Figures
Abstract
Background
Treponema pallidum subspecies pertenue (TPE) is the causative agent of human and nonhuman primate (NHP) yaws infection. The discovery of yaws bacterium in wild populations of NHPs opened the question of transmission mechanisms within NHPs, and this work aims to take a closer look at the transmission of the disease.
Methodology/Principal Findings
Our study determined eleven whole TPE genomes from NHP isolates collected from three national parks in Tanzania: Lake Manyara National Park (NP), Serengeti NP, and Ruaha NP. The bacteria were isolated from four species of NHPs: Chlorocebus pygerythrus (vervet monkey), Cercopithecus mitis (blue monkey), Papio anubis (olive baboon), and Papio cynocephalus (yellow baboon). Combined with previously generated genomes of TPE originating from NHPs in Tanzania (n = 11), 22 whole-genome TPE sequences have now been analyzed. Out of 231 possible combinations of genome-to-genome comparisons, five revealed an unexpectedly high degree of genetic similarity in samples collected from different NHP species, consistent with inter-species transmission of TPE among NHPs. We estimated a substitution rate of TPE of NHP origin, ranging between 1.77 × 10-7 and 3.43 × 10-7 per genomic site per year.
Conclusions/Significance
The model estimations predicted that the inter-species transmission happened recently, within decades, roughly in an order of magnitude shorter time compared to time needed for the natural diversification of all tested TPE of Tanzanian NHP origin. Moreover, the geographical separation of the sampling sites (NPs) does not preclude TPE transmission between and within NHP species.
Author summary
Treponema pallidum subsp. pertenue (TPE) is the causative agent of neglected tropical disease yaws. This bacterium infects humans in endemic regions in Africa, southern Asia and the Pacific region, and NHPs in Sub-Saharan Africa. The infection manifests as papillomas and/or ulcers, and can progress into deformation of bone and cartilage if untreated. Yaws is targeted by the currently ongoing second yaws eradication campaign, known as the Morges strategy, led by World Health Organization (WHO). In this work, we studied the transmission of the yaws bacterium in wild populations of NHPs, as they could serve as a possible source of reinfection of humans in the future, post-yaws eradication in humans. We have determined eleven whole TPE genomes from NHPs from different national parks in Tanzania. Together with previously determined genomes, we have analyzed 22 sequences, and we found a high degree of genetic similarity in samples isolated from different NHP species. This similarity is consistent with inter-species transmission of TPE among NHPs.
Citation: Janečková K, Roos C, Andrla P, Fedrová P, Tom N, Lueert S, et al. (2025) Whole-genome sequencing reveals evidence for inter-species transmission of the yaws bacterium among nonhuman primates in Tanzania. PLoS Negl Trop Dis 19(2): e0012887. https://doi.org/10.1371/journal.pntd.0012887
Editor: Stuart D. Blacksell, Mahidol Univ, Fac Trop Med, THAILAND
Received: May 29, 2024; Accepted: February 4, 2025; Published: February 26, 2025
Copyright: © 2025 Janečková et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The raw sequencing reads were uploaded under BioProject No. PRJNA1062612. S1 Table summarizes the sequencing runs and the GenBank SRA accession numbers, as well as accession numbers for each draft genome sequence.
Funding: The study was funded by the German Research Foundation (DFG): KN1097/3–1 [SK],KN1097/3–2 [SK], RO3055/2–1 [CR] and RO3055/2–2 [CR], and by the National Institute of Virology and Bacteriology (Program EXCELES, ID Project No. LX22NPO5103, Funded by the European Union - Next Generation EU) to DS. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Treponema pallidum subspecies pertenue (TPE) is the causative agent of yaws in human and nonhuman primates (NHPs) [1–3]. Human yaws is known to be endemic in countries spanning Africa, southern Asia, and the Pacific region [4]. Overall, there are 16 countries currently considered as being endemic for yaws (WHO, https://www.who.int/data/gho/data/themes/topics/topic-details/GHO/yaws-(endemic-treponematoses)). In addition, NHP infection – based on molecular diagnosis and/or visible sightings [2,3,5–9] – has also been documented in several sub-Saharan Africa countries. All countries with infected NHPs were previously known to be endemic for human yaws. While human yaws mainly affects children between the age of 6 to 10, with the main transmission route via close skin contact with an infected person, infection in NHPs in East Africa is generally associated with severe genital lesions indicative of a sexually transmitted disease [10]. However, lesions on the face and extremities also occur, especially in other regions of sub-Saharan Africa, highlighting the potential for the TPE bacterium to also cause both syphilis and yaws-like lesions in its primate host (reviewed in [11]). Until now, TPE infections have been diagnosed based on direct detection of the spirochete using darkfield microscopy in Papio papio [12] or nucleic acid amplification tests (NAAT) in several species, including Papio anubis (olive baboon), Papio cynocephalus (yellow baboon), Chlorocebus pygerythrus (vervet monkey), Chlorocebus aethiops (grivet monkey), Cercopithecus mitis (blue monkey), Gorilla gorilla gorilla (western lowland gorilla) [3], Cercocebus atys (sooty mangabey) [7], and Pan troglotydes verus (western chimpanzee) [8]. A previous molecular typing study from Tanzania – the country with the highest number of NHPs sampled – suggested possible inter-species transmission of TPE among NHPs based on strain identity [3]. This was further supported by the geographical clustering of samples and the lack of clustering according to species. However, the analysis was limited, with only two gene loci investigated.
In this work, we present eleven genomes of TPE isolates that originated from four species of NHPs in four national parks (NPs) in Tanzania. Combined with the 11 previously determined TPE genomes derived from Tanzanian NHPs, 22 TPE isolates (with determined whole genome sequences) were available for in-depth analysis. Out of 231 possible combinations of genome-to-genome comparisons, five revealed an unexpectedly high degree of genetic similarity between different NHP species, supporting the hypothesis of inter-species transmission of TPE among NHPs.
Materials and methods
Sample collection
No live animals were sampled for this study. All NHP samples were taken from our previous studies [1,13], and all the necessary ethical statements and DNA extraction details can be found there. Samples for whole-genome sequencing were selected based on (1) the availability of a sufficient treponemal DNA as determined by treponemal copy number via qPCR [2], and (2) the integrity of TPE DNA revealed by the amplification of long DNA amplicons [9]. TPE DNA integrity was determined by the positivity of long-range PCR that amplified a 4,835 bp-long region encoding the tprC gene using primers (ES-42F and TPI-11A-R) and previously published PCR conditions [14]. Selected samples were isolated from four NHP species: Chlorocebus pygerythrus (vervet monkey, n = 1), Cercopithecus mitis (blue monkey, n = 1), Papio cynocephalus (yellow baboon, n = 1) and Papio anubis (olive baboon, n = 8). Samples came from Lake Manyara NP (n = 7), Serengeti NP (n = 3), and Ruaha NP (n = 1). Further information about the samples can be found in the Results section.
DNA target enrichment and whole genome sequencing
For whole genome sequencing, samples with sufficient treponemal DNA and large fragments of TPE DNA, indicating the integrity of TPE DNA, were selected. Sample 09LMM2180815 was DpnI-enriched according to the protocol published previously by Grillová et al. [14] and sent for sequencing to Novogene (Novogene Company Limited, Cambridge, United Kingdom) with the Illumina HiSeq 2500 platform (150 bp paired-end library). All other samples were prepared as per the following protocol. Before sequencing, bacterial DNA was enriched using a Looxter Enrichment Kit (Analytik Jena, Jena, Germany) following the manufacturer’s protocol described in a previous study [9].
DNA target enrichment and library preparation followed the protocols described in a previous study [9]. Briefly, the amplified DNA was purified using SPRISelect beads (Beckman Coulter, Indianapolis, USA). We used the NEBNext Ultra II FS DNA Library Prep Kit for Illumina (New England Biolabs (NEB), Ipswich, USA) with a targeted fragment size of 300–700 bps during the enzymatic digest. Following the ligation of adapters, fragments were size-selected using AMPure XP beads according to the NEB protocol. The library was finalized through single index PCR enrichment of 15 µl adaptor-ligated DNA fragments using a Universal PCR primer (i5 Primer) in combination with the respective Index Primer (i7 Primer). The cycling conditions followed recommendations provided by NEB. Next, PCR reactions were cleaned using AMPure XP beads following standard protocols. Afterwards, quality was checked using a Bioanalyzer 2000 (Agilent, Palo Alto, USA) and quantified using a NEBNext Library Quant Kit for Illumina.
In the next step, TPE DNA was enriched using custom-made RNA baits (Arbor Biosciences, Ann Arbor, USA) using the procedure outlined in the myBaits Hybridization Capture for Targeted NGS Manual 4.0 (https://arborbiosci.com/mybaits-manuals/). The bait design is described elsewhere [2]. Following library denaturation and adapter-blocking, target library molecules were allowed to hybridize with their complementary bait. Details of the hybridization process were described previously [9]. In brief, bait-target hybrids were bound to streptavidin-coated magnetic beads that allowed us to separate the desired target DNA molecules from non-target DNA. After magnetic removal of the beads, the supernatant was purified using the SPRIselect clean-up method with NEBNext Sample Purification Beads (x 0.7; New England Biolabs (NEB), Ipswich, USA) according to the NEB protocol. The SPRIselect clean-up product was diluted 5-fold in 0.1X TE buffer and quantified on a Qubit 3.0 (ThermoFisher Scientific, Waltham, USA). In addition, 1 µl of the undiluted SPRI clean-up product was quality-checked on the Bioanalyzer.
Following NEB’s recommended protocol, library quantification was performed using a qPCR NEBNext Quant Kit for Illumina (NEB, Ipswich, USA). We reamplified the library from our first hybridization capture run to increase the yield of target DNA; we used KAPA HiFi Hot Start polymerase in combination with Illumina adapter-specific primers (sense 5’-AAT GAT ACG GCG ACC ACC GA-3’ and antisense 5’-CAA GCA GAA GAC GGC ATA CGA-3’). Cycling conditions were 2 min at 98 °C followed by five cycles of 20 sec at 98 °C, 30 sec at 60 °C, and 45 sec at 72 °C, with the last step being 5 min at 72 °C as the final phase.
The library products were quantified using Qubit 3.0 and cleaned using the SPRIselect method described above; a second round of hybridization capture was performed using the abovementioned method. The enriched 10 nM DNA libraries were pooled, and samples excluding sample 09LMM2180815 were sent for sequencing to the Transcriptome and Genome Analysis Laboratory at the University Medical School Göttingen, Germany. Our workflow included an initial 150 paired-end MiSeq run with the Nano Kit to test for the quality of the enriched libraries and their sequencing performance, prior to in-depth sequencing on Illumina’s HiSeq2500 platform. As a result, majority of samples (except sample 09LMM2180815, 13LMF5300415 and 21F8040407) were sequenced on both MiSeq and HiSeq platforms. S1 Table summarizes the sequencing runs and the GenBank SRA accession numbers.
Bioinformatic analysis
The bioinformatic analysis was performed using a previously described workflow [9,14]. The genome consensus sequence was determined based on at least three aligned reads (≥3x coverage). Chromosomal loci known to evolve under positive selection [15], genes with extensive intrastrain variability (e.g., tprK and other paralogous tpr genes [16,17], genes containing tandem repetitions (e.g., TP0433 and TP0470) and both t0012 and t0015 t-RNA genes (i.e., tRNA-Ala and tRNA-Ile [18]) were removed from the sequence alignments to omit sequences that can emerge either due to positive selection or recombination mechanisms [15,19,20]. These removed sites do not confer any phylogenetic information [14,15,20]. Sequences with these sites removed were used in all subsequent analyses.
Genome sequences were aligned in MEGAX (v.10.2.6, [21]) and used for computing pairwise distances with complete deletion of gap sections. The resulting number of differences (SNPs) in pairwise comparisons of the genomes were then used for further analyses. Bubble plot and boxplots were created in R (R Core Team; Vienna, Austria [22]) using packages readxl [23], ggplot2 [24] and cowplot [25].
Phylogenetic analysis
Phylogenetic trees were constructed in MEGAX (v.10.2.6). Positions of the alignment containing gaps and missing data were removed for further analyses. For the phylogenetic tree reconstruction of the Tanzanian NHP genomes we used the maximum-likelihood (ML) algorithm using the optimal substitution model as determined by MEGAX. Models used were HKY+G [26] and TN93+G+I [27] used with 1000 bootstrap replicates.
NHP TPE genomes from Tanzanian isolates were used (finished genomes, n = 9, shown in bold, and draft sequences, n = 13), as well as complete genomes from African, Indonesian, and Polynesian TPE strains isolated from humans [28–31]. Treponema pallidum subsp. pallidum (TPA) strains Nichols and SS14 [32] were used as outgroup. We inferred an evolutionary history by applying the initial trees for the heuristic search. Initial trees were obtained automatically by applying Neighbor-Joining (NJ) and BioNJ algorithms to a matrix of pairwise distances estimated using the HKY+G model and then selecting the topology with superior log-likelihood values. Additional analysis was performed using HKY+G model with using only 18 TPE NHP sequences after removing four draft genomes with the lowest coverage.
Phylogenetic tree (S1 Fig) including draft genome sequences of strains from other available TPE genomes of NHP from Cote d’Ivoire, Gambia, Senegal, and Tanzania [2,6,7], and also from TPE of human origin, i.e., from Papua New Guinea (specifically Lihir Island), Solomon Islands and Liberia [33–35] were constructed using HKY+G model. Genome sequences were produced from available SRA data using the same approach as described above, except for the genomes originating from the study by Mediannikov (2020) [6], where only draft assemblies are publicly available, and therefore, these draft genome sequences were used for phylogenetic analysis. TPA strains Nichols and SS14 [32] were again used as outgroup. Sequences with less than 57.8% genome coverage were not used for phylogenetic tree reconstructions.
Phylogenetic tree (S6 Fig) using TPE NHP sequences of two genes (TP0488 and TP0548) was constructed, using TN93+G+I model [27].
Evolutionary rate estimation
A ML tree was constructed from the sequence alignment of the TPE samples. This tree served as the basis for the subsequent root-to-tip regression analysis using TempEst [36]. We have used Minimizing the Mean of the Squares of the Residuals (RMS) and Maximizing R-Squared approach.
For Bayesian analysis of molecular sequences, BEAST v2.7.4 [37] was used to analyze all 22 TPE NHP sequences from Tanzania (with an additional outgroup of the syphilis-causing TPA strain SS14 [32] in models T6 and T15). Second analysis was performed on 17 TPE NHP sequences after removing four genomes with the lowest coverage and one determined by TempEst (with TPA strain SS14 as an outgroup in models T6 and T15). For the analysis, we removed all positions with gaps and missing sequence data from the alignment. Tip dates were specified based on the years of sample collection. To assess model fit, we employed Nested Sampling in BEAST2. To infer the evolutionary history, we employed the HKY model and the strict clock model in combination with three other different models, including the coalescent constant population demographic model (T5), the birth-death model (T6), and the Yule model (T15). The BEAST analysis was run for 10 million Markov chain Monte Carlo (MCMC) generations with a 10% burn-in. We combined results from three runs with LogCombiner 2.4 [37], resulting in 27 million computed generations. Parameter estimates were obtained by Tracer v1.7.2 [38], and a maximum clade credibility tree was generated with TreeAnnotator v2.7.4 [37]. The resulting tree was visually represented in FigTree v1.4.4 [39]. Two NHP TPE whole genome sequences (without putative recombinant sites, positively selected sequences, repetitions, and tpr sequences) of most closely genetically related strains were selected for the estimation of substitution rates in NHP TPE.
Results
Characteristics of eleven nonhuman primate TPE whole genomes
The Tanzanian NHP TPE-genomes newly generated in this study (n = 11) are shown in Table 1. The final coverage of these genomes ranged from 57.8% to 98.9% (with a sequencing depth of ≥3×), and the average sequencing depth ranged from 4× to 116× (Table 1).
A geographical representation of the TPE-genome origins (n = 11) and the previously published TPE genomes (n = 11) [2,9] are shown in Fig 1. Most of the samples originated from Lake Manyara NP (n = 14), followed by Serengeti NP (n = 5), Ruaha NP (n = 2) and Ngorongoro Conservation Area (n = 1).
National parks and the Ngorongoro Conservation Area are shown in green. Whole genome sequences of NHP TPE isolates from Tanzania are shown as circles. Circle fill color corresponds to NHP species. Inter-species comparisons with low levels of genetic diversity (as determined later) are shown in boxes. Made with Natural Earth (base layer available at https://www.naturalearthdata.com/features/).
A pairwise comparison of whole genome TPE sequences
A pairwise comparison of whole NHP TPE genome sequences of Tanzanian strains is shown in Fig 2. All possible pairwise combinations (n = 231; 22×21/2) were tested for nucleotide differences. The median number of detected nucleotide differences (n = 41) was relatively high; however, in some cases, the number of differences dropped to zero (Fig 2).
= 231) based on multiple sequence alignments of Tanzanian NHP TPE genome sequences. Complete gap deletion algorithm was used. There was a total of 386,092 positions in the final dataset. Exact numbers of detected nucleotide differences can be found in S3 Table. The size of the circles denotes the number of detected nucleotide differences between pairs of genomes, and the fill color corresponds to the population in a particular sampling area. The red outer circle color denotes the comparison of TPE isolates obtained from different species, while the blue outer circle denotes NHP intraspecies comparisons. Color of the axes labels corresponds to NHP species. Small circles with red rings correspond to interspecies comparisons with nearly identical genome sequences (boxed).
Values of pairwise combination differences discovered through whole genome sequence comparisons
All of the values of pairwise differences determined with complete deletion option (386,092 nt positions in the final dataset) are shown in S3 Table. The values that were obtained using pairwise deletion option can be seen in S4 Table. The length of the genomes used for the pairwise comparisons can be found in S5 Table. The median value of pairwise differences was 41 nucleotides between pairs of genomes with a range from 0–79 (Fig 3). Inter-species combinations showed a higher median value (med(x) = 46) compared to the intra-species median value (med(x) = 33). Similarly, a difference was found between pairwise differences derived from the same sampling site (med(x) = 34.5) compared to pairwise differences coming from different sites (different NPs; med(x) = 44). When both comparisons were combined, the differences were even higher (Fig 4), i.e., the median number of nucleotide differences for inter-species comparisons in different locations (med(x) = 43) was higher compared to intra-species comparisons from the same site (med(x) = 20).
While the median of all distances was 41 nucleotides, inter-species and inter-national park distances were higher compared to intra-species and within national park differences, respectively. The lower and upper hinges (edges) correspond to the first and third quartiles (i.e., the 25th and 75th percentiles). Data points beyond the whiskers are considered outliers and plotted individually.
NHP TPE genomes from Tanzanian isolates used were either finished genomes (n = 9, shown in bold) or partially determined (draft sequences, n = 13), as well as completely sequenced genomes from African, Indonesian, and Polynesian TPE strains isolated from humans [28–31]. TPA strains Nichols and SS14 [32] were used as an outgroup. The evolutionary history was inferred using the Maximum likelihood and HKY+G model [26]. All positions containing gaps and missing data were eliminated (complete deletion option). There were 1,981 variable sites in the original dataset. There were 385,875 positions in the final dataset (35% of the whole genome (1,093 kb)). Bootstrap support (1000 replicates) is shown next to the branches. The scale corresponds to the number of substitutions per nucleotide. The geographical and species origin of isolates is shown next to the tree. Finished genomes are shown in bold.
A pairwise comparison of TPE genomes obtained from different NHP species
The pairwise comparison of all analyzed TPE genomes isolated from different NHP species allowed us to identify genomes with the lowest number of nucleotide differences (Table 2). Among five such genome comparisons, the number of differences ranged between 3–6 nucleotides when comparisons were based on a complete gap deletion algorithm using multiple alignments where only nucleotide positions present in all genomes were compared (386,092 nt positions in the final dataset). In the pairwise comparisons based on the de novo alignment of every two sequences, the number of detected nucleotide differences was higher and ranged between 10–17 (corresponding to a genome sequence identity of 99.9991%–99.9985%), and the number of amino acid replacements in the corresponding proteome ranged between five and ten. The genes showing differences between highly related TPE samples and the corresponding amino acid changes are shown in S2 Table.
Phylogenetic analysis
The whole genome NHP TPE sequences newly generated in this study (n = 11), augmented with additional NHP TPE genomes from Tanzania (n = 11; [2,9]), were analyzed in the context of other available NHP and human TPE genomes. Despite the relatively low bootstrap support, all 22 isolates from Tanzania analyzed to date clustered separately from the twelve NHP TPE genome sequences from strain/isolates originating in West African countries, i.e., Senegal, Gambia, Guinea, and Côte d’Ivoire [2,5,6] (S1 Fig). The phylogeny of the Tanzanian NHP TPE-genomes together with the Fribourg-Blanc strain that infects NHPs in West Africa, as well as selected complete TPE genomes from African, Indonesian, and Polynesian strains of human origin, is shown in Fig 4. We note here that several whole genome sequences were nearly identical (e.g., 32LMM2190317 and 34LMM2190317, both derived from samples from Chlorocebus pygerythrus, with one nucleotide difference, corresponding to several years of separate evolution) and some were isolated from different NHP species (e.g., 6RUM2090716 (Chlorocebus pygerythrus) and 09LMM2180815 (Cercopithecus mitis)). Additional phylogenetic tree (S2 Fig) with four lowest coverage genomes removed has also been calculated.
Estimated substitution rates revealed by comparison of the most closely related genomes and BEAST analysis of genomic sequences
In this study, we used TempEst to perform root-to-tip regression analysis on a dataset comprising 22 TPE samples from Tanzania (S3 and S4 Figs). The dataset spans a 10-year collection period, allowing us to evaluate the consistency of the temporal signal across the samples. We conducted the BEAST analysis twice: first, with all 22 samples included, and second, with 17 samples after exclusion of four samples exhibiting the lowest average sequencing depth and one other outlying sample as determined by TempEst. These two analyses aimed to investigate the impact of low-quality data on the inferred evolutionary rates and the robustness of the temporal signal.
Two TPE whole genome sequences, including LMNP-1 (sample isolated in April 2007; [2]) and 19LMF8280815 (sample isolated in August 2015; [9]), both coming from olive baboons at Lake Manyara NP, were selected for the estimation of substitution rates in NHP TPE. Both genomes differed in two nucleotides when whole genomes without putative recombinant sites, positively selected sequences, repetitions, and tprK sequences were used, respectively. If both strains evolved from a common ancestor, the corresponding substitution rate would be equal to 2.20 × 10-7 per genomic site per year (2 nt differences between genomes; 1,092,306 and 1,092,317 nt genome length for 19LMF8280815 and LMNP-1, respectively; both TPE were isolated 8.33 years apart). In addition, BEAST analysis of all 22 NHP TPE sequences from Tanzania was used to estimate substitution rates using BEAST models T5, T6, and T15 (with an outgroup of TPA SS14 [32] used in models T6 and T15). In this analysis, the resulting length of analysed sequence was 386,063 nt (35% of the whole genome (1,093 kb)). This revealed estimated median substitution rates of 3.43 × 10-7 (95% HPD 5.16×10-10 – 8.64×10-7), 2.22 × 10-7 (95% HPD 7.41×10-9 – 5.17×10-7), and 2.76 × 10-7 (95% HPD 7.21×10-9 – 6.39×10-7), respectively (Table 3 and S5 Fig). Using BEAST models, these substitution rates were used to further estimate the elapsed time from the most recent common ancestor (Table 3). The lowest estimated times from the most recent common ancestor of TPE isolated from different species ranged between 51.3 and 99.3 years, while the estimated time from the most recent common ancestor of the most divergent TPE from Tanzanian NHPs ranged between 528.2 and 987.5 years.
To further extend our analysis, we performed the same BEAST analysis while removing five sequences of NHP TPE with the lowest coverage, resulting in 874,228 nt genome length alignment used for analysis (80% of the whole genome (1,093 kb)). This additional analysis revealed estimated median substitution rates of 1.77 × 10-7 (95% HPD 1.27×10-10 – 4.00×10-7), 1.81 × 10-7 (95% HPD 1.03×10-8 – 3.77×10-7), and 2.29 × 10-7 (95% HPD 2.23×10-8 – 4.73×10-7), respectively (Table 4). Using BEAST models, these substitution rates were used to further estimate the elapsed time from the most recent common ancestor (Table 4). The lowest estimated times from the most recent common ancestor of TPE isolated from different species ranged between 31.4 and 121.0 years, while the estimated time from the most recent common ancestor of the most divergent TPE from Tanzanian NHPs ranged between 393.5 and 1148.0 years.
Discussion
Today, Tanzanian NHP TPE genomes represent the majority of the available NHP TPE sequence data compared to the twelve NHP TPE draft genome sequences coming from NHPs sampled in Senegal, Gambia, Guinea, and Côte d’Ivoire [2,5,6]. Our recent work contributes another eleven NHP TPE genomes of Tanzanian origin. Combined with the previously determined NHP TPE genomes from the same region [2,9] a total of 22 TPE genome sequences from four NHP species (sampled at four different locations between 2007 and 2017) are now available. A previous multi-locus sequence and phylogenetic network analysis of TPE strains infecting African primates, including the above-stated areas, revealed a geographic clustering of analyzed TPE strains [2,3]. Together with the absence of clustering according to host species, it suggests ongoing inter-species transmission at the local level. To further test this hypothesis, TPE genome sequences from Tanzanian NHPs were analyzed regarding the presence and degree of genomic diversity. Unlike the previous work [9], where the TPE sample set was selected based on genetic diversity revealed from two gene loci (TP0488 and TP0548 [3]) and was therefore of limited importance to address this question, in this work we selected samples based on the availability of samples and their quality (i.e., DNA integrity). Despite the limited number of samples (n = 22), our current sample set represents an unbiased collection for investigating the genetic diversity among the different NHP-derived TPE genomes.
As revealed by the phylogenetic analysis, the geographical separation of sampling areas within Tanzania is not entirely unambiguous and, therefore, different geographical locations do not appear to limit transmission of TPE within Tanzania (e.g., Ruaha NP vs Lake Manyara NP, ~360 km distance). However, as expected, the genetic relationship within the Tanzanian NHP TPE strains is closer than that between Tanzanian NHP-derived genomes and those infecting NHPs in West Africa (S1 Fig). The same effect is equally visible at the local scale. Analyses of genetic diversity based on whole genome alignments within the four sampled national parks in Tanzania, for example, revealed lower genetic diversity within the geographic sampling locations compared to the genetic diversity calculated for TPE comparisons between different parks (Fig 3). This suggests that the epidemiological spread of TPE is fastest on the local scale, with a lower rate of spread across the sampling sites, and within the African continent.
The nature of the TPE-host system can explain our findings. The bacterium, for example, does not survive for long in a non-host environment [31], limiting the spread, e.g., through flies as potential mechanical vectors [40,41]. Although modern, well-characterized experimental data are missing, all current data, i.e., human infections, suggest close skin-to-skin (sexual and non-sexual) contact as the main route of infection. Moreover, NHPs are relatively resident, and migration rarely happens over longer distances [42], explaining the slower spread of the disease across the different regions in Tanzania. In addition to geographic barriers in TPE transmission, similar analysis within and between different NHP species revealed a similar difference in genetic diversity, supporting our previous work that the disease in Tanzanian NHPs is mainly sexually transmitted [1]. However, sexual behavior in NHPs is a porous species barrier to TPE transmission, as shown, for example, by the inverted intergeneric introgression between the critically endangered kipunjis (Rungwecebus kipunji) and yellow baboons (Papio cynocephalus) in two disjunct populations in Tanzania [43]. Additionally, different primate species can form mixed-species groups [44–46], and interaction between the individuals of different species, either aggressive (hitting, biting, leg or tail-pulling) or playful, including grooming activities, is a well-documented phenomenon [47–49].
As expected, highly related samples were found when the same species were sampled in one area at the same time (e.g., 32LMM2190317 and 34LMM2190317 differing in just one nucleotide, [9]; 2SNF2130815 and 6SNF2081115 that are genetically identical in the analyzed sequences, this study). Infection with the same (or a highly related) TPE strain is, therefore, common in a population at a given time. This aligns with our previous work, where we could show, based on MLST, that the number of strains infecting olive baboons is limited and, at Lake Manyara NP, ranges between six in 2007 and five in 2015 [3]. Surprisingly, genetically highly related samples were also found when different NHP species were sampled in one area simultaneously or in different NPs at different time points. Previous predictions on inter-species transmission between different NHP species were based on analyses of two chromosomal loci (TP0488 and TP0548) [3]; however, these loci were found to be recombinant and positively selected in TPA and Treponema pallidum subsp. endemicum (TEN) treponemes [15,19,20,50,51]. The phylogenetic tree based on analysis of these two loci differed from the tree based on whole genome sequences (S6 Fig); however, due to the low bootstrap support of the two-loci-based tree, it is not clear how the difference between the trees is relevant. The convergent evolution of these loci could explain the observed similarities based on these two chromosomal sequences. Since analyses based on whole genome sequences do not allow this explanation, highly related TPE isolates infecting different NHP species provide clear evidence for direct inter-species transmission of TPE strains.
We have used several methods and models to estimate the substitution rate of NHP TPE. These estimates ranged between 1.77 × 10-7 and 3.43 × 10-7 per genomic site per year, varying slightly more than by a two-fold difference. By analysis of genetically related TPE strains isolated at different time points and propagated in laboratory animals, substitution rates of human TPE isolates have been estimated to be 1.21 × 10−7 per nucleotide site per year or lower [52]; the substitution rate of TPA has been estimated to be 0.82 × 10−7 per nucleotide site per year or lower [53], using two TPA isolates one of which was propagated in rabbits. However, these values are very close to the rates estimated in this study. Moreover, the estimated substitution rates in other studies based on alignments of draft genomes of TPA sequences were equally similar to our estimated rates, i.e., 3.02 × 10−7 [54], 6.6 × 10−7 [55], and 1 × 10−7 [56] per nucleotide site per year. However, substitution rate estimates are likely subject to variation across different T. pallidum subspecies and strains that infect humans and, in the case of TPE, also NHPs. Nevertheless, our estimations of substitution rates allow us to predict that the inter-species transmission among Tanzanian NHPs happened in recent history, in the order of decades (31.4 – 121.0 years), which is about an order of magnitude faster than the time interval needed for natural diversification of all known TPE strains that infect Tanzanian NHPs (393.5 – 1148.0 years) as estimated by our BEAST analysis.
While this study provides further proof for inter-species transmission of TPE among NHPs, the potential transmission of the yaws bacterium from NHPs to humans remains hypothetical. Our investigations of humans in areas in Tanzania where NHPs are infected found no indication of yaws infection [57]. Unfortunately, no samples or data are available from the time when Tanzania reported its largest number of human infections, i.e., between 1930 and 1950 [58]. These data would help to trace the origin of human infections. A clear obstacle for current TPE research is the scarcity of available samples or data from one geographic area at a given period. Our work shows ongoing inter-species transmission in primates, a mammalian order taxon that also includes humans. This underlines the need to continue disease surveillance in humans that co-exist with infected NHPs even where yaws has been eliminated in humans. Any undetected – though rare – spillover could potentially result in a re-emergence of yaws in humans. More research, using complete genomes, is warranted in sub-Saharan Africa to investigate the relationship between NHPs and humans relative to TPE infections.
Supporting information
S1 Table. Overview of samples and sequencing platforms used, including GenBank SRA and Genome accession numbers of the BioProject No. PRJNA1062612.
https://doi.org/10.1371/journal.pntd.0012887.s001
(XLSX)
S2 Table. Comparison of TPE genomes isolated from different NHP species.
The number of detected single nucleotide variants (SNVs), amino acid changes, and affected genes are shown.
https://doi.org/10.1371/journal.pntd.0012887.s002
(XLSX)
S3 Table. Pairwise comparison of TPE genomes using the complete deletion algorithm.
Numbers of detected single nucleotide variants (SNVs) using complete deletion.
https://doi.org/10.1371/journal.pntd.0012887.s003
(XLSX)
S4 Table. Pairwise comparison of TPE genomes using the pairwise deletion algorithm.
Numbers of detected single nucleotide variants (SNVs) using pairwise deletion.
https://doi.org/10.1371/journal.pntd.0012887.s004
(XLSX)
S5 Table. Pairwise comparison of TPE genomes using the pairwise deletion algorithm – length of the genome used for each comparison.
https://doi.org/10.1371/journal.pntd.0012887.s005
(XLSX)
S1 Fig. Phylogenetic analysis of all available TPE genomes isolated from NHPs or humans.
Draft genomes available from other studies were used [2,6,7,33–35]. Sequences with less than 57.8% genome coverage were not used for phylogenetic tree reconstructions. All positions containing gaps and missing data were deleted. The evolutionary history was inferred using the Maximum Likelihood method and HKY+G model [26]. As an outgroup, human syphilis genomes of strains Nichols and SS14 [32] were used. There were 93 nucleotide sequences and a total of 256,792 positions in the final dataset. Bootstrap support (1000 replicates) is shown next to the branches. The scale corresponds to the number of substitutions per nucleotide. TPE genomes from Tanzanian NHP isolates clustered together.
https://doi.org/10.1371/journal.pntd.0012887.s006
(TIF)
S2 Fig. Phylogenetic analysis of 18 Tanzanian NHP TPE samples.
NHP TPE genomes from Tanzanian isolates. Completely sequenced genomes from African, Indonesian, and Polynesian TPE strains isolated from humans are also shown [28–31]. As an outgroup, human syphilis genomes of strains Nichols and SS14 [32] were used. The evolutionary history was inferred using the Maximum likelihood and HKY+G model [26]. All positions containing gaps and missing data were eliminated (complete deletion option). There were 873,701 positions in the final dataset. Bootstrap support (1000 replicates) is shown next to the branches. The scale corresponds to the number of substitutions per nucleotide.
https://doi.org/10.1371/journal.pntd.0012887.s007
(TIF)
S3 Fig. Root-to-Tip Regression with Minimizing RMS (Residual mean squared).
A. Phylogenetic tree. B. Root-to-tip regression graph.
https://doi.org/10.1371/journal.pntd.0012887.s008
(TIF)
S4 Fig. Root-to-Tip Regression with Maximizing R-Squared. A.
Phylogenetic tree. B. Root-to-tip regression graph.
https://doi.org/10.1371/journal.pntd.0012887.s009
(TIF)
S5 Fig. Phylogenetic analysis of Tanzanian TPE samples isolated from NHPs analyzed using the BEAST T5 model.
The human syphilis SS14 genome [32] was used as an outgroup. The estimated time elapsed from the most common ancestor is shown next to the branches. The length of the scale bar corresponds to 1000 years.
https://doi.org/10.1371/journal.pntd.0012887.s010
(TIF)
S6 Fig. Phylogenetic analysis of Tanzanian TPE samples isolated from NHPs based on two genes (TP0488 and TP0548).
All positions containing gaps and missing data were deleted. The evolutionary history was inferred using the Maximum Likelihood method and TN93+G+I model [27]. As an outgroup, the human syphilis strains Nichols and SS14 genome [32] were used. There were 33 nucleotide sequences and a total of 1,714 positions in the final dataset. Bootstrap support (1000 replicates) is shown next to the branches. The scale corresponds to the number of substitutions per nucleotide.
https://doi.org/10.1371/journal.pntd.0012887.s011
(TIF)
Acknowledgments
Computational resources were supplied by the project “e-Infrastruktura CZ” (e-INFRA LM2018140) provided within the program Projects of Large Research, Development, and Innovations Infrastructures. We thank Thomas Secrest (Secrest Editing, Ltd.) for his assistance with the English revision of the manuscript.
References
- 1. Knauf S, Batamuzi EK, Mlengeya T, Kilewo M, Lejora IAV, Nordhoff M, et al. Treponema infection associated with genital ulceration in wild baboons. Vet Pathol. 2012;49(2):292–303. pmid:21411621
- 2. Knauf S, Gogarten JF, Schuenemann VJ, De Nys HM, Düx A, Strouhal M, et al. Nonhuman primates across sub-Saharan Africa are infected with the yaws bacterium Treponema pallidum subsp. pertenue. Emerg Microbes Infect. 2018;7(1):157. pmid:30228266
- 3. Chuma IS, Roos C, Atickem A, Bohm T, Anthony Collins D, Grillová L, et al. Strain diversity of Treponema pallidum subsp. pertenue suggests rare interspecies transmission in African nonhuman primates. Sci Rep. 2019;9(1):14243. pmid:31578447
- 4. Asiedu K, Fitzpatrick C, Jannin J. Eradication of yaws: historical efforts and achieving WHO’s 2020 target. PLoS Negl Trop Dis. 2014;8(9):e3016. pmid:25254372
- 5. Zobaníková M, Strouhal M, Mikalová L, Čejková D, Ambrožová L, Pospíšilová P, et al. Whole genome sequence of the treponema fribourg-blanc: unspecified simian isolate is highly similar to the yaws subspecies. PLoS Negl Trop Dis. 2013;7(4):e2172.
- 6. Mediannikov O, Fenollar F, Davoust B, Amanzougaghene N, Lepidi H, Arzouni J-P, et al. Epidemic of venereal treponematosis in wild monkeys: a paradigm for syphilis origin. New Microbes New Infect. 2020;35:100670. pmid:32368345
- 7. Mubemba B, Gogarten JF, Schuenemann VJ, Düx A, Lang A, Nowak K, et al. Geographically structured genomic diversity of non-human primate-infecting Treponema pallidum subsp. pertenue. Microb Genom. 2020;6(11):mgen000463. pmid:33125317
- 8. Mubemba B, Chanove E, Mätz-Rensing K, Gogarten JF, Düx A, Merkel K, et al. Yaws disease caused by treponema pallidum subspecies pertenue in wild chimpanzee, Guinea, 2019. Emerg Infect Dis. 2020;26(6):1283–6. pmid:32441635
- 9. Janečková K, Roos C, Fedrová P, Tom N, Čejková D, Lueert S, et al. The genomes of the yaws bacterium, Treponema pallidum subsp. pertenue, of nonhuman primate and human origin are not genomically distinct. PLoS Negl Trop Dis. 2023;17(9):e0011602. pmid:37703251
- 10. Paciência FMD, Rushmore J, Chuma IS, Lipende IF, Caillaud D, Knauf S, et al. Mating avoidance in female olive baboons (Papio anubis) infected by Treponema pallidum. Sci Adv. 2019;5(12):eaaw9724. pmid:31840059
- 11.
Harper KN, Knauf S. Treponema pallidum Infection in Primates: Clinical Manifestations, Epidemiology, and Evolution of a Stealthy Pathogen. In: Brinkworth JF, Pechenkina K, editors. Primates, Pathogens, and Evolution. New York, NY: Springer; 2013. p. 189–219. https://doi.org/10.1007/978-1-4614-7181-3_7
- 12. Fribourg-Blanc A, Mollaret HH, Niel G. Serologic and microscopic confirmation of treponemosis in Guinea baboons. Bull Soc Pathol Exot Filiales. 1966;59(1):54–9. pmid:5333741
- 13. Chuma IS, Batamuzi EK, Collins DA, Fyumagwa RD, Hallmaier-Wacker LK, Kazwala RR, et al. Widespread treponema pallidum infection in nonhuman primates, Tanzania. Emerg Infect Dis. 2018;24(6):1002–9. pmid:29774840
- 14. Grillová L, Oppelt J, Mikalová L, Nováková M, Giacani L, Niesnerová A, et al. Directly sequenced genomes of contemporary strains of syphilis reveal recombination-driven diversity in genes encoding predicted surface-exposed antigens. Front Microbiol. 2019;10:1691. pmid:31417509
- 15. Maděránková D, Mikalová L, Strouhal M, Vadják Š, Kuklová I, Pospíšilová P, et al. Identification of positively selected genes in human pathogenic treponemes: syphilis-, yaws-, and bejel-causing strains differ in sets of genes showing adaptive evolution. PLoS Negl Trop Dis. 2019;13(6):e0007463. pmid:31216284
- 16. Čejková D, Strouhal M, Norris SJ, Weinstock GM, Šmajs D. A retrospective study on genetic heterogeneity within treponema strains: subpopulations are genetically distinct in a limited number of positions. PLoS Negl Trop Dis. 2015;9(10):e0004110. pmid:26436423
- 17. Šmajs D, Strouhal M, Knauf S. Genetics of human and animal uncultivable treponemal pathogens. Infect Genet Evol. 2018;61:92–107. pmid:29578082
- 18. Čejková D, Zobaníková M, Pospíšilová P, Strouhal M, Mikalová L, Weinstock GM, et al. Structure of rrn operons in pathogenic non-cultivable treponemes: sequence but not genomic position of intergenic spacers correlates with classification of Treponema pallidum and Treponema paraluiscuniculi strains. J Med Microbiol. 2013;62(Pt 2):196–207. pmid:23082031
- 19. Noda AA, Méndez M, Rodríguez I, Šmajs D. Genetic recombination in treponema pallidum: implications for diagnosis, epidemiology, and vaccine development. Sex Transm Dis. 2022;49(1):e7–10. pmid:34618784
- 20. Pla-Díaz M, Sánchez-Busó L, Giacani L, Šmajs D, Bosshard PP, Bagheri HC, et al. Evolutionary Processes in the Emergence and Recent Spread of the Syphilis Agent, Treponema pallidum. Mol Biol Evol. 2022;39(1):msab318. pmid:34791386
- 21. Kumar S, Stecher G, Li M, Knyaz C, Tamura K. MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol. 2018;35(6):1547–9. pmid:29722887
- 22. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. 2020. Available from: https://www.R-project.org/
- 23.
Wickham H, Bryan J. readxl: Read Excel Files. 2023. Available from: https://readxl.tidyverse.org, https://github.com/tidyverse/readxl
- 24.
Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York; 2016. Available from: https://ggplot2.tidyverse.org
- 25.
Wilke CO. cowplot – Streamlined plot theme and plot annotations for ggplot2. 2020. Available from: https://wilkelab.org/cowplot/
- 26. Hasegawa M, Kishino H, Yano T. Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. J Mol Evol. 1985;22(2):160–74. pmid:3934395
- 27. Tamura K, Nei M. Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evol. 1993;10(3):512–26. pmid:8336541
- 28. Gastinel P, Vaisman A, Hamelin A, Dunoyer F. Study of a recently isolated strain of Treponema pertenue. Prophyl Sanit Morale. 1963;35:182–8. pmid:13946770
- 29. Liska SL, Perine PL, Hunter EF, Crawford JA, Feeley JC. Isolation and transportation ofTreponema pertenue in golden hamsters. Curr Microbiol. 1982;7(1):41–3.
- 30. Noordhoek GT, Engelkens HJ, Judanarso J, van der Stek J, Aelbers GN, van der Sluis JJ, et al. Yaws in West Sumatra, Indonesia: clinical manifestations, serological findings and characterisation of new Treponema isolates by DNA probes. Eur J Clin Microbiol Infect Dis. 1991;10(1):12–9. pmid:2009873
- 31. Turner TB, Hollander DH. Biology of the treponematoses based on studies carried out at the international treponematosis laboratory center of the johns hopkins university under the auspices of the World Health Organization. Monogr Ser World Health Organ. 1957;(35):3–266. pmid:13423342
- 32. Pětrošová H, Pospíšilová P, Strouhal M, Čejková D, Zobaníková M, Mikalová L, et al. Resequencing of Treponema pallidum ssp. pallidum strains Nichols and SS14: correction of sequencing errors resulted in increased separation of syphilis treponeme subclusters. PLoS One. 2013;8(9):e74319. pmid:24058545
- 33. Beale MA, Noguera-Julian M, Godornes C, Casadellà M, González-Beiras C, Parera M, et al. Yaws re-emergence and bacterial drug resistance selection after mass administration of azithromycin: a genomic epidemiology investigation. Lancet Microbe. 2020;1(6):e263–71. pmid:35544222
- 34. Marks M, Fookes M, Wagner J, Butcher R, Ghinai R, Sokana O, et al. Diagnostics for yaws eradication: insights from direct next-generation sequencing of cutaneous strains of treponema pallidum. Clin Infect Dis. 2018;66(6):818–24. pmid:29045605
- 35. Timothy JWS, Beale MA, Rogers E, Zaizay Z, Halliday KE, Mulbah T, et al. Epidemiologic and genomic reidentification of yaws, liberia. Emerg Infect Dis. 2021;27(4):1123–32. pmid:33754988
- 36. Rambaut A, Lam TT, Max Carvalho L, Pybus OG. Exploring the temporal structure of heterochronous sequences using TempEst (formerly Path-O-Gen). Virus Evol. 2016;2(1):vew007. pmid:27774300
- 37. Bouckaert R, Heled J, Kühnert D, Vaughan T, Wu C-H, Xie D, et al. BEAST 2: a software platform for Bayesian evolutionary analysis. PLoS Comput Biol. 2014;10(4):e1003537. pmid:24722319
- 38. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior summarization in bayesian phylogenetics using tracer 1.7. Syst Biol. 2018;67(5):901–4. pmid:29718447
- 39.
Rambaut A. FigTree. Andrew Rambaut Group; 2018. Available from: https://github.com/rambaut/figtree/
- 40. Gogarten JF, Düx A, Mubemba B, Pléh K, Hoffmann C, Mielke A, et al. Tropical rainforest flies carrying pathogens form stable associations with social nonhuman primates. Mol Ecol. 2019;28(18):4242–58. pmid:31177585
- 41. Knauf S, Raphael J, Mitjà O, Lejora IAV, Chuma IS, Batamuzi EK, et al. Isolation of Treponema DNA from Necrophagous Flies in a Natural Ecosystem. EBioMedicine. 2016;11:85–90. pmid:27488881
- 42. Ferreira da Silva MJ, Kopp GH, Casanova C, Godinho R, Minhós T, Sá R, et al. Disrupted dispersal and its genetic consequences: comparing protected and threatened baboon populations (Papio papio) in West Africa. PLoS One. 2018;13(4):e0194189. pmid:29614097
- 43. Zinner D, Chuma IS, Knauf S, Roos C. Inverted intergeneric introgression between critically endangered kipunjis and yellow baboons in two disjunct populations. Biol Lett. 2018;14(1):20170729. pmid:29343565
- 44. Chapman CA, Chapman LJ. Interdemic variation in mixed-species association patterns: common diurnal primates of Kibale National Park, Uganda. Behavioral Ecology and Sociobiology. 2000;47(3):129–39.
- 45. Chapman CA, Chapman LJ. Mixed-species primate groups in the kibale forest: Ecological constraints on association. International Journal of Primatology. 1996;17(1):31–50.
- 46. Cords M. Vigilance and mixed-species association of some East African forest monkeys. Behav Ecol Sociobiol. 1990;26(4):297–300.
- 47. Falótico T, Mendonça-Furtado O, Fogaça MD, Tokuda M, Ottoni EB, Verderane MP. Wild robust capuchin monkey interactions with sympatric primates. Primates. 2021;62(4):659–66. pmid:33948760
- 48. Ihobe H. Interspecific interactions between wild pygmy chimpanzees (Pan paniscus) and red colobus (Colobus badius). Primates. 1990;31(1):109–12.
- 49. Rose LM, Perry S, Panger MA, Jack K, Manson JH, Gros-Louis J, et al. Interspecific interactions between cebus capucinus and other species: data from three costa rican sites. International Journal of Primatology. 2003;24(4):759–96.
- 50. Mikalová L, Strouhal M, Oppelt J, Grange PA, Janier M, Benhaddou N, et al. Human Treponema pallidum 11q/j isolate belongs to subsp. endemicum but contains two loci with a sequence in TP0548 and TP0488 similar to subsp. pertenue and subsp. pallidum, respectively. PLoS Negl Trop Dis. 2017;11(3):e0005434. pmid:28263990
- 51. Pětrošová H, Zobaníková M, Čejková D, Mikalová L, Pospíšilová P, Strouhal M, et al. Whole genome sequence of Treponema pallidum ssp. pallidum, strain Mexico A, suggests recombination between yaws and syphilis strains. PLoS Negl Trop Dis. 2012;6(9):e1832. pmid:23029591
- 52. Strouhal M, Mikalová L, Haviernik J, Knauf S, Bruisten S, Noordhoek GT, et al. Complete genome sequences of two strains of Treponema pallidum subsp. pertenue from Indonesia: modular structure of several treponemal genes. PLoS Negl Trop Dis. 2018;12(10):e0006867. pmid:30303967
- 53. Grillová L, Giacani L, Mikalová L, Strouhal M, Strnadel R, Marra C, et al. Sequencing of Treponema pallidum subsp. pallidum from isolate UZ1974 using anti-treponemal antibodies enrichment: first complete whole genome sequence obtained directly from human clinical material. PLoS One. 2018;13(8):e0202619. pmid:30130365
- 54. Lieberman NAP, Lin MJ, Xie H, Shrestha L, Nguyen T, Huang M-L, et al. Treponema pallidum genome sequencing from six continents reveals variability in vaccine candidate genes and dominance of Nichols clade strains in Madagascar. PLoS Negl Trop Dis. 2021;15(12):e0010063. pmid:34936652
- 55. Arora N, Schuenemann VJ, Jäger G, Peltzer A, Seitz A, Herbig A, et al. Origin of modern syphilis and emergence of a pandemic Treponema pallidum cluster. Nat Microbiol. 2016;216245. pmid:27918528
- 56. Majander K, Pla-Díaz M, du Plessis L, Arora N, Filippini J, Pezo-Lanfranco L, et al. Redefining the treponemal history through pre-Columbian genomes from Brazil. Nature. 2024;627(8002):182–8. pmid:38267579
- 57. Lubinza CKC, Lueert S, Hallmaier-Wacker LK, Ngadaya E, Chuma IS, Kazwala RR, et al. Serosurvey of Treponema pallidum infection among children with skin ulcers in the Tarangire-Manyara ecosystem, northern Tanzania. BMC Infect Dis. 2020;20(1):392. pmid:32493291
- 58. Hackett CJ. Extent and nature of the yaws problem in Africa. Bull World Health Organ. 1953;8(1–3):129–82; discussion 205-10. pmid:13042577