Beyond serology: saccharide profiling enables identification of antigenically similar and prompts re-evaluation of bacterial lipopolysaccharide evolution.

Leptospirosis is a zoonotic infectious disease of growing importance in both human and veterinary medicine. Gram-negative spirochetes of are traditionally classified into serovars based on their antigenic identity, which must be ascertained to design effective treatment procedures for humans and appropriate vaccination strategies in pets and livestock. Unfortunately, identifying serovars is challenging and currently requires access to a wide panel of reference strains, animal-derived antisera, or monoclonal antibodies. Here, we describe a new method for the identification of serovars that is based on monosaccharide composition analysis of the polysaccharide part of bacterial lipopolysaccharide (LPS) structures. Our approach requires no animal sacrifice and can be implemented in any laboratory equipped for chromatographic analysis. An LPS sugar fingerprint that is specific to each bacterial isolate that we studied can be generated. Importantly, sugar profiling of LPS enables distinguishing serovars that are antigenically very similar. Using our new approach, we discover that the LPS structures of two cattle pathogens belonging to two different species: and , and to one serovar: Hardjo, can be distinguished despite sharing major similarities. Through extensive phylogenetic analysis, we reveal which specific glycosyltransferases of the LPS biosynthesis locus likely drove the emergence of these similarities and identify a single glycosyltransferase that might have contributed to the formation of saccharide differences in the LPS structure. Our findings have implications for future work on the evolution of bacterial polysaccharide synthesis and highlight the importance of preventing horizontal gene transfer between pathogenic bacteria.