Molecular Mechanisms Involved in Insulin Resistance: Recent Updates and Future Challenges

Abstract

Insulin resistance (IR) is a condition in which insulin-mediated regulation of glucose metabolism in body tissues (primarily liver, adipose tissue and skeletal muscle) becomes disrupted. IR is a characteristic marker of type 2 diabetes and cardiovascular diseases. IR is generally associated with metabolic abnormalities, including hyperinsulinemia, impaired glucose homeostasis, hyperlipidemia and obesity. IR can arise from pathological, genetic and environmental factors or from a combination of these factors. Studies conducted in recent decades showcase the important role of adipose tissue in the development of IR via release of lipids and different circulating factors. These extracellular factors influence the intracellular levels of intermediates including ceramide and various lipids that influence the cell responsiveness to insulin. These intermediates are suggested to promote IR via inhibition of one or more components of insulin signaling pathway (e.g., insulin receptor, insulin receptor substrate proteins). This chapter will shed light on various molecular mechanisms and factors contributing to IR, which will help the researchers to design potential therapeutic strategies and interventions for efficiently managing IR and its related disorders.

1. Introduction

Following a meal, pancreatic β-cells produce insulin in response to increasing blood glucose and other metabolite levels for regulating systemic glucose homeostasis. Tissue insulin sensitivity, which characterizes the ability of a given concentration of insulin to correct blood glucose levels, is the driving force behind this homeostasis. Multiple processes in several organs are involved in this typically well-regulated homeostatic mechanism, including decreased glucose output from the liver (hepatic glucose output), increased glucose uptake into skeletal muscle and adipose tissue (where glucose is stored as glycogen), suppression of free fatty acid (FFA) release from adipocytes (suppression of lipolysis), and increased lipid accumulation in the liver and adipocytes. A sophisticated insulin-dependent signal transduction cascade controls these metabolic processes. Insulin resistance (IR) is defined as decreased insulin-stimulated glucose uptake into muscle and adipocytes and faulty insulin regulation of hepatic glucose production in patients with type 2 diabetes (T2D) and in many subjects affected by other conditions characterized by insulin resistance, such as obesity and polycystic ovary syndrome. The term insulin resistance was first coined to explain the considerable variability in the insulin dose necessary to lower high glucose levels in people with T2D, and then to characterize the magnitude of change in blood glucose level when a given amount of insulin and glucose was administered. The “defined quantity of insulin” is crucial because people with insulin resistance often have hyperinsulinemia, a condition in which insulin levels in the blood are higher than normal relative to the amount of blood glucose concentration under both fasting and fed conditions; this hyperinsulinemia compensates for IR in peripheral tissues to bring blood glucose levels back to normal [1].

When pancreas fails to supply excess insulin in humans with insulin resistance, a major defect in whole-body glucose homeostasis occurs, resulting in hyperglycemia and glucose intolerance (the latter including impaired fasting glucose and impaired glucose tolerance), which are the defining features of T2D. It is worth noting that, somewhat counterintuitively, patients with T2D frequently maintain “relative hyperinsulinemia” until the condition is at an advanced stage. IR is defined by insulin’s inability to induce glucose uptake into muscle and adipose cells due to a failure of the glucose transport mechanism mediated, at the molecular level, by glucose transporter type 4 (GLUT4) in those tissues. Furthermore, one of the hallmarks of IR is the inability to decrease hepatic glucose production, which is mostly due to a persistent increase in hepatic gluconeogenesis. IR has been linked to a variety of diseases. Indeed, IR represents a risk factor for various conditions, such as metabolic disorders (including T2D and obesity), heart disease, liver diseases (e.g., non-alcoholic fatty liver disease and non-alcoholic steatohepatitis), cancer, neurodegenerative diseases and frailty [2, 3, 4]. Despite the fact that IR is inextricably linked to T2D, an important factor involved in T2D pathophysiology is represented by the pancreas’ incapacity to function properly to compensate for the significant rise in blood glucose levels by secreting enough insulin to meet the increasing demand and help get blood glucose levels back to normal. IR is a key risk factor for T2D, yet it is not commonly recognized or treated in people without diabetes. The main reason for this phenomenon is that many people with insulin resistance do not have abnormal blood glucose levels. Therefore, diagnosis of IR is based on measuring insulin levels, which is not commonly done in clinical practice. Furthermore, only a small fraction of subjects with IR develop T2D, which is likely due to a propensity to β-cell failure in these subjects. There are no procedures to identify this susceptible subpopulation at this time. Individuals with IR are predisposed to significant disorders linked to T2D, including retinopathy, neuropathy and kidney disease, even if they do not have T2D [5]. In this chapter, the association between the early possible causes of IR is first discussed. Obesity is common in people with IR, but it is unclear whether concomitant hyperinsulinemia contributes to obesity development or whether it is a consequence of obesity-associated IR. We then look at how different metabolic tissues, such as muscle, adipose tissue, and the liver, communicate with one another. The mechanisms of impaired insulin signaling and the role of abnormal GLUT4 trafficking in the development of IR are also discussed. Extracellular factors that may contribute to IR are postulated. This discussion is then followed by a discussion of various intracellular molecular factors that contribute to IR. These factors have been considered as involved in processes that lead to IR. There are several ways for determining insulin action. Many laboratories have lately resorted to employing surrogate markers of insulin sensitivity and IR [6]. The “traditional” definition of IR is a condition in which blood glucose levels are abnormally high and insulin concentration needed to maintain glucose homeostasis is greater than predicted [7, 8].

2. Pathway to insulin resistance

Despite years of research, there is still a lot of uncertainty about the causative and temporal link between obesity, hyperinsulinemia, and IR. The proximal and distal parts of the insulin signaling system, which governs metabolism, can be arbitrarily partitioned. The classical components—which comprise the insulin receptor, insulin receptor substrate (IRS) proteins, phosphoinositide 3-kinase (PI3K) and AKT-constitute the proximal segment of the insulin signaling system. A common trait of the proximal components is their sparseness, which means that just a little part of each element is necessary to elicit a physiological signal. This guarantees signal amplification across the network. The proximal portion is also susceptible to very complex feedforward and feedback control, and is incorporated into a broader network that is dynamically regulated by combinatorial signaling inputs. The AKT substrates that are intimately related to the many physiological activities of insulin and are typically specialized to a particular cell type are referred to as the “distal segment” of the insulin signaling pathway. The distal elements are generally phosphorylated, which is a common trait. Insulin signaling begins with the hormone binding to its surface receptor, followed by activation of the receptor tyrosine kinase and tyrosine phosphorylation. IRS proteins are phosphorylated, causing them to create a signaling complex, which contains proteins with Src homology domains such as PI3K. As a result, phosphatidylinositol (3,4,5)-trisphosphate [PtdIns(3,4,5)P3 or PIP3] is produced. Serine/threonine (Ser/Thr) protein kinases like PDK1 and AKT, for example, are recruited to the inner leaflet of the plasma membrane. AKT is phosphorylated by PDK1 at one of its phosphorylation sites. Partially phosphorylated AKT activates mTORC2, while phosphorylation of AKT specifically at Ser473 results in complete AKT activation. Thus, AKT is a critical node in the insulin signaling pathway. AKT performs a variety of biological roles and is involved in the majority, if not all, of physiological metabolic processes. The Rab GTPase-Activating Protein (GAP) is an AKT substrate, which activates TBC1D4 (TBC1 Domain Family Member 4), a protein that regulates GLUT4 trafficking within the plasma membrane. The activation of glucose transport by insulin is the key mechanism that is disrupted in insulin-resistant muscle and fat cells. The GLUT4 is a facilitative glucose transporter, which is found in skeletal muscle, heart, adipocytes, and insulin-responsive neurons; it regulates muscle/fat glucose transfer. Unlike other transporters (like GLUT1), GLUT4 has a set of specific trafficking cues that let it migrate from endosomes and the trans-Golgi network (TGN) to a special intracellular population of vesicles known as “GLUT4 storage vesicles” (GSVs) [9, 10, 11]. GSVs act as a distinct controlled exocytic compartment that distributes GLUT4 to the cell surface in response to insulin and serves as a storage depot assuring low rates of glucose absorption in the fasting state. Although exercise increases GLUT4 translocation in muscle cells, it does so through a different mechanism than that regulated by insulin. AKT plays a critical role in the insulin-regulated GLUT4 translocation [12, 13]. These characteristics typically coexist, and there is strong evidence that each can cause the other two branches of the triad to emerge: obesity, hyperinsulinemia, and IR are caused by overnutrition in humans and animals; in humans, IR and obesity may also be caused by continuous insulin administration or by genetic factors; in addition, IR in humans may be caused by pharmacological interventions resulting in hyperinsulinemia [14, 15].

3. The trio-axis of obesity-hyperinsulinemia-insulin resistance

Obesity and IR are two topics that come up frequently. The long period during which obesity, IR and hyperinsulinemia develop, makes the determination of causative links between these conditions (which usually coexist in most persons with T2D at the time of diagnosis) particularly difficult. Obesity is common in people with IR, although it is unclear whether simultaneous hyperinsulinemia plays a role in obesity development or it is predominantly a result of obesity-dependent IR [16]. The study of first-degree relatives of people with T2D who only show some of these traits has shown to be one of the most effective strategies for addressing some unanswered questions in humans. As a result, a trait seen in relatives is more likely to appear early in the course of the disease. It has been found that these subjects can have considerable IR in skeletal muscle and liver (and possibly fat), along with modest hyperinsulinemia, even if they are not obese or glucose-intolerant [15, 17, 18]. Individuals who are lean and glucose-tolerant but exhibit IR have been identified in larger cross-sectional studies [19]. In these instances, obesity is unlikely to be the primary cause of tissue IR. However, the term “obesity” is defined differently depending on race and genetic background, and it should therefore used with caution. Body mass index may be more important in determining the risk of IR. Body weight, in general, and visceral fat (but not subcutaneous fat), in particular, should be considered for evaluation [20].

First-degree relatives of people with T2D had greater levels of circulating FFAs and intramuscular lipids than healthy control subjects [21], suggesting that intramyocellular lipid content represents an early abnormality in the pathogenesis of insulin resistance and that it may contribute to the impaired glucose uptake in skeletal muscle of insulin-resistant subjects to a greater extent than overall adiposity. This is in line with severe IR observed in patients with lipodystrophy syndromes, which are a heterogeneous group of diseases characterized by selective absence of adipose tissue, loss of functional adipocytes, ectopic steatosis, and severe dyslipidemia and IR [22, 23]. On the other hand, individuals with moderate or severe obesity can be “metabolically healthy” [24]. While it appears acceptable, based on this research, to conclude that obesity is not a risk factor essential for the development of IR, it is vital to highlight that the majority of subjects with IR are obese. As we will discuss later in the text, the amount and location of adiposity required to create IR varies greatly between subjects.

4. Insulin resistance and hyperinsulinemia

Defining the temporal link between hyperinsulinemia and IR is difficult since, as far as we know, IR does not exist in the absence of hyperinsulinemia in humans, and vice versa. Hyperinsulinemia can produce obesity and IR in humans, as evidenced by trials in which insulin is administered to induce hyperinsulinemia in otherwise healthy individuals or as it occurs naturally in people with insulinomas [25, 26]. Transgenic expression of multiple copies of the normal insulin gene causes hyperinsulinemia in mice, resulting in IR and glucose intolerance [27]. Inhibition of insulin secretion has also been shown to improve insulin sensitivity and to decrease body weight in rodents [28, 29, 30]. In mice, deletion of one copy of the insulin gene resulted in a reduction of the Western diet-induced hyperinsulinemia and in an improvement of insulin sensitivity [31]. Overall, the hypotheses that hyperinsulinemia causes IR and promotes obesity, or that IR associated with obesity causes hyperinsulinemia, both remain acceptable for the initial events involved in T2D pathophysiology. In actuality, IR and hyperinsulinemia coexist and lead to T2D in almost all cases [32]. Several data suggest a concept in which hyperinsulinemia is responsible for, or at least partly contributes to, many of the negative effects of IR; this implies that IR is a state in which many of the insulin actions are preserved, a condition known as “selective IR” [33, 34, 35]. This was first observed in the liver, where increased insulin levels are unable to decrease hepatic glucose output in people with T2D, although lipogenesis (a canonical insulin action in the liver) remains elevated [36, 37]. One explanation for this selectivity is that insulin signaling pathway in the liver splits into two arms, with IR affecting only the arm regulating hepatic gluconeogenesis but not the arm regulating lipid metabolism. Hepatic de novo lipogenesis is essentially a cell-autonomous phenomenon, whereas cell-nonautonomous suppression of hepatic glucose production by insulin depends upon the insulin-mediated decrease of adipocyte lipolysis and circulating FFAs [38]. There has also been evidence of selective IR in muscle and adipose tissue. Those insulin-regulated activities which are not affected by IR—such as lipogenesis, protein synthesis, or transcriptional control mediated by FOXO proteins—are hyperactivated in the context of hyperinsulinemia and are likely to worsen IR or its consequences [33, 34, 39, 40].

5. Heterogeneity in the development of insulin resistance and progression of metabolic disease and T2D

T2D patients are divided into different phenotypic clusters based on their symptoms and clinical features. Individuals in one of these groups share phenotypic traits. As a result, performing a comprehensive analysis of these groups will be of great importance in clinical settings. Phenotype data analysis and combination of phenotype data with genetic data are essential to gain a better understanding of the variability in the development and presentation of IR in humans [10, 11, 41].

6. Tissue-specific progression to insulin resistance

The appearance of IR occurs in various tissues in a specific order. The development of IR in several tissues—including skeletal muscle, liver, and adipose tissue—is a hallmark of fully developed T2D in humans [18, 19, 39, 40, 42]. Evidence shows a hierarchical progression of IR in skeletal muscle, liver and adipose tissue, whereby IR develops in one tissue and then spreads to other tissues via systemic circulating components. For example, IR in the liver and adipose tissue appears to occur prior to IR in muscle in C57Bl/6 mice fed a high-fat diet [43, 44, 45, 46]. An equivalent pattern in humans is unlikely, since first-degree relatives of persons with T2D who are in the early stages of the disease already have IR in both muscle and liver (and possibly fat) [47]. Since insulin sensitivity in humans is often measured as whole-body glucose consumption (to which adipose tissue contributes only to a small extent), the temporal development of IR in adipose tissue in humans is less obvious. Interestingly, multiple investigations show that insulin modulates hepatic glucose production via reducing adipocyte lipolysis in a non-cell-autonomous manner [45]. Given these findings, it is reasonable to believe that adipose tissue IR is a precursor to metabolic disease and T2D. However, there is a clear distinction between insulin action on the liver and insulin action on muscle: even in people with T2D, the defect in insulin sensitivity in the liver can be almost completely overcome by sufficiently high levels of insulin, whereas muscle (and fat) insulin sensitivity defects persist at higher insulin concentrations [40, 48, 49]. This indicates that the processes that cause IR in muscle and liver are distinct.

Tissue-specific insulin receptor gene knockouts in mice have provided persuasive evidence that IR in a particular tissue can at least spread to other organs. Experimenting with a specific deficiency in insulin action in muscle, fat, or liver has resulted in the spread of IR to other tissues in a number of cases [50]. However, depending on the tissue that is first targeted and/or in which a specific gene deletion occurs, the mechanism of inter-tissue communication varies. The deletion of GLUT4, which is essential for glucose uptake in adipose tissue and skeletal muscle, is one of the best examples of this inter-tissue communication. In mice, deletion of GLUT4 resulted in IR not only in the tissue from which the transporter was removed, but also in all metabolic tissues, including the liver. Surprisingly, normalization of blood glucose levels reverses IR in the liver and adipose tissue in muscle-specific Glut4 gene-knockout mice. This shows that glucotoxicity generated IR in this animal model, which is not the case in many other IR models, including the Western diet-fed C57BL/6J mice, which do not show considerable hyperglycemia [48, 49, 50, 51]. As a result, while these animal studies have been useful in uncovering mechanisms of IR in specific tissues, their clinical applicability is less evident because complete deletion of a gene preferentially in one tissue does not occur in humans. Nonetheless, these experiments have provided persuasive evidence that metabolic or signaling changes in one tissue can have systemic effects by influencing insulin activity in other organs, a phenomenon that has been well-validated by clinical findings [50, 51].

7. Impaired insulin signaling in insulin resistance

Over the past 40 years, much research has resulted in a precise understanding of the insulin signaling system, which mediates the insulin’s physiological activities. One popular theory is that IR is caused by a defect in one or more of these signaling components. Another viewpoint is that IR is only caused by a shift in metabolic flux. For example, since the 1960s fatty acids have been proven to impede cells’ ability to utilize carbohydrate by allosterically modifying crucial rate-limiting steps in carbohydrate metabolic pathways. Several pieces of evidence, however, refute this claim. IR can be seen in cells or tissues long after the animal tissues have been removed, implying that changes that contribute to IR are long-lasting and cannot be explained by the acute action of a systemic factor. Fatty acids decrease the insulin-dependent translocation of GLUT4 to the plasma membrane and limit glucose uptake, there is no indication that this inhibition is caused by an allosteric change of GLUT4. Finally, IR can persist even after significant changes in dietary intake and after changes in metabolic state induced by pharmacological interventions. Thus, based on this information, it is reasonable to believe that IR is caused by an alteration in insulin signaling, although the exact location of the defect in the insulin signaling pathway remains unknown. Many essential components of the insulin signaling system have been identified. These components are divided into two parts: (i) the proximal part, which represents the core canonical signaling pathway, which includes the insulin receptor, IRS, PI3K and AKT; and (ii) the distal part, which includes TBC1D4, GSK3 (glycogen synthase kinase-3) and PDE3B (phosphodiesterase 3B). IR has been linked to defects in proximal insulin signaling system, that are associated with cellular stress. Many of the intracellular stressors discussed in the next sections activate a variety of intracellular Ser/Thr kinases, including novel PKCs (protein kinase C), JNK (c-Jun amino-terminal kinase), mTOR (mammalian target of rapamycin)and S6 kinase, which phosphorylate either the insulin receptor or the insulin receptor-related protein (INSRR). This could be a negative-feedback route that inhibits insulin signaling, according to the theory. However, as it will be discussed later, mounting evidence suggests that proximal insulin signaling system is unaffected in IR, implying that IR is caused by abnormalities in distal components of the insulin signaling network [52, 53, 54].

8. Insulin resistance and insulin signaling at the proximal level

The current focus on proximal insulin signaling abnormalities as a cause of IR stems from research into rare, monogenic severe types of IR that were discovered to be caused by mutations in the insulin receptor gene or by the development of insulin receptor blocking antibodies. Because of the superficial parallels between these rare conditions and T2D, it is reasonable to conclude that both diseases are caused by abnormalities in insulin receptor function, with the degree of receptor failure varying only slightly. Despite early enthusiasm for this theory, subsequent research found that IR in most forms of T2D was caused by neither impaired insulin receptor activity nor changes in the expression or quantity of insulin receptors. Insulin-binding experiments in rat adipocytes found that only 2.4% of total insulin receptors are required for a full biological response, implying that metabolic cells like muscle, fat and liver cells have an abundance of insulin receptors; this finding became known as the “spare insulin receptor” hypothesis. Insulin-mediated glucose uptake is reduced in insulin-resistant skeletal muscle cells and adipocytes. Since a slight decrease in the number of insulin receptors could only diminish insulin sensitivity and not the maximal insulin response [54, 55, 56, 57, 58]. While some studies contradict the “spare insulin receptor” hypothesis, recent genetic studies in mice support the idea that insulin signaling is preserved when the number of insulin receptors is reduced: mice with heterozygous loss of the insulin receptor had normal glucose and insulin tolerance and no impairment in AKT signaling in muscle or adipose tissue [59, 60, 61, 62].

The concept of spare insulin receptors shifted focus to a “postreceptor defect”, which is represented by defects in signaling downstream intermediates of the insulin receptor as the cause of IR [57, 58, 63]. Loss-of-function mutations in a number of signaling genes—including TBC1D4, AKT2, and IRS1 in humans—have been linked to severe forms of IR and T2D; moreover, cancer treatments that block PI3K or AKT have been linked to IR and T2D in humans. IR is caused in mice by targeted deletion of these genes. In addition, IR results in reduction of skeletal muscle AKT phosphorylation in response to insulin stimulation [34, 64, 65].

Given evidence of “spareness” for IRS, PI3K and AKT, the possibility that abnormalities in proximal insulin signaling might be responsible for IR has to be questioned, in the same way that the “spare receptor” theory has to be questioned. Homozygous deletion of AKT2, the most prevalent AKT gene isoform, resulted in a 90% reduction in insulin-stimulated AKT phosphorylation, but with no discernible defect in phosphorylation of the AKT substrate, or protein synthesis in response to insulin. In this situation, there was a tiny quantity of AKT1 expression that was not influenced by the gene deletion and was enough to deliver a completely functional message as response to insulin [66, 67]. Similarly, whereas AKT2 accounts for 85% of total AKT in the liver, its ablation does not result in significant glucose intolerance because the remaining AKT1 compensates for this defect [68]. The insulin dose-response curve in adipocytes, where the curve for AKT phosphorylation is “shifted to the right” compared to that for AKT substrate phosphorylation or insulin action, indicates that partial phosphorylation of AKT is sufficient for maximal biological responses, providing additional evidence for “spareness” in proximal insulin signaling network. At “normal” insulin concentrations, phosphorylation of AKT substrates requires only 1% of the entire AKT pool to be activated [69, 70, 71]. Furthermore, AKT phosphorylation is reduced in muscle from T2D patients, while downstream substrate phosphorylation is unaffected. Importantly, studies in animals fed a Western diet have indicated that IR begins before any detectable insulin signaling defect. Only 42 days of Western diet feeding resulted in reduced insulin-stimulated AKT phosphorylation, but TBC1D4 phosphorylation remained normal. As a result, minor changes in phosphorylation of proximal insulin signaling components may result in insulin sensitivity, but they are unlikely to result in a reduction in the maximal physiologic response [53].

Thus, how can the predominance of abnormalities in proximal insulin signaling components observed in diverse IR models, such as lower AKT phosphorylation, be reconciled? It is possible that these defects are a result of defective glucose metabolism rather than the cause. This could be a direct effect secondary to compensatory hyperinsulinemia, a typical hallmark of IR (since persistent hyperinsulinemia can lead to degradation of proximal insulin signaling components); alternatively, it may be a cell-autonomous effect due to a reduction in AKT phosphorylation as a result of defective glycolysis. Many studies used insulin-stimulated AKT phosphorylation in mice (sometimes in response to a maximal, pharmacological dosage of insulin) as an indicator of insulin sensitivity [72, 73, 74, 75].

However, under physiological settings such as the response to a meal (with minimal insulin release), AKT phosphorylation is barely detectable, due to the non-linearity between AKT phosphorylation and phosphorylation of its substrates. As a result, when evaluating the physiological importance of insulin signaling, it is critical to look at the phosphorylation of a variety of AKT substrates to determine if there is a major deficiency in “AKT activity” in vivo. These findings suggest that a minor impairment in proximal insulin signaling network is unlikely to account for the significant reduction in insulin-stimulated glucose uptake observed in patients with T2D. Furthermore, these findings underline that lower AKT2 phosphorylation should not be used as a direct marker or even as a proxy measure of IR [71].

Negative feedback loops originating from Ser/Thr kinases that phosphorylate and limit the action of IRS proteins have also been proposed as a cause of IR. This theory is refuted by a number of studies. Since Platelet-derived growth factor (PDGF) by-passes these proteins to activate glucose uptake, mice bred to overexpress PDGF receptor (PDGFR) in muscle presented an ideal model to explore whether deficiencies in insulin receptor or IRS were implicated in experimental IR. In these mice, PDGF treatment resulted in increased glucose uptake in muscle [76]. Notably, when PDGFR transgenic rats were fed a Western diet, muscle glucose uptake in response to PDGF was decreased to the same degree as insulin-mediated uptake. This refutes a role for inhibitory Ser/Thr phosphorylation of the insulin receptor or IRS as a cause of IR, indicating that the deficiency in glucose uptake or IR does not involve the insulin receptor or IRS [8, 53, 54, 77, 78].

Furthermore, in mice, targeted mutation of one of the major putative inhibitory sites in IRS1 (Ser307), deletion of potential mediators of IR, such as PKC (which is reported to phosphorylate insulin receptor), and pharmacological blockade of key negative feedback pathways, such as mTOR (which is activated by insulin signaling and inhibits signaling by phosphorylating IRS through a negative feedback mechanism) [78, 79, 80, 81].

Finally, investigations in humans with IR or T2D revealed that insulin-stimulated muscle glucose uptake is reduced by 50–100% even at maximum insulin dosages [82, 83, 84, 85], with no change or reduction in AKT phosphorylation [86, 87, 88]. Only a few of these studies addressed the mechanism of AKT substrate phosphorylation in depth, and those that did found no deficiency or poorly linked with IR. These findings support the theory that the proximal insulin signaling network in human tissues has enough “spareness” to overcome even a moderate deficiency in AKT phosphorylation [87, 88, 89], and that lowered AKT phosphorylation is adequate to ensure a normal signal transduction. As previously stated, faulty proximal insulin signaling is most likely a result of IR rather than a cause of IR [90].

9. GLUT4 and insulin resistance

Insulin stimulates the transfer of intracellular GLUT4 storage vesicles to the cell surface, resulting in glucose uptake in skeletal muscle cells and adipocytes (Figure 1) [91, 92, 93, 94]. Insulin-dependent GLUT4 translocation has been linked to IR in both skeletal muscle and adipose tissue. This decrease in GLUT4 availability at the plasma membrane causes a reduced glucose uptake, which can lead to other IR-related consequences like reduced AKT phosphorylation, protein synthesis defects, and increased lipolysis [72, 95, 96]. GLUT4 does not show spareness, unlike proximal insulin signaling components such as IRS1 and AKT. The fact that heterozygous GLUT4 gene-knockout mice acquire metabolic disease exemplifies this concept [97].

Figure 1.

Translocation of glucose transporter type 4 (GLUT4) from GLUT4 storage vesicles (GSVs) to the plasma membrane of normal adipocytes and skeletal muscle cells (a). This process is altered in conditions characterized by insulin resistance (b).

However, while GLUT4 levels are lowered by 50% in human adipose tissue from patients with T2D, such levels remain unaltered in skeletal muscle, implying that GLUT4 levels cannot explain IR development in skeletal muscle [98]. Despite normal GLUT4 levels, insulin-stimulated GLUT4 translocation to the cell surface in skeletal muscle is faulty in both individuals with T2D [92] and in several rodent models of IR [99, 100]. Importantly, while exercise-modulated GLUT4 translocation to the cell surface is unaffected [101], the impairment in muscle GLUT4 trafficking in T2D is insulin signaling-specific. Insulin and exercise both cause GLUT4 translocation to the cell surface from discrete intracellular compartments [102].

The ultimate defect that defines IR is the impaired GLUT4 translocation to the plasma membrane. However, it is unknown how the numerous potential intracellular IR mediators mentioned later affect GLUT4 trafficking. Three options are discussed here. First, GLUT4 translocation requires that GLUT4 is localized in the appropriate intracellular compartment, the so-called GLUT4 storage vesicles (GSVs); GLUT4 targeting to GLUT4 GSVs has been hypothesized to be altered in IR [91, 100]. However, whereas this would likely result in GLUT4 degradation, GLUT4 levels in skeletal muscle from patients with IR remain unaffected. Second, given the importance of protein phosphorylation in insulin action [101, 102, 103, 104], it is possible that the defect is caused by a distal component of the insulin-regulated phosphorylation network such as TBC1D4, which regulates GLUT4 trafficking, although there is no convincing evidence for defective TBC1D4 phosphorylation in IR [105]. TBC1D4 is unlikely to be the only AKT target causing GLUT4 translocation, as cells lacking TBC1D4 still have some insulin-sensitive glucose transport [106]. Recent phosphoproteomics studies have revealed the existence of a wide range of insulin-responsive phosphoproteins in metabolic cells, allowing for the identification of insulin signaling targets in the distal part of the insulin signaling pathway that may be involved in the development of IR [104]. Indeed, IR is associated with massive alterations in the architecture of the entire insulin signaling pathway, according to examination of muscle cells from T2D patients [107]. Finally, a direct alteration of GLUT4 or a defect in a yet undiscovered protein that interacts with GLUT4 could cause the abnormalities in GLUT4 trafficking. This could include carbonylation and oxidation-induced inactivation of GLUT4, which have been observed in humans as a response to short-term overnutrition [108]. Protein carbonylation is linked to H₂O₂ production, lipid peroxidation and IR, suggesting a link between such molecular processes and the development of IR [109].

10. Adipose tissue and insulin resistance

While IR is regularly seen in lean first-degree relatives of patients with T2D, it is also found in many lean “healthy” individuals, suggesting that IR is more common than previously thought. In this regard, dietary habits, physical activity level and genetics are important factors that can significantly contribute to IR. Adipose tissue makes a significant contribution to the development of IR. Limitations in peripheral adipose tissue storage capacity and expansion in response to over nutrition (as it occurs in overweight and obesity) lead to increased circulating lipids, subsequent lipid accumulation in non-adipose tissues (ectopic lipid in liver, skeletal muscle, heart, and pancreas) and development of lipid-induced IR and metabolic derangements [110, 111]. Because of this, and since there is a clear link between IR and increased adipose tissue mass, we will discuss the role of adipose tissue mass and lipotoxicity as significant drivers of IR, as well as the emerging mechanisms by which adipocytes contribute to systemic IR.

10.1 Adipose tissue dysfunction

IR in adipocytes could be the first step in the progression of adipose tissue dysfunction, similar to IR in muscle and liver. In adipocytes from first-degree relatives of patients with T2D, there is a low expression of markers of insulin sensitivity such as GLUT4 and adiponectin (a crucial systemic insulin-sensitizing adipokine produced by adipose tissue), supporting this theory [112]. Furthermore, adipocyte hypertrophy (increase in adipocyte size) appears to precede T2D onset in Pima Indians, a group of Native Americans with a high incidence of IR and T2D [113]. Additionally, mouse models with adipose-specific IR also have IR in their muscle and liver. Notably, IR in the muscle of adipose-specific Glut4 gene-knockout mice was only present in vivo but not when muscles were isolated and assessed in vitro, implying a role for systemic factors (which did not include circulating FFAs or inflammatory cytokines) in the progression of IR from adipose tissue-specific pathology [114, 115].

Human genetic research has also suggested that adipose tissue plays a significant role in IR. Studies in identical twins or first-degree relatives of T2D patients have shown that inheritance has a substantial influence in IR and T2D [116]. More than 250 genetic loci have been linked to T2D so far, however they only account for 25% of T2D heritability [117]. While these investigations have generally discovered genes linked to beta-cell function and insulin secretion, deeper analysis of phenotypes more closely aligned with IR have begun to uncover genetic drivers of IR in other organs. Surprisingly, several of these drivers are involved in the function of adipose tissue [118]. Although subclinical lipodystrophy is a rare cause of severe IR, it has been suggested that milder forms of lipodystrophy are responsible for IR in general, supporting a model in which excessive lipid spillover into circulation is a proximal, mechanistic cause of altered insulin action. Specifically, when the individual’s capacity to store lipids in adipose tissue has been exceeded, lipid spillover into circulation leads to elevated plasma FFAs and triglyceride levels, which result in increased ectopic storage of these molecules in non-adipose tissues—such as liver and skeletal muscle—and subsequent metabolic derangements via lipotoxicity (lipid-induced toxicity). Surprisingly, genes in the insulin signaling system linked to IR (IRS1 and GRB14) are also linked to familial partial lipodystrophy [119].

PPARG (Peroxisome Proliferator-Activated Receptor Gamma, a master positive regulator of adipogenesis) and CCDC92, DNAH10, and L3MBTL3 (regulators of adipocyte differentiation) were among the 53 loci discovered in a study employing an integrated genomic approach to find genes related to IR. Thiazolidinediones are insulin-sensitizing peroxisome proliferator-activated receptor gamma agonists that are used in the treatment of T2D and act by promoting adipogenesis and adipose tissue growth (through cell size and cell number increase or adipocyte hypertrophy and hyperplasia) [119]. The availability of additional lipid storage induced by thiazolidinediones may therefore promote insulin sensitivity by alleviating lipotoxicity [120]. These drugs also improve insulin sensitivity in first-degree relatives of T2D patients, implying that adipose tissue hypertrophy and “unhealthy” lipid storage are critical regulators of insulin action and contributors to IR [121].

Adipose tissue’s primary function is to store fat and release it into circulation when needed, and it has the unique capacity to expand in response to nutrient overload. Lipids can be released into the bloodstream when the adipocyte capacity to store lipids has been exceeded [39]. There is compelling evidence that the accumulation of excess lipids in non-adipose tissues (e.g., skeletal muscle and liver), known as lipotoxicity (a.k.a. lipid-induced toxicity), plays a role in the development of muscle and liver IR [122]. As a result, studies aimed at understanding the cause and magnitude of increased circulating lipid levels in IR are now being pursued. Furthermore, intracellular lipid accumulation in cells and tissues—including pancreatic beta cells and liver—has been linked to the onset of cellular dysfunctions, such as secretory abnormalities and inflammation (Figure 2). Elevated circulating FFA levels have been linked to IR, and this has been proposed as a possible cause of lipotoxicity [123].

Figure 2.

Excessive adipocyte lipid storage in response to overnutrition, resulting in adipocyte hypertrophy, inflammation and increased release of free fatty acids (FFAs) into circulation, leading to ectopic fat accumulation, lipotoxicity and development of insulin resistance in non-adipose tissues, such as liver and skeletal muscle.

In humans and animals, lipid infusion causes muscle IR and enhanced hepatic gluconeogenesis, the latter attributable to changes in metabolic fluxes rather than to fat accumulation [123, 124, 125]. Furthermore, animals with increased circulating FFA levels due to increased lipolysis develop muscle and hepatic IR, whereas obese mice with reduced fat cell lipolysis are protected from glucose intolerance [126]. It is worth noting that, as discussed elsewhere [127], circulating FFA levels in patients with IR or T2D usually are not elevated. However, there are several confounders in this measurement, including the wide range of FFA levels in healthy adults and the fact that fasting FFAs are typically assessed rather than the more relevant postprandial FFAs. Nonetheless, there is strong evidence that serum FFA levels are elevated in first-degree relatives of patients with T2D [127, 128], implying that this elevation represents an early stage of the disease. It is unclear if the rise in circulating FFA levels is related to defects in insulin-mediated regulation of lipolysis, to alterations in fat storage capacity, or to an increase in adipose tissue mass without defects in lipolysis. Lipolysis per gram of adipose tissue mass is considerably lower in obese subjects, suggesting that enlargement of adipose tissue mass is the principal driver of abnormal FFA homeostasis [129].

Adipose tissue can grow in size by either hypertrophy, which involves the enlargement of existing adipocytes, or hyperplasia, which involves the generation of new fat cells from preadipocytes via adipogenesis, resulting in an increase in the number of tiny adipocytes [130]. Subcutaneous adipose tissue is more expandable than visceral adipose tissue in humans, whereas the opposite is true in C57BL/6J male mice [131]. Female mice, interestingly, show expandability of both adipose tissue depots in response to Western diet feeding, suggesting that sex hormones and other sex-dependent elements play a role in this process [131, 132]. Pathological adipose tissue expandability under situations of overnutrition, particularly adipose tissue hypertrophy, has got a lot of attention as a likely cause of IR. Indeed, first-degree relatives of patients with T2D have greater amounts of hypertrophic adipose tissue, implying that changes in cell size—presumably due to defective adipogenesis—represent an early event in the pathophysiology of T2D. Hypertrophic large adipocytes are linked to poor metabolic outcomes when compared to hyperplastic adipocytes [39, 133], which have been shown to confer metabolic health in obesity [134, 135, 136]. More importantly, hypertrophic adipocytes may contribute to an increase in circulating FFA levels due to their reduced FFA storage capacity. Reduced preadipocyte differentiation, diminished de novo lipogenesis or FFA uptake in hypertrophic adipocytes, and/or reduced adipose tissue expandability due to physical limits on expanding cell size may all contribute to decreased lipid storage capacity by the hypertrophic adipose tissue. Furthermore, adipogenesis abnormalities may result in decreased generation of beige adipocytes, thereby contributing to higher circulating FFA levels; indeed, beige adipocytes differentiate from a subpopulation of progenitors resident in white adipose tissue and have the ability to promote FFA oxidation through thermogenesis [137, 138].

10.2 Circulatory factors released from adipocytes

Adipose tissue secretes a number of factors (termed “adipokines”) into the bloodstream that regulate energy metabolism. These factors include cytokines, hormones, extracellular matrix proteins, as well as growth and vasoactive factors. The type of adipose tissue expansion has been demonstrated to impact the secretion of certain of these factors under IR conditions. Since the discovery of leptin as the first adipokine [139], a growing list of adipose tissue-secreted factors implicated in IR has been discovered, with roles in IR that are either protective or causative [20, 140].

Leptin, for example, regulates whole-body energy metabolism by acting on feeding centers in the brain to suppress food intake and increase energy expenditure; leptin deficiency causes obesity, hyperinsulinemia, IR and impaired glucose homeostasis [141]. Adiponectin, another well-known adipokine secreted from adipocytes, has been linked to regulation of cell insulin sensitivity. In humans, circulating adiponectin levels are favorably linked with whole-body insulin sensitivity; additionally, physical training increases circulating adiponectin levels and the expression of its receptors in muscle, which may mediate the improvement of IR in response to exercise [142]. Surprisingly, small and subcutaneous adipocytes release more adiponectin than visceral or large adipocytes [143]. Anti-atherogenic, anti-inflammatory, and insulin-sensitizing effects of adiponectin have also been discovered [144]. It is worth mentioning, however, that while adiponectin’s positive benefits in rats are outstanding, the role of this adipokine in humans is less obvious, and Mendelian randomization studies on adiponectin’s relationship with metabolic disease in humans have generated inconsistent results [145, 146].

Adipocytes release a variety of substances, including metabolites like lipids and extracellular vesicles that contain proteins and microRNAs. Branched fatty acid esters of hydroxy fatty acids (FAHFAs) are a unique class of lipids synthesized in adipocytes that have been shown to increase insulin sensitivity and reduce inflammation; accordingly, individuals with IR have lower circulating FAHFA levels [147]. As a result, further research into this metabolite class is necessary. Adipocytes, for example, release tiny lipid-encapsulated extracellular vesicles into the bloodstream. These vesicles may alter metabolic processes in other target tissues, such as the liver, according to increasing evidence based on mouse studies. MicroRNAs represent one of the components found in extracellular vesicles that have been linked to this mechanism. While investigations on microRNAs are intriguing, many fundamental aspects about the mechanism of their controlled secretion and their tissue targeting and entry into target cells remain unknown [148, 149].

Many circulating factors are also produced by other adipose tissue-resident cells, such as immune or vascular cells, rather than by adipocytes themselves (the so-called “stromal vascular fraction” of adipose tissue). Some of these adipokines, such as tumor necrosis factor (TNF), resistin or vascular endothelial growth factor (VEGF), are important regulators of tissue homeostasis and may be secreted as a result of adipose tissue enlargement during the development of obesity [150]. Nonetheless, inflammatory cytokines have been widely suggested as possible IR-inducing adipokines, and several of these factors have significant proinflammatory activities [151, 152, 153].

11. Inflammation and insulin resistance

It is now well recognized that cells of both innate and adaptive immunity, notably macrophages, infiltrate hypertrophic adipose tissue in most obesity models, and that this is accompanied by a loss of immunosuppressive regulatory T cells in visceral fat depots [154]. When macrophages in adipose tissue are activated in response to overnutrition, they polarize towards a proinflammatory phenotype and release cytokines that may trigger IR in all metabolic tissues [155]. Diet-induced obesity in mice and humans is unmistakably linked to elevated levels of systemic inflammatory markers, including C-reactive protein (CRP) and enhanced immune cell infiltration of adipose tissue and other organs [156]. In addition, inflammatory cytokines, such as TNF, can elicit IR in metabolic tissues when infused in humans [157]. Although macrophage infiltration into hypertrophic adipose tissue is well documented, the role of inflammation in IR is convoluted and controversial; for example, inflammatory markers are not elevated in first-degree relatives of T2D patients [158]. Furthermore, in Western diet-fed mice, tissue IR occurs before the adipose tissue infiltration by a considerable number of immune cells, and genetic or pharmacological anti-inflammatory methods do not prevent the development of Western diet-induced IR [159, 160]. The administration of a neutralizing antibody against interleukin-1 (IL-1), a proinflammatory cytokine implicated in IR, to approximately 4000 patients with T2D and almost 5000 subjects with prediabetes resulted in a significant decrease in CRP levels, as well as in a modest positive effect on cardiovascular outcomes, but without reducing the frequency of new-onset T2D or increasing fasting glucose levels [161, 162, 163].

Overall, evidence suggests that adipose tissue infiltration by macrophages is unlikely to be the major cause of IR. Macrophage infiltration into the growing adipose tissue may affect its function in addition to systemic inflammation, but the exact impact of this infiltration is unknown [164]. Anti-inflammatory macrophages (M2), on the other hand, have been shown to promote angiogenesis and preadipocyte differentiation, which aids adipose tissue expansion [165, 166]. The diversity of cytokines, their concentrations, and the timing of their release into the tissue are likely to have a considerable impact on the final biological response, contributing to the observed inconsistent results. The ability of genetically induced adipocyte IR to elicit adipose tissue inflammation adds to the growing body of evidence that inflammation may be a consequence rather than a cause of IR. Hyperinsulinemia has been shown to induce adipose tissue inflammation, implying that the latter is a late event in the IR pathophysiology [30].

12. Intracellular mediators and insulin resistance

Many extrinsic stimuli and genetic alterations can antagonize insulin action in vitro and in vivo, and their investigation has led to the identification of a series of molecules as putative intracellular mediators of IR. In the sections that follow, we will look at the role of a few intracellular components that have got a lot of attention as drivers of IR. It is worth noting that mechanisms of action of these components are not well-established yet, and further research is needed to better understand their role in IR development.

12.1 Accumulation of ceramides

Ceramides have been implicated as IR mediators by a large body of research. Ceramides are essential precursors of most of the complex sphingolipids localized in lipid bilayers, including sphingosine, sphingomyelins, and glucosylceramides. Ceramides accumulate in muscle, liver and adipose tissue of subjects with IR, according to human and animal studies [167, 168, 169, 170]. In insulin-resistant tissues, the levels of 16- or 18-carbon chain-length ceramides are raised, whereas the levels of other chain-length ceramides are not consistently changed [171, 172]. Indeed, in adipose tissue from obese subjects, the level of ceramide synthase isoform 6 (CERS6), which synthesizes C16 ceramide, is raised [171]. Surprisingly, the presence of a double bond in the ceramide backbone promotes IR, as ablation of the enzyme responsible for its formation (dihydroceramide desaturase 1) abrogates IR [173]. While it is unclear how specific extrinsic mediators of IR cause increased intracellular ceramide levels, it is possible that excess FFAs serve as a crucial substrates for ceramide biosynthesis [174, 175, 176].

Another theory connects intracellular ceramide to levels of circulating adiponectin. Ceramidase activity is found on adiponectin receptors, and lower adiponectin levels in IR may lead to decreased ceramidase activity and, consequently, to higher ceramide levels [177, 178]. AMP-activated protein kinase (AMPK), a major metabolic sensor that regulates mitochondrial biogenesis and metabolism, is activated by adiponectin, potentially regulating ceramide via increased mitochondrial lipid oxidation [179]. By using small-molecule inhibitors or genetic deletion of ceramide-producing enzymes to neutralize ceramide accumulation in metabolic organs, researchers were able to reverse or prevent IR induced by the Western diet in C57BL/6 mice with diet-induced obesity [122]. The relationship between ceramide and decreased insulin action is not univocal, as it is for many possible intracellular mediators of IR. In fact, ceramide suppresses AKT activity, although IR is unlikely to be caused by defects in AKT, which is a proximal arm of insulin signaling (as it has previously been mentioned). Ceramide could be part of a wider, IR-related stress mechanism that leads to mitochondrial dysfunction and to the production of reactive oxygen species (ROS). Ceramide has also been connected to the release of pro-inflammatory cytokines, which have been involved in IR, as it has previously been described [180, 181].

12.2 Accumulation of diacylglycerol (DAG)

Another popular theory for the cause of IR is the accumulation of diacylglycerols (DAGs) in muscle, adipocytes and liver, as a result of elevated serum FFA levels [182, 183]. Protein kinase C (PKC) is recruited to the plasma membrane by DAGs, where it phosphorylates and inhibits insulin receptor kinase activity. While it is quite plausible that DAG levels are elevated in insulin-resistant tissues, a scenario in which DAG-dependent phosphorylation of the insulin receptor is the major cause of IR raises a number of questions. Given the “spareness” of the insulin receptor and proximal signaling intermediates, it is doubtful that IR is caused solely by abnormalities in these components, at least in muscle. In contrast to other insulin-responsive proteins, the stoichiometry of insulin receptor phosphorylation at the region implicated in DAG-mediated IR is low and not detectable by conventional phosphopeptide analysis [79, 104, 184]. PKC deletion in the liver had little effect on whole-body insulin sensitivity in mice, indicating against PKC being a key target of DAG-induced IR in that tissue [79, 104], although this has since been challenged by studies in rats showing that acute knockdown of PKC in the liver protected animals from IR. However, antisense oligonucleotides were delivered systemically, which could target PKC expression in other organs. While technical differences between these studies and others have been suggested as a reason for the discrepancies observed [183], there appears to be enough disagreement about the role of the DAG-PKC-insulin receptor pathway in IR to warrant further investigation and, in particular, validation by multiple independent laboratories [185].

12.3 Mitochondrial dysfunction and reactive oxygen species (ROS)

IR has been linked to a decrease in mitochondrial function. Mitochondrial dysfunction is a term that has been used to describe a variety of mitochondrial phenotypes, including decreased respiratory capacity and ATP production, decreased number of mitochondria, accumulated mitochondrial damage due to defects in mitophagy, and altered mitochondrial morphology caused by changes in mitochondrial fission-fusion dynamics. Many of these alterations are also linked to an increase in mitochondrial ROS generation, which has long been linked to IR [186, 187, 188].

It is not unexpected that IR is linked to higher levels of reactive oxygen species (ROS). This is due to the fact that IR is frequently accompanied by a positive energy balance, which leads to an excess of reducing equivalents (NADH and FADH2). This determines a reductive stress on the mitochondrial respiratory electron transport chain, which invariably results in the formation of free electrons and, as a result, in an increased production of various forms of ROS [189]. Furthermore, enhanced ROS production has been found in response to a variety of extracellular stressors linked to IR, including inflammation [190]. Superoxide, H₂O₂, reactive nitrogen andoxidized lipids accumulate in insulin-resistant cells or tissues, and a mitochondria-targeted small molecule transiently produced mitochondrial ROS in muscle and adipocytes, causing IR. As a result, attempts to reduce ROS levels have been proven to reverse or prevent IR in mice [191, 192, 193, 194].

Reduced levels of coenzyme Q (CoQ) have recently been linked to IR in humans [44]. In mitochondria, CoQ is a key component of the electron transport chain, transferring electrons from complex I or II to complex III. Furthermore, unlike complex I, CoQ receives electrons directly from the electron-transferring flavoprotein, and this is unrelated to proton pumping or mitochondrial membrane potential, relying only on the availability of oxidized CoQ. Reduced CoQ accumulates, causing reductive stress in complex I, complex II and other dehydrogenases that feed electrons into the CoQ pool, resulting in increased ROS production [195]. As a result, lowering the total CoQ pool [44] will most likely lower the ROS production threshold at a given energy demand-supply ratio. It is also worth noting that FFA oxidation produces far more ROS than carbohydrate oxidation [195]. This is because the electron-transferring flavoprotein feeds a higher proportion of reducing equivalents straight into the CoQ pool during FFA oxidation. Therefore, as lipid metabolism increases, the supply of reducing equivalents outnumbers the demand, lowering the ratio of oxidized to reduced CoQ. This is likely worsened when total CoQ levels are low, as seen in IR [44], resulting in reductive stress and increased ROS production. The mechanism that regulates CoQ levels in IR is unknown. Intriguingly, statins, which are commonly used as cholesterol-lowering drugs, have been linked to IR in humans [196], with the possibility that this relationship is related to the statin-induced reductions in CoQ biosynthesis [44]. Unfortunately, given the low bioavailability of CoQ , oral supplements, which are frequently recommended as an antioxidant strategy, are unlikely to be successful in replenishing the mitochondrial CoQ pool in patients with IR or even in individuals who take statins. Other hazardous intermediates can be generated, in addition to ROS, as a result of mitochondrial respiration abnormalities. Acylcarnitine is an example of incompletely oxidized lipids produced by lipid overload. Acylcarnitine has been reported to accumulate in IR, indicating a deficiency in or an overabundance of the mitochondrial oxidative ability. In this regard, it has been postulated that lipid-induced mitochondrial stress mediates IR, although the exact mechanisms remain elusive [197].

12.4 Insulin resistance associated with stress pathway

Many of the pathways involved in IR pathophysiology, such as those involving ceramides, DAGs or ROS, are now being linked as part of what we call an “intracellular IR stress pathway”, according to new evidence. Ceramide, for example, promotes mitochondrial fission and ROS production [198, 199]. In subjects with IR, the quantity of mitochondrial ceramide is higher, and enzymes involved in ceramide biosynthesis have been found in mitochondria [185, 200, 201, 202, 203]. Ceramide is involved in apoptosis triggered by mitochondria in some cells, including insulin-producing pancreatic beta cells, but not in other metabolic tissues [204, 205, 206]. Ceramide also contributes to endoplasmic reticulum stress, which frequently co-occurs with mitochondrial stress and has been proposed as a driver of IR, where endoplasmic reticulum stress causes JNK activation, which, as previously described, affects the insulin signaling pathway via inhibitory IRS1 Ser/Thr phosphorylation [204, 205, 206]. Ceramide also induces PKC, a DAG-regulated kinase, to translocate to mitochondria, activating it and causing mitochondrial damage through an unknown mechanism [207]. Ceramides and DAGs are also biochemically connected; sphingomyelin synthase, for example, converts ceramide to DAG. Finally, in rats, reducing mitochondrial ROS levels with mitochondria-targeted catalase improved insulin sensitivity while lowering muscle DAG levels [208]. The potential connection of many of these suspected IR-causing elements into a dynamic network should help to resolve some of the current debates on this topic.

12.5 Signals from the mitochondria

Despite the interest in mitochondrial dysfunction in IR, it is unclear how intramitochondrial signals, like ceramide or ROS, may cause changes in insulin action, such as impaired GLUT4 translocation, which occurs mostly in the cytosol. The mitochondrial permeability transition pore (mPTP), a multiprotein complex located in the inner mitochondrial membrane, is a promising candidate for “inside-out” mitochondrial signaling because it opens under conditions of mitochondrial stress—most notably involving mitochondrial ROS—to allow molecules to be transported from mitochondria to the cytoplasm [209]. In L6 myotubes, inhibiting mPTP prevented ceramide- or palmitate-induced IR, and mice with defective mPTP opening were protected from diet-induced IR in skeletal muscle [210]. Although at least a part of the impact is attributable to its anti-obesogenic effect, deletion of mPTP in the liver has been shown to protect mice from liver steatosis and IR [211].

13. Conclusions and perspectives

The rising frequency of IR, as well as its crucial involvement in a variety of diseases, demands a greater understanding of the processes behind IR pathogenesis and how they interact with genetics and various surroundings, notably dietary factors. We have attempted to offer an overview of the main mechanisms hypothesized to contribute to IR in this chapter, highlighting both supportive and non-confirmatory evidence when appropriate. Many of the molecules and processes studied as causative in IR, in our opinion, function in series as a connected pathway or a loop rather than acting independently. Unfortunately, there has been a recent trend to describe IR as a dysfunction of insulin signaling, regardless of whether a simultaneous examination of insulin action on glucose metabolism has identified a defect in the latter process. We feel that this method has produced significant problems in the field, and we wish to send a message that simple, unitary errors in proximal insulin signaling are unlikely to be a major cause of IR. Rather, IR develops as a result of a variety of challenges that disrupt cellular homeostasis, resulting in cellular stress that can have a variety of deleterious consequences on insulin signal sensing and transmission.

The difficulty in translating findings from model organisms to humans, particularly in terms of differentiating IR causation from the multiplicity of effects, is a key roadblock in investigating the underpinnings of IR. By discovering causal genetic variants, human genetics holds a lot of promise for tackling this problem. However, genetics can only explain a portion of the pathophysiology of IR. Environmental variables play a crucial role in determining susceptibility to IR development and interact with genetics. Furthermore, the heterogeneity of metabolic diseases like T2D demands detailed phenotyping. Focusing on phenotypes that has better track with IR has proven difficult to achieve in the large cohorts. It is required to identify genetic polymorphisms that only explain a small proportion of disease in the human population. Despite these limitations, a number of genetic loci linked to human IR have been discovered, leading to a renewed focus on adipose tissue enlargement as a critical aspect of IR. However, since IR is a systemic condition, we expect future investigations to discover variations in genes governing multiple cellular processes throughout organs as linked to IR pathophysiology.

A more systematic approach involving large-scale omics to analyze the molecular landscape rather than relying on individual components as causal would be required to gain a better understanding of IR. Moreover, while knockout mice have been critical in characterizing the biochemistry of insulin action, they have also sparked numerous debates. One reason for this is that gene deletions typically result in adaptive processes that are difficult to define and may have limited physiological value, as indicated in a recent study with muscle-specific Akt gene-knockout mice [58]. In animals with both insulin and insulin-like growth factor 1 (IGF-1) receptors removed in muscle, similar adaptation mechanisms have been reported [212].

The ultimate goal of understanding mechanisms behind IR is to develop new, effective anti-IR therapeutic strategies. One key point to consider in this endeavor is whether such therapies would be beneficial if the initial insult—nutritional overload—persists. While IR is typically considered abnormal, as it is linked to a variety of disease outcomes, it is also a prevalent component of many normal physiological states, such as starvation, pregnancy, and hibernation. IR is believed to play a protective or adaptive role in such conditions, supporting survival by saving glucose for the brain and other vital tissues and organs or for the fetus during pregnancy. It is possible that IR has a similar function in metabolic disease. Since the primary metabolic tissues are frequently exposed to potentially harmful quantities of nutrients, IR could be a protective mechanism that helps to prevent tissue nutrition overload [190]. However, this comes at a price, namely concomitant hyperinsulinemia, which is the most serious pathophysiological consequence of IR. Insulin-sensitizing drugs may thus act as a “circuit breaker”, reducing hunger, inflammation and IR by suppressing hyperinsulinemia. As a result, we believe there is still a strong need to describe the molecular characteristics that drive IR in order to identify appropriate targets that can break the IR vicious cycle.

Acknowledgments

The authors are also thankful to Guru Nanak Dev University (Amritsar, Punjab, India) for providing various facilities to carry out the present work.

Conflict of interest

The authors declare no conflict of interest.