AlphaFold-based structural analysis of the Treponema pallidum proteome provides a clearer prediction of the organism’s biological and pathogenic attributes

Treponema pallidum subsp. pallidum ( TPA) , an obligate bacterial pathogen that causes the sexually transmitted disease syphilis, is an extreme auxotroph with a minimal genome and distinct parasitic lifestyle. Approximately 17% of its 977 annotated proteins lack a predicted function because they show little sequence similarity to proteins of known function from other organisms. We therefore utilized the robust structural modeling and matching capabilities of AlphaFold and FoldSeek to create a whole-proteome atlas for the Nichols type strain of this bacterium ( TPANIC ), with the goal of providing structure-based functional assignments for uncharacterized proteins and verifying the accuracy of annotations previously generated by protein sequence homology analysis alone. Using a combination of precomputed structures from the AlphaFold Database and newly generated AlphaFold 3 structural predictions, we produced high-confidence predicted structural models for approximately 92% of the TPANIC proteome, enabling us to identify structural homologs and assign potential functional annotations to 24 previously uncharacterized proteins. Beyond functional annotation, the updated genome-derived structural models revealed several major structural revisions resulting from corrected gene models. Most notably, integration of comparative structural analyses and AlphaFold-Multimer modeling redefined the composition of the TPA outer membrane biogenesis machinery, supporting the existence of a conserved spirochete-specific BAM-TAM complex containing BamA, TamB, BamD, and BamB. We anticipate that the resulting AlphaFold protein structure atlas, in combination with the recently developed tools of in vitro cultivation and genetic manipulation, will help elucidate the roles of TPANIC proteins in the organism’s biology, pathogenesis, and host immune interactions. IMPORTANCE Incidence of syphilis has been on the rise in the past decade, emphasizing the need to understand the mechanisms behind the persistence and virulence of its causative agent, Treponema pallidum . In this study, we employed a next-generation artificial intelligence-based protein structure modeling and matching pipeline, building upon prior structural conservation studies by expanding the functional predictions for hypothetical proteins. Consequently, this study provides a clearer picture of the metabolic and pathogenic capabilities of this elusive pathogen, while highlighting the intricacies of its regulatory landscape. We anticipate that our strategy will complement existing sequence-homology-based gene/genome annotation pipelines, assisting in both protein function validation and the identification of orthologous proteins lacking sequence homology.