High Methylation Rate of LMX1A, NKX6-1, PAX1, PTPRR, SOX1, and ZNF582 Genes in Cervical Adenocarcinoma

Transcription

1 ORIGINAL STUDY High Methylation Rate of LMX1A, NKX6-1, PAX1, PTPRR, SOX1, and ZNF582 Genes in Cervical Adenocarcinoma Cheng-Chang Chang, MD,*Þþ Rui-Lan Huang, PhD,Þ Hui-Chen Wang, BS,Þ Yu-Ping Liao, MS,Þ Mu-Hsien Yu, MD, PhD,* and Hung-Cheng Lai, MD, PhD*Þþ Objective: This study aimed to investigate the status of DNA ation of 6 genes, LMX1A, NKX6-1, PAX1, PTPRR, SOX1, and ZNF582, previously found from squamous cell carcinomas in adenocarcinomas (ACs) of the uterine cervix. Methods: We assessed the ation status of these genes in 40 ACs, cervical scrapings from 23 ACs, and 67 normal control cervices by real-time quantitative ation-specific polymerase chain reaction. The results were validated by bisulfite pyrosequencing. Results: The ation levels of all the 6 genes in the ACs were significantly higher than those in normal cervical tissues, especially for PAX1, PTPRR, SOX1, and ZNF582. The odds ratios and 95% confidence intervals (CIs) of high ation levels in PAX1, PTPRR, SOX1, and ZNF582 for the risk of developing an AC were 15.7 (95% CI, 7.0Y40.6), 16.9 (95% CI, 7.6Y43.0), 32.1 (95% CI, 12.1Y124.3), and 25.4 (95% CI, 10.4Y78.3), respectively (all P G 0.001). The ation indices of PAX1, PTPRR, SOX1, and ZNF582 recovered from scrapings of ACs were significantly higher than in normal controls. The odds ratios of these indices for the risk of developing an AC in PAX1, PTPRR, SOX1, and ZNF582 were 6.2 (95% CI, 2.6Y15.4), 12.1(95% CI, 3.8Y46.4), 6.2 (95% CI, 2.6Y15.8), and 20.6 (95% CI, 6.9Y77.5), respectively (all P G 0.001). Conclusions: Cervical ACs carry aberrantly high ation rates of PAX1, PTPRR, SOX1, and ZNF582Vcommonly ated in squamous cell carcinomasvwhich might help for AC screening. Key Words: Adenocarcinoma of the cervix, Bisulfite pyrosequencing, DNA ation, Real-time quantitative ation-specific PCR Received August 20, 2013, and in revised form October 29, Accepted for publication October 31, (Int J Gynecol Cancer 2014;24: 201Y209) *Department of Obstetrics and Gynecology, Tri-Service General Hospital; Laboratory of Epigenetics and Cancer Stem Cells, and Graduate Institute of Medical Sciences, National Defense Medical Center; and Department of Obstetrics and Gynecology, Shuang Ho Hospital, Taipei Medical University, Taipei, Taiwan, Republic of China. Address correspondence and reprint requests to Hung-Cheng Lai, MD, PhD, Department of Obstetrics and Gynecology, Tri-Service General Hospital, National Defense Medical Center, 5F, No. 325, Sec 2, Cheng-Gong Rd, Neihu District, Taipei City 114, Taiwan, Copyright * 2014 by IGCS and ESGO ISSN: X DOI: /IGC Republic of China. Department of Obstetrics and Gynecology, Shuang Ho Hospital, Taipei Medical University, No.291, Zhongzheng Rd., Zhonghe District, New Taipei City, 23561, Taiwan. hclai@ndmctsgh.edu.tw. This work was supported in part by the following grants: TSGC96-2-S01-4 from the Tri-Service General Hospital, NSC B and NSC B MY2 from the National Science Council, Republic of China, and the Teh-Tzer Study Group for Human Medical Research Foundation. Supplemental digital content is available for this article. Direct URL citation appears in the printed text and is provided in the HTML and PDF versions of this article on the journal s Web site ( The authors declare no conflicts of interest. International Journal of Gynecological Cancer & Volume 24, Number 2, February

2 Chang et al International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 ervical cancer remains one of the main causes of death of Cwomen worldwide. 1 Most cancers of the uterine cervix are squamous cell carcinomas (SCCs), with the rest being mostly adenocarcinomas (ACs). The incidence of SCC of the uterine cervix has been declining over time. However, the incidence of cervical AC has risen, especially in young women. Thus, in regions with good cervical cancer screening programs, the incidence has increased by 15% to 20% compared with unscreened populations. 2 Beyond this relative increase, absolute rates of cervical AC are thought to have increased in various countries over the past 2 to 3 decades. 3,4 The possible reasons might be that the precancerous phase of ACs is not well defined and that the sensitivity of cytology screening for ACs is inherently poor. These observations indicate that current screening practices might be insufficient to detect a substantial proportion of ACs. 5 Therefore, methods other than cytologic screening are needed. The main risk factor leading to cervical cancer is infection with oncogenic human papillomavirus (HPV), which is present in almost all cervical cancer tissues. Most previous studies have shown a strong association between cervical AC and HPV infection, as with SCCs. 6 The latest report using worldwide data confirmed that infection with HPV (type 16/18) is still the principal risk factor for an AC. 7 However, cofactors contributing to the progression to an AC seem distinct from those for an SCC. 7 The extent to which HPV infection and cofactors can explain the trend of increasing AC is unclear. The presence of HPV alone is inadequate for cervical oncogenesis, although it is a strong risk factor. The molecular mechanisms responsible for this inefficiency of HPV-initiated cervical carcinogenesis remain unclear. 8 Cofactors, genetic or environmental, might contribute a driving force to the progression of cervical lesions. The genetic component of cervical cancer, if any, is minimal, and the environmental factors are too subtle and complex to define. Epigenetic components such as DNA ation constitute an interface of the genome with the environment including viral infections. This leaves a heritable record of such interactions, which are ideal biomarkers for cancer detection. 8 In human cancers, epigenetic silencing of tumor suppressor genes by promoter hyperation is common, and this could be useful for early diagnosis and prognosis prediction for women with cancers. 8 The family of DNA transferases (DNMTs) can catalyze the ation of CpG sites in the human genome. DNMT1 is a preserved transferase with a preference for hemiated DNA, whereas DNMT3A and DNMT3B are de novo transferases with nearly equal preferences for ated and unated DNA. 8 Increasing expression of DNMT1 has been reported in cases of cervical neoplasia. 8 The RNA copy number and expression level of DNMT3B are also increased in HPV-infected immortalized cell lines and cancer tissues. 9 Epigenetic changes in cervical SCCs have been reported, 8,10 which supports the idea that DNA ation could be a marker for cervical cancer screening. 10Y12 Most previous studies of DNA ation in cervical cancers have focused on SCCs. Unlike SCCs, which show a declining trend, ACsVthe second most common cell type in cervical cancersvare increasing in frequency. 3,4 The analysis 202 of DNA ation in screening for ACs, although increasing, is limited. 13,14 Previously, we used a CpG island microarray and identified novel genes that were silenced by ation in cervical SCCs. 11 Quantitative analysis of these genes is effective in the detection of cervical intraepithelial neoplasia grade 3 and worse tumors. 12 To look for other genes for clinical applications, we used immunoprecipitation of ated DNA sequences coupled with microarray analysis and found that the genes for protein tyrosine phosphatase receptor type R (PTPRR) 15 and zinc finger protein 582 (ZNF582) 16 were highly ated in SCCs. Whether and to what extent the genes that are ated in SCCs are also ated in ACs remains to be determined. This will be critical for the clinical application of these DNA ation patterns as biomarkers for adjuvant cervical cancer screening. Here, we investigated the levels of DNA ation of these genes using omic approaches on SCCs recovered from ACs. MATERIALS AND METHODS Patients Patients with normal uterine cervices (n = 67) and with invasive AC tissues (n = 40) and scrapings (n = 23) of the uterine cervix participated in this study. The cervical scrapings of patients with normal cytology were used as normal control. The conditions of the patients were diagnosed and treated and their tissues were banked at the National Defense Medical Center in Taipei, Taiwan, as described. 17 All ACs were confirmed by histopathology. Controls were recruited from healthy women who underwent routine Papanicolaou screening during the same period. Informed consent was obtained from all patients and control subjects. Exclusion criteria included pregnancy, chronic or acute systemic viral infections, a history of cervical neoplasia, skin or genital warts, an immunocompromised state, presence of other cancers, or previous surgery of the uterine cervix. The institutional review board of the National Defense Medical Center approved the study. Real-Time Quantitative Methylation-Specific Polymerase Chain Reaction Amplification Genomic DNA was extracted from specimens collected using an established protocol for tissue banking. The concentration of DNA was determined using the PicoGreen fluorescence absorption method. Quantitative ationspecific polymerase chain reaction (QMSP) was performed after bisulfite treatment on denatured genomic DNA. Gene symbols, primers, and probes for QMSP will be provided on request. 12 The gene for COL2A was used as an internal reference for adjusting input DNA amount by amplifying non- CpG sequences in each sample. Quantitative ation-specific polymerase chain reaction was performed in a TaqMan probe system using the Applied Biosystems 7900HT Fast Real-Time PCR System (for sets A and B) and Roche LightCycler 480 system (for set C). The 5 -end and 3 -end of probes were separately labeled with 6-carboxyfluorescein and with a quencher dye. The 20-KL reaction contained 2 KL of bisulfite template * 2014 IGCS and ESGO

3 International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Methylation in Cervical Adenocarcinoma DNA, 250 nmol/l of each primer, 225 nmol/l TaqMan probe, and 10 KL of FastStart Universal Probe Master (ROX, Roche). For the TaqMan-based QMSP, each sample was analyzed in duplicate. The reactions were performed using an initial incubation at 95-C for 10 minutes, followed by 45 cycles of 95-C for 15 seconds and annealing and extending at the appropriate temperatures for 1 minute at 60-C. DNA ation level was ([Cp of assessed as the ation index (M-index), 10,000 2 COL2A] j [Cp of gene]). 16 The failed amplification of QMSP was the definite Cp value of COL2A higher than 36. Bisulfite Pyrosequencing We used bisulfite pyrosequencing (BPS) to detect and quantify the patterns and levels of DNA ation, which range from 0% to 100% at each CpG site. The BPS primers were designed using PyroMark Assay Design 2.0 software (Qiagen GmbH, Hilden, Germany). The PCR products were amplified using PyroMark PCR kits in a total volume of 20 KL containing 20 ng of bisulfite-converted DNA, 500 nmol/l of each primer, and 1 time of PCR Master Mix under the following conditions: denaturation at 95-C for 15 minutes, 49 polymerization cycles (95-C for 30 seconds; the appropriate annealing temperature for 40 seconds and 72-C for 45 seconds), and a final extension of 72-C for 5 minutes. The BPS followed the manufacturer s recommendations for operation on a PyroMark Q24 System. Statistical Analysis The sample size was calculated in advance to achieve a sufficient statistical power based on our previous ation results in SCCs. 12 Eighteen patients were sufficient to obtain a significant difference with an > level of 0.05 and a power of 95%. High and low ation levels identified from 50 percentiles of M-index values were used to analyze the association between DNA ation and diagnosis in each gene. Methylation index values for each gene were compared between cancer cases and controls using a 2-sided Mann-Whitney nonparametric U test. Associations between ation level and the risk of cervical AC were estimated using unconditional multivariate logistic regression models with adjustment for patient age. All analyses were carried out using R software (version 2.15; Differences were considered significant at P value less than Results are presented as the odds ratio (OR) and 95% confidence interval (CI) for each risk of developing an AC based on the M-index for each gene. RESULTS The demographic characteristics of patients supplying all samples are shown in Supplemental Table 1 available at We included 67 normal cervical scrapings, 40 ACs from cervical tissues, and 23 ACs FIGURE 1. Flow chart of the study sampling procedure. We studied 67 normal cervical scrapings, 40 ACs of cervical tissues, and 23 ACs of cervical scrapings. Set A is the prevalidation set (AC tissues from , set B is the validation set (AC tissues from 2005Y2008), and set C is the potential marker set (AC scrapings from 2009Y2012). The scrapings and tissues are not paired samples. * 2014 IGCS and ESGO 203

4 Chang et al International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 FIGURE 2. Real-time QMSP analysis of LMX1A, NKX6-1, PAX1, PTPRR, SOX1, and ZNF582 in normal cervical scrapings and cervical ACs in set A. Key: Nor.S, normal cervical scrapings (n = 24); AC.T, AC tissue of the cervix (n = 18). All genes showed significant hyperation in cervical ACs but not in normal cervical scrapings. Differences were evaluated using 2-sided Mann-Whitney nonparametric U tests. from cervical scrapings. We classified the samples into 3 sets as follows: prevalidation set A (AC tissues from 1999 to 2004), validation set B (AC tissues from 2005 to 2008), and potential marker set C (AC scrapings from 2009 to 2012). Set A included 24 normal cervical scrapings and 18 ACs of cervical tissues. Set B included 21 normal cervical scrapings and 22 ACs of cervical tissues. Set C included 22 normal cervical scrapings and 23 ACs of cervical scrapings (Fig. 1). In set A, we analyzed the ation status of LMX1A, NKX6-1, PAX1, PTPRR, SOX1, and ZNF582 in AC tissues and normal scrapings. High and low ation levels were based on the 50th percentile of M-index values and were used to analyze the association between DNA ation and diagnosis in each gene. Using an M-index of 50% as a cutoff value, QMSP confirmed significantly higher ation levels in NKX6-1, PAX1, PTPRR, SOX1, and ZNF582 in AC tissues than in normal cervical scrapings (P G 0.001, Fig. 2). In our prevalidation set A, NKX6-1 revealed wide variations in ation levels in normal control samples. LMX1A was weakly ated compared with controls in a metastatic stage of cervical cancer in our previous study (P = 0.028, Table 1). 18 We tested the ation levels of PAX1, PTPRR, SOX1, and ZNF582 in sets B and C. In set B, QMSP analysis validated significantly higher ation levels of PAX1, PTPRR, SOX1, andznf582 inactissuesthaninnormalcervical 204 scrapings (P G 0.001, Table 1 and Fig. 4A). In combining sets A and B, hyperation of PAX1, PTPRR, SOX1, and ZNF582 genes gave ORs of 15.7 (95% CI, 7.0Y40.6), 16.9 (95% CI, 7.6Y43.0), 32.1, (95% CI, 12.1Y124.3), and 25.4, (95% CI, 10.4Y78.3) for the risk of developing an AC, respectively (all P G 0.001, Table 1). The ation levels and patterns of these genes between normal cervical tissues and AC tissues were validated by BPS (Fig. 3). The ation levels of PAX1 CpG sites were 9% to 32% in AC tissues, whereas the range was 2% to 7% in normal cervical tissues. The ation levels of PTPRR CpG siteswere 6% to 28% in AC tissues versus 3% to 12% in normal cervical tissues. The ation levels of SOX1 CpG sites were very low (0%Y4%) in normal cervical tissues but were from 29% to 66% in cancer tissues. The ation levels of ZNF582 CpG sites were 10% to 54% in AC tissues compared with 4% to 7% in normal cervical tissues. For clinical applications, cervical scrapings rather than cervical tissues are more logical and feasible for ation analysis. We tested the ation levels of cervical scrapings in set C. The results confirmed that the ation levels of PAX1, PTPRR, SOX1, andznf582 genes in cervical scrapings were significantly different between ACs and normal controls. The ORs of PAX1, PTPRR, SOX1, andznf582 for the risk of developing an AC were 6.2 (95% CI, 2.6Y15.4), 12.1(95% CI, * 2014 IGCS and ESGO

5 International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Methylation in Cervical Adenocarcinoma TABLE 1. High ation levels of 6 genes and their related ORs in cervical AC tissue samples Set A Set B Set A + B Gene Normal Scraping (n = 24) AC.T (n = 18) OR (95% CI) P* Normal Scraping (n = 21) AC.T (n = 22) OR (95% CI) P* Normal Scraping (n = 45) AC.T (n = 40) OR (95% CI) P LMX1A 5.0 (1.2Y24.9) L NKX (7.7Y3617.0) L PAX (4.4Y317.8) ND ND G0.001 ND ND G (19.7Y15,049.3) G (7.0Y40.6) L PTPRR Inf G G (6.7Y864.5) (7.6Y43.0) L SOX1 Inf G G (19.7Y15,049.3) (12.2Y124.4) L ZNF582 Inf G G (11.0Y6655.6) (10.5Y78.4) L *Fisher exact test. Adjusted age. AC.T, AC of cervical tissue; ND, none detected. G0.001 G0.001 G0.001 G0.001 * 2014 IGCS and ESGO 205

6 Chang et al International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 FIGURE 3. Validation of ation status in AC tissues and normal cervical scrapings (normal cervix) by BPS. CpG islands of PAX1, PTPRR, SOX1, and ZNF582 were analyzed, and black circles represent completely ated genes. Arrows show transcription start sites. 206 * 2014 IGCS and ESGO

7 International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Methylation in Cervical Adenocarcinoma FIGURE 4. Methylation status of PAX1, PTPRR, SOX1, and ZNF582 in normal and AC cervical cells. A, Highly ated levels of these 4 genes were evident in ACs of cervical tissues (set A plus set B). B, Highly ated levels of the genes were found in scrapings of ACs but not in normal cervical tissues. Key: Nor.S, normal cervical scraping; AC.T, AC of cervical tissue; AC.S, ACs from cervical scrapings. 3.8Y46.4), 6.2, (95% CI, 2.6Y15.8), and 20.6 (95% CI, 6.9Y77.5), respectively (all P G 0.001, Fig. 4B and Table 2). DISCUSSION For decades, cytology-based tests to detect cervical cancers have been criticized because of unsatisfactory sensitivity, especially for ACs. 5 There have been frequent reports of faulty diagnoses on glandular lesions because of inadequate sample preparation, and there is an urgent need for an accurate method to detect cervical ACs. Although HPV testing is more sensitive than cytology for detecting ACs, only a small fraction of high-risk HPV-positive women will develop ACs. Therefore, further stratification of this risk group using additional biomarkers is needed to enable the gynecologist to decide whether to use endocervical curettage or a cone biopsy. This study corroborates previous findings that marker genes that are ated in SCCs are also ated in ACs. PAX1, PTPRR, SOX1, and ZNF582 were highly ated in ACs of cervical tissues and can be detected in cervical scrapings. PAX1, PTPRR, SOX1, and ZNF582, primarily identified from SCCs, 11,15,16 were all highly ated in ACs. The fact that these ation markers discovered from SCCs can also be detected in ACs at high frequency is important clinically and justifies the testing of these DNA ation patterns in future population-based studies for the detection of cervical lesions including ACs. The consistently high ation rates of PAX1, PTPRR, SOX1, and ZNF582 in both cell types not only extend the clinical applications of these biomarkers but also shed some light on the common origin of both cell types. It is common to see mixed SCCs and ACs (as adenosquamous carcinomas) in cervical cancers. The reasons for this concurrence remain largely unknown. The PAX1 and SOX1 proteins are transcription factors important in developmental processes. 19,20 Moreover, the expression of SOX1 attenuated the tumorigenic potential of neuronal precursors after neural stem cell transplantation, suggesting a role in stem cell differentiation. 21 Most members of the PAX family of proteins are important in self-renewal in embryogenesis. The concept of cancer * 2014 IGCS and ESGO 207

8 Chang et al International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 TABLE 2. High ation levels of 4 genes and their related OR in ACs of the cervix (from scrapings) Gene Normal Scraping (n = 22) AC Scraping (n = 23) OR* P* PAX (2.66Y15.49) G0.001 L PTPRR (3.84Y46.37) G0.001 L SOX (2.63Y15.79) G0.001 L ZNF (6.97Y77.5) G0.001 L *Adjusted age. stem cells or tumor-initiating cells suggests the ability of selfrenewal in tumors that prevents them from differentiating into heterogeneous nontumorigenic cancer cell types. 22 Although solid evidence is lacking to date, subcolumnar reserve cells emerged as the best candidate for cervical stem cells. 23 We speculate that ation of these differentiation-related genes at earlier stages of cervical carcinogenesis in reserve cells might explain why ation of the same genes can be detected in both cell types at high rates. The other possibility is that these genes might participate in some events in cervical carcinogenesis common to the development of both SCCs and ACs, especially late in the process of cancer invasion. 15,18 Further studies of DNA ation in cervical cancer stem cells might provide new insights into cervical carcinogenesis. 24 The functions of these genes in cancer biology are largely unexplored. PTPRR encodes a phosphatase that might counteract important kinase activities in cancer development. Indeed, our previous study revealed for the first time that PTPRR could target p44/42 MAP kinase and inhibit invasive phenotypes of cervical cancers. 15 SOX1 is a transcription factor. A recent report demonstrated its tumor-suppressor effects by direct binding to A-catenin and inhibiting Wnt signaling in liver cancer. 25 So far, there are no known functional roles for PAX1 and ZNF582 in human cancers. 20 The PAX gene family has been classified into 4 groups according to the sequencing motifs. PAX1 is categorized in group 1 without the homeodomain and is implicated as a tumor suppressor according to our studies. 11,12,26,27 Functional studies of PAX1 in cervical cancer are ongoing and might help to improve our understanding of cervical carcinogenesis. In summary, we report here 4 genes that are highly ated in SCCs and are also highly ated in ACs and can be detected in cervical scrapings. These results support the use of DNA ation as a biomarker for cervical cancer screening, either as an adjunct of cytology in developed countries or as a molecular method for cervical cancer screening in 208 developing or underdeveloped regions. More tests of these DNA ations in different infrastructures around the world, along with cytology, HPV testing, or self-sampling, are warranted. REFERENCES 1. Forouzanfar MH, Foreman KJ, Delossantos AM, et al. Breast and cervical cancer in 187 countries between 1980 and 2010: a systematic analysis. Lancet. 2011;378:1461Y Wang SS, Sherman ME, Hildesheim A, et al. Cervical adenocarcinoma and squamous cell carcinoma incidence trends among white women and black women in the United States for 1976Y2000. Cancer. 2004;100:1035Y Visioli CB, Zappa M, Ciatto S, et al. Increasing trends of cervical adenocarcinoma incidence in Central Italy despite Extensive Screening Programme, 1985Y2000. Cancer Detect Prev. 2004;28:461Y Bulk S, Visser O, Rozendaal L, et al. Cervical cancer in the Netherlands 1989Y1998: Decrease of squamous cell carcinoma in older women, increase of adenocarcinoma in younger women. Int J Cancer. 2005;113:1005Y Sasieni P, Castanon A, Cuzick J. Screening and adenocarcinoma of the cervix. Int J Cancer. 2009;125:525Y Bosch FX, Lorincz A, Muñoz N, et al. The causal relation between human papillomavirus and cervical cancer. J Clin Pathol. 2002;55:244Y Castellsagué X, Díaz M, de Sanjosé S, et al. Worldwide human papillomavirus etiology of cervical adenocarcinoma and its cofactors: implications for screening and prevention. J Natl Cancer Inst. 2006;98:303Y Szalmás A, Kónya J. Epigenetic alterations in cervical carcinogenesis. Semin Cancer Biol. 2009;19:144Y Wilting SM, Snijders PJ, Meijer GA, et al. Increased gene copy numbers at chromosome 20q are frequent in both squamous cell carcinomas and adenocarcinomas of the cervix. J Pathol. 2006;209:220Y Sova P, Feng Q, Geiss G, et al. Discovery of novel ation biomarkers in cervical carcinoma by global deation and * 2014 IGCS and ESGO

9 International Journal of Gynecological Cancer & Volume 24, Number 2, February 2014 Methylation in Cervical Adenocarcinoma microarray analysis. Cancer Epidemiol Biomarkers Prev. 2006;15:114Y Lai HC, Lin YW, Huang TH, et al. Identification of novel DNA ation markers in cervical cancer. Int J Cancer. 2008;123:161Y Lai HC, Lin YW, Huang RL, et al. Quantitative DNA ation analysis detects cervical intraepithelial neoplasms type 3 and worse. Cancer. 2010;116:4266Y Kang S, Kim JW, Kang GH, et al. Comparison of DNA hyperation patterns in different types of uterine cancer: cervical squamous cell carcinoma, cervical adenocarcinoma and endometrial adenocarcinoma. Int J Cancer. 2006;118:2168Y van der Meide WF, Snellenberg S, Meijer CJ, et al. Promoter ation analysis of WNT/A-catenin signaling pathway regulators to detect adenocarcinoma or its precursor lesion of the cervix. Gynecol Oncol. 2011;123:116Y Su PH, Lin YW, Huang RL, et al. Epigenetic silencing of PTPRR activates MAPK signaling, promotes metastasis and serves as a biomarker of invasive cervical cancer. Oncogene. 2013;32:15Y Huang RL, Chang CC, Su PH, et al. Methylomic analysis identifies frequent DNA ation of zinc finger protein 582 (ZNF582) in cervical neoplasms. PLoS One. 2012;7:e Chu TY, Hwang KS, Yu MH, et al. A research-based tumor tissue bank of gynecologic oncology: characteristics of nucleic acids extracted from normal and tumor tissues from different sites. Int J Gynecol Cancer. 2002;12:171Y Liu CY, Chao TK, Su PH, et al. Characterization of LMX-1A as a metastasis suppressor in cervical cancer. J Pathol. 2009;219:222Y Ekonomou A, Kazanis I, Malas S, et al. Neuronal migration and ventral subtype identity in the telencephalon depend on SOX1. PLoS Biol. 2005;3:e Robson EJ, He SJ, Eccles MR. A PANorama of PAX genes in cancer and development. Nat Rev Cancer. 2006;6:52Y Fukuda H, Takahashi J, Watanabe K, et al. Fluorescenceactivated cell sorting-based purification of embryonic stem cell-derived neural precursors averts tumor formation after transplantation. Stem Cells. 2006;24:763Y Nguyen LV, Vanner R, Dirks P, et al. Cancer stem cells: an evolving concept. Nat Rev Cancer. 2012;12:133Y Todaro M, Lombardo Y, Stassi G. Evidences of cervical cancer stem cells derived from established cell lines. Cell Cycle. 2010;9:1238Y Steenbergen RD, Ongenaert M, Snellenberg S, et al. Methylation-specific digital karyotyping of HPV16E6E7-expressing human keratinocytes identifies novel ation events in cervical carcinogenesis. J Pathol. 2013;231:53Y Tsao CM, Yan MD, Shih YL, et al. SOX1 functions as a tumor suppressor by antagonizing the WNT/A-catenin signaling pathway in hepatocellular carcinoma. Hepatology. 2012;56:2277Y Huang TH, Lai HC, Liu HW, et al. Quantitative analysis of ation status of the PAX1 gene for detection of cervical cancer. Int J Gynecol Cancer. 2010;20:513Y Chao TK, Ke FY, Liao YP, et al. Triage of cervical cytological diagnoses of atypical squamous cells by DNA ation of paired boxed gene 1 (PAX1). Diagn Cytopathol. 2013;41:41Y46. * 2014 IGCS and ESGO 209