Bioinformatic and Mechanistic Studies of Radical Enzymes in Human and Animal Gut Microbiomes

Abstract

The gut microbiomes of animals and humans consist of the trillions of microorganisms co-existing within and outside the host. These niches represent some of the densest microbial communities on Earth, with the microbes collectively encoding 130 times more genes than their hosts. This genetic diversity enables members of the gut microbiome to produce numerous small molecules that impact host health and disease, greatly expanding the chemistry within the body. Despite the vast number of correlations between gut microbiome composition and host health and disease reported in the field, few causal links are understood at the mechanistic or enzymatic level. Such efforts are hampered by gaps in our knowledge of what microbial enzymes perform important transformations in the gut and a lack of understanding of the roles of these enzymes in complex communities. Here, we identify and characterize novel radical enzymes from gut microbiomes, focusing on members of the glycyl radical enzyme (GRE) and radical S-adenosyl-L methionine (rSAM) enzyme superfamilies. Both classes of enzymes use single electron strategies to catalyze challenging chemical reactions inaccessible through typical two-electron mechanisms. This expanded chemical repertoire in the gut can have large impacts on host health. Chapter 2 describes our efforts to identify and characterize indoleacetate decarboxylase (IAD), a putative GRE responsible for catalyzing decarboxylation of the L-tryptophan catabolite indole-3-acetate (I3A) into 3-methylindole (skatole). We explored two IAD homologs from the microbes Olsenella uli (Ou) and Clostridium scatologenes, the latter of which is distinguished by the involvement of a second rSAM enzyme. Measurement of Michaelis–Menten kinetics, enzyme structure prediction, and site-directed mutagenesis of putative active site residues are consistent with IAD being a new member of the GRE family. In Chapter 3 we explored the catalytic mechanism of OuIAD to determine if it employs either: (1) a Kolbe-type decarboxylation reaction involving initial electron transfer and proton transfer steps or (2) hydrogen-atom transfer (HAT) from the methylene carbon of I3A to generate an initial benzylic substrate-based radical. Observation of a kinetic isotope effect with deuterated I3A, incorporation of two deuteria into skatole when the reaction is run in D2O, and computational modeling by collaborators from the Kulik Lab (Massachusetts Institute of Technology) support the proposal that IAD uses a HAT-dependent decarboxylation mechanism. Understanding enzyme function can help guide efforts to understand the biological consequences of metabolic activities in microbiomes. Skatole production by gut microbes in swine is associated with boar taint, while skatole synthesis in ruminants causes the respiratory disease fog fever. Given the deleterious effects of skatole in these contexts, we sought to develop small molecule inhibitors to modulate IAD activity. Chapter 4 describes our efforts to synthesize and assay a panel of substrate analogs as inhibitors of skatole production. We demonstrate that a previously identified inhibitor of skatole production in complex microbiome samples, indole-3-acetonitrile, is a promising inhibitor of OuIAD in vitro. We propose that this compound engages with the active site Cys through nucleophilic addition. Further understanding of this mechanism could inspire additional inhibitor design efforts. Our bioinformatic investigation of metagenomics and metatranscriptomics datasets from humans, swine, and ruminants revealed that IAD genes and transcripts are present at relatively high abundances. These data suggest that IAD is biologically relevant in the context of gut microbiomes and may be a viable enzymatic target for inhibitor development. Finally, Chapter 5 details our broader bioinformatic examination of the GRE family and identification of non-canonical members that substitute the conserved catalytic Gly residue for an Ala, Ser, or Thr. With collaborators from the Broderick Lab (Montana State University) and Hoffman Lab (Northwestern University), we confirmed via electron paramagnetic resonance (EPR) spectroscopy that all ‘X’REs can harbor stable α-carbon radicals on their cognate residue and engineered Gly variants. Altogether, these studies expand the known radical chemistry of living systems and bring to light evolutionary questions on the emergence of XREs. Future studies will be focused on understanding the evolution of these XREs, identifying the chemical transformations catalyzed by XREs, and elucidating the mechanistic roles of these alternatively substituted α carbon radicals in XRE function.