Molecular Subtyping of Treponema pallidum during a Local Syphilis Epidemic in Men Who Have Sex with Men in Melbourne, Australia

Abstract

Treponema pallidum is the causative agent of syphilis, a sexually transmitted infection of significant public health importance. Since 2000 there has been a marked increase in the number of cases of syphilis infections notified in Victoria, Australia, with the majority of cases occurring in men who have sex with men (MSM) and the highest incidence being in HIV-infected MSM. The molecular subtyping method described by Pillay et al. (A. Pillay et al., Sex. Transm. Dis. 25:408–414, 1998) has been used in this study to determine the diversity of T. pallidum subtypes circulating locally and to look for any relationship between T. pallidum subtypes and HIV status over a 6-year period (2004 to 2009). Treponema pallidum DNA was detected in 303 patient specimens (n = 3,652), and full subtyping profiles were obtained from 90 of these (from 88 patients). A total of 11 T. pallidum subtypes were identified: types 14e (28, 31.1%), 14d (15, 16.7%), 14k (13, 14.4%), 14p (12, 13.3%), 14i (7, 7.8%) 14b (6, 6.7%), 14l (5, 5.6%), and 12i, 13b, 13i, and 13e (1 each, 1.1%). This study showed a similar level of variation among circulating T. pallidum strains compared with that in other studies using the same methodology. A different mix of strains and different predominating strains have been found at each geographical study location, with type 14e emerging as the predominant local strain in Victoria. There was no detectable trend between T. pallidum subtypes and the specimen collection site or stage of syphilis (where known), nor was there any relationship between particular strains and HIV status.

INTRODUCTION

Treponema pallidum is the causative agent of syphilis, a sexually transmitted infection (STI) of significant public health importance. If left untreated, the progress of infection follows a fluctuating course where the patient may be intermittently infectious. A proportion of untreated patients will progress to tertiary syphilis and may present with gummatous, cardiovascular, or neurological disease. Syphilis in pregnant women can result in congenital infection with subsequent morbidity and mortality (20). For these reasons, persons with serological evidence of syphilis infection of uncertain duration are by convention regarded as latently infected, unless there is a known history of effective antibiotic therapy. In addition, syphilis infection has been linked to an increase in the risk of transmission and acquisition of the human immunodeficiency virus (HIV) (18).

In Victoria, Australia, Department of Health notifications of infectious syphilis have risen sharply over the last decade, from a low of 9 cases in 2000 to a peak of 423 cases in 2007 (8). The vast majority of cases have been in men who have sex with men (MSM). This epidemic has coincided with a sustained increase in HIV notifications in the state (3).

Over the decade 2000 to 2009, the Victorian Infectious Diseases Reference Laboratory (VIDRL), using a combination of serology and PCR testing, detected 1,395 episodes of infectious syphilis (primary, secondary, or reinfection) in 1,161 patients. There were 23 female patients and 1,138 male patients, of whom 537 were infected with HIV at some time during the decade. From the HIV-positive cohort, 68 patients presented with infectious syphilis and previously undiagnosed HIV infection; these may represent coinfections (Fig. 1). Similar trends have been recognized in other states of Australia, and outbreaks of syphilis in MSM have also been observed in other industrialized nations, such as the United States, Canada, and countries in Europe. There is now strong evidence that syphilis infection is contributing to increasing HIV incidence (7, 10, 17, 18).

Fig 1.

Infectious syphilis episodes by HIV status, VIDRL, 2000 to 2009. Combined serology and PCR data. Data are for 1,395 episodes from 1,161 patients, including 23 HIV-negative females. Numbers above the bars indicate simultaneous presentation with previously undiagnosed syphilis and HIV infection. Prev pos, previously positive, a category that includes patients with evidence of syphilis infection preceding an observed infectious episode during the study period.

Until recently, understanding the pathogenicity and epidemiology of T. pallidum has been complicated by its short survival outside the host and the inability to maintain culture of the organism in vitro. However, sequencing of the T. pallidum genome in 1998 has led to the development of new molecular techniques for the detection and characterization of this pathogen (5). Investigation of the molecular epidemiology of syphilis is desirable from a public health perspective as it allows monitoring of the prevalence and the geographical and temporal distribution of subtypes within various patient populations and allows further study of the relationship between syphilis and HIV transmission.

In 1998, Pillay et al. described a molecular subtyping system that can be used to determine the diversity of circulating syphilis strains (14). The method is based on the analysis of variation of the number of 60-bp repeats within the T. pallidum acidic repeat protein (arp) gene and sequence variation, as determined by restriction fragment length polymorphism (RFLP) profiling within the subfamily II of the T. pallidum repeat (tpr) genes.

The aim of this study was to examine the molecular epidemiology of syphilis in Melbourne, Australia, between 2004 and 2009. The study utilized the molecular subtyping system described above to determine the diversity of T. pallidum strains circulating in Melbourne, especially in relation to HIV status.

MATERIALS AND METHODS

Study population.

Routine diagnostic specimens referred to VIDRL between 2004 and 2009 for T. pallidum polA PCR were collected at general practice clinics and hospitals from several patient groups, including HIV-infected patients, MSM, and other groups with a high incidence of STIs. No specimens were directly sought by the laboratory, and no clinical information apart from clinical notes supplied by the requesting doctor was received.

HIV data.

As VIDRL acts as a state reference laboratory for HIV infection, patient HIV status was either known on the basis of a positive Western blot result or clinical notes or inferred on the basis of regular performance of T cell counts and HIV load assays. Patients with unknown HIV status were assumed to be negative.

Clinical specimens.

Specimens tested included oral, genital, and anorectal ulcer or lesion swab, cerebrospinal fluid (CSF), and tissue biopsy specimens. Swabs were collected by the attending doctor and transported to the laboratory either dry or in bacterial or viral transport medium. CSF specimens were sent fresh, and tissue samples were either fresh or formalin fixed and paraffin embedded. Samples were stored at 4°C for up to 7 days before undergoing DNA extraction.

DNA extraction.

Total DNA was extracted from fresh specimens on an Xtractor system (Qiagen Hilden, Germany) using a DX reagent pack, as per the manufacturer's instructions. DNA was extracted from paraffin-embedded tissue sections as described by Fyfe et al. (6). DNA extracts were stored at −20°C until required for testing in both the T. pallidum polA screening assay and the 2 molecular subtyping assays.

Real-time PCR for detection of T. pallidum.

A TaqMan real-time PCR assay targeting the polA gene of T. pallidum as described by Leslie et al. was used to screen for T. pallidum (9).

Molecular subtyping of T. pallidum from clinical samples.

Samples that tested positive for the polA gene of T. pallidum were retrospectively subtyped using the method described by Pillay et al. (14). As previously described, primers for the arp gene were selected from a unique sequence flanking the repeat region within the gene. The primers used were those described by Pillay et al. (14) but without the fluorescent dyes (Arp F, 5′-CAAGTCAGGACGGACTGTCC-3′; Arp R, 5′-GGTATCACCTGGGGATGC-3′). This primer pair amplified fragments of various sizes depending on the number of repeats present within the arp gene of the strain being typed. Amplification of the arp gene was performed as described by Pillay et al. (14) using the following modified reaction mix: 5 μl of Arp F and Arp R (final concentration, 1 μM), 5 μl 10× buffer (Qiagen, Hilden, Germany), 10 μl of Q solution (Qiagen), 0.5 μl of Taq enzyme (Qiagen), 0.5 μl of deoxynucleoside triphosphates (dNTPs; 20 nM [each] dATP, dCTP, dGTP, and dTTP; Thermo Fisher Scientific, Waltham, MA), and 5 μl of template DNA in a total volume of 50 μl. The amplification conditions were as follows: 94°C for 5 min, 40 cycles of 94°C for 1 min, 62°C for 1 min, and 72°C for 1 min, and a final extension step of 72°C for 5 min. The arp gene amplicons were electrophoresed through a 2% agarose gel with a 100-bp ladder at 100 V for 1 h.

A nested PCR described by Pillay et al. (14) was used to amplify the tpr gene. Primer pairs chosen from a conserved region in the tpr homologs were originally described by Pillay et al. (14) but modified in this study. The first pair consisted of a 20-mer forward primer, Tpr F1 (5′-ACTGGCTCTGCCACACTTGA-3′), and a 20-mer reverse primer, Tpr R1 (5′-CTACCAGGAGAGGGTGAAGC-3′), amplifying a 2,186-bp fragment of the region being examined. The second primer pair, designed to amplify a 1,836-bp region of the PCR product obtained in the first amplification of the tpr gene, was 18-mer forward primer Tpr F2 (5′-CAGGTTTTGCCGTTAAGC-3′) and 20-mer reverse primer Tpr R2 (5′-AATCAAGGGAGAATACCGTC-3′). Amplification of the tpr gene was performed as per the published method using a modified reaction mix (1st and 2nd rounds): 1 μl of forward and reverse primer (final concentration, 0.6 μM), 5 μl 10× buffer (Qiagen), 10 μl of Q solution (Qiagen), 0.5 μl of Taq enzyme, 0.5 μl of dNTPs (20 nM [each] dATP, dCTP, dGTP, and dTTP; Thermo Fisher Scientific), and 5 μl of template in a total volume of 50 μl. Electrophoresis of the tpr gene amplicon (2nd round) was performed using a 2% agarose gel together with a 100-bp ladder at 100 V for 1 h. The PCR product obtained was purified using a High Pure PCR product purification kit (Roche, Mannheim, Germany) as per the manufacturer's instructions and then digested with the restriction endonuclease MseI (Invitrogen, Carlsbad, CA). The digested PCR product was electrophoresed through a 2% agarose gel for 1 h at 100 V to determine the digestion pattern of the T. pallidum strain.

The molecular subtype of each T. pallidum strain was determined by combining the results of the arp gene size and the tpr gene banding pattern. Each sample was designated a number for the arp size which was determined by the number of repeats present in the arp gene and a lowercase letter for the tpr banding pattern (e.g., 14e) Fig. 2.

Fig 2.

(A and B) Examples of PCR amplicons of the arp gene region showing different sizes. (A) Lanes 1 and 16, 100-bp DNA ladder; lanes 2 and 15, 1-kb DNA ladder; lanes 5, 6, 7, 9, 12, and 14, 14 repeats; lane 13, 12 repeats; lanes 3, 4, 8, 10, 11, and 16, product not generated. (B) Lanes 1 and 16, 100-bp DNA ladder; lanes 3, 5, 7, 8, 9, 11, 13, 14, and 15, 14 repeats; lane 10, 13 repeats; lanes 3, 4, 8, 10, 11, and 16, product not generated. (C) MseI RFLP patterns of tpr amplicons showing example subtypes. Lanes 1 and 18, 100-bp DNA ladder; lane 2, T. pallidum Nichols strain; lanes 3, 7, 11, and 16, type e; lanes 4 and 10, type p; lane 6, type i; lane 8, type b; lane 13, type l; lanes 12, 15, and 17, type d; lanes 5, 9, and 14, no banding pattern generated.

RESULTS

Of the 3,652 samples tested in the screening polA real-time PCR assay, T. pallidum DNA was detected in 303 specimens (8.2%). Of these 303 patient samples, 90 specimens from 88 patients were positive in both the arp and tpr molecular subtyping assays, producing a full subtyping profile for 29% of the positive samples. For 87 of the 90 samples, there were concurrent syphilis serology results that were consistent with infectious syphilis. For the remaining 3 samples, serology results were not available. All 90 samples had a polA PCR threshold cycle (C_T) of <30 cycles (strongly positive). Of the 88 patients, 19 were HIV positive prior to contracting syphilis, 2 presented as HIV-syphilis copresentations, 2 contracted HIV after their episode of syphilis (during the study 1 of these patients had two typeable episodes of syphilis involving different strains), and 65 (including the single female) remained HIV negative or status unknown. Two positive samples collected at the same time from a patient with secondary syphilis showed the same profile (14e) (Table 1).

Table 1.

T. pallidum subtypes detected and HIV status

T. pallidum subtypea	No. of patients				Notes
	HIV infected before or coinfected at the time of the syphilis episode	HIV negative at time of syphilis episode but seroconverted later in the decade	HIV status negative or unknown	Total
14e	6	0	22	28	Includes concurrent oral and anal swabs from an HIV-negative patient with secondary syphilis
14d	5	1	9	15	Includes the first syphilis episode (positive penile swab, September 2006) in a patient who later seroconverted to HIV
14k	2	1	10	13	Includes the second syphilis episode (positive penile swab, December 2006) in a patient who later seroconverted to HIV; note change in T. pallidum strain
14p	1	1	10	12	Includes one HIV-negative female patient (all others in the study are male)
14i	1	0	6	7
14b	3	0	3	6
14l	2	0	3	5
13b	0	0	1	1
13e	0	0	1	1	Patient claims heterosexual orientation
13i	1	0	0	1
12i	0	0	1	1
Total	21	3	66	90b

^aNumerals refer to arp type; letters refer to trp type.

^bNinety specimens from 88 patients.

Eighty-three samples gave a molecular typing profile for the arp subtyping assay but no product could be amplified for the tpr region. Insufficient intact T. pallidum DNA was present in the remainder of the samples to amplify a product in either of the two subtyping PCR assays. It was noted that there was a correlation between the C_T value determined in the polA T. pallidum screening assay and the ability to identify full subtypes for those samples.

The 90 typeable samples were collected from the following sites: anorectal (2 HIV, 15 non-HIV), penile or scrotal (11 HIV, 35 non-HIV), oropharyngeal (4 HIV, 7 non-HIV), and site of collection unknown (4 HIV and 12 non-HIV, with the latter including the single female patient in the study). Although the stage of infection was not always stated on the request, swabs were identified as being from probable primary syphilis cases in 34 patients (6 HIV-infected patients), from probable secondary cases in six specimens from 5 patients (3 HIV-infected patients), and from cases of suspected reinfection, based on prior serological or clinical data, in 11 patients (4 HIV-infected patients).

A total of 11 subtypes were identified among patients with a full typing profile. These were 14e (28, 31.1%), 14d (15, 16.7%), 14k (13, 14.4%), 14p (12, 13.3%), 14i (7, 7.8%) 14b (6, 6.7%), 14l (5, 5.6%), and 12i, 13b, 13i, and 13e (1 each, 1.1%). Breakdown by HIV status is shown in Table 1. Results show that subtype 14e, possibly introduced in 2005, accounted for 32.2% of the fully typeable specimens; this became the dominant subtype circulating in Melbourne during the 6-year period between 2004 and 2009 (Fig. 3). Subtype 14d showed a decrease in prevalence between 2004 and 2009. All other subtypes were present in much smaller numbers across the 6-year period (Fig. 3), but the three subtypes detected in 2004 were still circulating in 2009. There was no obvious trend between the T. pallidum subtypes and the specimen site or the stage of syphilis infection (where known). All strains detected on more than one occasion were seen in both the HIV-infected and non-HIV-infected groups.

Fig 3.

Distribution of T. pallidum subtypes, Victoria, Australia, 2004 to 2009. Data are for 90 specimens from 88 patients.

Of patients with more than one specimen in the study, one HIV-negative male with clinical notes stating secondary syphilis with condylomata lata had type 14e detected at both oropharyngeal and anorectal sites. One patient had two distinct T. pallidum subtypes, 14d in mid-September 2006 and 14k 3.5 months later. In this case, typing was useful in demonstrating that the second episode was due to reinfection rather than treatment failure, as this patient remained both rapid plasma reagin (RPR) test and IgM positive between September 2006 and January 2007. This patient subsequently seroconverted to HIV positive in 2009. There was only one isolate typed from a female patient in this study; it was identified as the relatively uncommon 14b subtype. Additionally, a penile ulcer swab from an HIV-negative patient who claimed heterosexual orientation typed as the unique subtype 13e.

DISCUSSION

The inability to maintain T. pallidum in vitro has made direct detection and typing of the organism difficult. In recent years, the advent of molecular techniques has allowed scientists to gain new insights into both the pathogenicity and epidemiology of T. pallidum. We believe that this is the first study to attempt to examine the diversity of T. pallidum strains circulating in a major Australian urban center. Our study demonstrates that there are a limited number of T. pallidum strains circulating in Melbourne, with most strains represented in both HIV-infected and noninfected patient groups. Similar results have been described in other studies performed across Europe, the Americas, South Africa, and China (1, 2, 4, 11–16, 19).

Eleven subtypes were identified in this study. Subtype 14e was the predominant subtype circulating in Melbourne during this 6-year period. The increase of subtype 14e over the period may suggest that this subtype is more transmissible than the other subtypes described; however, this subtype was not commonly detected in the studies listed above. A similar range of strains has been identified within the MSM population in Scotland but with strain 14d predominating (2). Of note, subtype 14a, the predominate subtype in Portugal (4), and subtype 14f, the predominate subtype circulating in China (12) and North America (16, 19), were not identified in this study.

All other subtypes were present in smaller numbers across the 6-year period. The smaller numbers of the non-14e subtypes may suggest a more localized transmission network. However, it is not currently possible to identify the presence of such networks without access to contact-tracing data. Of particular interest, subtypes 12i, 13b, 13i, and 13e have not been described in any published T. pallidum subtyping studies to date. Further studies to investigate the distribution of T. pallidum subtypes across Australia would be useful in determining whether these may represent endemic Australian strains.

Of the 303 positive samples detected by real-time PCR, 212 samples had insufficient T. pallidum DNA for typing. It was noted that there was a correlation between the C_T values generated from the polA screening PCR and the ability to successfully type those strains. Previous T. pallidum typing studies have also demonstrated that this typing method has a relatively low sensitivity when applied to clinical samples that have tested positive in the polA syphilis screening PCR (2, 19).

In conclusion this study shows that the T. pallidum subtyping method described by Pillay et al. (14) is a useful, if insensitive, epidemiological tool. In this study, we did not have sufficient epidemiological data available to determine local trends or movement of strains. However, we were able to show that there is a measurable level of genetic variation among T. pallidum strains circulating in Melbourne. Apart from the single 13i strain, no subtype was found to be uniquely associated with HIV-infected patients and all arp subtype 14 strains were detected in both HIV-infected and non-HIV-infected patients. This observation counters claims by HIV-infected MSM that they are “serosorting,” i.e., engaging in unprotected sexual activities only with persons of the same HIV status.

To determine whether T. pallidum strains identified solely in this study are strictly Australian endemic strains would require wider studies. This could include testing of specimens sourced from indigenous communities and from surveillance programs in neighboring countries. Future studies, with the appropriate ethical approval, should also aim to incorporate local health department notification and contact-tracing information to further assess the accuracy and robustness of the typing system.

This study also emphasized the strong link between syphilis and HIV. In order to reduce the incidence of HIV, screening, rapid detection, and treatment of all infectious syphilis cases in MSM should become a public health priority.