Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Aberrant promoter methylation in human DAB2 interactive protein (hDAB2IP) gene in gastrointestinal tumour

Aberrant promoter methylation in human DAB2 interactive protein (hDAB2IP) gene in... British Journal of Cancer (2005) 92, 1117 – 1125 & 2005 Cancer Research UK All rights reserved 0007 – 0920/05 $30.00 www.bjcancer.com Aberrant promoter methylation in human DAB2 interactive protein (hDAB2IP) gene in gastrointestinal tumour 1 ,1 1 1 1 1 1 2,3 4,5 H Dote , S Toyooka , K Tsukuda , M Yano , T Ota , M Murakami , M Naito , M Toyota , AF Gazdar and N Shimizu 1 2 Department of Cancer and Thoracic Surgery, Okayama University Graduate School of Medicine and Dentistry, Okayama 700-8558, Japan; First Department of Internal Medicine, Cancer Research Institute, Sapporo Medical University, Sapporo 060-8543, Japan; Department of Molecular Biology, Cancer Research Institute, Sapporo Medical University, Sapporo 060-8543, Japan; Hamon Center for Therapeutic Oncology Research, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA; Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA The human DOC-2/DAB2 interactive protein (hDAB2IP) gene is a novel member of the Ras GTPase-activating family and has been demonstrated to be a tumour-suppressor gene inactivated by methylation in several cancers. In this study, we analysed the methylation and expression status of hDAB2IP in gastrointestinal tumours. The promoter region of hDAB2IP was divided into two regions (m2a and m2b) based on our previous report, and the methylation status was determined by bisulphite DNA sequencing in gastric cancer cell lines. The gene expression was semiquantified by real-time RT–PCR, and the results indicated that the m2b promoter region might be an authentic methylation-mediated key regulator of the gene expression. Based on the sequence data, we developed a methylation-specific PCR (MSP) for the m2a and m2b regions and applied it to the samples. Methylation-specific PCR revealed aberrant methylation in the m2a region in eight of 12 gastric cancer cell lines (67%), 16 of 35 gastric cancer tissues (46%) and 29 of 60 colorectal cancer tissues (48%), and in the m2b region in eight of 12 cell lines (67%), 15 of 35 gastric cancer tissues (43%) and 28 of 60 colorectal cancer tissues (47%). On the other hand, seven (12%) and 11 (19%) of 59 gastrointestinal nonmalignant mucosal specimens showed methylation in the m2a and m2b regions, respectively, suggesting that hDAB2IP methylation might play a causative role in carcinogenesis. The 5-aza-2 -deoxycytidine treatment restored the gene expression in the m2b-methylated cell lines, confirming that the methylation caused gene downregulation. We also examined the relationship between hDAB2IP methylation and the clinicopathological features in patients with primary tumours, and determined that methylation in the m2b region was associated with location of the tumour in the stomach. In summary, our results demonstrated that hDAB2IP methylation is frequently present in gastrointestinal tumours and that the resulting gene silencing plays an important role in gastrointestinal carcinogenesis. British Journal of Cancer (2005) 92, 1117–1125. doi:10.1038/sj.bjc.6602458 www.bjcancer.com Published online 15 March 2005 & 2005 Cancer Research UK Keywords: methylation; hDAB2IP; gastrointestinal tumour; histone acetylation; methylation specific PCR Aberrant methylation of normally unmethylated CpG islands form a unique protein complex and exert a negative regulatory located in the 5 promoter region of genes has been shown to be activity on the Ras-mediated signal pathway (Wang et al, 2002). associated with transcriptional inactivation of several genes in Recently, it was reported that the P2 region (598 to þ 44) of the human cancers, including gastrointestinal tumours, and appears to hDAB2IP gene shows promoter activity, and that transcriptional serve as an alternative to mutational inactivation (Jones and Laird, silencing by aberrant methylation of the P2 region is critical 1999; Baylin and Herman, 2000). tumorigenesis in prostate cancer (Chen et al, 2003) and breast The human DOC-2/DAB2 interactive protein gene (hDAB2IP), cancer (Dote et al, 2004). located on chromosome 9q33.1–q33.3, is a novel member of the In this study, we examined the methylation and expression Ras GTPase-activating family (Tseng et al, 1998; Chen et al, 2002; status of hDAB2IP in a series of gastric cancer cell lines, and Wang et al, 2002). It interacts directly with the DOC-2/DAB2 (also developed the methylation-specific PCR (MSP) assay for hDAB2IP known as differentially expressed in ovarian carcinoma-2, DOC-2) for application to gastrointestinal tumour samples. To clarify the protein that appears to be a tumour suppressor in malignant mechanism of gene silencing, the cell lines were treated with 5-aza- tumours, including mammary, prostate and ovarian cancers 2 -deoxycytidine (5-Aza-CdR) and/or trichostatin A (TSA) to (Fulop et al, 1998; Zhou and Hsieh, 2001). hDAB2IP and DAB2 examine the effect of methylation and histone deacetylation on the gene silencing. In addition, we analysed the relationship between the methylation status and the clinicopathological features in cases *Correspondence: Dr S Toyooka; E-mail: [email protected] with surgically resected gastric and colorectal cancers, in order to Received 16 August 2004; revised 5 January 2005; accepted 19 January investigate the clinical and pathogenetic importance of hDAB2IP 2005; published online 15 March 2005 methylation in gastrointestinal tumours. Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al MATERIALS AND METHODS DNA extraction and bisulphite treatment DNA from cell lines and specimens was subjected to bisulphite Cell lines treatment as described previously (Dote et al, 2004). Briefly for In all, 12 cell lines derived from human gastric cancer were used in bisulphite treatment, 1mg of genomic DNA was denatured by this study. Four gastric cancer cell lines of MKN series (MKN-1, NaOH and modified by sodium bisulphite, which converts all adenosquamous cell carcinoma; MKN-28 and MKN-74, well- unmethylated cytosines to uracils while methylated cytosines differentiated adenocarcinoma; MKN-45, poorly differentiated remain unchanged (Wang et al, 1980). The modified DNA was adenocarcinoma) and KATO-III (signet ring cell carcinoma) were purified using a Wizard DNA cleanup system (Promega, Madison, kindly provided from Professor K Shimizu (Department of WI, USA). Molecular Genetics, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan). NCI-SNU-5 and NCI- Map of 5 -flanking region of hDAB2IP and bisulphited SNU-16 (poorly differentiated adenocarcinoma) were obtained DNA sequencing analysis from the American Type Culture Collection (Manassaa, VA, USA). The other five cell lines (MKN-7, well-differentiated adenocarci- Our previous report identified the promoter region of hDAB2IP noma; NUGC-4, signet ring cell carcinoma; SH-10-TC, poorly (P2) that was divided into m2a (237 bp) and m2b (401 bp) regions differentiated adenocarcinoma; AZ521 and H-111-TC, well-differ- (Chen et al, 2003). The location of the CpG dinucleotides, P2, m2a entiated adenocarcinomas) were obtained from Cell Resource 0 and m2b regions in the 5 -flanking region of hDAB2IP (GenBank Center for Biomedical Research, Tohoku University. Most cells accession number AL365274) are shown in Figure 1A. The m2a were grown in RPMI1640 (Sigma Chemical Co., St Louis, MO, and m2b were amplified and PCR amplicons were cloned into USA) supplemented with 10% heat-inactivated fetal bovine serum pCR2.1-TOPO Vector and sequenced following our previous report and incubated in 5% CO . (Chen et al, 2003). To determine the methylation status of hDAB2IP, individual eight clones from each sample were sequenced by the Applied Biosystems PRISM dye terminator cycle Clinical samples sequencing method (Perkin-Elmer Corp., Foster City, CA, USA) using forward and reverse M13 primers. Surgically resected specimens of 35 primary gastric cancers, 60 primary colorectal cancers and 59 matched corresponding nonmalignant mucosa specimens were obtained from Okayama hDAB2IP mRNA expression by semiquantitative real-time University Hospital, Okayama, Japan after acquiring informed RT–PCR consent from each patient between 1998 and 2002. Patient demographics were shown in Table 1. Histological classification Total cellular RNA was isolated from cell lines and five and tumour staging were carried out according to Lauren nonmalignant gastric mucosa specimens with RNeasy minikit classification (Lauren, 1991) and the TNM Stage Grouping (Sobin (Qiagen Sciences, MD, USA) according to the manufacturer’s and Fleming, 1997). Peripheral blood lymphocytes were obtained instructions and then treated with 2 unitsml of DNase I (Ambion, Austin, TX, USA) for 30 min at 371C. RT reaction was performed from five healthy volunteers. The whole procedure was approved for 2mg of total RNA with the SuperScript II First-strand Synthesis by the Appropriate Institutional Review Board of our centre. using the Oligo (dT) primer system (Invitrogen Life Technologies, Inc., Carlsbad, CA, USA) in 20ml reaction mixture. cDNAs were semiquantified by fluorescence-based real-time RT–PCR using Table 1 Clinical characteristics of studied gastric and colorectal cancer TaqMan technology (Perkin Elmer Corp., Foster City, CA, USA) patients with the Gene Amp 5700 Sequence Detection System (Perkin Elmer Corp.). TATA box binding protein (TBP) was used to normalise No. of informative No. of colorectal the expression of hDAB2IP (Pawlowski et al, 2000; Toyooka et al, gastric cancer cases cancer informative 2001). The sequences of the primers and probe for hDAB2IP and Variables (%) cases (%) TBP expression were previously reported. We used serial dilutions Sex of the positive control cDNA to create a standard curve. The Male 21 (60) 37 (62) expression ratio was defined as the hDAB2IP PCR products Female 14 (40) 23 (38) compared to those of the TBP, multiplied by 100. All experiments were performed in duplicate. Tumour invasion T 1 6 (17) 4 (7) T 2 14 (40) 4 (7) MSP assay T 3 11 (31) 45 (75) T 4 4 (12) 6 (10) The methylation-specific primers were designed based on the bisulphited DNA sequencing data of gastric, breast (Dote et al, Nodal status 2004) and lung (data not shown) cancer cell lines, whose N 0 14 (40) 25 (42) expressions were remarkably downregulated (see below). The N 1 14 (40) 18 (30) location of MSP primers was shown in Figure 1A and B. The N 2 6 (17) 17 (28) methylation status of m2a and m2b regions in gastric and N 3 1 (3) — colorectal cancers was determined by MSP (Herman et al, 1996). The methylated and unmethylated alleles for m2a and m2b regions Distant metastasis were amplified using hDAB2IPm2a-MSP-F and hDAB2IPm2a- M 0 26 (74) 37 (62) M 1 9 (26) 23 (38) MSP-R for methylated m2a, hDAB2IPm2b-MSP-F and hDA- B2IPm2b-MSP-R for methylated m2b, hDAB2IPm2a-USP-F and TNM staging hDAB2IPm2a-USP-R for unmethylated m2a, and hDAB2IPm2b- Stage I 10 (29) 7 (12) USP-F and hDAB2IPm2b-USP-R for unmethylated m2b. The PCR Stage II 8 (23) 13 (22) products were resolved by electrophoresis in a 2% agarose gels Stage III 6 (17) 16 (27) containing ethidium bromide. The normal lymphocyte DNA that Stage IV 11 (31) 23 (38) was treated with Sss1 methyltransferase (New England BioLabs, British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK hDAB2IP methylation in gastrointestinal tumour H Dote et al TSS + 44 −598 + 79 Exon −608 P2 + 1053 I a m2a m2b 200 base −522 −302 −285 + 89 −522 + 89 RMSPm2a RMSPm2b 100 base Bisulphite sequence analysis for m2a Bisulphite sequence analysis for m2b m2a-MSP-F m2a-MSP-R m2b-MSP-F m2b-MSP-R MKN-1 ER = 1.81 M(−) M(−) MT = 2.95% MA = 5.0% MA = 1.91% MKN-7 M(±) M(−) ER = 3.03 MA = 15.4% MA = 33.3% MT = 27.2% NCI-SNU-5 M(±) ER = 1.28 M(±) MA = 71.2% MT = 54.4% MA = 21.3% NCI-SNU-16 M(+) ER = 2.24 M(+) MT = 87.9% MA =90.0% MA = 83.8% SH-10-TC M(−) ER = 13.8 M(±) MA = 2.75% MT = 7.72% MA = 17.5% Figure 1 Map of 5 flanking region of hDAB2IP gene and the bisulphited genomic DNA sequence. (A) Location of CpG dinucleotides in the genomic sequence (GenBank accession number AL365274) are indicated by thin vertical lines. The bent arrow indicates the transcription start site (TSS) ( þ 1). Five rectangular boxes, promoter region (P2), regions for the bisulphited genomic sequence (m2a: from 522 to 285, m2b: from 308 to þ 89) and regions for the amplicons of MSP (RMSPm2a, RMSPm2b). (B) Methylation status of individual cloned DNA fragments of five gastric cancer cell lines is shown. Each row represents one sequenced allele. Each circle represents a CpG dinucleotide. Filled circle, methylation; open circle, no methylation. Clonal numbers are 0 0 indicated by prefix C to the left. The numbers at the top indicate the CpG dinucleotide in the amplicon (5 to 3 ). The percentage of MA, MB and MT indicates the rate of methylated CpG dinucleotides in m2a, m2b and total regions of each sample, respectively. ER means expression ratio by semiquantitative real-time RT –PCR. The location of CpG dinucleotides included in MSP primers (m2a-MSP-F, m2a-MSP-R, m2b-MSP-F and m2b-MSP-R) are indicated by open boxes. M( þ ), positive for the hDAB2IP-methylated form by MSP; M(), negative for the hDAB2IP-methylated form by MSP; M(7), positive for both the hDAB2IP-methylated and -unmethylated forms by MSP. Bevely, MA, USA) and then subjected to bisulphite treatment was Statistical analysis used as a positive control for methylated alleles. Water blanks were The frequencies of hDAB2IP methylation between two groups were included in each assay. compared using the w test. The quantitative ratios of different groups were compared using the Mann–Whitney U-nonpara- metric test. Probability value less than 0.05 was defined as being 5-Aza-CdR and TSA treatment statistically significant. All data were analysed with StatView for Cell lines were treated with 5-Aza-CdR (Sigma-Aldrich Co., St Windows (SAS Institute Inc., Cary, NC, USA). Louis, MO, USA) at a concentration of 1–2mgml for 6 days with medium changes on days 1, 3 and 5 (Jones, 1985; Virmani et al, 2000). The treatment with histone deacetylase inhibitor, TSA RESULTS (Wako, Tokyo, Japan), was performed at a concentration of 150– 300 nM for 12–24 h (Yoshida et al, 1990; Toyooka et al, 2002a). For Bisulphited genomic DNA sequencing combination treatment, TSA was added to day 5 of 5-Aza-CdR- treated cell lines and cultured for additional 12–24 h. Treated or The m2a and m2b regions, covering the major part of the P2 untreated cells from individual triplicate flasks were harvested to region, were separately amplified by PCR to examine their be semiquantified by the gene expression level using real-time methylation status by direct sequencing in eight gastric cancer RT–PCR as described above. cell lines, and five cell lines that showed one of the four typical & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al methylation patterns were selected for more detailed analysis. The regions were almost consistently methylated in the expression- PCR amplicons of these five cell lines were cloned, and eight reduced cell lines (Figure 1B). Three and four cytosines were individual clones were sequenced (Figure 1B). MKN-1 showed included in the forward and reverse primers, respectively, for the almost no methylation in either the m2a or the m2b region. In m2a region producing a 163-bp amplicon, and five and three contrast, NCI-SNU-16 showed a heavy methylation pattern, with cytosines were included in the forward and reverse primers, methylation of over 80% of all the CpG dinucleotides in both the respectively, for the m2b region producing a 209-bp amplicon m2a and m2b regions. The other three cell lines showed a (Figure 1B). Methylation-specific PCR assay using these primers moderate methylation pattern, which was subdivided into two was performed to examine the methylation status of hDAB2IP in types, the m2a- and the m2b-dominant methylation pattern. Two the gastric cancer cell lines and in primary gastric and colorectal cell lines, namely, MKN-7 and NCI-SNU-5, showed an m2b- cancer tumours. Representative examples of the assay are dominant methylation pattern, and one cell line, namely, SH-10 illustrated in Figure 3A and B and the results are summarised in TC, showed an m2a-dominant methylation pattern. Table 2. Aberrant methylation of the m2a and m2b regions was found in 16 of 35 (46%) and 15 of 35 (43%) primary gastric cancer specimens, respectively, and in 29 of 60 (48%) and 28 of 60 (47%) hDAB2IP expression in gastric cancer cell lines primary colorectal cancer specimens, respectively. In the case of the gastric cancer cell lines, methylation of the gene was detected We examined the expression status of hDAB2IP by semiquantita- in eight of 12 cell lines (67%) for each of the m2a and m2b regions. tive real-time RT–PCR in eight cell lines, consisting of three The unmethylated form of hDAB2IP was always found in primary showing heavy methylation in both the m2a and m2b regions, one tumour samples because these had been grossly dissected and thus showing an m2a-dominant methylation pattern, two showing an showed some contamination with normal cells. In contrast, the m2b-dominant methylation pattern, and two showing no methyla- cancer cell lines consisted of a pure population of tumour cells, tion of the hDAB2IP gene. The semiquantitatively estimated mRNA and eight of the 12 cell lines (67%) had either methylated or expression values in each cell line are shown in Figure 2A. The unmethylated hDAB2IP alleles for both m2a and m2b regions. Four values in the two cell lines showing no methylation of the gene of the 12 cell lines (33%) showed both methylated and were as follows; the mean hDAB2IP expression ratio in NUGC-4 unmethylated forms for the m2a or the m2b region, and were was very high at 19.2, as expected, but the ratio in MKN-1 was regarded as showing partial methylation of the gene, and the quite unexpectedly low, with a value of 1.81. This transcriptional details of the CpG methylation status in these cell lines are shown downregulation of hDAB2IP in MKN-1 was attributed to treatment in Figure 1B. A total of 36 (38%) primary gastric and colorectal with a histone deacetylase inhibitor, as described below. As cancer specimens and seven (58%) cell lines showed aberrant compared with the value in NUGC-4, the expression ratios in all of methylation in both the m2a and m2b regions. On the other hand, the cell lines showing methylation of the hDAB2IP gene, with the 16 (17%) primary gastric and colorectal cancer specimens and two exception of SH-10-TC, were significantly lower, with the mean (17%) cell lines showed methylation of either the m2a or the m2b ratios in NCI-SNU-16, MKN-45, KATO-III, NCI-SNU-5 and MKN- region. There were no significant differences in the methylation 7 being 2.24, 3.23, 2.08, 1.28, and 3.03 respectively; the expression rate between the cancer cell lines and primary tumours in either ratio in SH-10-TC was only 13.8. SH-10-TC showed a moderate the m2a or the m2b region. In addition, seven (12%) and 11 (19%) methylation status with the m2a-dominant pattern, but its m2b of 59 gastrointestinal normal mucosal specimens showed methyla- region was almost completely unmethylated, as described above. tion in the m2a and m2b regions, respectively. Aberrant On the other hand, in MKN-7 and NCI-SNU-5, in which the methylation was not present in either of the two regions in expression ratios indicated significant downregulation of hDA- samples obtained from five peripheral blood lymphocytes. B2IP, the m2a-methylation rates were equal to that in SH-10-TC, but the m2b regions were additionally methylated (Figure 1B). These results indicate that the expression of hDAB2IP tended to be 5-Aza-CdR and TSA treatment for gastric cancer cell lines reduced depending on the methylation status in the m2b region, which is just adjacent to the transcriptional-start site, and suggest To confirm the involvement of DNA methylation in hDAB2IP that the m2b promoter region might be the methylation-mediated silencing, we treated five cell lines with 5-Aza-CdR. hDAB2IP key regulator of the expression of hDAB2IP in gastric cancer. In expression was upregulated following the 5-Aza-CdR treatment in the five nonmalignant gastric mucosal specimens examined, the the m2b-methylated cell lines. In contrast, no such upregulation mean expression ratio of hDAB2IP was 21.679.7, indicative of following 5-Aza-CdR treatment was observed in the m2b- normal expression. nonmethylated cell lines. This finding indicates that methylation in the m2b region plays an important role in regulating the gene silencing. Furthermore, we also treated cell lines with TSA, either MSP assay and aberrant methylation of hDAB2IP in singly or in combination with 5-Aza-CdR. Expression of hDAB2IP gastrointestinal tumours in MKN-7 was upregulated 3.1-fold by TSA treatment and by 6.9- fold by following combined treatment with TSA and 5-Aza-CdR. In We determined the forward and reverse primers for MSP of each of NCI-SNU-5, while hDAB2IP expression was upregulated 3.0-fold the m2a and m2b regions (Figure 1B). The CpG sites in these Figure 2 hDAB2IP mRNA expression using semiquantitative real-time RT –PCR. Triplicate sets of flasks before and after drug treatment were harvested and tested for gene expression using semiquantitative real-time PCR. Data shown are mean values7s.d., n ¼ 3. (A) The relative ratio of mRNA expression in each cell line. The relative expression ratio was remarkably repressed in m2b-methylated cell lines compared with m2b-unmethylated ones except MKN- 1. Following symbols of two plus and minus means positive and negative for methylation status by MSP, respectively. One plus means positive for both the methylated and unmethylated forms by MSP. (B) Effect of 5-Aza-CdR, TSA and combination on mRNA expression in m2b-methylated cell lines (MKN-7, NCI-SNU-5 and NCI-SNU-16). hDAB2IP expression of MKN-7 was upregulated 3.4-fold by 5-Aza-CdR, 3.1-fold by TSA and additively 6.9-fold by combination. In NCI-SNU-5, hDAB2IP expression was also upregulated 3.0-fold by 5-Aza-CdR and 3.0-fold by TSA, whereas TSA failed to make synergistic or additive effect for this cell line. In NCI-SNU-16, the expression was also upregulated 3.1-fold by 5-Aza-CdR, but TSA alone failed to make a positive effect, operating synergically to 6.6-fold on the expression with a combination of 5-Aza-CdR. (C) Effect of 5-Aza-CdR, TSA and combination on mRNA expression in m2b-unmethylated cell lines (MKN1 and SH-10-TC). In MKN-1, the downregulated expression was restored 3.1-fold by TSA alone. In SH-10-TC, the expression was decreased after any drug treatment, which may be caused by a cytotoxic effect from drug exposure. British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK hDAB2IP methylation in gastrointestinal tumour H Dote et al by treatment with TSA alone, combined treatment with TSA and 5- On the other hand, in SH-10-TC, an m2b-nonmethylated cell Aza-CdR failed to elicit a synergistic or additive effect. In NCI- line, treatment with the aforementioned drugs neither upregulated SNU-16, while treatment with TSA alone failed to have any positive nor downregulated the expression of hDAB2IP (0.1–0.4-fold), effect, combined treatment with 5-Aza-CdR elicited a positive because the number of SH-10-TC cells in culture was decreased by effect on hDAB2IP expression. treatment with 5-Aza-CdR, TSA or a combination of 5-Aza-CdR Cell line with unmethylated m2b region Cell line with partially methylated m2b region Cell line with heavily methylated m2b region Methylation status ++ ++ + + −− + − − of m2a region Methylation status ++ ++ ++ + −− − − of m2a region MKN-7 NCI-SNU- 5 NCI-SNU-16 P < 0.001 P = 0.005 P < 0.001 P = 0.008 P < 0.001 P = 0.022 P = 0.002 P = 0.002 P = 0.003 25 5 20 4 15 3 10 2 5 1 0 0 0 MKN-1 SH-10-TC P = N.S P < 0.001 P = 0.002 P = 0.001 P = N.S P < 0.001 0 0 & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Control (no treatment) 5-Aza-dC TSA Combination Control (no treatment) 5-Aza-dC Control (no treatment) TSA 5-Aza-dC Combination TSA Combination Control (no treatment) Control 5-Aza-dC (no treatment) TSA 5-Aza-dC Combination TSA Combination Semiquantitative expression ratio Semiquantitative expression ratio Semiquantitative expression ratio Semiquantitative NCI-SNU-16 expression ratio MKN-45 KATO-III Semiquantitative expression ratio NCI-SNU- 5 MKN-7 SH-10-TC Semiquantitative MKN-1 expression ratio NUGC-4 Normal mucosa Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al For m2a region Methylation-specific PCR Unmethylation-specific PCR For m2b region Methylation-specific PCR Unmethylation-specific PCR Gastric cancer For m2a region Methylation-specific PCR Unmethylation-specific PCR For m2b region Methylation-specific PCR Unmethylation-specific PCR Colorectal cancer For m2a region Methylation-specific PCR Unmethylation-specific PCR For m2b region Methylation-specific PCR Unmethylation-specific PCR British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK G1 G1 MKN-1 G2 G2 G3 G3 MKN-7 G4 G4 MKN-28 G5 G5 G6 G6 MKN-45 G7 G7 G8 MKN-74 G8 G9 G9 KATO-III G10 G10 G11 NCI-SNU-5 G11 G12 G12 NCI-SNU-16 G13 G13 G14 NUGC-4 G14 G15 G15 SH-10-TC G16 G16 G17 G17 AZ521 G18 G18 H-111-TC G19 G19 G20 G20 Positive control G21 G21 Water blank G22 G22 hDAB2IP methylation in gastrointestinal tumour H Dote et al and TSA (data not shown). This effect may be attributable to the sex, age, tumour size, extent of tumour invasion, lymph node cytotoxic effect of 5-Aza-CdR and TSA. Interestingly, in MKN-1, status, distant metastasis, TNM stage, histology and the tumour which showed no methylation in either the m2a or the m2b region, site in cases with gastrointestinal cancers. In cases of gastric the downregulated expression of hDAB2IP could be restored by cancer, while both the m2a- and m2b-methylation status were TSA treatment alone. This result suggests that the expression of the significantly associated with the tumour site, the m2b-methylation gene might be downregulated under the control of histone status showed a stronger correlation than the m2a-methylation deacetylation in this cell line alone. These results suggest that status (m2a, P ¼ 0.0172; m2b, P ¼ 0.0064; Table 3). In cases of both DNA methylation and histone deacetylation may act colorectal cancer, m2b methylation tended to be associated with cooperatively to silence hDAB2IP expression in some gastric cancer arising from the left side of the colon. (P ¼ 0.0926; Table 4). cancer cell lines, and that the methylation status of the m2b region hDAB2IP methylation was often observed in tumours arising from may act as the major gene silencing mechanism in most cell lines. the lower-third of the stomach and distal aspect of the colorectum. There were no significant differences between methylation-positive and methylation-negative groups with respect to the other factors hDAB2IP methylation and clinicopathological correlation examined in this study. Although accumulation of methylation appears to be age-related in a number of cancer-related genes, both We examined the relationship between the methylation status in the m2a- and m2b-methylation status were not significantly the tumour specimens and clinicopathological factors, including associated with the age of patients (gastric cancer: m2a, P ¼ 0.9369; m2b, P ¼ 0.9788; colorectal cancer: m2a, P ¼ 0.8484; m2b, P ¼ 0.5709; Table 5). Table 2 Aberrant promoter methylation of hDAB2IP in samples No. of No. of No. of both hDAB2IP hDAB2IP m2a and DISCUSSION m2a m2b m2b To determine the methylation status of the 5 CpG islands of methylated methylated methylated Samples samples (%) samples (%) samples (%) hDAB2IP in gastrointestinal tumours, we first performed bisul- phite genomic DNA sequencing in several gastric cancer cell lines. Gastric cancers Based on the sequence data, methylation-specific primers were Cell lines (n ¼ 12) 8 (67) 8 (67) 7 (58) designed to detect sensitively each methylated or unmethylated Primary tumours (n ¼ 35) 16 (46) 15 (43) 13 (37) allele (Toyooka et al, 2002b, 2003). These primers have also been shown to be applicable to lung and breast cancer cell lines (Dote Colorectal cancers et al, 2004). Using the MSP primers designed thus, we found Primary tumours (n ¼ 60) 29 (48) 28 (47) 23 (38) frequent methylation of the hDAB2IP gene in primary gastro- intestinal tumours, as well as in gastrointestinal tumour cell lines. Nonmalignant specimens (n ¼ 64) Normal gastric mucosa 2 (6) 7 (20) 2 (6) The higher frequency of methylation in the cell lines as compared (n ¼ 35) with that in the primary tumours might be explained by the Normal colorectal mucosa 5 (21) 4 (17) 3 (13) additional changes that the cells may acquire in culture, or by the (n ¼ 24) tumour cell lines being derived from more aggressive tumours and Peripheral lymphocytes 0 (0) 0 (0) 0 (0) therefore showing more malignant changes (Zochbauer-Muller (n ¼ 5) et al, 2001). Table 3 Clinicopathologic features and hDAB2IP promoter methylation in gastric cancer Frequency of hDAB2IP Frequency of hDAB2IP a b b Clinicopathologic features Variable m2a methylation (%) P-value m2b methylation (%) P-value Sex Male (n ¼ 21) 8 (38) 0.27 9 (43) 40.99 Female (n ¼ 14) 8 (57) 6 (43) Tumour invasion T1, 2 (n ¼ 20) 9 (45) 0.92 9 (45) 0.77 T3, 4 (n ¼ 15) 7 (47) 6 (40) Nodal metastasis N0 (n ¼ 14) 7 (50) 0.68 7 (50) 0.49 N1, 2 (n ¼ 21) 9 (43) 8 (38) Metastasis M0 (n ¼ 26) 12 (46) 0.93 11 (42) 0.91 M1 (n ¼ 9) 4 (44) 4 (44) TNM stage 1, 2 (n ¼ 18) 8 (44) 0.88 8 (44) 0.85 3, 4 (n ¼ 17) 8 (47) 7 (41) Histology Intestinal (n ¼ 17) 8 (47) 0.88 6 (35) 0.38 Diffuse (n ¼ 18) 8 (44) 9 (50) Tumour site U (n ¼ 7) 2 (22) 0.017 1 (14) 0.0064 M(n ¼ 13) 3 (23) 3 (23) L(n ¼ 15) 11 (73) 11 (73) a b 2 U ¼ upper site; M ¼ middle site; L ¼ lower site. Statistical significance was determined using w test. Figure 3 Representative examples of MSP in gastrointestinal tumours. (A) gastric cancer cell lines, (B) primary tumours. The unmethylated form of hDAB2IP was always found in primary tumours that had been grossly dissected and thus had at least some contamination with normal cells. Positive controls include peripheral blood lymphocytes from healthy individuals for the unmethylated form and normal lymphocyte DNA methylated by treatment with Sss I DNA methyltransferase for the methylated form. & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al Table 4 Clinicopathologic features and hDAB2IP promoter methylation in colorectal cancer Frequency of hDAB2IP Frequency of hDAB2IP a b b Clinicopathologic features Variable m2a methylation (%) P-value m2b methylation (%) P-value Sex Male (n ¼ 37) 18 (49) 0.95 15 (41) 0.23 Female (n ¼ 23) 11 (48) 13 (57) Tumour invasion T1, 2 (n ¼ 9) 4 (44) 0.80 5 (56) 0.56 T3, 4 (n ¼ 51) 25 (49) 23 (45) Nodal metastasis N0 (n ¼ 25) 9 (36) 0.11 11 (44) 0.73 N1, 2 (n ¼ 35) 20 (57) 17 (49) Metastasis M0 (n ¼ 53) 26 (49) 0.76 25 (47) 0.83 M1 (n ¼ 7) 3 (43) 3 (43) TNM stage 1, 2 (n ¼ 21) 7 (33) 0.088 9 (43) 0.66 3, 4 (n ¼ 39) 22 (56) 19 (49) Histological grade G1 (n ¼ 13) 7 (54) 0.65 6 (46) 0.97 G2, 3, 4 (n ¼ 47) 22 (47) 22 (47) Tumour site C, A, T (n ¼ 12) 4 (33) 0.25 3 (25) 0.093 D, S, R (n ¼ 48) 25 (52) 25 (52) a b 2 C ¼ cecum; A ¼ ascending colon; T ¼ transverse colon; D ¼ descending colon; S ¼ sigmoid colon; R ¼ rectum. Statistical significance was determined using w test. Table 5 Association between hDAB2IP promoter methylation status and examined the effect of the histone deacetylase inhibitor, TSA, age in gastrointestinal cancer as well as that of 5-Aza-CdR on the methylated cell lines, to distinguish the effect of histone deacetylation on the down- a b Methylation status Age P-value regulated gene expression of hDAB2IP. In MKN-7, the expression was upregulated by both 5-Aza-CdR and TSA treatment, Gastric cancer and combined treatment with both agents caused synergistic m2a-methylation (n ¼ 15) 66 (40 – 89) 0.9369 upregulation, with the extent of upregulation in this case m2a-unmethylation (n ¼ 20) 66 (36 – 85) being greater than that with either agent alone. In addition, in m2b-methylation (n ¼ 16) 66 (40 – 89) 0.9788 some other methylated cells, TSA enhanced hDAB2IP m2b-unmethylation (n ¼ 19) 66 (36 – 85) gene expression. These results indicate that in some cell lines Colorectal cancer showing gene methylation, there may be a synergistic and m2a-methylation (n ¼ 29) 66 (51 – 85) 0.8484 positive interaction between DNA methylation and histone m2a-unmethylation (n ¼ 31) 65 (26 – 86) deacetylation for transcriptional regulation, although this m2b-methylation (n ¼ 28) 66 (47 – 82) 0.5709 does not seem to be a common and major mechanism in gastric m2b-unmethylation (n ¼ 32) 65 (26 – 86) cancer cells. However, the number of examined cells was a b too limited to arrive at any definitive conclusion on the effect Mean age (range) in years. Statistical significance was determined using Mann – Whitney U-test. of TSA on hDAB2IP expression. Considering this effect of deacetylation on gene expression, it was reported that the expression of some methylation-repressed genes can be reactivated by TSA treatment (Yang et al, 2000) and that demethylation- We previously reported that the m2a region appeared to be the induced gene re-expression could be potentiated by TSA (Cameron et al, 1999). In addition, Suzuki et al (2002) analysed the key regulatory region for hDAB2IP expression in prostate cancer (Chen et al, 2003) and that methylation of both the m2a and m2b expression of over 10 000 genes by microarray analysis, and found regions occurred in breast cancer cell lines (Dote et al, 2004). that the expression of 74 genes was upregulated following However, in contrast to the observation in prostate and breast treatment with 5-Aza-CdR or TSA. They also showed that cancer, this study demonstrated that the downregulation of transcriptional silencing is mediated by both methylation and hDAB2IP expression correlated more with the methylation status histone deacetylation, although the effect of methylation was of not the m2a region, but of the m2b region, which is located just dominant for gene silencing. adjacent to exon Ia in gastric cancer cell lines (Figure 2A). hDAB2IP methylation was found to be present at similar Considering the close proximity of the m2b region to the frequencies in both early and advanced carcinomas, suggesting transcriptional-start site of the hDAB2IP gene, it stands to reason that the methylation actually occurs at an early stage of that the m2b region might act as the regulatory promoter in gastric carcinogenesis in gastrointestinal tumours. Furthermore, the result cancer cells. that hDAB2IP methylation was found in nonmalignant mucosa We demonstrated in this study that the expression of hDAB2IP suggested that hDAB2IP methylation might play a causative role in was highly suppressed in five m2b-methylated cell lines as carcinogenesis. Interestingly, we found a highly significant compared with that in two m2b-nonmethylated cell lines. This association between the tumour location in the stomach and result suggests that DNA methylation of the regulatory promoter methylation of the m2b region, which has been supposed to be the region may be critical for transcriptional regulation of the key regulatory sequence, as described above. Cancers showing hDAB2IP gene and play an important role in hDAB2IP inactiva- methylation of the m2b region showed a significant tendency tion. These results are consistent with the finding of restoration of towards being preferentially located in the lower-third of the hDAB2IP expression in methylated cell lines after 5-Aza-CdR stomach. Similar results were also described for p16, hMLH1 and treatment. TSLC1 methylation in gastric cancer (Schneider et al, 2000; Honda Recent data suggest that hypermethylated DNA interacts with et al, 2002; Sakata et al, 2002). Therefore, methylation of several several methyl-CpG-binding proteins, and that the interaction tumour-suppressor and tumour-related genes appears to be an facilitates the assembly of a repressive complex containing histone important pathogenetic mechanism of gastric cancers arising from deacetylase and formation of an inactive chromatin that leads to the antrum. Because intestinal metaplasia commonly arises in the gene silencing (Nan et al, 1998; Wade et al, 1999). Therefore, we antrum and then extends towards the body of the stomach, British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK hDAB2IP methylation in gastrointestinal tumour H Dote et al intestinal metaplasia may predispose to promoter region methyla- a right-sided location of the tumour, although this did not reach tion in these genes (Sato et al, 2001). In addition, a marked statistical significance. increase in the frequency of methylated genes from nonmetaplastic In conclusion, our results demonstrate that aberrant methyla- mucosa to intestinal metaplasia has been reported (Kang et al, tion drives gene downregulation, and supports the contention that 2001). In sporadic colorectal cancers, it has recently been shown the methylation-mediated transcriptional silencing of the hDAB2IP that CpG island methylation is associated with a right-sided gene may be a critical event in tumorigenesis of gastrointestinal location of the tumour (Hawkins et al, 2002). However, we also tumours, especially those arising from the antrum of the stomach found an inverse association between m2b-region methylation and in association with intestinal metaplasia. REFERENCES Baylin SB, Herman JG (2000) DNA hypermethylation in tumorigenesis: Sobin LH, Fleming ID (1997) TNM Classification of Malignant Tumors, epigenetics joins genetics. Trends Genet 16: 168–174 fifth edition (1997). Union Internationale Contre le Cancer and the Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB (1999) American Joint Committee on Cancer. Cancer 80: 1803–1804 Synergy of demethylation and histone deacetylase inhibition in the re- Suzuki H, Gabrielson E, Chen W, Anbazhagan R, van Engeland M, expression of genes silenced in cancer. Nat Genet 21: 103–107 Weijenberg MP, Herman JG, Baylin SB (2002) A genomic screen for Chen H, Pong RC, Wang Z, Hsieh JT (2002) Differential regulation of the genes upregulated by demethylation and histone deacetylase inhibition human gene DAB2IP in normal and malignant prostatic epithelia: in human colorectal cancer. Nat Genet 31: 141–149 cloning and characterization. Genomics 79: 573–581 Toyooka S, Carbone M, Toyooka KO, Bocchetta M, Shivapurkar N, Minna Chen H, Toyooka S, Gazdar AF, Hsieh JT (2003) Epigenetic regulation of a JD, Gazdar AF (2002a) Progressive aberrant methylation of the RASSF1A novel tumor suppressor gene (hDAB2IP) in prostate cancer cell lines. J gene in simian virus 40 infected human mesothelial cells. Oncogene 21: Biol Chem 278: 3121– 3130 4340– 4344 Dote H, Toyooka S, Tsukuda K, Yano M, Ouchida M, Doihara H, Suzuki M, Toyooka S, Toyooka KO, Harada K, Miyajima K, Makarla P, Sathyanar- Chen H, Hsieh JT, Gazdar AF, Shimizu N (2004) Aberrant promoter ayana UG, Yin J, Sato F, Shivapurkar N, Meltzer SJ, Gazdar AF (2002b) methylation in human DAB2 interactive protein (hDAB2IP) gene in Aberrant methylation of the CDH13 (H-cadherin) promoter region in breast cancer. Clin Cancer Res 10: 2082–2089 colorectal cancers and adenomas. Cancer Res 62: 3382–3386 Fulop V, Colitti CV, Genest D, Berkowitz RS, Yiu GK, Ng SW, Szepesi J, Toyooka S, Toyooka KO, Maruyama R, Virmani AK, Girard L, Miyajima K, Mok SC (1998) DOC-2/hDab2, a candidate tumor suppressor gene Harada K, Ariyoshi Y, Takahashi T, Sugio K, Brambilla E, Gilcrease M, involved in the development of gestational trophoblastic diseases. Minna JD, Gazdar AF (2001) DNA methylation profiles of lung tumors. Oncogene 17: 419–424 Mol Cancer Ther 1: 61–67 Hawkins N, Norrie M, Cheong K, Mokany E, Ku SL, Meagher A, O’Connor Toyooka S, Toyooka KO, Miyajima K, Reddy JL, Toyota M, Sathyanarayana T, Ward R (2002) CpG island methylation in sporadic colorectal cancers UG, Padar A, Tockman MS, Lam S, Shivapurkar N, Gazdar AF (2003) and its relationship to microsatellite instability. Gastroenterology 122: Epigenetic down-regulation of death-associated protein kinase in lung 1376– 1387 cancers. Clin Cancer Res 9: 3034– 3041 Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB (1996) Tseng CP, Ely BD, Li Y, Pong RC, Hsieh JT (1998) Regulation of rat DOC-2 Methylation-specific PCR: a novel PCR assay for methylation status of gene during castration-induced rat ventral prostate degeneration and its CpG islands. Proc Natl Acad Sci USA 93: 9821 –9826 growth inhibitory function in human prostatic carcinoma cells. Honda T, Tamura G, Waki T, Jin Z, Sato K, Motoyama T, Kawata S, Kimura Endocrinology 139: 3542–3553 W, Nishizuka S, Murakami Y (2002) Hypermethylation of the TSLC1 Virmani AK, Rathi A, Zochbauer-Muller S, Sacchi N, Fukuyama Y, Bryant gene promoter in primary gastric cancers and gastric cancer cell lines. D, Maitra A, Heda S, Fong KM, Thunnissen F, Minna JD, Gazdar AF Jpn J Cancer Res 93: 857–860 (2000) Promoter methylation and silencing of the retinoic acid receptor- Jones PA (1985) Altering gene expression with 5-azacytidine. Cell 40: 485–486 beta gene in lung carcinomas. J Natl Cancer Inst 92: 1303–1307 Jones PA, Laird PW (1999) Cancer epigenetics comes of age. Nat Genet 21: Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP (1999) Mi- 163–167 2 complex couples DNA methylation to chromatin remodelling and Kang GH, Shim YH, Jung HY, Kim WH, Ro JY, Rhyu MG (2001) CpG island histone deacetylation. Nat Genet 23: 62–66 methylation in premalignant stages of gastric carcinoma. Cancer Res 61: Wang RY, Gehrke CW, Ehrlich M (1980) Comparison of bisulfite 2847– 2851 modification of 5-methyldeoxycytidine and deoxycytidine residues. Lauren P (1991) Histogenesis of intestinal and diffuse types of gastric Nucleic Acids Res 8: 4777–4790 carcinoma. Scand J Gastroenterol Suppl 180: 160–164 Wang Z, Tseng CP, Pong RC, Chen H, McConnell JD, Navone N, Hsieh JT Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, (2002) The mechanism of growth-inhibitory effect of DOC-2/DAB2 in Bird A (1998) Transcriptional repression by the methyl-CpG-binding prostate cancer. Characterization of a novel GTPase-activating protein protein MeCP2 involves a histone deacetylase complex. Nature 393: associated with N-terminal domain of DOC-2/DAB2. J Biol Chem 277: 386–389 12622–12631 Pawlowski V, Revillion F, Hebbar M, Hornez L, Peyrat JP (2000) Prognostic Yang X, Ferguson AT, Nass SJ, Phillips DL, Butash KA, Wang SM, Herman value of the type I growth factor receptors in a large series of human JG, Davidson NE (2000) Transcriptional activation of estrogen receptor primary breast cancers quantified with a real-time reverse transcription- alpha in human breast cancer cells by histone deacetylase inhibition. polymerase chain reaction assay. Clin Cancer Res 6: 4217 –4225 Cancer Res 60: 6890–6894 Sakata K, Tamura G, Endoh Y, Ohmura K, Ogata S, Motoyama T Yoshida M, Kijima M, Akita M, Beppu T (1990) Potent and specific (2002) Hypermethylation of the hMLH1 gene promoter in solitary and inhibition of mammalian histone deacetylase both in vivo and in vitro by multiple gastric cancers with microsatellite instability. Br J Cancer 86: trichostatin A. J Biol Chem 265: 17174–17179 564–567 Zhou J, Hsieh JT (2001) The inhibitory role of DOC-2/DAB2 in growth Sato K, Tamura G, Tsuchiya T, Endoh Y, Usuba O, Kimura W, Motoyama T factor receptor-mediated signal cascade. DOC-2/DAB2-mediated inhibi- (2001) Frequent loss of expression without sequence mutations of the tion of ERK phosphorylation via binding to Grb2. J Biol Chem 276: DCC gene in primary gastric cancer. Br J Cancer 85: 199–203 27793–27798 Schneider BG, Gulley ML, Eagan P, Bravo JC, Mera R, Geradts J (2000) Loss Zochbauer-Muller S, Fong KM, Maitra A, Lam S, Geradts J, Ashfaq R, of p16/CDKN2A tumor suppressor protein in gastric adenocarcinoma is Virmani AK, Milchgrub S, Gazdar AF, Minna JD (2001) 5 CpG island associated with Epstein-Barr virus and anatomic location in the body of methylation of the FHIT gene is correlated with loss of gene expression the stomach. Hum Pathol 31: 45–50 in lung and breast cancer. Cancer Res 61: 3581– 3585 & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Molecular Diagnostics http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png British Journal of Cancer Springer Journals

Aberrant promoter methylation in human DAB2 interactive protein (hDAB2IP) gene in gastrointestinal tumour

Download PDF
9 pages

Loading next page...
 
/lp/springer-journals/aberrant-promoter-methylation-in-human-dab2-interactive-protein-pITpOc1aVO

References (41)

Publisher
Springer Journals
Copyright
Copyright © 2005 by The Author(s)
Subject
Biomedicine; Biomedicine, general; Cancer Research; Epidemiology; Molecular Medicine; Oncology; Drug Resistance
ISSN
0007-0920
eISSN
1532-1827
DOI
10.1038/sj.bjc.6602458
Publisher site
See Article on Publisher Site

Abstract

British Journal of Cancer (2005) 92, 1117 – 1125 & 2005 Cancer Research UK All rights reserved 0007 – 0920/05 $30.00 www.bjcancer.com Aberrant promoter methylation in human DAB2 interactive protein (hDAB2IP) gene in gastrointestinal tumour 1 ,1 1 1 1 1 1 2,3 4,5 H Dote , S Toyooka , K Tsukuda , M Yano , T Ota , M Murakami , M Naito , M Toyota , AF Gazdar and N Shimizu 1 2 Department of Cancer and Thoracic Surgery, Okayama University Graduate School of Medicine and Dentistry, Okayama 700-8558, Japan; First Department of Internal Medicine, Cancer Research Institute, Sapporo Medical University, Sapporo 060-8543, Japan; Department of Molecular Biology, Cancer Research Institute, Sapporo Medical University, Sapporo 060-8543, Japan; Hamon Center for Therapeutic Oncology Research, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA; Department of Pathology, University of Texas Southwestern Medical Center, Dallas, TX 75390, USA The human DOC-2/DAB2 interactive protein (hDAB2IP) gene is a novel member of the Ras GTPase-activating family and has been demonstrated to be a tumour-suppressor gene inactivated by methylation in several cancers. In this study, we analysed the methylation and expression status of hDAB2IP in gastrointestinal tumours. The promoter region of hDAB2IP was divided into two regions (m2a and m2b) based on our previous report, and the methylation status was determined by bisulphite DNA sequencing in gastric cancer cell lines. The gene expression was semiquantified by real-time RT–PCR, and the results indicated that the m2b promoter region might be an authentic methylation-mediated key regulator of the gene expression. Based on the sequence data, we developed a methylation-specific PCR (MSP) for the m2a and m2b regions and applied it to the samples. Methylation-specific PCR revealed aberrant methylation in the m2a region in eight of 12 gastric cancer cell lines (67%), 16 of 35 gastric cancer tissues (46%) and 29 of 60 colorectal cancer tissues (48%), and in the m2b region in eight of 12 cell lines (67%), 15 of 35 gastric cancer tissues (43%) and 28 of 60 colorectal cancer tissues (47%). On the other hand, seven (12%) and 11 (19%) of 59 gastrointestinal nonmalignant mucosal specimens showed methylation in the m2a and m2b regions, respectively, suggesting that hDAB2IP methylation might play a causative role in carcinogenesis. The 5-aza-2 -deoxycytidine treatment restored the gene expression in the m2b-methylated cell lines, confirming that the methylation caused gene downregulation. We also examined the relationship between hDAB2IP methylation and the clinicopathological features in patients with primary tumours, and determined that methylation in the m2b region was associated with location of the tumour in the stomach. In summary, our results demonstrated that hDAB2IP methylation is frequently present in gastrointestinal tumours and that the resulting gene silencing plays an important role in gastrointestinal carcinogenesis. British Journal of Cancer (2005) 92, 1117–1125. doi:10.1038/sj.bjc.6602458 www.bjcancer.com Published online 15 March 2005 & 2005 Cancer Research UK Keywords: methylation; hDAB2IP; gastrointestinal tumour; histone acetylation; methylation specific PCR Aberrant methylation of normally unmethylated CpG islands form a unique protein complex and exert a negative regulatory located in the 5 promoter region of genes has been shown to be activity on the Ras-mediated signal pathway (Wang et al, 2002). associated with transcriptional inactivation of several genes in Recently, it was reported that the P2 region (598 to þ 44) of the human cancers, including gastrointestinal tumours, and appears to hDAB2IP gene shows promoter activity, and that transcriptional serve as an alternative to mutational inactivation (Jones and Laird, silencing by aberrant methylation of the P2 region is critical 1999; Baylin and Herman, 2000). tumorigenesis in prostate cancer (Chen et al, 2003) and breast The human DOC-2/DAB2 interactive protein gene (hDAB2IP), cancer (Dote et al, 2004). located on chromosome 9q33.1–q33.3, is a novel member of the In this study, we examined the methylation and expression Ras GTPase-activating family (Tseng et al, 1998; Chen et al, 2002; status of hDAB2IP in a series of gastric cancer cell lines, and Wang et al, 2002). It interacts directly with the DOC-2/DAB2 (also developed the methylation-specific PCR (MSP) assay for hDAB2IP known as differentially expressed in ovarian carcinoma-2, DOC-2) for application to gastrointestinal tumour samples. To clarify the protein that appears to be a tumour suppressor in malignant mechanism of gene silencing, the cell lines were treated with 5-aza- tumours, including mammary, prostate and ovarian cancers 2 -deoxycytidine (5-Aza-CdR) and/or trichostatin A (TSA) to (Fulop et al, 1998; Zhou and Hsieh, 2001). hDAB2IP and DAB2 examine the effect of methylation and histone deacetylation on the gene silencing. In addition, we analysed the relationship between the methylation status and the clinicopathological features in cases *Correspondence: Dr S Toyooka; E-mail: [email protected] with surgically resected gastric and colorectal cancers, in order to Received 16 August 2004; revised 5 January 2005; accepted 19 January investigate the clinical and pathogenetic importance of hDAB2IP 2005; published online 15 March 2005 methylation in gastrointestinal tumours. Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al MATERIALS AND METHODS DNA extraction and bisulphite treatment DNA from cell lines and specimens was subjected to bisulphite Cell lines treatment as described previously (Dote et al, 2004). Briefly for In all, 12 cell lines derived from human gastric cancer were used in bisulphite treatment, 1mg of genomic DNA was denatured by this study. Four gastric cancer cell lines of MKN series (MKN-1, NaOH and modified by sodium bisulphite, which converts all adenosquamous cell carcinoma; MKN-28 and MKN-74, well- unmethylated cytosines to uracils while methylated cytosines differentiated adenocarcinoma; MKN-45, poorly differentiated remain unchanged (Wang et al, 1980). The modified DNA was adenocarcinoma) and KATO-III (signet ring cell carcinoma) were purified using a Wizard DNA cleanup system (Promega, Madison, kindly provided from Professor K Shimizu (Department of WI, USA). Molecular Genetics, Okayama University Graduate School of Medicine and Dentistry, Okayama, Japan). NCI-SNU-5 and NCI- Map of 5 -flanking region of hDAB2IP and bisulphited SNU-16 (poorly differentiated adenocarcinoma) were obtained DNA sequencing analysis from the American Type Culture Collection (Manassaa, VA, USA). The other five cell lines (MKN-7, well-differentiated adenocarci- Our previous report identified the promoter region of hDAB2IP noma; NUGC-4, signet ring cell carcinoma; SH-10-TC, poorly (P2) that was divided into m2a (237 bp) and m2b (401 bp) regions differentiated adenocarcinoma; AZ521 and H-111-TC, well-differ- (Chen et al, 2003). The location of the CpG dinucleotides, P2, m2a entiated adenocarcinomas) were obtained from Cell Resource 0 and m2b regions in the 5 -flanking region of hDAB2IP (GenBank Center for Biomedical Research, Tohoku University. Most cells accession number AL365274) are shown in Figure 1A. The m2a were grown in RPMI1640 (Sigma Chemical Co., St Louis, MO, and m2b were amplified and PCR amplicons were cloned into USA) supplemented with 10% heat-inactivated fetal bovine serum pCR2.1-TOPO Vector and sequenced following our previous report and incubated in 5% CO . (Chen et al, 2003). To determine the methylation status of hDAB2IP, individual eight clones from each sample were sequenced by the Applied Biosystems PRISM dye terminator cycle Clinical samples sequencing method (Perkin-Elmer Corp., Foster City, CA, USA) using forward and reverse M13 primers. Surgically resected specimens of 35 primary gastric cancers, 60 primary colorectal cancers and 59 matched corresponding nonmalignant mucosa specimens were obtained from Okayama hDAB2IP mRNA expression by semiquantitative real-time University Hospital, Okayama, Japan after acquiring informed RT–PCR consent from each patient between 1998 and 2002. Patient demographics were shown in Table 1. Histological classification Total cellular RNA was isolated from cell lines and five and tumour staging were carried out according to Lauren nonmalignant gastric mucosa specimens with RNeasy minikit classification (Lauren, 1991) and the TNM Stage Grouping (Sobin (Qiagen Sciences, MD, USA) according to the manufacturer’s and Fleming, 1997). Peripheral blood lymphocytes were obtained instructions and then treated with 2 unitsml of DNase I (Ambion, Austin, TX, USA) for 30 min at 371C. RT reaction was performed from five healthy volunteers. The whole procedure was approved for 2mg of total RNA with the SuperScript II First-strand Synthesis by the Appropriate Institutional Review Board of our centre. using the Oligo (dT) primer system (Invitrogen Life Technologies, Inc., Carlsbad, CA, USA) in 20ml reaction mixture. cDNAs were semiquantified by fluorescence-based real-time RT–PCR using Table 1 Clinical characteristics of studied gastric and colorectal cancer TaqMan technology (Perkin Elmer Corp., Foster City, CA, USA) patients with the Gene Amp 5700 Sequence Detection System (Perkin Elmer Corp.). TATA box binding protein (TBP) was used to normalise No. of informative No. of colorectal the expression of hDAB2IP (Pawlowski et al, 2000; Toyooka et al, gastric cancer cases cancer informative 2001). The sequences of the primers and probe for hDAB2IP and Variables (%) cases (%) TBP expression were previously reported. We used serial dilutions Sex of the positive control cDNA to create a standard curve. The Male 21 (60) 37 (62) expression ratio was defined as the hDAB2IP PCR products Female 14 (40) 23 (38) compared to those of the TBP, multiplied by 100. All experiments were performed in duplicate. Tumour invasion T 1 6 (17) 4 (7) T 2 14 (40) 4 (7) MSP assay T 3 11 (31) 45 (75) T 4 4 (12) 6 (10) The methylation-specific primers were designed based on the bisulphited DNA sequencing data of gastric, breast (Dote et al, Nodal status 2004) and lung (data not shown) cancer cell lines, whose N 0 14 (40) 25 (42) expressions were remarkably downregulated (see below). The N 1 14 (40) 18 (30) location of MSP primers was shown in Figure 1A and B. The N 2 6 (17) 17 (28) methylation status of m2a and m2b regions in gastric and N 3 1 (3) — colorectal cancers was determined by MSP (Herman et al, 1996). The methylated and unmethylated alleles for m2a and m2b regions Distant metastasis were amplified using hDAB2IPm2a-MSP-F and hDAB2IPm2a- M 0 26 (74) 37 (62) M 1 9 (26) 23 (38) MSP-R for methylated m2a, hDAB2IPm2b-MSP-F and hDA- B2IPm2b-MSP-R for methylated m2b, hDAB2IPm2a-USP-F and TNM staging hDAB2IPm2a-USP-R for unmethylated m2a, and hDAB2IPm2b- Stage I 10 (29) 7 (12) USP-F and hDAB2IPm2b-USP-R for unmethylated m2b. The PCR Stage II 8 (23) 13 (22) products were resolved by electrophoresis in a 2% agarose gels Stage III 6 (17) 16 (27) containing ethidium bromide. The normal lymphocyte DNA that Stage IV 11 (31) 23 (38) was treated with Sss1 methyltransferase (New England BioLabs, British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK hDAB2IP methylation in gastrointestinal tumour H Dote et al TSS + 44 −598 + 79 Exon −608 P2 + 1053 I a m2a m2b 200 base −522 −302 −285 + 89 −522 + 89 RMSPm2a RMSPm2b 100 base Bisulphite sequence analysis for m2a Bisulphite sequence analysis for m2b m2a-MSP-F m2a-MSP-R m2b-MSP-F m2b-MSP-R MKN-1 ER = 1.81 M(−) M(−) MT = 2.95% MA = 5.0% MA = 1.91% MKN-7 M(±) M(−) ER = 3.03 MA = 15.4% MA = 33.3% MT = 27.2% NCI-SNU-5 M(±) ER = 1.28 M(±) MA = 71.2% MT = 54.4% MA = 21.3% NCI-SNU-16 M(+) ER = 2.24 M(+) MT = 87.9% MA =90.0% MA = 83.8% SH-10-TC M(−) ER = 13.8 M(±) MA = 2.75% MT = 7.72% MA = 17.5% Figure 1 Map of 5 flanking region of hDAB2IP gene and the bisulphited genomic DNA sequence. (A) Location of CpG dinucleotides in the genomic sequence (GenBank accession number AL365274) are indicated by thin vertical lines. The bent arrow indicates the transcription start site (TSS) ( þ 1). Five rectangular boxes, promoter region (P2), regions for the bisulphited genomic sequence (m2a: from 522 to 285, m2b: from 308 to þ 89) and regions for the amplicons of MSP (RMSPm2a, RMSPm2b). (B) Methylation status of individual cloned DNA fragments of five gastric cancer cell lines is shown. Each row represents one sequenced allele. Each circle represents a CpG dinucleotide. Filled circle, methylation; open circle, no methylation. Clonal numbers are 0 0 indicated by prefix C to the left. The numbers at the top indicate the CpG dinucleotide in the amplicon (5 to 3 ). The percentage of MA, MB and MT indicates the rate of methylated CpG dinucleotides in m2a, m2b and total regions of each sample, respectively. ER means expression ratio by semiquantitative real-time RT –PCR. The location of CpG dinucleotides included in MSP primers (m2a-MSP-F, m2a-MSP-R, m2b-MSP-F and m2b-MSP-R) are indicated by open boxes. M( þ ), positive for the hDAB2IP-methylated form by MSP; M(), negative for the hDAB2IP-methylated form by MSP; M(7), positive for both the hDAB2IP-methylated and -unmethylated forms by MSP. Bevely, MA, USA) and then subjected to bisulphite treatment was Statistical analysis used as a positive control for methylated alleles. Water blanks were The frequencies of hDAB2IP methylation between two groups were included in each assay. compared using the w test. The quantitative ratios of different groups were compared using the Mann–Whitney U-nonpara- metric test. Probability value less than 0.05 was defined as being 5-Aza-CdR and TSA treatment statistically significant. All data were analysed with StatView for Cell lines were treated with 5-Aza-CdR (Sigma-Aldrich Co., St Windows (SAS Institute Inc., Cary, NC, USA). Louis, MO, USA) at a concentration of 1–2mgml for 6 days with medium changes on days 1, 3 and 5 (Jones, 1985; Virmani et al, 2000). The treatment with histone deacetylase inhibitor, TSA RESULTS (Wako, Tokyo, Japan), was performed at a concentration of 150– 300 nM for 12–24 h (Yoshida et al, 1990; Toyooka et al, 2002a). For Bisulphited genomic DNA sequencing combination treatment, TSA was added to day 5 of 5-Aza-CdR- treated cell lines and cultured for additional 12–24 h. Treated or The m2a and m2b regions, covering the major part of the P2 untreated cells from individual triplicate flasks were harvested to region, were separately amplified by PCR to examine their be semiquantified by the gene expression level using real-time methylation status by direct sequencing in eight gastric cancer RT–PCR as described above. cell lines, and five cell lines that showed one of the four typical & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al methylation patterns were selected for more detailed analysis. The regions were almost consistently methylated in the expression- PCR amplicons of these five cell lines were cloned, and eight reduced cell lines (Figure 1B). Three and four cytosines were individual clones were sequenced (Figure 1B). MKN-1 showed included in the forward and reverse primers, respectively, for the almost no methylation in either the m2a or the m2b region. In m2a region producing a 163-bp amplicon, and five and three contrast, NCI-SNU-16 showed a heavy methylation pattern, with cytosines were included in the forward and reverse primers, methylation of over 80% of all the CpG dinucleotides in both the respectively, for the m2b region producing a 209-bp amplicon m2a and m2b regions. The other three cell lines showed a (Figure 1B). Methylation-specific PCR assay using these primers moderate methylation pattern, which was subdivided into two was performed to examine the methylation status of hDAB2IP in types, the m2a- and the m2b-dominant methylation pattern. Two the gastric cancer cell lines and in primary gastric and colorectal cell lines, namely, MKN-7 and NCI-SNU-5, showed an m2b- cancer tumours. Representative examples of the assay are dominant methylation pattern, and one cell line, namely, SH-10 illustrated in Figure 3A and B and the results are summarised in TC, showed an m2a-dominant methylation pattern. Table 2. Aberrant methylation of the m2a and m2b regions was found in 16 of 35 (46%) and 15 of 35 (43%) primary gastric cancer specimens, respectively, and in 29 of 60 (48%) and 28 of 60 (47%) hDAB2IP expression in gastric cancer cell lines primary colorectal cancer specimens, respectively. In the case of the gastric cancer cell lines, methylation of the gene was detected We examined the expression status of hDAB2IP by semiquantita- in eight of 12 cell lines (67%) for each of the m2a and m2b regions. tive real-time RT–PCR in eight cell lines, consisting of three The unmethylated form of hDAB2IP was always found in primary showing heavy methylation in both the m2a and m2b regions, one tumour samples because these had been grossly dissected and thus showing an m2a-dominant methylation pattern, two showing an showed some contamination with normal cells. In contrast, the m2b-dominant methylation pattern, and two showing no methyla- cancer cell lines consisted of a pure population of tumour cells, tion of the hDAB2IP gene. The semiquantitatively estimated mRNA and eight of the 12 cell lines (67%) had either methylated or expression values in each cell line are shown in Figure 2A. The unmethylated hDAB2IP alleles for both m2a and m2b regions. Four values in the two cell lines showing no methylation of the gene of the 12 cell lines (33%) showed both methylated and were as follows; the mean hDAB2IP expression ratio in NUGC-4 unmethylated forms for the m2a or the m2b region, and were was very high at 19.2, as expected, but the ratio in MKN-1 was regarded as showing partial methylation of the gene, and the quite unexpectedly low, with a value of 1.81. This transcriptional details of the CpG methylation status in these cell lines are shown downregulation of hDAB2IP in MKN-1 was attributed to treatment in Figure 1B. A total of 36 (38%) primary gastric and colorectal with a histone deacetylase inhibitor, as described below. As cancer specimens and seven (58%) cell lines showed aberrant compared with the value in NUGC-4, the expression ratios in all of methylation in both the m2a and m2b regions. On the other hand, the cell lines showing methylation of the hDAB2IP gene, with the 16 (17%) primary gastric and colorectal cancer specimens and two exception of SH-10-TC, were significantly lower, with the mean (17%) cell lines showed methylation of either the m2a or the m2b ratios in NCI-SNU-16, MKN-45, KATO-III, NCI-SNU-5 and MKN- region. There were no significant differences in the methylation 7 being 2.24, 3.23, 2.08, 1.28, and 3.03 respectively; the expression rate between the cancer cell lines and primary tumours in either ratio in SH-10-TC was only 13.8. SH-10-TC showed a moderate the m2a or the m2b region. In addition, seven (12%) and 11 (19%) methylation status with the m2a-dominant pattern, but its m2b of 59 gastrointestinal normal mucosal specimens showed methyla- region was almost completely unmethylated, as described above. tion in the m2a and m2b regions, respectively. Aberrant On the other hand, in MKN-7 and NCI-SNU-5, in which the methylation was not present in either of the two regions in expression ratios indicated significant downregulation of hDA- samples obtained from five peripheral blood lymphocytes. B2IP, the m2a-methylation rates were equal to that in SH-10-TC, but the m2b regions were additionally methylated (Figure 1B). These results indicate that the expression of hDAB2IP tended to be 5-Aza-CdR and TSA treatment for gastric cancer cell lines reduced depending on the methylation status in the m2b region, which is just adjacent to the transcriptional-start site, and suggest To confirm the involvement of DNA methylation in hDAB2IP that the m2b promoter region might be the methylation-mediated silencing, we treated five cell lines with 5-Aza-CdR. hDAB2IP key regulator of the expression of hDAB2IP in gastric cancer. In expression was upregulated following the 5-Aza-CdR treatment in the five nonmalignant gastric mucosal specimens examined, the the m2b-methylated cell lines. In contrast, no such upregulation mean expression ratio of hDAB2IP was 21.679.7, indicative of following 5-Aza-CdR treatment was observed in the m2b- normal expression. nonmethylated cell lines. This finding indicates that methylation in the m2b region plays an important role in regulating the gene silencing. Furthermore, we also treated cell lines with TSA, either MSP assay and aberrant methylation of hDAB2IP in singly or in combination with 5-Aza-CdR. Expression of hDAB2IP gastrointestinal tumours in MKN-7 was upregulated 3.1-fold by TSA treatment and by 6.9- fold by following combined treatment with TSA and 5-Aza-CdR. In We determined the forward and reverse primers for MSP of each of NCI-SNU-5, while hDAB2IP expression was upregulated 3.0-fold the m2a and m2b regions (Figure 1B). The CpG sites in these Figure 2 hDAB2IP mRNA expression using semiquantitative real-time RT –PCR. Triplicate sets of flasks before and after drug treatment were harvested and tested for gene expression using semiquantitative real-time PCR. Data shown are mean values7s.d., n ¼ 3. (A) The relative ratio of mRNA expression in each cell line. The relative expression ratio was remarkably repressed in m2b-methylated cell lines compared with m2b-unmethylated ones except MKN- 1. Following symbols of two plus and minus means positive and negative for methylation status by MSP, respectively. One plus means positive for both the methylated and unmethylated forms by MSP. (B) Effect of 5-Aza-CdR, TSA and combination on mRNA expression in m2b-methylated cell lines (MKN-7, NCI-SNU-5 and NCI-SNU-16). hDAB2IP expression of MKN-7 was upregulated 3.4-fold by 5-Aza-CdR, 3.1-fold by TSA and additively 6.9-fold by combination. In NCI-SNU-5, hDAB2IP expression was also upregulated 3.0-fold by 5-Aza-CdR and 3.0-fold by TSA, whereas TSA failed to make synergistic or additive effect for this cell line. In NCI-SNU-16, the expression was also upregulated 3.1-fold by 5-Aza-CdR, but TSA alone failed to make a positive effect, operating synergically to 6.6-fold on the expression with a combination of 5-Aza-CdR. (C) Effect of 5-Aza-CdR, TSA and combination on mRNA expression in m2b-unmethylated cell lines (MKN1 and SH-10-TC). In MKN-1, the downregulated expression was restored 3.1-fold by TSA alone. In SH-10-TC, the expression was decreased after any drug treatment, which may be caused by a cytotoxic effect from drug exposure. British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK hDAB2IP methylation in gastrointestinal tumour H Dote et al by treatment with TSA alone, combined treatment with TSA and 5- On the other hand, in SH-10-TC, an m2b-nonmethylated cell Aza-CdR failed to elicit a synergistic or additive effect. In NCI- line, treatment with the aforementioned drugs neither upregulated SNU-16, while treatment with TSA alone failed to have any positive nor downregulated the expression of hDAB2IP (0.1–0.4-fold), effect, combined treatment with 5-Aza-CdR elicited a positive because the number of SH-10-TC cells in culture was decreased by effect on hDAB2IP expression. treatment with 5-Aza-CdR, TSA or a combination of 5-Aza-CdR Cell line with unmethylated m2b region Cell line with partially methylated m2b region Cell line with heavily methylated m2b region Methylation status ++ ++ + + −− + − − of m2a region Methylation status ++ ++ ++ + −− − − of m2a region MKN-7 NCI-SNU- 5 NCI-SNU-16 P < 0.001 P = 0.005 P < 0.001 P = 0.008 P < 0.001 P = 0.022 P = 0.002 P = 0.002 P = 0.003 25 5 20 4 15 3 10 2 5 1 0 0 0 MKN-1 SH-10-TC P = N.S P < 0.001 P = 0.002 P = 0.001 P = N.S P < 0.001 0 0 & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Control (no treatment) 5-Aza-dC TSA Combination Control (no treatment) 5-Aza-dC Control (no treatment) TSA 5-Aza-dC Combination TSA Combination Control (no treatment) Control 5-Aza-dC (no treatment) TSA 5-Aza-dC Combination TSA Combination Semiquantitative expression ratio Semiquantitative expression ratio Semiquantitative expression ratio Semiquantitative NCI-SNU-16 expression ratio MKN-45 KATO-III Semiquantitative expression ratio NCI-SNU- 5 MKN-7 SH-10-TC Semiquantitative MKN-1 expression ratio NUGC-4 Normal mucosa Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al For m2a region Methylation-specific PCR Unmethylation-specific PCR For m2b region Methylation-specific PCR Unmethylation-specific PCR Gastric cancer For m2a region Methylation-specific PCR Unmethylation-specific PCR For m2b region Methylation-specific PCR Unmethylation-specific PCR Colorectal cancer For m2a region Methylation-specific PCR Unmethylation-specific PCR For m2b region Methylation-specific PCR Unmethylation-specific PCR British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK G1 G1 MKN-1 G2 G2 G3 G3 MKN-7 G4 G4 MKN-28 G5 G5 G6 G6 MKN-45 G7 G7 G8 MKN-74 G8 G9 G9 KATO-III G10 G10 G11 NCI-SNU-5 G11 G12 G12 NCI-SNU-16 G13 G13 G14 NUGC-4 G14 G15 G15 SH-10-TC G16 G16 G17 G17 AZ521 G18 G18 H-111-TC G19 G19 G20 G20 Positive control G21 G21 Water blank G22 G22 hDAB2IP methylation in gastrointestinal tumour H Dote et al and TSA (data not shown). This effect may be attributable to the sex, age, tumour size, extent of tumour invasion, lymph node cytotoxic effect of 5-Aza-CdR and TSA. Interestingly, in MKN-1, status, distant metastasis, TNM stage, histology and the tumour which showed no methylation in either the m2a or the m2b region, site in cases with gastrointestinal cancers. In cases of gastric the downregulated expression of hDAB2IP could be restored by cancer, while both the m2a- and m2b-methylation status were TSA treatment alone. This result suggests that the expression of the significantly associated with the tumour site, the m2b-methylation gene might be downregulated under the control of histone status showed a stronger correlation than the m2a-methylation deacetylation in this cell line alone. These results suggest that status (m2a, P ¼ 0.0172; m2b, P ¼ 0.0064; Table 3). In cases of both DNA methylation and histone deacetylation may act colorectal cancer, m2b methylation tended to be associated with cooperatively to silence hDAB2IP expression in some gastric cancer arising from the left side of the colon. (P ¼ 0.0926; Table 4). cancer cell lines, and that the methylation status of the m2b region hDAB2IP methylation was often observed in tumours arising from may act as the major gene silencing mechanism in most cell lines. the lower-third of the stomach and distal aspect of the colorectum. There were no significant differences between methylation-positive and methylation-negative groups with respect to the other factors hDAB2IP methylation and clinicopathological correlation examined in this study. Although accumulation of methylation appears to be age-related in a number of cancer-related genes, both We examined the relationship between the methylation status in the m2a- and m2b-methylation status were not significantly the tumour specimens and clinicopathological factors, including associated with the age of patients (gastric cancer: m2a, P ¼ 0.9369; m2b, P ¼ 0.9788; colorectal cancer: m2a, P ¼ 0.8484; m2b, P ¼ 0.5709; Table 5). Table 2 Aberrant promoter methylation of hDAB2IP in samples No. of No. of No. of both hDAB2IP hDAB2IP m2a and DISCUSSION m2a m2b m2b To determine the methylation status of the 5 CpG islands of methylated methylated methylated Samples samples (%) samples (%) samples (%) hDAB2IP in gastrointestinal tumours, we first performed bisul- phite genomic DNA sequencing in several gastric cancer cell lines. Gastric cancers Based on the sequence data, methylation-specific primers were Cell lines (n ¼ 12) 8 (67) 8 (67) 7 (58) designed to detect sensitively each methylated or unmethylated Primary tumours (n ¼ 35) 16 (46) 15 (43) 13 (37) allele (Toyooka et al, 2002b, 2003). These primers have also been shown to be applicable to lung and breast cancer cell lines (Dote Colorectal cancers et al, 2004). Using the MSP primers designed thus, we found Primary tumours (n ¼ 60) 29 (48) 28 (47) 23 (38) frequent methylation of the hDAB2IP gene in primary gastro- intestinal tumours, as well as in gastrointestinal tumour cell lines. Nonmalignant specimens (n ¼ 64) Normal gastric mucosa 2 (6) 7 (20) 2 (6) The higher frequency of methylation in the cell lines as compared (n ¼ 35) with that in the primary tumours might be explained by the Normal colorectal mucosa 5 (21) 4 (17) 3 (13) additional changes that the cells may acquire in culture, or by the (n ¼ 24) tumour cell lines being derived from more aggressive tumours and Peripheral lymphocytes 0 (0) 0 (0) 0 (0) therefore showing more malignant changes (Zochbauer-Muller (n ¼ 5) et al, 2001). Table 3 Clinicopathologic features and hDAB2IP promoter methylation in gastric cancer Frequency of hDAB2IP Frequency of hDAB2IP a b b Clinicopathologic features Variable m2a methylation (%) P-value m2b methylation (%) P-value Sex Male (n ¼ 21) 8 (38) 0.27 9 (43) 40.99 Female (n ¼ 14) 8 (57) 6 (43) Tumour invasion T1, 2 (n ¼ 20) 9 (45) 0.92 9 (45) 0.77 T3, 4 (n ¼ 15) 7 (47) 6 (40) Nodal metastasis N0 (n ¼ 14) 7 (50) 0.68 7 (50) 0.49 N1, 2 (n ¼ 21) 9 (43) 8 (38) Metastasis M0 (n ¼ 26) 12 (46) 0.93 11 (42) 0.91 M1 (n ¼ 9) 4 (44) 4 (44) TNM stage 1, 2 (n ¼ 18) 8 (44) 0.88 8 (44) 0.85 3, 4 (n ¼ 17) 8 (47) 7 (41) Histology Intestinal (n ¼ 17) 8 (47) 0.88 6 (35) 0.38 Diffuse (n ¼ 18) 8 (44) 9 (50) Tumour site U (n ¼ 7) 2 (22) 0.017 1 (14) 0.0064 M(n ¼ 13) 3 (23) 3 (23) L(n ¼ 15) 11 (73) 11 (73) a b 2 U ¼ upper site; M ¼ middle site; L ¼ lower site. Statistical significance was determined using w test. Figure 3 Representative examples of MSP in gastrointestinal tumours. (A) gastric cancer cell lines, (B) primary tumours. The unmethylated form of hDAB2IP was always found in primary tumours that had been grossly dissected and thus had at least some contamination with normal cells. Positive controls include peripheral blood lymphocytes from healthy individuals for the unmethylated form and normal lymphocyte DNA methylated by treatment with Sss I DNA methyltransferase for the methylated form. & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Molecular Diagnostics Molecular Diagnostics hDAB2IP methylation in gastrointestinal tumour H Dote et al Table 4 Clinicopathologic features and hDAB2IP promoter methylation in colorectal cancer Frequency of hDAB2IP Frequency of hDAB2IP a b b Clinicopathologic features Variable m2a methylation (%) P-value m2b methylation (%) P-value Sex Male (n ¼ 37) 18 (49) 0.95 15 (41) 0.23 Female (n ¼ 23) 11 (48) 13 (57) Tumour invasion T1, 2 (n ¼ 9) 4 (44) 0.80 5 (56) 0.56 T3, 4 (n ¼ 51) 25 (49) 23 (45) Nodal metastasis N0 (n ¼ 25) 9 (36) 0.11 11 (44) 0.73 N1, 2 (n ¼ 35) 20 (57) 17 (49) Metastasis M0 (n ¼ 53) 26 (49) 0.76 25 (47) 0.83 M1 (n ¼ 7) 3 (43) 3 (43) TNM stage 1, 2 (n ¼ 21) 7 (33) 0.088 9 (43) 0.66 3, 4 (n ¼ 39) 22 (56) 19 (49) Histological grade G1 (n ¼ 13) 7 (54) 0.65 6 (46) 0.97 G2, 3, 4 (n ¼ 47) 22 (47) 22 (47) Tumour site C, A, T (n ¼ 12) 4 (33) 0.25 3 (25) 0.093 D, S, R (n ¼ 48) 25 (52) 25 (52) a b 2 C ¼ cecum; A ¼ ascending colon; T ¼ transverse colon; D ¼ descending colon; S ¼ sigmoid colon; R ¼ rectum. Statistical significance was determined using w test. Table 5 Association between hDAB2IP promoter methylation status and examined the effect of the histone deacetylase inhibitor, TSA, age in gastrointestinal cancer as well as that of 5-Aza-CdR on the methylated cell lines, to distinguish the effect of histone deacetylation on the down- a b Methylation status Age P-value regulated gene expression of hDAB2IP. In MKN-7, the expression was upregulated by both 5-Aza-CdR and TSA treatment, Gastric cancer and combined treatment with both agents caused synergistic m2a-methylation (n ¼ 15) 66 (40 – 89) 0.9369 upregulation, with the extent of upregulation in this case m2a-unmethylation (n ¼ 20) 66 (36 – 85) being greater than that with either agent alone. In addition, in m2b-methylation (n ¼ 16) 66 (40 – 89) 0.9788 some other methylated cells, TSA enhanced hDAB2IP m2b-unmethylation (n ¼ 19) 66 (36 – 85) gene expression. These results indicate that in some cell lines Colorectal cancer showing gene methylation, there may be a synergistic and m2a-methylation (n ¼ 29) 66 (51 – 85) 0.8484 positive interaction between DNA methylation and histone m2a-unmethylation (n ¼ 31) 65 (26 – 86) deacetylation for transcriptional regulation, although this m2b-methylation (n ¼ 28) 66 (47 – 82) 0.5709 does not seem to be a common and major mechanism in gastric m2b-unmethylation (n ¼ 32) 65 (26 – 86) cancer cells. However, the number of examined cells was a b too limited to arrive at any definitive conclusion on the effect Mean age (range) in years. Statistical significance was determined using Mann – Whitney U-test. of TSA on hDAB2IP expression. Considering this effect of deacetylation on gene expression, it was reported that the expression of some methylation-repressed genes can be reactivated by TSA treatment (Yang et al, 2000) and that demethylation- We previously reported that the m2a region appeared to be the induced gene re-expression could be potentiated by TSA (Cameron et al, 1999). In addition, Suzuki et al (2002) analysed the key regulatory region for hDAB2IP expression in prostate cancer (Chen et al, 2003) and that methylation of both the m2a and m2b expression of over 10 000 genes by microarray analysis, and found regions occurred in breast cancer cell lines (Dote et al, 2004). that the expression of 74 genes was upregulated following However, in contrast to the observation in prostate and breast treatment with 5-Aza-CdR or TSA. They also showed that cancer, this study demonstrated that the downregulation of transcriptional silencing is mediated by both methylation and hDAB2IP expression correlated more with the methylation status histone deacetylation, although the effect of methylation was of not the m2a region, but of the m2b region, which is located just dominant for gene silencing. adjacent to exon Ia in gastric cancer cell lines (Figure 2A). hDAB2IP methylation was found to be present at similar Considering the close proximity of the m2b region to the frequencies in both early and advanced carcinomas, suggesting transcriptional-start site of the hDAB2IP gene, it stands to reason that the methylation actually occurs at an early stage of that the m2b region might act as the regulatory promoter in gastric carcinogenesis in gastrointestinal tumours. Furthermore, the result cancer cells. that hDAB2IP methylation was found in nonmalignant mucosa We demonstrated in this study that the expression of hDAB2IP suggested that hDAB2IP methylation might play a causative role in was highly suppressed in five m2b-methylated cell lines as carcinogenesis. Interestingly, we found a highly significant compared with that in two m2b-nonmethylated cell lines. This association between the tumour location in the stomach and result suggests that DNA methylation of the regulatory promoter methylation of the m2b region, which has been supposed to be the region may be critical for transcriptional regulation of the key regulatory sequence, as described above. Cancers showing hDAB2IP gene and play an important role in hDAB2IP inactiva- methylation of the m2b region showed a significant tendency tion. These results are consistent with the finding of restoration of towards being preferentially located in the lower-third of the hDAB2IP expression in methylated cell lines after 5-Aza-CdR stomach. Similar results were also described for p16, hMLH1 and treatment. TSLC1 methylation in gastric cancer (Schneider et al, 2000; Honda Recent data suggest that hypermethylated DNA interacts with et al, 2002; Sakata et al, 2002). Therefore, methylation of several several methyl-CpG-binding proteins, and that the interaction tumour-suppressor and tumour-related genes appears to be an facilitates the assembly of a repressive complex containing histone important pathogenetic mechanism of gastric cancers arising from deacetylase and formation of an inactive chromatin that leads to the antrum. Because intestinal metaplasia commonly arises in the gene silencing (Nan et al, 1998; Wade et al, 1999). Therefore, we antrum and then extends towards the body of the stomach, British Journal of Cancer (2005) 92(6), 1117 – 1125 & 2005 Cancer Research UK hDAB2IP methylation in gastrointestinal tumour H Dote et al intestinal metaplasia may predispose to promoter region methyla- a right-sided location of the tumour, although this did not reach tion in these genes (Sato et al, 2001). In addition, a marked statistical significance. increase in the frequency of methylated genes from nonmetaplastic In conclusion, our results demonstrate that aberrant methyla- mucosa to intestinal metaplasia has been reported (Kang et al, tion drives gene downregulation, and supports the contention that 2001). In sporadic colorectal cancers, it has recently been shown the methylation-mediated transcriptional silencing of the hDAB2IP that CpG island methylation is associated with a right-sided gene may be a critical event in tumorigenesis of gastrointestinal location of the tumour (Hawkins et al, 2002). However, we also tumours, especially those arising from the antrum of the stomach found an inverse association between m2b-region methylation and in association with intestinal metaplasia. REFERENCES Baylin SB, Herman JG (2000) DNA hypermethylation in tumorigenesis: Sobin LH, Fleming ID (1997) TNM Classification of Malignant Tumors, epigenetics joins genetics. Trends Genet 16: 168–174 fifth edition (1997). Union Internationale Contre le Cancer and the Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB (1999) American Joint Committee on Cancer. Cancer 80: 1803–1804 Synergy of demethylation and histone deacetylase inhibition in the re- Suzuki H, Gabrielson E, Chen W, Anbazhagan R, van Engeland M, expression of genes silenced in cancer. Nat Genet 21: 103–107 Weijenberg MP, Herman JG, Baylin SB (2002) A genomic screen for Chen H, Pong RC, Wang Z, Hsieh JT (2002) Differential regulation of the genes upregulated by demethylation and histone deacetylase inhibition human gene DAB2IP in normal and malignant prostatic epithelia: in human colorectal cancer. Nat Genet 31: 141–149 cloning and characterization. Genomics 79: 573–581 Toyooka S, Carbone M, Toyooka KO, Bocchetta M, Shivapurkar N, Minna Chen H, Toyooka S, Gazdar AF, Hsieh JT (2003) Epigenetic regulation of a JD, Gazdar AF (2002a) Progressive aberrant methylation of the RASSF1A novel tumor suppressor gene (hDAB2IP) in prostate cancer cell lines. J gene in simian virus 40 infected human mesothelial cells. Oncogene 21: Biol Chem 278: 3121– 3130 4340– 4344 Dote H, Toyooka S, Tsukuda K, Yano M, Ouchida M, Doihara H, Suzuki M, Toyooka S, Toyooka KO, Harada K, Miyajima K, Makarla P, Sathyanar- Chen H, Hsieh JT, Gazdar AF, Shimizu N (2004) Aberrant promoter ayana UG, Yin J, Sato F, Shivapurkar N, Meltzer SJ, Gazdar AF (2002b) methylation in human DAB2 interactive protein (hDAB2IP) gene in Aberrant methylation of the CDH13 (H-cadherin) promoter region in breast cancer. Clin Cancer Res 10: 2082–2089 colorectal cancers and adenomas. Cancer Res 62: 3382–3386 Fulop V, Colitti CV, Genest D, Berkowitz RS, Yiu GK, Ng SW, Szepesi J, Toyooka S, Toyooka KO, Maruyama R, Virmani AK, Girard L, Miyajima K, Mok SC (1998) DOC-2/hDab2, a candidate tumor suppressor gene Harada K, Ariyoshi Y, Takahashi T, Sugio K, Brambilla E, Gilcrease M, involved in the development of gestational trophoblastic diseases. Minna JD, Gazdar AF (2001) DNA methylation profiles of lung tumors. Oncogene 17: 419–424 Mol Cancer Ther 1: 61–67 Hawkins N, Norrie M, Cheong K, Mokany E, Ku SL, Meagher A, O’Connor Toyooka S, Toyooka KO, Miyajima K, Reddy JL, Toyota M, Sathyanarayana T, Ward R (2002) CpG island methylation in sporadic colorectal cancers UG, Padar A, Tockman MS, Lam S, Shivapurkar N, Gazdar AF (2003) and its relationship to microsatellite instability. Gastroenterology 122: Epigenetic down-regulation of death-associated protein kinase in lung 1376– 1387 cancers. Clin Cancer Res 9: 3034– 3041 Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB (1996) Tseng CP, Ely BD, Li Y, Pong RC, Hsieh JT (1998) Regulation of rat DOC-2 Methylation-specific PCR: a novel PCR assay for methylation status of gene during castration-induced rat ventral prostate degeneration and its CpG islands. Proc Natl Acad Sci USA 93: 9821 –9826 growth inhibitory function in human prostatic carcinoma cells. Honda T, Tamura G, Waki T, Jin Z, Sato K, Motoyama T, Kawata S, Kimura Endocrinology 139: 3542–3553 W, Nishizuka S, Murakami Y (2002) Hypermethylation of the TSLC1 Virmani AK, Rathi A, Zochbauer-Muller S, Sacchi N, Fukuyama Y, Bryant gene promoter in primary gastric cancers and gastric cancer cell lines. D, Maitra A, Heda S, Fong KM, Thunnissen F, Minna JD, Gazdar AF Jpn J Cancer Res 93: 857–860 (2000) Promoter methylation and silencing of the retinoic acid receptor- Jones PA (1985) Altering gene expression with 5-azacytidine. Cell 40: 485–486 beta gene in lung carcinomas. J Natl Cancer Inst 92: 1303–1307 Jones PA, Laird PW (1999) Cancer epigenetics comes of age. Nat Genet 21: Wade PA, Gegonne A, Jones PL, Ballestar E, Aubry F, Wolffe AP (1999) Mi- 163–167 2 complex couples DNA methylation to chromatin remodelling and Kang GH, Shim YH, Jung HY, Kim WH, Ro JY, Rhyu MG (2001) CpG island histone deacetylation. Nat Genet 23: 62–66 methylation in premalignant stages of gastric carcinoma. Cancer Res 61: Wang RY, Gehrke CW, Ehrlich M (1980) Comparison of bisulfite 2847– 2851 modification of 5-methyldeoxycytidine and deoxycytidine residues. Lauren P (1991) Histogenesis of intestinal and diffuse types of gastric Nucleic Acids Res 8: 4777–4790 carcinoma. Scand J Gastroenterol Suppl 180: 160–164 Wang Z, Tseng CP, Pong RC, Chen H, McConnell JD, Navone N, Hsieh JT Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, (2002) The mechanism of growth-inhibitory effect of DOC-2/DAB2 in Bird A (1998) Transcriptional repression by the methyl-CpG-binding prostate cancer. Characterization of a novel GTPase-activating protein protein MeCP2 involves a histone deacetylase complex. Nature 393: associated with N-terminal domain of DOC-2/DAB2. J Biol Chem 277: 386–389 12622–12631 Pawlowski V, Revillion F, Hebbar M, Hornez L, Peyrat JP (2000) Prognostic Yang X, Ferguson AT, Nass SJ, Phillips DL, Butash KA, Wang SM, Herman value of the type I growth factor receptors in a large series of human JG, Davidson NE (2000) Transcriptional activation of estrogen receptor primary breast cancers quantified with a real-time reverse transcription- alpha in human breast cancer cells by histone deacetylase inhibition. polymerase chain reaction assay. Clin Cancer Res 6: 4217 –4225 Cancer Res 60: 6890–6894 Sakata K, Tamura G, Endoh Y, Ohmura K, Ogata S, Motoyama T Yoshida M, Kijima M, Akita M, Beppu T (1990) Potent and specific (2002) Hypermethylation of the hMLH1 gene promoter in solitary and inhibition of mammalian histone deacetylase both in vivo and in vitro by multiple gastric cancers with microsatellite instability. Br J Cancer 86: trichostatin A. J Biol Chem 265: 17174–17179 564–567 Zhou J, Hsieh JT (2001) The inhibitory role of DOC-2/DAB2 in growth Sato K, Tamura G, Tsuchiya T, Endoh Y, Usuba O, Kimura W, Motoyama T factor receptor-mediated signal cascade. DOC-2/DAB2-mediated inhibi- (2001) Frequent loss of expression without sequence mutations of the tion of ERK phosphorylation via binding to Grb2. J Biol Chem 276: DCC gene in primary gastric cancer. Br J Cancer 85: 199–203 27793–27798 Schneider BG, Gulley ML, Eagan P, Bravo JC, Mera R, Geradts J (2000) Loss Zochbauer-Muller S, Fong KM, Maitra A, Lam S, Geradts J, Ashfaq R, of p16/CDKN2A tumor suppressor protein in gastric adenocarcinoma is Virmani AK, Milchgrub S, Gazdar AF, Minna JD (2001) 5 CpG island associated with Epstein-Barr virus and anatomic location in the body of methylation of the FHIT gene is correlated with loss of gene expression the stomach. Hum Pathol 31: 45–50 in lung and breast cancer. Cancer Res 61: 3581– 3585 & 2005 Cancer Research UK British Journal of Cancer (2005) 92(6), 1117 – 1125 Molecular Diagnostics

Journal

British Journal of Cancer – Springer Journals

Published: Mar 15, 2005

There are no references for this article.