Deep shifts in Evolutionary Rate Trajectories of Ancient Bacterial Genes
Reconstruction of ancestral gene repertoires from extant genomes captures only the genes that survived; those lost from the record are invisible. Among the genes that did persist, selective pressures change not only across lineages but across time. Here, we resolve evolutionary rate trajectories across 528 genes in the Last Bacterial Common Ancestor (LBCA). We hypothesized that LBCA genes would show distinct evolutionary rate trajectories across bacterial history and tested this by resolving normalized branch lengths across five calibrated taxonomic intervals (phylum, class, order, family, and genus), spanning approximately 2.1 billion years of bacterial diversification. The distribution of rate trajectories is continuous, but four clusters capture the major patterns: Decelerating, Class-Peaking, Constant, and Accelerating. The Decelerating and Class-Peaking clusters are composed predominantly of Genetic Information Processing genes, whereas the Constant and Accelerating clusters are enriched for Metabolic genes. Specifically, the Decelerating cluster is enriched for core informational machinery, including translation initiation factors and components of the expressome, the molecular complex physically coupling transcription and translation, suggesting that transcription-translation interfaces locked in early in bacterial history. Cofactor-dependence and cofactor biosynthesis are decoupled: biosynthetic pathways producing metallocofactors such as heme, molybdopterin, and cobalamin, concentrated in the Constant and Accelerating clusters but proportionally more proteins use inorganic cofactors in the Decelerating and Class-Peaking clusters. This offset coincides with the shift in metal bioavailability associated with the Great Oxidation Event and reveals genomic fingerprints of the co-evolution of bacterial metabolism with planetary geochemistry.