Glycoxidative Stress and Cardiovascular Complications in Experimentally-Induced Diabetes: Effects of Antioxidant Treatment

Abstract

Diabetes mellitus (DM) is a common metabolic disease, representing a serious risk factor for the development of cardiovascular complications, such as coronary heart disease, peripheral arterial disease and hypertension. Oxidative stress (OS), a feature of DM, is defined as an increase in the steady-state levels of reactive oxygen species (ROS) and may occur as a result of increased free radical generation and/or decreased anti-oxidant defense mechanisms. Increasing evidence indicates that hyperglycemia is the initiating cause of the tissue damage in DM, either through repeated acute changes in cellular glucose metabolism, or through long-term accumulation of glycated biomolecules and advanced glycation end products (AGEs). AGEs are formed by the Maillard process, a non-enzymatic reaction between ketone group of the glucose molecule or aldehydes and the amino groups of proteins that contributes to the aging of proteins and to the pathological complications of DM. In the presence of uncontrolled hyperglycemia, the increased formation of AGEs and lipid peroxidation products exacerbate intracellular OS and results in a loss of molecular integrity, disruption in cellular signaling and homeostasis, followed by inflammation and tissue injury such as endothelium dysfunction, arterial stiffening and microvascular complications. In addition to increased AGE production, there is also evidence of multiple pathways elevating ROS generation in DM, including; enhanced glucose auto-oxidation, increased mitochondrial superoxide production, protein kinase C-dependent activation of NADPH oxidase, uncoupled endothelial nitric oxide synthase (eNOS) activity, increased substrate flux through the polyol pathway and stimulation of eicosanoid metabolism. It is, therefore, not surprising that the correction of these variables can result in amelioration of diabetic cardiovascular abnormalities. A linking element between these phenomena is cellular redox imbalance due to glycoxidative stress (GOS). Thus, recent interest has focused on strategies to prevent, reverse or retard GOS in order to modify the natural history of diabetic cardiovascular abnormalities. This review will discuss the links between GOS and diabetes-induced cardiovascular disorders and the effect of antioxidant therapy on altering the development of cardiovascular complications in diabetic animal models.

INTRODUCTION

In the past few decades, diabetes mellitus (DM) has rapidly increased in the world. It has been estimated that the number of diabetic patients will more than double within 20 years. DM is mainly characterized by the development of increased morbidity and mortality due to cardiovascular disease (CVD) [1-3]. Diabetic cardiovascular complications appear to be multifactorial in origin, but in particular, glycoxidative stress (GOS) has been suggested to be the unifying link between the various molecular disorders in DM. The biochemical process of advanced glycation end products (AGEs), is accelerated in DM due to chronic hyperglycemia-induced oxidative stress (OS), which has been postulated to play a central role in this disorder [4-6]. DM has been suggested to be a major source of AGEs in vivo [7]. The consequences of GOS include; damage to DNA, lipids, proteins, disruption in cellular homeostasis and accumulation of damaged molecules. These damaged molecules disrupt endothelial cells and cardiomyocytes and impair cardiovascular reactivity [7]. Nitric oxide (NO) regulates vasodilatation, anticoagulation, leukocyte adhesion, smooth muscle proliferation and the antioxidative capacity of endothelial cells [8]. Hyperglycemia has been reported to decrease the availability of NO in animals and humans, which may contribute to the hemodynamic and physiological changes occurring in DM [9]. G protein-adenylyl cyclase signaling, diacylglycerol (DAG)/ protein kinase C (PKC) pathway and calcium current/ entry, which play an important role in the regulation of cardiovascular functions has also been reported to be impaired in diabetes-induced GOS [10]. The harmful effects of hyperglycemia-induced GOS on cardiovascular function and the link to pathophysiological mechanisms underlying diabetic complications have lead to the development of new pharmacotherapeutic strategies. Although, aldose reductase inhibitors and PKC inhibitors represent a potential therapeutic approach to prevent the onset or the progression of these complications, recent interest has focused on strategies to prevent firstly GOS and then its harmful effects [11, 12]. Indeed, there is accumulating evidence that an interaction between AGE and its receptor (RAGE) generates OS and subsequently evokes vascular inflammation and thrombosis, thereby playing a central role in diabetic vascular complications [13]. This review presents evidence of the pathophysiology and cellular mechanisms of cardiovascular complications in experimental DM; discuss the therapeutic strategies with antioxidant compounds and evaluates the pyridoindole derivatives as potential remedy for diabetes-induced metabolic, structural and functional abnormalities in the cardiovascular system.

EXPERIMENTAL DIABETES-INDUCED CARDIOVASCULAR COMPLICATIONS

The Mechanisms of Abnormal Cardiac Contractility in DM

Patients with DM are more prone to develop cardiomyopathy than nondiabetic subjects [14]. This complication is serious and is a major cause of mortality due to DM. According to the Framingham study, the incidence of cardiovascular morbidity per year is 39.1% in diabetic males and 17.2% in diabetic females; chronic heart failure afflicts 7.6% of diabetic males and 11.4% of diabetic females. Diabetic cardiomyopathy is a specific syndrome, consisting of cardiomegaly, left ventricular dysfunction, electrical remodeling of the ventricle, and symptoms of congestive heart failure [15, 16]. Hypertension is recognized as an important cofactor in the development of fatal congestive heart failure in diabetics [17]. The pathophysiology of diabetic cardiomyopathy is incompletely understood and several mechanistic approaches are under debate. Decreased left ventricular compliance and increased interstitial connective tissue have been observed in chronically diabetic animals [17]. In contrast, ventricular myocardium from diabetic rats exhibit a reversible decrease in the speed of contraction, prolongation of contraction, and a delay in relaxation [17, 18]. These mechanical changes are associated with changes in the responses to alpha- and beta-adrenergic and cholinergic stimulation [18]. We have shown that the changes in myocardial beta-adrenergic responsiveness may contribute to heart failure in streptozotocin (STZ)-diabetic rats [19]). Both insulin-dependent diabetes mellitus (IDDM) [20] and non-insulin-dependent diabetes mellitus (NIDDM) [21] causes significant alterations in myocardial contractility induced by adrenergic agonists. In our laboratory, it has been showed that experimental diabetes produces a marked decrease in the rate of spontaneously beating right atria and a significant increase in basal contractile force of left ventricular papillary muscle. In diabetic rat atria, the positive chronotropic effect of noradrenaline was markedly increased, while the positive chronotropic effect of isoprenaline was significantly decreased. STZ-diabetes also induced an increased positive inotropic response of papillary muscle to both noradrenaline and isoprenaline [19, 21]. In addition, the ultrastructural degeneration of the diabetic myocardium (i.e. intracellular oedema, myofibrillar fragmentation, condensed pleomorphic mitochondria, thick capillary irregular basement membrane and swollen endothelial cells) was associated with mechanical alterations caused by adrenergic agonists [21]. Long-term experimental IDDM, characterized by ventricular remodeling, overt resting left ventricular sys systolic dysfunction and decreased left ventricular responsiveness to beta-adrenergic receptor stimulation is associated with histopathological changes indicating dilatational remodeling as increased capillary tortuosity and type IV collagen in capillaries and cardiomyocytes [22]. Insulin replacement therapy in STZ-diabetic rats can prevent these changes [19, 20], and thyroid hormones mediate the effect of insulin therapy [19]. These findings were recently confirmed in a study using conscious diabetic rats, which showed a decrease in basal mean arterial pressure, heart rate, +dP/dtmax and -dP/dtmax. In that study, insulin prevented the hemodynamic and myocardial function alterations observed in STZ-diabetic rats [23]. Another study indicated that male and female rats are affected differently by DM in terms of left ventricular developed pressure (LVDP) responses to beta-adrenergic receptors (beta-ARs) stimulation, probably due to the difference between their beta-ARs induced cyclic adenosine monophosphate (cAMP) responses [24]. Alterations in the myocardial beta-AR signaling pathway in two different models of rat NIDDM, has also been addressed; it showed that the beta-AR signaling system of the insulin deficient model is altered more than the hyperinsulinemic model and the exaggerated cAMP response of NIDDM [25, 26]. Hearts from diabetic db/db mice, a model of NIDDM, exhibited left ventricular failure and altered metabolism of exogenous substrates, was prevented by PPAR-alpha and -gamma agonists [27]. Diabetic cardiomyopathy was expressed by a decreased beta1-AR mRNA levels, and an increased beta2- and beta3-ARs mRNA levels in STZ-treated diabetic rat hearts [28, 29]. Insulin-deficient diabetes was associated with a reduced expression and concomitant functional loss of G inhibitor protein in ventricular cardiomyocytes [30]. Long-term DM results in a marked decrease in the positive inotropic response of left atria to electrical field stimulation due to the impairment of noradrenaline release from sympathetic nerve terminals [31]. On the other hand, it has been shown that an alpha-adrenergic receptors (alpha-ARs)-mediated electrophysiological response is unmasked when the beta-AR-mediated response is desensitized in the papillary muscle of STZ-induced diabetic rats [32]. Alpha1-adrenergic responsiveness of rat myocardium was reported to be higher in experimental diabetes and partly reversed by insulin treatment [32]. This increase has been suggested to serve as a compensatory mechanism for the lower beta-adrenergic responsiveness in this organ [32]. The results showed that diabetes-induced impairment in cardiac alpha 1-AR signaling is closely linked with the abnormal activation of cardiac PKC [33, 34]. Other studies also indicated that presynaptic alpha 2-ARs and muscarinic receptors that are linked to the inhibition of noradrenaline release during nerve stimulation are functionally impaired in diabetic animals [30].

The altered cholineacetyltransferase, cholinesterase activities and choline concentration has been suggested to be the reasons for the supersensitivity to muscarinic agonists in the myocardium of STZ-diabetic rats [35]. In addition, cardiac M2-muscarinic receptor (M2-MR) gene expression was increased in STZ-diabetic rats [35, 36]. Alterations in the atrial muscarinic system in STZ-diabetic rats have been suggested to be a consequence of impaired functional muscarinic receptor-G protein coupling [37, 38]. In fact, aldose reductase inhibitors have been successfully used to treat and prevent muscarinic receptor-mediated cardiac complications in experimental DM [35]. In the light of previous reports, diabetes-induced functional abnormalities in the heart may be attributed to altered adenosine action. The purinergic response of rat atria was increased in STZ-diabetes [39]. It has been demonstrated that the changes in adenosine receptor gene expression and protein content correlated with some of the abnormalities characterized in the diabetic heart [39]. An increase in adenosine A1 receptor protein content, without a change in mRNA level was observed in isolated cardiac myocytes. DM caused an increase of adenosine A3 receptor mRNA and protein content in the heart and cardiac myocytes. The level of adenosine A2a receptor mRNA was increased in the whole diabetic heart, but it decreased in cardiac myocytes with undetectable changes in protein content. Interestingly, the administration of insulin to diabetic rats for four days resulted in the adenosine receptors, mRNA and protein content returning to levels observed in the heart of normal rats [39, 40].

Hyperglycemia-induced up-regulation of the endothelin (ET)-system is another important event in the pathogenesis of diabetic CVD. The measurement of ET-1, ET (A) and ET (B) receptors and mRNAs showed a significant increase in mRNA levels in the hearts of diabetic rats. ET-1 was detected in the cardiomyocytes, endothelium and smooth muscle cells of the larger blood vessels by immunohistochemistry, which diabetes increased. Autoradiographic localization of ET-1 receptors showed increased binding in the endothelium and myocardium of diabetic animals [41]. ET-1 was also elevated in atherosclerotic plaques. Diabetic patients have accelerated atherosclerosis and also show elevated plasma levels of ET-1. Inhibiting the actions of ET-1 and the signaling pathways are thought to represent potential therapeutic targets for the prevention of atherosclerotic diabetic disease [42].

Metabolic impairment such as hyperglycemia, hyperlipidemia, hypoinsulinemia, and alterations in cardiac metabolism are the main reasons for vascular structural and functional changes, which instigate the cellular effects leading to redox imbalance, interstitial fibrosis, myocyte death, and disturbances in ion transport and homeostasis [43]. A decrease in myosin ATPase, a shift in myosin isoenzyme distribution, alterations in a variety of Ca²⁺ fluxes associated with cardiac mechanical alterations are observed in experimental DM [44]. A deficiency in sarcoplasmic reticulum Ca²⁺ uptake contributes to a lower beta-ARs-stimulated inotropic response of diabetic myocardium [45]. Chronic STZ-diabetes leads to Ca²⁺ accumulation in cardiac myocytes [46]. In accordance with this, the changes in intracellular Ca²⁺ mobilization has been suggested to account for the depressed IGF-I-induced inotropic response in the diabetic heart [47]. Indeed, alterations in intracellular Ca²⁺ handling play a pivotal role in the impairment of cardiac function in diabetes. The reduced function of the sarcoplasmic reticulum Ca²⁺-ATPase and Ca²⁺-release in response to beta-AR challenge, is thought to have a causative role in the depressed hemodynamic performance of the challenged heart in the early stage of DM [47]. The observed changes in the contractility, together with Ca²⁺ handling of the diabetic heart are most likely attributable to functional disturbances of SERCA2a, RyR2 and NO/ cyclic guanosine monophosphate (cGMP) systems in response to beta-adrenergic activation [48,49]. The changes in cardiac mechanical parameters correlate with decreased ATP and increased lactate in the inner layer of the left ventricle in diabetic rats. But, of course, this cannot be explained by the rate of decrease in total ATP and lactate accumulation alone [50]. Recent evidence implicates disturbances in cardiac energy metabolism; in the uncontrolled diabetic state, cardiac myocytes use fatty acids almost exclusively to support ATP synthesis [51].

IMPAIRED VASCULAR REACTIVITY IN DM: ROLE OF ENDOTHELIAL CELLS

Vascular disease, a complicating feature of DM, is characterized mainly by atherosclerotic and coronary artery abnormalities [52]. Diabetic patients have a 2- to 4-fold increase in the incidence for coronary artery disease and 10-fold increase in peripheral vascular diseases, due in part, to accelerated atherogenesis [53]. Alterations of vascular smooth muscle function have been widely implicated in the development of vascular complications and circulatory dysfunction in DM [54]. However, the influence of DM on catecholamine-induced vasopressor activity under in vitro conditions is contradictory. Although, a decreased responsiveness of the rat aorta to noradrenaline and phenylephrine has been noted in STZ-diabetes [55-57], many investigators have demonstrated an increased responsiveness of the rat aorta to alpha-AR stimulation in the same model [58, 59]. Similarly, we found an increased responsiveness of isolated aorta to alpha-AR agonists in rats with at least six weeks diabetes, which was reversed by insulin treatment [60-64]. The duration of DM might be the main reason for this discrepancy [65, 66]. Our observations using diabetic rat aorta were subsequently confirmed by the finding of an increase in the contractility of the internal mammary artery and saphenous vein obtained from diabetic patients undergoing coronary artery bypass surgery [67]. The increased responsiveness of the diabetic aorta to vasoconstrictor agents was associated with the abnormal structure of the smooth muscle and endothelial cells [62, 63].

We demonstrated that the entry of Ca⁺² into the cytosolic membrane of diabetic rat aorta was augmented by the increased activation of voltage (L-type) and receptor-operated channels, since there was an increased contraction to L-type Ca²⁺ channel activator BayK-8644 and AR-receptor agonist phenylephrine [68, 69]. Moreover, Ca²⁺ overload and/or impaired Ca²⁺,Mg²⁺-ATPase activity might also been responsible for these abnormalities [45]. Indeed, other investigators suggested that changes in functional Ca²⁺ store sizes and the Ca²⁺ entry seems to be responsible for the alterations in contractile responses to phenylephrine following DM [65, 70]. In fact, diabetes-induced contractile disturbances are not only related to the activity of Ca²⁺ channels or Ca²⁺ current but also the involvement of K⁺ channels. A diminished relaxation response to cromakalim, a K_ATP channel opener, has been reported in diabetic vessels by us, as well as others [71-73].

In 1980, Furchgott discovered that vascular endothelial cells produce an endothelium-derived relaxing factor (EDRF) in response to stimulation by ACh [74]. Vascular relaxation to ACh and a number of other agonists was found to be dependent on the presence of an intact endothelium. In 1987, Moncada et al. and Ignarro et al. separately, proved that EDRF is NO [75, 76]. A year later, Moncada also demonstrated that NO is synthesized from the amino acid L-arginine [77]. Earlier, Murad had shown that the relaxing effect of nitrovasodilators depended on the activation of soluble guanylate cyclase to increase cGMP levels in the arteries [78]. All these works have contributed to the establishment of NO as a signaling molecule in the cardiovascular system. In 1992, NO was described as the molecule of the year [79] and the importance of the NO discovery was recognized with the Nobel Prize in Physiology and Medicine awarded to Furchgott, Ignarro and Murad. Subsequently, NO had been implicated in the pathogenesis of diseases such as DM, atherosclerosis and hypertension. Today, it is hard to find a disease which is not associated with altered NO homeostasis. In 1989 it was reported that endothelial cells contain a cytosolic enzyme that is directly or indirectly regulated by Ca²⁺, which converts L-arginine into a compound that stimulates guanylate cyclase and behaves similarly to EDRF [80]. In the same year we reported, for the first time, that endothelium-dependent relaxations produced by ACh and histamine in aortic rings precontracted with noradrenaline was significantly increased in 6 weeks STZ-induced, IDDM [60]. In contrast, the relaxations elicited by those agents were significantly attenuated in aorta of NIDDM rats. [60]. Previously, an attenuated response of isolated aorta to endothelium-dependent vasodilators had been reported by Oyama et al. in STZ-diabetic rats [81] and by Durante et al. in Bio Bred (BB) rats [82]. However, we found that the abnormal endothelial function in diabetic aorta was especially dependent on hyperglycemic control and the duration of DM [61, 62].

Over the years, many investigators have demonstrated some abnormalities in endothelium-dependent relaxation in arteries from a number of experimental animal models of DM [83-87], while others have found no changes in response to endothelium-dependent vasodilators [88]. In this context, the endothelium of diabetic aorta exhibits decreased, increased or unchanged responsiveness to ACh depending on the model, severity and/or duration of DM [88]. Now, endothelium-dependent vasodilatation is recognized as a reproducible and accessible parameter to probe endothelial function in different pathophysiological conditions, including DM, aging and CVD. Similarly, we confirmed endothelial dysfunction in studies of human NIDDM [67]. The impaired endothelium-dependent vasodilatation might arise from decreased NO production, enhanced inactivation of NO, impaired diffusion of NO to the underlying smooth muscle cells, decreased responsiveness of the smooth muscle to NO and enhanced generation of endothelium-derived constricting factors. As a common observation, we and many others reported that the relaxation of diabetic aorta induced by sodium nitroprusside (an endothelium-independent and cGMP-mediated relaxant agent) was comparable to control tissue, suggesting that vascular smooth muscle exhibit a normal responsiveness to a guanylate cyclase stimulator such as NO, in the diabetic state [60-62, 84, 85]. Despite a growing understanding of the mechanisms by which NO is released from the endothelium, the precise molecular mechanisms by which NO relaxes vascular smooth muscle are not fully understood. Previous studies have provided evidence that a major action of NO in normal vascular smooth muscle is to reduce intracellular free Ca²⁺ levels via cGMP-dependent and-independent mechanisms. It was also proposed that NO can activate SERCA by a cGMP-independent process [89, 90]. NO influences vascular homeostasis in many ways beyond modulation of vasomotion, such as inhibition of smooth muscle cell proliferation, platelet aggregation, platelet and monocyte adhesion to the endothelium, LDL oxidation, expression of adhesion molecules and endothelin production [91].

Besides cyclooxygenase and nitric oxide synthase (NOS), another distinct endothelial pathway is endothelium-dependent hyperpolarization (EDHF), which is also involved in the relaxation of the vascular smooth muscle cells. EDHF has been demonstrated unequivocally in various blood vessels from different species, including human and is likely to play an important role in cardiovascular physiology. This alternative pathway involves the activation of two populations of endothelial K⁺ channels, the small conductance and intermediate conductance Ca²⁺-activated K⁺ channels, which are altered in DM [92, 93]. On the other hand, the interaction between endothelial cells and monocytes is thought to play an important role in normal vascular biology, while an increased/prolonged interaction is implicated in the pathogenesis of vascular disease [91]. Monocyte adhesion to the endothelium suppresses the release of biologically active NO in monocyte in a concentration-and time-dependent manner, via an IL-1B dependent pathway [91, 94]. This event would predispose the vessel to develop atherosclerotic lesions due to the potential consequence of increased macromolecular permeability and intimal proliferation via the development of vasospasm, thrombosis and recruitment of platelets and leukocytes [91, 94]. In this respect, we confirmed the monocyte adhesion to the endothelial cells and impaired morphology of diabetic aorta by electron microscopy [63, 64, 68, 95]. Beside monocytes, there are many other inhibitors of the biological activity of NO, such as decreased L-arginine uptake, decreased co-factors (Ca²⁺, calmodulin, BH4), inhibition of electron flow (nicotinamide adenine dinucleotide phosphate; NADPH, flavins), inhibition of NOS expression, inhibition of substrate binding to NOS, and NO scavengers [97, 98].

GLYCOXIDATIVE STRESS IN THE NATURAL HISTORY OF DIABETES-INDUCED CARDIOVASCULAR COMPLICATIONS

Although the mechanisms responsible for mediating endothelial dysfunction have not been completely defined, elevated OS plays a key role in the development of endothelial dysfunction, as well as, atherosclerosis in diabetes-associated vascular disease. Increased basal production of superoxide (O₂^–), and hydrogen peroxide (H₂O₂) in the aorta has been shown in early studies using diabetic animals [98, 99]. The main reasons for the diabetes- induced increase in OS are; an increased generation of oxygen-derived free radicals through autoxidation of glucose, AGE-formation, increased substrate flux through the polyol pathway, and stimulation of eicosanoid metabolism, sorbitol-DAG metabolism and NOS [4-7, 11, 13, 92, 96, 100, 101]. Glucose-induced changes in the activity of NADPH oxidase and endothelial nitric oxide synthase (eNOS) can cause vascular endothelial cell dysfunction via disregulation of eNOS and/or changes in the expression of the subunits of NADPH oxidase [102]. Generally, the development of OS is due to the excessive formation and/or insufficient removal of highly reactive molecules such as, reactive oxygen species (ROS) and reactive nitrogen species (RNS) [100-103]. ROS includes; free radicals such as O₂^–, hydroxyl (OH), peroxyl (RO₂), hydroperoxyl (HRO₂^–), as well as, non-radical species such as H₂O₂ and hydrochlorous acid (HOCl). RNS include free radicals such as; NO and nitrogen dioxide (NO₂^–), as well as non-radicals such as peroxynitrite (ONOO^–), nitrous oxide (HNO₂) and alkyl peroxynitrates (RONOO). There are multiple sources of OS in DM, which include non-enzymatic, enzymatic and mitochondrial pathways [4-7, 11, 101, 103]. Non-enzymatic sources of OS originate from the oxidative biochemistry of glucose. Hyperglycemia can directly cause an increase in ROS generation. Glucose can undergo autoxidation and generate OH radicals [103]. In addition, glucose can react with proteins in a non-enzymatic manner leading to the development of Amadori products followed by the formation of advanced glycation end-products AGEs; ROS is generated at multiple steps during this process. The formation of AGEs progressively increases with normal aging, even in the absence of disease. However, they are formed at an accelerated rate in DM due to uncontrolled hyperglycemia. DM is a major source of AGEs production in vivo and is not only a marker but also an important causative factor for the pathogenesis of the disease. AGE modification of proteins leads to alterations in their normal function, abnormal signal transduction pathways and tissue injury all contribute to the development of serious complications, which is responsible for the morbidity and mortality observed in diabetic patients [6, 7, 12, 13, 104-106]. Binding and activation of cellular RAGE by AGEs or similar ligands can also promote OS and AGE formation via the NAD(P)H oxidase and the myeloperoxidase pathways. Another potential mechanism of AGE formation is the polyol pathway. Glucose entering the polyol pathway may form AGEs via reactive intermediates, i.e. glyoxal, methylglyoxal or 3-deoxyglucosone, as well as, via depletion of NAD(P)H or glutathione raising intracellular OS, all of which indirectly result in the increased formation of AGEs [104-106]. There are two types of cell surface AGE receptors, those that bind AGEs and initiate cell activation and those that bind and degrade AGEs. RAGE is the best studied receptor of the first category; it recognizes AGEs and initiates OS. The second group of receptors includes; AGER1, AGER3 and CD36. The best evaluated receptor in this category is AGER1, which has been found to have marked anti-oxidant properties [12, 13, 106, 107]. In the presence of chronic hyperglycemia, there is enhanced metabolism of glucose through the polyol (sorbitol) pathway, which also results in enhanced production of O₂^– [106]. Enzymatic sources for increased generation of ROS in DM include mainly NOS, NAD(P)H oxidase and xanthine oxidase [13,104,108,109]. All isoforms of NOS require five cofactors/prosthetic groups such as flavin adenine dinucleotide (FAD), flavin mononucleotide (FMN), heme, BH4 and Ca²⁺-calmodulin. If NOS lacks its substrate L-arginine or one of its cofactors, NOS may produce O₂⁻instead of NO and this is referred to as the uncoupled state of NOS [109, 110]. NAD(P)H oxidase is a membrane associated enzyme that consists of five subunits and is a major source of O₂^– production [109]. It has been demonstrated that the enhanced production of O₂^– in the diabetic vasculature, predominantly mediated by the increased levels of NAD(P)H oxidize also stimulates PKC activity [109]. It has been suggested that mitochondrial O₂^– is the initiating “snowball” that turns OS into an avalanche in DM by stimulating more ROS and RNS production via downstream activation of nuclear factor-kappaB (NF-KB)-mediated cytokine production, PKC and NAD(P)H oxidase [111-113].

In light of the OS hypothesis, we previously investigated the time course for changes in endothelium-dependent relaxations using isolated aorta rings. It was found that endothelium-dependent and -independent relaxant responses of diabetic aorta are more transient than those of non-diabetic control vessels and the time required to reach peak relaxation after the addition of ACh was shorter in aortic rings isolated from 6-month STZ-diabetic rats compared to non-diabetic rats [114]. In our study, pretreatment of diabetic vessels with superoxide dismutase (SOD) normalized the recovery phases of endothelium-dependent and -independent relaxations, but had no effect on the peak responses to ACh and sodium nitroprusside. In the presence of diethyldithiocarbamate, an inhibitor of SOD, the transient nature of the relaxant response to ACh or sodium nitroprusside was found to be more marked and the peak relaxations were inhibited; these effects of diethyldithiocarbamate were more pronounced in diabetic than in control rings. Catalase caused a decrease in the peak relaxant response to ACh and accelerated the fading of the relaxation in diabetic aorta. While, pretreatment with 3-amino-1,2,4 triazole, (a catalase inhibitor), inhibited the peak relaxant response to ACh in diabetic rings. The combination of SOD plus 3-amino-1,2,4 triazole produced an increase in the transient nature of endothelium-dependent relaxation of diabetic aorta, which was greater than that with 3-amino-1,2,4 triazole alone. Neither catalase nor 3-amino-1,2,4 triazole affected the characteristics of sodium nitroprusside-induced relaxation. Desferrioxamine, an inhibitor of OH radical production, or mannitol, a OH radical scavenger, had no effect on the characteristics of either ACh- or sodium nitroprusside-induced relaxation in control and diabetic aorta [114]. Similarly, others have reported that the rapid fade of the endothelium-dependent responses of diabetic vessels is significantly suppressed by pretreatment with SOD but not with catalase, deferoxamine, allopurinol, or indomethacin [115]. The OS-related rapid inactivation of NO has been recognized as an important factor leading to diabetes-induced endothelial dysfunction [92, 114-117].

We, as well as, others have found that the diabetic aorta is more sensitive to ROS exposure when compared to non-diabetic, healthy aorta [118, 119]. H₂O₂ induced a brief contraction before relaxation of aorta rings with an intact endothelium that was more pronounced in diabetic vessels. Removal of the endothelium or pretreatment of the rings with L-nitro-arginine methyl ester (L-NAME), a NO synthase inhibitor, immediately abolished the H₂O₂-induced response with a transient increase in tone, while pre-incubation with indomethacin had no significant effect on aortic contractions induced by H₂O₂ in long-term diabetic and healthy control animals. In addition, long-term STZ-diabetes caused a significant increase in the maximum relaxation induced by H₂O₂ in aorta already preconstricted with alpha-AR agonists. Preincubation with SOD had no effect on the maximum relaxation induced by H₂O₂, in both non-diabetic and diabetic animals but led to an increase in H₂O₂-induced contractions in non-diabetic rats. Aortic rings preincubated with diethyldithiocarbamate; H₂O₂ produced only contractions in non-diabetic rats, while H₂O₂-induced relaxations were markedly depressed in diabetic animals. Catalase did not affect the smooth muscle tone of intact or endothelium-denuded aortic rings, stimulated by H₂O₂ and aminotriazole failed to affect H₂O₂-induced contractions or relaxations of diabetic and non-diabetic vessels. These series of experiments suggest that long-term DM causes a decrease in vessel SOD activity that may lead to a change in the dismutation reaction of O₂⁻ to H₂O_2, causing an increase in endogenous O₂^– anions. The diabetes-induced increase in catalase activity could account for the increased production of harmful OH in the aorta. The enhanced endogenous formation of OH might be an important factor in the development of vascular disorders associated with chronic DM [118].

The excessive production of O₂⁻ in DM plays a significant role in the inhibition of EDRF/NO bioavailability under basal physiological conditions. The increased vasorelaxation caused by exogenous stimulants of NOS in the early stages of experimental DM [60-62] might compensate for an environment where the basal endothelial relaxation is decreased. Moreover, the involvement of endogenous H₂O₂ during ACh-stimulated endothelium-dependent relaxation might be increased, which is likely to account for the maintenance of the vasorelaxant effect of ACh in the later stage of DM. This phenomenon of an enhanced contribution of H₂O₂ in the establishment of stimulated endothelium-dependent relaxation could account for the maintenance of NOS-mediated relaxation in diabetic arteries [114, 118, 119], is also supported by Matoba et al. These authors reported that in some pathological conditions eNOS generates O₂⁻, which is converted by SOD to H₂O₂ facilitating endothelium-dependent relaxation [120]. It has also been reported that acute administration of scavengers of O₂⁻, including SOD [113, 120], the combination of SOD with catalase [121] or OH radical scavengers, restored or normalized the abnormal endothelium-dependent responses in different models of DM [115, 121-123]. Nevertheless, an increased production of OH might also mediate the increased vasoconstriction tendency of diabetic vessels under basal condition. This is supported by the observation that OH can catalyze the production of vasoconstrictor prostanoids [122-124]. We subsequently revealed an increased SOD-induced relaxation of diabetic aorta when compared with non-diabetic aorta. Pretreatment of aortic rings with L-NAME inhibited SOD-induced relaxation, which was greater in diabetic rings compared to control aorta. While, catalase inhibited SOD-induced relaxation only in diabetic rings. Pretreatment with the cyclooxygenase inhibitor, indomethacin, or the catalase inhibitor, aminotriazole, had no effect on SOD-induced relaxation in both control and diabetic vessels. These findings suggested that: (i) under normal physiological conditions, the relaxant effect of SOD is related to the inhibition of O₂⁻-induced EDRF/NO destruction in the rat aorta; (ii) in the diabetic state, an excess of O₂⁻ increasingly inhibits basal EDRF/NO, and exogenously added SOD causing an enhancement of the dismutation reaction resulting in the excess O₂⁻ forming H₂O₂. The production of H₂O₂, a vasorelaxant molecule, probably accounts for the increased responsiveness of diabetic rings to exogenous SOD [119]. On the other hand, the responsiveness of the diabetic aorta may reflect an important facet of H₂O₂ linked to its specific modulatory role that might compensate for an increase in vascular reactivity in response to alpha-AR stimulation due to OS. It is well known that the vasorelaxation of alpha-AR agonist-precontracted artery rings by H₂O₂ requires the formation of cGMP, and is mediated by NO [125]. H₂O₂ also augments the secretion of other endothelium-derived relaxing peptides, e.g., adrenomodulin and C-type natriuretic peptide, which are thought to compensate for vasoconstriction due to increased OS, in conditions such as, hypertension, atherosclerosis as well as DM [126]. In fact, under hyperglycemic conditions the excessive production of O₂⁻ interacts with the basal production of EDRF/NO, leading to an increase in the generation of the highly reactive molecule ONOO⁻. In this situation, the chemical environment of NO (i.e. an increase in O₂⁻), may be a determining factor whether NO exerts protective or harmful effects [127]. Accordingly, it has been reported that a greater production of O₂⁻ and a greater formation of nitrate (NO is metabolized by O₂⁻ to NO₃⁻) occurs in aortic rings from diabetic rats [128]. It is also known that ONOO⁻ can modify tyrosine residues in various proteins to form nitrotyrosine, which can lead to damage that alters protein function and stability [128, 129]. Thus, nitrotyrosine is a marker of OS [129, 130], in diseases such as atherosclerosis, heart disease and DM [129]. Several studies have shown that the formation of nitrotyrosine and ONOO⁻ are factors in NO responsiveness and NO production in blood vessels [129, 130]. In this regard, we compared the vasodilatory effect of exogenous ONOO⁻ with sodium nitrite in precontracted aorta isolated from STZ-induced diabetic and age-matched control rats [131]. ONOO⁻ produced a concentration-dependent relaxation of aortic rings with/without endothelium. Vasorelaxation occurred with higher concentrations of the decomposition product ONOO⁻ or sodium nitrite, although these relaxations were considerably slower with reduced sensitivity. Endothelium intact rings were less sensitive to the vasorelaxant effect of ONOO⁻ than endothelium-denuded rings, in control but not in diabetic aorta. The maximum relaxation to ONOO⁻ was also increased in diabetic vessels, which was unaffected by the removal of the endothelium either in diabetic or control rings. DM did not alter the relaxations elicited by both decomposed ONOO⁻ and sodium nitrite. ONOO⁻-induced relaxation was not inhibited by diethylenetriaminepentaacetic acid, an inhibitor of OH formation. While, pretreatment with ONOO⁻ significantly suppressed the phenylephrine-induced tone and ACh-stimulated endothelium-dependent relaxation; both effects were more pronounced in diabetic than in control aorta. Taken together, these findings suggest that there is a relationship between the increased responsiveness to exogenous ONOO⁻ and the depressed basal NO bioavailability in diabetic aorta [131]. The reported inhibitory effect of ONOO⁻ on NO bioavailability and the elevation in endothelial dysfunction by angiotensin II in diabetic animals is consistent with the findings. Angiotensin II has been found to increase the generation of ROS within vascular smooth muscle cells, primarily through activation of membrane-bound NADPH/NADH oxidase, which promotes a progressive deterioration of the cardiovascular system in STZ-induced diabetic rats [132, 133]. Diabetes-induced over production of angiotensin II may also in turn, trigger the dysfunction of vascular smooth muscle SERCA by the generation of ONOO⁻ [134]. In fact, the physiological actions of NO and O₂⁻ may depend crucially on their local concentrations and their intracellular site of formation. In order to accumulate as H₂O₂, endothelial O₂⁻ must first escape interacting with NO molecules; under normal conditions, at least in some blood vessels, it appears that a small amount of O₂⁻ survives long enough to be acted upon by SOD to generate H₂O₂. However, under pathological conditions, the production of O₂⁻ (largely from sources other than eNOS) may overwhelm the production of NO in many blood vessels. Under these circumstances, H₂O₂ production will be limited by the availability of SOD, and endothelium derived ONOO⁻, prostanoids, and possibly isoprostanes may tip the balance toward vasoconstriction [92, 124, 135].

Although the mechanisms underlying diabetes-induced endothelial dysfunction are likely to be multi-factorial, in-vivo and in-vitro studies indicate a crucial role for the DAG-PKC pathway in mediating this phenomenon. PKC may have adverse effects on vascular function, including the activation of O₂⁻-producing enzymes such as NADPH oxidase, as well as, increased expression of a dysfunctional O₂⁻ producing uncoupled (eNOS III). PKC-mediated O₂⁻ production may inactivate NO derived from eNOS III, and inhibit the activity and/or expression of the NO downstream target; soluble guanylyl cyclase [124, 135]. During the development of DM a number of biochemical and mechanical factors converge on the endothelium, resulting in endothelial dysfunction and vascular inflammation. Hyperglycaemia leads to increased formation of AGE, which quenches NO and impair endothelial function. AGEs act directly to induce cross-linking of proteins such as collagen to promote vascular stiffness, and, thus, alter vascular structure and function. Receptor-mediated actions of AGEs can induce intracellular signaling that leads to enhanced OS and increase key proinflammatory and prosclerotic cytokines [6, 7, 12, 104]. Increased vascular inflammation, including enhanced expression of interleukin- 6 (IL-6), vascular cellular adhesion molecule-1 (VCAM-1) and monocyte chemoattractant protein (MCP-1) are observed, as is a marked decrease in NO bioavailability [136].

GLYCOXIDATIVE STRESS: A KEY THERAPEUTIC TARGET?

The Role of Antioxidants in Prevention and Delaying Cardiovascular Complications Associated with Experimental DM

A possible target for the management of the multiple risk factors in diabetic CVD could be the treatment of endothelial dysfunction, since all the risk factors are related to endothelial dysfunction [135-139]. As mentioned above, various interventions have proven effective in restoring impaired endothelium-dependent vasodilatation in certain vascular beds in animal and human DM. In this regard, optimizing eNOS function is a reasonable approach; this can be realized by using co-factors, e.g., BH4, L-arginine, or by increasing cGMP availability via phosphodiesterase type 5 (PDE5) inhibitors, or by enhancing eNOS expression. [138]. Several studies have reported on the modulating effect L-arginine supplementation has on diabetic in vivo NO production, under conditions known to be associated with endothelial dysfunction [139]. Angiotensin-converting enzyme (ACE) inhibition, which is conventionally used for the management of CVD, has been shown to; decrease angiotensin II-induced NADH oxidase activity and vascular production of O₂⁻, to stimulate basal NO production by suppression of bradykinin breakdown or perhaps by potentiating the vascular effects of insulin [140]. In addition to ACE inhibitors, physical exercise, calcium channel blockers and angiotensin II receptor antagonists have been found to ameliorate flow-evoked endothelium-dependent vasodilatation in diabetic patients and animals [141]. We and others, showed an improvement in endothelium-dependent vasodilation, as well as, myocardial structure and function following the administration of lipid-lowering statins to diabetic rats [142, 143]. More recently, endothelin A receptor antagonists, atresantan or 1-methylnicotinamide, a primary metabolite of nicotinamide preserved endothelium-dependent relaxation in diabetic vessels [144, 145]. Interestingly, a considerable body of evidence suggests that increasing NO bioavailability via antioxidant therapy is an important intervention not only for endothelial dysfunction but also other cardiovascular abnormalities induced by DM [22, 63, 68, 119].

The inhibition of intracellular free radical formation would provide a therapeutic strategy to prevent OS and the related diabetic vascular complications. In this regard, the determination of ROS production is an important evaluation of the intracellular and extracellular redox balance and homeostasis. It would be possible determine the level of ROS generating enzymes in the extracellular and intracellular compartments depending on location, e.g., NAD(P)H/NADH oxidase in the cell membrane, xanthine oxidase, myeloperoxidase, cytochrome P450 in cytoplasm, and mitochondrial oxidation and antioxidant enzymes such as, SODs, glutathione peroxidase, heme oxygenase, thioredoxin peroxidase/peroxiredoxin, catalase, and paraoxonase [112, 146]. Although there is controversy about the antioxidant status in DM, our biochemical measurements revealed that SOD activity is inhibited in diabetic aorta together with the activation of catalase [114]. In addition to this, hearts from 12 weeks diabetic animals showed about a four-fold increase in the level of thiobarbituric acid reactive substances (TBARS), indicative of increased lipid peroxidation. This was accompanied by approximately 100 % increase in both catalase and glutathione peroxidase (GSHPx) enzyme activities [147]. We have also found; increased GSHPx and catalase activity, decreased glutathione reductase (GR) activity and SOD activity unchanged in the diabetic heart. The activity of pentose phosphate pathway enzymes, glucose-6-phosphate dehydrogenase (G-6PD) and 6-phosphogluconate dehydrogenase (6-PGD) were significantly decreased in diabetic aorta [148]. In an attempt to neutralise OS, cells utilize antioxidant defenses comprised of enzymatic and non-enzymatic compounds that determine redox balance. Enzymes include; SOD, catalase, thioredoxin, and GSHPx, while, non-enzymatic compounds include; glutathione, ascorbate, tocopherol and polyphenolic compounds [109, 149]. SOD immediately converts O₂^– to H₂O₂, which is then detoxified to water either by catalase in the lysosomes or GSHPx, in the mitochondria. GR, which regenerates glutathione, is used as a hydrogen donor by GSHPx during the elimination of H₂O₂. Similarly, SOD and GSHPx expression/activities were decreased whereas; catalase was increased in experimental models of DM [101, 150, 151]. The modulation of these enzymes in target organs prone to diabetic complications such as the heart may prove beneficial in the prevention and management of heart failure. Accordingly, the use of gene transfer of extracellular SOD to study the endothelial effects of OS has recently been reported. Gene transfer of normal extracellular SOD improved endothelial dysfunction in several disease states, thus, this technique may be useful to study the mechanism by which gene variants predispose vessels to endothelial dysfunction and vascular disease [152]. More recently, Tempol, a cell permeable SOD mimetic, was found to increase endothelium-dependent vasodilatation in arteries from hypertensive animals, most likely through the lowering of ROS, making it a potential therapeutic tool for a disease characterized by decreased NO bioavailability [141].

Several reports using experimental diabetic models have shown that improving glycemic control does not sufficiently improve the accompanying OS [153]. We found that in vivo insulin replacement therapy alone was unable to prevent or restore metabolic, functional and/or structural abnormalities in the heart and aorta of STZ-diabetic rats [19-21, 61, 62]. However, the combination of insulin with non-enzymatic antioxidants showed more effectiveness in the prevention and amelioration of vascular [154] and cardiac [147] abnormalities in STZ-diabetic rats. In this context, to avoid interference from glucose toxicity, the increased availability of AGE detoxification systems, which include; enzymatic mechanisms such as the glyoxalase-I and II system, anti-oxidant defenses, circulating proteins that trap AGEs, receptor-dependent intracellular uptake and destruction, as well as, urinary excretion have stimulated great interest. Thus, methylglyoxal, a highly reactive AGE precursor molecule, is detoxified by glyoxalase I and II. Its reduction can be catalyzed by aldehyde reductase, aldose reductase and carbonyl reductase. While, a group of circulating proteins such as lysozyme, defensins and lactoferrin can bind AGEs, preventing them from causing cellular toxicity or from binding to other molecules [12, 104, 106]. Consequently, the use of antiglycation agents, AGEs crosslink breakers, including alagebrium, benfotiamin, aminoguanidine, carnosine may be considered as future therapeutic approaches to improve endothelial function, vascular reactivity and cardiac hemodynamics in diabetic patients [106, 155-157]. Another treatment strategy has been recently suggested that involves the increase in AGER1 function (the chief endocytic AGE receptor), which inhibits the function of RAGEs and may provide protection against the destructive effect of a chronically high OS state such as in DM [12, 13, 106].

Over the past three decades, many in vivo studies have been performed utilizing exogenous antioxidants in experimental diabetic models. The beneficial effects of this form of treatment on OS can be measured by certain observable biomarkers. These markers include the enzymatic activities of catalase, SOD, GSH-Px, GR and TBARS levels. The indirect measurement of free-radical production can also be undertaken, i.e., malondialdehyde (MDA) and F₂-Isoprostanes, and more recently 4-hydroxynonenal, which has been shown to be consistently elevated in DM [11, 108, 117, 158]. Normalization of the activity of any of these markers, and ultimately, the balance of free-radical production/removal, would be an effective method to reduce ROS-induced damage. Many animal studies have been conducted with this aim in mind and indeed have shown that diabetes-induced alterations of OS indicators can be reversed when the animals are treated with various antioxidants.

Experiments from “the ADIC Study Group”, observed the effectiveness of exogenous antioxidant treatments in the recovery of OS markers and endogenous antioxidant enzyme activities in STZ-diabetic rats [22, 63, 64, 154, 159-161]. These recoveries were associated with the amelioration in vascular and cardiac metabolism/function [22, 46, 63, 64, 69, 95, 119, 142, 147, 148, 154, 159-161]. We have previously focused on investigating the effects of dietary supplementation of the antioxidants vitamin E, alpha-lipoic acid and vitamin A. 0.5% dietary vitamin E (DL-alpha tocopheryl acetate) supplementation given to STZ-diabetic rats eliminated the accumulation of lipid peroxides and returned plasma triglycerides toward normal levels. In addition, diabetes-induced abnormal aortic contractility and endothelial dysfunction were significantly but not completely prevented by vitamin E treatment. It also protected the morphological integrity of the diabetic aorta [63]. Vitamin E treatment (400-500 IU/kg/day) prevented protein glycation, and Ca2+accumulation in STZ-diabetic rats [46] and gluthation-dependent antioxidant defense in the aorta and/or heart and did not affect the decreased activity of G-6PD, 6-PGD in diabetic aorta [148]. It also restored isoprenaline-stimulated relaxation, as well as, structural and metabolic alterations in the diabetic aorta [95]. Other actions of this antioxidant include; the prevention of (1) capillary basement membrane thickening, vascular endothelial growth factor expression and aldose reductase enzyme activity in retina of diabetic rats [161], (2) Ca²⁺-ATPase activity, protein glycosylation and/or Ca²⁺ levels in the brain [162] and in the kidney [163], ameliorates impaired sympathetic neurotransmission and the contractility of isolated vas deferens [160], protects nerve blood flow and conduction velocity [164, 165]. In general, studies with vitamin E showed that it protects vascular endothelium and cardiac function in experimental models of DM [166, 167]. Studies on the effect of vitamin E supplementation and deprivation in diabetic patients have also been undertaken [168]. Although plasma vitamin E levels was significantly lower in poorly controlled NIDDM patients with coronary heart disease, small clinical trials revealed the beneficial cardiovascular effect of vitamin E supplementation [169-172]. In a double-blind, placebo-controlled, randomized study, vitamin E supplementation (1000 IU/day) given to IDDM patients for three months significantly improved endothelium-dependent vasorelaxation [170]. Similarly, in another study the administration of vitamin E (800 IU/day) and C (1000 mg/day) in combination for six months had a positive effect on endothelium-dependent vasorelaxation in IDDM patients but had no effect in NIDDM [171]. Results from the HOPE trial, investigating myocardial infarction, stroke and death from CVD in diabetic patients 55 years of age or older was published in 2000, it found no significant difference in the primary outcome between vitamin E and placebo treated groups [172]. A recent randomized trial reported that there was no significant overall effects of treatment with vitamin C + vitamin E + beta-carotene on the risk of developing NIDDM in patients at high risk of CVD [173]. However, another recent clinical study suggest that supplementation with vitamin E accompanied by a change in lifestyle might help to reduce the damage brought about by free radical toxicity in DM [174]. Another randomized clinical trial comparing the effectiveness of alpha and gamma isomers of vitamin E demonstrated that at every dose of alpha vitamin E, gamma vitamin E concentration was significantly suppressed. It was thought that if a standardized preparation of gamma vitamin E (without the alpha isomer) became available it may be useful in the prevention of atherosclerosis in NIDDM patients [175].

Alpha lipoic acid (ALA; thioctic acid) is another important chemical with antioxidant properties that can be used in DM studies. ALA scavenges OH radical, HOCl and singlet oxygen. It can also exert antioxidant effects in biological systems through transitional metal chelation. In its reduced form dihydrolipoic acid (DHLA) it is not only an antioxidant but also has pro-oxidant properties in systems where OH radicals are generated. Thus, the ALA/DHLA ratio helps to recycle endogenous antioxidants such as vitamin E [176,177]. We found that, ALA treatment (50 mg/kg/day) effectively reversed body weight, blood glucose, plasma insulin, cholesterol, triglycerides and lipid peroxidation levels of 12 weeks STZ-diabetic rats. In addition, it partially improves blood pressure, catalase activity and tissue morphology and completely reversed the contractile effect of phenylephrine in diabetic aorta [64]. In another study, we showed that ALA treatment also partially reverse SOD-induced relaxation, which is probably related to the elimination of O₂⁻ and H₂O₂, mediating the recovery of basal EDRF/NO availability [119]. The potency of ALA in reversing hypertension due to an effect on vascular reactivity and tissue morphology, as well as, its general metabolic actions in diabetic rats, confirms the importance of hyperglycemia-induced OS in the development of diabetes-induced vascular complications [178, 179]. The metabolic and hemodynamic sequela insulin-deficiency in the heart is also corrected by ALA [180]. ALA reduces OS in STZ-induced diabetic rats by modulating the thiol status of the cells [177, 180], thus, preventing high glucose-induced receptor death and mitochondriondependent cardiac apoptosis [181], it also increases HSP60 and decreases 4-HNE in diabetic heart [182]. Clinical trials with ALA have focused on the treatment of diabetic neuropathy, the results obtained are more promising than those with vitamin E. Pharmacodynamic studies have shown that ALA favorably influences the vascular abnormalities of diabetic polyneuropathy, such as impaired microcirculation, the increased indices of OS and increased levels of markers for vascular dysfunction, such as thrombomodulin, albuminuria, and NF-KB [183, 184]. Thus far, seven controlled randomized clinical trials giving ALA to patients with diabetic peripheral and cardiac autonomic neuropathy have been completed; (Alpha-Lipoic Acid in Diabetic Neuropathy [ALADIN I-III] [185, 186], Deutsche Kardiale Autonome Neuropathie [DEKAN] [187], Oral Pilot [ORPIL] [188], Symptomatic Diabetic Neuropathy [SYDNEY] [189], Neurological Assessment of Thioctic Acid in Neuropathy [NATHAN I- II] [190]. These studies have used different; study designs, treatment duration, doses, sample sizes, and patient population. They have revealed the efficacy and safety of thioctic acid (600 mg intravenously for 3 weeks) in diabetic patients with symptomatic polyneuropathy [184, 191].

We demonstrated that 12 weeks STZ-diabetes caused a reduction in plasma retinol and retinol-binding protein (RBP), which was significantly improved by insulin therapy (8-10 IU/rat/day s.c), whereas, vitamin A (retinol acetate, 30 mg/kg/day i.o) treatment failed to increase plasma retinol and RBP levels. These findings suggest that the combination of insulin with vitamin A provides more benefit than either agent alone in the treatment of the general characteristics of DM and that the maintenance of the antioxidant defense of the diabetic heart can reduce peroxidative stress-induced cardiac injury [147]. In addition, our findings suggest that the maintenance of ACh-stimulated endothelium-dependent vasorelaxant tone within normal physiological levels depends largely on the prevention and/or inhibition of peroxidative stress, and that the use of vitamin A together with insulin provides better metabolic control with a reduction in diabetes-induced vascular complications than when insulin is used alone [154]. Vitamin A is also an ingredient of the Cod Liver Oil (CLO). CLO is mainly comprised of; omega-3 polyunsaturated fatty acids (PUFAs), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) as well as vitamin A. CLO treatment caused significant improvements in hyperlypidemia and glucose metabolism, allowing the recovery of the normal growth rate of STZ-diabetic rats. It also ameliorated OS/peroxidative stress and endogenous antioxidant enzyme activities in the heart and blood vessels [22, 159]. In addition, CLO treatment also completely prevented endothelial deficiency and partially corrected the phenylephrine-induced vasoconstriction of diabetic aorta. The impaired responsiveness of the right atria to the positive chronotropic effect of isoprenaline and noradrenaline and the positive inotropic effect of the papillary muscle to both agents, and abnormal ultrastructure of cardiomyoctes were also significantly ameliorated by CLO [22, 159]. Several reports have shown that a combination of antioxidants may be superior to the use of monotherapy.

The Steno-2 trial compared the effect of a multifactorial intensive therapy (n=80) with that of conventional treatment (n=80) on modifiable risk factors for CVD in patients with NIDDM. The patients received pharmacotherapy that targeted hyperglycemia, dyslipidemia, hypertension and microalbuminuria including daily supplementation of vitamin C (250 mg), E (100 mg), folic acid (400 mg) and chromium picolinate (100 mg), as well as, behavioral modification including; low-fat diet, exercise and smoking cessation. The control group received conventional therapy as recommended by national guidelines. The intensive therapy resulted in almost a 50% decrease in the risk of cardiovascular events, providing evidence that a multifactorial approach is superior to conventional therapy for the prevention of OS-induced vascular complications in DM [192]. However, a more recent trial (POPADAD) does not provide evidence to support the use of aspirin or antioxidants in the primary prevention of cardiovascular events and mortality in the diabetic population studied [193]. Treatment with Zycose, a new compound, provided therapeutic benefit believed to be due to its additive effects with other drugs; lowering homocysteine levels (folic acid), reducing the production of advanced glycation end products (benfotiamine), improving endothelial function (folic acid, benfotiamine, ALA), reducing oxidative stress (ALA, vitamin E) and reducing carbonyl stress (benzamine) [194].

Probucol is another antioxidant compound used in experimental DM to determine its effect on cardiovascular function. Treatment with this antioxidant prevented lipid abnormalities, sorbitol pathway enzymes and nerve blood flow and conduction velocity deficits in diabetic rats. The effectiveness of probucol was more apparent when used in combination with essential fatty acids (evening primrose oil) [164]. Short term probucol treatment (300 mg/kg/day, i.o.) also resulted in partial restoration of endothelial function and noradrenaline-induced contraction in aorta without any alterations in morphological abnormalities in STZ-diabetic rats [68]. Essentially, probucol is a diphenolic compound that reduces atherosclerosis and restenosis probably through anti-hyperlypidemic, anti-inflammatory and anti-oxidant properties [150]. It also preserves pancreatic beta-cell function through a reduction of OS in diabetic patients [195, 196].

In addition to the antioxidants mentioned above, a number of commonly used drugs have been reported to have antioxidant activity, in addition to their primary pharmacological property. For example, gemfibrozil, a lipid lowering fibrate, was previously reported to have antioxidant actions [197]. However, although it reduced lipid peroxidation, it did not alter blood glucose or tissue antioxidant enzymes in chronically diabetic rats [198]. This suggests that the beneficial effect of short-term gemfibrozil treatment in reducing lipid peroxidation, endothelial dysfunction and vascular reactivity are attributed to its effect on circulating lipids [198, 199]. Antihyperlipidemic statins are thought to exert an antioxidant effects via a variety of mechanisms. Statin therapy (10 mg/kg/day,orally) markedly reduces OS markers and ultra structural impairments in diabetic rat myocardium [142]. It also ameliorated diabetes-induced vascular abnormal vasoconstriction and endothelial dysfunction via an affect on; oxidizing metabolism, NO bioavailability and intracellular Ca²⁺ mobilization [143]. In addition, we have determined the vascular effect of thiazolidinediones, using human internal mammary artery, human saphenous vein and diabetic rat aorta in vitro [73, 200, 201], the beneficial effect were, at least in part linked with their antioxidant properties, similar to those shown in diabetic animals [202].

BENEFICIAL EFFECTS OF PYRIDOINDOL ANTIOXIDANTS IN DM

Numerous studies have proven that stobadine (STB), a novel synthetic pyridoindole, is an efficient antioxidant [203]. During the last decade, in collaboration with the Slovak Academy of Sciences, we examined the antioxidant properties of STB and related pyridoindole derivatives. The protective effect of STB against glycoxidative damage using an in vitro experimental glycation model has been reported [204]. STB also attenuated diabetes-induced pathological changes in various organs, decreased albuminuria, enzymuria, lipid peroxidation and matrix collagen cross-linking [205]. In addition, it reduced plasma cholesterol and triglyceride levels in STZ-diabetic rats [46, 206] Recently, we showed that STB caused a significant decrease in protein carbonylation, advanced oxidation protein products and normalized protein thiol, total thiol, non-protein thiol groups in the liver of diabetic animals [207]. The combination of STB with vitamin E was beneficial in reducing hyperglycemia and lipid peroxidation in diabetic animals. The antihyperglyceamic effect of these antioxidants may be due to the early stabilization of the pancreas beta cell membrane. These antioxidants may partially protect insulin-secreting cells against the detrimental effects of free radicals when given at the onset of diabetes or promptly after STZ injection [46]. STB was also able to control the glycation of hemoglobine, determined by HbA1c levels, which was significantly decreased by long-term STB treatment of chronically STZ-diabetic rats [160]. STZ-diabetic rats, orally treated with a low dose of STB (24.7 mg/kg/day), showed an inhibition in cardiac protein glycation and Ca²⁺accumulation. These effects were associated with decreased plasma lipid peroxidation. However, STB treatment did not influence the diabetes-induced inhibition of cardiac Ca²⁺,Mg²⁺-ATPase activity [46]. This is supported by the finding that long term STB treatment prevents diabetes-induced deterioration of cardiac Na⁺,K⁺-ATPase [208] and reduces oxidative damage of myocardial tissue, as measured by conjugated dienes. It also reverses myocardial levels of alpha-tocopherol and coenzyme Q9 to near control values. Moreover, it attenuates the elevated activity of SOD and diminishes angiopathic and atherogenic processes in diabetic myocardium [206-208]. Diabetes-induced stimulation of cardiac GSH-Px activity and inhibition of cardiac GR activity was prevented by STB treatment [148]. These findings suggest that the cardioprotective effect of STB is more complex than its preservation of mitochondrial function [209].

The combination of STB and vitamin E provided more benefits than each individual therapy alone in the reduction of lipid peroxidation and cardiac abnormalities in diabetic animals [46]. This was evident in the; kidney, brain [148], liver [46], peripheral nerves [160, 164] and retina of diabetic rats [161]. Interestingly, the increase in retinal capillary basement membrane thickness (RCBMT) was 12.34% in diabetic rats treated with STB and 23.07% in diabetic rats treated with vitamin E. While, the increase in RCBMT was just 4.38% following combination therapy [161]. However, combination therapy did not protect against leukocyte free radical production, while STB alone had a significant protective effect on leukocyte function [210]. Although STB and vitamin E produces parallel effects on the preservation/restoration of the pro-oxidant feature of DM and generally has similar characteristics in the protection of tissue function; some differences have also been observed. For instance, while the in vivo administration of vitamin E alone had no effect, STB treatment completely returned noradrenaline-induced contractions to basal levels in diabetic vas deferens [160]. The effect of long term treatment with STB or vitamin E in the development of cataracts in diabetic animals has also been investigated. Interestingly, at the end of the experimental period, the visual cataract score was significantly decreased in the diabetic groups treated with STB, while vitamin E had no significant effect [211]. It is conceivable that an antioxidant given alone in vivo would be more effective when other antioxidants are present. It is well known that alpha-tocopherol can recycle stobadine [212], thus the antioxidant potency of stobadine in vivo may be increased by its interaction with other antioxidants [213]. We anticipated that stobadine given in a mixture with vitamin E would have enhanced antioxidant and anticataract action when given to diabetic rats. However, contrary to our expectation, combined treatment with STB + vitamin E accelerated the progression of cataract, without affecting the overall visual cataract score. Moreover, biochemical analysis showed a further decrease of lens protein free sylfhydryls in diabetic rats treated with the mixture of STB + vitamin E compared to the untreated diabetic group, indicating a prooxidant effect of the STB + vitamin E cocktail [212, 213]. Similar observations are in the literature showing that individually taken vitamin E, C, or A was effective in reducing cataract risk, while the combination of the three supplements potentiated cataract formation [214].

The effects of STB on vascular function have been largely studied at different stages of STZ-diabetes. Satnikova et al. reported that 8-month STB supplementation to diabetic rats resulted in the protection of aortic function and ultrastructure; evaluated by aortic reactivity to noradrenaline, ACh and H₂O₂ under isometric conditions and transmission electron microscopy, respectively [215]. We found that STB treatment significantly reduced phenylephrine and BayK 8644-induced contractions in diabetic rats compared to control and untreated-diabetic rats. Neither DM nor STB treatment changed the vascular reactivity to KCl and CaCl₂. Diabetes-induced decrease in vascular relaxation to ACh was ameliorated by STB treatment and STB partially controlled mean arterial blood pressure in diabetic rats. These observations support the findings that increased entry of Ca⁺² mediates elevated contraction to BayK 8644 and phenylephrine in diabetic aorta and the inhibitory effect of in vivo STB treatment on Ca⁺² entry through cytosolic membrane-bound Ca⁺² channels may account for its reducing effectiveness on the vasoconstriction and blood pressure. A23187-induced vasorelaxation was markedly inhibited in STB-treated diabetic rats compared with control and untreated-diabetic rats. The inhibitory effect of STB on Ca⁺² influxes in endothelial cells may account for its reduced effect on A23187-stimulated relaxations [69]. These results are also in agreement with the Ca⁺² mobilization regulating effect of STB on membranes [216-218] and support our previous findings, which demonstrated an inhibitory effect of STB on intracellular Ca⁺² accumulation in diabetic heart and liver [56]. STB also exerts a protective effect through free radical-scavenging and anti-oxidant activities in the ischemia/reperfusion injury model, as previously reported [218-220]. More recently, we have focused on evaluating the effects of STB and other antioxidant pyridoindoles [221] using transformed and non-transformed cells. Pre-treating cells with STB significantly increased cell viability and decreased apoptosis rate; determined by flow cytometric analysis and by measuring caspase-3 and caspase-9 activities in doxorubicin-induced, free radical-mediated apoptosis model of P815 cells [222]. We have also suggested that STB and its analogues protect insulin release in the INS 1E-cell line exposed to H₂O₂ or cytokine cocktail and that the antioxidant effects of pyridoindoles also mediate antiapoptotic processes via intrinsic as well as extrinsic pathway of programmed cell death in insulin releasing cells. This may provide an appropriate explanation for the decrease in severity of hyperglycemia caused by STB, which might have therapeutic implications in the prevention of autoimmune DM [223]. Interestingly, a derivative of STB, suppressed the intracellular generation of ROS, reduced cell swelling, improved the reduced cell proliferation and viability was not able to suppress protein carbonyl formation in HT22 neuronal cells under glucose toxicity [224].

CONCLUSION

Several mechanisms have been proposed to explain the harmful effects of glucose. Glucose could be directly toxic to cell components because it can promote non-enzymatic glycosylation and the accumulation of AGEs, which impair cellular functions and mediate an increase in the production of ROS. In the presence of elevated glucose metabolism, the excessive production of electrons can promote the generation of ROS; if there is not a matching increase in the efficiency of electron transport system, this can lead to GOS and cellular injury in the long term. Although, multiple biochemical pathways are likely to be responsible for diabetic cardiovascular complications, substantial evidence suggests a key role for GOS. Therefore, recent interest has focused on strategies to prevent, reverse or retard GOS in order to modify the natural history of diabetic CVD. Even though, there is an increasing amount of experimental evidence showing the beneficial effects of antiglycoxidative (antioxidant) agents in the prevention/treatment of diabetic CV abnormalities, some clinical trials conducted to date, failed to provide adequate support for the use of antioxidant therapy in DM. However, it is still too early to reach a definitive conclusion on the clinical benefits of antioxidants. It is possible that antioxidants would be more demonstrably effective in a patient population chosen on the basis of elevated levels of GOS. Unfortunately, none of the studies to date effectively assessed the baseline GOS at the time of patient recruitment. The clinical trials generally used endpoints that were not directly related to GOS, but rather gross markers of overall cardiovascular health, such as effect on mortality. In the evaluation of GOS/OS and the effects of antioxidants thereon, specific markers should be measured. We strongly believe that more research on the use of antioxidants (antiglycoxidative agents) in the prevention of diabetic cardiovascular complications is necessary and should be encouraged. Pharmacological intervention using different antioxidants may have significant implications in the prevention of the pro-oxidant feature of DM and protect the redox status of the cardiovascular cells. Finally, the search for novel antioxidant/antiglycoxidative agents with enhanced efficiency, optimized bioavailability, reduced adverse effects and decreased toxicity should be continued, with the prospect of preventing, or delaying long-term cardiovascular diabetic complications.

ABBREVIATIONS

Ach

 = Acetylcholine

AR

 = Adrenergic receptor

AGEs

 = Advanced glycation end products

ACE

 = Angiotensin-converting enzyme

alpha AR

 = Alpha-adrenergic receptor

ALA

 = Alpha lipoic acid

RONOO

 = Alkyl peroxynitrates

beta-Ars

 = Beta-adrenergic receptors

CLO

 = Cod Liver Oil

cAMP

 = Cyclic adenosine monophosphate

cGMP

 = Cyclic guanosine monophosphate

DM

 = Diabetes mellitus

DAG

 = Diacylglycerol

DHLA

 = Dihydrolipoic acid

DHA

 = ocosahexaenoic acid

EPA

 = Eicosapentaenoic acid

eNOS

 = Endothelial nitric oxide synthase

ET

 = Endothelin

EDHF

 = Endothelium-dependent hyperpolarization

EDRF

 = Endothelium-derived relaxing factor

FAD

 = Flavin adenine dinucleotide

FMN

 = Flavin mononucleotide

G-6PD

 = Glucose-6-phosphate dehydrogenase

GOS

 = Glycoxidative stress

GSH-Px

 = Glutathione peroxidase

GR

 = Glutathione reductase

HOCl

 = hydrochlorous acid

H₂O₂

 = Hydrogen peroxide

HRO₂

 = Hydroperoxyl

HNO₂

 = Nitrous oxide

IDDM

 = Insulin-dependent diabetes mellitus

IL-6

 = interleukin- 6

LVDP

 = Left ventricular developed pressure

L-NAME

 = L-nitro-arginine methyl ester

MDA

 = Malondialdehyde

MCP-1

 = Monocyte chemoattractant protein

NADPH

 = Nicotinamide adenine dinucleotide phosphate

NO

 = Nitric oxide

NOS

 = Nitric oxide synthase

NO₂

 = Nitrogen dioxide

NIDDM

 = Non-insulin-dependent diabetes mellitus

NF-KB

 = Nuclear factor-kappaB

OS

 = Oxidative stress

ONOO

 = Peroxynitrite

OH

 = Hydroxyl

PDE5

 = Phosphodiesterase type 5

PUFAs

 = Omega-3 polyunsaturated fatty acids

PKC

 = Protein kinase C

ROS

 = Reactive oxygen species

RNS

 = Reactive nitrogen species

RAGE

 = Receptor advanced glycation end products

RCBMT

 = Retinal capillary basement membrane thickness

RBP

 = Retinol-binding protein

RO₂

 = Peroxyl

STB

 = Stobadine

STZ

 = Streptozotocin

SOD

 = Superoxide dismutase

O₂

 = Superoxide

TBARS

 = Thiobarbituric acid reactive substances

VCAM-1

 = Vascular cellular adhesion molecule-1

6-PGD

 = 6-phosphogluconate dehydrogenase