Free Radical Biology and Medicine 53 (2012) 2008–2016

Contents lists available at SciVerse ScienceDirect

Free Radical Biology and Medicine

0891-58

http://d

n Corr

E-m

journal homepage: www.elsevier.com/locate/freeradbiomed

Original Contribution

Ankaflavin: a natural novel PPARg agonist upregulates Nrf2 to attenuatemethylglyoxal-induced diabetes in vivo

Bao-Hong Lee, Wei-Hsuan Hsu, Yu-Ying Chang, Hsuan-Fu Kuo, Ya-Wen Hsu, Tzu-Ming Pan n

Department of Biochemical Science & Technology, College of Life Science, National Taiwan University, Taipei 10617, Taiwan

a r t i c l e i n f o

Article history:

Received 21 May 2012

Received in revised form

4 August 2012

Accepted 15 September 2012Available online 26 September 2012

Keywords:

Ankaflavin

Advanced glycation end products

Methylglyoxal

Peroxisome proliferator-activated

receptor-gNuclear factor erythroid-related factor 2

Free radicals

49/$ - see front matter & 2012 Elsevier Inc. A

x.doi.org/10.1016/j.freeradbiomed.2012.09.02

esponding author. Fax: þ886 2 33663838.

ail address: tmpan@ntu.edu.tw (T.-M. Pan).

a b s t r a c t

Ankaflavin (AK) is an active compound having anti-inflammatory, anti-cancer, antiatherosclerotic, and

hypolipidemic effects. We have previously reported that AK acts as an antioxidant and antidiabetic

drug; however, the mechanism by which AK prevents diabetes remains unknown. Hyperglycemia is

associated with protein glycation, which produces advanced glycation end-products (AGEs). Methyl-

glyoxal (MG)—a metabolite of carbohydrates—is believed to cause insulin resistance by inducing

inflammation and pancreas damage. In this work, diabetes was induced in Wistar rats (4 weeks of age)

by treating them with MG (600 mg/kg bw) for 4 weeks. We observed that AK (10 mg/kg bw) exerted

peroxisome proliferator-activated receptor-g (PPARg) agonist activity, thereby enhancing insulin

sensitivity (as indicated by hepatic GLUT2 translocation, PTP1B suppression, and glucose uptake) by

downregulating blood glucose and upregulating pancreatic and duodenal homeobox-1 and Maf-A

expression and increasing insulin production in MG-induced rats. However, these effects were

abolished by the administration of GW9662 (PPARg antagonist), but the expression of hepatic heme

oxygenase-1 (HO-1) and glutamate–cysteine ligase (GCL) was not suppressed in MG-induced rats.

Therefore, the nuclear factor erythroid-related factor-2 (Nrf2) activation was investigated. AK did not

affect hepatic Nrf2 mRNA or protein expression but significantly increased Nrf2 phosphorylation

(serine 40), which was accompanied by increased transcriptional activation of hepatic HO-1 and GCL.

These data indicated that AK protected rats from oxidative stress resulting from MG-induced insulin

resistance. In contrast, these effects were not detected when the rats were treated with the antidiabetic

drug rosiglitazone (10 mg/kg bw). Moreover, we found that AK did not inhibit the generation of AGEs

in vitro; however, the glutathione (GSH) levels in liver and pancreas of MG-induced rats were elevated

in rats administered AK. Therefore, we believe that GSH may lower the MG level, which attenuates the

formation of AGEs in the serum, kidney, liver, and pancreas of MG-induced rats. We also found that AK

treatment reduced the production of inflammatory factors, such as tumor necrosis factor-a and

interleukin-1b. Taken together, the results of our mechanistic study of MG-induced rats suggest that

the protective effects of AK against diabetes are mediated by the upregulation of the signaling pathway

of Nrf2, which enhances antioxidant activity and serves as a PPARg agonist to enhance insulin

sensitivity.

& 2012 Elsevier Inc. All rights reserved.

Hyperglycemia is associated with protein glycation, andadvanced glycation end-products (AGEs) are generated by none-nzymatic reactions between carbohydrates and proteins [1–3].AGEs have a propensity to generate free radicals and undergoautoxidation to generate other reactive intermediates, therebyinducing diabetes [4]. Methylglyoxal (MG) is also known as2-oxopropanal, pyruvaldehyde, pyruvic aldehyde, 2-ketopropio-naldehyde, acetylformaldehyde, propanedione, or propionalde-hyde. MG is a highly reactive dicarbonyl metabolite producedduring glucose metabolism [5], and it is a major precursor of AGEs

ll rights reserved.

5

that are involved in the pathogenesis of diabetes and inflamma-tion. Studies have suggested that AGEs and MG can generate largeamounts of proinflammatory cytokines; these findings are report-edly related to the modulation of inflammatory cytokines throughoxidative stress [6–10]. Oxidative stress is increased duringdiabetes and hyperinsulinemia; reactive oxygen species havebeen reported to be generated as a result of hyperglycemia, whichcauses many of the secondary complications of diabetes [5,8,10].

Recent studies show that acute MG administration to Sprague–Dawley rats causes glucose intolerance and reduces adiposetissue insulin-stimulated glucose uptake (MG 50 mg/kg bw; iv)and results in pancreatic dysfunction (MG 60 mg/kg bw; infusion)[11,12]. However, the in vitro studies have not establishedwhether MG is the cause or an effect of diabetes [12], although

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–2016 2009

MG invariably induces inflammation by the formation of AGEsand activation of the receptor for AGEs (RAGE) [8–10].

Under physiological conditions, glyoxalase converts MG toD-lactate in a reaction dependent upon reduction by glutathione.MG is also detoxified by the conversion to D-lactoylglutathioneand D-lactate, which is catalyzed in the cytosol of all cells byglyoxalases I and II, suggesting that glyoxalase I completelyinhibits the hyperglycemia-induced formation of AGEs; thus, therole of MG in the formation of AGEs is vital [13]. In healthyhumans, plasma levels of MG are r1 mM; however, these levelsare elevated two- to fourfold in patients with diabetes [14].

MG (60 mg/kg bw, infusion) efficiently induces diabetes in SDrats [12]. The administration of MG by gavage is consideredsuitable for inducing chronic diabetes and serves as a method toassess the anti-inflammatory activity of test drugs. However, toour knowledge, the induction of inflammation by orally adminis-tered MG has not been studied. Therefore, in this study, weattempted to induce diabetes through oral administration of MG(600 mg/kg bw). Further, the ability of various antioxidants, suchas quercetin and phenolic acid, to activate nuclear factorerythroid-related factor 2 (Nrf2) and attenuate oxidative damagehas been evaluated in previous studies [15,16]; the antioxidantsilymarin was found to inhibit the generation of AGEs, therebyimproving the symptoms of diabetes [17].

Rosiglitazone (Rosi)—a peroxisome proliferator-activatedreceptor-g (PPARg) ligand—belongs to the thiazolidinedione(TZD) class of antidiabetic drugs. The activation of PPARg isknown to affect pancreatic b-cell function and insulin production[18]. However, TZD use is limited by side effects that aremediated through ectopic activation of PPARg, including weightgain, edema, and increased risk of fractures. These side effects ofTZDs have limited their potential application for a variety ofinflammatory and obesity-related metabolic diseases. The futurewidespread use of TZDs as insulin sensitizers and for treatment ofother metabolic and inflammatory diseases clearly requires abreakthrough that would allow better therapeutic profiles [19].

Monascus species have been used as traditional food fungi ineastern Asia for several centuries. Monascus-fermented rice wasgradually developed as a popular functional food for hypolipide-mia. Ankaflavin (AK) is a water-insoluble polyketide metaboliteisolated and identified from Monascus. This yellow pigment hasan azaphilonoid structure. Recently, it was proven to be thefunctional ingredient for anti-inflammatory actions [20]. AK hasbeen reported to downregulate hyperlipidemia and lipid perox-idation; and no toxicity or side effects were found in our previousstudy [21]. However, the interactions of AK with other com-pounds are as yet unknown. In this study, we found that AK actsas a novel natural PPARg agonist. However, the influence of AK onPPARg activity involved in the prevention of MG-induced diabetesby PPARg agonists in vivo remained unclear.

Experimental procedures

Chemicals

Glucose, MG, insulin, rosiglitazone, and lipopolysaccharide(LPS) were purchased from Sigma–Aldrich (St. Louis, MO, USA).2-[N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino]-2-deoxy-D-glu-cose (2-NBDG) was from Invitrogen (Carlsbad, CA, USA). PPARgantagonist (GW9662) and glutathione (GSH) ELISA kit werepurchased from Cayman Chemical (Ann Arbor, MI, USA). Anti-glucose transporter 2 (GLUT2) antibody, anti-protein-tyrosinephosphatase 1B (PTP1B) antibody, anti-pancreatic and duodenalhomeobox-1 (PDX-1) antibody, anti-Maf-A antibody, and anti-CCAAT/enhancer binding protein b (C/EBPb) antibody for rat were

purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).The anti-p-Nrf2, anti-Nrf2, and anti-heme oxygenase-1 (HO-1)antibodies for rat were purchased from Bioss (Woburn, MA, USA).Anti-glutamate–cysteine ligase (GCL) antibody for rat was pur-chased from Epitomics (Burlingame, CA, USA). Fetal bovine serumwas purchased from Hyclone (Logan, UT, USA). RPMI 1640medium, L-glutamine, sodium pyruvate, and antibiotics (penicil-lin/streptomycin) were purchased from Gibco (Grand Island, NY,USA). Preparation of AK (495% purity) was identified by nuclearmagnetic resonance (Varian Gemini, 200 MHz, FT-NMR; Varian,Palo Alto, CA, USA) and electrospray ionization–mass spectro-metry (Thermo Electron, Waltham, MA, USA) analysis.

Agonist activity of AK for PPARg

The PPARg agonist activity of AK was confirmed by LanthaScreenTR-FRET PPARg coactivator assay kit (Invitrogen). The assay wascarried out following the normative manual.

Animals and diabetes induction

Male Wistar rats (4 weeks of age) were obtained from theNational Laboratory Animal Breeding and Research Center (Taipei,Taiwan). Animals were acclimatized for 1 week before use; theywere divided at random into six treatment groups (six rats pergroup) and provided with food and water ad libitum. Animalswere subjected to a 12-h light/dark cycle with a maintainedrelative humidity of 60% and a temperature at 25 1C (protocolcomplied with guidelines described in the Animal Protection Law,

amended on 17 January 2001, Hua-Zong-(1)-Yi-Tzi-9000007530,Council of Agriculture, Executive Yuan, Taiwan, ROC). Rats weredivided into the following treatment groups: (1) control (salineadministration), (2) MG (600 mg/kg bw), (3) MG þ rosiglitazone(a synthetic PPARg agonist; 10 mg/kg bw), (4) MG þ rosiglita-zone þ GW9662 (a synthetic PPARg antagonist; 10 mg/kg bw),(5) MG þ ankaflavin (a novel natural PPARg agonist; 10 mg/kgbw), and (6) MGþankaflavinþGW9662. MG, Rosi, and AK wereadministered to Wistar rats by oral administration for 28 days,and GW9662 was administered by intraperitoneal (ip) injectionfor 28 days.

Oral glucose tolerance test (OGTT) and insulin tolerance test (ITT)

OGTT and ITT were performed at day 24 and day 26. Afterovernight fasting, an OGTT was performed. Briefly, rats wereanesthetized, a basal blood sample was collected, and an oralglucose load (2 g/kg bw) was given by oral administration.Subsequently, blood was collected (0–120 min), and the bloodglucose levels were determined using a glucose assay kit (BioAs-say Systems, Hayward, CA, USA) and insulin levels were measuredwith a rat insulin ELISA kit (Mercodia, Winston Salem, NC, USA).For ITT, rats were given an ip injection of insulin solution(0.5 U/kg bw) after 4 h of fasting. Blood glucose level was determined.

Glucose uptake of hepatic cells

Hepatic cell glucose uptake was performed according a pre-vious study [22]. The 2-NBDG was chosen as a glucose uptakeindicator.

Isolation of hepatic plasma membrane for GLUT2

Plasma membrane extracts of liver were isolated using anisolation kit (BioVision, Mountain View, CA, USA).

Fig. 1. The structure of AK, which is a polyketide metabolite from Monascus and

has an azaphilonoid structure.

Fig. 2. Agonist activity of AK for PPARg. The LanthaScreen TR-FRET PPARgcoactivator assay kit provides a sensitive and robust method for high-throughput

screening of potential PPARg ligands as agonists or antagonists of ligand-

dependent coactivator recruitment. The kit uses a terbium (Tb)-labeled anti-

glutathione S-transferase (GST) antibody, a fluorescein-labeled coactivator

peptide, and a PPARg ligand-binding domain (LBD). To run the LanthaScreen

TR-FRET PPARg coactivator assay in agonist mode (to identify agonist compounds),

the PPARg LBD is added to ligand test compounds followed by addition of a

mixture of the fluorescein-coactivator peptide and Tb–anti-GST antibody. After an

incubation period at room temperature, the TR-FRET 520:495 emission ratio is

calculated. Data are shown as means7SEM (n¼4).

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–20162010

AGE, TNF-a, IL-1b, and pancreatic insulin assays

An AGE ELISA kit (New England Biolabs, Beverly, MA, USA),TNF-a ELISA kit (eBioscience, San Diego, CA, USA), and IL-1b ELISAkit (Peprotech, Rocky Hill, NJ, USA) were purchased to measurethe levels in tissues and serum. The insulin level of pancreas wasassayed using a rat insulin ELISA kit (Mercodia).

Peripheral blood mononuclear cell (PBMC) culture

Blood was collected and red blood cells (RBCs) were lysed withRBC buffer; subsequently, the remaining cells were treated withLPS (2 mg/ml) in RPMI 1640 medium supplemented with 10% fetalbovine serum, 2 mM L-glutamine, 1 mM sodium pyruvate, andantibiotics (100 U/ml penicillin and 100 mg/ml streptomycin) for24 h, and the inflammatory factors were measured by ELISA. TheRAGE expression without treatment with LPS was investigated byreal-time PCR.

Real-time PCR

Total RNA was isolated using Trizol (Life Technologies, Carlsbad,CA, USA) according to the manufacturer’s instructions. cDNA from3 mg of RNA was generated using the SuperScript III First-StrandSynthesis System for RT-PCR (Life Technologies) according to themanufacturer’s instructions. The reverse-transcription product wasdiluted in water and a volume corresponding to 30 ng of original RNAwas used for real-time PCR. Real-time PCR amplification and detec-tion were performed using the SYBR Green qPCR SuperMix-UDG withROX (Life Technologies) in a fluorescence thermal cycler (StepOnereal-time PCR system; Life Technologies) according to the manufac-turer’s protocol. The primers used were as follows: glyoxalase I [23]forward, 50-ATTTGGCCACATTGGGATTGC-30, and reverse, 50-TTCAATC-CAGTAGCCATCAGG-30; Nrf2 [24] forward, 50-GAGACGGCCATGACT-GAT-30, and reverse, 50-GTGAGGGGATCGATGAGTAA-30; PPARg [25]forward, 50-CCCTGGCAAAGCATTTGTAT-30, and reverse, 50-ACTGG-CACCCTTGAAAAATG-30; RAGE [26] forward, 50-CTACCTATTCCTG-CAGCTTC-30, and reverse, 50-CTGATGTTGACAGGAGGGCTTTCC-30;b-actin forward, 50-CTTTCTACAATGAGCTGCGTG-30, and reverse,50-TCATGAGGTAGTCTGTCAGG-30 [27].

Western immunoblotting

Pancreas and liver tissues were homogenized in RIPA buffer(Cell Signaling Technology, Beverly, MA, USA) using a homogeni-zer. Protein was resolved by 10% SDS–PAGE and transferred to apolyvinylidene difluoride membrane. The membranes wereblocked with 5% nonfat dry milk solution for 1 h and incubatedovernight with primary antibodies for GLUT2 (plasma mem-brane), PDX-1, Maf-A, and C/EBPb for 4 h; subsequently, themembrane was washed three times each for 5 min inphosphate-buffered saline–Tween 20 (PBST) and shaken in asolution of horseradish peroxidase-linked secondary antibodyfor 1 h and washed three more times each for 5 min in PBST.The expression of proteins was detected by enhanced chemilu-minescence reagent (Millipore, Billerica, MA, USA).

Statistical analysis

Experimental results were averaged by triplicate analysis.The data were recorded as means7SD and analysis was doneusing the Statistical Analysis System (SAS Inc., Cary, NC, USA).One-way analysis of variance was performed using ANOVAprocedures. Significant differences between means were deter-mined by Duncan’s multiple range tests. Results were consideredstatistically significant at po0.05.

Results

The PPARg agonist activity of AK

AK was isolated from Monascus-fermented rice—a traditionalChinese medicine; its structure is shown in Fig. 1. We found thatAK exerted PPARg agonist activity greater than that exerted bypioglitazone, but weaker than that exerted by Rosi (Fig. 2).Therefore, we investigated the ability of AK to regulate bloodglucose levels.

The antidiabetic activity of AK in MG-induced rats

The level of serum insulin was markedly decreased in MG(600 mg/kg bw)-induced rats compared to the controls. However,the serum insulin level was increased by the administration ofAK; this difference was not statistically significant. Further,administration of GW9662—a PPARg antagonist—effectivelyinhibited insulin secretion in rats treated with AKþGW9662and RosiþGW9662 (Fig. 3A). The hypoglycemic activities of AKand Rosi were abolished by the administration of GW9662(Fig. 3B), suggesting that AK and Rosi both improved bloodglucose level in MG-induced rats.

The levels of fasting blood glucose in MG-induced rats, asassessed by the OGTT (glucose, 2 g/kg bw), were measured on day

Fig. 3. The (A) serum insulin and (B) blood glucose levels of rats were measured at day 28 after MG induction. (C) The OGTT (0–120 min) of MG-induced rats was carried

out by measuring blood glucose after administration of glucose (2 g/kg bw), and (D) the serum insulin was also quantified using an insulin ELISA kit at each time point

(0–120 min) at day 24. (E) Moreover, the attenuation of insulin resistance by AK treatment was evaluated by ITT assay. Briefly, the MG-induced rats were administered

insulin (0.5 U/kg bw) after 4 h of fasting, and the blood glucose was determined at day 26. MG, AK, and Rosi were administered orally for 28 days and GW9662 was

administered by ip injection for 28 days. MG, methylglyoxal; AK, ankaflavin; Rosi, rosiglitazone; GW, GW9662. Data are shown as means7SD (n¼6). Groups with different

letters are significantly different (po0.05).

Fig. 4. (A) The translocation of hepatic GLUT2 from cytosol to membrane and PTP1B expression were investigated by Western blot. (B) The promotion of AK for insulin

sensitivity was evaluated; glucose uptake of liver cells was measured by flow cytometry using 2-NBDG as a fluorescent indicator (n¼6). (C) Transcription factor for

antioxidant enzyme expression was determined by hepatic Nrf2 mRNA level (real-time PCR) (n¼6). (D) The activation of Nrf2 by phosphorylation and GCL and HO-1

expression was confirmed by Western blot. (E) The hepatic glyoxalase mRNA, a downstream target of Nrf2, was determined in MG-induced rats by real-time PCR (n¼6).

(F) The hepatic PPARg mRNA was determined in MG-induced rats by real-time PCR (n¼6). (G) The GSH level in liver of MG-induced rats was quantified (n¼6). MG, AK, and

Rosi were administered orally for 28 days and GW9662 was administered by intraperitoneal injection for 28 days. MG, methylglyoxal; AK, ankaflavin; Rosi, rosiglitazone.

Groups with different letters are significantly different (po0.05). Data are shown as means7SD.

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–2016 2011

24 after overnight fasting and indicated that AK and Rosiboth prevented MG-induced hyperglycemia compared with theMG treatment group; however, AK significantly lowered blood

glucose compared with the Rosi group after administration ofglucose for 60 min. The administration of GW9662 abolishedthis antidiabetic activity in both AK and Rosi treatment groups

Fig. 5. The level of pancreatic insulin was impaired by MG induction and AK, and

the preventive mechanism of AK for pancreas damage was investigated. (A) The

insulin production of the pancreas was quantified by insulin ELISA (n¼6). (B) The

expression of pancreatic C/EBPb, Maf-A, and PDX-1 in MG-induced rats was

confirmed by Western blot. (C) The GSH level in pancreas of MG-induced rats was

quantified (n¼6). MG, AK, and Rosi were administered orally for 28 days and

GW9662 was administered by intraperitoneal injection for 28 days. MG, methyl-

glyoxal; AK, ankaflavin; Rosi, rosiglitazone. Groups with different letters are

significantly different (po0.05). Data are shown as means7SD.

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–20162012

(Fig. 3C). We also determined the levels of fasting serum insulin inMG-induced rats at various times using the OGTT assay. Theresults indicated a hyperinsulinemic effect in the MG andMGþRosiþGW treatment groups (Fig. 3D). However, AK andRosi both lowered serum insulin levels in MG-induced rats anddownregulated blood glucose levels.

Twenty-six days after MG treatment, the ITT was administered(Fig. 3E). MG-induced rats were given an ip injection of insulin(0.5 U/kg bw) after 4 h of fasting. We found that AK and Rosi bothsignificantly lowered the level of blood glucose 15 min after insulinadministration, suggesting that AK and Rosi decreased insulinresistance in MG-induced rats. Moreover, we also found thatGW9662 attenuated the hypoglycemic activity in the MGþRo-siþGW treatment group and that this effect was not detected inthe MGþAKþGW group. These findings indicate that the effect ofAK as an antidiabetic may depend not only on PPARg activity.

AK improves liver function of MG-induced rats

MG treatment markedly decreased the level of hepatic totaland membrane GLUT2 and increased hepatic PTP1B expression(Fig. 4A and Supplementary Fig. S1). However, AK and Rosi bothrestored hepatic GLUT2 expression and translocation to themembrane and inhibited PTP1B to enhance insulin sensitivity inMG-induced rats. GW9662 treatment abolished the translocationof GLUT2 and suppression of PTP1B expression. Thus, these dataindicate that the effects of AK and Rosi improved insulin sensi-tivity. Similarly, the capacity of liver cells to take up the glucoseanalogue 2-NBDG was also inhibited by MG treatment for 28days; however, treatment with AK and Rosi significantlyenhanced the ability of cells to take up 2-NBDG (Fig. 4B).

In contrast, MG, AK, and Rosi did not affect hepatic Nrf2 mRNAlevels (Fig. 4C) or protein expression (Fig. 4D; Supplementary Fig.S2A). However, we observed that AK induced the phosphorylationof Nrf2 in the liver of MG-induced rats (Fig. 4D; SupplementaryFig. S2B). In addition, the elevation of Nrf2 phosphorylation wasnot observed in the Rosi treatment group. Thus, the expression ofhepatic HO-1 and GCL (antioxidant enzyme) was evaluated byimmunoblotting. There was no detectable effect on GCL expres-sion in AK-treated (with or without GW9662) or Rosi-treated(with or without GW9662) rats (Fig. 4D; Supplementary Fig. S3A);however, AK markedly increased HO-1 expression in the liver ofMG-induced rats (Fig. 4D; Supplementary Fig. S3B).

We also investigated the mRNA levels of the MG scavengerhepatic glyoxalase by p-Nrf2 regulation. As shown in Fig. 4E,AK with or without GW9662 significantly increased glyoxalasemRNA levels. However, this was not detected in the Rosi treat-ment group. PPARg possesses anti-inflammatory activity, andhepatic PPARg expression was suppressed by MG induction; inaddition, we found that AK and Rosi both prevented a decrease inhepatic PPARg mRNA in response to MG (Fig. 4F). These resultsindicated that AK treatment increased the levels of Nrf2 phos-phorylation and promoted downstream signaling, including HO-1and glyoxalase. Further, we detected an increase in hepatic GSHlevels in MG-induced rats treated with AK (Fig. 4G).

AK improves pancreatic function in MG-induced rats

The production of pancreatic insulin was decreased by MGinduction; similarly, GW9662 treatment also inhibited insulinproduction. However, we found that AK and Rosi both restoredthe level of pancreatic insulin (Fig. 5A). We next investigated thepotential mechanism responsible for insulin production in MG-induced rats treated with AK and Rosi. The results suggested thatthe pancreatic level of C/EBPb in MG-induced rats was increasedbut was attenuated significantly by AK and Rosi treatments. The

pancreatic levels of Maf-A and PDX-1 were increased in MG-induced rats treated with AK and Rosi; however, GW9662 treat-ment abolished this increase in pancreatic Maf-A and PDX-1levels (Fig. 5B; Supplementary Fig. S4). These data indicated thatthe increase in Maf-A and PDX-1 expression was dependent uponPPARg activation by treatment with AK and Rosi.

In addition, we found that oxidative stress was evident inthe pancreases of MG-induced rats as reflected by GSH levels.As shown in Fig. 5C, AK significantly elevated pancreatic GSHlevels to a greater extent than Rosi. These results indicated thatAK and Rosi could prevent pancreatic damage caused by oxidativestress, thereby resulting in elevations in Maf-A and PDX-1expression and increasing insulin production.

Inhibition of AGEs by AK treatment

MG induction markedly increased the levels of AGEs in serum,liver, and pancreas (Fig. 6). AK treatment inhibited this effect; this

Fig. 6. The glycation production was measured by assaying AGE levels in (A) serum, (B) liver, and (C) pancreas of MG-induced rats with the AGE ELISA kit (n¼6). MG, AK,

and Rosi were administered orally for 28 days and GW9662 was administered by intraperitoneal injection for 28 days. MG, methylglyoxal; AK, ankaflavin; Rosi,

rosiglitazone. Groups with different letters are significantly different (po0.05). Data are shown as means7SD.

Fig. 7. (A) The production of inflammatory factors from culture medium of PBMCs

incubated with LPS for 24 h was quantified by IL-1b and TNF-a ELISA. (B) RAGE

expression without LPS treatment was investigated by real-time PCR. MG, AK, and

Rosi were administered orally for 28 days and GW9662 was administered by

intraperitoneal injection for 28 days. MG, methylglyoxal; AK, ankaflavin; Rosi,

rosiglitazone. Groups with different letters are significantly different (po0.05).

Data are shown as means7SD (n¼6).

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–2016 2013

inhibition was greater than that observed when Rosi was admi-nistered. However, whether the suppression of AGE generation byAK treatment resulted from MG metabolism or inhibition of AGEsynthesis remains unknown. Therefore, we investigated theinhibitory activity of AK on the formation of AGEs using anin vitro assay. Supplementary Fig. S5 shows that AK did notinhibit the formation of AGEs. These results indicated thatdecreased levels of AGEs in serum, liver, and pancreas resultingfrom AK treatment were dependent upon MG metabolism; more-over, the degradation of MG was probably due to downstreamsignaling of Nrf2, including glyoxalase and GSH.

Anti-inflammatory activity of AK

AGEs activate RAGE, thereby elevating the production ofinflammatory cytokines and inducing insulin resistance. As indi-cated above, AK enhanced the metabolism of MG and lowered thelevels of AGEs. In accord with these findings, AK significantlysuppressed RAGE expression in PBMCs and decreased IL-1b andTNF-a production when PBMCs were stimulated with LPS for 24 h(Fig. 7). Therefore, we hypothesized that AK suppressed inflam-mation, which then improved insulin sensitivity in MG-inducedrats. Moreover, the IL-1b and TNF-a levels in the kidneys, liver,pancreas, and serum of MG-induced rats were lowered signifi-cantly by AK and Rosi treatment (Fig. 8).

Discussion

It is well documented that glycation of proteins contributes tothe pathology of a number of chronic diseases such as diabetes.Considerable effort has been focused on identifying clinicallyuseful inhibitors of protein AGEs to delay or prevent glycationand alleviate insulin resistance [5,7]; candidates include silymarin[17], alagebrium [11], quercetin [15], and rutin [28]. An in vitromodel of an anti-protein glycation assay has been used to identifyinhibitors of glycation. Glyoxal and MG are glycating agents thatgenerate AGEs and glucose [29,30]. In this study, we chose MG toinduce insulin resistance in rats through oral administration.

Fig. 8. The levels of IL-1b in (A) kidney, (B) liver, (C) pancreas, and (D) serum of MG-induced rats were measured by ELISA and the levels of TNF-a in (E) kidney, (F) liver,

(G) pancreas, and (H) serum of MG-induced rats were measured by ELISA. MG, AK, and Rosi were administered orally for 28 days and GW9662 was administered by

intraperitoneal injection for 28 days. MG, methylglyoxal; AK, ankaflavin; Rosi, rosiglitazone. Groups with different letters are significantly different (po0.05). Data are

shown as means7SD (n¼6).

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–20162014

Recently, we described the antioxidative and anti-inflammatoryactivities of secondary metabolites isolated from Monascus-fer-mented products [31], and another study also supported theseactivities [32]. Furthermore, we have previously reported thatthese products could act as adjuvants for treating diabetes,reducing blood glucose, improving insulin resistance, and pre-venting pancreatic damage in streptozotocin-induced rats [33,34].Therefore, the antidiabetic activity of AK—a major anti-inflammatory and antioxidative agent present in Monascus-fer-mented products—was evaluated in this study.

Several PPARg agonists such as TZD (Rosi) have been reportedto improve insulin sensitivity in diabetic patients [35] and exertanti-inflammatory activity [36]. In this study, we found that AKwas similar to Rosi with respect to its PPARg agonist activity, andits EC50 was stronger than that of Rosi (Fig. 2). AGEs and MG cangenerate large amounts of proinflammatory cytokines [8–10].In addition, inflammatory factors involved in insulin resistancehave been reported in previous studies [37]. In the present study,we found that AK improved OGTT (Fig. 3B) and ITT (Fig. 3D)results by activating PPARg and promoting insulin sensitivity inMG-induced rats. In addition, AK also inhibited RAGE expressionin the PBMCs of MG-treated rats (Fig. 7) and the productionof inflammatory cytokines (Fig. 8), which led to attenuation ofinsulin resistance and elevation of glucose uptake (Fig. 4B).GW9662 abolished the insulin sensitivity-enhancing effects ofAK and Rosi; however, GW9662 was unable to completely inhibitthe regulation of blood glucose levels and anti-inflammatoryeffects of AK. These results suggest that a different mechanismwas responsible for the antidiabetic effects of AK.

The abnormal accumulation of MG in cells is known to occur inpatients with diabetes. MG decreases glucose tolerance in rodents[11,38], suggesting that postprandial MG production in normo-glycemic individuals could result in glucose intolerance andconsequently greater accumulation of MG. In addition, fructosemetabolism generates excess MG and aldehydes in the liver [39].These metabolites of fructose could be detoxified by the sequen-tial action of two thiol-dependent enzymes: (a) glyoxalase I,which catalyzes the formation of S-D-lactoyl glutathione from2-oxoaldehydes and GSH, and (b) glyoxalase II, which hydrolyzesthe thiol esters to produce D-lactate and GSH [5–10,13,40].Previous studies have reported that HO-1 is suppressed bytreatment with AGEs in vitro [5,7,17], and glyoxalase and HO-1

are upregulated by Nrf2 [15,16,41,42]. Therefore, in this study, wedetermined the alterations in the glyoxalase and antioxidantsystems by AK treatment in MG-induced rats.

Thus, we found that the attenuation of inflammation by AKresulted from decreased levels of MG. This depended on the levelsof glyoxalase, HO-1, and GSH, which were upregulated by hepaticp-Nrf2 (Fig. 4). The ultimate outcome of these events was areduction in the formation of AGEs (Fig. 6). The HO-1 expression,Nrf2 activation, GSH elevation, and detoxifying effects of AKagainst MG-induced damage were not observed in rats treatedwith Rosi. Moreover, Nrf2 phosphorylation, glyoxalase expression,and HO-1 expression were not inhibited by GW9662. Thesefindings suggest that AK can metabolize MG, thus reducinginflammation and blood glucose level and exerting PPARg agonistactivity (Fig. 9).

Phenolic compounds and antioxidants promote Nrf2 activa-tion, thereby reducing oxidative damage and insulin resistancein vitro and in vivo [15,16,43–45]. Nrf2 may play an importantrole in improving glucose tolerance and insulin resistance in Nrf2-knockout mice fed a high-fat diet for 180 days [46]. Our findingsled us to suggest that AK may activate Nrf2, thus avoidinginflammatory cytokine-induced insulin resistance, even thoughAK was found to act as a PPARg agonist in our study. Muchevidence indicates that antioxidants, such as silymarin [17],g-tocopherol [47], and other phytochemicals [48], may inhibitthe formation of AGEs or lower MG levels. The data presentedhere indicate that AK reduced MG levels in MG-induced rats bythe action of glyoxalase but did not exert antiglycative activity(Supplementary Fig. S5).

Both AK and Rosi treatments elevated pancreatic insulinproduction in MG-induced rats; however, this activity was abol-ished by treatment with GW9662 (PPARg antagonist), suggestingthat the elevation of insulin production by AK and Rosi adminis-tration depended on the activation of PPARg. To investigate howMG reduced the levels of pancreatic insulin, we investigatedinsulin biosynthesis. Maf-A and PDX-1 are positive regulatorsfor insulin synthesis, whereas C/EBPb is a negative regulator ofinsulin gene transcription. Inactivation of Maf-A and PDX-1 inpancreatic b cells results in dysfunctional insulin expression andcauses diabetes [49,50]. PPARg is required for Maf-A and PDX-1transcription and inhibits C/EBPb activity [51]. Taking theseresults altogether, we conclude that AK and Rosi promoted PPARg

Fig. 9. The protective mechanisms of AK against MG-induced inflammation and

diabetes.

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–2016 2015

activation and increased Maf-A and PDX-1 expression, therebyelevating pancreatic insulin synthesis (Figs. 5 and 9).

In contrast, GSH and other antioxidants play a central role inthe degradation of MG and prevention of injury to the pancreas.Oxidative stress resulting in decreased insulin secretion has beeninvestigated [52]. The antioxidant quercetin potentiates insulinsecretion and protects INS-1 pancreatic b cells against oxidativedamage [53]. In the present study, we found that the level of GSHwas reduced in the pancreas of MG-induced rats, and AK admin-istration significantly restored pancreatic GSH expression (Fig. 5).Thus, we suggest that AK stimulated MG metabolism in the liverthrough glyoxalase activity, consequently reducing MG and AGElevels in the pancreas of MG-induced rats.

Thus, we have confirmed the antidiabetic benefits of AK inMG-induced rats. We conclude that AK activated both Nrf2 andPPARg to enhance the scavenging of MG, thereby reducing thelevels of AGEs and promoting insulin sensitivity by inhibitinginflammation. However, several issues remain to be confirmed asfollows. (a) The clearance of AGEs: whether macrophages, Kupffercells, or hepatic sinusoidal endothelial cells phagocytize glycativesoluble macromolecules is a matter of much debate [54]; how-ever, we are unable to confirm whether AK promoted the abilityof Kupffer cells or macrophages to take up and clear AGEs inMG-induced rats. (b) The effect of AK on PPARa activationis unclear. One study has shown that some natural PPARgagonists also activate PPARa; they are called selective peroxisomeproliferator-activated receptor-g modulators (SPPARMs). TheseSPPARMs exert greater hypoglycemic and hypolipidemic abilitiesand are safer when used in diabetic patients than TZD synthesisby chemical strategy [55]. We found that AK is a natural PPARgagonist identified from Monascus-fermented products; our resultsindicated that AK exerted OGTT improvements and insulinsensitivity in MG-induced rats better than Rosi; however, it isunclear whether AK is a SPPARM. On the other hand, AK exertsanti-cancer [56], anti-inflammatory [57], and antihyperlipidemic[58] activities. (c) The potential mechanisms of Nrf2 phosphor-ylation by antioxidants have been reported as a function ofsignaling by p38 mitogen-activated kinase (MAPK), protein kinaseC, and extracellular signal-regulated kinases [59–61]. In thisstudy, AK enhanced Nrf2 phosphorylation, resulting in beneficialeffects in MG-induced rats; however, the kinase(s) activated by

AK remains to be identified. (d) Inflammatory factors may induceMAPKs. These kinases, in turn, inhibit PPARg function by directphosphorylation of its serine residues, which affects DNA-bindingactivity and results in its degradation by the ubiquitin–proteasome-dependent pathway [62]. However, TZDs can bindto PPARg and change its structure, thereby preventing its phos-phorylation. This suggests that TZDs enhance insulin sensitivityby preventing PPARg degradation [62]. However, whether thisprevents the phosphorylation of PPARg by altering its structure isnot known.

Conclusion

In conclusion our experimental data demonstrate that AKeffectively reduces AGE levels in serum, liver, and pancreas ofMG-induced rats. This inhibitory effect is dependent on theactions of hepatic glyoxalase and HO-1 by Nrf2 regulation. More-over, AK promotes insulin production and GSH levels, therebyattenuating insulin resistance caused by inflammation. Ourresults also suggest that AK may activate PPARg to promoteinsulin sensitivity in MG-induced rats (oral administration with600 mg/kg bw; Fig. 9). AK may potentially be supplied in foodsupplements, or developed as a ‘‘functional food,’’ to minimize thedevelopment of diabetes.

Acknowledgment

This research work and subsidiary spending were supportedby the National Science Council (Taiwan) (NSC 99-2628-B-002-004-MY2).

Appendix A. Supplementary material

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.freeradbiomed.2012.09.025.

References

[1] Day, J. F.; Thorpe, S. R.; Baynes, J. W. Nonenzymatically glucosylated albumin:in vitro preparation and isolation from normal human serum. J. Biol. Chem.254:595–597; 1979.

[2] Negre-Salvayre, A.; Salvayre, R.; Auge, N.; Pamplona, R.; Portero-Otin, M.Hyperglycemia and glycation in diabetic complications. Antioxid. RedoxSignaling 11:3071–3109; 2009.

[3] Schiekofer, S.; Andrassy, M.; Chen, J.; Rudofsky, G.; Schneider, J.; Wendt, T.;Stefan, N.; Humpert, P.; Fritsche, A.; Stumvoll, M.; Schleicher, E.; Haring, H. U.;Nawroth, P. P.; Bierhaus, A. Acute hyperglycemia causes intracellular forma-tion of CML and activation of ras p42/44 MAPK and nuclear factor kappaB inPBMCs. Diabetes 52:621–633; 2003.

[4] Bonnefont-Rousselot, D. Glucose and reactive oxygen species. Curr. Opin. Clin.Nutr. Metab. Care 5:561–568; 2002.

[5] Wu, C. H.; Huang, S. M.; Lin, J. A.; Yen, G. C. Inhibition of advanced glycationendproduct formation by foodstuffs. Food Funct. 2:224–234; 2011.

[6] Huang, S. M.; Wu, C. H.; Yen, G. C. Effects of flavonoids on the expression ofthe pro-inflammatory response in human monocytes induced by ligation ofthe receptor for AGEs. Mol. Nutr. Food Res. 50:1129–1139; 2006.

[7] Lin, J. A.; Fang, S. C.; Wu, C. H.; Huang, S. M.; Yen, G. C. Anti-inflammatoryeffect of the 5,7,40-trihydroxy-6-geranyl flavanone isolated from the fruit ofArtocarpus communis in S100B-induced human monocytes. J. Agric. FoodChem. 59:105–111; 2011.

[8] Guha, M.; Bai, W.; Nadler, J. L.; Natarajan, R. Molecular mechanisms of tumornecrosis factor alpha gene expression in monocytic cells via hyperglycemia-induced oxidant stress-dependent and -independent pathways. J. Biol. Chem.275:17728–17739; 2000.

[9] Shanmugam, N.; Kim, Y. S.; Lanting, L.; Natarajan, R. Regulation ofcyclooxygenase-2 expression in monocytes by ligation of the receptor foradvanced glycation end products. J. Biol. Chem. 278:34834–34844; 2003.

B.-H. Lee et al. / Free Radical Biology and Medicine 53 (2012) 2008–20162016

[10] Shanmugam, N.; Reddy, M. A.; Guha, M.; Natarajan, R. High glucose-inducedexpression of proinflammatory cytokine and chemokine genes in monocyticcells. Diabetes 52:1256–1264; 2003.

[11] Dhar, A.; Desai, K. M.; Wu, L. Alagebrium attenuates acute methylglyoxal-induced glucose intolerance in Sprague–Dawley rats. Br. J. Pharmacol.159:166–175; 2010.

[12] Dhar, A.; Dhar, I.; Jiang, B.; Desai, K. M.; Wu, L. Chronic methylglyoxalinfusion by minipump causes pancreatic beta-cell dysfunction and inducestype 2 diabetes in Sprague–Dawley rats. Diabetes 60:899–908; 2011.

[13] Desai, K.; Wu, L. Methylglyoxal and advanced glycation endproducts: newtherapeutic horizons? Recent Patents Cardiovasc Drug Discovery 2:89–99; 2007.

[14] Wang, H.; Meng, Q. H.; Gordon, J. R.; Khandwala, H.; Wu, L. Proinflammatoryand proapoptotic effects of methylglyoxal on neutrophils from patients withtype 2 diabetes mellitus. Clin. Biochem. 40:1232–1239; 2007.

[15] Weng, C. J.; Chen, M. J.; Yeh, C. T.; Yen, G. C. Hepatoprotection of quercetinagainst oxidative stress by induction of metallothionein expression throughactivating MAPK and PI3K pathways and enhancing Nrf2 DNA-bindingactivity. New Biotechnol. 28:767–777; 2011.

[16] Yeh, C. T.; Yen, G. C. Induction of hepatic antioxidant enzymes by phenolicacids in rats is accompanied by increased levels of multidrug resistance-associated protein 3 mRNA expression. J. Nutr. 136:11–15; 2006.

[17] Wu, C. H.; Huang, S. M.; Yen, G. C. Silymarin: a novel antioxidant withantiglycation and antiinflammatory properties in vitro and in vivo. Antioxid.Redox Signaling 14:353–366; 2011.

[18] Irwin, N.; McKinney, J. M.; Bailey, C. J.; McClenaghan, N. H.; Flatt, P. R. Acuteand long-term effects of peroxisome proliferator-activated receptor-g activa-tion on the function and insulin secretory responsiveness of clonal beta-cells.Horm. Metab. Res. 43:244–249; 2011.

[19] Kahn, S. E.; Zinman, B.; Lachin, J. M.; Haffner, S. M.; Herman, W. H.;Holman, R. R.; Kravitz, B. G.; Yu, D.; Heise, M. A.; Aftring, R. P.; Viberti, G.Rosiglitazone-associated fractures in type 2 diabetes. Diabetes Care31:845–851; 2008.

[20] Hsu, W. H.; Lee, B. H.; Lu, I. J.; Pan, T. M. Ankaflavin and monascin regulateendothelial adhesion molecules and endothelial NO synthase (eNOS) expres-sion induced by tumor necrosis factor-a (TNF-a) in human umbilical veinendothelial cells (HUVECs). J. Agric. Food Chem. 60:1666–1672; 2012.

[21] Lee, C. L.; Kung, Y. H.; Wu, C. L.; Hsu, Y. W.; Pan, T. M. Monascin andankaflavin act as novel hypolipidemic and high-density lipoproteincholesterol-raising agents in red mold dioscorea. J. Agric. Food Chem.58:9013–9019; 2010.

[22] Jia, X.; Wu, L. Accumulation of endogenous methylglyoxal impaired insulinsignaling in adipose tissue of fructose-fed rats. Mol. Cell. Biochem.306:133–139; 2007.

[23] Kumagai, T.; Nangaku, M.; Kojima, I.; Nagai, R.; Ingelfinger, J. R.; Miyata, T.;Fujita, T.; Inagi, R. Glyoxalase I overexpression ameliorates renal ischemia–reperfusion injury in rats. Am. J. Physiol. Renal Physiol. 296:F912–F921; 2009.

[24] Ueda, K.; Ueyama, T.; Yoshida, K. I.; Kimura, H.; Ito, T.; Shimizu, Y.; Oka, M.;Tsuruo, Y.; Ichinose, M. Adaptive HNE–Nrf2–HO-1 pathway against oxidativestress is associated with acute gastric mucosal lesions. Am. J. Physiol.Gastrointest. Liver Physiol. 295:G460–G469; 2008.

[25] Yang, H.; Zhao, L. F.; Zhao, Z. F.; Wang, Y.; Zhao, J. J.; Zhang, L. Heme-oxygenase-1 prevents liver fibrosis in rats by regulating the expression ofPPARg and NFkB. World J. Gastroenterol. 18:1680–1688; 2012.

[26] Giron, M. D.; Vargas, A. M.; Suarez, M. D.; Salto, R. Sequencing of twoalternatively spliced mRNAs corresponding to the extracellular domain of therat receptor for advanced glycosylation end products (RAGE). Biochem.Biophys. Res. Commun. 251:230–234; 1998.

[27] Son, D.; Kojima, I.; Inaqi, R.; Matsumoto, M.; Fujita, T.; Nanqaku, M. Chronichypoxia aggravates renal injury via suppression of Cu/Zn-SOD: a proteomicanalysis. Am. J. Physiol. Renal Physiol. 294:F62–F72; 2008.

[28] Pashikanti, S.; de Alba, D. R.; Boissonneault, G. A.; Cervantes-Laurean, D.Rutin metabolites: novel inhibitors of nonoxidative advanced glycation endproducts. Free Radic. Biol. Med. 48:656–663; 2010.

[29] Brownlee, M. The pathological implications of protein glycation. Clin. Invest.Med. 18:275–281; 1995.

[30] Wondrak, G. T.; Cervantes-Laurean, D.; Roberts, M. J.; Qasem, J. G.; Kim, M.;Jacobson, E. L.; Jacobson, M. K. Identification of alpha-dicarbonyl scavengers forcellular protection against carbonyl stress. Biochem. Pharmacol. 63:361–373; 2002.

[31] Hsu, L. C.; Hsu, Y. W.; Liang, Y. H.; Kuo, Y. H.; Pan, T. M. Anti-tumor and anti-inflammatory properties of ankaflavin and monaphilone A from Monascuspurpureus NTU 568. J. Agric. Food Chem. 59:1124–1130; 2011.

[32] Kuo, C. F.; Chyau, C. C.; Wang, T. S.; Li, C. R.; Hu, T. J. Enhanced antioxidant andanti-inflammatory activities of Monascus pilosus fermented products by additionof turmeric to the medium. J. Agric. Food Chem. 57:11397–11405; 2009.

[33] Shi, Y. C.; Pan, T. M. Anti-diabetic effects of Monascus purpureus NTU 568fermented products on streptozotocin-induced diabetic rats. J. Agric. FoodChem. 58:7634–7640; 2010.

[34] Shi, Y. C.; Pan, T. M. Antioxidant and pancreas-protective effect of red moldfermented products on streptozotocin-induced diabetic rats. J. Sci. Food Agric.90:2519–2525; 2010.

[35] Diaz-Delfin, J.; Morales, M.; Caelles, C. Hypoglycemic action of thiazolidine-diones/peroxisome proliferator-activated receptor-g by inhibition of thec-Jun NH2-terminal kinase pathway. Diabetes 56:1865–1871; 2007.

[36] Hirai, S.; Takahashi, N.; Tsuyosh Ackerman, W. E.; Zhang, X. L.; Rovin, B. H.;Kniss, D. A. Modulation of cytokine-induced cyclooxygenase 2 expression by

PPARG ligands through NFkappaB signal disruption in human WISH andamnion cells. Biol. Reprod. 73:527–535; 2005.

[37] Henkel, J.; Neuschafer-Rube, F.; Pathe-Neuschafer-Rube, A.; Puschel, G.Aggravation by prostaglandin E2 of interleukin-6-dependent insulin resis-tance in hepatocytes. Hepatology 50:781–790; 2009.

[38] Ankrah, N. A.; Appiah-Opong, R. Toxicity of low levels of methylglyoxal:depletion of blood glutathione and adverse effect on glucose tolerance inmice. Toxicol. Lett. 109:61–67; 1999.

[39] Kelley, G. L.; Allan, G.; Azhar, S. High dietary fructose induces a hepatic stressresponse resulting in cholesterol and lipid dysregulation. Endocrinology145:548–555; 2004.

[40] Thornalley, P. J. Methylglyoxal glyoxalases and the development of diabeticcomplications. Amino Acids 6:15–23; 1994.

[41] He, M.; Siow, R. C.; Sugden, D.; Gao, L.; Cheng, X.; Mann, G. E. Induction ofHO-1 and redox signaling in endothelial cells by advanced glycation endproducts: a role for Nrf2 in vascular protection in diabetes. Nutr. Metab.Cardiovasc. Dis. 21:277–285; 2011.

[42] Mingzhan, X. U. E.; Rabbani, N.; Momiji, H.; Imbasi, P.; Anwar, M. M.;Kitteringham, N.; Park, B. K.; Souma, T.; Moriguchi, T.; Yamamoto, M.;Thornalley, P. J. Transcriptional control of glyoxalase 1 by Nrf2 providesa stress responsive defence against dicarbonyl glycation. Biochem. J.443:213–222; 2012.

[43] Ho, C. Y.; Cheng, Y. T.; Chau, C. F.; Yen, G. C. Effect of diallyl sulfide on in vitroand in vivo Nrf2-mediated pulmonic antioxidant enzyme expression viaactivation ERK/p38 signaling pathway. J. Agric. Food Chem. 60:100–107; 2012.

[44] Yeh, C. T.; Chiu, H. F.; Yen, G. C. Protective effect of sulforaphane onindomethacin-induced cytotoxicity via heme oxygenase-1 expression inhuman intestinal Int 407 cells. Mol. Nutr. Food Res. 53:1166–1176; 2009.

[45] Shih, P. H.; Hwang, S. L.; Yeh, C. T.; Yen, G. C. Synergistic effect of cyanidinand PPAR agonist against nonalcoholic steatohepatitis-mediated oxidativestress-induced cytotoxicity through MAPK and Nrf2 transduction pathways.J. Agric. Food Chem. 60:2924–2933; 2012.

[46] Chartoumpekis, D. V.; Ziros, P. G.; Psyrogiannis, A. I.; Papavassiliou, A. G.;Kyriazopoulou, V. E.; Sykiotis, G. P.; Habeos, I. G. Nrf2 represses FGF21 duringlong-term high-fat diet-induced obesity in mice. Diabetes 60:2465–2473; 2011.

[47] Masterjohn, C.; Mah, E.; Guo, Y.; Koo, S. I.; Bruno, R. S. g-Tocopherol abolishespostprandial increases in plasma methylglyoxal following an oral dose ofglucose in healthy, college-aged men. J. Nutr. Biochem. 23:292–298; 2012.

[48] Wang, S. H.; Chang, J. C.; Pokkaew, R.; Lee, J. F.; Chiou, R. Y. Y. Modified fastprocedure for the detection and screening of antiglycative phytochemicals.J. Agric. Food Chem. 59:6906–6912; 2011.

[49] Brissova, M.; Blaha, M.; Spear, C. Reduced PDX-1 expression impairs isletresponse to insulin resistance and worsens glucose homeostasis. Am. J.Physiol. Endocrinol. Metab. 288:E707–E714; 2005.

[50] Kaneto, H.; Miyatsuka, T. Kawamori, D. PDX-1 and Maf-A play a crucial role inpancreatic beta-cell differentiation and maintenance of mature beta-cellfunction. Endocr. J. 55:235–252; 2008.

[51] Moibi, J. A.; Gupta, D.; Jetton, T. L.; Peshavaria, M.; Desai, R.; Leahy, J. L.Peroxisome proliferator-activated receptor-g regulates expression of PDX-1and NKX6.1 in INS-1 cells. Diabetes 56:88–95; 2007.

[52] Youl, E.; Bardy, G.; Magous, R.; Cros, G.; Sejalon, F.; Virsolvy, A.; Richard, S.;Quignard, J. F.; Gross, R.; Petit, P.; Bataille, D.; Oiry, C. Quercetin potentiatesinsulin secretion and protects INS-1 pancreatic b-cells against oxidativedamage via the ERK1/2 pathway. Br. J. Pharmacol. 161:799–814; 2010.

[53] Xiong, F. L.; Sun, X. H.; Gan, L.; Yang, X. L.; Xu, H. B. Puerarin protects ratpancreatic islets from damage by hydrogen peroxide. Eur. J. Pharmacol.529:1–7; 2006.

[54] Svistounov, D.; Smedsrod, B. Hepatic clearance of advanced glycation endproducts (AGEs)—myth or truth? J. Hepatol. 41:1038–1040; 2004.

[55] Higgins, L. S.; DePaoli, A. M. Selective peroxisome proliferator-activatedreceptor g (PPARg) modulation as a strategy for safer therapeutic PPARgactivation. Am. J. Clin. Nutr. 91:267S–272S; 2010.

[56] Hsu, L. C.; Hsu, Y. W.; Liang, Y. H.; Kuo, Y. H.; Pan, T. M. Anti-tumor and anti-inflammatory properties of ankaflavin and monaphilone A from Monascuspurpureus NTU 568. J. Agric. Food Chem. 59:1124–1130; 2011.

[57] Hsu, W. H.; Lee, B. H.; Lu, I. J.; Pan, T. M. Ankaflavin and monascin regulateendothelial adhesion molecules and endothelial NO synthase (eNOS) expres-sion induced by tumor necrosis factor-a (TNF-a) in human umbilical veinendothelial cells (HUVECs). J. Agric. Food Chem. 60:1666–1672; 2012.

[58] Lee, C. L.; Kung, Y. H.; Wu, C. L.; Hsu, Y. W.; Pan, T. M. Monascin andankaflavin act as novel hypolipidemic and high-density lipoproteincholesterol-raising agents in red mold dioscorea. J. Agric. Food Chem.58:9013–9019; 2010.

[59] Yeh, C. T.; Yen, G. C. Involvement of p38 MAPK and Nrf2 in phenolic acid-induced P-form phenol sulfotransferase expression in human hepatoma HepG2 cells. Carcinogenesis 27:1008–1017; 2006.

[60] Ogborne, R. M.; Rushworth, S. A.; O’Connell, M. A. Epigallocatechin activateshaem oxygenase-1 expression via protein kinase Cd and Nrf2. Biochem.Biophys. Res. Commun. 373:584–588; 2008.

[61] Wu, C. C.; Hsu, M. C.; Hsieh, C. W.; Lin, J. B.; Lai, P. H.; Wung, B. S. Upregulation ofheme oxygenase-1 by epigallocatechin-3-gallate via the phosphatidylinositol3-kinase/Akt and ERK pathways. Life Sci. 78:2889–2897; 2006.

[62] Choi, J. H.; Banks, A. S.; Estall, J. L.; Kajimura, S.; Bostrom, P.; Laznik, D.; Ruas,J. L.; Chalmers, M. J.; Kamenecka, T. M.; Bluher, M.; Griffin, P. R.; Spiegelman,B. M. Anti-diabetic drugs inhibit obesity-linked phosphorylation of PPARg byCdk5. Nature 466:451–456; 2010.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具