Oncogenic Signaling Pathways Activated in DMBA-Induced Mouse Mammary Tumors

Abstract

Only about 5% of human breast cancers can be attributed to inheritance of breast cancer susceptibility genes, while the balance are considered to be sporadic in origin. Breast cancer incidence varies with diet and other environmental influences, including carcinogen exposure. However, the effects of environmental carcinogens on cell growth control pathways are poorly understood. Here we have examined oncogenic signaling pathways that are activated in mammary tumors in mice treated with the prototypical polycyclic aromatic hydrocarbon (PAH) 7,12-dimethylbenz[a]anthracene (DMBA). In female FVB mice given 6 doses of 1 mg of DMBA by weekly gavage beginning at 5 weeks of age, all of the mice developed tumors by 34 weeks of age (median 20 weeks after beginning DMBA); 75% of the mice had mammary tumors. DMBA-induced mammary tumors exhibited elevated expression of the aryl hydrocarbon receptor (AhR), c-myc, cyclin D1, and hyperphosphorylated retinoblastoma (Rb) protein. Because of this, the activation of upstream regulatory pathways was assessed, and elements of the Wnt signaling pathway, the NF-κB pathway, and the prolyl isomerase Pin-1 were found to be frequently up-regulated in the tumors when compared to normal mammary gland controls. These data suggest that environmental carcinogens can produce long-lasting alterations in growth and anti-apoptotic pathways, leading to mammary tumorigenesis.

Introduction

Although the second leading cause of cancer mortality in women is breast cancer, we still do not have a very complete understanding of the mechanisms of mammary tumorigenesis. Extensive investigation has identified breast cancer susceptibility genes that can be inherited, but these appear to play a minor role in most cases of breast cancer. The geographic variation in incidence of breast cancer suggests that environmental and dietary factors play a significant role. A major class of environmental carcinogens is a family of structurally related chemicals, the polycyclic aromatic hydrocarbons (PAH). These carcinogens and related halogenated compounds have been implicated in mammary tumorigenesis by epidemiological and laboratory studies. DMBA (7,12-dimethylbenz[a]anthracene) is a prototypical PAH that has been used to promote tumors in laboratory animals (Medina, 1974). Mammary tumors can be produced in rodents following administration of DMBA by oral gavage, leading to up-regulation of the cellular cytosolic receptor for DMBA, the aryl hydrocarbon receptor (AhR) (Trombino et al., 2000).

Upon ligand activation, the AhR translocates into the nucleus and associates with the cofactor ARNT, the AhR nuclear translocation protein (Denison and Nagy, 2003; Swanson et al., 1995). This activated AhR/ARNT complex then binds to specific DNA recognition sites upstream of AhR responsive genes and induces gene transcription (Denison and Nagy, 2003). The early steps in tumorigenesis involve AhR-dependent up-regulation of cytochrome P450 enzymes, which metabolize DMBA into a mutagenic epoxide intermediate that readily forms DNA adducts. These adducts are associated with DNA mutations and the malignant transformation that is thought to be involved in PAH-mediated carcinogenesis (Nebert et al., 1990; Rundle et al., 2000). However, there is growing evidence that the AhR also regulates cell growth (Abdelrahim et al., 2003; Ge and Elferink, 1998; Holcomb and Safe, 1994; Ma and Whitlock, 1996; Puga et al., 2000) and survival (Caruso et al., 2004; Matikainen et al., 2001), suggesting that there are additional mechanisms through which the AhR may contribute to mammary tumorigenesis. Furthermore, while PAH-mediated DNA damage may be an important initiation factor, it has also been suggested that DNA damage may not be sufficient for the development of tumors (Miller and Miller, 1981).

We hypothesized that epigenetic changes in oncogene expression and intracellular signaling may play important roles. To test whether additional biochemical changes may occur that could lead to mammary tumor formation after PAH exposure, we established an oral dose regimen of DMBA that resulted in mammary tumors in 75% of FVB mice by 30 weeks of age, associated with tonic up-regulation of the AhR. RNA transcripts for cyclin D1 and c-myc, important oncogenes in human breast cancer (Bartkova et al., 1994; Jamerson et al., 2004), were increased in many of the DMBA-induced mammary tumors. These results led us to examine the effects of DMBA treatment on upstream regulators of c-myc and cyclin D1, including the NF-κB and Wnt signal transduction pathways, and the prolyl isomerase Pin1, an important integrator of multiple signaling pathways.

NF-κB/Rel is a structurally and evolutionary conserved family of transcription factors. NF-κB is an important regulator of mammary gland development, where it controls proliferation and branching (Brantley et al., 2001; Cao et al., 2001), and also protects the epithelium during apoptotic alveolar involution (Clarkson et al., 2000). We and others have demonstrated constitutive activation of NF-κB factors in breast cancer: high levels of nuclear NF-κB were found in the majority of primary human breast and DMBA-induced rat mammary tumor specimens, breast cancer cell lines, and carcinogen-transformed human mammary epithelial cells (Sovak et al., 1997). NF-κB up-regulation can precede carcinogen-induced rat mammary tumor formation (Kim et al., 2000), and overexpression of the NF-κB family member c-Rel in the mammary gland promotes breast cancer in transgenic mice (Romieu-Mourez et al., 2003). NF-κB can activate transcription of cyclin D1 (Joyce et al., 2001), and NF-κB and AhR act cooperatively to transactivate the murine proto-oncogene c-myc (Kim et al., 2000).

Another pathway that has been implicated in carcinogen-induced tumorigenesis, and may also lead to transactivation of the c-myc and cyclin D1 oncogenes is the Wnt pathway. The Wnt pathway was originally discovered in Drosophila as an important signaling pathway in development (Sharma and Chopra, 1976). Later it was appreciated to have a role in tumor development (Nusse and Varmus, 1982; Rijsewijk et al., 1987). The discovery of mutations in the adenomatous polyposis coli (APC) gene, an important tumor suppressor protein in the canonical Wnt pathway, as the underlying cause of the familial adenomatous polyposis colon cancer syndrome, provided molecular evidence for the role of Wnt signaling in human tumors (Nagase and Nakamura, 1993). Wnts are a family of secreted glycoproteins that bind to Frizzled (Fz) 7 transmembrane cell surface receptors. Activation of Fz leads to inhibition of glycogen synthase kinase 3β (GSK3β) via the Disheveled (Dvl) protein. Inhibition of GSK3β prevents N-terminal phosphorylation, and subsequent ubiquitination and degradation of β-catenin. The failure to degrade β-catenin then leads to cytoplasmic accumulation and nuclear translocation. Once in the nucleus, β-catenin binds transcription factors of the T cell (TCF) and lymphoid enhancing (LEF) families and stimulates transactivation of various genes such as cyclin D1 and c-myc, which are believed to be central to the tumorigenic potential of Wnt signaling (Polakis, 2000).

Cyclin D1 promotes cell growth by activating cyclin-dependent kinases (Cdks) 4 and 6, which phosphorylate the retinoblastoma protein (Rb), a critical step in regulation of the restriction point at the G1/S transition (Weinberg, 1995). Recent work from our lab has identified CK2, a ubiquitously expressed serine/threonine kinase consisting of 2 α or α′ catalytic subunits and 2 regulatory β subunits, as a positive regulator of β-catenin stability and Wnt signaling in both mammalian cells (Song et al., 2003) and in Xenopus embryos, where it has Wnt-like axis determining activity (Dominguez et al., 2004). We and others have demonstrated that constitutive activation of canonical Wnt signaling (through overexpression of CK2α (Landesman-Bollag et al., 2001) kinase inactive-GSK (Farrago et al., Cancer Research, in press), β-catenin (Michaelson and Leder, 2001) or Wnts (Li et al., 2000)) leads to mammary tumorigenesis, and squamous metaplasia (Miyoshi et al., 2002) in mouse models. Moreover, aberrant constitutive activation of canonical Wnt signaling has been frequently found in primary human breast tumors and in human breast cancer cell lines (Brown, 2001; Smalley and Dale, 2001). Activation of the Wnt pathway has been frequently seen in carcinogen-induced tumors, particularly those of the gastrointestinal tract in rodents (Koesters et al., 2001; Nozaki et al., 2003; Takahashi et al., 1998; Takahashi and Wakabayashi, 2004; Ubagai et al., 2002); however, there are few studies of the contribution of Wnt to mammary tumorigenesis. In a recent paper, the frequency of Wnt pathway mutations induced in mice by 1,3-butadiene was only 18% (Zhuang et al., 2002).

The prolyl isomerase Pin1, also shown to be up-regulated in human mammary tumors, has also been implicated in activation of Wnt signaling through various mechanisms. Pin1 is a prolyl isomerase that catalyzes the cis-trans isomerization of pSer/Thr-Pro motifs, and thereby facilitates conformational changes that affect a wide variety of protein functions including subcellular localization, catalytic activity, protein-protein interactions, and protein turnover. Up-regulation of Pin1 has been observed in a variety of human cancers (Bao et al., 2004), including breast cancer, where a strong correlation has been observed between Pin-1 and β-catenin up-regulation (Ryo et al., 2001). Pin1 directly potentiates Wnt signaling and expression of the target genes cyclin-D1 and c-myc by facilitating β-catenin release from APC, the adenomatous polyposis coli gene product (Ryo et al., 2001). Pin-1 also directly binds to and stabilizes cyclin D1, and Pin-1 knockout mice demonstrate many of the same phenotypes as cyclin D1 null mice, i.e., impairment of normal mammary gland development and retinal hypoplasia (Liou et al., 2002).

To begin to elucidate the mechanism leading to the DMBA-induced increase in cyclin D1 and c-myc, we assessed pathways implicated in the regulation of these genes by comparing tumor samples to normal mammary glands. Here, we demonstrate activation of signaling through the AhR, NF-κB, the Wnt signaling pathway and Pin-1.

Materials and Methods

DMBA Treatment

Virgin female FVB/N mice were treated according to a protocol approved by the Boston University Institutional Animal Care and Use Committee. Mice were housed in a 2-way barrier at the Boston University School of Medicine animal facility in accordance with the regulations of the American Association for the Accreditation of Laboratory Animal Care. Twenty mice were each given 6 weekly 1.0 mg doses of DMBA in 0.2 ml of sesame oil by oral gavage, beginning at 5 weeks of age. Mice were then mated continuously to provide an oscillating hormonal environment and followed until either tumors developed or the mice died. Mice bearing tumors >0.5 cm were euthanized by CO₂ inhalation and necropsied. Mammary tumors and grossly normal mammary glands from parous age-matched control FVB mice were excised and portions of the tissues were prepared for histology. The remaining tissue was frozen on dry ice for molecular and biochemical studies. Frozen tissues were stored at −80°C.

Histology

Upon necropsy, tumors and organs were removed and immediately fixed in Optimal Fix (American Histology Reagent Company, Inc). The tissues were processed, embedded in paraffin, and sectioned at 7 μ. The sections were mounted on glass slides and stained with hematoxylin and eosin using routine laboratory procedures in the Mutant Mouse Pathology Laboratory at the University of California, Davis. Sections were compared with other specimens in the extensive mouse mammary tumor database at 〈http://imagearchive.compmed.ucdavis.edu〉.

Immunoblot Analysis

Whole cell protein extracts were prepared by homogenizing frozen tumors or mammary gland specimens in lysis buffer containing a cocktail of protease inhibitors (50 mM Tris-HCl pH 8.0, 1% Nonidet P-40, 125 mM NaCl, 1 mM NaF, 1mM phenylmethylsulfonyl fluoride (PMSF), 1 ug/ml aprotinin, 1 ug/ml pepstatin, 1 ug/ml leupeptin, 1 mM Na3VO4, and 10 mM sodium pyrophosphate) (see EMSA methods for nuclear extraction methods). Equal amounts of protein as determined by BCA protein assay (Pierce, Rockford, IL) were diluted with 4X sample loading buffer (100 mM Tris-HCl, pH 6.8, 200 mM dithiothretol, 4% SDS, 0.2% bromophenol blue, 20% glycerol), boiled, and loaded onto 8% polyacrylamide gels.

Electrophoresis was performed in a Bio-Rad Mini Protean II gel system at 110V for 2 hours. After electrophoresis, gels were transferred onto PVDF membranes (Millipore Inc.). Membranes were blocked in 5% milk in 1X TBS, 0.05% Tween, incubated with an appropriate primary antibody, washed, incubated with HRP-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA), washed again and visualized by ECL (Pierce, Rockford, IL) according to the manufacturer’s instructions. Primary antibodies were the following antibodies: anti-β-catenin (BD, Upstate), anti-β-actin (SIGMA or C-11 Santa Cruz), anti-CK2α (BD), anti-GSK3β (Stressgen), anti-AhR (BioMol), Pin1 (Ab-1, Oncogene), Phospho-specific Rb (p-780, p-795, p-807/811, Cell Signaling), pan-Rb antibody (G3-245, BD Pharmingen). For quantitative analysis of each band, integrated pixel density minus background density was determined using a Fluor-S MultiImager and analysis was done using Quantity One software (Bio-Rad, Hercules, CA).

RNA Extraction and Reverse Transcription

For expression analysis, total RNA was extracted from mouse tissues by homogenization of frozen tissue in 4M Guanidine Hydrochloride (American Bioanalytical), followed by ultracentrifugation (16 hours, 42,000 rpm) in a 5.7 M cesium chloride (American Bioanalytical), 0.025 M NaOAc (American Bioanalytical) gradient. After DNase treatment (Roche, Indianapolis, IN), RNA was reextracted using buffered phenol/choloform (American Bioanalytical), and ethanol precipitated. One μg RNA was then reverse transcribed using the ProSTAR First Strand RT-PCR kit (Stratagene, La Jolla, CA).

Quantitative Real-Time PCR (QPCR)

Twenty-five μl reactions were prepared by mixing 8 ng of cDNA, 12.5 μl of Taqman Universal PCR Mastermix (Applied Biosystems, Foster City, CA), and 1.25 μl of an Assay on Demand Gene Expression Reagent (Applied Biosystems, Foster City, CA) for the gene of interest. These include cyclin D1 (CCND1), β-catenin (CATNB), c-myc (MYC), aryl-hydrocarbon receptor (AHR), or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an endogenous control. QPCR was performed in an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA). The initial step was for 10 min at 95°C followed by 40 cycles (95°C for 15 seconds, 60°C for 60 seconds). Background signal was eliminated and Ct values were determined using the SDS version 1.1 analysis software (Applied Biosystems, Foster City, CA). Standard curves within the linear range for GAPDH and gene of interest were performed.

Electromobility Shift Assay (EMSA) and Nuclear Extraction

Frozen tumor tissue was pulverized in liquid nitrogen with a mortar and pestle and resuspended at a concentration of 1 g/ml in homogenization buffer (10 mM Hepes pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 50 mM sucrose, 1 mM DTT, 0.5 mM PMSF, 5 μg/ml leupeptin, and 5 μg/ml aprotinin, Sigma Chemical Co., St. Louis, MO). Samples were Dounce-homogenized for 20 strokes with a loose pestle and then 20 strokes with a tight pestle. The KCl concentration was then adjusted to 100 mM and the nuclei were washed twice. Nuclear proteins were extracted on ice for 30 minutes in 2 packed nuclear volumes of 10 mM Hepes pH 7.9, 400 mM NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 20% glycerol, 1 mM DTT, 0.5 mM PMSF, 0.5 μg/ml leupeptin, and 5 μg/ml aprotinin. Protein concentration was determined using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA). The sequence of the upstream regulatory element (URE) NF-κB-binding oligonucleotide from the c-myc gene is as follows: 5′-GATCCAAGTCCGGGTTTTCCCCAACC-3′ (Duyao et al., 1990), where the core element is underlined. The Octomer-1 (Oct-1) binding oligonucleotide has the following sequence: 5′-TGTCGAATGCAAATCACTAGAA-3′. Nuclear extracts samples (5 μg) were subjected to EMSA, as described (Sovak et al., 1997).

Statistics

For all quantifiable measurements, p-values were obtained using the Student’s t-test comparing values from the tumor set versus values from the set of normal samples, using the statistical analysis package in Excel.

Results

Tumor-Incidence and Histology in FVB/N Mice Treated with DMBA

Female mice were treated with 6 weekly 1 mg doses of DMBA by oral gavage beginning at 5 weeks of age. Beginning 4 weeks after the final dose, at 15 weeks of age, mice began to develop evidence of tumors (Figure 1, Kaplan-Meier survival plot) and were then euthanized for analysis. By 34 weeks of age (29 weeks after beginning DMBA treatment), all mice had developed tumors. 75% of the mice had mammary tumors, 15% of the mice had lung tumors, 10% had lymphomas, and 5% had skin cancers; some mice had more than 1 malignancy. In addition, 30% of the mice had myeloid hyperplasia.

These findings can be compared with large reported series of female FVB mice, in which 15% of mice develop bronchoalveolar lung adenomas and carcinomas by 14 m of age; lymphomas (in 6%) and skin tumors (in 3%) have been reported at 24 m of age (Mahler et al., 1996). Mammary gland hyperplasia and carcinomas (both associated with and independent of pituitary hyperplasia) also can occur, but have not been reported in the first year of life (Nieto et al., 2003; Wakefield et al., 2003). The most common histologies of the DMBA-induced mammary tumors were squamous or adenosquamous carcinomas, accounting for 85% of the mammary tumors, with other tumors being of the Wnt type (Rosner et al., 2002), or scirrhous tubular, spindle cell, or papillary carcinomas (Table 1 and Figure 2).

DMBA-Induced Mammary Tumors Have Increased Expression of AhR mRNA, Protein

If the AhR plays an ongoing role in mammary tumorigenesis, we predicted it would be up-regulated in carcinogen-induced tumors. To test this hypothesis, expression levels of the AhR protein and mRNA transcript were measured by immunoblot and quantitative PCR analysis, respectively. Proteins were extracted from tumors arising in DMBA-treated mice and the normal mammary glands from age-matched controls, and immunoblotting was performed for AhR. The blots were stripped and re-probed for β-actin to control for protein loading. AhR protein expression in 7 of 11 mammary tumors was considerably higher than in the controls (Figure 3a, 3b). The bands were quantitated by phosphoimager analysis, and the average level of AhR protein (expressed as a normalized ratio of Fluor-S units compared to actin) in tumors was 0.98 ± 0.09 (mean ± S.E.) as compared to 0.08 ± 0.03 in the controls (p = 0.011). Similar results were obtained in 4 separate experiments using 2 different AhR antibodies.

To measure the level of AhR mRNA, quantitative PCR (QPCR) was performed in tumor and normal samples, with GAPDH expression used for normalization. The ratio of AhR mRNA in each individual sample as compared to the mean of the normal tissues is presented as the fold-change. Similar to AhR protein results, AhR mRNA levels were higher in the majority of DMBA-induced tumors, as compared to normal mammary tissue (Figure 3c), although the magnitude of AhR mRNA and protein levels did not always directly correlate (e.g., Tumor T3). These results suggest that AhR up-regulation in mammary tumors may be regulated at multiple levels, but are consistent with a role for the AhR in mammary tumorigenesis.

Cyclin D1 and c-myc mRNA Levels in DMBA-Induced Mammary Tumors

Two cell cycle regulators that have been shown to be important for the progression of mammary tumorigenesis are the cyclin D1 and c-myc oncogenes. Thus, in our initial studies, quantitative PCR was performed to compare mRNA levels of these genes in tumor and normal samples. The ratio of cyclin D1 mRNA in each individual sample as compared to the mean of the normal tissues is presented as the fold-change, normalized to GAPDH expression (Figure 4a). Six out of 12 tumor samples tested showed increased cyclin D1 mRNA levels, which represents a statistically significant increase in the average cyclin D1 mRNA in the tumors versus controls [9.30 ± 1.27 vs. 1.00 ± 0.00 (p = 0.086)]. In contrast to cyclin D1, an increase in c-myc mRNA was observed only in a few of the tumors (Figure 4b).

Hyperphosphorylated Rb Protein Expression in DMBA-Induced Mammary Tumors

G1 kinases, including cyclin D1-associated kinases, directly phosphorylate Rb, and the hyperphosphorylated Rb dissociates from E2F, which is critical in promoting cell cycle progression (Weinberg, 1995). Thus, Rb hyperphosphorylation is a measure of inactivation of Rb-mediated growth suppression and deregulated cell cycle control. Therefore, we examined the phosphorylation status of Rb protein in DMBA-induced mammary tumors. Increased amounts of phosphoRb and an increased proportion of Rb in the phosphorylated form was readily detected in DMBA tumors, in contrast to normal mammary gland (Figure 5a, 5b). Hyperphosphorylated Rb protein levels in tumors were 0.34 ± 0.03 compared with 0.03 ± 0.01 in normal controls (p = 0.013), indicating that Rb is inactivated in DMBA-induced mammary tumors.

DMBA-Induced Mammary Tumors Exhibit Increased Nuclear NF-κB Binding

The NF-κB pathway is capable of regulating transcription of the c-myc and cyclin D1 genes. To determine whether DMBA-induced mouse mammary tumors have constitutively active nuclear-NF-κB, EMSAs were performed using the c-myc promoter URE as a probe (Duyao et al., 1990). Nuclear extracts from the DMBA-induced mouse mammary tumors were compared with those from mammary glands of age-matched female FVB/N mice or from a nulliparous animal (Figure 6). NF-κB bands were identified using supershift analysis (data not shown). While NF-κB binding in normal mammary gland fluctuates with the estrus cycle, the predominant complex seen was made up of p50 homodimers (Band 5) and a minor amount of p50/p65 heterodimers (Band 1), as seen previously (Clarkson et al., 2002). In addition, some clipping of p50 occurred (Band 6), as we have observed previously (Pianetti et al., 2001). In contrast, DMBA-induced tumors displayed increased levels of the potent transactivating complexes p50/p65, p50/c-Rel (Band 2), p52/RelB (Band 3), in addition to p50 or p52 (Band 2) homodimers. Equal loading and protein integrity of the nuclear extracts was confirmed by analysis of Oct-1 binding. Thus, DMBA-induced mammary tumors are characterized by elevated levels of NF-κB binding complexes known to be able to activate transcription.

DMBA-Induced Mammary Tumors Have Elevated Expression of Canonical Wnt Signaling Components

To evaluate the Wnt signaling pathway in DMBA-induced mammary tumors, we initially measured whole cell and nuclear levels of β-catenin protein, the essential transcriptional cofactor for canonical Wnt signaling. In the whole cell extracts, substantial up-regulation of β-catenin was found in 6 of 12 tumor samples tested, when compared to mammary glands from untreated, age-matched FVB mice (Figure 7a, 7b). The average level of β-catenin, expressed as a normalized ratio of Fluor-S units compared to β-actin, was significantly higher in the tumors, as compared to the normal controls [0.36 ± 0.35 vs. 0.03 ± 0.01 (p = 0.008)]. Similar results were obtained in 3 separate experiments using 2 different β-catenin antibodies (data not shown). We also examined β-catenin expression in the nuclear extracts prepared for EMSA, and found a similar pattern of increased nuclear β-catenin (data not shown).

One of the positive regulators of β-catenin protein stability is the serine-threonine kinase CK2, which phosphorylates β-catenin in the armadillo repeat region and stabilizes it. Interestingly, all but 1 tumor demonstrated a significant up-regulation of the major catalytic CK2α subunit (Figure 8a, 8b) as compared to the normal tissues. Mean tumor and normal values were [0.24 ± 0.014 vs. 0.003 ± 0.001 (p = 0.001)] respectively. The up-regulation of CK2α correlated well with the up-regulation of β-catenin (Figure 8c), consistent with the putative role of CK2 as a positive Wnt regulator. These results demonstrate activation of the Wnt signaling pathway in DMBA-induced mammary tumors.

Increased Pin1 Protein Expression in DMBA-Induced Mammary Tumors

Expression of the prolyl-isomerase Pin1 is often increased in human cancers, including breast cancer (Bao et al., 2004). Therefore, we examined Pin1 expression in the DMBA-induced mammary tumors. As shown in Figure 9, Pin1 protein expression was significantly increased in many of the mammary tumors, compared to age-matched normal FVB mice mammary glands. The average level of Pin1, normalized ratio of Fluor-S units compared to β-actin, was 0.25 ± 0.02 in the tumors compared with 0.08 ± 0.06 in the controls (p = 0.032) (Figure 9a, 9b). Interestingly, QPCR analysis of Pin1 mRNA levels showed no significant differences between tumors and the age-matched normal controls (data not shown), suggesting that elevated Pin1 expression in DMBA-induced mammary tumors is likely due to a posttranscriptional mechanism. A correlation between the de-regulation of Pin1 protein and several important cell cycle regulators was observed (Table 2). In many of the tumor samples, increased Pin1 expression correlated well with an increase β-catenin (Figure 9c), and also with the level of hyperphosphorylated Rb and cyclin D1 mRNA (Table 2).

Discussion

The present study describes an oral dose regimen of the environmental carcinogen DMBA capable of generating mammary tumors in FVB strain mice. We hypothesized that, in addition to its genotoxic properties, DMBA would also alter oncogenic signaling pathways in order to transform cells. Here we have shown evidence of up-regulation of cell cycle regulators that have well-established roles in mammary tumorigenesis and of the signaling pathways capable of up-regulating them. These data provide evidence of specific biochemical changes that accompany tumor formation in PAH-induced mammary tumors, and activation of critical regulatory pathways that can contribute to growth and oppose apoptosis in cancer in the process of carcinogen-induced mammary tumorigenesis in mice. These observations expand upon previous reports of DNA damaging events such as the induction of point mutations in genes such as c-H-ras by DMBA (Cardiff et al., 1988).

As it is believed that both genotoxic and mitogenic activities of environmental carcinogens may be mediated through the aryl hydrocarbon receptor (AhR), our initial experiments involved measuring levels of the AhR protein and transcript. In tumors arising in the carcinogen-treated mice, as we have previously seen in rats treated with a single dose of DMBA (Trombino et al., 2000), there was evidence of constitutive up-regulation of the AhR itself. Preliminary data identified frequent up-regulation of an AhR target gene, CYP1B1, although CYP1B1 levels did not always correlate with AhR levels in any given tumor (data not shown). Since it has been well established that CYP1B1 is involved in activation of DMBA to its genotoxic metabolites and in metabolism of estradiol into mutagenic intermediates, our data suggest that AhR up-regulation in tumors may contribute to ongoing genetic changes through CYP1B1 induction.

As a means to evaluate fundamental cell cycle regulators known to be important in mammary tumorigenesis, initial studies measured mRNA levels of the cyclin D1 and c-myc oncogenes, and levels of hyperphosphorylated Rb. Quantitative PCR results suggested that cyclin D1 mRNA is frequently elevated, and c-myc is occasionally elevated. Western blot analysis demonstrated increased amounts of phosphoRb and an increased proportion of Rb in the phosphorylated form. When we examined upstream regulatory pathways, we found evidence of constitutive NF-κB and β-catenin signaling.

The NF-κB pathway blocks apoptosis in many cancers, and we have demonstrated that constitutive expression of c-rel in the mammary gland of transgenic mice produces mammary tumors (Romieu-Mourez et al., 2003). In the DMBA-induced mammary tumors, there were functional nuclear NF-κB complexes that included the transactivators p65, c-Rel, RelB, and partners p50 and p52. The c-myc and cyclin D1 genes are transactivated by NF-κB heterodimeric complexes composed of p65/p50 and c-Rel/p50, but not by homodimers of p50 or p52. Thus, the relative composition of subunits regulates the activation of these genes, suggesting a possible mechanism for the observed up-regulation of cyclin D1 and c-myc oncogenes.

Our data also provide evidence of activation the Wnt pathway in the majority of PAH-induced mammary tumors. A second pathway capable of c-myc and cyclin D1 transactivation. Wnt signaling can be assessed by total levels of the co-transactivator β-catenin in cells and tissues, but particularly by measurement of nuclear levels. Both total and nuclear β-catenin levels were elevated in most of the tumors. Most tumors also had elevated protein levels of the positive kinase regulator, CK2. CK2 may be contributing to the increase in tumor β-catenin, as well as stabilizing other pro-oncogenic proteins. The majority of the tumors with significantly elevated levels of β-catenin (T1, T9, T11) also have histological characteristics associated with Wnt pathway mammary tumors (Rosner et al., 2002). Additionally, squamous cell carcinoma, seen histologically in the majority of the DMBA-induced mammary tumors, has been associated with Wnt-related tumors of the mammary gland (Rosner et al., 2002).

DMBA-induced mammary tumors also expressed elevated levels of Pin1, a prolyl isomerase that has been shown to be up-regulated in a variety of human cancers, and to function as a signaling integrator in the regulation of β-catenin (Pang et al., 2004; Ryo et al., 2001), cyclin D1 (Liou et al., 2002; Wulf et al., 2001), c-myc (Yeh et al., 2004), and NF-κB (Ryo et al., 2003). Consistent with these observations, a concurrent increase in Pin1, β-catenin, and Rb hyperphosphorylation was seen in many of the DMBA-induced mammary tumors (Table 2). Interestingly, while elevated cyclin D1 correlated well with hyperphosphorylated Rb in some tumors, in others it did not, suggesting that there may be activation of other non-cyclin D1-dependent Rb kinases or inhibition of Rb phosphatases in some of the DMBA-induced tumors. While previous studies have demonstrated an interaction between the AhR and Rb (Ge and Elferink, 1998; Puga et al., 2000) and shown that dioxin-induced cell cycle arrest may occur through an AhR-mediated down-regulation of cyclin D1 and concomitant increases in hypophosphorylated Rb (Barnes-Ellerbe et al., 2004), ours is the first to demonstrate in vivo up-regulation of these 3 factors following carcinogen treatment and suggests that in this system the AhR may be stimulating the cell cycle. This is consistent with in vitro observations made of mammary epithelial cells transfected with the AhR (Brooks and Eltom, 2005).

Our observation of activation of these known oncogenic pathways not only correlates well with previous data from our lab using Sprague–Dawley rats (Landesman-Bollag et al., 2001; Sovak et al., 1997; Trombino et al., 2000), but with other studies of carcinogen-induced mammary tumors in which gene expression patterns were analyzed by microarray. In particular, various studies describe up-regulation of cyclin D1 transcripts in DMBA and 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) induced mammary tumors (Kuramoto et al., 2002; Shan et al., 2002). QPCR analysis of Pin1 mRNA levels showed no significant difference between tumors and the age-matched normal controls (data not shown), suggesting that Pin1 up-regulation in DMBA-induced mammary tumors is likely due to a posttranscriptional mechanism. This observation is consistent with the lack of Pin-1 mRNA up-regulation identified by microarray. Posttranscriptional mechanisms may also contribute to the activation of β-catenin and NF-κB.

While the current studies do not determine whether the signaling cascades are activated sequentially or in parallel in the course of transformation, this can be tested in the future using transgenic mouse models. For instance, by treating AhR-null mice with DMBA by this protocol, we can assess the role of this receptor in the activation of the downstream mediators. Understanding the interaction of these signaling pathways may provide targets for novel therapeutics in the future. Furthermore, since many transgenic models exist in the FVB background in which oncogenes produce mammary tumorigenesis in a multistep stochastic fashion, these studies provide us with the capability of examining the interaction of carcinogens and genetic predisposition to breast cancer.