Introduction

The phenolic antioxidant butylated hydroxyanisole (BHA) is a commonly used food preservative with broad biological activities (1), including protection against acute toxicity of chemicals, modulation of macromolecule synthesis and immune response, induction of phase II detoxifying enzymes, and, indeed, its potential tumor-promoting activities. Whereas the potential cytotoxicity of BHA has been partially attributed to reactive intermediates (1,2), BHA has also been shown to shift cell death from necrosis to apoptosis (3,4) and to inhibit mitochondrial complex I and lipoxygenases (3). A chemopreventive role for BHA is reiterated by the induction of A5 subunit of GST in rat liver immunoblotting experiments (5). BHA has also been reported to increase the levels of liver glutathione and the activity of hepatic cytosolic gamma-glutamylcysteine synthetase (3). Moreover, BHA has been shown to be an effective inhibitor of 7,12-dimethylbenz(a)anthracene-induced mammary carcinogenesis (6) in Sprague–Dawley rats; and is effective in the chemoprevention (7) of 1,2-dimethylhydrazine-induced large bowel neoplasms. In addition, BHA in diet has been demonstrated (8) to inhibit the initiation phase of 2-acetylaminofluorene and aflatoxin B1 hepatocarcinogenesis in rats. We have previously demonstrated (9) that the cytotoxicity of BHA is due to the induction of apoptosis that is mediated by the direct release of cytochrome c and the subsequent activation of caspases.

Pivotal to the antioxidant response (10–13) typical in mammalian homeostasis and oxidative stress is the important transcription factor Nrf2 or Nuclear Factor-E2-related factor 2 that has been extensively studied by many research groups (10–13) including this laboratory (14–17). Under homeostatic conditions, Nrf2 is mainly sequestered in the cytoplasm by a cytoskeleton-binding protein called Kelch-like erythroid CNC homologue (ECH)-associated protein 1 (Keap1; (14,18,19). When challenged with oxidative stress, Nrf2 is quickly released from Keap1 retention and translocates to the nucleus (14,20). We have recently identified (14) a canonical redox-insensitive nuclear export signal (NES; 537LKKQLSTLYL546) located in the leucine zipper (ZIP) domain of the Nrf2 protein. Once in the nucleus, Nrf2 not only binds to the specific consensus cis-element called antioxidant response element (ARE) present in the promoter region of many cytoprotective genes (15,19,21), but also to other trans-acting factors such as small Maf (MafG and MafK; (22) that can coordinately regulate gene transcription with Nrf2. We have previously demonstrated (1,23) that BHA is capable of activating distinct mitogen-activated protein kinases (MAPKs) such as extracellular signal-regulated protein kinase 2 (ERK2), and c-Jun N-terminal kinase 1 (JNK1). We have also reported (15) that different segments of Nrf2 transactivation domain have different transactivation potential; and that different MAPKs have differential effects on Nrf2 transcriptional activity with ERK and JNK pathways playing an unequivocal role in positive regulation of Nrf2 transactivation domain activity (15). To better understand the biological basis of signaling through Nrf2, it has also become imperative to identify possible interacting partners of Nrf2 such as coactivators or corepressors apart from trans-acting factors such as small Maf (22).

Nrf2 knockout mice are greatly predisposed to chemical-induced DNA damage and exhibit higher susceptibility towards cancer development in several models of chemical carcinogenesis (21). Observations that Nrf2-deficient mice are refractory to the protective actions of some chemopreventive agents (21), indeed, highlight the importance of the Keap1-Nrf2-ARE signaling pathway as a molecular target for prevention. In the present study, we have investigated, by microarray expression profiling, the global gene expression profiles elicited by oral administration of BHA in small intestine and liver of Nrf2 knockout (C57BL/6J/Nrf2−/−) and wild type (C57BL/6J) mice to enhance our understanding of BHA-regulated cancer chemopreventive effects mediated through Nrf2. We have identified clusters of BHA-modulated genes that are Nrf2-dependent in small intestine and liver and categorized them based on their biological functions. The identification of BHA-regulated Nrf2-dependent genes will yield valuable insights into the role of Nrf2 in BHA-modulated gene regulation and cancer chemopreventive effects. This study also enables the identification of novel molecular targets for BHA-mediated chemoprevention that are regulated by Nrf2. The current study is also the first to investigate the global gene expression profiles elicited by BHA in an in vivo murine model where the role of Nrf2 is also examined.

Materials and Methods

Animals and Dosing. The protocol for animal studies was approved by the Rutgers University Institutional Animal Care and Use Committee (IACUC). Nrf2 knockout mice Nrf2 (−/−) (C57BL/SV129) have been described previously (24). Nrf2 (−/−) mice were backcrossed with C57BL/6J mice (The Jackson Laboratory, Bar Harbor, ME, USA). DNA was extracted from the tail of each mouse and genotype of the mouse was confirmed by polymerase chain reaction (PCR) by using primers (3′-primer, 5′-GGA ATG GAA AAT AGC TCC TGC C-3′; 5′-primer, 5′-GCC TGA GAG CTG TAG GCC C-3′; and lacZ primer, 5′-GGG TTT TCC CAG TCA CGA C-3′). Nrf2(−/−) mice-derived PCR products showed only one band of ∼200 bp, Nrf2 (+/+) mice-derived PCR products showed a band of ∼300 bp while both bands appeared in Nrf2(+/−) mice PCR products. Female C57BL/6J/Nrf2(−/−) mice from third generation of backcrossing were used in this study. Age-matched female C57BL/6J mice were purchased from The Jackson Laboratory. Mice in the age-group of 9–12 weeks were housed at Rutgers Animal Facility with free access to water and food under 12 h light/dark cycles. After one week of acclimatization, the mice were put on AIN-76 A diet (Research Diets Inc. New Jersey USA) for another week. The mice were then administered BHA (Sigma-Aldrich, St. Louis, MO) at a dose of 200 mg/kg (dissolved in 50% PEG 400 solution at a concentration of 20 mg/ml) by oral gavage. The control group animals were administered only vehicle (50% PEG 400 solution). Each treatment was administered to a group of four animals for both C57BL/6J and C57BL/6J/Nrf2(−/−) mice. Mice were sacrificed 3 h after BHA treatment or 3 h after vehicle treatment (control group). Livers and small intestines were retrieved and stored in RNA Later (Ambion, Austin, TX) solution.

Sample Preparation for Microarray Analyses. Total RNA from liver and small intestine tissues were isolated by using a method of TRIzol (Invitrogen, Carlsbad, CA) extraction coupled with the RNeasy kit from Qiagen (Valencia, CA). Briefly, tissues were homogenized in trizol and then extracted with chloroform by vortexing. A small volume (1.2 ml) of aqueous phase after chloroform extraction and centrifugation was adjusted to 35% ethanol and loaded onto an RNeasy column. The column was washed, and RNA was eluted following the manufacturer’s recommendations. RNA integrity was examined by electrophoresis, and concentrations were determined by UV spectrophotometry.

Microarray Hybridization and Data Analysis. Affymetrix (Affymetrix, Santa Clara, CA) mouse genome 430 2.0 array was used to probe the global gene expression profiles in mice following BHA treatment. The mouse genome 430 2.0 array is a high-density oligonucleotide array comprised of over 45,101 probe sets representing over 34,000 well-substantiated mouse genes. The library file for the array is available at http://www.affymetrix.com/support/technical/libraryfilesmain.affx. After RNA isolation, all the subsequent technical procedures including quality control and concentration measurement of RNA, cDNA synthesis and biotin-labeling of cRNA, hybridization and scanning of the arrays, were performed at CINJ core expression array facility of Robert Wood Johnson Medical School (New Brunswick, NJ). Each chip was hybridized with cRNA derived from a pooled total RNA sample from four mice per treatment group, per organ, and per genotype (a total of eight chips were used in this study; Fig. 1). Briefly, double-stranded cDNA was synthesized from 5 μg of total RNA and labeled using the ENZO BioArray RNA transcript labeling kit (Enzo Life Sciences, Inc., Farmingdale, NY, USA) to generate biotinylated cRNA. Biotin-labeled cRNA was purified and fragmented randomly according to Affymetrix’s protocol. Two hundred microliters of sample cocktail containing 15 μg of fragmented and biotin-labeled cRNA was loaded onto each chip. Chips were hybridized at 45°C for 16 h and washed with fluidics protocol EukGE-WS2v5 according to Affymetrix’s recommendation. At the completion of the fluidics protocol, the chips were placed into the Affymetrix GeneChip Scanner where the intensity of the fluorescence for each feature was measured. The expression value (average difference) for each gene was determined by calculating the average of differences in intensity (perfect match intensity minus mismatch intensity) between its probe pairs. The expression analysis file created from each sample (chip) was imported into GeneSpring 7.2 (Agilent Technologies, Inc., Palo Alto, CA) for further data characterization. Briefly, a new experiment was generated after importing data from the same organ in which data was normalized by array to the 50th percentile of all measurements on that array. Data filtration based on flags present in at least one of the samples was first performed, and a corresponding gene list based on those flags was generated. Lists of genes that were either induced or suppressed more than two fold between treated versus vehicle group of same genotype were created by filtration-on-fold function within the presented flag list. By use of color-by-Venn-Diagram function, lists of genes that were regulated more than twofold only in C57BL/6J mice in both liver and small intestine were created. Similarly, lists of gene that were regulated over twofold regardless of genotype were also generated.

Fig. 1
figure 1

Schematic representation of experimental design; SIT, small intestine.

Full size image

Quantitative Real-Time PCR for Microarray Data Validation. To validate the microarray data, 13 genes of interest were selected from various categories for quantitative real-time PCR analyses. Glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) served as the “housekeeping” gene. The specific primers for these genes listed in Table I were designed by using Primer Express 2.0 software (Applied Biosystems, Foster City, CA) and were obtained from Integrated DNA Technologies, Coralville, IA. The specificity of the primers was examined by a National Center for Biotechnology information blast search of the mouse genome. Instead of using pooled RNA from each group, RNA samples isolated from individual mice as described earlier were used in real-time PCR analyses. For the real-time PCR assays, briefly, first-strand cDNA was synthesized using 4 μg of total RNA following the protocol of SuperScript III first-strand cDNA synthesis system (Invitrogen) in a 40 μl reaction volume. The PCR reactions based on SYBR Green chemistry were carried out using 100 times diluted cDNA product, 60 nM of each primer, and SYBR Green master mix (Applied Biosystems) in 10 μl reactions. The PCR parameters were set using SDS 2.1 software (Applied Biosystems) and involved the following stages: 50°C for 2 min, 1 cycle; 95°C for 10 min, 1 cycle; 95°C for 15 s→55°C for 30 s→72°C for 30 s, 40 cycles; and 72°C for 10 min, 1 cycle. Incorporation of the SYBR Green dye into the PCR products was monitored in real time with an ABI Prism 7900 HT sequence detection system, resulting in the calculation of a threshold cycle (C T) that defines the PCR cycle at which exponential growth of PCR products begins. The carboxy-X-rhodamine (ROX) passive reference dye was used to account for well and pipetting variability. A control cDNA dilution series was created for each gene to establish a standard curve. After conclusion of the reaction, amplicon specificity was verified by first-derivative melting curve analysis using the ABI software and the integrity of the PCR reaction product and absence of primer dimers was ascertained. The gene expression was determined by normalization with control gene GAPDH.

Table I Oligonucleotide Primers Used in Quantitative Real-Time PCR
Full size table

Statistics. In order to validate the results, the correlation between corresponding microarray data and real-time PCR data was evaluated by the ‘coefficient of determination’, r 2.

Results

BHA-Modulated Gene Expression Patterns in Mouse Small Intestine and Liver

Subsequent to data normalization, 50.5% (22,779) of the probes passed the filtration based on flags present in at least one of four small intestine sample arrays depicted in Fig. 1. Amongst these probes, 13.86 and 11.69% of probes were induced and suppressed over twofold respectively regardless of genotype. Expression levels of 2,580 probes were either elevated (1,490) or suppressed (1,090) over two fold by BHA only in the wild-type mice, while 3,243 probes were either induced (1,669) or inhibited (1,574) over twofold by BHA only in the Nrf2(−/−) mice small intestine (Fig. 2a). Similarly, changes in gene expression profiles were also observed in mice liver. Overall, the expression levels of 52.79% (23,809) probes were detected in least in one of four liver sample arrays depicted in Fig. 1. Amongst these probes, 6.29 and 5.94% of probes were induced and suppressed over twofold respectively regardless of genotype. In comparison with the results from small intestine sample arrays, a smaller proportion (1,317) of well-defined genes were either elevated (493) or suppressed (824) over two fold by BHA in wild-type mice liver alone; whereas 1,596 well-defined genes were induced (1,005) or inhibited (591) in Nrf2(−/−) mice liver (Fig. 2b).

Fig. 2
figure 2

Regulation of Nrf2-dependent gene expression by BHA in mouse small intestine and liver. Gene expression patterns were analyzed at 3 h after administration of a 200 mg/kg single oral dose of BHA; Nrf2-dependent genes that were either induced or suppressed over twofold were listed. The positive numbers on the y-axis refer to the number of genes being induced; the negative numbers on the y-axis refer to the number of genes being suppressed.

Full size image

Quantitative Real-Time PCR Validation of Microarray Data

To validate the data generated from the microarray studies, several genes from different categories (Table I) were selected to confirm the BHA-regulative effects by the use of quantitative real-time PCR analyses as described in detail under Materials and Methods. After ascertaining the amplicon specificity by first-derivative melting curve analysis, the values obtained for each gene were normalized by the values of corresponding GAPDH expression levels. The fold changes in expression levels of treated samples over control samples were computed by assigning unit value to the control (vehicle) samples. Computation of the correlation statistic showed that the data generated from the microarray analyses are well-correlated with the results obtained from quantitative real-time PCR (coefficient of determination, r 2 = 0.89; Fig. 3).

Fig. 3
figure 3

Correlation of microarray data with quantitative real-time (qRT) PCR data. Fold changes in gene expression measured by quantitative real-time PCR for each sample in triplicate (n = 3) were plotted against corresponding fold changes from microarray data (coefficient of determination, r 2 = 0.89

Full size image

BHA-Induced Nrf2-Dependent Genes in Small Intestine and Liver

Genes that were induced only in wild-type mice, but not in Nrf2(−/−) mice, by BHA were designated as BHA-induced Nrf2-dependent genes. Based on their biological functions, these genes were classified into categories, including ubiquitination and proteolysis, electron transport, detoxification enzymes, transport, apoptosis and cell cycle control, cell adhesion, kinases and phosphatases, transcription factors and interacting partners, RNA/protein processing and nuclear assembly, biosynthesis and metabolism, cell growth and differentiation, and G protein-coupled receptors (Table II lists a subset of these genes relevant to our interest).

Table II BHA-Induced Nrf2-Dependent Genes in Small Intestine and Liver
Full size table

Gene expression in small intestine in response to BHA treatment was more sensitive than that elicited in the liver with a larger number of Nrf2-dependent genes being upregulated in the former. The category of transcription factors and interacting partners predominated the upregulated genes followed by kinases and phosphatases. In the former category, a number of interesting transcription factors were identified as BHA-regulated Nrf2-dependent genes. In the small intestine, these primarily included insulin-like growth factor 2 (Igf2), Jun oncogene (Jun), Notch gene homolog 4 (Drosophila, Notch 4), nuclear receptor co-repressor (Ncor1), nuclear receptor interacting protein 1 (Nrip1), serum response factor binding protein 1 (Srfbp1), Spred-1 (Spred1), suppressor of cytokine signaling 5 (Socs5), thyroid hormone receptor beta (Thrb), transforming growth factor, beta receptor 1(Tgfbr1), transducer of ERBB2, 2 (Tob2), members 19 and 23 of tumor necrosis factor superfamily (Tnfrsf19 and Tnfrsf23), v-maf musculoaponeurotic fibrosacroma oncogene family, protein G (avian, MafG), and wingless-type MMTV integration site 9 B (Wnt9b). Similarly, the major BHA-regulated Nrf2-dependent transcription factors identified in the liver included activating signal cointegrator 1 complex subunit 2 (Ascc2), Eph receptor A3 (Epha3), Eph receptor B1 (Ephb1), fos-like antigen 2 (Fosl2), insulin-like growth factor 2 receptor (Igf2r), nuclear receptor interacting protein 1 (Nrip1), RAB4A member RAS oncogene family (Rab4a), reticuloendotheliosis oncogene (Rel), and transcription factor AP-2 beta (Tcfap2b). Interestingly, induction of Nrip1 was observed in both small intestine and liver suggesting that the Nrf2/ARE pathway may play a dominant role in BHA-elicited regulation of this gene.

Amongst the kinases and phosphatases in small intestine, the highest expression was seen with microtubule associated serine/threonine kinase-like (Mastl). Several important members of the mitogen-activated protein kinase (Mapk) pathway activating the Nrf2/ARE signaling were also induced in the small intestine including Mapk8, Mapk6, Map3k9, Map4k4 and Map4k5 with strongest induction seen with Mapk8. Moreover, induction was also observed of Janus kinase 2 (Jak2), types 1 and 2 of neurotrophic tyrosine kinase, receptor (Ntrk1 and Ntrk2), and tyrosine kinase, non-receptor, 1 (Tnk1). Comparable expression was noted of different isoforms of protein kinases including epsilon, eta, and cAMP dependent regulatory, type II alpha (Prkce, Prkch and Prkar2a respectively). Regulatory subunits of protein phosphatase 1 (Ppp1r14a) and 2 (Ppp2r5e) and catalytic subunit of protein phosphatase 3, beta isoform (Ppp3cb) were also upregulated in the small intestine as BHA-regulated and Nrf2-dependent genes. In the liver, the genes induced in the same category included diacylglycerol kinase kappa (Dagkκ), inhibitor of kappaB kinase gamma (Iκbkg), protein kinase, cAMP dependent regulatory, type I, alpha (Prkar1a), protein tyrosine phosphatase (Ptp), and putative membrane-associated guanylate kinase 1 (Magi-1) mRNA, alternatively spliced c form (Baiap1).

Representative genes induced by BHA in an Nrf2-dependent manner in the category of apoptosis and cell cycle control genes included members of the caspase cascade as well as cyclins G1 and T2 in small intestine; and growth arrest and DNA-damage-inducible 45 alpha (Gadd45a) in liver. Several important Nrf2-dependent detoxifying genes were also upregulated by BHA including glutamate-cysteine ligase, catalytic subunit (Gclc) in both small intestine and liver, glutathione S-transferase, mu 1 (Gstm1) and mu 3 (Gstm3) isoforms in small intestine and liver respectively, and heme oxygenase (decycling) 1 (Hmox1) in liver. BHA could also modulate many other important categories of genes in an Nrf2-dependent manner. Salient amongst them were the biosynthesis and metabolism genes, G-protein coupled receptors, RNA/protein processing and nuclear assembly, ubiquitination and proteolysis, and transport genes. Major transporters induced included the multidrug resistance associated proteins Mrp1 and 3 and Mdr1. Also induced were the sodium/glucose cotransporter (Slc5a12) and ion channels for sodium, potassium and chloride ions.

BHA-Suppressed Nrf2-Dependent Genes in Small Intestine and Liver

As shown in Table III which lists a subset of genes relevant to our interest, BHA treatment also inhibited the expression of many genes falling into similar functional categories in an Nrf2-dependent manner, although the number of genes was smaller. Notably, the category of transcription factors and interacting partners remained the largest amongst these genes with nuclear receptor coactivator 3 (Ncoa3), and src family associated phosphoprotein 1 (Scap1) being among the genes suppressed by BHA and co-regulated with Nrf2 in small intestine; and epidermal growth factor receptor pathway substrate 15 (Eps15) and hypoxia inducible factor 1, alpha subunit (Hif1a) being suppressed in the liver.

Table III BHA-Suppressed Nrf2-Dependent Genes in Small Intestine and Liver
Full size table

Amongst the kinases and phosphatases, BHA suppressed, in small intestine, the expression of G protein-coupled receptor kinase 5 (Grk5), glycogen synthase kinase 3 beta (Gsk3β), Mapk-activated protein kinase 5 (Mapkapk5), Mapk associated protein 1 (Mapkap1), p21-activated kinase 3 (Pak3), and ribosomal protein S6 kinase, polypeptide 4 (Rps6ka4). In the liver, BHA suppressed, in an Nrf2-dependent manner, genes such as Map2k7, phosphoinositide-3-kinase, regulatory subunit 5, p101 (Pik3r5), Ribosomal protein S6 kinase, polypeptide 5 (Rps6ka5) and microtubule associated serine/threonine kinase family member 4 (Mast4) amongst others.

Major genes down-regulated by BHA in an Nrf2-dependent manner in the category of apoptosis and cell cycle control included B-cell leukemia/lymphoma 2 (Bcl2), breast cancer 1 (Brca1), and Cyclin I in liver. Among the transport genes, the fatty acid binding protein 6, ileal (gastrotropin; Fabp6) was suppressed in small intestine by BHA. In the category of electron transport genes, representative genes included the cytochrome c oxidase, subunit VIIa 1 (Cox7a1) which was inhibited in both small intestine and liver, and thioredoxin reductase 2 (Txnrd2) which was suppressed in liver. Besides, members of the cytochrome P450 family such as Cyp3a44 in small intestine, and Cyp21a1 and Cyp7a1 in liver were also down-regulated by BHA in an Nrf2-dependent manner in the same category. Other categories of BHA-suppressed genes including ubiquitination and proteolysis, RNA/protein processing and nuclear assembly, biosynthesis and metabolism, and cell growth and differentiation were also identified as regulated through Nrf2.

Discussion

Since BHA was first introduced as a food preservative back in the 1960s, it has attracted a lot of attention and debate because of its potentially diverse biological effects on the health of humans including its potential cancer chemopreventive effects. Although extensive studies have been conducted to define the biological activities of BHA, and a growing body of evidence has been accumulated, a comprehensive definition of the potential cellular targets of BHA that trigger important signal transduction pathways remains a challenge that has yet to be undertaken. Indeed, the molecular basis and the mechanisms of action of BHA are not quite yet fully understood (1). Transcription factor Nrf2 or Nuclear Factor-E2-related factor 2 is indispensable to cellular defense against many chemical insults of endogenous and exogenous origin (21), which play major roles in the etiopathogenesis of many cancers. Pivotal to this role of Nrf2 is the antioxidant response element (ARE) present in the promoter regions of many cytoprotective genes (19,21). Indeed, Nrf2 (−/−; deficient) mice, which are highly susceptible to cancer development, are known to be refractory to the protective actions of some cancer chemopreventive agents (21). It is, therefore, of interest to investigate the role of Nrf2 in BHA-elicited global gene expression profiles in vivo in order to extend the latitude of current understanding on the Nrf2-ARE signaling pathway that has emerged as an important molecular target for cancer chemoprevention. To our knowledge, this is the first attempt to elucidate, by microarray expression profiling, the gestalt genomic basis of BHA-regulated Nrf2-dependent cancer chemoprevention in vivo in an Nrf2-deficient murine model.

In the continuing quest to unravel the complex secrets of the biology of Nrf2 in cancer chemoprevention, there is renewed interest in dissecting the interacting partners of Nrf2 such as coactivators and corepressors which are co-regulated with Nrf2. In a recent microarray study (25), we have reported that CREB-binding protein (CBP) was upregulated in mice liver on treatment with (−)epigallocatechin-3-gallate (EGCG) in an Nrf2-dependent manner. We have also demonstrated (15) previously, using a Gal4-Luc reporter co-transfection assay system in HepG2 cells, that the nuclear transcriptional coactivator CBP, which can bind to Nrf2 transactivation domain and can be activated by extracellular signal-regulated protein kinase (ERK) cascade, showed synergistic stimulation with Raf on the transactivation activities of both the chimera Gal4-Nrf2 (1–370) and the full-length Nrf2. In the current study, we observed the upregulation of the trans-acting factor v-maf musculoaponeurotic fibrosarcoma oncogene family, protein G, avian (MafG), nuclear receptor co-repressor 1 (Ncor1) and nuclear receptor co-repressor interacting protein (Nrip1); as well as downregulation of the nuclear receptor co-activator 3 (Ncoa3) in an Nrf2-dependent manner. Although microarray expression profiling cannot provide evidence of binding between partners, this is the first investigation to potentially suggest that co-repressors Ncor1 and Nrip1 and co-activator Ncoa3, such as CBP in our previous studies, may serve as putative BHA-regulated nuclear interacting partners of Nrf2 in eliciting the cancer chemopreventive effects of BHA. Furthermore, induction of Nrip1 was observed in both small intestine and liver suggesting that the Nrf2/ARE pathway may play an important role in BHA-elicited regulation of this gene. Taken together, it is tempting to speculate that the BHA-regulated chemopreventive effects through the ARE may be regulated by a multimolecular complex, which involves Nrf2 along with the transcriptional co-repressors Ncor1 and Nrip1 and the transcriptional co-activator Ncoa3, in addition to the currently known trans-acting factors such as MafG (22), with multiple interactions between the members of the putative complex as we have shown recently with the p160 family of proteins (27). Further studies of a biochemical nature would be needed to substantiate this hypothesis and extend our current understanding of Nrf2 regulation in chemoprevention with BHA.

The major detoxification genes induced in this study included glutamate-cysteine ligase, catalytic subunit (Gclc) in both small intestine and liver, glutathione S-transferase, mu 1 (Gstm1) and mu 3 (Gstm3) isoforms in small intestine and liver respectively, and heme oxygenase (decycling) 1 (Hmox1) in liver. Since Nrf2 binds to the cis-acting ARE and induces the expression of many important Phase II detoxification and antioxidant genes (10,19–21), and since BHA is known to induce Phase II genes, the current study on spatial regulation in small intestine and liver of BHA-regulated chemopreventive effects via Nrf2 showing the upregulation of Phase II detoxification genes is, indeed, consistent with previous reports (11, 28–31) and also validates our results from a biological perspective. Indeed, we were able to detect the presence of NAD(P)H:quinone oxidoreductase (NQO1) gene induction in qRT-PCR experiments performed in another study at a 12 h time-point (data not shown) suggesting that there may be a relatively delayed induction of the NQO1 gene compared with the other Phase 2 genes in response to BHA and possible differential kinetics of BHA-regulated Phase 2 gene response. Interestingly, genes involved in Phase I drug metabolism as well as Phase III drug transport were also regulated by BHA via Nrf2. The Phase I drug-metabolizing enzymes (DMEs) identified here included cytochrome P450 family members Cyp7a1 and Cyp21a1 suppressed in liver and Cyp3a44 suppressed in small intestine. The roles played by these Phase I DMEs in BHA-elicited cancer chemopreventive effects, however, remain presently unknown. Several transport-related genes were also identified in this study for the first time as both BHA-regulated and Nrf2-dependent. Amongst the upregulated transporters were the ATP-binding cassette superfamily members belonging to MDR/TAP, MRP or ALD subfamilies such as MDR1a/P-glycoprotein (Abcb1a) and MRP 1 (Abcc1) in small intestine; and MRP3 (Abcc3) in the liver. The ALD member responsible for peroxisomal import (32) of fatty acids (Abcd3) was upregulated in the liver, whereas the fatty acid binding protein 6, ileal (gastrotropin; Fabp6) was suppressed in small intestine, suggesting a putative role for BHA and Nrf2 co-regulation of lipid pathways in chemoprevention that has never been examined. Also identified for the first time as transport genes upregulated by BHA via Nrf2 were members of the solute carrier family such as genes encoding for zinc transport, sodium/glucose cotransport, sodium-dependent phosphate transport and organic anion transport (Slc39a10, Slc5a12, Slc20a2 and Slco2a1 respectively). Although the involvement of water channels (aquaporins) in cell migration, fat metabolism, epidermal biology and neural signal transduction point to a role in the pathophysiology of cancers (33), the current study is the first to show BHA-elicited regulation via Nrf2 of aquaporins 1, 3 and 8 which may play a role in the overall cancer chemopreventive effects of BHA. Taken together, the current study suggests that BHA could coordinately regulate the Phase I, II, and III xenobiotic metabolizing enzyme genes as well as antioxidative stress genes through Nrf2-dependent pathways in vivo. The regulation of these genes could have significant effects on prevention of tumor initiation by enhancing the cellular defense system, preventing the activation of procarcinogens/reactive intermediates, and increasing the excretion/efflux of reactive carcinogens or metabolites (26).

Bone morphogenetic proteins (BMPs) are multifunctional signaling molecules regulating growth, differentiation, and apoptosis in various target cells (34,35). BMP receptor 1A (Bmpr1a) is a serine/threonine kinase receptor that mediates the osteogenic effects of the BMPs and is co-ordinately regulated with transforming growth factor, beta (Tgfβ) in vitro (36). Here, we show that Bmpr1a was upregulated along with a strong induction of Tgfβ (>14-fold) in an Nrf2-dependent manner in small intestine. Therefore, the current study is the first to identify a role for Nrf2 in co-regulation of BMP and Tgfβ pathways in BHA-elicited chemopreventive effects in vivo which may be, in part, due to induction of apoptosis caused by activation of BMP (34). Interestingly, Bmp6 was suppressed in the liver suggesting a spatial regulation of BHA-regulated chemoprevention via Nrf2 in the liver and small intestine. A recent study (37) reported that BMP-2 modulates the expression of molecules involved in Wnt signaling, and activates the canonical Wnt pathway in normal human keratinocytes. In our study, we also observed an Nrf2-dependent upregulation of wingless-type MMTV integration site 9 B (Wnt9b) along with a down-regulation of skin hornerin (Hrnr) in small intestine, and an upregulation (greater than fivefold) of stratum corneum (epidermis) development genes such as keratin associated protein 6-1 (Krtap6-1) and keratin complex 2, basic, gene 18 (Krt2-18) in liver. Since BMP-2 and Wnt are involved in the development of skin and skin appendages (37) as morphogens, and since Nrf2 has been implicated in hyperproliferation of keratinocytes (38), our results are the first identification of a putative in vivo cross-talk regulated by Nrf2 and modulated by BHA between BMP and Wnt pathway members in the etiopathophysiology and chemoprevention of skin cancers. Further studies in an appropriate in vivo model as well as in vitro mechanistic studies will be necessary to enhance our current understanding of regulation of skin cancer by Nrf2.

We have demonstrated previously (1) in vitro that BHA is capable of activating distinct mitogen-activated protein kinases (MAPKs) including extracellular signal-regulated protein kinase 2 (ERK2), and c-Jun N-terminal kinase 1 (JNK1). The current study elucidates the Nrf2-dependent, BHA-modulated regulation in vivo of many members of the MAPK family including Mapk6, Mapk8, Map3k9, Map4k4, Map4k5, Map2k7, Mapkap1, and Mapkapk5; as well as the Jun oncogene, thus, validating the physiological relevance of our results. BHA could also alter the expression of many important signaling biomolecules in discrete signal transduction pathways in an Nrf2-dependent manner including those of the JAK/STAT pathway (Janus kinase 2, Jak2, and protein inhibitor of activated Stat4, Pias4). In mammalian cells, insulin-induced PI3K (phosphoinositide 3-kinase) activation, generates the lipid second messenger PtdIns (P 3), which is thought (39) to play a key role in triggering the activation of p70 ribosomal S6 protein kinase (S6K). The identification in the current study of phosphoinositide-3-kinase, regulatory subunit 5, p101 (Pik3r5), insulin-like growth factor 1 (Igf1), Insulin-like growth factor 2 receptor (Igf2r), and ribosomal protein S6 kinase, polypeptide 5 (Rps6ka5) as Nrf2-dependent and BHA-regulated genes is interesting as this is the first identification of Igf2r as a target of BHA in vivo, and may be another putative mechanism by which BHA elicits its Nrf2-mediated chemopreventive effects. We also observed a BHA-elicited, Nrf2-dependent stimulation of diacylglycerolkinase kappa, and epsilon and eta isoforms of protein kinase C (Prkce and Prkch respectively) which is consistent with reports of PKC-activation by BHA and diacylglycerol (40,41). In addition, G protein-coupled receptor kinase 5 (GRK5) and glycogen synthase kinase 3 beta (Gsk3β), which were down-regulated in small intestine in an Nrf2-dependent manner, were identified for the first time as putative targets for BHA-mediated chemoprevention.

BHA could also modulate the expression of many genes involved in apoptosis and cell cycle control in an Nrf2-dependent manner including cyclin G1 (Ccng1), cyclin T2 (Ccnt2), cyclin I (Ccni), G0/G1 switch gene 2 (G0s2), growth arrest and DNA-damage-inducible 45 alpha and beta (Gadd45a, Gadd45b), CASP8 and FADD-like apoptosis regulator (Cflar), growth arrest specific 1 (Gas1), G1 to S phase transition 1 (Gspt1), breast cancer 1 (Brca1), and p21 (CDKN1A)-activated kinase 3 (Pak3). Several other important categories of genes were identified as Nrf2-dependent and BHA-regulated such as cell adhesion, biosynthesis and metabolism, ubiquitination and proteolysis, RNA/protein processing and nuclear assembly, cell growth and differentiation, DNA replication and G-protein coupled receptors. The current study, thus, addresses the spatial regulation in mouse small intestine and liver of global gene expression profiles elicited by BHA in exerting its chemopreventive effects via Nrf2. Since a greater number of genes in this study were altered in the small intestine as compared to the liver, and since the phenolic compound BHA (Fig. 4) is a lipophilic molecule with a KowWin (http://www.syrres.com/Esc/est_kowdemo.htm) estimated log octanol/water partition coefficient (log P) as high as 3.50, the spatial regulation of gene expression profiles may be attributed to a complex of physiological factors including partitioning across the gastrointestinal tract, intestinal transit time, uptake into the hepatobiliary circulation, exposure parameters such as Cmax, Tmax and AUC, and pharmacokinetics of disposition after oral administration of BHA. Further studies will be necessary to address the effect(s) of temporal dependence on pharmacokinetic parameters and gene expression profiles to further enhance our current understanding of BHA-mediated chemoprevention mechanisms.

Fig. 4
figure 4

Chemical structure of butylated hydroxyanisole (BHA).

Full size image

In conclusion, our microarray expression profiling study provides some novel insights into the pharmacogenomics and spatial regulation of global gene expression profiles elicited in the mouse small intestine and liver by BHA in an Nrf2-dependent manner from a gestalt biological perspective. Amongst these BHA-regulated genes, clusters of Nrf2-dependent genes were identified by comparing gene expression profiles between C57BL/6J Nrf2(+/+) and C57BL/6J/Nrf2(−/−) mice. The identification of novel molecular targets that are regulated by BHA via Nrf2 underscores the ineluctable importance of the Nrf2/ARE signaling pathway in cancer chemoprevention. This study clearly extends the current latitude of thought on the molecular mechanisms underlying BHA’s cancer chemopreventive effects as well as the role(s) of Nrf2 in its biological functions. Future in vivo and in vitro mechanistic studies exploring the germane molecular targets or signaling pathways as well as Nrf2-dependent genes related to the significant functional categories uncovered in the current study would inexorably extend our current understanding of cancer chemoprevention.