Coronary microvascular dysfunction in diabetes mellitus: A review

Abstract

The exploration of coronary microcirculatory dysfunction in diabetes has accelerated in recent years. Cardiac function is compromised in diabetes. Diabetic patients manifest accelerated atherosclerosis in coronary arteries. These data are confirmed in diabetic animal models, where lesions of small coronary arteries have been described. These concepts are epitomized in the classic microvascular complications of diabetes, i.e. blindness, kidney failure and distal dry gangrene. Most importantly, accumulating data indicate that insights gained from the link between inflammation and diabetes can yield predictive and prognostic information of considerable clinical utility. This review summarizes the evidence for the predisposing factors and the mechanisms involved in diabetes, and assesses the current state of knowledge regarding the triggers for inflammation in this disease. We evaluate the roles of hyperglycemia, oxidative stress, polyol pathway, protein kinase C, advanced glycation end products, insulin resistance, peroxisome proliferator-activated receptor-γ, inflammation, and diabetic cardiomyopathy as a “stem cell disease”. Furthermore, we discuss the mechanisms responsible for impaired coronary arteriole function. Finally, we consider how new insights in diabetes may provide innovative therapeutic strategies.

Key Words: Coronary artery; Diabetes; Endothelial dysfunction; Hyperglycemia; Inflammation; Insulin; Microcirculation; Nitric oxide; Oxidative stress

INTRODUCTION

Diabetes and coronary artery disease

Cardiovascular diseases are significantly increased in patients with metabolic syndrome and type 2 diabetes. In particular, coronary artery disease (CAD) causes much of the serious morbidity and mortality in patients with diabetes, who have a 2- to 4-fold increase in the risk of CAD[1]. In one population-based study[2], the 7-year incidence of first myocardial infarction or death was 20% for diabetic patients, but only 3.5% for non-diabetic patients. The aging of the population and an increasing prevalence of obesity and sedentary life habits in the United States contribute to an increasing prevalence of diabetes.

Metabolic and vascular dysfunction

Factors such as chronic hyperglycemia, lipid abnormalities, inflammation, oxidative stress, endothelium dysfunction, increased thrombosis and decreased fibrinolysis are likely to promote cardiovascular events in patients with metabolic syndrome and type 2 diabetes[3]. Clinical and experimental studies have demonstrated that cardiac function is compromised in type 2 diabetes, suggesting that alterations in myocardial metabolism in the diabetic state are responsible for this impairment[4]. Also, changes in coronary vessel function can lead to a mismatch of myocardial supply and demand, thereby provoking ischemic episodes in the diabetic heart.

Functional evidence precedes morphological changes

Vascular lesions of small coronary arteries have been described in diabetic patients and experimental animals. Characteristic morphological features include thickening of the arterial wall[5] and capillary basement membrane, periodic acid-Schiff positive deposits in the vessel wall of small arteries[6], microaneurysms, perivascular and interstitial fibrosis, and fibrosis in the wall of small coronary arteries[7]. However, previous studies[8,9] in small arteries and arterioles of diabetic subjects have demonstrated that before the appearance of morphological changes, a vasomotor dysfunction develops in microvessels, affecting both smooth muscle- and endothelium-mediated regulatory mechanisms. The first studies on abnormal nitric oxide (NO) production were performed in rats and then confirmed in diabetic humans. The relaxation of coronary arteries in response to pharmacological stimuli was reduced or suppressed in diabetic rats[10]. In humans, the vasodilation of coronary arteries was also altered after pharmacological [acetylcholine (ACh)] or mechanical (cold test) stimuli, but these abnormalities of large vessels were not associated with angiographic lesions, and were independent of other cardiovascular risk factors[11], suggesting impaired endothelial function without any anatomical lesions. Bagi et al[12] have described the effects of diabetes mellitus on coronary arterioles. In type 2 diabetic mice, agonist- and flow-induced dilation of coronary arterioles was reduced. Furthermore, Miura et al[13] confirmed the impairment of coronary microvascular function in human coronary arterioles isolated from patients affected by type 1 or type 2 diabetes mellitus: coronary arterioles showed a reduced vasodilation to hypoxia due to a decreased activity of ATP-sensitive potassium channels. The mechanisms underlying type 2 diabetes-induced impaired vasodilation and scientific consensus will be evaluated in this review.

ETIOLOGY OF CORONARY MICROVASCULAR DYSFUNCTION IN DIABETES

Role of hyperglycemia

Hyperglycemia suppresses flow-mediated endothelial dependent vasodilation, and impairs endothelial-dependent vasodilation in diabetic and healthy human subjects[14-17]. Hyperglycemia is clearly recognized as the primary culprit in the pathogenesis of diabetic complications, inducing repeated acute changes in intracellular metabolism (activation of the polyol pathway, activation of diacylglycerol (DAG)-protein kinase C (PKC), increased oxidative stress, endothelial cell glycocalyx perturbation[18]), as well as cumulative long-term changes in the structure and function of macromolecules through formation of advanced glycation end products (AGEs). Coronary microvascular endothelial cells exposed to hyperglycemia exhibit increased oxidative stress that may arise from enhanced pro-oxidant enzyme activity and diminished generation of antioxidant glutathione[19] (Figure 1).

Figure 1 Etiology of mechanisms involved in coronary vascular dysfunction in diabetes. Increased oxidative stress is the unifying element common to all pathways through which the various mechanisms described interact to cause endothelial dysfunction and cardiomyopathy in diabetes. Hyperglycemia acts directly through the glycation of the scavenger enzyme superoxide dismutase (SOD) and other antioxidant enzymes, and indirectly through the cross activation of advanced glycation end product (AGE)/advanced glycation end product receptor (RAGE) interaction and of the polyol pathway. The latter raises the NADH/NAD⁺ ratio, modifying the redox state of the cells and leading to the production of superoxide anions. Tumor necrosis factor-α (TNF-α) has been shown to affect intracellular insulin signaling, promoting insulin resistance and consequently impairing the insulin-mediated endothelial function in coronary arterioles. The insulin-resistance mediated endothelial dysfunction can be restored by the activation of peroxisome proliferator-activated receptor-γ by thiazolidinediones and rosiglitazone. Then, if we consider diabetes as a chronic, subclinical inflammatory disease, it acts through TNF-α and Interleukin (IL)-6, increasing the cellular oxidative stress. AGE/RAGE interaction and hyperglycemia can in turn activate the diacylglycerol (DAG)-protein kinase C (PKC) pathway, thus promoting fibrosis, myofilament desensitization, and apoptosis, and finally leads to diabetic cardiomyopathy. TnI: Troponin I; TnT: Troponin T.

Blood glucose control is not significantly different between type 1 and type 2 diabetic patients, suggesting[20] that other factors are involved. A longer period of undetected blood glucose abnormalities, lipid alterations or decreased insulin sensitivity could be involved in type 2 diabetes. Furthermore, the initial ACh-induced endothelium-dependent dysfunction can be improved by the normalization of blood glucose control in type 1 diabetes mellitus[21], whereas such remediation has never been observed in type 2 diabetes mellitus. Consequently, hyperglycemia is probably not the sole mechanism by which diabetes mellitus induces vascular endothelial dysfunction.

Role of insulin resistance

Role of insulin and insulin resistance in endothelium-dependent vasodilation: Endothelial cells express insulin receptors. Insulin is known to elicit NO-dependent vasodilation in human skeletal muscle[22]; it may stimulate basal NO production directly by itself or indirectly by a second messenger. In vitro evidence indicates that insulin activates the L-arginine-NO vasodilator pathway in men[23] and stimulates the release of endothelium-derived relaxing factors (NO and PGI₂). The L-arginine availability for endothelial NO synthase (NOS) could be reduced by either diminished L-arginine recycling or increased L-arginine metabolism. The study by Hein et al[24] suggests that one mechanism by which ischemia/reperfusion (I/R) inhibits NO-mediated arteriolar dilation is through increased arginase activity, which limits the availability of l-arginine to NOS for NO production. ACh-induced vasodilation is correlated with insulin sensitivity in healthy subjects, suggesting that insulin plays an important role in the early processes of endothelial dysfunction[25]. Insulin resistance precedes the development of type 2 diabetes mellitus and is associated with increased plasma concentrations of endothelin and von Willebrand factor (vWF) in obese subjects, even in the absence of diabetes mellitus[26]. Insulin vascular action may be blunted in insulin resistant states, such as obesity[26], hypertension, impaired glucose tolerance and type 2 diabetes mellitus[27]. Wang et al[28] showed that, in cultured endothelial cells, free fatty acid, which is increased in metabolic syndrome, induces insulin resistance, inhibits eNOS activation and consequently causes endothelial dysfunction. Avogaro et al[29] confirmed this hypothesis in uncomplicated type 2 diabetic patients by linking cellular glucose disposal defects with insulin resistance.

Mechanisms of insulin resistance-mediated impaired vasodilation: The mechanisms by which hyperinsulinemia acts on endothelial function have been studied by Rask-Madsen et al[30], showing the role of tumor necrosis factor-α (TNF-α) in insulin-stimulated endothelial function in humans. They found that the forearm blood flow response to ACh was inhibited by TNF-α and the inhibitory effect of TNF-α was larger during co-infusion of insulin. The results demonstrate that TNF-α plays a pivotal role in insulin mediated endothelial dysfunction. TNF-α affects intracellular insulin signaling in fat, skeletal muscle and other insulin sensitive tissues by inhibiting kinase activity in the proximal part of the insulin signaling pathway[31,32]. A similar signaling pathway in vascular endothelium results in production of NO[33], which is necessary for insulin-stimulated vasodilation[34]. In a rat model of obesity-associated insulin resistance, vascular insulin signaling is compromised[35]. In cultured endothelial cells, 10 min of exposure to TNF-α inhibits insulin signaling and NO production.

Role of peroxisome proliferator-activated receptor-γ

Peroxisome proliferator-activated receptor-γ and the amelioration of insulin resistance: Peroxisome proliferator-activated receptor-γ (PPAR-γ) is a ligand-activated transcription factor belonging to the nuclear receptor family[36]. PPAR-γ is a regulator of lipid and glucose metabolism and is the target of insulin-sensitizing drugs such as thiazolidinediones, which are frequently used to treat metabolic complications associated with type 2 diabetes mellitus[37]. Long-term activation of PPAR-γ by thiazolidinediones in type 2 diabetic subjects reduces plasma levels of insulin and glucose with the consequent attenuation of vascular dysfunction[38] (Figure 1). PPAR-γ is also expressed in vascular tissues, specifically in vascular smooth muscle cells[39] and endothelial cells[40]. Activators of PPAR-γ, via a mechanism that is unrelated to lipid and carbohydrate metabolism, may also protect vascular function in diabetes mellitus[41]. Furthermore, PPAR-γ has been demonstrated to increase the release of NO in porcine pulmonary artery and human umbilical vein endothelial cells in culture[42].

PPAR-γ antagonists also provided encouraging results in in-vivo human studies. Thiazolidinediones improved coronary vasomotor abnormalities in Mexican Americans with insulin resistance[43]. The addition of pioglitazone to conventional lipid-lowering therapy in non-diabetic but insulin-resistant patients affected by familial combined hyperlipidemia significantly enhanced not only myocardial glucose uptake, but also myocardial blood flow measured by positron emission tomography[44].

PPAR-γ activation improves endothelial dysfunction: Both experimental and clinical evidence has revealed the role of PPAR-γ in diabetic endothelial dysfunction. Bagi et al[45] have shown that short-term treatment of type 2 diabetic mice with the PPAR-γ activator, rosiglitazone, augments NO-mediated, flow-dependent dilation of coronary arterioles, despite the presence of hyperglycemia and hyperinsulinemia. These changes are associated with a reduction in vascular NAD(P)H oxidase activity and enhancement of vascular catalase activity. The study demonstrates functionally important antioxidant activity of the PPAR-γ ligand. In fact, rosiglitazone, by activating vascular PPAR-γ, prevents the impairment of NO mediation of coronary arteriolar dilations. The underlying mechanism is most likely the enhancement of NO bioavailability via a reduction in the level of vascular reactive oxygen species (ROS). In the clinical environment, Esposito et al[46] showed that treatment for over one year with rosiglitazone reduces circulating inflammatory markers and improves endothelial function in patients with metabolic syndrome. These observations show that rosiglitazone increases the number and migratory activity of cultured endothelial progenitor cells in type 2 diabetic patients[47].

Role of inflammation

Insulin resistance and diabetes mellitus in inflammatory disorders: Inflammation is a condition that underscores many cardiovascular pathologies including endothelial dysfunction, insulin resistance and, consequently metabolic syndrome and diabetes mellitus. The inflammatory biomarkers, C reactive protein (CRP), interleukin-6 (IL-6) and TNF-α are potentially informative given their involvement in biologically plausible mechanisms of insulin resistance. Among these cytokines, TNF-α is likely to play a pivotal role because it is one of the key inflammatory mediators expressed during a variety of inflammatory conditions[30] and is capable of initiating the expression of an entire spectrum of inflammatory cytokines ranging from interleukins to interferons.

Several studies have shown the potential application of these cytokines in the prediction of diabetes risk. Yudkin et al[48] showed that CRP, IL-6 and TNF-α are elevated in the insulin resistance syndrome, from quantitative measures of insulin resistance. Hak et al[49] found similar relationships among healthy middle-aged women. In the insulin resistance and atherosclerosis study[50], among 1008 non-diabetic subjects with no prior history of CAD, CRP levels were independent of insulin sensitivity as measured by a frequently sampled intravenous glucose tolerance test. They also found that higher geometric mean CRP levels were linearly related to an increase in a number of components of the metabolic syndrome. These observations suggest that an enhanced acute phase response is associated with insulin resistance and may presage the development of type 2 diabetes.

These epidemiological findings are strengthened by experimental studies that demonstrate the hyperglycemic effects of several proinflammatory cytokines including IL-6 and TNF-α, both of which are partly derived from adipose tissues. In rodent models of glucose homeostasis, IL-6 modifies glucose-stimulated insulin release from isolated pancreatic β cells and diminishes insulin-stimulated glycogen synthesis by hepatocytes in culture[51,52]. In humans, the exogenous administration of recombinant IL-6 induces dose-dependent hyperglycemia and concordant elevations in circulating levels of glucagons[53]. TNF-α may induce insulin resistance through a variety of mechanisms, including direct inhibitory effects on the glucose transporter protein GLUT4, the insulin receptor, and insulin receptor substrates[54]. TNF-α has been shown to alter the intracellular insulin signaling in fat, skeletal muscle, and other insulin-responsive tissues by inhibiting kinase activity in the proximal part of the insulin-signaling pathway[31,32]. A similar signaling pathway in vascular endothelium results in the production of NO, which is necessary for insulin-stimulated vasodilation[33,34]. Jiang et al[35] demonstrated in a comparable rat model that vascular signaling is compromised. In cultured endothelial cells, 10 min of exposure to TNF-α inhibits insulin signaling and NO production.

The role of TNF-α in coronary microvascular dysfunction was first identified in the setting of ischemia reperfusion-injury by using a murine genetic model (TNF-α overexpression mice, TNF^++/++)[55]; recent studies have shown that TNF-α is also likely to be involved in coronary microvascular dysfunction occurring in the setting of both pre-diabetic metabolic syndrome and overt diabetes. Picchi et al[56] recently demonstrated in an animal model of metabolic syndrome (i.e. Zucker fatty rats) that impairment in coronary endothelial function is caused by TNF-α overexpression; we demonstrated that endothelial dysfunction occurring in obesity is the result of the effects of the inflammatory cytokine TNF-α and subsequent production of superoxide (O₂˙^-). We also used genetic models of obesity and type 2 diabetes (Lepr^db mouse), heterozygote lean controls (m Lepr^db), and Lepr^db mice null for TNF-α (db^TNF-/db ^TNF-)[57]. Our results revealed that endothelial function is normal in TNF-deficient diabetic mice and that TNF-α overexpression impairs endothelium-dependent vasodilation which can be restored toward normal by administration of TNF-α antibodies. The mechanism by which TNF-α affects endothelial function is through increased superoxide production by NAD(P)H oxidases, which in turn leads to a reduced NO bioactivity by direct scavenging. Moreover, we observed that AGE receptor (RAGE) for these products seem to amplify TNF-α expression in diabetes; thus, TNF-α and the AGE/RAGE signaling pathway play pivotal roles in endothelial dysfunction in type 2 diabetes[58].

In order to document the extent of endothelial dysfunction at the different stages of type 2 diabetes, we also studied type 2 diabetic (db/db) mice aged 12, 18 and 24 wk[59]. We demonstrated that TNF-α is the key factor in producing O₂˙^- and induces endothelial dysfunction in db/db (a model of obesity and type 2 diabetes) and Db/db (lean control) mice as they age. Furthermore, we found that in younger mice (12-18 wk) the mechanism by which TNF-α affects endothelial function is through an increased O₂˙^- production by NAD(P)H oxidases which in turn leads to reduced NO bioactivity; on the contrary, in older mice (18-24 wk), TNF-α impairs endothelial function through an increased O₂˙^- production by mitochondria, which in turn leads to a reduced NO bioactivity by direct scavenging. Further support for the role of TNF-α in endothelial dysfunction comes from a recent study where aorta rings from diabetic mice instead of coronary arterioles were used: polyphenol (resveratrol) is capable of exerting a protective effect against vascular oxidative stress by inhibiting the activation of vascular NAD(P)H oxidase, which leads to a downregulation of eNOS phosphorylation induced by TNF-α[60]. In conclusion, these recent findings contribute to understanding the link between inflammation, insulin resistance and coronary microvascular dysfunction and highlight the key role of TNF-α as a common denominator in these pathologies.

Should anti-inflammatory treatment be considered among diabetes therapies? Despite significant evidence, prospective data for evaluation of the relationship between chronic subclinical inflammation and the incidence of type 2 diabetes are sparse. In the Atherosclerosis Risk Communities Study, markers of inflammation (such as white blood cell count, fibrinogen, and low serum albumin)[61] and inflammation-associated hemostasis variables (such as factor VIII and vWF)[62] were associated with the risk of type 2 diabetes. However, these relationships were largely abolished after adjustment for obesity. Pradhan et al[63] showed the association between elevated levels of CRP and IL-6 and the risk of type 2 diabetes in otherwise healthy middle-aged women. Among participants of the Women’s Health Study followed for 4 years, through age-matched analyses on obesity control, family history of diabetes, and other clinical risk factors, the authors found that elevated CRP levels are associated with a 4-fold increase in risk for presenting with diabetes. Recent reports on reduction in the incidence of type 2 diabetes accompanying pharmacological interventions for coronary heart disease prevention offer further support for links between inflammation, diabetogenesis, and atherosclerosis. In two large intervention trials of angiotensin-converting enzyme (ACE) inhibitors for the prevention of cardiovascular disease, treatment assignment to captopril and ramipril[64] was associated with a statistically significant reduction in the incidence of type 2 diabetes. Similarly, the use of pravastatin in the primary prevention of coronary heart disease was also associated with a 30% reduction in the risk of type 2 diabetes[65]. One compelling hypothesis that may account for these effects is mitigation of subclinical inflammation. Angiotensin II induces IL-6 expression from both macrophages and smooth muscle cells, and is co-localized with IL-6 in human atheroma[66]. Furthermore, long-term ACE-inhibition lowers CRP levels among individuals with CAD[67]. In addition, statin therapy in general appears to lower CRP levels and exhibits beneficial adjunctive effects on the restoration of endothelial function[68].

Role of oxidative stress

Oxidative stress is defined as an increase in the steady-state levels of ROS and may occur as a result of increased free radical generation and/or decreased antioxidant defense mechanisms. This seems to be the common final pathway that leads to endothelial dysfunction in diabetes mellitus (Figure 1). There are multiple intracellular sources for the formation of oxygen free radicals [e.g. mitochondria, xanthine oxidase, NAD(P)H oxidase etc.]. Our results[57] suggest that the primary proximate route to radical production in type 2 diabetes is through NAD(P)H oxidase activation by TNF-α.

Mechanisms that decrease antioxidant mechanisms: Hyperglycemia promotes glycation and inactivation of antioxidant proteins such as Cu/Zn superoxide dismutase (SOD), leading to inactivation and reduction in antioxidant defense for these proteins[69]. Experimental studies in streptozotocin-induced diabetic rats have shown decreased concentrations of antioxidants such as vitamin E, SOD and catalase[70]. For example, the consumption of NAD(P)H leads to decreased glutathione activity, which is efficient for capturing free radicals[71]. Experimentally, when the activities of SOD (which captures O₂.) and catalase (H₂O₂ inhibitor) were maintained, endothelial function was not altered even in hyperglycemic patients.

Mechanisms of increased generation of oxygen free radicals: Many experimental studies suggest that increased superoxide production accounts for a significant proportion of the NO deficit in diabetic vessels. Potential sources of vascular superoxide production include NAD(P)H-dependent oxidases[72,73], xanthine oxidase[74], lipoxygenase, mitochondrial oxidase and NOS[75]. Hyperglycemia increases oxidative stress through ROS overproduction at the mitochondrial transport chain level; Piconi et al[76] proposed that the mitochondrial oxidative activity may be a therapeutic target in diabetes. However, NAD(P)H oxidase appears to be the principal source of superoxide production in several animal models of vascular disease, including diabetes[77]. An increase in TNF-α expression induces activation of NAD(P)H oxidase and production of ROS, leading to endothelial dysfunction in type 2 diabetes[57].

Guzik et al[78] have described the mechanisms of increased superoxide production in human diabetes mellitus. First they found that basal superoxide release is significantly elevated in vessels from patients with diabetes. They demonstrated that endothelium is a net contributor to total vascular superoxide production. In fact, in arteries of non-diabetic patients, endothelium removal resulted in a significant increase in superoxide release, which suggests that in these vessels the net contribution of the endothelium is to reduce vascular superoxide release by production of NO. In marked contrast, endothelium removal in artery segments from diabetic patients significantly reduces superoxide release, suggesting a key role of the endothelium in superoxide production. Similarly, NOS inhibition in diabetic vessels decreases superoxide release, suggesting that the net effect of NOS activity in these vessels is superoxide production rather than NO production. Furthermore, they found that sepiapterin (a BH₄ precursor) significantly reduces vascular superoxide production in vessels from patients with diabetes. Therefore, NAD(P)H oxidase seems to be the most important source of superoxide production in diabetes mellitus and the superoxide anion is likely to reduce NO bioactivity by direct scavenging. Nevertheless, in diabetic vessels, the endothelium is a significant net source of superoxide because of a profound loss of normal eNOS function, characterized by a transition from NO production to superoxide production. In fact, peroxynitrite, generated from NO and superoxide, directly oxidizes BH₄ to BH₂ (dihydrobiopterin), a biopterin that does not support eNOS enzymatic activity[79]. Indeed, some data suggest that competition between BH₂ and BH₄ for eNOS binding may increase eNOS uncoupling. Therefore, upregulation of vascular superoxide production by NAD(P)H oxidases may in turn lead to eNOS uncoupling through oxidation of BH₄. This reduces NO production and further increases endothelial superoxide production. Bagi et al[80] confirmed these results: they found that in vitro administration of the NAD(P)H oxidase inhibitor apocynin restores flow-induced coronary arteriolar dilation in type 2 diabetes mice, suggesting that NAD(P)H oxidase is likely to be the main source of the enhanced superoxide production in coronary microvessels. A recent study supports the role of NAD(P)H oxidase[81]: the authors demonstrated that endothelial dysfunction in atherosclerosis is mediated, at least in part, via the interaction of oxidizing low density lipoproteins (Ox-LDL) with its receptor, LOX-1, which in turn stimulates endothelial generation of superoxide by activation of NAD(P)H oxidase.

However, different sources of superoxide anions have been described in recent reports[82,83], suggesting that an enhanced level of mitochondrial superoxide, likely produced during the enhanced rate of glucose metabolism, might be responsible for all high glucose related processes. Bagi et al[45] reported that in carotid arteries, either the presence of 2-DG (a competitive inhibitor of glycolysis) or the presence of TTFA (inhibitor of the mitochondrial complex II) significantly reduces hyperglycemia-induced superoxide production. Correspondingly, in skeletal muscle arterioles, the presence of 2-DG during hyperglycemia prevents the reduction of flow-induced dilation. Moreover, the presence of TTFA substantially moderates the hyperglycemia-induced reduction in flow-induced dilation of arterioles. These results indicate that during hyperglycemia a higher rate of glucose metabolism (glycolysis and mitochondrial utilization) is likely to elicit enhanced production of superoxide in mitochondria of skeletal muscle arterioles.

Role of the polyol pathway

High blood glucose levels increase activity in the polyol pathway. In the polyol pathway, glucose is reduced to sorbitol by aldose reductase, leading to depletion in cellular stores of NAD(P)H[84]. Reduced NAD(P)H is required for the functioning of many endothelial enzymes, including NOS and cytochrome P450, as well as for the antioxidant activity of glutathione reductase. Sorbitol is then oxidized to fructose by sorbitol dehydrogenase. This reaction uses NAD⁺ and raises the NADH/NAD⁺ ratio, which modifies the redox state of the cells and results in the production of superoxide anions (Figure 1). Alternatively, a high polyol pathway flux consumes large amounts of ATP and may thus provide the energy supply required for endothelial-derived relaxing factor production[85].

Role of PKC

PKC and its physiological role: PKC is a family of serine/threonine kinases that consists of at least 12 members[86] and can be classified into three groups: conventional PKC, novel PKC, and typical PKC. PKC can be activated through multiple pathways in response to a wide array of stimuli, including cytokines, mechanical shears, stresses, hormones, and even glucose. With exposure to hyperglycemia in diabetes, accumulation of the glycolytic intermediate, glycerol-3-phosphate, stimulates the de novo synthesis of DAG, which in turn activates specific isoforms of PKC. In addition, chronic hyperglycemia can also increase the production of AGEs and generate ROS, which have been shown to activate the DAG-PKC pathway[87] (Figure 1).

Hyperglycemia increases circulating cytokines, growth factors, and hormones such as endothelin-1 and angiotensin II; these secreted cytokines can also activate PKCs by binding to their cell surface receptors[88,89]. Diabetes is also associated with severe dyslipidemia, in which increased levels of circulating free fatty acids have been reported to activate PKC either directly or through de novo synthesis of DAG in endothelial cells[90,91].

PKC isoforms and diabetic cardiomyopathy: In the myocardium, several isoforms of PKC are activated to the membrane fraction of the heart by hyperglycemia: PKC-βII and -δ are the major isoforms to be activated by chronic hyperglycemia. Targeted overexpression of the βII isoform in transgenic mice results in a cardiac phenotype reminiscent of that seen in diabetic cardiomyopathy, characterized by early diastolic dysfunction, small vessel disease, myocardial hypertrophy and loss of cardiac contractility and cardiomyocytes[92,93]. This is probably due to PKC-βII-mediated phosphorylation of troponin I, which may decrease myofilament Ca²⁺ responsiveness[94]. This observation is consistent with a previous finding that PKC activation can induce phosphorylation of troponin I and T and downregulation of calcium-stimulated ATPase in actomyosin, with subsequent inhibition of cardiac sarcoplasmic reticulum Ca²⁺ accumulation which in turn reduces cardiac contractility[95]. Overexpression of PKC-βII resulted in extensive cardiac fibrosis, probably due to upregulation of the expression of fibrosis-promoting factors such as transforming growth factor-βI and connective tissue growth factor[93]. These factors can further result in the transcription and deposition of extracellular matrix components such as collagens and fibronectins. This phenomenon is consistent with the observation that collagen and fibronectin deposition is increased in myocardial tissues from diabetic humans and animals[96]. Loss of cardiomyocytes has been blamed, in part, for the ventricular dysfunction in diabetic hearts and is probably caused by hyperglycemia-induced cardiomyocyte apoptosis[97]. Activation of PKC-delta is known to induce cellular apoptosis in many cell types including neutrophils, keratinocytes, neuronal cells, fibroblasts, transformed cells and cardiomyocytes[98].

PKC targeted therapies for diabetic cardiomyopathy: Because pathologic activation of the DAG-PKC pathway has been shown to play key roles in the onset and progression of cardiovascular complications, efforts have been focused on developing effective approaches to regulate PKC activities, and therefore, to reverse or even prevent these lethal complications. Several PKC inhibitors have been developed, but few of them demonstrated PKC isoform specificities, with the exception of a PKC-βII-selective inhibitor (LY333531) that has been demonstrated to ameliorate diabetes-induced PKC-βII activation and its related vascular abnormalities in cell culture, animal models and clinical trials[99]. Specific receptors for activated C kinases (RACKs) regulate PKC translocation and confer PKC isoform-selective-interactions. Peptides have been designed according to the protein sequence of RACK and have been shown to inhibit the translocation of specific isoforms of PKC[100]. By a similar strategy, peptide activators of specific PKC isoforms have been developed from a pseudo RACK sequence[101]. These peptide inhibitors and activators have been used successfully to evaluate cardiac functions both in vitro and in vivo [102,103].

Role of AGEs

Glucose is known to bind to free amino groups in proteins or to lipids. Through a series of oxidative and non-oxidative reactions, AGEs are formed irreversibly and accumulate in tissues over time. Although AGE formation occurs during the normal aging process, it is markedly accelerated during diabetes, as a consequence of an increase in substrates such as glucose and in the prevailing oxidant stress in diabetes[104]. AGEs promote atherogenesis by Ox-LDL and by causing changes in the intimal collagen. A major contribution of AGEs to atherogenesis, however, emerges from important studies that have led to the isolation of a RAGE on the cell surface. RAGE acts as a signal transduction receptor and also binds with non-AGE-related proinflammatory molecules such as S100/calgranulins and amphoterins[105]. The overlapping accumulation and expression of RAGE and its ligands at sites of tissue lesions sustain RAGE-mediated cellular activation and the induction of multiple signaling pathways.

The endothelium is exposed to AGEs localized on circulating proteins or cells (for example diabetic red blood cells), as well as to AGEs present in the underlying subendothelial matrix. Receptors for AGEs have been found on the endothelial cell surface, and mediate both the uptake and transcytosis of AGEs, as well as the internal signal transduction. AGE-RAGE interaction causes alteration of the barrier function that has been documented by an increased permeability of endothelial cells incubated with AGEs and increased transit of macromolecules through the endothelial monolayer. The increase in permeability is accompanied by alterations in the physical integrity of the endothelium, as shown by the destruction of organized actin structures and alterations in cellular morphology[106]. Binding of AGEs to endothelial RAGE also results in the depletion of antioxidant defense mechanisms (e.g. glutathione, vitamin C)[107] and the generation of ROS[108], leading to increased oxidative stress (Figure 1). As a consequence, NF-κB activation occurs and thus promotes the expression of NF-κB regulated genes, including procoagulant tissue factor and adhesion molecules such as E-selectin, intercellular adhesion molecule-1 and vascular adhesion molecule-1[109,110]. In addition, interaction of RAGEs leads to an increase in thrombomodulin and also activates the receptors for the cytokines, IL-1, TNF-α and growth factors, causing the migration and proliferation of smooth muscle cells.

AGEs linked to the vascular matrix can chemically interfere with the bioavailability of NO. AGEs, when added to NO in vitro, block NO activity in a concentration-dependent manner. Studies using animal models with experimentally-induced diabetes demonstrated that an alteration of endothelium-dependent dilation occurs quickly within 2 mo from diabetes induction[111]. Presumably, the inactivation of NO occurs through a direct reaction of the NO radical with other free radicals that are formed during the reactions of advanced glycation. Interestingly, AGEs impair endothelial NOS activity in rabbit aorta and femoral artery[112]. In human saphenous vein endothelial cells, PCR upregulates RAGE expression in a dose and time-dependent manner, enhancing the binding ability of RAGE with its endogenous ligands[113]. In parallel, AGEs induce the expression of the potent vasoconstrictor, endothelin-1, changing endothelial function towards vasoconstriction[114]. Zhang et al[115] have reported that AGEs induce phosphorylation of 2 signal transduction kinases (ERK1/2 and JNK) and produce inflammatory responses in adventitial cells of porcine coronary arteries. Antioxidants and inhibitors of NAD(P)H oxidase may attenuate this inflammatory response. Moreover, the authors hypothesized that AGE signaling inhibition may produce a vascular protection: soluble RAGE, which prevents AGE-mediated signaling, has been shown to inhibit vascular inflammation and lesion formation in diabetic apolipoprotein E-deficient mice[116,117].

Diabetic cardiomyopathy as a “stem cell” disease

Diabetic cardiomyopathy is described as the structural and functional changes caused by death of cardiac cells, with chronic loss of myocytes and vascular cells, which lead to a decrease in muscle mass, chamber dilation, impaired ventricular function, and finally, to symptoms of heart failure. Recent and accumulating evidence supports the concept that the heart possesses a store of multipotent progenitor cells (stem cells), which have the ability to differentiate into myocytes, endothelial cells and smooth muscle cells, both in vivo and in vitro[118-120]. Based on these observations, it has been postulated that under diabetic conditions, an imbalance between myocytes death and their regeneration may occur, with defects in both growth and survival of the store of multipotent cells in the heart that can be mediated by at least two underlying different mechanisms: the enzymatic o-glycosylation of proteins and oxidative stress-mediated stem cell damage. Ventricular acute hyperglycemia can promote apoptosis of ventricular myocytes via enzymatic glycosylation and the activation of the transcription factor p53 and effector responses involving the local renin-angiotensin system[97]. Moreover, in the same study, angiotensin II secretion and its receptor binding can promote the generation of ROS and initiate an oxidative-stress mediated cell death. This demonstrates that hyperglycemia can act through both direct and indirect mechanisms. Previously, the targeted deletion of the p66^shc gene was shown to decrease ROS and improve cell resistance to oxidative damage, prolonging their life[121]. Based on this observation, Rota et al[122] used an insulin dependent-streptozotocin induced diabetes model in wild type and p66^shc Knockout mice, in order to test the hypothesis that oxidative stress in diabetes can alter the store of multipotent progenitor cells homeostasis, producing aging, loss of growth and death of stem cells, leading to diabetic cardiomyopathy. They found that p66^shc deletion was followed by a reduction in oxidative stress, in cardiac stem cell senescence and that it favors the activation of cell growth mechanisms. However, the most important finding was that it prevented the increase in cardiac size and shape, diastolic and systolic dysfunction, which are the main features of the diabetic-failing heart. If we can consider diabetic cardiomyopathy as a “cardiac stem cell compartment disease”, it will open new previously unexpected therapeutic options in the near future.

MECHANISMS RESPONSIBLE FOR IMPAIRED CORONARY ARTERIOLAR VASOMOTION

Myogenic response

Microvessels respond to an increase or a decrease in transmural pressure by constriction and dilation, respectively. Active myogenic responses are present in both subendocardial and subepicardial arterioles: Kuo et al[123] also demonstrated that subepicardial arterioles exhibited greater vasodilatory responses at low pressure and augmented constriction at higher pressures. Changes in local regulatory mechanisms, intrinsic to the vascular wall, such as pressure sensitive myogenic response, have been proposed to contribute to the decreased dilator capacity of skeletal vessels in type 2 diabetes mellitus; because coronary vascular resistance is influenced by myogenic reactivity, enhanced myogenic tone could adversely affect vasodilator function of arterioles. In type 2 diabetic db/db mice, Lagaud et al[124] found that in mesenteric arterioles, there is enhanced pressure-induced myogenic tone due to the upregulation and activation of smooth muscle PKC. In obese Zucker rats, Frisbee et al[125] also reported enhanced myogenic tone in skeletal muscle arterioles. In contrast, Bagi et al[12] found that no significant difference between active and passive diameters of coronary arterioles of db/db and control mice developed at 80-mmHg intraluminal pressure. Moreover, the myogenic tone of arterioles in response to stepwise increases in intraluminal pressure from 20 to 120 mmHg was also not significantly different in the two groups, indicating that in db/db mice, enhanced myogenic constriction is unlikely to be responsible for the decreased vasodilation of coronary arterioles (Figure 2).

Figure 2 Possible mechanisms involved in vasomotion of impaired coronary arterioles in diabetes. Several mechanisms have been postulated to be responsible for impaired vasomotion in coronary artery disease. There is strong evidence accumulating in favor of each impairment category as causal. We hypothesize that they interact in as yet unspecified ways rather than operate through separate pathways to cause diabetes. Major challenges in this field include better understanding each of these mechanisms, but the greatest opportunity for seminal breakthroughs may reside in reconciling our understanding among these mechanisms and their roles in diabetes.

Endothelium-dependent NO mediated-dilation

One of the primary in vivo physiological stimuli for local regulation of arteriolar diameter is the presence of intraluminal blood flow. Increases in intraluminal flow elicit endothelium-dependent vasodilation via the release of vasodilator substances, such as NO[126]. In vivo flow-mediated dilation in skeletal muscle microvessels was significantly reduced in type 2 diabetic obese Zucker rats compared with controls[125]. Lagaud et al[124] and Pannirselvam et al[127] previously demonstrated that in mesenteric arteries of db/db mice, dilations in response to ACh were reduced, suggesting impaired endothelium-dependent NO-mediated dilation. Interestingly, they found unaltered dilation in response to the NO donor, sodium nitroprusside (SNP); therefore, they speculated that the reduced NO-mediated dilation might be due to a decreased synthesis of NO caused by reduced availability of the eNOS substrate, L-arginine, or reduced levels of tetrahydrobiopterin (BH₄). These results have been confirmed by Bagi et al[45], who studied the effects of acute hyperglycemia on skeletal muscle arterioles of healthy rats. They found that transient elevation of glucose concentrations resulted in the reduction of NO-mediation of flow-induced dilation; acute hyperglycemia was likely to elicit enhanced production of superoxide which reduced the bioavailability of NO and the level of the NOS cofactor, BH₄, thereby eliciting a reduction in flow-induced arteriolar dilation.

Little information is available in the literature regarding the effect of diabetes mellitus on endothelial function in coronary arterioles. Ammar et al[128] demonstrated that, in an in vivo beating heart, dilation of epicardial coronary arterioles to the endothelium-dependent vasodilator, ACh, was impaired in diabetic animals, while responses to adenosine and SNP were intact. Topical application of SOD and catalase restored ACh vasodilatory responses suggesting a pivotal role of ROS in destroying NO. However, which ROS was responsible for endothelial dysfunction was unclear. Bagi et al[45] showed that in coronary arterioles isolated from diabetic mice, NO mediation of flow- and agonist-induced dilation was reduced. Nevertheless, the dilation in response to the NO donor, NONOate, was also decreased suggesting that an alteration in NO synthesis, due to reduced levels of L-arginine or BH₄, is unlikely to be the main cause of the decreased dilation. In fact, the authors found enhanced vascular production of superoxide anions which is likely to interfere with the mediation by NO of flow- and agonist-induced dilation. The reduced dilations to flow, ACh, and NONOate could be reversed by administration of SOD to the organ chamber, suggesting that in coronary arterioles of db/db mice, an enhanced level oxidative stress is present in both the endothelial and smooth muscle layers of microvessels. Interestingly, Zhang et al[129] showed that even if endothelial dysfunction is a main characteristic of I/R models, the underlying mechanisms may be different. They found that the process is partially mediated by increased arginase activity that leads to reduced availability of L-arginine, and partially mediated by the downregulation of endothelial NOS production. They also obtained the same results in hypertension-mediated endothelial dysfunction[129]. Bagi et al[45] showed that NAD(P)H is likely to be the main source of the enhanced superoxide production in coronary microvessels. In fact, in vivo administration of the NAD(P)H inhibitor, apocynin, restores flow-induced coronary arteriolar dilation in mice with type 2 diabetes. They also demonstrated that short-term treatment of type 2 diabetic mice with the PPAR-γ activator, rosiglitazone, augments NO-mediated flow-dependent dilations of coronary arterioles by reducing vascular superoxide production via a favorable alteration of oxidant/antioxidant enzyme activities (Figure 2).

Hypoxia-induced vasodilation

Hypoxia is known to induce potent endothelium-dependent vasodilation. Several animal studies have reported that ATP-sensitive potassium channels (K_ATP) play a pivotal role in mediating such a vasodilation in conduit and resistance arteries[130]. It is not well established how hypoxia elicits vasodilation, but interestingly, Quayle et al[131] demonstrated that anoxia, but not hypoxia, is enough to activate the K_ATP current in rat femoral artery, suggesting that part of the relaxant effect of hypoxia may be mediated by changes in intracellular [Ca²⁺] through modulation of calcium channel activity. However, Miura et al[13] described the vasodilation to hypoxia and the role of K_ATP in human coronary arteries from patients with diabetes. They confirmed that the mechanism of hypoxia-induced vasodilation involves opening of K_ATP and they found that vasodilation to both hypoxia and K_ATP stimulation is impaired in both type 1 and type 2 diabetes mellitus in the human coronary microcirculation. The exact mechanism of impaired vasodilation is not completely understood. The impairment seems to be specific to the K_ATP mechanism, because no reduction is observed in vascular smooth muscle cell relaxation followed by a decrease in intracellular Ca²⁺ concentration either by cGMP production attributable to NO (SNP) or by membrane hyperpolarization through Ca²⁺-activated K⁺ channel activation. Therefore, the reduced hypoxia-induced dilation in coronary arterioles could be responsible for the impaired myocardial perfusion in patients affected by diabetes mellitus by hindering vasodilator responses during ischemia (Figure 2).

Vasodilation during increased metabolic demand

Coronary arterioles are responsible for the close coupling of coronary blood flow and myocardial oxygen consumption (MVO₂)[132], although the exact coupling mechanism remains unclear. Coronary blood flow may increase five- to six-fold from baseline[133]. The primary means by which oxygen delivery to the myocardium may be increased during higher demand is via coronary dilation. Microvascular dilation during raised MVO₂ is heterogeneous according to the vessel size[132], with the greatest magnitude of dilation being inversely related to baseline coronary diameter. Adenosine has received most attention as a potential mediator of metabolic vasodilation, but its role has been seriously questioned. Jones et al[134] reported that NO participates in coronary microvascular dilation during increases in metabolic demand with rapid atrial pacing. Furthermore, Embrey et al[135] demonstrated that coronary microvascular dilation during increase in MVO₂ by a dobutamine and pacing control is virtually abolished by NOS inhibition. In contrast, several authors reported that glibenclamide prevents the increase in coronary blood flow associated with increase in MVO₂, demonstrating a possible role for K_ATP channels in mediating faster coronary blood flow during larger oxygen consumption[136].

Ammar et al[137] demonstrated that coronary blood flow during increase in MVO₂ is impaired in hyperglycemic dogs. Coronary arteriolar dilation to ACh is impaired, while dilatory response remains unaltered during administration of the NO-donor, SNP, demonstrating that impaired dilation during hyperglycemia is selective for endothelium-mediated dilation. The authors also reported that both endothelium-derived contracting factors, such as thromboxane A and prostaglandin H₂, and free radicals are likely to cause impaired metabolic dilation during hyperglycemia, because reversal of either mechanism permits normal vasodilation to raise the metabolic demand (Figure 2).

The work by Xu et al[138] on the role of LOX-1 in atherosclerosis provides direct evidence that endothelial dysfunction in atherosclerosis is mediated, at least in part, via the interaction of Ox-LDL with its receptor, LOX-1, which in turn stimulates endothelial generation of superoxide radicals by activation of NAD(P)H oxidase. The results of this study contribute to the development of novel adjunctive therapies using anti-Ox-LDL and/or anti-LOX-1 antibodies or soluble receptors to prevent endothelial dysfunction following atherosclerosis. This work is an example of how basic research can provide important insight into the underlying mechanisms of ischemic heart disease, obesity, type 2 diabetes and atherosclerosis and provide valuable guidance on therapeutic design.

CONCLUSION

Our understanding of type 2 diabetes has begun to recognize many factors involved in etiology, pathophysiology and clinical and microvascular manifestations of this common disease. The mechanisms involved in the etiology of microvascular complications have been described separately, but the signaling pathways always interact to amplify one another and induce endothelial dysfunction in diabetes. However, hyperglycemia is known to be the primary culprit in the pathogenesis of diabetic microvascular complications. This induces acute changes in cellular metabolism such as glycation and consequent inactivation of protein involved in the control of microvascular function. Hyperglycemia also activates and being activated by several other mechanisms, particularly the generation of AGEs, polyol, activation of DAG PKC pathways, and chronic or subclinical inflammation. The common final pathway is, however, the increase in oxidative stress, which seems to play a pivotal role in diabetic endothelial dysfunction and cardiomyopathy (Figure 1). Among the mechanisms responsible for impaired vasomotion, endothelial dysfunction is most likely to play the primary role (Figure 2). Knowledge gained from these studies will help further understand the increased cardiovascular risk and development of chronic vascular disease in type 2 diabetes. Furthermore, the quest to identify proximal stimuli for diabetes may provide a solution to the specific problem and new approaches for aiding development of therapeutic strategies. Future studies will gauge their utility as guides to monitor therapy.