Gut Peptides Are Novel Regulators of Intestinal Lipoprotein Secretion: Experimental and Pharmacological Manipulation of Lipoprotein Metabolism

Individuals with metabolic syndrome or frank type 2 diabetes (T2D) are at increased risk for cardiovascular disease (CVD) contributed to by the atherogenic dyslipidemia that frequently accompanies these conditions (1,2). The typical lipid abnormalities in these conditions include elevated plasma triglycerides (TG) carried predominantly in triglyceride-rich lipoproteins (TRLs) (e.g., VLDL and chylomicrons), low HDL cholesterol (HDL-C) and HDL particle numbers, and qualitative changes in lipoproteins (e.g., predominance of small, dense LDL and glycation and oxidation of lipoprotein particles) (3). Elevated TRL plays a pathophysiological role in HDL lowering and the generation of small, dense LDL via lipid exchange and modification of the composition of these denser lipoprotein classes (4–6). In addition to the well-documented, impaired clearance of TRL particles from the circulation, both intestinal (apolipoprotein [apoB]-48–containing) and hepatic (apoB-100–containing) TRL production are increased in insulin-resistant states and T2D (7). Understanding the molecular regulation of lipoprotein production by the liver and intestine may provide clues for the development of therapeutic strategies to attenuate the atherogenic dyslipidemia. Intestinal lipoprotein (chylomicron) production, in particular, is no longer viewed as being regulated only by fat ingestion and luminal fat content, although ingested fat is still unarguably the dominant determinant of intestinal lipoprotein production. Instead, it is now appreciated that chylomicron production, even in the fasted state, is subject to a variety of hormonal and nutritional effectors and is dysregulated in insulin-resistant and diabetic states (8). More recently, studies of the role of gut hormones, glucagon-like peptide-1 (GLP-1) and glucagon-like peptide-2 (GLP-2), and pharmacological regulators of their plasma concentration have provided new insights into the regulation of chylomicron production, which will be discussed in this Perspective.

Incretin-based therapies (i.e., GLP-1 receptor [GLP-1R] agonists and dipeptidyl peptidase-4 [DPP-4] inhibitors) are antidiabetes agents with proven efficacy in glucose lowering (9,10). The mechanism of action of these agents centers on modulation of pancreatic hormone secretion by GLP-1. GLP-1 is secreted by the intestinal L cells in response to meal ingestion. It enhances glucose-dependent insulin secretion and inhibits glucagon secretion, among other biological functions. GLP-1R agonists activate the GLP-1R, thus providing an exogenous source of GLP-1 activity. DPP-4 inhibitors preserve the endogenously secreted GLP-1 by inhibiting the enzyme DPP-4 that rapidly degrades GLP-1 and many other peptides (11). These compounds offer several additional benefits compared with other classes of antidiabetes medications, including a low risk of hypoglycemia. With long-term use, incretin-based therapies also improve insulin sensitivity and pancreatic function and are weight neutral (DPP-4 inhibitors) or induce weight loss (GLP-1R agonists), which contribute to effective glycemic control. Besides glucose lowering, evidence is accumulating that incretin-based therapies provide pleiotropic benefits, one of which is improvement in lipid parameters.

Improvement in both fasting and postprandial lipid profiles has been documented in clinical trials examining the glucose-lowering efficacy of several GLP-1R agonists and DPP-4 inhibitors in patients with T2D. Several recent meta-analyses indicated modest reductions in fasting LDL-C, total cholesterol, and TG with small or no significant improvement in HDL-C following chronic treatments with GLP-1R agonists (12–14) or DPP-4 inhibitors (13,15). For example, treatment with exenatide or liraglutide for several weeks to 3 years reduced fasting levels of TG, total cholesterol and LDL-C, increased HDL-C (16–19); and reduced fasting apoB (19–22), apoB-48 (a surrogate measure of intestinal lipoprotein particle numbers) (23), and free fatty acid (FFA) (24–26). Compared with GLP-1R agonists, DPP-4 inhibitors are less effective in lowering fasting lipids. In some but not all studies, sitagliptin decreased fasting TG (21) and decreased both fasting TG and apoB-48 (27). Alogliptin decreased fasting TG, apoB-48, and remnant-like particles (28), and anagliptin reduced fasting TG, apoB, and LDL-C (29). These effects may be related to the efficacy of these therapies in improving glycemic control with both GLP-1R agonists and DPP-4 inhibitors (12,30,31) and in reducing weight that is more notable with GLP-1R agonists than DPP-4 inhibitors (12,32,33). Future studies are required to determine the contribution of weight loss and improved glycemic control to fasting and postprandial lipids and to delineate the effects of these compounds on lipids independent of improved glycemic control and weight loss.

GLP-1R agonists are consistently associated with improvements in postprandial lipids (Table 1). Postprandial TG was lowered with exenatide (16,26,34,35) or liraglutide treatments (23) of varied treatment length, ranging from several weeks to 1 year. Besides postprandial TG, 51 weeks of treatment with exenatide also decreased apoB-48, VLDL cholesterol, and FFA during mixed-meal tests (16). Similarly, liraglutide reduced postprandial apoB-48 concentrations, along with reductions in total cholesterol and LDL-C (23). In healthy humans, a 390-min infusion of native GLP-1 markedly reduced postprandial TG excursion (36). In individuals with impaired glucose tolerance or recent-onset T2D, a single subcutaneous dose of exenatide also attenuated postprandial TG excursion (37).

GLP-1R agonists are also associated with qualitative changes in lipoprotein particles and atherosclerosis-related markers. For instance, exenatide treatment decreased oxidative stress markers (e.g., malondialdehyde and oxidized LDL-to-LDL-C ratio [16]), while liraglutide shifted the composition of LDL particles away from small, dense LDL particles and decreased the ratio of apoB/apoA-I (19).

Improvements in postprandial lipids by DPP-4 inhibitors were more notable compared with fasting lipids (Table 1). For instance, vildagliptin treatment for 4 weeks lowered postprandial chylomicron TG (CM-TG) and apoB-48 (38), and sitagliptin treatment for 2 weeks decreased postprandial TG (26). Sitagliptin treatment for 6 weeks decreased postprandial TG and apoB-48, total apoB, VLDL cholesterol, and FFA during an oral lipid tolerance test (39). Beneficial effects on postprandial lipids were also observed with alogliptin, thus 16 weeks of alogliptin treatment decreased postprandial TG in plasma and TRL, along with decreased chylomicron apoB-48 and cholesterol (28), and 1-week treatment decreased postprandial TG and apoB in subjects without diabetes (40). Postprandial TG, apoB, and LDL-C were reduced with anagliptin (29). Vildagliptin therapy for 4 weeks decreased postprandial remnant-like particles and increased LDL size (41), qualitative changes in lipoprotein size and composition that have been associated with reduced risk of atherosclerosis. In addition, sitagliptin was found to reduce plasma markers of low-grade inflammation and cell adhesion molecules (42), which are also potentially antiatherosclerotic changes.

No clear contribution of fasting TG to postprandial TG excursion can be established from the chronic studies that examined both. Decreased postprandial TG response to 3-week treatment with liraglutide was not accompanied by decreased fasting TG (23). With regard to DPP-4 inhibitors, both postprandial TG excursion and fasting TG were reduced with alogliptin (28) or anagliptin (29). However, postprandial TG excursion was attenuated by sitagliptin (39) and vildagliptin (38) in T2D or alogliptin in healthy individuals (40) without significantly lower fasting TG. Future studies are needed to specifically examine to what extent changes in fasting TG contribute to the attenuation of postprandial TG excursion by GLP-1R agonists and DPP-4 inhibitors. The difference in fasting lipids with GLP-1R agonists and DPP-4 inhibitors may be attributed to the level of biological GLP-1 activity achieved with these two therapeutic strategies. Compared with subcutaneous injection of GLP-1R agonists, which raise GLP-1 plasma concentration to pharmacological levels, oral DPP-4 inhibitors only modestly elevate circulating GLP-1 within the physiological range (43). GLP-1R agonists induce weight loss and delay gastric emptying, whereas DPP-4 inhibitors do not. It is also possible that, due to the broad spectrum of substrates of DPP-4, DPP-4 inhibitors affect lipid metabolism through mechanisms beyond GLP-1 activity.

Improvement in postprandial lipemia by GLP-1R agonists and DPP-4 inhibitors may be related to less effective dietary fat handling, decreased secretion, and/or increased clearance of intestinal lipoproteins. GLP-1 is well known to slow gastric emptying (44,45), which is expected to slow the passage of dietary fat to the small intestine. GLP-1 was also shown to inhibit gastric lipase secretion (46) and intestinal motility (47). These effects may translate into diminished efficiency with regard to processing of dietary fat. Exendin-4 treatment lowered jejunal microsomal triglyceride transfer protein activity and diminished jejunal TG availability in hamsters (48). This was associated with an increase in TG levels in the luminal content of the intestine 2 h post fat load, suggesting impairment in lipid absorption. Alternatively, exendin-4 slows gastric emptying, which could limit the amount of lipids available for absorption and chylomicron assembly in the small intestine. Whether GLP-1R agonist treatments are accompanied by increased fecal lipid excretion or compensatory absorption in the distal gut remains unknown. As neither GLP-1R agonists nor DPP-4 inhibitors cause clinical steatorrhea, these compounds may be associated with slowed digestion and absorption rate without clinically apparent fat malabsorption.

Reduced postprandial lipemia may be the result of decreased secretion and/or increased clearance of chylomicrons. In the very few studies so far that have examined intestinal lipoprotein kinetics in humans in response to GLP-1 agonists (exenatide, single dose in healthy subjects) (49) or DPP-4 inhibitors (sitagliptin, single dose in healthy subjects or 6-week treatment in T2D patients) (27,50), fractional catabolic rates were not significantly affected. The current available evidence thus supports the view that the reduced postprandial chylomicron concentration is primarily due to reduced production rather than an increase in particle clearance. With more chronic treatment, clearance of chylomicrons may be increased, but this has not yet been examined. Chronic use of GLP-1R agonists induces weight loss and both GLP-1R agonists and DPP-4 inhibitors attenuate insulin resistance and augment β-cell function. As both hepatic and intestinal lipoprotein production are stimulated by elevated circulation FFA in humans (51), greater antilipolytic effects of insulin on adipose tissue in this setting result in less FFA substrate delivery to the liver and intestine, which could potentially lead to attenuated VLDL-TG and/or CM-TG synthesis and secretion. Insulin resistance, in the whole body and at the intestinal level, is associated with overproduction of intestinal lipoproteins (52–54). Several lines of evidence also suggest that GLP-1 may modulate intestinal lipoprotein synthesis and secretion beyond gastric emptying, lipid digestion, and gut motility. In rats, when lipids were infused directly into the duodenum, GLP-1 infusion decreased lipid absorption, along with reduced lymph flow and apoB output (55). This indicates that the GLP-1 effect on intestinal lipid handling may occur between the intestinal lumen and lymph output. In hamsters, administration of sitagliptin or exendin-4 (a GLP-1R agonist) prior to oral fat load decreased TRL TG and apoB-48 accumulation in the circulation following triton injection (56). However, when exendin-4 was given 1 h after oral gavage of olive oil to allow lipids to pass through the stomach into the duodenum (i.e., to circumvent the slowing of gastric emptying by GLP-1), the suppression in TG and apoB-48 was not abolished (56).

Our recent studies in healthy humans support an acute, direct inhibitory effect of incretin-based therapies on intestinal lipoprotein production (49,50). Intestinal lipoprotein synthesis and secretion are complex processes (8), with GLP-1 directly or indirectly modifying a number of the regulatory factors. For example, GLP-1 stimulates glucose-dependent insulin secretion and insulin acutely inhibits intestinal lipoprotein production (57). In addition, intestinal lipoprotein production is stimulated by elevated circulating FFA (51). Short-term GLP-1 infusion or exenatide injection suppressed circulating levels of FFA in humans (36,37), and sitagliptin treatment for 4 weeks also decreased postprandial levels of FFA (39), which is expected to contribute to the reduced apoB-48 production. The experimental design of our mechanistic study in humans allowed us to isolate the intestine-specific effect of the incretin-based therapies from gastric emptying, changes in pancreatic hormones or FFA, glycemic control, or weight reduction. To overcome the expected inhibition of gastric emptying with exenatide, nutrients were infused directly into the duodenum. TRL kinetics was studied under the conditions of a pancreatic clamp to minimize the changes in pancreatic hormone secretion. As a result, insulin levels were only modestly and transiently increased following exenatide injection. Under such conditions, intestinal lipoprotein production was significantly decreased by exenatide (49), suggesting a direct inhibitory effect, possibly at the intestinal enterocyte level. This notion was further supported by a subsequent study in humans in which a single dose of sitagliptin decreased intestinal lipoprotein production while insulin levels were matched in the treatment and placebo arms of the study (50). Plasma FFA levels were similar between exenatide or sitagliptin and placebo in both studies; therefore, the reduction in apoB-48 production was not due to changes in FFA levels in these experimental settings. Neither exenatide nor sitagliptin affected TRL apoB-100 production in acute studies in healthy humans (49,50). This might be due to the lack of GLP-1R expression in the liver (58). More recently, longer-term (6 weeks) treatment with sitagliptin suppressed both TRL apoB-48 and apoB-100 production rates in patients with T2D (27). The mechanism of improvement in hepatic lipoprotein production in this study remains to be further elucidated, although it is likely explained by indirect factors, such as improvement in glycemia with chronic use, rather than direct inhibition of hepatic lipoprotein synthesis and secretion. In contrast, a direct action of GLP-1 on intestinal lipoprotein production is supported by ex vivo studies in isolated hamster intestinal enterocytes, where exendin-4 inhibited apoB-48 secretion into the medium (56). Consistent with this, GLP-1R expression has been identified in the small intestine of humans (59). The molecular mechanisms where GLP-1 regulates chylomicron synthesis, assembly, and secretion remain to be elucidated. Interestingly, exendin-4 administered directly into the central nervous system also suppressed intestinal lipoprotein secretion in hamsters (60). GLP-1R is expressed in certain regions in the brain, which underlies the mechanism whereby GLP-1 induces weight loss (61). The relative contribution of the central nervous system to the regulation of intestinal lipoprotein production has not been fully established, has not yet been demonstrated in humans, and warrants further study.

Published data provide some clues but are not conclusive regarding particle size change with incretin-based therapy. GLP-1R agonists have been shown to lower postprandial TG concentration in healthy and insulin-resistant humans following ingestion of a high-fat meal (36,37), as well as in rats, hamsters, and mice following oral lipid administration (56,62). A reduction in TRL TG concentration could be explained by fewer particles of unchanged size, smaller particles (carrying less TG per particle) but unchanged number, or the combination in which there are fewer particles that are also reduced in size. Given that GLP-1R agonists have been shown to reduce TRL apoB-48 concentration and secretion rate, this suggests that fewer particles are present that are either normal in size or are smaller, with reduced lipidation. We have observed that acute intraperitoneal exendin-4 treatment of healthy hamsters reduced the number of large TRL particles, suggesting a possible shift toward the formation of smaller chylomicrons (48). Taken together, these data suggest that in addition to fewer chylomicrons being secreted, it is possible that these particles are also smaller in size. In contrast, it should be noted that when lipid was infused intraduodenally (i.e., the effect of the GLP-1R agonist in slowing gastric emptying was experimentally “eliminated”) intraperitoneal exendin-4 lowered TRL apoB-48 concentration without affecting TRL TG concentration in healthy hamsters. These results mirror those from healthy humans receiving intraduodenal lipid infusion (49) and were found to result from a shift toward the secretion of larger, more highly lipidated particles during exendin-4 treatment (as assessed by fast protein liquid chromatography in the hamster; particle size not directly assessed in humans) (S. Farr, K. Adeli, unpublished data). This may suggest that GLP-1 has direct effects on intestinal apoB-48 output, while its ability to lower postprandial TG concentration and particle lipidation (size) may rely on proximal processes within the stomach or upper gastrointestinal tract. Further studies are necessary to confirm these findings and offer a mechanistic explanation.

Native GLP-1, GLP-1R agonists, and DPP-4 inhibitors have been suggested to provide cardiovascular risk benefits (63–65). In addition, as DPP-4 cleaves several other peptides, many of which directly affect the heart and blood vessels, DPP-4 inhibitors may offer additional cardiovascular benefits beyond elevating GLP-1 (65). A less appreciated potential for these compounds to provide cardiovascular benefits is their capacity to improve lipid profiles, as discussed above. Prospective, controlled, hard CVD outcomes trials are required to investigate the true CVD risk benefits of incretin-based therapies, if they indeed exist. Two outcomes trials examined the DPP-4 inhibitors alogliptin in patients with T2D after acute coronary syndrome (66) and saxagliptin in patients with T2D who had a history or were at increased risk of CVD events (67). These trials demonstrated noninferiority over placebo but did not provide evidence of CVD risk reduction. It remains to be determined in ongoing and future clinical trials, some of which are soon to be released, whether GLP-1R agonists and/or DPP-4 inhibitors elicit CVD benefits.

GLP-2, a gut peptide that is also derived from posttranslational processing of the proglucagon gene in intestinal L cells (68), as is GLP-1, has also been implicated in intestinal lipid handling. It is cosecreted with GLP-1 in a 1:1 molar ratio in response to nutrient ingestion. GLP-2 promotes intestinal growth and nutrient absorption. It increases bowel mass by promoting proliferation and decreasing apoptosis of the enterocytes. GLP-2 also provides cytoprotection and plays a critical role in the adaptive response to intestinal injury and stress (69). This biological effect of GLP-2 is currently being used for treatment of short bowel syndrome and related gastrointestinal abnormalities, such as malabsorption, inflammation, and mucosal damage. Not surprising, with its enhancement in intestinal growth and function, GLP-2 is known to enhance intestinal absorption of several macronutrients, including carbohydrates and amino acids. This is associated with enhanced mucosal hexose transport, increased expression of genes encoding nutrient transporters, and increased expression of digestive enzymes along the gastrointestinal tract (69).

GLP-2 regulation of intestinal lipid handling has recently been investigated. In animal models, GLP-2 enhanced intestinal luminal lipid absorption and increased secretion of intestinal lipoproteins in vivo, as well as TRL apoB-48 secretion ex vivo, in cultured jejunal fragments derived from hamsters (70). In healthy humans receiving an intravenous infusion of GLP-2 for 6.5 h, postprandial plasma TG and FFA concentrations were increased (71). The mechanism whereby GLP-2 exacerbates postprandial lipemia is not fully understood. In hamsters, GLP-2 administration did not stimulate apoB-48 synthesis or increase mRNA expression of Mttp that encodes microsomal triglyceride transfer protein. GLP-2 increased the expression of fully glycosylated CD36 in the jejunum, and its enhancement of chylomicron secretion was lost in CD36^−/− mice (70). The role of CD36 in GLP-2 accentuation of postprandial lipemia in humans requires further elucidation as chylomicron remnants are increased in the postprandial state in humans with CD36 deficiency (72). GLP-2 infusion stimulated glucagon secretion and inhibited the secretion of gastric acid (71) and ghrelin (73) and modestly inhibited gastric emptying (74). The contribution of these factors to intestinal lipid uptake and secretion is not clear, but experimental hyperglucagonemia did not promote intestinal lipoprotein production in humans (75). In our recent study in healthy humans receiving intraduodenal infusion of mixed macronutrients, GLP-2 significantly increased TRL TG and apoB-48 concentrations within 30 min of a single subcutaneous injection during a pancreatic clamp (76). Such an effect was rapid and transient, peaking at approximately 1 h following GLP-2 administration and returned to baseline at around 3 h. This effect was not likely due to increased TRL particle production or decreased clearance, as suggested by mathematical modeling. Instead, it was most plausible that GLP-2 resulted in the release of stored, “preformed” particles from the enterocytes or lymph vessels. Indeed, 7 h after ingestion of a meal containing retinyl palmitate, which labels chylomicrons, GLP-2 injection resulted in a rapid rise in plasma and TRL retinyl palmitate as well as TRL apoB-48, indicating that GLP-2 promoted the release of intestinally stored lipids and lipoproteins.

The effect of GLP-2 on chylomicron particle size has not been directly examined in humans. In healthy individuals receiving intraduodenal lipid infusion, GLP-2 increased TRL apoB-48 more robustly than TRL TG under pancreatic clamp conditions (76), suggesting release of smaller, less lipidated particles. Acute GLP-2 treatment of healthy hamsters raised both postprandial TG and apoB-48 concentrations, in part via enhanced CD36-mediated fatty acid uptake (56). An increase in TG accumulation in the chylomicron fractions of the plasma was observed with GLP-2 treatment, as assessed by fast-protein liquid chromatography (S. Farr, K. Adeli, unpublished data); however, further studies are needed to assess whether this can be attributed to increased particle number, size, or both.

While the rapid mobilization of stored apoB-48 by GLP-2 is a distinct feature, it is surprising considering the location of GLP-2 receptor (GLP-2R) expression. The GLP-2 receptor is expressed in various tissues, including the stomach, small and large bowel, brain, and lung, with the intestine having the highest expression of GLP-2Rs. Although the L cells that secrete GLP-2 are adjacent to the enterocytes, the GLP-2R is not expressed on the cell surface of enterocytes, the cells responsible for chylomicron synthesis, assembly, and secretion (77). Instead, GLP-2Rs are expressed in the enteroendocrine cells of the proximal small intestine (jejunum). In porcine intestine, in addition to expression in enteroendocrine cells, GLP-2Rs are also expressed in endothelial nitric oxide synthase–expressing neurons and vasoactive intestinal polypeptide-positive enteric neurons (78). GLP-2 infusion upregulated endothelial nitric oxide synthase expression and activity and stimulated intestinal blood flow in pigs (78). In humans, GLP-2 increased blood flow primarily in the superior mesenteric artery (79). It is possible that GLP-2 stimulates chylomicron release by increasing local blood flow associated with enhanced nitric oxide production. This hypothesis remains to be tested in future studies.

The different actions of GLP-1 and GLP-2 on intestinal lipoprotein secretion may be related to their difference in site of action (i.e., neural stimulation, indirect endocrine effect, or paracrine effect), which requires further elucidation. The net effect of these gut peptides on chylomicron secretion is further complicated by their different rates of clearance from the circulation. GLP-1 and GLP-2 are cosecreted in a 1:1 molar ratio from the L cells and both are substrates of DPP-4. However, their degradation by DPP-4 occurs at different rates. GLP-1 degradation is rapid, with a half-life of ∼1.5 min, whereas GLP-2 is more stable with a half-life of ∼7 min (80,81). This results in more sustained circulating levels of GLP-2 compared with GLP-1. In hamsters, short-term, simultaneous infusion of GLP-1 and GLP-2 at a 1:1 molar ratio increased intestinal lipid absorption and elevated TRL TG and apoB-48. With more prolonged (120 min) infusion or with sitagliptin inhibition of DPP-4 activity, lipid response to oral fat load was reduced (62). In humans, as discussed above, sitagliptin treatment inhibited intestinal lipoprotein production in a pattern similar to exenatide treatment (49,50). Although the effects described above are pharmacological, it is likely that GLP-2 facilitates dietary fat absorption under physiological conditions. With sustained GLP-1 activity, induced by treatments with long-lasting GLP-1R agonists or DPP-4 inhibitors, postprandial lipemia is attenuated by GLP-1.

Recent advances in the study of gut peptides GLP-1 and GLP-2 have provided new insight into the regulation of intestinal lipoprotein secretion (Fig. 1). Besides glycemic control, incretin-based therapies improve fasting lipid profiles in clinical trials, more so for GLP-1R agonists than DPP-4 inhibitors, and ameliorate postprandial lipemia during meal tests. This is achieved via multiple pathways, including GLP-1 action on pancreatic hormone secretion, gastric emptying, gut motility, weight control, and improvement in glycemic control and metabolic status. In addition, GLP-1R agonists and DPP-4 inhibitors have been shown to directly inhibit intestinal lipoprotein production. The related gut peptide GLP-2, on the other hand, enhances intestinal lipid absorption and the release of intestinally derived lipoprotein particles. With the increased use of incretin-based therapies in glycemic control and the GLP-2 analogs in the treatment of intestinal disorders, their actions in relation to the modulation of postprandial lipemia should be recognized and examined further. Several critical questions remain, such as the mechanisms of GLP-1R agonists and DPP-4 inhibitors on intestinal lipoprotein production, the long-term CVD outcomes of incretin-based therapies, the mechanisms of GLP-2 regulating intestinal lipoprotein release, the contribution of central GLP-1R signaling to intestinal lipoprotein metabolism, and long-term consequences of GLP-2 analog use on lipids and atherosclerosis. The coordinated secretion yet distinct physiological roles of GLP-1 and GLP-2 require more in-depth study in health and disease. Pharmacological manipulation of gut hormone actions may provide beneficial effects on atherogenic lipoproteins. A more complete understanding of the regulation of intestinal lipoprotein secretion by gut hormones may provide new opportunities for developing novel strategies to reduce CVD risk.

Funding. S.D. and C.M. are recipients of postdoctoral fellowship awards from the Banting & Best Diabetes Centre, University of Toronto, and S.D. is the recipient of a Focus on Stroke 12 Fellowship Award from the Heart and Stroke Foundation of Canada. The authors acknowledge funding from the Heart and Stroke Foundation of Ontario (to K.A.) and the Canadian Institutes of Health Research and the Heart and Stroke Foundation of Ontario (to G.F.L.). G.F.L. holds the Sun Life Financial Chair in Diabetes and the Drucker Family Chair in Diabetes Research.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.