Susceptibility to Oxidative Stress, Insulin Resistance, and Insulin Secretory Response in the Development of Diabetes from Obesity

Background/Aim. Oxidative stress plays a critical role in the pathogenesis of various diseases. Recent reports indicate that obesity may induce systemic oxidative stress. The aim of the study was to potentiate oxidative stress as a factor which may aggravate peripheral insulin sensitivity and insulin-secretory response in obesity in this way to potentiate development of diabetes. The aim of the study was also to establish whether insulin-secretory response after glucagon-stimulated insulin secretion is susceptible to prooxi-dant/antioxidant homeostasis status, as well as to determine the extent of these changes. Methods. A mathematical model of glucose/insulin interactions and C-peptide was used to indicate the degree of insulin resistance and to assess their possible relationship with altered antioxidant/prooxidant homeostasis. The study included 24 obese healthy and 16 obese newly diagnozed non-insulin dependent diabetic patients (NIDDM) as well as 20 control healthy subjects, matched in age. Results. Total plasma antioxidative capacity, erythrocyte and plasma reduced glutathione level were significantly decreased in obese diabetic patients, but also in obese healthy subjects, compared to the values in controls. The plasma lipid peroxidation products and protein carbonyl groups were significantly higher in obese diabetics, more than in obese healthy subjects, compared to the control healthy subjects. The increase of erythrocyte lipid peroxidation at basal state was shown to be more pronounced in obese daibetics, but the apparent difference was obtained in both the obese healthy subjects and obese diabetics, compared to the control values, after exposing of erythrocytes to oxidative stress induced by H 2 O 2. Positive correlation was found between the malondial-dehyde (MDA) level and index of insulin sensitivity (FIRI). Conclusion. Increased oxidative stress together with the decreased antioxidative defence seems to contribute to decreased insulin sensitivity and impaired insulin secretory response in obese diabetics, and may be hypothesized to favour the development of diabetes during obesity.


Introduction
Oxidative stress plays a critical role in the pathogenesis of various diseases 1 .Recent reports indicate that obesity may induce systemic oxidative stress 2 .Increased free radical production in accumulated fat, because of increased nicotinanide adenine dinucleotide phosphate (NADPH) oxidase and decreased antioxidant enzymes activity, causes dysregulated production of adipocytokines locally, such as plasminogen activator inhibitor-1 (PAI-1), tumor necrosis factor α (TNFα), resistin, leptin and adiponectin 3−8 .Increased oxidative stress in blood can affect other organs such as endothelium, liver and skeletal muscle, and may lead to decreased insulin sensitivity and development of insulin resistance in obese subjects 8−10 .
In diabetes, oxidative stress impairs glucose uptake in muscle and fat and decreases insulin secretion from pancreatic β cells 11−13 .Increased oxidative stress also underlies the pathophysiology of late diabetic complications directly affecting vascular wall 14 .
The life time risk to develop impaired glucose tolerance and diabetes is increased among obese subjects compared to normal-weight matched subjects 15,16 .Complex interactions between obesity and insulin resistance largely account for the pathogenesis of non-insulin dependent diabetes.There are many factors that can aggravate insulin resistance of obesity.Elevated insulin level itself is atherogenic by stimulating sympathetic nerve activity.Increased insulin secretion is associated with the progressive loss of β-cell function as well.Insulin resistance results in reduced peripheral disposal of glucose in muscle and increased hepatic glucose output in the fasting state 17,18 .Decreased concentration of reduced thiols together with increased lipid peroxidation products, may play a role in the development of insulin resistance, because an optimal antioxidant/prooxidant state is required for the stabilization of nitric oxide available for insulin action 19,20 .
The study was performed with the aim to estimate the relation between susceptibility to oxidative stress and the antioxidative defense system as possible risk factors of importance in the development of diabetes during obesity; to establish whether insulin-secretory response was susceptible to prooxidant/antioxidant status of the homeostasis, the extent of their changes in the course of stimulated insulin secretion was performed.A mathematical model of glucose/insulin interactions was used to indicate the degree of insulin resistance and to asses the possible relationship with the altered antioxidant/prooxidant homeostasis.

Methods
The study included 24 obese healthy and 16 obese newly diagnozed non-insulin dependent diabetic patients (NIDDM), as well as 20 control healthy subjects, blood donor volunteers, matched in age.The degree of obesity was defined in terms of body mass index−BMI (kg/m 2 ).Patients and control subjects taking antioxidative vitamins or other drugs with antioxidative properties were excluded from the study.
Heparinized venous blood was obtained by peripheral venopuncture.The first blood sample was always collected at 8.00 am after 12 h of fasting, the second was collected after the glucagon test.All samples were immediately centrifuged at 2 000g for 15 min, plasma was carefully removed.Sedimented erythrocytes were washed three times in physiological saline.Received plasma and erythrocytes were immediately processed for the analyses described below.
Residual β-cell function was assessed by using the intravenous glucagon stimulation test.Obese subjects and obese diabetics received 1mg intravenously (i.v.) glucagon and blood samples were collected at basal state (0 min) and 6 min after.The values for the residual β-cell function was assessed from the insulin and glucose concentrations by the formula 21 : β-cell function (%) = 20 × insulin/glucose-3.5 (before and after the glucagon stimulation).Insulin resistance was expreassed as an empirical fasting insulin resistance index (FIRI) consisting of the product of plasma insulin and glucose that was obtained by the formula 22 : FIRI = fasting glucose × fasting insulin/25 (normally around 1 unit).
The reduced glutathione level (GSH) in plasma was measured by spectrophotometric assay using DTNB 23 .The level of plasma GSH was expressed as μmol/l.
The packed erythrocytes (0.2 ml) were treated with the same volume of ice-cold H 2 O and 1.3 ml of 5% TCA-1.0 mM EDTA immediately after the drawing.The supernatant was analyzed for sulphydryl group content at 412 nm with DTNB in 0.1 M phosphate buffer 23 .The level of erythrocyte reduced glutathione (RBC GSH) was expressed as μmol/ml RBC.
Plasma antioxidative capacity was tested by the modification of procedure of Asakawa and Matsushita 24 using an in vitro model of Fe 2+ -induced peroxidation of liposome suspension (from egg yolks).For this purpose 1.5 ml of liposome suspension at the concentration of 0.5 mg of total egg yolk phospholipids/ml, 0.2 ml of 0.5 M phosphate buffer pH 7.4 and 0.3 ml of tested plasma sample were incubated at 37 o C for 30 min by adding 0.2 ml of 1 mM/l FeCl 2 .Then 1 ml of 30% TCA was added to the reaction mixture followed by centrifugation for 15 min at 10 000 g.One milliliter of 0.67% TBA was added to 1ml of the received supernatant and the mixture was boiled at 97 o C for 20 min.Optical absorbance was measured at 532 nm.In this way, under prooxidant conditions (in the presence of traces of iron) incubation was carried out and the capacity of plasma to scavenge free oxygen radical production was determined as malondialdehyde (MDA) level.The antioxidative capacity was expressed as the percent of inhibition of MDA formation com-Strana 393 Kocić R, et al.Vojnosanit Pregl 2007; 64(6): 391-397.pared with the control samples which contained phosphate buffer instead of plasma.
Plasma lipid peroxidation was determined by thiobarbituric acid (TBA) reactivity.A colored complex has a maximum absorbance at 532 nm.The level of MDA was expressed as μmol/l.The activity of xanthine oxidase was determined by the increase of uric acid production when xanthine was used as substrate 25 .
The level of lipid peroxidation products in erythrocytes was estimated by measuring TBA reactivity 26 .For this purpose 0.4 ml of packed cells were suspended in 1.6 ml of phosphate buffer saline pH 7.4 (as a 20% suspension of red cells).To this suspension 0.5 ml of 30% TCA was added and 2 ml of the obtained supernatant was transferred into another tube, and 0.15 ml of 0.1 M/l EDTA and 0.5 ml of 1% TBA (dissolved in 0.05 M NaOH) was added.Tubes were mixed and kept in a boiling water for 15 minutes.After thad the tubes were cooled to room temperature, absorbance was read at 532 nm and 600 nm.Absorbance at 600 nm was substracted from absorbance at 532 nm.Malondialdehyde values in nmol/ml of erythrocytes (RBC) were determined using the extinction coefficient of MDA/TBA complex at 532 nm (1.56 × 10 5 per cm per molar solution).The results were expressed as nmol MDA/ml RBC.
For estimation of erythrocyte susceptibility to H 2 O 2 induced oxidative stress, erythrocytes were washed with an isotonic saline solution containing 2.0 mM sodium azide to inhibit catalase activity 26 .Washed erythrocytes were resuspended in saline-azide potassium phosphate buffer pH 7.4.One mililiter of 20% cell suspension was then preincubated for 10 min at 37 o C. Peroxidative challenge was induced by addition of an equal volume (1.0 ml) of H 2 O 2 in isotonic saline-azide solution (final H 2 O 2 concentration 3.0 mM).The reaction was terminated by the addition of 0.5 ml of 30% TCA and mixture was centrifuged.Malondialdehyde was detected in the obtained supernatant as a basic TBA reacting agent and it was used as an indicator of lipid peroxidation.The obtained results were expressed as nmolMDA/ml RBC.
Protein carbonyls were measured in plasma proteins and the remaining protein pellet according to the method explained in our previous report 27 .
Heparinised blood samples were collected for the assay procedures.After preparing the hemolysate, where the labile fraction is eliminated, hemoglobins are retained by cationic exchange resin and chromatographic (ion exchange)spectrophotometric method was used for the measurement of hemoglobin A 1 c (BioSystems).Plasma fructosamine was measured by the method of Bake et al 28 .
Plasma glucose was monitored by an ILAB 300 analyzer.
Plasma C-peptide and insulin were assayed by using the radioimmunoassay kits (Inep-Vinca-Serbia).
All presented values were given as mean + SD for statistical evaluation of the results between two groups.For multiple comparisons of means the ANOVA analysis of variance was used.The coefficient of the correlation (r) between the investigated parameters was obtained by the linear regression analysis and by the regression equation.

Results
The clinical characteristics of 24 obese healthy subjects and 16 obese − newly diagnosed diabetics, at the time of the study, are shown in Table 1.The mean age of obese subjects was 48.52±6.4yr (mean ±SD; range 36−61 yr) and of obese diabetics 50.3±6.2yr (range 42−61 yr).There was a predominance of female (n = 17) over male obese subjects (n = 7) and of obese diabetics 9 were male and 7 female.Plasma HbA1c and fructosamine level were significantly higher in obese diabetics compared with the level in obese healthy subjects.Significant difference was found between functional activity of β-cells (expressed as percents of Cpeptide secretion during normal function).The obtained results point to significantly reduced residual β-cell function in obese diabetics at basal state and after glucagon load (Figures 1 and 2).
Significantly lower total plasma antioxidative capacity was found in both groups of the obese diabetics and the obese healthy subjects compared with the control values (Figure 3).The values of total plasma antioxidative capacity, as well as plasma and erythrocyte GSH (glutathione) concentration progessively decreased from obese healthy sub- jects to obese diabetics (Figures 3 and 4).The level of plasma proten oxidative modification products-protein car-bonyls was significantly increased in both groups compared to controls (Figure 4).Plasma MDA level of obese diabetics had the highest value, but in obese healthy subjects it was also increased compared with the control values.Statistically significant increase in the concentrations of erythrocyte MDA products was found in obese diabetics and obese healthy subjects compared to control healthy subjects.The changes in erythrocyte MDA concentration after H 2 O 2 -induced oxidative stress, which are also documented in Figure 5, indicate that the cells of obese diabetics are the most susceptible ones to oxidative stress.The activity of xanthine oxidase was significantly increased in obese subjects but in obese NIDDM it was almost twice as of the control (Figure 6).The level of plasma MDA positivelly correlated with FIRI, indicating that increased oxidative stress may contribute to increased insulin resistance (Figure 7).

Control
Obese healthy subjects Obese diabetic patients 0 Plasma antioxidative capacity was tested by the modification of procedure of Asakawa and Matsushita 24 using an in vitro model of Fe 2+ -induced peroxidation of liposome suspension (from egg yolks).Protein carbonyls was measured in plasma proteins according to the method explained in our previous report 27 Obese healthy subjects Obese diabetic patients 0 0,2

Discussion
The accumulation of fat and BMI closely correlates with the markers of systemic oxidative stress 2,29,30 .It was documented that the adipose tissue represents a major source of the elevated plasma free radicals in obesity, because of the increased NADPH oxidase activity and elevated level of fatty acids.A high level of mRNA expression of the transcription factor PU.1, which upregulates the transcription of the NADPH oxidase gene in adipose tissue of obese mice was documented 30 .Byproducts of lipid peroxidation, such as trans-4-hydroxy-2-nonenal and malondialdehyde, are documented as potent chemoattractants, which may cause the infiltration of macrophages and inflammation in adipose tissue during obesity 2, 31−33 .They may be the important source of inflammatory cytokines, are also known to produce free radicals via NADPH oxidase 33,34 .In the present study, it was documented that obesity may serve as an independent risk factor for the decreased activities of antioxidants and for the associated susceptibility to systemic oxidative stress.The generation of reactive oxygen species has been suggested to occur via xanthine oxidase activity as well.
Plasma and cellular level of antioxidant defense system represents an antioxidant endogenous potential that can protect cells from the toxic effects of free radicals.Fall in antioxidant plasma and cellular defense (plasma antioxidative capacity, plasma GSH as well as RBC GSH) was the most obvious in obese diabetics (Figures 3 and 4).A possible reason for the decreased level in obese subjects and obese diabetics may be related to decreased glutathione reductase activity, decreased GSH synthesis by glutathionesynthetase, γ-glutamyl-cysteine synthetase, glutathione-Stransferase and the impaired transport of thiol compounds in erythrocytes 35−39 .The exogenous GSH can potentiate and modulate glucose-induced insulin secretion from rat pancreatic islets in a dose-related manner.Reduced plasma antioxidative activity which was found in obese subjects and especially obese diabetics, could be also due to the structural modifications, such as glycosylation, of antioxidative enzymes 40 .
Recent results have indicated that a local increase in oxidative stress in accumulated fat causes dysregulated production of adipocytokines.It was also shown that nuclear translocation of PPARγ was inhibited in conditions of increased generation of free radicals 41,42 .Therefore, downregulation of adiponectin expression may be partially attributed to the decreased gene expression and smaller amount of nuclear PPARγ under increased oxidative stress.Oxidative stress is known to impair both insulin secretion by pancreatic β cells 13 and glucose transport in muscle and adipose tissue 11, 12 .The reactive oxygen species (ROS), such as H 2 O 2, are produced transiently in the response to insulin stimulation and also act as a second messenger for insulin signaling in adipocytes via NADPH oxidase 43−45 .A transient increase of intracellular ROS is important for the insulin signaling pathway, while excessive and long term exposure to ROS reduces insulin sensitivity and impairs glucose and lipid metabolism.An increased oxidative stress is capable of decreasing insulin sensitivity and peripheral insulin effectivity.Insulin is effective through the increased nitric oxide (NO) production, but in the presence of free oxygen radicals the activating system NO-guanylate cyclase-c-GMP may have impaired functional activity 46 .The results which are summarized in Figure 7 indicate that the increased oxidative stress positively correlated with FIRI.The erythrocytes of obese diabetics are more susceptible to lipid peroxidation with a free radical initiator than the erythrocytes of obese healthy subjects or the controls (Figure 5).Newly diagnosed NIDDM patients showed the more pronounced increase of circulating level of MDA in plasma and erythrocytes compared with obese healthy subjects or the control subjects.The results of the San Antonio Heart Study 47 demonstrated that hyperglycaemia itself can cause the increased oxidative stress and susceptibility to oxidation of macromolecules, including plasma lipids and proteins.Abnormalities in membrane function, including the decrease of insulin receptors and glucose transport have been already shown to exist in experimental and human obesity and diabetes, but together with the enhanced production of free oxygen radicals, they are able to damage cell membrane systems leading to the mechanical disruption and increased permeability 37,48 .Enhanced production of free oxygen radicals could cause the damage of cells leading to the decreased secretory potential of islet cells, because they have very low free radical scavenging enzyme activities 48 .Xanthine oxidase (XO) reaction is the key catabolic reaction of purine metabolism, representing at the same time a central mechanism of oxidative tissue injury through the production of superoxide anion radical.Its activity is significantly increased in obese healthy subjects and obese NIDDM patients (Figure 6).
Our previous results indicated that more aggressive insulin-sensitizing therapy is the therapy of choice in reducing oxidative stress in non-insulin-dependent diabetes 49 .The results of the present study may confirm current suggestions that insulin-sensitizing therapy, given together with antioxi-dants, may delay the onset and/or development of noninsulin dependent diabetes during obesity 50 .

Conclusion
The disbalance between free radical-induced increase in lipid and protein oxidative modification, together with the decreased antioxidant plasma and cellular defences, may aggravate insulin resistance and decrease insulin-secretory response in obese subjects.

Fig. 5 −%Fig. 4 −
Fig. 5 − Plasma and erythrocyte (RBC) MDA at basal state (I) and after H 2 O 2 -induced oxidative stress (II) Plasma lipid peroxidation was determined by thiobarbituric acid (TBA) reactivity 25 .The level of MDA was expressed as μmol/l.The level of lipid peroxidation products in erythrocytes was estimated by measuring thyobarbituric acid (TBA) reactivity 26 .For the estimation of erythrocyte susceptibility to H 2 O 2 -induced oxidative stress, the peroxidative challenge was induced by addition of an equal volume (1.0 ml) of H 2 O 2 in isotonic saline-azide solution (final H 2 O 2 concentration 3.0 mM).Results were expressed as nmol MDA/ml RBC

Fig. 1 −Fig. 2 − 21 ControlFig. 3 −Fig. 6 −Fig. 7 −
Fig. 1 − Fasting and glucagon-stimulated C-peptide in obese healthy subjects and diabetic patients Obese subjects and obese diabetics received 1mg iv glucagon and blood samples were collected at basal state (0 min) and 6 min after.C-peptide was measured in both samples