Methylation changes in promoter and enhancer regions of the WT1 gene in Wilms’ tumours

Cancer Letters 166 (2001) 165±171

www.elsevier.com/locate/canlet

Methylation changes in promoter and enhancer regions of the WT1 gene in Wilms' tumours Jaroslav MaresÏ a,*, VõÂteÏzslav KrÏÂõ zÏ a, Andreas WeinhaÈusel b, SÏaÂrka VodicÏkova c, Roman Kodet d, Oskar A. Haas b, ZdeneÏk SedlaÂcÏek a, Petr Goetz a a

Institute of Biology and Medical Genetics, Second Medical School, Charles University, V uÂvalu 84, 15018 Prague 5, Czech Republic b Children's Cancer Research Institute, St. Anna Children's Hospital, Kinderspitalgasse 6, A-1090 Vienna, Austria c Department of Paediatric Oncology, University Hospital Motol, V uÂvalu 84, 15018 Prague, Czech Republic d Institute of Pathological Anatomy, Second Medical School, Charles University, V uÂvalu 84, 15018 Prague, Czech Republic Received 11 October 2000; received in revised form 10 January 2001; accepted 13 January 2001

Abstract Although the WT1 gene has been implicated in the aetiology of Wilms' tumour, mutations in WT1 are found only in minority of the tumours. DNA methylation of regulatory elements represents another possibility of modulation of gene expression. We studied methylation in the promoter and enhancer regions of the WT1 gene in 34 Wilms' tumour patients by the polymerase chain reaction on HpaII-digested DNA and by the bisulphite method. No methylation was detected in the promoter region in either tumour or normal kidney or blood DNA samples. In contrast, a HpaII site in the enhancer region was at least partially methylated in normal kidney and blood DNA samples and in about one-third of the tumours, while the majority of tumours showed no methylation. The differential methylation in the enhancer region of the WT1 gene may indicate that methylation of this element can play a role in the regulation of this gene. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: DNA methylation; WT1 gene; Promoter; Enhancer; Wilms' tumor

1. Introduction Childhood malignancies represent a small but very speci®c subgroup of human tumours. Because of their early onset, special mechanisms are thought to be responsible for their initiation and progression. If understood in more detail, these mechanisms can reveal important clues about general processes of cancerogenesis. Wilms' tumour, a neoplasm of the kidney, is one of the most common solid tumours of childhood, affect* Corresponding author. Tel.: 1420-2-2443-5991; fax: 1420-22443-5994. E-mail address: [email protected] (J. MaresÏ).

ing approximately 1 in 10 000 children during the ®rst years of life. The tumour is caused by deviations from the normal pattern of cell differentiation during early development of the kidney [1]. The tumour suppressor gene WT1 in 11p13 [2,3] and several other loci have been implicated in the aetiology of the malignancy in both its sporadic and familial forms [1]. WT1 encodes a zinc-®nger DNA-binding protein which most likely acts as a transcriptional regulator, but other additional roles including one in RNA splicing have recently been proposed [4,5]. Unexpectedly, mutations in the coding region of the WT1 gene were identi®ed in only about 10±20% of Wilms' tumours [1,5,6] and no nucleotide alterations were detected in the promoter of the gene [7]. DNA

0304-3835/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(01)00402-5

166

J. MaresÏ et al. / Cancer Letters 166 (2001) 165±171

methylation can be another possible mechanism which may participate in the modulation of gene expression. The aim of our study was to elucidate whether methylation of the WT1 gene plays a role in the aetiology of Wilms' tumours. The link between DNA methylation of CpG dinucleotides and cancer is now well established. In general, ®ve different phenomena have been described. (i) Due to frequent transition from methylated cytosine to thymine, there is an increased mutation rate at CpG sites, including mutations in cancer genes. (ii) Overall genome hypomethylation is observed in some tumours and hypomethylation of speci®c protooncogenes may lead to their enhanced expression. (iii) Hypermethylation of some tumour suppressor genes (e.g. RB1, p16 or VHL) may in some cases be the mechanism of their inactivation. (iv) Methylation may be associated with genomic instability on the chromosome level. (v) In imprinted genes involved in cancer, aberrant methylation may lead to disturbances in their parental expression [8±10]. The strong link between DNA methylation and cancer is not only interesting from the point of view of basic research, but may also lead to novel therapies based on modulation of the methylation/demethylation process. Aberrant genomic imprinting in the 11p15 region was associated with the development of Wilms' tumour [1], and parental imprinting in 11p13 may also play a role in the expression of the WT1 gene [11,12]. The regulatory elements of the WT1 gene are well known [13,14]. We have therefore focused our study of DNA methylation changes in a collection of Wilms' tumours to the promoter and enhancer regions of the WT1 gene. We show that while no methylation differences can be detected in several CpG sites within the promoter region, one CpG site within the enhancer region is differentially methylated. We could not, however, ®nd any signi®cant correlation between the WT1 enhancer methylation and clinical parameters of the tumours.

2. Materials and methods 2.1. Clinical samples Thirty-four patients with Wilms' tumour were

studied after obtaining the informed consent from their parents. The study was approved by the local ethical committee. All patients were treated according to the current SIOP (Societe Internationale d'Oncologie PeÂdiatrique) protocol including 4 weeks of preoperative chemotherapy. Tumours were classi®ed according to the Stockholm criteria. Samples were frozen in liquid nitrogen and DNA was extracted using standard protocols with proteinase K digestion and phenol extraction. Normal kidney samples from six patients and blood from all patients were also processed as above. 2.2. Polymerase chain reaction (PCR) after digestion with methylsensitive enzyme HpaII Two hundred nanograms of genomic DNA from the tumour from each patient as well as the normal kidney and blood samples were digested with 50 units of methylsensitive restriction endonuclease HpaII for 16 h and then for additional 2 h with 20 units of fresh enzyme. The digests were precipitated, redissolved and used as templates for PCR. Uncut DNAs processed in the same way were always ampli®ed as controls. Methylation of the HpaII sites (CCGG) within the ampli®ed segment protects the DNA from cleavage and the DNA can be ampli®ed. Absence of methylation allows cleavage and the cleaved DNA cannot serve as a template for PCR. The promoter region (Fig. 1A) was ampli®ed in an ampli®cation reaction with 25 ng of DNA, 15 pmol of each primer, 250 mM of each dNTP and 0.6 units of Taq polymerase in a total volume of 25 ml in 1 £ PCR buffer for 36 cycles at 948C, 608C and 728C for 1 min each, preceded by extended denaturation at 968C for 5 min and followed by ®nal extension at 728C for 5 min. The primers used were TATAACTGGTGCAACTCCCGGC and AAAGAGTGGTTTGGAGGGAGGG for the distal promoter region (ampli®cation product of 265 bp) [7] and CAACCCCATCTCTACTCCCA and CGCTGCCTTGAACTCCTTAC for the proximal promoter region (391 bp product). The enhancer region (Fig. 1B) was ampli®ed under the same conditions except that 30 cycles of 948C for 30 s, 588C for 30 s and 728C for 40 s were used with primers TTCTGGTGTATGGTTTTTGA and CTTCTCCAGAGAAATGCTGT (348 bp product). Ten

J. MaresÏ et al. / Cancer Letters 166 (2001) 165±171

167

Fig. 1. The location and sequence of WT1 promoter and enhancer regions. (A) Sequence of the 5 0 ¯anking region, promoter and exon 1 of the WT1 gene. The translation of the 5 0 end of the coding region is shown under the sequence. The sequence of intron 1 is in lowercase. The 104 bp minimal promoter of Fraizer et al. [14] is underlined. Possible transcription start sites compiled from Refs. [13,14,21] are marked with §. The primers used to amplify the distal and proximal parts of the promoter region are indicated. HpaII sites within the ampli®ed fragments are boxed. Primers (WT-c, WT-d) enabling ampli®cation of methylated and deaminated DNA by methylation-sensitive PCR are shown in italics. (B) Sequence of the end of exon 10 and 3 0 ¯anking region of the WT1 gene with the enhancer. The poly(A) signal is marked with #, the poly(A) addition site with a poly(A) stretch. The 250 bp enhancer of Fraizer et al. [14] is underlined. Primers and HpaII sites are indicated as in (A).

microlitres of the PCR were examined on 2% agarose gels stained with ethidium bromide or Sybr Green. 2.3. Methylation-speci®c PCR based on bisulphite deamination All samples were analyzed for methylation of the 5 0 region of the WT1 gene using bisulphite deamination-

based methylation-sensitive PCR [15]. Genomic DNA from one normal individual was methylated in vitro with SssI-DNA-methylase (NEB) and used as a positive control. All DNA samples were deaminated with sodium bisulphite. Sequences of primers WT-c TTAGAGTCGGGACGGTAGTTTAGGC and WT-d CTAAAAACGTTCAACGCTAACCTCG (Fig. 1A) were deduced from genomic DNA sequence based

168

J. MaresÏ et al. / Cancer Letters 166 (2001) 165±171

on the assumption that cytosins are only methylated in CpG dinucleotides and, after deamination, all unmethylated cytosins are converted into uracil, whereas methylated cytosins are not. The PCR reactions were performed on deaminated DNA in 25 ml of 1 £ reaction buffer, 200 mM of each dNTP, 6.25 pmol of both primers WT-c and WT-d and 1 unit of Taq polymerase. After initial denaturation for 5 min at 958C the ampli®cation was performed for 35 cycles of 958C for 25 s, 608C for 25 s and 708C for 40 s, and a ®nal extension for 7 min at 728C. Only methylated DNA yields a speci®c PCR product of 233 bp which can be sequenced to analyze also the methylation of the internal CpG sites. 3. Results 3.1. Collection of Wilms' tumour patient samples The group of 34 Wilms' tumour patients included 15 boys and 19 girls with age at diagnosis ranging from 6 months to 17 years. The clinical stage ranged from 1 to 5 and the grade of the tumours ranged from I to III (Table 1). Genomic DNA from both Wilms' tumour and peripheral blood lymphocytes, as well as the clinical and histological data, were obtained from all 34 patients. 3.2. Methylation of the promoter region of the WT1 gene All samples of Wilms' tumour DNA showed absence of methylation of HpaII sites in both distal and proximal promoter regions, as re¯ected by the absence of PCR products on HpaII-digested tumour DNA. All six samples of normal kidney DNA and blood DNA from all patients also showed absence of methylation in both regions examined, indicating no inter-individual variability in the methylation status. The control DNA not digested with HpaII always yielded a PCR band, as expected. The absence of methylation of the 5 0 region of the WT1 gene was con®rmed in all samples by the bisulphite deamination-based methylation-speci®c PCR. While the in vitro methylated positive control DNA (after deamination and ampli®cation with primers speci®c for methylated DNA) yielded the expected 233-bp-long PCR product, there was no ampli®cation in any of

the patient samples tested (data not shown). This indicates that the DNA samples were not methylated in the primer binding sites and did not allow the annealing of the primers speci®c for methylated sequences. 3.3. Methylation in the enhancer region of the WT1 gene In contrast to the 5 0 region, the single HpaII site analyzed in the 3 0 enhancer region showed variable methylation between different tumours. Eleven out of the 34 samples yielded a PCR band on HpaII-digested tumour DNA indicating at least partial methylation of the HpaII site, while the remaining 23 samples showed absence of methylation (Table 1 and Fig. 2). A band was also always present in PCR reactions on HpaII-digested normal kidney and blood DNA, indicating at least partial methylation of this site in normal kidney samples and in lymphocytes of all patients. A PCR product was also observed in all undigested control samples. 3.4. Correlation between methylation and clinical parameters of the tumours There was no signi®cant correlation between the methylation differences observed in the enhancer region of the WT1 gene and the clinical parameters including sex, age of onset and tumour stage and grade of the respective Wilms' tumour patients (Table 1). 4. Discussion The WT1 gene is an outstanding member of the group of tumour suppressor genes. In contrast to the majority of these genes represented, e.g., by p53 or RB1, the expression of WT1 is developmentally regulated and limited only to a subset of cell types, mainly in the developing kidney and the urogenital tract. In addition to that, mutations in WT1 were identi®ed in only a small fraction (10±20%) of Wilms' tumours, often just on one allele [1,5,6]. On the other hand, WT1 gene mutations also contribute to several urogenital anomaly syndromes and the expression of the gene is involved in other cancers including leukaemias and breast cancer. Some of the phenotype effects

J. MaresÏ et al. / Cancer Letters 166 (2001) 165±171

169

Table 1 Clinical and histological parameters of the Wilms' tumours examined and methylation status of the tumour DNA a Patient no.

Sex

Age at diagnosis

Clinical stage

Histological grade

Distal promoter methylation

Proximal promoter methylation

Enhancer methylation

1 4 9 23 32 36 47 60 61 77 103 122 138 152 153 186 200 219 287 201 295 329 330 331 333 352 364 413 429 432 433 469 471 479

F M M M F M M M F F F F F M F F F F F F M F F M M F M M F F F M M M

12 y. 2 y. 3 y. 1 y. 0 y. 1 y. 2 y. 3 y. 4 y. 7 y. 17 y. 4 y. 4 y. 1 y. 5 y. 4 y. 1 y. 6 y. 4 y. 2 y. 4 y. 4 y. 0 y. 2 y. 2 y. 3 y. 3 y. 2 y. 9 y. 2 y. 9 y. 17 y. 5 y. 0 y.

4 1 3 1 1 1 5 4 1 1 1 4 1 1 1 1 1 5 5 1 1 1 1 1 1 1 4 1 1 1 4 3 1 1

II II III II II II II II II II II II II II II III II II II II II II II II II II II II II II II II II I

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

2 2 2 1 2 2 2 2 2 2 2 2 2 2 2 1 1 2 2 2 1 1 2 2 1 2 2 1 1 1 2 2 1 1

7 mo. 10 mo. 4 mo. 5 mo. 8 mo. 2 mo. 11 mo. 9 mo. 4 mo. 2 mo. 2 mo. 6 mo. 6 mo. 5 mo. 7 mo. 6 mo. 3 mo. 7 mo. 9 mo. 10 mo. 5 mo. 10 mo. 6 mo. 2 mo. 4 mo. 3 mo. 6 mo. 7 mo. 2 mo. 11 mo. 6 mo.

a y, years; mo., months. 1 marks the presence of a band in PCR on HpaII-digested template, which indicates methylation of at least of a fraction of the genomic DNA analyzed. Within the distal and proximal promoter, all normal kidney and blood samples were not methylated, while within the enhancer all showed methylation.

may be based on aberrant dose or aberrant ratios of different WT1 splicing isoforms [5]. The WT1 gene is therefore a good candidate for being modulated by an epigenetic mechanism like DNA methylation. The methylation of the WT1 gene region in Wilms' tumours was analyzed previously using methylationsensitive rare cutter enzymes and pulse-®eld gel electrophoresis [16]. Although the method is not fully comparable with the method used in our study, both yielded similar results. The rare cutter enzyme sites in

the CpG island in the 5 0 end of the WT1 gene were unmethylated both in blood and in most (27 out of 29) samples of Wilms' tumour DNA, which is in accord with our data generated by the analysis of several HpaII sites in this region. Rare cutter sites downstream of the WT1 gene were also analyzed in the experiment, but these were located about 50 kb more distal to the enhancer region studied by us. These distal sites were differentially methylated, with the highest methylation identi®ed in tumours from patients who had not under-

170

J. MaresÏ et al. / Cancer Letters 166 (2001) 165±171

Fig. 2. Ampli®cation of the WT1 enhancer region after HpaII digestion. Analyses of Wilms' tumour and blood DNA samples from two patients. The 348-bp-long PCR product is generated with primers indicated in Fig. 1B. Absence of the PCR product from the HpaIIdigested tumour DNA in the ®rst patient indicates demethylation of the HpaII site within the enhancer, which abolishes the protection of the DNA against the HpaII cleavage.

gone preoperative chemotherapy. There was no correlation, however, between the DNA methylation and WT1 expression [16]. Methylation of the WT1 tumour suppressor gene was also studied in other types of cancer. Sequencing of bisulphite-modi®ed DNA was used to study cytosine methylation of the WT1 promoter in colon adenomas and carcinomas [17]. Normal mucosa was unmethylated on all cytosines, but the majority of tumour tissues showed increased methylation. The study pointed out, however, that the methylation was site-speci®c and varied signi®cantly across the promoter region, with the Sp1 sites (CCCGCCC) being signi®cantly less methylated. The methylation status did not have any in¯uence on the level of mRNA expression in the tumours and the authors suggested that the WT1 gene might not be regulated by promoter methylation [17]. In another study, methylation of the WT1 gene in malignant mesothelioma was analyzed in the rat model [18]. Here speci®c methylation of one site in intron 1 of the rat wt1 gene was observed in majority of tumours, but not in normal rat tissues. The HpaII sites in the 5 0 region of the gene showed no methylation, and, again, there was no correlation between the methylation of

this site and wt1 gene expression, even when the methylation was experimentally removed [18]. A recent study of breast cancer has found extensive methylation of the CpG island of the WT1 gene in two cell lines and 5 out of 20 primary tumours [19]. In the cell lines methylation was associated with WT1 silencing which could be removed by experimental demethylation. Methylation in one of the cell lines was accompanied by a shorter WT1 transcript. The expression in tumours was not examined [19]. Generally, the analysis based on PCR on HpaIIdigested DNA may be negatively in¯uenced by contamination of the tumour material with normal cells. Our results show that the normal kidney and blood samples produce no PCR band which indicates they are unmethylated in the sites tested in the WT1 promoter region. Normal cells are therefore unlikely to distort the analysis of the tumour DNAs, no matter if these are methylated or not. The normal cells produce ampli®cation in the enhancer region which must therefore be methylated. Contamination of the tumour sample with these cells would lead to a PCR product and a possible false conclusion that the tumour DNA is methylated as well. Here, however, 23 out of 34 tumour DNAs produced no ampli®cation and must be unmethylated. Even if the remaining 11 tumours showed a band due to contamination with normal cells, it would still mean all tumours are unmethylated in the enhancer region, in contrast to normal kidney. Because of the lack of correlation between the WT1 methylation pattern and clinical status of the tumours in the published reports and in our study, the role of differential methylation of the WT1 gene in cancerogenesis remains unclear. The elucidation of the role of DNA methylation in tumour suppressor gene inactivation is further complicated by the fact that the methylation-based mechanism of gene silencing may be limited only to a small subset of tumours of the particular type, as in the case of the RB1 gene methylation pattern observed in retinoblastoma [20]. The identi®cation of differential methylation patterns in the enhancer region of the WT1 gene in Wilms' tumours suggests that this regulatory element may play a more important role in possible methylation-based control of the WT1 gene than the promoter region. This is in accord with the lack of correlation between WT1 promoter methylation and altered gene expression in the majority of published reports [16±

J. MaresÏ et al. / Cancer Letters 166 (2001) 165±171

18]. It is also in accord with the observation that the WT1 promoter is similar to housekeeping gene promoters and functioning in many different cell types [14]. The tissue speci®city of the WT1 gene is therefore controlled by other elements, one of them being the enhancer [7,14]. It remains to be determined whether or not there is a correlation between the methylation status of the enhancer region and WT1 expression in Wilms' tumours. Acknowledgements We thank Nina Heiss for critical reading of the manuscript. This work was supported by Grant No. 4348-3 from the Internal Grant Agency of the Czech Ministry of Health and by private donations to the Children's Cancer Research Institute in Vienna.

[9] [10] [11] [12]

[13]

[14]

[15]

References [1] V. Huff, Wilms' tumor genetics, Am. J. Med. Genet. 79 (1998) 260±267. [2] M. Gessler, A. Poustka, W. Cavenee, R.L. Neve, S.H. Orkin, G.A. Bruns, Homozygous deletion in Wilms tumours of a zinc-®nger gene identi®ed by chromosome jumping, Nature 343 (1990) 774±778. [3] K.M. Call, T. Glaser, C.Y. Ito, A.J. Buckler, J. Pelletier, D.A. Haber, E.A. Rose, A. Kral, H. Yeger, W.H. Lewis, C. Jores, D.E. Hausman, Isolation and characterization of a zinc ®nger polypeptide gene at the human chromosome 11 Wilms' tumor locus, Cell 60 (1990) 509±520. [4] R. Davies, A. Moore, A. Schedl, E. Bratt, K. Miyahawa, M. Ladomery, C. Miles, A. Menke, V. van Heyningen, N. Hastie, Multiple roles for the Wilms' tumor suppressor WT1, Cancer Res 59 (1999) 1747s±1750s. [5] M. Little, G. Holmes, P. Walsh, WT1: what has the last decade told us? Bioessays 21 (1999) 191±202. [6] M. Gessler, A. Konig, K. Arden, P. Grundy, S. Orkin, S. Sallan, C. Peters, S. Ruyle, J. Mandell, F. Li, W. Cavenee, G. Bruns, Infrequent mutation of the WT1 gene in 77 Wilms' tumors, Hum. Mutat. 3 (1994) 212±222. [7] G.R. Grubb, K. Yun, A.E. Reeve, M.R. Eccles, Exclusion of the Wilms' tumour gene (WT1) promoter as a site of frequent mutation in Wilms' tumour, Oncogene 10 (1995) 1677± 1681. [8] P.A. Jones, M.L. Gonzalgo, Altered DNA methylation and

[16] [17]

[18] [19]

[20]

[21]

171

genome instability: a new pathway to cancer? Proc. Natl. Acad. Sci. USA 94 (1997) 2103±2105. C. Schmutte, P.A. Jones, Involvement of DNA methylation in human carcinogenesis, Biol. Chem. 379 (1998) 377±388. J.G. Herman, Hypermethylation of tumor suppressor genes in cancer, Semin. Cancer Biol. 9 (1999) 359±367. Y. Jinno, K. Yun, K. Nishiwaki, T. Kubota, O. Ogawa, A.E. Reeve, N. Niikawa, Mosaic and polymorphic imprinting of the WT1 gene in humans, Nat. Genet. 6 (1994) 305±309. K. Mitsuya, H. Sui, M. Meguro, H. Kugoh, Y. Jinno, N. Niikawa, M. Oshimura, Paternal expression of WT1 in human ®broblasts and lymphocytes, Hum. Mol. Genet. 6 (1997) 2243±2246. W. Hofmann, H.D. Royer, M. Drechsler, S. Schneider, B. Royer-Pokora, Characterization of the transcriptional regulatory region of the human WT1 gene, Oncogene 8 (1993) 3123±3132. G.C. Fraizer, Y.J. Wu, S.M. Hewitt, T. Maity, C.C. Ton, V. Huff, G.F. Saunders, Transcriptional regulation of the human Wilms' tumor gene (WT1): Cell type-speci®c enhancer and promiscuous promoter, J. Biol. Chem. 269 (1994) 8892±8900. M. Zeschnigk, C. Lich, K. Buiting, W. Doer¯er, B. Horsthemke, A single-tube PCR test for the diagnosis of Angelman and Prader-Willi syndrome based on allelic methylation differences at the SNRPN locus, Eur. J. Hum. Genet. 5 (1997) 94±98. B. Royer-Pokora, S. Schneider, Wilms' tumor-speci®c methylation pattern in 11p13 detected by PFGE, Genes Chromosomes Cancer 5 (1992) 132±140. M.O. Hiltunen, J. Koistinaho, L. Alhonen, S. Myohanen, S. Marin, V.M. Kosma, M. Paakkonen, J. Janne, Hypermethylation of the WT1 and calcitonin gene promoter regions at chromosome 11p in human colorectal cancer, Br. J. Cancer 76 (1997) 1124±1130. E.V. Kleymenova, X. Yuan, M.E. LaBate, C.L. Walker, Identi®cation of a tumor-speci®c methylation site in the Wilms tumor suppressor gene, Oncogene 16 (1998) 713±720. D.E. Laux, E.M. Curran, W.V. Welshons, D.B. Lubahn, T.H. Huang, Hypermethylation of the Wilms' tumor suppressor gene CpG island in human breast carcinomas, Breast Cancer Res. Treat. 56 (1999) 35±43. N. Ohtani-Fujita, T.P. Dryja, J.M. Rapaport, T. Fujita, S. Matsumura, K. Ozasa, Y. Watanabe, K. Hayashi, K. Maeda, S. Kinoshita, T. Matsumura, Y. Ohnishi, Y. Holta, R. Takahashi, M.V. Kato, K. Ishizoki, M.S. Sasaki, B. Horsthemke, K. Minoda, T. Sakai, Hypermethylation in the retinoblastoma gene is associated with unilateral sporadic retinoblastoma, Cancer Genet. Cytogenet. 98 (1997) 43±49. C.E. Campbell, A. Huang, A.L. Gurney, P.M. Kessler, J.A. Hewitt, B.R. Williams, Antisense transcripts and protein binding motifs within the Wilms' tumour (WT1) locus, Oncogene 9 (1994) 583±595.