Distribution Study of Chlamydia trachomatis Serovars among High-Risk Women in China Performed Using PCR-Restriction Fragment Length Polymorphism Genotyping

Abstract

This was one of the first epidemiological studies in China focused on genital Chlamydia trachomatis serotype distribution in high-risk female populations using omp1 gene-based restriction fragment length polymorphism analysis. One thousand seven hundred seventy cervical swab samples from women attending sexually transmitted disease clinics and female sex workers in six cities in China (Shenzhen and Guangzhou in southern China, Nanjing and Shanghai in eastern China, and Nanning and Chengdu in southwestern China) were subjected to serovar genotyping. The proportion of omp1 genes successfully amplified in 240 C. trachomatis plasmid-positive samples was 94.2% (226/240). Serotypes E (n = 63; 27.9%), F (n = 53; 23.5%), G (n = 28; 12.4%), and D (n = 25; 11.1%) were most prevalent. Though there was no significant difference in the geographic distribution of C. trachomatis, serotype E was predominant in the South (32.1%) and East (27.1%), while serotype F was predominant in the Southwest (28.3%). Serotype F infection was associated with young age and single status. Serovar G was associated with lower abdominal pain; 47.5% of asymptomatic patients were infected with serovar E. These results provide information on distribution of genital C. trachomatis serotypes among high-risk women in China and indicate that high-risk women, including those who are asymptomatic, can be infected with multiple serovars of C. trachomatis, revealing exposure to multiple sources of infection. Although the scope for generalizations is limited by our small sample size, our results showing clinical correlations with genotypes are informative.

Genital Chlamydia trachomatis infections (caused by serovars D through K) have been recognized as the most prevalent bacterial sexually transmitted infections throughout the world. The World Health Organization (WHO) estimated that 92 million new cases of C. trachomatis infection occurred throughout the world in 1999 (23). In China, chlamydial infection has not yet been reported as a separate notifiable sexually transmitted infection (2, 19), but a rapid increase in reported cases of nongonococcal urethritis and cervicitis, the majority of which are believed to be caused by C. trachomatis, has been observed in the National STD Surveillance System (4). Besides causing nongonococcal urethritis and cervicitis, chlamydial infections, if undiagnosed and not treated in a timely manner, may result in serious secondary complications and sequelae, including pelvic inflammatory disease, ectopic pregnancy, tubal infertility (14, 18, 22), and increased risk of human immunodeficiency virus transmission and acquisition (9).

Currently, C. trachomatis is classified into 15 different serovars based on immunogenic epitope analysis of the major outer membrane protein (MOMP) with polyclonal and monoclonal antibodies (20). The MOMP is the principal immunodominant surface antigen of C. trachomatis, with antigenic determinants located across four symmetrically spaced variable domains (VDI to VDIV), which are flanked and interspaced by five constant domains. Variable domains are coded by the omp1 gene, and their nucleotide sequences exhibit distinct variations in different serovars. Subsequently, they have become widely used for the genotyping of C. trachomatis isolates (1, 12). Typically, C. trachomatis serovars A through C are found associated with trachoma, serovars D through K are associated with urogenital infections, and serovars L1 through L3 are associated with the systemic disease lymphogranuloma venereum (13, 18, 26). Genotypic characterization of C. trachomatis isolates not only can provide valuable insight into the C. trachomatis serovars circulating within a given community but also can improve understanding of their epidemiology, which may assist in developing strategies for improved sexually transmitted disease (STD) control.

The aims of the present study were to characterize the C. trachomatis strains detected in samples collected from women recruited from different settings in China by use of genotyping and to determine potential correlations between sociodemographic characteristics, geographic distribution, clinical symptoms, and the infecting C. trachomatis genotypes.

MATERIALS AND METHODS

Study population and data collection.

The study population was women recruited from sexually transmitted disease clinics (STDC), female reeducation centers, and sex entertainment venues in six cities, namely, Shanghai and Nanjing (eastern China), Chengdu and Nanning (southwestern China), and Guangzhou and Shenzhen (southern China). The study was approved by the WHO Research Ethics Review Committee and the Medical Ethics Committee of the National Center for STD Control, China Center for Disease Prevention and Control. Informed consent was obtained from subjects prior to participation. Participants were interviewed following a structured questionnaire on sociodemographic background, risk behaviors, and clinical symptoms. All participants were provided with a routine physical examination and health education and counseling.

Sample collection.

Vaginal swab and endocervical swab specimens were collected from each participant. The vaginal swabs were collected using a sterile Dacron-tipped plastic swab (Unipath, Ltd., Bedford, United Kingdom) before the collection of the cervical swab specimens in order to avoid contamination of the vaginal canal with cervical discharge. The vaginal swabs were used for rapid C. trachomatis tests, and results were used locally for clinical services and clinic-based evaluation (data were reported elsewhere [25]). The endocervical swabs were collected with the following steps: the swab was inserted into the endocervical canal until the tip of the swab was no longer visible; the swab was then rotated for 10 s and withdrawn, avoiding contact with vaginal surfaces. Endocervical swabs were temporarily stored at −20°C at the local laboratory before being transported to the National STD Reference Laboratory at the National Center for STD Control in Nanjing for further testing.

Upon arrival of specimens, swabs were thawed at room temperature for about 20 min and were then added to 1.2 ml double-distilled water in a 2-ml sterile polypropylene tube and agitated for 15 s. Each swab was pressed to the side of the tube to express the liquid, and then about 1 ml specimen liquid was prepared.

Plasmid DNA extraction and PCR.

C. trachomatis infection was established through plasmid PCR testing. C. trachomatis plasmid DNA extraction was performed using Amplicor CT/NG specimen preparation kits (Roche Molecular Systems, Inc., Branchburg, NJ) according to the manufacturer's instructions. Briefly, 100 μl CT/NG LYS liquid was added to a 2-ml sterile tube after incubation at room temperature for 30 min. One hundred microliters of specimen was added to the tube and incubated for 10 min to lyse cells, and 200 μl CT/NG DIL was then added and incubated at room temperature for 10 min. Plasmid PCR was performed according to the CT/NG AMPLICOR kit manufacturer's instructions. The kit carried the CT/NG internal control, and C. trachomatis-positive and -negative controls were used at each amplification.

omp1 gene DNA extraction and PCR.

The omp1 gene DNA extraction was performed with plasmid-positive samples using the filter tube-based isolation method (UIQ-10 clinical sample genomic DNA isolation mini kit; Sangon, Shanghai, China). Briefly, an 800-μl specimen of each plasmid PCR-positive sample was incubated at 37°C for 10 min and then centrifuged at 8,000 rpm for 10 min. The supernatant was discarded, and 200 μl Tris-EDTA (pH 8.0) and 400 μl DCL buffer were added to the pellet and vortexed for 15 s. Samples were combined with 100 μg/ml proteinase K, incubated at 55°C for 15 min, and then boiled for 5 min. Two hundred sixty microliters of 100% ethanol was added to processed specimens, and DNA was allowed to precipitate out for 5 min; samples were transferred to the UNIQ-10 spin column, centrifuged at 10,000 rpm for 1 min at 37°C, and washed twice with washing buffer. Finally, DNA absorbed on the column was washed down using 100 μl elution buffer and collected for omp1 PCR.

For omp1 gene amplification, 10 μl of extracted DNA was added to a reaction tube with the primary omp1 PCR mixture containing 50 pmol (each) of primers NLO (5′-ATG AAA AAA CTC TTG AAA TCG-3′) and NRO (5′-CTC AAC TGT AAC TGC GTA TTT-3′), 0.2 mM (each) of dATP, dTTP, dGTP, and dCTP, 1× PCR buffer (20 mM Tris-HCl, pH 8.4, and 50 mM KCl), 1.5 mM MgCl₂, and 4 U Taq DNA polymerase (BioReady Taq pac; BioFlux) (13).

Nested omp1 PCRs were run using 5 μl of omp1 primary PCR products. The 50-μl nested PCR mixture contained 50 pmol (each) of primers NLI (5′-TTT GCC GCT TTG AGT TCT GCT-3′) and NRI (5′-CCG CAA GAT TTT CTA GAT TTC-3′), 0.2 mM (each) of dATP, dTTP, dGTP, and dCTP, 1× PCR buffer (20 mM Tris-HCl, pH 8.4, and 50 mM KCl), 1.5 mM MgCl₂, and 3 U Taq DNA polymerase (BioReady Taq pac; BioFlux). Primary and nested omp1 PCRs were performed using the GeneAmp PCR System 9600 thermocycler under the following conditions: 6 min of denaturation at 95°C, followed by 49 cycles of amplification at 95°C for 1 min, 45°C for 2 min, and 72°C for 2 min. DNA extracted from reference serovar L2 strain culture (U.S. CDC L2 440R) was used as a positive control, while double-distilled water was used as a negative control in each omp1 amplification. The omp1 nested PCR products were visualized after electrophoresis in 1% agarose gels by ethidium bromide straining.

Restriction fragment length polymorphism (RFLP) and genotyping.

Nested omp1 PCR products were digested with 20 μl AluI reaction system. Products were electrophoresed through a polyacrylamide gel (acrylamide/bisacrylamide, 29:1; 12 V/cm for 1.5 h) to enable identification of serovars B/Ba, D, E, F, G, K, L1, L2, and C complex (C, J, H, I, and L3). Serotypes C and J were differentiated after digestion with HinfI (16). Serotypes H, I, and L3 were separated with CfoI digestion. The digested products were electrophoresed through a 7% polyacrylamide gel (acrylamide/bisacrylamide, 29:1; 1× Tris-borate-EDTA; 5 V/cm, 1.5 h) to differentiate serotypes.

Statistical analysis.

Data were entered into a database using the EpiData 3.1 software (EpiData Association, Odense, Denmark). SPSS for Windows 11.0 (SPSS, Inc., Chicago, IL) and MedCalc for Windows 8.1 (MedCalc Software, Mariakerke, Belgium) were used to analyze the data. Statistical analysis of the data was performed using the chi-square or Fisher exact test. A P value of <0.05 was considered to be statistically significant.

RESULTS

The 1,770 endocervical swabs were collected from women, including 1,111 (62.8%) STD patients and 659 (33.2%) female sex workers (FSW); the mean age was 30.34 or 28.16 years, respectively (P < 0.001). One thousand seventy-seven participants were single, 667 were married, and 26 participants refused to give marriage status.

Out of the 1,770 endocervical swabs, 240 were found to be positive for C. trachomatis plasmid PCR, 1,529 were negative, and 1 sample gave an inhibited result. The omp1 gene amplification was successfully performed with 226 (94.2%) C. trachomatis-positive samples. Of the omp1 gene-amplified products, 10 samples produced a multiband result after AluI digestion which could not be identified and 3 samples were illegible after the digestion, possibly because of the low copy numbers of omp1 PCR products. The proportion of plasmid PCR-positive samples successfully genotyped by PCR-RFLP analysis was 88.75% (213/240).

The most prevalent omp1 genotypes corresponded to serovar E (n = 63 [27.9%]), followed by serovars F (n = 53 [23.5%]), G (n = 28 [12.4%]), and D (n = 25 [11.1%]), while serovars I, H, and B/Ba were less prevalent (n = 7 [6.6%], n = 6 [2.7%], and n = 3 [1.3%], respectively) (Table 1). Infection with serovar F was associated with young age and unmarried status (χ² = 4.30 and P = 0.04 and χ² = 4.44 and P = 0.03, respectively). The average age of serovar F-infected individuals was 24.9 years, and more than half (31/53) of serovar F-infected individuals were unmarried. No significant difference was found in distribution of genotypes of C. trachomatis among samples collected from different geographic areas, although serovar E was predominant in the southern cities (32.1%) and the eastern cities (27.1%). Serovar F was most prevalent in the southwestern cities (28.3%). The distribution of serovars among FSW recruited from the sexual entertainment venues or female reeducation centers was not different from that among females attending STDC.

TABLE 1.

Association of genotypes of C. trachomatis with sociodemographic characteristics, geographic distribution, and clinical symptoms

Characteristic of patients	No. of samples	% of samples positive for C. trachomatis genotype
		Ba	D	E	F	G	H	I	J	K	Me	Unf
Age
<25 yr	93	2.2	8.6	29.0	31.2b	11.8	2.2	1.1	4.3	5.4	2.2	2.2
≥25 yr	131	0.8	13.0	27.5	18.3	13.0	3.1	3.8	6.9	7.6	5.3	0.8
Marital status
Married	121	0.8	9.9	28.1	18.2c	15.7	3.3	3.3	8.3	6.6	4.1	1.7
Unmarried	99	2.0	11.1	27.3	31.3	9.1	2.0	3.0	3.0	6.1	4.0	1.0
Area in China
Southern
Shenzhen	56	1.8	12.5	32.1	26.8	7.1	1.8	1.8	3.6	3.6	7.1	1.8
Guangzhou	25	0	16.0	32.0	20.0	12.0	0	4.0	4.0	8.0	4.0	0
Subtotal	81	1.2	13.6	32.1	24.7	8.6	1.2	2.5	3.7	4.9	6.2	1.2
Eastern
Nanjing	20	5.0	10.0	30.0	20.0	15.0	5.0	0	5.0	5.0	0	5.0
Shanghai	65	1.5	9.2	26.2	18.5	15.4	3.1	3.1	12.3	7.7	1.5	1.5
Subtotal	85	2.4	9.4	27.1	18.8	15.3	3.5	2.4	10.6	7.1	1.2	2.4
Southwestern
Nanning	43	0	11.6	25.6	27.9	11.6	2.3	4.7	2.3	7.0	7.0	0
Chengdu	17	0	5.9	17.6	29.4	17.6	5.9	5.9	0	11.8	5.9	0
Subtotal	60	0	10.0	23.3	28.3	13.3	3.3	5.0	1.7	8.3	6.7	0
Group
STDC	154	1.3	11.7	26.0	24.0	13.6	2.6	2.6	7.1	5.8	3.2	1.9
FSW	72	1.4	9.7	31.9	22.2	9.7	2.8	4.2	2.8	8.3	6.9	0
Symptoma
LAP
Yes	33	0	6.1	15.2	15.2	54.5d	3.0	0	0	3.0	3.0	0
No	191	1.6	12.0	30.4	24.6	5.2	2.6	3.7	6.8	7.3	4.2	1.6
AVD
Yes	154	1.9	13.0	19.5d	24.0	12.3	2.6	2.6	7.8	9.1	5.2	1.9
No	69	0	7.2	46.4	21.7	13.0	2.9	4.3	1.4	1.4	1.4	0
Total	226	1.3	11.1	27.9	23.5	12.4	2.7	3.1	5.8	6.6	4.4	1.3

^aLAP, low abdominal pain; AVD, abnormal vaginal discharge.

^bStatistically significant difference (P = 0.04, chi-square test) in proportions of serovar between patients <25 years old and those ≥25 years old.

^cStatistically significant difference (P = 0.04, chi-square test) in proportions of serovar between married and unmarried patients.

^dStatistically significant difference (P < 0.001, chi-square test) in proportions of serovar between symptomatic and asymptomatic patients.

^eMultiple bands after RFLP.

^fUnable to identify after RFLP.

Of the patients who complained of lower abdominal pain, 54.5% (18/33) were infected with serovar G, and the proportion of serovar G infection was significantly higher than that for patients without the complaint (χ² = 58.26; P < 0.001). The majority of patients without abnormal vaginal discharge were more likely to be infected with C. trachomatis serovar E (n = 32; 46.4%; χ² = 15.86; P < 0.001) (Table 1). Nearly half (47.5%; 28/59) of asymptomatic patients (who had neither lower abdominal pain nor abnormal vaginal discharge) were infected with C. trachomatis serovar E. Serovar K was associated with the symptoms of abnormal vaginal discharge at a borderline-significant level (χ² = 3.30; P = 0.06).

DISCUSSION

The genotyping methods are more sensitive and specific than serotyping for C. trachomatis serovar identification (12, 13). Many studies have shown the feasibility of deducing the serotypes of C. trachomatis clinical isolates using PCR-based RFLP or sequencing of the amplified omp1 gene, which encodes the MOMP (8, 10, 15). This technology has provided a valuable and sensitive means for molecular epidemiological analysis to identify high-risk groups and track sexual networks (6, 7). To the best of our knowledge, this is one of the first studies on genotype distribution of C. trachomatis in mainland China (10) and a study with a relative large sample size compared with previous studies (Table 2). The failure rate (5.6%) of nested omp1 PCR in our study was comparable to those reported in many previous studies (5, 11) but much lower than in other studies (10).

TABLE 2.

Distribution of genotypes of C. trachomatis in different populations in Asian countries

Group	Country or area	Reference	Typing method	No. of samples	Genotype distribution
STDC	Taiwan	Hsu et al. (10)	Sequencing	102	E (22%), D (19%), F (16%), J (15%), G (11%), K (11%), H (6%), B (2%)
	Thailand	Yamazaki et al. (24)	RFLP	15	F (60%), E (13%), H (13%), D (7%), K (7%)
	Mainland China	This study	RFLP	154	E (26%), F (24%), G (14%), D (12%), J (7%), K (6%), M (3%), H (3%), I (3%)
FSW	Korea	Lee, et al. (15)	Sequencing	40	E (45%), F (20%), G (15%), D (5%), H (5%), J (2.5%)
	Thailand	Yamazaki et al. (24)	RFLP	56	F (29%), E (20%), K (18%), D (14%), G (7%), H (5%), J (5%), I (2%), Ba (2%)
	Mainland China	This study	RFLP	72	E (32%), F (22%), D (10%), G (10%), K (8%), M (7%), I (4%), H (3%), J (3%)

In a comparison of our results with those for some other Asian countries (Table 2), it can be found that the distribution of C. trachomatis genotypes in mainland China is similar to those observed in the Taiwan area and in Korea (Table 2). However, the distribution pattern of genotypes in southwestern China, where serovar F is the predominant genotype, is more likely to be similar to the distribution in Thailand. Our finding that genotype B was rarely involved in urogenital infections is consistent with previous findings (Table 2).

Although we reported that the proportions of women recruited from different settings testing positive for chlamydial infection were significantly different (3), we did not observe population clustering of serovars with the recruiting sites or any difference in serovar distribution among women from that among men (data not shown), suggesting the exposure risk of these populations might be similar. It was found that the serovar distribution of C. trachomatis infections among the homosexual network (men having sex with men [MSM]) differed from that among the heterosexual network. Lymphogranuloma venereum caused by C. trachomatis serovars L1 to L3 has been a significant problem among MSM in recent years (17, 21). However, as shown in Table 2, no L serotypes were found in these studies. Unfortunately, the current study did not include any subjects recruited from the homosexual network. Further study is needed to investigate the serotype distribution in MSM populations in China.

Analysis of circulating C. trachomatis genotypes among individuals exhibiting clinical symptoms in China has not been described before, despite the serious complications of C. trachomatis infection. In our study population, serovar E was associated with asymptomatic infection, suggesting that the symptom-based screening and preventive strategies are likely to miss the detection of C. trachomatis infections caused by serovar E. The association of serovar K with abnormal vaginal discharge in women reported by Molano et al. (16) was also observed with a statistically borderline significance (P = 0.06). Serovar G was the third-most-common serovar and was significantly associated with low abdominal pain. The previous study also found a significant relationship between the detection of serovar Ga and having clinical symptoms (16).

Some limitations of this study should be acknowledged. First, the selection bias (i.e., the participants were consecutively recruited from the selected clinics or venues) may limit the generalizability of the findings. However, this possibility can be diminished by the fact that the randomized recruitment of participants may not be the principal focus of such studies. Second, the representative distribution of genotypes may be affected due to the small sample sizes of the study in the target populations. Third, the genotyping methods (PCR-RFLP) used to identify the genotypes of C. trachomatis were certainly suboptimal insofar as we were unable to identify some samples with multiple infections and/or low copy numbers of the target gene. Finally, regarding the studies on the relationship between genotypes of chlamydial infection and sexual behavior (network), the information biases, particularly those related to self-reported sexual behaviors and drug use, must also be considered.

In conclusion, the genotype distributions were similar to those reported for other countries in the same regions, and C. trachomatis omp1 genotyping may provide some clues for social and sexual network tracking. Further studies with large sample sizes and careful collection of clinical data are needed to explore the correlations between the infecting genotypes and clinical outcomes.