Codon bias, nucleotide selection, and genome size predict in situ bacterial growth rate and transcription in rewetted soil

Abstract

In soils, the first rain after a prolonged dry period greatly impacts soil microbial community function, yet we lack a full understanding of the genomic traits associated with the microbial response to rewetting. Genomic traits such as codon usage bias and genome size have been linked to bacterial growth in soils--however this is often through measurements in culture. Here, we used metagenome-assembled genomes in combination with metatranscriptomics and 18O- water stable isotope probing to track genomic traits associated with transcriptional activity and growth of soil microorganisms over the course of one week following rewetting of a grassland soil. We found that the codon bias in ribosomal protein genes was the strongest predictor of growth rate. We also observed higher growth rates in bacteria with smaller genomes, demonstrating that reduced genome size contributes to bacterial growth responses to sudden changes in water or nutrient availability--potentially explaining why smaller genomes are more prevalent in arid and carbon poor systems. High levels of codon bias corresponded to faster transcriptional upregulation of ribosomal protein genes. In early transcribing taxa, nucleotides requiring less energy to produce were more common at synonymous substitution sites--where nucleotide substitutions did not change the encoded amino acid. We found several of these relationships also existed within a phylum, suggesting that association between genomic traits and activity could be a generalized characteristic of soil bacteria. These results provide in situ evidence that following rewetting, certain genomic characteristics affect soil microbial growth rate and transcription, and points towards the fitness advantages that these traits might pose for bacteria under changing conditions in soil.