Reduced Endometrial Ascension and Enhanced Reinfection Associated With Immunoglobulin G Antibodies to Specific Chlamydia trachomatis Proteins in Women at Risk for Chlamydia.

Background Previous research revealed antibodies targeting Chlamydia trachomatis elementary bodies was not associated with reduced endometrial or incident infection in C. trachomatis–exposed women. However, data on the role of C. trachomatis protein–specific antibodies in protection are limited.

A whole-proteome C. trachomatis array screening serum pools from C. trachomatis–exposed women identified 121 immunoprevalent proteins. Individual serum samples were probed using a focused array. Immunoglobulin (Ig) G antibody frequencies and endometrial or incident infection relationships were examined using Wilcoxon rank sum test. The impact of the breadth and magnitude of protein-specific IgGs on ascension and incident infection were examined using multivariable stepwise logistic regression. Complementary RNA sequencing quantified C. trachomatis gene transcripts in cervical swab samples from infected women.

IgG to pGP3 and CT_005 were associated with reduced endometrial infection; anti-CT_443, anti–CT_486, and anti–CT_123 were associated with increased incident infection. Increased breadth of protein recognition did not however predict protection from endometrial or incident infection. Messenger RNAs for immunoprevalent C. trachomatis proteins were highly abundant in the cervix.

Protein-specific C. trachomatis antibodies are not sufficient to protect against ascending or incident infection. However, cervical C. trachomatis gene transcript abundance positively correlates with C. trachomatis protein immunogenicity. These abundant and broadly recognized antigens are viable vaccine candidates.

Chlamydia, genital tract, antibody, serology, endometrium, immunoproteome Protein-specific C. trachomatisantibodies are not sufficient to protect against ascending or incident infection. However, cervical C. trachomatisgene transcript abundance positively correlates with C. trachomatisprotein immunogenicity. These abundant and broadly recognized antigens are viable vaccine candidates. Chlamydia trachomatis genital infections are a major global health burden. Ascension of C. trachomatis from the infected cervix to the uterus and fallopian tubes can result in pelvic inflammatory disease and other sequelae, including chronic pelvic pain, infertility, and ectopic pregnancy. Chlamydial infections are treated with antibiotics, and prior infection provides a detectable level of natural immunity. Partial immunity is reflected by decreased infection concordance between older sex partners, lower bacterial loads with prior infection [1], and reduced reinfection for women in whom infection previously cleared spontaneously [2]. Repeated infections are associated with increased sequelae [3], demonstrating the critical need for an effective vaccine. Chlamydiae attach to cervical epithelial cells through multiple mechanisms [4–7] and stimulate host cell membrane invagination and engulfment into a protective inclusion. There is minimal evidence for antibody-mediated neutralization [8, 9], likely because adhesin-receptor pairs are not essential for entry [10]. CD4 T cells are key to the protective immune response through release of molecules inhibiting chlamydial growth and replication, including the T-helper 1–defining protein, interferon γ (IFN-γ). Nevertheless, identification of C. trachomatis proteins that elicit anti–C. trachomatis antibodies in high percentages of exposed individuals (immunoprevalence), and/or evoke high levels of antibody (immunodominance), reveals C. trachomatis proteins that are readily processed and presented to the immune system and could thus serve as important vaccine antigens. Furthermore, antibody can augment antigen presentation to T cells [11], and when IFN-γ is present antibody can limit C. trachomatis infection by enhancing bacterial opsonophagocytosis and degradation [12]. We previously investigated a role for immunoglobulin (Ig) G and IgA in limiting chlamydial infection to the cervix in infected women and determined that C. trachomatis–specific antibodies in serum and cervical secretions were not associated with reduced odds of endometrial infection [13]. Both serum and cervical anti–C. trachomatis IgG were associated with significantly increased risk of repeated infection during a year of follow-up. However, our study was limited by the lack of assessment of C. trachomatis protein–specific antibodies. We now extend our investigation by exploiting proteome microarrays [14] to identify immunoprevalent C. trachomatis proteins, using pooled serum samples from a C. trachomatis–exposed cohort of women [15]. Array specificity was validated using serum pools from C. trachomatis–seronegative volunteer women. The 121 most reactive proteins identified by whole-proteome arrays were used to probe serum samples from 222 sexually active women with well-defined biological and behavioral factors related to risk of C. trachomatis infection spread and subsequent reinfection [15]. The relationships of antigen-specific antibodies to ascending and incident infection were determined. Finally, the breadth of C. trachomatis–specific antibody recognition within the cohort was correlated with the relative abundance of C. trachomatis transcripts detected in cervical swab samples from 14 C. trachomatis–infected women.

Patient Populations Institutional review boards for human subjects research at the University of Pittsburgh and the University of North Carolina approved the studies, and all participants provided written informed consent before enrollment. Serum samples from women recruited at the University of Pittsburgh between 2011–2015 into the T cell Response Against Chlamydia (TRAC) study [15] were analyzed by immunoproteome assay for antibodies to chlamydial proteins. At enrollment, demographic, clinical, cervical, and endometrial microbiology data were obtained, along with self-report of sexually transmitted infections and sexual histories [15]. Women testing positive for cervical and endometrial infection were defined as Endo+ (n=77), while those testing positively for cervical infection only were defined as Endo− (n=67). Enrollment seropositivity to C. trachomatis and C. pneumoniae was determined by microimmunofluorescence assay (MIF) [16], at titers ≥1:16. Participants returned for follow-up at 1, 4, 8, and 12 months for repeated evaluation and sample collection. Women with C. trachomatis during follow-up received azithromycin, and their partners were referred for treatment. Women having C. trachomatis at any follow-up visit or reporting C. trachomatis infection between visits were defined as F/U+. Those who completed ≥3 follow-up visits without detected/reported C. trachomatis infection were defined as F/U−. Among the women with C. trachomatis at enrollment, 69 of 116 (59%) were F/U− and 47 (41%) were F/U+. Among 51 without C. trachomatis infection at enrollment, 9 (18%) were F/U+ and 42 (82%) were F/U−. The remainder were excluded from reinfection analyses because they did not complete ≥3 follow-up visits. A second cohort with identical recruitment criteria and sites in Pittsburgh is ongoing (TRAC2). This includes cervical sampling for RNA sequencing. Enrollment specimens from 17 participants, stored at −80°C in tubes containing zircon beads and preservative (RNA/DNA Shield; Zymo Research), were processed and analyzed for absolute abundance of C. trachomatis–derived mRNA transcripts. RNA Sequencing of Cervical Swab Samples Total RNA and DNA was simultaneously extracted from cervical cytobrush specimens using a ZymoBIOMICS DNA/RNA Miniprep (Zymo Research) kit. Host and bacterial ribosomal RNAs were depleted and libraries constructed, as described elsewhere [17]. The HTSeq count [18] was used to quantify C. trachomatis–derived transcripts after raw reads were demultiplexed, quality trimmed, and filtered to remove residual rRNAs. C. trachomatis–Specific Serum IgG Profiling by Proteome Immunoassay Whole-proteome microarrays (C. trachomatis serovar D) comprising 895 proteins were generated [14]. Supplementing these C. trachomatis antigens were multiple negative and positive controls including Epstein–Barr virus (EBV) viral capsid antigen p18. Confirmation of on-chip protein synthesis was achieved by probing with fluorescence-conjugated antibodies directed against N- and C-terminal fusion tags [19]. Whole-proteome microarray slides were used to analyze 6 serum pools (n=5 per pool) reflecting high-range (1:512–1:8192), midrange (1:64–1:256), and low-range (1:16–1:32) titers of anti–C. trachomatis IgG by MIF from TRAC enrollment samples. Six pools (n=5 per pool) from 20 C. trachomatis–seronegative female volunteers were also probed to assess specificity. Antigen-specific IgG binding was detected by a secondary fluorescence-conjugated antibody, and the signal intensity obtained was considered proportional to the amount of primary antibody already bound. Quality control was performed [19], and mean fluorescence intensity values for each assay imported to R [20]. Antigen-specific seropositivity was determined by neighborhood averaging, with positivity thresholds determined from mean fluorescence intensity values of the closest 50 spots, including negative polymerase chain reaction controls but excluding positive EBV controls [19]. The 118 most reactive proteins identified using serum pools on whole-proteome microarrays were selected for minimized microarrays. CT_043 and CT_305, and CT_682 were also included because they were frequently recognized by T cells from TRAC participants [21]. This yielded 121 proteins for individualized recognition profiling. Individual antigen seropositivity thresholds were calculated based on distance-weighted negative polymerase chain reaction control intensities (n=15 per slide) [19]. IgG amounts were also analyzed as continuous variables from the relative fold increase of a protein-antibody signal compared with negative controls. Statistical Analysis Antigen-specific IgG antibodies were evaluated as binary and continuous variables affecting the risk of ascending infection or reinfection. Fold change values were log-transformed and tested for normal distribution with the Shapiro-Wilk normality test before continuous analysis. Association of positive antibody frequencies with self-reported prior C. trachomatis infections were analyzed by analysis of variance. The correlation between frequencies and magnitudes of IgG responses was determined with the Pearson correlation test. Antibody reactivity to 121 chlamydial proteins was compared between Endo− and Endo+ groups and between F/U− and F/U+ groups using Wilcoxon rank sum test. Analysis was stratified to assess incident infection for women uninfected or infected at enrollment. With our goal to identify C. trachomatis–specific antibodies protecting against ascension and/or reinfection, 55 proteins with low reactive frequency (recognized by <10% of the cohort) and an additional 3 proteins with low reactive level (average fold change in intensity expression above negative controls in reactive women, <1.5) were filtered from further analysis. A univariable regression model was used to calculate unadjusted odds ratios (OR) for altered risk of endometrial ascension or incident infection. Stepwise multivariable logistic regression determined the best subset of proteins associated with altered risk using the Akaike information criterion, with the α to enter model selection set at univariable P≤.1. For binary data, we used Firth logistic regression to avoid predictive variables that separate completely with respect to outcome. Previously identified risk factors for ascension (age, current or incident gonorrhea, chlamydial infection at enrollment; and sex with new, uncircumcised, or infected partners) [15], were included in stepwise regression analyses, with reporting of adjusted OR for variables retained in the final model.

Baseline Characteristics of Participants Immunoproteome data from 222 TRAC women, (144 infected and 78 uninfected) were analyzed; 84% were seropositive to C. trachomatis by prior MIF assay. Participants’ sociodemographic characteristics have been previously reported for the TRAC cohort; they were young, single, African American, cisgender women, high school graduates or with some college [15]. Identification of Immunoprevalent and Immunodominant Chlamydial Proteins Six MIF-positive serum pools and 6 MIF-negative serum pools from C. trachomatis–seronegative female volunteers were screened for reactivity to 860 chlamydial proteins successfully expressed on microarrays. No data were removed after quality control. MIF-negative pools were reactive to EBV protein controls only (data not shown). Frequencies of positivity and average IgG fold increases to each protein among all participants and among participant subgroups are reported in Supplementary Table 1. Thirty proteins bound IgG from >25% of participants, and 11 were recognized by >50% of participants. This latter group included CT_858, chlamydial proteaselike activity factor (CPAF), which elicited the most frequent (75%) and dominant fold increases; CT_443, outer membrane complex protein B (OmcB); CT_813, an inclusion membrane protein (InaC); CT_142 and CT_143, secreted proteins; pGP3, a plasmid-encoded virulence factor; CT_681, the major outer membrane protein (MOMP); CT_795, a secreted protein; CT_529, a hypothetical protein; CT_798, glycogen synthase (GlgA); and CT_841, a protease. Strong concordance between immunoprevalent and immunodominant proteins was detected for the top 15 recognized proteins (Figure 1). Using all 121 proteins, the correlation coefficient rwas0.89 (P<.001). Chlamydial Gene Expression During Cervical Infection Chlamydial gene expression during cervical infection was assessed by Total RNAseq using libraries generated from 14 C. trachomatis–positive TRAC2 enrollees. In combination, transcripts detected covered almost the entire chlamydial genome (approximately 841 loci), while transcripts mapping to 268 genes were detected in >50% of infected participants (Supplementary Table 2). Of the top 50 proteins predicted by RNA abundance, 15 (30%) were among the top 50 proteins binding antibody, including CPAF, OmcB, CT_813, MOMP, and GlgA. Association of Antigen Immunoprevalence With Prior Chlamydial Infection For TRAC participants, regardless of enrollment infection status, IgG binding frequency, or breadth of C. trachomatis protein recognition tended to increase with the number of self-reported prior chlamydial infections (P=.95 [uninfected] and 0.21 [infected]) (Figure 2A and 2B). Failure to achieve significance may be influenced by a small subset of participants who did not report prior chlamydial infection but exhibited high antibody-binding frequencies, suggesting that they were unaware of prior exposure. Although there was a trend for infected participants to have a greater frequency of binding antibodies than those who were uninfected, 2-way analysis of variance comparison revealed no significant increase for infected women overall (P=.12). Distribution of Chlamydial Antigen Seropositivity Figure 3 depicts the distributions of antibody-binding proteins (per participant) between Endo− and Endo+ and between F/U− and F/U+ participants who were uninfected or infected at enrollment. Although a broader distribution of binding antibodies was detected for infected women, frequencies were high in many uninfected women (Figure 3B), likely reflecting prior chlamydial infection (55% of the cohort) [15]. The breadth of antigen recognition did not differ between Endo− and Endo+ women (P=.28; Figure 3A), indicating that lack of ascension was not associated with an overall higher frequency of antichlamydial antibodies. Antibodies specific for >40 antigens were detected in approximately 10% of Endo+ women. Breadth of protein recognition did not differ between F/U− and F/U+ women who were uninfected (P=.63; Figure 3B) or infected (P=.99; Figure 3C) at enrollment, indicating that an overall higher frequency of binding antibodies was not associated with reduced incident infection. For F/U+ women, >10 and >40 protein-specific antibodies were detected in approximately 10% of those who were uninfected or infected at enrollment, respectively. Association of Antigen-Specific Antibodies With Ascension Although overall antibody frequencies measured at enrollment were not associated with absence of endometrial infection, we investigated whether individual protein-specific antibodies predicted altered odds of ascension by regression analyses, incorporating previously determined factors associated with endometrial infection status [15] (Table 1). Table 1. | Protein IDa or Biologic Factor | Gene | Function | Univariable Analysisb | Multivariable Model | || |---|---|---|---|---|---|---| | ORc (95% CI) | P Value | Adjusted ORd (95% CI) | P Value | ||| | Binary Analysis | |||||| | pGP3 | pgp3 | Plasmid-encoded virulence factor | 0.38 (.18–.77) | .007 | 0.41 (.19–.86) | .02 | | CT_123a | accB | Acetyl coenzyme A carboxylase, biotin carboxyl carrier | 0.40 (.17–.85) | .02 | … | … | | CT_732a | ribE | 6,7-Dimethyl-8-ribityllumazine synthase | 0.34 (.12–.85) | .02 | … | … | | CT_619 | … | T3SS effector | 0.36 (.12–.92) | .03 | 0.45 (.14–1.24) | .14 | | CT_005a | incV | Inclusion membrane protein | 0.33 (.09–.97) | .04 | … | … | | CT_189a | … | DNA gyrase subunit A | 0.47 (.2–1.07) | .07 | … | … | | CT_017a | … | Hypothetical protein | 0.54 (.25–1.11) | .10 | … | … | | Neisseria gonorrhoeae infection | … | … | … | … | 2.37 (.83–7.04) | .11 | | Oral contraceptive pills | … | … | … | … | 2.70 (.90–9.25) | .09 | | Continuous Analysis | |||||| | CT_005a | incV | Inclusion membrane protein | 0.47 (.23–.84) | .02 | 0.50 (.24–.89) | .03 | | CT_732a | ribE | 6,7-Dimethyl-8-ribityllumazine synthase | 0.52 (.29–.89) | .02 | … | … | | CT_017a | Hypothetical protein | 0.75 (.55–.99) | .055 | … | … | | | CT_123a | accB | Acetyl coenzyme A carboxylase, biotin carboxyl carrier | 0.73 (.51–1) | .06 | … | … | | CT_189a | … | DNA gyrase subunit A | 0.7 (.47–1.01) | .07 | … | … | | CT_621 | … | Hypothetical protein | 0.76 (.54–1.03) | .10 | … | … | | Oral contraceptive pills | … | … | … | … | 2.79 (.96–9.27) | .071 | Abbreviations: CI, confidence interval; ID, identifier; OR, odds ratio; T3SS, type 3 secretion system. Proteins were determined in both binary and continuous analyses to be moderately associated with reduced endometrial infection. Univariable analysis shows proteins with unadjusted Pvalues <.1. Ratio of the odds of testing positive for cervical and endometrial infection with antibody to the protein to the odds of testing positive without the antibody. Unless otherwise indicated, reference groups are composed of all women who did not report or have the specified factor. OR after adjustment for all covariates in the final multivariable model. Binary analysis identified antibodies to 7 proteins (pGP3, CT_123, CT_732, CT_619, CT_005, CT_189, and CT_017) that were associated with decreased odds of endometrial infection (P<.1) (Table 1). Antibody to 2 proteins, pGP3 and CT_619, were retained in the final multivariable model associated with decreased odds, while biologic factors, Neisseria gonorrhoeae infection, and OCP use were associated with enhanced ascension, as reported previously [15]. pGP3, a virulence factor secreted into the host cell cytosol, was one of the most highly recognized proteins in the entire cohort, with a response frequency of 65% overall, 78% in Endo− women and 57% in Endo+ women. CT_619, a type 3 secretion system effector, was recognized by only 16% of the cohort, with 22% of Endo− and 9% of Endo+ women having antibodies to this protein (Supplementary Table 1). Antibody to MOMP, which comprises 65% of the C. trachomatis outer membrane complex and a known adhesin, was associated with a nonsignificant decreased odds of ascension (OR,.59 [confidence interval,.29–1.18]; P=.14; not shown). Univariable continuous analysis of the fold increase in IgG reactivity above negative controls identified antibodies associated with decreased odds of ascension (P<.1 at the individual protein level), and 5 of 7 proteins (CT_005, CT_732, CT_017, CT_123, and CT_189) overlapped with proteins detected by binary analysis (Table 1). Stepwise regression retained CT_005 (IncV) with biologic factor oral contraceptive pill use in the final model. Antibody binding CT_005 was associated with decreased odds of ascending infection, and the protein was recognized by 10% of the cohort, 17% of Endo− and 6% of Endo+ women (Supplementary Table 1). OCP use was again associated with increased odds of ascension [15]. Association of Antigen-Specific Antibodies With Repeated Infection Since active infection promoted increased breadth of antibodies, we analyzed the association of protein-specific antibodies with repeated infection for women according to infection status at enrollment. Stepwise regression included age, current or incident gonorrhea, chlamydial infection at enrollment, and sex with new, uncircumcised, or infected partners, biologic and behavioral factors modulating the risk of repeated infection in this cohort [15]. Among uninfected women, IgG to CT_828, ribonucleoside-diphosphate reductase beta chain, was associated with increased odds of incident infection by binary analysis but was not retained in the final model, and continuous analysis did not detect antigen-specific antibodies altering the odds of incident infection (Table 2). Stepwise regression retained only behavioral factors in the multivariable model, associated with increased odds of repeated infection [15]. Table 2. | Protein ID or Behavioral or Biological Factor | Gene | Function | Univariable Analysisa | Multivariable Model | || |---|---|---|---|---|---|---| | ORb (95% CI) | P Value | Adjusted ORc (95% CI) | P Value | ||| | Women Uninfected at Enrollment | |||||| | Binary analysis | |||||| | CT_828 | nrdB | Ribonucleoside-diphosphate reductase beta chain | 3.35 (.81–14.62) | .09 | … | … | | Sex during follow-up | |||||| | With new male partner | … | … | … | … | 9.30 (.54–172.46) | .12 | | With Chlamydia trachomatis–infected male partner | … | … | … | … | 2.41 × 104 (.04 to 5.00 × 108) | .13 | | With uncircumcised male partner | … | … | … | … | 36.13 (.95 to 3.87 × 103) | .06 | | Continuous analysis | |||||| | Sex during follow-up | |||||| | With new male partner | … | … | … | … | 9.30 (.54–172.46) | .12 | | With C. trachomatis–infected male partner | … | … | … | … | 2.41 × 104 (.04 to 5.00 × 108) | .13 | | With uncircumcised male partner | … | … | … | … | 36.13 (.95 to 3.87 × 103) | .06 | | Women Infected at Enrollment | |||||| | Binary analysis | |||||| | CT_443 | omcB | Outer membrane complex protein B | 2.2 (.92–5.63) | .08 | 3.95 (1.12–17.18) | .046 | | CT_486 | fliY | ABC transporter | 2.27 (.86–6.23) | .10 | 4.09 (1.05–17.48) | .046 | | CT_322 | tufA | Elongation factor Tu | 0.31 (.05–1.14) | .08 | || | Neisseria gonorrhoeae infection | … | … | … | … | 9.00 (2.01–51.78) | .007 | | C. trachomatis endometrial infection at enrollment | … | … | … | … | 0.26 (.08–.72) | .01 | | Age | … | … | … | … | 0.77 (.62–.92) | .01 | | Sex during follow-up | |||||| | With new male partner | … | … | … | … | 2.36 (.82–7.50) | .12 | | With C. trachomatis–infected male partner | … | … | … | … | 9.05 × 105 (75.14 to 4.86 × 1011) | .01 | | With uncircumcised male partner | … | … | … | … | 15.21 (1.80–176.39) | .02 | | Continuous analysis | |||||| | CT_123 | accB | Acetyl coenzyme A carboxylase | 1.33 (.961.92) | .10 | 1.55 (1.03–2.45) | .04 | | Gonorrhea coinfection | … | … | … | … | 8.30 (2.00–45.50) | .006 | | C. trachomatis endometrial infection at enrollment | … | … | … | … | 0.34 (.12–.90) | .04 | | Age | … | … | … | … | 0.78 (.64–.92) | .006 | | Sex during follow-up | |||||| | With C. trachomatis–infected male partner | … | … | … | … | 3.85 × 104 (9.94 to 8.63 × 109) | .03 | | With uncircumcised male partner | … | … | … | … | 17.49 (2.16–220.26) | .01 | Abbreviations: CI, confidence interval; ID, identifier; OR, odds ratio. Univariable analysis shows proteins with unadjusted Pvalues <.1. Ratio of the odds of reinfection with the antibody to the odds without the antibody. Unless otherwise indicated, reference groups are composed of all women who did not report or have the specified factor. OR after adjustment for all covariates in the final multivariable model. Univariable analysis of binary data revealed that antibodies to 3 proteins, OmcB, CT_486, and CT_322, were associated with altered odds of repeated infection for infected participants (P<.1). Continuous analysis revealed anti–CT_123 IgG associated with enhanced odds of repeated infection (Table 2). Stepwise regression retained antibodies binding 3 proteins: OmcB, CT_486 (FliY), an ABC transporter, and CT_123, acetyl-coenzyme A carboxylase, all associated with enhanced reinfection. OmcB, immunoprevalent and immunodominant in the cohort (Figure 1), was recognized by 84% of F/U+ and 69% of F/U− women; while CT_486 was recognized by 19% of the cohort, including 24% of F/U+ and 13% of F/U− women; CT_123 was recognized by 26% of the cohort, 33% of F/U+ and 22% of F/U− women (Supplementary Table 1). Antibody binding MOMP was associated with a nonsignificant increased odds of incident infection with both binary (OR,1.90 [confidence interval,.87–4.31]; P=.12) and continuous analysis (1.14 [.87–1.50]; P=.35). Gonorrhea coinfection, sex with a new male partner, sex with a C. trachomatis–infected male partner, and sex with an uncircumcised male partner during follow-up were associated with enhanced odds of reinfection. Increasing age and C. trachomatis endometrial infection at enrollment were associated with decreased odds of repeated infection, as previously reported for this cohort [15].

A whole C. trachomatis proteome microarray was used to identify serologically immunoprevalent and immunodominant chlamydial proteins in highly C. trachomatis–exposed women. The C. trachomatis proteome was not recognized by serum samples from C. trachomatis–seronegative volunteers. Next, 222 individual serum samples from this cohort were probed for recognition of the most immunoprevalent C. trachomatis antigens using minimized arrays. This provided a unique opportunity to determine whether IgGs to specific chlamydial proteins were associated with protection against C. trachomatis ascension and/or subsequent reinfection, circumstances that increase reproductive morbidity rates. Antibody specificity is suggested by the high correlation between antibody breadth and fold increases among women in the TRAC cohort who are highly exposed to C. trachomatis. Furthermore, of the 50 most highly recognized proteins, RNA sequencing analysis of cervical cytobrush samples from 14 women determined that 15 (30%) were among the most highly abundant C. trachomatis genes in samples for women with C. trachomatis. Finally, C. trachomatis–seronegative volunteers did not have detectable antibody reactivity to the C. trachomatis proteome. Overall, the breadth of C. trachomatis antigen-specific antibodies did not differ between Endo− and Endo+ women, or between F/U− and F/U+ women, regardless of infection status at enrollment. The breadth of C. trachomatis antibodies in Endo+ and F/U+ women (>40 C. trachomatis proteins) suggests that higher frequencies of binding antibodies fail to protect women from endometrial ascension or repeated infection. We noted nonsignificant trends for increased antibody breadth among women who reported prior infections and women with active infection. These data indicate that C. trachomatis exposure likely drives increased C. trachomatis antibody breadth without benefit. Antibodies to specific C. trachomatis proteins did not predict protection from ascension or repeated infection even when the magnitude of the response was considered. Despite inclusion of input proteins for stepwise regression analyses to those with P values≤.1, IgGs to proteins that remained in the final multivariable model were rare, and none approached the effect sizes of behavioral and biological factors that influence ascension or repeated infection. Nevertheless, multivariable stepwise regression identified antibodies specific for 3 C. trachomatis proteins—pGP3, CT_619, and IncV—associated with reduced odds of endometrial infection. Of these, pGP3 is involved in host cell adherence and invasion [22], neutralization of antimicrobial peptides [23], and stimulation of host inflammatory responses [24, 25]. It is highly immunogenic [26] and promotes C. trachomatis infection. Thus, antibody-mediated inhibition of pGP3’s functions could impair C. trachomatis invasion. However, 57% of Endo+ women had anti-pGP3 antibodies, and average fold increases were comparable to those detected in 78% of Endo− women (2.48 in Endo+ vs 2.63 in Endo− women; Supplementary Table 1). The mechanism by which antibodies to CT_619, a putative type 3 secretion system effector, could be protective is undetermined because its function is unknown. IncV tethers the inclusion membrane to the host’s endoplasmic reticulum, potentially influencing nonvesicular lipid or ion trafficking to the inclusion [27]. The biological plausibility for an inhibitory role for anti-Inc protein antibodies derives from reduced infection in vitro and in mice after incubation of Chlamydia muridarum with anti-IncA [28]. Analysis of F/U− versus F/U+ women identified antibodies specific to 3 proteins associating with increased odds of repeated infection. These included antibodies to OmcB, FliY (previously reported as an immunodominant B-cell antigen in nonhuman primates [29] and mice [30]), and CT_123. Although not significant, antibodies to MOMP also trended toward increased odds of reinfection. These data suggest that antibodies to outer membrane proteins are minimally neutralizing on C. trachomatis reexposure. OmcB is specific for C. trachomatis but FliY and CT_123 homologues are abundant in eubacteria. The biological significance of these responses is questionable because their effect sizes were lower than those of biological or behavioral factors. Previous analyses of T-cell responses determined that protein-specific CD4 production of IFN-γ correlated with protection from incident infection [21]. Taken together, our data suggest that increased anti-C. trachomatis antibodies identify women with compromised CD4 T-cell responses that increase the risk of chronic and/or repeated infections while eliciting futile antibody production. The primary strength of this study is the highly characterized cohort of C. trachomatis–exposed women providing adequate numbers of defined susceptible and protected participants for subgroup analyses. Additional strengths include use of a whole-proteome microarray to identify immunoprevalent antigens for subsequent individualized profiling and use of binary and continuous data for stepwise regression analyses, increasing sensitivity to detect differences between subgroups. A potential study limitation is that IgA responses and cervical responses were not measured. However, serum IgG is transported into female genital tract mucus secretions by the major histocompatibility complex class I–related neonatal Fc receptor and predominates in cervicovaginal mucus [31]. Second, our assay identifies binding antibody but does not provide information regarding function. If any C. trachomatis–specific antibody had neutralizing properties, we would expect an association with protection. Third, selection of immunoprevalent proteins for individualized analysis may have excluded rarely recognized proteins contributing to protection. Finally, in vitro translated proteins may not fold to conformations recognized by serum antibodies, while in vitro transcription/translation does not support posttranslational modifications (eg, glycosylation) needed for optimal protein binding. The most immunoprevalent proteins identified for TRAC2 participants overlap with 70% of the most recognized proteins in an independent cohort of previously C. trachomatis–infected women recruited in San Antonio, Texas [32], and 50% are included in the top 50 proteins recognized by a Dutch cohort of infected women [19]. Overall, CPAF, pGP3, InaC, CT_142, CT_143, and CT_841 are among the top 10 recognized proteins in each cohort. These data indicate significant conservation of antibody recognition, and thus antigen presentation, across individuals who are likely highly genetically diverse, suggesting these antigens are viable vaccine candidates. Both CPAF [33] and pGP3 [34] have been found to induce protective immunity in mouse models of chlamydial infection that is dependent on CD4 T-cell IFN-γ production, highlighting the importance of using a delivery system that stimulates CD4 T-cell immunity. Supplementary Data Supplementary materials are available at The Journal of Infectious Diseases online. Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author. Notes Acknowledgments. We thank the women who agreed to participate in this study; Ingrid Macio and Melinda Petrina, for their efforts in the clinic and the microbiology laboratory; Pam Kozlowski for her expertise and advice with enzyme-linked immunosorbent assays; and the staff at the Allegheny County Health Department STD Clinic, for their efforts. Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases (grants/awards R01AI119164, U19AI084024, and U19AI144181). Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

- 1. Schillinger JA, Katz BP, Markowitz LE, et al. Genotype-specific concordance of Chlamydia trachomatis genital infection within heterosexual partnerships. Sex Transm Dis 2016; 43:741–9. [DOI] [PMC free article] [PubMed] [Google Scholar] - 2. Geisler WM, Lensing SY, Press CG, HookEW 3rd. Spontaneous resolution of genital Chlamydia trachomatis infection in women and protection from reinfection. J Infect Dis 2013; 207:1850–6. [DOI] [PMC free article] [PubMed] [Google Scholar] - 3. Stamm WE, Roddy RE, Barnes RC, et al. Repeated genital infections with Chlamydia trachomatis: prevalence and risk factors. Chlamydial infections. Cambridge, United Kingdom: Cambridge University Press, 1986:499–502. [Google Scholar] - 4. Su H, Raymond L, Rockey DD, Fischer E, Hackstadt T, Caldwell HD.. A recombinant Chlamydia trachomatis major outer membrane protein binds to heparan sulfate receptors on epithelial cells. Proc Natl Acad Sci U S A 1996; 93:11143–8. [DOI] [PMC free article] [PubMed] [Google Scholar] - 5. Ajonuma LC, Fok KL, Ho LS, et al. CFTR is required for cellular entry and internalization of Chlamydia trachomatis. Cell Biol Int 2010; 34:593–600. [DOI] [PubMed] [Google Scholar] - 6. Stallmann S, Hegemann JH.. The Chlamydia trachomatis Ctad1 invasin exploits the human integrin β1 receptor for host cell entry. Cell Microbiol 2016; 18:761–75. [DOI] [PubMed] [Google Scholar] - 7. Elwell CA, Ceesay A, Kim JH, Kalman D, Engel JN.. RNA interference screen identifies Abl kinase and PDGFR signaling in Chlamydia trachomatis entry. PLoS Pathog 2008; 4:e1000021. [DOI] [PMC free article] [PubMed] [Google Scholar] - 8. Rank RG, Batteiger BE.. Protective role of serum antibody in immunity to chlamydial genital infection. Infect Immun 1989; 57:299–301. [DOI] [PMC free article] [PubMed] [Google Scholar] - 9. Olsen AW, Follmann F, Erneholm K, Rosenkrands I, Andersen P.. Protection against Chlamydia trachomatis infection and upper genital tract pathological changes by vaccine-promoted neutralizing antibodies directed to the VD4 of the major outer membrane protein. J Infect Dis 2015; 212:978–89. [DOI] [PubMed] [Google Scholar] - 10. Nans A, Ford C, Hayward RD.. Host-pathogen reorganisation during host cell entry by Chlamydia trachomatis. Microbes Infect 2015; 17:727–31. [DOI] [PMC free article] [PubMed] [Google Scholar] - 11. Igietseme JU, Eko FO, He Q, Black CM.. Antibody regulation of T-cell immunity: implications for vaccine strategies against intracellular pathogens. Expert Rev Vaccines 2004; 3:23–34. [DOI] [PubMed] [Google Scholar] - 12. Naglak EK, Morrison SG, Morrison RP.. IFNγ is required for optimal antibody-mediated immunity against genital Chlamydia infection. Infect Immun 2016; 84:3232–42. [DOI] [PMC free article] [PubMed] [Google Scholar] - 13. Darville T, Albritton HL, Zhong W, et al. Anti-chlamydia IgG and IgA are insufficient to prevent endometrial chlamydia infection in women, and increased anti-chlamydia IgG is associated with enhanced risk for incident infection. Am J Reprod Immunol 2019; 81:e13103. [DOI] [PMC free article] [PubMed] [Google Scholar] - 14. Hufnagel K, Lueong S, Willhauck-Fleckenstein M, et al. Immunoprofiling of Chlamydia trachomatis using whole-proteome microarrays generated by on-chip in situ expression. Sci Rep 2018; 8:7503. [DOI] [PMC free article] [PubMed] [Google Scholar] - 15. Russell AN, Zheng X, O’Connell CM, et al. Analysis of factors driving incident and ascending infection and the role of serum antibody in Chlamydia trachomatis genital tract infection. J Infect Dis 2016; 213:523–31. [DOI] [PMC free article] [PubMed] [Google Scholar] - 16. Wang SP, Grayston JT.. Human serology in Chlamydia trachomatis infection with microimmunofluorescence. J Infect Dis 1974; 130:388–97. [DOI] [PubMed] [Google Scholar] - 17. O’Connell CM, Brochu H, Girardi J, et al. Simultaneous profiling of sexually transmitted bacterial pathogens, microbiome, and concordant host response in cervical samples using whole transcriptome sequencing analysis. Microb Cell 2019; 6:177–83. [DOI] [PMC free article] [PubMed] [Google Scholar] - 18. Anders S, Pyl PT, Huber W.. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics 2015; 31:166–9. [DOI] [PMC free article] [PubMed] [Google Scholar] - 19. Hufnagel K, Hoenderboom B, Harmel C, et al. Chlamydia trachomatis whole-proteome microarray analysis of the Netherlands Chlamydia cohort study. Microorganisms 2019; 7:1–18. [DOI] [PMC free article] [PubMed] [Google Scholar] - 20. R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2014. http://www.R-project.org/ [Google Scholar] - 21. Russell AN, Zheng X, O’Connell CM, et al. Identification of Chlamydia trachomatis antigens recognized by T cells from highly exposed women who limit or resist genital tract infection. J Infect Dis 2016; 214:1884–92. [DOI] [PMC free article] [PubMed] [Google Scholar] - 22. Galaleldeen A, Taylor AB, Chen D, et al. Structure of the Chlamydia trachomatis immunodominant antigen Pgp3. J Biol Chem 2013; 288:22068–79. [DOI] [PMC free article] [PubMed] [Google Scholar] - 23. Hou S, Dong X, Yang Z, Li Z, Liu Q, Zhong G.. Chlamydial plasmid-encoded virulence factor Pgp3 neutralizes the antichlamydial activity of human cathelicidin LL-37. Infect Immun 2015; 83:4701–9. [DOI] [PMC free article] [PubMed] [Google Scholar] - 24. Liu Y, Huang Y, Yang Z, et al. Plasmid-encoded Pgp3 is a major virulence factor for Chlamydia muridarum to induce hydrosalpinx in mice. Infect Immun 2014; 82:5327–35. [DOI] [PMC free article] [PubMed] [Google Scholar] - 25. O’Connell CM, AbdelRahman YM, Green E, et al. Toll-like receptor 2 activation by Chlamydia trachomatis is plasmid dependent, and plasmid-responsive chromosomal loci are coordinately regulated in response to glucose limitation by C. trachomatis but not by C. muridarum. Infect Immun 2011; 79:1044–56. [DOI] [PMC free article] [PubMed] [Google Scholar] - 26. Chen D, Lei L, Lu C, Galaleldeen A, Hart PJ, Zhong G.. Characterization of Pgp3, a Chlamydia trachomatis plasmid-encoded immunodominant antigen. J Bacteriol 2010; 192:6017–24. [DOI] [PMC free article] [PubMed] [Google Scholar] - 27. Stanhope R, Flora E, Bayne C, Derré I.. IncV, a FFAT motif-containing Chlamydia protein, tethers the endoplasmic reticulum to the pathogen-containing vacuole. Proc Natl Acad Sci U S A 2017; 114:12039–44. [DOI] [PMC free article] [PubMed] [Google Scholar] - 28. Armitage CW, O’Meara CP, Harvie MC, Timms P, Blumberg RS, Beagley KW.. Divergent outcomes following transcytosis of IgG targeting intracellular and extracellular chlamydial antigens. Immunol Cell Biol 2014; 92:417–26. [DOI] [PMC free article] [PubMed] [Google Scholar] - 29. Randall A, Teng A, Liang X, et al. A primary Chlamydia trachomatis genital infection of rhesus macaques identifies new immunodominant B-cell antigens. PLoS One 2021; 16:e0250317. [DOI] [PMC free article] [PubMed] [Google Scholar] - 30. Teng A, Cruz-Fisher MI, Cheng C, et al. Proteomic identification of immunodominant chlamydial antigens in a mouse model. J Proteomics 2012; 77:176–86. [DOI] [PMC free article] [PubMed] [Google Scholar] - 31. Li Z, Palaniyandi S, Zeng R, Tuo W, Roopenian DC, Zhu X.. Transfer of IgG in the female genital tract by MHC class I-related neonatal Fc receptor (FcRn) confers protective immunity to vaginal infection. Proc Natl Acad Sci U S A 2011; 108:4388–93. [DOI] [PMC free article] [PubMed] [Google Scholar] - 32. Wang J, Zhang Y, Lu C, Lei L, Yu P, Zhong G.. A genome-wide profiling of the humoral immune response to Chlamydia trachomatis infection reveals vaccine candidate antigens expressed in humans. J Immunol 2010; 185:1670–80. [DOI] [PubMed] [Google Scholar] - 33. Murthy AK, Chambers JP, Meier PA, Zhong G, Arulanandam BP.. Intranasal vaccination with a secreted chlamydial protein enhances resolution of genital Chlamydia muridarum infection, protects against oviduct pathology, and is highly dependent upon endogenous gamma interferon production. Infect Immun 2007; 75:666–76. [DOI] [PMC free article] [PubMed] [Google Scholar] - 34. Mosolygó T, Szabó AM, Balogh EP, et al. Protection promoted by pGP3 or pGP4 against Chlamydia muridarum is mediated by CD4+ cells in C57BL/6N mice. Vaccine 2014; 32:5228–33. [DOI] [PubMed] [Google Scholar] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article.