Nitric oxide (NO·), an important mediator of inflammation, and β-catenin, a component of the Wnt–adenomatous polyposis coli signaling pathway, contribute to the development of cancer. We have identified two T-cell factor 4 (Tcf-4)-binding elements (TBE1 and TBE2) in the promoter of human inducible NO synthase 2 (NOS2). We tested the hypothesis that β-catenin regulates human NOS2 gene. Mutation in either of the two TBE sites decreased the basal and cytokine-induced NOS2 promoter activity in different cell lines. The promoter activity was significantly reduced when both TBE1 and TBE2 sites were mutated (P < 0.01). Nuclear extract from HCT116, HepG2, or DLD1 cells bound to NOS2 TBE1 or TBE2 oligonucleotides in electrophoretic mobility shift assays and the specific protein-DNA complexes were supershifted with anti-β-catenin or anti-Tcf-4 antibody. Overexpression of β-catenin and Tcf-4 significantly increased both basal and cytokine-induced NOS2 promoter activity (P < 0.01), and the induction was dependent on intact TBE sites. Overexpression of β-catenin or Tcf-4 increased NOS2 mRNA and protein expression in HCT116 cells. Lithium chloride (LiCl), an inhibitor of glycogen synthase kinase-3β, increased cytosolic and nuclear β-catenin level, NOS2 expression, and NO· production in primary human and rat hepatocytes and cancer cell lines. Treatment with Wnt-3A-conditioned medium increased β-catenin and NOS2 expression in fetal human hepatocytes. When administered in vivo, LiCl increased hepatic β-catenin level in a dose-dependent manner with simultaneous increase in NOS2 expression. These data are consistent with the hypothesis that β-catenin up-regulates NOS2 and suggest a novel mechanism by which the Wnt/β-catenin signaling pathway may contribute to cancer by increasing NO· production. (Cancer Res 2006; 66(14): 7024-31)
Nitric oxide (NO·) is a diverse biological mediator with important roles in vascular biology, inflammation, and cancer (1–4). Increased expression of nitric oxide synthase 2 (NOS2) is reported in a variety of different cancers as well as in the cancer-prone oxyradical overload diseases (e.g., ulcerative colitis, viral hepatitis, hemochromatosis, and Wilson disease; reviewed in ref. 4). These observations have led to strategies of chemoprevention using NOS2 inhibitors to block experimental tumor growth (5, 6). Furthermore, NO· can cause post-translational modification in proteins involved in cell cycle, DNA repair, and apoptosis (7, 8). The expression of NOS2 is regulated by inflammatory cytokines predominantly at the transcriptional level (9–14). However, post-transcriptional mechanisms have also been shown (15–17). We have earlier reported p53-mediated transrepression of NOS2 and existence of a negative feedback loop in vitro and in vivo (18, 19).
β-Catenin is an important component of the Wnt signaling pathway and is involved in diverse cellular processes, including cell adhesion, growth, differentiation, and transcription of Wnt-responsive genes (20–22). Alteration in Wnt/β-catenin signaling can contribute to the development of cancer (23–25). Interestingly, a quite unique function of β-catenin is described recently in the activation of FOXO transcription factor to promote cellular dormance under oxidative stress (26, 27). In the absence of Wnt, intracellular β-catenin is either bound to cadherins at cell adhesion junctions or, if free in the cytosol, rapidly degraded by a multiprotein complex consisting of β-catenin, glycogen synthase kinase-3β (GSK-3β), adenomatous polyposis coli (APC), and Axin (20). In the presence of Wnt binding to the cell surface receptor, GSK-3β is inactivated, thereby releasing β-catenin that translocates to the nucleus, binds to T-cell factor 4 (Tcf-4)/lymphocyte enhancer factor, and targets Wnt-responsive genes, including c-myc, cyclin D1, peroxisome proliferator-activated receptor-δ, and cyclooxygenase-2 (COX-2; 21, 28). Based on the evidence that β-catenin regulates genes involved in carcinogenesis, including those associated with inflammation, and that NOS2 is overexpressed in a variety of human cancers, we studied the involvement of NOS2 in Wnt signaling pathway and tested the hypothesis that β-catenin regulates NOS2 using both in vitro and in vivo models.
Materials and Methods
Cell lines and reagents. The human colorectal cancer cell lines DLD1 and HCT116, lung cancer cell line A549, 293 embryonic kidney cells, and hepatoblastoma cell line HepG2 were obtained from the American Type Culture Collection (Rockville, MD) and cultured at 37°C in 5% CO2 in medium containing 10% fetal bovine serum (Clontech, Mountain View, CA), 100 units/mL penicillin, 100 μg/mL streptomycin, and 15 mmol/L HEPES (pH 7.4). Unless indicated, cells were stimulated with a cytokine mixture consisting of 1,000 units/mL human tumor necrosis factor-α (TNF-α; R&D Systems, Minneapolis, MN), 100 units/mL interleukin (IL)-1β (provided by C. Reynolds, National Cancer Institute, NIH, Bethesda, MD), and human 250 units/mL IFN-γ (R&D Systems or Roche Pharmaceuticals, Nutley, NJ), which were purified recombinant proteins.
Hepatocyte isolation and culture. Human hepatocytes were isolated from histologically normal liver and were kindly provided by Drs. Steven Strom and Ken Dorko (University of Pittsburgh Core Pathology Facility, Pittsburgh, PA) according to an Institutional Review Board–approved protocol. Human hepatocytes were prepared by a three-step collagenase perfusion technique (29). Detail is provided as Supplementary information.
Plasmid constructs. The human NOS2 promoter reporter plasmid piNOS(7.2)Luc contains −7.2 kb of upstream 5′-flanking DNA linked to the luciferase reporter gene and has been described earlier (12, 14). Mutations of the −7.2 kb Tcf-4-binding elements (TBE) were generated from the piNOS(7.2)Luc reporter plasmid by using the QuikChange mutagenesis kit according to the manufacturer's recommendations (Stratagene, La Jolla, CA; Table 1). The reporter plasmids pTOP-FLASH and pFOP-FLASH, β-catenin expression vector, and Tcf-4 and dominant-negative Tcf-4 (ΔNTcf-4) expression vectors were kindly provided by Bert Vogelstein (The Johns Hopkins University Medical Institutions, Baltimore, MD; refs. 23, 24). Wnt-3A-conditioned medium was kindly provided by Dr. Satdarshan P.S. Monga (University of Pittsburgh; ref. 30).
|Oligo/plasmid .||Position .||Sequence (5′ to 3′) .|
|WT||−3,839 to −3,833||GGTGCTCTTTGATTTCTCCC|
|Mutation||−3,839 to −3,833||GGTGCTCTTTGGCTTCTCCC|
|WT||−6,136 to −6,129||TCCAAGAGCATCAAAGACCA|
|Mutation||−6,136 to −6,129||TCCAAGAGCGCCAAAGACCA|
|Oligo/plasmid .||Position .||Sequence (5′ to 3′) .|
|WT||−3,839 to −3,833||GGTGCTCTTTGATTTCTCCC|
|Mutation||−3,839 to −3,833||GGTGCTCTTTGGCTTCTCCC|
|WT||−6,136 to −6,129||TCCAAGAGCATCAAAGACCA|
|Mutation||−6,136 to −6,129||TCCAAGAGCGCCAAAGACCA|
NOTE: Binding sites are shown in bold type. Mutations are italicized.
Transient transfections and activity assays. DNA transfections of cells were carried out in six-well plates (Corning, Corning, NY) by using LipofectAMINE plus (Invitrogen, Carlsbad, CA) and Mirus TransIT reagent (Mirus, Madison, WI). The experimental details are provided as Supplementary information.
Western blot analysis. SDS-PAGE was done according to the method of Towbin et al. (31). The following specific antibodies were used for immunodetection with appropriate dilutions: rabbit anti-NOS2 polyclonal (BD Biosciences, San Jose, CA), rabbit anti-β-catenin polyclonal (Santa Cruz Biotechnology, Santa Cruz, CA), and mouse anti-β-actin monoclonal (Sigma, St. Louis, MO).
Preparation of nuclear protein and electrophoretic mobility shift assay. Preparation of nuclear extract from HepG2 cells and electrophoretic mobility shift assay (EMSA) were done as described earlier (32). The details are provided as Supplementary information.
Chromatin immunoprecipitation assay. Chromatin immunoprecipitation (ChIP) was done following the introduction manual of ChIP-IT kit (Active Motif, Carlsbad, CA) and published protocols (33). The details are provided as Supplementary information.
NO· production assessment. Culture supernatants were collected and assayed for nitrite, the stable end products of NO· oxidation, using the Greiss reaction as described (34).
Animals. Inbred male Lewis (LEW) rats weighing 300 to 350 g were obtained from Harlan Sprague Dawley, Inc. (Indianapolis, IN). Animals were maintained in laminar flow cages in a specific pathogen-free animal facility at the University of Pittsburgh with a standard diet and water ad libitum. Animal protocols were approved by the Animal Care and Use Committee of the University of Pittsburgh, and the experiments were done according to the guidelines of the Council on Animal Care at the University of Pittsburgh and the National Research Council's Guide for the Humane Care and Use of Laboratory Animals. Briefly, rats were injected with lithium chloride (LiCl) through the tail vein at low dose of 4.3 mg/kg and high dose of 12.9 mg/kg body weight every 12 hours for 2 days. Control rats were injected with the same volume of vehicle (0.9% saline solution). Forty-eight hours following LiCl injection, rats were sacrificed and livers were removed.
Immunofluorescent staining for NOS2 and β-catenin. Immunofluorescent staining was done as described (35). Detail optimized protocol is described as Supplementary information.
Statistical methods. Data are the mean ± SD. Data were analyzed by the Student's t test or ANOVA, and P < 0.05 was considered statistically significant.
Two TBEs are identified in the NOS2 promoter. Sequence analysis of the 5′-flanking sequence of the NOS2 gene identified two putative TBEs (designated as TBE1 and TBE2) upstream in the NOS2 promoter. TBE1 (CTTTGAT) is located −3.8 kb upstream of the TATA box and perfectly matched the consensus CTTTG(A/T)(A/T) sequence for Tcf-4 binding (36), whereas TBE2 was located −6.1 kb upstream of the TATA box and contained an inverted perfect match (ATCAAAG; Table 1; Supplementary Fig. S7).
Mutation of the TBE abrogates basal and cytokine-induced NOS2 promoter activity. To determine the role of cis-acting TBE motifs in regulating NOS2 gene transcription, site-directed mutation of one or both (double mutant) TBE sites was generated in the context of a −7.2 kb NOS2 luciferase promoter plasmid (Supplementary Fig. S7). Mutation in either of the TBE sites decreased basal and cytokine-induced NOS2 promoter activity in DLD1 and HCT116 cell lines (P < 0.05 and P < 0.01; Fig. 1). However, mutation of both TBE1 and TBE2 sites markedly decreased basal and cytokine-stimulated NOS2 promoter activity (P < 0.01; Fig. 1). Similar results were observed in human AKN-1 liver and A549 lung cells (data not shown).
β-catenin/Tcf-4 specifically binds to the TBEs of the NOS2 promoter in vitro and in vivo. To establish that β-catenin and Tcf-4 nuclear proteins can bind to the NOS2 promoter-specific TBE DNA elements at −3.8 and −6.1 kb, gel shift assays were done using nuclear extracts from several human colorectal cancer cell lines or HepG2 liver cancer cell line. Nuclear extracts from HepG2 cells showed strong protein binding to the wild-type (WT) TBE1 oligonucleotide (Fig. 2A,, lanes 1-7) and slightly weaker binding to the TBE2 oligonucleotide (Fig. 2B,, lanes 1-7). Specificity for binding was confirmed by cold competition with excess unlabeled WT TBE oligonucleotides (Fig. 2A, and B, lane 2); however, cold mutant TBE oligonucleotides did not show any effect (Fig. 2A, and B, lane 3). Supershift of the complexes with anti-Tcf-4 or anti-β-catenin antibodies confirms the presence of Tcf-4 and/or β-catenin in the protein-DNA complex (Fig. 2A, and B, lanes 4-6). A control antibody against E-cadherin failed to elicit a supershift (Fig. 2A, and B, lane 7), supporting the notion that the protein-DNA complex was not contaminated by cytoplasmic β-catenin bound to E-cadherin. WT oligonucleotide without any nuclear extract served as negative control (Fig. 2A , lane 8). Additional EMSA experiments using glutathione S-transferase (GST)-Tcf-4 fusion protein showed specific protein-DNA binding to NOS2 TBE1 or TBE2, and the complex was supershifted by anti-GST antibody (Supplementary Fig. S8A-C), corroborating the binding observed with the nuclear protein extracts.
We then did ChIP assay in HepG2 cells to confirm in vivo binding of β-catenin/Tcf-4 to the NOS2 promoter. The DNA-β-catenin/Tcf-4 complexes were immunoprecipitated with antibodies shown in Fig. 2C followed by reversal of cross-linking and PCR amplification using primers flanking the TBE1 (−3.8 kb) and TBE2 (−6.1 kb) and nuclear factor-κB (NF-κB)/signal transducers and activators of transcription 1 (STAT1; −5.8 to −5.2 kb) sites in the NOS2 promoter (Fig. 2C). Agarose gel analysis of PCR products showed that the TBE1 or TBE2 containing DNA was immunoprecipitated by anti-β-catenin (Fig. 2C,, lane B) or anti-Tcf-4 (Fig. 2C,, lane T) antibodies, and only a weak background was observed by IgG (Fig. 2C,, lane I). Transcription factor IIB (TFIIB; Fig. 2C,, lane F) antibodies were used as positive control (Fig. 2C). Input lane (Fig. 2C,, lane In) is shown as binding to HepG2 genomic DNA used as template for PCR. As an additional negative control to show that the anti-β-catenin or anti-Tcf-4 antibodies were not binding in a nonspecific manner, we did ChIP assay using PCR primers spanning 600 bp adjacent promoter sequence at −5.8 to −5.2 kb in the NOS2 promoter. This region does not contain TBE1 or TBE2 sites but contains active NF-κB (−5.8 kb) and STAT1 (−5.8 and −5.2 kb) sites that we recently documented to bind NF-κB and STAT1 proteins by in vivo ChIP assay (37). As expected, this region did not show any binding with anti-β-catenin (Fig. 2C,, lane B) or anti-Tcf-4 (Fig. 2C,, lane T) antibodies. IgG (Fig. 2C,, lane I) and TFIIB (Fig. 2C , lane F) antibodies were used as negative and positive controls, respectively, for binding. These evidences indicate that β-catenin/Tcf-4 specifically binds to the NOS2 promoter TBE1 and TBE2 regions in vivo.
Overexpression of β-catenin/Tcf-4 increased basal NOS2 promoter activity. We determined if overexpression of β-catenin/Tcf-4 induced basal NOS2 transcription. Cotransfection of β-catenin and Tcf-4 expression plasmids with the WT −7.2 kb NOS2 promoter (containing both TBE1 and TBE2 sites) elicited a 4- to 5-fold increase in the basal promoter activity compared with empty vector control (P < 0.001) in 293 embryonic kidney cells, and this induction was significantly decreased by addition of a ΔNTcf-4 expression plasmid (Fig. 3A). Overexpression of β-catenin and Tcf-4 still increased basal NOS2 promoter activity when one of the two TBE sites was deleted; however, the promoter activity was almost completely abrogated when both TBEs were deleted (Fig. 3A). Cotransfection with pFOP or pTOP reporter plasmids (23, 24) along with β-catenin and Tcf-4 expression plasmids served as negative and positive controls, respectively (Fig. 3A).
To further examine if the increased NOS2 promoter activity driven by β-catenin overexpression required intact TBE motifs, additional cotransfection experiments were done using the −7.2 kb NOS2 promoter harboring the mutant TBE sites. Cotransfection of β-catenin and the WT −7.2 kb NOS2 promoter displayed ∼6-fold increase in basal transcriptional activity in 293 embryonic kidney cells (P < 0.001), and this required intact TBE1 or TBE2 sites for full induction (Fig. 3B). Furthermore, coexpression of ΔNTcf-4 significantly decreased the promoter activity (P < 0.001). Overexpression of Tcf-4 also increased basal NOS2 promoter activity in HepG2 cells in a dose-dependent manner (Fig. 3C). These findings indicate that binding of β-catenin/Tcf-4 complex to TBE leads to the basal up-regulation of NOS2 gene transcription and is consistent with a similar role of TBEs in constitutive expression of the COX-2, c-myc, and IL-8 genes (28, 36, 38). Overexpression of β-catenin and Tcf-4 also increased basal NOS2 promoter activity in A549 lung cancer cell line (Fig. 3D , column 2).
Overexpression of β-catenin/Tcf-4 increased cytokine-induced NOS2 promoter activity. To determine if overexpression of β-catenin/Tcf-4 also affected cytokine-induced NOS2 promoter activity, cotransfections were done in the presence of cytokine stimulation in A549 lung cancer cell line. As expected, the cytokine combination of TNF-α plus IL-1β plus IFN-γ (cytokine mixture) induced a 4-fold increase in NOS2 promoter activity (P < 0.01), and this was further increased ∼2-fold by overexpression of β-catenin and Tcf-4 (Fig. 3D). However, cotransfection of ΔNTcf-4 abrogated this increase in promoter activity. When IFN-γ was used alone as a suboptimal NOS2 promoter stimulus, overexpression of β-catenin and Tcf-4 increased promoter activity ∼10-fold (Supplementary Fig. S9), which was partially decreased by the overexpression of ΔNTcf-4 (Supplementary Fig. S9).
Overexpression of β-catenin and Tcf-4 increased endogenous NOS2 mRNA and protein expression. To determine that the effects of β-catenin/Tcf-4 overexpression on NOS2 promoter activity yield meaningful results for NOS2 transcription and translation, we evaluated the influence of β-catenin or Tcf-4 overexpression on endogenous NOS2 mRNA and protein expression in HCT116 cells (Fig. 4A). Transfection with control Tcf-4 (Fig. 4A,, lane 1) or β-catenin (Fig. 4A,, lane 2) empty vectors did not induce NOS2 mRNA or protein, whereas induction of both endogenous mRNA and protein was observed with overexpression of Tcf-4 (Fig. 4A,, lane 3) or β-catenin (Fig. 4A,, lane 4). Cytokine mixture–induced NOS2 mRNA and protein are shown as positive control (Fig. 4A , lane 5). These results indicate that the endogenous NOS2 gene transcription and subsequent translation are regulated by β-catenin and Tcf-4 overexpression.
Transfection of ΔNTcf-4 decreases endogenous NOS2 mRNA and protein expression. To show that β-catenin signaling actually influences endogenous NOS2 expression, HCT116 cells were transfected with either empty (Fig. 4B,, lane 2) or ΔNTcf-4 expression (Fig. 4B,, lane 3) vectors. To better visualize basal NOS2 expression, lanes were loaded with twice the amount of RNA [reverse transcription-PCR (RT-PCR)] or protein (Western blot) than what is used in Fig. 4A. Transfection of ΔNTcf-4 decreased both basal endogenous NOS2 mRNA and protein expression (Fig. 4B). These data provide further evidence of β-catenin-mediated regulation of endogenous NOS2.
LiCl increased spontaneous NOS2 protein expression. GSK-3β, a component of Wnt β-catenin signaling pathway, plays an important role in phosphorylation of β-catenin, leading to its ubiquitination and degradation by the proteasomal pathway (39). To further strengthen our results, we determined if blockade of GSK-3β would affect NOS2 expression. We used LiCl, a widely used inhibitor of GSK-3β (38), to abrogate β-catenin phosphorylation and allow its stabilization and accumulation. Primary human hepatocytes were incubated with 20 mmol/L LiCl for 1 to 24 hours. We found that treatment with LiCl increased the NOS2 protein expression in primary human hepatocytes in a time-dependent manner (Fig. 5A). Protein lysate from cytokine mixture–treated cells (TNF-α plus IL-1β plus IFN-γ) was used as positive controls (Fig. 5A,, lane 2). To document that LiCl was modulating β-catenin levels in the human hepatocytes, Western blot for β-catenin was done using nuclear protein. As expected, LiCl increased nuclear β-catenin accumulation in a time-dependent manner (Fig. 5A). Interestingly, LiCl induced NOS2 expression relatively early compared with its effect on IL-8 protein expression that was seen 48 hours after LiCl exposure (38). LiCl treatment also enhanced the cytosolic and nuclear expression of β-catenin and NOS2 in the rat hepatocytes as determined by immunofluorescence (Supplementary Fig. S10). However, untreated rat hepatocytes showed predominantly membrane-bound β-catenin without any detectable NOS2 expression.
LiCl increased nitrite production via inhibition of GSK-3β in cancer cell lines. We also determined NO· production by measuring nitrite level following LiCl treatment in HepG2 cells. Nitrite level was increased in a time-dependent manner following exposure to LiCl (Fig. 5B). The increase in nitrite production was relatively modest compared with nitrite levels induced by cytokines (34). Similar results for nitrite synthesis in response to LiCl were also obtained from cancer cell lines HuH7, SW480, and DLD1 cells (data not shown).
LiCl induced hepatic NOS2 expression in vivo. To test our hypothesis in vivo, LiCl was injected into rats, and β-catenin and NOS2 expressions were examined in liver by immunofluorescence (Fig. 5C). β-Catenin (FITC; Fig. 5C,, green) was localized predominantly on the hepatocellular membrane in the absence of LiCl. Furthermore, NOS2 protein (Cy3; Fig. 5C,, red) was weakly expressed in cytosol in the absence of LiCl. Injection of the higher dose of LiCl (12.9 mg/kg body weight) increased both the cytoplasmic and the nuclear β-catenin and cytoplasmic NOS2 protein expression. The nucleus was counterstained with Hoechst dye (Fig. 5C , blue).
Wnt-3A protein increases β-catenin and NOS2 protein expression in fetal human hepatocytes. To provide further evidence that the Wnt/β-catenin pathway can induce NOS2 expression in vivo, Wnt-3A-conditioned medium (30) was used to stimulate primary human fetal hepatocytes. The Wnt-3A-conditioned medium increased intracellular expression of β-catenin and NOS2 proteins as determined by immunofluorescence (Fig. 6).
In the present study, we have provided strong evidence that NOS2 is a novel downstream target of β-catenin, suggesting the involvement of NO· in Wnt/β-catenin signaling pathway that could contribute to human malignancy. Wnt/β-catenin signaling pathway and NO· are both implicated in the development of human cancer (4, 24). Based on the evidence provided in the present study, the addition of NOS2, as one of the targets of β-catenin/Tcf-4 complex, further emphasizes the importance and complexity of the Wnt pathway in physiologic and pathophysiologic processes, including carcinogenesis. Whereas NO· is an important molecule with diverse functions, including immune modulation, neurotransmission, vasodilation, and inflammation, a sustained and increased level of NO· can be deleterious and can contribute to the development of cancer (4, 40). Expression of NOS2 has been shown in human colon polyps, suggesting that NOS2 may contribute to the transition from benign adenoma to carcinoma in situ (41). Furthermore, NO· can form nitrosamines (42), and the observation that NOS2 is expressed in ulcerative colitis in humans (43, 44) has led to the hypothesis that chronic NOS2 expression contributes to direct damage and long-term colon cancer development in chronic inflammatory oxyradical overload diseases. A similar hypothesis has been speculated for NOS2 expression seen in liver biopsies from patients with chronic viral hepatitis (45), as this condition is a major risk factor for the development of hepatocellular carcinoma.
The tumor-suppressive function of APC in colon cancer is mediated by its ability to bind β-catenin, which favors the formation of a complex with GSK-3β, leading to the degradation of β-catenin (23, 24). Stabilization of β-catenin leads to its translocation into nucleus and subsequent activation of several target genes by β-catenin/Tcf-4 complex (a comprehensive list of target genes are available on http://www.stanford.edu/~rnusse/pathways/targets.html). Activating mutations in the regulatory motifs of β-catenin is described in many human cancers (reviewed in ref. 21). Transactivation of genes by β-catenin/Tcf-4 complex involves its binding to a consensus TBE in the promoter region (22, 36) of the target genes. Consistent with these observations, in the present study, we found two TBE sites, TBE1 and TBE2, in human NOS2 promoter that are a seven of seven nucleotide match to the core consensus TBE response element. We and others have earlier reported that NOS2 promoter contains functionally active consensus sequences for NF-κB, AP1, STAT1, AABS, NRE, KLF6, and other cis-acting DNA elements (12–14, 32, 33, 46–48). These cis-acting DNA elements and their cognate transcription factors span more than −10 kb upstream in the 5′-flanking sequence, indicating that the human NOS2 promoter is under complex control. The identification of active TBE sites in the present study further expands the critically important mechanisms in place that regulate NOS2 transcription.
In the present study, we have described the role of Wnt/β-catenin signaling in the regulation of NOS2 in colorectal, hepatocellular, and lung cancer cell lines as well as primary human and rodent hepatocytes and animal model. Site-directed mutagenesis provided direct evidence of a functional role of the TBE sites of NOS2 promoter in the regulation of NOS2 expression. EMSA and ChIP assays confirmed binding of β-catenin/Tcf-4 proteins to the TBE motifs in the promoter of NOS2. Furthermore, overexpression of β-catenin-induced NOS2 promoter activity, which was dependent on intact TBE sites, also led to an increased NOS2 expression followed by an increase in the synthesis of NO·. Cotransfection of a ΔNTcf-4 partially abrogated basal NOS2 promoter activity as well as endogenous NOS2 mRNA and protein expression, which further confirms the specificity of this pathway in the regulation of NOS2. Further evidence, suggesting the involvement of Wnt signaling in NOS2 expression in our study, comes from treating the primary human fetal hepatocytes with Wnt-3A-conditioned medium, which caused an increase in both β-catenin and NOS2 expressions.
Serine-threonine kinase, GSK-3β, is a critical component in the maintenance of β-catenin pool in the cytoplasm. GSK-3β phosphorylates several specific serine and threonine residues in the NH2-terminal region of β-catenin and is essential for subsequent ubiquitin-mediated proteasomal degradation. Inactivation of GSK-3β leads to the accumulation of β-catenin in the cytoplasm. Therefore, we analyzed β-catenin-induced alterations of gene expression in primary human hepatocytes using LiCl, a widely used inhibitor of GSK-3β, which induces strong accumulation of nonphosphorylated β-catenin (38). Blockade of GSK-3β resulted in accumulation of β-catenin and spontaneous increase in NOS2 expression in primary hepatocytes, hepatic tissue in vivo, and cultured cell lines. However, despite the wide-scale use of LiCl as an inhibitor of GSK-3β, it is valid to recognize the limitations in its specificity and potential effects on NOS2 signaling independent of β-catenin. Our results are consistent with a recent report showing enhanced NOS2 expression in a trophoblast cell line in response to kenpaullone, another GSK-3β inhibitor (49).
The findings in the present study, describing the up-regulation of NOS2 by β-catenin, further expand the existing and rather extensive regulatory network of Wnt signaling pathway. Together with our earlier report, showing the β-catenin-mediated up-regulation of COX-2 and subsequent production of prostaglandin, the present results indicate a possible role of Wnt/β-catenin pathway in inflammation and inflammation-associated cancer (28). Given the wide-ranging function of NO·, Wnt/β-catenin pathway seems to be involved in many more important physiologic and pathophysiologic processes than earlier conceived. The continuously growing evidence, supporting the involvement of NO· in the cancer development, further strengthen the tumorigenic contributions of β-catenin and thereby establishes a critical Wnt/β-catenin/NO· pathway with important implications in carcinogenesis. NO·, being a critical mediator of inflammation with its involvement in cancer development, future studies are warranted to study the role of Wnt/β-catenin/NO pathway in inflammation-induced cancer.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
Grant support: NIH Intramural Research Program, National Cancer Institute, Center for Cancer Research, and NIH grant R01-GM52021 (D.A. Geller).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Dr. Curtis Harris for his critical evaluation of the article, Karen Macphearson for assistance with reference manager, Dr. Satdarshan P.S. Monga for providing Wnt-3A and control medium, and Drs. Stephen Strom, Ken Dorko, and Hongbo Cai for kindly providing cultures of primary fetal human hepatocytes.