Trichomonas vaginalis strain diversity among female sex workers in Ecuador using DNA sequence-based typing

Abstract

Background: Molecular methods to track the spread of Trichomonas vaginalis (TV) infection, the most common curable non-viral sexually transmitted infection globally, associated with poor reproductive health outcomes and low socio-economic status are challenging, as ultra-long repetitive DNA sequences in TV make whole genome sequencing difficult. We undertook multilocus sequence typing (MLST) of TV using nested-PCR from clinical samples, to describe strain diversity among at-risk female sex-workers (FSWs) in Ecuador.

Methods: Sociodemographic data and vulvo-vaginal swabs were collected from two groups of FSWs, street-based workers (SBWs) and brothel-based workers (BBWs). DNA extracts, positive for TV by real-time PCR, were amplified by two-step nested-PCR for seven TV genes and MLST-amplicon libraries sequenced using Illumina MiSeq. Sequence types (STs) were clustered into clonal complexes using goeBURST and population structure investigated using STRUCTURE.

Results: Of 250 FSWs, 58 were positive for TV by real-time PCR. Subsets of TV positive vaginal DNA extracts were sequence-typed from 15 SBWs and 17 BBWs, alongside a non-sex worker sample collected from the same region, and a positive control. Compared with BBWs, SBWs were older (p < 0.001) and earnt less for sex work. TV-MLST revealed new STs and two major population subtypes. No associations were found between ST and behaviouralcharacteristics. goeBURST analysis of study samples identified four clonal complexes in which the largest complex comprised primarily of BBWs. When combined with a larger international dataset, goeBURST revealed 9 clonal complexes and 24 separate STs or nodes. FSWs with the same ancestral TV population structure were not displaced by the added STs.

Conclusion: TV-MLST revealed high strain diversity among Ecuadorian FSWs and a two-type sub-population. The preservation of links between STs associated with some FSWs when adding a larger set of archived STs, suggests potential for use as an aid to TV associated sexual network identification.

Supplementary Information: The online version contains supplementary material available at 10.1186/s12879-025-12185-7.

Keywords: Ecuador; Sequence-typing; Sex workers; Sexual networks; Trichomonas vaginalis.

Fig. 1

Initial goeBURST analysis of STs identified from samples in this study using stringMLST. All circles represent an ST, larger circles (131 and 133) correspond to that ST belonging to two individuals. Two FSWs represented by ST-131 were BBWs from location A and B, whilst the 133 represented the only NSW recruited from location B and a SBW from location C. Lines connecting each ST within CCs represent the rules applied using the goeBURST algorithm to link those two STs. Black lines refer to no recourse to tie break rules, blue shows a link corresponding to tie break rule one 1 and yellow refers to tie break rule 4 or 5 (frequency found on the data set and ST number, respectively) [16]

Fig. 2

goeBURST analysis with samples reported from studies and stored within PubMLST. Samples were combined from this study and any available in PubMLST to understand how connected each CC would be when combined with a larger data set. As previously described, from the Ecuador dataset ST-131 and ST-133 both represent two individuals. In the larger dataset, 9 STs are represented by more than one individual (17, 32, 34, 37, 38, 39, 44, and 58). The samples ST-62 from Spain and ST-123 from Ecuador are represented by the same loci but only labelled as 123* in this diagram. All strains represented by STs here were collected from women between 1936–2021 from the USA, UK, Spain, Austria, and Ecuador. All non-Ecuador samples are presented in red

Fig. 3

Two type population structure as determined by STRUCTURE. Type 1 I represented in red, and type 2 represented in green. Y-axis scale represents ancestry likelihood. Each bar represents an ST and the likelihood of ancestry to either type 1 (red) or type 2 (green). Samples with at least 0.8 (80%) ancestry to one of these population groups were used as a cut off to assign STs to that group, any of which had less than 0.8 ancestry were classed as unassigned samples. Blue circles represent STs from the Ecuador sample set, any STs without a blue circle represent samples taken from PubMLST

Fig. 4

goeBURST of TV samples collected from FSWs and reported previously, coloured by reported ancestry to a subpopulation. In this instance all STs are coloured according to percentage (%) likelihood of their assignment to Type I or II populations