Clinical and genomic diversity of Treponema pallidum subspecies pallidum to inform vaccine research: an international, molecular epidemiology study

Abstract

Background: The increase in syphilis rates worldwide necessitates development of a vaccine with global efficacy. We aimed to explore Treponema pallidum subspecies pallidum (TPA) molecular epidemiology essential for vaccine research by analysing clinical data and specimens from early syphilis patients using whole-genome sequencing (WGS) and publicly available WGS data.

Methods: In this multicentre, cross-sectional, molecular epidemiology study, we enrolled patients with primary, secondary, or early latent syphilis from clinics in China, Colombia, Malawi, and the USA between Nov 28, 2019, and May 27, 2022. Participants aged 18 years or older with laboratory confirmation of syphilis by direct detection methods or serological testing, or both, were included. Patients were excluded from enrolment if they were unwilling or unable to give informed consent, did not understand the study purpose or nature of their participation, or received antibiotics active against syphilis in the past 30 days. TPA detection and WGS were conducted on lesion swabs, skin biopsies, skin scrapings, whole blood, or rabbit-passaged isolates. We compared our WGS data to publicly available genomes and analysed TPA populations to identify mutations associated with lineage and geography.

Findings: We screened 2802 patients and enrolled 233 participants, of whom 77 (33%) had primary syphilis, 154 (66%) had secondary syphilis, and two (1%) had early latent syphilis. The median age of participants was 28 years (IQR 22-35); 154 (66%) participants were cisgender men, 77 (33%) were cisgender women, and two (1%) were transgender women. Of the cisgender men, 66 (43%) identified as gay, bisexual, or other sexuality. Among all participants, 56 (24%) had HIV co-infection. WGS data from 113 participants showed a predominance of SS14-lineage strains with geographical clustering. Phylogenomic analyses confirmed that Nichols-lineage strains were more genetically diverse than SS14-lineage strains and clustered into more distinct subclades. Differences in single nucleotide variants (SNVs) were evident by TPA lineage and geography. Mapping of highly differentiated SNVs to three-dimensional protein models showed population-specific substitutions, some in outer membrane proteins (OMPs) of interest.

Interpretation: Our study substantiates the global diversity of TPA strains. Additional analyses to explore TPA OMP variability within strains is vital for vaccine development and understanding syphilis pathogenesis on a population level.

Funding: US National Institutes of Health National Institute for Allergy and Infectious Disease, the Bill & Melinda Gates Foundation, Connecticut Children's, and the Czech Republic National Institute of Virology and Bacteriology.

Figure 1

Screening, enrolment, and specimen testing algorithm for participants with early syphilis in this study TPA=Treponema pallidum subspecies pallidum. ∗Screening based on suspected primary, secondary, or early latent syphilis among patients from clinical sites in China, Colombia, Malawi, and USA. †Includes individuals who refused screening procedures in Malawi. ‡One participant was enrolled twice, initially with secondary syphilis and then with primary syphilis. §One participant initially staged with secondary syphilis had a PCR-positive skin biopsy but, upon later review, was restaged as having a healed primary lesion. ¶Each participant can contribute multiple specimen types. ||166 total of unique genomes generated, which includes sequences from fully enrolled study participants; 43 participants from Malawi consented for screening who had darkfield-negative lesions but were PCR-positive, and a convenience sample of 10 individuals from Colombia enrolled in a longitudinal study with available specimens. ∗∗In general, available specimens with 40 or more polA copies per μL underwent TPA enrichment and whole-genome sequencing. ††Lesion exudates from primary syphilis and whole blood from secondary syphilis were passaged through rabbits to enrich for treponemes.

Figure 2

Clinical manifestations among study participants with primary and secondary syphilis, by TPA clade, geographical location, and quantitative PCR result (copies per μL) (A) Multiple, shallow penile ulcers (China; lesion swab = 22 copies per μL). (B) Multiple, deep perineal ulcers (Colombia; lesion swab = 2117 copies per μL). (C) Multiple vulvar chancres (Malawi; 8764 copies per μL). (D) Penile ulcer underneath the foreskin with a purulent discharge (Colombia; lesion swab = 3034 copies per μL). (E) Penile chancres with purulent base (Malawi; lesion swabs = 6897 copies per μL). (F) Penile chancres with purulent base (Malawi; lesion swabs = 2281 copies per μL). (G) Diffuse scaly, macular rash involving testicles (Colombia, skin biopsy = 53 copies per μL). (H) Hyperpigmented macules involving soles (China; skin biopsy = 305 copies per μL). (I) Extensive condyloma lata involving labia and perineum (Malawi; lesion swab = 2281 copies per μL). (J) Erythematous macular rash in pregnancy (Colombia; skin biopsy = 2460 copies per μL). (K) Diffuse rash with crusting lesions consistent with lues maligna (China; skin biopsy = 250 copies per μL). (L) Moist, condyloma lata involving vulvar area and perineum (Malawi; lesion swab = 6897 copies per μL).

Figure 3

Recombination-masked TPA whole-genome phylogeny Recombination-masked TPA whole-genome phylogeny derived from 166 individuals in this study, a convenience sample of 62 genomes published in 2021, and five reference genomes (TPA [red], Treponema pallidum subspecies pertenue [blue], and Treponema pallidum subspecies endemicum [green]). Nodes with more than 80% bootstrap support are indicated with a black circle. Manual subclade assignments are included to facilitate comparison to recent published literature (subclade), alongside TPA population clusters determined using baps Bayesian modelling (figure 4) but with maximum likelihood phylogeny used as a prior for partitioning (ML Baps). ML=maximum likelihood.TPA=Treponema pallidum subspecies pallidum.

Figure 4

Global TPA population structure (A) PCAs of global TPA strains, overall and by lineage. Analysis of 1413 TPA genomes derived from diverse studies in global sites confirms that membership in Nichols (n=350) versus SS14 (n=1063) clade accounts for most of the genetic variation (excluding tpr family, tp0470, and arp genes). When PCA was restricted to members of the same lineage, more distinct population structure is evident among Nichols-like strains. (B) PCA annotated by TPA population determined using baps Bayesian modelling. (C) Composition of the five Nichols-lineage and three SS14-lineage populations by geography and sample collection dates. Median year of sample collection and number of samples are included. PC1=principal component 1. PC2=principal component 2. PCA=principal component analysis. TPA=Treponema pallidum subspecies pallidum.

Figure 5

Accumulation of lineage-associated and population-associated missense mutations in select TPA proteins Evaluation of predicted protein structures for genes with multiple lineage-informative missense SNVs confirms population-specific mutations across global TPA isolates. Genes with lineage-informative fixed (orange) and highly differentiated (blue) SNVs identified during comparison of 1413 TPA genomes are highlighted in the circular plot (excluding tpr family, tp0470, and arp genes). Starting from the grey bar: (1) coloured lines highlight genes affected by fixed (outer; orange) or highly differentiated (inner; blue) SNVs, and (2) histograms that depict the frequency of fixed (orange background) and highly differentiated (blue background) missense mutations by gene. Nichols-lineage three-dimensional structural protein models were previously predicted by Hawley and colleagues (TP0515, TP0858, TP0865, and TP0966) or by AlphaFold2 (TP0136, TP0179, and TP0462). Boxes highlight differences by TPA population for the FadLs TP0865 and TP0858; within-population allele frequencies of more than 1% are annotated. Frequencies of lineage-informative fixed and highly informative SNVs are provided in appendix 2 (pp 4–5). Genomic coordinates correspond to the Nichols reference strain. OMF=outer membrane factor. SNV=single nucleotide variants. TPA=Treponema pallidum subspecies pallidum.