Genomics in the taxonomy and evolution in Lactobacillus and related genera

Bacterial taxonomy is a fundamental discipline in microbiology since it provides an operative and predictive classification system, and constitutes a framework for application (i.e., reliable identification of microorganisms), scientific communication and knowledge exchange. According to the prokaryotic species definition, DNA-DNA hybridization and analysis of 16S rRNA gene sequence are currently being considered the golden standard molecular criteria for prokaryotic species delineation. However, the recent availability of complete sequence of prokaryotic genomes allows, for the first time, the fulfill investigation of the genomic relatedness between organisms and gives the opportunity to reconstruct events of genome evolution and understand the natural relationships between microorganisms. In the present research project the role of genomics was investigated for the taxonomy and evolution of Lactobacillus genus and related genera in different taxonomic ranks, through the application of several techniques. The interest in these bacteria derives from the presence of the highest number of GRAS species (Generally Recognised As Safe) and for their role in fermented food production. Moreover, certain strains of human origins are being exploited as probiotics or vaccine carriers. In the first part of the dissertation, the taxonomic analysis of Lactobacillus genus based on the comparative analysis on 16S rRNA gene sequence confirmed the heterogeneity of this genus, which was intermixed with Pediococcus genus, and characterized by a complex evolutionary history. Furthermore, the majority of Lactobacillus species was found to be included in 15 robust subgeneric groups, but species belonging to the same cluster did not always share the same phenotypic properties, especially data coming from the analysis of carbohydrate metabolism, upon which traditional taxonomic analysis is based. Finally, two species, Lactobacillus catenaformis and L. vitulinus were found to be part of to Clostridium subphylum cluster XVII, therefore they were reclassified as Eggerthia catenaformis and Kandleria vitulina, respectively. The presence of consistent phylogenetic groups and the discrepancy between metabolic and phylogenetic data constituted the appropriate taxonomic framework for genome data integration, which was applied in the second part of the present study. The genetic background of the metabolic pathways related to carbohydrate metabolism, i.e. glycolysis and pentose phosphate pathway, was analysed in 42 strains belonging to Lactobacillus and related genera of order Lactobacillales: the investigation of gene order and distribution provided new information concerning the genetic bases of heterofermentative/homofermentative metabolism in lactic acid bacteria and lineage specific trends of gene loss/gain within the two metabolic pathways, in particular for Lactobacillus delbrueckii-group. The detection of peculiar glycolytic gene distribution in Lactobacillus species belonging to different metabolic groups in terms of presence/absence of five putative markers (pdhA, pdhB, pdhC, pdhD and pfkA) motivated the taxonomic validation of these data on the type strains of species belonging to L. delbrueckii-group, as the first step towards a genome-based taxonomy of Lactobacillus. A coherent genetic background (absence of pdhA, pdhB, pdhC, pdhD, and presence of pfkA) was detected for species of L. delbrueckii-group despite their metabolic heterogeneity. These data represent a key-step for the resolution of Lactobacillus taxonomy, since a matching between the genotypic background related to carbohydrate metabolism and phylogenetic data was found for L. delbrueckii-group, denoted here as Lactobacillus sensu stricto, due to its diversity from the rest of lactobacilli, as it was revealed from comparative genomic analysis. Furthermore these results demonstrated that the proposed approach, based on comparative genomics and evolutionary analysis and then taxonomic validation on the type strains, can represent a successful modus operandi to resolve taxonomic issues and for the correct characterization of taxa. In the last part of the study, a single sequenced genome was shown to constitute a fundamental resource for the development of a Multilocus Seqeunce Typing (MLST) scheme, thus demonstrating that the availability of a single sequenced strain (even if a non type-strain) is a valuable support for the investigation of the intraspecific analysis of a panel of strains and a genome-based indication for the sequencing of other strains. Another important outcome of this project is the methodological contribution derived from genome data, which were integrated with different combinations involving molecular phylogeny, analysis of pathway gene content and order, split decomposition and MLST. In conclusion, the taxonomic analysis of Lactobacillus genus and related genera has shown that genome data integration is a powerful support for the creation of a tri-dimensional description of taxa, which creates a link between current taxonomic criteria and harmonizes genotypic, phylogenetic and phenotypic data.