Peripheral and central regulation of insulin by the intestine and microbiome

Abstract

Blood glucose and insulin homeostasis is disrupted during the progression of type 2 diabetes. Insulin levels and action are regulated by both peripheral and central responses that involve the intestine and microbiome. The intestine and its microbiota process nutrients and generate molecules that influence blood glucose and insulin. Peripheral insulin regulation is regulated by gut-segment-dependent nutrient sensing and microbial factors such as short-chain fatty acids and bile acids that engage G-protein-coupled receptors. Innate immune sensing of gut-derived bacterial cell wall components and lipopolysaccharides also alter insulin homeostasis. These bacterial metabolites and postbiotics influence insulin secretion and insulin clearance in part by altering endocrine responses such as glucagon-like peptide-1. Gut-derived bacterial factors can promote inflammation and insulin resistance, but other postbiotics can be insulin sensitizers. In parallel, activation of small intestinal sirtuin 1 increases insulin sensitivity by reversing high fat-induced hypothalamic insulin resistance through a gut-brain neuronal axis, whereas high fat-feeding alters small intestinal microbiome and increases taurochenodeoxycholic acid in the plasma and the dorsal vagal complex to induce insulin resistance. In summary, emerging evidence indicates that intestinal molecular signaling involving nutrient sensing and the host-microbe symbiosis alters insulin homeostasis and action. Gut-derived host endocrine and paracrine factors as well as microbial metabolites act on the liver, pancreas, and the brain, and in parallel on the gut-brain neuronal axis. Understanding common nodes of peripheral and central insulin homeostasis and action may reveal new ways to target the intestinal host-microbe relationship in obesity, metabolic disease, and type 2 diabetes.

INTRODUCTION

An influx of nutrients from the intestinal lumen into the circulation after a meal increases insulin secretion from the pancreatic β-cells. A rise in plasma insulin levels lowers blood glucose levels by increasing glucose uptake into peripheral tissues such as muscle and fat and lowers glucose production by the liver. This well-defined insulin response maintains blood glucose, but the control of insulin homeostasis and action is disrupted during the progression of type 2 diabetes (T2D). For example, an excess of calorie intake and/or obesity promotes ineffective glucose lowering by a specific level of blood insulin, which is often termed insulin resistance (1). Blood glucose control is primarily balanced by insulin levels and action, and insulin resistance is linked to hyperinsulinemia. Hyperinsulinemia can promote obesity and be both a cause and consequence of insulin resistance (2, 3). Higher insulin levels can act as a trigger or compensate for insulin resistance, and this integrated response is often able to delay or prevent the development of overt diabetes, which is characterized by increased blood glucose. Thus, several classes of therapeutics in obesity, prediabetes, and T2D aim to enhance insulin action and secretion to maintain or improve blood glucose homeostasis.

Since the discovery of insulin, a major research focus has been to characterize pancreatic insulin secretion and insulin action/signaling cascades in the liver, muscle, and fat that regulate glucose and lipid metabolism in healthy, obese, and/or diabetic conditions (4–6). Nutrient-stimulation of pancreatic beta cells increases blood insulin, which reaches the liver, muscle, and fat and these insulin-sensitive organs impact nutrient metabolism and blood glucose control. It is well established that the intestine regulates peripheral insulin levels and action and blood glucose levels. The intestine regulates glucose absorption, barrier function, and endocrine communication with the pancreas through incretins. In parallel to the investigation of peripheral insulin control and action, a seminal work described central control of insulin action. A stand-alone and landmark study published in 1979 (7) reported that intracerebroventricular administration of insulin in baboons affects energy balance by lowering food intake and body weight, and thereby preliminary extending the biological function of insulin to the regulation of energy balance by the brain. Together with the discovery of leptin that was documented in 1994, an influx of literature in the past 20 years has reported that both insulin and leptin activate their respective receptors in various regions of the brain to regulate feeding and nutrient metabolism (8–17). Although these studies are mostly conducted in mice and rats owing to feasibility issues, emerging studies conducted in humans have reported that intranasal delivery of insulin or leptin increases peripheral insulin sensitivity and regulates glucose metabolism through central-dependent mechanisms (18–21). The purpose of this perspective is to highlight the role of the intestine in peripheral and central control of insulin homeostasis and action (Fig. 1). We aim to provide historical context and potential future directions related to how the intestine can influence insulin levels and action, and the connection to blood glucose control in healthy condition as well as in progression to T2D.

Figure 1.

The intestine and microbiota contribute to peripheral and central insulin homeostasis and action. Mediobasal hypothalamus (MBH), dorsal vagal complex (DVC), taurochenodexycholic acids (TCDCA), NAD+-dependent deacetylase sirtuin 1 (Sirt1), and glucagon-like peptide 1 (GLP-1).

PERIPHERAL REGULATION OF INSULIN

It is well established that the intestine participates in the control of blood insulin responses. The intestine is the site of nutrient absorption. Food constituents, intestinal transit time, and intestinal transport of nutrients can all influence the magnitude and duration of postprandial blood glucose responses and consequent pancreatic insulin secretion (22). Detection of nutrients such as glucose in the gut lumen also evokes endocrine response that signals to the pancreas to promote insulin secretion. For example, sensing of glucose in the small intestine promotes enteroendocrine l-cell-mediated glucagon-like peptide-1 (GLP-1) secretion, which can augment glucose-stimulated insulin secretion from pancreatic beta cells (23). Oral xenobiotics can also influence blood glucose and insulin responses (24, 25). Historically, there has been a focus on how intestinal mediators alter insulin secretion. Blood insulin homeostasis is also controlled by insulin clearance and first-pass hepatic insulin clearance removes over half of the insulin that is secreted into the portal circulation (26). Seminal work showed that insulin has a specialized blood clearance mechanism involving carcinoembryonic antigen-related cell adhesion molecule (Ceacam-1) and insulin-degrading enzyme (IDE) (27, 28). Insulin receptor-mediated sequestration of extracellular insulin is also positioned to regulate blood insulin clearance (29). The intestine can absorb insulin and alter insulin clearance (30). For example, GLP-1 lowers insulin clearance (31). Hence an integrated response of increased insulin secretion and lower insulin clearance is relayed from gut hormones to raise blood insulin levels (Fig. 1).

The intestine harbors many bacteria and changes in these microbial communities, microbial metabolites and the host-microbe relationship are positioned to influence how the gut alters insulin homeostasis and action. Akin to the bidirectional relationship between blood glucose and gut microbes, the symbiotic relationships between intestinal microbiota and host hormones can influence insulin levels and action (32). Intestinal microbes help process dietary carbohydrates, which can influence the glycemic effect of food and consequent insulin responses (33). Gut microbes participate in an integrated response that is influenced by host genetic and environmental factors such as diet, which combine to modify blood glucose and pancreatic β-cell insulin secretion (34, 35). The exact microbial mediators and mechanisms that link features of different gut microbial communities to changes in insulin secretion are not yet known. However, there are two intriguing possibilities the link host-microbe symbiosis to insulin homeostasis. It is possible that microbes or their components penetrate the gut barrier and interact with receptors within the pancreas or insulin-responsive tissues. There is evidence for gut-derived bacterial cell wall components (i.e., postbiotics) such as muropeptides altering insulin secretion, whole-body insulin sensitivity, and cell autonomous insulin action (36–39). Diet-induced obesity can increase circulating muropeptides in mice. The type of muropeptide is critical, as those that activate Nod1 promote lipolysis and insulin resistance, where Nod2 ligands can promote insulin sensitivity (40–42). The ability of Nod1-activating muropeptides to promote inflammation in multiple metabolic tissues that control blood glucose is one way that Nod1 influences immunometabolism differently from Nod2 (37, 38). Further, Nod1 can synergize with other bacterial components such as lipopolysaccharides (LPSs) to promote profound inflammation, whereas Nod2 can tolerize immune response when combined with LPSs (37, 41). This is relevant to metabolic disease because LPSs derived from the gut microbiota can also penetrate the gut barrier during metabolic endotoxemia (43, 44). LPSs engage Toll-like receptor 4, and this innate immune response promotes inflammation in metabolic tissue that controls blood glucose. Even low doses of LPSs can compromise endocrine control of metabolism and promote hepatic and peripheral insulin resistance, thereby worsening blood glucose control (43). LPSs can also increase GLP-1 and insulin secretion (45, 46). It is possible that a feature of T2D such as hyperglycemia alters gut barrier function and promotes tissue-specific colonization of microbes or their metabolites (47, 48). Compartmentalization of microbes in different tissues and gut segments is an important consideration and site-specific changes in immune responses and microbial defenses should be considered in the regulation of insulin beyond “chronic inflammation” during obesity or T2D (49–51). Gut microbes can also modify insulin clearance. It is not yet clear how specific microbial components or tissue-specific microbial colonization influence insulin clearance. It is known that a small cluster of five related bacteria taxa explained most of the microbe influence insulin clearance during diet-induced obesity in mice (52). It was found that Enterococcaceae, Clostridiaceae, and Peptostreptococcaceae accounted for >90% of microbe-transmissible impaired insulin clearance and lowered levels of Ceacam-1 in the liver (52) (Fig. 1). Future work should focus on identifying how specific strains or molecules derived from gut bacteria alter insulin clearance, especially in the liver, which contains bacterial innate immune sensors that regulate lipid and glucose metabolism and gut dysbiosis (53).

The intestinal host-microbe relationship can influence insulin through changes in gut hormones. Again, compartmentalization of host-microbe response is important because microbes participate in the gut segment-dependent effects of incretins. Glucose is a key signal for l-cell-mediated GLP-1 release via a sodium glucose cotransporter 1-dependent pathway in the upper/small intestine and consequent changes in insulin secretion (54). Microbial metabolites such as short chain fatty acids (SCFAs) and bile acids interact with G-protein-coupled receptors and G-protein-coupled bile acid receptors and participate in upper/small intestine GLP-1 secretion and the consequent effect on insulin secretion (54). However, in the colon, these same microbial metabolites appear to regulate l-cell-mediated GLP-1 release and intestinal transit time, independent of glucose sensing (54, 55). A key future goal is to define specific mediators in each gut segment that contribute to insulin and glucose control and whether they depend on the host-microbe symbiosis. It is not yet clear if diet or pre-existing features of obesity and metabolic syndrome are the key drivers of intestinal changes that alter blood insulin. There is evidence that ingestion of specific dietary components is sufficient to coax microbes to generate endocrine responses that alter insulin homeostasis. For example, oligofructose and inulin fiber supplementation stimulates GLP-1 and peptide tyrosine tyrosine (PYY) in the ileum and colon, where the preclinical and clinical evidence has been expertly reviewed (56). Intestinal microbiota fermentation of a diet supplemented with inulin produces short-chain fatty acids, which act on the free fatty acid receptor 2 to increase the density of enteroendocrine cells that secrete PYY (57). Consistently, short-chain fatty acids activate free fatty acid receptor 2 in the ileum to regulate glucose homeostasis via a GLP-1-dependent pathway, whereas the ability of short-chain fatty acids to activate free fatty acid receptor 2 in the colon and secrete GLP-1 is also blunted by the activation of the farnesoid X receptor (58, 59). Finally, the ability of oligofructose ingestion to improve blood glucose control, increase glucose-stimulated insulin secretion, and reduce body weight gain is dependent on the action of the fermentable fiber on GLP-1 in mice as well (60).

CENTRAL REGULATION OF INSULIN ACTION

Insulin binds to the insulin receptor expressed in the hypothalamus and the dorsal vagal complex of the brain and triggers signaling cascades to regulate systemic glucose and energy homeostasis (9, 13, 61–65). Insulin resistance in both the hypothalamus and dorsal vagal complex is detected in association with excess calorie intake, obesity, and/or T2D, and are partly responsible for a disruption in metabolic homeostasis (61–63, 66–68). A direct molecular intervention in the hypothalamus and dorsal vagal complex can affect insulin action and contribute to improved blood glucose control (61, 66, 67). Alternatively, we put forward a working hypothesis that insulin resistance in the brain can be reversed by targeting the gut with small intestinal signaling molecules and/or components of the microbiota.

This working hypothesis stems from the fact that small intestinal-derived peptides such as CCK and GLP-1 act in a paracrine (via vagal afferent) and an endocrine (via the brain) fashion to trigger a gut-brain axis to regulate systemic metabolic homeostasis (69–75). Similarly, we postulate that brain insulin resistance can be reversed by targeting the small intestine via a gut-brain axis (Fig. 1). In this regard, short-term high fat (HF)-feeding induces insulin resistance in parallel to a reduction of NAD⁺-dependent deacetylase sirtuin 1 (Sirt1) in the small intestine of rats (76). Genetic knockdown of Sirt1 in the upper small intestinal mucosal of healthy rats for 14 days induces insulin resistance as the ability hyperinsulinemia to suppress hepatic glucose production is impaired when assessed by pancreatic-euglycemic clamps (76). Conversely, administration of insulin-sensitizer resveratrol directly into the upper small intestine of HF rats increases Sirt1 protein and activity in the small intestine mucosal, and consequently, not only rescues the ability of circulating hyperinsulinemia but also restores that ability of insulin infusion into the hypothalamus to lower hepatic glucose production during pancreatic-euglycemic clamps via a gut-brain neuronal axis (76). Importantly, the insulin-sensitizing effect of activating Sirt1 in the upper small intestine is retained in HF-induced obesity as well as in hyperglycemic diabetic rats (76). The underlying signaling mechanism within the upper small intestine of how Sirt1 signals to the hypothalamus remains unknown. Is the action of gut-derived peptides as well as any signaling cascades within the vagus necessary for small intestinal Sirt1 to increase insulin sensitivity? Is the neuronal relay between the nucleus of the solitary tract and the hypothalamus necessary for small intestinal Sirt1 to increase insulin sensitivity? If yes, which neurotransmitter(s) is involved?

Although much remains to be explored, a set of studies has demonstrated that insulin resistance in the dorsal vagal complex can also be reversed by targeting the small intestine (77). In parallel to the development of insulin resistance, short-term HF induces changes of the upper small intestinal microbiome in rats that led to an increase in small intestinal taurochenodeoxycholic acid (TCDCA) as well as in the plasma and the dorsal vagal complex (77). This is in parallel to the fact that the serum level of taurine-conjugated bile acids (i.e., TCDCA and tauroursodeoxycholic acid) is elevated and positively correlated with whole-body insulin resistance in people with T2D (78). Translation of a healthy microbiota into the upper small intestine of rats lowers TCDCA levels and also increases the ability of circulating hyperinsulinemia or insulin infusion into the dorsal vagal complex to lower hepatic glucose production in HF rats (77). Further, direct infusion of TCDCA into the dorsal vagal complex of HF rats with healthy microbiome transplant prevents the ability of the microbiome transplant to increase insulin sensitivity in HF rats (77). These findings collectively suggest that HF-induced changes in small intestinal microbiome increase TCDCA levels in the plasma and subsequently in the dorsal vagal complex to induce insulin resistance in rats (Fig. 1). It remains unknown whether changes in TCDCA levels will be found in model of obesity and diabetes, and whether the changes are small intestinal microbiome-dependent. Second, the underlying mechanism responsible for the elevation of TCDCA induced by changes in small intestinal microbiome remains elusive. Finally, the translational relevance of targeting small intestinal Sirt1 as well as changes in microbiome on insulin sensitivity in HF, obese, and/or diabetic conditions warrants further investigation.

CONCLUSIONS

The intestine relays peripheral and central signals that regulate blood insulin levels and action as well as glucose levels. Detection of nutrients in the intestine generates signals that alter blood insulin levels and action through responses in the pancreas, liver, and the brain. The intestinal microbiota participates in the paracrine and endocrine communication between the gut and other tissues involved in insulin and glucose control. Understanding how host-microbe symbiosis integrates peripheral and central control of insulin in both animal models and humans is an important future direction.

GRANTS

J.D.S is supported by a CIHR Foundation Grant (FDN-154295) and a Tier 2 Canada Research Chair in Metabolic Inflammation at McMaster University. T.K.T.L. is supported by a CIHR Foundation Grant (FDN-143204) and holds the John Kitson McIvor (1915-1942) Endowed Chair in Diabetes Research & a Tier 1 Canada Research Chair in Diabetes and Obesity at the Toronto General Hospital Research Institute and the University of Toronto. Both J.D.S and T.K.T.L. are supported by a CIHR Team Grant (MRT-168045) that focuses on investigating the metabolic impact of intestinal microbes from humans.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.D.S. and T.K.T.L. drafted manuscript; J.D.S. and T.K.T.L. prepared figures; J.D.S. and T.K.T.L. edited and revised manuscript; J.D.S. and T.K.T.L. approved final version of manuscript.