High glucose induces antioxidant enzymes in human endothelial cells in culture. Evidence linking hyperglycemia and oxidative stress.

It has been suggested that oxidative stress may play an important role in the pathogenesis of diabetic complications. Hyperglycemia may cause increased production of free radicals, and evidence supports a prominent role for these reactive molecules as mediators of endothelial cell dysfunction in diabetes. It has been demonstrated that active oxygen species induce antioxidant enzyme expression in some tissues, and this phenomenon is considered proof of an existing oxygen-dependent toxicity. In this study, human endothelial cells from umbilical vein, immortalized human endothelial cells, and immortalized human endothelial cells transfected to express high glutathione peroxidase levels were grown in normal and high-glucose conditions. High glucose delayed replication after 7 and 14 days of culture of human endothelial cells, both from umbilical vein and immortalized, while transfected cells were not affected. The activity and the mRNA expression of the antioxidant enzymes CuZn-superoxide-dismutase, Mn-superoxide-dismutase, catalase, and glutathione peroxidase were evaluated after 2, 7, and 14 days of culture. High glucose at days 7 and 14 induced an overexpression of CuZn-superoxide-dismutase, catalase, and glutathione peroxidase in both human endothelial cells from umbilical vein and immortalized human endothelial cells, while in transfected cells it did not. This study demonstrates that high glucose induces an increase in antioxidant enzyme levels in human endothelial cells, suggesting that elevated glucose levels may produce an oxidative stress in the cells.

Sod and GSH inhibit the high glucose-induced oxidative damage and the PDGF increased secretion in cultured human endothelial cells.

Poor control of blood glucose has been established as a key pathogenetic mechanism in the vascular complications of diabetes. It has been reported that glucose may autooxidize generating free radicals which have been suggested to delay proliferation, to modify mobility, to influence platelet-derived growth factor and other secretory protein production in a variety of cell systems. Platelet-derived growth factor, in turn, may induce proliferation and migration of vascular smooth muscle cells and thus play a role in atherogenesis. In the present study the effects of antioxidants on the high glucose-dependent oxidative cell damage and increased platelet-derived growth factor secretion have been investigated using cultured human endothelial cells. Our findings show that rising the glucose concentration in the culture medium from 5 mM to 20 mM, increased the production of free radicals cell damage markers, such as malondialdehyde and conjugated dienes, as well as the production of platelet-derived growth factor. The addition of superoxide dismutase or glutathione prevents both such effects. These results suggest that antioxidants may be a helpful therapeutic adjuvant to reduce the vascular complications of diabetes.

The defence against free radicals protects endothelial cells from hyperglycaemia-induced plasminogen activator inhibitor 1 over-production.

It has been shown that hyperglycaemia may stimulate plasminogen activator inhibitor 1 (PAI-1) over-production in human endothelial cells in culture. At the same time, it has been shown that glucose may enolize, producing free radicals. In this study, the possibility that hyperglycaemia stimulates PAI-1 over-production in human endothelial cells in culture by generating free radicals has been evaluated. For this purpose two experimental models were used: human endothelial cells transfected to express high glutathione peroxidase levels cultured in hyperglycaemic media, and human endothelial cells cultured in hyperglycaemic media with the antioxidant GSH. Cells grown in 20 mM glucose produced higher values of PAI-1 with respect to controls. The production of PAI-1 was not influenced by hyperglycaemia in transfected cells. GSH in the medium reduced hyperglycaemia-induced PAI-1 over-production, but also reduces the basal production of PAI-1 in the cells grown in normal glucose concentration. These data show that antioxidant defences may reduce hyperglycaemia-induced PAI-1 over-production in human endothelial cells in culture. The hypothesis that oxidative stress may play an important role in the pathogenesis of diabetic complications is then supported by this study.

Metabolic control may influence the increased superoxide generation in diabetic serum.

Superoxide anion (O2-) generation in serum from 10 Type 1 diabetic patients and 10 normal subjects was measured ex vivo. The amount of O2- production was significantly increased in diabetic serum 0.41 +/- 0.04 (+/- SD) vs 0.14 +/- 0.04 mumol l-1 min-1, p less than 0.001) and correlated with fasting plasma glucose and glycosylated protein levels in both diabetic (r = 0.72, p less than 0.01, and r = 0.62, p less than 0.05) and normal r = 0.75, p less than 0.01 and r = 0.64, p less than 0.05) subjects. Improved metabolic control in the diabetic patients was associated with a reduction of serum O2- production (0.28 +/- 0.06 mumol l-1 min-1, p less than 0.01), but the correlation between O2- levels and fasting plasma glucose and glycosylated protein concentrations was retained (r = 0.86 and r = 0.72, respectively, both p less than 0.01).

Protein C deficiency in insulin-dependent diabetes: a hyperglycemia-related phenomenon.

In 30 insulin-dependent diabetic patients protein C (PC) antigen and PC activity were significantly lower than those of matched control healthy subjects. An inverse correlation between fasting plasma glucose and both PC concentration and activity was present in diabetics, while a direct correlation between PC concentration and PC activity was observed. Induced hyperglycemia in diabetic and normal subjects was able to decrease both PC antigen levels and PC activity, and heparin reversed in part this effect. In diabetic patients euglycemia obtained by insulin infusion restored to normal the depressed PC levels. Heparin did not alter both the basal PC concentration and activity in healthy controls. These data stress the major role of hyperglycemia in determining PC decrease in diabetics, and suggest that PC reduction is probably associated to hyperglycemia-enhanced thrombin formation.

The role of hyperglycaemia-induced alterations of antithrombin III and factor X activation in the thrombin hyperactivity of diabetes mellitus.

Factor X concentration and factor X activation, antithrombin III anti-Xa activity and plasma concentration, and fibrinopeptide A were measured in 20 diabetic patients and 20 normal subjects. Although factor X activation (81.3 +/- 2.2 vs 97.3 +/- 2.1%, p less than 0.01; mean +/- SE) and antithrombin III activity (76.5 +/- 2.2 vs 96.3 +/- 1.8%, p less than 0.01) were reduced in the diabetic patients, fibrinopeptide A concentration was increased (3.7 +/- 0.4 vs 1.7 +/- 0.2 ng ml-1, p less than 0.01). The ratio of factor X activation to antithrombin III anti-factor Xa activity was increased in the diabetic patients (1.10 +/- 0.01 vs 1.01 +/- 0.02, p less than 0.01). Induced hyperglycaemia was able to mimic all these abnormalities, without changing factor X or antithrombin III concentration. The results suggest that in vivo hyperglycaemia produces a decrease of factor X activation, but at the same time increases fibrinopeptide A formation due to a greater decrease of antithrombin III anti-Xa activity.

Evidence for a reduced heparin cofactor II biological activity in diabetes.

A reduction of heparin cofactor II (HCII) biological activity, despite its normal plasma concentration, is reported in insulin-dependent diabetic patients. A good linear correlation between HCII activity and concentration is present in normal controls but not in diabetics. In these subjects HCII activity correlates inversely with fasting blood glucose and glycated proteins but not with Hb A1. These data demonstrate the presence of a depressed HCII activity in the presence of its normal plasma concentration in insulin-dependent diabetics and suggest a role for short-term metabolic control in conditioning this phenomenon.

Hyperglycemia may determine fibrinopeptide A plasma level increase in humans.

The effects of hyperglycemia on plasma fibrinopeptide A (FPA) levels in normal subjects are reported. An increase of FPA concentration parallel to sustained hyperglycemia was observed; when the glycemia returned to basal values, FPA showed values in normal range. Heparin infusion was able to significantly decrease the hyperglycemia-induced augment of FPA levels. Isovolumic-isotonic NaCl solution infusion produced a slight (NS) increase in FPA levels; however, mild hyperglycemia, achieved by glucagon, was also able to produce a significant increase in FPA concentration. These data demonstrate the direct role of hyperglycemia in conditioning FPA level, and suggest that hyperglycemia, by itself, is a sufficient stimulus to produce thrombin activation in humans.

Increased alpha 2-macroglobulin in diabetes: a hyperglycemia related phenomenon associated with reduced antithrombin III activity.

Increased alpha 2-macroglobulin (alpha 2M) activity and concentration, and decreased antithrombin III (ATIII) plasma concentration are reported in diabetic subjects. In diabetes an inverse correlation between ATIII activity and blood glucose, HbA1, alpha 2M activity and alpha 2M concentration, and a direct correlation between both alpha 2M activity and alpha 2M concentration with blood glucose and HbA1 are found. Moreover, a direct correlation between alpha 2M activity and alpha 2M concentration fails. In both diabetic and normal subjects induced hyperglycemia increases alpha 2M activity and alpha 2M concentration reduces ATIII activity, while ATIII concentration is not affected. These data which show that hyperglycemia may increase alpha 2M molecule levels while altering only the biological function of ATIII, provide evidence that hyperglycemia may decrease, directly, the biological function of some proteins and may condition the levels of some risk factors for the development of diabetic complications such as alpha 2M.

Hyperglycemia-conditioned increase in alpha-2-macroglobulin in healthy normal subjects: a phenomenon correlated with deficient antithrombin III activity.

Induced hyperglycemia in normal subjects increases alpha 2-macroglobulin (alpha 2M) activity and alpha 2M concentration and reduces antithrombin III (ATIII) activity, while it does not affect ATIII plasma concentration. Hyperglycemia-determined variations in ATIII activity and alpha 2M molecules are correlated in an inverse and parallel fashion. A compensatory role for the increase in alpha 2M in the regulation of the coagulation system may be hypothesized. Moreover, these data provide evidence that hyperglycemia may decrease, directly, the biological function of some proteins and may influence the levels of some risk factors for the development of complications in diabetes.

Blood glucose may condition factor VII levels in diabetic and normal subjects.

Increased factor VII levels have been reported in Type 1 (insulin-dependent) diabetic subjects. A direct correlation between fasting plasma glucose and factor VII level was found to exist in both diabetic and normal subjects. Induced-hyperglycaemia was able to increase factor VII levels in both diabetic patients and normal control subjects while, when euglycaemia was achieved in diabetic patients, factor VII values returned to normal range. This study shows that the level of factor VII may be directly conditioned by circulating blood glucose and, therefore, stresses the role of hyperglycaemia in conditioning coagulation abnormalities in diabetes mellitus.

A preliminary note on inhibiting effect of alpha-tocopherol (vit. E) on protein glycation.

Human plasma albumin was incubated in 25 mM/l glucose at 37 degrees C for up to sevent days. Aliquots of this mixture were also incubated with alpha-tocopherol (Vitamin E) at 2 mg/dl, 4 mg/dl and 8 mg/dl, respectively. Vitamin E inhibited protein glycation and, furthermore, shows a dose-dependent effect. This report suggests the possibility of using Vitamin E, at therapeutic doses, to obtain the inhibition of non-enzymatic glycation.

Impaired glucose metabolism in heroin and methadone users.

Plasma glucose and insulin responses to both oral and intravenous glucose stimulation were evaluated in heroin and methadone addicts, compared to healthy control subjects. Both groups of addicts had an altered response to oral and intravenous glucose load. These phenomena were linked to a reduced insulin response. Moreover, increased fasting insulin levels in both groups of addicts were observed. These data show that both heroin and methadone addiction may alter glucose metabolism, and, furthermore, stress the findings of similarities between opiate addicts and non-insulin dependent diabetics.

Induced hyperglycemia alters antithrombin III activity but not its plasma concentration in healthy normal subjects.

The effects of induced hyperglycemia on both antithrombin III (ATIII) biologic activity and plasma concentration in normal subjects are reported. A decrease in ATIII activity parallel to hyperglycemia was observed, while ATIII concentration was unchanged. When the glycemia returned to basal values ATIII activity concomitantly showed values in the basal range. Heparin infusion was able to significantly preserve ATIII activity from glycemia-induced alterations. These data demonstrate that hyperglycemia by itself may alter ATIII biologic activity. Moreover, the effect of heparin administration suggests that both glucose and heparin compete in vivo for the same functional site. Our study, showing the possible role of hyperglycemia in altering the biologic function of some proteins, stresses the role of increased blood glucose in the development of some complications in diabetes.

Daily rapid blood glucose variations may condition antithrombin III biologic activity but not its plasma concentration in insulin-dependent diabetes. A possible role for labile non-enzymatic glycation.

The effect of rapid daily variation of glycemia and labile HbA1 on both antithrombin III (ATIII) activity and plasma concentration in ten insulin dependent diabetics has been evaluated. The variations of both plasma glucose and labile HbA1 were inversely correlated to the alterations of ATIII activity (r = -0.71 and r = -0.73, respectively, p less than 0.001), while no change in ATIII plasma concentration was present. These data suggest a direct role of glucose in determining rapid alteration of ATIII biologic activity, in vivo, in diabetes, probably mediated by labile non-enzymatic glycation.

Heparin preserves antithrombin III biological activity from hyperglycemia-induced alterations in insulin-dependent diabetics.

Alteration in antithrombin III (ATIII) biological activity, despite its normal plasma concentration, in diabetic subjects is shown in this report. This alteration is glycemia level-dependent, there existing an inverse correlation between fluctuations of daily blood glucose level, labile glycosylated hemoglobin and ATIII activity. The subcutaneous and endovenous heparin administration restores ATIII activity, but does not modify its plasma concentration in diabetics. Moreover, heparin treatment preserves ATIII activity from glycemia-induced alterations. These data suggest a role for glucose, probably through a labile nonenzymatic glycation process, in determining the alteration of ATIII biological activity. Moreover, showing the possibility by heparin administration to restore ATIII activity and preserve its biological function from effects of glycemia variations, stress the hypothesis that glucose and heparin compete in vivo, both against the same catalytic residue of ATIII.

Metabolic control may alter antithrombin III activity but not its plasma concentration in diabetes: a possible role for nonenzymatic glycosylation.

The effects of metabolic control on both antithrombin III (AT III) activity and AT III plasma concentration in 20 insulin-treated diabetic subjects have been evaluated. Basal AT III activity was significantly lower in diabetic subjects versus healthy controls (P less than 0.001), whereas no difference was found in AT III concentration. A good correlation was found between AT III activity and AT III concentration (r = 0.81; P less than 0.001) in healthy controls, but this correlation was not significant in diabetic subjects (r = 0.12; P = NS). In those subjects a linear inverse correlation was found to exist between AT III activity and level of glycosylated proteins (r = -0.43; P less than 0.05). Diabetic subjects were also examined after 1 and 2 mo of restored metabolic control, obtained by human insulin (DNA-recombinant) therapy. Improved metabolic control was characterized by an increase of AT III activity (P less than 0.05), a decrease of mean daily blood glucose, and stable HbA1 and glycosylated proteins (P less than 0.05), while AT III concentration did not vary. On the other hand, a significant inverse correlation between AT III activity and glycosylated proteins was found during both the first and second months (r = -0.54 and r = -0.53, respectively; P less than 0.01). Moreover, no correlation between AT III activity and AT III concentration was found. These data suggest that impaired metabolic control may alter the biologic activity of AT III in diabetes, but not its plasma concentration.