ARTHRITIS & RHEUMATOLOGY Vol. 67, No. 6, June 2015, pp 1535–1547 DOI 10.1002/art.39041 C 2015, American College of Rheumatology V

High Chlamydia Burden Promotes Tumor Necrosis Factor–Dependent Reactive Arthritis in SKG Mice Athan C. Baillet,1 Linda M. Rehaume,1 Helen Benham,1 Connor P. O’Meara,2 Charles W. Armitage,2 Roland Ruscher,1 Geraldine Brizard,3 Marina C. G. Harvie,2 Jared Velasco,1 Phillip M. Hansbro,4 John V. Forrester,5 Mariapia A. Degli-Esposti,6 Kenneth W. Beagley,2 and Ranjeny Thomas1 Objective. Chlamydia trachomatis is a sexually transmitted obligate intracellular pathogen that causes inflammatory reactive arthritis, spondylitis, psoriasiform dermatitis, and conjunctivitis in some individuals after genital infection. The immunologic basis for this inflammatory response in susceptible hosts is poorly understood. As ZAP-70W163C–mutant BALB/c (SKG) mice are susceptible to spondyloarthritis after systemic exposure to microbial b-glucan, we undertook the present study to compare responses

to infection with Chlamydia muridarum in SKG mice and BALB/c mice. Methods. After genital or respiratory infection with C muridarum, conjunctivitis and arthritis were assessed clinically, and eye, skin, and joint specimens were analyzed histologically. Chlamydial major outer membrane protein antigen–specific responses were assessed in splenocytes. Treg cells were depleted from FoxP3-DTR BALB/c or SKG mice, and chlamydial DNA was quantified by polymerase chain reaction. Results. Five weeks after vaginal infection with live C muridarum, arthritis, spondylitis, and psoriasiform dermatitis developed in female SKG mice, but not in BALB/c mice. Inflammatory bowel disease did not occur in mice of either strain. The severity of inflammatory disease was correlated with C muridarum inoculum size and vaginal burden postinoculation. Treatment with combination antibiotics starting 1 day postinoculation prevented disease. Chlamydial antigen was present in macrophages and spread from the infection site to lymphoid organs and peripheral tissue. In response to chlamydial antigen, production of interferon-g and interleukin-17 was impaired in T cells from SKG mice but tumor necrosis factor (TNF) responses were exaggerated, compared to findings in T cells from BALB/c mice. Unlike previous observations in arthritis triggered by b-glucan, no autoantibodies developed. Accelerated disease triggered by depletion of Treg cells was TNF dependent. Conclusion. In the susceptible SKG strain, Chlamydia-induced reactive arthritis develops as a result of deficient intracellular pathogen control, with antigen-specific TNF production upon dissemination of antigen, and TNF-dependent inflammatory disease.

Supported by National Health and Medical Research Council grants 1027657 and 569938 and by the Princess Alexandra Hospital Foundation. Dr. Baillet’s work was supported by a grant from the Societe Franc ¸aise de Rhumatologie. Dr. Benham’s work was supported by an Australian Postgraduate Award from the Australian Government, Department of Education and Training and by Arthritis Queensland. Dr. Thomas’ work was supported by a Future Fellowship from the Australian Research Council and by Arthritis Queensland. 1 Athan C. Baillet, MD, PhD, Linda M. Rehaume, PhD, Helen Benham, MBBS, PhD, FRACP, Roland Ruscher, PhD, Jared Velasco, BSc (Hons), Ranjeny Thomas, MBBS, MD: University of Queensland Diamantina Institute, Translational Research Institute, and Princess Alexandra Hospital, Brisbane, Queensland, Australia; 2 Connor P. O’Meara, PhD, Charles W. Armitage, PhD, Marina C. G. Harvie, PhD, Kenneth W. Beagley, PhD: Queensland University of Technology, Brisbane, Queensland, Australia; 3Geraldine Brizard, MSc: Lions Eye Institute, Nedlands, West Australia, Australia; 4Phillip M. Hansbro, PhD: Hunter Medical Research Institute and University of Newcastle, Newcastle, New South Wales, Australia; 5John V. Forrester, MD, FRS(E), FMedSci: Lions Eye Institute, Nedlands, West Australia, Australia, and University of Aberdeen Medical School, Aberdeen, Scotland; 6Mariapia A. Degli-Esposti, PhD: Lions Eye Institute, Nedlands, West Australia, Australia, and University of West Australia, Crawley, West Australia, Australia. Address correspondence to Ranjeny Thomas, MBBS, MD, University of Queensland Diamantina Institute, Translational Research Institute, 37 Kent Street, Woolloongabba, Queensland 4102, Australia. E-mail: [email protected]. Submitted for publication September 10, 2014; accepted in revised form January 15, 2015. 1535

1536

BAILLET ET AL

Chlamydia trachomatis, an obligate intracellular pathogen, is among the most frequent causes of bacterial sexually transmitted disease. Approximately 4–15% of individuals with genital C trachomatis infection subsequently develop reactive arthritis (1). This form of spondyloarthritis (SpA) is characterized by arthritis, enthesitis, inflammation of the axial skeleton, conjunctivitis, and a pustular psoriasis-like skin rash called keratoderma blenorrhagica. The mechanism of reactive arthritis is poorly understood, but tumor necrosis factor (TNF) inhibition has been shown to be a safe and effective treatment (2). One suggested mechanism of reactive arthritis is immune dysregulation, in which reduced clearance promotes dissemination and persistence of C trachomatis (3). The adjuvant effect of residual chlamydial pathogen– associated molecular patterns (PAMPs) is proposed to then stimulate an inflammatory response within affected organs. Alternatively, Chlamydia antigens may induce autoimmune responses to self antigens by molecular mimicry, and/or the dissemination of chlamydial PAMPs may skew immunity, e.g., toward Th17 polarization (4). Dissemination of C trachomatis into synovial tissue through infected macrophages is suggested by the presence of inclusion bodies and chlamydial DNA (5). Progress toward a mechanistic understanding has been limited by a lack of suitable animal models. T cells in ZAP-70W163C–mutant BALB/c (SKG) mice exhibit attenuated T cell receptor (TCR) signaling, with increased autoreactivity and production of interleukin-17 (IL-17). SKG mice display an increased proportion of peripherally derived Treg cells accompanied by reduced interferon-g (IFNg) production by T cells (6). They develop T cell– and IL-23–dependent clinical, histologic, and radiographic features of SpA after intraperitoneal injection of microbial 1,3-b-glucan (curdlan) (4). Chlamydia muridarum is a murine pathogen that is closely similar to the human pathogen C trachomatis, which infects the vaginal epithelium. In mice, vaginal infection ascends to the upper genital tract causing scarring and occlusion of the fallopian tube (oviduct), with the infection eventually resolving within 4 weeks. In BALB/c mice, infection control requires CD41 T cells, IL-12, and IFNg. Since SKG mice are susceptible to SpA, we compared responses to urogenital C muridarum infection in these mice versus responses in BALB/c mice. MATERIALS AND METHODS C muridarum and major outer membrane protein (MOMP) recombinant purification. C muridarum (Weiss) was grown and C muridarum MOMP expressed and purified as previously described (7,8).

Mice. Male and female BALB/c mice ages 6–9 weeks were bred under specific pathogen–free conditions at the University of Queensland. Experiments were approved by the University of Queensland Animal Ethics Committee (UQDI/ 467/12/NHMRC). TLR-22/2 BALB/c mice (4) were crossed with SKG mice. FoxP3-DTR mice, expressing a diphtheria toxin receptor–enhanced green fluorescent protein (GFP) fusion protein under the control of the foxp3 locus (9,10) were backcrossed with BALB/c mice (kindly provided by Dr. Ian van Driel, University of Melbourne, Melbourne, Victoria, Australia) and then crossed with SKG mice. One day after genital inoculation, female mice were administered a single intraperitoneal injection of diphtheria toxin (1 mg; SigmaAldrich). Treg cell depletion in the peripheral blood and the spleen was confirmed by flow cytometry at 1 day and 7 days, respectively, after diphtheria toxin administration. Infection of mice, measurement of chlamydial shedding, and tape stripping. Female mice were primed by subcutaneous administration of 2.5 mg medroxyprogesterone acetate (Depo-Provera; Pfizer) and infected with 7 3 103 inclusion-forming units (IFU) of C muridarum intravaginally, as described previously (7). In other sets of experiments, groups of female mice were infected with 5 3 102, 5 3 104, and 106 IFU C muridarum intravaginally. In some experiments, SKG mice were infected with 7 3 103 IFU ultravioletinactivated C muridarum together with 10 mg lipopolysaccharide (Escherichia coli O111:B4; Sigma-Aldrich) (11). On days 3, 7, 14, 21, 35, 55, and 70 after inoculation, cervicovaginal swabs were collected (12). Chlamydial vaginal load was determined by direct inoculation of dilutions of vaginal swabs onto McCoy cell monolayers (7). Male mice were infected with either 106 IFU C muridarum per urethra (13) or 3.5 3 102 IFU intranasally (14). As a Chlamydia-positive control, knee joints were injected intraarticularly with 2 3 104 IFU C muridarum (15). Following removal of hair from the dorsal skin, tape stripping (20 strokes) of dorsal skin was performed, using transparent tape (16). Antibodies. Experiments used the following antibodies: Alexa Fluor 488–conjugated anti–IL-17A (TC11-18H10.1), PerCP/Cy5.5-conjugated anti-IFNg (4S.B3), allophycocyanin (APC)–conjugated anti-TNF (MP6-XT22), Pacific Blue–conjugated anti-CD3 (145-2C11), phycoerythrin (PE)– or Pacific Blue–conjugated anti-CD11b (M1/70), PE/Cy7-conjugated antiCD4 (GK1.5), PerCP/Cy5.5-conjugated anti–Ly-6G (1A8), APC/Cy7-conjugated anti-F4/80 (BM8) (all from BioLegend), PE-conjugated Ki-67 (B56; BD Biosciences), PE/Texas Red– conjugated anti-CD8 (53-6.7; Abcam), Alexa Fluor 700– conjugated anti-FoxP3 (FJK-16s; eBioscience), and relevant isotype controls. Anti-CD16/32 antibody (93; BioLegend) was used to block Fc receptors. Antibiotic and soluble TNF receptor (sTNFR) treatment. From day 1 through day 14 after vaginal inoculation, rifampin (0.4 mg/day intraperitoneally; Sanofi-Aventis) (17) and doxycycline (0.3 mg/day by gavage; Alphapharm) were administered. Female C muridarum–infected FoxP3DTR SKG mice were administered sTNFR (etanercept; 250 mg/100 ml) subcutaneously 1 day before, 2 days after, and 5 days after vaginal inoculation. Clinical and histopathologic assessment. Mice were weighed weekly. The same observer, who was blinded with regard to the treatment, measured the thickness of both fore

TNF-DEPENDENT REACTIVE ARTHRITIS IN CHLAMYDIA-INFECTED SKG MICE

paws and hind paws using a digital caliper (7), at various time points over a 12-week period. Conjunctivitis was defined as conjunctival redness with eyelid edema/thickening and discharge or tearing (18). Eye swabs were cultured and surveyed for C muridarum by immunofluorescence. Twelve weeks after inoculation, mice were killed and eyes, ears, hind paws, fore paws, sacroiliac joints, and tails were collected, fixed in formalin, embedded, and sectioned.

1537

Joints were decalcified in EDTA. Hematoxylin and eosin– stained sections were assessed and assigned histologic scores (19) by 2 independent readers who were blinded with regard to the treatment. Skin was scored on a scale of 0–8 and joints on a scale of 0–5 (details on the scoring parameters are available at https://espace.library.uq.edu.au/view/UQ: 353054). Mice with histologic scores of .1 were considered diseased.

Figure 1. Genital infection with Chlamydia muridarum (Cmu) induces enthesitis, synovitis, spondylitis, and sacroiliitis in female SKG mice. A, Hind paws of a female BALB/c mouse (top) and a female SKG mouse (bottom) 12 weeks after genital inoculation with C muridarum. B and C, Inflammation index (change in thickness from baseline) of the hind paws (B) and fore paws (C) after genital inoculation with C muridarum. D, Hind paw asymmetry (absolute difference, left hind paw thickness minus right hind paw thickness) after genital inoculation with C muridarum. In B–D, values are the mean 6 SEM in SKG mice (red) and BALB/c mice (black). * 5 P , 0.05; *** 5 P , 0.0001 versus BALB/c mice, by twoway analysis of variance. E, Histologic scores in the hind paw, fore paw, and spine 12 weeks after inoculation. Each symbol represents an individual mouse; bars show the mean 6 SEM. ** 5 P , 0.001 by Mann-Whitney test. F–I, Hematoxylin and eosin–stained sections from the sacroiliac joint (F), hind paw joint (G), fore paw joint (H), and intervertebral disk (I) of BALB/c and SKG mice, obtained 12 weeks after inoculation. Asymmetric sacroiliitis and plantar fasciitis are seen in the SKG mouse specimens (arrows in F and G, respectively). Original magnification 3 4 in F–H; 3 10 in I. Bars 5 0.5 mm. Results shown are representative of 43 mice in 2 independent experiments.

1538

Flow cytometric analysis. Cells were stimulated for 2 hours in the presence of MOMP (30 mg/ml) or phorbol myristate acetate (50 ng/ml) plus ionomycin (1 mg/ml), and then with brefeldin A. Cells were stained and analyzed on a Gallios flow cytometer (Beckman Coulter). Assessment of cell proliferation and cytokine production. Splenocytes, harvested 7 days postinoculation, were incubated in RPMI/10% fetal calf serum (FCS), with or without MOMP (5 mg/well). After 48 hours, supernatants were removed for cytokine analysis using Th1/Th2/Th17 CBA kits (BD Biosciences). Proliferation was assessed according to uptake of tritiated thymidine (PerkinElmer). MOMP-specific and antiproteoglycan antibodies. Enzyme-linked immunosorbent assay (ELISA) plates (Greiner Bio-One) were coated overnight with 2 mg/well MOMP protein, washed, blocked with 5% FCS in phosphate buffered saline, and then incubated with serially diluted serum. After washing, plates were incubated with horseradish peroxidase– conjugated secondary detection antibodies against mouse IgA, IgG, IgG1, or IgG2a and then developed with tetramethylbenzidine (Sigma-Aldrich) and read at 450 nm. End point titer was defined as 2 standard deviations above the mean absorbance in negative controls. Positive anti-MOMP IgG was considered as a priori evidence of infection. Sera from C muridarum–infected BALB/c and SKG mice were analyzed for IgG reactivity to proteoglycan by ELISA, as previously described (7). Chlamydia-specific ompA quantitative polymerase chain reaction. C muridarum present in tissue at low levels was measured by ompA-specific quantitative reverse transcription–polymerase chain reaction as previously described (7). Primers used to amplify the ompA gene encoding C muridarum MOMP were 50 -GCCGTTTTGGGTTCTGCTT-30 and 50 -CGAGACGTAGGCTGATGGC-30 . Myeloid cell sorting. Cell pellets were obtained from lumbar lymph nodes, spleens, and blood collected 1, 3, or 7 days postinoculation and from hind paw joints harvested 12 weeks postinoculation. CD11b-enriched cells were sorted from cell pellets by magnetic separation using CD11b MicroBeads (Miltenyi Biotec). The proportion of CD11b cells among splenocytes increased from 10% before sorting to 62% after sorting. Skin and joint explant culture. Ears and hind paw joints were harvested in saline. Ears were split longitudinally and the dermis cultured for 24 hours in Dulbecco’s modified Eagle’s medium with 1% sodium pyruvate/10% FCS, with supernatant discarded after 1 hour and 3 hours to remove cytokines released by necrotic cells. Ankle joints were chopped into 3 pieces and cultured in medium. Statistical analysis. The normality of distribution was assessed by Kolmogorov-Smirnov test. Student’s t-test and analysis of variance (ANOVA), respectively, were used to assess the significance of quantitative differences between 2 groups and $3 groups for normally distributed data, and the Mann-Whitney and Kruskal-Wallis tests, respectively, were used for non-normally distributed data. The number of animals needed to achieve statistical power was estimated from previous experiments (4). The significance of differences between time points was determined by two-way ANOVA with Tukey’s post hoc comparison. Fisher’s exact test was used to assess differences in binary outcomes between groups. C muridarum load in the early course of the genital infection was expressed based on the area under the curve (AUC) of C

BAILLET ET AL

muridarum IFU in the swabs from days 3–55, and correlations with hind paw swelling 7 weeks postinoculation were determined by Pearson’s R test. P values (2-tailed) less than 0.05 were considered significant.

RESULTS Induction of reactive arthritis in SKG mice by infection with C muridarum. By 5 weeks after genital infection with 7 3 103 IFU C muridarum, arthritis of both fore paws and hind paws had developed in 85% of the female SKG mice and in none of the female BALB/c mice (Figures 1A–C). Joint swelling was progressive and bilateral, but severity was asymmetric (Figure 1D) in C muridarum–infected SKG mice. Histologic analysis revealed inflammation of the axial and appendicular skeleton (Figure 1E) with sacroiliitis (Figure 1F), enthesitis (Figure 1G), synovitis (Figure 1H), and spondylitis (Figure 1I). No dactylitis developed in either group. Psoriasis-like inflammatory infiltration of the dermis, with epidermal thickening and parakeratosis, occurred in all male and female C muridarum–infected SKG mice but was absent in BALB/c mice (Figures 2A–E). Tape stripping of the abdominal skin of C muridarum–infected SKG mice did not trigger the K€ obner phenomenon or psoriasis-related features. However, when female mice were infected with higher doses of C muridarum (106 or 5 3 104 IFU), they exhibited features of pustular psoriasis skip lesions, including apoptotic bodies, spongiosis, and subcorneal neutrophilic pustules resembling keratoderma blenorrhagica (histologic images available at https://espace.library.uq.edu. au/view/UQ:353054). Fifty percent of the male SKG or BALB/c mice administered a urethral inoculum of 106 IFU C muridarum developed an infection, based on the presence of seroconversion (anti-MOMP IgG) as assessed 12 weeks postinoculation. Infected mice did not develop subsequent paw swelling, potentially because of the immuneprivileged status of the male mouse genital tract (20). We therefore infected male SKG mice intranasally with 5 3 102 IFU C muridarum. This resulted in a 50% incidence of arthritis. Paw swelling was mild in male infected SKG mice. In contrast, conjunctivitis was significantly more frequent among male than female SKG mice (Figures 2F and G). Swabs from affected conjunctivae were negative for C muridarum by culture (results not shown). There was no increase in conjunctivitis incidence among naive SKG mice relative to BALB/c mice. Histologic analysis revealed mild cellular infiltration of the joints and moderate infiltration of the skin of male mice. Inflammation and histologic scoring data

TNF-DEPENDENT REACTIVE ARTHRITIS IN CHLAMYDIA-INFECTED SKG MICE

1539

Figure 2. Genital infection with Chlamydia muridarum (Cmu) triggers skin and eye inflammation in SKG mice. A–D, Hematoxylin and eosin– stained skin sections from a female uninfected BALB/c mouse (A), a female uninfected SKG mouse (B), a female C muridarum (7 3 103 inclusion-forming units [IFU])–infected BALB/c mouse (C), and a female C muridarum (7 3 103 IFU)–infected SKG mouse (D), showing epidermal thickening (arrow), dermal infiltration (arrowhead), and parakeratosis (asterisk) in the infected SKG mouse. Original magnification 3 10. Bars 5 0.5 mm. E, Histologic scores in skin samples 12 weeks after inoculation. Each symbol represents an individual mouse; bars show the mean 6 SEM. ** 5 P , 0.001 by Mann-Whitney test. F, Conjunctivitis in a male SKG mouse following genital inoculation with C muridarum. G, Conjunctivitis incidence following C muridarum infection in male and female BALB/c and SKG mice. Bars show the standard error of the percentage survival without conjunctivitis. * 5 P , 0.05 versus infected male BALB/c mice, by Mantel-Cox log rank test. Results shown are representative of 30 mice in 2 independent experiments.

are available at https://espace.library.uq.edu.au/view/ UQ:353054. We did not observe weight loss, diarrhea, ileitis, or colitis in male C muridarum–infected BALB/c or SKG mice infected via the respiratory tract or in female mice infected via the reproductive tract (results not shown). With respect to clinical and histologic features and organs involved, these results indicate that after genital or respiratory infection with C muridarum, SKG mice develop signs of reactive arthritis similar to those caused by C trachomatis and Chlamydia pneumoniae in humans. The pathologic features differed between male and female animals, with a predilection for conjunctivitis in males and more severe arthritis, spondylitis, and enthesitis in females. Correlation between C muridarum load in the genital tract of female SKG mice and severity of arthritis. In contrast to the results obtained after infection of SKG mice with live C muridarum, challenge with ultraviolet-inactivated C muridarum did not induce paw swelling (Figure 3A), suggesting that replicating infec-

tion is required for immunopathology. Treatment with rifampin and doxycycline starting on day 1 postinfection decreased C muridarum load in the vagina by 1,000-fold by day 5 postinfection and significantly reduced arthritis severity, consistent with an effect of live C muridarum load on arthritis severity (Figures 3B and C). To further assess the impact of C muridarum load, we infected female SKG and BALB/c mice with increasing doses of C muridarum, ranging from 5 3 102 IFU to 106 IFU. Paw swelling was detected in SKG mice inoculated with low doses of C muridarum (5 3 102 IFU); paw swelling in these mice was significantly greater than that observed after infection of BALB/c mice with 106 IFU C muridarum (Figure 3D). The magnitude of chlamydial vaginal shedding (i.e., the AUC) correlated with the severity of the hind paw arthritis measured 7 weeks postinfection (Figure 3E). Three days after inoculation with C muridarum at 106 IFU, the load of C muridarum in the genital tract was significantly lower in BALB/c mice than in SKG mice (Figure 3F); similar results were observed

1540

BAILLET ET AL

Figure 3. Chlamydia muridarum (Cmu) load correlates with arthritis severity in female SKG mice. A, Hind paw inflammation index after genital infection of female SKG mice with C muridarum or with ultraviolet (UV)–inactivated C muridarum. B, C muridarum load in genital swabs from female C muridarum (7 3 103 inclusion-forming units [IFU])–infected SKG mice with or without rifampin and doxycycline treatment from day 1 to day 14 postinoculation. C, Hind paw inflammation index in female C muridarum–infected SKG mice with or without rifampin and doxycycline treatment. D, Hind paw inflammation index in female BALB/c and SKG mice following genital infection with C muridarum at 5 3 102, 5 3 104, or 106 IFU. E, Correlation between chlamydial load as determined by the area under the curve (AUC) of C muridarum IFU from baseline to day-14 swabs and the inflammation index at 7 weeks postinoculation (wpi) in female SKG mice. Statistical significance was calculated by Pearson’s linear regression. F, C muridarum load in swabs from female SKG mice and BALB/c mice. Values in A–D and F are the mean 6 SEM. * 5 P , 0.05; ** 5 P , 0.001; *** 5 P , 0.0001 (with versus without UV inactivation [A], with versus without antibiotic treatment [B and C], or SKG versus BALB/c mice [D and F]), by two-way analysis of variance. Results shown are representative of 14 mice (A and E), 33 mice (B), 17 mice (C), 24 mice (D), and 47 mice (F), in 2 independent experiments.

when both groups of mice were inoculated with C muridarum at 5 3 102, 5 3 104, and 7 3 103 IFU (data not shown). These data indicate that the development of reactive arthritis in SKG mice is associated with greater bacterial burden and with less efficient clearance of C muridarum than in BALB/c mice 3 days after challenge.

Increased antigen-specific TNF levels, reduced IFNg and IL-17A production, and altered innate immune response after administration of C muridarum. In view of the observed difference in C muridarum clearance, we compared innate and adaptive immune responses in female SKG and BALB/c mice 7 days after genital inoculation with C muridarum. The C muridarum–

TNF-DEPENDENT REACTIVE ARTHRITIS IN CHLAMYDIA-INFECTED SKG MICE

1541

Figure 4. Deficient interferon-g (IFNg) and interleukin-17A (IL-17A) production and increased myeloid cell proportion in female Chlamydia muridarum (Cmu)–infected SKG mice compared to BALB/c mice. A–D, Concentrations of tumor necrosis factor (TNF) (A), IFNg (B), IL-17A (C), and IL-6 (D) in culture supernatant after 2 days of culture with major outer membrane protein. E and F, Gating strategy (E) and proportions of monocyte/macrophages (CD11b1Ly-6Glow cells) in female C muridarum–infected BALB/c and SKG mice (F) 7 days postinoculation. G, C muridarum genomic DNA in CD11b1 cells from female C muridarum–infected BALB/c mouse spleens and in CD11b1 and CD11b2 cells from female C muridarum–infected SKG mouse spleens 7 days after vaginal inoculation with C muridarum. Dotted line represents background amplification. In A–D, F, and G, each symbol represents an individual mouse; bars show the mean 6 SEM. * 5 P , 0.05; ** 5 P , 0.001; *** 5 P , 0.0001, by one-way analysis of variance (ANOVA) followed by Tukey-Kramer test in A–D and by Mann-Whitney test in F and G. H, Hind paw inflammation index following genital infection of SKG mice and TLR2–/– SKG mice with C muridarum. Values are the mean 6 SEM. * 5 P , 0.05; ** 5 P , 0.001; 2/2 SKG mice, by two-way ANOVA. Results shown are representative of 22 mice (A–D), 28 mice (E and F), 6 mice *** 5 P , 0.0001 versus TLR2 (G), and 15 mice (H), in 2 independent experiments.

1542

specific IgG2a:IgG1 ratio was .1 with similar specific IgG and IgA titers in both infected BALB/c and infected SKG mice, suggesting Th1 polarization in both strains. Chlamydial MOMP antigen–specific splenocyte proliferation in female BALB/c and SKG mice increased significantly and to an equivalent extent after C muridarum infection (detailed results available at https://espace.library.uq.edu. au/view/UQ:353054). However, MOMP-specific TNF production was increased, and IFNg and IL-17A decreased, in female C muridarum–infected SKG mice relative to BALB/c mice, with no difference in levels of IL-6 (Figures 4A–D). Compared to BALB/c mice, MOMP-specific IFNg production by CD41 T lymphocytes from C muridarum–infected mice, measured by flow cytometry, was also reduced in female SKG mice, whereas MOMP-specific TNF production by CD3– cells was increased (detailed results available at https://espace.library.uq.edu.au/view/ UQ:353054). Unlike previous observations in curdlan-treated SKG mice, which develop CD41 T cell–mediated autoimmunity with production of autoantibodies and a systemic cytokine response (7), the concentration of IL-6, TNF, IFNg, and IL-17A in serum was low in female C muridarum–infected BALB/c and SKG mice and was not significantly different between the 2 strains. In contrast to MOMP-specific CD41 splenic T cells, higher levels of IL-6 were found in skin and joint explant cultures from female C muridarum–infected SKG mice compared to BALB/c mice, while TNF levels were relatively low (potentially related to low sensitivity for detection of T cell–derived TNF in such cultures, or the late timing of analysis in the joints). Antiproteoglycan antibody was not detected in infected mice 12 weeks postinfection. Thus, T cell cytokine production in response to C muridarum antigen presentation is altered soon after infection in SKG mice, and we found no evidence of autoimmune cross-reactivity with proteoglycan. Detailed results of these analyses are available at https://espace. library.uq.edu.au/view/UQ:353054. Increased systemic load of C muridarum DNA in myeloid cells of SKG mice compared to BALB/c mice. It has been suggested that monocytes and macrophages disseminate proinflammatory C pneumoniae and C trachomatis PAMPs systemically (14,21). To determine whether C muridarum was present in antigen-presenting cells outside the genital tract of female BALB/c and SKG mice that had been infected vaginally, we quantified C muridarum ompA DNA. One week postinoculation, the proportions of splenic CD11b1 myeloid cells and CD11b1Ly-6Glow monocyte/macrophages in C muridarum–infected female SKG mice were increased relative to those in C muri-

BAILLET ET AL

darum–infected BALB/c mice (Figure 4E). We found C muridarum DNA in spleen and lumbar lymph nodes of both BALB/c and SKG mice 7 days postinoculation (data available at https://espace.library.uq.edu.au/view/ UQ:353054). After sorting of splenocytes, C muridarum–specific DNA was found in CD11b1 myeloid cells from mice of each strain 7 days postinoculation, but not in CD11b– cells (Figure 4G). The high proportion of CD11b1 myeloid cells, including monocytes and macrophages, in infected SKG mice (Figure 4F) implies a high load of C muridarum PAMPs systemically and at the site of inflammation, associated with development of reactive arthritis. We hypothesized that dissemination of C muridarum PAMPs triggers tissue inflammation through the activation of innate immune receptors. As Toll-like receptor 2 (TLR-2) plays a major role in innate immune activation by Chlamydia (4,22), we infected TLR-22/2 mice with C muridarum. Consistent with reduced impact of C muridarum PAMPs, clinical arthritis severity was significantly decreased after C muridarum infection of TLR-22/2 SKG mice relative to wild-type SKG mice (Figure 4H). Prevention by Treg cells of skin and joint inflammation triggered by genital infection with C muridarum. Since SKG mice have an increased proportion of Treg cells (https://espace.library.uq.edu.au/ view/UQ:353054) and a reduced capacity for TCR signaling compared to BALB/c mice, and we found a lower capacity for early clearance of C muridarum, we reasoned that Treg cells may impede T cell–mediated control of C muridarum after genital infection of female SKG mice. To test this hypothesis, we depleted Treg cells from FoxP3-DTR SKG mice using diphtheria toxin 1 day after genital inoculation with C muridarum. The proportion of GFP-positive Treg cells in peripheral blood declined from a mean 6 SEM of 14.8 6 3% pretreatment (n 5 4) to 0.4 6 0.1% 1 day after diphtheria toxin treatment. One week later, Treg cell proportions in the spleens of SKG and BALB/c mice remained significantly reduced (Figure 5A). In contrast to non–Treg cell–depleted SKG mice, in which C muridarum load in genital swabs decreased from day 3 to day 7 postinoculation (Figure 3F), C muridarum load increased in Treg cell–depleted female SKG mice over the same time period (Figure 5B). Moreover, at the site of inflammation, C muridarum DNA was found in footpads of Treg cell–depleted mice 1 week postinoculation. Consistent with this increase in C muridarum load, 40% of female Treg cell–depleted C muridarum–infected SKG mice developed florid arthritis, conjunctivitis, and skin inflammation within 4

TNF-DEPENDENT REACTIVE ARTHRITIS IN CHLAMYDIA-INFECTED SKG MICE

1543

Figure 5. Treg cells control Chlamydia muridarum (Cmu)–induced reactive arthritis in female SKG mice. A and B, Proportion of Treg cells in the spleens of female FoxP3-DTR SKG mice 7 days after diphtheria toxin injection (A), and chlamydial load (in mean 6 SEM inclusion-forming units [IFUs]) in genital swabs from female C muridarum–infected SKG mice and female Treg cell–depleted C muridarum–infected SKG mice (B). C, Eye disease, skin disease, and arthritis (inset) in a female C muridarum–infected SKG mice 7 days after Treg cell depletion. D and E, Incidence of arthritis (D) and conjunctivitis (E) in female C muridarum–infected SKG mice after Treg cell depletion. Bars show the standard error of the percentage survival without arthritis (D) or conjunctivitis (E). F–J, Hematoxylin and eosin–stained samples of skin (F), hind paw (H), and fore paw (I) and histologic scores in skin samples (G) and hind paw samples (J) from female Treg cell–depleted C muridarum–infected SKG mice 7 days postinoculation. Original magnification 3 10 in F; 3 4 in H and I. Bars 5 0.25 mm. K–M, C muridarum major outer membrane protein–specific splenocyte production of interferon-g (IFNg) (K), interleukin-6 (IL-6) (L), and tumor necrosis factor (TNF) (M) in female Treg cell–depleted SKG mice with histologically evident arthritis 7 days postinoculation. N, Histologic scores of ankle specimens from Treg cell–depleted C muridarum–infected SKG mice with or without etanercept (ETN) treatment for 7 days after diphtheria toxin injection. In A, G, and J–N, each symbol represents an individual mouse; bars show the mean 6 SEM. * 5 P , 0.05; ** 5 P , 0.001; *** 5 P , 0.0001 (versus female Treg cell–depleted C muridarum–infected SKG mice in B, versus female C muridarum–infected SKG mice in D and E), by one-way analysis of variance (ANOVA) followed by Tukey-Kramer test (A), two-way ANOVA (B), Montel-Cox log rank test (D and E), or MannWhitney test (G and J–N). Results shown are representative of 47 mice (A), 32 mice (B–E), 24 mice (F–M), and 20 mice (N), in 3 independent experiments.

1544

BAILLET ET AL

days (Figures 5C–J). Despite Treg cell depletion, BALB/c mice did not develop a similar inflammatory syndrome (Figures 5D and E). MOMP-specific splenocyte production of IFNg, IL-6, and TNF was increased 7 days postinoculation in mice with arthritis, relative to mice without arthritis (Figures 5K–M). There was no increase in IL-17A production (data not shown). Arthritis severity was TNF dependent, as histologic severity was reduced in female Treg cell–depleted SKG mice treated with sTNFR (Figure 5N). These data indicate that Treg cells normally restrain the production of TNF in Chlamydia-infected SKG mice and that TNF production in response to C muridarum does not clear, and may even exacerbate, the chlamydial load, and provokes inflammatory reactive arthritis. DISCUSSION In the present study we found that the TCR ZAP-70W163C mutation predisposed SKG mice to characteristic features of reactive arthritis, including asymmetric arthritis, enthesitis, spondylitis, sacroiliitis, conjunctivitis, and psoriasis-like skin disease, upon C muridarum infection. Although SKG mice are prone to develop autoimmune diseases, Chlamydia-induced reactive arthritis in these animals appears to be driven by altered host immunity to Chlamydia, rather than by self cross-reactivity (7,23,24). While high levels of antiproteoglycan autoantibodies develop in SKG mice with curdlan-induced SpA, these autoantibodies were absent in mice with C muridarum–induced reactive arthritis. In contrast, impaired antigen-specific IFNg and IL-17 production by SKG mouse T cells was accompanied by elevated TNF levels and an increased C muridarum burden in the genital tract. Reactive arthritis development was dependent on live infection, and its severity correlated with bacterial load. Finally, our results demonstrated that Treg cells in SKG mice restrain TNF production, and in their absence, chlamydial infection triggers TNF-dependent inflammatory arthritis associated with a reduced rate of bacterial clearance (Figure 6). Thus, the inflammatory response in SKG mice may be detrimental to local control of infection and restriction of tissue dissemination of chlamydial antigen. How does impaired control of C muridarum infection trigger arthritis, conjunctivitis, and skin inflammation after intravaginal inoculation? The reduced capacity of SKG mice for T cell IFNg production in response to chlamydial antigen is critical, given that CD41 T cells and IFNg are required for bacterial killing, through macrophage activation and reactive oxygen species production (25,26). Similar to the cytokine differences in

Figure 6. Model of putative disease mechanisms involved in Chlamydia muridarum–induced reactive arthritis in SKG mice. IFNg 5 interferon-g; IL-17A 5 interleukin-17A; PAMPs 5 pathogen-associated molecular patterns; TNFa 5 tumor necrosis factor a.

BALB/c and SKG mice, IFNg-producing CD41 T cells have been shown to protect against C muridarum genital tract infection and IFNg-deficient mice were unable to clear infection (27), while TNF1CD81 T cells, neutrophils, IL-1b, and TLR-2 promoted oviduct inflammatory pathology (28). We hypothesize that local migration of myeloid cells activated by chlamydial PAMPs and carrying chlamydial antigen spreads the inflammatory response from genital tract to diseased tissue, and that myeloid antigen-presenting cells present chlamydial antigens to antigen-specific TNF-producing T cells locally. This hypothesis is consistent with detection of chlamydial antigen–specific CD41 and CD81 T cells in the joints of patients with Chlamydia-induced reactive arthritis (29), detection of Chlamydia DNA in the footpads of infected SKG mice in the highly inflammatory setting of Treg cell depletion, and detection of both MOMPspecific TNF-producing T cells and chlamydial DNA within myeloid cells in the spleen. Furthermore, IL-17A acts synergistically with IFNg to inhibit the growth of C muridarum in the early stage of infection (30), and production of both cytokines by MOMP-

TNF-DEPENDENT REACTIVE ARTHRITIS IN CHLAMYDIA-INFECTED SKG MICE

specific T cells was reduced in SKG mice compared to BALB/c mice. The increased production of TNF by SKG mouse T cells compared to BALB/c mouse T cells in response to chlamydial antigen, and the TNF dependence of reactive arthritis when acutely triggered by Treg cell depletion, are evidence of a pathogenic role of TNF in this model, which is in turn supported by the response to TNF inhibitor treatment in humans with reactive arthritis (2). Chlamydial PAMPs, such as MOMP, Hsp60, and macrophage infectivity potentiator, strongly stimulate TLRs, especially TLR-2, to activate myeloid cells, which take up and disseminate Chlamydia (22,28). TLR-2 also drives production of cytokines, such as IFNg and IL-17A, that are required for bacterial clearance and inflammatory cell recruitment to the genital tract (28,31). TLR-2 agonists activate the adaptive immune response and expansion of Treg cells (32). Myeloid cells are increased in infected SKG mice, predisposing them to higher levels of inflammation. However, TNF-family ligands can confer resistance of highly activated effector T cells to Treg cell–mediated suppression (“tuned suppression”), a mechanism that may drive chronic inflammation and prolonged pathology, including arthritis (33). Impaired function of Treg cells in SKG mice has been described (33), and the severe inflammatory response to Treg cell depletion demonstrates their importance in inflammation control. Similar to C muridarum–induced reactive arthritis in SKG mice, increased numbers of Treg cells with defective IL-10 production have been observed in humans with peripheral SpA (34). The differences in C muridarum shedding between BALB/c and SKG mice observed on day 3 suggest innate, rather than adaptive, immune mechanisms. It was recently reported that SKG mice were deficient in a ZAP-70–dependent novel innate-like IL-17A–producing g/d T cell subset (35). Chlamydiainfected macrophages also promote T cell apoptosis through paracrine effects (36), suggesting that Chlamydia-mediated immunomodulatory programmed cell death renders SKG mice even more immunodeficient. Interventions that reduced bacterial load and thus PAMP load, including antibiotics, also reduced arthritis in SKG mice. In patients with chronic C trachomatis–induced arthritis, antibiotic combinations of rifampin and azithromycin, or rifampin and doxycycline, reduced disease activity and bacterial load (37). In contrast, antibiotics barely impacted postenteric reactive arthritis, associated with direct disturbance of the gut microbe–mucosal barrier, as in SKG mice administered systemic b-glucan (38).

1545

In human reactive arthritis, it has not been clear whether chronic infection with live and infectious C trachomatis induces recurrent joint flares as a recurrent source of bacterial antigens or by pathogenic polarization of immune responses. C trachomatis can be found in both healthy and arthritic joints (39–41). In both mouse strains studied herein, we also found that chlamydial DNA could be detected outside the genital tract and that MOMP antigen–specific T cells were present in the spleen. Previous studies showed that C trachomatis up-regulates expression of multiple genes to adopt a persistent (rather than replicative) state (42,43), similar to Mycobacterium tuberculosis during persistent infection. Furthermore, detection of C muridarum DNA does not imply the presence of live bacteria, and the amplification may have detected remnant DNA after C muridarum phagocytosis. While mice of each strain cleared live Chlamydia from the genital tract, albeit more slowly in the SKG mice, development of reactive arthritis pathology was not associated with chronic infection of live Chlamydia in the genital tract. C muridarum DNA could not be detected in the paw, except when Treg cells were depleted and severe paw inflammation was present. Taken together, our data suggest that in SKG mice, it is the T cell inflammatory response to disseminated chlamydial antigen presentation that distinguishes their susceptibility to inflammatory disease, rather than inflammation arising from unusual chronic infection, or specific dissemination of Chlamydia and associated PAMPs in this mouse strain. This conclusion is consistent with previous observations of an intense synovitis triggered by injection of C trachomatis– infected synovial fibroblasts into the knees of Lewis rats and HLA–B27–transgenic Lewis rats (44). Those investigators also demonstrated that the synoviocyte is a suitable host cell for C trachomatis, and can function as a reservoir of microbial antigens that is sufficient to perpetuate joint inflammation through a Th1 cytokine imbalance (45). The current studies demonstrate that reactive arthritis can be triggered by Chlamydia independently of HLA–B27. Epidemiologic studies revealed no association between the risk of acquiring reactive arthritis and B27 positivity, but showed that B27-positive patients are more likely to have a severe and prolonged disease course (46). An increased incidence of reactive arthritis has been described in B27-negative patients infected with human immunodeficiency virus (HIV) (47). Similar to findings in SKG mice, activation and Th1 cytokine production of CD41 T cells is impaired in HIV-infected patients.

1546

BAILLET ET AL

Collectively, our findings in SKG mice strongly support the notion that susceptibility to reactive arthritis results from a genetic predisposition of the SKG TCR to respond, with TNF production, to antigen derived from the intracellular obligate pathogen Chlamydia, leading to TNF-dependent inflammatory disease. Either reduction in the burden of Chlamydia through treatment with combination antibiotics or reduction of inflammation through use of TNF inhibitors reduced disease in SKG mice. The importance of this with regard to human chlamydial reactive arthritis is supported by evidence that combination antibiotics and TNF inhibitors each reduced reactive arthritis disease activity. Indeed, only combination rifampin and azithromycin can clear C trachomatis in vitro from infected monocytes (48–50). Importantly, TNF inhibitors suppressed disease without increasing the risk of uncontrolled infection. The development of reactive arthritis in HIV-infected patients further supports the paradigm that reactive arthritis is driven by altered host T cell immunity, and in the example of HIV, virus-driven immune impairment impacts broader pathogen-associated chronic inflammation.

6.

7.

8.

9.

10.

11.

12.

13.

AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Thomas had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Baillet, Rehaume, Benham, Velasco, Beagley, Thomas. Acquisition of data. Baillet, Rehaume, Benham, O’Meara, Armitage, Ruscher, Brizard, Harvie, Velasco, Hansbro, Beagley. Analysis and interpretation of data. Baillet, Rehaume, Benham, O’Meara, Armitage, Ruscher, Brizard, Velasco, Hansbro, Forrester, Degli-Esposti, Beagley, Thomas.

14.

REFERENCES

18.

1. Rich E, Hook EW III, Alarcon GS, Moreland LW. Reactive arthritis in patients attending an urban sexually transmitted diseases clinic. Arthritis Rheum 1996;39:1172–7. 2. Meyer A, Chatelus E, Wendling D, Berthelot JM, Dernis E, Houvenagel E, et al, on behalf of the Club Rhumatisme et Inflammation. Safety and efficacy of anti–tumor necrosis factor a therapy in ten patients with recent-onset refractory reactive arthritis. Arthritis Rheum 2011;63:1274–80. 3. Bas S, Kvien TK, Buchs N, Fulpius T, Gabay C. Lower level of synovial fluid interferon-g in HLA-B27-positive than in HLAB27-negative patients with Chlamydia trachomatis reactive arthritis. Rheumatology (Oxford) 2003;42:461–7. 4. Beckett EL, Phipps S, Starkey MR, Horvat JC, Beagley KW, Foster PS, et al. TLR2, but not TLR4, is required for effective host defence against Chlamydia respiratory tract infection in early life. PloS One 2012;7:e39460. 5. Nanagara R, Li F, Beutler A, Hudson A, Schumacher HR Jr. Alteration of Chlamydia trachomatis biologic behavior in syno-

15. 16.

17.

19.

20. 21. 22.

23.

vial membranes: suppression of surface antigen production in reactive arthritis and Reiter’s syndrome. Arthritis Rheum 1995; 38:1410–7. Hirota K, Hashimoto M, Yoshitomi H, Tanaka S, Nomura T, Yamaguchi T, et al. T cell self-reactivity forms a cytokine milieu for spontaneous development of IL-171 Th cells that cause autoimmune arthritis. J Exp Med 2007;204:41–7. O’Meara CP, Armitage CW, Harvie MC, Timms P, Lycke NY, Beagley KW. Immunization with a MOMP-based vaccine protects mice against a pulmonary Chlamydia challenge and identifies a disconnection between infection and pathology. PloS One 2013;8:e61962. Cunningham KA, Carey AJ, Lycke N, Timms P, Beagley KW. CTA1-DD is an effective adjuvant for targeting anti-chlamydial immunity to the murine genital mucosa. J Reprod Immunol 2009;81:34–8. Chen Y, Haines CJ, Gutcher I, Hochweller K, Blumenschein WM, McClanahan T, et al. Foxp31 regulatory T cells promote T helper 17 cell development in vivo through regulation of interleukin-2. Immunity 2011;34:409–21. Kim J, Lahl K, Hori S, Loddenkemper C, Chaudhry A, deRoos P, et al. Depletion of Foxp31 cells leads to induction of autoimmunity by specific ablation of regulatory T cells in genetically targeted mice. J Immunol 2009;183:7631–4. Ito T, Hoshiai K, Tanabe K, Nakamura A, Funamoto K, Aoyagi A, et al. Maternal undernutrition with vaginal inflammation impairs the neonatal oligodendrogenesis in mice. Tohoku J Exp Med 2011;223:215–22. Carey AJ, Cunningham KA, Hafner LM, Timms P, Beagley KW. Effects of inoculating dose on the kinetics of Chlamydia muridarum genital infection in female mice. Immunol Cell Biol 2009;87:337–43. Pal S, Peterson EM, de la Maza LM. New murine model for the study of Chlamydia trachomatis genitourinary tract infections in males. Infect Immun 2004;72:4210–6. Moazed TC, Kuo CC, Grayston JT, Campbell LA. Evidence of systemic dissemination of Chlamydia pneumoniae via macrophages in the mouse. J Infect Dis 1998;177:1322–5. Hough AJ Jr, Rank RG. Induction of arthritis in C57B1/6 mice by chlamydial antigen: effect of prior immunization or infection. Am J Pathol 1988;130:163–72. Sano S, Chan KS, Carbajal S, Clifford J, Peavey M, Kiguchi K, et al. Stat3 links activated keratinocytes and immunocytes required for development of psoriasis in a novel transgenic mouse model. Nat Med 2005;11:43–9. Zhang X, Gao L, Lei L, Zhong Y, Dube P, Berton MT, et al. A MyD88-dependent early IL-17 production protects mice against airway infection with the obligate intracellular pathogen Chlamydia muridarum. J Immunol 2009;183:1291–300. Nakamura T, Toda M, Ohbayashi M, Ono SJ. Detailed criteria for the assessment of clinical symptoms in a new murine model of severe allergic conjunctivitis. Cornea 2003;22:S13–8. Ruutu M, Thomas G, Steck R, Degli-Esposti MA, Zinkernagel MS, Alexander K, et al. b-glucan triggers spondylarthritis and Crohn’s disease–like ileitis in SKG mice. Arthritis Rheum 2012; 64:2211–22. Pate MS, Hedges SR, Sibley DA, Russell MW, Hook EW III, Mestecky J. Urethral cytokine and immune responses in Chlamydia trachomatis-infected males. Infect Immun 2001;69:7178–81. Igietseme JU, Perry LL, Ananaba GA, Uriri IM, Ojior OO, Kumar SN, et al. Chlamydial infection in inducible nitric oxide synthase knockout mice. Infect Immun 1998;66:1282–6. Vabulas RM, Ahmad-Nejad P, da Costa C, Miethke T, Kirschning CJ, Hacker H, et al. Endocytosed HSP60s use Tolllike receptor 2 (TLR2) and TLR4 to activate the Toll/interleukin-1 receptor signaling pathway in innate immune cells. J Biol Chem 2001;276:31332–9. Yoshitomi H, Sakaguchi N, Kobayashi K, Brown GD, Tagami T, Sakihama T, et al. A role for fungal b-glucans and their receptor

TNF-DEPENDENT REACTIVE ARTHRITIS IN CHLAMYDIA-INFECTED SKG MICE

24.

25.

26. 27.

28.

29.

30.

31.

32.

33.

34. 35.

36.

Dectin-1 in the induction of autoimmune arthritis in genetically susceptible mice. J Exp Med 2005;201:949–60. Sakaguchi N, Takahashi T, Hata H, Nomura T, Tagami T, Yamazaki S, et al. Altered thymic T-cell selection due to a mutation of the ZAP-70 gene causes autoimmune arthritis in mice. Nature 2003;426:454–60. Cotter TW, Ramsey KH, Miranpuri GS, Poulsen CE, Byrne GI. Dissemination of Chlamydia trachomatis chronic genital tract infection in g interferon gene knockout mice. Infect Immun 1997;65:2145–52. Perry LL, Feilzer K, Caldwell HD. Immunity to Chlamydia trachomatis is mediated by T helper 1 cells through IFN-g-dependent and-independent pathways. J Immunol 1997;158:3344–52. Scurlock AM, Frazer LC, Andrews CW Jr, O’Connell CM, Foote IP, Bailey SL, et al. Interleukin-17 contributes to generation of Th1 immunity and neutrophil recruitment during Chlamydia muridarum genital tract infection but is not required for macrophage influx or normal resolution of infection. Infect Immun 2011;79:1349–62. Darville T, O’Neill JM, Andrews CW Jr, Nagarajan UM, Stahl L, Ojcius DM. Toll-like receptor-2, but not Toll-like receptor-4, is essential for development of oviduct pathology in chlamydial genital tract infection. J Immunol 2003;171:6187–97. Gerard HC, Carter JD, Hudson AP. Chlamydia trachomatis is present and metabolically active during the remitting phase in synovial tissues from patients with chronic Chlamydia-induced reactive arthritis. Am J Med Sci 2013;346:22–5. Zhang Y, Wang H, Ren J, Tang X, Jing Y, Xing D, et al. IL-17A synergizes with IFN-g to upregulate iNOS and NO production and inhibit chlamydial growth. PloS One 2012;7: e39214. He X, Nair A, Mekasha S, Alroy J, O’Connell CM, Ingalls RR. Enhanced virulence of Chlamydia muridarum respiratory infections in the absence of TLR2 activation. PloS One 2011;6: e20846. Yamazaki S, Okada K, Maruyama A, Matsumoto M, Yagita H, Seya T. TLR2-dependent induction of IL-10 and Foxp31CD251 CD41 regulatory T cells prevents effective anti-tumor immunity induced by Pam2 lipopeptides in vivo. PloS One 2011;6:e18833. Tanaka S, Maeda S, Hashimoto M, Fujimori C, Ito Y, Teradaira S, et al. Graded attenuation of TCR signaling elicits distinct autoimmune diseases by altering thymic T cell selection and regulatory T cell function. J Immunol 2010;185:2295–305. Appel H, Wu P, Scheer R, Kedor C, Sawitzki B, Thiel A, et al. Synovial and peripheral blood CD41FoxP31 T cells in spondyloarthritis. J Rheumatol 2011;38:2445–51. Wencker M, Turchinovich G, Di Marco Barros R, Deban L, Jandke A, Cope A, et al. Innate-like T cells straddle innate and adaptive immunity by altering antigen-receptor responsiveness. Nat Immunol 2014;15:80–7. Miyairi I, Byrne GI. Chlamydia and programmed cell death. Curr Opin Microbiol 2006;9:102–8.

1547

37. Carter JD, Espinoza LR, Inman RD, Sneed KB, Ricca LR, Vasey FB, et al. Combination antibiotics as a treatment for chronic Chlamydia-induced reactive arthritis: a double-blind, placebo-controlled, prospective trial. Arthritis Rheum 2010;62: 1298–307. 38. Rehaume LM, Mondot S, Aguirre de Carcer D, Velasco J, Benham H, Hasnain SZ, et al. ZAP-70 genotype disrupts the relationship between microbiota and host, leading to spondyloarthritis and ileitis in SKG mice. Arthritis Rheumatol 2014;66: 2780–92. 39. Wilkinson NZ, Kingsley GH, Jones HW, Sieper J, Braun J, Ward ME. The detection of DNA from a range of bacterial species in the joints of patients with a variety of arthritides using a nested, broad-range polymerase chain reaction. Rheumatology (Oxford) 1999;38:260–6. 40. Schumacher HR Jr, Arayssi T, Crane M, Lee J, Gerard H, Hudson AP, et al. Chlamydia trachomatis nucleic acids can be found in the synovium of some asymptomatic subjects. Arthritis Rheum 1999;42:1281–4. 41. Carter JD, Gerard HC, Espinoza LR, Ricca LR, Valeriano J, Snelgrove J, et al. Chlamydiae as etiologic agents in chronic undifferentiated spondylarthritis. Arthritis Rheum 2009;60:1311–6. 42. Gerard HC, Freise J, Wang Z, Roberts G, Rudy D, KraussOpatz B, et al. Chlamydia trachomatis genes whose products are related to energy metabolism are expressed differentially in active vs. persistent infection. Microbes Infect 2002;4:13–22. 43. Gerard HC, Whittum-Hudson JA, Schumacher HR, Hudson AP. Synovial Chlamydia trachomatis up regulates expression of a panel of genes similar to that transcribed by Mycobacterium tuberculosis during persistent infection. Ann Rheum Dis 2006;65:321–7. 44. Inman RD, Chiu B, Johnston ME, Falk J. Molecular mimicry in Reiter’s syndrome: cytotoxicity and ELISA studies of HLAmicrobial relationships. Immunology 1986;58:501–6. 45. Inman RD, Chiu B. Early cytokine profiles in the joint define pathogen clearance and severity of arthritis in Chlamydiainduced arthritis in rats. Arthritis Rheum 2006;54:499–507. 46. Hannu T. Reactive arthritis. Best Pract Res Clinical Rheumatol 2011;25:347–57. 47. Davis P, Stein M, Latif A, Emmanuel J. Acute arthritis in Zimbabwean patients: possible relationship to human immunodeficiency virus infection. J Rheumatol 1989;16:346–8. 48. Barber CE, Kim J, Inman RD, Esdaile JM, James MT. Antibiotics for treatment of reactive arthritis: a systematic review and metaanalysis. J Rheumatol 2013;40:916–28. 49. Dreses-Werringloer U, Padubrin I, Jurgens-Saathoff B, Hudson AP, Zeidler H, Kohler L. Persistence of Chlamydia trachomatis is induced by ciprofloxacin and ofloxacin in vitro. Antimicrob Agents Chemother 2000;44:3288–97. 50. Dreses-Werringloer U, Padubrin I, Zeidler H, Kohler L. Effects of azithromycin and rifampin on Chlamydia trachomatis infection in vitro. Antimicrob Agents Chemother 2001;45:3001–8.

Copyright of Arthritis & Rheumatology is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.