Genomics of the pathogenic clostridia

The earliest DNA-based studies of the clostridia investigated the guanine + cytosine (G+C) content of DNA and measured hybridization as an indication of the relatedness of strains (1-3). Since those early “genome” studies, a range of other DNA-based typing methods, such as pulsed field gel electrophoresis, amplified fragment-length polymorphism analysis, 16S rRNA gene sequence comparisons, multilocus variable-number tandem-repeat analysis, and multilocus sequence typing (MLST), have been used to classify, group, and differentiate clostridial isolates (4). However, these methods have now been largely superseded. With the rapid advances in DNA sequencing technologies and the subsequent drop in the price and effort required to acquire whole-genome sequence (WGS) data, the comparative analysis of strains is now commonly based on core genome single-nucleotide polymorphisms (SNPs), gene content analysis, MLST on a large group of genes, or average nucleotide identity (ANI) across whole genomes. Whole-genome analysis is the tool of choice for comparing isolates and informs our understanding of the pathogenic mechanisms that are deployed by the clostridia, the epidemiology of disease outbreaks, and the evolutionary histories of the organisms. Analysis of WGS allows the identification of metabolic pathways or the lack thereof (e.g., amino acid metabolism and substrate degradation pathways) and thus indicates how clostridia are adapted to their lifestyles within the ecological niches that they occupy. The quality of WGS information available in sequence databases is variable, so some caution is needed in how it is interpreted when investigating genome structure and absolute gene content. Complete closed genomes are particularly valuable and can be used as scaffolds around which the more usual draft assemblies can be made. Like WGS data from all species, the clostridial WGS annotations need to undergo continual refinement as gene functions are experimentally confirmed and extended. Commonly, a quarter to a third of the open reading frames identified in genomes encode putative proteins of unknown function. Whole-genome analysis provides a much better tool for taxonomic classification than the traditional biochemical, phenotypic, and limited molecular characterization methods (e.g., 16S rRNA gene sequencing) that were previously applied.