The 3p21.3 tumour suppressor gene (TSG) RASSF1A is inactivated predominantly by promoter methylation and rarely by somatic mutations. Recently we demonstrated that epigenetic inactivation of RASSF1A is frequent in both clear cell and papillary adult renal cell carcinomas (even though 3p21.3 allele loss is rare in papillary tumours). Wilms' tumour is the most common childhood kidney tumour, but relatively little is known about its molecular pathogenesis. Thus TSGs such as WT1, p16(CDKN2a) and p53 are inactivated in only a minority of cases. In view of the involvement of RASSF1A in adult renal cancers we investigated RASSF1A as a candidate Wilms' TSG. We detected RASSF1A hypermethylation in 21 of 39 (54%) primary Wilms' tumours. 3p21.3 allele loss was not detected in nine informative Wilms' tumours (five with RASSF1A methylation). In contrast to RASSF1A, only a minority (10.3%) of Wilms' tumours demonstrated p16 promoter methylation. As chromosome 3p allele loss is frequent in colorectal cancer, we proceeded to investigate RASSF1A promoter methylation in colorectal cancer and detected RASSF1A methylation in 80% (4/5) colorectal cancer cell lines and 45% (13/29) primary colorectal cancers. There was no correlation between RASSF1A and p16 methylation in colorectal cancer. We have demonstrated that RASSF1A inactivation is the most frequent genetic or epigenetic event yet reported in Wilms' tumourigenesis and that allelotyping studies may fail to identify regions containing important TSGs.
Alteration in genomic methylation is now established as an important mechanism in the pathogenesis of human cancers. Most commonly, methylation promotes tumourigenesis by promotor methylation and transcriptional silencing of tumour suppressor genes such as p16, CDH1and VHL (Baylin et al., 1998; Jones, 1999; Tycko, 2000; Costello and Plass, 2001 and references within). However, promoter methylation may also provoke genomic instability by epigenetic inactivation of mismatch repair genes such as MLH1 (Kane et al., 1997; Herman et al., 1998). In addition, alterations in methylation may also promote tumourigenesis through altered expression of oncogenes. Thus the methylation of regulatory elements may result in abnormal biallelic expression of the imprinted growth promoter insulin-like growth factor 2 (IGF2) (and silencing of expression of the closely linked H19 gene) in human cancers such as Wilms' tumour and colorectal cancer (Steenman et al., 1994; Cui et al., 1998). Initial studies of genes such as VHL and RB1 suggested that inactivation of tumour suppressor genes by promoter methylation was a less frequent mechanism of inactivation than somatic mutations (Herman et al., 1994; Stirzaker et al., 1997). However, it is now recognized that for some tumour suppressor genes, such as RASSF1A, promoter methylation is the major mechanism of inactivation (Dammann et al., 2000; Agathanggelou et al., 2001, Burbee et al., 2001). Furthermore, although allele loss and de novo methylation often account for the ‘two hits’ resulting in tumour suppressor gene inactivation, methylation can also occur commonly at loci at which allele loss is rare (e.g. RASSF1A methylation in papillary RCC (Morrissey et al., 2001).
The short arm of chromosome 3 contains multiple TSGs and 3p allele loss is a frequent finding in many human cancer types. In addition to the VHL and FHIT TSGs at 3p25 and 3p14 respectively, we and others have reported recently that the 3p21.3 TSG, RASSF1A, is methylated in many tumour types including lung, breast and renal cancers (Dammann et al., 2000; Agathanggelou et al., 2001; Burbee et al., 2001; Morrissey et al., 2001).
Wilms' tumour is the most common childhood kidney tumour and is thought to be derived from primitive nephroblasts. Colorectal cancer is the second most common adult cancer in the Western world. Alterations in genomic methylation have been implicated in the pathogenesis of both tumour types, although Wilms' tumour has been investigated much less frequently. Thus, alterations in methylation affecting IGF2 and H19 are well described in Wilms' tumour and recently p16 (CDKN2a) promoter methylation was described in some tumours (Steenman et al., 1994; Arcellana-Panlilio et al., 2000). Changes in genomic methylation in colorectal cancer have been the subject of intense enquiry. In particular, a subgroup of colorectal cancers is reported to demonstrate a methylator phenotype with frequent promoter methylation of TSGs such as p16, MLH1, CDH1 and hypermethylation of IGF2/H19 (Toyota et al., 1999). Chromosome 3p allele loss is frequent in colorectal cancer, but VHL inactivation appears to be rare and the involvement of FHIT is variable (Thiagalingam et al., 1996; Hao et al., 2000). Consequently we wished to investigate RASSF1A as a candidate colorectal cancer TSG. In addition, as epigenetic silencing of RASSF1A is frequent in adult renal cancers (including papillary RCC in which 3p allele loss is rare) we also investigated RASSF1A status in Wilms' tumour. Furthermore we sought to investigate the relationship between RASSF1A promoter methylation and methylation of other genes (e.g. p16 and H19) in Wilms' tumours and colorectal cancers.
RASSF1A methylation status in Wilms' tumour
Twenty-one of 39 (54%) primary Wilms' tumours demonstrated RASSF1A hypermethylation by methylation specific PCR (MSP) (Figure 1a). To confirm the presence of promoter hypermethylation and to determine the precise pattern of CpG methylation within the RASSF1A CpG island we directly sequenced the methylated fragment from each of the 21 Wilms' tumours with RASSF1A methylation from nucleotide −110 bp to +41 bp after sodium bisulphite modification. Eight tumour samples demonstrated methylation of all 16 CpGs within the amplified fragment and seven had complete methylation at 15 of 16 CpGs (in each case partial methylation was detected at CpG 8 (see Figure 1b)). Two samples demonstrated complete methylation of 13 of 16 CpGs and partial methylation at CpGs 4 and 8 plus one other CpG. Another sample showed complete methylation of 14 CpGs, with partial methylation of CpG 9 and was completely unmethylated at CpG 8. Most tumours (19/24) with RASSF1A methylation also contained unmethylated RASSF1A alleles which might be attributable to the presence of contaminating normal tissue (tumour samples were not microdissected). Corresponding normal tissue DNA was available for 13 tumours with RASSF1A methylation. No RASSF1A methylation was detected in the matched blood (n=4) and normal kidney (n=7) DNA samples for 11 of the 13 cases. However in two cases normal tissue RASSF1A methylation was detected: (a) in one case the normal kidney DNA demonstrated unmethylated and methylated RASSF1A alleles, whereas the tumour DNA was completely methylated and (b) in another case both tumour and normal kidney DNA demonstrated partial RASSF1A methylation.
Clinical information on tumour stage was available for 32 Wilms' tumours (WTs). The frequency of RASSF1A methylation, 68% (13/19), was higher in more advanced stage 3 and 4 tumours than in stage 1 and 2 tumours (50%, 7/14, P=0.47) but this was not significantly different. Seven patients relapsed following treatment of the primary tumour, however the frequency of RASSF1A methylation in these primary tumours (57%) was similar to the overall frequency of RASSF1A methylation in Wilms' tumour.
Nine Wilms' tumours were informative for 3p allele loss studies at D3S4604 (see Figure 1C), however none showed evidence of 3p allele loss (5/9 had RASSF1A methylation).
To confirm that RASSF1A is expressed in foetal kidney RT–PCR analysis of RASSF1A transcript expression was undertaken in human foetal tissue samples and demonstrated similar levels of RASSF1A transcript in human foetal kidney, lung, brain, liver and heart (see Figure 3b).
Relationship between RASSF1A, H19 and p16 (CDKN2a) methylation status in Wilms' tumours
We investigated the relationship between RASSF1A methylation in WT and p16 and H19 methylation. Four of 39 (10.3%) Wilms' tumours demonstrated p16 promoter methylation (Figure 2a). The frequency of p16 methylation in Wilm's tumours with RASSF1A methylation was not different to that in tumours without RASSF1A methylation (3/21 versus 1/18 respectively). p16 methylation was detected in 7% (1/14) stage 1 and 2 tumours, 15.4% (2/13) stage 3 tumours and 25% (1/4) stage 4 tumours.
Twenty-five of 39 WTs were informative for assessment of 11p15.5 allele loss, and 19 demonstrated no allele loss. The frequency of RASSF1A methylation in tumours without 11p LOH was higher than in those tumours with 11p allele loss (15/19 (79%) and 2/6 (33%) respectively) but this difference did not quite reach statistical significance (P=0.06). Analysis of H19 promoter methylation in 19 WTs without 11p15.5 allele loss demonstrated an abnormal biallelic methylation pattern in five tumours (Figure 2b). There was no association between RASSF1A methylation and H19 hypermethylation: the frequency of RASSF1A methylation in Wilms' tumours without 11p allele loss and normal H19 methylation pattern was 71% (10/14) compared to 80% (4/5) in tumours with H19 hypermethylation.
RASSF1A methylation in colorectal cancer
We detected RASSF1A promoter methylation in 13 of 29 (45%) primary colorectal cancers and four of five colorectal cancer cell lines analysed. RASSF1A methylation was also detected in four of 11 matching normal colorectal mucosa samples available for tumours with RASSF1A methylation.
The RASSF1A methylated fragments amplified from 10 colorectal cancers and four normal tissue samples were sequenced. Five of the 10 tumour and one of four normal tissue RASSF1A methylated fragments demonstrated methylation at all 16 CpGs. The remaining tumour and normal samples methylated RASSF1A fragments amplified were completely methylated at a minimum of 12 of the 16 CpGs.
We investigated whether RASSF1A methylation correlated with TNM status, but did not detect any significant associations. Fifteen colorectal cancers were informative for 3p allele loss studies at D3S4604, and 2/15 demonstrated 3p allele loss (0/2 with and 4/13 without 3p21 allele loss had RASSF1A methylation).
Relationship between RASSF1A, H19 and p16 methylation status in colorectal cancer
We investigated the relationship between RASSF1A and p16 methylation in colorectal cancer. Six of 29 (21%) primary colorectal cancers and 5/6 colorectal cancer cell lines demonstrated p16 promoter methylation. The frequency of p16 methylation in primary colorectal cancers with RASSF1A methylation was similar to that in tumours without RASSF1A methylation (2/13 versus 4/16 respectively).
RASSF1A methylation and expression in colorectal cancer cell lines
To confirm that RASSF1A promoter methylation is associated with reduced transcription, the colorectal cancer cell line SW48 was treated with demethylating agent 5′ aza-2deoxy-cytidine for 5 days. RASSF1A expression increased after 5′ aza-2deoxy-cytidine treatment while there was no change in expression of RASSF1C (previous studies have demonstrated that the RASSF1 isoform C is widely expressed and not susceptible to methylation silencing) (Figure 3a).
Epigenetic silencing of RASSF1A in Wilms' tumour and colorectal cancer
We have demonstrated that RASSF1A promoter methylation is frequent in both Wilms' tumour and colorectal cancer. Although epigenetic silencing of tumour suppressor genes is a common and important event in many human cancer types (Costello and Plass, 2001), the role of epigenetic events in Wilms' tumour has not been investigated extensively. Thus, although hypermethylation of the H19 promoter associated with loss of imprinting (and biallelic expression) of IGF2 is well described (Steenman et al., 1994), there are few studies of tumour suppressor gene methylation in Wilms' tumours. Recently, Arcellana-Panlilio et al. (2000) reported p16 methylation in 23% of Wilms' tumours, although the frequency varied from ∼10% of stage 1 tumours to 40% of stage IV tumours. We detected p16 methylation in ∼10% of Wilms' tumours and the frequency of p16 methylation increased with advanced tumour stage. However, the frequency of p16 methylation in Wilms' tumours was much lower than that for RASSF1A methylation. We and others have demonstrated previously that RASSF1A promoter methylation is associated with loss of mRNA expression in a variety of cancer cells (Dammann et al., 2000; Agathanggelou et al., 2001; Burbee et al., 2001; Morrissey et al., 2001). Furthermore, RASSF1A suppresses the growth of tumour cell lines in both in vivo and in vitro studies (Dammann et al., 2000; Burbee et al., 2001). As WT1 inactivation occurs in only a minority of sporadic WTs (Davies et al., 1999; Lee and Haber, 2001) and p53 mutations are uncommon (Waber et al., 1993; Bardeesy et al., 1994), epigenetic inactivation of RASSF1A is the most frequent cause of tumour suppressor gene inactivation yet described in Wilms' tumour. In a subset of patients with Wilms' tumour and H19 hypermethylation, the H19 epimutation can also be detected in the non-neoplastic renal parenchyma (Moulton et al., 1994; Okamoto et al., 1997). We found that in most cases RASSF1A methylation was restricted to tumour tissue, however in one case the adjacent renal parenchyma was mosaic for RASSF1A methylation while the tumour was completely methylated. In this case RASSF1A methylation may have been an early event in Wilms' tumour precursor cells (although contamination of normal tissue with cancer cells cannot be excluded completely). For certain tumours, such as colorectal cancer, a subset of tumours display frequent methylation of many genes (methylator phenotype) (Toyota et al., 1999). We found a higher frequency of p16 methylation in tumours with RASSF1A methylation than those without RASSF1A methylation. However this association was not statistically significant and larger studies are required to determine if this association is real and indicative of a methylator phenotype in a subset of Wilms' tumours.
Although the presence of homozygous deletions in lung and breast cancer defined the critical region from which RASSF1A was isolated (Lerman and Minna, 2000), RASSF1A methylation has been detected frequently in tumour types in which 3p21 allele loss is common (e.g. small cell lung cancer, clear cell renal cell carcinoma) or rare (e.g. papillary renal cell carcinoma (Dammann et al., 2000; Agathanggelou et al., 2001; Burbee et al., 2001; Morrissey et al., 2001). Thus, as illustrated by RASSF1A involvement in Wilms tumour, allelotyping studies may fail to identify regions containing important TSGs.
The RASSF1 gene has several major isoforms due to alternative splicing and promoter usage, but epigenetic silencing of the longer isoform, RASSF1A, is specifically associated with cancer. The RASSF1A 39 KDa predicted peptide contains a RAS association domain, a diacylglycerol (DAG) binding domain and a region that is a putative substrate for ATM phosphorylation (Vos et al., 2000). In both in vitro and in vivo studies RASSF1 was reported to be an effector of RAS (Vos et al., 2000). The finding of frequent RASSF1A inactivation in lung cancers is consistent with previous reports implicating mutations or alterations in activity of the RAS signal transduction pathways in the pathogenesis of these tumours (Dammann et al., 2000; Agathanggelou et al., 2001). In contrast, RAS mutations are reported to be rare in Wilms' tumour (Waber et al., 1993), although WT1 has been suggested to suppress ras-mediated transformation in NIH3T3 cells (Luo et al., 1995). The finding of frequent RASSF1A inactivation in Wilms tumour offers possibilities for developing novel therapeutic interventions targeted at reversing RASSF1A silencing or the downstream consequences of RASSF1A inactivation. It is interesting to note that although the molecular genetics of Wilms' tumour, clear cell and papillary RCC differ considerably, epigenetic inactivation of RASSF1A is common to all three tumour types.
We detected RASSF1A methylation in 45% of primary colorectal cancers and 80% of colorectal cancer cell lines. For 36% of the primary colorectal cancers with RASSF1A methylation, methylation was also detected in the corresponding non-neoplastic colon tissue. Tissue-specific variations in gene methylation status are well recognized, and the frequency of gene methylation in normal colonic mucosa may increase with age (Issa et al., 1996; Ahuja et al., 1998; Pao et al., 2001). Although one explanation for RASSF1A methylation in normal tissue would be contamination by tumour cells, it has also been suggested gene methylation in normal tissues may be part of a premalignant ‘field change’. An intriguing possibility is that methylation of RASSF1A in colonic tissue might be a marker for tumour susceptibility but further studies (including analysis of colonic DNA samples from a group of age-matched controls with normal colonoscopy) would be required to investigate this. Recently, van Engeland et al. (2002) also reported frequent RASSF1A methylation (20%) in sporadic colorectal cancers and detected an inverse relationship between the presence of K-ras mutations and RASSF1A methylation in colorectal cancers.
Agathanggelou A, Honorio S, Macartney DP, Martinez A, Dallol A, Rader J, Fullwood P, Chauhan A, Walker R, Shaw JA, Hosoe S, Lerman MI, Minna JD, Maher ER, Latif F . 2001 Oncogene 20: 1509–1518
Ahuja N, Li Q, Mohan AL, Baylin SB, Issa JP . 1998 Cancer Res. 58: 5489–5494
Arcellana-Panlilio MY, Egeler RM, Ujack E, Pinto A, Demetrick DJ, Robbins SM, Coppes MJ . 2000 Gene Chromosome Cancer 29: 63–69
Astuti D, Agathanggelou A, Honorio S, Dallol A, Martinsson T, Kogner P, Cummins C, Neumann HP, Voutilainen R, Dahia P, Eng C, Maher ER, Latif F . 2001 Oncogene 20: 7573–7577
Bardeesy N, Falkoff D, Petruzzi MJ, Nowak N, Zabel B, Adam M, Aguiar MC, Grundy P, Shows T, Pelletier J . 1994 Nat. Genet. 7: 91–97
Baylin SB, Herman JG, Graff JR, Vertino PM, Issa JP . 1998 Adv. Cancer Res. 72: 141–196
Burbee DG, Forgacs E, Zöchbauer-Müller S, Shivakumar L, Gao B, Randle D, Virmani A, Bader S, Sekido Y, Latif F, Fong K, Gazdar AF, Lerman MI, White M, Minna JD . 2001 J. Natl. Cancer Inst. 93: 691–699
Costello JF, Plass C . 2001 J. Med. Genet. 38: 285–303
Cui H, Horon IL, Ohlsson R, Hamilton SR, Feinberg AP . 1998 Nat. Med. 4: 1276–1280
Dammann R, Li C, Yoon JH, Chin PL, Bates S, Pfeifer GP . 2000 Nat. Genet. 25: 315–319
Davies R, Moore A, Schedl A, Bratt E, Miyahawa K, Ladomery M, Miles C, Menke A, van Heyningen V, Hastie N . 1999 Cancer Res. 59: Suppl S1747–S1750 discussion S1751
Hao X, Willis IE, Pretlow TG, Rao JS, MacLennan GT, Talbot IC, Pretlow TR . 2000 Cancer Res. 60: 18–21
Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB . 1996 Proc. Natl. Acad. Sci. USA 93: 9821–9826
Herman JG, Latif F, Weng Y, Lerman MI, Zbar B, Liu S, Samid D, Duan DS, Gnarra JR, Linehan WM . 1994 Proc. Natl. Acad. Sci. USA 91: 9700–9704
Herman JG, Umar A, Polyak K, Graff JR, Ahuja N, Issa JP, Markowitz S, Willson JK, Hamilton SR, Kinzler KW, Kane MF, Kolodner RD, Vogelstein B, Kunkel TA, Baylin SB . 1998 Proc. Natl. Acad. Sci. USA 95: 6870–6875
Issa JP, Vertino PM, Boehm CD, Newsham IF, Baylin SB . 1996 Proc. Natl. Acad. Sci. USA 93: 11757–11762
Jones PA . 1999 Trends Genet. 15: 34–37
Kane MF, Loda M, Gaida GM, Lipman J, Mishra R, Goldman H, Jessup JM, Kolodner R . 1997 Cancer Res. 57: 808–811
Lee SB, Haber DA . 2001 Exp. Cell Res. 264: 74–99
Lerman MI, Minna JD . 2000 Cancer Res. 60: 6116–6133
Luo XN, Reddy JC, Yeyati PL, Idris AH, Hosono S, Haber DA, Licht JD, Atweh GF . 1995 Oncogene 11: 743–750
Morrissey C, Martinez A, Zatyka M, Agathanggelou A, Honorio S, Astuti D, Morgan NV, Moch H, Richards FM, Kishida T, Yao M, Schraml P, Latif F, Maher ER . 2001 Cancer Res. 61: 7277–7281
Moulton T, Crenshaw T, Hao Y, Moosikasuwan J, Lin N, Dembitzer F, Hensle T, Weiss L, McMorrow L, Loew T, Kraus W, Gerald W, Tycko B . 1994 Nat. Genet. 7: 440–447
Okamoto K, Morison IM, Taniguchi T, Reeve AE . 1997 Proc. Natl. Acad. Sci. USA 94: 5367–5371
Pao MM, Tsutsumi M, Liang G, Uzvolgyi E, Gonzales FA, Jones PA . 2001 Hum. Mol. Genet. 10: 903–910
Richards FM, Schofield PN, Fleming S, Maher ER . 1996 Hum. Mol. Genet. 5: 639–644
Steenman MJ, Rainier S, Dobry CJ, Grundy P, Horon IL, Feinberg AP . 1994 Nat. Genet. 7: 433–439
Stirzaker C, Millar DS, Paul CL, Warnecke PM, Harrison J, Vincent PC, Frommer M, Clark SJ . 1997 Cancer Res. 57: 2229–2237
Thiagalingam S, Lisitsyn NA, Hamaguchi M, Wigler MH, Willson JK, Markowitz SD, Leach FS, Kinzler KW, Vogelstein B . 1996 Cancer Res. 56: 2936–2939
Toyota M, Ohe-Toyota M, Ahuja N, Issa JPJ . 2000 Proc. Natl. Acad. Sci. USA 97: 710–715
Tycko B . 2000 J. Clin. Invest. 105: 401–407
van Engeland M, Roemen GM, Brink M, Pachen MM, Weijenberg MP, de Bruine AP, Arends JW, van den Brandt PA, de Goeij AF, Herman JG . 2002 Oncogene 21: 3792–3795
Vos MD, Ellis CA, Bell A, Birrer MJ, Clark GJ . 2000 J. Biol. Chem. 275: 35669–35672
Waber PG, Chen J, Nisen PD . 1993 Cancer 72: 3732–3738
We thank Carol Hardy for 11p LOH analysis and Dr FM Richards, Dr J Kingdom and Dr L Wong for human foetal RNA. We also thank the Association for International Cancer Research, Cancer Research Campaign and SPARKS for financial support.
About this article
Cite this article
Wagner, K., Cooper, W., Grundy, R. et al. Frequent RASSF1A tumour suppressor gene promoter methylation in Wilms' tumour and colorectal cancer. Oncogene 21, 7277–7282 (2002). https://doi.org/10.1038/sj.onc.1205922
- colorectal cancer
- Wilms' tumour
Immunohistochemical Profile of Tumor Suppressor Proteins RASSF1A and LATS1/2 in Relation to p73 and YAP Expression, of Human Inflammatory Bowel Disease and Normal Intestine
Pathology & Oncology Research (2020)
Potential of RASSF1A promoter methylation as a biomarker for colorectal cancer: Meta-analysis and TCGA analysis
Pathology - Research and Practice (2020)
Cancer Letters (2019)
Interrelations of Apoptotic and Cellular Senescence Genes Methylation in Inflammatory Bowel Disease Subtypes and Colorectal Carcinoma in Egyptians Patients
Applied Biochemistry and Biotechnology (2019)
LncRNAs with miRNAs in regulation of gastric, liver, and colorectal cancers: updates in recent years
Applied Microbiology and Biotechnology (2019)