Correlation of Metabolic Variables with the Number of ORFs in Human Pathogenic and Phylogenetically Related Non- or Less-Pathogenic Bacteria.

Abstract

To date, a few works have performed a correlation of metabolic variables in bacteria; however specific correlations with these variables have not been reported. In this work, we included 36 human pathogenic bacteria and 18 non- or less-pathogenic-related bacteria and obtained all metabolic variables, including enzymes, metabolic pathways, enzymatic steps and specific metabolic pathways, and enzymatic steps of particular metabolic processes, from a reliable metabolic database (KEGG). Then, we correlated the number of the open reading frames (ORF) with these variables and with the proportions of these variables, and we observed a negative correlation with the proportion of enzymes (r=-0.506, p<0.0001), metabolic pathways (r=-0.871, p<00.0001), enzymatic reactions (r=-0.749, p<00.0001), and with the proportions of central metabolism variables as well as a positive correlation with the proportions of multistep reactions (r=0.650, p<00.0001) and secondary metabolism variables. The proportion of multifunctional reactions (r: -0.114, p=0.41) and the proportion of enzymatic steps (r: -0.205, p=0.14) did not present a significant correlation. These correlations indicate that as the size of a genome (measured in the number of ORFs) increases, the proportion of genes that encode enzymes significantly diminishes (especially those related to central metabolism), suggesting that when essential metabolic pathways are complete, an increase in the number of ORFs does not require a similar increase in the metabolic pathways and enzymes, but only a slight increase is sufficient to cope with a large genome.