In vitro models to detect in vivo bile acid changes induced by antibiotics

Abstract

Bile acid homeostasis plays an important role in many biological activities through the bile–liver–gut axis. In this study, two in vitro models were applied to further elucidate the mode of action underlying reported in vivo bile acid changes induced by antibiotics (colistin sulfate, tobramycin, meropenem trihydrate, and doripenem hydrate). 16S rRNA analysis of rat fecal samples anaerobically incubated with these antibiotics showed that especially tobramycin induced changes in the gut microbiota. Furthermore, tobramycin was shown to inhibit the microbial deconjugation of taurocholic acid (TCA) and the transport of TCA over an in vitro Caco-2 cell layer used as a model to mimic intestinal bile acid reuptake. The effects induced by the antibiotics in the in vitro model systems provide novel and complementary insight explaining the effects of the antibiotics on microbiota and fecal bile acid levels upon 28-day in vivo treatment of rats. In particular, our results provide insight in the mode(s) of action underlying the increased levels of TCA in the feces upon tobramycin exposure. Altogether, the results of the present study provide a proof-of-principle on how in vitro models can be used to elucidate in vivo effects on bile acid homeostasis, and to obtain insight in the mode(s) of action underlying the effect of an antibiotic, in this case tobramycin, on bile acid homeostasis via effects on intestinal bile acid metabolism and reuptake.

Connecting Gut Microbial Diversity with Plasma Metabolome and Fecal Bile Acid Changes Induced by the Antibiotics Tobramycin and Colistin Sulfate

“…The accumulation of taurine-conjugated 1° BAs in the fecal samples of tobramycin-treated male and female rats shows that the deconjugation of these taurine-conjugated 1° BAs is reduced in these treatment groups. Moreover, the profound increase in TCA appears to be related not only to reduced deconjugation but is also associated with reduced passage through the intestinal wall mediated by downregulation of genes coding for BA transporters . Finally, the 1° BA fecal accumulation appears to be a robust characteristic response of microbiome disruptions, as it has been observed also for other antibiotics, such as clindamycin, lincomycin, and vancomycin, , and this has been seen in other situations such as in disease states like cancer, lupus, non-alcoholic fatty liver disease, cirrhosis, and many others …”

“…Moreover, the profound increase in TCA appears to be related not only to reduced deconjugation but is also associated with reduced passage through the intestinal wall mediated by downregulation of genes coding for BA transporters. 49 Finally, the 1°BA fecal accumulation appears to be a robust characteristic response of microbiome disruptions, as it has been observed also for other antibiotics, such as clindamycin, lincomycin, and vancomycin, 27,37 and this has been seen in other situations such as in disease states like cancer, lupus, non-alcoholic fatty liver disease, cirrhosis, and many others. 50 Given the strong effects of tobramycin on fecal 16S composition metabolic functions, it was interesting to find by coordination analysis of changes in the blood plasma metabolites that previously established plasma metabolites known to be associated with perturbations in the gut microbiome were altered, but that the fold-change levels were moderate.…”

“…Additionally, gut bacteria are known to deconjugate taurine- or glycine-conjugated 1° BAs and further dehydroxylate them to produce 2° BAs. This inhibitory effect was also observed in vitro, supporting that tobramycin significantly reduced those bacteria responsible for these reactions. Taurine- and glycine-conjugated 2° BAs are also reduced in the feces of both the sexes, demonstrating the attenuation of 2° BAs conjugation with taurine or glycine amino acid.…”

“…Any adverse or toxicological outcomes of the metabolome changes and their consequences for host health either with or without co-administration with other drugs are not known. To address this, further studies, need to be performed . Furthermore, taking into account the observed variations in control animals, the establishment of a baseline analysis of the interindividual variability in the intestinal microbiota is highly recommended to understand if changes are within biological variation or not.…”

“…To address this, further studies, need to be performed. 49 Furthermore, taking into account the observed variations in control animals, the establishment of a baseline analysis of the interindividual variability in the intestinal microbiota is highly recommended to understand if changes are within biological variation or not. To establish pertinent correlations between particular micro-organisms and metabolites, the number of antibiotics in the present study was too low to account for the number of variables (gut microbiome as well as the number of fecal and plasma metabolites) and was too small to derive correlations which would pass a statistical test.…”