Gut Reactions: An ADME-Centric View of the Gut Microbiome; Part II

Gut Reactions, Part II

Peter Spanogiannopoulos’s (UCSF) title was “A Prevalent Operon from the Human Gut Microbiome is Responsible for the Inactivation of Fluoropyrimidine Anticancer Drugs.” Host-targeted drugs impact the growth of gut bacteria and the composition of the gut microbiome. 5-Fluorouracil (5FU) is a thymidylate synthase inhibitor and anticancer drug that acts by interfering with pyrimidine (thymidine) synthesis. 5FU is inactivated by reduction to dihydrofluorouracil (DHFU). 5FU has antibacterial activity by the same mechanism as the anti-cancer activity in the host (pyrimidine metabolism is highly conserved from bacteria through mammals). Closely related bacteria have very different susceptibility to 5-FU; one mechanism of 5FU resistance is inactivation of the drug. Proteobacteria (e.g., E. coli, Salmonella enterica) and Firmicutes can inactivate 5FU to DHFU. It was demonstrated conclusively and convincingly that a specific E. coli operon is responsible for 5FU inactivation, but I will not go into detail because the work is unpublished. This is just one of many examples of unintended consequences involving gut bacteria…an anticancer drug that kills some gut bacteria and is inactivated by others (thereby shifting the balance toward the latter), which in turn impacts its efficacy.

Joe Dempsey’s (University of Washington) title was “A Multi-omic Approach to Understand the Development of the Gut-Liver Axis.” The neonatal human gut microbiome is relatively aerobic and high in Lactobacillus, Difidobacterium, Staphylococcus, and Enterococcus; in adults it is more diverse, anaerobic, and high in Firmicutes (with increases in Proteobacteria and Bacteroidetes after the age of 70). Multi-omic approach: genomics, metabolomics, transcriptomics, and proteomics. Age-related changes in the gut microbiome modulate the metabolome within the gut-liver axis, altering the expression and function of xenobiotic-processing genes. Primary bile acids are synthesized (including conjugation with taurine or glycine) in the liver, and deconjugated and converted to secondary bile acids in the gut, and age-related changes in bile acid composition correlate with changes in the function and composition of the gut microbiome. PXR in the host is activated by the secondary bile acid lithocholic acid (LCA), which is a product of the gut microbiome. Not surprisingly, the expression and activity of Cyp3a11 are much lower in germ-free mice (no gut bacteria, no LCA) than in conventional mice. Products of host metabolism (bile acids) impact the gut bacteria (see David Shen’s talk), and products of gut bacterial metabolism (LCA and others) impact the host (PXR activation, leading to induction of CYPs, etc.). It’s a recurring theme, and it can no longer be ignored. Based on work from this lab (Julia Cui’s group), I’m pretty sure that variations in gut bacteria account for the tremendous variation in CYP3A4 activity among humans.

David Shen’s (Penn) title was “Physiologic Implications of Co-metabolism Between the Gut Microbiome and its Host.” The host produces urea and primary bile acids, which are metabolized to ammonia and secondary bile acids, respectively, by gut bacteria. Bacterial products of bile acid metabolism impact the host, for example, as FXR agonists, and FXR activation, in turn, modulates the small intestinal microbiome. Bile acids are toxic to gram-positive bacteria. Obeticholic acid (OCA), derived from chenodeoxycholic acid (CDCA) is a potent, selective FXR agonist. In clinical trials of OCA in primary biliary cholangitis, increases in Gram-positive bacteria (due to decreased bile acids) were seen in the stool. This bacterial taxonomic signature in response to OCA in humans may be due to changes in the small intestine (especially the proximal small intestine), which has a different population of bacteria than the large intestine. Manipulation of gut bacteria to treat disease: bacteria in the colon convert urea to ammonia via urease, which is not expressed by mammals. In a properly prepared mouse host, inoculation with a urease-deficient consortium of bacteria results in long-term reduction in fecal ammonia, and reduces morbidity and mortality in a mouse model of liver disease. Another example of bilateral gut bacteria↔host interactions. It also makes me wonder about the impact of bacteria in the small intestine (as opposed to the large intestine, which is where we normally think of gut bacteria because they are more abundant there) on drug metabolism.

Julia Cui’s (University of Washington) title was “Gut-Liver Axis and Environmental Chemical Exposures.” Polybrominated diphenyl ethers (PBDEs) are fat-soluble, highly persistent, and bioaccumulative. Very high levels of PBDEs are found in human breast milk and in the blood of adult humans. Neonatal exposure to PBDEs is associated with an increased risk of disease (hypothyroid, low IQ, reproductive tox, hepatotox, cancer, diabetes) later in life. The gut microbiome may inactivate PBDEs (conventional vs. germ-free mice). Persistent environmental chemicals such as bisphenol A (BPA), PBDEs, and polychlorinated biphenyls (PCBs) alter the gut microbiome with corresponding changes in the levels of bacterial metabolites such as succinate, acetate, and lactate in the liver, resulting in persistent epigenetic and transcriptomic changes in adult mice that were exposed as neonates. The results suggest that the hepatic reprogramming amounts to microbial metabolite-mediated epigenetic regulation of the host. More on the theme of bilateral gut bacteria↔host interactions.

Seth Walk’s (Montana State University) title was “Microbiome Determinants of Arsenic Toxicity,” and he also presented some recently published results on gut bacterial residency. Arsenic in drinking water is a major global health threat (ranked #1 on the Toxic Substance and Disease Registry since 1997). Gut microbiome has a highly beneficial effect with regard to arsenic toxicity, due to expression of arsenic (As, +3 oxidation state) methyltransferase (AS3MT). With a small set of human stool donors, there was a correlation between bacterial species and protection vs. arsenic toxicity in mice, and the effect of individual bacterial species is detectable in the hypersensitive AS3MT knockout mouse model. Note: some bacteria cannot colonize the gut of germ-free mice by themselves and require a partner such as E. coli. Bioaccumulation of arsenic in gut bacteria appears to reduce toxicity. Gut bacterial residency: The stability of resident gut bacteria depends on how you look at it. In the human gut, some bacterial clones are extremely stable, others are changing all the time. The E. coli genome is highly diverse; only 20% of genes are present in all strains. The resolution of 16S RNA is low; therefore, at the level of 16S RNA sequencing (relevant to higher taxonomic levels), resident gut bacteria appear to be extremely stable. However, at the clonal level of one common gut bacterial family, there is very little stability. I have to admit that the terminology and computational methods used to evaluate gut bacterial residency is well outside of my comfort zone. But one conclusion is that you shouldn’t have to take probiotics every day, as the manufacturers recommend…if you do, they’re not working the way they’re supposed to.