LUMINAL BILE SALT HYDROLASE ACTIVITY IMPROVES HOST GLUCOSE HOMEOSTASIS

BACKGROUND: Alterations in the gut microbiome have significant implications for metabolism and can modulate gene expression in different host tissues, thereby contributing to the occurrence of insulin resistance and type 2 diabetes. However, the contributions of specific bacterial functions in glucose homeostasis of the host are poorly understood. We previously demonstrated that time-restricted feeding is associated with increased luminal bile salt hydrolase (BSH) activity, which may be a mechanism by which the microbiome improves host glucose homeostasis. Moreover, we demonstrated that we can functionally manipulate the gut microbiome using engineered native bacteria, a novel technique for introducing transgenes of interest to perform mechanistic microbiome/host studies. By using this method, we tested the hypothesis that bile acid deconjugation can exert beneficial effects on glucose homeostasis.

METHODS: Conventionally raised C57BL/6 mice were gavaged with vehicle (PBS), engineered native E. coli with BSH transgene (EcAZ-2^BSH) or without (EcAZ-2) (n = 8-9 per group in two separate cohorts) and kept on a normal chow diet. Colonization and BSH-expressing function of gavaged bacteria in vivo were confirmed by the plating of stool samples. After the single gavage, food intake and body weight of mice were measured every week; body composition was monitored at week 1, 3, 7 and 10; fasted and postprandial glucose and insulin levels were tested weekly; oral glucose tolerance and insulin tolerance tests were also performed to assess the glucose homeostasis. In the end, mice were sacrificed to harvest portal vein blood for the hormonal level analysis, cecal contents for the analysis of metabolome and metagenome, and liver and distal ileum for the transcriptome analysis.

RESULTS: EcAZ-2 and EcAZ-2^BSH colonized stably in conventionally raised mice and the BSH function of the bacteria was retained during the entire time of engraftment. The engineered bacteria treatment didn’t affect the body weight, fat mass, lean mass and food intake. However, EcAZ-2^BSH significantly decreased the postprandial insulin levels without changing the postprandial glucose levels within one month after treatment and for the entire length of the experiment after the initial single gavage (EcAZ-2^BSH vs EcAZ-2, P = 0.018, 0.043, 0.025, 0.033 and 0.042 on week 5, 6, 7, 8 and 9 respectively). These metabolic changes occurred in the setting of altered luminal bile acid pool and host gene expression.

CONCLUSIONS: A single treatment with engineered bacteria overexpressing the BSH function increased insulin sensitivity and improved host glucose homeostasis, and this occurred in the setting of altered luminal bile acid pool, indicating engineered native bacteria are a promising technique to understand the role of the microbe-host interaction in glucose homeostasis.