MUC2 expression is regulated by histone H3 modification and DNA methylation in pancreatic cancer

Authors

  • Norishige Yamada,

    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    Search for more papers by this author
  • Tomofumi Hamada,

    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    Search for more papers by this author
  • Masamichi Goto,

    Corresponding author
    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    • Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
    Search for more papers by this author
    • Fax: +81-99-265-7235

  • Hideaki Tsutsumida,

    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    Search for more papers by this author
  • Michiyo Higashi,

    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    Search for more papers by this author
  • Mitsuharu Nomoto,

    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    Search for more papers by this author
  • Suguru Yonezawa

    1. Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
    Search for more papers by this author

Abstract

Mucins are highly glycosylated proteins that play important roles in carcinogenesis. In pancreatic neoplasia, MUC2 mucin has been demonstrated as a tumor suppressor and we have reported that MUC2 is a favorable prognostic factor. Regulation of MUC2 gene expression is known to be controlled by DNA methylation, but the role of histone modification for MUC2 gene expression has yet to be clarified. Herein, we provide the first report that the histone H3 modification of the MUC2 promoter region regulates MUC2 gene expression. To investigate the histone modification and DNA methylation of the promoter region of the MUC2 gene, we treated 2 human pancreatic cancer cell lines, PANC1 (MUC2-negative) and BxPC3 (MUC2-positive) with the DNA methyltransferase inhibitor 5-azacytidine (5-aza), the histone deacetylase inhibitor trichostatin A (TSA), and a combination of these agents. The DNA methylation level of PANC1 cells was decreased by all 3 treatments, whereas histone H3-K4/K9 methylation and H3-K9/K27 acetylation in PANC1 cells was changed to the level in BxPC3 cells by treatment with TSA alone and with the 5-aza/TSA combination. The expression level of MUC2 mRNA in PANC1 cells exhibited a definite increase when treated with TSA and 5-aza/TSA, whereas 5-aza alone induced only a slight increase. Our results suggest that histone H3 modification in the 5′ flanking region play an important role in MUC2 gene expression, possibly affecting DNA methylation. An understanding of these intimately correlated epigenetic changes may be of importance for predicting the outcome of patients with pancreatic neoplasms. © 2006 Wiley-Liss, Inc.

MUC2 is a gel-forming secretory mucin that is expressed in many organs, including the colon, small intestine and respiratory tract. Altered expression of MUC2 has been described in many neoplasms, and our previous studies have shown that MUC2 expression in neoplastic cells is related to clinical outcome in various neoplasms.1, 2, 3, 4, 5 Production of MUC2 in the pancreas (in which it is not normally expressed) may have a close relationship with patient prognosis in pancreatic cancer,4, 5, 6 and MUC2 has been suggested to be related to carcinogenesis in pancreatic and colorectal cancer.7 Muc2 knock-out mice frequently developed adenomas in the small intestine that progressed to invasive adenocarcinoma.8 Thus, a study of MUC2 regulation may contribute to both clinical management of various cancers and to the understanding of the cancer biology of the pancreas.

MUC2 is the major mucin in the intestinal epithelium, and the corresponding gene has been mapped to human chromosome 11p 15.9 Regulation of MUC2 expression has been extensively studied, but the exact regulatory mechanism remains not fully understood. In the pancreatic cell line PANC1, which normally does not express MUC2, Siedow et al. have shown that de novo expression of MUC2 is triggered by treatment of the cells with 5-aza-2′-deoxycytidine.10 Ho et al. have demonstrated that expression of MUC2 gene products is correlated with methylation status in the proximal region of the promoter,11 and suppression of the MUC2 gene in colorectal carcinoma cells in vitro and in vivo has been shown to be associated with methylation of the MUC2 promoter region.12, 13, 14 Methylation of cytosine residues at CpG dinucleotides is an important epigenetic change that has been linked to transcriptional repression and regulation of chromatin structure.12, 13, 14 Hamada et al. have determined the detailed methylation status of a wide area of the MUC2 promoter region in pancreatic cancer cell lines, and suggested that methylation of certain CpG sites may play a particularly important role in the regulation of MUC2 transcription.15

Posttranslational modification of histone tails also plays a critical role in epigenetic silencing.16, 17 Histone proteins are subject to many different chemical modifications, including phosphorylation, acetylation and methylation.18 Such modifications affect the access of regulatory factors and alter the histone complexes with chromatin, thereby influencing gene expression. Acetylation of lysine residues on histone H3 leads to the formation of an open chromatin structure, whereas methylation of K9 on histone H3 (H3-MeK9) has recently been shown to be a marker of heterochromatin from yeast to mouse.19 The H3-MeK9 modification serves as a cis-acting binding site for heterochromatin protein 1 (HP 1), and it has been shown that modifications of histone H3 also contribute to euchromatin gene silencing by switching between acetylation and methylation of K9.20 Another histone modification, methylation of K4 of histone H3, localizes to sites of active transcription, and this modification may be stimulatory for transcription. Combinations of histone modifications at different residues may act synergistically or antagonistically to affect gene expression,18 but a detailed analysis of histone modification associated with the MUC2 promoter region in pancreatic cancers has not been performed.

To further understand the DNA methylation and histone modification status of the MUC2 gene, methylation-specific PCR (MSP) and chromatin immunoprecipitation (ChIP) assays were performed in PANC1 and BxPC3 cells. We also treated the PANC1 cells with 5-aza and/or TSA to examine the relationship among DNA methylation, histone modification and MUC2 gene expression. In this report, we describe an epigenetic mechanism through which MUC2 gene expression is regulated by a tightly-related combination of DNA methylation and histone modification associated with the 5′ flanking region of the MUC2 promoter.

Material and methods

Cell lines

Human pancreatic carcinoma cell lines PANC1, BxPC3, human breast cancer cell line MCF-7 (MUC2−) and human colon adenocarcinoma cell line LS174T (MUC2+) were obtained from American Type Culture Collection (Manassas, VA). PANC1, a poorly differentiated tumor cell line, shows no expression of MUC2, whereas BxPC3, a moderately differentiated tumor cell line, expresses MUC2. PANC1 cells were cultured in D-MEM and 10 mM HEPES, and BxPC3 cells were cultured in RPMI 1640 and 10 mM HEPES. LS174T cells and MCF-7 cells were cultured in Eagle's Minimum Essential Medium and 10 mM HEPES. All media were supplemented with 10% (v/v) fetal bovine serum and 5 ml penicillin/streptomycin.

Western blot analysis of MUC2 protein expression

Preparation of cell lysates and Western blotting procedures have been described previously.21 Equal aliquots of total protein (50 μg per lane) were electrophoresed on a 3–8% SDS-polyacrylamide gel (Invitrogen, Carlsbad, CA), transferred to nitrocellulose membranes (Invitrogen), and blotted using primary antibody directed against human MUC2 (rabbit 1:1,000) (TaKaRa, Tokyo, Japan). After incubation with the appropriate horseradish peroxidase-conjugated secondary antibody (1:1,000 Amersham Biosciences, Piscataway, NJ), immune complexes were visualized using an enhanced chemiluminescence (ECL) detection system (Amersham Biosciences), with protein markers (Bio-Rad, Hercules, CA) used as molecular size standards.

DNA extraction and DNA methylation-specific PCR (MSP) analysis

DNA from cell lines was extracted using a DNeasy Tissue System (Qiagen, Valencia, CA), according to the manufacturer's instructions. Bisulfite modification of the genomic DNA was carried out using a CpGenome DNA Modification Kit (Chemicon International, Temecula, CA). The modified DNA was amplified by PCR. The primers were designed for the MUC2 promoter (GenBank accession number U67167) using MethPrimer,22 a program for designing bisulfite-conversion-based methylation PCR primers (Table I). Each PCR mixture contained genomic DNA, 1.25 units TaKaRa Ex Taq HS (TakaRa), 1× Ex Taq buffer, a 2 mM deoxynucleotide triphosphate mixture, and 50 pmol sense and antisense primers in a volume of 25 μl. The PCR conditions were 94°C for 5 min, 36 cycles of 94°C for 20 sec, 62°C for 30 sec and 72°C for 45 sec, with a final extension reaction at 72°C for 5 min.

Table I. MSP and Chip Primers Designed for MUC2 Promoter Analysis
NamePrimer sequencePosition in U67167Size (bp)
  • *

    M indicates the primer for methylated alleles.

  • **

    U indicates the primer for unmethylated alleles.

  • ***

    Negative control (NAC) primers flank a region of genomic cDNA between the GAPDH gene and the chromosome condensation-related SMC-associated protein (CNAP1) gene.

MSP primers
MSP 1_M*_SenseTGGGGGTGTTAGTGATTTTATAGTC−1,871 to −1,847100
MSP 1_M_AntisenseACTAAACCCCATTCCTAACGAA−1,793 to −1,772 
MSP 1_U**_SenseTGGGGGTGTTAGTGATTTTATAGTT−1,871 to −1,847102
MSP 1_U_AntisenseAAACTAAACCCCATTCCTAACAAA−1,795 to −1,772 
MSP 2_M_SenseTTTATATAAGTTAGTGGTTTTTTCGG−531 to −506217
MSP 2_M_AntisenseAATCTAATCAAACTCCTTAACCCG−338 to −315 
MSP 2_U_SenseTTATATAAGTTAGTGGTTTTTTTGG−530 to −506216
MSP 2_U_AntisenseAATCTAATCAAACTCCTTAACCCAC−339 to −315 
MSP 3_M_SenseGTTGTTTTATTTTGAAGAAGGTTGC−207 to −193229
MSP 3_M_AntisenseCGATATAAATTACGCCCGAA−8 to +12 
MSP 3_U_SenseGTTGTTTTATTTTGAAGAAGGTTGTG−207 to −192238
MSP 3_U_AntisenseTAACAAAAACAATATAAATTACACCCAAA−8 to +21 
MSP 4_M_SenseTGTTTTCGGGCGTAATTTATATC−12 to +11123
MSP 4_M_AntisenseCTCTCTCACCTATCTAAAACTCCGA+88 to +112 
MSP 4_U_SenseTTTGTTTTTGGGTGTAATTTATATT−10 to 11131
MSP 4_U_AntisenseATCTACTCTCTCACCTATCTAAAACTCCA+88 to +117 
ChIP primers
ChIP 1_SenseAGTAGATGCTGCGAATCTGG−1,942 to −1,923224
ChIP 1_AntisenseTAGGAGCCCTGTCTGAGGTTA−1,739 to −1,719 
ChIP 2_SenseACACACTGAGCTTCCTCCAC−1,330 to −1,311182
ChIP 2_AntisenseACCCCAACAGTTATGGAGACT−1,169 to −1,149 
ChIP 3_SenseAGAGTCCCACACAAGTCAGTG−536 to −516223
ChIP 3_AntisenseAAGTCTGGTCAGGCTCCTTAG−334 to −314 
ChIP 4_SenseCCTGACCAGACTTGCTTCTG−326 to −307173
ChIP 4_AntisenseGTTATCGGAGGTCATCTCTTTATG−177 to −154 
ChIP 5_SenseCAGCTGTCTTCTTGATCATCTG−116 to −95233
ChIP 5_AntisenseGTCTGCTCTCTCACCTGTCTG+97 to +112 
NAC_SenseATGGTTGCCACTGGGGATCT***174
NAC_AntisenseTGCCAAAGCCTAGGGGAAGA  

Chromatin immunoprecipitation (ChIP) assays

ChIP assays were carried out using a ChIP-IT Kit (Active Motif, Carlsbad, CA). Briefly, cells were harvested and their proteins were crosslinked to DNA by incubation in 1.0% formaldehyde for 5 min at room temperature. After stopping the fixation reaction by addition of a glycine solution, the formaldehyde-fixed cells were spun down by brief centrifugation, after which the supernatant was carefully aspirated. The cells were then resuspended in lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl (pH 8.0), and protease inhibitor). The nucleoprotein complexes were sonicated to reduce the sizes of DNA fragments to 200–500 bp, and then immunoprecipitated for 16 hrs at 4°C with rotation, using anti-dimethyl histone H3-K4, anti-acetyl histone H3-K9, anti-acetyl histone H3-K14, anti-acetyl histone H3-K27 or anti-dimethyl histone H3-K27 antibody (Upstate Biotechnologies, Lake Placid, NY), anti-trimethyl histone H3-K4, anti-dimethyl histone H3-K9 or anti-trimethyl histone H3-K9 antibody (Abcam, Cambridge, UK) and negative control IgG (Active Motif). The resultant immune complexes were collected using protein A-agarose beads, after which the DNA was purified by phenol–chloroform extraction, precipitated with ethanol, and resuspended in distilled water. About a 1:100 solution of the precipitated DNA was used for PCR, and a 1:100 solution of the DNA before addition of antibody was used as an internal control for quantitative accuracy. PCR was performed in a solution containing 1× PCR buffer (TaKaRa), 1 μM ChIP or negative control primers (Table I), a 0.25 mM deoxynucleoside triphosphate mixture, 10% dimethyl sulfoxide (Sigma-Aldrich, St. Louis, MO) and 1.0 units of TaKaRa Ex Taq HS (TaKaRa). The amplified products were subjected to agarose gel electrophoresis.

5-Azacytidine and TSA treatment

PANC1 cells were split 24 hrs before treatment. The cells were incubated for 5 days with 1 mM 5-azacytidine (Nacalai tesque, Kyoto, Japan), a methyltransferse inhibitor, and/or for 5 days with 10 μM trichostatin A (Wako Pure Chemical Industries, Osaka, Japan), a histone deacetylase inhibitor. Media were changed every 24 hrs.

Real-time PCR analysis of immunoprecipitated DNA

Quantitative PCR analysis was performed using the SYBR Green PCR Core Reagents Kit (Applied Biosystems, Foster City, CA). Real-time detection of the emission intensity of SYBR green bound to double-stranded DNAs was achieved with an ABI PRISM 7000 Sequence Detection System (Applied Biosystems). All PCRs were carried out under the same conditions: 94°C for 20 sec, 59°C for 30 sec and 72°C for 30 sec for 45 cycles (+ final extension). Titrations of known amounts of DNA were included in each reaction as a standard for quantitation. DNA from the ChIP samples immunoprecipitated with anti-acetylated H3-K9/K14/K27, anti-dimethylated H3-K4/K9/K27 and anti-trimethylated H3-K4/K9, and from ChIP samples immunoprecipitated with a nonspecific antibody control (NAC) were included in each PCR set. Data were analyzed by Comparative CT methods.23 The fraction of immunoprecipitated DNA was calculated as a ratio of immunoprecipitated DNA to the total amount of input DNA in the PCR reaction.

Quantitative RT-PCR analysis of pancreatic carcinoma cell lines

The mRNA from cells that had or had not undergone 5-aza, TSA or 5-aza/TSA combination treatment was purified with a QuickPrep micro mRNA purification kit (Amersham Biosciences). Of a total of 200 μl mRNA, 20 μl was reverse-transcribed with hexamers (Applied Biosystems). A 1-μl cDNA aliquot was amplified in 25 μl of 2xTaqMan Universal Master Mix, 2.5 μl of 20xTarget Assay Mix, and 2.5 μl of 20X Control Assay Mix (Applied Biosystems) under the following PCR conditions: 2 min at 50°C, 10 min at 95°C, 60 cycles of 15 sec at 95°C and 1 min at 60°C. The primers and probes were designed and synthesized by Applied Biosystems. The product number of the Target Assay Mix used for MUC2 was Hs00159374. Human GAPDH (product number 4310884E) was used to calibrate the original concentration of mRNA, i.e. the concentration of mRNA in the tissue was defined as the ratio of target mRNA copies versus GAPDH mRNA copies. In this analysis, data from 3 separate experiments were averaged.

Immunohistochemistry

MUC2 protein expression levels in PANC1 cells after 5-aza, TSA and 5-aza/TSA treatment for 4 days were assessed by immunohistochemistry. MUC2 was detected using a rabbit polyclonal anti-MUC2 antibody (TaKaRa; dilution, 1:1,000 for cell culture; incubation period, 1 hr at 37°C). Immunohistochemical staining was performed by the immunoperoxidase method using a Vectastain Elite ABC Kit (Vector Laboratories, Burlingame, CA), as described previously.15

Results

Selection of MSP primers for different regions of the MUC2 gene promoter

Four primers for methylation-specific PCR (MSP) were designed to target different regions of the MUC2 promoter (Fig. 1), based on the results of bisulfite genomic sequencing of the promoter region.15 Before MSP was performed, the expression of MUC2 protein in the human pancreatic carcinoma cell lines PANC1 and BxPC3, the human colon adenocarcinoma cell line LS174T and the human breast cancer cell line MCF7 was examined using Western blot analysis. As shown in Figure 2a, BxPC3 and LS174T cells expressed MUC2, while PANC1 and MCF7 cells did not do so.

Figure 1.

Schematic representation of the MUC2 gene promoter region, which spans positions −2,000 to +186 with respect to the transcription start site. The relative positions of the sequence-binding motifs of transcription factors, CpG sites and the MSP and ChIP primers used in this study are indicated. There are 59 CpG sites in this region.

Figure 2.

Analysis of CpG methylation in pancreas cancer cell lines. (a), Western blot analysis of MUC2 in 4 cell lines. BxPC3 and LS174T cells showed high expression of MUC2, whereas PANC1 and MCF7 cells showed no expression. The same membrane was stripped and reprobed with an antibody that recognizes β-actin to show equivalent protein loading in each lane. (b), Methylation-specific PCR (MSP) analysis of the MUC2 gene promoter region in PANC1 and BxPC3 cells, using MSP 1 to 4 primers. The PCR-products labeled “M” (methylated) were generated by methylation-specific primers, and those labeled “U” (unmethylated) by primers specific for unmethylated DNA. The PCR product sizes are summarized in Table I. (c), DNA methylation status of MUC2 in PANC1, BxPC3, LS174T and MCF7 cells in the MSP 3 region, detected by MSP. The methylated allele was detected in PANC1 and MCF7 cells, whereas the unmethylated allele was detected in BxPC3 and LS174T cells.

Methylation of the MUC2 promoter region in PANC1 and BxPC3 cells was examined using MSP and the 4 primer sets MSP 1 to 4 (Fig. 2b). The MSP 1 primer was designed to target the upstream region of the MUC2 promoter (Fig. 1), and with this primer methylation bands (lanes indicated by M, Fig. 2b) were observed in both the MUC2-negative cell line PANC1 and the MUC2-positive cell line BxPC3. The MSP 2 and MSP 3 primers clearly showed methylation in PANC1 cells, but indicated the presence of unmethylated DNA (lanes indicated by U, Fig. 2b) in BxPC3 cells. This result indicates that hypomethylation of the promoter correlates with MUC2 expression. The contrast between the M and U bands was more remarkable with MSP 3 than with MSP 2. The MSP 4 primer was designed to include a part of the MUC2 exon region (Fig. 1), and this primer indicated methylation of this region in both cell types, suggesting that the DNA is always methylated in the MUC2 exon region. Overall, the MSP results were consistent with the results of our previous bisulfite sequencing,15 which demonstrated that the methylation status in the 2 cell lines differs greatest in the downstream region of the MUC2 promoter.

To ensure the reliability of the MSP 3 primer, we performed MSP on the 4 cell lines using the MSP3 primer set (Fig. 2c). An unmethylated band was clearly obtained in the MUC2-positive cell lines BxPC3 and LS174T, whereas methylation was observed in PANC1 and MCF7 cells. These results suggest that suppression of MUC2 gene expression is associated with methylation of the MUC2 promoter region, as described in previous reports.12, 13, 14 Since these results showed the utility of the MSP3 primer, this primer set was used in the following experiments.

For the precise comparison of our MSP results and actual methylation status, here we show bisulfate genomic sequencing data obtained in our previous study.15 AP2-binding CpG site No. 5 (refer Hamada et al.15 for detail of numbering) detected by MSP primer set 1 (MSP 1) showed 60% methylation in PANC1 and 50% methylation in BxPC3. CpG site No. 7 showed 100% methylation in PANC1 and 70% in BxPC3. Similarly, AP2-binding CpG sites detected by MSP 4 showed 100% methylation in PANC1 (No. 45 and 46), 70% in BxPC3 (No. 45), and 30% in BxPC3 (No. 46). CpG site No. 47 showed 100% methylation in PANC1 and 30% in BxPC3. CpG site No. 50 showed 100% methylation in PANC1 and 50% in BxPC3. These data indicate that CpG sites included by MSP 1 and 4 are partially methylated. For MSP 3, Cdx2-binding CpG site No. 41 showed 95% methylation in PANC1 and 50% methylation in BxPC3. AP2-binding CpG sites detected by MSP 3 showed 100% methylation in PANC1 (No. 45 and 46), 70% in BxPC3 (No. 45) and 30% in BxPC3 (No. 46). CpG site No. 47 showed 100% methylation in PANC1 and 30% in BxPC3. These data indicate that CpG sites included by MSP 3 are also partially methylated.

Modification of histone H3 associated with the 5′ flanking region of MUC2 promoter shows a clear difference between PANC1 and BxPC3 cells

To elucidate the relationship between DNA methylation and histone modification, 5 ChIP primers were designed to target regions somewhat similar to those targeted by the MSP primers (Fig. 1). To determine the profiles of histone H3 modification in PANC1 (MUC2-negative) and BxPC3 (MUC2-positive) cells, ChIP assays were performed for the 5 regions, using anti-dimethyl-H3-K4/K9/K27, anti-trimethyl-H3-K4/K9 and anti-acetyl-H3-K9/K14/K27 antibodies.

Methylation of histone H3 lysine 4 (H3-MeK4) is associated with active gene transcription, and elevated levels of H3-MeK4 have been observed at active gene promoters.24 Figure 3 shows that di- and tri-methylation of lysine 4 of histone H3 (H3-Me2K4 and H3-Me3K4) in PANC1 cells were quite similar to those in BxPC3 cells for ChIP regions 1 to 4. However, in ChIP region 5, H3-Me2K4 and H3-Me3K4 were more highly methylated in BxPC3 cells than in PANC1 cells.

Figure 3.

ChIP analysis of histone H3 modification in the MUC2 promoter region (ChIP regions 1 to 5) in PANC1 and BxPC3 cells. Reactions were controlled by the PCR reactions using input DNA. Negative control reaction using aliquot precipitated with negative control IgG antibody (NAC) was shown. The unbound and bound fractions in each chromatin immunoprecipitation experiment are shown. The position of clear difference between PANC1 and BxPC3 cells were indicated by closed and open arrowhead, respectively.

Methylation of lysine 9 of histone H3 (H3-MeK9) facilitates formation of heterochromatin, and elevated levels of H3-MeK9 at promoter sequences are associated with suppression of gene expression.25, 26 As seen in Figure 3, all 5 ChIP regions of the MUC2 gene promoter showed a higher degree of di- and tri-methylation of lysine 9 histone H3 (H3-Me2K9 and H3-Me3K9) in PANC1 cells than in BxPC3 cells.

Acetylation of lysines 9, 14 and 27 of histone H3 (H3-AceK9, H3-AceK14 and H3-AceK27) is associated with euchromatin formation, and acetylation of promoter-proximal histones is associated with gene expression.27 Figure 3 indicates that the acetylation levels in BxPC3 cells was high at H3-AceK9 in ChIP region 5 and at H3-AceK27 in ChIP regions 4 and 5, compared to those in PANC1 cells, but that H3-AceK14 showed no clear difference between PANC1 and BxPC3 cells.

Methylation of lysine 27 of histone H3 (H3-MeK27) is a marker for constitutive and facultative heterochromatin.28 In our study, H3-Me2K27 in ChIP regions 3 to 5 showed higher methylation levels in PANC1 cells than in BxPC3 cells.

In summary, ChIP region 5 showed clear differences between PANC1 and BxPC3 cells in H3-K4 methylation, H3-K9 methylation, H3-K9 acetylation and H3-K27 acetylation. Thus, chromatin modification associated with ChIP region 5 correlated well with the DNA methylation status obtained with MSP 3.

MUC2 mRNA and protein levels are restored by TSA and 5-aza/TSA treatment

To confirm that DNA methylation and histone H3 modification suppress the expression level of MUC2 mRNA, quantitative RT-PCR analysis was performed in PANC1 cells treated with 5-aza, TSA or 5-aza/TSA in combination (Fig. 4a). It is known that 5-aza treatment alone decreases DNA methylation29 and that TSA inhibits histone deacetylase.30 Although MUC2 gene suppression is thought to be an effect of DNA methylation in PANC1 cells, treatment with 5-aza did not dramatically restore the MUC2 mRNA level in these cells (Fig. 4a). However, treatment with TSA or 5-aza/TSA significantly restored the MUC2 mRNA level, compared to treatment with 5-aza alone. To determine the actual amount of MUC2 protein, immunohistochemistry was performed using anti-MUC2 antibody (Fig. 4b). Consistent with the RT-PCR analysis, expression of MUC2 was not observed in PANC1 cells. Treatment with 5-aza slightly increased the expression of MUC2, but TSA and 5-aza/TSA treatment showed definite restoration of MUC2 expression.

Figure 4.

Expression of the MUC2 gene examined by quantitative RT-PCR and immunoreactivity. (a), Examples of quantitative RT-PCR results. PANC1 cells were treated with 5-aza, TSA and 5-aza/TSA in combination. BxPC3 cells were used as MUC2-positive controls. The graph shows gene expression levels relative to those in nontreated PANC1 cells, as log10 values. (b), MUC2 immunoreactivity in PANC1 cells. BxPC3 cells were used as MUC2-positive controls. No expression of MUC2 was noted in PANC1 cells without treatment. After 5-aza treatment, PANC1 cells showed slight restoration of MUC2 expression, whereas TSA- or 5-aza/TSA-treated PANC1 cells showed definite restoration of MUC2 expression.

Trichostatin A decreases DNA methylation and changes histone H3 modification in the 5′ flanking region of the MUC2 promoter

Our results suggest that the 5′ flanking region (about −300 bp to 0 bp) of the MUC2 promoter plays an important role in the expression of the MUC2 gene. To determine the relationship between DNA methylation and histone H3 modification, DNA methylation analysis was performed by MSP using the MSP 3 primer (Fig. 5). All treatments partially induced demethylation in PANC1 cells. Furthermore, a ChIP analysis was performed to examine changes in histone modification in ChIP region 5 after treatment of PANC1 cells (Fig. 6a). Treatment with 5-aza alone had no effect on K4 and K9 dimethylation in histones associated with the MUC2 gene in PANC1 cells, but treatment with TSA and 5-aza/TSA increased K4 dimethylation and decreased K9 dimethylation (Fig. 6b). Acetylation of K9 and K27 of histone H3 associated with ChIP region 5 of the MUC2 promoter was not increased in PANC1 cells in response to 5-aza alone, whereas acetylation was significantly increased by TSA and 5-aza/TSA (Fig. 6b). Hence, 5-aza had minimal effects on histone H3 modification, whereas TSA and 5-aza/TSA had strong effects. Overall, our results suggest that DNA demethylation, histone H3-K4 methylation, histone H3-K9 dimethylation and histone H3-K9/K27 acetylation in the 5′ flanking region of the MUC2 promoter might be all necessary for MUC2 gene expression.

Figure 5.

MSP analysis of the MUC2 promoter region before and after treatment of PANC1 cells. PANC1 cells were treated with 5-aza, TSA, or 5-aza/TSA in combination. After bisulfite treatment, genomic DNA was subjected to MSP analysis using the MSP 3 primer. Unmethylated bands were detected following all 3 treatments.

Figure 6.

ChIP analysis in the promoter region before and after treatment of PANC1 cells. (a) ChIP analysis of ChIP region 5 of the MUC2 promoter was performed using antibodies against dimethylated H3-K4, dimethylated H3-K9, acetylated H3-K9, acetylated H3-K27. BxPC3 cells were used as MUC2-positive controls. Negative control reaction using aliquot precipitated with negative control IgG antibody was shown (NAC). (b) Each immunoprecipitate was quantified by quantitative PCR analysis and normalized relative to a control included in each experiment.

Discussion

The close relationship between DNA methylation and histone H3 modification in regulation of MUC2 gene expression has not been shown previously. Our data reveal the details of H3 methylation and acetylation in 2 pancreatic cancer cell lines (PANC1 and BxPC3) in which the methylation status of the MUC2 promoter is different. In addition, because MSP primers and ChIP primers were designed for almost the same regions, we were able to confirm the connection between DNA methylation and histone modification in gene regulation. Changes in DNA methylation had little effect on histone modification, whereas histone modification had a strong effect on DNA methylation. This observation and the relative changes in MUC2 expression after treatment with DNA methyltransferase and histone deacetylase inhibitors suggest that histone modification may be more important than DNA methylation in regulation of MUC2. Finally, our results showed that neither DNA methylation nor histone modification alone fully determine expression of the MUC2 gene, suggesting that MUC2 expression is regulated by a combination of DNA methylation and histone H3 modification in the 5′ flanking region of the MUC2 promoter.

Methylation of cytosine in genomic DNA is known to play an important role in gene regulation and especially in gene silencing.31, 32, 33 Generally, the promoter region of a transcribed gene is hypomethylated.34, 35 In the current study, we designed 4 sets of MSP primers, based on the results of bisulfite genomic sequencing of the MUC2 promoter region between −1,989 to +288 upstream.15 These data indicate that CpG sites included by MSP 1 and MSP 4 are partially methylated, which resulted in the presence of methylation band in both cell lines (BxPC3 and PANC1) with MSP 1 and 4. Of these primers (Table I), experiments using the MSP 3 primer set showed clear differences in methylation in PANC1 and BxPC3 cells. These results indicate that the 5′ flanking region of the promoter, and especially the region from −300 to 0 bp, may play an important role in methylation-related gene silencing of MUC2.

In addition to DNA methylation, it has become increasingly evident that histone modification can contribute to gene regulation.36, 37 Therefore, ChIP primers were designed to target somewhat similar regions to those targeted by the MSP primers, and a variety of anti-histone H3 antibodies were used for immunoprecipitation. Of the ChIP primers, ChIP 5, which was designed to target a region similar to that targeted by the MSP 3 primer, showed a relatively clear difference in chromatin modification between PANC1 and BxPC3 cells. Based on these results, we focused on ChIP 5 in further ChIP assays using anti-dimethyl-H3-K4/K9/K27 and anti-acetyl-H3-K9/K27 antibodies.

There was a significant difference in the H3-K9 methylation level between PANC1 and BxPC3 cells in all 5 ChIP analyses (Fig. 3), suggesting that H3-K9 methylation is closely related to DNA methylation and acts as an epigenetic marker for silencing of the MUC2 gene. Methylation at histone H3-K9 that is specifically catalyzed by Suv39h has recently been shown to be a marker of heterochromatin,19, 25, 26, 38 and histone H3-K9 methylation contributes to establishing and maintaining heterochromatin by providing a binding site for heterochromatin protein 1 (HP 1).39, 40

Strahl and Allis have proposed that distinct histone modifications on 1 or more tails act sequentially or in combination to form a “histone code” that is read by other proteins to bring about distinct downstream events.41 According to our ChIP 5 results, this view of the “histone code” might apply to the regulation of MUC2. However, further analyses of all the histone termini (H2 to H4) are needed to confirm this hypothesis.18, 42

TSA has been reported to act synergistically with 5-aza to reactivate DNA methylation-silenced genes,43, 44, 45, 46 and Xiong et al. have reported that histone deacetylase (HDAC) inhibitors decrease DNA methyltransferase-3B (DNMT3B) mRNA stability and down regulate de novo DNA methyltransferase.47 To examine the relationship between histone modification and DNA methylation, we treated PANC1 cells with 5-aza and/or TSA. The level of DNA methylation in PANC1 cells was decreased by TSA alone, which suggests that the 2 epigenetic phenomena might occur in PANC1 cells simultaneously. In contrast, 5-aza alone did not restore acetylation of histone H3, which ChIP assays showed to be deacetylated in ChIP region 5 in untreated PANC1 cells. Expression of MUC2 only showed a slight increase upon treatment with 5-aza alone, despite decreased DNA methylation upon 5-aza treatment. These results suggest that MUC2 gene expression cannot be restored by promoter region demethylation alone. In contrast, TSA and 5-aza/TSA treatment restored MUC2 gene transcription dramatically. Thus, coordinated changes in DNA methylation and histone modification are suggested to be important for epigenetic regulation of the MUC2 gene.

Augenlicht et al.48 also studied the histone status of MUC2 promotor region using TSA. They reported that sodium butyrate (NaB) specifically represses the expression of the MUC2 gene in MUC2-induced HT29 cells and goblet cell-differentiated variant C1.16E cells, and interpreted that NaB repression is independent of the nature of the stimulus that triggers MUC2 expression. However, there is a possibility that NaB and TSA treatment at the cytotoxic concentration may have reduced MUC2 mRNA and protein in MUC2 expressing cell lines. On the contrary, we demonstrated that treatment with TSA significantly restored the MUC2 mRNA and protein in MUC2-negative PANC1 cells. In addition, the authors suggest that inhibition of MUC2 is linked to the ability of butyrate to repress histone deacetylase activity, since TSA, another inhibitor of histone deacetylases, also inhibited MUC2 expression in MUC2-induced HT29 cells. However, the technology used for the determination of histone H4 modification seems unsuitable for the precise assessment of correlation between histone H4 modification and MUC2 gene expression. We performed ChIP assays for the distinct region of MUC2 promoter to elucidate the relationship between histone H3 modification and MUC2 expression.

Various epigenetic changes were found in the 5′ flanking region of the MUC2 promoter. There are numerous transcription factor binding sites in the promoter region, including an AP2 site (methylation-receptive transcription factors) and an Sp1 site (methylation-nonreceptive transcription factors).15, 49 In addition, a CDX2 binding site is present downstream of the MUC2 promoter,50,51 and Yamamoto et al. have reported that the homeodomain protein CDX2 interacts with the MUC2-WT cis element (bases −201 to −162) in the MUC2 promoter and that ectopic expression of the CDX2 protein activates transcription of MUC2.52 In our previous bisulfite genomic sequencing study,15 CDX2-binding CpG site No. 41 detected by MSP 3 showed 95% methylation in PANC1 and 50% methylation in BxPC3. Thus, MUC2 expression might be regulated by epigenetic changes in the promoter region containing the CDX2 binding site.

Recent studies have shown that RNA interference (RNAi) also plays an important role in gene silencing.53, 54 Volpe et al. have reported that heterochromatic silencing and histone H3 K9 methylation are regulated by RNAi,55 and it has been suggested that small interfering RNAs (siRNAs) regulate DNA methylation of promoter regions.55, 56 Hence, it is apparent that DNA methylation, histone modification and RNAi are intimately related to each other,54, 57 and comprehensive epigenetic studies in which these 3 mechanisms are examined will be necessary to confirm the details of regulation of MUC2 gene expression.

Acknowledgements

The authors thank Ms. Yoshiko Arimura and Ms. Yukari Nishimura for their excellent technical assistance and advice. This study was supported by the Ministry of Education, Science, Sports, Culture and Technology, Japan to S. Yonezawa.

Ancillary