Investigation into mediators of metabolic health: A role for the gut microbiota, bile acids and Angptl4

The global prevalence of obesity has increased substantially over the past decades. As a consequence, there has been a rise in obesity-related comorbidities, such as diabetes mellitus, non-alcoholic fatty liver disease (NAFLD), and dyslipidemia, each of which serves as an independent risk factor for cardiovascular diseases. While the increase in obesity rates and related metabolic diseases are believed to be primarily the result of an increased consumption of calorie-dense foods in combination with reduced levels of physical activity, various factors have been suggested to influence the progression of these metabolic diseases, including the gut microbiota, bile acids and angiopoietin-like protein 4 (ANGPTL4).

The gut microbiota live in a close relationship with the host and have an essential role in several aspects of host physiology, such as conferring protection against invading pathogens and facilitating the digestion of complex carbohydrates. The composition of the gut microbiota varies greatly between individuals, which is caused by differences in host genome and environmental factors, and may also be affected by specific disease states. Emerging evidence suggests that changes in intestinal microbiota are not merely a consequence of metabolic health, but may in fact contribute to certain diseases. To link changes in gut microbiota to metabolic health, several mechanisms have been proposed, including lipopolysaccharides, short-chain fatty acids, bile acids and ANGPTL4, which are extensively described in Chapter 2 and Chapter 3. Bile acids are primarily known for their role in lipid absorption. By emulsifying triglycerides and stimulating pancreatic lipase activity, bile acids facilitate the digestion of dietary lipids and promote their absorption, although it is becoming increasingly apparent that bile acids also have major regulatory roles in the control of lipid, glucose and energy metabolism. Another important mediator of metabolic health is ANGPTL4. ANGPTL4 is a potent inhibitor of lipoprotein lipase (LPL), and is active under various physiological conditions, including fasting, exercise and cold exposure, thereby ensuring the proper distribution of plasma triglycerides over different tissues. In addition to inhibiting LPL, ANGPTL4 is also able to inhibit intestinal pancreatic lipase, thereby limiting dietary lipid absorption. Next to its role in lipid metabolism, ANGPTL4 has also been suggested to influence glucose metabolism. Although it is well-established that the gut microbiota, bile acids and ANGPTL4 influence metabolic health, the exact mechanisms are far from being completely understood. Therefore, the main of this thesis is to increase our understanding regarding the role of these mediators in metabolic health.

The first part of this thesis focuses on the role of the gut microbiota in NAFLD. NAFLD describes a spectrum of liver diseases ranging from simple steatosis to non-alcoholic steatohepatitis, liver fibrosis and cirrhosis. In Chapter 4 we studied the influence of modulation of the gut microbiota in the development of NAFLD. To that end, in a mouse model of NAFLD, we stimulated the gut bacteria by feeding mice the highly fermentable fiber guar gum (GG) and suppressed the gut bacteria via oral administration of antibiotics. We found that stimulation of the gut microbiota using GG enhanced hepatic inflammation and fibrosis, concurrent with elevated plasma and hepatic bile acid levels. In contrast, suppression of the gut bacteria using antibiotics decreased portal secondary bile acid levels and attenuated hepatic inflammation and fibrosis. Together, these data suggest a causal link between disturbances in gut microbial community and NAFLD. To investigate if bile acids are causally involved in the progression of NAFLD upon GG feeding, we fed mice chow supplemented with the primary bile acid taurocholic acid. Provision of taurocholic acid raised plasma bile acid levels and stimulated hepatic inflammation and fibrosis. Taken together, our data indicate that the gut microbiota have a marked impact on NAFLD, possibly via changes in the portal delivery of bile acids. Accordingly, in Chapter 5, we hypothesized that targeting the gut microbiota may provide therapeutic tools to treat NAFLD. As impairments of the intestinal barrier are also suggested to mediate the effect of the gut microbiota on NAFLD, we administrated Akkermansia muciniphila and Bacteroides thetaiotaomicron to obese mice with NAFLD, which have both been suggested to improve the intestinal barrier. It was found that daily administration of A. muciniphila or B. thetaiotaomicron for 5 weeks to obese mice with NAFLD did not influence hepatic steatosis, inflammation or fibrosis. Importantly, we also found no difference in intestinal permeability as well as portal lipopolysaccharide levels. In conclusion, although we demonstrate that the gut microbiota have a marked impact on NAFLD, in our experimental setup, administration of A. muciniphila or B. thetaiotaomicron to obese mice with NAFLD did not improve features of NAFLD.

The liver receives about 70% of its blood supply from the intestine via the portal vein and is therefore the first organ that may be exposed to potentially harmful metabolites released by the intestinal microbiota. Identification of these harmful metabolites are very important as it might reveal novel therapeutic targets for NAFLD. To investigate if certain metabolites are harmful to the liver, it is important to utilize in vivo or ex vivo systems that reflect the human liver. In the second part of this thesis, we investigated whether precision-cut liver slices (PCLS) may be a suitable ex vivo model to study the human liver. Accordingly, in Chapter 6, we used whole genome gene expression profiling to compare the effect of activation of the nuclear receptor PPARα between human PCLS and primary human hepatocytes. We show that the induction of gene expression by PPARα is in general well captured by both the human primary hepatocytes and human PCLS, as activation of PPARα consistently upregulates genes involved in lipid and xenobiotic metabolism in both models. By contrast, downregulation of gene expression by PPARα activation was almost exclusively observed in human PCLS, which was largely connected to cells of the immune system, which are present in PCLS but absent in the primary hepatocytes. Taken together, we show that human PCLS are a suitable and superior model over primary hepatocytes to study PPARα activation in human liver. Subsequently, in Chapter 7, we show, for the first time, the response of the human liver to FXR activation by obeticholic acid (OCA), a promising drug for the treatment of non-alcoholic steatohepatitis that is now in phase 3 clinical trials. It was found that human PCLS are highly and consistently responsive to FXR activation. In addition, by comparing the response of human PCLS to wild-type and FXR-/- mice receiving obeticholic acid, we were also able to identify putative novel FXR target genes. Taken together, these results indicate that human PCLS are a suitable model to study human liver and represent a very useful tool for testing the response of human liver to intestinal-derived metabolites.

In the third part of this thesis, we focused on the metabolic effects of ANGPTL4. ANGPTL4 raises plasma triglyceride levels by inhibiting the enzyme lipoprotein lipase. Interestingly, bile acids and bile acids resins also have been shown to influence plasma triglyceride levels. Because we previously observed that bile acids lower ANGPTL4 secretion by intestinal cells, in Chapter 8, we hypothesized that the triglyceride-lowering effect of bile acids is mediated by ANGPTL4. To test this hypothesis, wild-type and Angptl4-/- mice were fed chow supplemented with the primary bile acid taurocholic acid. It was found that the triglyceride-lowering effect of bile acids was not mediated by ANGPTL4, but may be mediated by downregulation of the LPL inhibitors ANGPTL3 and APOC3. Intriguingly, plasma and hepatic BA concentrations were significantly lower in TCA-supplemented Angptl4-/- mice than in TCA-supplemented wild-type mice, suggesting that loss of ANGPTL4 impairs intestinal bile acid absorption. Since the gut microbiota converts primary bile acids into secondary bile acids, we hypothesized that the elevated excretion of primary bile acids and decreased bile acid absorption in Angptl4-/- mice may be dependent on the gut microbiota. Indeed, suppression of the gut bacteria using antibiotics abolished the differences in bile acid absorption. In conclusion, our data indicate that ANGPTL4 is not involved in the triglyceride-lowering effect of bile acids but promotes bile acid absorption during taurocholic acid supplementation via a mechanism dependent on the gut microbiota.

The observation that Angptl4-/- mice develop a massive acute phase response, lethal chylous ascites and peritonitis when fed a high fat diet rich in saturated fat, has hampered the study of ANGPTL4 in diet-induced obesity and related metabolic dysfunction. To overcome that obstacle, in Chapter 9, we investigated the influence of ANGPTL4 on diet-induced obesity and metabolic dysfunction by feeding wild-type and Angptl4-/- mice a high fat diet rich in unsaturated fat, combined with fructose and cholesterol. It was found that Angptl4-/- mice displayed increased bodyweight gain, adipose tissue mass and adipose tissue inflammation, but unexpectedly had markedly improved glucose tolerance, which was accompanied by elevated plasma insulin levels. Inasmuch as the gut microbiota has been suggested to influence insulin secretion and as ANGPTL4 has been proposed to link the gut microbiota to host metabolism, we hypothesized a potential role for the gut bacteria. Indeed, the gut bacterial composition was profoundly different between wild-type and Angptl4-/- mice. Furthermore, suppression of the gut bacteria using antibiotics largely abolished the differences in glucose tolerance and insulin levels between wild-type and Angptl4-/- mice. Taken together, loss of ANGPTL4 improves glucose tolerance at least partly via changes in the gut microbiota. Nevertheless, future studies are warranted to investigate how ANGPTL4 alters the gut bacterial composition and how this may influence insulin levels and glucose tolerance.

In summary, the studies presented in this thesis have clarified and strengthened our understanding regarding the role of the gut microbiota, bile acids and ANGPTL4 in metabolic health. We showed that the gut microbiota have an important role in the pathogenesis of NAFLD, which is likely mediated via changes in portal delivery of bile acids. Subsequently, we demonstrated that human PCLS provide a suitable model to study the human liver and provide a pivotal tool for testing the effect of intestinal metabolites on human liver. Finally, we show that the triglyceride-lowering effects of bile acids is not mediated by ANGPTL4 and reveal two novel functions of ANGPTL4 in mediating bile acids absorption and glucose metabolism. Although our studies extended our knowledge of the role of these mediators in metabolic diseases, several questions remain. Since the intestinal microbiota, bile acids and ANGPTL4 are all interconnected and are all involved in lipid and glucose metabolism, it is important to realize that modulation of one of these three mediators likely also affects the other two mediators, including their metabolic responses, reflecting a strong interdependence.