A 5500-year-old Treponema pallidum genome from Sabana de Bogotá, Colombia

Abstract

INTRODUCTION: The origins and early diversification of Treponema pallidum subspecies, which cause syphilis, yaws, and bejel, remain poorly resolved despite paleopathological evidence for treponemal disease in pre-Columbian populations. Sparse genomic data from ancient, precolonial contexts create a multimillennial disconnect between osteological evidence of treponematosis and its molecular record. RATIONALE: To better resolve the evolutionary history of T. pallidum, we performed genomic reconstruction and analysis of pathogen DNA detected in shotgun-sequenced DNA data for a human population genomics project. Although the skeleton showed no macroscopic signs of treponematosis, our metagenomic screening identified T. pallidum in an individual who died ~5500 years ago [5464 to 5309 calendar years before the present (cal yr B.P.), 2σ] and was excavated from Middle Holocene contexts of the Tequendama I rock shelter in Sabana de Bogotá, Colombia. We applied multiple methods (competitive mapping, phylogenetics reconstructions, genome-wide average nucleotide identity statistics, and patristic distance) to confidently identify this pathogen as T. pallidum, reconstruct the genome (TE1-3), and evaluate its relationship to other genomically characterized Treponema. With this genome, we inferred divergence estimates of TE1-3 from other T. pallidum to reconstruct the early diversification of T. pallidum subspecies. RESULTS: Our analyses place TE1-3 as an early-diverging sister lineage to all modern T. pallidum subspecies, unveiling a previously unknown subspecies. Bayesian molecular clock analyses estimate the divergence between TE1-3 and other T. pallidum lineages to ~13,700 years ago [95% highest posterior density (HPD): 6768 to 20,592 cal yr B.P.]. This suggests that the divergence between TE1-3 and the modern, genomically characterized T. pallidum subspecies occurred during the Late Pleistocene to Early Holocene, closely following the peopling of the Americas, whereas the diversification of the known subspecies themselves took place more recently, within the Holocene ~6000 cal yr B.P. (3622 to 9452 cal yr B.P. 95% HPD). After assessing the breadth of coverage, we determined that TE1-3 likely possessed virulenceassociated genes found in modern T. pallidum strains, suggesting similar genetic potential for virulence. CONCLUSION: Our results show the presence and previously unknown diversity of T. pallidum from a Middle Holocene hunter-gatherer context in Sabana de Bogotá, Colombia, extending the genomic record of treponematoses by more than three millennia. TE1-3’s early divergence supports models in which treponemal pathogens had already diversified in the Americas during the Holocene. The availability of this ancient South American genome provides a rare calibration point for reconstructing the early diversification of T. pallidum subspecies. This study also underscores the value of metagenomic screening of ancient remains to uncover previously undetectable infectious disease histories in deep time. These findings open new questions about the timing, routes, and drivers of treponemal spread, and on the longstanding interplay among Treponema pallidum, human hosts, and the broader socioecological landscapes in which these diseases have evolved, persisted, and spread to affect human populations.