Bioactivity of Flavonoids on Insulin-Secreting Cells

Authors


Abstract

ABSTRACT:  Flavonoids are usually found in fruits and other plant organs and therefore widely consumed. They are antioxidants, anti-inflammatory, anticarcinogenic, and protective against coronary disease and metabolic disorders. These beneficial effects make them good candidates for the development of new functional foods with potential protective/preventive properties against several diseases. We must consider that this fact could lead to a higher intake of some of these flavonoids. Most of the studies concerning their beneficial effects showed peripheral activity of these molecules, but there is no clear information about their central effects on a key organ on metabolic control: the endocrine pancreas. The pancreas has an endocrine function of major importance to regulate nutrient metabolism, such as control of glucose homeostasis via insulin and glucagon secretion. Its importance in whole body nutrient equilibrium is highlighted by the fact that several pathologies, such as type 1 and/or 2 diabetes, are related at some point to a pancreatic cell deregulation. In this review, we compile the most relevant results concerning the effects of flavonoids on several aspects of pancreatic functionality. Studies using animals with drug-induced diabetes support the hypothesis that flavonoids can ameliorate this pathogenesis. The great diversity of flavonoid structures makes it difficult to establish common effects in the pancreas. Published data suggest that there might be direct effects of flavonoids on insulin secretion, as well as on prevention of beta-cell apoptosis, and they could even act via modulation of proliferation. The mechanisms of action involve mainly their antioxidant properties, but other pathways might also take place.

Introduction

It is now generally accepted that food can have health-promoting properties that go beyond their traditional nutritional value (Saris and others 1998). Recently, much attention has been paid to some food factors that may be beneficial for the prevention of body fat accumulation and possibly reduce the risk of diabetes and heart diseases (Kovacs and Mela 2006). Flavonoids are a class of such bioactive compounds, usually found in fruits and other plant organs and therefore widely consumed. Their roles as antioxidants (Williams and others 2004; Scalbert and others 2005), anti-inflammatories (Li and others 2001), anticarcinogenics (Lu and others 2004), and protective agents against coronary disease (Bagchi and others 2003) and metabolic syndrome are widely accepted. These beneficial effects make them good candidates for the development of new functional foods with potential protective/preventive properties against several diseases. We must consider that this fact could lead to a higher intake in some of these flavonoids. Most of the studies concerning the beneficial effects of flavonoids have shown peripheral activity of these molecules (Pinent and others 2006; Koo and Noh 2007; Vafeiadou and others 2007; Meeran and Katiyar 2008), but there is no clear information about their central effects on a key organ on metabolic control: the endocrine pancreas. The pancreas has an endocrine function of major importance to regulate nutrient metabolism, such as control of glucose homeostasis via insulin and glucagon secretion. Its importance in whole body nutrient equilibrium is highlighted by the fact that several pathologies, such as type 1 and/or 2 diabetes, involved in nutrient metabolism, are related at some point to a pancreatic cell deregulation. In addition, the pancreas is located first in line after enteric absorption, so it could be reached by high concentrations of the bioactive-absorbed flavonoids. Therefore, it might be a target for the effects of flavonoids.

Given the great importance of an optimal functionality of the pancreatic tissue and the abundance of flavonoids in the human diet (including their potential use as bioactive molecules), this review discusses the most relevant results concerning the effects of flavonoids on several aspects of pancreatic functionality, including insulin secreting capacity, beta-cell apoptosis, and oxidative stress (also summarized in Table 1 and 2). We also revise the related effects of resveratrol because it is as well a phenolic compound and the literature describing its healthy effects has increased enormously in recent years (summarized in Table 3).

Table 1—.  Summary of in vitro effects of flavonoids.
FlavonoidConcentrationCell typeEffectReference
Genistein10 to 100 μmol/LMouse isletsIncrease in glucose-stimulated insulin secretion (GSIS)Jonas and others (1995)
100 μmol/LRat islets Sorenson and others (1994)
 MIN6 Ohno and others (1993)
>100 μmol/LRat isletsIncrease in GSISPersaud and others (1999)
50 μmol/LRat isletsInhibition of GSIS 
0.01 to 10 μmol/LINS-1Increase in GSISLiu and others (2006)
 MIN6  
 Mouse islets  
25 μmol/LHuman isletsReduction of NaF-induced apoptosisElliott and others (2002)
100 μmol/LHuman and rat isletsIncrease in apoptosis 
  Increase in NaF-induced apoptosis 
100 μmol/LRIN-m5FIncrease in apoptosis 
  Reduction in NaF-induced apoptosis 
100 μmol/LRat isletsInhibition proliferationSorenson and others (1994)
Anthocyanins and anthocyanidins50 μmol/LINS-1Increase in basal insulin secretionJayaprakasam and others (2005)
(cyanidin-3-galactoside and pelargonidin)  Increase in GSIS 
Puerarine50100 μmol/LRat isletsInhibition of the H2O2-induced apoptosisXiong and others (2006)
  Increase in the activities of antioxidant enzymes 
EGCG0 to 436.3 μmol/LRIN-m5FPrevention of cytokine-induced cell deathHan (2003)
 Mouse islets Song and others (2003)
10 to 100 μmol/LRat isletsReduction of the apoptosis induced by hypoxia-reperfusionHara and others (2006)
  Decrease in the markers of oxidative damage 
Sylimarin100 μmol/LRIN-m5FPrevention of cytokine-induced cell deathMatsuda and others (2005)
  Down-regulation of iNOS expression 
Table 2—.  Summary of in vivo effects of flavonoids.
FlavonoidTreatmentModelEffectReference
Soy protein30% in dietSprague-Dawley rats fed normal chow and high-fat diet  
Equol2.55 μM in plasma Inhibition of insulin secretionNoriega-Lopez and others (2007)
0.4 μM in plasmaSprague–Dawley rats  
Genistein Daidzein0.15 μM in plasma   
0.2 g/kg dietNonobese diabetic (NOD) miceReduction of plasma glucose levelsChoi and others (2008)
  Increase in plasma insulin levels 
  Increase in insulin-positive beta cells 
Quercetin10 to 15 mg/kg·day (10 d)Normoglycemic ratsIncrease in the number of pancreatic islet cellsVessal and others (2003)
 STZ-diabetic ratsIncrease in the number of pancreatic islet cells 
15 mg/kg·day Reduction of plasma glucose levelsCoskun and others (2005)
 (4 wk)STZ-diabetic ratsIncrease in plasma insulin levels 
  Reduction of plasma glucose levels 
  Protection of pancreatic beta-cell degeneration 
  Reduction of oxidative stress markers and increase in the activities of antioxidant enzymes in pancreas 
10 mg/kg·day (14 d)STZ-diabetic ratsIncrease in plasma glucose levelsColdiron and others (2002)
Rutin100 mg/kg·day (45 d)STZ-diabetic ratsIncrease in plasma insulin levelsStanley and Kamalakkannan (2006)
  Reduction of plasma glucose levels 
  Reduction of lipid peroxidative products and increase in the activities of antioxidant enzymes in pancreas 
EGCG5 mg/kg·day (4 d)STZ-diabetic ratsReduction of the plasma insulin levelsYun and others (2006)
  Reduction of the size of pancreatic islets 
100 mg/kg·day (10 d)STZ-diabetic ratsSuppression of the iNOS expression and prevention of islet damageSong and others (2003)
Epicatechin30 mg/kg twice a day (6 d)STZ-diabetic ratsPrevention of hyperglycemiaKim and others (2003)
  Prevention of hyperglycemia 
  Prevention of pancreas morphology 
  Increase in insulin release from isolated islets 
  Reduction of nitrite production 
Silymarin200 mg/kg (8 doses, 30 d)Alloxan-diabetic ratsIncrease in plasma insulin levelsSoto and others (2004)
  Reduction of plasma glucose levels 
  Induction of pancreas recovery 
GSPE50 to 100 mg/kg·day (1 to 3 d)Alloxan-diabetic ratsPrevention of hyperglycemia and hypoinsulinemiaEl-Alfy and others (2005)
  Restoration of oxidative status in the pancreas 
Catechins0.1% solutionHamstersProtection against N-nitrosobis(2-oxopropyl)amine- induced oxidative damageTakabayashi and others (1997)
600 mg/kg·day (10 d)STZ-diabetic ratsIncrease in serum insulin levelsSingab and others (2005)
  Reduction of serum glucose levels 
  Reduction of serum lipid peroxides 
Table 3—.  Summary of resveratrol effects.
In vitro studies
ConcentrationCell typeEffectReference
219.1 μmol/LINS-1No effect on insulin secretionZhang and others (2004)
3, 10, 30, 100 μmol/LRIN-m5F  
10, 30, 100 μmol/LHit-T15Increase in insulin secretionChen and others (2007)
10, 30 μmol/LMIN6  
1 to 100 μmol/LRat isletsInhibition of GSIS (dose-dependent)Szkudelski (2006, 2007, 2008)
In vivo studies
TreatmentModelEffectReference
3 mg/kg (1 dose)Normal wistar ratsIncrease in plasma insulin levelsChen and others (2007)
 Reduction of plasma glucose levels 
0.04% in diet (18 mo)HFD- fed miceReduction of hyperinsulinemiaBaur and others (2006)
400 mg/kg·dayHFD- fed miceReduction of fasting plasma insulinLagouge and others (2006)
50 mg/kg (1 dose)Normoglycemic ratsReduction of blood insulinSzkudelski (2008)
0.5 mg/kg, 3 times a day (14 d)STZ-nicotinamide-diabetic miceReduction of plasma insulinSu and others (2006)
3 and 10 mg/kg, (1 dose)Normal miceIncrease in plasma insulin levels, reduction of plasma glucose levelsChi and other (2007)
STZ-nicotinamide-diabetic miceIncrease in plasma insulin levels, reduction of plasma glucose levels 
STZ-diabetic miceNo effect in plasma insulin levels, reduction of plasma glucose levels 

Flavonoid classification and metabolism

Flavonoids are a class of phenolic compounds of plant origin. They are found in fruits, vegetables, and drinks such as wine, tea, and chocolate (Bhagwat and others 2003; Aron and Kennedy 2008). Their basic structure consists of 3 phenolic rings: the benzene ring (A) is condensed with a 6-member ring (C), which in the 2-position carries a phenyl benzene ring (B) as a substituent (Aherne and O'Brien 2002). Depending on the structure and oxidation level of the ring C, flavonoids are further divided into several subclasses. The main groups of flavonoids are anthocyanidins, flavonols, flavones, flavanones, isoflavones, and flavanols (Table 4). In addition, the basic flavonoid structures can be modified and have several kinds of substitutions (including glycosylation, hydrogenation, hydroxylation, malonylation, methylation, glucuronidation, and sulfatation) following different patterns, which confers them different physical properties. Furthermore, the degree of polymerization adds more variability to the flavonoids. So, the term flavonoid includes thousands of structures with different chemical, physical, and biological properties.

Table 4—.  Classification, representative structure, and food source of main flavonoids.
inline image

Some researchers support the view that the form in which flavonoids are found in plants is not their bioactive form (Manach and others 2005; Stevenson and Hurst 2007). It has been suggested that most aglycone flavonoids are absorbed by passive diffusion (Crespy and others 2003). One hypothesis points out that flavonoid glycosides must be modified (deglycosilated) by intestinal enzymes or colonic microflora in the intestine prior to absorption (Williamson and others 2000; Crespy and others 2003; Manach and others 2005). Next, absorbed flavonoids can be further metabolized in the enterocyte and in the liver (Spencer and others 2004; Walle 2004; Silberberg and others 2006; Walle 2007; Zhang and others 2007). All this high diversity of structures makes its detailed identification and plasma measurement very much complicated, but despite all these difficulties, several forms of modified flavonoids have been found in plasma (Manach and others 2005; Lotito and Frei 2006).

Furthermore, some flavonoids have also been found within tissues: stomach, small intestine, colon, liver, pancreas, spleen, kidney, muscle, heart, endothelial cells, lung, brain, thyroid, bone, skin, bladder, prostate, testes, vagina, uterus, ovary, mammary gland, fat, adrenal gland, esophagus, eyes, lymph nodes, and pituitary gland (Ueno and others 1983; Suganuma and others 1998; Chang and others 2000; Coldham and Sauer 2000; Kim and others 2000; Youdim and others 2000; Datla and others 2001; Abd El Mohsen and others 2002; Meng and others 2002; Mullen and others 2002; Swezey and others 2003; Vitrac and others 2003; Graf and others 2005). Concerning flavonoids in the pancreas, epigallocatechin gallate (EGCG) was detected in this tissue as soon as 1 h after oral administration and reached about 3 times higher levels after 24 h. A 2nd oral administration 6 h later resulted in a 4-fold increase of EGCG levels in the pancreas (Suganuma and others 1998).

Brief highlight of the role of the pancreas in disease

The pancreas is an organ highly involved in regulation of nutrient metabolism. Its function is regulated by nutrients (mainly glucose but also fatty acids and amino acids) and hormones (insulin itself, as well as others such as incretins). In turn, pancreas has an endocrine function of major importance to regulate nutrient metabolism, that is, the control of glucose homeostasis via insulin and glucagon secretion (by beta and alpha cells, respectively). In addition, pancreas has an exocrine function, since it secretes digestive enzymes into the gut. Its importance in whole body nutrient equilibrium is highlighted by the fact that several pathologies involved in nutrient metabolism are related at some point to a pancreatic cell deregulation (Muoio and Newgard 2006). It has been shown that during obesity there are changes in pancreatic beta cells to adapt to the increased energy intake and therefore glucose availability (Muoio and Newgard 2006). Obesity might also cause insulin resistance, which stimulates further insulin secretion by the pancreas. And this can lead to type 2 diabetes, where loss of effective beta-cell function is the key determinant of deteriorating glycemic control (Bagust and Beale 2003). Type 1 diabetes results also from loss of beta cells due to an immune assault (Ichinose and others 2007).

In fact, diabetes poses a major challenge to the present population globally. It is a major threat to global public health that is rapidly reaching epidemic scale, with the biggest impact on adults at working age in developing countries. According to the World Health Organization, at least 171 million people worldwide have diabetes. This figure is likely to be more than double by 2030 and reach 366 million (Wild and others 2004). It is postulated that most of this increase will occur as a result of a 150% rise in developing countries (Wild and others 2004). This estimation is likely to be even higher taking into consideration population growth, aging, urbanization, declining levels of physical activity, which—combined with unbalanced diet—increases the risk of obesity linked to the onset of type 2 diabetes. These facts imply that the pancreas is an important tissue for nutrient regulation and therefore a clear target for the action of putative bioactive compounds (Feick and others 2007).

Effects of Flavonoids on Insulin-Secreting Cells

Effects of flavonoids on insulin secretion

As already mentioned, a proper pancreatic function implies a well-regulated insulin secretion. Several studies describe the effects of flavonoids in insulin secretion.

In vitro studies demonstrate that some flavonoids modify the insulin-secreting capacity of the cell. Genistein is the most studied isoflavone. Early studies showed that this compound increases glucose-stimulated insulin secretion in the beta-pancreatic cell line MIN6 (mouse-derived) (Ohno and others 1993) as well as in cultured islets from mice (Jonas and others 1995) and rats (Sorenson and others 1994) at concentrations up to 100 μmol/L. Higher concentration of genistein, on the other hand, inhibited insulin secretion in rat islets (Persaud and others 1999). More recently, acute genistein treatment at physiologically achievable concentrations was shown to potentiate glucose-stimulated insulin secretion both in cell lines and isolated mouse islets (Liu and others 2006).

Several molecules from the class of anthocyanins and anthocyanidins are also effective insulin secretagogues when tested in pancreatic cell lines, with different efficiencies depending on the structure. It seems that the number of hydroxyl groups in ring B of anthocyanins played an important role in their ability to secrete insulin. Among the anthocyanidins tested, pelargonidin was the most active one (Jayaprakasam and others 2005).

Resveratrol, a plant phenolic compound with potential beneficial effects including prevention of diabetes and attenuation of some diabetic complications, has been shown to interfere with pancreatic beta-cell insulin secretion, although results are controversial. In 1 study where components of grape skin were extracted, the fraction corresponding to the compound resveratrol did not show any activity on insulin secretion by INS-1 cells (Zhang and others 2004). The concentrations used for this assay were high (50 μg/mL, which is about 219 μmol/L)(Zhang and others 2004). In fact, the concentration of resveratrol used for the experiments seems to be critical, since other experiments showed no effect of 100 μmol/L resveratrol on insulin secretion in MIN6 cells while lower concentrations (10 and 30 μmol/L) effectively activated it (Chen and others 2007). Instead, in other cell lines (Hamster-derived Hit-T15 and rat-derived RIN-m5F), resveratrol was insulin-inducer at all 3 concentrations tested (10, 30, 100 μmol/L) (Chen and others 2007). In pancreatic islets isolated from normal rats, incubation with resveratrol (1 to 100 μmol/L) inhibited glucose-stimulated insulin secretion in a dose-dependent way (Szkudelski 2006, 2008), whereas the basal hormone release was not modified (Szkudelski 2008).

Concerning in vivo effects, long-term studies of the effects of soy protein containing genistein and daidzein were performed in models fed a normal and a high-fat diet (Noriega-Lopez and others 2007). The study showed that chronic consumption of saturated fat increased insulin secretion associated with an increase in pancreatic islets, and soy protein ameliorated this situation. Thus, soy protein (due to its amino acid pattern as well as its isoflavones) reduces blood insulin. This was analyzed by hyperglycemic clamps, in which it was observed that rats fed soy protein secreted less insulin to maintain glucose at normal levels. Interestingly, this effect was observed also in normal-fed rats. Furthermore, a hyperglycemic clamp of rats infused with phytoestrogens (genistein, daidzein, and equol) showed that these compounds rapidly inhibit the release of insulin, suggesting a short-term mechanism regulating this process involving down-regulation of peroxisome proliferator-activated receptor gamma (PPARγ) and glucose transporter type 2 (Glut2) mRNA expression (Noriega-Lopez and others 2007). On the other hand, genistein and daidzein have been shown to elevate plasma insulin levels in nonobese diabetic (NOD) mice, an animal model that spontaneously develops autoimmune diabetes (Choi and others 2008). A 9-wk treatment with genistein or daidzein (0.2 g/kg diet, animals fed freely) suppressed the blood glucose rising in NOD mice by elevating plasma insulin levels. Such effects were accompanied by an increase in insulin-positive beta cells although it remained unresolved whether there was more insulin secretion from remaining beta cells or increased beta-cell mass in isoflavone-treated mice.

Regarding resveratrol, different results are found. A 3 mg/kg body weight dose (intraperitoneally administrated) increased fasting plasma insulin in normal rats, in association with a glucose reduction (Chen and others 2007). Such effect was observed at 20 min and it was sustained for 60 min. In contrast, blood insulin was reduced in normal rats by a 50 mg/kg body weight dose of resveratrol (Szkudelski 2008). Reduction in plasma insulin was also found in hyperinsulinemic mice models. The hyperinsulinemia induced by a high fat diet was reduced by resveratrol treatment (0.04% in diet, 18 mo) (Baur and others 2006). Similarly, another experiment with HFD-fed mice showed reduction in fasting plasma insulin by 400 mg/kg·day resveratrol (Lagouge and others 2006).

Many studies on diabetes make use of animal models in which pancreatic function is impaired due to drug (streptozotocin, alloxan) treatment and there are a few studies of the action of flavonoids using these models. Since in these animal models insulin secretion is impaired due to enhancement of beta-cell death, flavonoid effects will be discussed in another section.

Effects of flavonoids on beta-cell apoptosis

Complex diseases such as diabetes lead to loss of functional beta-cell mass. Actually, beta-cell apoptosis may be a common feature of type 1 and type 2 diabetes, although the mechanism activating it in each case may be different (reviewed by Cnop and others (2005)). Thus, the potential antiapoptotic effects of flavonoids, already demonstrated in other cell types (Bagchi and others 2002; Du and others 2007), must be carefully considered when analyzing their potential beneficial effects against these pathologies.

The flavonoid genistein modulates beta-cell apoptosis. Both in pancreatic islets and in a beta-cell line, low (25 μmol/L) genistein doses reduced NaF-induced apoptosis, whereas high (100 μmol/L) genistein doses caused apoptosis (Elliott and others 2002). The isoflavone puerarin was shown to significantly suppress the apoptosis of H2O2-induced rat islet cells. This effect was achieved by pretreating the cells with either 50 or 100 μmol/L puerarin, and was accompanied by an increase in the activities of antioxidant enzymes (Xiong and others 2006).

In type 1 diabetes; beta cells are destroyed by autoimmune responses mediated by cytokines. The monomer epigallocatechin gallate (EGCG) can prevent such cytokine-induced cell death both in insulinoma cell line (Han 2003) and in isolated mice islet cells (Song and others 2003). In both studies, the flavonoid effects were dose-dependent (maximal tested dose: 200 μg/mL), and achieved by cotreatment of the EGCG together with the cytokines for 24 h. Such effects could be mediated by down-regulation of nitric oxide synthase inducible (iNOS) expression through inhibition of nuclear factor-kappa B (NF-κB) activation (Han 2003). Furthermore, apoptosis can also be induced by hypoxia-reperfusion, a condition that, for example, takes place after transplantation of pancreatic islets. EGCG in rat pancreatic islets, reduced as well the apoptosis induced by hypoxia-reperfusion (Hara and others 2006). At the same time, EGCG decreased the markers of oxidative damage that had been induced by hypoxia (Hara and others 2006).

The cytokine-induced cell death is also prevented by the flavonoid silymarin (Matsuda and others 2005). In this case, the experiment was performed in an insulinoma cell line (RINm5F). The effect was mediated by suppression of Jun N-terminal kinase (JNK) and signal transducer and activator of transcription (STAT), although it is not clear how this effect is achieved, and whether it is only due to antioxidant activity (Matsuda and others 2005).

Thus, it seems that some flavonoids can exert antiapoptotic effects in pancreatic beta cells. More experiments are, however, required to elucidate whether such antiapoptotic effects can also take place in vivo and if this could ameliorate or prevent pancreatic dysfunctions linked to apoptosis of beta-cell mass.

Effects of flavonoids on beta-cell proliferation

Another important aspect of pancreatic integrity is regeneration of the endocrine pancreas. Beta-cell mass expansion in adult animals is slow. However, under circumstances of insulin demand adult animals can adapt beta-cell mass. The inability of the endocrine pancreas to undergo this adaptation can result in hyperglycemia and diabetes (Dhawan and others 2007). Thus, acting on mechanisms that regulate plasticity of beta cells can be an important step in developing effective strategies to treat the above-mentioned diseases. Furthermore, after beta-cell loss due to either type 1 or type 2 diabetes, beta-cell regeneration can occur (Tourrel and others 2002); (Meier and others 2005). Some hypotheses point out to neogenesis from pancreatic ducts as a mechanism for beta-cell regeneration, although others suggest that the regeneration of beta cells is mediated by replication of beta cells rather than neogenesis (reviewed by Levine and Itkin-Ansari (2007).

Several flavonoids modulate cell proliferation. Procyanidins (Pinent and others 2005; Lizarraga and others 2007), monomers such as quercetin (Notoya and others 2004) and gallic acid (Veluri and others 2006), and other flavonoids such as genistein (Harmon and Harp 2001) modulate cell proliferation in many cell types, including among others fibroblasts (Harmon and Harp 2001), osteoblasts (Notoya and others 2004), and several carcinoma cell lines (Veluri and others 2006; Lizarraga and others 2007). Studies performed in pancreatic beta cells are scarce, but the fact that there are key cell cycle regulators common in all these cell types suggests that flavonoids might have effects on beta-cell proliferation. In fact, genistein is a potent inhibitor of proliferation of cultured islet cell, when added at high concentrations (100 μmol/L) for 4 d (Sorenson and others 1994).

Oxidative stress diminishes beta-cell proliferation, and the protective effect of some phenolics on proliferation after oxidative stress in pancreatic beta cells has been tested. Some of such compounds, chrysin, quercetin, catechin, and caffeic acid, protected the cells significantly, whereas gallic acid and hesperitin did not (Lapidot and others 2002).

Concerning in vivo effects on beta-cell regeneration, the monomer quercetin (10 to 15 mg/kg per day for 10 d) increased the number of pancreatic islet cells in streptozotocin (STZ)-diabetic rats. Curiously, an increased number of beta cells was also found in normal (nondiabetic) rats (Vessal and others 2003). In any case, the rest of the studies concerning the effects of flavonoids as a potential preventive against pancreas destruction in drug-treated animals (discussed subsequently) did not show effects on nontreated animals.

Thus, although some evidence points out the potential role of flavonoids as modulators of beta-cell proliferation, their exact effects and the implications that this could have against the development of diabetes are still quite unexplored.

Mechanisms of Action

Flavonoid antioxidant effects on insulin secretion

Some of the most used models for diabetes are the animals with drug-induced diabetes, either by STZ treatment or by alloxan treatment. Both diabetogens lead to beta-cell loss, and thus lowered insulin secretion, mimicking human type 1 diabetes. The mechanism of action of such drugs is, however, not well understood. In alloxan diabetes, reactive oxygen mediates the selective necrosis of beta cells (Lenzen 2008). In STZ diabetes, DNA alkylation might mediate the toxic action of this drug, although production of reactive oxygen species may also be involved (Lenzen 2008). Since flavonoids can act as potent antioxidant molecules, some studies have focused on their effects in drug-induced diabetes.

Quercetin is a well-studied flavonoid. Its antioxidant beneficial effects are extended to protection of pancreatic beta-cell destruction by STZ (Coskun and others 2005). In this study, quercetin (15 mg/kg/d) was injected intraperitoneally 3 d prior to induction of diabetes with STZ. The effects of STZ treatment on lowering plasma insulin levels as well as in increasing the glucose levels were partly prevented by quercetin treatment. Immunohistochemical examination showed that quercetin ameliorates the degeneration of beta cells and decreases oxidative stress markers. However, in a previous study, the same flavonoid quercetin did not show modification of oxidative stress status in STZ-diabetic rats (Coldiron and others 2002). Rutin, a glycosidic form of quercetin, ameliorated the decrease in fasting insulin plasma levels induced by STZ-treatment in rats, while it had no effect in normal rats when administered at 100 mg/kg for 45 d. The same treatment also showed amelioration of STZ-induced oxidative stress markers, by lowering pancreatic concentrations of thiobarbituric acid-reactive substances (TBARS) and lipid hydroperoxides and increasing the activities of antioxidant enzymes (Stanley and Kamalakkannan 2006).

The monomer EGCG in islet cells obtained after multiple low doses of STZ-treated mice suppressed the expression of iNOS, the enzyme involved in cytokine-induced pancreatic beta-cell damage. This antioxidant effect was accompanied by a reduction of the STZ-induced hyperglycemia (Song and others 2003). On the other hand, EGCG showed also prooxidant effects in an STZ-diabetic model. When administered for 4 d at a concentration (5 mg/kg·day), which leads to nanomolar plasma levels, EGCG did not ameliorate but further reduced the lowering insulin secretion provoked by the STZ treatment. In normal rats, EGCG showed no effect on insulin secretion (Yun and others 2006).

Green tea (–)-epicatechin administered in rats at 30 mg/kg twice a day for 6 d protected them from the hyperglycemia induced by STZ, prevented pancreas morphology, and increased insulin release of isolated pancreas of such STZ-treated rats, concomitant with a reduction in nitrite production (Kim and others 2003).

The flavonoid silymarin was shown to induce pancreas recovery against alloxan-induced diabetes, both when administered during alloxan treatment or after it. The reduction in serum insulin levels induced by alloxan was inhibited by simultaneous administration with silymarin and pancreas tissue morphology was normal (Soto and others 2004). Considering previous literature, the researchers suggest that this effect could be due to the antioxidant properties of silymarin, although no markers of oxidative stress were analyzed in this study (Soto and others 2004).

In addition to pure monomers, plant extracts that contain a mixture of flavonoids have also been studied. Grape seed procyanidins are effective in preventing the hyperglycemia and hypoinsulinemia in alloxan-induced diabetic rats. The concomitant pancreatic damage induced by alloxan in terms of lipid peroxidation was also prevented by a grape seed procyanidin extract (GSPE), suggesting therefore a protective effect of GSPE against alloxan action via restoration of oxidative status in the pancreas (El-Alfy and others 2005).

Tea catechins might also be protective against induced-oxidative damage, as suggested by experiments with hamsters where catechins reduced the amount of oxidative markers (peroxides and 8-hydroxydeoxyguanosine) (Takabayashi and others 1997).

The flavonoid-rich fraction of 70% alcoholic extract of Egyptian moru alba root has also been claimed to be protective against STZ-induced damage. At a high concentration, (600 mg/kg/d, for 10 d) this fraction reduced serum glucose levels and increased serum insulin levels in STZ-diabetic rats. At the same time, it improved cellular oxidative damage, decreasing lipid peroxides (Singab and others 2005).

Finally, resveratrol is another phenolic compound with antioxidant activity (Jia and others 2008), whose antioxidant capacity could also contribute to its effects on modifying insulin secretion. Reduction of glucose plasma levels due to resveratrol treatment has been shown in different diabetic models, although the effects on insulin secretion are diverse. STZ-nicotinamide diabetic mice is an animal model with moderate hyperglycemia, reduced pancreatic insulin stores but significant glucose-stimulated insulin secretion (Masiello and others 1998). In such model, resveratrol was shown to reduce plasma insulin levels, likely due to the reduction of hyperglycemia (Su and others 2006). Thus the researchers pointed out that the hypoglycemic effects of resveratrol were not due to modulation of insulin secretion by the pancreas but due to its effects on other tissues, and, although the study did not investigate whether such effects were due to resveratrol antioxidant properties, they suggested the possibility that resveratrol changes the oxidative stress of the diabetic tissues. On the contrary, other researchers found out opposite effects on insulin plasma levels in the STZ-nicotinamide diabetic model (Chi and others 2007). In that study, reduction of glycemia was at least in part due to increased insulin secretion. The same study was also carried out in STZ-diabetic mice (where the pancreas insulin-secreting capacity is impaired) but in that case resveratrol treatment did not modify insulin levels while it reduced hyperglycemia. Thus, resveratrol showed no antioxidant effects on pancreas but enhancement of glucose uptake in periferial tissues. Resveratrol showed as well no effects on insulin secretion in STZ-diabetic rats (Chen and others 2007), in opposite to the above-mentioned effects in normal rats.

Therefore, some flavonoids might prevent beta-cell destruction in models of drug-induced diabetes (mimicking human type 1 diabetes), and this could be attributed, at least in part, to their antioxidant properties. The concrete effects, however, depend on the flavonoid and the concentration and/or experimental conditions. Since oxidative stress is also suggested to take part in type 2 diabetes, the antioxidant properties of flavonoids could be a general mechanism of preventing this pathology.

Other mechanisms of action of flavonoids on insulin secretion

Although flavonoids have been widely studied because of their antioxidant properties, they can also act through other mechanisms, by interacting with protein function, modulating intracellular cascades, and modulating gene expression (Del Bas and others 2005; Pinent and others 2006; Stevenson and Hurst 2007). Concerning the mechanisms that may lead to beta-cell dysfunction, these are not well understood. The insulin and insulin-like growth factor (IGF) signaling pathways are critical for proper maintenance and functioning of beta cells (Dhawan and others 2007). Mice models with inactivation of the insulin receptor in beta cells exhibit a selective loss of insulin secretion and a progressive impairment of glucose tolerance (Kulkarni and others 1999). Flavonoids such as procyanidins have been shown to interact with insulin signaling pathways in other cell types (Pinent and others 2004). Therefore, they might modulate beta-cell function, insulin secretion, and proliferation in part via interaction with insulin signaling cascade.

In fact, other flavonoids have been demonstrated to interfere with insulin signaling cascade in pancreas. In pancreatic cell lines and mouse islets, the insulin-secreting activity of genistein is mediated at least in part via cAMP accumulation and protein kinase A (PKA) activation (Liu and others 2006), and is independent of estrogen receptor mechanisms protein tyrosine kinases, and nitric oxide (NO)-signaling pathways.

Some flavonoids interfere with the glucose-induced depolarization of the cell membrane that initiates firing of action potentials that result on insulin secretion. Of green tea catechins, (–)-epigallocatechin-3-gallate (EGCG) and (–)-epicatechin-3-gallate (ECG), but not (–)-epicatechin and (–)-epigallocatechin, inhibit the activity of ATP-sensitive potassium KATP channels at tens of micromolar concentrations, ECG being 3 times more effective than EGCG. Further, by using cloned beta-cell-type KATP channels, these researchers showed that only EGCG at 1 μmol/L, a readily achievable plasma concentration by oral intake in humans, but not other epicatechins, significantly blocked channel reactivation after ATP washout, suggesting that interaction of phosphatidylinositol polyphosphates (PIP) with the channel was impaired by EGCG (Jin and others 2007). Previously, Baek and others (2005), also reported that the gallate-ester moiety of epicatechins may be critical for inhibiting the KATP channel activity via the pore-forming subunit Kir6.2 and this may be a possible mechanism by which green tea extracts or EGCG may cause unexpected side effects at the micromolar plasma level. Less agreement exists around resveratrol effects. Chen and others (2007) stated that resveratrol increases insulin secretion by blocking KATP and Kv channels in mouse beta-cell lines, and this could contribute to the increased plasma insulin in normal rats. Opposite effects were found by Szkudelski (2006, 2007, 2008) in pancreatic islets of normal rats, where resveratrol inhibited insulin release. The researchers found that interaction of resveratrol with KATP was not involved in such inhibitory effects. Further, the researchers suggested that the impairment of mitochondrial metabolism is involved in resveratrol based on their findings: (1) resveratrol inhibits as well insulin-secretion stimulated with the secretagogues leucine, glutamine, and succinate (2) resveratrol reduced glucose oxidation, enhanced lactate release, decreased ATP and attenuated glucose-induced hyperpolarisation of the inner mitochondrial membrane.

The above-mentioned mechanism would be in addition to the antioxidant effects described for plant compounds, which would modify the oxidative stress and/or modulate the action of reactive species that are in part responsible for the pancreatic dysfunction (Soto and others 2004; Coskun and others 2005; El-Alfy and others 2005; Yun and others 2006; Feick and others 2007).

Conclusions

The great diversity of flavonoid structures makes it difficult to establish their common effects in the pancreas. In any case, it is clear that the effects depend not only on the flavonoid structure but also on the experimental design, and remarkably so on the concentrations tested. Studies using animals with drug-induced diabetes support the hypothesis that flavonoids can ameliorate this pathogenesis. Published data suggest that there might be direct effects of flavonoids on insulin secretion, as well as preventing beta-cell apoptosis, and they could even act through modulation of proliferation. The mechanisms of action involve mainly their antioxidant properties, although other pathways might also be involved. To date, the contribution of all these flavonoid effects to the overall improvement of experimental diabetes remains obscure.

Acknowledgments

This study was supported by grant nr AGL2005-04889/ALI from the Comisión Interministerial de Ciencia y Tecnología (CICYT) of the Spanish Government. Anna Castell is the recipient of a fellowship from the autonomous government of Catalonia (Generalitat de Catalunya). Gemma Montagut is the recipient of a fellowship from the Rovira i Virgili Univ. in Tarragona. We are grateful to Montserrat Vaqué for kindly providing the flavonoid structure figures.

Ancillary