Advertisement
Molecular Biology, Pathobiology, and Genetics

Discovery of Aberrant Expression of R-RAS by Cancer-Linked DNA Hypomethylation in Gastric Cancer Using Microarrays

Michiko Nishigaki, Kazuhiko Aoyagi, Inaho Danjoh, Masahide Fukaya, Kazuyoshi Yanagihara, Hiromi Sakamoto, Teruhiko Yoshida and Hiroki Sasaki
DOI: 10.1158/0008-5472.CAN-04-3340 Published March 2005

Abstract

Although hypomethylation was the originally identified epigenetic change in cancer, it was overlooked for many years in preference to hypermethylation. Recently, gene activation by cancer-linked hypomethylation has been rediscovered. However, in gastric cancer, genome-wide screening of the activated genes has not been found. By using microarrays, we identified 1,383 gene candidates reactivated in at least one cell line of eight gastric cancer cell lines after treatment with 5-aza-2′deoxycytidine and trichostatin A. Of the 1,383 genes, 159 genes, including oncogenes ELK1, FRAT2, R-RAS, RHOB, and RHO6, were further selected as gene candidates that are silenced by DNA methylation in normal stomach mucosa but are activated by DNA demethylation in a subset of gastric cancers. Next, we showed that demethylation of specific CpG sites within the first intron of R-RAS causes activation in more than half of gastric cancers. Introduction of siRNA into R-RAS-expressing cells resulted in the disappearance of the adhered cells, suggesting that functional blocking of the R-RAS-signaling pathway has great potential for gastric cancer therapy. Our extensive gene list provides other candidates for this class of oncogene.

  • hypomethylation
  • R-RAS
  • oncogene
  • gastric cancer
  • microarray

Introduction

DNA hypermethylation in cancer has been observed most often in CpG islands in gene regions and leads to down-regulation of tumor suppressor genes (1, 2) . Recently, histone modifications and silencing before DNA methylation of the p16/INK4A gene has been reported (3). Therefore, whether DNA hypermethylation is a cause of tumorigenesis remains unestablished. In contrast, very frequent hypomethylation is seen in both highly and moderately repeated DNA sequences in many human cancers, including heterochromatic DNA repeats, dispersed retrotransposons, and endogenous retroviral elements (4). Whether genome-wide DNA hypomethylation, however, is a cause or consequence of tumorigenesis has been unclear. As possible mechanisms, insertional activation of proto-oncogenes by induction of endogenous retroviral elements (5) and induction of genomic instability (6) were previously considered. Recently, mutant mice carrying a hypomorphic DNA methyltransferase 1 (Dnmt 1) have been reported to develop aggressive T-cell lymphomas that displayed a high frequency of chromosome 15 trisomy (7). This article shows that DNA hypomethylation plays a causal role in tumor formation, possibly by promoting chromosomal instability. For gene activation by cancer-linked hypomethylation, in 1983 Feinberg and Vogelstein (8) first reported that human growth hormone, α-globin, and β-globin are methylated in normal tissues and become hypomethylated in cancers. It has been rediscovered recently that the cancer-linked hypomethylation leads to activation of genes that are important in cancer (4). A correlation between hypomethylation and overexpression has been shown for MN/CA9 encoding a tumor antigen in renal cell carcinomas (9), MDR1 in myeloid leukemias (10), BCL-2 in chronic lymphocytic leukemias (11), MAGE-1 in melanomas (12), and SNCG/BCSG1 in breast carcinomas and ovarian carcinomas (13). SNCG, also referred to as BCSG1, is a member of a neuronal protein family, synuclein, and its expression is highly tissue-specific (14, 15) . As described above, it was clearly shown that hypomethylation of the SNCG gene CpG island promotes its aberrant expression in breast and ovarian carcinomas (13). Extensive recent study of pancreatic cancer by microarray analysis showed that frequent hypomethylation of seven genes (claudin4, lipocalin2, 14-3-3σ, trefoil facter2, S100A4, mesothelin, and prostate stem cell antigen) caused overexpression (16). In the early studies, c-MYC, H-RAS, c-FOS, and AFP genes have been reported to be less hypomethylated in gastric cancer tissues than in noncancerous tissues (17). However, in gastric cancer, genome-wide screening of the activated genes has not been found. In this study, using genome-wide expression analysis, we screened genes activated by cancer-linked hypomethylation in gastric cancer.

Materials and Methods

Tissue Samples and 5-Aza-2′Deoxycytidine and Trichostatin A Treatment of Gastric Cancer Cells. Thirty-three gastric cancer tissues and six noncancerous tissues were obtained with informed consent from patients who underwent a gastrectomy at the National Cancer Center Hospital (Tokyo, Japan). Eight cell lines (HSC39, HSC43, HSC44, HSC58, HSC59, HSC60, KATOIII, and OCUM2M) that were derived from gastric cancers were used. The treatment of the eight cell lines consisted of 5-aza-2′ deoxycytidine (5Aza-dC, 1 μmol/L) for 24 hours, followed by addition of trichostatin A (TSA), to a final concentration of 1 μmol/L for a further 24 hours. Treatments with TSA alone and 5Aza-dC alone were also done by using the same amount of the drugs.

Microarray Analysis. For total RNA isolation, surgical specimens and cultured cells were mixed with IsoGen lysis buffer (Nippon Gene Co., Ltd., Toyama, Japan) at room temperature, extracted with chloroform, and precipitated with a glycogen carrier (20 μg/μL) in isopropanol. The RNA pellet was washed with 70% ethanol prepared and dissolved in RNase-free water. We used human U95A version 2 arrays (Affymetrix, Santa Clara, CA) for analysis of mRNA expression levels corresponding to 12,626 transcripts. The procedures were conducted according to the protocols of the supplier. Briefly, 10 μg of fragmented cRNA were hybridized to the microarrays in 200 μL of a hybridization cocktail at 45°C for 16 hours in a rotisserie oven set at 60 rpm. The arrays were then washed with a nonstringent wash buffer (6× saline-sodium phosphate-EDTA) at 25°C, followed by a stringent wash buffer [100 mmol/L MES (pH 6.7), 0.1 mol/L NaCl, and 0.01% Tween 20] at 50°C, stained with streptavidin phycoerythrin (Molecular Probes, Eugene, OR), washed again with 6× saline-sodium phosphate-EDTA, stained with biotinylated anti-streptavidin immunoglobulin G, followed by a second staining with streptavidin phycoerythrin and a third wash with 6× saline-sodium phosphate-EDTA. The arrays were scanned using a GeneArray scanner (Affymetrix) at 3-μm resolution, and the expression value (average difference) of each gene was calculated using GeneChip Analysis Suite version 4.0 software (Affymetrix). The mean of average difference values in each experiment was 1,000 to reliably compare variable multiple arrays. By use of human U95A version 2 arrays, we analyzed the eight treated and untreated gastric cancer cell lines twice and the 33 surgical specimens once. To select reactivated genes in at least one cell line of the eight gastric cancer cell lines after treatment with 5Aza-dC and TSA, we compared the mean of average difference values of duplicated experiments between untreated cells and 24-hour-treated cells. More than 3-fold overexpressed genes were selected by Microsoft Excel. The dendrogram of the clustering analysis was generated using the programs Cluster and Treeview (18).

Reverse Transcription-PCR and cRNA Slot-Blot Analyses. First, we produced tens of micrograms of cRNA from 1 to 5 μg total RNA prepared from the gastric cancer cell lines and the surgical specimen of gastric cancer by T7 transcription–mediated RNA amplification. Single stranded cDNAs were synthesized from 5 μg cRNA by use of First-strand synthesis kit (Amersham Biosciences, Piscataway, NJ) with random hexamers. We carried out PCR in a volume of 25 μL containing 1× Accuprime PCR buffer I (Invitrogen Corp., Carlsbad, CA), 0.5 μmol/L of each primer, and Accuprime Taq DNA polymerase (Invitrogen, Corpo, Carlsbad, CA). The thermal profile consisted of an initial denaturation at 95°C for 5 minutes followed by repetitions at 95°C for 1 minute, 56°C for 1 minute, and 72°C for 1 minute, with a final extension step at 72°C for 10 minutes. We amplified all of the genes from 50 ng of cDNA template with multiple cycle numbers (20-30 cycles) to determine the appropriate conditions for obtaining semiquantitative differences in expression levels. Primers for ELK1, R-RAS, RHOB, RHO6, and MSX2 were ELK1, 5′-ATGGGGCTTTTCAATAGGGG-3′ and 5′-CCAGGAGTCTTTTGAACCCA-3′; R-RAS, 5′-TCTGACTACGACCCCACTATTG-3′ and 5′-AAGGGTGGGTATGTGATGTGTC-3′; RHOB, 5′-GCCTGCTGATCGTGTTCAGT-3′ and 5′-GTCGTAGGCTTGGATGCGCA-3′; RHO6, 5′-CGCTCTGAACTCATCTCTTC-3′ and 5′-GGACTTAGGTGAGAAGCATGTG-3′; and MSX2, 5′-CCTGTTGAGAGGAATTGATGG-3′ and 5′-AAAGGTATACCGGAGGGAGG-3′. Two micrograms of cRNA for each sample were incubated at 65C for 10 minutes and then blotted onto a nylon membrane, Hybond-N+ (Amersham). Probes were obtained by the above reverse transcription-PCR (RT-PCR) method and then labeled with [32P]dCTP using a DNA labeling kit (Roche Diagnostics, Mannheim, Germany).

Methylation Analysis by Genomic Southern Blot. Five micrograms of genomic DNA from gastric cancer cell lines were digested to completion with HpaII or MspI. The restriction products were separated on 2% agarose gels and transferred to the nylon membranes. Two probes were obtained by genomic PCR using two primer sets, 5′-CTGGCTCATGGATTAGGAAT-3′ and 5′-CATAATCATGAGCTCTGGCA-3′ for probe 1, and 5′-ACCATCCAGTTCATCCAGGT-3′ and 5′-GGTGGAATCTCAAAGGTGCT-3′ for probe 2.

Bisulfite Sequencing Analysis. A region containing site S1 ( Fig. 4) was amplified from bisulfite-treated genomic DNA by PCR using two primer sets, 5′-TTTAAGTAGTTGAGATTATAGG-3′ and 5′-CTCTCTTAAAACTAAATAATCCC-3′ for site S1 analysis. The PCR products were applied to the ABI 310 DNA analyzer (Applied Biosystems) using the above primers or nested primers.

Methylation-Specific PCR. Bisulfite-treated genomic DNA was amplified with either a methylation-specific or unmethylation-specific primer set at 40 cycles: at 95°C for 5 minutes followed by repetitions at 95°C for 45 seconds, 52°C (methylated) and 52°C (unmethylated) for 45 seconds, and 72°C for 1 minute, with a final extension step at 72°C for 10 minutes. The methylation-specific primers for site S1 were designed using 5′-TTTAAGTAGTTGAGATTATAGG-3′ as the forward primer and 5′-AAAAACTAAAAATCAAACGTAAAACCG-3′ as the reverse primer. The unmethylation-specific primers for site S1 were designed using 5′-TTTAAGTAGTTGAGATTATAGG-3′ as the forward primer and 5′-AAAAAAAACTAAAAATCAAACATAAAACCA-3′ as the reverse primer.

RNAi Analysis. All siRNAs used for RNA interference of the R-RAS gene product were designed with an online siRNA design tool provided by Qiagen (http://www.qiagen.com/siRNA). Among the three siRNAs designed, which had no sequence identities with any other known genes including RAS family genes, two had a suppressive effect on R-RAS mRNA expression (data not shown). The target sequence used for subsequent analysis was ACACGAAGATCTGCAGTGTGGAT. The negative control siRNA of which sequence has no known homology to mammalian genes was also purchased from Qiagen. HSC58 and HSC59 cells were inoculated in 24-well plates at a density of 3 × 104 for each well. Twenty-four hours after inoculation, 76 pmol of siRNA were transfected into each well with RNAi Fect Transfection Reagent according to the instructions of the manufacturer (Qiagen). Fluorescein-labeled negative control siRNA was used to monitor the efficiency of transfection. Transfected cells were trypsinized, stained with trypan blue, and counted viable cells with a hemocytometer at the indicated time.

Results

Genome-wide Screening of Genes Reactivated by Pharmacologic Unmasking of Methylation and Deacetylation. To identify genes induced by treatment with 5Aza-dC and TSA, we used eight gastric cancer cell lines, HSC39, HSC43, HSC44, HSC58, HSC59, HSC60, KATOIII, and OCUM2M. In previous studies, five tumor suppressor genes, including CDH1 (19), TIMP2 (20), CDKN2D (21), CDKN1C (22), and RUNX3 (23, 24) , have been reported to be silenced by DNA methylation in gastric cancer. First, as a preliminary test we showed the results of the above four genes, except for RUNX3, of a microarray analysis of less than 1,000 genes ( Fig. 1 ). CDH1, TIMP2, CDKN2D, and CDKN1C were found to be induced in some of the eight gastric cancer cell lines by treatment with 5Aza-dC and TSA. We also confirmed the microarray data by RNA slot-blot analysis (data not shown). Next, all the RNA samples of untreated and treated cells were analyzed using microarrays containing 12,625 transcripts. In each cell line, ∼300 to 500 genes were found to be transcriptionally activated by the pharmacologic unmasking of methylation and deacetylation.

Figure 1.
Figure 1.

Microarray results of CDH1, TIMP2, CDKN2D, and CDKN1C genes in eight gastric cancer cell lines, HSC39, HSC43, HSC44, HSC58, HSC59, HSC60, KATOIII, and OCUM2M. Induction by treatment with 5Aza-dC and TSA is found in the four genes in some gastric cancer cell lines. The mean of expression values on duplicated experiments in untreated cells (white) and treated cells (black).

Selection of Transcriptionally Activated Genes in Gastric Cancer by Cancer-Linked DNA Hypomethylation. To select genes showing abnormal expression in gastric cancer due to cancer-linked DNA hypomethylation, we compared expression profiles of 33 primary gastric cancers and 6 normal gastric mucosae. Of the 1,383 genes, 159 genes, which are suppressed in 6 normal tissues but expressed in at least one gastric cancer tissue, were selected ( Table 1 ). We showed the location of CpG islands near the first exon of the 159 genes in Table 1. In concordance with a previous report (25), four tumor antigen genes, MAGEA1, MAGEA4, MAGEA9, and CTAG1, were found. Gastric cancer mainly divides into two types, diffuse and intestinal. In a manner similar to a recent report for disruption of cell-type-specific methylation at the Maspin gene promoter in undifferentiated thyroid cancers (26), tumor suppressor genes (MSH5 and CDH1 encoding E-cadherin) or tumor suppressor gene-associated genes (BAP1 encoding BRCA1-associated protein-1, and three BCL2-associated protein genes, BIK, BBC3, and BAX) were also found to be activated by DNA hypermethylation. To investigate whether type-specific or subtype-specific DNA hypomethylation occurs, we did a hierarchical clustering analysis of the 159 genes in the 33 gastric cancer tissues including 16 diffuse type and 17 intestinal type. As shown in Fig. 2 , 7 of 17 intestinal type cancers (horizontal green bar) and 8 of 16 diffuse type cancers (horizontal red bar) were clustered. We also provided three representative gene clusters in an inset in Fig. 2. The gene cluster of the diffuse type (vertical red bar) contains two known oncogenes, R-RAS and ARHB/RHOB, whereas that of the intestinal type (vertical green bar) contains an oncogene, ELK1.

View this table:
Table 1.

Gene candidates activated by cancer-associated hypomethylation

Figure 2.
Figure 2.

Hierarchical clustering analysis with expression data of the 159 demethylated gene candidates in primary gastric cancers. Expression data of the 159 genes in 33 primary gastric cancers including the 16 diffuse type and 17 intestinal type were analyzed by the programs Cluster and Treeview (16). The two clusters consist of 7 of 17 intestinal type cancers (horizontal green bar) and 8 of 16 diffuse type cancers (horizontal red bar). Vertical red bar, gene cluster of the diffuse type; vertical green bar, intestinal type. D-P, diffuse type (red); I, intestinal type (green).

In the gene list ( Table 1), we found six known oncogenes including a member of the ETS oncogene family, ELK1, an oncogene in WNT signaling, FRAT2, an oncogene amplified in gastric cancers, FGFR2, together with three oncogenes in the RAS/RHO family, R-RAS, RHOB, and RHO6. A Notch ligand, JAG1, and a target for the RAS signaling, MSX2, are also potential oncogene candidates. By RT-PCR analysis, we next investigated reactivation of the six oncogenes and the two oncogene candidates in eight gastric cancer cell lines by treatment with 5Aza-dC and/or TSA. As shown in Fig. 3A , five genes (ELK1, R-RAS, RHOB, RHO6, and MSX2) of the eight genes were reactivated by treatment with 5Aza-dC alone, TSA alone, and both 5Aza-dC and TSA. Although we have not done analyses to confirm reactivation, at least TK1, STAT1, RABGGTA, ABCC10, MMP10, CDC34, FZR1, and MAPK8IP1 are considered to be cancer-related genes. Among the five genes, we further analyzed the R-RAS gene because a critical promoter region has been previously reported (27). First, we did RT-PCR analysis of R-RAS in 6 noncancerous tissues and 33 primary gastric cancers. R-RAS was suppressed in the 6 normal tissues but expressed in 18 (55%) of 33 gastric cancers (9 in 16 diffuse type cancers, 9 in 17 intestinal type cancers; Fig. 3B).

Figure 3.
Figure 3.

A, RT-PCR analysis of four oncogenes, ELK1, R-RAS, RHOB, and RHO6, and one potential oncogene, MSX2, in untreated and treated cells with 5Aza-dC and/or TSA. B, RT-PCR of R-RAS in 6 noncancerous tissues and 33 primary gastric cancer tissues. R-RAS is suppressed in the 6 normal tissues but expressed in 18 of 33 (55%) primary gastric cancers. C, RT-PCR of R-RAS in three laser-captured regions, the pit, isthmus/neck, and gland, in the gastric mucosa. No R-RAS mRNA is detected in the three regions, whereas H-RAS mRNA is expressed in all of the three regions.

The gastric epithelium consists of tubular units of epithelial cells (28). Each unit is divided into three successive regions from surface to base: pit, isthmus/neck, and gland. The isthmus or neck contains stem cells from which mature epithelial cells are derived and moved to the other two regions. Mucus-secreting pit cells migrate up from progenitor cells toward the gastric lumen. The neck cells migrate down toward the base of the gland where they give rise to pepsinogen-producing chief cells. To conclude that there is no R-RAS expression in all of the three regions in the gastric mucosa, we carried out RT-PCR analysis combined with a laser-captured microdissection procedure on the human gastric mucosa as described in our previous report (29). With this procedure, no R-RAS mRNA was detected in the three regions whereas H-RAS mRNA was expressed in all of the three regions ( Fig. 3C). The expression of a pit cell marker, MUC5AC (30, 31) , was significantly higher in the pit cell-enriched sample whereas a proliferating cell marker, PCNA, was detected preferentially in the neck cell-enriched sample. This indicates the procedure was appropriately done for our purpose. These results show that R-RAS is silenced in the gastric epithelium but activated in more than half of gastric cancers.

Aberrant Expression of R-RAS in Gastric Cancers by Demethylation of Specific CpG Sites within the First Intron. We found two CpG islands (GC content >50%, length >200 bp, and ratio of CpG to GpC >0.6) surrounding the first exon by searching with a UCSC Genome Browser (32, 33) . A schematic diagram of the structure of the 5′ region of the R-RAS gene is shown in Fig. 4A . We first investigated the methylation status in the 5′ region of R-RAS in eight gastric cancer cell lines by genomic Southern blot analysis with a methylation-sensitive enzyme, HpaII, and a resistant enzyme, MspI. Two probes for the Southern blots are shown in Fig. 4A, and results of Southern blot analysis with probe 2 and RNA slot-blot analysis for quantifying R-RAS mRNA levels in each cell line are shown in Fig. 4B. Southern blot analysis with probe 1 showed no correlation between the methylation status of the MspI/HapII sites and R-RAS expression levels (data not shown), whereas Southern blot analysis with probe 2 clearly showed two types of correlation ( Fig. 4A and B). Four bands, termed B1-4, were detected in Southern blot analysis with probe 2. Increased R-RAS mRNA levels correlated with a shift from B3 to B1 in HSC39, KATOIII, HSC43 or OCUM2M, HSC60, and HSC58 (type I correlation in Fig. 4B). Another correlation (type II) is that R-RAS mRNA levels increased depending on the shift from B2 to B4 in OCUM2M or HSC43, HSC59, and HSC44 ( Fig. 4B).

Figure 4.
Figure 4.

A, schematic diagram of the structure of the 5′ region of the R-RAS gene. Genomic locus of the first exon was identified by UCSC Genome Browser and the 5′ CpG island of R-RAS was detected by a CpG Island Searcher (42). Two probes for genomic Southern blot analysis with HpaII and MspI are indicated (probe 1 and probe 2). MspI/Hap/I sites including sites S1-4 and four bands, termed B1-4, which were detected in Southern blot analysis with probe 2 (B and C), are also indicated. B, genomic Southern blot analysis with HpaII and MspI in eight gastric cancer cell lines using probe 2. H and M, blots with HpaII and MspI, respectively. Each result of RNA slot-blot analysis in eight gastric cancer cell lines is shown (bottom). C, genomic Southern blot analysis with HpaII and MspI in two normal tissues and two primary gastric cancer tissues with R-RAS expression using probe 2. H and M, blots with HpaII and MspI, respectively. Each result of RT-PCR in the above four samples is shown (bottom). Two normal tissues (N1 and N2); two primary gastric cancer tissues [T1 (I-W5) and T2 (D-P(2)2s)]. D, methylation status of the MspI/HapII site S1 and its surrounding CpG sites in gastric cancer cell lines. R-RAS mRNA levels increased depending on the levels of demethylation on site S1 and its surrounding CpG sites in HSC39, KATOIII, HSC43 or OCUM2M, HSC60, and HSC58. E, methylation-specific PCR analysis on site S1 in normal tissues, gastric cancer cell lines, and primary gastric cancers. Two noncancerous tissues (N1 and N2); two primary gastric cancer tissues without R-RAS expression [T4 (I-P(1)11) and T5 (D-PS59)]; and three R-RAS–expressing cancer tissues [T1 (I-W5), T2 (D-P(2)2S), and T3 (I-W80)]. Demethylation on site S1 is found only in cancer cell lines and primary gastric cancers that express R-RAS mRNA. The PCR products in lanes U and M indicate the presence of unmethylated and methylated templates, respectively.

Thus, in gastric cancer cell lines, DNA demethylation for R-RAS activation occurred in a bidirectional manner or in two types; however, in primary gastric cancers, one of the two directions or the two types was only found by Southern blot analysis (data not shown). Representative data are shown in Fig. 4C. In two normal tissues (N1 and N2) and two primary gastric cancers with R-RAS mRNA expression (T1 and T2), R-RAS mRNA levels also increased depending mainly on the shift from B3 to B1 ( Fig. 4C). These data indicated that cancer-linked demethylation on the MspI/HapII site S1 mainly leads to abnormal expression of the R-RAS gene in gastric carcinogenesis.

By genomic bisulfite sequencing in eight gastric cancer cell lines, we confirmed the methylation status of the MspI/HapII site S1 and its surrounding CpG sites in gastric cancer. As shown in Fig. 4D, R-RAS mRNA levels increased depending on the levels of demethylation on site S1 and its surrounding CpG sites in HSC39, KATOIII, HSC43, HSC60, and HSC58.

We further did methylation-specific PCR using bisulfite-modified genomic DNAs in gastric cancer cell lines and primary gastric cancers. In the above five gastric cancer cell lines (HSC39, KATOIII, HSC43, HSC60, and HSC58), demethylation of site S1 was only found in HSC60 and HSC58 depending on the levels of R-RAS mRNA ( Fig. 4E). We also found a correlation between the demethylation status of site S1 and R-RAS expression was observed in primary gastric cancers. Representative data were shown in Fig. 4E. Two noncancerous tissues (N1 and N2) and two primary gastric cancer tissues [T4 (I-P(1)11) and T5 (D-PS59)] without R-RAS expression showed no demethylation, whereas three R-RAS-expressing cancer tissues [T1 (I-W5), T2 (D-P(2)2S), and T3 (I-W80)] showed partial demethylation.

Suppression of R-RAS by RNAi Inhibited Cell Survival in R-RAS-Expressing Gastric Cancer Cells. By mutations at codons 38 or 87 analogous to positions 12 and 61 of H-RAS for its oncogenic activation, R-RAS has been reported to have oncogenic activity, although its activity is weaker than that of H-RAS or K-RAS (34). Therefore, we investigated whether oncogenic mutation of R-RAS occurs in gastric carcinogenesis. Direct sequencing analysis of genomic and cDNA fragments containing the above two sites showed no mutation in 22 gastric cancers with R-RAS mRNA expression (data not shown).

To clarify the biological meanings of aberrant expression of R-RAS mRNA in gastric cancer, we tried suppressing R-RAS in HSC58 cells with the highest level of R-RAS mRNA and HSC59 with no R-RAS mRNA by RNAi. The efficiencies of the siRNA introduction that were determined by use of a fluorescein-labeled negative control siRNA were 80% to 90% in HSC58 cells and 60% to 70% in HSC59 cells (data not shown). In R-RAS-expressing HSC58 cells, the introduction of R-RAS-targeted siRNA significantly decreased the cell number that adhered to a plate at 48 hours after the introduction, but the control siRNA did not ( Fig. 5B ), whereas no siRNA effect was observed in HSC59 cells (data not shown). Along with this effect in HSC58 cells, aggregated cells disappeared, and shrunken cells were frequently observed ( Fig. 5C). Because a decreased level of R-RAS mRNA was detected at 7.5 hours after siRNA transfection ( Fig. 5A), phenotypic alterations might be induced by some downstream events rather than a direct effect of decreased expression of R-RAS mRNA. In any case, these results suggest that R-RAS is essential for the survival of the gastric cancer cells that express R-RAS by DNA demethylation.

Figure 5.
Figure 5.

A, real-time PCR analysis of R-RAS mRNA after siRNA transfection. Decreased R-RAS mRNA was detected at 7.5 hours after siRNA transfection, whereas no change of GAPDH mRNA level was observed even at 24 hours after the transfection. B, effect on cell growth and survival after siRNA transfection. The survived cell number clearly decreased at 48 hours after R-RAS siRNA transfection. C, morphologic changes of cells by siRNA transfection. Accompanied by decreasing cell number, aggregated cells disappeared and shrunken cells were observed at 48 hours after transfection of R-RAS siRNA.

Discussion

In humans, the Ras family of small GTPases consists of nine closely related proteins and can be further divided into two subgroups in terms of structural similarity. The first of these contains the four classic p21 Ras proteins, H-RAS, N-RAS, K-RAS, and E-RAS that was previously recognized as a pseudogene of H-RAS but has now become a new member (35). The other five subgroups consist of RAP, RAL, R-RAS, R-RAS2, and M-RAS. Each member has different effects on biological phenotypes.

The R-RAS gene was initially cloned as a ras-related gene (27) and showed interaction with the Bcl-2 gene product (36). Owing to an actively mutated R-RAS inhibiting Bcl-2-mediated rescue of apoptosis, it is thought that R-RAS promotes apoptosis signaling pathway(s) (37). On the other hand, as McCormick (38) pointed out, interaction between R-RAS and Bcl-2 might have a biological function other than apoptosis because the Bcl-2 protein preferentially binds to the nucleotide-depleted form of R-RAS in vitro. Cell transformed by activated H-RAS can survive in the absence of serum whereas R-RAS-transformed cells seem to die by an apoptotic-like mechanism in response to the removal of the serum (39). It has also been shown that R-RAS activates integrin-mediated and integrin-independent cell adhesion (40, 41) .

Suppression of R-RAS expression by RNA interference induced morphologic alteration and cell death ( Fig. 5). Because of the results of terminal deoxyribonucleotidyl transferase–mediated dUTP nick end labeling assay showing that the proportion of apoptotic cells within the adherent cell population did not alter even after siRNA transfection (data not shown), cell death might arise immediately after cell detachment from the substrate by means of R-RAS interference by siRNA. Our results are consistent with the model for regulation of cell adhesion by R-RAS. The biological significance of the R-RAS function in tumor tissues should be analyzed in detail.

This study suggests that functional blocking of R-RAS itself and/or the R-RAS-signaling pathway has great potential for gastric cancer therapy. Moreover, methylation-specific PCR analysis of the R-RAS gene could become a conventional tool for identifying target patients for any new drug that may be developed in the near future.

In the same manner, the gene list containing 159 genes also provides many candidates for cancer-related genes that are transcriptionally activated by DNA hypomethylation in human cancers. Further identification of such genes may provide insight into tumor biology and provide novel diagnostic and therapeutic targets.

Acknowledgments

Grant support: Program for the Promotion of Fundamental Studies in Health Sciences of the Pharmaceuticals and Medical Devices Agency; Grant-in-Aid for the 3rd Comprehensive 10-Year Strategy for Cancer Control and for Cancer Research (15-5 and 16-16) from the Ministry of Health, Labour and Welfare of Japan; Research Grant of the Princess Takamatsu Cancer Research Fund; and Research Resident Fellowship award from the Foundation for Promotion of Cancer Research (M. Fukaya).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Footnotes

    • Received September 15, 2004.
    • Revision received December 22, 2004.
    • Accepted January 11, 2005.

    References

    1. ↵
      Baylin SB, Herman JG. DNA hypermethylation in tumorigenesis: epigenetics joins genetics. Trends Genet 2000; 16: 168–74.
      OpenUrlCrossRefPubMed
    2. ↵
      Costello JF, Fruhwald MC, Smiraglia DJ, et al. Aberrant CpG-island methylation has non-random and tumour-type-specific patterns. Nat Genet 2000; 24: 132–8.
      OpenUrlCrossRefPubMed
    3. ↵
      Bachman KE, Park BH, Rhee I, et al. Histone modifications and silencing prior to DNA methylation of a tumor suppressor gene. Cancer Cell 2003; 3: 89–95.
      OpenUrlCrossRefPubMed
    4. ↵
      Feinberg AP, Tycko B. The history of cancer epigenetics. Nat Rev Cancer 2004; 4: 143–53.
      OpenUrlCrossRefPubMed
    5. ↵
      Jaenisch R, Schnieke A, Harbers K. Treatment of mice with 5-azacytidine efficiently activates silent retroviral genomes in different tissues. Proc Natl Acad Sci U S A 1985; 82: 1451–5.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      Chen RZ, Pettersson U, Beard C, Jackson-Grusby L, Jaenisch R. DNA hypomethylation leads to elevated mutation rates. Nature 1998; 395: 89–93.
      OpenUrlCrossRefPubMed
    7. ↵
      Gaudet F, Hodgson JG, Eden A, et al. Induction of tumors in mice by genomic hypomethylation. Science 2003; 300: 489–92.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      Feinberg AP, Vogelstein B. Hypomethylation distinguishes genes of some human cancers from their counterpart. Nature 1983; 301: 89–91.
      OpenUrlCrossRefPubMed
    9. ↵
      Cho M, Uemura H, Kim S, et al. Hypomethylation of the MN/CA9 promoter and up-regulated MN/CA9 expression in human renal cell carcinoma. Br J Cancer 2001; 85: 563–7.
      OpenUrlCrossRefPubMed
    10. ↵
      Nakayama M, Wada M, Harada T, et al. Hypomethylation status of CpG sites at the promoter region and overexpression of the human MDR1 gene in acute myeloid leukemias. Blood 1998; 92: 4296–307.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      Hanada M, Delia D, Aiello A, Stadtmauer E, Reed J. BCL-2 gene hypomethylation and high level expression in B-cell chronic lymphocytic leukemia. Blood 1993; 82: 1820–8.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      De Smet C, De Backer O, Faraoni I, Lurquin C, Brassuer F, Boon T. The activation of human gene MAGE-1 in tumor cells is correlated with genome-wide demethylation. Proc Natl Acad Sci U S A 1996; 93: 7149–53.
      OpenUrlAbstract/FREE Full Text
    13. ↵
      Gupta A, Godwin AK, Vanderveer L, Lu A, Liu J. Hypomethylation of the synuclein γ gene CpG island promotes its aberrant expression in breast carcinoma and ovarian carcinoma. Cancer Res 2003; 63: 664–73.
      OpenUrlAbstract/FREE Full Text
    14. ↵
      Ueda K, Fukushima H, Masliah E, et al. Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease. Proc Natl Acad Sci U S A 1993; 90: 11282–6.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      Jakes R, Sillantini M, Goedert M. Identification of two distinct synucleins from human brain. FEBS Lett 1994; 345: 27–34.
      OpenUrlCrossRefPubMed
    16. ↵
      Sato N, Maitra A, Fukushima N, et al. Frequent hypomethylation of multiple genes overexpressed in pancreatic ductal adenocarcinoma. Cancer Res 2003; 63: 4158–66.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      Fang JY, Xiao SD. Alteration of DNA methylation in gastrointestinal carcinogenesis. J Gastroenterol Hepatol 2001; 16: 960–8.
      OpenUrlCrossRefPubMed
    18. ↵
      Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 1998; 95: 14863–8.
      OpenUrlAbstract/FREE Full Text
    19. ↵
      Grady WM, Willis J, Guilford PJ, et al. Methylation of the CDH1 promoter as the second genetic hit in hereditary diffuse gastric cancer. Nat Genet 2000; 26: 16–7.
      OpenUrlCrossRefPubMed
    20. ↵
      Suzuki H, Gabrielson E, Chen W, et al. A genomic screen for genes up-regulated by demethylation and histone deacetylase inhibition in human colorectal cancer. Nat Genet 2002; 31: 141–9.
      OpenUrlCrossRefPubMed
    21. ↵
      Song S-H, Jong H-S, Choi H-H, et al. Methylation of specific CpG sites in the promoter region could significantly down-regulate p16(INK4a) expression in gastric adenocarcinoma. Int J Cancer 2000; 87: 236–40.
      OpenUrlCrossRefPubMed
    22. ↵
      Shin JY, Kim H-S, Park J, Park J-B, Lee JY. Mechanism for inactivation of the KIP family cyclin-dependent kinase inhibitor genes in gastric cancer cells. Cancer Res 2000; 60: 262–5.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      Li Q-L, Ito K, Sakakura C, et al. Causal relationship between the loss of RUNX3 expression and gastric cancer. Cell 2002; 109: 113–24.
      OpenUrlCrossRefPubMed
    24. ↵
      Waki T, Tamura G, Sato M, Terashima M, Nishizuka S, Motoyama T. Promoter methylation status of DAP-kinase and RUNX3 genes in neoplastic and non-neoplastic gastric epithelia. Cancer Sci 2003; 94: 360–4.
      OpenUrlCrossRefPubMed
    25. ↵
      Li J, Yang Y, Fujie T, et al. Expression of BAGE, GAGE, and MAGE genes in human gastric carcinoma. Clin Cancer Res 1996; 2: 1619–25.
      OpenUrlAbstract
    26. ↵
      Ogasawara S, Maesawa C, Yamamoto M, et al. Disruption of cell-type-specific methylation at the Maspin gene promoter is frequently involved in undifferentiated thyroid cancers. Oncogene 2004; 23: 1117–24.
      OpenUrlCrossRefPubMed
    27. ↵
      Lowe DG, Capon DJ, Delwart E, Sakaguchi A, Naylor SL, Goeddel DV. Structure of the human and murine R-ras genes, novel genes closely related to ras proto-oncogenes. Cell 1987; 48: 137–46.
      OpenUrlCrossRefPubMed
    28. ↵
      Karam SM, Leblond CP. Origin and migratory pathways of the eleven epithelial cell types present in the body of the mouse stomach. Microsc Res Tech 1995; 31: 193–214.
      OpenUrlCrossRefPubMed
    29. ↵
      Aoyagi K, Tatsuta T, Nishigaki M, et al. A faithful method for PCR-mediated global mRNA amplification and its integration into microarray analysis on laser-captured cells. Biochem Biophys Res Commun 2003; 300: 915–20.
      OpenUrlCrossRefPubMed
    30. ↵
      De Bolos C, Garrido M, Real FX. MUC6 apomucin shows a distinct normal tissue distribution that correlates with Lewis antigen expression in the human stomach. Gastroenterology 1995; 109: 723–34.
      OpenUrlCrossRefPubMed
    31. ↵
      Ho SB, Roberton AM, Shekels LL, Lyftogt CT, Niehans GA, Toribara NW. Expression cloning of gastric mucin complementary DNA and localization of mucin gene expression. Gastroenterology 1995; 109: 735–47.
      OpenUrlCrossRefPubMed
    32. ↵
      Kent WJ, Sugnet CW, Furey TS, et al. The human genome browser at UCSC. Genome Res 2002; 12: 996–1006.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      Karolchik D, Baertsch R, Diekhans M, et al. The UCSC Genome Browser Database. Nucleic Acids Res 2003; 31: 51–4.
      OpenUrlAbstract/FREE Full Text
    34. ↵
      Saez R, Chan AM, Miki T, Aaronson SA. Oncogenic activation of human R-ras by point mutations analogous to those of prototype H-ras oncogenes. Oncogene 1994; 9: 2977–82.
      OpenUrlPubMed
    35. ↵
      Takahashi K, Mitsui K, Yamanaka S. Role of ERas in promoting tumour-like properties in mouse embryonic stem cells. Nature 2003; 423: 541–5.
      OpenUrlCrossRefPubMed
    36. ↵
      Fernandez-Sarabia MJ, Bishoff JR. Bcl-2 associates with the ras-related protein R-ras p23. Nature 1993; 366: 274–5.
      OpenUrlCrossRefPubMed
    37. ↵
      Troppmair J, Rapp UR. Apoptosis regulation by Raf, Bcl-2, and R-RAS. Recent Results Cancer Res 1997; 143: 245–9.
      OpenUrlPubMed
    38. ↵
      McCormick F. Ras-related proteins in signal transduction and growth control. Mol Reprod Dev 1995; 42: 500–6.
      OpenUrlCrossRefPubMed
    39. ↵
      Hansen M, Rusyn EV, Hughes PE, Ginsberg MH, Cox AD, Willumsen BM. R-Ras C-terminal sequences are sufficient to confer R-Ras specificity to H-Ras. Oncogene 2002; 21: 4448–61.
      OpenUrlCrossRefPubMed
    40. ↵
      Zhang Z, Vuori K, Wang H-G, Reed JC, Ruoslahti E. Integrin activation by R-RAS. Cell 1996; 85: 61–9.
      OpenUrlCrossRefPubMed
    41. ↵
      Kwong L, Wozniak MA, Collins AS, Wilson SD, Keely PJ. R-RAS promotes focal adhesion formation through focal adhesion kinase and 130 (Cas) by a novel mechanism that differs from integrins. Mol Cell Biol 2003; 23: 933–49.
      OpenUrlAbstract/FREE Full Text
    42. ↵
      Takai D, Jones PA. Comprehensive analysis of CpG islands in human chromosomes 21 and 22. Proc Natl Acad Sci U S A 2002; 99: 3740–5.
      OpenUrlAbstract/FREE Full Text
    View Abstract
    Back to top
    Cancer Research: 65 (6)
    March 2005
    Volume 65, Issue 6

    Sign up for alerts

    View this article with LENS

    Open full page PDF
    Article Alerts
    Email Article
    Citation Tools
    Discovery of Aberrant Expression of R-RAS by Cancer-Linked DNA Hypomethylation in Gastric Cancer Using Microarrays
    Michiko Nishigaki, Kazuhiko Aoyagi, Inaho Danjoh, Masahide Fukaya, Kazuyoshi Yanagihara, Hiromi Sakamoto, Teruhiko Yoshida and Hiroki Sasaki
    Cancer Res March 15 2005 (65) (6) 2115-2124; DOI: 10.1158/0008-5472.CAN-04-3340
    del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

    Jump to section

    Advertisement