Chromosome formation in bacteria

Objective: The genome of bacteria is classically separated into the essential, stable and slow evolving chromosomes and the accessory, mobile and rapidly evolving plasmids. This paradigm is being questioned since the discovery of extra-chromosomal essential replicons (EER), be they "megaplasmids", "secondary chromosomes" or "chromids". These genomic elements are generally thought to be plasmidic in origin, based on their harboring essential genes, their having a GC content identical to that of the chromosome, and the structure of their replication origin. However, none of these criteria are universally applicable and the true nature of these replicons remains to be formally determined. Here we explore the relationships of chromosomes and plasmids with reference to their genetic information inheritance systems (GIIS), under the assumption that the inheritance of EERs is integrated to the cell cycle and highly constrained in contrast to that of standard plasmids. Methodology: We performed a global comparative genomic analysis including all bacterial complete genome sequences available in NCBI RefSeq. Using ACLAME and KEGG GIIS-related proteins as input, we first identified GIIS functional homologs from all bacterial genomes. These homologs were then clustered according to function and sequences homology. The relationships between all bacterial replicons were then investigated using the GIIS clusters as parameters and unsupervised analyses (projections and graphs, and clustering. Finally, identification of putative EERs and of trends in GIIS usage were performed using supervised analyses (classification, regression). Result: This strategy clearly discriminated between chromosomes and plasmids with respect to their GIIS usage, and revealed another class of genomic elements that corresponds to EERannotated replicons. They are characterized by a specific GIIS usage that testifies of the continuity of the genomic material in bacteria. Furthermore, whereas some are plasmidic in origin, others derive from the cleavage of an ancestral chromosome. Conclusion: Our study provides insights into the formation of "neo-chromosomes" and the emergence of multipartite genomes in bacteria, as well as clues about forces shaping the genome.