Genomic diversity and host-specificity in Corynebacterium pseudotuberculosis using comparative population genomics

Corynebacterium <i>pseudotuberculosis</i> is a facultative intracellular pathogen responsible for chronic infections in livestock, primarily small ruminants and horses, with occasional zoonotic transmission. To investigate the genomic diversity, evolutionary stability, and host adaptation of this species, we analyzed 788 high-quality genomes representing isolates from diverse hosts, geographic regions, and time periods. Comparative population genomics revealed remarkably conserved genome architecture, supporting a closed pangenome with minimal accessory gene variation. Virulence and antimicrobial resistance (AMR) screening across multiple databases confirmed the universal presence of phospholipase D (pld) and the absence of major horizontally acquired AMR determinants, except for APH(3')-IIa, TEM-116, and APH(3')-IIIa in a few goat isolates from Brazil. Distinct metabolic features between biovars were conserved, notably nitrate reduction and molybdenum cofactor biosynthesis in biovar equi. However, gene presence/absence alone did not explain host specificity. Instead, machine learning applied to 8,028 core-genome SNPs identified allelic variants associated with host origin, particularly in genes linked to amino-acid biosynthesis and peptide transport (Opp system). These findings demonstrate that host adaptation in this species is driven by fine-scale SNP variation within core metabolic pathways, rather than acquisition of classical virulence or resistance genes, highlighting the species' exceptional genomic stability and narrow evolutionary flexibility.