Int. J. Cancer (Pred. Oncol.): 79, 215–220 (1998) r 1998 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer CORRELATION BETWEEN METHYLATION STATUS OF THE p16/CDKN2 GENE AND THE EXPRESSION OF p16 AND Rb PROTEINS IN PRIMARY NON-SMALL CELL LUNG CANCERS Kenji KASHIWABARA*, Tetsunari OYAMA, Takaaki SANO, Toshio FUKUDA and Takashi NAKAJIMA 2nd Department of Pathology, Gunma University School of Medicine, Gunma, Japan In order to clarify the frequency of p16 gene inactivation and its relationship with Rb expression, immunohistochemical analysis of p16 and Rb proteins was carried out on 82 paraffin-embedded sections of primary non-small cell lung cancers (NSCLCs). From immunohistochemical results, abnormal p16 expression was observed in 66% of NSCLCs, 80% in squamous cell carcinomas and 46% in adenocarcinomas. An inverse correlation between p16 and Rb expressions was noted. Moreover, the methylation status of the p16 gene was investigated by the methylation-specific polymerase chain reaction (MS-PCR) using 29 frozen samples of NSCLCs. MS-PCR revealed the methylation of the p16 gene in 10 (34%) of 29 NSCLCs. All NSCLCs exhibiting methylation exhibited abnormal p16 expression and were positive for Rb. In NSCLCs, no difference in methylation status was observed with respect to clinico-pathological characteristics including histological subtype and tumor stage. Our results demonstrate that abnormality of p16 expression is frequent in primary NSCLCs and methylation of the promoter of the p16 gene occurs in 34% of primary NSCLCs, which might play a significant role in the inactivation of the p16 gene. Int. J. Cancer (Pred. Oncol.) 79:215–220, 1998. r 1998 Wiley-Liss, Inc. Protein p16, the gene product of CDKN2A/p16INK4A, binds cyclin-dependent kinase 4 (Cdk4) and inhibits the formation of Cdk4/cyclin D complexes, resulting in the inhibition of the cell cycle-dependent phosphorylation of the retinoblastoma (Rb) gene product (Serrano et al., 1993). Alterations of the p16 gene, such as homozygous deletion and point mutations, have been observed in various cancers, and the loss of the cell cycle regulatory function of the p16 protein is thought to be implicated in the development and progression of cancers (Kamb et al., 1994; Nobori et al., 1994). In lung cancers, p16 gene alterations have been frequently detected, particularly in lung cancer cell lines and metastatic cancers (Hayashi et al., 1994; Okamoto et al., 1995). Aberrant expression of the p16 gene has frequently been observed in non-small cell lung cancers (NSCLCs) (Geradts et al., 1995; Kinoshita et al., 1996; Kratzke et al., 1996; Sakaguchi et al., 1996). In primary lung cancers, however, recent studies have clarified that point mutations and homozygous deletion of the p16 gene are not as frequent as previously expected (Hayashi et al., 1994; Okamoto et al., 1995). Therefore, another mechanism of p16 gene inactivation in primary lung cancers is suspected. Blocking the transcription of genes by methylation of 58-CpG islands around gene regulatory regions has been identified as an alternative mechanism of gene inactivation. To date, the silencing of genes by aberrant methylation has been described in various genes such as Rb, VHL, E-cadherin, and p15. Aberrant methylation of 58-CpG islands around promoter regions of the p16 gene has been suggested to be an alternative mechanism of gene inactivation in primary lung cancers (Merlo et al., 1995; Otterson et al., 1995). Moreover, the inactivation of the p16 gene by this mechanism has been reported in various other cancers such as breast, prostate, kidney, and colon (Herman et al., 1995). To investigate the expression of p16 protein in primary NSCLCs, we studied immunohistochemically p16 protein expression using a novel monoclonal antibody (MAb) for the p16 protein. Moreover, we investigated the methylation status of the p16 gene by the methylation-specific polymerase chain reaction (MS-PCR) and examined the interrelationships between aberrant p16 expression, the methylation status of the p16 gene and Rb expression in primary NSCLCs. MATERIAL AND METHODS Tissue samples Samples of a total of 82 non-small cell lung cancers (NSCLC) were collected from the pathological files of the Gunma University Hospital during 1990 to 1994. Paraffin blocks of sections of these cancers were used for immunohistochemical analysis of the p16 protein. For 29 of the 82 NSCLCs used for immunohistochemical study, fresh tumor tissues were also kept at -80°C and were used for methylation analysis of the promoter region of the p16 gene. Histological subtypes were clarified and clinical information was obtained from the pathological files or medical records of the patients. Immunohistochemical analyses of p16 and Rb proteins Sections 4 µm thick were cut from each paraffin blocks. After dewaxing with xylene and rehydration through a graded series of ethanol, the sections were microwaved in 10 mM phosphate citrate buffer, pH 8.0, at 90°C for 15 min and were treated with 0.3% H2O2-methanol solution in order to reduce endogenous peroxidase activity. Then they were incubated with normal horse serum to reduce non-specific antibody binding and were subsequently subjected to the primary MAb reaction. The primary MAb for p16, JC 8, was produced by Dr. J. Koh at the Massachusetts General Hospital Cancer Center. The JC 8 is an anti-p16 murine MAb IgG2a which recognizes an epitope in the first ankylin repeat (amino acids 1–32) of human p16 protein (Burns, et al., in press). The antibody detects a single 16 kDa band on Western blots of human tissues (data not shown). Mouse MAb against the Rb protein was purchased from PharMingen (G3-245: San Diego, CA). The primary antibodies for p16 and Rb proteins were diluted to 1:500 and 1:200, respectively, and reacted with the sections overnight at 4°C. Detection of immunoreactivity was carried out by the avidin-biotin-peroxidase (ABC) method according to the manufacturer instructions (Vectastain, Vector, Burlingame, CA). The sections were subjected to a color reaction with 0.02% 3,38diaminobenzidine tetrahydrochloride containing 0.005% H2O2 in 50 mM ammonium acetate-citric acid buffer, pH 6.0, and were lightly counterstained with hematoxylin. Nuclear staining was considered specific for p16 and Rb expression regardless of cytoplasmic staining. For the evaluation of p16 immunohistochemistry, we used the criteria of Geradts et al. (1995) with modification and the nuclear staining pattern of p16 protein were classified into 3 categories: (1) positive (1); whereby entire tumor cells exhibit distinct nuclear staining, compatible with normal pattern of Geradts’ criteria, (2) heterogeneous (6); whereby unstained clusters of tumor cells are intermingled with stained cells or occupy certain areas of the same tumor, compatible with abnormal pattern of Geradts’ criteria, (3) negative (-); whereby *Correspondence to: Second Department of Pathology, Gunma University School of Medicine, 3-39-22, Showa-machi, Maebashi, Gunma 371, Japan. Fax: (81)272-20-7981. E-mail: kashy@news. sb. gunmau.u.ac.jp Received 25 September 1997; Revised 27 November 1997 KASHIWABARA ET AL. 216 entire tumor cells are unstained, compatible with abnormal pattern of Geradts’ criteria. Methylation analysis of the p16 gene Tumor DNA was extracted from frozen cancer tissues by the conventional phenol-chloroform method after homogenization and proteinase K digestion. Purified cancer DNA was chemically modified by sodium bisulfite according to the Frommer’s method with a slight modification (Clark et al., 1994). Basically, methylated cytosine is not prone to modification by sodium bisulfite and primers complimentary to modified DNA can detect methylation. One microgram of tumor DNA was denatured in 0.3 M NaOH for 15 min at 37°C. Then, freshly prepared sodium bisulfite, pH 5.0 (Kokusan, Tokyo, Japan) and hydroquinone (Wako, Tokyo, Japan) were added at final concentrations of 3.0 M and 0.5 mM, respectively, and the mixture was incubated under mineral oil for 16 hr at 50°C. Treated DNA was desalted using the Wizard DNA Clean-up system (Promega, Madison, WI) and eluted with 50 µl of distilled water. The modified DNA was desulphonated by NaOH (final concentration 0.3 M) at room temperature for 5 min and finally purified using the Qiagen tissue kit (Qiagen GmbH, Hilden, Germany) and extracted with 50 µl of distilled water. The methylation-specific polymerase chain reaction (MS-PCR) was performed according to the method of Herman et al. (1996a). The DNA sequence of primers, conditions of the reaction, and sizes of PCR products in MS-PCR are described in Table I. The PCR TABLE I – PCR PRIMERS USED FOR MS-PCR Genomic Sequence p16 methylated ( p16-M) p16 unmethylated ( p16-U) Sense 58-38 Antisense 58-38 TCACCAGAGGGTGGGGCGGACCGC * ** * TTATTAGAGGGTGGGGCGGATCGC * ** * ** * TTATTAGAGGGTGGGGTGGATTGT CGACCCCGGGCCGCGGCCGTGG ** * ** GACCCCGAACCGCGACCGTAA * *** * ** * ** CAACCCCAAACCACAACCATAA Anneal size position 1167 65°C 150bp 60°C 151bp The * above nucleotides indicate mismatches between genomic DNA and p16-M p16-U primers. Sequence differences between p16-M and p16-U are underlined. FIGURE 1 – Representative results of p16 and Rb immunohistochemical analysis. (a) and (b) A p16-negative and Rb-positive squamous cell carcinoma, respectively. Neoplastic nuclei are not stained with p16 MAb, whereas scattered non-neoplastic nuclei stained are stained (Arrows) (a). Scale bar, 50 µm. Rb stain shows distinct reactivity (b) Scale bar, 50 µm. (c) and (d) A p16-positive and Rb-negative adenocarcinoma, respectively. Neoplastic cells exhibit diffuse nuclear reactivity with cytoplasmic staining for p16 (c) but Rb is not detected (d). Scale bar, 50 µm. THE p16 ABNORMALITY IN LUNG CANCERS mixture consisted of 16.6 mM ammonium sulfate, 67 mM TrisHCl,pH 8.8, 6.7 mM MgCl2, 10 mM 2-mercaptoethanol, 1.25 mM of each dNTP, 2.5 µM of each primer and 0.125 units of Taq polymerase (Perkin-Elmer/Cetus, Norwalk, CT). The total reaction volume was 20 µl including 2 µl of chemically modified tumor DNA solution as a template. The first step was denaturation (95°C, 2 min), followed by 35 cycles of denaturation (95°C, 45 sec), annealing (60°C or 65°C, 30 sec) and extension (72°C, 30 sec) with a final step (72°C, 4 min). Ten microliters of the reaction mixture were electrophoresed on 8% non-denaturing polyacrylamide gel containing ethidium bromide. Ten microliters of PCR products were digested by 10 units of Bst UI (New England Bio Labs, Beverly, MA) according to the manufacturer instructions to confirm the identity of the products. RESULTS Immunohistochemical analysis of p16 and Rb proteins Using MAb JC8, immunohistochemical analysis of the p16 protein was performed on paraffin sections of 82 NSCLCs. Immunoreactivity for the p16 protein was observed both in the nuclei and cytoplasm of cancer cells as well as those of nonTABLE II – IMMUNOHISTOCHEMICAL RESULTS OF p16 PROTEINS IN 82 NON-SMALL CELL LUNG CARCINOMAS Immunohistochemical results for p16 protein (1) Squamous cell carcinoma (41 cases) Well differentiated (3) Moderately differentiated (30) Poorly differentiated (8) Adenocarcinoma (35 cases) Well differentiated (13) Moderately differentiated (14) Poorly differentiated (8) Large cell carcinoma (6 cases) Stage I (33) II (13) III (28) IV (8) Total 82 cases (6) 4 (10%) 29 (70%) 0 6 0 3 3 21 1 4 (12%) 2 1 5 12 (34%) 3 4 2 1 12 3 10 3 28 (34%) 1 2 4 1 4 1 10 (12%) neoplastic mesenchymal cells (Fig. 1). The presence of nuclear reactivity in non-neoplastic cells was regarded as an internal control for p16 expression in paraffin-embedded sections. As shown in Table II, the p16 protein expression pattern was immunohistochemically evaluated as being positive in 28, heterogeneous in 10 and negative in 44 NSCLCs. Half of the adenocarcinomas were immunohistochemically positive for the p16 protein, which is a higher percentage than that in squamous cell carcinomas. In adenocarcinomas, p16 immunoreactivity appeared inversely related to the degree of tumor cell differentiation (Fig. 2), but such a trend was not seen in squamous cell carcinomas. In 6 large cell carcinomas, 5 were p16-negative or heterogeneous and only one was 16-positive. There was no correlation between p16 expression status and clinical stage of NSCLCs. In addition, 3 small cell lung carcinomas investigated in our study were immunohistochemically positive for p16 protein (data not shown). Rb expression was immunohistochemically examined in 82 NSCLCs. Of the 82 NSCLCs, 54 and 28 were positive and negative for Rb protein, respectively. Rb immunoreactivity was seen exclusively in the nuclei of both cancer cells and non-neoplastic cells. 0Rb expression was inversely related to p16 expression. Of 54 NSCLCs positive for Rb protein, 47 NSCLCs showed abnormal p16 immunoreactivity. Of 28 NSCLCs negative for Rb protein, inversely, 21 NSCLCs showed normal p16 expression. The number of NSCLCs both positive and negative for p16 and Rb expression was 7 and 3, respectively. (2) 8 (20%) 2 19 (54%) 8 9 217 5 3 17 9 14 4 44 (54%) Methylation status of the p16 gene To clarify the methylation status of the promoter region of the p16 gene, 29 fresh frozen NSCLC tissues, comprising 16 squamous cell carcinomas, 11 adenocarcinomas and 2 large cell carcinomas, were studied by the MS-PCR method. The size of the p16-U PCR product was 151 bp and the 151 bp band was observed for all of the cancers investigated (Fig. 3a). The MS-PCR product, 150 bp in size, can be cut by Bst UI endonuclease, but p16-U PCR products cannot be digested by this endonuclease because the restriction site for Bst UI was abrogated after bisulfite modification (Fig. 3b). As shown in Table III, methylation of the p16 gene was observed in 10 (34%) of 29 NSCLCs. Moreover, MS-PCR failed to detect the methylation-specific band for 2 small cell carcinomas (data not shown). Histologically, methylation of the p16 gene was detected in 6 squamous cell carcinomas (6/16, 38%), 3 adenocarcinomas (3/11, 27% ) and 1 large cell carcinoma (1/2, 50%). There was no correlation observed between methylation-status and clinico- FIGURE 2 – (a) p16 is diffusely present in the well-differentiated papillary adenocarcinoma. Bar, 50 µm. (b) p16 is not detected in the poorly differentiated solid adenocarcinoma. Bar, 50 µm. KASHIWABARA ET AL. 218 FIGURE 3 – MS-PCR of lung cancers. (a) Note strong bands between the 118 bp and 194 bp markers in lane M for cases 4, 12, 13, 18, 19, 23, 26, 28 and 30, which are MS-PCR-positive, whereas no amplification was observed in lane M for cases 11 and 34. (b) BstUI digestion of PCR products; p16-M products are digested with BstUI but p16-U products are not. *Indicates f 174-Hae III digest as m.w. markers. TABLE III – CLINICO-PATHOLOGICAL ANALYSIS OF NON-SMALL CELL LUNG CARCINOMAS ACCORDING TO THE METHYLATION STATUS OF THE p16 GENE Clinico-pathological background Age Sex Male (23) Female (6) Histological subtype Squamous cell carcinoma (16) Adenocarcinoma (11) Large cell carcinoma (2) Pathological tumor stage I (12) II (2) III (10) IV (5) Total 29 cases TABLE IV – COMPARISON OF METHYLATION STATUS OF THE p16 GENE WITH IMMUNOHISTOCHEMISTRY FOR p16 AND Rb PROTEINS IN NON-SMALL CELL LUNG CARCINOMAS Methylation status Present 67 6 3.2 64.7 6 8.9 9 1 14 5 6 3 1 10 8 1 3 1 4 2 10 (34%) Immunohistochemical results of p16 and Rb proteins Absent 9 1 6 3 19 (66%) pathological characteristics, such as age distribution, sex, histological subtype and pathological tumor stage. Relationship between methylation status and p16 and Rb expression The results of the MS-PCR and p16 and RB immunohistochemical analysis for 29 NSCLCs are summarized in Table IV. Of 10 MS-PCR-positive NSCLCs, immunohistochemical analysis for the detection of p16 protein failed to demonstrate any immunoreactivity in 7 cases. Another 3 cases showed heterogeneous p16 expression (Fig. 4). Therefore, abnormal p16 expression was observed in all 10 NSCLCs exhibiting methylation of the p16 gene as determined by MS-PCR. In the remaining 19 MS-PCR-negative NSCLCs, on the contrary, 7 cancers were p16-positive and the other 12 cancers were judged as being heterogeneous or negative for p16 immunoreactivity. On the other hand, all MS-PCR-positive NSCLCs were immunohistochemically positive for Rb protein. Methylation status MS-PCR positive (10 cases) MS-PCR negative (19 cases) (1) (6) (2) Rb(1) Rb(2) Rb(1) Rb(2) Rb(1) Rb(2) 0 0 3 0 7 0 0 7 1 3 6 2 DISCUSSION Our immunohistochemical studies on p16 and Rb proteins not only revealed the frequent aberration of p16 protein and heterogeneous p16 expression in primary NSCLCs, but also confirmed an inverse relationship between p16 and Rb expression as shown in previous studies. Several immunohistochemical studies have shown abnormal p16 expression at frequency rates from 27% to 51% in NSCLCs (Geradts et al., 1995; Kinoshita et al., 1996; Kratzke et al., 1996; Sakaguchi et al., 1996). Of 82 NSCLCs, 54% completely lacked p16 immunoreactivity and 12% showed heterogeneous staining of the p16 protein. Our immunohistochemical study using a novel MAb for the p16 protein indicated a higher frequency of p16 abnormalities than that in previous reports. Moreover, heterogeneous staining of the p16 protein, namely the presence of unstained cells among stained cells in a single tumor, was clearly demonstrated as previously pointed out by several immunohistochemical surveys of the p16 protein (Reed et al., 1995; Geradts et al., 1995; Kinoshita et al., 1996; Kratzke et al., 1996). While it has been pointed out that this may be caused by the fluctuation of p16 expression during the cell cycle and others have suggested an inactivation of the p16 gene by homozygous deletion in unstained cells (Reed et al., 1995; Geradts et al., 1995). As shown in Table IV, THE p16 ABNORMALITY IN LUNG CANCERS 219 FIGURE 4 – A p16-heterogeneous large cell carcinoma, which is MS-PCR-positive. In a tumor cell cluster in the central field, most tumor cells show p16 reactivity with scattered negative cells, whereas anaplastic and sarcomatous tumor cells around it exhibit no p16 reactivity. (a) Hematoxylin and eosin stain; (b) p16. Bar: 100 µm). our concomitant MS-PCR and immunohistochemical studies clearly demonstrated heterogeneous staining of the p16 protein in 3 NSCLCs exhibiting methylational inactivation of the p16 gene. This indicates that heterogeneous p16 expression might be due to the coexistence of methylated and unmethylated sub-populations of cancer cells in a single tumor. A similar suggestion has been made about heterogeneous expression in relation to methylation status of the hMLH1 gene in colon cancers (Kane et al., 1997). However, its precise biological meaning remains unclear. From a clinico-pathological viewpoint, abnormal p16 expression was more frequent in squamous cell carcinomas than in adenocarcinomas. A similar trend has been reported in previous immunohistochemical studies of the p16 protein in NSCLCs (Kinoshita et al., 1996; Kratzke et al., 1996; Sakaguchi et al., 1996). Indeed, frequent alterations of the p16 gene, resulting in aberrant p16 expression, have been reported in squamous cell carcinomas of the head and neck (Reed et al., 1996). In addition, our immunohistochemical study revealed that lack of p16 expression is more common in poorly differentiated adenocarcinomas than in well or moderately differentiated ones. These findings suggest that p16 expression might be closely related to the histological degree of differentiation or to progression of adenocarcinomas. Previous studies have shown that inactivation of the p16 gene is associated with tumor dissemination, metastatic potential and poor prognosis in NSCLCs (Nakagawa et al., 1995; Kratzke et al., 1996) Recent advances in molecular biology have confirmed that aberration of p16 expression occurred not only due to point mutations or homozygous deletions of the gene, but also to the methylation of the promoter of the gene (Hayashi et al., 1994; Okamoto et al., 1995; Merlo et al., 1995; Otterson et al., 1995). In primary NSCLCs, mutation or deletion of the p16 gene is a quite rare event, in contrast with established cell lines and metastatic lesions of NSCLCs (Hayashi et al., 1994; Okamoto et al., 1995). The loss of p16 expression due to methylation has been demonstrated in lung cancer cell lines (Otterson et al., 1995). Even in primary NSCLCs, Merlo et al. (1995) have described methylation of the p16 gene in 7(26%) of 27 NSCLCs and Herman et al. (1996b) reported p16 gene methylation in 53% of NSCLCs. However, simultaneous analysis of the methylation status and expression of the p16 gene has never been carried out on primary NSCLCs. Our results clearly show that about one third of NSCLCs exhibited methylation of the p16 gene, resulting in abnormal p16 expression with intact Rb expression. These findings indicate that methylation of the gene promoter is an important and significant inactivation mechanism of the p16 gene in primary NSCLCs The precise role of methylation in cancer development could not be elucidated by our simultaneous study on the methylation status and expression of the p16 gene. However, there were no differences in the methylation status of the p16 gene among histological subtypes, which contrasted markedly with the immunohistochemical differences in p16 expression among squamous cell carcinomas and adenocarcinomas. This observation may indicate that genetic alterations other than methylation frequently occur in squamous cell carcinomas of the lung, similar to head and neck cancers (Reed et al., 1996). Moreover, methylation of the p16 gene was observed irrespective of the pathological tumor stage in NSCLCs, suggesting that p16 gene methylation might occur early in lung carcinogenesis. As shown in Table IV, abnormal p16 expression was observed in more than one half of NSCLCs despite the absence of methylation of the gene. False-negative results in methylation analysis, as a result of the unsuccessful amplification of DNA in the MS-PCR due to incomplete bisulfite modification, were unlikely, because the amplification with p16-U primers, which have more mismatches with genomic sequences than p16-M primers, was observed in all cases due to contaminating non-neoplastic cells. Although genetic alteration of the p16 gene has not been examined in our study, homozygous deletions and mutations are reported to be rare in primary NSCLCs (Hayashi et al., 1994; Okamoto et al., 1995). The absence of p16 protein without any genetic alterations and promoter methylation has been reported in ovarian tumors (Marchini et al., 1997). Therefore, another mechanisms of down-regulation of the p16 gene activity are likely to be present in NSCLCs, and further investigations are necessary to resolve this issue. In conclusion, lack of p16 expression was frequent in primary NSCLCs and methylation of the promoter of the p16 gene is observed in about one third of all NSCLCs. ACKNOWLEDGEMENTS We express particular thanks to Dr. J. Koh, Massachusetts General Hospital Cancer Center, Boston, MA, and Dr. D.L. Louis, Kublik Laboratory for Neuropathology and Molecular NeuroOncology Laboratory, Massachusetts General Hospital, Boston, MA, for kindly supplying the anti-p16 MAb. 220 KASHIWABARA ET AL. REFERENCES BURNS, K.L., UEKI, K., JHUNG, S.L., KOH, J. and LOUIS, D.N., Molecular genetic correlates of p16, cdk4 and PRB immunohistochemistry in giloblastomas. J. Neuropathol. Exp. Neurol. (1997) (In press). CLARK, S.J., HARRISON, J., PAUL, C.L. and FROMMER, M., High sensitivity mapping of methylated cytosines. Nucleic Acids Res., 22, 2990–2997 (1994). GERADTS, J., KRATZKE, R.A., NIEHANS, G.A. and LINCOLN, C.E., Immunohistochemical detection of the cyclin-dependent kinase inhibitor 2/multiple tumor suppressor gene 1 (CDKN2/MST1) product gene p16INK4A in archival human solid tumors: Correlation with retinoblastoma protein expression. Cancer Res., 55, 6006–6011 (1995). HAYASHI, N., SUGIMOTO, Y., TSUCHIYA, E., OGAWA, M. and NAKAMURA, Y., Somatic mutation of the MTS (multiple tumor suppressor) gene 1/CDKN4I cyclin-dependent kinase 4 inhibitor) gene in human primary non-small cell lung carcinomas. Biochem. Biophys. Res. Comm., 202, 1426–1436 (1994). HERMAN, J.G., GRAFF, J.R., MYOHANEN, S., NELKIN, B.D. and BAYLIN, S.B., Methylation-specific PCR: A novel PCR assay for methylation status of CpG island. Proc. natl. Acad. Sci. (Wash.), 93, 9821–9826 (1996a) HERMAN, J.G., JEN, J., MERLO, A. and BAYLIN, S.B., Hypermethylationassociated inactivation indicates a tumor suppressor role for p15ink4B1. Cancer Res., 56, 722–727 (1996b). HERMAN, J.G., MERLO, A., MAO, L., LAPIDUS, R.G., ISSA, J.-P.J., DAVIDSON, N.E., SIDRANSKY, D. and BAYLIN, S.B., Inactivation of the CDKN2/p16/ MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers. Cancer Res., 55, 4525–4530 (1995). KAMB, A., GRUIS, N.A., WEAVER-FELDHAUS, J., LIU, Q., HARSHMAN, K., TAVTIGIAN, S.V., STOCKERT, E., DAY, R.S. III, JOHNSON, B.E. and SKOLNICK, M.H., A cell cycle regulator potentially involved in genesis of many tumor types. Science, 264, 436–440 (1994). KANE, M.F., LODA, M., GAIDA, G.M., LIPMAN, J., MISHARA, R., GOLDMAN, H., JESSUP, J.M. and KOLODNER, R., Methylation of the hMLH1 promoter correlated with lack of expression of hMLH1 in sporadic colon tumor cell lines. Cancer Res., 57, 808–811 (1997). KINOSHITA, I., DOSAKA-AKITA, H., MISHIMA, T., AKIE, K., NISHI, M., HIROUMI, H., HOMMURA, F. and KAWAKAMI, Y., Altered p16INK4 and retinoblastoma protein status in non-small cell lung cancers: Potential synergistic effect with altered p53 protein on proliferative activity. Cancer Res., 56, 5557–5562 (1996). KRATZKE, R.A., GREATENS, T.M., RUBINS, J.B., MADDAUS, M.A., NIEWOEHNER, D.E., NIEHANS, G.A. and GERADTS, J., Rb and p16INK4a expression in resected non-small cell lung tumors. Cancer Res., 56, 3415–3420 (1996). MARCHINI, S., CODEGONI, A.M., BONAZZI, C., CHIARI, S. and BROGGINI, M. Absence of deletions but frequent loss of expression of p16INK4 in human ovarian tumors. Brit. J. Cancer, 76, 146–149 (1997). MERLO, A., HERMAN, J.G., MAO, L., LEE, D.J., GABRIELSON, E., BURGER, P.C., BAYLIN, S.B. and SIDRANSKY, D., 58CpG island methylation is associated with transcriptional silencing of the tumor suppressor p16/ CDKN2/MTS1 in human cancers. Nature (Med.), 1, 686–692 (1995). NAKAGAWA, K., CONRAD, N., WILLIAMS, J.P., JOHNSON, B.E. and KELLY, M. Mechanism of inactivation of CDKN2 and MTS2 in non-small cell lung cancer and association with advanced stage. Oncogene, 11, 1843–1851 (1995). NOBORI, T., MIURA, K., WU, D.J., LOIS, A., TAKABAYASHI, K. and CARSON, D.A., Deletions of the cyclin-dependent kinase-a inhibitor gene in multiple human cancers. Nature (Lond.), 368, 753–756 (1994). OKAMOTO, A. and 13 OTHERS, Mutations in the p16INK4/MTS1/, p15/MTS2 and p18 Genes in Primary and metastatic lung cancer. Cancer Res., 55, 1448–1451 (1995). OTTERSON, G.A., KHLEIF, S.N., CHEN, W., COXON, A.B. and KAYE, F., CDKN2 gene silencing in lung cancer by DNA hypermethylation and kinetics of p16INK4 protein induction by 5-aza-28deoxycystidine. Oncogene, 11, 1211–1216 (1995). REED, A.L., CALIFANO, J., CAIRNS, P., WESTRA, W.H., JONES, R.M., KOCH, W., AHRENDT, S., EBY, Y., SEWELL, D., NAWROZ, H., BARTEK, J. and SIDRANSKY, D.A., High frequency of p16 (CDKN2/MTS-1/INK4A) in activation in head and neck squamous cell carcinoma. Cancer Res., 56, 3630–3633 (1996). REED, J.A., LOGANZO, F., SHEA, C.R., WALKER, G.J., FLORES, J.F., GLENDENING, J.M., BODGANY, J.K., SHIEL, M.J., HALUSAKA, F.G., FOUNTAIN, J.W. and ALBINO, A.P., Loss of the p16/Cyclin-dependent kinase inhibitor 2 tumor suppressor gene in melanocytic lesions correlates with invasive stage of tumor progression. Cancer Res. 55, 2713–2718 (1995). SAKAGUCHI, M., FUJII, Y., HIRABAYASHI, H., YOON, H.-E., KOMOT, Y., OUE, T., KUSAFUKA, T., OKADA, A. and MATSUDA, H., Inversely correlated expression of p16 and Rb protein in non-small cell lung cancers: an immunohistochemical study. Int. J. Cancer, 65, 442–445 (1996). SERRANO, M., HANNON, G.J., and BEACH, D., A new regulatory motif in cell-cycle control causing specific inhibition of cyclin D/CDK 4. Nature (Lond.), 336, 704–707 (1993).