Genetic Variation in the MBL2 Gene Is Associated with Chlamydia trachomatis Infection and Host Humoral Response to Chlamydia trachomatis Infection

This study aims to assess the potential association of MBL2 gene single nucleotide polymorphisms (SNPs) to Chlamydia trachomatis infection. We analysed a selected sample of 492 DNA and serum specimens from Dutch Caucasian women. Women were categorized into four groups of infection status based on the results of DNA and antibody tests for C. trachomatis: Ct-DNA+/IgG+, Ct-DNA+/IgG−, Ct-DNA−/IgG+, and Ct-DNA−/IgG−. We compared six MBL2 SNPs (−619G > C (H/L), −290G > C (Y/X), −66C > T (P/Q), +154C > T (A/D), +161A > G (A/B), and +170A > G (A/C)) and their respective haplotypes in relation to these different subgroups. The −619C (L) allele was less present within the Ct-DNA−/IgG+ group compared with the Ct-DNA−/IgG− group (OR = 0.49; 95% CI: 0.28–0.83), while the +170G (C) allele was observed more in the Ct-DNA+/IgG+ group as compared with the Ct-DNA−/IgG− group (OR = 2.4; 95% CI: 1.1–5.4). The HYA/HYA haplotype was more often present in the Ct-DNA−/IgG− group compared with the Ct-DNA+/IgG+ group (OR = 0.37; 95% CI: 0.16–0.87). The +170G (C) allele was associated with increased IgG production (p = 0.048) in C. trachomatis PCR-positive women. This study shows associations for MBL in immune reactions to C. trachomatis. We showed clear associations between MBL2 genotypes, haplotypes, and individuals’ stages of C. trachomatis DNA and IgG positivity.


Introduction
Chlamydia trachomatis is the most prevalent bacterial sexually transmitted infection (STI) worldwide. The course of infection is variable; a recently acquired infection can have an active course, may be asymptomatic, or can be self-limiting. Untreated infections have a severe impact on the health of a patient owing to the possible development of late complications such as pelvic inflammatory disease, tubal pathology, and ectopic pregnancy [1].
The innate immune response is, similar to any other infectious disease, the first line of defence against infection by C. trachomatis [2]. Mannose-binding lectin (MBL) is an acute 2 of 11 phase protein produced by the liver and has a central role in the innate immune response; MBL recognises and binds to patterns of glycoproteins present on microorganisms enabling opsonisation [3]. This C-type serum lectin binds to the 40 kDa major outer membrane protein of C. trachomatis, hampering invasion and infection of the host cell [4]. After binding of MBL to the outer member protein, a conformational change in mannose associated serine proteases (MASP-1 and MASP-2) occurs, which activates the lectin pathway of the complement system [5]. The MASP proteases cleave C4 and C2, generating C4b2a complexes that have C3 convertase activity [6,7]. MBL functions in close proximity to immunoglobulins (Ig) and facilitates opsonisation by macrophages [8].
The gene encoding the human MBL protein, MBL2, is located on chromosome 10 (10q11. 2-q21). This MBL2 gene incorporates four exons encoding a cysteine-rich region as well as a glycine-rich, collagen-like region (exon 1 and 2); a "neck" region (exon 3); and a carbohydrate-binding domain (exon 4) [9,10]. Six single-nucleotide polymorphisms (SNPs) of MBL2 have been described [11]; that is, two SNPs in the promoter region: −619 G > C (H/L), −290 G > C (Y/X); one SNP at the 5' untranslated region within the leader sequence: −66 C > T (P/Q); and three SNPs in exon 1: +154 C > T (A/D), +161 A > G (A/B), and +170 A > G (A/C). Figure 1 shows the rs-numbers and the relative positions of the SNPs on the gene. The promoter alleles are in strong linkage disequilibrium with the SNPs in exon 1, which generally results in seven haplotypes: HYPA, LYPA, LYQA, LXPA, HYPD, LYPB, and LYQC [12]. The fourth position in the haplotype, A, B, C, or D, is a combination of the three polymorphisms in exon 1, with A being the wild-type allele in all three positions and B, C, or D representing an SNP in the corresponding locus.
The innate immune response is, similar to any other infectious disease, the first line of defence against infection by C. trachomatis [2]. Mannose-binding lectin (MBL) is an acute phase protein produced by the liver and has a central role in the innate immune response; MBL recognises and binds to patterns of glycoproteins present on microorganisms enabling opsonisation [3]. This C-type serum lectin binds to the 40 kDa major outer membrane protein of C. trachomatis, hampering invasion and infection of the host cell [4]. After binding of MBL to the outer member protein, a conformational change in mannose associated serine proteases (MASP-1 and MASP-2) occurs, which activates the lectin pathway of the complement system [5]. The MASP proteases cleave C4 and C2, generating C4b2a complexes that have C3 convertase activity [6,7]. MBL functions in close proximity to immunoglobulins (Ig) and facilitates opsonisation by macrophages [8].
The gene encoding the human MBL protein, MBL2, is located on chromosome 10 (10q11.2-q21). This MBL2 gene incorporates four exons encoding a cysteine-rich region as well as a glycine-rich, collagen-like region (exon 1 and 2); a "neck" region (exon 3); and a carbohydrate-binding domain (exon 4) [9,10]. Six single-nucleotide polymorphisms (SNPs) of MBL2 have been described [11]; that is, two SNPs in the promoter region: −619 G > C (H/L), −290 G > C (Y/X); one SNP at the 5' untranslated region within the leader sequence: −66 C > T (P/Q); and three SNPs in exon 1: +154 C > T (A/D), +161 A > G (A/B), and +170 A > G (A/C). Figure 1 shows the rs-numbers and the relative positions of the SNPs on the gene. The promoter alleles are in strong linkage disequilibrium with the SNPs in exon 1, which generally results in seven haplotypes: HYPA, LYPA, LYQA, LXPA, HYPD, LYPB, and LYQC [12]. The fourth position in the haplotype, A, B, C, or D, is a combination of the three polymorphisms in exon 1, with A being the wild-type allele in all three positions and B, C, or D representing an SNP in the corresponding locus. There is a strong association between the genotype of the MBL2 gene and the level of MBL protein production. The haplotypes HYPA and LYQA are associated with high concentrations of MBL; LYPA and LXPA with intermediate/low concentrations; and HYPD, LYQC, and LYPB with MBL deficiency [11,13,14]. It has been shown that SNPs in exon 1 reduce the functionality of the protein and decrease MBL concentrations, thereby greatly reducing its complement-activating ability [15,16]. Serum MBL concentration is a determinant of susceptibility to infectious diseases and of disease outcome, and shows a strong correlation with allelic variants of the gene [11,13,17,18]. For example, it has been shown that children with exon 1 variants of the MBL2 gene were more susceptible to meningococcal disease than children with wildtype alleles [19].
Owing to the effect of MBL2 polymorphisms on susceptibility to infectious diseases, we aim to evaluate the role of these six known SNPs in the MBL2 gene to a C. trachomatis infection. We will divide our study population into four biological subgroups and assess There is a strong association between the genotype of the MBL2 gene and the level of MBL protein production. The haplotypes HYPA and LYQA are associated with high concentrations of MBL; LYPA and LXPA with intermediate/low concentrations; and HYPD, LYQC, and LYPB with MBL deficiency [11,13,14]. It has been shown that SNPs in exon 1 reduce the functionality of the protein and decrease MBL concentrations, thereby greatly reducing its complement-activating ability [15,16]. Serum MBL concentration is a determinant of susceptibility to infectious diseases and of disease outcome, and shows a strong correlation with allelic variants of the gene [11,13,17,18]. For example, it has been shown that children with exon 1 variants of the MBL2 gene were more susceptible to meningococcal disease than children with wildtype alleles [19].
Owing to the effect of MBL2 polymorphisms on susceptibility to infectious diseases, we aim to evaluate the role of these six known SNPs in the MBL2 gene to a C. trachomatis infection. We will divide our study population into four biological subgroups and assess the role of these six SNPs, and we will determine the role of the SNPs in the production of IgG.

Study Sample Characteristics
The median age of women in this analysis was 23 years (15-32 years). A total number of 65 (13%) samples were PCR positive, 73 (15%) samples were IgG positive, 139 (28%) sam-ples were positive for both parameters, and 215 (44%) samples were negative for both C. trachomatis DNA and IgG. The median IgG response was 100 (50-1600). In total, 216 (44%) samples were obtained from women who reported STI-related symptoms and 159 (32%) had co-infection with other microorganisms upon inclusion; in particular, Candida albicans was prevalent (n = 144).

SNP Distribution
All genotypes analysed in this study were in Hardy-Weinberg equilibrium. Table 1 shows the overall SNP distribution and distribution of SNPs in relation to C. trachomatis infection.

Association of SNPs and Stadium of Infection
We observed a significant difference in the carriage of the C (L) allele of the −619 SNP between the Ct-DNA−/IgG+ and Ct-DNA−/IgG− groups (p = 0.027). This observation was also shown in multivariate analysis for the Ct-DNA−/IgG+ group (OR: 1.5; 95% CI: Distribution analysis between the Ct-DNA+/IgG− group and Ct-DNA−/IgG− group shows reduced carriage of the −290 C (X) allele in the Ct-DNA−/IgG− group (p trend = 0.067).
Logistic regression analysis shows increased carriage of the +170 G (C) allele in the Ct-DNA+/IgG+ group compared with the Ct-DNA−/IgG− group, although this did not reach statistical significance (p = 0.06).
The SNP distribution did not differ significantly between the Ct-DNA−/IgG+ group and the Ct-DNA+/IgG− group. No significant differences in SNP distribution were observed when comparing the Ct-DNA+/IgG− group and the Ct-DNA+/IgG+ group.
No differences were observed when the Ct-DNA+/IgG− and Ct-DNA+/IgG+ groups were combined and compared to the Ct-DNA−/IgG+ group or Ct-DNA−/IgG− group, nor when these latter groups were combined in both univariate and multivariate analyses.

Effect of Homozygous and Heterozygous Carriage of Alleles and Susceptibility to Infection
Comparisons were made between homozygous and heterozygous carriage of alleles in order to assess susceptibility to infection. The following observations were made: the −619 C (L) allele was significantly more often present in the Ct-DNA−/IgG+ group compared with the Ct-DNA−/IgG− group (OR: 0.49, 95% CI: 0.28-0.83; p = 0.009). Another difference was observed comparing carriage of the +170 G (C) allele in the Ct-DNA+/IgG+ group compared with the Ct-DNA−/IgG− group (OR: 2.4, 95% CI: 1.1-5.4; p = 0.027).

Haplotype Frequencies and Associations with Susceptibility
The haplotype distribution among all subgroups is summarized in

Association of SNPs with Immune Response
All PCR-positive women (n = 204) were included in this sub-analysis; 65 were IgG antibody negative and 139 were IgG positive. When comparing the distribution of SNPs with presence of IgG titre in C. trachomatis-positive women, carriage of the +170 G (C) allele was associated with presence of IgG in those C. trachomatis-positive women (p = 0.048, 3% vs. 12%), but median titre did not differ significantly between +170 AA (AA) and *G (*C) alleles (median IgG titre +170 AA (AA): 100 (range 0-1600), median IgG titre +170 *G (*C): 100 (range 0-400); Mann-Whitney U p = 0.32). No other associations for SNPs and IgG response were observed. Coinfection with other diagnosed microorganisms proved not to be a confounding factor.

Discussion
This study confirms an effect of polymorphisms in the MBL2 gene in susceptibility to infection and the humoral IgG response to C. trachomatis infection. We divided our study population into four biological subgroups (Ct-DNA+/IgG+, Ct-DNA+/IgG−, Ct-DNA−/IgG+, and Ct-DNA−/IgG−) and have observed different associations between these subgroups. By introducing these subgroups, we were able to investigate the role of the MBL2 SNPs in different possible stages of infection. It is known that genetic variants of the gene and associated variation in MBL concentration influence the susceptibility to and outcome of a wide variety of infectious diseases [17,19,20]. However, the associations described in this report are, especially with regard to the humoral response, the first for C. trachomatis infection.
The observed SNP frequency distribution was similar to that observed in other studies performed in Europe [12,14]. We observed a difference in the distribution of the +170 G (C) allele; that is, the G (C) allele was more often present in women included in the Ct-DNA+/IgG+ group than those in the Ct-DNA−/IgG− group. The G (C) allele corresponds to an inadequate MBL and low MBL-producing haplotype. When combining the haplotypes corresponding to high-producing, low-producing, and deficient haplotypes, we did not observe any difference in distribution comparing the subgroups to the Ct-DNA−/IgG− group. This may be because of the division of ten haplotype combinations over the four subgroups, resulting in low We observed a difference in the distribution of the +170 G (C) allele; that is, the G (C) allele was more often present in women included in the Ct-DNA+/IgG+ group than those in the Ct-DNA−/IgG− group. The G (C) allele corresponds to an inadequate MBL and low MBL-producing haplotype. When combining the haplotypes corresponding to highproducing, low-producing, and deficient haplotypes, we did not observe any difference in distribution comparing the subgroups to the Ct-DNA−/IgG− group. This may be because of the division of ten haplotype combinations over the four subgroups, resulting in low haplotype frequencies per subgroup. Our third observation was the relation between the G (C) allele at position +170 and host IgG antibody production against C. trachomatis. This is relevant to know because increased susceptibility to C. trachomatis and prolonged infection may increase the chances of developing late complications [21]. Finally, we observed statistical trends in the Ct-DNA+/IgG− group and the Ct-DNA+/IgG+ group.
In vitro models support the role of MBL2 in susceptibility to C. trachomatis infection. A previous in vitro study has shown the inhibitive role of MBL to Chlamydia spp. infections, including to C. trachomatis and Chlamydia pneumoniae infections, suggesting MBL has an influence on immunity to these infections [4]. Sziller et al. [22] observed, in a group of Hungarian women with proven tubal infertility, a significantly higher frequency of the B allele (+161 G) than in healthy controls. They hypothesize that a defect in first-line defence due to the polymorphism in MBL2 contributes to persistence of the bacterium, which leads to damage of the Fallopian tubes. We do not find any association with this allele. This is not contradictory, as we did not assess the fertility of women in our analysis. It is possible that the role and/or mechanism of MBL is different for acquiring the infection than in complicating the course of disease. Two studies performed by Laisk and colleagues [23,24] investigated haplotypes in relation to C. trachomatis-induced tubal factor infertility (TFI) and observed that low-producing haplotypes (LXA/LXA, HYA/O, and LYA/O) were risk factors for developing TFI. They did not discover any association between MBL-deficient haplotypes or the very-low-producing haplotypes and TFI.
A relation between MBL deficiency and IgG has been observed previously by Roos et al. [25]. They pre-incubated mannose-coated plates with purified IgG or IgM antibodies and measured C4 deposition, a complement protein, upon addition of MBL-deficient serum. C4 concentrations were similar to that of MBL-sufficient serum, indicating a restorative role of antibodies for complement activation. This immunological redundancy may explain the variable inter-individual clinical outcome of disease in MBL-deficient persons, and has been proposed previously [26,27].
Carriage of an SNP in exon 1 reduces the functionality of MBL and additionally decreases MBL serum concentrations [15,16]. We have shown that patients carrying the G (C) allele at +170 were more likely to produce IgG antibodies than patients who had well-functioning MBL. Although this study has a relatively low number of patients with this allele, we believe that the observation is biologically plausible as the structural defect in MBL can be compensated for in vitro [28]. Laisk et al. [24] have also observed that the highproducing haplotype HYA/HYA was associated with TFI independent of C. trachomatis infection, but they could not confirm their previous results [23] when analysing MBL2 genotypes and C. trachomatis-induced TFI [24]. We also find an association with this haplotype, albeit a protective one. Taking these results together, it seems that a highproducing haplotype is protective for C. trachomatis infection because of its complement activating ability, but also increases the risk of tubal pathology. This is an indication that MBL is important in immunity for C. trachomatis, but needs to be tightly regulated to prevent collateral damage.
The strengths of this study are that we used clearly defined subgroups on the basis of the presence of C. trachomatis DNA and/or specific IgG serum titres. We show associations of SNPs in different stadia of the infection, which indicates the significant immunological role of MBL to C. trachomatis infection, and may possibly have an effect on the clinical outcome.
This study has several limitations. Despite the sample size, the frequency of variants of exon 1 is relatively low, making it difficult to link susceptibility to C. trachomatis to one of these mutations. Additionally, their impact at the population level is expected to be limited because of the low frequency of these exon 1 SNPs in this population [29].
Furthermore, data on other confounding factors, such as Mycoplasma genitalum infection, bacterial vaginosis, birth control, and C. trachomatis virulence and load, are not available and may have influenced the results.
Creating biological subgroups was preferred over combining individuals based on parameters defining C. trachomatis infection, despite its potential limitations. For example, it is unknown when the individuals from the Ct-DNA−/IgG+ group were infected with C. trachomatis, or whether individuals from the Ct-DNA−/IgG− group actually had exposure. Moreover, not every infected individual will generate an antibody response, so this may have introduced some bias.
More research should be performed to assess the role of MBL to susceptibility to C. trachomatis. Larger, prospective studies should be conducted to gain further insight into the hypothesized immunological redundancy of antibodies and its effect on infection when an individual has an MBL2 genotype coding for low or deficient MBL. Although an association between MBL and IgG production has not been observed, studies assessing this can be of interest because C. trachomatis serovars can induce different serological responses [30,31].
The results obtained from immunogenetic studies such as this one are of high relevance for public health and healthcare in general. Our results contribute to the understanding of disease pathogenesis of C. trachomatis infection, and our findings provide new insights into the immunological pathways that may contribute to the variable clinical course of the infection. Studies investigating SNPs in, for example, interleukin pathways are of similar importance and enhance the knowledge of chlamydial infection. Furthermore, our results may be integrated with existing immunogenetic knowledge, possibly aiding in the development of targeted and personalized approaches in the prevention, diagnosis, and treatment of the infection [28,32,33].
To conclude, this study suggests a role for MBL in immunological response to C. trachomatis. We observed associations of MBL2 genotype and the stage of infection, and a clue for possible immunological redundancy was observed.

Sample Collection
A total of 492 samples were randomly selected from a previous case-control study [20]. Those samples were obtained from Dutch Caucasian women (age 15-35 years old) attending the STI outpatient clinic in Amsterdam, the Netherlands. Ethnicity (Dutch Caucasian) was self-reported and by means of questionnaires, which has been shown to be highly valid and representative in this context [34]. Upon inclusion, cervical swabs and serum samples were obtained from these women. Moreover, demographical, clinical, and laboratory data were available including age, symptoms, presence of coinfection systematically tested (including Candida albicans, Neisseria gonorrhoeae, and Trichomonas vaginalis), results of PCR test for C. trachomatis (COBAS Amplicor), and the C. trachomatis-specific serum IgG titre (medac Diagnostika).

Ethical Approval
The Medical Research Involving Human Subjects Act (WMO, Dutch Law) stating official approval of the study by the Medical Ethical Committee does not apply to our collected anonymous human material (MEC Letter reference: #10.17.0046). The previous case-control study [35], from which the samples in this study were selected, was approved by the University's Medical Ethical Committee. All participants of that study provided informed consent to use their samples anonymously for future research.

Laboratory Tests
The methods used for the detection and extraction of C. trachomatis DNA and determination of IgG serum titres have been described elsewhere [35]. In short, C. trachomatis detection was performed with COBAS Amplicor (Hoffman-La Roche, Basel, Switzerland) from DNA extracted from the cervical swab and was determined positive with when the Ct value was below 40. DNA for the MBL genotyping assay was extracted from peripheral blood mononuclear cells (PBMCs) by means of the isopropanol isolation method; a mixture of PBMC in PBS, nuclisense lysisbuffer, and glycogen was incubated for 30 min at 65 degrees Celsius and left to cool at room temperature. Isopropanol was added to the mixture and the samples were centrifuged. The supernatant was discarded and the remaining pellets were washed twice with 75% EtOH. The pellets were dissolved in T10 and stored for later analysis. The presence of IgG was determined with a C. trachomatis-specific ELISA (medac Diagnostika, Hamburg, Germany). Samples with a titre of ≥1:50 were considered positive.

Subgroups for Analyses
We classified the patients in this analysis into four subgroups to assess susceptibility for C. trachomatis. The first group includes women with both positive PCR and IgG determinations (Ct-DNA+/IgG+). The second subgroup includes women with a positive C. trachomatis PCR and negative IgG titre (Ct-DNA+/IgG−). The third subgroup includes women with a positive C. trachomatis-specific IgG titre and negative PCR result. The fourth subgroup represents women with both negative PCR and a negative C. trachomatis IgG titre (Ct-DNA−/IgG−).  To assess the role of MBL genotypes and haplotypes in initiating a humoral immune response, the following analysis was conducted: genotype and haplotype distributions were compared among women with a positive PCR result with and without IgG response to assess the potential associations of MBL2 genotypes and haplotypes and serum IgG response to C. trachomatis infection. The classification by Steffensen  Allele O in this context is any promoter combination with any mutant allele of exon 1, whereas A in this context is any promoter combination with the wildtype alleles of the three exon 1 SNPs. The role of the P/Q allele is limited for MBL concentrations, so we did not include it in our haplotype [13]. To assess the role of MBL genotypes and haplotypes in initiating a humoral immune response, the following analysis was conducted: genotype and haplotype distributions were compared among women with a positive PCR result with and without IgG response to assess the potential associations of MBL2 genotypes and haplotypes and serum IgG response to C. trachomatis Allele O in this context is any promoter combination with any mutant allele of exon 1, whereas A in this context is any promoter combination with the wildtype alleles of the three exon 1 SNPs. The role of the P/Q allele is limited for MBL concentrations, so we did not include it in our haplotype [13].

Statistical Analyses
Descriptive statistics are provided and presented as the number (%) and median (range). All SNPs were assessed for Hardy-Weinberg equilibrium to test for deviation of Mendelian inheritance. Cross-tabulation of SNPs and haplotypes in women with and without C. trachomatis infection was performed, including χ 2 statistics. Forward conditional multivariate regression analysis including all SNPs was used to observe associations found in univariate analysis. Finally, χ 2 statistics and multivariate regression analyses were performed to assess the relation between the individual SNPs and dichotomised IgG production. Owing to the limited number of SNPs, correction for multiple testing was not performed to prevent underestimation of possible associations [40]. Analyses were performed using SPSS 13.0 (SPSS Inc., Chicago, IL, USA). A p-value of less than 0.05 was considered statistically significant, whereas 0.05 < p < 0.07 was considered a statistical trend.

Informed Consent Statement:
The previous case-control study [35], from which the samples in this study were selected, was approved by the University's Medical Ethical Committee. All participants of that study provided informed consent to use their samples anonymously for future research.
Data Availability Statement: Data can be provided on reasonable request.