Persistence drives gene clustering in bacterial genomes

Background: Gene clustering plays an important role in the organization of the bacterial chromosome and several mechanisms have been proposed to explain its extent. However, the controversies raised about the validity of each of these mechanisms remind us that the cause of this gene organization remains an open question. Models proposed to explain clustering did not take into account the function of the gene products nor the likely presence or absence of a given gene in a genome. However, genomes harbor two very different categories of genes: those genes present in a majority of organisms - persistent genes - and those present in very few organisms rare genes. Results: We show that two classes of genes are significantly clustered in bacterial genomes: the highly persistent and the rare genes. The clustering of rare genes is readily explained by the selfish operon theory. Yet, genes persistently present in bacterial genomes are also clustered and we try to understand why. We propose a model accounting specifically for such clustering, and show that indispensability in a genome with frequent gene deletion and insertion leads to the transient clustering of these genes. The model describes how clusters are created via the gene flux that continuously introduces new genes while deleting others. We then test if known selective processes, such as co-transcription, physical interaction or functional neighborhood, account for the stabilization of these clusters. Conclusion: We show that the strong selective pressure acting on the function of persistent genes, in a permanent state of flux of genes in bacterial genomes, maintaining their size fairly constant, that drives persistent genes clustering. A further selective stabilization process might contribute to maintaining the clustering.