Article Text
Abstract
Objective: To examine the molecular epidemiology of syphilis in Scotland.
Methods: Ulcer specimens were collected from 85 patients with infectious syphilis. Typing of Treponema pallidum was performed using a method that examines variation in two loci; the number of 60-basepair repeats within the arp gene and sequence variation in the tpr genes.
Results: Patients were predominately white men who have sex with men (MSM). Treponemal DNA was detected in 75 specimens and a total of six subtypes were identified from 58 typeable specimens (77%). The most common subtypes were 14d (44/58, 76%), followed by 14e (7/58, 12%), 14j (3/58, 5%), 14b (2/58, 3%), 14p and 14k (1/58, 2%).
Conclusions: This study shows that subtype 14d is the predominant subtype circulating in Scotland and there is a surprising level of genetic diversity within the Scottish MSM community.
Statistics from Altmetric.com
Although syphilis is a relatively rare disease in the UK, diagnoses of infectious syphilis have increased since 1997; rising from 301 cases to 3749 cases in 2007.1 2 This has been largely due to outbreaks, frequently among men who have sex with men (MSM), in a number of cities.1 In Scotland, diagnoses have increased annually since the outbreak started in 2000/1.3 There was a total of 235 and 248 cases in 2006 and 2007, respectively, with the majority of cases in MSM (81%).4
If left untreated, syphilis can lead to devastating diseases including cardiovascular syphilis, benign gummatous syphilis and neurosyphilis.5 The vertical transmission of Treponema pallidum to the unborn child may result in the distressing condition of congenital syphilis. In addition, syphilis infection increases the risk of both the transmission and acquisition of HIV6 and co-infection with HIV is a frequent occurrence. The severe consequences of syphilis highlight the importance of implementing effective measures to control its dissemination.
The availability of a molecular typing system allows strain variation to be identified in the absence of an in-vitro cultivation system for T pallidum. The typing system for T pallidum7 is based on analysis of the arp (acidic repeat protein) gene and subfamily II (tpr E, G and J) of the tpr (T pallidum repeat) genes. Investigation into the molecular epidemiology of syphilis combines molecular typing with demographic or behavioural data and enables the identification, characterisation and geographical distribution of molecular subtypes, which in turn could inform public health intervention programmes.
This is the first study in the UK, and the second one in Europe, to examine the molecular epidemiology of syphilis, and presents evidence of multiple subtypes circulating in this predominantly MSM population diagnosed with infectious syphilis in Scotland. This study was performed as part of the European Surveillance of Sexually Transmitted Infections project (http://www.essti.org).
Methods
Clinical specimens
Between August 2006 and December 2007 genital, anal or oral ulcer swabs were collected from patients with suspected infectious syphilis who were attending genitourinary medicine clinics in Scotland. The specimens were sent to the Scottish Bacterial Sexually Transmitted Infections Reference Laboratory (SBSTIRL), Royal Infirmary of Edinburgh, for molecular testing. Extracted DNA for samples that were positive for T pallidum PCR were then assigned a study number and clinic number and patient identifiers were removed to allow blinded molecular typing at the Sexually Transmitted Bacterial Reference Laboratory (STBRL), Health Protection Agency, London. The DNA samples were transported from SBSTIRL to STBRL overnight on dry ice.
Epidemiological patient data
Epidemiological data were collected at the clinic visit following an interview by healthcare staff using a proforma designed as part of the National Enhanced Surveillance of Infectious Syphilis in Scotland (NESISS) programme.
DNA extraction and detection of T pallidum DNA by real-time PCR
DNA was extracted by centrifuging 1 ml M4RT media (Remel, Lenexa, USA) or AL lysis buffer (Qiagen, Crawley, UK) at 13 000 rpm for 1 minute and 800 μl of the supernatant was removed. The remaining 200 μl and any pelleted cellular material was then processed using the QIAamp blood and body fluids spin protocol (Qiagen) according to the manufacturer’s instructions, and the DNA was eluted into 100 μl sterile water. Treponemal DNA was detected by a real-time PCR assay that targets the DNA polymerase A gene (polA) of T pallidum.8 The presence of treponemal DNA was re-confirmed on receipt at STBRL by the same T pallidum polA real-time PCR.8
Molecular typing of T pallidum
Molecular subtyping was performed based on a previously described method7 9 that examines variation in two loci; the number of 60 basepair (bp) repeats in the acidic repeat protein (arp) and sequence variation in the T pallidum repeat (tpr) genes. The final molecular subtype is defined based on the two loci (eg, 14d equals 14 repeats in the arp gene and tpr restriction fragment-length polymorphism (RFLP) pattern “d”). PCR of the arp gene was performed in 50 μl reactions containing 1.67 U expand high fidelity PCR system Taq polymerase (Roche Diagnostics, Mannheim, Germany), 2.0 mmol magnesium chloride (Roche Diagnostics), 2 mmol (each) of dTTP, dCTP, dATP and dGTP (Invitrogen, Paisley, UK), 0.2 μmol of primers ARP-F (5′–CAAGTCAGGACGGACTGTCCCTTGC–3′) and ARP-R (5′–GGTATCACCTGGGGATGCGCACG–3′) and 10 μl DNA template. Thermocycling parameters were as follows: one cycle at 94°C for 5 minutes, 45 cycles at 94°C for 50 s, 62°C for 1 minute and 68°C for 2.5 minutes and a final extension step of 68°C for 15 minutes. All thermocycling was performed on the MJ Research PTC-225 thermal cycler (Waltham, USA). The arp PCR products were electrophoresed in a DNA7500 LapChip on an Agilent Bioanalyser (Agilent Technologies, Wokingham, UK) and the number of arp repeats was determined by comparing the size of the amplicons with the Nichols strain of T pallidum, which has 14 repeats.7
A nested PCR was used to amplify the tpr genes. The first round of the nested PCR was performed in 100 μl reactions, which contained 2.5 U expand high fidelity PCR system Taq polymerase with 1.5 mmol magnesium chloride (Roche Diagnostics), 2 mmol (each) of dTTP, dCTP, dATP and dGTP (Invitrogen) and 0.2 μmol primers B1-F (5′–ACTGGCTCTGCCACACTTGA–3′) and A2-R (5′–CTACCAGGAGAGGGTGACGC–3′) and 6 μl of DNA template. Thermocycling parameters were as follows: one cycle at 94°C for 5 minutes, 35 cycles of 94°C for 1 minute, 60°C for 2 minutes and 68°C for 2.5 minutes, and a final extension step of 68°C for 15 minutes. The second round of the tpr nested PCR was performed in 100 μl reactions as described for the first round, except that 0.2 μmol of a second set of primers, IP6-F (5′–CAGGTTTTGCCGTTAAGC–3′) and IP7-R (5′–AATCAAGGGAGAATACCGTC–3′), was used. Thermocycling parameters were as follows: one cycle at 94°C for 5 minutes, 40 cycles of 94°C for 1 minute, 59°C for 1 minute and 68°C for 2 minutes, and a final extension step of 68°C for 15 minutes.
The tpr nested amplicon (16.2 μl) was digested at 37°C for 1 h in a 20 μl reaction containing 4 × bovine serum albumin (New England BioLabs, Hitchin, UK) and 10 U of the restriction enzyme Mse I (New England BioLabs). Restriction fragments were electrophoresed using the Agilent Bioanalyser (Agilent Technologies). The RFLP patterns were compared with the T pallidum Nichols strain and sized using the ladder, and profiles were determined by the original publication7 and additional profiles were kindly supplied by the lead author (A Pillay, personal communication).
The number of 60-bp arp repeats and trp RFLP pattern were combined to create the molecular subtype as previously described.7 The arp and the trp assays were repeated on all specimens that initially gave a negative result.
Statistical analysis
The Student’s t test was used to determine any significant difference between the mean polA PCR cycle threshold (Ct) values for fully typeable specimens and non-typeable specimens. Fisher’s exact test was used to assess the significance of differences of demographic and epidemiological data for individuals with common subtypes and the remaining study population.
Results
There were 88 treponemal DNA-positive ulcer specimens (excluding duplicates) submitted to SBSTIRL, and from these a total of 87 specimens from 85 individuals were available for molecular typing during the 17-month study period. Twenty specimens were collected in 2006 and 67 were collected in 2007. Other than sex and age, there was no epidemiological information available for 10 patients. The most common epidemiological and demographic characteristics of the patient population are described in table 1. To summarise, the patient population was predominantly British white homosexual men with a median and mean age of 32 and 35 years, respectively (range 16–72).
The presence of treponemal DNA was re-confirmed using the polA PCR in 75 of the 87 specimens. Of these 75 specimens, 58 were positive by both the tpr and arp assays to produce a full subtype in 77% (58/75) of specimens. Of the 58 fully typeable specimens, a total of six subtypes was identified. The most common subtype was 14d (44/58, 76%), followed by 14e (7/58, 12%), 14j (3/58, 5%), 14b (2/58, 3%), 14p and 14k (each 1/58, 2%).
Six specimens were positive by the tpr assay only and in total there were seven different tpr RFLP patterns (b, d, e, j, k, l and p). A schematic representation of the tpr RFLP patterns identified in this study is shown in fig 1. Three specimens gave a positive arp result of 14 repeats, but no tpr profile could be determined. One of these specimens was a duplicate patient sample; the other sample gave a positive “e” tpr pattern only. Fourteen repeats were identified in all but one of the positive arp results. Insufficient amplicon was generated to determine accurately the size of the remaining fragment, but the larger size indicated that it contained more than 14 repeats. This specimen had an “e” tpr pattern. There were seven specimens that were negative for both the arp and the tpr assays. Using the Student’s t test, there was a significant difference (p<0.001) between the mean polA PCR Ct values from specimens that were typeable (overall lower Ct values) and non-typeable (overall higher Ct values).
Subtype 14d appeared throughout the whole 17-month collection period, whereas subtype 14e was collected over a 10-month period (August 2006 to May 2007), 14j over a 13-month period (October 2006 to September 2007) and two specimens with the 14b subtype were identified in April 2007 and November 2007.
Table 1 compares demographic and epidemiological data for individuals with 14d and 14e subtypes with the remaining study population. Using Fisher’s exact test there was no significant difference between any epidemiological or demographic factor of those patients with subtypes 14d or 14e and the remaining study population (for all comparisons p⩾0.2).
The place of acquisition was known for 73 patients (86%). Most patients (79%) recorded acquiring infection in an area or city in Scotland, or elsewhere in the UK (7%) in cities where outbreaks of syphilis have been described.2 Patients also described acquiring syphilis from contacts outside the UK (7%) or quoted multiple possible locations (7%), all of which included an area of Scotland. The most common subtype, 14d, was distributed throughout Scotland and other locations throughout England (table 2). There were six places of possible acquisition declared outside of the UK, but a full subtype was only retrievable from one specimen, Pakistan (14j). The other specimens from individuals who declared possible places of acquisition did not reveal a full subtype; South Africa (tpr l), Canary Islands (arp 14), Australia, Thailand and Spain (no arp or tpr result).
Discussion
This is the first UK study to examine the molecular epidemiology of syphilis and it demonstrated variation in the subtypes circulating in Scotland and the rest of the UK. Of the 347 cases of infectious syphilis identified throughout Scotland over the study period, 88 (25%) were diagnosed by serology as well as amplification of T pallidum DNA in ulcer specimens and so were available for this study. Fifty-eight of these (66%) were fully typed (17% of all syphilis cases). Six full subtypes (14b, 14d, 14e, 14k, 14j and 14p) were identified in this Scottish study. Subtype 14d accounted for 76% of the fully typeable specimens and was the dominant subtype circulating in the Scottish syphilis outbreak during 2006 and 2007. In the only other European typing study performed in Lisbon, Portugal,10 a lower genetic diversity was observed; three full subtypes from 42 specimens (14a, 50%; 14d, 45%; 14f, 5%). The most prevalent subtype (14a) was not found in Scotland. Subtype 14d was predominant in South Africa and Madagascar and was also demonstrated in the USA.7 9 11
Key messages
This is the first study to describe the molecular epidemiology of circulating syphilis strains in Scotland and in the UK.
The majority of syphilis cases are diagnosed in white MSM.
Subtype 14d predominates and the identification of six different subtypes reveals a surprising level of genetic diversity given the low prevalence of syphilis in Scotland.
Epidemiological and molecular typing data have shown that syphilis is becoming endemic within the Scottish MSM community.
The much smaller numbers of the other non-14d subtypes may indicate smaller and more localised transmission networks and/or outbreaks that are most probably related to the importation of syphilis strains into the Scottish MSM community. However, it is currently not possible to determine the presence of such networks using the available epidemiological information. There appears to be a higher level of strain diversity when syphilis is endemic, such as in South Africa, Madagascar and the southeastern USA.7 9 12 Subtypes 14j, 14k and 14p have not been described in the six published studies to date.7 9 10 11 12 13 Further work to investigate the distribution of syphilis subtypes throughout England and Wales, and any correlation with Scotland, would be useful.
Subtype 14d predominated for the duration of this study, although a more discriminatory typing method may be required before it can be concluded that the strains characterised as subtype 14d are indistinguishable from each other. The predominance of a single subtype may be a reflection of the low prevalence (3.3 per 100 000) of syphilis in Scotland, with a limited number of subtypes circulating in a relatively small core group. However, the possibility that subtype 14d represents a more transmissible or even more virulent strain cannot be excluded. Without typing data from when the outbreak started in 2000/1,4 it is not possible to say whether 14d is the original outbreak strain in Scotland. Further prospective typing studies would determine if subtype 14d continues to predominate and perpetuate sustained transmission and endemic spread of syphilis throughout Scotland. In spite of this subtype predominance, this study has demonstrated that there is variation in the molecular epidemiology of syphilis in Scotland. The identification of six different subtypes and one additional different tpr profile (“l”) is surprising given the low prevalence of syphilis in Scotland. It is interesting that there is such diversity when there was little evidence of overseas travel and most patients declared that the most likely place of acquisition was within Scotland or elsewhere within the UK. This may suggest that syphilis is becoming endemic within the Scottish MSM community; this is supported by the stabilising epidemiology that shows a similar number of cases from 2005 to 2007.4
The major demographic of the study population was white British homosexual men, which also matches the published surveillance data,4 so although only 17% of strains from cases of infectious syphilis in 2006 and 2007 were typed, the dataset is most probably representative. There was no evidence to suggest that subtype 14d was associated with any particular patient population characteristics or locations; however, this does not exclude the presence of other subtypes that are related to particular epidemiological, demographic or risk factors. It was not possible to establish this during this study because of the low numbers of the other subtypes. When the data are available, the number of patients co-infected with HIV is high, 13% (10/75), and this highlights the need to limit the transmission of syphilis with great exigency. Greater understanding of the disease molecular epidemiology may allow a more focused intervention strategy with the aim of limiting syphilis and HIV infection.
The presence of treponemal DNA was not reconfirmed in all specimens (14%). This loss of sensitivity may be due to DNA degradation as a result of prolonged storage or repeat freeze–thawing. The overall sensitivity of this typing method may increase if fresh samples were used. Of the 75 specimens available for typing, only 77% were fully typeable. DNA quality and/or poor assay sensitivity, as indicated by the significant difference of the means of the polA PCR Ct values from the typeable and non-typeable specimens, may explain why seven specimens were negative with this typing assay. The sensitivity of the arp gene PCR in particular was low, and in one case the observed weak amplification prevented an accurate assessment of the band size. The three instances in which an arp result was obtained but no tpr profile was determined could be caused by primer–template mismatch or insufficient DNA quality to permit amplification of this larger (2186 bp) product. Insufficient specimen volume was available for further investigation. Previous syphilis typing studies have also demonstrated the poor sensitivity of this typing method, ranging from 87% to as low as 49%, when applied to specimens that have been confirmed to contain treponemal DNA.9 10 11 12 13
This report is timely following the first European study that was performed in Portugal.10 The introduction of routine typing using this method would allow the creation of a database of syphilis subtypes to monitor the prevalence and distribution of subtypes throughout Europe. However, it is clear that further work is required to establish a more sensitive, robust and more discriminatory typing strategy. Continued molecular typing of syphilis would allow: (1) the identification of new subtypes; (2) the detection of outbreak strains; (3) the identification of associations between particular subtypes, their virulence and disease outcome; and (4) would also determine whether endemic strain levels are reduced by elimination strategies.
Acknowledgments
The authors would like to thank all the ESSTI collaborators for their contribution to the ESSTI project (www.essti.org). Thanks to all staff at the genitourinary medicine clinics throughout Scotland who contribute to the NESISS programme and many thanks to Tom Nichols for his statistical advice.
Footnotes
Funding Financial support for this study was provided by the European Commission (DG SANCO), agreement no 2004210: ESSTI European Surveillance of Sexually Transmitted Infections.
Competing interests None.
Contributors MJC was the project microbiologist, performed the laboratory work, analysed the data and is the lead author for the paper. SAC contributed to the design of the study and assisted in the laboratory work and data analysis. HMP and LAW provided the specimens and epidemiological information and contributed to the design of the study. CAI contributed to the design of the study, supervised the project and was the project lead. All authors contributed to the write up and reviewed the final draft of the manuscript.
Provenance and peer review Not commissioned; externally peer reviewed.
Linked Articles
- Whistletop tour