The cyclin-dependent kinase inhibitors p15INK4b and p57KIP2 are important regulators of the cell cycle, and their abnormal expression has been detected in various tumors. However, little is known about the role of p15INK4b and p57KIP2 in the pathogenesis of vulvar carcinoma, and the prognostic impact is still unknown. In our current study, we examined the expression of p15INK4b and p57KIP2 in a large series of vulvar squamous cell carcinomas to elucidate the prognostic impact.
Expression of p15INK4b and p57KIP2 were examined in 297 vulvar squamous cell carcinomas using immunohistochemistry. Both uni- and multivariate analysis of prognostic factors were performed, and correlations with clinicopathologic parameters were examined.
Compared to the high levels of p15INK4b and p57KIP2 in normal vulvar squamous epithelium, low levels of p15INK4b and p57KIP2 were found in 82% and 44% of vulvar carcinomas, respectively. Low levels of p15INK4b and p57KIP2 correlated significantly with malignant features, including large tumor diameter (p = 0.03 and p = 0.001, respectively) and increased invasiveness (p = 0.003 and p = 0.04, respectively). Although p15INK4b and p57KIP2 levels could not be identified as prognostic markers, combined analysis of p14ARF/p15INK4b/p16INK4a showed that patients whose tumors expressed low levels of two or three of these INK4 proteins had a worse prognosis than those with only low levels of one or no protein (univariate analysis p = 0.02). The independent prognostic significance of these INK4 proteins was confirmed by multivariate analysis (p = 0.008).
Citation: Holm R, Førsund M, Nguyen MT, Nesland JM, Trope CG (2013) Expression of p15INK4b and p57KIP2 and Relationship with Clinicopathological Features and Prognosis in Patients with Vulvar Squamous Cell Carcinoma. PLoS ONE 8(4): e61273. https://doi.org/10.1371/journal.pone.0061273
Editor: John Matthew Koomen, Moffitt Cancer Center, United States of America
Received: August 15, 2012; Accepted: March 9, 2013; Published: April 8, 2013
Copyright: © 2013 Holm et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This project was supported by grants from The Norwegian Cancer Society. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Vulvar carcinoma is a rare female genital malignancy with an incidence ranging from 1 to 2 per 100 000 person-years worldwide , . Vulvar cancer has mainly been linked to elderly women but an increasing incidence among younger women has been reported recently , . The standard treatment with radical surgery is associated with a considerable morbidity . Thus, it is important to make individualized treatment procedures in order to reduce negative effects for patients with a good prognosis. The identification of new biomarkers could possibly improve the prediction of clinical outcome and may also be important for development of better treatment strategies.
Cyclin-dependent kinase inhibitors (CDKI), the major inhibitors of the cell cycle, are divided into the INK4 and CIP/KIP family. The INK4 members includes p15INK4b, p16INK4a, p18INK4c and p19INK4d/ARF (p14ARF in humans), whereas the CIP/KIP members includes p21CIP1, p27KIP2 and p57KIP2 , . INK4 protein binds specifically to the CDK4 and CDK6 complexes, causing G1 arrest. The CIP/KIP family has a broader specificity for CDKs also inhibiting the other cyclin-CDK complexes at a later stage of the cell cycle , . Previously p14ARF, p15INK4b, p16INK4a, p21CIP1 and p27KIP2 have been found to be involved in the neoplastic process of vulvar carcinomas –, however only a limited number of cases has been investigated for p15INK4b . Furthermore, to our knowledge, there has been no study of p57KIP2 in vulvar carcinomas. The loss of p15INK4b and p57KIP2 function occurs frequently in a variety of human cancers suggesting that its down-regulation may be important in neoplastic transformation –. Furthermore, an association has been found between abnormality of p15INK4b and p57KIP2 and unfavorable outcome , , –. The aim of our study was to investigate the expression of p15INK4b and p57KIP2 in a large series of vulvar squamous cell carcinomas to clarify their potential prognostic values.
Between 1977 and 2006, 297 patients had been diagnosed with vulvar squamous cell carcinoma at The Norwegian Radium Hospital. The median age at diagnosis was 74 years (range 35–96 years). Surgical radicality in the vulvar specimen (no rest tumor) was obtained in 198 (67%) of the patients and the remaining 99 (33%) patients did not obtain surgical radicality. Prior to surgery, radiotherapy was given to six patients and three cases received radiotherapy/chemotherapy. Postoperative irradiation was administered to 63, chemotherapy to three and irradiation/chemotherapy to four of the patients. After treatment, the patients have been followed at The Norwegian Radium Hospital or at a local hospital. Follow-up information is available for all patients until 1. September, 2009. During follow-up, 122 (40%) patients died of vulvar cancer. The median follow-up time for patients still alive was 151 months (range, 43 to 378 months). The tumors were all staged based on the International Federation of Gynecology and the Obstetrics (FIGO) classification from 2009 . The Regional Committee for Medical Research Ethics South of Norway (S-06012), The Social and Health Directorate (04/2639 and 06/1478) and The Data Inspectorate (04/01043) approved the current study protocol. In this study we have used paraffin embedded tumor tissue from vulvar cancer patients diagnosed between 1977 and 2006. Many of these patients are either dead or very old. Therefore, we have not been able to obtain patient consent. Permission has been obtained from The Social and Health Directorate (04/2639) to perform this study without patient consent.
The histological specimens were reevaluated by an experienced pathologist (J.M.N) according to World Health Organization recommendations . Two hundred and eighty (94%) tumors were keratinizing/nonkeratinizing, 13 (5%) were basaloid and 4 (1%) were veruccoid. Thirty-six samples of normal vulva form patients, undergoing surgery for benign gynecological diseases, were included as controls. Results obtained from our previous studies on cell cycle proteins in this same cohort of vulvar carcinomas , ,  were co-analyzed with those of the current study.
Sections from formalin-fixed, paraffin-embedded tissues were immunostained using the Dako EnVision™ Flex+ System (K8012; Dako, Glostrup, Denmark) and the Dako Autostainer. Deparaffinization and the unmasking of epitopes were performed using PT-Link (Dako) and EnVision™ Flex target retrieval solution at a high pH for p15INK4b and low pH for p57KIP2. To block endogeneous peroxidase the sections were treated with 0.3% hydrogen peroxide (H2O2) for 5 min. Sections were incubated overnight at 4°C with monoclonal antibody p15INK4b (clone 15P06, 1∶500, 0.4 µg IgG1/ml, Thermo Fischer Scientific, Fremont, CA, USA) and polyclonal rabbit antibody p57KIP2 (1∶1000, 0.14 µg IgG/ml, Sigma, St. Lous, MO, USA). The specimens were subsequently treated with goat anti-mouse IgG or goat anti-rabbit IgG for 30 min, EnVision™ Flex/HRP enzyme for 30 min, 3′3-diaminobenzidine tetrahydrochloride (DAB) for 10 min, counterstained with hematoxylin, dehydrated and mounted in Richard-Allan Scientific Cyto seal XYL (Thermo Scientific, Waltham, MA, USA). All of the sample series included appropriate positive controls, which included normal vulva (p15INK4b) and ovarian serous cystadenoma (p57KIP2). Negative controls included substitution of the monoclonal antibody with mouse myceloma protein of the same subclass and concentration as the monoclonal antibody, or normal rabbit IgG of the same concentration as the polyclonal antibody.
Semiquantitative classes were used to describe the extent of staining (percent of positive tumor cells: absent, 0; <10%, 1; 10–50%, 2; >50%, 3) and intensity (absent, 0; weak, 1; moderate, 2; strong, 3). By multiplying the extent and the intensity of the signal, product scores for nuclear staining were produced which ranging from 0 to 9. Protein levels for p15INK4b and p57KIP2 were classified as high when a score of 9 and low when a score of <9. The cutoff value for the immunoreactivity was based on staining pattern observed in normal vulvar epithelium. Examination of immunostaining was performed in a blinded fashion by two observers (R.H. and J.M.N) with no knowledge of the clinicopathological variables, patient outcomes and cell cycle protein results from our previous studies , , . All discordant scores were reviewed until a final agreement was obtained.
The Pearson's chi-square (χ2) and Linear-by-linear association were performed to fine associations between protein expression and clinicopathologic variables. Survival analysis was processed using the Kaplan and Meier estimation and log-rank test. Disease-specific survival was calculated from the date of diagnosis to vulvar cancer related death. A Cox proportional hazards regression model was used for both univariate and multivariate evaluation of survival rates. In the multivariate analysis, a backward stepwise regression was performed with a p≤0.05 as the inclusion criterion for variables in the univariate analysis. All calculations were processed using SPSS 18.0 statistical software package (SPSS, Chicago, IL, USA) and statistical significance was considered as p≤0.05.
The vulvar carcinoma tissues in our cohort have been collected over an extensive period from 1977–2006. Due to the large variation in storage time and given that the fixation protocol for these tissues up to 1987 was acid formalin, whereas from 1987–2006 was buffered formalin, Mann-Whitney U test was used to evaluate whether this has any influence on the p15INK4b and p57KIP2 immunostaining. The Mann-Whitney U test showed that the distribution of p15INK4b and p57KIP2 expression was the same between samples processed before and after 1987.
In 36 cases of normal vulvar squamous epithelium, nuclear staining for p15INK4b and p57KIP2 was identified in basal, parabasal, middle and top layers with a score of 9 (Figure 1A and B). The immunostaining results in vulvar carcinomas are summarized in Table 1. High p15INK4b and p57KIP2 immunostaining (score 9) in the nucleus was observed in 53/297 (18%) and 165/297 (56%) of the cases, respectively (Figure 1C–D), whereas low expression (score <9) was identified in 244/297 (82%) and 132/297 (44%), respectively (Figure 1E–F).
High nuclear expression of p15INK4b (A) and p57KIP2 (B) in normal vulvar epithelium. High nuclear expression of p15INK4b (C) and p57KIP2 (D) and low nuclear expression of p15INK4b (E) and p57KIP2 (F) in vulvar carcinomas.
The levels of p15INK4b and p57KIP2 in relation to clinicopathological parameters are shown in Table 2. Low expression of p15INK4b and p57KIP2 were significantly correlated with large tumor diameter (p = 0.03 and p = 0.001, respectively) and deep invasion (p = 0.003 and p = 0.04, respectively). In addition, low expression of p57KIP2 significantly correlated with younger age (p = 0.03) and a low level of cyclin D3 (p = 0.05).
By univariate analysis neither p15INK4b nor p57KIP2 were associated with disease-specific survival (p = 0.29 and p = 0.94). Because the prognosis of patients most likely do not dependent on the expression of one single member of the INK4 and CIP/KIP families but on an interplay between the different proteins, the previously determined p16INK4a and p14ARF status ,  were combined with the expression of p15INK4b and the previously identified p27KIP1 and p21CIP1 data  were combined with the expression of p57KIP2 for each patient. Patients whose tumors expressed low levels of two or three of these INK4 proteins have a worse prognosis than those with only low levels of one or no protein (p = 0.02) (Figure 2). Hovewer, the combination of the CIP/KIP members were not significantly correlated to clinical outcome (p = 0.78). By multivariat analysis, lymph node metastasis, vessel infiltration, tumor diameter and the combined p14ARF/p15INK4b/p16INK4a status were of independent prognostic significance (Table 3).
The Kaplan-Meier curve of disease-specific survival in relation to the combined analysis of p14ARF/p15INK4b/p16INK4a showed that patients whose tumors expressed low levels of two or three of these INK4 proteins had a worse prognosis than those with only low levels of one or no protein (p = 0.02).
The results from the univariate analysis combining p14ARF/p15INK4b/p16INK4a status in a subgroup of patients who obtained surgical radicality (no rest tumor) is showed in Figure 3. In this subgroup, multivariate analysis showed that the combined p14ARF/p15INK4b/p16INK4a status were of independent prognostic significance (Table 3). However, p15INK4b, p57KIP2 and combined p21CIP1/p27KIP2/p57KIP2 status were not significantly correlated to prognosis (data not shown). We also performed analyses of the subgroup of patients presenting with rest tumor. The parameters used are the same as those used in the analyses of patients who obtained surgical radically (no rest tumor). We were not able to reveal any prognostic significance (data not shown).
The Kaplan-Meier curve of disease-specific survival in relation to the combined analysis of p14ARF/p15INK4b/p16INK4a showed that patients whose tumors expressed low levels of two or three of these INK4 proteins had a worse prognosis than those with only low levels of one or no protein (p = 0.06).
Compared to the high level of p57KIP2 in normal vulvar squamous epithelium, low p57KIP2 expression was identified in 44% of vulvar carcinomas. Low p57KIP2 expression has been reported in 89% of non small cell lung cancer , 74% of bladder cancer , 72% of breast cancer , 71% of colorectal carcinoma , 60% of ovarian carcinoma , 50% of esophageal squamous cell carcinoma  and 45% of hepatocellular carcinoma , . The wide range of the occurrence of low p57KIP2 expression may be due to the various tumor types studied. In our current study, reduced p57KIP2 level significantly correlated with a large tumor diameter and increased invasiveness. The decrease in p57KIP2 expression with increased tumor size is in line with the findings in some other malignancies, including laryngeal , liver , oral  and pancreatic cancer . Thus the reduced expression of p57KIP2 may be involved in progression of different tumors, including vulvar carcinoma.
Previously, homozygous deletion of p15INK4b has been detected in 3 of 6 (50%) vulvar carcinomas . In the current study, 82% of the vulvar carcinomas had low levels of p15INK4b expression compared to the normal counterpart. The low level of p15 protein in a higher number of vulvar carcinomas than cases with deleted p15INK4b may reflect that loss of p15 expression are not only due to deletion, but also to mutations and methylation of p15INK4b gene. Decreased expression of p15INK4b protein has been observed in 54% of malignant peripheral nerve sheat tumor . The p15INK4b gene has been reported to be homozygously deleted in 50% of T-cell lymphoma , 33% of ovarian cancer  and 23% of nonsmall cell lung cancer , whereas mutations of the p15INK4b gene were detected in 12% of non small cell lung cancer . In addition, p15INK4b promoter methylation has been found in 65% of head and neck squamous cell carcinoma , 47% of hepatocellular carcinoma , 36% of T-cell lymphoma  and 30% of ovarian cancer . However, no p15INK4b deletion and/or mutations and/or methylation were detected in cervical cancer  and uveal melanoma . Thus, abnormality of p15INK4b may be tumor specific. Our present results showed a significantly correlation between low p15INK4b expression and malignancy of vulvar carcinomas, including a large tumor diameter and increased invasiveness. Endo et al.  have reported in an earlier study of malignant peripheral nerve sheath tumors that protein levels of p15INK4b are significantly lower in large tumors compared with small tumors. Taken together, the low expression of p15INK4b in the majority of vulvar carcinomas and the association with malignant features suggest that p15INK4b may be important in the pathogenesis and/or progression of vulvar carcinomas.
We found no prognostic significance for p57KIP2. Similar findings were previously reported in esophageal squamous cell carcinoma , colorectal cancer , pancreatic carcinoma  and ovarian cancer . In contrast, reduced expression of p57KIP2 has been correlated with poor outcome in univariate as well as in multivariate analysis in patients with carcinomas of the laryngeal  and breast . For patients with ovarian cancer  and hepatocellular carcinoma  low p57 expression was significantly correlated with poor prognosis in univariate but not in multivariate analysis.
Abnormal expression of p15INK4b has been linked to unfavorable outcome in univariate  and multivariate  analysis in patients with ovarian cancers, whereas Endo et al.  found that in patients with malignant peripheral nerve sheath tumors a decreased expression of p15INK4b was associated with an unfavorable prognosis. We can not confirm the prognostic significance in the present series of vulvar carcinoma patients. However, in both univariate- and multivariate tests a combined analysis of p14ARF/p15INK4b/p16INK4a showed that patients with tumors expressing low levels of two or three of these INK4 proteins had a worse prognosis than those with only low levels of one or no protein. The same phenomenom has been found in malignant peripheral nerve sheath tumors , indicating that a synergetic effect of the combined deficiency for p14ARF, p15INK4b and p16INK4a induced high-grade malignancy, not only in malignant peripheral nerve sheath tumors, but also in vulvar carcinomas. The combined analysis of p14ARF/p15INK4b/p16INK4a may in the future be used as prognostic marker for patients with vulvar carcinomas, but future studies must be performed to eventually confirm a role for these INK4 proteins as predictor for therapy.
In the current study, low p57KIP2 expression was significantly associated with low expression of cyclin D3. This is in line with the positive association between p57KIP2 and cyclin D1 in esophageal squamous cell carcinoma . In contrast, an inverse correlation has been seen between p57KIP2 and cyclin D1 expression in hepatocellular carcinoma , whereas such association has not been identified in ovarian cancer . Furthermore, a positive correlation between p57KIP2 and cyclin A in colorectal cancer  and an inverse association between p57KIP2 and cyclin E in ovarian cancer  have been reported, which was not found in our study (data not shown). Taken together, these findings suggest that cell cycle proteins regulate the cell growth through different mechanisms in the various tumor types.
Compared to the high levels of p15INK4b and p57KIP2 in normal vulvar squamous epithelium, low levels of p15INK4b and p57KIP2 were found in 82% and 44% of vulvar carcinomas, respectively. Furthermore, low levels of p15INK4b and p57KIP2 correlated significantly with large tumor diameter and increased invasiveness. These findings suggest that p15INK4b and p57KIP2 may be involved in the progression of vulvar carcinomas. Although p15INK4b and p57KIP2 levels could not be identified as prognostic markers, combined analysis of p14ARF/p15INK4b/p16INK4a showed that patients whose tumors expressed low levels of two or three of these INK4 proteins had a worse prognosis than those with only low levels of one or no protein.
Conceived and designed the experiments: RH JMN CGT. Performed the experiments: MF MTN. Analyzed the data: RH JMN. Contributed reagents/materials/analysis tools: RH CGT. Wrote the paper: RH JMN CGT.
- 1. Coulter J, Gleeson N (2003) Local and regional recurrence of vulval cancer: management dilemmas. Best Pract Res Clin Obstet Gynaecol 17: 663–681.
- 2. Giles GG, Kneale BL (1995) Vulvar cancer: the Cinderella of gynaecological oncology. Aust N Z J Obstet Gynaecol 35: 71–75.
- 3. Jones RW, Baranyai J, Stables S (1997) Trends in squamous cell carcinoma of the vulva: the influence of vulvar intraepithelial neoplasia. Obstet Gynecol 90: 448–452.
- 4. Messing MJ, Gallup DG (1995) Carcinoma of the vulva in young women. Obstet Gynecol 86: 51–54.
- 5. Sherr CJ, Roberts JM (1999) CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev 13: 1501–1512.
- 6. Pateras IS, Apostolopoulou K, Niforou K, Kotsinas A, Gorgoulis VG (2009) p57KIP2: “Kip”ing the cell under control. Mol Cancer Res 7: 1902–1919.
- 7. Knopp S, Bjorge T, Nesland JM, Trope C, Scheistroen M, et al. (2004) p16INK4a and p21Waf1/Cip1 expression correlates with clinical outcome in vulvar carcinomas. Gynecol Oncol 95: 37–45.
- 8. Knopp S, Nesland JM, Trope C, Holm R (2006) p14ARF, a prognostic predictor in HPV-negative vulvar carcinoma. Am J Clin Pathol 126: 266–276.
- 9. Wong YF, Chung TK, Cheung TH, Nobori T, Yim SF, et al. (1997) p16INK4 and p15INK4B alterations in primary gynecologic malignancy. Gynecol Oncol 65: 319–324.
- 10. Okamoto A, Hussain SP, Hagiwara K, Spillare EA, Rusin MR, et al. (1995) Mutations in the p16INK4/MTS1/CDKN2, p15INK4B/MTS2, and p18 genes in primary and metastatic lung cancer. Cancer Res 55: 1448–1451.
- 11. Chang HW, Ling GS, Wei WI, Yuen AP (2004) Smoking and drinking can induce p15 methylation in the upper aerodigestive tract of healthy individuals and patients with head and neck squamous cell carcinoma. Cancer 101: 125–132.
- 12. Yang B, Guo M, Herman JG, Clark DP (2003) Aberrant promoter methylation profiles of tumor suppressor genes in hepatocellular carcinoma. Am J Pathol 163: 1101–1107.
- 13. Bozdogan O, Atasoy P, Batislam E, Basar MM, Basar H (2008) Significance of p57(Kip2) down-regulation in oncogenesis of bladder carcinoma: an immunohistochemical study. Tumori 94: 556–562.
- 14. Fan GK, Xu F, Yang B, Fujieda S (2006) p57(kip2) expression is related to carcinogenesis and tumor progression in laryngeal tissues. Acta Otolaryngol 126: 301–305.
- 15. Matsumoto M, Furihata M, Ohtsuki Y, Sasaguri S, Ogoshi S (2000) Immunohistochemical characterization of p57KIP2 expression in human esophageal squamous cell carcinoma. Anticancer Res 20: 1947–1952.
- 16. Pateras IS, Apostolopoulou K, Koutsami M, Evangelou K, Tsantoulis P, et al. (2006) Downregulation of the KIP family members p27(KIP1) and p57(KIP2) by SKP2 and the role of methylation in p57(KIP2) inactivation in nonsmall cell lung cancer. Int J Cancer 119: 2546–2556.
- 17. Sui L, Dong Y, Ohno M, Watanabe Y, Sugimoto K, et al. (2002) Expression of p57kip2 and its clinical relevance in epithelial ovarian tumors. Anticancer Res 22: 3191–3196.
- 18. Endo M, Kobayashi C, Setsu N, Takahashi Y, Kohashi K, et al. (2011) Prognostic significance of p14ARF, p15INK4b, and p16INK4a inactivation in malignant peripheral nerve sheath tumors. Clin Cancer Res 17: 3771–3782.
- 19. Kudoh K, Ichikawa Y, Yoshida S, Hirai M, Kikuchi Y, et al. (2002) Inactivation of p16/CDKN2 and p15/MTS2 is associated with prognosis and response to chemotherapy in ovarian cancer. Int J Cancer 99: 579–582.
- 20. Liu Z, Wang LE, Wang L, Lu KH, Mills GB, et al. (2005) Methylation and messenger RNA expression of p15INK4b but not p16INK4a are independent risk factors for ovarian cancer. Clin Cancer Res 11: 4968–4976.
- 21. Tsihlias J, Kapusta L, Slingerland J (1999) The prognostic significance of altered cyclin-dependent kinase inhibitors in human cancer. Annu Rev Med 50: 401–423.
- 22. Ito Y, Takeda T, Sakon M, Tsujimoto M, Monden M, et al. (2001) Expression of p57/Kip2 protein in hepatocellular carcinoma. Oncology 61: 221–225.
- 23. Li JQ, Wu F, Usuki H, Kubo A, Masaki T, et al. (2003) Loss of p57KIP2 is associated with colorectal carcinogenesis. Int J Oncol 23: 1537–1543.
- 24. Pecorelli S (2009) Revised FIGO staging for carcinoma of the vulva, cervix, and endometrium. Int J Gynaecol Obstet 105: 103–104.
- 25. Kleihues PSL (2003) Pathology and genetics of tumors of the breast and femal genital organs. In World Health Organization of Tumours. Edited by Tavassoli FA DJ. IARC Press 2003: 316–321.
- 26. Knopp S, Bjorge T, Nesland JM, Trope C, Holm R (2005) Cyclins D1, D3, E, and A in vulvar carcinoma patients. Gynecol Oncol 97: 733–739.
- 27. Larson PS, Schlechter BL, King CL, Yang Q, Glass CN, et al. (2008) CDKN1C/p57kip2 is a candidate tumor suppressor gene in human breast cancer. BMC Cancer 8: 68.
- 28. Nan KJ, Guo H, Ruan ZP, Jing Z, Liu SX (2005) Expression of p57(kip2) and its relationship with clinicopathology, PCNA and p53 in primary hepatocellular carcinoma. World J Gastroenterol 11: 1237–1240.
- 29. Fan GK, Chen J, Ping F, Geng Y (2006) Immunohistochemical analysis of P57(kip2), p53 and hsp60 expressions in premalignant and malignant oral tissues. Oral Oncol 42: 147–153.
- 30. Ito Y, Takeda T, Wakasa K, Tsujimoto M, Matsuura N (2001) Expression of p57/Kip2 protein in pancreatic adenocarcinoma. Pancreas 23: 246–250.
- 31. Drexler HG (1998) Review of alterations of the cyclin-dependent kinase inhibitor INK4 family genes p15, p16, p18 and p19 in human leukemia-lymphoma cells. Leukemia 12: 845–859.
- 32. Kim JW, Namkoong SE, Ryu SW, Kim HS, Shin JW, et al. (1998) Absence of p15INK4B and p16INK4A gene alterations in primary cervical carcinoma tissues and cell lines with human papillomavirus infection. Gynecol Oncol 70: 75–79.
- 33. Hearle N, Damato BE, Humphreys J, Wixey J, Green H, et al. (2003) Contribution of germline mutations in BRCA2, P16(INK4A), P14(ARF) and P15 to uveal melanoma. Invest Ophthalmol Vis Sci 44: 458–462.
- 34. Rosenberg E, Demopoulos RI, Zeleniuch-Jacquotte A, Yee H, Sorich J, et al. (2001) Expression of cell cycle regulators p57(KIP2), cyclin D1, and cyclin E in epithelial ovarian tumors and survival. Hum Pathol 32: 808–813.
- 35. Yang X, Karuturi RK, Sun F, Aau M, Yu K, et al. (2009) CDKN1C (p57) is a direct target of EZH2 and suppressed by multiple epigenetic mechanisms in breast cancer cells. PLoS One 4: e5011.