Varying the type of carbohydrate in the diet can influence glucose metabolism and insulin action. For example, rats consuming a high fructose diet develop insulin resistance, hypertriglyceridemia and hypertension (Thorburn et al. 1989). Several lines of evidence suggest that this diet leads to the metabolic changes observed in syndrome X, in which insulin resistance, hypertension and dyslipidemia are observed in glucose intolerant and prediabetic patients (Reaven 1988). Such metabolic modifications have been associated with a high incidence of cardiovascular disease (DeFronzo and Ferranini 1991, Reaven and Laws 1994). On the other hand, the role of free radical attack in diabetes mellitus and in the cardiovascular complications of the disease has been documented largely through the effects of free radicals on lipids and proteins (Giugliano et al. 1996, Jain et al. 1989, Oberley 1988, Ozdemirler et al. 1995). In these disease conditions, lipophilic antioxidants, such as  tocopherol, have been shown to be efficient in the protection of lipid and cell membranes against free radical attack (Esterbauher et al. 1989). Moreover,  tocopherol supplementation in patients with non-insulin-dependent diabetes mellitus (NIDDM)³ improves insulin action (Paolisso et al. 1993). At this time, few data are available concerning some of the antioxidant system components during insulin-resistance states, without hyperglycemia, as observed in rats fed a high fructose diet. In light of such evidence, the present study investigated the free radical activity of rats fed a high fructose diet. The effects of vitamin E supplementation on insulin action was also investigated to understand the involvement of free radical attack or protection (i.e., lipid peroxidation, antioxidant enzyme activity, and trace element and glutathione concentrations) on the impaired insulin activity of rats fed a high fructose diet. These results are of interest in human nutrition, because of the increasing consumption of fructose in the western population (Economic Research Service 1989).

MATERIALS AND METHODS

Table 1. Diet compositions of the control (C), high fructose (FT) and high fructose + vitamin E (FVE) groups [View Table]

During insulin administration, the increase in glucose uptake by insulin-sensitive tissues was measured by the increase in the rate of disappearance (Rd). At a high insulin infusion rate (13,600 pmol/min), Rd can be determined by the glucose infusion rate (GIR) because hepatic glucose production is completely inhibited at this insulin level (Faure et al. 1994). At steady state, the rate of glucose appearance (Ra) is equal to Rd and is given by the GIR. Under these conditions, the GIR reflects the insulin sensitivity of the peripheral tissues. Because hepatic glucose production was completely inhibited at this insulin infusion rate, we did not use labeled glucose in this study (Faure et al. 1994).

Trace element analysis was performed using atomic absorption spectrophotometry. For trace element measurements, plasma was removed by centrifugation (3000 × g, 15 min). Zinc and copper concentrations were determined using flame atomic absorption spectrophotometry (Perkin-Elmer model 460, Norwalk, CT), as described previously (Arnaud et al. 1984 and 1985). Zinc and copper concentrations were calculated using an exogenous calibration curve.

Selenium concentrations were determined by flameless atomic absorption spectroscopy after a sampling dilution procedure, with a Perkin-Elmer model 5100 fitted with a HGA 600 graphite furnace. Plasma selenium was determined by standard addition calibration curves (Neve et al. 1987). Se-glutathione peroxidase activity (Se-GSH-Px, EC 1.11.1.19) was measured by the modified method of Gunzler et al. (1974), using tert-butyl hydroperoxide as substrate. The results were expressed as µmol of NADPH oxidized per minute per gram of hemoglobin for erythrocyte and as international units per liter for plasma Se-GSH-Px.

Cu-Zn Superoxide dismutase activity (Cu-Zn SOD, EC 1.15.1.1) was determined by monitoring the auto-oxidation of pyrogallol according to the method of Marklund and Marklund (1974). One unit of Cu-Zn SOD activity is defined as the amount of the enzyme required to inhibit the rate of pyrogallol autoxidation by 50% and is given in µg/g hemoglobin (Hb).

RESULTS

Assessment of insulin sensitivity and metabolic status. The three groups exhibited different insulin sensitivities. In comparison with the C group which had a GIR of 172.36 ± 10.45 µmol/kg·min, a marked insulin resistance was observed in the FT group which had a GIR of 83.96 ± 8.1 µmol/kg·min) (P < 0.001). The FVE-supplemented group had a significantly higher GIR (118.98 ± 9.2) than the FT group, but it was significantly lower than the GIR of the C group.

In comparison with the C group, the FT group had significantly higher plasma triglycerides (Table 2) which were not significantly different from plasma triglycerides of the FVE group. Glycemia and insulinemia were not significantly different among the groups (Table 2). Plasma fructosamine was significantly higher in the FT group than in the other groups which did not differ from one another (Table 2).

Table 2. Plasma glucose, fructosamine, triglycerides and insulin in control (C), fructose (FT)- and fructose + vitamin E (FVE)-supplemented rats1,2,3 [View Table]

Table 3. Oxidative defense system components, lipid peroxidation intermediates and trace elements in control (C), fructose (FT)- and fructose + vitamin E (FVE)-supplemented rats1-4 [View Table]

DISCUSSION

This study confirms the results of previous studies demonstrating that fructose feeding leads to glucose intolerance and decreases insulin sensitivity in intact animals (Thorburn et al. 1989). Several metabolic hypotheses have been advanced to explain insulin resistance in fructose-fed rats. It has been shown that chronic fructose feeding alters the activity of several enzymes regulating hepatic carbohydrate metabolism, including decreasing the activity of glucokinase and increasing glucose-6-phosphatase activity (Van Den Bergue 1986) leading to hepatic insulin resistance. In the present study, we observed a lower glucose uptake by the peripheral tissues as described previously (Thorburn et al. 1989). Glycemia was not significantly different in FT rats compared with both other groups. Triglycerides of the FT and the FVE groups were fourfold greater compared with the C group. This could be linked to a high formation of glycerol-3-phosphate leading to an increased synthesis of VLDL by liver (Beck-Nielsen et al. 1978). The metabolic defects of this animal model are different from those of the NIDDM model and closer to that designated syndrome X which includes dyslipidemia, hypertension, hyperinsulinemia and glucose intolerance, whereas glycemia following overnight food deprivation remains normal (Reaven 1988). Moreover, FT and FVE rats did not have significantly different body weights than the C rats, which is different than other experimental models of insulin resistance using genetically obese rats. This observation suggests that in the absence of hyperglycemia in rats food-deprived overnight, an enhanced oxidative stress can also be associated with insulin resistance (or its metabolic consequences), as shown by the increased plasma TBARS and the increased blood GSSG-GSH ratio. Previous studies have shown enhanced lipid peroxidation as a consequence of experimentally induced diabetes mellitus and indicated that oxidative stress may be involved in the genesis of diabetic complications (Young et al. 1995). Furthermore, hyperglycemia per se could have a direct effect on oxidative lipid and protein modifications through the formation of glucose-derived free radicals, during the protein glycation process (Wolff et al. 1990).

Considering the role of free radical activity in insulin sensitivity, Paolisso et al. (1993) demonstrated that the administration of high doses of vitamin E during NIDDM in men is a useful tool to reduce oxidative stress and improve insulin action. In our experimental model, vitamin E supplementation also had a beneficial effect on insulin action. The effect of such a lipophilic antioxidant on insulin action originates from several mechanisms. It could be linked to a decrease in the blood GSSG-GSH ratio as previously hypothesized (Paolisso et al. 1993), leading to a better physicochemical protection of cell membranes. As previously shown (Ammon et al. 1989), cell membrane fluidity must be preserved to maintain insulin activity. Moreover, the FT rats had a higher plasma fructosamine concentration than the two other groups. In other words, enhanced plasma protein glycation occurred in high fructose-fed rats even in the absence of hyperglycemia. We observed that vitamin E supplementation also had a beneficial effect on plasma fructosamine concentration through its role as a free radical scavenger. Membrane modifications by free radical attack do affect not only phospholipids but also proteins (Mikaelian et al. 1994) which could also reduce insulin activity (Ceriellio et al. 1991). Contrary to the Se-GSH-Px activity of red cells, the Cu-Zn-SOD activity of red cells was significantly lower in high fructose-fed rats. Although plasma copper was not different than in the control group, this observation could be linked to a decrease in tissue copper including liver, muscles and red cells associated with a fructose-enriched diet (Fields et al. 1984, Wapnir and Devas 1995). Because vitamin E supplementation leads to a normalization of red blood cell Cu-Zn-SOD activity, the hypothesis that the protein could be damaged by oxidative stress and/or glycation can be advanced (Hunt and Wolff 1991, Hunt et al. 1988). On the other hand, the lower red cell Cu-Zn-SOD activity in FT-fed rats could partly affect insulin sensitivity because this antioxidant enzyme has a key role in the cell protection against the deleterious effects of the superoxide anion (Nath et al. 1984). Some trace elements have an important role in free radical protection. In this study, plasma zinc and selenium concentrations were significantly lower in the FT and FVE groups. Zinc is a biological antioxidant (Bray et al. 1990), and its depletion can lead to oxidative stress (Faure et al. 1991b) and a reduced insulin sensitivity (Faure et al. 1991a). In particular, this metal is associated with the apoprotein of superoxide dismutase; thus its depletion could alter the protein as previously shown (Coudray et al. 1992). Selenium exerts its antioxidant effects through its role as a cofactor of Se-GSH-Px. In comparison with other studies (Fields et al. 1984), the diet fructose concentration was lower and the duration of the diet and treatment was shorter in the present study. Tissue selenium likely was not lowered in this study but other studies are necessary to investigate levels of tissue trace elements in our rat model.

In conclusion, the current results provide additional evidence that fructose feeding of rats leads to insulin resistance. It demonstrates for the first time that this diet has a deleterious effect on the antioxidant defense systems. Furthermore, our results provide direct evidence that vitamin E has a beneficial effect on insulin sensitivity. Further investigation is warranted to determine the roles of antioxidants and the oxygen free radicals on insulin action. Our findings are relevant in the field of human nutrition, particularly because of the increasing consumption of dietary fructose which is promoted as a healthy food for diabetic, prediabetic patients or in the general population. The increase in oxidative stress accompanying this diet could have adverse effects through an enhanced risk of cardiovascular lesions.