Curcumin (diferuloylmethane), a component of turmeric, has been shown to have therapeutic properties. Induction of phase 2 detoxifying enzymes is a potential mechanism through which some of the actions of curcumin could proceed. Heme oxygenase-1 (HO-1), an antioxidant phase 2 enzyme, has been reported to have cytoprotective effects in pancreatic β-cells. Curcumin on further purification yields demethoxy curcumin (DMC) and bisdemethoxy curcumin (BDMC). The objective of the present study was to determine the mechanism by which these purified curcuminoids induce HO-1 in MIN6 cells, a mouse β-cell line. Demethoxy curcuminoids induced HO-1 promoter linked to the luciferase reporter gene more effectively than curcumin. The induction was dependent on the presence of antioxidant response element (ARE) sites containing enhancer regions (E1 and E2) in HO-1 promoter and nuclear translocation of nuclear factor-E2-related factor (Nrf2), the transcription factor that binds to ARE. Curcuminoids stimulated multiple signaling pathways that are known to induce HO-1. Inhibition of specific signaling pathways with pharmacological inhibitors and cotransfection experiments suggested the involvement of phosphotidylinositol 3-kinase and Akt. Real-time quantitative RT-PCR analysis showed significant elevation in the mRNA levels of HO-1 and two other phase 2 enzymes, the regulatory subunit of glutamyl cysteine ligase, which is needed for the synthesis of glutathione, and NAD(P)H:quinone oxidoreductase, which detoxifies quinones. DMC and BDMC induced the expression of HO-1 and translocated Nrf2 to nucleus in β-cells of mouse islets. Our observations suggest that demethoxy curcuminoids could be used to induce a cellular defense mechanism in β-cells under conditions of stress as seen in diabetes.
- phase 2 enzymes
- pancreatic β-cells
- oxidative stress
phase 1 and phase 2 enzymes are involved in cellular defense against oxidative stress and xenobiotic insult (42). The electrophiles generated by phase 1 enzymes (such as cytochrome P-450s) are scavenged by phase 2 enzymes including heme oxygenase (HO)-1, γ-glutamylcysteine ligase (GCL), glutathione S-transferase, and NAD(P)H:quinone oxidoreductase (NQO1) (38). Heme oxygenases catalyze the rate-limiting step of heme degradation, producing equimolar quantities of carbon monoxide (CO), Fe2+, and biliverdin. Biliverdin is then converted to bilirubin by bilirubin reductase. HO-2 is the constitutively active form, and HO-1 is the inducible form. In addition to its role in the metabolism of heme, HO-1 has emerged as an important mediator of cellular defense against wide-ranging tissue injuries (44). Embryonic fibroblasts isolated from mice deficient in HO-1 have been shown to be susceptible to oxidative injury resulting from accumulation of free radicals (37). The antioxidant and antiapoptotic actions of HO-1 have been also attributed to the by-products of heme degradation, namely CO and biliverdin (44). HO-1 has been suggested to be an important therapeutic target in various disease models (8, 44, 51). Several studies have demonstrated the cytoprotective effects of HO-1 in pancreatic β-cells. For example, HO-1 upregulation led to protection of β-cells from various apoptotic stimuli, including cytokines and Fas (36, 40, 55). Induction of HO-1 in mouse islets by protoporphyrin led to improved islet function and survival after transplantation (36). In a recent study, Li et al. (30) demonstrated that rat islets transduced with adenoviral HO-1 survived longer after transplantation and lymphocyte infiltration was reduced in engrafted islets. Furthermore, studies in HO-1-deficient mice have demonstrated the anti-inflammatory action of HO-1 (24). Thus HO-1 is an important target to enhance β-cell survival under conditions of stress.
Inducers of phase 2 enzymes are considered to have therapeutic potential in treating diseases associated with oxidative stress and inflammation (23, 53). One such inducer is curcumin (diferuloylmethane), the active principle in turmeric, a yellow spice from the rhizome of Curcuma longa L, which is being widely used in foods and folk medicine. Curcumin constitutes ∼3–5% of the composition of turmeric. Curcumin has been shown to have anticarcinogenic, antioxidant, and anti-inflammatory actions by various in vitro and in vivo studies (22). Curcumin inhibits lipid peroxidation (48) and scavenges NO (49). Studies in human subjects have indicated that curcumin exerts biological actions without causing toxic side effects (45). Commercial-grade curcumin, which has been generally used in experimental studies, contains a mixture of three curcuminoids, namely, curcumin (77%), demethoxy curcumin (DMC; 17%), and bisdemethoxycurcumin (BDMC; 4%).
Although curcumin has been shown to induce HO-1 in other cell types (3, 15, 35), the mechanism of action of individual curcuminoids on HO-1 expression is not established especially in pancreatic β-cells, a cell type known for low-level expression of antioxidant enzymes (41, 54). The demethoxy curcuminoids are minor components of curcumin, and it will be of interest to determine their antioxidant effects independently. The main objective of this study was to purify curcumin to its constituent curcuminoids and to determine the pathways leading to their induction of HO-1. In this study, we demonstrate that DMC and BDMC induce the expression of HO-1 by a pathway involving the transcription factor nuclear factor-E2-related factor (Nrf2) and phosphotidylinositol 3-kinase (PI3-kinase)/Akt-mediated signaling.
MATERIALS AND METHODS
Cell culture media and supplies were purchased from Gemini Bio Products (Woodland, CA), Sigma Chemical (St. Louis, MO), and Invitrogen-Life Technologies (Rockville, MD). Antibodies specific for ERK, phospho ERK, Akt, phospho Akt (serine 473), p38 subunit of MAPK (p38MAPK), phospho p38MAPK, and β-actin were from Cell Signaling (Beverly, MA). HO-1 antibody was purchased from Calbiochem (La Jolla, CA), and Nrf2 (H-300) antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-rabbit secondary antibody linked to cyanine-3 (Cy3) was purchased from Jackson ImmunoResearch (West Grove, PA). Pharmacological inhibitors, wortmannin, U-0126, SB-203580, and rottlerin were purchased from Biomol (Plymouth Meeting, PA). The inhibitor of Akt, 5-(2-benzothiazolyl)-3-ethyl-2-[2-(methylphenylamino)ethenyl]-1-phenyl-1H-benzimidazolium iodide (Akt inhibitor IV) was purchased from Calbiochem. Plasmids for transfection experiments were purified using Qiagen's (Valencia, CA) Maxi kit. The dual-luciferase assay kit was purchased from Promega (Madison, WI). The plasmid encoding dominant negative mutant form of Akt (T308A; S473A) was provided by Dr. Emmanuel Van Obberghen (Hopital Pasteur, Nice, France).
Promoter constructs of HO-1 linked to a firefly luciferase reporter gene were generated as described previously (1). The plasmid pHO15luc was generated by cloning 15-kb promoter fragment of mouse HO-1 gene into luciferase reporter gene vector pSK1luc. The full-length promoter contains multiple ARE sites at enhancer regions E1 and E2. The plasmid pHOluc (ΔE1) was generated by deletion of 600-bp (SacI/SacI) fragment of pHO15kluc. A 161-bp AflII/BsrBI fragment (E2) was removed to generate the plasmid pHOluc (Δ E2). Deletion of both enhancer elements resulted in the construct, pHOluc (Δ E1+Δ E2). The deletion mutant pHO7luc was generated by digesting pHO15luc with XhoI. A constitutively active Renilla luciferase (pRL-TK-luc) was purchased from Strategene (La Jolla, CA).
Isolation of curcuminoids.
Curcumin, purchased from Sigma Chemical, was purified by column chromatography on silica gel (230–400 mesh) with a solvent system of chloroform- ethanol (25:1 vol/vol). The eluent fractions were tested by thin-layer chromatography. Pure fractions were pooled and concentrated. The purity of each curcuminoid was confirmed by HPLC. Each of the three isolated curcuminoids gave a single peak in HPLC. The structure of these compounds was confirmed by NMR spectral data analysis.
Culture of pancreatic β-cell line and isolation of mouse islets.
MIN6 cells, a mouse pancreatic β-cell line (33), obtained from Dr. Jun-ichi Miyazaki (Kyoto University, Japan) were cultured in DMEM containing 5.6 mM glucose, 10% FBS, 100 μg/ml streptomycin, 100 U/ml penicillin, and 50 μM β-mercaptoethanol (BME) at 37°C in a humidified atmosphere of 5% CO2. Islets were isolated by collagenase digestion from BALB/c mouse at the Islet Core Facility, Barbara Davis Center for Childhood Diabetes, Aurora, Colorado, as described previously (39). Islets were incubated at 37°C in 1 ml of DMEM medium with 0.5% FBS and 5.6 mM glucose at a density of 100 islets per well in 6-well dishes.
Transient transfection in cultured MIN6 cells was carried out by the procedure described earlier using LipofectAMINE 2000 reagent (Invitrogen-Life Technologies, Carlsbad, CA) (21). MIN6 cells cultured in 12-well dishes to ∼70% confluence were used. Plasmids (2 μg) and LipofectAMINE reagent 2000 (4 μl) were diluted separately in 100 μl of Opti-MEM I, mixed, incubated at room temperature (RT; 20 min) and added to the cells. A constitutively active Renilla luciferase (pRL-TK-luc) was included to correct promoter activation for transfection efficiency. The transfected MIN6 cells were cultured in low-serum (0.1%) medium containing BME with appropriate treatment. The treated cells were washed with cold PBS and lysed with 100 μl of reporter lysis buffer. After freezing and thawing, the lysate was centrifuged (10,600 g; 20 min) to collect the supernatant. The activities of firefly luciferase and Renilla luciferase were measured using a dual luciferase assay kit (Promega). The ratios of these two luciferase activities were taken as the measure of promoter activation.
MIN6 cells were cultured in chamber slides to 70% confluence. After exposure of the cells to curcuminoids, they were fixed in 4% paraformaldehyde for 30 min at RT and washed with PBS. The fixed cells were permeabilized in PBS containing 0.2% Triton X-100 and 5% BSA for 90 min at RT. This was followed by exposure to the primary antibody (anti-Nrf2; 1:500) at 4°C overnight. After washing the cells in PBS, the secondary antibody linked to Cy3 (anti-rabbit) was added along with 4,6-diamidino-2-phenylindole (DAPI; 2 μg/ml; nuclear staining) for 90 min at RT. The cells were then washed in PBS, mounted on slides with mounting medium, and examined by fluorescent microscopy.
Immunocytochemical analysis of islets.
Mouse islets incubated in the absence and presence of demethoxy curcuminoids were fixed in 4% paraformaldehyde for 15 min, suspended in 30% sucrose for another 30 min, embedded in Tissue-Tek Optimal Cutting Temperature compound, and frozen. Immunocytochemistry was carried out with 7-μm-thick sections. The slides were heated for 5 min in 10 mM citrate buffer (pH 6.0) in a steamer for antigen retrieval and cooled for 15 min. After being washed in PBS, the slides were incubated in blocking solution (5% normal goat serum and 0.2% Triton in PBS) for 1 h in a humidified chamber. The slides were exposed to primary antibodies at the dilution of 1:2,000 (insulin) and 1:500 (HO-1 and Nrf2) in 3% BSA at 4°C overnight in a humidified chamber, washed in PBS, and exposed to secondary antibodies linked to Cy3 or FITC in 3% BSA for 1 h in the dark. After PBS wash, the slides were incubated with DAPI (2 μg/ml) for 10 min, washed in PBS, and mounted with glycerol mounting medium. The sections were examined by fluorescent microscopy using a Zeiss Axioplan 2 microscope fitted with Cooke SensiCamQE high-performance CCD. For quantitation, the mean integrated fluorescence intensity of the images was calculated using Slide Book Application software (Intelligent Imaging Innovations, Denver, CO).
Western blot analysis.
MIN6 cells incubated under different conditions were washed with ice-cold PBS, and lysed with mammalian protein extraction reagent (M-PER, Pierce, Rockford, IL) supplemented with phosphatase inhibitors (20 mM of sodium fluoride, 1 mM of sodium orthovanadate and 500 nM of okadaic acid) and protease inhibitor cocktail. The protein content of 20,800 g supernatant of the lysates was measured (4). Diluted samples containing equal amounts of protein were mixed with 2× Laemmli sample buffer. The proteins were resolved on a 12% SDS-polyacrylamide gels and transferred to polyvinylidene difluoride membranes. The blots were blocked with TBST [20 mM Tris·HCl (pH 7.9), 8.5% NaCl, and 0.1% Tween 20] containing 5% nonfat dry milk at RT for 1 h and exposed to primary antibodies for ERK, phospho ERK, Akt, phospho Akt (serine 473), p38 subunit of MAPK (p38MAPK), phospho p38MAPK (1:1,000) in TBST containing 5.0% BSA at 4°C overnight. After washing with TBST, anti-rabbit IgG conjugated to alkaline phosphatase was added for 1 h at RT. Blots were then rinsed with washing buffer [10 mM Tris·HCl (pH 9.5), 10 mM NaCl, and 1 mM MgCl2], developed with CDP-Star reagent (New England Biolabs, Beverly, MA), and exposed to X-ray film. The intensity of bands was measured using Fluor-S MultiImager and Quantity One software from Bio-Rad.
RNA isolation and real-time quantitative RT-PCR.
MIN6 cells cultured in 100-mm dishes were exposed to 20 μM of curcumin, DMC, or BDMC for 24 h. RNA was isolated by Qiagen's RNAeasy column method. Quality of RNA was assessed in Agilent 2100 Bioanalyzer using RNA 6000 Nano Labchip kit. The expression of phase 2 enzymes, HO-1, the modulatory subunit of γ-GCL (GCLM) and NQO1 at mRNA levels was examined by real-time quantitative RT-PCR using Taqman probes. The sequence for primers and probes used are given below. The PCR reactions were monitored in real time in an ABI Prism 7700 sequence detector (Perkin Elmer/Applied Biosystems). After amplification, real-time data acquisition and analysis were performed.
The sequence for primers and probes are as follows:
HO-1: Forward primer: GTGATGGAGCGTCCACAGC; Reverse primer: TGGTGGCCTCCTTCAAGG; Probe: 5′-6FAM-CGACAGCATGCCCCAGGATTTGTC-TAMRA-3′
GCLM: Forward primer: CGCCTGCGAAAAAAGTGC; Reverse primer: TCATTCAAGGTCTTTTGGATACAGTC; Probe: 5′-6FAM-CGT CCACGCACAGCGAGGAGCT-TAMRA-3′
NQO-1: Forward primer: GGAAGCTGCAGACCTGGTGA; Reverse primer: CCTTTCAGAATGGCTGGCA; Probe: 5′-6FAM-TTCAGTTCCCATTGCAGTGGTTTGGG-TAMRA-3′
Statistical analysis was performed by one-way ANOVA with Dunnett's multiple comparison test.
ARE site-dependent induction of HO-1 promoter by curcuminoids in β-cells.
The different HO-1 promoter constructs linked to firefly luciferase reporter gene, used in this study, are shown in Fig. 1A. The full-length promoter contains multiple ARE sites in the enhancer regions, E1 and E2. Several constructs with E1 and/or E2 deleted were also used. Each of the purified curcuminoids increased the activity of full length promoter by two- to fourfold with demethoxy curcuminoids being more active than curcumin. This finding suggested that the absence of methoxy groups as in DMC and BDMC results in enhanced induction of the HO-1 promoter. Induction of HO-1 promoter by curcuminoids was completely lost when both enhancer elements containing ARE sites were deleted, suggesting that the induction is ARE dependent. Removal of E1 reduced the induction partially, whereas E2 was not found to be critical for induction. It is also interesting to note that the truncated construct, HO-1-7k-luc that contains E1 was more active than full-length promoter (Fig. 1B), which could be due to removal of negative regulatory elements in the upstream region of HO-1 gene (16).
Curcuminoids induce HO-1 promoter through the transcription factor Nrf2.
Although several transcription factors have been reported to be involved in the regulation of HO-1 expression (44), the critical regulation seems to proceed through the transcription factor Nrf2 that binds to ARE sites during HO-1 induction by curcumin and its derivatives. Nrf2 is normally sequestered in the cytoplasm by the inhibitory protein Keap1. In response to an inducer, Keap1 is directed to proteosomal degradation and Nrf2 is translocated to the nucleus. To test this, we carried out cotransfection experiments. Overexpression of Keap1, which binds to Nrf2 in cytoplasm resulted in a 40–50% decrease in HO-1 induction by curcuminoids (Fig. 2). Similarly, overexpression of dominant negative Nrf2 with deleted transactivation domain also led to a 55–60% decrease suggesting that curcuminoids induce HO-1 promoter through Nrf2 (Fig. 2).
Curcuminoid-mediated induction of phase 2 enzymes by RT-PCR analysis.
Next we examined the mRNA levels of HO-1 by real-time RT-PCR in MIN6 cells exposed to curcuminoids. As shown in Fig. 3A, the curcumin, DMC, and BDMC increased HO-1 mRNA levels by 2.7- to 5.7-fold. The demethoxy curcuminoids were more active than curcumin. These findings are comparable to the extent of induction of the HO-1 promoter. Next, we wanted to determine whether curcuminoids could induce the expression of other antioxidant enzymes that are regulated by a common Keap1-Nrf2-ARE pathway. First we examined the expression of an enzyme involved in the synthesis of glutathione (GSH), a tripeptide composed of glutamate, cysteine, and glycine, which regulates the redox state of the cell. The rate-limiting step in GSH synthesis is catalyzed by the heterodimeric enzyme γ-GCL, which consists of a catalytic subunit (GCLC) and GCLM. The expression of both subunits is regulated by the transcription factor Nrf2. Indeed, the curcuminoids increased the mRNA levels of GCLM by two- to fourfold (Fig. 3B). Finally, we examined the expression of NQO1, another Nrf2-regulated phase 2 enzyme that plays an important role in detoxifying quinones. Curcuminoids induced the expression of NQO1 at the mRNA level by 75–170% (Fig. 3C). The extent of induction of GCLM and NQO1 was less compared with that of HO-1. Differences in fold induction could be due to the number of ARE sites in 5′ flanking regions of phase 2 enzymes. For example, NQO1 promoter contains two ARE sites (20), whereas HO-1 promoter has two enhancer regions E1 and E2, each with several ARE sites.
Nuclear translocation of Nrf2 by curcuminoids.
Next, we used two approaches to determine whether Nrf2 undergoes nuclear translocation in MIN6 cells exposed to curcuminoids. Immunocytochemical analysis showed Nrf2 to be mostly in the cytoplasm in the untreated condition (Fig. 4A). Treatment of MIN6 cells with curcuminoids for 6 h resulted in increased Cy3 signal for Nrf2 in the nucleus, which overlapped with the DAPI stain. For quantitation, the mean integrated fluorescence of Cy3 in 10 different fields was calculated using Slide Book Application software (Intelligent Imaging Innovations, Denver, CO). There was two- to threefold increase in nuclear Nrf2 levels in treated MIN6 cells, with demethoxy curcuminoids being more active than curcumin. In addition to the nuclear localization, there was a significant increase in Nrf2 levels in MIN6 cells exposed to curcuminoids, probably as a result of enhanced stabilization of this transcription factor. Therefore, a decrease in cytosolic Nrf2 was not observed in treated cells. Western blot analysis of Nrf2 in nuclear fractions showed increases of 90, 150, and 180% for curcumin, DMC, and BDMC respectively (Fig. 4B).
Induction of HO-1 in β-cells of mouse islets by demethoxy curcuminoids.
A combination of DMC and BDMC (10 μM each) induced the expression of HO-1 by sevenfold as shown by RT-PCR analysis (Fig. 5A). To determine whether this induction takes place in β-cells of islets, an immunofluorescent staining approach was taken. There was significant increase in the levels of HO-1 (Cy3 staining) in β-cells stained for insulin with FITC compared with untreated islets (Fig. 5B). Furthermore, we observed nuclear translocation of Nrf2 along with overall increase in Nrf2 levels in β-cells of islets incubated in the presence of DMC and BDMC (Fig. 5C). Observations from this set of experiments suggest that demethoxy curcuminoids induce HO-1 in mouse primary β-cells.
Curcuminoids stimulate multiple signaling pathways.
One of the mechanisms by which Nrf2 dissociates from Keap1 and translocates to nucleus is by phosphorylation at serine 40 (17). Previous studies have suggested the involvement of multiple signaling pathways in Nrf2-mediated induction of HO-1 (1, 31). Therefore, first, we examined the signaling pathways stimulated by curcuminoids in MIN6 cells. There were significant increases in the active phosphorylated forms of Akt (a downstream target of PI3-kinase), ERK1/2, and p38MAPK (Fig. 6). The phosphorylated Akt increased by 165–250% (P < 0.001), phosphorylated ERK1/2 by 225–250% (P < 0.001), and phosphorylated p38MAPK by 250–300% (P < 0.001) at the respective peak time point for different curcumin analogs. The total protein content of these targets did not change significantly during the 4-h period examined. The stimulation of signaling by three curcuminoids was similar with some variations in the time course of activation. BDMC caused very transient activation of ERK1/2 and p38MAPK, with peak phosphorylation after 1 h of treatment. In contrast, activation of these kinases in response to curcumin and DMC treatment was slower and more sustained. Given that BDMC was the most potent activator of HO-1 expression following 24 h of treatment, these observations suggested that ERK1/2 and p38MAPK activation was not integral to this response.
PI3-kinase/Akt pathway is involved in the induction of HO-1 promoter by curcuminoids.
To determine the role of different signaling pathways in the induction of the HO-1 promoter, we used pharmacological inhibitors that block a specific signaling pathway; wortmannin for PI3-kinase, 5-(2-Benzothiazolyl)-3-ethyl-2-[2-(methylphenylamino)ethenyl]-1-phenyl-1H-benzimidazolium iodide (Akt inhibitor IV) for Akt, U-0126 for MEK/ERK, and SB-203580 for p38MAPK and rottlerin for PKCδ. We observed that wortmannin and Akt inhibitor IV inhibited HO-1 induction by curcuminoids significantly, by 40–55%, suggesting that signaling mediated by PI 3-kinase/Akt could play a significant role in HO-1 induction (Fig. 7A). We further confirmed the role of Akt by a cotransfection approach. When HO-1 promoter linked to luciferase reporter gene was cotransfected with dominant negative mutant forms of the regulatory subunit of PI3-kinase (Δp85) or Akt (T308A; S473A), 34–42% (P < 0.01) decreases in HO-1 induction by curcuminoids were observed (Fig. 7B). U-0126 and SB-203580 on the other hand did not decrease HO-1 induction by any of the curcuminoids, suggesting that ERK1/2 and p38MAPK are not likely to be involved (Fig. 7C). It is interesting to note that another study had previously demonstrated in rat renal epithelial cells that curcumin activates HO-1 through p38MAPK pathway (3). Therefore, the pathway involved in curcuminoid action might be cell-type dependent. The p38MAPK pathway regulates HO-1 expression with activation as well inhibition in an isozyme-dependent manner (25). Next, the results presented in Fig. 7C suggest that PKCδ plays a minor role in the induction of HO-1 by DMC and BDMC since rottlerin caused a marginal and significant decrease of 25% (P < 0.05) in HO-1 promoter activity. This observation is in agreement with the findings of Rushworth et al. (43), who demonstrated the role of PKCδ in curcumin-mediated induction of HO-1 expression in human monocytes.
Curcuminoid-mediated induction of HO-1 expression by Western blot analysis: role of signaling pathways.
Next, we examined the expression of HO-1 at the protein level in MIN6 treated with curcuminoids after exposing them to inhibitors of signaling pathways. Western blot analyses (Fig. 8) showed that induction of HO-1 expression by curcuminoids in the absence of inhibitors paralleled the findings in promoter assays by transient transfection. Significant decreases (40–60%; P < 0.001) in HO-1 protein content were seen in MIN6 cells treated with curcuminoids in the presence of wortmannin and Akt inhibitor. On the other hand, the inhibitors U-0126 (MEK/ERK), and SB-203580 (p38MAPK) did not influence the induction of HO-1 by curcuminoids. Treatment with rottlerin (PKCδ) resulted in marginal decreases (25%; P < 0.05 for demethoxy curcuminoids). These results are comparable to the inhibition of HO-1 promoter activation by rottlerin (Fig. 7C). As such, inhibition of PI3-kinase pathway does not block curcuminoid action completely and therefore involvement of other pathways cannot be ruled out. Thus induction of HO-1 by curcuminoids proceeds primarily through PI3-kinase/Akt pathway with minor contribution from PKCδ.
Potential therapeutic actions of curcumin have been the subject of several recent reports (reviewed in Refs. 22, 46). Curcumin on further purification yields demethoxy curcuminoids, DMC and BDMC. In the present study, we demonstrate that demethoxy curcuminoids, minor components of curcumin, induce the activity of HO-1 promoter in MIN6 cells, a mouse β-cell line. The induction proceeded through nuclear translocation of the transcription factor, Nrf2. The signaling pathway mediated by PI 3-kinase and Akt played an important role in the induction. DMC and BDMC induced the expression of HO-1 and translocated Nrf2 to nucleus in β-cells of mouse islets. We also demonstrate that curcuminoids induce the expression of two other cytoprotective phase 2 enzymes, namely GCLM and NQO1, suggesting a coordinated defense mechanism against stress. These findings are of significance in the context of oxidative stress-induced β-cell loss in diabetes.
Previous studies have compared the effects of curcumin, DMC and BDMC with respect to different end points. Presence of hydroxyl groups at ortho position of the aromatic ring and β-diketone moiety were found to be essential for quinone reductase activity of natural and synthetic curcuminoids in murine hepatoma cells (9). These structural requirements are met by curcumin, DMC, and BDMC, which differ in the presence of methoxy group(s). Huang et al. (18) demonstrated that curcumin is more potent than BDMC in the inhibition of tumor promotion in mouse skin. However, another study (2) reported that BDMC is a more potent antimutagenic agent. We observed BDMC to be more active than curcumin in the induction of HO-1. It appears that the activities of curcuminoids could differ depending on the biological action examined and the cell type. In the present study, there were minor differences in the mechanism of action of the three compounds examined. There were variations in the peak time point of activation of ERK and p38MAPK (Fig. 6). In addition, PKCδ seems to contribute to HO-1 induction by DMC and BDMC but not curcumin (Fig. 7).
Genetic polymorphism in human HO-1 promoter, which modulates its transcriptional activity, has been shown to be associated with changes in risk profile for cardiovascular disease (10). Mammalian HO-1 promoter undergoes complex regulation with the involvement of several response elements in the promoter regions spanning −10 to −15 kb (44). This regulatory region seems to be the convergence point for a broad spectrum of agents acting through various transcription factors, including Jun, Fos, Ets, cAMP-response element, Maf, and Cap′n′collar (CNC) family of transcription factors (44). Among these, Nrf2 (NF-E2-related factor 2), that belongs to the CNC family of bZIP transcription factors, plays a critical role (34). It binds to the cis-acting ARE, also known as electrophile response element, 5′-TGACnnnGC-3′ in the promoter regions of many phase 2 enzymes, including HO-1 (11). Nrf2 binds to ARE as a heterodimer with one of the small Maf proteins. ARE sites seemed to play a major role in the induction of HO-1 promoter by curcuminoids because deletion of ARE-site containing E1 and E2 regions blunts the induction (Fig. 1). The role of Keap1-Nrf2-ARE pathway is further supported by cotransfection experiments with Keap1 and a dominant negative form of Nrf2 (Fig. 2) and immunocytochemical analysis of nuclear localization of Nrf2 (Fig. 4). However, other studies (12, 15, 16) have reported the regulation of HO-1 induction by pathways not involving Nrf2. A recent study by Hock et al. (16) has characterized the differential regulation of HO-1 promoter by Jun B and Jun D in renal epithelial cells.
Translocation of Nrf2 takes place when it is phosphorylated on serine 40 (19). Previous studies have suggested the involvement of different signaling pathways in Nrf2 phosphorylation and HO-1 induction. For example, PI3-kinase pathway mediates HO-1 induction by carnosal, a polyphenol isolated form the herb rosemary (31). Oxidative stress uses MAPK pathways to induce this gene (26, 56). Cadmium-mediated induction of HO-1 involves p38MAPK (1). Some previous studies have shown the involvement of PKCδ and p38MAPK in HO-1 induction by curcumin in monocytes and in renal epithelial cells respectively (3, 43). On the other hand, we find curcuminoids to exert HO-1 induction primarily through PI3-kinase/Akt-mediated pathway with minor contribution from PKCδ (Figs. 7 and 8). It is likely that curcuminoids could use different pathways depending on the cell type. Other inducers of HO-1 through PI 3-kinase/Akt pathway include insulin (13), epigallocatechin-3-gallate (57), and peroxynitrite (29). The mechanism by which Akt translocates Nrf2 into nucleus is not clearly understood. Akt is not likely to translocate Nrf2 by direct phosphorylation because of lack of consensus phosphorylation sequence. There are at least two possible mechanisms through which Akt could facilitate Nrf2-mediated HO-1 induction. First, Akt inhibits glycogen synthase kinase-3β, which causes export of Nrf2 from nucleus to cytoplasm by a phosphorylation-dependent mechanism. Second, Akt could stabilize Nrf2 by decreasing its proteosomal degradation because it is known to inhibit ubiquitine ligases (32, 50). We did observe an increase in nuclear Nrf2 content without a parallel decrease in cytosol in MIN6 cells exposed to curcuminoids (Fig. 4A).
Induction of antioxidant enzymes could be an important strategy to improve β-cell survival in diabetes. Pancreatic β-cells are particularly vulnerable to oxidative stress-induced injury due to low-level expression of antioxidant enzymes (41, 54). Chronic oxidative stress is an important cause of glucose toxicity in β-cells in Type 2 diabetes (41). Proinflammatory cytokines, important mediators of β-cell death in Type 1 diabetes, induce the expression of inducible NO synthase, leading to the generation of NO (6). When NO combines with superoxide generated by macrophages, there is generation of highly toxic peroxynitrite. Several studies have demonstrated that peroxynitrite is one of the mediators of cytokine-induced β-cell death (7, 28, 52). In a recent study, our laboratory demonstrated that β-cell apoptosis could be significantly reduced in nonobese diabetic mice, an autoimmune diabetic animal model, after administration of a mimetic of manganese superoxide dismutase, which is known to scavenge free radicals (14). In the present study, we demonstrate that curcumin and its derivatives can induce a group of antioxidant enzymes, HO-1, GCLM, and NQO1 which play a significant role in cellular defense against oxidative stress and xenobiotic insult. In addition, curcumin has been previously reported to have potent anti-inflammatory effects that are beneficial in autoimmune diabetes (5, 27, 47). These reports and our current findings suggest that curcuminoids have potential therapeutic values in the treatment of diabetes.
This work was supported by Juvenile Diabetes Research Foundation Grant 5-2005-1104 (to S. Pugazhenthi), American Diabetes Association Grant 1-06-JF-40 (to S. Pugazhenthi), and by Pilot and Feasibility Grant P30 DK-057516 (to S. Pugazhenthi) from the Diabetes and Endocrinology Research Center.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 by American Physiological Society