Two dimensions of chemical variation of the human microbiome across body sites and in COVID-19 patients (preprint)

ABSTRACT

A better understanding of dysbiosis is a major goal of human microbiome studies, but more knowledge about chemical effects on microbial communities is needed. Oxidation-reduction and hydration-dehydration reactions are chemical processes that are important for physiological functions and, it is hypothesized here, may also influence the elemental composition of microbial proteins. Chemical metrics of biomolecules relevant to these processes are carbon oxidation state (ZC) and stoichiometric hydration state (nH2O). I calculated these metrics for protein sequences derived from microbial genomes (multiplied by 16S rRNA-based taxonomic abundances to obtain community reference proteomes), shotgun metagenomes, and metaproteomes. Metaproteomes of gut communities are reduced (i.e., have lower ZC) compared to oral communities. In contrast, community reference proteomes have lower nH2O in gut compared to nasal, skin, and oral communities, and metagenomes for gut and oral communities exhibit the same trend. The chemical differences for metaproteomes may be explained by physiological adjustment of protein expression levels to anaerobic, reducing conditions in the gut, whereas metagenomes and reference proteomes may reflect evolutionary adaptation to dehydrating conditions brought on by intestinal absorption of water. Community reference proteomes, metagenome-assembled genomes (MAGs), and metaproteomes compiled from various studies yield a common trend of more reduced proteins in gut communities of COVID-19 patients compared to controls. These chemical differences imply more reducing conditions in the guts of COVID-19 patients, a finding that contrasts with oxidative conditions that have been previously associated with dysbiosis in inflammatory bowel disease and HIV infection. These results reveal how the human microbiome is shaped by multiple chemical factors over a range of timescales and suggest a new strategy for using multi-omics data to infer changes in gut redox conditions in COVID-19 patients.