Comparative genomics of Lactobacillus curvatus enables prediction of traits relating to adaptation and strategies of assertiveness in sausage fermentation

Lactobacillus (L.) curvatus reaches high numbers in a variety of habitats, which suggests a high (genomic) diversity within this species. Empirically selected strains are used as starter cultures in sausage fermentation. Determinants for the assertiveness of a strain in this environment are assumed to be multifactorial. We used comparative genomics and in silico proteomics of 10 L. curvatus strains, which were representative of its genetic and physiological biodiversity, to possibly derive genetic determinants for strain or group specific assertiveness in sausage fermentation. Their genome sizes ranged from 1.7 Mb up to 2.0 Mb. The estimated pan- and core genomes were 3.0 Mb and 1.4 Mb, respectively. The accessory genome, GC-content and coding density revealed a significant genomic diversity within this species. Plasmids were found, which were either closely related or unique in several strains. Putative assertiveness determinants including CRISPR/Cas systems, prophages, bacteriocin production, or specific metabolic settings were detected. Such traits of the accessory genome could not be correlated with the source of isolation. Pathways, which previously have been predicted for a relation with adaptation to meat of L. sakei, are part of the core genome of L curvatus. Intraspecies differences in the accessory genome of L. curvatus comprise ribose metabolism, enzymes involved in nucleotide metabolism (nucleoside phosphorylases, phosphopentomutase, adenosine deaminase, ribose transporters), and tyrosine decarboxylases, ornithine decarboxylases. One group of the strains encoded a phosphotransferase system (PTS) as ribose transporter, whereas the second group encoded an ATP binding cassette (ABC) transporter. Analysis of the ribose uptake by HPLC analysis revealed different efficiencies of both transporter systems. Except for bacteriocin formation, no strain specific traits were identified predicting assertiveness of single strains. This fits our previous observation that single strains of L. curvatus could not override others in a competitive setting. Rather, pairs or sets of strains, comprising metabolically synergistic or non-assertive partner strains were able to dominate the fermentation. Indeed, this work suggests that assertive partnerships can be predicted along their complementary accessory genomes.