiration of cd T cells, the remaining (adherent) macrophages were washed repeatedly to remove any remaining non-adherent cells. This process allowed determination of the transcriptional responses by both cell populations via RT-PCR. After 5 h of co-culture with autologous B. abortus-infected bovine macrophages, cd T cell mRNA transcripts for IL-8, IL-1b, GM-CSF, MIP-1a, and CD25 were up-regulated ( Figure 6A). A slight up-regulation of IL-17 after 5 and 24 h of co-culture was also observed in cd T cells (data not shown). Later in infection, particularly after 72 h of co-culture, cd T cells produced markedly stronger granzyme B, RANTES, IFNc, and CD36 transcripts ( Figure 6A). To corroborate the transcriptional analysis at the protein level, surface expression for CD25 (IL-2Ra) by cd T cells was assayed via flow cytometry. Co-culture of cd T cells with B. abortus-infected macrophages enhanced CD25 expression by cd T cells within 5 h, and by 72 h, the percentage of cd T cells expressing CD25 had increased .15 fold ( Figure 7). The presence of cd T cells also altered the transcriptional profile of macrophages in the co-culture. Bovine macrophages produced elevated IL-6 mRNAs following infection with B. abortus; however, this response was accelerated and amplified by cd T cells ( Figure 6B). cd T cells also enhanced the expression of IL-8 and IL-23p19 by macrophages, particularly, early in infection. IL-8 was also detected within 5 h of cd T cells/ macrophage culture, but was not detected in wells containing macrophages only ( Figure S2). iNOS mRNA expression was only detected at 5 h post-infection, a response unaltered by the presence of cd T cells. Significant changes in mRNA levels for IL-4, IL-10, IL-12p35, IL-12p40, TNF-a, FoxP3, and TGF-b mRNA from cd T cells or macrophages were not observed under the conditions tested (data not shown).

Discussion
With over 500,000 new human cases a year [28], brucellosis is the most common zoonotic infection in the world [29]. The pathological manifestations of brucellosis are diverse and include arthritis, endocarditis, and meningitis in humans, while animal brucellosis is characterized by spontaneous abortion [2]. While abortus infection. C57BL/6 mice were infected i.p. with 5610 4 CFUs of B. abortus 2308, and two weeks later cd T cells (.95% purity) and an enriched TCRab (,55% CD4 + , 25% CD8 + ) cell fraction were isolated from the spleens of infected mice. Cells were stimulated with 500 ng/ml ionomycin and 50 ng/ml PMA for three days, and cell-free supernatants from triplicate wells were assayed for cytokine production via ELISA. The mean 6 SD is shown; * P,0.05 versus the enriched TCRab cells. Results are representative of two independent experiments. doi:10.1371/journal.pone.0021978.g002 much emphasis has been placed on identifying new vaccine candidates for human and animal brucellosis, relatively little attention has been paid to effective coordination of the innate immune response by the host. cd T cells represent a small percentage of circulating T cells in adult humans and mice, [7]; however, cd T cells are a major lymphocyte subset in adult ruminants and make up a majority (up to 70%) of circulating lymphocytes in neonatal calves [8]. cd T cells play an active role in the regulation and resolution of pathogen induced immune responses [30]; however, different subsets of cd T cells may have different functions. Transcriptional analyses of bovine cd T cells suggest that while CD8 2 cd T cells are activated, proliferative, and inflammatory, the CD8 + subset suppressed genes are consistent with quiescence trafficking to the mucosa and immune suppression [7]. While human Vc9d2 T cells exhibit strong cytolytic activity against Brucella-infected cells and are able to impair intracellular growth of B. suis in autologous macrophages [17], this subset of cd T cells is absent in nonhuman primates, and the role of cd T cells in an in vivo model of brucellosis has not been determined.
Here we report mice deficient in cd T cells had impaired innate immunity to B. abortus. The protective role of cd T cells during infection appeared to be temporal, as TCRd 2/2 mice were more susceptible to B. abortus colonization at 7 days post-infection; however, the absence of cd T cells did not impair immunity, particularly, at time points after 2 wks post-infection. Vaccination of mice with RB51 before challenge with wild-type B. abortus 2308 also resulted in reduced colonization in wt and TCRd 2/2 mice, suggesting that the protective function of cd T cells may be limited to innate immunity. In addition, the temporal protection conferred by cd T cells correlates with the theory that cd T cells may be a link between innate and adaptive immunity [31]. Studies in which cd T cells are protective against the organism Listeria monocytogenes [32] have found the magnitude and timing of the cd T cell response are related to the infectious dose of bacteria used, with cd T cell expansion being quicker and of a larger magnitude in affected organs following a higher dose of bacteria [33]. Therefore, as our initial studies used a high dose of B. abortus (5610 5 CFUs), we queried whether the protection mediated by cd T cells in our  abortus between wt and TCRd 2/2 mice was also similar, regardless of whether 5610 4 or 5610 5 CFUs were used as an infectious dose. Notably, regardless of whether TCRd 2/2 mice or wt mice neutralized of cd T cells via mAb (clone UC7-13D5) treatment was used, either model showed similar impairment of immunity to B. abortus, indicating that the effect of cd T cells is not due to a developmental defect in TCRd 2/2 mice. A recent study has shown that treatment of mice with mAb against the TCRcd (such as UC7-13D5) may not actually deplete cells, but rather block TCRcd signaling [34], which according to the results presented here would indicate the protective effect of murine cd T cells against B. abortus requires TCRcd-specific signaling.
To determine if cd T cells expand in response to infection, the lymphocyte composition of spleens from B. abortus-infected C57BL/6 mice was characterized at several time points postinfection. At two weeks post-infection, the proportion of cd T cells amongst lymphocytes in the spleens from infected mice increased nearly three-fold. The total number of lymphocytes in spleens increased following B. abortus infection, indicating the enhanced percentage of cd T cells was due to an increase in the total number of cd T cells rather than a reduction of other T cell subsets. The finding that cd T cells expanded or were recruited to the spleens of infected mice lent further support to the importance of cd T cells in resolving infection by Brucella. The observed spike in splenic cd T cells occurring after the peak of Brucella infection was similar to what has been found in mice infected with Nocardia asteroides or Mycobacterium bovis BCG [9,35]. Different mechanisms have been proposed for the protection conferred by cd T cells in murine models. Klebsiella-infected TCRd 2/2 mice are found to have decreased levels of IFN-c and TNF-a, which has been postulated to result in enhanced susceptibility to infection [10]. cd T cells are the main producer of IL-17 in naïve mice [14], and IL-17 production by cd T cells has been shown to be augmented in response to infection [11,13,14]. In addition, the protection conferred by cd T cells against infection with L. monocytogenes, E. coli, and F. tularensis LVS has been shown to be dependent on IL-17 [11,13,36]. Therefore, the cd T cell cytokine profile was assessed following infection with B. abortus to ascertain the mechanism by which these cells confer protection. Similar to what others have found, cd T cells were the main producer of IL-17 in infected mice. Interestingly, IL-17Ra 2/2 mice were not impaired in their ability to control B. abortus infection, and IL-17Ra signaling was not required for cd T cell-mediated protection. Intracellular cytokine staining revealed that while cd T cells were the major source of IL-17 in both naïve and B. abortusinfected mice, IL-17 production was not enhanced by Brucella infection (Figure S1A-B). This finding explains why IL-17Ra is dispensable for protection against systemic challenge with B. abortus. While IFN-c was produced by purified cd T cells from B. abortus- infected mice, production of IFN-c by an enriched TCRab cell fraction was much more robust. B. abortus-induced IFN-c production by CD4 + T cells was confirmed by intracellular cytokine staining ( Figure S1A-B). Subsequent secretion analysis of cd T cells sorted from naïve and B. abortus infected TCRa 2/2 mice (which have a higher proportion of cd T cells) revealed that B. abortus infection may actually suppress IL-17 and IFN-c production by cd T cells ( Figure S1B). However, it is important to consider the function of cd T cells may be different in TCRa 2/2 mice than in wt mice, and we must not over-interpret these results. The protective effects of IFN-c and cd T cells against B. abortus were found to be independent of each other and, in fact, could be compensatory, as depletion of cd T cells from IFN-c 2/2 mice resulted in a greater impairment of immunity to B. abortus than did depletion of cd T cells from wt mice. Indeed, depletion of TNF-a and/or cd T cells from B. abortus-infected mice was found to enhance IFN-c production by splenocytes. This finding further suggested the host may enhance IFN-c production to compensate for the loss of protection conferred by TNF-a and/or cd T cells and may also indicate the protective effects of IFN-c in C57BL/6 mice may be independent of TNF-a and/or cd T cells.
cd T cells were not found to be a major producer of TNF-a, and neutralization of cd T cells from B. abortus-infected mice did not significantly reduce TNF-a production. Others have shown that macrophages from TCRd 2/2 mice display impaired TNF-a production upon in vitro stimulation [27], yet no differences in the ability of peritoneal macrophages from either wt or TCRd 2/2 mice to control brucellae infection or produce TNF-a were observed. However, TNF-a was required for cd T cells to confer protection against B. abortus infection. While it has been shown that TNF-a preferentially activates cd T cells as measured by up-regulation of surface CD69 expression [26], TNF-a neutralization had no effect upon CD69 expression by cd T cells or NK cells in B. abortus-infected mice. TNF-a is also known to enhance T cell cytotoxicity [37]; thus, future studies will investigate whether cd T cells mediate protection against B. abortus infection via TNF-adependent cytotoxic effects. Interestingly, B. abortus infection enhanced CD69 expression more robustly in cd T cells than in NK cells, which may help explain work by others indicating that NK cells are dispensable for innate immunity to B. abortus [38].
In order to corroborate our murine findings using bovine lymphocytes, a co-culture system was adopted and purified bovine cd T cells were cultured with autologous B. abortus-infected macrophages. Bovine cd T cells could impair the intramacrophage replication of B. abortus by co-cultured macrophages, but this effect varied depending on the donor of cd T cells, which was not entirely surprising, since cattle exhibit wide variability in their immune response to infection [39]. While the addition of cd T cells from all donors to autologous B. abortus-infected macrophages resulted in enhanced IFN-c production in cell culture supernatants, cd T cells only conferred significant protection when high levels (.40 ng/ml) of IFN-c were produced. Neutralization of IFN-c in vitro abrogated the protective effect of bovine cd T cells in our co-culture system, further demonstrating the requirement of IFN-c for bovine cd T cell-mediated protection. IFN-c has been shown to enhance the resistance of bovine macrophages to bacterial infection by inducing apoptosis [40]; therefore, future studies will assess whether cd T cells and IFN-c affect bovine macrophage apoptosis during B. abortus infection.
Transcriptional analyses of bovine co-cultured with Brucellainfected macrophages suggest cd T cells follow a ''priming'' model of activation [41]. Early in infection (,5 h), IL-8, MIP-1a (CCL3), GM-CSF, IL-1b, IL-17, and CD25 mRNAs were upregulated by cd T cells. At later time points after infection (,72 h), cd T cells enhanced expression of granzyme B, RANTES, and IFN-c mRNAs. This biphasic immune response indicates that cd T cells initially respond to infection by expressing signals that lead to direct enhancement of macrophage function, along with an enhanced responsiveness of the cd T cells to further secondary signals, such as antigen or cytokines, while later in infection, cd T cells express signals indicative of an ''effector'' state [41]. In addition, the presence of cd T cells seemed to alter the early mRNA profile of infected macrophages as increased levels of IL-8, IL-6, and IL23p19 were observed in infected macrophages cocultured with cd T cells relative to infected macrophages alone. These cytokines are all considered part of the Th17 response [42], suggesting that the presence of cd T cells may shift the subsequent immune response to infection. TNF-a mRNA was not found to be induced in bovine cells under the conditions tested, which is not surprising as others have found that intravenous or subcutaneous vaccination of cattle with B. abortus RB51 does not increase serum TNF-a levels [43].
The need for model systems to analyze the immune system and pathogenesis of disease in cattle is great since studies in cattle are difficult owing to the limited availability of genetically similar animals and the high costs involved in animal purchase and housing [44]. In vivo bovine studies with virulent Brucella species are particularly difficult to perform due to the limited availability of large animal BSL-3 facilities. Therefore, to determine if bovine cd T cells that were protective in our co-culture system could protect in an in vivo model, Rag-1 2/2 mice were depleted of NK cells and reconstituted with bovine macrophages alone, macrophages plus autologous cd T cells, or macrophages plus autologous CD4 + T cells prior to infection with B. abortus. When mice were sacrificed seven days later, adoptive transfer of macrophages with cd T cells, but not macrophages only nor macrophages with CD4 + T cells, resulted in a significant reduction in splenic colonization by B. abortus. These findings further demonstrate bovine cd T cells can protect against B. abortus infection, and reconstitution of Rag-1 2/2 mice with bovine cells is a viable method, which can be used to assay the role of bovine leukocytes during infection.
In this study, we showed that both murine and bovine cd T cells rapidly responded and could provide protection against, infection with B. abortus. While murine and bovine cd T cells utilized different mechanisms of protection to confer protection against B. abortus, cd T cells from both species were found to produce IFN-c, enhance surface expression of activation markers, and help coordinate the host cytokine response following B. abortus infection. Characterization of the transcriptional profile of bovine cd T cells subsequent to infection revealed a biphasic immune response. Initially, bovine cd T cells expressed signals leading to direct enhancement of macrophage function along with enhanced cd T cells responsiveness to further secondary signals, while later in infection cd T cells expressed signals indicative of an effector state. Collectively, these findings demonstrate cd T cells are important for controlling B. abortus infection and provide insight into the response of both murine and bovine cd T cells in response to infection. In addition, these results indicate immunotherapeutic strategies that target cd T cells could be viable means to augment innate immunity against brucellosis.

Ethics Statement
All animal care and procedures were in accordance with institutional policies for animal health and well-being, and approved by MSU Institutional Animal Care and Use Committee under protocol 38.

Mice
Breeder pairs of cd T cell-deficient (TCRd 2/2 ), TCRabdeficient (TCRa 2/2 ), T and B cell-deficient (Rag-1 2/2 ), and GM-CSF (GM-CSF 2/2 ) deficient mice on a C57BL/6 background were obtained from The Jackson Laboratory (Bar Harbor, ME), and breeder pairs IL-17Ra 2/2 mice on B6 background were a gift from Amgen (Seattle, WA); mice were bred and maintained at the Montana State University Animal Resource Center (Bozeman, MT). C57BL/6 and IFN-c 2/2 mice (B6 background; Jackson Laboratory) were used at 7 to 11 weeks of age. All mice were maintained at Montana State University Animal Resource Center under pathogen-free conditions in individually ventilated cages under HEPA-filtered barrier conditions and were fed sterile food and water ad libitum. In some experiments, cd T cells and/or TNFa were neutralized in vivo via intraperitoneal (i.p.) administration of 0.5 mg of anti-cd TCR mAb (clone UC7-13D5; BioXcell) and/or 0.5 mg of anti-TNF-a mAb (clone XT3.11; BioXcell) on days -1 and 3 post-infection. For challenge studies with B. abortus strain 2308, mice were maintained under similar isolation conditions, in our institutional ABSL-3 facilities. All animal care and procedures were in accordance with institutional policies for animal health and well-being.

Bacterial Strains and Growth Conditions
B. abortus strain 2308 and the B. abortus vaccine strain RB51 were obtained from the National Veterinary Services Laboratory, USDA (Ames, IA). Bacteria were grown under aerobic conditions in potato infusion agar for 72 h (Difco Laboratories) at 37uC and 5% CO 2 . For inoculation, a colony was chosen and incubated overnight at 37uC with shaking in Brucella broth (Difco); the bacterial suspension was adjusted spectrophotometrically to an optical density at 600 nm corresponding to desired inoculum concentration. All experiments with live brucellae were performed in our institutional biosafety level 3 facilities.

Tissue Colony Counts
At selected time points post-infection, spleens and livers were harvested for CFU determinations. Organs were dounce homogenized, and serial 10-fold dilutions in triplicate of homogenates in sterile water were grown on Brucella agar (BA). After incubation for 3 to 5 days at 37uC with 5% CO 2 , Brucella colonies were enumerated, and the number of CFU per tissue was calculated from the dilutions.

Isolation and Infection of Murine Peritoneal Macrophages
Peritoneal macrophages were isolated, as previously described [48,49]. Briefly, mice were given a single i.p. injection of 1.0 ml of expired thioglycolate medium (Difco), and 3 days later, the peritoneum of each mouse was washed with RPMI 1640 (Gibco BRL-Life Technologies [Life Technologies] containing 2% fetal calf serum (Life Technologies) without antibiotics. Peritoneal cells were washed twice in the same medium without antibiotics and allowed to adhere overnight to 24 well microtiter plates. Macrophages (1610 6 /ml) were infected with B. abortus (30 bacteria:1 macrophage) for 1 hour at 37uC/5% CO 2 . Macrophages were then washed with PBS, fresh complete medium (CM): RPMI 1640 medium supplemented with 1 mM sodium pyruvate, 1 mM nonessential amino acids, penicillin/streptomycin (10 U/ ml), and 10% FBS (Atlanta Biologicals) containing 50 mg/ml gentamicin (Sigma-Aldrich) were added, and cells were incubated for 30 min at 37uC/5% CO 2 . After washing twice, fresh CM containing gentamicin (2.5 mg/ml) was added, and cells were incubated 37C/5% CO 2 . At various time points post-infection, the wells were washed three times, macrophages were lysed with sterile water, and intracellular bacterial burden was determined by serial dilution of macrophage lysates on BA.

Cytokine ELISAs
Spleens were aseptically removed from C57BL/6 mice at various time points after challenge with B. abortus 2308. For whole cell cultures, murine splenocytes (5610 6 cells/ml) were cultured following the removal of erythrocytes at for 72 h in CM. To culture purified populations, mononuclear cells were prepared from a Lympholyte M (Cedarlane Labs) gradient, and cd T cells along with an enriched T cell fraction were purified using a cd T cell isolation kit (Miltenyi Biotec), according to the manufacturer's instructions. Cells were cultured at 5610 5 cells/ml in CM alone or in the presence of ionomycin (500 ng/ml)/PMA (50 ng/ml) or plate bound anti-TCRcd mAb (UC7-13D5; 100 mg/ml) for 72 h at 37uC/5%CO 2 . Supernatants were also harvested from murine peritoneal macrophages and bovine macrophages (in some instances co-cultured with autologous cd T cells), which were infected with B. abortus. All supernatants were filtered through a 0.4 mm filter and stored at -80uC. Capture ELISA for IL-4, IL-6, IL-10, IL-17, IFN-c, and TNF-a was used to quantify cytokine levels, as previously described [45,46,49].

Generation of Bovine Monocyte-Derived Macrophages
Whole blood was collected from 6-10 month old Holstein calves into sodium heparin tubes (BD Biosciences). Leukocytes were separated from whole blood using Histopaque-1077 (Sigma-Aldrich) for bovine cells, as previously described [47]. Monocyte-derived macrophages were generated via modification of a previous protocol [50]. PBMCs (1610 7 /ml) were cultured in CM without antibiotics containing 12.5% autologous serum. Cells were allowed to adhere to tissue culture flasks for 2 h at 37uC/5%CO 2 when the flasks were agitated. The cells were then cultured at 37uC/5%CO 2 for 72 h when flasks were agitated again, and the supernatant was aspirated and fresh CM without antibiotics containing 12.5% autologous serum was added. Adherent cells were cultured for an additional 72 h at 37uC/5%CO 2 , when the flasks were again agitated, and the supernatant removed. Adherent cells were then briefly trypsinized and the flasks agitated to dislodge the cells, which were washed in fresh CM without antibiotics containing 10% FBS. Monocyte-derived macrophages were then cultured at 2610 5

Purification of Bovine T Cell Subsets
Bovine PBMCs were isolated as described above and then incubated with biotinylated mAbs against CD4 + T cells (clone CC30) [51] or the TCRcd (clone GD3.8) [52]. Washed cells were incubated with streptavidin microbeads (Miltenyi). Cells were washed again and positively sorted using magnetic LS columns (Miltenyi) according to manufacturer's instructions. The resulting purity was greater than 95%. Purified T cells were allowed to rest overnight and were resuspended in fresh CM the following day immediately prior to being added to wells containing B. abortus-infected macrophages.

Infection of Bovine Macrophage and cd T Cell Co-Cultures
Bovine macrophages (prepared as described above) were infected with B. abortus in a manner similar to that described previously [5]. Macrophages (2610 5 /ml) were infected with B. abortus (30 bacteria:macrophage) for 1 hour at 37uC/5% CO 2 . Macrophages were then washed with PBS, fresh CM containing 50 mg/ml gentamicin (Sigma-Aldrich) was added, and cells were incubated for 30 min at 37uC/5% CO 2 . After washing twice, as described above, fresh CM containing gentamicin (2.5 mg/ml) with or without autologous CD4 + cells or cd T cells (2610 6 /ml prepared as described above,) was added, and cells were incubated 37uC/5% CO 2 . A neutralizing antibovine IFN-c mAb (clone CC302, Serotec, 50 ng/ml: [53]) was also added to some wells. At various time points post-infection, the wells were washed three times, macrophages were lysed with sterile water, and intracellular bacterial burden was determined by serial dilutions of macrophage lysates on BA. The nonadherent T cells were aspirated from the co-culture, centrifuged, and stored in Tri Reagent (Sigma-Aldrich) for RNA extraction. In some instances, after the aspiration of T cells (or CM only) and three washes with PBS, adherent macrophages were lysed in Tri Reagent for RNA extraction. Supernatants were also harvested and 0.4 mm-filtered and stored at -80uC to determine IFN-c levels using mAbs from ABD Serotec: the capture mAb was clone CC330, and the detecting mAb was clone CC302. IL-8 concentrations were also measured using the anti-human CXCL-8/IL-8 DuoSet Kit from R&D Systems, which cross-reacts with bovine IL-8.

Reconstitution of Rag-1 2/2 Mice with Bovine Leukocytes
Rag-1 2/2 mice were depleted of NK cells via i.p. administration of 0.5 mg of an anti-NK1.1 mAb (clone PK136, BioXCell) on days -2 and day 3 post-infection. One day before infection, PBS, macrophages (5610 5 /mouse) only or macrophages plus T cells (1610 7 /mouse) were adoptively transferred into mice intraperitoneally. Macrophages and T cells were purified, as described above. One day later, mice were infected i.p. with 1610 4 B. abortus 2308. Splenic colonization was determined seven days later.

RNA Extraction and RT-PCR Analysis
RNA was extracted from cells in Tri Reagent according to manufacturer's guidelines. RNA was then further purified via extraction with an RNeasy Mini Kit (Qiagen). cDNA was generated using the Superscript III First Strand Synthesis System (Invitrogen). Primers for immune-related genes along with b-actin (endogenous control) were designed using the PrimerQuest application from IDTDNA.com and purchased from IDT. Amplicons were visualized under UV illumination on a 2% agarose gel containing GelRed (Biotium).

Statistical Analysis
The Student t test was used to evaluate the differences in colonization, splenic weights, cytokine production, intracellular bacterial burden, and lymphocyte populations when two groups were compared. When more than two groups were compared, an ANOVA followed by Tukey's test was used. A p value of ,0.05 was considered significant. Figure S1 B. abortus infection does not induce IL-17 or IFN-c production by cd T cells. A. Splenocytes from naïve or B. abortus-infected mice (7 dpi) were stimulated overnight with PMA/Ionomycin and brefeldin A was added for the last 3 h of culture. Following surface staining, cells were permeabilized and stained for intracellular IL-17 or IFN-c. Top panel, the proportion of IL-17 producing cd T cells was determined following gating on lymphocytes. Second panel from top, cells were gated on CD4 + (CD3 + ) T cells and assayed for IL-17 production. Third panel from top, cells were gated on cd T cells (CD3 + /TCR cd + ) and assayed for IFN-c production. Bottom panel, cells were gated on CD4 + (CD3 + ) T cells and assayed for IFN-c production. Depicted is the mean 6 SD of 5 mice/group and is representative of two independent experiments. B. cd T cells were sorted from naïve or B. abortus-infected (7 dpi) mice and stimulated for 72 h with PMA/ Ionomycin. Cytokine levels in supernatant were determined by ELISA. Depicted is the mean 6 SD of triplicate wells. *P,0.05 versus cytokine production by cd T cells from naïve mice. (TIF) Figure S2 IL-8 production is augmented by cd T cells when co-cultured with bovine macrophage during infection with B. abortus. IL-8 concentrations were measured by ELISA in supernatants from B. abortus-infected bovine macrophages cultured with or without autologous T cells at various time points after infection. Data depict the mean 6 S.D. of triplicate measurements/group. *P,0.05 versus wells containing macrophages only. (TIF)