Mucin alterations are a common feature of colonic neoplasia, and alterations in MUC2 mucin have been associated with tumor progression in the colon. Bile acids have been linked to colorectal carcinogenesis and mucin secretion, but their effects on mucin gene expression in human colon carcinoma cells is unknown
Human colon carcinoma cells were treated ≤ 6 hours with 10–200 μM deoxycholate, chenodeoxycholate, or ursodeoxycholate. MUC2 protein was assayed by Western blot analysis and MUC2 transcription was assayed using a MUC2 promoter reporter luciferase construct. Transcription activator protein 1 (AP-1) activity was measured using an AP-1 reporter construct and confirmed by Western blot analysis for c-Jun/AP-1.
MUC2 transcription and MUC2 protein expression were increased three to fourfold by bile acids in a time and dose-dependent manner with no effect on cell viability. AP-1 activity was also increased (deoxycholate > chenodeoxycholate > ursodeoxycholate). Treatment with the putative chemopreventive agent curcumin, which decreased AP-1 activity, also decreased MUC2 transcription. Cotransfection with a dominant negative AP-1 vector decreased MUC2 transcription, confirming the significance of AP-1 in MUC2 induction by deoxycholate. Calphostin C, a specific inhibitor of protein kinase C (PKC), greatly decreased bile acid-induced MUC2 transcription and AP-1 activity, whereas inhibitors of MAP kinase had no effect.
Epithelial surfaces, including the colonic epithelium, are protected by mucins, large carbohydrate-rich glycoproteins that are the major secreted glycoproteins of the gastrointestinal tract. In addition to their protective function in the normal colon, quantitative and qualitative alterations in mucins are a common feature of colonic neoplasia.1 MUC2, a high molecular weight glycoprotein with an O-linked carbohydrate, is the major secreted mucin in the large and small intestine.2, 3 Human colon carcinomas and cell lines derived from tumors can differ significantly in the amount of MUC2 mucin synthesized and these differences correlate with altered biochemical and biologic properties including those with relevance to invasion and metastasis.4–6 MUC2 is strongly expressed in mucinous carcinomas, and patients with mucinous colorectal carcinomas characteristically present with advanced-stage disease. Relatively little is known, however, about the mechanisms responsible for regulation of MUC2 expression in the colon.
Bile acids, amphophilic derivatives of cholesterol, have been reported to promote colorectal carcinogenesis.7 The high fat intake typical part of the Western diet is associated with increased levels of fecal bile acids and a higher incidence of colorectal carcinomas.8–10 Primary bile acids produced in the liver are metabolized by colonic bacteria to produce secondary bile acids. These secondary bile acids, primarily deoxycholic acid (DCA), are cytotoxic to colonocytes and are established tumor promoters in animal models.7, 11 Bile acids also have been reported to stimulate invasion and metastasis of colon carcinoma cells12, 13 via activation of multiple signaling pathways. Previous reports have indicated that bile acids can modulate mucin secretion in colon carcinomas,14, 15 but the effects of bile acids on mucin gene expression in human colon carcinoma cells are unknown.
In the current study, we sought to determine the effects of bile acids on MUC2 expression in colon carcinoma cells and the molecular mechanisms involved. We find that bile acids induce MUC2 expression in human colon carcinoma cells at the level of transcription through a process that involves primarily a MAP kinase-independent, protein kinase C (PKC)-dependent activation of transcription activator protein 1 (AP-1).
MATERIALS AND METHODS
DCA, chenodeoxycholic acid (CDCA), ursodeoxycholic acid (UDCA), and phorbol 12-myristate 13-acetate (PMA) were obtained from Sigma (St. Louis, MO). Calphostin C, PD98059 (2′-amino-3′-methoxyflavone), U0126 (1,4-diamino-2, 3-dicyano-1,4-bis(2-aminophenylthio)butadiene), and H-8 (N-[2-(methylamino)ethyl]-5-isoquinolinesulfonamide dihydrochloride) were purchased from Calbiochem (San Diego, CA). Mouse monoclonal antibody (MoAb) CCP-58, specific for MUC2 glycoprotein, was obtained from Novocastra (Newcastle, UK). Antibodies for extracellular signal-regulated kinase (ERK1/2) and phospho-ERK1/2 were obtained from Cell Signaling Technology (Beverly, MA), anti-c-jun/AP-1 was obtained from Santa Cruz Biotechnology (Santa Cruz, CA), and anti-beta-actin MoAb was obtained from Sigma. Secondary antibodies and the enhanced chemiluminescence (ECL) Western blotting detection system were from Amersham Pharmacia Biotech (Piscataway, NJ). FuGENE 6 transfection reagent was from Roche (Indianapolis, IN). Precast 3–8% tris-acetate gradient gels were purchased from Invitrogen (Carlsbad, CA).
Cell Culture and Treatment
LiM6, a derivative of the LS174T colon carcinoma cell line that has high liver-metastasizing ability in vivo,4 was maintained at 37 °C in 5% CO2 atmosphere in Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal bovine serum (FBS) with penicillin and streptomycin.
LiM6 cells were plated in regular medium for 24 hours. The medium was then replaced with 0.5% FBS for an additional 24 hours. Cultures were then treated with bile acids or PMA. For inhibitor assays, LiM6 cells were pretreated with inhibitors for 1 hour before exposure to DCA for an additional 16 hours. Calphostin C was used under a fluorescent lamp of 13 W located 15 cm above the plates.
For determining the effects of bile acids on viability, cells were treated for 24 hours with ≤ 100 μM DCA, CDCA, or UDCA, then labeled for 3 hours with MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, purchased from Promega, Madison, WI) and tetrazolium reduction was quantified by measuring absorbance at 490 nm.
Protein Extraction and Western Blotting
Cellular proteins from treated LiM6 cells were prepared in 30 mM Tris-HCI, pH 6.8, 150 mM NaCl, 2 mM ethylenediaminetetraacetic acid, 100 mM sodium fluoride, 10 mM sodium pyrophosphate, 2 mM orthovanadate, 1% Triton X-100, 1% Nonidert P40 (NP-40), 0.2 mM phenylmethanesulfonyl fluoride, and 1 mini tablet protein inhibitor (Boehringer Mannheim, Indianapolis, IN). The protein concentration of the supernatant was determined by using the bicinchoninic acid-based BCA protein assay kit (Pierce, Rockford, IL). Equal amounts of protein were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis on either 3–8% Tris-acetate gradient gels for MUC2 detection or 10% Tris-glycine gels for detection of other proteins. After gel electrophoresis and transfer to nitrocellulose, the membranes were stained in 0.5% Ponceau S with 1% acetic acid to confirm the equal loading and transfer efficiency. Membranes were incubated at 4 °C overnight in a blocking solution containing 5% bovine skim milk and 0.1% Tween 20 (Fischer Scientific, Pittsburgh, PA) in TBS (10 mM Tris-HCl with 150 mM NaCl, pH 7.6), then probed with specific primary and secondary antibodies conjugated to horseradish peroxidase. Immunoreactive bands were visualized by chemiluminescence solution and exposure to X-ray film.
Plasmids, Transient Transfection, and Luciferase Reporter Assays
Methods to measure the MUC2 promoter activity with luciferase as a reporter have been reported previously.16, 17 Upstream fragments of 2665 base pair (bp) from immediately adjacent to the 5′ translation start site of human MUC2 (GenBank accession number U67167)16 were generated by routine polymerase chain reaction from human genomic DNA, using the following primer pairs (forward: 5′-GAGGCTAGCCCGGGCTTCCTGGTGAGTC-3′, and reverse: 5′-GAGCTCGAGCATGGTGGCTGGCAGGGGC-3′). The 2665-bp fragment was then inserted upstream of the luciferase reporter in the pGL3-basic vector, according to the instructions provided by the manufacturer (Promega). A fragment of 2205 bp of the MUC2 promoter was prepared by digestion of the 2665-bp fragment of the MUC2 promoter luciferase reporter in the pGL3-basic vector with exonuclease (Erase-a-Base system; Promega). DNA sequencing was performed to verify the correct clone.
Plasmids were prepared using the Genopure plasmid midi kit from Roche (Indianapolis, IN). Dominant-negative AP-1 (blunted TAM67)18 and the AP-1 promoter-luciferase reporter19 were kindly provided by Shrikanth A. Reddy and Jonathan Kurie of The University of Texas M. D. Anderson Cancer Center (Houston, TX). Beta-galactosidase expression vector pCH110, used as an internal control for transfection efficiency, was purchased from Amersham Pharmacia Biotech.
For transient transfection, LiM6 cells were seeded at a concentration of 4 × 105 cells per well in 6-well plates. After overnight culture, the cells in each well were transfected with DNA (1 μg of MUC2 or AP-1-luciferase reporter plasmid and 0.2 μg of pCH110) by using 3 μL of FuGENE6 (Roche) by following the manufacturer's protocol. After a 24-hour exposure to the transfection mixture, the cells were incubated in medium containing 10% FBS and different concentrations of DCA or other bile acids or inhibitors for an additional 16 hours and then harvested for measurement of beta-galactosidase activity and luciferase activity. The latter was measured by using the Promega luciferase assay system according to the manufacturer's protocol (technical manual TM033) using a TD-20/20 luminometer (Turner Designs; Sunnyvale, CA). Luciferase was normalized to the beta-galactosidase activity to account for differences in transfection efficiency. All experiments were performed in triplicate and repeated at least twice.
Effects of Bile Acids on MUC2 Expression and Transcription
To determine whether MUC2 glycoprotein is induced by bile acids in human colon carcinoma cells, we evaluated MUC2 protein levels in the colon carcinoma cell line LiM6 treated with DCA, CDCA, and UDCA. Viability of LiM6 cells was not significantly affected by treatment for 24 hours with 100 μM DCA (treated/control = 1.00 ± 0.06 [mean ± the standard error of the mean], n = 3), CDCA (0.99 ± 0.03), or UDCA (0.99 ± 0.01). The addition of DCA in 10–200 μM for 6 hours to LiM6 increased MUC2 protein expression in a dose-dependent manner with a maximum increase of approximately 4-fold (Fig. 1). Different bile acids may have different biologic effects.20–22
To examine the effects of different bile acids on MUC2 induction, we treated LiM6 with DCA, CDCA, and UDCA for 6 hours at a concentration of 100 μM. As shown in Figure 1, all three bile acids increased MUC2 protein expression, with DCA having the strongest effect. In agreement with previous reports,23 100 ng/mL PMA also strongly induced MUC2 expression (data not shown).
To determine whether bile acids increase MUC2 gene transcription, we employed an MUC2 promoter luciferase construct, which contains a 2205-bp human MUC2 5′-flanking region fused to a luciferase reporter gene. After transient transfection, cells were treated with different concentrations of bile acids and luciferase activities were determined (Fig. 1). DCA induced MUC2 promoter-driven luciferase activities in a dose-dependent manner in LiM6 cells, with a maximum 2–3-fold increase with 200 μM DCA. All 3 tested bile acids induced MUC2 transcription activity to a varying extent at a concentration of 100 μM.
Requirement of Transcription Activator Protein 1 for Induction of MUC2 by Bile Acids
Previous reports have suggested that AP-1 transcription factor is important in mediating bile acid signaling.24–26 However, it is not known whether MUC2 induction by bile acids involves AP-1. We therefore assessed the ability of bile acids to induce AP-1 expression and transcriptional activation in LiM6 colon carcinoma cells. There was a dose-dependent expression of c-jun/AP-1 protein in LiM6 cells treated for 6 hours with 10–200 μM DCA (Fig. 2A, left), and an increased expression with 100 μM DCA, CDCA, or UDCA (Fig. 2). This coincided with the MUC2 induction by these bile acids (Fig. 1).
To test whether bile acid induction of c-jun/AP-1 protein expression was accompanied by functional activation of AP-1, a luciferase reporter gene driven by the c-jun/AP-1 promoter was transiently transfected into LiM6. After 24 hours, the transfected cells were treated with different bile acids (100 μM) and with increasing doses of DCA (Fig. 2). Luciferase activities from these cells were stimulated by all three bile acids tested, with DCA having the strongest effect (Fig. 2). DCA induced AP-1 transcription activity in a dose-dependent manner with a maximum 2–3-fold increase at a concentration of 200 μM, a result consistent with induction of MUC2 expression (Fig. 1).
To confirm the role of AP-1 in regulation of MUC2 induction by DCA, we cotransfected a MUC2 promoter construct (PM2PL-2205) and a cytomegalovirus (CMV)-driven dominant-negative AP-1 vector (pCMV-TAM67) into LiM6. Cells were then treated with 100 μM DCA for an additional 16 hours. Dominant-negative AP-1 blocked the induction of MUC2 transcription induced by deoxycholate and also decreased the basal level of MUC2 transcriptional activity (Fig. 3). This result demonstrates that AP-1 activity is necessary for the bile acid-dependent induction of MUC2.
To further investigate the requirement of AP-1 for bile acid induction of MUC2, we examined the effect of curcumin, a dietary pigment that is of interest as a potential chemopreventive agent.27 Although the precise mechanism by which curcumin inhibits colon tumorigenesis remains to be elucidated, it has been reported to inhibit c-Jun expression and to block AP-1 activation.28 In LiM6 colon carcinoma cells, the basal levels of MUC2 and AP-1 protein were decreased by treatment with 10 μM curcumin. There was also partial inhibition of the bile acid-dependent induction of MUC2 and AP-1 (Fig. 4). In assays of MUC2 transcription and AP-1 transcriptional activity, curcumin had relatively little effect on basal activity, but inhibited the bile acid-dependent induction of MUC2 and AP-1 activity (Fig. 4).
Involvement of Protein Kinase C in Induction of MUC2 by Bile Acids
Previous results have shown that the PKC inducer PMA increases MUC2 expression, and that this is prevented by the PKC inhibitor, calphostin C.23 It has been reported that bile acids activate PKC in colorectal carcinoma tissue specimens and cells.29, 30 However, it is unclear whether induction of MUC2 by bile acids is mediated by PKC. When LiM6 cells were treated with 5 nM calphostin C, there was almost complete inhibition of basal and bile acid-induced MUC2 protein and c-jun/AP-1 protein expression (Fig. 5). This was accompanied by suppression of MUC2 transcription and of AP-1 transcriptional activity (Fig. 5). These results indicate that PKC is involved in the AP-1–dependent induction of MUC2 expression by bile acids.
Because forskolin, a PKA activator, has been reported to stimulate MUC2 expression,31 we also examined the effect on MUC2 induction by bile acids of a less specific protein kinase inhibitor, H-8 (N-[2-(methylamino)ethyl]-5-isoquinolinesulfonamide dihydrochloride). Previous reports32 indicate that H-8 inhibits cyclic adenosine monophosphate-dependent protein kinase (PKA, Ki = 1.2 μM) to a greater extent than Ca++-phospholipid–dependent protein kinase (PKC, Ki = 14.4 μM). H-8 decreased levels of MUC2 protein and c-jun/AP-1 protein both in the presence and absence of bile acid (Fig. 6). There was a modest decrease in MUC2 transcription and in AP-1 transcriptional activity (Fig. 6), which was much less complete than that seen for calphostin C (Fig. 5). These results suggest that the bile acid-dependent induction of MUC2 depends less on PKA than on PKC.
MUC2 Induction by Bile Acids is Independent of MAP Kinase
A previous study23 found that specific inhibitors of MAP kinase, U0126 and PD98059, inhibited the induction of MUC2 by PMA. To investigate the role of MAP kinases in the induction of MUC2 by bile acids, we examined phosphorylation of ERK1/2 after treatment of LIM6 cells with 100 μM DCA, CDCA, or UDCA. The 3 bile acids had no effect on the levels of phosphorylated ERK1/2 or total ERK1/2 (Fig. 7A). DCA (100 μM) failed to induce phosphorylation of ERK1/2, JNK, and P-38 kinases at the early and late time points tested (0.5–6 hours; Fig. 7B). To further exclude the involvement of MAP kinase on the induction of MUC2 by bile acids, we used PD98059 and U0126 to selectively block the activity of MEK. We found that these inhibitors did not block the DCA-dependent increase in MUC2 protein and MUC2 transcription, although both PD98059 and U0126 did suppress phosphorylation of ERK1/2 (data not shown). Additional Western analyses (data not shown) indicate that bile acids also have no effect on levels of COX-2, FXR, or nuclear factor (NF)-kappa-B proteins in LiM6 cells.
The effects of bile acids on mucin gene expression in colon carcinoma cells have not been well studied. DCA has been reported to increase the secretion of mucin in LS174T colon carcinoma cells, but mucin synthesis or MUC2 gene expression was not studied.15 Another study found that long-term treatment of Caco-2 colon carcinoma cells decreased total mucin, but those cells do not express MUC2.14 In the current study, we find that bile acids increase MUC2 expression in LiM6 colon carcinoma cells. The transcriptional activity of MUC2 promoter/reporter construct transiently transfected into LiM6 was increased by DCA and other bile acids in a dose-dependent fashion, indicating that bile acid-induced MUC2 upregulation occurs at the transcriptional level.
AP-1 is an important transcription factor that mediates expression of multiple genes in important biologic processes including cell growth, apoptosis, and transformation.33–35 Because AP-1 acts through binding to the consensus sequence [TGA(C/G)TCA] and the MUC2 proximal promoter harbors 3 potential AP-1 binding sites (deduced by inspection of GenBank U67167 using http://www.genomatrix.de/cgi-bin/matinspector-prof), we postulated that AP-1 could play a role in the induction of MUC2 by bile acids. Our data indicate that AP-1 is essential to efficiently upregulate MUC2 after bile acid treatment, because c-jun/AP-1 expression and transcription activity coincided with MUC2 induction, and inhibition of AP-1 expression and activity efficiently suppressed bile acid-mediated upregulation of MUC2. Several transcription factors have been reported to regulate MUC2 expression in other systems. For example, MUC2 can be upregulated by lipopolysaccharide, Haemophilus influenzae, and PMA via NF-kappa-B.36–38 Other transcription factors that have been shown to regulate MUC2 expression in other contexts include SP-1, CDX-2, and GATA-5.39–41 It is likely that AP-1 activity is required for MUC2 expression in cells other than colon carcinoma, but this has not yet been established.
Activation of PKC by bile acids is well documented, and may be one mechanism of bile acid-induced carcinogenesis.29, 30 We found that the PKC inhibitor calphostin C strongly blocked c-jun/AP-1 and MUC2 induction by DCA, indicating that PKC is involved in the bile acid-dependent induction of MUC2. Although forskolin, a PKA activator, stimulates MUC2 expression,31 the bile acid-dependent induction of MUC2 depends less on PKA than on PKC, because the protein kinase inhibitor, H-8, which preferentially inhibits PKA compared with PKC,32 is much less effective than calphostin C in blocking bile acid-dependent induction of MUC2. In contrast to previous work on MUC2 induction by PMA via the Ras/Raf/ERK cascade, we find that the induction of MUC2 by bile acids is independent of ERK1/2 MAP kinases in LiM6 colon carcinoma cells. Treatment with DCA did not affect the phosphorylation of ERK1/2, and inhibitors of MAP kinase, PD5089 and U0126, did not block MUC2 induction by DCA in LiM6 cells. Available data also suggest that MUC2 induction by bile acids does not involve JNK, p38 Kinase, COX-2, FXR, or NF-kappa-B, although much of the signaling pathway remains to be elucidated.
We conclude that treatment of human colon carcinoma cells with bile acids in physiologic concentrations42, 43 upregulates MUC2 transcription by activation of AP-1 via PKC, independent of MAP kinase. The biologic consequences of the induction of MUC2 expression by bile acids are unclear. Further studies are needed to confirm that the secretion of mature MUC2 glycoprotein is increased by treatment of colon carcinoma cells with bile acids, and to determine whether induction of MUC2 by bile acids can increase the invasion potential of cells in vitro and their metastatic potential in vivo. A more detailed understanding of the precise mechanisms by which bile acids induce MUC2 could also facilitate the development of chemopreventive strategies to diminish the risk of carcinogenesis and metastasis, particularly in mucinous carcinomas.