Curcumin has been shown to possess anti-initiating and anti-promoting activity in experimental systems. However, the mechanisms of its actions are not fully elucidated in vivo . In the present study, mechanisms of curcumin-mediated anti-initiation were investigated in mice employing benzo[ a ]pyrene (B[ a ]P) as a model carcinogen. Dietary pretreatment of mice with chemopreventive doses of curcumin showed significant inhibition of B[ a ]P-induced enzyme activity, protein and messenger RNA (mRNA) levels of cytochrome P450 1A1/1A2 in liver and lungs. Although curcumin alone did not alter the basal levels of aryl hydrocarbon receptor (AhR), it significantly decreased the B[ a ]P-induced AhR protein levels, its phosphorylation, nuclear translocation and subsequent binding to DNA, thereby decreasing the transactivation of CYP1A . Dietary curcumin led to increase in NF-E2-related factor-2 (Nrf2) protein levels and enhanced its nuclear translocation in liver and lungs of mice as compared with controls. Additionally, increased binding of Nrf2 to antioxidant response element occurred in nuclear extracts from liver and lungs of mice pretreated with dietary curcumin. Induction of activity, protein and mRNA levels of glutathione S -transferase, its isoforms and NAD(P)H:quinone oxidoreductase-1 by dietary curcumin in mice paralleled the curcumin-mediated activation of Nrf2, leading to increased detoxification of B[ a ]P. In agreement with the observed curcumin-mediated decrease in B[ a ]P-induced phase I enzyme and concomitant induction of phase II enzymes, pretreatment with dietary curcumin resulted in significant reduction of B[ a ]P-induced DNA adduct, oxidative damage and inflammation. To conclude, curcumin exhibits anti-initiating effects via modulating the transcriptional regulators of phase I and phase II enzymes in mice.
Polycyclic aromatic hydrocarbons are toxic environmental contaminants and have been implicated in the etiology of number of diseases including cancer ( 1 ). Benzo[ a ]pyrene (B[ a ]P), a well-known ubiquitous carcinogen belonging to polycyclic aromatic hydrocarbons group of compounds, is metabolically activated by CYP1A class of cytochrome P450 (CYP450) enzymes to form a highly mutagenic reactive electrophile, benzo[ a ]pyrenediol-epoxide (BPDE) ( 2 ). Though phase II enzymes catalyze the detoxification of BPDE, some of the reactive electrophile interacts covalently with DNA to form adducts. Unrepaired/misrepaired adduct leads to mutation in genes involved in proliferation, growth and apoptosis and finally to a disease condition like cancer. This implicates that phase I and II enzymes play an important role in carcinogen metabolism and hence could be an important target for chemoprevention.
Plant-derived natural compounds are receiving increasing attention as chemopreventives because of low toxicity and high tolerability. Studies from several groups on the chemopreventive efficacy of turmeric/curcumin have shown them to inhibit chemical carcinogen-induced mutagenesis and tumorigenesis in experimental systems ( 3–6 ). Although the chemopreventive potential of curcumin has been established at both initiation and promotion stages of carcinogenesis ( 7 , 8 ), mechanisms of its actions are still not well elucidated.
It has previously been shown that turmeric pretreatment decreased the B[ a ]P-induced CYP1A1/1A2 activity in tissues of rat ( 9 ), whereas it enhanced the activity of glutathione S -transferase (GST) in mouse liver ( 10 , 11 ), resulting in significant decrease in levels of B[ a ]P/7,12-dimethylbenz[a]anthracene-derived DNA adducts in target and/or non-target tissues ( 11–14 ). Nevertheless, it remained unclear as to how curcumin modulates phase I/II enzymes in the course of its action as an anti-initiating agent.
CYP450 1A, in general, is regulated by a basic helix-loop-helix cytosolic protein, aryl hydrocarbon receptor (AhR). Upon ligand binding, AhR translocates to the nucleus ( 15 ), where it heterodimerises with AhR nuclear translocator protein and binds to the xenobiotic response element (XRE) flanking CYP1A1 gene, thereby activating its transcription. Likewise, induction of conjugating enzymes occur in part by NF-E2-related factor-2 (Nrf2), a redox-sensitive member of the cap ‘n’ collar basic leucine-zipper family of transcription factors via an antioxidant response element (ARE) ( 16 ). Curcumin has been shown to increase CYP1A1 activity and AhR–DNA binding in MCF7 cells ( 17 ). Also, studies in renal epithelial and MDA-MB468 breast cells showed the parallel increase in Nrf2 nuclear protein and hemoxygenase-1 in response to curcumin ( 18 , 19 ). Since these investigations have been carried out in cell lines in vitro employing comparatively very high doses of curcumin; absorption, metabolism and pharmacokinetics of curcumin may differ in vivo . The present study aimed to delineate in vivo anti-initiating mechanisms of dietary curcumin (with relatively low and chemopreventive doses) in target (lungs) and non-target (liver) tissues employing B[ a ]P as a model carcinogen.
We demonstrate the inhibitory effects of dietary curcumin on B[ a ]P-induced AhR activation, nuclear translocation and DNA binding, which subsequently result in decreased transcriptional activation of CYP1A and hence reduced protein expression and enzyme activity of CYP1A1/1A2 in vivo . The study also highlights that treatment with dietary curcumin enhanced the Nrf2 protein expression, its nuclear accumulation and DNA binding in mice, together with the induction of phase II enzymes [GST, its isoforms and NAD(P)H:quinone oxidoreductase-1 (NQO1)] activity, protein and messenger RNA (mRNA) levels. These modulations by curcumin further result in decreased carcinogen-induced stress and DNA adducts.
Materials and methods
B[ a ]P, curcumin, ethoxyresorufin, methoxyresorufin, resorufin, 1-chloro-2,4-dinitrobenzene, 3,4-dichloronitrobenzene, ethacrynic acid, hydroxyapatite, poly[dIdC], primers, oligonucleotide probes and antibodies for laminin, phospho-serine and phospho-threonine were obtained from Sigma Chemical Company (St Louis, MO). [ 3 H]-B[ a ]P (specific activity 5455 mCi/mmol) and [α- 32 P]/[γ- 32 P] deoxyadenosine triphosphate (specific activity >3800 Ci/mmol) were purchased from board of radiation and isotope technology (Hyderabad/Mumbai, India). Antibodies for CYP1A1, CYP1A2, AhR, β-actin, tubulin, NQO1, Nrf2, p38 protein kinase, extracellular signal-regulated protein kinase (ERK) 1/2 and Jun N-terminal protein kinase (JNK) 2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) and phospho-p38, phospho-ERK and phospho-JNK were from Cell Signaling Technology (Beverly, MA). Monoclonal antibody for BPDE–DNA adduct clone 5D11 was from Hycult Biotechnology (Uden, The Netherlands). The anti-GST and anti-rabbit/anti-mouse horseradish peroxidase-conjugated secondary antibodies were obtained from Amersham Biosciences (Buckinghamshire, UK) and antibodies for GST isoforms (alpha, mu and pi) were from Alpha Diagnostics (San Antonio, TX).
All animal studies were conducted after approval from the Institutional Animal Ethics Committee as per Committee for the Purpose of Control and Supervision of Experiments on Animals, Government of India guidelines. Inbred male Swiss albino mice (6–8 weeks old) (Animal house, Advanced Centre for Treatment, Research and Education in Cancer, India) were randomized and housed under standard conditions of 22 ± 2°C, 45 ± 10% relative humidity and 12 h light–dark cycles. Animals received drinking water ad libitum during the experimental period and were fed with either powdered standard laboratory diet (control diet) or experimental diet (0.01 and 0.05% curcumin in standard laboratory diet) prepared as described ( 4 ) for 16 days. Mice were given 1 mg B[ a ]P in corn oil or corn oil as vehicle by gavage on 15th day of dietary pretreatment and were killed 24 h after B[ a ]P administration, and their liver and lungs were perfused and excised. Doses of curcumin (0.01/0.05%) employed in the study were equivalent to those known to be present in turmeric diet (0.2/1%) that has been shown to reduce multiplicity, volume and/or incidence of carcinogen-induced tumors at various sites in experimental animals ( 4 , 6 , 20 ). In the present study, only higher dose of curcumin was employed as curcumin control, whereas in experiments wherein cellular distribution of Nrf2, AhR and its ligand binding activity were compared, both 0.01 and 0.05% curcumin doses were used as curcumin control.
Phase I and phase II enzyme activity
Liver and lung microsomal and cytosolic fractions were prepared by homogenizing the tissues in 0.15 M KCl and subsequent differential centrifugation ( 21 ). Protein concentrations of cell fractions were determined ( 22 ). Microsomes were used for measuring activities of CYP1A1 and 1A2 employing isozyme-specific probe drugs—ethoxyresorufin and methoxyresorufin, respectively, as detailed ( 23 ). The product resorufin formed from O -dealkylation of ethoxyresorufin and methoxyresorufin was measured fluorimetrically at Ex = 550 nm and Em = 585 nm.
The activities of GST and its isoforms alpha, mu and pi were assayed in cytosols using isoform-specific substrates such as 1-chloro-2,4-dinitrobenzene, cumene hydroperoxide, 3,4-dichloronitrobenzene and ethacrynic acid, respectively, and measuring the product by spectrometry exactly as described ( 24 , 25 ). Since measurements of GST alpha by this method ( 25 ) also reflect the activity of selenium-dependent glutathione peroxidase, activity with hydrogen peroxide (H 2 O 2 ) as a substrate was also calculated. Contribution of GST to activity was then evaluated by subtracting the enzyme activity measured with H 2 O 2 as substrate from that with cumene hydroperoxide as substrate. A complete assay mixture without enzyme served as the control. The NQO1 activity was measured spectrophotometrically in cytosolic fractions from the reduced nicotinamide adenine dinucleotide phosphate and menadione-dependent dicoumarol-inhibitable reduction of cytochrome c , the terminal electron acceptor, at 550 nm as described ( 26 ). Dicoumarol-sensitive NQO1 activity was calculated from the difference in reaction rates obtained with and without dicoumarol.
Total cell and cytosolic extracts were prepared by previously detailed cell fractionation procedure ( 27 ). Nuclei from mice tissues were isolated by following the standard protocol ( 28 ). Lysates were aliquotized, their protein content was determined ( 22 ) and stored at −80°C. Microsomal (50 μg), cytosolic (100 μg), nuclear (200 μg) as well as total cell proteins (50 μg) were separated on an 8–12% sodium dodecyl sulfate–polyacrylamide gel. Proteins were then transferred to polyvinylidene difluoride membrane and after blocking with 5% non-fat skimmed milk in Tris-buffered saline (pH 7.4) containing 0.1% Tween 20, membranes were probed with 1:2000 dilutions of polyclonal antibodies recognizing CYP1A1, CYP1A2, AhR, GST, NQO1, Nrf2, ERK1/2, p38 protein kinase, JNK2, phospho-p38, p-ERK and p-JNK or with 1:8000 dilutions of antibodies for GST isoforms (alpha, mu and pi) overnight at 4°C. Blots were washed thrice with Tris-buffered saline (pH 7.4) containing 0.1% Tween 20 and incubated with 1:8000 dilutions of anti-rabbit or anti-goat horseradish peroxidase-conjugated secondary antibodies. Immunoreactive bands were visualized with enhanced chemiluminescence reagent from Amersham Biosciences followed by autoradiography.
RNA extraction and semiquantitative reverse transcription–polymerase chain reaction
Total RNA was isolated using Trizol reagent from Invitrogen (Carlsbad, CA) as per the manufacturer’s guidelines. Chromosomal DNA contamination was removed by DNase I digestion and 2 μg of RNA was reverse transcribed to complementary DNA using MBI Fermentas cDNA synthesis kit (Hanover, MA). The resulting complementary DNAs were subjected to polymerase chain reaction for CYP1A1 (forward: 5′-CAGATGATAAGGTCATCACGA-3′; reverse: 5′-TTGGGGATATAGAAGCCATTC-3′), CYP1A2 (forward: 5′-CAGTATCCAAGACATCACAAG-3′; reverse: 5′-TGTGTATCGGTAGATCTCCAG-3′), GST Ya (forward: 5′-AAGCCAGGACTCTCACTA-3′; reverse: 5′-AAGGCAGTCTTGGCTTCT-3′), NQO1 (forward: 5′-CATTCTGAAAGGCTGGTTTGA-3′; reverse: 5′-TTTCTTCCATCCTTCCAGGAT-3′) and β-actin (forward: 5′-GTGGGCCGCTCTAGGCACCA-3′; reverse: 5′-CGGTTGGCCTTAGGGTTCAGG-3′) following the amplification conditions as detailed ( 29–31 ). Polymerase chain reaction products were electrophorized on 1.8% agarose gel and visualized after ethidium bromide staining.
Electrophoretic mobility shift assay
Electrophoretic mobility shift assay was performed using XRE: (5′-agcttGATCTGAGCTCGGAGTTGCGTGAGAAGAGCCG-3′) oligonucleotide probe for AhR and GST Ya–ARE for Nrf2 (5′-TAGCTTGGAAATGACATTGCTAATGGTGACAAAGCAACTTT-3′). XRE/ARE oligonucleotide probes were labeled with [α- 32 P] adenosine triphosphate using klenow enzyme and with [γ- 32 P] adenosine triphosphate using T4 polynucleotide kinase, respectively, and purified on a sephadex G-25 column. The binding reaction was carried out at 25°C for 10 min with or without 5 μg of nuclear protein, 1 μg poly[dIdC] and 10 6 c.p.m. of labeled probe in a final volume of 20 μl binding buffer [10X buffer: 100 mM Tris–HCl (pH 7.5), 500 mM NaCl, 50 mM MgCl 2 , 100 mM ethylenediaminetetraacetic acid, 10 mM dithiothreitol, 1% Triton X-100 and 50% glycerol]. Binding specificity was confirmed by cold competition analysis wherein 25- and 50-fold molar excess of specific unlabeled (XRE for AhR/ARE for Nrf2) or non-specific unlabeled probe (Sp-1) were added to the binding reaction. DNA–protein complexes were separated on a 6% non-denaturing polyacrylamide gel at 200 V. The gel was fixed, dried and DNA–protein complexes were visualized by autoradiography.
Ligand binding assay
AhR ligand binding was studied in liver cytosols of mice employing hydroxyapatite column chromatography as described ( 17 ). Since administering 1 mg [ 3 H]-B[ a ]P to animals (in order to simulate the conditions followed in all in vivo experiments) was practically not possible, cytosols from curcumin or control diet-fed animals were incubated with 10 nM [ 3 H]-B[ a ]P for 16 h at 4°C. In parallel, cytosols from control animals were incubated with 10 nM [ 3 H]-B[ a ]P in the presence of 0.5 or 5 μM curcumin added from outside, served as the positive controls as reported previously in MCF7 cells ( 17 ). [ 3 H]-B[ a ]P–AhR was separated from [ 3 H]-B[ a ]P by eluting the complexes on a pre-equilibrated hydroxyapatite column and radioactivity in the fractions was determined by liquid scintillation counting. Binding specificity was confirmed by incubating cytosols with [ 3 H]-B[ a ]P in presence of 1000-fold excess cold B[ a ]P.
For immunoprecipitation, total cell lysate (500 μg/ml) was precleared by incubating with 20 μl of protein A sepharose beads (Amersham Biosciences) for 1 h at 4°C with gentle rocking. The lysate was centrifuged; cleared extract was then incubated with anti-phospho-serine/anti-phospho-threonine antibody overnight at 4°C with gentle rocking. After addition of 20 μl protein A sepharose beads, the complexes were rotated for additional 2 h at 4°C. Immunoprecipitated proteins were washed thrice with Tris-buffered saline containing 0.05% Tween 20 and resuspended in sodium dodecyl sulfate sample buffer. The samples were boiled, resolved through sodium dodecyl sulfate–polyacrylamide gel electrophoresis and subsequently immunoblotted with anti-AhR as described above.
B[a]P–DNA adduct measurement
The interaction of BPDE with DNA was studied by immunohistochemical staining for BPDE–DNA adducts in formalin-fixed, paraffin-embedded 5 μm tissue sections as described previously ( 32 ). Sections were incubated with anti-BPDE antibody (1:25 dilution). Detection was done using Vectastain ABC kit from Vector Laboratories (Burlingame, CA). Diaminobenzidine was employed as the chromogen and slides were counterstained with Mayor’s hematoxylin. Images were captured with Zeiss Microscope (Imager Z1) with Axiocam MRc5 digital camera attached.
Levels of 8-hydroxy-2′-deoxyguanosine (8-OH-dG) were measured in DNA isolated from mouse liver and lungs. DNA samples were digested with nuclease P1 and alkaline phosphatase and purified before they were quantitatively measured for the oxidative DNA adduct 8-OH-dG employing an ELISA kit (JAICA, Fukuroi City, Japan) as per the manufacturer’s instructions.
Prostaglandin E2 measurement
Prostaglandin E2 (PGE2) levels were determined in mouse liver and lungs employing PGE2-EIA Kit from Cayman Chemical Company (Ann Arbor, MI) according to the manufacturer’s protocol.
Statistical analysis was performed using the SPSS 14.0 software. Data are presented as mean ± standard error. Means of all data were compared by analysis of variance with post hoc testing. P ≤ 0.05 was considered statistically significant.
Based on the net body weight gain and histopathological evaluation of tissues, no toxicity and mortality was observed in animals belonging to the different treatment groups during the experimental period.
Curcumin pretreatment inhibits B[a]P-induced CYP450 isozymes
It was observed that mice receiving control diet and single gavage of 1 mg B[ a ]P showed significant induction of CYP1A1 and CYP1A2 enzyme activity as compared with the basal activity in liver ( Figure 1A ) and lungs ( Figure 1B ). Although dietary curcumin alone did not alter the basal activity in either of the tissues, pretreatment with dietary curcumin showed significant decrease in B[ a ]P-induced CYP1A1/1A2 activities in both liver and lungs. Decrease in B[ a ]P-induced CYP1A1/1A2 activities was higher with 0.05% curcumin diet (43–50% in liver and 59–65% in lungs) as compared with 0.01% curcumin diet (20–30% in liver and 26–34% in lungs). We also determined if the observed curcumin-mediated decrease in B[ a ]P-induced activity of CYP450 was due to altered level of proteins or mRNA. Basal protein expressions of CYP1A1/1A2 were detected in tissues of mice treated with either control or 0.05% curcumin diet alone ( Figure 1C ). It was noted that mice receiving single dose of 1 mg B[ a ]P and control diet showed significant induction of CYP1A1/1A2 proteins in both liver and lungs, although the extent of induction was different. As observed in enzyme activity, dietary pretreatment with 0.01/0.05% curcumin significantly reduced the B[ a ]P-induced levels of CYP1A1/1A2 proteins in mouse tissues. Further, assessment of CYP1A1/1A2 mRNA profile implied that although 0.01% curcumin pretreatment showed significant inhibition in only B[ a ]P-induced CYP1A1 mRNA in lungs and CYP1A2 mRNA in liver, 0.05% curcumin significantly decreased the B[ a ]P-induced CYP1A1/1A2 mRNA levels in both tissues ( Figure 1D ) without altering the basal levels. Together, these findings suggest that since pretreatment with dietary curcumin resulted in decreased B[ a ]P-induced CYP1A1/1A2 activity, protein and mRNA expression, it could be due to the transcriptional regulation of CYP1A class of phase I enzymes by curcumin in vivo .
Dietary curcumin modulates B[a]P-induced AhR by decreasing its activation, nuclear translocation and DNA binding
Since inhibition of B[ a ]P-induced CYP1A1/1A2 activity, protein and mRNA in both liver and lungs were observed with 0.05% dietary curcumin, comparative evaluations of binding of B[ a ]P–AhR complex to XRE were carried out with nuclear extracts from tissues of mice receiving 0.05% curcumin/control diet employing electrophoretic mobility shift assay. As seen in Figure 2A , pretreatment with dietary curcumin decreased the B[ a ]P-induced AhR–DNA binding in nuclear extracts from mouse tissues. Further, in the cold competition analysis, the intensity of signal was found to decrease with 25- and 50-fold molar excess cold XRE for AhR, whereas the signal did not diminish with non-specific cold Sp-1, confirming binding specificity of AhR ( Figure 2B ). It was observed that AhR protein levels in different cellular compartments were comparable in animals treated with either control or 0.05% curcumin diet alone, thus implying that curcumin alone did not influence the basal AhR levels ( Figure 2C ). However, 0.05% curcumin pretreatment significantly decreased the B[ a ]P-induced total AhR protein expression in tissues of mice ( Figure 2D ). Comparisons of AhR protein expressions in cytosolic and nuclear compartments revealed that curcumin pretreatment led to significant and dose-dependent decrease in B[ a ]P-induced AhR nuclear translocation in liver and lungs of mice ( Figure 2D ). Since dietary curcumin inhibited the B[ a ]P-induced nuclear translocation of ligand-activated AhR, we also ascertained if curcumin at the doses employed could compete with B[ a ]P for ligand binding. Ligand binding was significantly inhibited in the positive test control, wherein cytosolic proteins from animals receiving control diet were incubated with 10 nM [ 3 H]-B[ a ]P and 0.5–5 μM curcumin; however, we could not observe significant inhibition in binding when cytosolic proteins from curcumin (0.01/0.05%) -fed animals were incubated with [ 3 H]-B[ a ]P ( Figure 2E ). Decreased binding was observed with 1000-fold excess cold competitor, confirming the binding specificity.
Studies have shown that phosphorylation of AhR/AhR nuclear translocator is crucial for transformation of liganded AhR to a DNA-binding form ( 33 ). In the present study, we noted that B[ a ]P treatment resulted in significant stimulation of AhR phosphorylation in liver and lungs of mice as compared with controls ( Figure 2F ). Importantly, B[ a ]P-induced levels of phosphorylated AhR were significantly diminished upon pretreatment with dietary curcumin ( Figure 2F ).
Overall, the results advocate that dietary curcumin modulates AhR in vivo by decreasing carcinogen-induced (i) phosphorylation/activation; (ii) nuclear uptake and (iii) DNA binding.
Curcumin induces phase II enzyme activity, protein and mRNA
To further elucidate the anti-initiating mechanisms of curcumin-mediated chemoprevention, we studied the effect of dietary curcumin on phase II enzymes. It was observed that mice receiving 0.05% curcumin diet showed significant induction of GST and NQO1 activity ( Figure 3A and B ), protein ( Figure 3C ) and mRNA expressions ( Figure 3D ) in liver and lungs as compared with those treated with control diet. Dietary curcumin (0.05%) also led to significant increase in activity ( Figure 3A ) and protein levels ( Figure 3C ) of GST isoforms (alpha, mu and pi) in liver when compared with controls. However, it was noted that in lungs, pretreatment with 0.05% curcumin significantly induced GST alpha and pi, whereas GST mu remained unaltered ( Figure 3B and C ). Furthermore, among the three isoforms studied, curcumin-mediated increase was higher in GST alpha in liver and in GST pi in lungs. Animals pretreated with dietary curcumin and then challenged with B[ a ]P also showed significant enhancement in GSTs and NQO1 activities ( Figure 3A and B ), protein ( Figure 3C ) and mRNA ( Figure 3D ) expressions when compared with mice receiving control diet and B[ a ]P; nevertheless, the enhancement was relatively less as compared with that in mice receiving dietary curcumin alone. These results imply that dietary curcumin induces phase II detoxifying enzymes involved in the detoxification of carcinogen.
Dietary curcumin augments the Nrf2 protein expression, its nuclear accumulation and ARE binding in mice
We next addressed whether curcumin-mediated enhancement of phase II enzymes in vivo paralleled the increase in Nrf2–DNA binding. It was observed that animals receiving 0.05% dietary curcumin and/or B[ a ]P showed enhanced binding of nuclear extracts from liver and lungs of mice to GST Ya–ARE compared with controls ( Figure 4A ), thus accounting for the increased transcription of GSTs. Binding specificity was confirmed by competitive assay, wherein decrease in signal was observed with 25- and 50-fold molar excess cold ARE for Nrf2, whereas band intensity remained unaltered when cold Sp-1 (non-specific competitor) was employed in the assay ( Figure 4B ). We further investigated the cellular distribution of Nrf2 in mice belonging to the various treatment groups by immunoblotting. Mice receiving curcumin pretreatment and/or B[ a ]P showed significantly increased total Nrf2 protein in liver/lungs as compared with controls ( Figure 4C and D ). Additionally, significant increase in nuclear Nrf2 protein expression with concomitant decrease in cytosolic fraction was observed in tissues of mice pretreated with dietary curcumin alone ( Figure 4C ) and/or subsequently exposed to B[ a ]P ( Figure 4D ).
Curcumin decreases B[a]P-induced DNA adducts, oxidative damage, inflammation and stress-responsive kinases
The observed inhibition of B[ a ]P-induced CYP450 isozymes and enhancement of phase II enzymes by curcumin was further complemented with immunohistochemical measurements of enzyme-catalyzed B[ a ]P-induced DNA adducts in tissue sections of liver and lungs from animals belonging to the various treatment groups. Animals receiving control or 0.05% curcumin diet alone did not show any carcinogen adduct formation ( supplementary Figure 5A and B is available at Carcinogenesis Online). Single exposure of 1 mg B[ a ]P resulted in detection of exposure-specific DNA adduct formation in liver and lungs of mice (indicated by brown-stained nuclei) as compared with vehicle-treated animals. Curcumin pretreatment resulted in significant decrease in B[ a ]P–DNA adduct formation. Reduction in B[ a ]P-induced DNA adducts was higher with 0.05% curcumin diet (41% in liver and 63% in lungs) as compared with 0.01% curcumin diet (13% in liver and 28% in lungs) ( supplementary Figure 5A and B is available at Carcinogenesis Online). Since curcumin-mediated decrease was observed in B[ a ]P-induced CYP450 isozymes, we further investigated the effects of dietary curcumin on B[ a ]P-induced inflammation and oxidative damage produced in response to free radicals generated during oxidative metabolism of B[ a ]P by CYP450s. As shown in Figure 5C , B[ a ]P administration resulted in elevated levels of oxidative DNA lesion, 8-OH-dG in liver and lungs of mice, although the extent of damage was higher in lungs which can be ascribed to the fact that B[ a ]P is a lung-specific carcinogen. Dietary curcumin (0.01 and 0.05%) pretreatment attenuated B[ a ]P-induced oxidative damage by 19–44% in liver and 38–61% in lungs, thus perhaps protecting against B[ a ]P-induced electrophilic/oxidative stress. Basal levels of 8-OH-dG were observed in animals receiving control or 0.05% dietary curcumin alone. Pretreatment with 0.01/0.05% curcumin significantly decreased the B[ a ]P-induced PGE2 levels in lungs; however, the PGE2 levels were comparable in liver tissue from animals belonging to the various treatment groups ( Figure 5D ).
The stress-responsive mitogen-activated protein kinase (MAPK) has been documented to play a crucial role in regulation of the AhR-mediated cellular functions ( 34 ). We noted that curcumin pretreatment inhibited B[ a ]P-stimulated phosphorylation of AhR. To get further insights into the molecular events involved, modulatory effects of curcumin pretreatment on MAPKs were investigated. Results ( Figure 5E ) showed a significant and dose-dependent reduction in B[ a ]P-induced activation of phospho-p38 and p-JNK in tissues of mice upon pretreatment with dietary curcumin. Conversely, B[ a ]P-induced phosphorylation of ERK in mouse tissues was significantly inhibited only at 0.05% curcumin dose. Dietary curcumin alone also led to the activation of these kinases, but that was marginal when compared with the upregulation mediated by B[ a ]P treatment. The levels of total form of each kinase remained unaltered in various treatment groups under the experimental conditions.
In the last decade, a number of chemopreventive agents that have emerged highly promising after preclinical safety and efficacy studies have failed in clinical trials ( 35 ). These failures have channalized the major interest on understanding the mechanistic aspects of a putative chemopreventive agent, before it is implemented in clinical trials. Chemopreventive efficacy of curcumin has been established in various experimental models ( 3–7 , 12 , 20 ). Though numerous mechanistic studies have been carried out with curcumin in vitro (reviewed in ref. 36 ), much remains to be elucidated in vivo to understand the mechanisms of observed chemoprevention. The major aim of this study was to delineate the anti-initiating mechanisms of curcumin-mediated chemoprevention in vivo .
Our results demonstrate that pretreatment with dietary curcumin (0.01/0.05%) significantly inhibited B[ a ]P-induced CYP1A1/1A2 enzyme activities in tissues of mice, as reported earlier with 1% turmeric diet ( 37 ). However, dietary curcumin alone did not affect the basal activity as reported for turmeric ( 37 , 38 ). In contrast to this, curcumin administered at higher concentrations (2% in diet and/or 200–400 mg/kg p.o.), by itself, has been shown to inhibit CYP1A1 activity in mouse liver ( 10 , 39 ). The findings in present study also illustrate significant inhibition of carcinogen-induced CYP1A1/1A2 protein and mRNA expressions in parallel to the observed decrease in enzyme activity upon pretreatment with chemopreventive doses of dietary curcumin, though basal levels of protein/mRNA remained unaltered. Together, the results suggest that curcumin mediates inhibition of carcinogen-induced CYP450s at transcriptional level.
AhR is a ligand-dependent transcription factor, which binds to XRE to activate the transcription of a battery of xenobiotic-metabolizing enzymes ( 15 , 40 ). The present study shows that pretreatment with dietary curcumin attenuated the B[ a ]P-induced increase in AhR–DNA binding, thus accounting for curcumin-mediated transcriptional repression of CYP1Ain vivo . In addition, dietary curcumin decreased the B[ a ]P-induced new synthesis of AhR protein as well as nuclear translocation of AhR which in turn would be responsible for the observed curcumin-mediated decrease in AhR–DNA binding and subsequent CYP1A transactivation. However, 0.01 and 0.05% dietary curcumin did not influence AhR ligand binding in vivo as against that observed with in vitro addition of curcumin at higher concentrations (0.5–5 μM) ( 17 ). This could possibly be due to the lack of sensitivity of the assay system employed or due to the low circulating levels of curcumin achieved after administration of 0.01 and 0.05% dietary curcumin. Dietary curcumin-mediated reduction in B[ a ]P-induced AhR phosphorylation was observed in mouse tissues. Several lines of evidence indicate that phosphorylation of AhR is one of the crucial events required in transformation of liganded AhR to a DNA-binding form ( 33 ) and the process may be modulated by MAPK and/or protein kinase c ( 41–43 ). Results herein indicate significant suppression of B[ a ]P-induced phosphorylated levels of MAPK (ERK, p38 protein kinase and JNK) upon dietary curcumin pretreatment, thus hinting at the possible involvement of B[ a ]P-induced MAPK in mediating phosphorylation of AhR in vivo , although the role of other kinases like protein kinase c or tyrosine kinase cannot be ruled out. It may be important to mention that dietary curcumin-mediated decrease in B[ a ]P-induced nuclear translocation of AhR and DNA binding could be either due to direct effect of curcumin on AhR protein itself or due to the probable in vivo interaction of curcumin with other AhR-associated proteins like hsp90, XAP2 and p23 or AhR nuclear translocator, the heterodimeric partner of AhR. Effects of curcumin on AhR ligand binding and AhR–DNA binding have been shown in several cell lines where the doses of curcumin employed were very high and metabolic competence of cells is likely to be very different from those in our in vivo study ( 17 , 41 ); however, the in vivo studies wherein higher doses of curcumin have shown enhanced CYP1A1 activity, transcriptional modulation by curcumin have not been reported ( 10 , 39 ).
During the course of xenobiotic metabolism, phase I enzymes predominantly CYP450 metabolize the xenobiotics to more reactive electrophilic moieties ( 2 ), which in turn are detoxified by phase II enzymes ( 44 ). Therefore, enhancement in the activity of detoxifying enzymes by chemopreventives would play an important role in blocking the initiation process. Interestingly, data in the present study indicate dietary curcumin-mediated enhancement in the activity of phase II enzyme, GST in mice, as reported with turmeric ( 11 ) and curcumin ( 10 , 39 , 45–47 ) The observed curcumin-mediated increase in hepatic GST isoform activities at low doses (0.01/0.05%) were in agreement with the increase observed with 2% dietary curcumin ( 10 ); however, there were differences in the relative induction of isoforms studied. Additionally, our study reports induction of GST isoforms in extrahepatic tissue, the lungs as well as of another important detoxifying enzyme NQO1 in liver and lungs of mice. Further, mice pretreated with dietary curcumin and subsequently challenged with B[ a ]P also showed enhanced activity of GST, its isoforms and NQO1, suggesting increased detoxification of B[ a ]P by dietary curcumin-induced conjugating enzymes in vivo . Results herein also signify dietary curcumin-mediated increase in protein as well as mRNA expressions of GSTs and NQO1 in mouse tissues, suggesting the role of curcumin in transcriptional regulation of phase II enzymes. To date, effects of curcumin on GST/NQO1 mRNA in vivo are not reported; however, a recent study but in cultured cells employing micromolar concentrations has shown curcumin-mediated induction of GST pi mRNA ( 46 ).
The basic leucine zipper transcription factor Nrf2 plays a key role in regulating the gene expression of cytoprotective enzymes through ARE ( 16 ). Importantly, increased hepatic and pulmonary Nrf2 protein were observed in tissues of mice pretreated with chemopreventive doses of curcumin alone and/or subsequently exposed to B[ a ]P. Additionally, we report enhanced nuclear translocation of Nrf2 and increased Nrf2–ARE binding by dietary curcumin in vivo . A cause–effect relationship between Nrf2 and GST/NQO1 induction, however, cannot be established in the present in vivo experimental conditions and warrants further investigations with Nrf2 knockout mice. Based on the observations in vitro in cultured cells ( 18 , 19 , 46 ), we speculate that Nrf2 might be similarly regulating the observed curcumin-mediated increased transactivation of phase II enzymes in vivo . To our knowledge, this is the first report showing parallel increase in nuclear accumulation of Nrf2, Nrf2–ARE binding and increased activities of phase II enzymes like GSTs and NQO1 in vivo upon curcumin pretreatment. However, while this manuscript was in preparation, a study in agreement with our findings reported induction of hemoxygenase-1 (another important detoxifying enzyme) by orally administered curcumin (200 mg/kg) in rat liver probably via Nrf2–ARE-driven pathway ( 48 ).
The electrophilic intermediate formed as a result of metabolic activation of carcinogen interacts with cellular biomolecules. Since DNA adduct formation marks the process of initiation, the observed curcumin-mediated decrease in DNA adducts in target/non-target tissues appears to play an important role in blocking the initiation process. However, the role of curcumin in enhanced repair of DNA adduct in vivo still remains unexplored and needs further investigations to add to the anti-initiating mechanisms of this polyphenol. Furthermore, curcumin pretreatment decreased the B[ a ]P-induced oxidative damage, which could either be due to the observed curcumin-mediated decrease in B[ a ]P-induced CYP450s or reported free radical scavenging activity of curcumin ( 49 ). In the present study, decrease in B[ a ]P-induced PGE2 levels in lungs were also observed upon curcumin pretreatment, suggestive of decreased inflammation usually associated with stress and DNA damage. B[ a ]P being a lung-specific carcinogen does not lead to hepatocarcinogenicity; however, we have observed almost similar anti-initiating effects of curcumin against B[ a ]P in target (lungs) as well as non-target tissue (liver, the hub of metabolizing enzymes). It is known that the types, amounts and location of carcinogen DNA adducts and cell turnover in a tissue may be the significant factors in tumor development and determining the target organ.
Together, the observed curcumin-mediated decrease in carcinogen-induced AhR induction and CYP1A1 transcription and concomitant induction of Nrf2 and GST/NQO1 resulting in decreased carcinogen-derived DNA adducts put forth the probable involvement of curcumin in modulating the AhR–Nrf2 interaction. The cross talk of these two transcription factors has been shown to be important for efficient coupling of phase I and II xenobiotic-metabolizing enzymes, subsequently ameliorating the oxidative/electrophilic stress ( 40 , 50 ). The strength of this study in showing the biological effects of curcumin in vivo at 50–200 times less dose than that reported in literature ( 10 , 39 , 45 , 47 , 48 ), despite its poor solubility and absorption, would have been further enhanced if circulating levels of curcumin and/or its metabolites were measured. Curcumin-mediated enhancement in Nrf2 is likely to protect against electrophilic insults from different chemical classes of carcinogens, whereas modulation of carcinogen-induced AhR may protect against selective carcinogens only.
To surmise, dietary curcumin inhibits carcinogen-induced CYP450 isozymes by modulating the transcriptional regulator of CYP1A while it induces phase II enzymes through Nrf2, to decrease the carcinogen-induced DNA damage (BPDE adduct and oxidative DNA lesion), thereby further reducing the carcinogen-induced inflammatory and stress response to exhibit its anti-initiating activity as illustrated in Figure 6 .
Indian Council of Medical Research (5/13/12/2001-NCD III); Council of Scientific and Industrial Research (No.9/513 (59)/2004-EMR-I to R. G.)
aryl hydrocarbon receptor
antioxidant response element
- B[ a ]P
benzo[ a ]pyrene
extracellular signal-regulated protein kinase
glutathione S -transferase
Jun N-terminal protein kinase
mitogen-activated protein kinase
xenobiotic response element
The authors thank Mrs Sadhana Kannan for assisting in statistical analysis; Dr A.Ramchandani and Dr M.Krishna for critical reading of the manuscript and Mr Prasad Phase and Mr M.Jagtap for technical assistance.
Conflict of Interest Statement: None declared.