Plasma F₂ Isoprostanes: Direct evidence of increased free radical damage during acute hyperglycemia in type 2 diabetes

After local ethical committee approval and after obtaining written informed consent, we studied 21 patients with type 2 diabetes. Patients were recruited from the Bertram Diabetes Center, Norwich, U.K. (a secondary care diabetes facility), or from local primary care services if they were nonsmokers aged between 40 and 70 years and treated with diet or oral hypoglycemic agents alone. Patients were considered to have type 2 diabetes if they had been diagnosed after the age of 40 years, had no history of ketosis, and had stable glycemic control on diet or oral hypoglycemic agents for at least 6 months. Patients with clinical evidence of coronary artery disease (history of previous myocardial infarction or angina) and those receiving insulin were excluded. Patients with microalbuminuria (defined as an elevated urine albumin-to-creatinine ratio >2.5 for men and >3.5 for women) or macroproteinuria (defined as albustix-positive proteinuria) were excluded. Patients omitted all oral hypoglycemic agents or other medication on the morning of testing and fasted for 15 h before glucose tolerance tests were undertaken. All type 2 diabetic patients managed with diet alone had a fasting venous plasma glucose >7.0 mmol/l. Clinical details are shown in Table 1.

All patients underwent a single glucose tolerance test (75 g oral anhydrous glucose) at 0800, with peripheral blood samples taken at 0, 60, 90, and 120 min from an indwelling venous cannula. Patients remained seated throughout the glucose tolerance test. HbA_1c was assessed using a commercially available kit (Roche, Welwyn, U.K.) on an automated biochemistry analyzer (Cobas Mira; Roche), with the normal range quoted as 4.5–5.7%.

The intracellular glutathione (GSH) redox cycle between reduced GSH and oxidized GSH is an effective mechanism in protecting against intracellular oxidative damage because GSH acts as substrate for antioxidant enzymes and as a free radical scavenger, and the ratio of reduced-to-oxidized GSH by mass can be used as an index of intracellular oxidative stress (13). A 15-ml sample of whole blood was collected into EDTA, diluted to twice the original volume with PBS, and layered onto a 15 ml lymphocyte separation medium (ICN Biomedicals, Basingstoke, UK). It was then centrifuged at 390g for 30 min at 20°C. The lymphocyte layer was removed and washed twice with PBS, and the cell pellet was resuspended and an aliquot removed for cell counting. GSH was extracted from the lymphocyte pellet and measured using an enzymatic recycling method (14) modified for an automated biochemistry analyzer (Cobas Mira). This ratio was measured at 0 and 90 min during each glucose tolerance test and expressed as a ratio of reduced to oxidized GSH. The 90-min time point was chosen as being peak hyperglycemia and, by extension, the time point most likely to demonstrate changes in hyperglycemia-associated oxidative stress

The determination of 8-epi-prostaglandin F_2α (8-epi-PGF _2α) was based on previously described methodology (11,15). Peripheral venous blood (10 ml) was collected into polyethylene tubes containing a 3.8% (wt/vol) trisodium citrate solution (blood-to-anticoagulant ratio of 9:1) with indomethacin (as a cyclooxygenase inhibitor) and butylated hydroxytoluene (BHT; as a free radical scavenger) at final concentrations of 14 and 20 μmol/l, respectively. The sample was allowed to stand for 30 min at 4°C to enable complete inhibition of cyclooxygenase enzymes. Platelet-poor plasma was obtained by centrifugation at 1,120g for 15 min at 4°C. The plasma was transferred to a polypropylene screw-cap tube, and BHT was added at a final concentration of 20 μmol/l. The sample was then stored at −80°C until analysis. For determination of plasma total antioxidant status (TAOS), blood (5 ml) was collected in EDTA tubes and centrifuged at 1,120g for 15 min at 4°C. The plasma was transferred to screw-cap polypropylene tubes and stored at −80°C until analysis. Plasma samples were subjected to alkaline hydrolysis for the measurement of total (sum of free plus esterified) 8-epi-PGF_2α by GC/MS. Plasma (0.5 ml) was transferred to a glass tube, followed by the addition of 8-epi-PGF₂-d4 as an internal standard (2 ng in 20 ml ethanol). The sample was hydrolyzed with 1.0 mol/l aqueous potassium hydroxide (0.5 ml) for 30 min at 40°C. Hydrolysis was terminated by the addition of 0.1 mol/l hydrochloric acid (4.25 ml), and the pH of the sample was adjusted to 7.4 using 0.05 mol/l sodium phosphate buffer (4.5 ml). Isolation of 8-epi-PGF was carried out by immunoaffinity extraction of the hydrolyzed plasma. Samples were loaded on an immunoaffinity cartridge (prepared with an anti–8-epi-PGF₂ antiserum) preconditioned with 16 ml sodium phosphate buffer (0.05 mol/l, pH 7.4). The cartridge was washed with water (20 ml) to remove nonretained components, and 8-epi-PGF₂ was eluted using an acetone and water mixture (95:5 dilution, 5.5 ml). The immunoaffinity extraction steps were programmed into an Aspec XL sample processor (Gilson Medical Electronics, Villiers-le-Bel, France) and run automatically. The final eluate from the immunoaffinity extraction was dried under nitrogen and the sample converted to a perfluorobutyl/trimethylsilyl (PFB/TMS) derivative. Samples were analyzed by GC/negative ion chemical ionization/MS using an Autosystem XL GC coupled to a TurboMass MS (Perkin-Elmer, Beaconsfield, U.K.), with ammonia as reagent gas. The GC/MS assay has a limit of detection of ∼10 pg/ml (28 pmol/l) in plasma, with an intra- and interassay coefficient of variation of 4.4 and 7.6%, respectively. Analysis was performed using selected ion recording of the carboxylate anion [M-181] at m/z 569 for 8-epi-PGF_2α and m/z 573 for 8-epi-PGF-d4. Quantitative determination was based on the peak height ratio of 8-epi-PGF_2α against the internal standard. The isolation of 8-epi-PGF_2α from plasma is based on the specific interaction between 8-epi-PGF_2α and polyclonal anti–8-epi-PGF_2α antibodies, prepared by raising antisera against 8-epi-PGF_2α. After derivatization of the immunoextracted material, quantitation of 8-epi-PGF_2α was carried out by stable-isotope dilution GC/MS with selected ion recording. The measurement of 8-epi-PGF_2α provides a direct index of lipid peroxidation on phospholipid membranes in vivo and reflects oxidative processes within this environment. Oxidative stress due to hyperglycemia is associated with the increased generation of oxygen-derived radicals, and it is this change in free radical–mediated oxidation that is measured through the analysis of F₂-isoprostanes. The nature of this assay makes it exceptionally unlikely (if not impossible) that direct interference by plasma glucose or insulin concentrations in vivo can occur; indeed, these types of assays are particularly valuable simply because they are not subject to interference from other constituents present in the assay mix. Plasma concentrations of 8-epi-PGF_2α can be expressed in absolute amounts or normalized to plasma total arachidonic acid. Normalized concentrations of 8-epi-PGF_2α are useful in situations where significant changes in the lipid profile or plasma fatty acids are expected, and they are unlikely to provide further information where intervention is limited to an isolated oral glucose load because even a high-calorie/high–saturated fat meal does not influence postprandial plasma total arachidonic levels (16).

The total antioxidant status of plasma was determined by its capacity to inhibit the peroxidase-mediated formation of the 2,2-azino-bis-3-ethylbensthiazoline-6-sulfonic acid (ABTS ⁺) radical. In this assay, the relative inhibition of ABTS ⁺ formation in the presence of plasma is proportional to the antioxidant capacity of the sample. Briefly, plasma (2.5 μl) was incubated for 3 min at 37°C in a 96-well plate with a reaction mixture made up of (final concentrations) 20 μl ABTS (20 mmol/l), 20 μl horseradish peroxidase (30 mU/ml), and 37.5 μl PBS (pH 7.4). The reaction was started by the addition of 20 ml hydrogen peroxide (final concentration 0.1 mmol/l), and the increase in absorbance over 6 min was monitored at 405 nm, using a microplate reader (model 12605; Anthos Labtech, Salzburg, Austria). At the end of 6 min, the absorbance due to the accumulation of ABTS ⁺ in the test sample was read along with a control (containing 2.5 μl PBS instead of plasma). The difference in absorbance (control absorbance minus test absorbance), divided by the control absorbance (expressed as a percentage) was used to represent the percentage inhibition of the reaction.

Plasma lipid profiles were measured at time 0 and 90 min during each glucose tolerance test. Lipid profiles were assessed using commercially available kits (Roche) on an automated biochemistry analyzer (Cobas Mira; Roche), with estimation of LDL cholesterol (17).

Data are shown as the means ±1 SD, and all variables were normally distributed. Differences in individual variables measured more than twice during glucose tolerance test were analyzed by repeated-measure one-way ANOVA, with paired t tests where a significant difference (P < 0.05) was found. Otherwise, paired t tests were used for paired measurements. Relationships between variables were analyzed by simple linear regression or stepped multiple regression analysis with entry at P < 0.1. Data were analyzed using Apple Macintosh Statview software (1996).

Clinical details are shown in Table 1. Of the diabetes group, 17 (81%) were taking metformin or a metformin-sulfonylurea combination. Seven patients (33%) were taking gliclazide, either alone or in combination with metformin, and seven (33%) were taking antihypertensive medication (Tables 1 and 2).

There were no significant changes in the intracellular oxidative balance measured by reduced-to-oxidized GSH ratio or in the plasma total antioxidant status (P=0.1) (Table 2).

There was a highly significant (P=0.0102) rise in plasma 8-epi-F_2α isoprostane concentrations between baseline and 90 min. There were no significant relationships between baseline plasma 8-epi-F_2α isoprostane concentrations and HbA_1c (r=0.32, P=0.15), fasting plasma glucose (r=0.33, P=0.13), or measures of intracellular oxidative balance and total antioxidant capacity (both P > 0.2). Peak plasma 8-epi-F_2α isoprostane concentrations were directly related only to TAOS at 90 min (r=−0.495, P=0.025). Stepwise multiple regression demonstrated that only TAOS at 90 min was independently and inversely related to peak plasma 8-epi-F_2α isoprostane (R²=0.248, P=0.025). Finally, the only variable independently inversely related to the absolute change in plasma 8-epi-F_2α isoprostane concentrations during the glucose tolerance test was TAOS at 90 min (r=−0.45, P=0.041), and when the upper tertile of the 8-epi-F_2α isoprostane increment was compared with the lower tertile, only TAOS at 90 min differed between tertiles (47.8± 10.4 vs. 62.9± 15.7%, respectively; P=0.023) (Table 2).

The main finding of this study is that acute hyperglycemia after a glucose load in type 2 diabetes is associated with an acute increase in plasma concentrations of 8-epi-F_2α isoprostane. This must indicate increased free radical–mediated generation of these compounds from arachidonic acid in membrane and lipoprotein phospholipids (8–12). This provides sensitive and direct evidence for a link between acute rather than chronic hyperglycemia and free radical damage in type 2 diabetes.

Changes in oxidative balance, or in antioxidant defenses, during acute hyperglycemia have been demonstrated before in subjects with and without type 2 diabetes (18–21). Increased LDL oxidizability induced by copper ions (22,23) has also been reported postprandially in type 2 diabetes, although the difficulties with this methodology have been reviewed (7), as has the confounding effect of dietary oxidized lipids (24). However, these measurements are surrogates, and the present study demonstrates directly increased plasma levels of a free radical–generated oxidation product of membrane arachidonic acid during acute hyperglycemia, independent of dietary intake of oxidized lipids or dietary isoprostane intake (25). Increased generation of reactive oxygen species is a feature of hyperglycemia in type 2 diabetes and impaired glucose tolerance (5,26), and the increased free radical damage during acute hyperglycemia demonstrated in this study occurred without significant changes in intracellular oxidative balance or plasma antioxidant capacity. That plasma total antioxidant capacity was the sole independent determinant of the increase in 8-epi-F_2α isoprostane may suggest that acute increases in free radical generation during hyperglycemia do not influence TAOS, but that it is limited by the variety of antioxidant defenses contributing to the TAOS measurements. It is also possible that the TAOS and GSH assays are less sensitive for detecting changes in oxidative balance or stress compared with the 8-epi-F_2α isoprostane assay used here, or that changes in TAOS and intracellular oxidative balance occurred before the 90-min sampling point used in this study.

Plasma F2 isoprostane concentrations increased by 34% during acute hyperglycemia, and this is similar to observations in other models of increased oxidative damage. For example, Morrow et al. (9) demonstrated increased plasma esterified 8-epi-F_2α isoprostane in heavy smokers, and smoking cessation led to a 24% reduction in mean 8-epi-F_2α isoprostane concentration within a few weeks. 8-Epi-F_2α isoprostane may also have biologically important proatherogenic actions, as well as being a marker for free radical damage. In vitro, 8-epi-F_2α isoprostane at physiological concentrations promotes increased message and protein for endothelin-1 (27), promotes platelet adhesion to collagen in a dose-dependent manner (28), and antagonizes some actions of nitric oxide (29). Also, increased levels of 8-epi-F_2α isoprostane are detectable in human coronary atherosclerotic plaque (29), particularly in smokers with increased oxidative damage rather than patients with treated hypertension or dyslipidemia (30), suggesting that 8-epi-F_2α isoprostane generation may occur within coronary plaque.

Metanalysis of available epidemiological and prospective studies has shown a consistent direct relationship between blood glucose levels and predominantly cardiovascular mortality (31) in type 2 diabetes, but some data suggest postprandial hyperglycemia may be an independent predictor of cardiovascular mortality in type 2 diabetes (1–3), and surrogates for macrovascular disease, such as carotid intimal-medial thickness, are more closely related to acute rather chronic hyperglycemia (32), and much of the controversy over the diagnostic classification of diabetes was based on the predictive power of post–glucose load hyperglycemia as a marker for increased vascular risk (33). The present data suggest one possible pathway for this association, by demonstrating increased free radical damage during acute hyperglycemia. This could promote an increase in vascular event rates through some of the mechanisms outlined above, or through proatherogenic processes sensitive to reactive oxygen species, such as increased adhesion molecule expression or coronary plaque metalloproteinase expression (34,35).

The present study demonstrates that acute hyperglycemia in type 2 diabetes is associated with a significant increase in free radical–mediated damage to membrane components, measured by plasma 8-epi-F_2α isoprostane concentrations, and this may be a link between acute hyperglycemia, increased free radical damage, and macrovascular risk in type 2 diabetes.

We are grateful to Drs. Greenwood, Heyburn, and Temple for allowing access to their patients, and to the Norwich and Norfolk Diabetes Trust for funding salary costs associated with this project.