ANTICANCER RESEARCH 33: 4651-4656 (2013)

Serum DNA Hypermethylation in Patients with Kidney Cancer: Results of a Prospective Study

STEFAN HAUSER, TOBIAS ZAHALKA, GUIDO FECHNER, STEFAN C. MÜLLER and JÖRG ELLINGER

Department of Urology, University of Bonn, Bonn, Germany

Abstract. Aim: No reliable biomarker for renal cell The incidence of renal cell carcinoma (RCC) in Europe is more carcinoma (RCC) exists. The purpose of this study was to than 40,000 per year (1). RCC is accountable for more than analyze the value of CpG island hypermethylation of cell- 20,000 deaths per year in Europe (1). Since 1980, the incidence free (cf) circulating serum DNA in patients with RCC as a of kidney cancer has doubled, whereas the mortality rate has potential biomarker. Patients and Methods: In total 35 decreased (2). Today, most tumors are found coincidentally patients with RCC and 54 healthy individuals were enrolled during radiological investigation; a reliable biomarker does not in this study. Cell-free DNA (cFDNA) in serum was isolated yet exist. With the new armamentarium of multi-tyrosine kinase and digested with methylation-sensitive restriction enzymes inhibitor and mammalian Target of Rapamycin (mTOR)- (Bsh1236I, HpaII and HinP1I) to quantify the amount of inhibitors, it is often not possible to determine treatment methylated Adenomatosis-poliposis-coli gene (APC), response, therefore a biomarker could help to distinguish Gluthation-a-transferase-protein 1 gene (GSTP1), ARF between patients with response from those with progressive tumor suppressor protein gene (p14(ARF)), cyclin-dependent disease. Changes in the CpG island methylation pattern are kinase inhibitor 2A (p16), Retinoid-acid-receptor-beta gene widely found in cancer cells, both generalized DNA (RAR-B), RAS-association domain family-1 gene (RASSF1), hypomethylation and regional CpG island hypermethylation are Tissue inhibitor of metalloproteinase- gene (TIMP3) and common findings (3). Earlier studies demonstrated that the Prostaglandin-endoperoxid synthase 2 (PTGS2) DNA detection of DNA hypermethylation allows for normal tissue fragments. Results: In 30 of 35 investigated patients with to be distinguished from malignant renal tissue, and also allows RCC, at least one gene was methylated within the serum prognostic conclusions to be drawn (4-8). In a prior article cfDNA. The methylation frequency ranged from 14.3% for published in this journal, we showed that cell-free DNA p14(ARF) to 54.3% for APC. All genes, except p16 and (cfDNA) in serum is elevated in patients with RCC (9); an TIMP3, were significantly more frequently methylated in increased cfDNA level seems to be a hallmark of human patients with RCC compared to healthy individuals. Receiver malignancies. The origin of cfDNA remains to be clarified, but operator characteristic analysis showed a high specificity for at least a small amount of this DNA is derived from the tumor serum cfDNA methylation [between 85.2% for RAR-B and itself (10). Thus, in patients with increased cfDNA levels, the 100% for p14(ARF)], but the sensitivity was low in single- analysis of tumor-specific alterations may help to distinguish gene analysis [range-14.3% for p14(ARF) to 54.3% for between various tumor entities, and between patients with APC]. The combined analysis of multiple genes increased the benign and malignant disease. We, therefore, analyzed whether diagnostic sensitivity (i.e. APC, PTGS2 and GSTP1, 62.9%) cfDNA is hypermethylated in eight genes namely methylated at a high specificity (87%). DNA hypermethylation of APC Adenomatosis-poliposis-coli gene (APC), Gluthation-A- was correlated with advanced tumor stage. Conclusion: The Transferase-Protein 1 gene (GSTP1), ARF tumor suppressor detection of hypermethylated cfDNA in serum may be helpful protein gene (p14(ARF)), cyclin-dependent kinase inhibitor 2A for the identification of RCC; the combinatorial analysis of (p16), Retinoid-Acid-Receptor-Beta gene (RAR-B), RAS- multiple genes may increase the diagnostic accuracy. association domain family 1 gene (RASSF1), Tissue inhibitor of metalloproteinase 3 gene (TIMP3) and Prostaglandin- Endoperoxid Synthase 2 (PTGS2) which are involved in the development and progression of RCC (11-23). Correspondence to: Dr. Stefan Hauser, Department of Urology, University of Bonn, Sigmund-Freud-Str. 25, 53105 Bonn, Germany. E-mail: [email protected] Patients and Methods

Key Words: Cell-free DNA, CpG-island, hypermethylation, kidney Our study was planned and conducted as a prospective non- cancer, biomarker, serum. interventional study. We included 35 patients with organ-confined

0250-7005/2013 $2.00+.40 4651 ANTICANCER RESEARCH 33: 4651-4656 (2013)

Table I. Clinicopathological parameters of patients with kidney cancer Table II. Primer design for Adenomatosis-poliposis-coli gene (APC), and healthy individuals. Gluthation-a-transferase-protein 1 gene (GSTP1), ARF tumor suppressor protein gene (p14(ARF)), cyclin-dependent kinase inhibitor 2A (p16), Renal cell carcinoma Healthy controls Retinoid-acid-receptor-beta gene (RAR-B), RAS-association domain (n=35) (n=54) family 1 gene (RASSF1), Tissue inhibitor of metalloproteinase-3 gene (TIMP3) and Prostaglandin-endoperoxid synthase-2 (PTGS2) genes. Median age (years, range) 66 (31-80) 28.5 (18-56) Male/female 28/7 37/17 Gene Primer sequence Histology Clear cell 29 82.8% APC Forward GGA GAG AGA AGC AGC TGT GTA AT Papillary 4 11.4% Reverse CAG CCA CAT GTC GGT CAC Chromophobe 2 5.5% GSTP1 Forward GGG ACC CTC CAG AAG AGC TNM Reverse ACT CAC TGG TGG CGA AGA CT pT1a 14 40% p14 Forward AGT TAA GGG GGC AGG AGT G pT1b 7 20% Reverse GGA GGG TCA CCA AGA ACC TG pT2 1 3% p16 Forward AGC ACC GGA GGA AGA AAG AG pT3a 4 11% Reverse CTG CCT GCT CTA CCC CTC TC pT3b 6 17% PTGS2 Forward GGA GAG GAA GCC AAG TGT CC pT3c 1 3% Reverse GGT TTC CGC CAG ATG TCT TT n.a. 2 5.7% RAR-B Forward ATG CGA GCT GTT TGA GGA CT N+ 2 5.7% Reverse AAT GCG TTC CGG ATC CTA C M+ 3 8.5% RASSF1A Forward GCT TGC TAG CGC CCA AAG R1 3 8.5% Reverse CAG CTC CCG CAG CTC AAT TIMP3 Forward GCA CGG CAA CTT TGG AGA G Reverse GAG CCA AGG GGT CAT TGC

RCC treated at the Department of Urology at the University Hospital Bonn: 15 patients underwent open radical nephrectomy and 20 patients underwent open nephron-sparing surgery. Fifty-four according to the manufacturer’s recommendations. DNA healthy individuals served as the control group. The patient hypermethylation was detected by PCR as described earlier (26): The characteristics are shown in Table I. All patients gave written PCR primers covered at least one restriction site of the three used informed consent prior to enrolment into the study. The Ethics restriction enzymes in the promoter CpG island of APC, GSTP1, Committee of the University Hospital Bonn approved the study p14(ARF), p16, RAR-B, RASSF1A, TIMP3 and PTGS2 (for primer (vote number 19/03). Serum samples were collected one day before design see Table II). The amount of input DNA was determined using surgery (Serum-S monovette with clotting activator Sarstedt, an actin-beta (ACTB) primer without any restriction site. The ratio of Nürnbrecht, Germany). Clotting occurred for 30-240 min prior to methylated target templates to the amount of ACTB was defined as the centrifugation at 1,800 ×g (10 min). relative DNA methylation level. Quantitative real-time PCR was carried out in triplicate on an ABIPrism7900HT (Applied Biosystems, Foster DNA isolation and hypermethylation analysis. Cell-free DNA was City, CA, USA). Each 10 μl reaction consisted of 1× SYBRGreenER isolated from 1 ml serum by using the ChargeSwitch gDNA Kit (Invitrogen, Carlsbad, USA), 200 nmol forward/reverse primer and (Invitrogen, Paisley, Scotland, UK) according to the manufacturer’s 1 μl of digested DNA. PCR was carried out by the following recommendations (final elution volume 50 μl). To detect CpG procedure: 95˚C for 10 min, followed by 40 cycles at 95˚C for 15 s hypermethylation, a restriction endonuclease-based assay was used as and 60˚C for 60 s. Melting curve analysis was performed to confirm described in detail elsewhere (24). The methylation-sensitive restriction the specificity of the PCR products. The copy number was determined enzymes cut unmethylated DNA, while methylated DNA sequences using a standard curve. Each run included a positive control, a negative block the restriction enzyme activity. Thus, only fully- methylated target control and water blanks. sequences will deliver an amplification signal on subsequently performed real-time polymerase chain reaction (PCR), whereas Statistical analysis. Methylation levels and clinicopathological unmethylated DNA does not. The three methylation-sensitive restriction parameters were correlated using the Mann-Whitney test or the Chi- enzymes were used Bsh1236I (recognition site CGCG), Hpall (CCGG) square-test, as appropriate. The area under the curve (AUC), and Hin6I (GCGC) (Fermentas, St. Leon-Rot, Germany). DNA was sensitivity and specificity were determined by receiver operating digested with 20 U of Bsh1236I, HpaII and Hin6I in a total volume of characteristic (ROC) analysis. Statistical tests were performed using 45 μl 1× Tango Buffer at 37˚C (4 hours). To ensure complete digestion, SPSS Statistics v20 (IBM Corperation, Armonk, NY, USA). an additional 5 U of each enzyme were added and incubated for 14 h Significance was concluded at p<0.05. followed by inactivation at 65˚C for 20 min. Buffy coat DNA from a healthy volunteer was used to create a positive and a negative control for the PCR reaction. Complete unmethylated DNA was synthesized Results using the GenomiPhi Amplification Kit (Amersham Bioscience, Freiberg, Germany) (25). A fully-methylated positive control was Clear cell RCC was found in the majority of patients (29/35), created by treating the buffy coat DNA with SssI CpG whereas papillary RCC was found in four and chromophobe methyltransferase (New England Biolabs, Ipswich, MA, USA) RCC was found in two; detailed clinico-pathological

4652 Hauser et al: Serum DNA Hypermethylation in Kidney Cancer

Table III. Cell-free DNA methylation frequency in patients with renal Table IV. Receiver operator characteristic analysis for the cell carcinoma and healthy controls. discrimination of renal cell carcinoma patients and healthy subjects using cell-free serum DNA methylation of various genes alone and in Renal cell carcinoma Healthy controls Chi2 combination. (n=35) (n=54) Cut-off Sensitivity Specificity AUC 95%CI Positive % Positive % p-Value APC 0.37 54.3% 90.7% 0.72 0.60-0.83 APC 19 54.3 5 9.3 <0.0001 GSTP1 0.75 17.1% 98.1% 0.57 0.45-0.70 GSTP1 6 17.1 1 1.9 0.01 p14(ARF) 0.26 14.3% 100% 0.57 0.45-0.70 p14(ARF) 5 14.3 0 0 0.005 PTGS2 0.47 22.9% 96.3% 0.59 0.47-0.72 p16 9 25.7 9 16.7 0.49 RAR-B 0.19 40.0% 85.2% 0.61 0.49-0.73 PTGS2 8 22.9 2 3.7 0.004 RASSF1A 0.09 22.9% 98.2% 0.60 0.48-0.73 RAR-B 14 40 8 14.8 0.009 APC or GSTP1 57.1% 88.9% 0.73 0.62-0.84 RASSF1A 8 22.9 1 1.9 0.001 APC or PTGS2 60.0% 87.0% 0.74 0.62-0.85 TIMP3 20 57.1 21 38.9 0.126 APC or RAR-β 74.3% 77.8% 0.76 0.65-0.87 PTGS2 or GSTP1 62.9% 87.0% 0.75 0.50-0.75 Adenomatosis-poliposis-coli gene (APC), Gluthation-a-transferase- protein 1 gene (GSTP1), ARF tumor suppressor protein gene AUC, Area under the curve; 95% CI, 95% confidence interval for AUC. (p14(ARF)), cyclin-dependent kinase inhibitor 2A (p16), Retinoid-acid- Adenomatosis-poliposis-coli gene (APC), Gluthation-a-transferase- receptor-beta gene (RAR-B), RAS-association domain family 1 gene protein 1 gene (GSTP1), ARF tumor suppressor protein gene (RASSF1), Tissue inhibitor of metalloproteinase 3 gene (TIMP3) and (p14(ARF)), cyclin-dependent kinase inhibitor 2A (p16), Retinoid-acid- Prostaglandin-endoperoxid synthase 2 (PTGS2). receptor-beta gene (RAR-B), RAS-association domain family 1 gene (RASSF1), Tissue inhibitor of metalloproteinase 3 gene (TIMP3) and Prostaglandin-endoperoxid synthase 2 (PTGS2). parameters are presented in Table I. In 30 out of 35 investigated patients with kidney cancer, at least one gene was methylated within the serum cfDNA. A broad variation of Discussion methylation frequency was detected in the eight genes e.g. methylation of TIMP3 (57.1%) and APC (54.3%) were From the first description of cfDNA in 1948 (27) to frequent, in contrast in p14(ARF) (14.3%) and GSTP1 (17.1%) investigation of epigenetic alterations in cfDNA, nearly 5 methylation was uncommon in patients with RCC; see Table decades have passed. In patients with RCC, multiple gene III for detailed methylation frequencies. Except for p16 and sites have been analyzed for their methylation status, this TIMP3 all genes were significantly more frequently was performed for tissue as well as serum (11-22). For methylated in patients with RCC compared to healthy Drosophila melanogaster wingless (WNT) antagonist family individuals. The cfDNA methylation profile is shown in Figure genes, Urakami and collegues found the same methylation 1. The ROC analysis for the investigated genes (all except p16 patterns in tissue and serum in 72% patients with renal and TIMP3) showed a high specificity ranging from 85.2% for cancer whereas serum DNA in healthy controls showed no RAR-B to 100% for p14(ARF), nevertheless the sensitivity was aberrant methylation (28). Analyzing epigenetic alterations low in the single-gene analysis, ranging between 14.3% for in circulating cfDNA allows conclusions to be drawn on the p14(ARF) and 54.3% for APC, corresponding to an AUC from characteristics of the primary tumor and thus may be useful 0.57 to 0.72 (Table IV). In contrast a combined analysis using in defining biomarker. In the current study we report on the the or-disjunction as a paired analysis for APC with PTGS2, methylation patterns of eight genes in RCC, APC, GSTP1, RARβ and GSTP1 resulted in an increased sensitivity (60- p14(ARF), p16, RAR-B, RASSF1A and TIMP3. De Martino 74.3%) while the specificity remained high (77.8-88.8%). The et al. analyzed the methylation patterns of PTGS2, RASSF1A combination of APC, PTGS2 and GSTP1 as paired analysis and p16 in serum cfDNA in patients with RCC and healthy using the disjunction led to 62.9% sensitivity and 87% controls (23). RASSF1A was methylated in 45.9% of the specificity (see Table IV). All other combinations revealed no patients with RCC, in contrast 22.9% of the patients in our additional diagnostic information. An increased APC cohort were positive for methylation. Despite this methylation level was observed in patients with locally discrepancy in percentages for methylation of RASSF1A the advanced disease: the methylation ratio was significantly difference from healthy controls was significant (p=0.001) increased in patients with pT3 compared to patients with pT1 in both studies. In both studies, methylation of p16 differed stage RCC (0.16 vs. 0.034, p=0.035). There was no correlation non-significantly between patients with RCC and healthy of cfDNA hypermethylation and other investigated controls. The AUC for RASSF1A was comparable between clinicopathological parameters (grade, lymph node metastasis, the studies (0.69 vs. 0.60) with a high specificity (93% vs. smoking, age, sex; all p>0.05). 98.2%) but only low sensitivity (45.9% vs. 22.9%). In

4653 ANTICANCER RESEARCH 33: 4651-4656 (2013)

Figure 1. Quantitative analysis of serum cell-free DNA hypermethylation at the promoter region of Adenomatosis-poliposis-coli gene (APC), Gluthation-a-transferase-protein 1 gene (GSTP1), ARF tumor suppressor protein gene (p14(ARF)), cyclin-dependent kinase inhibitor 2A (p16), Retinoid-acid-receptor-beta gene (RAR-B), RAS-association domain family 1 gene (RASSF1), Tissue inhibitor of metalloproteinase 3 gene (TIMP3) and Prostaglandin-endoperoxid synthase 2 (PTGS2). Right, healthy individuals; left, patients with renal cell carcinoma. The methylation ratio was scaled as no methylated DNA (white) to fully-methylated DNA (black).

4654 Hauser et al: Serum DNA Hypermethylation in Kidney Cancer contrast to our study population, PTGS2 methylation was References observed at similar frequencies among patients with RCC and healthy controls (p=0.689), whereas this difference in 1 Ljungberg B: Prognostic factors in renal cell carcinoma. Scand J our cohort reached statistical significance (p=0.004). The Surg 93(2): 118-125, 2004. AUC for PTGS2 was comparable (0.517 vs. 0.47), whereas 2 Pantuck AJ, Zisman A and Belldegrun AS: The changing natural the specificity was higher in our analysis (65.1% vs. 96.3%); history of renal cell carcinoma. J Urol 166(5): 1611-1623, 2001. 3 Herman JG and Baylin SB: Gene silencing in cancer in the sensitivity remains low for both studies (38.2% vs. association with promoter hypermethylation. N Engl J Med 22.9%). They found no correlation between 349(21): 2042-2054, 2003. clinicopathological parameters with serum cfDNA 4 Esteller M, Garcia-Foncillas J, Andion E, Goodman SN, Hidalgo methylation for the investigated genes, this is also supported OF, Vanaclocha V, Baylin SB and Herman JG: Inactivation of the by our data (see Table IV). Hoque et al. investigated DNA-repair gene MGMT and the clinical response of gliomas to RASSF1A (11%), TIMP3 (17%), RARβ (11%), p16 (22%), alkylating agents. N Engl J Med 343(19): 1350-1354, 2000. p14(ARF) (6%), GSTP1 (6%) and APC (6%) methylation 5 Ohno T, Hiraga J, Ohashi H, Sugisaki C, Li E, Asano H, Ito T, Nagai H, Yamashita Y, Mori N, Kinoshita T and Naoe T: Loss frequencies in serum samples of patients with RCC (14). of O 6-methylguanine-DNA methyltransferase protein Compared to our data (Table III), these frequencies are much expression is a favorable prognostic marker in diffuse large B- lower. A possible explanation is the use of a bisulfite-based cell lymphoma. Int J Hematol 83(4): 341-347, 2006. assay compared to the methylation-sensitive restriction 6 Morrissey C, Martinez A, Zatyka M, Agathanggelou A, Honorio enzyme assay, which may be more sensitive. It was shown S, Astuti D, Morgan NV, Moch H, Richards FM, Kishida T, Yao that bisulfite treatment leads to the loss of up to 90% of M, Schraml P, Latif F and Maher ER: Epigenetic inactivation of DNA during sample preparation (29), which may lead to the RASSF1A 3p21.3 tumor suppressor gene in both clear cell and papillary renal cell carcinoma. Cancer Res 61(19): 7277- underestimation of methylation frequencies in samples with 7281, 2001. limited amounts of DNA. One drawback for all investigated 7 Yoon JH, Dammann R and Pfeifer GP: Hypermethylation of the genes was the relative low sensitivity (range: 14.3%-54.3%). CpG island of the RASSF1A gene in ovarian and renal cell By combining the use of multiple genes it was possible to carcinomas. Int J Cancer 94(2): 212-217, 2001. increase the sensitivity (Table IV). For the investigated 8 Bachman KE, Herman JG, Corn PG, Merlo A, Costello JF, genes, the combination analysis seems to deliver more Cavenee WK, Baylin SB and Graff JR: Methylation-associated clinically robust results compared to single-gene analysis, silencing of the tissue inhibitor of metalloproteinase-3 gene suggest a suppressor role in kidney, brain, and other human similarly to prostate cancer for which published data show cancers. Cancer Res 59(4): 798-802, 1999. that the investigation of multiple genes increases the 9 Hauser S, Zahalka T, Ellinger J, Fechner G, Heukamp LC, A diagnostic and prognostic information, our data support this VONR, Muller SC and Bastian PJ: Cell-free circulating DNA: finding for kidney cancer (30, 31). The analysis of cfDNA Diagnostic value in patients with renal cell cancer. Anticancer methylation may also be indicative of a poor prognosis: APC Res 30(7): 2785-2789, 2010. methylation levels were increased in patients with pT3 10 Ellinger J, Muller DC, Muller SC, Hauser S, Heukamp LC, von compared to pT1 tumor. However, follow-up information Ruecker A, Bastian PJ and Walgenbach-Brunagel G: Circulating mitochondrial DNA in serum: a universal diagnostic biomarker was not available and therefore we were not able to correlate for patients with urological malignancies. Urol Oncol 30(4): cfDNA methylation and survival data in order to determine 509-515, 2012. the prognostic relevance. It should be mentioned that other 11 Battagli C, Uzzo RG, Dulaimi E, Ibanez de Caceres I, researchers were not able to demonstrate an association of Krassenstein R, Al-Saleem T, Greenberg RE and Cairns P: cfDNA methylation and prognostic-relevant parameters in Promoter hypermethylation of tumor suppressor genes in urine RCC; this has also been described for other urological from kidney cancer patients. Cancer Res 63(24): 8695-8699, 2003. malignancies. It is important to emphasize that DNA 12 Dulaimi E, Ibanez de Caceres I, Uzzo RG, Al-Saleem T, Greenberg RE, Polascik TJ, Babb JS, Grizzle WE and Cairns P: hypermethylation also occurs in non-malignant diseases, and Promoter hypermethylation profile of kidney cancer. Clin Cancer furthermore, aging, nutrition and lifestyle factors contribute Res 10(12 Pt 1): 3972-3979, 2004. to altered DNA methylation. For example, methylated serum 13 Esteller M, Corn PG, Baylin SB and Herman JG: A gene cfDNA fragments (e.g. RASSF1A and p16) were also found hypermethylation profile of human cancer. Cancer Res 61(8): in patients with hepatitis C (32, 33). 3225-3229, 2001. In conclusion analyzing serum cfDNA methylation may 14 Hoque MO, Begum S, Topaloglu O, Jeronimo C, Mambo E, be a promising biomarker in RCC but further investigations Westra WH, Califano JA and Sidransky D: Quantitative detection of promoter hypermethylation of multiple genes in the tumor, in the future need to be performed to define its actual role. urine, and serum DNA of patients with renal cancer. Cancer Res 64(15): 5511-5517, 2004. Acknowledgements 15 Chung WB, Hong SH, Kim JA, Sohn YK, Kim BW and Kim JW: Hypermethylation of tumor-related genes in genitourinary We thank Ms. Doris Schmidt for her excellent technical assistance. cancer cell lines. J Korean Med Sci 16(6): 756-761, 2001.

4655 ANTICANCER RESEARCH 33: 4651-4656 (2013)

16 Esteller M, Corn PG, Urena JM, Gabrielson E, Baylin SB and 26 Ellinger J, Albers P, Perabo FG, Muller SC, von Ruecker A and Herman JG: Inactivation of glutathione S-transferase P1 gene by Bastian PJ: CpG island hypermethylation of cell-free circulating promoter hypermethylation in human neoplasia. Cancer Res serum DNA in patients with testicular cancer. J Urol 182(1): 58(20): 4515-4518, 1998. 324-329, 2009. 17 Arai E, Kanai Y, Ushijima S, Fujimoto H, Mukai K and 27 Mandel: Les acides nucleiques du plasma sanguin chez Hirohashi S: Regional DNA hypermethylation and DNA l’homme. CR Acad Sci Paris 142: 241-243, 1948. methyltransferase (DNMT) 1 protein overexpression in both 28 Urakami S, Shiina H, Enokida H, Hirata H, Kawamoto K, renal tumors and corresponding nontumorous renal tissues. Int J Kawakami T, Kikuno N, Tanaka Y, Majid S, Nakagawa M, Cancer 119(2): 288-296, 2006. Igawa M and Dahiya R: WNT antagonist family genes as 18 Herman JG, Merlo A, Mao L, Lapidus RG, Issa JP, Davidson biomarkers for diagnosis, staging, and prognosis of renal cell NE, Sidransky D and Baylin SB: Inactivation of the carcinoma using tumor and serum DNA. Clin Cancer Res CDKN2/p16/MTS1 gene is frequently associated with aberrant 12(23): 6989-6997, 2006. DNA methylation in all common human cancers. Cancer Res 29 Grunau C, Clark SJ and Rosenthal A: Bisulfite genomic 55(20): 4525-4530, 1995. sequencing: Systematic investigation of critical experimental 19 Kawakami T, Okamoto K, Ogawa O and Okada Y: Multipoint parameters. Nucleic Acids Res 29(13): E65-65, 2001. methylation and expression analysis of tumor suppressor genes 30 Tokumaru Y, Harden SV, Sun DI, Yamashita K, Epstein JI and in human renal cancer cells. Urology 61(1): 226-230, 2003. Sidransky D: Optimal use of a panel of methylation markers 20 Morris MR, Hesson LB, Wagner KJ, Morgan NV, Astuti D, Lees with GSTP1 hypermethylation in the diagnosis of prostate RD, Cooper WN, Lee J, Gentle D, Macdonald F, Kishida T, adenocarcinoma. Clin Cancer Res 10(16): 5518-5522, 2004. Grundy R, Yao M, Latif F and Maher ER: Multigene 31 Ellinger J, Bastian PJ, Jurgan T, Biermann K, Kahl P, Heukamp methylation analysis of Wilms' tumour and adult renal cell LC, Wernert N, Muller SC and von Ruecker A: CpG island carcinoma. Oncogene 22(43): 6794-6801, 2003. hypermethylation at multiple gene sites in diagnosis and 21 Dreijerink K, Braga E, Kuzmin I, Geil L, Duh FM, Angeloni D, prognosis of prostate cancer. Urology 71(1): 161-167, 2008. Zbar B, Lerman MI, Stanbridge EJ, Minna JD, Protopopov A, 32 Mohamed NA, Swify EM, Amin NF, Soliman MM, Tag-Eldin Li J, Kashuba V, Klein G and Zabarovsky ER: The candidate LM and Elsherbiny NM: Is serum level of methylated RASSF1A tumor suppressor gene, RASSF1A, from human chromosome valuable in diagnosing hepatocellular carcinoma in patients with 3p21.3 is involved in kidney tumorigenesis. Proc Natl Acad Sci chronic viral hepatitis C? Arab J Gastroenterol 13(3): 111-115, USA 98(13): 7504-7509, 2001. 2012. 22 Gonzalgo ML, Yegnasubramanian S, Yan G, Rogers CG, Nicol 33 Ahmed R, Salama H, Fouad A, Sabry D, AbdAlah el S and TL, Nelson WG and Pavlovich CP: Molecular profiling and Kamal M: Detection of aberrant p16INK4A methylation in sera of classification of sporadic renal cell carcinoma by quantitative patients with HCV-related liver diseases: An Egyptian study. methylation analysis. Clin Cancer Res 10(21): 7276-7283, 2004. Med Sci Monit 16(9): CR410-415, 2010. 23 de Martino M, Klatte T, Haitel A and Marberger M: Serum cell- free DNA in renal cell carcinoma: A diagnostic and prognostic marker. Cancer 118(1): 82-90, 2012. 24 Ellinger J, Bastian PJ, Haan KI, Heukamp LC, Buettner R, Fimmers R, Mueller SC and von Ruecker A: Noncancerous PTGS2 DNA fragments of apoptotic origin in sera of prostate cancer patients qualify as diagnostic and prognostic indicators. Int J Cancer 122(1): 138-143, 2008. 25 Umetani N, de Maat MF, Mori T, Takeuchi H and Hoon DS: Synthesis of universal unmethylated control DNA by nested Received August 11, 2013 whole-genome amplification with phi29 DNA polymerase. Revised September 16, 2013 Biochem Biophys Res Commun 329(1): 219-223, 2005. Accepted September 19, 2013

4656