Introduction

Diabetes mellitus is a chronic metabolic disorder characterized by hyperglycemia due to insufficiency of endogenous insulin secretion or action or post-receptor events affecting carbohydrate, fat and protein metabolism [34]. Chronic hyperglycemia in diabetes leads to auto-oxidation of glucose, non-enzymatic protein glycosylation, impaired glutathione metabolism, alteration in antioxidant enzymes and formation of lipid peroxides; these events accelerate production of free radicals and weaken the antioxidant defense [4, 39, 41, 61]. Increased production of reactive oxygen species may play a major role in the destruction of the pancreas and the progression of β-cell dysfunction [22]. Therefore, oxidative stress is believed to play a pivotal role in the etiology and pathogenesis of diabetes mellitus and its long-term complications [48]. Furthermore, it has been hypothesized that reduction of hyperglycemia and improvement in the control of blood sugar could reduce oxidative stress and ameliorate the metabolic functions of pancreatic β-cells [10]. Although numerous synthetic oral hypoglycemic drugs have been developed to combat diabetes, the management of diabetes without any side effects is still a challenge. Renewed attention to alternative medicines and natural therapies has stimulated a new wave of research interest in traditional practices, and there is a need to look for more efficacious agents with fewer side effects [25]. One of the most studied groups of natural products has been the flavonoids, which are found in several fruits and vegetables and constitute part of daily food consumption [7].

Naringenin is a flavanone compound found in citrus fruits, such as grapes and oranges, and also in tomato skin. Naringenin has been pharmacologically evaluated for antioxidant [5], anti-inflammatory [58], anticancerous [50], antiatherosclerotic [66], hepatoprotective [68], nephroprotective [3] and immunomodulatory [15] activities. Naringenin effectively quenches free radicals due to a 4'hydroxyl group in its B ring [64]. Recently, Ortiz-Andrade et al. [45] reported that naringenin is non-toxic, with a high (5,000 mg/kg) LD50, and prevents the absorption of glucose from the intestine in rats with non-insulin-dependent diabetes mellitus. However, a review of the literature reveals a paucity of data on the protective effect of naringenin on the pancreas. In the present investigation, we attempted to evaluate the putative antihyperglycemic and pancreatic-protective effect of naringenin in rats with experimental streptozotocin (STZ)–nicotinamide-induced diabetes mellitus.

Materials and methods

Chemicals

(±) Naringenin (95%), streptozotocin and nicotinamide were procured from Sigma Chemical Co. (St. Louis, MO, USA), stored at 2–4°C and protected from direct sunlight. All other chemicals used were purchased from standard local commercial suppliers and were of analytical grade.

Experimental animals

Male albino rats of the Wistar strain (150–180 g) were used in the present study. The animals were housed in clean polypropylene cages under conditions of controlled temperature (25 ± 2°C) with a 12/12-h day–night cycle, during which time they had free access to food and water ad libitum. Animal experiments were conducted per national guidelines and protocols approved by the Institutional Animal Ethical Committee.

Experimental induction of diabetes mellitus

Diabetes was induced experimentally in 12-h-fasted rats by a single intraperitoneal injection of streptozotocin (50 mg/kg) dissolved in 0.1 M of cold citrate buffer (pH 4.5) [52], followed by intraperitoneal administration of nicotinamide (110 mg/kg) after 15 min [36]. Since STZ is capable of inducing fatal hypoglycemia due to marked release of insulin from the pancreas, the STZ-administered rats were provided after 6 h with a 10% glucose solution for 24 h so as to prevent hypoglycemia. After 72 h, rats with a blood glucose concentration above 250 mg/dl were considered to be diabetic and used for further investigation.

Experimental design

The experimental rats were divided into five groups, each comprising six rats:

  1. Group I:

    Normal rats that received vehicle alone (0.5 ml of 0.1 M of cold citrate buffer [pH 4.5])

  2. Group II:

    Normal rats that were treated with naringenin (50 mg/kg b.w./day) in 0.5% carboxymethyl cellulose orally for 21 days

  3. Group III:

    STZ–nicotinamide-induced diabetic rats that were untreated

  4. Group IV:

    STZ–nicotinamide-induced diabetic rats that were treated with naringenin (50 mg/kg b.w./day) in 0.5% carboxymethyl cellulose orally for 21 days

  5. Group V:

    STZ–nicotinamide-induced diabetic rats that were treated with glyclazide (5 mg/kg b.w./day) in aqueous suspension orally for 21 days

In the present investigation, naringenin was administered in a concentration of 50 mg/kg b.w. since the same concentration has been used in several previous studies [27, 45, 55]. Similarly, in the present study, naringenin was administered for a period of 21 days because, in previous studies, the duration of administration has varied from 5 days to 4 weeks. At the end of the experimental period, the rats were fasted overnight, anesthetized and finally sacrificed by cervical decapitation. From each rat, blood samples were collected; subsequently, serum and plasma were separated. The pancreatic tissues were excised and stored at -80°C until analysis.

Preparation of pancreatic tissue homogenate

Pancreatic tissue from each experimental animal was homogenized (100 mg/ml buffer) with 50-mM phosphate buffer (pH 7.0) and centrifuged at 12,000 g for 15 min at 4°C. The supernatant thus obtained was used for biochemical assays. The protein concentration of each fraction was determined by the method of Bradford [8], using crystalline bovine serum albumin as standard.

Biochemical assays

Determination of blood glucose, glycosylated hemoglobin and serum insulin

Blood glucose was assayed by the glucose oxidase–peroxidase (GOD-POD) method using a commercial kit (ERBA, Mumbai, India). Glycosylated hemoglobin was assayed in the blood by the chromatographic–spectrophotometric ion-exchange method using a commercial diagnostic kit (Biosystems S.A., Barcelona, Spain). Serum insulin levels were assayed by solid phase enzyme-linked immunosorbent assay using a commercial kit (UBI Magiwel, Mountain View, Canada).

Determination of activities of hepatic marker enzymes in serum

Activities of AST and ALT were determined [30] and expressed in terms of micromoles of pyruvate liberated/min/mg of protein at 37°C. ALP activity was assayed using disodium phenyl phosphate as substrate[29] and expressed as micromoles of phenol liberated/min/mg of protein. LDH was assayed by the method of King [29]. The method is based on the ability of LDH to convert lactate to pyruvate with the help of the coenzyme nicotinamide adenine dinucleotide (NAD). The pyruvate formed was made to react with dinitrophenyl hydrazine in HCl. The hydrazone formed turned into an orange color complex in alkaline medium, which was measured at 420 nm.

Determination of activities of antioxidant enzymes in pancreatic tissue

Superoxide dismutase (SOD) (units/min/mg protein) activity was assayed by the method of Marklund and Marklund [35]. In this test, the degree of inhibition of pyrogallol auto-oxidation by the supernatant of the tissue homogenate was measured. The change in absorbance was read at 470 nm against blank every minute for 3 min on a spectrophotometer. The enzyme activity was expressed as units per milligram protein.

Catalase (CAT) activity was assayed by the method of Sinha [59]. In this test, dichromatic acetic acid is reduced to chromic acetate when heated in the presence of hydrogen peroxide (H2O2), with the formation of perchloric acid as an unstable intermediate. In the test, the green color developed was read at 590 nm against blank on a spectrophotometer. The activity of catalase was expressed as micromoles of H2O2 consumed/min/mg protein.

The activity of glutathione peroxidase (Gpx) was determined essentially as described by Rotruck et al. [56]. The principle of this method is that the rate of glutathione oxidation by (H2O2), as catalyzed by the Gpx present in the supernatant, is determined; the color that develops is read against a reagent blank at 412 nm on a spectrophotometer. The activity of Gpx was expressed in terms of microgram of reduced glutathione (GSH) consumed/min/mg protein.

The activity of glutathione-S-tranferase (GST) was assayed by the method of Habig and Jacoby [18]. The conjugation of GSH with 1 chloro, 2-4 dinitrobenzene (CDNB), a hydrophilic substrate, was observed spectrophotometrically at 340 nm to measure the activity of GST. The enzyme activity was expressed as micromoles of CDNB formed/min/mg protein.

Determination of the levels of non-enzymatic antioxidants in plasma

The concentration of reduced glutathione (μg of GSH /mg protein) was assayed by the method of Moron et al. [40]. To the supernatant of the plasma, 0.5 ml of 10% trichloroacetic acid was added, and the mix was re-centrifuged. To the protein-free supernatant, 4 ml of 0.3-M Na2HPO4 (pH 8.0) and 0.5 ml of 0.04% (w/v) 5,5-dithiobis-2-nitrobenzoic acid were added. The absorbance of the resulting yellow color was read spectrophotometrically at 412 nm. The results were expressed as microgram per milligram protein.

Vitamin C was measured by the method of Omaye et al. [44]. Vitamin C is oxidized by copper to form dehydroascorbic acid which reacts with 2,4-dinitrophenyl hydrazine to form bis-2,4-dinitrophenyl hydrazine; this undergoes further rearrangement to form a product with an absorption maximum at 520 nm. The results were expressed as microgram per milligram protein.

Vitamin E was estimated by the method of Desai [13]. In this method, ferric ions are reduced to ferrous ions in the presence of tocopherol, resulting in the formation of a pink color, read spectrophotometrically at 536 nm. The results were expressed as microgram per milligram protein.

Determination of pancreatic lipid peroxidation

Malondialdehyde (MDA) (nmol/mg protein), a measure of lipid peroxidation, was assayed in the form of thiobarbituric acid-reacting substances (TBARS) by the method of Ohkawa et al. [43]. Briefly, to 0.2 ml of 8.1% sodium dodecyl sulphate, 1.5 ml of 20% acetic acid (pH 3.5) and 1.5 ml of 0.81% thiobarbituric acid aqueous solution were added in succession. To this reaction mixture, 0.2 ml of the pancreatic tissue sample was added. The mixture was then heated in boiling water for 60 min. After cooling to room temperature, 5 ml of butanol: pyridine (15: 1 v/v) solutions were added. The mixture was then centrifuged at 2,000 g for 15 min. The upper organic layer was separated, and the intensity of the resulting pink color was read at 532 nm. Tetramethoxypropane was used as an external standard. The level of lipid peroxides was expressed as nanomole of MDA formed/mg protein.

Histological studies

Conventional techniques of paraffin wax sectioning and hematoxylin-eosin (HE) staining were used for histological studies [14]. Briefly, a portion of each pancreas was fixed in 10% formalin for a week at room temperature, followed by dehydration in a graded series of ethanol, clearing in xylene and embedding in paraffin wax at 57°C. Sections of 4–6 μm thickness were cut, stained by aqueous hematoxylin and alcoholic eosin and were examined by bright-field microscopy (Carl Zeiss Axioskop 2 plus; Jena, Gera, Germany).

Statistical analysis

The values listed in Tables 1, 2 and 3 and Figs. (bar diagrams) 1 and 2 represent the mean ± standard deviation (SD) of the observations made on six animals in each group. The statistical significance of differences between all of the groups was assessed by one-way analysis of variance (ANOVA) using SPSS software package for Windows (Version 16.0; SPSS Inc., Chicago, IL, USA). Post-hoc testing was performed for inter-group comparisons (between two groups) using the least significant difference (LSD) test. Values were considered statistically significant when P < 0.05.

Table 1 Mean blood glucose levels, glycosylated hemoglobin percentages and serum insulin levels in Wistar rats
Full size table
Table 2 Mean activities of enzymatic antioxidants in pancreatic tissue samples of Wistar rats
Full size table
Table 3 Mean levels of non-enzymatic antioxidants in plasma of Wistar rats
Full size table
Fig. 1
figure 1

Mean activities of hepatic marker enzymes in serum of Wistar rats. Values represent the mean ± SD of the observations made on six rats in each group. The enzyme activities are thus expressed: (1) ALT and AST as micromoles of pyruvate liberated/h/mg of protein, (2) ALP as micromoles of phenol liberated/min/mg of protein, and (3) LDH as micromoles of pyruvate formed/ h/ mg protein. AST aspartate aminotransferase, ALT alanine aminotransferase, ALP alkaline phosphatase, LDH lactate dehydrogenase. Group I normal untreated, Group II normal, naringenin-treated, Group III diabetic untreated, Group IV diabetic naringenin-treated, Group V diabetic glyclazide-treated. Statistical analysis: one-way analysis of variance [ANOVA] with post-hoc testing [least significant difference]. aStatistically significant difference (P < 0.05) when compared with Group I and Group II values. bStatistically significant difference (P < 0.05) when compared with Group III values. cStatistically significant difference (P < 0.05) when compared with Group V values

Full size image
Fig. 2
figure 2

Mean level of MDA in pancreatic tissue samples of Wistar rats. Values represent the mean ± SD of observations made on six rats in each group. MDA malondialdehyde. Group I normal untreated, Group II normal, naringenin-treated, Group III diabetic untreated, Group IV diabetic naringenin-treated, Group V diabetic glyclazide-treated. Statistical analysis: one-way analysis of variance [ANOVA] with post-hoc testing [least significant difference]. aStatistically significant difference (P < 0.05) when compared with Group I and Group II values. bStatistically significant difference (P < 0.05) when compared with Group III values

Full size image

Results

Assay of blood glucose levels, glycosylated hemoglobin percentages and serum insulin levels in Wistar rats

Table 1 lists the mean blood glucose levels, glycosylated hemoglobin percentages and serum insulin levels in each of the five groups of rats. Normal rats treated with naringenin (Group II) did not exhibit significant differences in the test parameters when compared to untreated normal (Group I) rats. Diabetic, untreated (Group III) rats showed significantly (P < 0.05) higher mean levels of blood glucose and glycosylated hemoglobin and significantly (P < 0.05) lower mean levels of serum insulin, when compared to untreated normal (Group I) and naringenin-treated normal (Group II) rats. Conversely, diabetic rats administered naringenin (Group IV) and those administered glyclazide (Group V) exhibited significantly (P < 0.05) lower mean levels of blood glucose and glycosylated hemoglobin and significantly (P < 0.05) higher mean levels of serum insulin, when compared to diabetic, untreated (Group III) rats. No significant differences were noted between the values obtained in naringenin-treated (Group IV) rats and those in glyclazide-treated (Group V) rats. However, although there were significant improvements in these parameters in Group IV and Group V rats (compared to the levels in diabetic untreated rats), the values in Group IV and Group V rats still differed significantly from those observed in the normal untreated (Group I) and normal naringen-treated (Group II) rats.

Activities of hepatic marker enzymes in serum of Wistar rats

The mean activities of serum ALT, AST, ALP and LDH were found to be significantly (P < 0.05) higher in Group III rats than those in Group I and Group II rats. Although rats in Groups IV and V exhibited significantly (P < 0.05) lower mean levels of these enzymes than Group III rats, the mean activities were still significantly higher than those in Group I and Group II rats. The mean activities of these enzymes in serum were significantly higher in Group IV rats than those in Group V rats. No significant differences were observed in these test parameters between Group I and Group II rats (Fig. 1).

Activities of enzymatic antioxidants in pancreatic tissue of Wistar rats

There were no significant differences between the mean activities of SOD, CAT, Gpx and GST in the pancreatic tissue of normal naringenin-treated (Group II) rats and the mean activities in pancreatic tissue of normal untreated (Group I) rats. Significantly (P < 0.05) lower mean activities of these enzymatic antioxidants were observed in the pancreatic tissue of diabetic untreated (Group III) rats than those in Group I and Group II rats (Table 2). However, significantly (P < 0.05) higher mean activities of these enzymes were observed in diabetic rats treated with naringenin (Group IV) and diabetic rats treated with glyclazide (Group V) than those in Group III rats (Table 2). Interestingly, there were no significant differences in the mean values of these parameters between naringenin-treated diabetic (Group IV) rats and glyclazide-treated diabetic (Group V) rats; however, the mean values of CAT, Gpx and GST in Group IV and Group V rats were significantly lower than those in untreated normal and naringenin-treated normal rats (Table 2).

Concentrations of non-enzymatic antioxidants in plasma of Wistar rats

The mean levels of GSH, vitamin C and vitamin E in the plasma samples of diabetic untreated rats were significantly (P < 0.05) lower than those in normal untreated and normal naringenin-treated rats (Table 3). In diabetic rats treated with naringenin and those treated with glyclazide, significantly (P < 0.05) higher mean levels of GSH, vitamin C and vitamin E were noted when compared to the values in diabetic untreated rats; however, they remained significantly lower than those in normal untreated and normal naringenin-treated rats (Table 3). No significant differences in the mean values of these parameters were noted between diabetic naringenin-treated rats and diabetic glyclazide-treated rats. There were also no significant differences between the mean values of these parameters in normal naringenin-treated rats and normal untreated rats (Table 3).

Malondialdehyde (MDA) concentrations in pancreatic tissues of Wistar rats

There was no significant difference between the mean values of pancreatic tissue MDA in Group II (normal naringenin-treated) rats and Group I (normal untreated) rats. The mean concentration of MDA in the pancreatic tissue of Group III (diabetic untreated) rats was significantly (P < 0.05) higher than that in Group I (normal, untreated) and Group II (normal naringenin-treated) rats. Interestingly, although the mean levels of pancreatic tissue MDA in diabetic rats treated with naringenin (Group IV) and in those treated with glyclazide (Group V) were significantly (P < 0.05) lower than that in diabetic untreated (Group III) rats, these levels were not significantly different to those in normal untreated (Group I) and normal naringenin-treated (Group II) rats. There was no significant difference between the mean values in Group IV and Group V rats (Fig. 2).

Histopathological examination of pancreatic tissue of Wistar rats

The histoarchitecture of the pancreatic tissue in three groups of Wistar rats is shown in Fig. 3a–c. Normal islets with clusters of β-cells were noted in the pancreatic tissue of normal untreated (Group I) rats (Fig. 3a). Degranulation of β-cells and vacuolation of the islets were noted in the pancreatic tissue of diabetic untreated (Group II) rats (Fig. 3b). Interestingly, regeneration of β-cells and reduction in the vacuolation of the pancreatic islets were observed in the pancreatic tissue of diabetic naringenin-treated (Group IV) rats (Fig. 3c).

Fig. 3
figure 3

Histoarchitecture of pancreatic tissues of Wistar rats. Hematoxylin-eosin staining. ×200 magnification. a Normal untreated (Group I) rats showing normal pancreatic islets with cluster of β-cells as indicated by arrows. b Streptozotocin–nicotinamide-induced diabetic untreated (Group III) rats showing degranulation of β-cells and severe vacuolation of the pancreatic islet as indicated by arrows. c Diabetic naringenin-treated (Group IV) rats showing preservation of almost normal profile of the β-cells and minimal vacuolation as indicated by arrows

Full size image

Discussion

Streptozotocin (STZ), N-(methyl nitro carbamoyl)-d-glucosamine is an antibiotic derived from Streptomyces achromogenes. This compound is universally used to induce diabetes mellitus in experimental animals due to its cytotoxic effects on pancreatic β-cells [22]. Nicotinamide can partially reverse the inhibition of insulin secretion to prevent aggravation of experimental diabetes following the administration of STZ [32]. STZ–nicotinamide-induced diabetic animals exhibit a number of attributes that are similar to those found in humans suffering from non-insulin-dependent diabetes mellitus, such as persistent hyperglycemia, glucose intolerance and significantly altered glucose-stimulated secretion of insulin [36]. Hence, in the present study, STZ–nicotinamide-induced diabetic male Wistar rats were used as the experimental model to evaluate the antihyperglycemic and pancreatic protective effect of naringenin.

In STZ–nicotinamide-induced diabetic untreated (Group III) rats, significantly (P < 0.05) higher mean blood glucose levels and significantly (P < 0.05) lower mean serum insulin levels were noted than the values in normal (Group I) rats (Table 1). However, administration of naringenin to STZ–nicotinamide-induced diabetic rats appeared to maintain blood glucose and serum insulin levels at near normal levels (Table 1). Naringenin possibly stimulated insulin secretion from the remaining pancreatic β-cells, thereby exerting its antihyperglycemic action in experimental diabetes. Interestingly, it has recently been shown that naringenin improves insulin signaling and sensitivity in a dietary model of metabolic syndrome [27]; moreover, naringenin has been found to prevent hepatic steatosis and to improve insulin sensitivity in animals fed a high fat diet [42].

In persistent hyperglycemia, the excess blood glucose reacts non-enzymatically with hemoglobin to form glycosylated hemoglobin, which is a reliable marker of ambient glycemia over a period of 90 days [1]. In the present investigation, significantly higher mean percentages of glycosylated hemoglobin were observed in diabetic untreated rats than in normal untreated rats (Table 1). However, lower mean blood percentages of glycosylated hemoglobin were observed in naringenin-treated diabetic rats, probably due to the reduction in mean blood glucose levels brought about by naringenin. Other investigators have also hypothesized that lowered levels of glycosylated hemoglobin are the consequence of reduced blood glucose levels [23, 46, 53]. Punithavathi et al. [51] recently reported that in streptozotocin-induced diabetic rats, administration of gallic acid resulted in decreased levels of blood glucose which, in turn, resulted in lowered levels of glycosylated hemoglobin.

The liver, a major site of insulin clearance and production of inflammatory cytokines, plays an important role in maintaining normal glucose concentrations in fasting and post-prandial states [31]. Elevated levels of serum enzymes, such as ALT, AST, ALP and LDH, are well-known markers of hepatic damage; these enzymes are believed to leak from the cytosol into the bloodstream as a consequence of damage to hepatic tissue [69]. In the present study, significantly (P < 0.05) higher mean serum levels of ALT, AST, ALP and LDH were observed in STZ–nicotinamide-induced diabetic untreated (Group III) rats than in untreated normal (Group I) rats (Fig. 1). However, in naringenin-treated diabetic (Group IV) rats, significantly (P < 0.05) lower mean levels of these serum enzymes were noted (Fig. 1), suggesting that naringenin may protect the hepatic tissue from oxidative stress-mediated cellular damage and may increase insulin action.

West [65] postulated that the chronic hyperglycemia occurring in diabetes mellitus could lead to an increased production of reactive oxygen species (ROS), this possibly resulting in depletion of certain non-enzymatic or enzymatic scavengers. Superoxide dismutase (SOD) is a natural cellular antioxidant enzyme which catalyzes the dismutation of superoxide radicals into hydrogen peroxide (H2O2) [38], while catalase (CAT) is a hemeprotein which converts H2O2 to water and oxygen [11]. Glutathione peroxidase (Gpx), being a selenium-containing enzyme, aids the removal of H2O2, therein preventing the formation of hydroxyl radical [OH-•] [20]; it also degrades hydroperoxide, a potentially toxic molecule [62]. In the present study, the mean activities of SOD, CAT and Gpx in pancreatic tissue samples were found to be significantly lower in diabetic untreated rats than in untreated normal rats (Table 2). The lower mean SOD activity was possibly due to accumulated superoxide radicals, H2O2-mediated inactivation of SOD or glycosylation of SOD [54]. Similar mechanisms possibly accounted for the lower mean CAT activity as well [37, 67]. The decreased mean Gpx activity observed in pancreatic tissue of diabetic untreated rats possibly represented an important adaptive response to increased peroxidative stress, as hypothesized by Kinalski et al. [28].

Recent studies have suggested that lowering of SOD, CAT and Gpx activity in diabetic rats could be prevented by treatment with antioxidants, such as cysteinyl metformin [33], aucubin [21] and S-allylcysteine [57], all of which decrease oxidative stress in pancreatic tissue. So also, in the present investigation, significantly (P < 0.05) higher mean activities of SOD, CAT and Gpx were noted in the pancreatic tissue of naringenin-treated diabetic rats than in diabetic untreated rats. Naringenin possibly protects pancreatic cells against oxidative insult.

Glutathione-S-transferase (GST), a glutathione-dependent enzyme, protects cells from ROS by utilizing a wide variety of products of oxidative stress as substrates [6]. In the present investigation, the mean activity of GST was found to be significantly lower in untreated diabetic rats than in untreated normal rats (Table 2). These data are consistent with those of earlier workers [47]. In naringenin-treated diabetic rats, the mean GST activity was markedly higher than that in untreated diabetic rats, possibly due to its antioxidant potential.

Reduced glutathione (GSH) is a major non-protein thiol in living organisms. Thiol groups are believed to play an important role in maintaining the intracellular and membrane redox state of the secretory function of β-pancreatic cells [48]. In the present study, the observed lower mean plasma levels of GSH in untreated diabetic rats (Table 3) were possibly due to increased utilization of GSH as an antioxidant defense against ROS, as suggested by others [17]. However, naringenin-treated diabetic rats exhibited significantly higher mean levels of GSH in plasma than untreated diabetic rats (Table 3). Ardestani et al. [2] reported that treatment with an extract of Teucrium polium restored the GSH content to near normal levels in diabetic rats and suggested that this was due to reduction in oxidative stress. In a similar way, naringenin may alleviate the intensity of oxidative stress, thereby reducing the degradation of GSH.

Ascorbic acid, a water-soluble antioxidant, prevents oxidative damage to the cell membrane induced by aqueous radicals and facilitates the maintenance of vitamin E levels at optimal concentrations [63]. α-tocopherol (vitamin E) is the most important lipid-soluble antioxidant, acting as a chain terminator of lipid peroxidation and protecting cellular structures from attack by free radicals [9]. In the present study, significantly (P < 0.05) lower mean levels of vitamins C and E were noted in plasma of untreated diabetic rats than in plasma of untreated normal rats (Table 3), observations which are consistent with those of earlier workers [24, 60]. However, the mean plasma levels of vitamins C and E were found to be significantly higher in naringenin-treated diabetic rats than those in diabetic untreated rats (Table 3). Naringenin may act by reducing hyperglycemia-mediated oxidative stress probably by decreasing the consumption of free radical scavengers.

Lipid peroxidation, a free-radical mediated propagation of oxidative insult to polyunsaturated fatty acids, is a characteristic feature of chronic diabetes; it impairs cell membrane fluidity and alters the activity of membrane-bound enzymes and receptors, resulting in membrane malfunction [19]. Malondialdehyde (MDA), a secondary product of lipid peroxidation, is used as an indicator of tissue damage [43]. The observed elevated mean level of MDA in diabetic untreated (Group III) rats in the present investigation possibly resulted from increased intensity of lipid peroxidation occurring due to intensified free radical production, as suggested earlier [34]. Oral administration of naringenin to diabetic (Group IV) rats brought about a significant decrease in the mean level of MDA. Naringenin possibly scavenges free radicals, therein stabilizing the endogenous antioxidant defense network and decreasing the level of lipid peroxidation, as has been suggested for other antioxidants such as quercetin [12], tetrahydrocurcumin [49] and gallic acid [51].

In the present investigation, shrinkage and vacuolization of pancreatic islets and reduction in β-cell mass were noted in pancreatic tissue of diabetic untreated (Group III) rats. However, oral administration of naringenin to diabetic (Group IV) rats appeared to preserve the remaining β-cell mass and to reduce vacuolization of pancreatic islets. Naringenin possibly protected the pancreatic islets from free radical-mediated oxidative stress and preserved the integrity of pancreatic β-cells, thereby stimulating the remaining pancreatic β-cells to synthesize and secrete additional insulin to maintain glucose homeostasis. Such a sequence of events has been proposed by other investigators to bolster the antioxidant defense system [16, 26]. These histological findings in diabetic rats treated with naringenin (Group IV rats) in the present study appear to correlate with increased levels of serum insulin, which would have brought about lowered levels of blood glucose and lowered percentages of glycosylated hemoglobin in this group of rats. Recently, Sivakumar and Subramanian [60] reported the same phenomenon in streptozotocin-induced diabetic rats treated with an antioxidant D-pinitol.

In conclusion, it may be suggested that naringenin confers protection against experimental diabetes through its antihyperglycemic and antioxidant properties.