Role of insulin action in the pathogenesis of diabetic complications

Abstract

Among the various pathological conditions associated with type 2 diabetes, insulin resistance has long been reported to be a potent risk factor for diabetic complications. The liver, skeletal muscle, and adipose tissue are the major organs of action of insulin in systemic glucose metabolism, but insulin receptors and their downstream insulin signaling molecules are also constitutively expressed in vascular endothelial cells, vascular smooth muscle, and monocytes/macrophages. Forkhead box class O family member proteins (FoxOs) of transcription factors are essential regulators of cellular homeostasis, including glucose and lipid metabolism, oxidative stress response and redox signaling, cell cycle progression and apoptosis. In vascular endothelial cells, FoxOs strongly promote atherosclerosis via suppressing nitric oxide production and enhancing inflammatory responses. In liver sinusoidal endothelial cells, FoxOs induces hepatic insulin resistance by inducing nitration of insulin receptor in hepatocytes. Insulin resistance in adipose tissue limits capacity of lipid accumulation in adipose tissue, which promotes ectopic lipid accumulation and organ dysfunction in liver, vascular, and kidney. Modulation of insulin sensitivity in adipose tissue to induce healthy adipose expansion is expected to be a promising strategy for diabetic complications.

Introduction

According to the “National Health and Nutrition Survey,” published by the Ministry of Health, Labor and Welfare of Japan in 2019, the number of persons strongly suspected of having diabetes, which was estimated to be 10.8 and 19.7% women and men, respectively, has been significantly increasing since 2009. In addition to the lower insulin secretion capacity of the Japanese people than that of the western people as a genetic factor, insulin resistance (a condition in which insulin is less effective) has developed owing to obesity and visceral fat accumulation caused by environmental factors, such as the rapid westernization of diet and a lack of exercise [1, 2]. Type 2 diabetes is the most common type of diabetes and an important risk factor for atherosclerotic diseases, such as stroke, ischemic heart disease, and peripheral arterial disease; the risk of these diseases is predicted to be 2–4 times higher in type 2 diabetes patients than in non-diabetic individuals [3–6].

Among the various pathological conditions associated with type 2 diabetes, insulin resistance is a potent risk factor for diabetic complications. Epidemiological studies, such as Diabetes Epidemiology: Collaborative analysis of Diagnostic criteria in Europe (DECODE) [7] and Funagata [8], have shown that microvascular complications associated with type 2 diabetes are closely related to the development of long-term persistent hyperglycemia, whereas macrovascular complications such as atherosclerosis occur due to the impairment of glucose tolerance at a pre-diabetic stage. Interestingly, a strong correlation has been established between insulin resistance and the risk of developing cardiovascular diseases (CVDs) [9]. Several molecular mechanisms contribute to the association between insulin resistance and CVDs [10–13]. These mechanisms include insulin resistance in atherosclerosis development, vascular function, hypertension, and macrophage accumulation [10.

This review focuses on the interactions between insulin resistance and diabetic complications, especially in macrovascular diseases, and the molecular mechanisms involved.

Insulin signaling in vascular cells

The liver, skeletal muscle, and adipose tissue are the major organs of action of insulin in systemic glucose metabolism; however, insulin receptors and their downstream insulin signaling molecules are also constitutively expressed in vascular endothelial cells [14–16], vascular smooth muscles [17], and monocytes/macrophages [18].The physiological roles of insulin receptors and their downstream insulin signaling-related molecules have attracted increasing attention. Since the 1990s, it has been shown that insulin promotes endothelial nitric oxide synthase (eNOS) phosphorylation in an Akt-dependent and PI3-kinase-dependent manner in vascular endothelial cells [15, 19]; furthermore, a relationship between insulin signaling and vascular function in vascular endothelial cells has been established (Fig. 1). In addition, it has been reported that obese (ob/ob) rats develop insulin resistance in blood vessels as well as in the liver, skeletal muscle, and adipose tissue [14]. The study of the mechanism of insulin signaling in blood vessels has begun to gain increasing attention so that its correlation with the pathological conditions, such as atherosclerosis, could be unraveled.

Fig. 1.

Regulation of FoxOs by insulin signaling. Upper: a model for FoxOs regulation through insulin-induced and phosphorylation-dependent degradation. Lower: mice phenotypes in myeloid and endothelial cells-specific FoxO1/3a/4 knockout mice. ROS Reactive oxygen species, iNOS Inducible nitric oxide synthase, eNOS Endothelial nitric oxide synthase.

Vascular endothelial cells

Vascular endothelial cells consist of a single layer of epithelial cells lining the vessel walls. They not only act as a barrier between the vascular lumen and arterial wall but also release a variety of vasoactive factors that regulate numerous functions, including vascular contraction, cell adhesion, platelet adhesion and aggregation, inflammation, and cell proliferation. Atherosclerosis is accompanied by morphological changes in blood vessels (occlusion, stenosis, enlargement, rupture, and dissection), and the abnormalities in vascular endothelial cell functions that precede these changes are considered as the early lesions of atherosclerosis. When mice lacking vascular endothelial cell-specific insulin receptors are crossed with apolipoprotein E (ApoE)-deficient mice, atherosclerosis is accelerated, although there is no difference in glucose tolerance, insulin levels, lipids, or blood pressure when compared to those in the controls [16]. In the controls of vascular endothelial cells, insulin phosphorylates Ser1177 of eNOS (eNOS activation), suppresses vascular cell adhesion molecule (VCAM)-1 expression, and acts in an anti-atherosclerotic manner.

In vascular endothelial cell-specific insulin receptor-deficient mice, these effects are reduced, while the nitric oxide (NO)-induced vasodilation is inhibited, and the VCAM-1-dependent leukocyte adhesion is increased. Furthermore, it has been reported that systemic glucose metabolism and blood pressure-independent endothelial cell-specific vasorelaxation responses are attenuated in the vascular endothelial cell-specific insulin receptor dominant-negative transgenic mice [20]. Furthermore, mice deficient in Akt1, a member of the Akt subfamily, a downstream molecule of insulin signaling, show reduced eNOS phosphorylation and accelerated atherosclerosis in the ApoE-deficient mice [21]. In contrast, mice overexpressing insulin receptor substrate 1 (IRS1) specifically in vascular endothelial cells show atherosclerosis suppression [22]. Based on these reports, we suggest that insulin signaling in vascular endothelial cells can prevent atherosclerosis, while insulin resistance accelerates atherosclerosis almost consistently.

Vascular smooth muscle cells (VSMCs)

The effects of insulin on VSMCs and cell proliferation through p21 rat sarcoma virus (p21Ras) activation [23, 24] and mitogen activated protein (MAP) kinase activation [25, 26] have been previously reported. These proliferative effects are dependent on the concentration of insulin; however, when cultured smooth muscle cells were exposed to the physiological concentrations of insulin, no significant effects on MAP kinase activation or DNA synthesis enhancement were observed [25, 27].

Insulin can activate both insulin receptor (IR) and insulin-like growth factor 1 receptor (IGF1R) at elevated concentrations, owing to which interpreting whether IR or IGF1R mediates its effects becomes difficult. Both IR and IGF1R are expressed in VSMCs. Structurally and functionally, both are dimers formed by two monomers conjugated by disulfide bonds. In addition to homozygous IR and IGF1R, a hybrid form of IR/IGF1R can also be formed. The abundance of IGF1R is approximately eight-to-tenfold higher than that of IR in most vascular cells, including VSMCs, resulting in approximately 80% hybrid IR/IGF1R and only 20% homozygous IR (homo IR) formation. At physiological levels, insulin (1–10 nM) can only bind to homo-IR, whereas the physiological levels of insulin-like growth factor 1(IGF1) can activate both homo IGF1R and hybrid IR/IGF1R. Therefore, it can be speculated that IGF1R inhibits insulin effects by binding with IR to form a hybrid IR/IGF1R, thereby reducing the levels of homo IR [28]. The pathophysiological significance of the hybrid form of IR/IGF1R in vascular smooth muscle cells for the development of diabetic vascular complications remains to be elucidated.

Macrophages

In humans with insulin resistance, the activity of IRs is attenuated in monocytes [29] and insulin-resistant macrophages have a strong proinflammatory effect. However, there have been inconsistent reports that high insulin concentrations in mononuclear cells and macrophages, including lymphocytes, have anti-inflammatory and pro-inflammatory effects.

In an ob/ob mice model of type 2 diabetes, insulin signaling was attenuated in the intraperitoneal macrophages, which promoted lipid uptake. The expression of CD36 is increased at the level of post-translational modifications, and CD36 exerts pro-atherosclerotic effects by increasing the binding and uptake of oxidized low density lipoproteins (LDL) [30]. In fact, it has been reported that in macrophage-specific IR-deficient mice, using LDL receptor-deficient mice as the background and lysozyme M bacteriophage cyclization recombination (LysM-Cre) mice, apoptosis and necrosis were induced through increasing the cholesterol uptake by macrophages in aortic plaque lesions, thereby promoting atherosclerosis [31]. These results suggest that insulin resistance in macrophages promotes atherosclerosis, independent of systemic insulin resistance. However, a contradictory report on macrophage-specific insulin receptor-deficient mice generated using LysM-Cre mice and ApoE-deficient mice showed that insulin resistance has an inhibitory effect on atherosclerosis [32]. Therefore, we consider that the attenuation of insulin signaling in macrophages suppresses the LPS-induced IL-6 and IL-1β expression, which contributes to the suppression of atherosclerosis. The significance of macrophage insulin signaling in atherosclerosis, including that of forkhead transcription factors (FoxOs), varies, and its role may not be consistent across the mouse models and stages of atherosclerosis.

Hyperinsulinemia and atherosclerosis

It has been suggested that persistent hyperinsulinemia promotes atherosclerosis; however, detailed studies using animal models are lacking. Recently, the effects of continuous subcutaneous administration of insulin on atherosclerosis were investigated in high-fat diet (HFD)-fed ApoE-deficient mice using insulin pellets [33]. In the HFD ApoE-deficient mice, continuous insulin administration suppressed atherosclerosis by reducing aortic VCAM-1 expression and increasing eNOS phosphorylation. As the high fat-fed ApoE-deficient mice became obese and insulin-resistant, this model was the one in which exogenous insulin contributed to hyperinsulinemia. As a treatment strategy for obesity and type 2 diabetes, exogenous insulin over insulin resistance and the activation of vascular insulin signaling may contribute to the suppression of atherosclerosis. In addition, the insulin receptor–heterozygous ApoE-deficient mice showed increased hyperinsulinemia than the ApoE-deficient mice; however, there was no difference in insulin sensitivity, serum lipids level, and blood pressure in the aorta, skeletal muscle, liver, and adipose tissue [34]. These results suggest that hyperinsulinemia by itself has no effect on atherosclerosis and that the changes in insulin signaling activity play an important role in the development and progression of atherosclerosis.

Forkhead transcription factors (FoxOs)

FoxOs are transcription factor with a forkhead domain and comprises a subfamily of FoxO1, FoxO3a, FoxO4, and FoxO6; FoxO6 is expressed almost exclusively in the central nervous system, and FoxO1, FoxO3a, and FoxO4 are expressed in vascular endothelial cells, vascular smooth muscle cells, and macrophages. The activity of FoxOs as a transcription factor is regulated by various post-translational modifications, and its regulation by the phosphorylation of Akt, a serine/threonine kinase, has been widely studied. When Akt is activated by growth factors, such as insulin, FoxOs are phosphorylated in the nucleus and translocated out of the nucleus, where it becomes inactive as a transcription factor. FoxOs are a multifunctional protein that regulates cell proliferation, differentiation, apoptosis, and stress resistance and has diverse pathophysiological and physiological implications at the individual level. For example, in the liver, FoxOs positively regulates the gene expression of G6Pase (G6pc) and PEPCK (Pck1), which promote glycogenesis [35–38]; in addition, Notch signaling is also involved in the regulation of G6pc expression [39].

FoxOs in vascular endothelial cells and atherosclerosis

It has been reported that the induction of atherosclerosis in LDL receptor-deficient mice induces insulin resistance in macro vessels and attenuates the phosphorylation of FoxO1 and 3a, suggesting that FoxO is activated in atherosclerotic vessels [40]. In vascular endothelial cells, FoxO1 and 3a negatively regulate the transcriptional activity of inducible nitric oxide synthase (iNOS and Nos2) and positively regulate the transcriptional activity of eNOS (Nos3) [15, 19]. Vascular endothelial cell-specific deletion of FoxO1, 3a, and 4, with the aid of Tie2 promoters, markedly suppressed the development of atherosclerosis in LDL receptor-deficient mice [40]. In the vascular endothelial cells of these mice, eNOS-derived nitric oxide production was increased, while iNOS expression was reduced. Nuclear factor kappa B (NF-κB) activity, oxidative stress production, cellular senescence, and apoptosis were suppressed. In addition, the adhesion factors intercellular adhesion molecule-1 and VCAM-1 were the target genes of FoxO1, indicating that FoxOs are the key factor that integrally regulates multiple mechanisms of atherosclerosis.

Furthermore, it has been reported that hyperglycemia induces the deacetylation of FoxO1 in vascular endothelial cells and promotes its nuclear translocation [19]. Knock-in mice with a homeostatic deacetylated FoxO1 mutant exhibit accelerated atherosclerosis, and the mechanism is bone marrow cell-independent [41]. In other words, insulin resistance and hyperglycemia, the main pathogenesis pathways of type 2 diabetes, promote atherosclerosis via FoxOs activation in vascular endothelial cells. However, it is unclear whether the target genes of dephosphorylated FoxOs activated by insulin resistance and deacetylated FoxOs activated by hyperglycemia are the same. This is an interesting topic that should be studied so that the significance of insulin resistance and hyperglycemia in the development and progression of atherosclerosis in type 2 diabetes can be elucidated.

FoxOs in macrophages and atherosclerosis

As in vascular endothelial cells, insulin resistance and FoxOs activation are also observed in the macrophages of obese and atherosclerotic mice [30]. However, atherosclerosis is promoted in systemic FoxO4-deficient mice [42] and in mice for which the bone marrow cell-specific FoxO1/3a/4 has been knocked out using LysM-Cre [43]. One possible mechanism is that the bone marrow cell-specific FoxO1/3a/4-deficient mice might have increased cell division in the bone marrow due to the presence of guanosine monophosphate (GMP), which led to neutrophil and monocyte hyperplasia with splenomegaly in the peripheral blood. In addition, the proportion of monocytes of Ly6C^hi, a subset of atherosclerosis-promoting cells, increased in the peripheral blood, and FoxOs in bone marrow cells acted as an atherosclerosis suppressor by regulating the number and expression of neutrophils and monocytes. In addition, LDL receptor deficiency was observed in bone marrow cells. Insulin resistance and the expression of nitration/nitrosation of insulin receptors were increased in the livers of myeloid cell-specific FoxOs-deficient mice in the background of LDL receptor-deficient mice. This suggests that nitric oxide (NO) derived from inflammatory macrophages in the liver might have possibly attenuated the insulin receptor activity nitration and nitrosation of the insulin receptor.

Thus, it is interesting to note that although FoxOs is activated in both the vascular endothelial cells and macrophages of animal models of insulin resistance, the pathophysiological significance of FoxOs in atherosclerosis is quite different. FoxOs may play a role in maintaining the balance at an individual level in the pathogenesis of atherosclerosis by exerting physiological effects that are unique to cells and organs.

FoxOs and glucose metabolism in liver sinusoidal endothelial cells

Mice deficient in endothelial cell-specific FoxO1/3a/4 exhibit hyperglycemia and insulin resistance in the liver when fed a normal diet [44]. The expression of eNOS also increases in the liver sinusoidal endothelial cells (LSECs) of the same mice, indicating that NO acts on hepatocytes through a paracrine mechanism from LSECs to induce insulin resistance via the induction of glycogenic enzyme expression in the liver and nitration of tyrosine residues in the insulin receptor. Sustained insulin stimulation of the wild-type LSECs results in the sustained inactivation of FoxOs and increased production of eNOS-dependent NO, which suggests that chronic hyperinsulinemia is involved in the development of hyperglycemia and insulin resistance in liver parenchymal cells via the sustained inactivation of FoxOs and increased production of NO in LSECs.

Adiopose insulin resistance and “healthy adipose expansion”

Type 2 diabetes and NAFLD

In recent decades, metabolic syndrome as well as non-alcoholic fatty liver disease (NAFLD) have become increasingly prevalent [45, 46]. NAFLD is a clinical and pathological term that describes a disease spectrum ranging from simple steatosis to nonalcoholic steatohepatitis (NASH), cirrhosis, and hepatocellular carcinoma [47]. Obesity and type 2 diabetes are recognized as important risk factors for NAFLD; the prevalence of NAFLD is 4.6 times higher in the obese population than in normal individuals [48], and 33–50% of type 2 diabetes patients are diagnosed with NAFLD [49]. In addition, diabetes is moderately associated with liver cancer-related deaths [50], which highlights the clinical significance of preventive interventions for NAFLD [49].

In a previous study, we treated ob/ob mice with the sodium glucose co-transporter 2 (SGLT2) inhibitor ipragliflozin (Ipra); Ipra administration improved hepatic steatosis in HFD-induced and leptin-deficient ob/ob mice irrespective of body weight reduction [51]. In ob/ob mice, the Ipra-induced hyperphagia increased energy intake and attenuated body weight reduction with an increased epididymal fat mass. There was an inverse correlation between the weights of the liver and epididymal fat in the Ipra-treated obese mice, suggesting that the Ipra treatment promotes normotopic fat accumulation in the epididymal fat and prevents ectopic fat accumulation in the liver. Notably, despite increased adiposity, Ipra ameliorates the obesity-associated inflammation and insulin resistance in epididymal fat, referred to as “healthy adipose expansion” [52]. Clinically, Ipra improves liver dysfunction in patients with type 2 diabetes, irrespective of body weight reduction.

The effects of SGLT2 inhibitors on NASH and NASH-associated hepatic carcinogenesis has been elucidated by treating a western diet (WD)-fed melanocortin 4 receptor-deficient (MC4R-KO) mouse model of human NASH and NASH-associated hepatocellular carcinoma [53] with the SGLT2 inhibitor canagliflozin (CANA) (Fig. 2). It was confirmed that an 8 week treatment with CANA could attenuate hepatic steatosis in WD-fed melanocortin 4 receptor deficient-knockout (MC4R-KO) mice, with an increased epididymal fat mass without inflammatory changes [54]. The indices of systemic and adipose insulin resistance [55, 56], namely, homeostasis model assessment-estimated insulin resistance (HOMA-IR) and adipose tissue insulin resistance (Adipo-IR), respectively, were significantly reduced upon CANA treatment. CANA treatment for 20 weeks inhibited the development of hepatic fibrosis in WD-fed MC4R-KO mice. After 1 year of CANA treatment, the number of hepatocellular carcinomas was significantly reduced in WD-fed MC4R-KO mice.

Fig. 2.

Healthy adipose expansion induced by a SGLT2 inhibitor. Possible mechanisms by which a SGLT2 inhibitor promote adipose expansion without deteriorating adipose inflammation, called as “healthy adipose expansion”. An SGLT2 inhibitor attenuates insulin resistance in adipose tissues. Restoring insulin action in adipose tissues inhibits adipocyte apoptosis, associated by increased capacity of lipid accumulation in adipocytes. When energy intake is enhanced by an SGLT2 inhibitor, it can be accumulated in adipocytes. Macrophage infiltration into adipose tissues is decreased probably due to decreased adipocytes apoptosis, resulting in decreased inflammatory cytokines from adipose tissues. Decreased adipose inflammation also attenuates insulin resistance in adipose tissues.

One possible mechanism to promote healthy adipose expansion is the improvement of insulin sensitivity in adipose tissue, which has been assessed using Adipo-IR in a previous study [54]. Based on these observations, we suggest that the phenotypes of CANA-treated mice are similar to those of adipocyte-specific inducible phosphatase and tensin homolog (PTEN)-knockout mice, which exhibit an enhanced adipocyte insulin signaling [57, adipose tissue and weight gain during high-fat diet feeding, enhanced insulin sensitivity, improved hepatic steatosis, and reduced adipose tissue inflammation. Thus, we suggest that the enhanced insulin sensitivity in adipose tissue can explain, at least a little, the increased lipid storage capacity of the adipocytes of SGLT2 inhibitor-treated mice.

Effects of healthy adipose expansion on other organs

Perivascular adipose tissue (PVAT) surrounds the vasculature, and it has been suggested that PVAT plays an important role in the pathogenesis of cardiovascular diseases [58]. PVAT not only stores triglycerides and functions as a structural support for vessels but also secretes numerous biologically active molecules to control vascular function and remodeling. It has been reported that in obesity or T2DM, PVAT predominantly secretes pro-inflammatory and pro-atherogenic cytokines, which can cause local endothelial dysfunction, thus contributing to the progression of systemic and local vascular diseases. Ipra administration increases adipocyte size and reduces the expression of proinflammatory and profibrotic genes in the abdominal PVAT of WD-fed wild-type (WT) mice [59]. Furthermore, Ipra suppresses the WD-induced macrophage accumulation and adipocyte death in the abdominal PVAT, and perivascular implantation of epididymal fat from Ipra-treated mice to ApoE-deficient mice attenuates the cuff-induced neointimal hyperplasia and vascular remodeling.

We further reviewed whether healthy adipose expansion affects diabetic nephropathy (DN). Ipra treatment reduces urinary albumin excretion (UAE) and glomerular hypertrophy in the HFD-fed mice [60].In the perirenal adipose tissue (PRAT) of the Ipra-treated mice, adipocyte size is increased and inflammation, fibrosis, and adipocyte death are suppressed. In a conditioned medium prepared from the PRAT (PRAT-CM) of Ipra-treated mice, the concentration of leptin is significantly lower than that in the PRAT-CM of mice without IPRA treatment. The serum leptin concentration in the renal vein is positively correlated with UAE. PRAT-CM from the HFD-fed mice shows greater cell proliferation signaling in mouse glomerular endothelial cells than PRAT-CM from standard diet-fed mice via the p38MAPK and leptin-dependent pathways, whereas the effects are significantly attenuated in PRAT-CM from the Ipra-treated mice. These findings suggest that the Ipra-induced healthy adipose expansion in PRAT may play an important role in the improvement of DN in the HFD-fed mice. Based on these observations, we suggest that the enhancement of insulin sensitivity in adipose tissue may serve as a novel approach to inhibit diabetic complications by inducing a healthy adipose tissue expansion.

Conclusions

The systemic and local insulin resistance are fundamental pathophysiological processes that promote atherosclerosis, and the activation of insulin signaling in these cells may antagonize the pathology. Although persistent hyperinsulinemia was initially thought to act in an atherosclerosis-promoting manner, recent reports at the animal level suggest that insulin administration to insulin-resistant individuals may also suppress atherosclerosis. These findings from basic research studies are expected to lead to further discussion and research on insulin signaling and atherosclerosis in humans. In addition, insulin resistance in adipose tissue is considered to play an important role in the development of vascular, hepatic, and renal complications associated with diabetes by limiting the capacity of lipid accumulation in adipose tissues. Therefore, modulation of insulin sensitivity in adipose tissue to induce healthy adipose expansion can be a promising therapeutic strategy for diabetic complications.

Declarations

Conflict of interest

Kyoichiro Tsuchiya declares that he has no conflict of interest.

Statement of animal and/or human participants

This article does not contain any studies with human or animal subjects.