Clinical Biochemistry

Clinical Biochemistry 43 (2010) 1443 1448 Contents lists available at ScienceDirect Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem Rapid and reliable detection of LINE-1 hypomethylation using high-resolution melting analysis Stefanie Stanzer a, Marija Balic a, Jasmin Strutz a, Ellen Heitzer a, Florian Obermair a, Cornelia Hauser-Kronberger b, Hellmut Samonigg a, Nadia Dandachi a, a Division of Oncology, Department of Internal Medicine, Medical University of Graz, Auenbruggerplatz 15, 8036, Graz, Austria b Department of Pathology, General Hospital and Paracelsus Medical University Salzburg, Muellner Hauptstrasse 48, A-5020 Salzburg, Austria article info abstract Article history: Received 15 July 2010 Received in revised form 19 August 2010 Accepted 17 September 2010 Available online 27 September 2010 Keywords: Hypomethylation LINE-1 Repetitive elements High-resolution melting MethyLight Pyrosequencing Archival prostate tumor samples Objectives: The aim of the present study was to evaluate the precision and reproducibility of the LINE-1 high-resolution melting (HRM) assay to detect LINE-1 hypomethylation. Design and methods: We first evaluated a methylated DNA dilution matrix and a panel of human cancer cell lines. We then applied this LINE-1 HRM assay to a set of 37 archival prostate cancer tissue samples. Results: Our LINE-1 HRM assay revealed small and reproducible run-to-run and bisulfite-to-bisulfite variations. As expected, we found a large variation in methylation levels between different cancer cell lines. All results were confirmed with MethyLight and pyrosequencing as indicated by the high correlation coefficient. Finally, we successfully applied the LINE-1 HRM assay to archival prostate cancer tissues. Conclusions: The present LINE-1 HRM assay represents a novel, accurate, and cost-effective method to measure global hypomethylation, which makes it suitable for high- and low-throughput laboratories. 2010 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. Introduction Cancer is a complex disease arising from both genetic and epigenetic alterations. Most molecular mechanisms leading to genetic alterations in cancer are well identified. Although it is known that epigenetic changes play an important role in cancer diseases, epigenetic mechanisms are still poorly understood. One of the best studied epigenetic DNA modifications is DNA methylation where CpG (cytosine phosphate guanine) dinucleotides are often methylated via the action of DNA methyltransferases. Two types of aberrant DNA methylation patterns coexist in human cancer cells including global hypomethylation and local gene-specific hypermethylation [1 3]. In the human genome, repetitive elements (e.g., LINE-1 elements) comprise about 45% of human genomic DNA and are commonly highly methylated. Global hypomethylation of these repetitive elements is a common characteristic in human cancer cells [3 5].LINE-1 elements are long interspersed nucleotide elements (=LINEs) of up to 6 kb and comprise almost 21% of the human genome. These elements are usually methylated, and their transcription and retrotransposition are suppressed by a variety of control mechanisms including methylation or small This work was supported by the Austrian National Bank Fund (grant 12680 to N.D.). Corresponding author. Fax: +43 316 385 4167. E-mail address: nadia.dandachi@medunigraz.at (N. Dandachi). interfering RNA [6]. LINE-1 elements are frequently hypomethylated in various human cancers including colorectal, prostate, lung, or gastric cancer [7 12]. In human cancer, hypomethylation of LINE-1 elements can lead to its reactivation which can induce genomic instability [13,14]. The methylation status of LINE-1 elements is thought to provide a good indicator to measure global hypomethylation in tumor cells and may therefore serve as a biomarker for early cancer diagnostics, prediction of prognosis, and response to treatment. Thus, there is great demand for reliable and broadly accessible techniques to detect hypomethylation in human cancers. Several methods have been used to study the methylation status of LINE-1 elements, including Southern blot analyses, high-performance liquid chromatography, COBRA (Combined Bisulfite Restriction Analysis), pyrosequencing, and MethyLight [14 18]. All these methods have several advantages and disadvantages, and the choice of assay depends on the requirements. Some of these methods are very labor- and time-intensive, and some require large amounts of genomic DNA. Pyrosequencing has the advantage of analyzing single CpG sites in a DNA pool, but it requires specialized equipment. MethyLight, a fluorescence-based real-time polymerase chain reaction (PCR) analysis technique, has the advantage of being a closedtube method. This reduces the risk of contamination. However, MethyLight is expensive due to the requirement of TaqMan probes and the need for a reference gene for normalization. High Resolution Melting (HRM) technology is a novel approach that has recently been reported as a rapid and robust analysis tool for the 0009-9120/$ see front matter 2010 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2010.09.013

1444 S. Stanzer et al. / Clinical Biochemistry 43 (2010) 1443 1448 detection of DNA hypermethylation [19,20]. HRM offers several advantages over the widely used MethyLight assay as shown recently by our group [20]. In this study, we demonstrated that HRM is a reliable and broadly accessible method for determining DNA hypermethylation, and it is adjustable to analyze any gene region and sample type. In the current study, we extended and applied the HRM technology to analyze DNA methylation of the LINE-1 element. Therefore, we comprehensively evaluated the LINE-1 HRM assay in a methylated DNA dilution matrix and in a panel of human cancer cell lines. All results generated by HRM were validated with the Methy- Light and the pyrosequencing assay. Finally, we applied this LINE-1 HRM assay to a set of FFPE (formalin-fixed and paraffin-embedded) prostate tissue samples. To our knowledge, this is the first report using HRM technology to detect the methylation status of LINE-1 elements. Materials and methods Controls, cancer cell lines and tumor samples Universal methylated DNA (Chemicon, Temecula, CA) was arbitrarily used as 100% methylated control DNA. However, it is noteworthy that pyrosequencing analysis revealed only a mean methylation level of 70% (data not shown). A standard dilution series of 100%, 75%, 50%, 25%, 10%, and 0% methylated DNA in a background of universal unmethylated DNA (Qiagen, Hilden, Germany) was constructed by serially diluting the methylated control DNA into the unmethylated control. The standard dilution series from 100% to 0% methylated DNA was also used to assess precision and performance characteristics of the assays. In addition, this study includes a panel of cancer cell lines for validation experiments (MDA-MB-231, SK-BR-3, T-47D, MDA-MB-453, MCF-7, HT-29, SW-480, SK-MEL-30, SK-MEL-28, PC-3, DU 145, LNCaP, and Hep G2). Two non-tumorigenic breast cell lines (MCF-10A and MCF-12A) were also used. Besides SUM-159 and SUM-1315, which were obtained from Asterand (Detroit, MI, USA), all other cell lines were purchased from American Type Culture Collection (Manassas, VA, USA). All cell lines were cultured according to supplier's recommendations. We used cell lines obtained either freshly harvested from cultures or after formalin fixation and paraffin embedding, as adapted from the published protocol by Kerstens et al. [21]. The methylation level of each sample was evaluated by comparing the PCR product melting profiles of all samples and the PCR product melting profiles of the standards. Aliquots from the same DNA were utilized for the MethyLight and pyrosequencing assays. FFPE blocks from 37 prostate cancer patients were collected between 1998 and 2009 in the Department of Pathology, Hospital and Paracelsus Medical University Salzburg and used as a set. Only patients with pt3 tumors (N0 and N1) were included. The study was approved by the Institutional Review Board. Extraction of genomic DNA High molecular weight genomic DNA from cultured cell lines was extracted using the QIAamp DNA Blood Midi Kit (Qiagen) according to the manufacturer's protocol. Hematoxylin-and-eosin-stained sections of FFPE tissue were reviewed for each sample to identify representative tumor areas. FFPE tissue blocks were microdissected using a 4 mm punch to ensure areas with N70% tumor cells for each case. Genomic DNA was extracted using the QIAamp DNA FFPE tissue kit (Qiagen) according to the manufacturer's protocol. Sodium bisulfite modification For the methylation analysis, 1 μg of genomic DNA was subjected to bisulfite conversion with the EpiTec Bisulfite Kit (Qiagen) according to the manufacturer's instructions. The purified bisulfite converted samples were eluted in 40 μl volume and stored at 20 C. HRM assay We used 5 UTR of LINE 1.2 sequence from NCBI (National Center for Biotechnology Information) accession number M80343 (Table 1) representing genome-wide LINE-1 sequences [4]. LINE-1 HRM Primers were designed using Methyl Primer Express Software (Applied Biosystems, Foster City, CA, USA) and as recommended by Wojdacz et al. [22]. PCR amplification and HRM were performed on the LightCycler 480 (Roche Applied Science, Penzberg, Germany). PCR was performed in a 20 μl volume consisting of 1 LightCycler 480 High Resolution Melting Master Mix (Roche), 200 nmol/l of each primer for LINE-1 and 10 ng of bisulfite-treated DNA template. The final MgCl 2 concentration was adjusted to 3 mmol/l. The cycling protocol started with one cycle of 95 C for 10 min followed by 50 cycles of 95 C for 10 s, a touch down of 63 C to 58 C for 10 s (1 C/cycle), 72 C for 20 s, and a HRM step of 95 C for 1 min, 40 C for 1 min, and 70 C for 5 s. The melting step was performed using a continuous acquisition to 90 C at 25 acquisitions per 1 C. Each plate included multiple water blanks. A standard curve with known methylation ratios was included in each assay. Methylation standards and cell lines were performed in duplicates. FFPE samples were performed in triplicates. HRM data were analyzed using the Gene Scanning Software (Roche). Data processing included normalization and temperature shifting using the LightCycler Software. Raw data were exported from the LightCycler and used to generate a standard curve by linear regression analysis. This standard curve was then used to determine the methylation level of the samples. We applied Microsoft Excel 2007 software for this analysis. MethyLight assay Aliquots from the same DNA were analyzed by the MethyLight assay. The PCR primers and probes are listed in Table 1. ALU primers and probes have been used as previously published [17] and were used to normalize for the amount of input DNA. The MethyLight PCR for ALU was performed in a 20 μl reaction volume with 1 LightCycler 480 Probe Master, 0.4 ng bisulfite modified DNA, 500 nmol/l forward and reverse PCR primers and 200 nmol/l probe. The LINE-1 MethyLight assay consisted of 1 LightCycler 480 Probe Master, 10 ng bisulfite-modified DNA, 400 nmol/l forward and reverse PCR primers and 200 nmol/l probe. The cycling conditions were as follows: mono-color hydrolyses probes detection format, 1 cycle of 95 C for 10 min, 45 cycles of 95 C for 10 s, 60 C (for LINE-1) or 52 C (for ALU) for 30 s, and 72 C for 1 s. According to the LINE-1 HRM assay, methylation standards and cell lines were performed in duplicates and FFPE samples in triplicates. Pyrosequencing PCR and pyrosequencing for LINE-1 methylation were performed as previously published by Daskalos et al. [7] using PyroMark PCR Kit Table 1 LINE-1 HRM primers. HRM LINE-1 ML LINE-1 Primer sequence 5 3 FW GGGGGAGGAGTTAAGATGGT RV AAATACAAAAATCACCCGTCTTCTA ML Fw GTTAGATAGTGGGCGTAGGTTATTG ML Rv TTACGCTTCCCAAATAAAACAATACC ML Probe FAM CCTACTTCGACTCGCGCACGATACG TAMRA GenBank accession number M80343 M80343

S. Stanzer et al. / Clinical Biochemistry 43 (2010) 1443 1448 1445 (Qiagen). Briefly, PCR was carried out in a 25 μl reaction containing 20 ng of bisulfite-treated genomic DNA, PyroMark PCR Master Mix, 200 nmol/l Primer, and 1.5 mmol/l MgCl 2. Cycling conditions were as follows: 15 min 95 C followed by 45 cycles of 30 s 94 C, 30 s 56 C, 30 s 72 C, and a final extension step at 72 C for 10 min. The PCR products (5 μl) were analyzed by electrophoresis in a 2% agarose gel to confirm successful amplification. The remaining 20 μl was purified and made single-stranded to act as a template in a pyrosequencing reaction as recommended by the manufacturer using the PyroMark Q24 Vacuum Workstation (Qiagen). PCR products were immobilized using Streptavidin Sepharose High Performance (GE Healthcare), washed, and denaturated. Annealing of 300 nmol/l pyrosequencing Primer was performed at 80 C for 2 min; pyrosequencing was done in a PyroMark Q24 (Qiagen). The ratio of C to T nucleotides at each analyzed CpG site was evaluated for six CpGs, reflecting the proportion of methylated LINE-1 DNA. Furthermore, we used built-in controls for the bisulfite treatment to prevent false-positive methylation detection. Statistical analysis Descriptive statistics, such as mean, standard deviation (SD), and range, were calculated for each experiment. Student's t-test and Pearson correlation were used to compare test results. Analyses were performed with the SPSS v.17 Statistical Package (SPSS, Inc., Chicago, IL). Results LINE-1 HRM assay and dilution matrix We designed LINE-1 HRM primers that amplify an 89-bp product of bisulfite-treated DNA from LINE-1 elements. Short amplicon size is a critical factor in HRM analysis, especially when analyzing FFPE samples. The amplified PCR product consists of a pool of unmethylated and methylated amplicons, which can be discriminated by HRM due to their distinct melting profiles. First, we validated the reliability and reproducibility of our LINE-1 HRM assay by analyzing different sets of bisulfite converted methylation standards (n=4), tested in independent HRM assays on different days (n =14). We obtained good reproducibility in methylation standards of different bisulfite treatments with small inter-assay variations. Linear regression analysis (Fig. 1) revealed that LINE-1 methylation levels were highly correlated with levels predicted by input fraction of methylated DNA (mean r=0.994, pb0.001). However, linear regression also showed that the linear dynamic range of the LINE-1 HRM assay was moderate. As shown in Fig. 1, 100% methylated DNA only gave an average normalized relative signal of 19.3 (±1.8). This modest dynamic range is also demonstrated by decreased and consolidated melting curves (Fig. 3). We also found this underestimation of fully methylated DNA when using MethyLight and pyrosequencing (data not shown). Cancer cell lines In a second step, we evaluated our LINE-1 HRM assay by analyzing a panel of cancer cell lines, two non-tumorigenic breast cell lines (MCF-10A, MCF-12A) and peripheral blood leukocytes from seven healthy volunteers (Fig. 2). DNA from healthy volunteers, MCF-10A and MCF-12A, showed similar high LINE-1 methylation levels (mean 94.3, range 89.5 99.6). In contrast and as expected, cancer cell lines revealed high variability in LINE-1 methylation, ranging from 24.4% to 101.5% methylation. In some cancer cell lines (e.g., DU-145, 101.5%), LINE-1 methylation levels resembled that of the peripheral blood cells, while other cancer cell lines showed LINE-1 hypomethylation (e.g., SK-BR-3, 24.4%). Representative melting curves for methylation standards and several cancer cell lines are presented in Fig. 3. Fig. 1. Linear regression analysis shows that LINE-1 methylation levels were highly associated with levelspredicted by input fraction ofmethylated DNA (r=0.994,pb0.001). To validate the HRM LINE-1 assay, all samples were also analyzed with MethyLight and pyrosequencing assays. Pearson correlation indicated a high and a statistically significant agreement between all three assays (Table 2). Next, we wanted to test the influence of formalin fixation and paraffin embedding on measuring LINE-1 methylation using HRM analysis. To test this influence, we compared results from genomic DNA extracted from freshly harvested and FFPE cells from 10 of the analyzed cell lines. Linear regression analysis showed an excellent correlation between genomic DNA extracted from freshly harvested and FFPE cells (Fig. 4). Prostate cancer samples Finally, to evaluate the application of our LINE-1 HRM assay, we selected a panel of primary prostate cancer samples (n=37) to analyze LINE-1 hypomethylation in lymph node negative and lymph node positive primary prostate tumor samples. The mean LINE-1 methylation level was 87.6 (±6.3) in lymph-node-negative and 84.2 (±9.9) in lymph-node-positive tumor samples. We found no statistically significant difference in LINE-1 methylation levels between lymph-node-negative and lymph-node-positive primary prostate tumors using both LINE-1 HRM (data not shown) and MethyLight assays. We achieved a good correlation between Methy- Light and HRM results (r=0.673, pb0.001). Discussion In this study, we report on an extended application of HRM to reliably and quickly detect global hypomethylation changes. To our knowledge, this is the first report demonstrating the applicability of HRM to detect LINE-1 methylation. This HRM assay can also be expanded to analyze other repetitive elements such as ALU or SAT2 to detect global hypomethylation. We evaluated variability of bisulfite treatment and HRM assay by comparing methylation results obtained from different sets of bisulfite converted methylation standards, tested in independent assays on different days. Resulting data showed small and reproducible run-to-run and bisulfite-to-bisulfite variations. Low impact of bisulfite variation on LINE-1 methylation was also shown by Irahara et al. [15]. Clinical samples undergo different bisulfite treatments making low bisulfite-to-bisulfite variations between assays important to reliably detect LINE-1 methylation. Precision and reproducibility are also important, because differences in LINE-1 methylation levels between distinct subgroups can be small. We observed a low dynamic range for both HRM and MethyLight assays in detecting LINE-1 methylation. This observation has also been made by Richards et al. [14]. The explanation is that spontaneous

1446 S. Stanzer et al. / Clinical Biochemistry 43 (2010) 1443 1448 Fig. 2. Variable LINE-1 methylation levels in various cancer cell lines (n=15), two non-tumorigenic breast cell lines (n=2), and peripheral blood leukocytes from healthy volunteers (n=7). Fig. 3. Melting curves for (A) methylation standards and (B) methylation standards and representative cancer cell lines.

S. Stanzer et al. / Clinical Biochemistry 43 (2010) 1443 1448 1447 Table 2 Pearson correlation between LINE-1 methylation results detected by three different methylation assays using cancer cell lines. Pearson correlation coefficient HRM MethyLight Pyrosequencing HRM 1 MethyLight 0.949* 1 Pyrosequencing 0.943* 0.937* 1 *pb0.001. deamination of 5-methylcytosine to thymine occurs frequently in the human genome [23]. Methylation assays that are based on bisulfite conversion, as our present HRM assay, cannot distinguish between TpG from either mutation of 5-methylcytosine or conversion of an unmethylated cytosine. Therefore, sequence variations exist among different individuals resulting in higher background noise. This fact should be kept in mind when analyzing LINE-1 methylation. Using HRM to detect LINE-1 hypomethylation, we found a large variation in methylation levels between different cancer cell lines. Some cell lines showed methylation levels similar to those detected in normal cell lines and peripheral blood. In contrast, other cell lines showed high levels of hypomethylation. This large variation in global methylation has already been shown by other groups, both in cell lines and in tumor tissues [4,13,14] and severe hypomethylation has been suggested to represent a selective advantage for tumors [24]. LINE-1 HRM results were confirmed with MethyLight and pyrosequencing as indicated by the high correlation coefficient. In an attempt to evaluate the influence of formalin fixation and paraffin embedding on methylation analysis, we demonstrated that the reproducibility of LINE-1 methylation results was comparable between DNA from freshly harvested and FFPE cells. Our data show that normalized melting curves from both freshly harvested and FFPE cells reveal the same quality. However, genomic DNA extracted from FFPE cells do not fully reflect the situation in archival FFPE tissue samples, because formalin-induced DNA damage is caused by both formalin fixation and storage conditions. It has been shown that DNA hypomethylation occurs late and heterogeneously in prostate cancer progression [11].Additionally,LINE- 1 hypomethylation has been significantly associated with lymph node involvement [5,25]. Therefore, we evaluated the applicability of the LINE-1 HRM assay in prostate cancer. We analyzed a set of 37 pt3 prostate lymph-node-positive and -negative tumor samples. To underpin the value of LINE-1 HRM results, we concomitantly performed the analysis by MethyLight. We found a good concordance between MethyLight and HRM results. The lower correlation coefficient between Fig. 4. Pearson correlation of LINE-1 HRM results obtained from DNA extracted from freshly harvested and FFPE cells (r=0.980, pb0.001). MethyLight and LINE-1 HRM assay in primary tumor samples may be partly caused by contamination with normal tissue despite the fact that microdissection has been performed. Also, it is generally known that analysis of archival FFPE tissue samples compared to cell line model systems is more susceptible to measurement variations. Although we were able to show that our LINE-1 HRM assay can precisely detect small changes in hypomethylation, we could not find a statistically significant difference between lymph-node-negative and -positive tumor samples. One reason could be the small sample size. Another reason could be that methylation levels of LINE-1 elements are different at different loci [16]. Thus, analyzing a different locus could have revealed different results. However, we are aware of the limitation of the present analysis due to the lack of comparison of the tumor methylation levels to the corresponding normal tissue. Such analysis will be of utmost importance in future analysis of the biological significance of hypomethylation using HRM-based assays, which was not the goal of the present study. Several methods have been applied to detect global hypomethylation, and all of them have advantages and disadvantages. One important disadvantage of COBRA and pyrosequencing is that both methods require post-pcr manipulation. In contrast, MethyLight and HRM assay do not require post-pcr processing, thereby reducing the potential contamination problem. An advantage of HRM assays is the use of a saturating fluorescent dye, making the analysis cost-effective compared to alternative approaches using labeled probes such as MethyLight. A disadvantage of our MethyLight assay is that partial LINE-1 methylation may not be identified, since primer and probes specifically target sequences containing multiple fully methylated CpGs. In contrast, with HRM even heterogeneous methylation patterns can be identified as seen by the shape of the melting curve [20,22]. Conclusion The present LINE-1 HRM assay represents a novel, accurate, and cost-effective method to measure global hypomethylation, which makes it suitable for high- and low-throughput laboratories. This approach can easily be adapted to the analysis of other repetitive elements and sample types. The accessibility and ease of this approach will facilitate further global methylation studies. Acknowledgments The authors wish to thank Lucy Smith for critical reading of the manuscript. References [1] Esteller M. Epigenetics in cancer. N Engl J Med 2008;358(11):1148 59. [2] Kristensen LS, Nielsen HM, Hansen LL. Epigenetics and cancer treatment. Eur J Pharmacol 2009;625(1 3):131 42. [3] Wilson AS, Power BE, Molloy PL. DNA hypomethylation and human diseases. Biochim Biophys Acta 2007;1775(1):138 62. [4] Chalitchagorn K, Shuangshoti S, Hourpai N, Kongruttanachok N, Tangkijvanich P, Thong-ngam D, et al. Distinctive pattern of LINE-1 methylation level in normal tissues and the association with carcinogenesis. Oncogene 2004;23(54):8841 6. [5] Ehrlich M. DNA hypomethylation in cancer cells. Epigenomics 2009;1(2):239 59. [6] Baylin SB, Ohm JE. Epigenetic gene silencing in cancer a mechanism for early oncogenic pathway addiction? Nat Rev Cancer 2006;6(2):107 16. [7] Daskalos A, Nikolaidis G, Xinarianos G, Savvari P, Cassidy A, Zakopoulou R, et al. Hypomethylation of retrotransposable elements correlates with genomic instability in non-small cell lung cancer. Int J Cancer 2009;124(1):81 7. [8] Ogino S, Kawasaki T, Nosho K, Ohnishi M, Suemoto Y, Kirkner GJ, et al. LINE-1 hypomethylation is inversely associated with microsatellite instability and CpG island methylator phenotype in colorectal cancer. Int J Cancer 2008;122(12): 2767 73. [9] Park SY, Yoo EJ, Cho NY, Kim N, Kang GH. Comparison of CpG island hypermethylation and repetitive DNA hypomethylation in premalignant stages of gastric cancer, stratified for Helicobacter pylori infection. J Pathol 2009;219(4): 410 6. [10] Saito K, Kawakami K, Matsumoto I, Oda M, Watanabe G, Minamoto T. Long interspersed nuclear element 1 hypomethylation is a marker of poor prognosis in stage IA non-small cell lung cancer. Clin Cancer Res 2010;16(8):2418 26.

1448 S. Stanzer et al. / Clinical Biochemistry 43 (2010) 1443 1448 [11] Yegnasubramanian S, Haffner MC, Zhang Y, Gurel B, Cornish TC, Wu Z, et al. DNA hypomethylation arises later in prostate cancer progression than CpG island hypermethylation and contributes to metastatic tumor heterogeneity. Cancer Res 2008;68(21):8954 67. [12] Ogino S, Nosho K, Kirkner GJ, Kawasaki T, Chan AT, Schernhammer ES, et al. A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer. J Natl Cancer Inst 2008;100(23):1734 8. [13] Estecio MR, Gharibyan V, Shen L, Ibrahim AE, Doshi K, He R, et al. LINE-1 hypomethylation in cancer is highly variable and inversely correlated with microsatellite instability. PLoS ONE 2007;2(5):e399. [14] Richards KL, Zhang B, Baggerly KA, Colella S, Lang JC, Schuller DE, et al. Genome-wide hypomethylation in head and neck cancer is more pronounced in HPV-negative tumors and is associated with genomic instability. PLoS ONE 2009;4(3):e4941. [15] Irahara N, Nosho K, Baba Y, Shima K, Lindeman NI, Hazra A, et al. Precision of pyrosequencing assay to measure LINE-1 methylation in colon cancer, normal colonic mucosa, and peripheral blood cells. J Mol Diagn 2010;12(2):177 83. [16] Phokaew C, Kowudtitham S, Subbalekha K, Shuangshoti S, Mutirangura A. LINE-1 methylation patterns of different loci in normal and cancerous cells. Nucleic Acids Res 2008;36(17):5704 12. [17] Weisenberger DJ, Campan M, Long TI, Kim M, Woods C, Fiala E, et al. Analysis of repetitive element DNA methylation by MethyLight. Nucleic Acids Res 2005;33 (21):6823 36. [18] Yang AS, Estecio MR, Doshi K, Kondo Y, Tajara EH, Issa JP. A simple method for estimating global DNA methylation using bisulfite PCR of repetitive DNA elements. Nucleic Acids Res 2004;32(3):e38. [19] Wojdacz TK, Dobrovic A. Methylation-Sensitive High Resolution Melting (MS-HRM): a new approach for sensitive and high-throughput assessment of methylation. Nucleic Acids Res 2007;35(6):e41. [20] Balic M, Pichler M, Strutz J, Heitzer E, Ausch C, Samonigg H, et al. High quality assessment of DNA methylation in archival tissues from colorectal cancer patients using quantitative high-resolution melting analysis. J Mol Diagn 2009;11(2):102 8. [21] Kerstens HM, Robben JC, Poddighe PJ, Melchers WJ, Boonstra H, de Wilde PC, et al. AgarCyto: a novel cell-processing method for multiple molecular diagnostic analyses of the uterine cervix. J Histochem Cytochem 2000;48(5):709 18. [22] Wojdacz TK, Hansen LL, Dobrovic A. A new approach to primer design for the control of PCR bias in methylation studies. BMC Res Notes 2008;1:54. [23] Laird PW. The power and the promise of DNA methylation markers. Nat Rev Cancer 2003;3(4):253 66. [24] Cho NY, Kim JH, Moon KC, Kang GH. Genomic hypomethylation and CpG island hypermethylation in prostatic intraepithelial neoplasm. Virchows Arch 2009;454 (1):17 23. [25] Santourlidis S, Florl A, Ackermann R, Wirtz HC, Schulz WA. High frequency of alterations in DNA methylation in adenocarcinoma of the prostate. Prostate 1999;39 (3):166 74.