RB1 and CDKN2A Functional Defects Resulting in Retinoblastoma O

Molecular Biology, Vol. 36, No. 5, 2002, pp. 625Ð630. Translated from Molekulyarnaya Biologiya, Vol. 36, No. 5, 2002, pp. 777Ð783. Original Russian Text Copyright © 2002 by Babenko, Zemlyakova, Saakyan, Brovkina, Strelnikov, Zaletaev, Nemtsova.

GENOMICS. TRANSCRIPTOMICS. PROTEOMICS UDC 577.218 RB1 and CDKN2A Functional Defects Resulting in Retinoblastoma O. V. Babenko1, V. V. Zemlyakova1, S. V. Saakyan2, A. F. Brovkina2, V. V. Strelnikov1, D. V. Zaletaev1, and M. V. Nemtsova1 1 Medical Genetic Research Center, Russian Academy of Medical Sciences, Moscow, 115478 Russia: E-mail: [email protected] 2 Moscow Institute of Eye Diseases, Ministry of Health of the Russian Federation, Moscow, 103064 Russia Received October 25, 2001

Abstract—Multiplex methylation-sensitive PCR was employed in studying the methylation of the RB1 and CDKN2A/p16 promoter regions in 52 retinoblastomas. Aberrant methylation inactivating RB1 was detected in 14 (27%) tumors. Methylation of p16 was for the ﬁrst time observed in retinoblastoma (9 tumors, 17%). Both promoters proved to be methylated in two tumors. In four tumors, aberrant methylation was combined with structural defects of both RB1 alleles. Aberrant methylation of the p16 promoter was the second mutation event in two tumors and was not accompanied by RB1 defects in one tumor. Complex testing for RB1 mutations, loss of heterozygosity, and functional inactivation of the two genes revealed molecular defects in at least one allele in 51 (98%) tumors.

Key words: retinoblastoma, RB1, CDKN2A/p16, promoter region, CpG methylation, methylation-sensitive PCR

INTRODUCTION tural organization of chromatin and in regulation of the gene functional activity [4]. The role of genetic factors in carcinogenesis in now beyond doubt. As shown by numerous studies, By now, ample data have been accumulated on the malignant cells arise upon structural alterations of role of epigenetic events, including disturbed DNA certain genes. Increasing importance is also attached methylation, in carcinogenesis [1, 5]. Methylation dis- to epigenetic gene regulation, whereby the expression balance has been observed in virtually all types of of a gene is altered without changes in its nucleotide neoplastic cells. The average genome methylation is sequence [1]. An example is DNA methylation, i.e., low (hypomethylated are single CpG dinucleotides enzymatic addition of the methyl group to C5 of the scattered through the genome), while hypermethyla- pyrimidine ring in cytosine. Only cytosines in CpG tion occurs in CpG islands located in the promoter dinucleotides are thus methylated. The modification is regions of several genes involved in controlling the stable and heritable, although demethylating agents or cell cycle, differentiation, and apoptosis [4, 6]. In car- enzymes may cause reversion. cinogenesis, substantial changes in DNA methylation are maintained through tumor cell generations and, DNA methylation plays an important role in regu- what is more, progressively extend to other genes. lating the mammalian genome and underlies such bio- This problem has received increasing attention in logical phenomena as inactivation of the X chromo- recent years. Methylation of CpG islands in the pro- some, monoallelic expression in imprinting, and sup- moter region and subsequent gene inactivation have pression of foreign DNA (e.g., methylation of been demonstrated for numerous tumor suppressor, transposons suppresses their transcriptional activity metastasis-associated, and DNA repair genes [7]. In and protects the host cell) [1, 3]. Methylation and certain tumors, aberrant methylation involves several demethylation alternate in a regular fashion during genes to produce a certain methylation profile, which may serve as a marker of malignancy and may corre- embryo development, resulting in a strictly specified late with disease severity, rapid metastatic spread, or methylation profile of individual genes and the whole with resistance to chemo- and radiotherapy [1, 8]. genome. Normally, the profile is maintained stable through generations of a given somatic cell lineage. In The contribution of DNA methylation to carcino- the mature organism, methylation substantially affects genesis is not restricted to the epigenetic effect. Being DNAÐprotein interactions without changing the unstable, 5-methylcytosine may be spontaneously genetic information, and thereby participates in struc- deaminated to yield thymine (about 30% point muta-

626 BABENKO et al. tions), which is unrecognizable for repair systems. RESULTS AND DISCUSSION The resulting ë í transition may change the struc- Detection of Aberrant Methylation ture and function of the corresponding protein [9]. in the RB1 Promoter Region As a process inactivating tumor suppressor genes, Along with structural mutations and loss of het- methylation has first been described for RB1 [10]. erozygosity, methylation of the promoter region and Methylation of the RB1 promoter is involved in subsequent gene inactivation initiate carcinogenesis. Epigenetic alterations have been found to inactivate retinoblastoma development [10Ð12]. The normal various genes in tumor cells. For example, aberrant function of RB1 is essential for the cell-cycle control. methylation is the major inactivation mechanism of However, rather than acting alone, RB1 is functionally p16 in lung carcinoma and in various leukemias. The associated with other suppressor genes and protoon- gene for E-cadherin, which is involved in cell-to-cell cogenes, which code for modifiers of the RB1 activity. contacts, is methylated in mammary and bladder carcinomas. MLH1, which participates in repair of Nuclear protein p16/CDKN2A inhibits cyclin- unpaired DNA, is methylated in almost all colorectal dependent kinases (CDK4/6) and prevents their com- carcinomas associated with microsatellite instability. plexation with cyclins D, which phosphorylate and The estrogen receptor gene (ER) is aberrantly methy- thereby inactivate RB1 to release transcription factor lated in 20Ð30% mammary carcinomas and in 60Ð E2F from its complex with RB1. As a result, E2F is 70% acute leukemias. The promoter region of the activated, transcription of various cell genes transcription factor WT1 gene is methylated in 90% enhanced, and the cell proceeds to the S phase. Inhibi- mammary carcinomas, 50% colorectal carcinomas, tion of cyclin-dependent kinases by p16 is the best- and in 10% Wilms tumors [17]. studied mechanism controlling the RB1 activity in the Methylation of the RB1 promoter leads to retino- cell [13]. blastoma [12, 18, 19]. We tested 52 tumors for this alteration by means of MS-PCR. The gist of the tech- In this work, we studied the methylation of the RB1 nique is selective hydrolysis of DNA sites containing and p16/CDKN2A promoter regions in 52 retinoblas- 5-methylcytosine with methylation-sensitive restric- toma specimens, and determined the spectrum of tion enzymes. DNA regions devoid of modified molecular defects leading to tumorigenesis. cytosines are cleaved, and subsequent PCR yields no product. When 5-methylcytosine is in the recognition site of a restriction enzyme, the region is not cleaved, EXPERIMENTAL and a PCR product of a certain size can be detected in the gel. The method is highly sensitive and allows Tumor specimens were obtained from patients detection of rare methylated alleles in the presence of with various forms of retinoblastoma, who were sub- excess wild-type alleles. The GC-rich RB1 promoter jected to surgery in the Oncoophthalmological fragment under study contains four HpaII (CCGG) Department, Moscow Institute of Eye Diseases. and four HhaI (CGCG) sites. MS-PCR revealed RB1 Genomic DNA was isolated from the surgery material promoter methylation in 14 (27%) tumors. Methyla- and from peripheral blood lymphocytes according to tion was combined with loss of heterozygosity in ten the standard protocol [14]. tumors and with RB1 mutations in three tumors. No RB1 defects other than methylation were found in one Methylation of CG-rich promoter regions of RB1 tumor. and p16 was probed by methylation-sensitive PCR The RB1 promoter region has 27 CpG dinucle- (MS-PCR). Genomic DNA (1 µg) was digested with otides, which can be methylated throughout the CpG 10 units of HpaII in 10 µl of the reaction mixture overnight. island to produce a specimen-specific pattern [19]. The product (150 ng) was amplified in multiplex PCR [15] The number of methylated cytosines is not associated with primers prRBF (CTGGACCCACGCCAGGTTTC), with the extent of gene inactivation [20]. Methylation prRBR (ATTGGTACCCGACTCCCGTTACAAAAT), mostly occurs in sporadic retinoblastoma (10Ð15% pr16F (AGCCAGCCCCTCCTCTTTCTTC), and pr16R tumors) [21] and is less frequent (6Ð9%) in congenital retinoblastoma [18, 22]. In addition to aberrant meth- (GAACGCACTCAAACACGC). For a positive internal ylation of the promoter, methylated cytosines have control of amplification, multiplex PCR was run been found in CGA codons of exons 8 and 14. The with primers Ext2-8F (TCTAGTTTTCCCACTCT- resulting ë í transition yields the TGA stop GTCTC) and Ext2-8R (TTCCTTCCACCCACCCT- codon and leads to premature termination of RB1 syn- GAC), which are directed to EXT2 exon 8 lacking the thesis [20]. This mechanism has not been observed for HpaII site. The amplification product was resolved in other RB1 regions containing CGA codons, suggest- 8% PAG and silver stained [16] or in 2% agarose gel ing that methylation of individual CpG in exons and stained with ethidium bromide. makes only a minor contribution to the RB1 inactiva-

MOLECULAR BIOLOGY Vol. 36 No. 5 2002

GENE FUNCTIONAL DEFECTS IN RETINOBLASTOMA 627 tion. The extent of RB1 promoter methylation in our tumor sample was 27%, far higher than considered bp before. This may be explained by the more sensitive 472 Internal method used to detect aberrant methylation or by control (Ext2-8) some specific features of our patients. 349 pr16/CDKN2A Analyses of mutations, loss of heterozygosity, and methylation revealed the causes of inactivation for 249 prRB1 both RB1 alleles in 46 tumors (table). Three tumors had one mutation each. The remaining three showed 12345678 no RB1 molecular defects. Since genes are known to cooperate in carcinogenesis, we assumed a contribu- Methylation of the RB1 and p16 promoters as revealed by tion of other genes to retinoblastoma development. MS-PCR with HpaII. The patterns indicate (1) methylation of the p16 promoter, (2, 5, 7) methylation of the RB1 promoter, (3, 6, 8) no methylation, and (4) methylation of both Detection of Aberrant Methylation promoters. in the p16/CDKN2A Promoter Region Taking account of the effect of p16 on the RB1 activity, we studied p16 methylation in the 52 tumors. mechanisms of cell malignization even in the case of The p16 promoter region contains three HpaII sites. monogenic disorders, such as retinoblastoma. To simultaneously assess the methylation of p16 and RB1, we elaborated a multiplex MS-PCR protocol Aberrant p16 methylation combined with inactiva- (figure). tion of both RB1 alleles is of special interest. Since the two genes affect one another, inactivation of either Aberrant methylation of the p16 promoter region may be sufficient for carcinogenesis. The p16 pro- was detected in 9 (17%) tumors (table). In six tumors, moter might be methylated as a result of cell methylation of the p16 promoter was combined with malignization. Yet it is difficult to establish the cause defects of both RB1 alleles. The defects were RB1 and the effect, because RB1 and p16 are involved in methylation in two cases, a mutation and loss of het- one mechanism of cell-cycle control. erozygosity in three cases, and two mutations in one case. Thus, p16 methylation was the third mutation Thus, RB1 is inactivated by structural (mutations, events in these six tumors. In two tumors, aberrant loss of heterozygosity) or functional (promoter meth- methylation of the p16 promoter was the second event ylation) alterations. Aberrant methylation of the p16 combined with loss of heterozygosity at intragene promoter might also contribute to retinoblastoma. RB1 markers or with a mutation in exon 15. Only p16 Methylation of tumor suppressor genes occurs methylation was observed in one tumor. early in the genome of the neoplastic cell and may be As already noted, p16 modulates the activity of detected before clinical manifestation of the disease. RB1 [23]. Simultaneous inactivation of RB1 and p16 For instance, p16 is methylated in hyperplastic bron- has been shown to play an important role in prostate chial epithelium; MLH1, in colorectal adenomas; etc. carcinoma [24]. Both genes affect the telomerase [13, 26]. Methylation of these genes has also been activity, which is essential for tumor cell immortaliza- observed in most lung and colorectal carcinomas. tion [25]. Thus, epigenetic factors have been experimentally shown to play a role in hereditary disorders and in We have no grounds to state that aberrant methyla- cancer, suggesting a novel type of molecular defects. tion of the p16 promoter is a cause of retinoblastoma, Aberrant methylation/demethylation may serve as a while p16 inactivation does contribute to tumorigene- diagnostic or prognostic marker and allow differenti- sis. It is noteworthy that this defect alone, without ation of certain tumors. Gene methylation profiling of changes in RB1, was observed in one sporadic retino- various tissues and tumors will provide further insight blastoma. Cell malignant transformation is multistep into the regulation of cell proliferation, differentia- and involves accumulation of mutations along with tion, and malignization. other structural or functional alterations of the genome. Inactivation of tumor suppressor genes and To summarize, we identified the defects leading to subsequent distortion of regulatory mechanisms are of retinoblastoma in 42 (81%) patients. A defect of at particular importance. Since RB1 is functionally asso- least one allele of RB1 or p16 was revealed in 98% ciated not only with p16 but also with other tumor patients. The RB1 promoter proved to be methylated suppressor genes and protooncogenes coding for in about one-third of the tumors. Still, there were a modulators of RB1 activity, their inactivation may few tumors without structural and functional defects also play a role in retinoblastoma. Hence our results of RB1. Complex interactions of various genes and suggest that complex analysis of the methylation pro- their alterations in neoplastic cells make it possible to file of tumor suppressor genes is required to study the assume that other, still unknown mechanisms also

MOLECULAR BIOLOGY Vol. 36 No. 5 2002 628 BABENKO et al.

Spectrum of molecular defects in retinoblastoma CDKN2A Tumor RB1 region Defect Nucleotide sequence Retinoblastoma type methylation T2 Exon 13 Nonsense mutation G1267T Ð Sporadic unilateral LOH … T3 Exon 20 Missense mutation C1982T Ð Familial unilateral LOH … T4 Exon 11 2-bp deletion 1061del(ga) Ð Sporadic unilateral LOH T5 Exon 16 Nonsense mutation C1466A Ð " Intron 9 Splicing site defect 945-3T G T6 Exon 4 27-bp insertion 490ins27 Ð " Exon 15 10-bp insertion 1424ins9 T7 Exon 15 2-bp deletion 1438del(ac) + " No LOH T8 Ð Ð Ð + " Not informative T9 Exon 17 Nonsense mutation C1510T Ð " LOH … T10 Promoter Methylation … + Sporadic bilateral LOH … T11 Exon 3 Nonsense mutation G297A Ð Sporadic unilateral LOH … T12 Exon 13 Deletion/insertion AA1235C Ð Sporadic bilateral LOH … T13 Promoter Methylation … + Sporadic unilateral LOH … T14 Promoter Methylation … Ð Sporadic bilateral Exon 17 Nonsense mutation C1666T T15 Exon 23 Nonsense mutation C 2359 T + Sporadic unilateral LOH … T16 Exon 10 Missense mutation A956G Ð " Exon 20 Missense mutation C1984T T18 Promoter Methylation … Ð " LOH … T19 Exon 7 1-bp insertion 624insT + Familial unilateral Exon 18 Missense mutation A1676C T20 Exon 4 Nonsense mutation G409T Ð Sporadic unilateral Not informative T22 Exon 7 1-bp insertion 624insT Ð " LOH … T24 Intron 23 Splicing site defect 2489-2C A Ð " LOH T25 Promoter Methylation … Ð " LOH … T26 Promoter Methylation … Ð " LOH … T27 Exon 6 Missense mutation T587C Ð " LOH … T28 Intron 12 Splicing site defect 1215-2G A Ð " LOH … T29 Exon 18 Nonsense mutation C1735T Ð " LOH … T30 Exon 7 Nonsense mutation C644A Ð " Not informative

MOLECULAR BIOLOGY Vol. 36 No. 5 2002 GENE FUNCTIONAL DEFECTS IN RETINOBLASTOMA 629

Table (Contd.) CDKN2A Tumor RB1 region Defect Nucleotide sequence Retinoblastoma type methylation T33 Promoter Methylation … Ð " LOH T34 Exon 8 Nonsense mutation C751T Ð " LOH … T37 Exon 16 Nonsense mutation C1466A Ð Familial unilateral Intron 6 Splicing site defect 606 + 2G C T38 Exon 25 Missense mutation C2535A Ð Sporadic unilateral LOH T39 Exon 21 Nonsense mutation A2143T Ð " LOH … T40 Ð Ð Ð Ð " Not informative T41 Exon 1 Nonsense mutation G88T Ð " Not informative T42 Intron 12 Splicing site defect 1126-2insAGAA + " LOH T47 Intron 22 Splicing site defect 2326-3T G + " LOH … T48 Exon 16 Nonsense mutation C1466A Ð " No LOH T50 Exon 17 Nonsense mutation C1654T Ð " LOH … T51 Exon 18 2-bp insertion 1765insGG Ð Sporadic unilateral Promoter Methylation … T52 Promoter Methylation … Ð Sporadic unilateral LOH … T53 Ð Ð - + Familial unilateral LOH T54 Exon 16 35-bp insertion 1457ins36 Ð Sporadic unilateral No LOH T55 Exon 16 43-bp insertion 1471ins42 Ð " No LOH T56 Intron 17 Splicing site defect 1731 + 2insG Ð " Exon 18 Missense mutation A1676C T57 Exon 4 8-bp deletion 423del(accaaagt) Ð " Promoter Methylation T58 Exon 12 Nonsense mutation C1150T Ð " LOH … T59 Promoter Methylation … Ð " LOH … T60 Promoter Methylation … Ð " LOH … T62 Promoter Methylation …–" No LOH T63 Promoter Methylation … Ð " LOH … T64 Exon 22 Nonsense mutation G2236T Ð Familial unilateral LOH T66 Intron 23 Splicing site defect 2489-2C A Ð Sporadic unilateral LOH T67 Exon 9 Missense mutation T908A Ð Familial unilateral Exon 22 Nonsense mutation G2242T Note: LOH is loss of heterozygosity. Cases are indicated when testing for zygosity with microsatellite markers was not informative.

MOLECULAR BIOLOGY Vol. 36 No. 5 2002 630 BABENKO et al. affect the RB1 activity and play a role in retinoblas- 12. Greger, V., Debus, N., Lohmann, D., et al., Hum. Genet., toma. 1994, vol. 94, pp. 491Ð496. 13. Belinsky, S., Nikula, K., Palmisano, W., et al., Proc. Natl. Acad. Sci. USA, 1998, vol. 95, pp. 11891Ð11896. ACKNOWLEDGMENT 14. Sambrook, J., Fritsch, E.F., and Maniatis, T., Molecular This work was supported by the Russian program Cloning: A Laboratory Manual, Cold Spring Harbor, Human Genome. N.Y.: Cold Spring Harbor Lab. Press, 1989. 15. Strelnikov, V.V., Nemtsova, M.V., Chesnokova, G.G., REFERENCES et al., Mol. Biol., 1999, vol. 33, pp. 330Ð336. 16. Landegren, U., Laboratory Protocols for Mutation 1. Zaletaev, D.V., Sbornik nauchnykh trudov RAEN (Col- Detection, Oxford: Oxford Univ. Press, 1996. lection of Works of the Russian Academy of Natural Sci- ences), Moscow, 2000, pp. 303Ð309. 17. Zaletaev, D.V., Mol. Biol., 2000, vol. 34, pp. 671Ð683. 2. Tazi, J. and Bird, A., Cell, 1990, vol. 58, pp. 909Ð920. 18. Ohtani-Fujita, N., Dryja, T.P., Rapaport, J.M., et al., 3. Toyota, M., Ho, S., Ahuja, N., et al., Cancer Res., 1999, Cancer Genet., 1997, vol. 98, pp. 43Ð49. vol. 59, pp. 2307Ð2312. 19. Stirzaker, C., Millar, D., Paul, C., et al., Cancer Res., 4. Likhtenshtein, A.V. and Kisseljova, N.P., Biokhimiya, 1997, vol. 57, pp. 2229Ð2237. 2001, vol. 66, pp. 293Ð317. 20. Mancini, D., Singh, S., Ainsworth, P., and Rodenhiser, D., 5. Kisseljova, N.P. and Kisseljov, F.L., Kantserogenez Hum. Genet., 1997, vol. 61, pp. 80Ð87. (Carcinogenesis), Zaridze, D.G., Ed., Moscow: Nauch- 21. Lohmann, D.R., Gerick, M., Brandt, B., et al., Am. J. nyi Mir, 2000. Hum. Genet., 1997, vol. 61, pp. 282Ð294. 6. Baylin, S.B., Trends Genet., 2000, vol. 16, pp. 168Ð174. 22. Zeschnigk, M., Lohmann, D., and Horsthemke, B., 7. Laird, P.W. and Jaenisch, R., Annu. Rev. Genet., 1996, J. Med. Genet., 1999, vol. 36, pp. 793Ð794. vol. 30, pp. 441Ð464. 8. Baylin, S.B., Hum. Mol. Genet., 2001, vol. 10, pp. 687Ð 23. Lukas, J., Parry, D., Aagard, L., et al., Nature, 1995, 692. vol. 375, pp. 503Ð506. 9. Singal, R. and Ginder, G.D., Blood, 1999, vol. 93, 24. Jarrard, D.F., Sarkar, S., Shi, et al., Cancer Res., 1999, pp. 4059Ð4070. vol. 59, pp. 2957Ð2964. 10. Sakai, T., Toguchida, J., Ohtani, N., et al., Am. J. Hum. 25. Kiono, T., Foster, S.A., Koop, J.I., et al., Nature, 1998, Genet., 1991, vol. 48, pp. 880Ð888. vol. 396, pp. 84Ð89. 11. Greger, V., Passarge, E., Hopping, W., et al., Hum. 26. Herman, J.G., Umar, A., Polyak, K., et al., Proc. Natl. Genet., 1989, vol. 83, pp. 155Ð158. Acad. Sci. USA, 1998, vol. 95, pp. 6870Ð6875.

MOLECULAR BIOLOGY Vol. 36 No. 5 2002