Plasmid Copy Number and Disease Severity in Naturally Occurring Ocular Chlamydia trachomatis Infection

The Chlamydia trachomatis plasmid is a virulence factor. Plasmid copy number, C. trachomatis load and disease severity were assessed in a treatment-naive population where trachoma is hyperendemic. By using droplet digital PCR, plasmid copy number was found to be stable (median, 5.34 [range, 1 to 18]) and there were no associations with C. trachomatis load or disease severity.

The Chlamydia trachomatis plasmid is a virulence factor. Plasmid copy number, C. trachomatis load and disease severity were assessed in a treatment-naive population where trachoma is hyperendemic. By using droplet digital PCR, plasmid copy number was found to be stable (median, 5.34 [range, 1 to 18]) and there were no associations with C. trachomatis load or disease severity. T rachoma is caused by infection with ocular strains of Chlamydia trachomatis. The 7.5-kb C. trachomatis plasmid has been shown to function as a virulence factor in animal models (1,2). Phenotypic differences exist between plasmid-cured and wildtype C. trachomatis strains with respect to infectivity, glycogen accumulation, induction of inflammation, and activation of Tolllike-receptor pathways (3,4). Plasmid deletion mutagenesis studies showed that deletion of the plasmid-borne pgp4 gene results in an in vitro phenotype identical to that of a plasmid-free strain (5). This supports bacterial transcriptome analysis showing a decrease in transcript levels of a subset of chromosomal genes in a naturally occurring plasmid-free strain of C. trachomatis, demonstrating that the plasmid is a transcriptional regulator of virulence-associated genes (6).
There is little information in the literature relating plasmid copy number (per genome) to virulence (7)(8)(9). The mechanisms of plasmid virulence are not clearly defined, particularly in naturally occurring infections. We assessed plasmid copy number variation and its association with disease severity in ocular C. trachomatis infection from a treatment-naive population on the Bijagós Archipelago of Guinea Bissau where trachoma is hyperendemic.
This study was conducted in accordance with the declaration of Helsinki. Ethical approval was obtained from the Comitê Nacional de Ética e Saúde (Guinea Bissau), the LSHTM Ethics Committee (United Kingdom), and the Gambia Government/MRC Joint Ethics Committee (The Gambia). Written (thumbprint or signature) informed consent was obtained from all study participants or their guardians as appropriate. Following the survey all communities on the study islands were treated with azithromycin in line with WHO and national protocols.
Individuals from 300 randomly selected households from 38 villages on four islands were examined by a single trained examiner using the simplified WHO and modified FPC grading systems (10,11). In the FPC system, follicles (F), papillae (P), and conjunctival scarring (C) are separately scored on a scale of 0 to 3. Active disease (TF [follicular trachoma] or TI [inflammatory trachoma] according to the simplified WHO system) equates to F2/3 and P3, respectively. C2/3 (and in some cases C1) is equivalent to TS (trachomatous scarring). Both systems were used to provide detailed phenotypic information and comparability with other studies. Individuals' age, sex, and ethnicity were recorded.
Swabs were taken from the left upper tarsal conjunctiva of each participant using a validated procedure (12,13). Swabs were col-lected dry into microcentrifuge tubes (Simport, Canada), kept on ice in the field, and frozen to Ϫ80°C within 8 h of collection. Measures were taken to avoid cross-contamination in the field and in the laboratory (13).
DNA extraction and droplet digital PCR (ddPCR) for detection of C. trachomatis plasmid were conducted as described previously (14). A second duplex assay was used to estimate plasmid and chromosome (omcB) target concentrations within the same reaction in plasmid-positive samples. We used published primerprobe target sequences appropriate for quantitation of all genovars of C. trachomatis (7,14). We used a modified omcB probe to improve quenching efficiency and reduce background fluorescence (Table 1). Methods for master mix preparation, droplet generation, thermal cycling conditions, droplet reading, target DNA concentration calculation, and retesting of saturated samples are described elsewhere (14). Estimated quantities of omcB and plasmid are expressed as copies/swab. C. trachomatis load refers to omcB copies/swab. Plasmid copy number (per genome) was calculated using the plasmid/genome ratio.
Raw quantitation data were processed as previously described (14). Geometric mean omcB load and linear and logistic regression analyses (with odds ratios [OR]) were conducted in STATA 12 (Stata Corporation, College Station, TX) to examine associations between plasmid copy number, load, and detailed clinical phenotype. C. trachomatis load and plasmid copy number data were log(e) transformed, and robust standard error was used where indicated.
The theoretical advantages of ddPCR are presented by Hindson et al. (15). These include nanoliter-sized droplet partitioning of the reaction, which promotes optimal primer-template interaction conditions robust to variation in PCR efficiency, thus enabling accurate estimation of both plasmid and omcB copy numbers within the same reaction. We have discussed the precision and accuracy of our diagnostic ddPCR assay elsewhere (14).   There are a few published studies examining plasmid copy number in reference strains of C. trachomatis (7-9, 16, 17). Pickett et al. showed that across 12 C. trachomatis serovars, the plasmid copy number was not significantly different, but there were variations depending on growth phase and condition during in vitro culture (7). Seth-Smith et al. showed an increased plasmid copy number in ocular relative to urogenital strains (8). We demonstrate a stable plasmid copy number distribution in naturally occurring ocular C. trachomatis infection that does not vary with geographic location, clinical phenotype, or C. trachomatis load. Our data show that ddPCR may have limitations in measuring plasmid copy number in very-low-load infections (Ͻ200 omcB copies/swab), where plasmid copy number variance is greatest. This observation may reflect a breakdown in the assumptions required to apply the Poisson distribution to accurately estimate load with ddPCR. Despite the caveats, our data suggest plasmid copy number stability in naturally occurring ocular C. trachomatis infection.
Maintenance of the plasmid at low copy numbers carries an inherent risk during cell partition (18), but naturally occurring plasmid-free strains are rare (19)(20)(21). A lower-risk, higher-copynumber system is metabolically expensive but may confer a fitness advantage. Thus, the maintenance of 5 or 6 plasmids per genome may maximize infectivity or intracellular survival while provoking minimal host immune response.
Though there is convincing evidence that the chlamydial plasmid is a virulence factor (3,4,6,(22)(23)(24), our data suggest that plasmid copy number is not associated with disease severity and that additive gene dosage effects do not appear to correlate with pathogen virulence in vivo. This supports in vitro work showing no association between plasmid copy number and tissue tropism (9). Previous work in vitro and in animal models suggests that subtle genomic differences between chlamydial isolates are associated with differences in growth kinetics, immune responses, and pathology (25,26). Further epidemiological and in vitro studies using comparative pathogen genomics to examine these associations are required to fully understand the relationship between disease severity and chlamydial virulence.