IFNγ and Antibody Synergize To Enhance Protective Immunity against Chlamydia Dissemination and Female Reproductive Tract Reinfections

ABSTRACT

CD4 T cell-dependent IFNγ production and antibody are the two best known effectors for protective immunity against Chlamydia female reproductive tract (FRT) infection. Nevertheless, mice lacking either IFNγ or B cells can clear the vast majority of Chlamydia from the FRT, while suffering from varying degrees of disseminated infection. In this study, we investigated whether IFNγ and B cells play complementary roles in host defense against Chlamydia and evaluated their relative contributions in systemic and mucosal tissues. Using mice deficient in both IFNγ and B cells (IFNγ^−/− x μMT), we showed that mice lacking both effectors were highly susceptible to lethal systemic bacterial dissemination following Chlamydia muridarum intravaginal infection. Passive transfer of immune convalescent serum, but not recombinant IFNγ, reduced bacterial burden in both systemic and mucosal tissues in IFNγ^−/− x μMT mice. Notably, over the course of primary infection, we observed a reduction of bacterial shedding of more than 2 orders of magnitude in IFNγ^−/− x μMT mice following both C. muridarum and C. trachomatis FRT infections. In contrast, no protective immunity against C. muridarum reinfection was detected in the absence of IFNγ and B cells. Together, our results suggest that IFNγ and B cells synergize to combat systemic Chlamydia dissemination, while additional IFNγ and B cell-independent mechanisms exist for host resistance to Chlamydia in the lower FRT.

INTRODUCTION

The obligate intracellular bacterium Chlamydia causes a variety of human and animal diseases by invading multiple mucosal tissues. Depending on the strains and serovars, Chlamydia may target the epithelium of the eyes, the respiratory tract, the reproductive tract, or further invade the local draining lymph nodes and cause systemic infections. Chlamydia trachomatis is the etiological agent for the most prevalent sexually transmitted infection with nearly 130 million cases worldwide annually (1). C. trachomatis infections cause major public health concerns due to the severe disease sequelae, such as pelvic inflammatory disease, ectopic pregnancy, and infertility. In the past 50 years, only one Chlamydia vaccine candidate successfully proceeded to a Phase I human clinical trial (2), and there is no human Chlamydia vaccine available to date.

Chlamydia muridarum is a murine pathogen that shares over 98% sequence similarity with the human strain C. trachomatis (3). Although Chlamydia infections can be spontaneously resolved in immunocompetent hosts, C. muridarum and C. trachomatis can naturally ascend to the upper female reproductive tract (FRT) and cause analogous immune-mediated pathology in mouse and human, respectively. Protective immunity against Chlamydia is conferred primarily by CD4 cells. The importance of CD4 Th1 cells in Chlamydia resistance is manifested by varying degrees of defects in bacterial control in MHCII-, TCRα/β-, interleukin 12-, IFNγ-, and iNOS-deficient mice following C. muridarum intravaginal infection (4–7). Specifically, mice deficient in MHCII or TCRα/β exhibit long-term, high levels of Chlamydia shedding from the FRT, demonstrating an absolute requirement of CD4 T cells at the site of infection (4, 5). In contrast, IFNγ-deficient mice are able to clear majority of C. muridarum from the FRT, followed by a chronic phase of low-grade bacterial shedding, and a proportion of these mice succumb to disseminated infection (5, 6). The distinct phenotypes in T cell-deficient and IFNγ-deficient mouse models suggest that while CD4 T cell responses are indispensable, host resistance to Chlamydia does not strictly rely on Th1-dependent IFNγ-production.

Previous studies showed that convalescent antibodies are protective against Chlamydia reinfections, but B cells do not appear to participate in Chlamydia primary clearance given that identical kinetics of bacterial shedding was observed in WT and B cell-deficient (μMT) mice (8, 9). More recently, we reported that both μMT and antibody-deficient mice (AID^−/− x μS^−/−) experience transient systemic Chlamydia dissemination at the early stage of FRT infection, with concomitant increase in Chlamydia-specific CD4 T cell expansion in secondary lymphoid organs (10, 11). Therefore, while B cells are dispensable for bacterial clearance from the FRT, they are essential for Chlamydia containment within the local mucosa. Whether Chlamydia-specific antibodies are detrimental or beneficial in human Chlamydia infection remains a topic of debate. High anti-Chlamydia antibody titers appear to correlate with severe immunopathology in women (12), yet both could presumably be consequences of repeated infections. Given that causal relationships among antibody levels, immunopathology and protective immunity are difficult to establish in human infection, animal models serve as valuable tools for understanding the contribution of B cells and antibodies to Chlamydia resistance.

Although IFNγ and antibody have long been recognized as the two major players in anti-Chlamydia immunity, neither seems to be absolutely required for bacterial resistance in the FRT. It is conceivable that the relatively mild defects in bacterial clearance in IFNγ- and B cell-deficient mice are due to functional redundancy between these two effectors. To test this possibility, and also seek the potential synergistic effects between IFNγ and B cells, we generated mice deficient in both effectors (IFNγ^−/− x μMT) and examined their susceptibility to Chlamydia FRT infection. The contributions of IFNγ and B cells during Chlamydia secondary challenge were also investigated.

RESULTS

IFNγ and B cell double knockout mice succumb to lethal disseminated C. muridarum infection.

Previous studies have demonstrated that IFNγ^−/− mice were highly susceptible to Chlamydia dissemination with mortality rates ranging from ∼30% to 100%, following C. muridarum intravaginal infections (5, 6, 13). In contrast, B cell-deficient μMT mice experience only transient, non-lethal Chlamydia dissemination at the early stage of C. muridarum FRT infection (10). To determine whether IFNγ and B cells elicit complementary effector functions in Chlamydia containment within the FRT mucosa, we generated IFNγ and B cell double deficient mice (IFNγ^−/− x μMT), and infected these mice intravaginally with C. muridarum (strain Nigg II). We observed that IFNγ^−/− x μMT mice succumb to infection quickly, with a median survival of 13 days (range 12 to 15 days) (Fig. 1A). Although 100% of IFNγ^−/− mice also succumb to C. muridarum infection, these mice survived much longer than IFNγ^−/− x μMT mice, with a median survival of 27 days (range 15 to 45 days) (Fig. 1A). To examine the kinetics and extent of bacterial dissemination in IFNγ^−/− x μMT mice, we harvested various tissues from WT and IFNγ^−/− x μMT mice at 4, 8, and 12 days postinfection (dpi) and measured systemic bacterial burdens. Although no significant difference was observed at day 4-postinfection (Fig. 1B), a trend of increase in bacterial burden in systemic tissues, including spleen, intraperitoneal cavity (IP lavage), lungs, and kidneys, was evident in IFNγ^−/− x μMT mice at day 8 (Fig. 1B). By day 12-postinfection, high bacterial burdens of > 10⁶ C. muridarum were detected in both FRT mucosal and systemic tissues of IFNγ^−/− x μMT mice, and these mice quickly succumb to infection after this time point (Fig. 1A and B). These results suggest that mice deficient in both IFNγ and B cells are significantly more susceptible to bacterial dissemination than IFNγ or B cell single deficient mice following C. muridarum intravaginal infection.

FIG 1.

Mice deficient in both IFNγ and B cells are highly susceptible to lethal disseminated C. muridarum intravaginal infection. C57BL/6 (B6), IFNγ^−/−, μMT, and IFNγ^−/− x μMT mice were infected intravaginally with 1 × 10⁵ C. muridarum. (A) Survival curve following infection. (B) Bacterial burdens in lower female reproductive tract (FRT), upper FRT, spleen, intraperitoneal (IP) lavage, lungs, and kidneys on 4-, 8-, and 12- days postinfection (dpi) as measured by IFU assay. Each data point represents an individual mouse. Data shown are pooled results of 2 independent experiments. Data are mean ± SEM, *, P < 0.05; ***, P < 0.001; ****, P < 0.0001 as calculated using Log-rank test (A) and Mann-Whitney U Test (B).

Immune serum and exogenous IFNγ partially restore C. muridarum resistance in IFNγ^−/− x μMT mice.

During primary infection, immune serum transfer was sufficient to revert C. muridarum systemic dissemination in B cell-deficient μMT mice (11). We therefore asked whether replenishing IFNγ^−/− x μMT mice with Chlamydia-specific antibody and/or exogenous IFNγ could restore efficient control of systemic infection in these mice. μMT, IFNγ^−/− and IFNγ^−/− x μMT mice were treated with immune convalescent-phase serum and/or recombinant mouse IFNγ (rIFNγ) over the first 11 days of C. muridarum intravaginal infection, and tissues were harvested at day 12 for bacterial burdens (Fig. 2A). Consistent with our previous findings, transfer of immune serum reduced bacterial burden in μMT mice in systemic tissues including spleen, IP cavity, lungs, and kidneys, but not in the upper or lower FRT mucosa (Fig. 2B). Likewise, replenishing circulating IFNγ, but not immune serum, to IFNγ^−/− mice resulted in small but significant reductions in bacterial burden in IP cavity, lungs, and kidneys (Fig. 2B). Notably, in IFNγ^−/− x μMT mice, treatment of rIFNγ had no detectable effect on bacterial burdens, while transfer of immune serum was able to significantly reduce bacterial counts to levels comparable to IFNγ^−/− mice in both systemic and mucosal tissues. Lastly, treating IFNγ^−/− x μMT mice with rIFNγ in conjunction with immune serum had no additive effect in reducing bacteremia (Fig. 2B). Together, these data suggest that IFNγ and antibody likely elicit distinct effector functions that do not compensate for the loss of one another in host resistance to Chlamydia dissemination.

FIG 2.

Immune serum, but not systemic IFNγ treatment, reduces tissue bacterial burden in IFNγ^−/− x μMT mice. B6, IFNγ^−/−, μMT, and IFNγ^−/− x μMT mice were infected intravaginally with 1 × 10⁵ C. muridarum. Cohorts of mice were treated systemically with convalescent immune sera or recombinant mouse IFNγ (rIFNγ) throughout the course of infection. (A) Schematic depicting timeline for infection, immune sera and/or rIFNγ treatment, and tissue harvest. (B) Bacterial burdens of lower FRT, upper FRT, spleen, IP lavage, lungs and kidneys on day 12-postinfection as measured by IFU assay. Each data point represents an individual mouse. Data shown are pooled results of 3 independent experiments. Data are mean ± SEM, *, P < 0.05; **, P < 0.01; ***, P < 0.001 as calculated using Mann-Whitney U Test (B6 and μMT groups) and One-way ANOVA (IFNγ^−/− and IFNγ^−/− x μMT groups).

Chlamydia shedding from the FRT is reduced in the absence of both IFNγ and B cells.

We next determined whether IFNγ and B cells are absolutely required for reducing Chlamydia shedding from the lower FRT. It has been well documented that IFNγ^−/− mice are capable of clearing >99% of C. muridarum from the lower FRT while not capable of completely eradicating the bacteria (5, 6, 13). On the other hand, B cell-deficient μMT mice clear C. muridarum from the FRT with the same efficiency as WT controls (8, 10). Since natural resolution of C. muridarum from WT mice takes 4 to 5 weeks, the early death of IFNγ^−/− x μMT mice from disseminated infection in previous experiments prevented us from evaluating the full kinetics of bacterial shedding. To overcome this limitation, we infected cohorts of WT and IFNγ^−/− x μMT mice with 3 commonly used laboratory strains of C. muridarum (Nigg II, Weiss and Nigg) exhibiting a spectrum of virulence (14). As suggested by survival experiments in Fig. 3A, C. muridarum strain Weiss showed slightly reduced virulence compared to C. muridarum Nigg II, and C. muridarum Nigg appeared to be the least virulent strain in this infection model. Strikingly, when bacterial shedding was monitored by vaginal swabs in surviving mice, significant reductions were observed in both WT and IFNγ^−/− x μMT mice between days 7 and 14, regardless of C. muridarum strain used (Fig. 3B to D). For the 5 of 12 IFNγ^−/− x μMT mice that survived C. muridarum Nigg infection for over 26 days, bacterial burdens were more than 2 orders of magnitude lower on days 26 to 28 compared to day 7 (Fig. 3D). These results demonstrated that although IFNγ and B cells are required for bacterial eradication from the lower FRT, mice are still capable of significantly reducing C. muridarum burdens in the absence of both effectors, indicating that alternative protective mechanisms exist for mucosal defense against Chlamydia in the lower FRT mucosa.

FIG 3.

C. muridarum shedding from the FRT is reduced in the absence of both IFNγ and B cells. Cohorts of B6 and IFNγ^−/− x μMT mice were infected intravaginally with various C. muridarum strains (Nigg II, Weiss or Nigg) at 1 × 10⁵ IFU. (A) Survival curve following infection. (B) to (D) Bacterial shedding from the lower FRT as measured by weekly vaginal swabs and IFU assay. Numbers above the bars depicts the number of surviving mice at the time point/total number of mice in the group. Each data point represents an individual mouse. Data shown are pooled results of 2 independent experiments. Data are mean ± SEM, *, P < 0.05; **, P < 0.01; ***, P < 0.001 as calculated using Log-rank test (A) and Mann-Whitney U Test (B) to (D). ND, not determined.

B cells are dispensable for C. trachomatis primary clearance in mice.

Intravaginal inoculation of human C. trachomatis in mice leads to spontaneous clearance in the absence of adaptive immunity (15). In contrast, direct deposit of C. trachomatis into the upper FRT requires CD4 T cells for clearance (16). Thus, intrauterine C. trachomatis infection serves as another useful model to investigate the necessity of immune components in host resistance to Chlamydia. To investigate whether loss of both IFNγ and B cells affects host resistance against C. trachomatis, we performed transcervical infection of WT, μMT, IFNγ^−/− and IFNγ^−/− x μMT mice with 10⁶ C. trachomatis serovar D, and monitored bacterial shedding by vaginal swabs. WT and μMT mice quickly cleared C. trachomatis from FRT with very low numbers of C. trachomatis recovered by vaginal swabs at all time points. In contrast, IFNγ^−/− and IFNγ^−/− x μMT mice experienced prolonged bacterial shedding with similar kinetics over the first 100 days (Fig. 4A). Although total bacteria shed from IFNγ^−/− x μMT mice appeared to be slightly more than IFNγ^−/− mice, differences between these 2 strains were not statistically significant (Fig. 4B). These observations supported previous findings that IFNγ is the predominant effector for C. trachomatis control in the murine model, whereas B cells play a minimal role. Curiously, bacterial burdens in both IFNγ^−/− and IFNγ^−/− x μMT peaked around 14 dpi at ∼10⁴ IFU and diminished to ∼10 IFU around day 80, indicating that additional effectors other than IFNγ and B cells also exist for protective immunity against C. trachomatis.

FIG 4.

Susceptibility to C. trachomatis transcervical infection is not altered by B cell deficiency. B6, IFNγ^−/− and IFNγ^−/− x μMT mice were infected transcervically with 1 × 10⁶ C. trachomatis serovar D. (A) Bacterial shedding from the lower FRT was measured for the first 100 days of primary infection. (B) Area under the curve of the IFNγ^−/− and IFNγ^−/− x μMT curves in (A). Data shown are combined results of 2 independent experiments with 6–12 mice per group. Error bars represent the mean ± SEM, ****, P < 0.0001 as calculated using Two-way ANOVA (mixed-effects analysis) (A) and un-paired t test (B).

IFNγ and B cell double deficiency results in complete loss of protective immunity against C. muridarum secondary infection.

In contrast to the absolute requirement of CD4 T cells during Chlamydia primary infections, either CD4 T cell or antibody can facilitate secondary Chlamydia clearance in mice (9). We next asked whether IFNγ and B cells play complementary roles in host resistance to C. muridarum reinfections, as IFNγ and B cell single deficient mice do not exhibit any major defect in secondary clearance (8, 17). To do this, we first infected μMT mice intravaginally with C. muridarum and allowed infections to resolve spontaneously. We then rechallenged these μMT memory mice with or without antibody (Ab) treatment to either neutralize IFNγ, IL-17A or deplete CD4 T cells (Fig. 5A). As shown in Fig. 5B, μMT memory mice exhibit marked protection against reinfection as bacterial burdens in isotype control treated group were close to 10,000-fold lower than naive μMT mice at both day 3 and day 9. While anti-IL-17A treatment had no detectable effect on bacterial burden in μMT memory mice, either IFNγ neutralization or CD4 T cell depletion resulted in complete loss of protective immunity, and bacterial burdens in these mice were comparable to unimmunized naive μMT controls. Additionally, neutralizing IL-17A in conjunction with IFNγ had no additive effect on bacterial burden. Taken together, we concluded that, unlike in primary infection, IFNγ and B cells are the two immune effectors required for protective immunity against C. muridarum reinfection in mice.

FIG 5.

Complete loss of protective immunity against C. muridarum secondary infection in the absence of IFNγ and B cells. Immune μMT mice were reinfected with 1 × 10⁵ C. muridarum with or without depleting antibody treatment. (A) Schematic depicting timeline for 1° and 2° infections, antibody (Ab) treatment, and sampling. (B) Bacterial shedding from the lower FRT as measured by vaginal swabs on days 3 and 9 and IFU assay. Data shown are combined results of four independent experiments with 6 to 12 mice per treatment group. Each data point represents individual mouse. Error bars represent the mean ± SEM, ***, P < 0.001 when compared to isotype control group, ^###,P < 0.001 when compared to naive μMT group. Data were analyzed using Two-way ANOVA followed by multiple Mann-Whitney U Tests.

DISCUSSION

Understanding host defense mechanisms against Chlamydia is an essential step for developing a much-needed Chlamydia vaccine. To date, IFNγ and antibody are the 2 best known protective immune effectors in Chlamydia immunity (18). The major defects in both IFNγ- and B cell-deficient mice following C. muridarum intravaginal infection, however, are disseminated bacterial infections, rather than uncontrolled Chlamydia shedding from the FRT (5, 6, 10). It is, therefore, our speculation that IFNγ and antibody may compensate for the loss of one another in the FRT during Chlamydia infection. Using IFNγ and B cell double deficient mice (IFNγ^−/− x μMT), we showed definitively in this study that the major synergistic effect of IFNγ and antibody is to prevent lethal bacterial dissemination during C. muridarum primary infection, while additional immune effector(s) exist for Chlamydia control in the lower FRT. In contrast, deficiency in both effectors results in a complete loss of protective immunity against C. muridarum secondary challenge.

Following intravaginal inoculation, IFNγ^−/− x μMT mice exhibit a gradual increase of systemic Chlamydia burden over the first 12 days which leads to lethality. This phenotype closely resembles those of the Rag2^−/−γc^−/− mice reported recently (13). Given that Rag2^−/−γc^−/− mice are highly defective in both innate and adaptive immunity, these results suggest that IFNγ and B cells are the 2 immune components necessary and sufficient to control Chlamydia dissemination. A prior study by Poston et al. showed that Rag1^−/− mice succumb to Chlamydia dissemination following intravaginal infection with a lethal clone of C. muridarum Nigg (CM001). Transfer of immune serum extended the survival time of Rag1^−/− mice, and transfer of naive B cells was able to completely rescue the lethal phenotype. These observations led to their conclusion that T cell-independent IFNγ (from Rag1^−/− mice) and B cells cooperate to prevent mortality associated with C. muridarum infection (19). Our study corroborates their findings and provides direct evidence that IFNγ and antibody synergize to prevent C. muridarum dissemination. The non-redundant roles of IFNγ and antibody is manifested by: (i) partial defects (survival and bacterial burden) observed in single knockout animals compared to IFNγ^−/− x μMT double deficient mice; (ii) transfer of rIFNγ, but not immune serum, reduces systemic bacteria in IFNγ^−/− mice; (iii) transfer of immune serum completely eradicates systemic bacteria in μMT mice, whereas the same treatment only partially reduces bacterial counts in IFNγ^−/− x μMT mice. It is noteworthy that replenishing rIFNγ intraperitoneally only slightly reduced Chlamydia burdens in IFNγ^−/− mice, and no reduction was detected in IFNγ^−/− x μMT mice (Fig. 2). Similar observations were made by Williams et al. that while IFNγ depletion in immunocompetent Nu/+ mice increased mortality and bacterial burden in the lung following MoPn (C. muridarum) intranasal infection, IFNγ repletion in immunocompromised Nude mice failed to confer any reproducible protection (20). The suboptimal effect of systemic IFNγ treatment suggests that the effector function of IFNγ may require a local cytokine gradient and/or cognate interaction between IFNγ-producing and -responding cells (21). Supporting this notion, IFNγ is detected in genital secretion but not in serum or spleen after intravaginal Chlamydia infection (22) (data not shown). Intriguingly, several recent studies by Zhong and colleagues demonstrated that IFNγ may be derived from different cell types in order to combat Chlamydia at different mucosal tissues, highlighting the necessity of more nuanced investigation of discrete roles of IFNγ at various stages of Chlamydia infection (23–25). In contrast to IFNγ treatment, we observed that immune serum was able to completely eradicate systemic bacteria from μMT mice. This is consistent with our previous reports that antibody is essential for bacterial containment within the FRT (10, 11). Curiously, small but statistically significant reductions in FRT bacterial burdens were observed in IFNγ^−/− x μMT mice after immune serum treatment, especially in the upper FRT (Fig. 2). We speculate that the hyper-inflammatory environment in IFNγ^−/− x μMT mice may cause tissue leakage at both upper and lower FRT that facilitates diffusion of protective antibodies into the tissue without the need of transcytosis. Future experiments will be needed to test this possibility.

Examining the full time course of infection in IFNγ^−/− x μMT mice was not possible due to the lethal dissemination phenotype following C. muridarum Nigg II intravaginal infection. Using 3 C. muridarum strains (Nigg II, Weiss and Nigg) with varying degrees of virulence, we showed that IFNγ^−/− x μMT mice were able to partially reduce bacterial burdens in the lower FRT, especially during the first 14 days of infection. It is evident that this bacterial control mechanism is IFNγ- and B cell-independent, but likely to be CD4 T-cell dependent, since T cell-deficient mice constantly shed high levels of C. muridarum without any reduction for at least 50 days (5, 13, 19), and the time frame between days 7 to 14 also coincide with robust clonal expansion, homing and accumulation of Chlamydia-specific CD4 T cells in the FRT (10). In parallel to our findings, a recent study by McSorley and colleagues demonstrated that T-bet/Th1 cells are dispensable for primary clearance of C. muridarum from the FRT, although a compensatory effector pathway for protective immunity remains to be discovered (26). It is imperative to note that since IFNγ^−/− x μMT mice failed to completely eradicate bacteria from the FRT, IFNγ and B cells remain to be the two key effectors for Chlamydia clearance in mice.

Despite the remarkable genetic similarities between C. muridarum and C. trachomatis, these two species exhibit strong host tropism during infections (27). In contrast to the results observed by Williams et al. (20), early studies by Zhong and de la Maza showed that recombinant IFNγ was highly effective in reducing systemic bacterial burden following C. trachomatis serovar L1 intravenous infection in mice (28). The differential sensitivities of C. muridarum and C. trachomatis to murine IFNγ were later reconciled by delineating the distinct IFNγ evasion strategies used by these two strains in their respective hosts. While C. trachomatis is capable of salvaging indole to evade human IFNγ induced tryptophan deprivation, it is susceptible to murine IFNγ induced immunity-related GTPases, thereby quickly eliminated from IFNγ-competent mice (29–32). IFNγ^−/− and IFNγR^−/− mice manifest prolonged bacterial shedding for at least 50 days following C. trachomatis FRT infection (27, 33). Notably, antibody responses in these mice were comparable to, if not higher than their WT counterparts, despite the fact that augmented Ab responses afford no protective immunity against secondary challenge (33). By comparing the kinetics of bacterial shedding in B cell-deficient mice with or without IFNγ (μMT and IFNγ^−/− x μMT), we showed in this current study that IFNγ is the dominant effector for C. trachomatis control in mice, and B cells/antibodies do not contribute to the reduced C. trachomatis burdens at the late stage of primary infection. Therefore, much like C. muridarum infection in mice, a CD4 T cell-dependent, but IFNγ-independent effector must be responsible for reducing C. trachomatis burdens in IFNγ^−/− and IFNγ^−/− x μMT mice. Unfortunately, little information is available to date concerning alternative protective mechanisms. We argue that identification of such mechanisms is pivotal since the IFNγ-dependent host tropism of C. muridarum and C. trachomatis strongly suggests that IFNγ resistance is a natural outcome of host-pathogen coevolution, therefore an IFNγ-independent protective mechanism would be essential for the host to combat emerging IFNγ-resistant Chlamydia strains.

C. muridarum primary infections in mice generates robust natural immunity against secondary challenges, providing a valuable system for interrogating immune effector pathways essential for host protection. A major difference between protective responses to Chlamydia primary and secondary infections is whether antibody can confer protection in the absence of CD4 T cells (34). Recent studies by Morrison and colleagues highlighted that the effector function of antibody relies on host IFNγ production and/or presence of neutrophils (17, 35). Our data add to their findings by showing that in the absence of B cells/antibody, IFNγ is also pivotal for protection. Anti-IFNγ treatment in μMT mice results in a complete loss of protective immunity, a phenotype that mirrors the effect of anti-CD4 treatment (Fig. 5B). Thus, IFNγ likely represents the key effector cytokine produced by memory CD4 T cells to combat reinfection. An alternative, but not mutually exclusive explanation is that IFNγ produced by innate cell types is essential for inducing effective antigen presentation and/or polarization/recruitment of memory CD4 T cells. A more detailed investigation is needed to dissect the precise mechanism. It is worth noting that although a protective role of Th17 cells was reported in BALB/c mouse models following C. muridarum intravaginal infection (36), and a shift to Th17 response was observed in mice lacking classical Th1 responses (26), we found no difference in bacterial burden following anti-IL-17A treatment in the presence or absence of IFNγ, indicating that IL-17A mediated signals are neither necessary nor sufficient for protective immunity against secondary infections in this setting.

The search for immune protective mechanisms against Chlamydia was impeded by lack of major phenotypes in majority of gene-deficient mouse models following Chlamydia FRT infection (34). On the contrary, redundancy of immune effectors in protective immunity is beneficial for the hosts, which may explain why Chlamydia infection in human, albeit prevalent, are mostly asymptomatic and spontaneously resolved. Our study highlighted the need for further interrogation of the interplay between known immune effectors, and exploring unknown immune components beyond IFNγ and antibody. A deeper understanding of the differences between the systemic versus mucosa, and primary versus secondary immune responses to Chlamydia is essential for developing a protective Chlamydia vaccine in humans.

MATERIALS AND METHODS

Mice.

C57BL/6 (B6), IFNγ^−/− (B6.129S7-Ifng^tm1Ts/J), and μMT (B6.129S2-Ighm^tm1Cgn/J) mice were purchased from The Jackson Laboratory (Bar Harbor, ME). IFNγ^−/− x μMT mice were generated by crossing IFNγ^−/− mice with μMT mice. All mice used for experiments were 6 to 24 weeks old. Mice were maintained under SPF conditions and all mouse experiments were approved by University of Arkansas for Medical Sciences Institutional Animal Care and Use Committee (IACUC).

Bacteria.

C. muridarum strain Nigg II was purchased from ATCC (VR-123). C. muridarum strain Weiss and C. trachomatis serovar D strain UW-3/Cx were kind gifts from Dr. Richard Morrison (UAMS). C. muridarum strain Nigg was a kind gift from Drs. Roger Rank and V. Laxmi Yeruva (UAMS). Chlamydia strains were propagated in McCoy cells or HeLa 229 cells, elementary bodies (EBs) purified by discontinuous density gradient centrifugation and titrated on HeLa229 cells as previously described (10).

Chlamydia infection and enumeration.

Mice were synchronized for estrus by subcutaneous injection of 2.5 mg medroxyprogesterone (Depo-provera) 5 to 7 days prior to intravaginal infection. For intravaginal infection, 1 × 10⁵ C. muridarum in SPG buffer were deposited directly into the vaginal vault using a pipet tip. For transcervical infection, 1 × 10⁶ C. trachomatis in SPG buffer were inoculated directly into the upper FRT using a NEST device as previous described (16). To enumerate bacterial shedding from the lower FRT, vaginal swabs were collected, suspended in SPG buffer, and disrupted with glass beads. Inclusion forming units (IFUs) were determined by plating serial dilutions of swab samples on HeLa 229 cells, staining with anti-MOMP MAb and enumerated microscopically. To enumerate bacteria burden within tissues, intraperitoneal (IP) lavage was collected in SPG buffer, upper FRT (ovaries, oviducts, and upper 1/3 of uterine horn), lower FRT (vagina, cervix, and lower 1/3 of uterine horn), spleen, kidney and lung were homogenized in SPG buffer. Tissue homogenates were disrupted with glass beads, centrifuged at 500 g for 10 min, supernatants collected, and serial dilutions plated on HeLa 229 cells for IFU counts.

In vivo immune serum, recombinant cytokine and MAb treatment.

Immune convalescent-phase serum was collected from B6 mice >90 days after C. muridarum Nigg II intravaginal infection. Sera were pooled, sterile filtered and aliquoted before use in vivo treatment. Each mouse received 50 μL of immune serum by IP injection once every 4 days, starting 1 day prior to infection. For recombinant mouse IFNγ (rIFNγ) treatment, each mouse was injected IP with 1 μg rIFNγ (PeproTech) in 500 μL PBS every other day beginning 1 day prior to infection. Antibodies used for in vivo treatment, including anti-IFNγ (clone XMG1.2, neutralizing), anti-IL-17A (clone 17F3, neutralizing), and anti-CD4 (clone GK1.5, depleting), were obtained from BioXcell. Antibody treatment was performed by IP injection of 0.25 mg of each antibody on days -1, 4, and 7 after secondary challenge.

Statistical analysis.

Statistical analysis was performed with GraphPad Prism 9. Log-rank Mantel-Cox test was used for survival curves; unpaired t test was used for normally distributed continuous-variable comparisons; Mann-Whitney U test was used for nonparametric comparisons; Repeated measures two-way ANOVA was used to compare bacterial shedding over time.