Antioxidant vitamins C and E ameliorate hyperglycaemia‐induced oxidative stress in coronary endothelial cells

Objective:  Vitamins C and E have protective features in many disease states associated with enhanced oxidative stress. The aim of this study was to investigate whether vitamin(s) C and/or E modulate hyperglycaemia‐induced oxidative stress by regulating enzymatic activities of prooxidant, i.e. NAD(P)H oxidase and/or antioxidant enzymes, namely endothelial nitric oxide synthase (eNOS), superoxide dismutase, catalase and glutathione peroxidase, using coronary microvascular endothelial cells (CMEC).


Introduction
Nitric oxide (NO), generated from amino acid L-arginine within healthy endothelium by endothelial type of nitric oxide synthase (eNOS), plays a pivotal role in the regulation of normal vascular tone [1]. The characteristics of endothelium change in several pathological conditions, including diabetes mellitus, leading to a phenomenon called 'endothelial dysfunction', which is characterized by impaired endothelial cell function. Although the actual causes of this pathology are unknown, several mechanisms have thus far been proposed. These include inefficient utilization of substrate L-arginine by NOS [2], an abnormal NOS activity due to inadequate availability of cofactor tetrahydrobiopterin (BH 4 ) [3], concurrent release of endothelium-derived vasoconstrictors by cyclooxygenase (COX) pathway [4] and scavenging of NO by advanced glycation end products [5].
In recent years, enhanced oxidative stress status arising from the excess release of reactive oxygen species (ROS) has been documented in diabetic animals and cells cultured under high glucose conditions [6,7] and associated with the pathogenesis of diabetic endothelial dysfunction. ROS may be generated as a result of prolonged exposure of cells to hyperglycaemia that results in non-enzymatic glycation of plasma proteins [8], which then undergo further spontaneous reactions to produce free radicals, such as superoxide anion (O 2 -), the foundation molecule of other ROS [9]. O 2 is also formed by enzymatic activities of COX, xanthine oxidase (XO), uncoupled eNOS and NAD(P)H oxidase [3,10,11]. Amongst these enzymes, NAD(P)H oxidase enzyme system has attracted much of the attention. It has been characterized as the main source of ROS in coronary microvascular endothelial cells (CMEC) [12] and has later been coupled to oxidative stress-mediated endothelial dysfunction in the central retinas of type II diabetic rats [13]. Under physiological conditions, O 2 is converted to hydrogen peroxide (H 2 O 2 ) by superoxide dismutases (SOD) [14] and upon generation, H 2 O 2 itself is further metabolized to H 2 O by catalase and glutathione peroxidase (GPx) [14]. However, under hyperglycaemic conditions, the levels of O 2 may be elevated as a consequence of glycosylation and hence inactivation of SOD [15]. O 2 readily scavenges NO to diminish its vasoprotective effects and produces vascular smoothmuscle (VSM) contractions in the absence of intracellular antioxidants such as glutathione and cysteine [16,17]. Depletion of these antioxidants in diabetic conditions is a common occurrence as demonstrated in aortic endothelial cells isolated from diabetic rabbits [18] and in CMEC cultured with 22-mM D-glucose-containing media for 7 days [19]. Taken together, the currently available data imply that the elevation of antioxidant levels in diabetic states may be critical in suppressing hyperglycaemia-induced oxidative stress generation and thus initiation and/or progression of endothelial dysfunction. Several recent studies including our own have supported this hypothesis in that treatments of diabetic patients with antioxidants, such as allopurinol, and the incubation of CMEC grown under hyperglycaemic conditions with free-radical scavengers, e.g. Tiron, displayed beneficial in vivo and in vitro effects respectively [19,20]. A growing body of evidence has revealed that vitamins C and E, in addition to other intracellular antioxidants such as glutathione, also improve endothelial function in human subjects and in animal models of diabetes mellitus [21][22][23]. Moreover, they are associated with increases in NO generation and total antioxidant status and also with reduction in blood pressure [24,25]. Although both antioxidant vitamins are known to stimulate NO generation and have prominent ROS-scavenging effects, the underlying mechanisms of their putative beneficial actions remain to be determined. The present study was therefore set out to investigate whether vitamins C and/or E modulate redox state in CMEC exposed to hyperglycaemic conditions via regulation of the enzymatic activities of prooxidant [i.e. NAD(P)H oxidase] and antioxidant enzymes (i.e. eNOS, total SOD, catalase and GPx).

Materials and Methods
Isolation and Characterization of CMEC CMEC were isolated from 12-to 14-week-old Wistar rat hearts as previously described [26]. Briefly, two hearts were mounted and perfused retrogradely on a constantflow Langendorff system with 0.04% collagenase. The ventricles were then chopped, and collagenase digestion was quenched by the addition of bovine serum albumin to the perfusate. CMEC were obtained by sedimentation of myocytes and incubated in 0.01% trypsin at 37 C for the prevention of non-endothelial cell attachment. Cells were then activated by washing in calcium and suspended in Medium 199 (Life Technologies, UK) supplemented with 10% foetal calf serum, 10% newborn calf serum, benzylpenicillin 250 U/ml, streptomycin 250 mg/ ml, amphotericin B 12.5 mg/ml and gentamicin 50 mg/ml. Cell suspensions were plated and incubated at 37 C under 5% CO 2 . After 1-h incubation, unattached cells were washed off with saline and remaining cells were cultured to confluence. For different experiments, CMEC were cultured for 7 days in the growth medium containing either 5.5 mM (normal) or 22 mM (high) D-glucose. CMEC were also cultured with 5.5 mM D-glucose þ 16.5 mM L-glucose and 5.5 mM D-glucose þ 16.5 mM mannitol in order to investigate the effects of extracellular glycation and increased osmolarity. Growth medium was changed on a daily basis and all experiments were performed on CMEC up to and including passage number 5. Cultured CMEC were characterized by their typical 'cobblestone' morphology and their ability to form capillary-like tubes on the Matrigel [27]. To assess the contribution of inducible NOS (iNOS; calcium-independent isoform) to overall NOS activity, Ca 2þ was replaced with EGTA (1 mM). Reactions were terminated by the addition of 1 ml HEPES (20 mM, pH 5.5) containing EDTA (1 mM) and EGTA (1 mM). Newly formed L-[ 3 H]-citrulline, neutral at pH 5.5, was separated from the incubation mixture by cation exchange resin (Dowex AG 50 W-X8; Bio-Rad, UK) and quantified using a liquid scintillation counter. The results were expressed as pmol L-citrulline/mg protein/min.

Nitrite Detection
Nitrite levels were measured by Griess reaction as an index of NO generation following conversion of nitrate to nitrite by nitrate reductase [28]. An aliquot of the cellular homogenate was mixed with an equal volume of Griess reagent (sulfanilamide 1% w/v, naphthylethylenediamine dihydrochloride 0.1% w/v and orthophosphoric acid 2.5% v/v) and incubated at room temperature for 10 min prior to the measurement of absorbances at 540 nm. The amount of nitrite formed was compared to those of known concentrations of sodium nitrite and normalized to the protein content of the respective flask.

SOD Assay
SOD activity was measured by a reaction, dependent upon the inhibition of cytochrome c by endogenous SOD in cellular homogenates using a Cobas-Fara centrifugal analyser. The O 2 -, required for reduction, was generated by a reaction of xanthine-XO. One unit of XO activity was defined by the amount of homogenate required to inhibit, by 50%, the rate of cytochrome c reduction. For assay of total SOD activity, 0.1 mM xanthine was dissolved in 50 mM NaCO 3 buffer. A dilute stock solution was added to a 10 mM solution of cytochrome c, 50mM xanthine, 0.1 mM EDTA and 50 mM sodium carbonate to produce a change in the absorbance of 0.0250/min at 550 nm at pH 10.

Catalase Assay
The activity of catalase was determined by a photometric method where the activity was determined by monitoring the decomposition of H 2 O 2 at 240 nm in the presence of methanol which produces formaldehyde which in turn reacts with Purpald (4-amino-3-hydazino-5mercapto-1,2,4-triazole) and potassium periodate to produce a chromophore. Quantification was performed in comparison with the results obtained with catalase solutions of known activities (31.2, 15.6, 7.8 and 3.9 U/ ml) and formaldehyde standards (25, 50, 100 and 200 mM).

GPx Assay
The activity of GPx was determined in cellular homogenates using a method developed by McMaster et al. [29]. Briefly, a fresh solution containing 0.3 U/ml glutathione reductase, 1.25 mM reduced glutathione and 0.19 mM NADPH in 50 mM potassium buffer (pH 7.4) was prepared. Homogenates of 100-mg total protein were added to this solution and incubated for 3 min prior to addition of 12-mM t-butylhydroperoxide to commence the reaction. Absorbances were read at 340 nm for 4 min. Activities were calculated as nmole glutathione/mg.

Evaluation of Cell Viability
A small aliquot of cells cultured under different conditions was incubated with 0.1% trypan blue for a few minutes and viewed under a light microscope. Dead cells were permeable to trypan blue and thus become coloured. By counting 100 cells, the percentage of viable cells was calculated.

Statistical Analysis
The results were presented as mean AE standard error mean. Numbers (n) indicated throughout the article denote the numbers of separate CMEC isolations and individual experiments. Statistical analyses were performed by both Student's t-test and ANOVA where appropriate. A p-value of <0.05 was considered statistically significant.

Effects of Vitamins C and E on Prooxidant and Antioxidant Enzyme Activities
The current study revealed greater prooxidant [i.e. NAD(P)H oxidase] and antioxidant (i.e. total SOD, catalase and GPx enzyme) activities in CMEC cultured with high (22 mM), compared to normal (5.5 mM), glucose concentrations for 7 days (p < 0.05 for each enzyme). However, the treatment of CMEC with antioxidant vitamins C (0.1-1 mM) and E (0.1-1 mM) alone or in combinations (0.1/0.1 mM and 1/1 mM, vitamins C and E respectively) significantly reduced these enzyme activities (table 1). Equimolar concentrations of L-glucose or mannitol did not have any impact on enzyme activities when compared to cells grown under normoglycaemic conditions (p > 0.05) (table 1). Hyperglycaemia failed to alter eNOS and iNOS activities in CMEC as assessed by L-[ 3 H]-arginine to L-[ 3 H]citrulline conversion assay (p > 0.05). However, the addition of vitamins C and E alone or in combination with the culture medium significantly enhanced eNOS activity without altering iNOS activity in both normoglycaemic and hyperglycaemic CMEC (p < 0.05) ( figure  1a,b). Equimolar concentrations of L-glucose or mannitol did not have any impact on eNOS or iNOS activities compared to cells cultured under normoglycaemic conditions (p > 0.05) (data not shown).   (table 2).

Effects of Hyperglycaemia on CMEC viability
There were no significant differences in CMEC viability between cells cultured in different concentrations of glucose as assessed by trypan blue exclusion assay. Approximately, 86 AE 11% vs. 78 AE 9% of normoglycaemic and hyperglycaemic cells were viable respectively (p > 0.05).

Discussion
The endothelium releases a large number of vasoactive substances including NO to maintain normal vascular tone [1]. eNOS is associated with production of moderate levels of NO in healthy endothelium, while iNOS is coupled to excess generation of NO, endothelial cell damage and atherosclerosis in a number of disease settings including diabetes mellitus [30,31]. The inhibitors of iNOS, such as aminoguanidine, have therefore proved to be critical in preventing diabetic endothelial dysfunction [32]. Despite being a constitutively expressed isoform, the expression and activity of eNOS are affected by many pathological conditions associated with enhanced oxidative stress status such as genetic hypertension [33]. However, the regulation of eNOS in diabetes mellitus remains ambiguous, as both enhanced and diminished to reduce its biological half-life [16]. Although lower concentrations of free radicals may be beneficial in endothelial adaptation to ensure vasomotion control, their higher concentrations may induce several intracellular pathways, such as phosphatases and transcription factors (e.g. NFkB), to disrupt endothelial integrity by producing other potent ROS like the hydroxyl radical via Fenton reaction [34]. NAD(P)H oxidase has recently been characterized and shown as the main source of free radicals in CMEC [12]. Hyperglycaemia-mediated oxidative stress generation may be further exacerbated by the inadequacy of antioxidant enzymes -SOD that dimutate O 2 to H 2 O 2 , and catalase and/or GPx that metabolize H 2 O 2 to H 2 O [14].
In the light of the currently available data, the present study aimed to investigate whether vitamins C and E alone or in combination maintain a well-balanced oxidative status, a prerequisite for normal endothelial cell function, by regulating prooxidant and antioxidant enzyme activities in CMEC. To study, the combinations of vitamins C and E were important in relation to the findings indicating that vitamin C, apart from being a free-radical scavenger, is required for regeneration of vitamin E to its active form [35]. Putative beneficial effects of antioxidant vitamins may be attributed to their ability to (i) scavenge free radicals, (ii) regulate NO synthesis or release, (iii) regulate ROS generation and (iv) regulate antioxidant enzyme activities that metabolize ROS.
It is known that low molecular weight antioxidants, such as urate and thiols, along with vitamins C and E constitute the first line of defence against oxidative stress in the extracellular environment by scavenging free oxygen radicals and hence preventing oxidation of proteins and lipids [36]. Although consistent beneficial effects of vitamin C have been reported in diabetic patients, animal models of diabetes mellitus and cell culture models, similar studies with vitamin E have produced contradictory findings [21][22][23]. For instance, the Cambridge Heart Antioxidant Study revealed a marked reduction in non-fatal myocardial infarction, in patients received 400-800 IU of vitamin E/day compared with patients receiving placebo [21]. However, the subsequent Heart Outcomes Prevention Evaluation Study failed to confirm the beneficial effects of vitamin E [37].
The present study has demonstrated similar rates of cellular viability, eNOS and iNOS activities as well as nitrite (the stable end product of NO) production between CMEC cultured with high (22 mM) and normal (5.5 mM) glucose concentrations for 7 days. These findings are in support of a recent study demonstrating similar levels of eNOS and iNOS protein expressions and nitrite generation in CMEC cultured under identical conditions that were used in the current study [19]. In the present study, the incubation of CMEC with combinations of vitamins C and E or solely with vitamin C or E significantly enhanced nitrite generation and eNOS activity but failed to alter iNOS activity in both sets of cells without dramatically altering cellular viability rates. Similar results with vitamin C have also been reported under in vivo conditions in that long-term dietary intake of vitamin C by apolipoprotein E (apoE)- deficient mice have been associated with significant increases in eNOS but not iNOS activities and nitrite generation in apoE-deficient mice aortas [38]. These increases may in part be assigned to the ability of vitamin C to spare intracellular thiols to stabilize NO through controlling the formation of biologically active S-nitrosothiols and also to its ability to increase the intracellular levels of eNOS cofactor, i.e. BH 4 [39,40]. Vitamin E-mediated significant elevations in endothelial NO release and endothelial function have been documented in diabetic rat aorta [23]. The increases in these parameters were more likely to be a consequence of inhibitory effects of vitamin E on LDL oxidation via suppression of protein kinase C-mediated phosphorylation of muscarinic receptors on endothelial cells rather than its regulatory action on eNOS [41].
In the present study, marked increases in basal levels of O 2 and NAD(P)H oxidase activity have been determined in CMEC cultured in hyperglycaemic vs. normoglycaemic medium for 7 days. As ROS, under hyperglycaemic conditions, may also be generated by the polyol pathway, eicosanoid synthesis, protein kinase C activation and non-enzymatic glycation of plasma proteins [8,9], it was critical to investigate the extent of NAD(P)H oxidase activity to overall ROS generation in CMEC grown under high glucose concentrations. Hence, two structurally distinct and specific NAD(P)H oxidase inhibitors, namely PAO (0.1-3 mM) and AEBSF conditions in the absence and presence of vitamins C and E alone or in combination. Data from four separate experiments are expressed as mean AE s.e.m. *p < 0.05 difference compared to NG group and **p < 0.05 difference compared to HG group (5-100 mM), were used in this study. Both agents reduced antioxidant enzyme activities and O 2 levels but increased nitrite production in CMEC cultured with high glucose concentrations thereby indicating NAD(P)H oxidase as the main source of ROS in CMEC exposed to hyperglycaemic conditions. These data are in good agreement with those of several previous studies showing that antioxidants and free-radical scavengers such as probucol and Tiron improve endothelial function in hyperglycaemic CMEC and in the thoracic aortic rings from spontaneously hypertensive rats (SHR) [19,25]. The impacts of vitamins C and E on free-radical generation were also investigated in this study. Treatments with both vitamins C and E reduced both O 2 production and NAD(P)H oxidase activity selectively in CMEC cultured with high glucose concentrations. These data confirm the previous findings pertaining to regulatory effects of these vitamins on NAD(P)H oxidase activity in SHR aortas and also their well-known free-radicalscavenging effects [25]. Although the mechanisms by which vitamins C and E may modulate NAD(P)H oxidase activity have not been investigated in the current study, both the transcriptional and post-translational modification of NAD(P)H oxidase by vitamins have previously been associated with its regulation [42]. In addition to these, vitamin E may mediate the interactions between membrane-bound and cytosolic components of NAD(P)H oxidase to form fully active enzyme upon induction [42].
The current study has shown that hyperglycaemia elicits significant increases in total SOD, GPx and catalase activities in rat CMEC. The increases in these antioxidant enzyme activities are in keeping with their well-known induction in response to oxidative stress and have also been previously reported in human endothelial cells [43] and in patients with type II diabetes mellitus [44]. The results of the current study are also supportive of our recent study demonstrating enhanced expression of CuZn-SOD, Mn-SOD and catalase protein levels in rat CMEC cultured with high, compared to normal, concentrations of D-glucose for 7 days [19]. The increases in antioxidant enzyme protein expressions in the former study and in activities in the current study are independent of changes in osmolarity and extracellular glycation as assessed in parallel experiments where equimolar concentration of mannitol or L-glucose were substituted with D-glucose. Although a recent study has shown that gene transfer of CuZn-SOD failed to improve the endothelium-dependent vascular relaxation in carotid arteries from diabetic rabbits [45], another study using a cell-permeable SOD has shown to enhance basal and agonist-stimulated endothelium-dependent vascular relaxant responses in diabetic rat aorta [6]. These data indicate that the former findings may be due to inability of CuZn-SOD to penetrate VSM layer or due to glycosylation and therefore inactivation of SOD by high glucose levels [46]. Although putative mechanisms by which vitamins C and E may modulate antioxidant enzyme activities remain largely unknown, a recent study has revealed that transcriptional, translational and post-translational regulations as major determinants of local antioxidant enzyme levels in the renal cortex of diabetic rats [47].
In conclusion, the present study indicates that hyperglycaemia-mediated oxidative stress in CMEC does not appear to arise from alterations in eNOS activity or NO availability. However, exaggerated synthesis and release of ROS in particular O 2 may contribute to the pathogenesis of this phenomenon, and increases in the activity of antioxidant enzymes may be an adaptive response of CMEC to meet the biological demand exerted by hyperglycaemic oxidative stress. Our data demonstrate that the elevation of intracellular levels of antioxidant vitamins C and E to the levels that can effectively scavenge O 2 levels [48] may provide beneficial cellular effects by regulating pro-and antioxidant enzyme activities in diabetic states.