Genomic insights into habitat adaptation of Lactobacillus species

Abstract

Lactobacillus is one of the most important genera within the lactic acid bacteria group, due to its importance in the food industry and the health field. This diversity can be explained either by their radiation in different environments or by the domestication process in artificial habitats, such as fermented foods. In this study, we performed a comparative genomic analysis of 1020 Lactobacillus genomes, categorizing them into five broad habitats: insects, vertebrates (including humans and animals), vegetables, free-living environments, and dairy products. Utilizing phylogenetic relationships, genomic distances, and gene presence/absence profiles, we identified distinct clustering patterns associated with specific environmental adaptations. Notably, species within the Lactobacillus delbrueckii clade exhibited GC content variations fivefold greater than those observed in other bacterial genera, indicating significant genomic divergence. Insect-associated species showed a strong correlation between genes for carbohydrate utilization and those for amino acid biosynthesis across all habitats. However, individual gene analyses revealed no consistent correlation between habitat adaptation and phylogenetic proximity, suggesting that Lactobacillus employs strain-specific adaptive mechanisms rather than universal genetic markers. Notably, around 50% of the genes associated with specific habitats are hypothetical. Our findings highlight the genomic complexity of Lactobacillus, driven by diverse adaptive strategies, and underscore the need for more comprehensive sampling to fully elucidate the evolutionary dynamics within this important genus.

Keywords: Comparative genomics; Domestication; Habitat Adaptation; Lactobacillus; Pangenomics.

Fig. 1

a Maximum likelihood phylogenetic tree of the 43 reference genomes of the Lactobacillus genus. L. plantarum ATCC 8014 and L. pentosus ZFM94 were used as outgroup. The size and the %GC are represented from the inner to the outer circle. b Bar chart of the percentage of habitats in species of the genus Lactobacillus. We used ITol v5 for the visualization and modification of the tree (Letunic and Bork 2021b)

Fig. 2

COG category of genes with a positive statistical correlation between phenotype and genotype determined by Scoary with a cutoff of 95% specificity

Fig. 3

Bubble plot comparing key enzymes of carbohydrate metabolic utilization genes and amino acid biosynthetic pathways between species. The size of the bubble indicates the percentage presence of each enzyme within each species

Fig. 4

Principal Coordinate Analysis (PCoA) performed on Lactobacillus genomes to construct a Bray–Curtis dissimilarity matrix. The occurrence of taxonomic species in each genome is color-coded. For visualization purposes, data were transformed to the cubic root, based on a total CAZyme module counts and b total Glycoside Hydrolases (GHs). c Heat map of the most abundant carbohydrate-activated enzyme (CAZymes) modules found in Lactobacillus species on average per species. In parentheses, the number of genomes from each species is indicated

Fig. 5

a Comparative phylogeny between the PrtB-based Maximum Likelihood tree (left) and the complete genome-based maximum likelihood tree of L. delbueckii species (right). We use S. aureus ATCC 12600 as an outgroup. Visualization and tree modification were performed with ITol v5 (Letunic and Bork 2021b). The color in the tips indicates the different genus. b Maximum likelihood phylogenetic tree of L. delbueckii in the context of key species of the genus Lactobacillus, Streptococcus, Floricoccus, and Vagococcus (left), compared with the genomic context of proteinase PrtB (right). The genomic context was made with GeneSpy software (Garcia et al. 2019)