Fatty acids exert both acute and long-term effects on insulin secretion. Fatty acids are actively taken up and metabolized by P-cells, and can regulate P-cell enzymes and ion channels (27). It has long been recognized that FFAs acutely (i.e., when elevated for less than about 6 to 12 hours) increase glucose-stimulated insulin secretion (GSIS) (172,178,179). Conversely, acute lowering of plasma FFAs with nicotinic acid results in a reduction in basal plasma insulin in both nonobese and obese healthy, fasted individuals (144) and in patients with type 2 diabetes (26,144). Fatty acylCoA and possibly DAGs accumulated within the P-cells may stimulate protein kinase C and stimulate exocytosis of insulin granules (180). The recently discovered GPR40 receptor is highly expressed in P-cells (181,182) and may be involved in the FFA-mediated insulin secretion. FFAs are ligands for this cell-surface G-protein coupled receptor (183), and binding has been shown to promote insulin secretion in vitro (184-187). This occurs via a series of actions, from protein kinase A activation and increased AMP/ATP ratio, which antagonizes voltage-gated K+ channels, leading to opening of voltage-dependent Ca2+ channels, increasing the intracellular Ca2+ concentration, resulting in exocytosis of insulin-containing secretory granules (185). Previous studies have also shown that FFA binding to GPR40 may also induce K+-ATP channel-independent mobilization of intracellular Ca2+ pool (188,189).
In contrast to acute exposure, prolonged intravenous infusion of a synthetic lipid emulsion infusion (>12-24 hours) results in reduced GSIS and P-cell mass in vitro (190) and reduced GSIS in vivo (172,191,192). Several in vitro studies in P-cell lines and in rodent and human islets have subsequently confirmed that insulin secretion at high glucose concentrations is impaired in a time-dependent fashion by exposure to FFAs (193-197). Islets from prediabetic ZDF rats and from fructose-fed insulin resistant rats appear to be more susceptible to this FFA-mediated desensitization of GSIS (195,196). Some controversy exists, however, because basal insulin secretion at low glucose concentrations was elevated in normal rodent islets and islet cell lines in most studies (28,193-195,198). Furthermore, insulin secretion at low glucose concentration is either unchanged or decreased by FFAs in islets from ZDF prediabetic rats or prediabetic OLEFT rats (195,199).
P-cell lipotoxicity, a term coined by Unger in 1995, describes lipid-induced functional impairments in GSIS as well as reduction in P-cell mass, and is also linked to, but not necessarily caused by, intracellular TG accumulation (137). Insulin secretion is mainly regulated by glucose through the closure of ATP-sensitive K+ channels, leading to membrane depolarization, opening of voltage-dependent Ca2+ channels, increased intracellular Ca2+ concentration, subsequent activation of kinases, and exocytosis of secretory granules. A potential mechanism lies in the stimulation by FFAs of the ATP-sensitive K+ channels (200,201) leading to impaired mitochondrial function. Ongoing accumulation of FFAs may chronically prevent K+ channels from closure, thus contributing to the resistance. Intracellular stores of triglycerides can be hydrolyzed by hormone-sensitive lipase, which is expressed and active in P-cells (202) and, therefore, may constitute an additional in situ supply of long-chain fatty acids. FFAs may induce expression of uncoupling protein(UCP)2, thus decreasing the ATP pool generated from glucose, and insulin secretion (203). Although no amelioration has been seen after adenovirus-mediated UCP-2 overexpression in P-cells derived from Zucker diabetic rats (139), UCP2 expression is increased in animal models of type 2 diabetes (204-206). Fatty acid accumulation causes induction of oxidative stress (197,207) via elevated synthesis of ceramides, which in turn induce the expression of the inducible NO synthase iNOS (208). Superoxide radical, which been shown to activate UCP2, is increased in ß-cells from diabetic mice (206) and Zucker diabetic rats (209). NO and oxygenated free radicals activate some caspases responsible for apoptosis, thus leading to reduced ß-cell mass (207,210-212).
An alternative hypothesis has been proposed in which FFAs may modulate the expression of certain genes involved in glucose or fatty acid metabolism. Exposure of ß-cells to high levels of FFAs leads to decreased expression of the glucose transporter Glut-2 and glucokinase with subsequent decreased utilization of glucose (213). In addition, FFAs decrease insulin biosynthesis (193,214-216), alter proinsulin processing, and decrease insulin gene transcription by unclear mechanisms (217,218). GPR40 has been suggested to mediate not only acute but also chronic effects of FFAs, because loss of GPR40 decreases insulin secretion by ß-cells in response to FFAs, and GPR40-deficient mice are protected against high fat diet-induced hyperinsulinemia, hepatic steatosis, and hypertriglyceridemia, as well as increased hepatic glucose output, hyperglycemia, and glucose intolerance (219). Conversely, overexpression of GPR40 results in impaired ß-cell function, hypoinsulinemia, and diabetes (219). FFAs-mediated downregulation of PKC or inhibition of specific PKC isoforms may also be involved.
In summary, there is convincing evidence from in vitro and some in vivo studies in animals and humans that chronically elevated fatty acids impair various aspects of pancreatic ß-cell function. It is not yet known, however, whether a chronic elevation of plasma FFAs contributes to the ß-cell dysfunction that is characteristic of the progression from prediabetes to type 2 diabetes in humans or how important this factor is in relation to other causative factors.
Effects of FFAs on Lipid Oxidation and Mitochondrial Function Effects of FFAs on Muscle Lipid Oxidation and Mitochondrial Oxidative Phosphorylation
Skeletal muscle has been shown to have the capacity to switch between fat and glucose as fuel. In lean, insulin sensitive people, a switch from fasted to fed state is reflected by a pronounced decrease of FA uptake and oxidation whereas glucose is preferentially used as substrate. This capacity has been termed "metabolic flexibility" (220), compared to the "inflexibility" of insulin resistant muscle to make this transition. In obese persons, fasted muscle metabolism is characterized by partially blunted fat oxidation and less suppression of glucose oxidation, and the switch to fed state is accompanied by only a slight decrease in fat oxidation and partial increase in glucose utilization (221). Defects in skeletal muscle mitochondrial oxidative capacity (the process which produces ATP from fuel oxidation) and fat metabolism are correlated with, and may contribute to, insulin resistance (220,222,223). In a recent study, Ukropcova et al. reported that insulin sensitivity was linked to the capacity of the muscle to oxidize fat, and that this relationship was retained ex vivo by cultured myocytes (224).
Studies have linked defects in mitochondrial oxidative phosphorylation and insulin resistance in elderly subjects and in healthy individual with family history of type 2 diabetes (123,124). In both cases, defects in insulin-stimulated muscle glucose metabolism were associated with lipid accumulation within the muscle, and with markedly reduced muscle mitochondrial ATP synthesis and tricarboxylic acid flux, reflecting altered mitochondrial oxidative and phosphorylative capacity. Another report has shown reduced mitochondrial size in obese, insulin resistant subjects with or without type 2 diabetes (222). Two mechanisms have been invoked to explain these mitochondrial defects, which include mitochondrial dysfunction and a loss of mitochondria, potentially owing to impaired biogenesis. PGC1 (PPAR^co-activator 1) is a transcription factor known to control the adaptative thermo-genesis process in muscle to enhance mitochondrial oxidative phosphorylation, and is involved in mitochondrial biogenesis (225). Interestingly, expression of PGCla and/or ß is reduced in obese Caucasian subjects with glucose intolerance and type 2 diabetes (226), and in obese diabetic and overweight nondiabetic Mexican-Americans (227). Forced expression of PGCla in muscle leads to increased oxidative type I muscle fibers and expression of mitochondrial markers (228). Conversely, PGCla-deficient mice have lower mitochondria number and respiratory capacity, but normal mitochondrial function, and impairment of muscle PGCla signalling may contribute to systemic insulin resistance (229). Interaction between PGCla and other transcription factors, including the estrogen-related receptor (ERR) and PPARa (230,231), may also be involved in the upregulation of muscle mitochondrial oxidative phosphorylation and FA oxidation, and inhibition of glucose oxidation (232). However, when fed a high-fat diet, PGCla-deficient mice are protected against insulin resistance, despite impairment in skeletal oxidative phosphorylation, an observation which may be explained by the fact that PGCla may have deleterious effects in other organs, such as liver, where it potentially increases HGP, and p-cells, where it may decrease GSIS by impairing ATP-sensing K+channel activity (232). Further studies are awaited to delineate the role of these nuclear receptors and co-factors in linking fat metabolism and mitochondrial function.
Decreased mitochondrial fatty acid oxidation, caused by mitochondrial dysfunction or biogenesis, may generate increased intracellular fatty acyl CoA and DAGs which, in turn, impair insulin signalling via altered IRS phosphorylation and PI3K activity, leading ultimately to altered glucose uptake. Conversely, elevated FFAs may amplify this defect by directly or indirectly impairing mitochondrial function or activity of transcriptional factors involved in mitochondrial function and biogenesis. For instance, lipid infusion to raise plasma FFA concentration in healthy men during hyperinsulinemic clamp conditions leads to decreased insulin-stimulated ATP synthesis and concomitant insulin resistance (233). On the other hand, FFAs have been shown to decrease PGC1 expression and mitochondrial oxidative phosphorylation (also named OxPhos) (234). These results were corroborated by the report from Sparks and colleagues showing that, after 3 days, feeding a high fat diet to humans led to decreased muscle PGC1 and OxPhos gene expression (235). It remains, however, to be determined whether chronic elevation of FFA in humans results in the same defects.
Effects of FFAs on Pancreatic 3-Cell Mitochondrial Function
In line with the above data, it has become clear that mitochondrial function is also required for pancreatic p-cells to secrete insulin in response to glucose, and increased p-cell mass is necessary to respond to the increased demand for insulin. As discussed above, increased ATP/AMP ratio within the p-cells triggers a series of events, involving K+-ATP-dependent channels and Ca2+ voltage-dependent channels, leading to exocytosis of insulin secretory granules. It is therefore possible that, based on what is observed in the muscle, impaired mitochondrial function may impede glucose-stimulated insulin secretion. In support of this hypothesis, FFAs may induce expression of the mitochondrial inner membrane protein UCP2 that uncouples glucose oxidative metabolism from ATP synthesis, thereby decreasing the ATP pool generated from glucose and impairing insulin secretion (203).
Was this article helpful?