Early Colonization of the Upper Genital Tract by Chlamydia muridarum Is Associated with Enhanced Inflammation Later in Infection

Chlamydia trachomatis is the most commonly reported bacterial sexually transmitted infection in the United States. Modeling infection in animals can be challenging, as mice naturally clear C. trachomatis when it is deposited in the lower genital tract. However, C. trachomatis can productively infect mice when the lower genital tract is bypassed and bacteria are deposited directly into the upper genital tract via transcervical inoculation.

ABSTRACT

Chlamydia trachomatis is the most commonly reported bacterial sexually transmitted infection in the United States. Modeling infection in animals can be challenging, as mice naturally clear C. trachomatis when it is deposited in the lower genital tract. However, C. trachomatis can productively infect mice when the lower genital tract is bypassed and bacteria are deposited directly into the upper genital tract via transcervical inoculation. Interestingly, the mouse-adapted Chlamydia species C. muridarum can infect mice both by transcervical inoculation and by natural ascension if introduced into the vaginal vault. In this study, we investigated whether the route of infection plays a role in the downstream immune responses to C. muridarum infection. We found that transcervical infection with C. muridarum results in higher bacterial burdens in the upper genital tract at earlier time points, correlating with levels of innate immune cells. When bacterial burdens were equivalent in intravaginally and transcervically infected mice at later time points, we observed substantially higher levels of adaptive immune cells in transcervically infected mice. Our data suggest that different routes of infection with the same organism can elicit different immune responses in the same tissue.

INTRODUCTION

Chlamydia trachomatis is the most prevalent bacterial sexually transmitted infection in the United States, responsible for over one million newly reported cases each year (1). While C. trachomatis infection can be cleared with antibiotics, infection is often asymptomatic. If left untreated, infection can lead to downstream diseases in the upper genital tract such as pelvic inflammatory disease, ectopic pregnancies, and infertility. In humans, C. trachomatis is typically transmitted into the lower genital tract, where it can then ascend into the upper genital tract.

Mouse models of infection have been widely implemented in order to better understand Chlamydia infection and the resulting host immune response (2–5). In order to survive, the Chlamydia species have each evolved to evade host immune detection specific to their host. In mice, gamma interferon (IFN-γ) production induces a family of GTPases called the immunity-related GTPases (IRGs), which are able to mediate rapid clearance of human-pathogenic C. trachomatis (6, 7), whereas the mouse-adapted pathogen C. muridarum is able to evade clearance by the IRGs (8). As a result, when C. trachomatis is deposited into the lower genital tract of mice, the presence of IRGs and other mouse-specific factors leads to clearance of C. trachomatis and failure to ascend (5). In contrast, C. muridarum naturally ascends into the upper genital tract of mice when deposited intravaginally, establishes a productive infection in mice, and induces robust immunopathology that is reminiscent of human infection (9).

Despite the fairly well characterized host response to both C. trachomatis and C. muridarum infection in mice, it is unknown if the route of infection plays a substantial role in driving particular downstream immune responses. In this study, we took advantage of the ability of mouse-adapted C. muridarum to infect mice through intravaginal infection or by direct inoculation of the upper genital tract by the use of transcervical infection. Using the two routes of infection, we determined differences with respect to bacterial burden, innate and adaptive immunity, and resulting immunopathology in the upper genital tract. Overall, we found that transcervical infection resulted in a more robust innate and adaptive immune response that correlated with enhanced immunopathology. These responses, however, were not entirely dependent on bacterial burden. Our data suggest that while the two methods of infection ultimately allow similar antigen loads in the upper genital tract of mice, they can result in significantly different immune responses.

RESULTS

C. muridarum transcervical infection results in rapid colonization of the upper genital tract of mice.

In this study, we sought to determine if the route of infection with Chlamydia muridarum impacts the downstream immune response in the upper genital tract of mice. We infected mice either intravaginally, depositing bacteria into the lower genital tract, or transcervically, depositing bacteria directly into the upper genital tract. At different days postinfection, the entire upper genital tract (uterine horns and ovaries) was harvested and bacterial burden quantified using quantitative PCR (qPCR). We found that at the beginning of infection (day 2 and day 4 postinfection), transcervically infected mice had substantially higher bacterial burdens than mice infected intravaginally (Fig. 1). By the end of the time course (day 8 postinfection), the transcervically infected mice and intravaginally infected mice exhibited the same levels of bacterial burdens (Fig. 1). Taken together, these data indicate that transcervical infection with C. muridarum resulted in a higher rate of colonization of the upper genital tract of mice but that bacteria deposited intravaginally were able to ascend into the upper genital tract at later time points and establish infection to the same extent as when bacteria were deposited directly into the upper genital tract.

FIG 1.

C. muridarum transcervical infection results in rapid colonization of the upper genital tract of mice. Mice were infected transcervically (tc) or intravaginally (ivag) with 10⁶ IFU of C. muridarum. At various time points postinfection, mice were sacrificed and upper genital tracts harvested. Bacterial burden was assessed through qPCR. Data are compiled from time points of three independent experiments. d2, day 2 postinfection; d4, day 4 postinfection; d6, day 6 postinfection; d8, day 8 postinfection.

Innate immune cells traffic to the upper genital tract more rapidly following transcervical infection with C. muridarum.

Given the differences in bacterial burdens at early time points between the two routes of infection, we hypothesized that there might be differences in the innate immune cell populations recruited to the upper genital tract. We infected mice intravaginally or transcervically and at different time points postinfection, harvested the upper genital tract, and characterized innate immune cell populations using flow cytometry. At early time points, there were substantially more neutrophils (Fig. 2A), natural killer (NK) cells (Fig. 2B), and dendritic cells (Fig. 2C) that had trafficked to the genital tract in transcervically infected mice than in intravaginally infected mice. This spike in innate immune cell infiltrates in transcervically infected mice correlated with the higher bacterial burden that we had previously observed (Fig. 1). Following day 2 (neutrophils) or day 4 (NK cells and dendritic cells [DCs]), there were no longer any differences in the levels of the innate immune cell populations between the two routes of infection. While not surprising, these data do confirm that the magnitude of the innate immune response to C. muridarum in the upper genital tract of mice can be at least partially attributed to the quantity of bacteria localized in the upper genital tract. Given that transcervical infection leads to a higher bacterial burden at day 2 and day 4 postinfection, it is logical that the number of innate immune cells that first respond to infection would also be greater.

FIG 2.

Innate immune cells traffic to the upper genital tract more rapidly following transcervical infection with C. muridarum. Mice were infected transcervically or intravaginally with 10⁶ IFU of C. muridarum. At various time points postinfection, mice were sacrificed and upper genital tracts harvested. Organs were assessed by flow cytometry for the absolute number of (A) neutrophils (CD11b⁺ Gr1^hi), (B) NK cells (CD3⁻ NK1.1⁺), and (C) dendritic cells (CD11c⁺). Data are compiled from time points of two independent experiments.

Adaptive immune cells are enriched in the upper genital tract following C. muridarum transcervical infection.

While innate immune cells are the first responders at the site of infection, adaptive immune cells are then critical for promoting bacterial clearance and long-lasting immunity. Using flow cytometry, we compared the levels of recruitment of adaptive immune populations in the upper genital tract of mice infected either intravaginally to the levels seen with those infected transcervically. We found that there were significantly more B cells (Fig. 3A), CD4⁺ T cells (Fig. 3B), and CD8⁺ T cells (Fig. 3C) in the upper genital tract of transcervically infected mice than in that of intravaginally infected mice at almost every time point examined. We also observed that the numbers of B cells and T cells increased as time went on, further confirming that the adaptive immune response to C. muridarum builds over time. The bulk of the adaptive immune response occurred at later time points (day 6 and day 8), but the differences that we observed in B cells and CD4⁺ T cells between the two routes of infection did not correlate with bacterial burden (Fig. 1). By those later time points, the bacterial burden exhibited by intravaginally infected mice was the same as or higher than that seen with transcervically infected mice (Fig. 1) despite substantially lower levels of B cells and CD4⁺ T cells. In contrast to what we observed with innate immune cell infiltrates, these data suggest that the strength of the adaptive immune response does not necessarily correspond to the antigen load and that the route of infection plays a critical role in determining the quantity of recruited immune cells.

FIG 3.

Adaptive immune cells are enriched in the upper genital tract following C. muridarum transcervical infection. Mice were infected transcervically or intravaginally with 10⁶ IFU of C. muridarum. At various time points postinfection, mice were sacrificed and upper genital tracts harvested. Organs were assessed by flow cytometry for the absolute number of (A) B cells (B220⁺), (B) CD4⁺ T cells (CD4⁺ CD3⁺), and (C) CD8⁺ T cells (CD8⁺ CD3⁺). Data are compiled from time points of two independent experiments.

IFN-γ production is heightened following C. muridarum transcervical infection.

Given the enhanced immune response following transcervical infection, we tested whether it also corresponded to increased production of IFN-γ. IFN-γ has been shown to be critical in restricting Chlamydia growth and enhancing clearance (10–12). At both day 2 and day 4 postinfection, there were substantially higher levels of IFN-γ in the upper genital tract of transcervically infected mice (Fig. 4A). It was previously reported that NK cells produce IFN-γ during the initial stages of infection (13). We found substantially higher NK cell numbers at day 2 and day 4 postinfection in transcervically infected mice (Fig. 2B), correlating with increased levels of IFN-γ production (Fig. 4A). However, it is also possible that other cell types in the upper genital tract may also have contributed to the levels of IFN-γ that we observed. For example, previous reports have demonstrated that Th1-producing CD4⁺ T cells are critical for bacterial clearance and are primary producers of IFN-γ following the innate immune response (2, 5). Therefore, it is possible that the heightened level of IFN-γ production seen at day 4 could also correspond to a larger CD4⁺ T cell population in the upper genital tract (Fig. 3B).

FIG 4.

IFN-γ production is heightened following C. muridarum transcervical infection. Mice were infected transcervically or intravaginally with 10⁶ IFU of C. muridarum. At various time points postinfection, mice were sacrificed and upper genital tracts harvested. (A) Upper genital tracts were homogenized and supernatants used to assess IFN-γ levels by ELISA. (B) Flow cytometry was used to determine absolute numbers of cytokine-producing CD4⁺ T cells. Data are pooled from two independent experiments.

Given that we observed greater numbers of CD4⁺ T cells in the genital tract of mice following transcervical infection at later time points, we hypothesized that there would similarly be greater numbers of cytokine-producing CD4⁺ T cells. Indeed, there were significantly more IFN-γ-producing and tumor necrosis factor alpha (TNF-α)-producing CD4⁺ T cells in the transcervically infected mice at day 6 and day 8 postinfection than in the intravaginally infected mice (Fig. 4B).

Transcervical infection elicits greater immunopathology than intravaginal infection.

It was previously established that, reminiscent of human infection, C. muridarum ascension into the upper genital tract correlates with enhanced disease pathology. However, given the enhanced immune response in terms of both cell infiltrates and IFN-γ production, we wondered whether or not transcervical infection could induce greater levels of immunopathology than intravaginal infection. Mice were infected intravaginally or transcervically, and at day 4 or day 8 postinfection, the upper genital tracts were harvested and assessed for immunopathology through scoring and hematoxylin-eosin (H&E) staining of sections. Mice infected transcervically had significantly greater immunopathology in both the uterine horns and ovaries than intravaginally infected mice at both time points, consistent with the observed differences in immune cell recruitment and cytokine production (Fig. 5A). Intravaginally infected mice exhibited little pathology, as the lumen of the uterine horns was clear (Fig. 5B, top left) and the ovaries were not surrounded by inflammatory infiltrates (Fig. 5B, top right). Transcervically infected mice exhibited more-severe pathology, consisting of entry of immune cell infiltrates into the lumen of the uterine horns (Fig. 5B, bottom left) and inflammation in both the ovaries and the surrounding bursa membrane (Fig. 5B, bottom right). Taken together, these data suggest that higher bacterial burden at early time points leads to greater levels of innate immune infiltrates and higher cytokine production, which in turn can induce enhanced immunopathology.

FIG 5.

Transcervical infection elicits greater immunopathology than intravaginal infection. Mice were infected transcervically or intravaginally with 10⁶ IFU of C. muridarum. At various time points postinfection, mice were sacrificed and upper genital tracts were harvested, fixed, sectioned, and assessed for immunopathology by H&E staining. (A) Quantification of immunopathology. (B) Representative images of H&E-stained upper genital tracts of mice. Data are pooled from two independent experiments. *, immune infiltrates.

DISCUSSION

There currently exists no vaccine against infection by C. trachomatis despite its prevalence in the United States and worldwide. Modeling infection in animals is necessary in order to develop an effective vaccine. There are two commonly used animal models for genital Chlamydia infection, differing in the Chlamydia species used and the method of inoculation. Because Chlamydia species are extremely highly host adapted, the human-pathogenic Chlamydia species C. trachomatis is able to infect mice only when deposited directly into the upper genital tract (5). In contrast, the mouse-adapted C. muridarum species can naturally ascend into the upper genital tract of mice when deposited intravaginally (9). While the ability to ascend can be partially attributed to host genetics, it is also possible that the microbiota of the lower genital tract plays a role. Multiple human studies have indicated that Chlamydia infection in women correlates with certain commensal microbes (14–16). However, it is unclear if the microbiota of the lower genital tract of laboratory mice influences the ability of different Chlamydia species to ascend into the upper genital tract.

In this study, we sought to investigate if differences in the infection route would have an impact on the timing or nature of the immune response while keeping the Chlamydia species constant and without manipulating the vaginal microbiome. Given that C. muridarum can productively infect both the lower and upper genital tracts, we sought to do a comprehensive comparison of mice infected intravaginally or transcervically with C. muridarum. There have been several studies comparing different routes of infection with C. muridarum, but those studies mainly focused on examining bacterial burden and hydrosalpinx score differences resulting from different host genetic backgrounds (17–19). Comparing different mouse models of Chlamydia infection is absolutely essential in order to further define host factors that are critical for controlling infection, including the compromises made by using different routes of inoculation. To this end, we assessed downstream differences in infiltrating innate and adaptive immune cells in mice infected intravaginally or transcervically with C. muridarum. Despite infecting with the same quantity of bacteria, we found that transcervical infection with C. muridarum elicited a more robust innate (Fig. 2) and adaptive (Fig. 3) immune response than intravaginal infection. The differences in innate immune infiltrates largely correlated with bacterial burden (Fig. 1), suggesting that the strength of the innate immune response in the upper genital tract of mice is directly dependent on the quantity of bacteria. However, the adaptive immune response was still greater in transcervically infected mice at later time points even when the bacterial burdens administered by the two infection routes were identical. This disparity suggests that the strength of the adaptive immune response is not completely dependent on the quantity of antigen present in the tissue. Given that the transcervically infected mice had previously had higher bacterial burdens and greater innate immune responses at early time points, it is possible that these differences at early time points are what dictate the strength of the adaptive immune response.

Consequences of an enhanced immune response include immunopathology and tissue damage. These outcomes are predominant causes of Chlamydia-related disease in humans. We found that enhanced levels of innate immune responses and bacterial burden correlated with levels of IFN-γ and immunopathology and were dependent on the transcervical infection route at early time points (day 4 postinfection). We also observed that at day 8 postinfection, the transcervically infected mice still exhibited greater immunopathology and higher levels of IFN-γ in the upper genital tract. While these data do not correspond to differences in bacterial burden, they do correlate with numbers of adaptive immune cells (Fig. 3 and 4).

Chlamydia-specific CD4⁺ T cells have been demonstrated to home in on the genital tract and produce high levels of IFN-γ in response to infection (20, 21). Given that there are greater numbers of CD4⁺ T cells in the genital tracts of mice following transcervical infection, it is likely that this enhanced population contributes significantly to higher levels of IFN-γ in the upper genital tract. While CD4⁺ T cells are necessary for Chlamydia clearance (5, 22), antigen-nonspecific bystander CD4⁺ T cells have been found to contribute to immunopathology in mice infected with C. trachomatis serovar D (23). Furthermore, it has been recently demonstrated that CD4⁺ T cells can induce pathology in mice deficient in CD8⁺ T cells, following C. muridarum infection (24). It is possible that the influx of CD4⁺ T cells to the upper genital tract in mice following transcervical infection contributes not only to IFN-γ production but also to immunopathology.

Modeling Chlamydia infection in mice is necessary to determine what factors are important for pathogenesis and host immunity. While transcervical infection with C. muridarum might be considered less physiologically relevant than intravaginal infection, it is important to understand the consequences of different inoculation routes when employing the various models. Ultimately, we find that the route of inoculation with C. muridarum can play an important role in determining the downstream immune responses. Understanding how routes of infection differ in their downstream consequences can help us build better models for studying Chlamydia infection and, in turn, develop an effective vaccine.

MATERIALS AND METHODS

Growth and isolation of bacteria.

Chlamydia muridarum Nigg was propagated in McCoy cells as previously described (25, 26). Purified elementary bodies were divided into aliquots and stored at −80°C in SPG buffer (250 mM sucrose, 10 mM sodium phosphate and 5 mM l-glutamic acid) and thawed immediately prior to use.

Mice, infections, and preparation of tissue.

Six-to-8-week-old female C57BL/6 mice were purchased from The Jackson Laboratory (Bar Harbor, ME). All mice were housed in the Harvard Medical School Center for Animal Resources and Comparative Medicine, and all experiments were approved by Harvard’s Institutional Animal Care and Use Committee. At 1 week prior to infection, mice were treated with 2.5 mg medroxyprogesterone subcutaneously. For intravaginal infection, 10⁶ inclusion-forming units (IFU) of C. muridarum in 10 μl SPG buffer was deposited in the lower genital tract. For transcervical infection, 10⁶ C. muridarum IFU in 10 μl SPG buffer was deposited in the upper genital tract using an NSET pipette tip (ParaTechs, Lexington, KY) as described previously (5). At specific times postinfection, mice were sacrificed and the upper genital tract (uterine horns and ovaries) was harvested.

Quantitative PCR.

Bacterial burden was determined through quantitative PCR as previously described (27). Briefly, DNA was isolated from tissue samples using a QIAamp DNA minikit (Qiagen, Valencia, CA) and Chlamydia 16S DNA and mouse GAPDH (glyceraldehyde-3-phosphate dehydrogenase) were quantified using primer pairs and dually labeled probes (IDT [San Jose, CA] or Applied Biosciences).

Flow cytometry.

Prior to staining, upper genital tracts were minced with scalpels and enzymatically dissociated in HBSS/Ca2⁺/Mg2⁺ containing 1 mg/ml type XI collagenase and 50 Kunitz units/ml DNase for 30 min at 37°C, washed in Ca2⁺/Mg2⁺-free phosphate-buffered saline (PBS) containing 5 mM EDTA, and then ground between frosted microscope slides prior to filtration through a 70-μm-pore-size mesh. For intracellular cytokine staining, cells were stimulated for 4 to 5 h with 50 ng/ml phorbol myristate acetate (PMA) (Enzo Life Sciences, Farmingdale, NY) and 500 ng/ml ionomycin (EMD Millipore, Darmstadt, Germany) in the presence of brefeldin A (BioLegend, San Diego, CA). All antibodies were purchased from BioLegend except where noted. Cells were incubated using fluorochrome-conjugated antibodies against mouse CD3 (clone 17A2), CD4 (clone RM4-5; Invitrogen, Carlsbad, CA), CD8 (clone 53-6.7), Gr1 (clone RB6-8C5), CD11b (clone M1/70), CD11c (clone N418), B220 (clone RA3-6B2), and NK1.1 (clone PK136); a live/dead Fixable Aqua dead cell stain kit to exclude dead cells (Invitrogen); and anti-FcRγ (Bio X Cell, West Lebanon, NH). For intracellular staining, cells were fixed and permeabilized according to the instructions of the manufacturer (BioLegend) and stained with anti-IFN-γ (clone XMG 1.2) and anti-TNF-α (clone MP6-XT22). AccuCheck counting beads (Invitrogen) were used to determine the absolute cell number. Data were collected on an LSR II flow cytometer (BD Biosciences, San Jose, CA) and analyzed using FlowJo (Tree Star, Ashland, OR).

ELISA.

Upper genital tracts were mechanically homogenized in 2 ml PBS. Tissue homogenates were frozen at −20°C and thawed immediately prior to use. Thawed homogenates were spun down, and the resulting supernatant was used for IFN-γ enzyme-linked immunosorbent assay (ELISA) as described previously (28).

Pathology.

For histopathology sectioning and scoring, upper genital tracts were excised from mice and fixed in 4% paraformaldehyde. Tissues were sent to the Rodent Histopathology Core Facility at Harvard Medical School for sectioning and hematoxylin and eosin (H&E) staining. Sections were assessed for immunopathology as described previously (23).

Statistical analysis.

Statistical analysis was performed using Prism (GraphPad). All data were analyzed using unpaired t tests. Differences were considered statistically significant if the P value was less than 0.05 (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). Data are represented as means ± standard errors of the means (SEM).