Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Single-CpG-resolution methylome analysis identifies clinicopathologically aggressive CpG island methylator phenotype clear cell renal cell carcinomas

Single-CpG-resolution methylome analysis identifies clinicopathologically aggressive CpG island... Carcinogenesis vol.33 no.8 pp.1487–1493, 2012 doi:10.1093/carcin/bgs177 Advance Access Publication May 18, 2012 Single-CpG-resolution methylome analysis identifies clinicopathologically aggressive CpG island methylator phenotype clear cell renal cell carcinomas 1 2 1 efforts (6). Such efforts have revealed that renal carcinogenesis Eri Arai , Suenori Chiku , Taisuke Mori , 1 3 involves inactivation of histone-modifying genes, such as SETD2 (7), Masahiro Gotoh , Tohru Nakagawa , 3 1, a histone H3 lysine 36 methyltransferase, JARIDIC (KDM5C (7)), Hiroyuki Fujimoto and Yae Kanai * a histone H3 lysine 4 demethylase, and UTX (KMD6A (8)), a histone Division of Molecular Pathology, National Cancer Center Research Institute, H3 lysine 27 demethylase, as well as the SWI/SNF chromatin-remod- Tokyo 104-0045, Japan, Science Solutions Division, Mizuho Information eling complex gene, PBRM1 (9). Non-synonymous mutations of the and Research Institute, Inc., Tokyo 101-8443, Japan and Department of NF2 gene and truncating mutations of the MLL2 gene have also been Urology, National Cancer Center Hospital, Tokyo 104-0045, Japan reported (7). However, such gene mutations cannot fully explain the *To whom correspondence should be addressed. Tel: +81 3 3542 2511; clinicopathological diversity of clear cell RCCs. Fax: +81 3 3248 2463; Not only genetic, but also epigenetic events appear to accumulate Email: ykanai@ncc.go.jp during carcinogenesis, and both types of event reflect the clinicopatho- To clarify the significance of DNA methylation alterations dur - logical diversity of cancers in various organs in association with each ing renal carcinogenesis, methylome analysis using single-CpG- other (10–12). DNA methylation alterations are one of the most con- resolution Infinium array was performed on 29 normal renal sistent epigenetic changes in human cancers (13–16). In fact, on the cortex tissue (C) samples, 107 non-cancerous renal cortex tis- basis of methylation-specific PCR (MSP), combined bisulfite restric- sue (N) samples obtained from patients with clear cell renal cell tion enzyme analysis (17,18) and bacterial artificial chromosome array- carcinomas (RCCs) and 109 tumorous tissue (T) samples. DNA based methylated CpG island amplification (BAMCA (19,20)), we have methylation levels at 4830 CpG sites were already altered in N suggested that non-cancerous renal cortex tissue obtained from patients samples compared with C samples. Unsupervised hierarchical with RCCs is already at the precancerous stage associated with DNA clustering analysis based on DNA methylation levels at the 801 methylation alterations, even though no remarkable histological changes CpG sites, where DNA methylation alterations had occurred in are evident and there is no association with chronic inflammation or N samples and were inherited by and strengthened in T samples, persistent infection with viruses or other pathogenic microorganisms. clustered clear cell RCCs into Cluster A (n = 90) and Cluster B Genome-wide analysis using BAMCA revealed that DNA methylation (n  =  14). Clinicopathologically aggressive tumors were accumu- status in non-cancerous renal cortex tissue at the precancerous stage lated in Cluster B, and the cancer-free and overall survival rates was basically inherited by the corresponding clear cell RCC in indi- of patients in this cluster were significantly lower than those of vidual patients (19). DNA methylation alterations at the precancerous patients in Cluster A. Clear cell RCCs in Cluster B were char- stage may confer further susceptibility to genetic and epigenetic alter- acterized by accumulation of DNA hypermethylation on CpG ations and generate more malignant clear cell RCCs (2,13). However, islands and considered to be CpG island methylator phenotype in our previous studies using BAMCA, the resolution and the number (CIMP)-positive cancers. DNA hypermethylation of the CpG sites of probes were limited. Therefore, further analysis is needed to clarify on the FAM150A, GRM6, ZNF540, ZFP42, ZNF154, RIMS4, the significance of DNA methylation alterations in renal carcinogenesis. PCDHAC1, KHDRBS2, ASCL2, KCNQ1, PRAC, WNT3A, Recently, methylome analysis using the Infinium array has made TRH, FAM78A, ZNF671, SLC13A5 and NKX6-2 genes became it possible to interrogate 27 000 highly informative CpG sites at sin- hallmarks of CIMP in RCCs. On the other hand, Cluster A was gle-CpG resolution (21). In order to clarify the significance of DNA characterized by genome-wide DNA hypomethylation. These methylation alterations during renal carcinogenesis, we used the data indicated that DNA methylation alterations at precancerous Infinium BeadChip system to perform genome-wide DNA methyla- stages may determine tumor aggressiveness and patient outcome. tion analysis of 29 samples of normal renal cortex tissue (C) obtained Accumulation of DNA hypermethylation on CpG islands and from patients without any primary renal tumors, 107 samples of non- genome-wide DNA hypomethylation may each underlie distinct cancerous renal cortex tissue (N) from patients with clear cell RCCs pathways of renal carcinogenesis. and 109 samples of tissue from the tumors (T) themselves. Correla- tions between the genome-wide DNA methylation profiles and clin- icopathological parameters were then examined. Introduction Materials and methods Clear cell renal cell carcinoma (RCC) is the most common histological Patients and tissue samples subtype of adult kidney cancer and frequently affects working-age adults in midlife. In general, RCCs at an early stage are curable by The 109 T samples and corresponding 107 N samples showing no remarkable histological changes were obtained from materials that had been surgically nephrectomy. However, some RCCs relapse and metastasize to dis- resected from 110 patients with primary clear cell RCCs. These patients did tant organs, even if the resection has been considered complete (1). not receive preoperative treatment and underwent nephrectomy at the National Such clinicopathological diversity may be attributable to distinct path- Cancer Center Hospital, Tokyo, Japan. There were 79 men and 31 women with a ways of renal carcinogenesis (2). It is well known that clear cell RCCs mean (±SD) age of 62.8 ± 10.3 years (range 36–85 years). Histological diagnosis are characterized by inactivation of the Von Hippel–Lindau tumor- was made in accordance with the World Health Organization classification (22) suppressor gene (3). In addition, systematic resequencing and exome (Supplementary Figure S1, available at Carcinogenesis Online). All the tumors analysis of RCCs are now being performed by The Cancer Genome were graded on the basis of criteria described previously (23) and classified Atlas (4), The Cancer Genome Project (5) and other international according to the pathological Tumor-Node-Metastasis (TNM) classification (24). The criteria for macroscopic configuration of RCC (17–19) followed those Abbreviations: BAMCA, bacterial artificial chromosome array-based meth- established for hepatocellular carcinoma (HCC): type 3 (contiguous multinodular ylated CpG island amplification; C, normal renal cortex tissue obtained from type) HCCs show poorer histological differentiation and a higher incidence of patients without any primary renal tumor; CIMP, CpG island methylator phe- intrahepatic metastasis than type 1 (single nodular type) and type 2 (single notype; HCC, hepatocellular carcinoma; N, non-cancerous renal cortex tissue nodular type with extranodular growth) HCCs (25). The presence or absence obtained from patients with clear cell renal cell carcinomas; NCBI, National of vascular involvement was examined microscopically on slides stained with Center for Biotechnology Information; RCC, renal cell carcinoma; T, tumor- hematoxylin–eosin and elastica van Gieson. The presence or absence of tumor ous tissue; TNM, Tumor-Node-Metastasis. thrombi in the main trunk of the renal vein was examined macroscopically. © The Author 2012. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons. org/licenses/by-nc/3.0), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. E.Arai et al. RCC is usually enclosed by a fibrous capsule and well demarcated, and hardly pyrosequencing method (Supplementary Figure S2, available at Car- ever contains fibrous stroma between cancer cells. Therefore, we were able to cinogenesis Online). With regard to the well-known methylation- obtain cancer cells from surgical specimens, avoiding contamination with both silencing Von Hippel–Lindau tumor-suppressor gene (probe Target ID: non-cancerous epithelial cells and stromal cells. cg22782492), DNA hypermethylation (Δβ > 0.1) was detected in 12 T–N For comparison, 29 samples of normal renal cortex tissue (C1–C29) were (12%) of 104 patients, for whom both N and T samples were assayed obtained from materials that had been surgically resected from 29 patients in the same experimental batch. This incidence corresponded to that in without any primary renal tumor. These patients included 18 men and previous reports (28,29). Taken together, the data confirmed the reliabil- 11 women with a mean (±SD) age of 61.4 ± 10.8 years (range 31–81 years). Of ity of the present Infinium assay. these patients, 22 had undergone nephroureterectomy for urothelial carcinomas of the renal pelvis and ureter, 6 had undergone nephrectomy with resection of retro- Although precancerous conditions in the kidney have been rarely peritoneal sarcoma around the kidney, and the remaining 1 had undergone para- described, our previous study suggested that N samples are already at aortic lymph node dissection for metastatic germ cell tumor, which resulted in precancerous stages, from the viewpoint of altered DNA methylation, simultaneous nephrectomy because it was not possible to preserve the renal artery. despite the absence of any remarkable histological changes and the All patients included in this study provided written informed consent, and lack of association with chronic inflammation and persistent infection the study was approved by the Ethics Committee of the National Cancer Cen- with viruses or other pathogenic microorganisms (17–20). (a) In ter, Tokyo, Japan. fact, the logistic model adjusted by sex, age and experimental batch Infinium assay revealed that DNA methylation levels on 4830 probes were already High-molecular-weight DNA from fresh frozen tissue samples was extracted altered in N samples compared with those in C samples (FDR, using phenol–chloroform, followed by dialysis (26). Five-hundred-nanogram q = 0.01, Table I). (b) In order to reveal DNA methylation alterations aliquots of DNA were subjected to bisulfite conversion using an EZ DNA inherited by clear cell RCCs themselves, ordered differences of DNA TM Methylation-Gold Kit (Zymo Research, Irvine, CA). Subsequently DNA methylation level from C to N and then to T samples were examined methylation status at 27 578 CpG loci was examined at single-CpG resolution by the cumulative logit model adjusted by sex, age and experimental using the Infinium HumanMethylation27 Bead Array (Illumina, San Diego, batch. Ordered differences from C to N and then to T samples were CA). This array contains CpG sites located within the proximal promoter observed on 11  089 probes (FDR, q  =  0.01, Table I). (c) In order regions of the transcription start sites of 14 475 consensus coding sequences in to reveal the cancer-prone DNA methylation alterations, differences the National Center for Biotechnology Information Database. On average, two assays were selected per gene, and from 3 to 20 CpG sites for more than 200 in DNA methylation levels between 104 paired samples of N and cancer-related and imprinted genes. Forty control probes were employed for T assayed in the same experimental batch were examined using the each array; these included staining, hybridization, extension, bisulfite conver - Wilcoxon matched pairs test. Significant differences between N and sion and negative controls. An Evo robot (Tecan, Switzerland) was used for the corresponding clear cell RCCs themselves were observed on automated sample processing. Whole-genome amplification was performed 10 870 probes (FDR, q = 0.01, Table I). using the Infinium Assay Kit (Illumina (21)). After hybridization, the specifi- DNA hypermethylation frequently occurred at the very early stages cally hybridized DNA was fluorescence labeled by a single-base extension of renal carcinogenesis [(a) in Table I], whereas DNA hypometh- reaction and detected using a BeadScan reader (Illumina) in accordance with ylation was also observed during progression to established cancers the manufacturer’s protocols. The data were then assembled using GenomeS- [(b) and (c) in Table I]. Eight hundred and one probes satisfied all of tudio methylation software (Illumina). At each CpG site, the ratio of the fluo- rescent signal was measured using a methylated probe relative to the sum of the the above criteria (a)–(c) (Table I): DNA methylation alterations on methylated and unmethylated probes, i.e. the so-called β-value, which ranges these 801 probes (Supplementary Table S2, available at Carcinogen- from 0.00 to 1.00, reflecting the methylation level of an individual CpG site. esis Online) were already evident in N samples, and were inherited by and strengthened in T samples. Statistics In the Infinium assay, the call proportions (P-values for detection of signals above the background <0.01) for 32 probes (shown in Supplementary Table S1, available at Carcinogenesis Online) in all of the tissue samples exam- ined were less than 90%. Since such a low proportion may be attributable to Table I. DNA methylation alterations during renal carcinogenesis polymorphism at the probe CpG sites, these 32 probes were excluded from the present assay. In addition, all CpG sites on chromosomes X and Y were The number of probes showing DNA hypermethylation and DNA excluded, to avoid any gender-specific methylation bias, leaving a final total hypomethylation of 26 454 autosomal CpG sites. (a) The probes on which DNA methylation levels were altered in samples of Infinium probes showing significant differences in DNA methylation lev- non-cancerous renal cortex tissue (N) obtained from patients with clear els between the 29 C and 107 N samples were identified by a logistic model cell RCCs relative to those in samples of normal renal cortex tissue (C) adjusted by sex, age and experimental batch. Ordered differences from 29 obtained from patients without any primary renal tumor. (Logistic model C to 107 N and then to 109 T samples themselves were examined by the adjusted by sex, age and experimental batch; FDR, q = 0.01.) cumulative logit model adjusted by sex, age and experimental batch. Dif- DNA hypermethylation (β > β ) N C ferences of DNA methylation status between 104 paired samples of N and DNA hypomethylation (β < β ) the corresponding T obtained from a single patient and assayed in the same N C Total 4830 experimental batch were examined by Wilcoxon matched pairs test. A false (b) The probes on which DNA methylation levels showed ordered differences discovery rate (FDR) of q = 0.01 was considered significant. Unsupervised from C to N, and then to tumorous tissue (T) samples. hierarchical clustering (Euclidan distance, Ward method) based on DNA (Cumulative logit model adjusted by sex, age and experimental batch; FDR, methylation levels (Δβ ) was performed in patients with clear cell RCCs. T–N q = 0.01.) Correlations between clusters of patients and clinicopathological parameters DNA hypermethylation (β < β < β , C N T were examined using Wilcoxon rank sum test and Fisher’s exact test. Survival . . β < β =. β or β =. β < β ) curves of patients belonging to each cluster were calculated by the Kaplan– C N T C N T DNA hypomethylation (β > β > β , Meier method, and the differences were compared by the log-rank test. The C N T . . number of Infinium assay probes showing DNA hyper- or hypomethylation β > β =. β or β =. β > β ) C N T C N T in each cluster and the average DNA methylation levels (Δβ ) of each clus- Total 11 089 T–N ter were examined using Wilcoxon rank sum test at a significance level of (c) The probes on which DNA methylation levels differed between T and the P < 0.05. The CpG sites discriminating the clusters were identified by Fish- corresponding N samples (Wilcoxon matched pairs test; FDR, q = 0.01) er’s exact test and random forest analysis (27). 5408 DNA hypermethylation (Δβ > 0) T–N DNA hypomethylation (Δβ < 0) 5462 T–N Total 10 870 Results Among the 4589 probes, 2675 showed DNA hypermethylation in T samples DNA methylation alterations during renal carcinogenesis than in C samples (β > β ; FDR, q = 0.01). T C First, DNA methylation levels of representative CpG sites based on Among the 241 probes, 126 showed DNA hypomethylation in T samples the Infinium assay were clearly verified using a highly quantitative than in C samples (β < β ; FDR, q = 0.01). T C 1488 Single-CpG-resolution methylome analysis of RCCs Fig. 1. Unsupervised hierarchical clustering using DNA methylation levels (Δβ ) on the 801 probes in 104 patients with clear cell RCCs. The 801 probes T–N satisfied all of the criteria (a), (b) and (c) in ‘DNA methylation alterations during renal carcinogenesis’ in Results and Table I. On the 801 probes, DNA methylation alterations occurred at the precancerous stages and were inherited by and strengthened in clear cell RCCs themselves. (A) 104 patients with clear cell RCCs were hierarchically clustered into Cluster A (n = 90) and Cluster B (n = 14). The DNA methylation levels (Δβ ) are shown in the color range maps. T–N −6 The cluster trees for patients and probes are shown at the top and left of the panel, respectively. (B) The cancer-free (P = 3.59 × 10 ) survival rates of Stage I–III −2 patients in Cluster B were significantly lower (log-rank test) than those of patients in Cluster A. Overall (P = 1.32 × 10 ) survival rates of all patients in Cluster B were significantly lower (log-rank test) than those of patients in Cluster A. 1489 E.Arai et al. Epigenetic clustering of clear cell RCCs DNA methylation profiles of clear cell RCCs belonging to each cluster Unsupervised hierarchical clustering using DNA methylation levels The distribution of DNA methylation levels (Δβ ) in all 26  454 T–N (Δβ ) on the above 801 probes, on which DNA methylation altera- probes for 104 clear cell RCCs belonging to Cluster A or B is sum- T–N tions occurred at the precancerous stages and may continuously par- marized along chromosomes in Figure 2A. Clear cell RCCs belonging ticipate in renal carcinogenesis, subclustered 104 patients with clear to Cluster B clearly showed accumulation of DNA hypermethylation cell RCCs, of whom both N and T samples were assayed in the same (Δβ > 0.1) relative to DNA hypomethylation, whereas clear cell T–N experimental batch, into Cluster A (n  =  90) and Cluster B (n  =  14, RCCs belonging to Cluster A showed greater DNA hypomethylation Figure 1A). The clinicopathological parameters of clear cell RCCs (Δβ < −0.1) relative to DNA hypermethylation (Figure 2A). T–N belonging to Clusters A and B are summarized in Table II. (The The proportions of the probes showing the various degrees of DNA number of samples for each TNM stage is also described in Supple- hypermethylation in T samples compared with the corresponding N mentary Table S3, available at Carcinogenesis Online.) Epigenetic samples (Δβ > 0.1, 0.2, 0.3, 0.4 or 0.5) for all 26 454 probes, and T–N clustering of clear cell RCCs was dependent on neither age nor sex of the proportions of the probes showing various degrees of DNA hypo- the patients (Table II). Clear cell RCCs belonging to Cluster B had a methylation in T samples compared with the corresponding N samples larger diameter, more frequent macroscopically evident extranodular (Δβ < −0.1, −0.2, −0.3, −0.4 or −0.5) for all 26 454 probes in clear T–N (type 2) or multinodular (type 3) growth, vascular involvement, renal cell RCCs belonging to Clusters A and B are summarized in Figure vein tumor thrombi, infiltrating growth, tumor necrosis and renal pel- 2B. Although the probes showing prominent DNA hypomethylation vis invasion, and also had higher histological grades and pathologi- (Δβ < −0.5) were accumulated slightly more in Cluster B than in T–N cal TNM stages than those in Cluster A (Table II). Figure 1B shows Cluster A, the incidence of DNA hypomethylation in Clusters A and B the Kaplan–Meier survival curves of patients belonging to Clusters did not reach a statistically significant difference (Δβ < −0.1, −0.2, T–N A and B. The period covered ranged from 42 to 4024  days (mean, −0.3 or −0.4, Figure 2B). On the other hand, the probes showing DNA 1821 days). The cancer-free and overall survival rates of patients in hypermethylation were markedly accumulated in Cluster B relative to Cluster B were significantly lower than those of patients in Cluster A Cluster A, regardless of the degree of DNA hypermethylation (Δβ T–N −6 −2 (P = 3.59 × 10 and P = 1.32 × 10 , respectively, Figure 1B). > 0.1, 0.2, 0.3, 0.4 or 0.5, Figure 2B). These data indicate that clear cell RCCs belonging to Cluster B are characterized by accumulation of DNA hypermethylation. Table II. Correlation between the subclassification of patients with clear The top 61 probes (including the 60th and 61st, which showed cell RRCs based on DNA methylation profiles and the clinicopathological equivalent P-values) on which DNA methylation levels (Δβ ) dif- parameters T–N −6 fered markedly between Clusters A and B (P < 1.056  × 10 , Wil- Clinicopathological parameters Cluster A Cluster B P coxon rank sum test) are listed in Supplementary Table S4, available (n = 90) (n = 14) at Carcinogenesis Online. Although only 19 246 probes (72.8%) out −2 b of the total of 26  454 were located within CpG islands, 60 (98.4%) Age 62.08 ± 67.36 ± 8.36 × 10 10.08 11.06 of the top 61 probes located within CpG islands showed DNA hyper- −1 c Sex Male 63 11 5.47 × 10 methylation in clear cell RCCs belonging to Cluster B (Δβ > 0.097, T–N Female 27 3 Supplementary Table S4, available at Carcinogenesis Online): only −4 b Tumor 5.10 ± 8.75 ± 1.07 × 10 1 probe among the top 61 was located within a non-CpG island and diameter (cm) 3.19 2.85 showed DNA hypomethylation (Δβ   =  −0.425 ± 0.096 in Cluster −4 c T–N Macroscopic Type 1 37 1 6.29 × 10 B). Taken together, the data indicated that Cluster B is significantly configuration Type 2 29 2 correlated with clinicopathological phenotype and characterized by Type 3 24 11 −6 c frequent DNA hypermethylation on CpG islands. Such characteristics Predominant G1 47 1 8.33 × 10 of clear cell RCCs in Cluster B are similar to those of CpG island histological G2 35 4 grades G3 7 7 methylator phenotype (CIMP)-positive cancers (30,31) in other well- G4 1 2 studied organs, such as those of the colon (32) and stomach (33). In −4 c Highest G1 8 0 5.67 × 10 other words, our single-CpG-resolution methylome analysis identi- histological G2 43 1 fied, for the first time, CIMP-positive clear cell RCCs as Cluster B. grades G3 24 4 G4 15 9 −4 c Vascular Negative 54 1 2.45 × 10 Hallmark CpG sites of CIMP-positive clear cell RCCs involvement Positive 36 13 Scattergrams of DNA methylation levels (β values) in T and the corre- −3 c Renal vein Negative 69 5 3.38 × 10 sponding N samples from representative patients with clear cell RCCs tumor thrombi Positive 21 9 −4 c belonging to Clusters A and B (Supplementary Figure S3, available at Predominant Expansive 84 7 1.86 × 10 Carcinogenesis Online) indicated that probes for which DNA meth- growth pattern Infiltrative 6 7 −3 c ylation levels were low in the N samples and for which the degree of Most aggressive Expansive 57 4 2.06 × 10 growth pattern Infiltrative 33 10 DNA hypermethylation in T samples relative to the corresponding N −6 c Tumor necrosis Negative 71 2 4.86 × 10 samples was prominent (marked by red circles in panels E to H in Positive 19 12 Supplementary Figure S3, available at Carcinogenesis Online) were −2 c Invasion to Negative 83 10 3.98 × 10 obvious only in Cluster B, and not in Cluster A. renal pelvis Positive 7 4 Therefore, in order to discriminate clear cell RCCs belonging to −5 c Pathological Stage I 50 0 5.41 × 10 Cluster B from those belonging to Cluster A, we first focused on the TNM stage Stage II 1 1 probes for which the average β value in all N samples was less than Stage III 23 9 0.2 and the incidence of more than 0.4Δβ was markedly higher in Stage IV 16 4 T–N −6 Cluster B relative to Cluster A (P < 1.98 × 10 , Fisher’s exact test). The number of samples in each TNM stage was described in Supplementary Among such probes, 16 (the FAM150A, GRM6, ZNF540, ZFP42, Table S3, available at Carcinogenesis Online. ZNF154, RIMS4, PCDHAC1, KHDRBS2, ASCL2, KCNQ1, PRAC, P-values of <0.05 are in italics. ZNF154, WNT3A, TRH, FAM78A and ZNF671 genes) showed Wilcoxon rank sum test. >0.4Δβ in 6 (42.8%) or more RCCs among the 14 belonging to T–N Fisher’s exact test. Cluster B, but only in 2 (2.2%) or fewer RCCs among the 90 belong- If the tumor showed heterogeneity, findings in the predominant area were ing to Cluster A (Table IIIA). DNA methylation levels (Δβ ) on the described. T–N 16 CpG sites differed completely between Clusters A and B (Supple- If the tumor showed heterogeneity, the most aggressive features of the tumor were described. mentary Figure S4, available at Carcinogenesis Online). In addition, 1490 Single-CpG-resolution methylome analysis of RCCs random forest analysis (27) (Supplementary Figure S5, available at DNA methylation status in N samples were significantly altered relative Carcinogenesis Online) using the 869 probes on which DNA meth- to those in C samples. The single-CpG-resolution analysis revealed that ylation levels (Δβ ) differed significantly between Clusters A and the DNA methylation status of 4830 CpG sites was actually altered T–N B [FDR (q  =  0.01)] identified the top four probes that were able to at the precancerous stage in comparison to normal renal cortex tissue discriminate Cluster B from Cluster A in Table IIIB. Two probes were samples. In addition, it was revealed that alterations at the precancerous shared by Tables IIIA and IIIB. Thus, CpG sites on the 18 probes stages tended to involve DNA hypermethylation [Table I(a)]. Among can be considered as hallmarks of CIMP-positive clear cell RCCs, i.e. the 801 probes we selected, DNA methylation alterations occurred at clear cell RCCs belonging to Cluster B. the precancerous stage and were inherited by, and strengthened in, clear cell RCCs themselves, indicating that DNA methylation alterations on the 801 probes may participate continuously in renal carcinogenesis Discussion from the precancerous stage until cancers have become established. The DNA methylation profiles of these 801 probes clustered clear cell RCCs Here, we have reported the results of methylome analysis of 245 renal into clinicopathologically valid subclusters: clear cell RCCs belonging tissue samples at single-CpG resolution. To our knowledge, no study to Cluster B showed clinicopathological parameters reflecting tumor involving Infinium analysis of such a large number of renal tissue aggressiveness, and patients with Cluster B tumors showed a poorer samples has been reported to date. We have been focusing on DNA outcome. Quantitative reverse transcription–PCR analysis indicated that methylation alterations at the precancerous stage: our previous studies DNA hypermethylation may result in significantly reduced expression using methylation-specific PCR, combined bisulfite restriction enzyme of representative genes listed in Tables IIIA and IIIB and Supplementary analysis and bacterial artificial chromosome arrays suggested that N Table S4, available at Carcinogenesis Online (Supplementary Table S5, samples are already at the precancerous stage associated with DNA available at Carcinogenesis Online). These findings suggest that DNA methylation alterations (17–20). First, we identified the probes on which Fig. 2. (A) Distribution of DNA methylation levels (Δβ ) in all 26 454 probes in 104 clear cell RCCs belonging to Cluster A or B. The DNA methylation T–N levels are shown in the color range maps. Clear cell RCCs belonging to Cluster A are skewed toward DNA hypomethylation (Δβ < −0.1, cold color) relative T–N to DNA hypermethylation (warm color). Clear cell RCCs belonging to Cluster B clearly showed accumulation of DNA hypermethylation (Δβ > 0.1, warm T–N color) relative to DNA hypomethylation (cold color). (B) The proportions of the probes showing the various degrees of DNA hypermethylation, when the tumor tissue (T) sample was compared with the corresponding non-cancerous renal cortex tissue (N) sample (Δβ > 0.1, 0.2, 0.3, 0.4 or 0.5, warm color), to all probes, T–N and the proportions of the probes showing the various degrees of DNA hypomethylation, when the T sample was compared with the corresponding N sample (Δβ < −0.1, −0.2, −0.3, −0.4 or −0.5, cold color), to all probes in Clusters A and B. Bar, standard deviation. The probes showing DNA hypermethylation T–N were markedly accumulated in Cluster B relative to Cluster A, regardless of the degree of DNA hypermethylation (Δβ > 0.1, 0.2, 0.3, 0.4 or 0.5). The probes T–N showing prominent DNA hypomethylation (Δβ < −0.5) were slightly accumulated in Cluster B compared with Cluster A. These data indicated that clear cell T–N RCCs belonging to Cluster B are mainly characterized by accumulation of DNA hypermethylation. 1491 E.Arai et al. Table IIIA. CpG sites as hallmarks of the CpG island methylator phenotype of clear cell RRCs a b c d f Target ID Chr Position CpG island Gene symbol P The number of tumors whose Δβ > 0.4 (%) T–N Cluster A (n = 90) Cluster B (n = 14) −12 cg17162024 8 53,478,454 Y FAM150A 2 (2.2) 12 (85.7) 4.60 × 10 −11 cg14859460 5 178,422,244 Y GRM6 0 (0) 10 (71.4) 3.84 × 10 −8 cg03975694 19 38,042,472 Y ZNF540 2 (2.2) 9 (64.3) 3.64 × 10 −8 cg06274159 4 188,916,867 Y ZFP42 1 (1.1) 8 (57.1) 9.91 × 10 −8 cg08668790 19 58,220,662 Y ZNF154 1 (1.1) 8 (57.1) 9.91 × 10 −7 cg19332710 20 43,438,865 Y RIMS4 2 (2.2) 8 (57.1) 4.68 × 10 −6 cg12629325 5 140,306,458 Y PCDHAC1 2 (2.2) 7 (50) 5.10 × 10 −6 cg18239753 6 62,995,963 Y KHDRBS2 2 (2.2) 7 (50) 5.10 × 10 −6 cg06263495 11 2,292,004 Y ASCL2 2 (2.2) 7 (50) 5.10 × 10 −6 cg17575811 11 2,466,409 Y KCNQ1 1 (1.1) 7 (50) 1.21 × 10 −6 cg12374721 17 46,799,640 Y PRAC 2 (2.2) 7 (50) 5.10 × 10 −7 cg21790626 19 58,220,494 Y ZNF154 0 (0) 7 (50) 1.62 × 10 −6 cg01322134 1 228,194,448 Y WNT3A 0 (0) 6 (42.9) 1.98 × 10 −6 cg01009664 3 129,693,613 Y TRH 0 (0) 6 (42.9) 1.98 × 10 −6 cg12998491 9 134,152,531 Y FAM78A 0 (0) 6 (42.9) 1.98 × 10 −6 cg19246110 19 58,238,928 Y ZNF671 0 (0) 6 (42.9) 1.98 × 10 Probe ID for the Infinium HumanMethylation27 Bead Array. Chromosome. National Center for Biotechnology Information (NCBI) Database (Genome Build 37). Y means CpG island. The probes satisfied the following criteria: (i) the average β value for all samples of non-cancerous renal cortex tissue (N) was <0.2, e e (ii) >0.4Δβ was observed in six or more clear cell RCCs (≥42.9%) in Cluster B, whereas >0.4Δβ in two or fewer clear cell RCCs (≤2.2%) in Cluster A and T–N T–N −6 f (iii) the incidence of >0.4Δβ was markedly higher in Cluster B than in Cluster A (P < 1.98 × 10 , Fisher’s exact test ). T–N Table IIIB. CpG sites as hallmarks of the CpG island methylator phenotype of clear cell RRCs a b c d e Target ID Chr Position CpG island Gene symbol P Δβ (mean ± SD) T–N Cluster A (n = 90) Cluster B (n = 14) f −7 cg17162024 8 53,478,454 Y FAM150A 0.126 ± 0.120 0.499 ± 0.184 3.40 × 10 −7 cg22040627 17 6,617,030 Y SLC13A5 0.045 ± 0.072 0.283 ± 0.103 2.64 × 10 f −7 cg14859460 5 178,422,244 Y GRM6 0.077 ± 0.105 0.434 ± 0.184 1.10 × 10 −7 cg09260089 10 134,599,860 Y NKX6-2 0.078 ± 0.083 0.372 ± 0.150 2.26 × 10 Probe ID of the Infinium HumanMethylation27 Bead Array. Chromosome. NCBI Database (Genome Build 37). Y means CpG island. Top four probes capable of discriminating Cluster B from Cluster A identified by random forest analysis (Supplementary Figure S5, available at Carcinogenesis Online) using the 869 probes on which the DNA methylation levels (Δβ ) were differed significantly between Clusters A and B (Wilcoxon rank sum test). T–N The FAM150A and GRM6 genes were shared by Tables IIIA and IIIB. methylation alterations occurring at the precancerous stage determine CpG islands. Since Cluster B was significantly associated with both both the aggressiveness of RCCs and the outcome of affected patients frequent DNA hypermethylation on CpG islands and distinct clinico- through alterations of gene expression levels. pathological phenotypes of clear cell RCCs, RCCs belonging to Clus- Unsupervised hierarchical clustering based on our previous study ter B can be recognized as CIMP-positive clear cell RCCs on the basis using bacterial artificial chromosome arrays also clustered clear cell of the definition of well-studied CIMP-positive cancers (30,31) such RCCs into clinicopathologically valid subclusters: 14% of examined as colorectal cancer (32) and stomach cancer (33), although Morris RCCs belonged to a subcluster showing clinicopathological param- et al. (28) previously considered that the relevance of the CIMP-pos- eters reflecting tumor aggressiveness and poorer patient outcome itive phenotype to RCCs had not yet been clearly defined. Although (19). DNA methylation profiles in N samples based on BAMCA data McRonald et al. (29) suggested that a subset of RCCs might display were also inherited by the corresponding clear cell RCC developing CIMP based on findings indicating that the distribution of the number in the same patient. In this study, 14% of the clear cell RCCs sub- of methylated CpGs in individual tumors differed from the expected jected to Infinium analysis belonged to Cluster B. BAMCA is suitable Poisson distribution, they did not identify distinct CpG sites that could for detecting DNA methylation alterations occurring in a coordinated become hallmarks for CIMP in the kidney. It has been suggested that, manner on individual large regions of chromosomes (34–37), whereas in order to identify CIMP-positive cancers in specific organs, marker 27 000 Infinium array is suitable for detecting DNA methylation alter - CpG sites or genes that are specific to each organ or histological type ations on promoter regions of specific genes. Different methodologies of tumor should be used (38), rather than classical CIMP marker identified similar clinicopathologically valid subclusters of RCCs, genes (30,31) that were originally identified in colorectal cancers. indicating that such clustering based on DNA methylation profiles is The present single-CpG-resolution analysis identified such hallmark not accidental but reproducible, and may reflect the distinct epigenetic CpG sites for the first time. Using the 18 CpG sites in Tables IIIA and pathway of renal carcinogenesis. IIIB, CIMP-positive RCCs or RCCs equivalent to those in the present In contrast to Cluster A, which appeared to be characterized by Cluster B could be reproducibly identified. These 18 CpG sites may accumulation of DNA hypomethylation (Figure 2A), Cluster B was be useful for further clarifying the molecular basis of the epigenetic clearly characterized by accumulations of DNA hypermethylation on pathway of renal carcinogenesis. 1492 Single-CpG-resolution methylome analysis of RCCs 14. Arai,E. et al. (2010) DNA methylation profiles in precancerous tissue and DNA methylation alterations are known to result in chromosomal cancers: carcinogenetic risk estimation and prognostication based on DNA instability through chromatin configuration changes (39). In fact, ger - methylation status. Epigenomics, 2, 467–481. mline mutations of the de novo DNA methyltransferase DNMT3B 15. Kanai,Y. (2008) Alterations of DNA methylation and clinicopathological gene have been reported in patients with immunodeficiency, centro- diversity of human cancers. Pathol. Int., 58, 544–558. meric instability and facial anomalies (ICF) syndrome, a rare recessive 16. Kanai,Y. et al. (2007) Alterations of DNA methylation associated with abnor- autosomal disorder characterized by DNA hypomethylation of peri- malities of DNA methyltransferases in human cancers during transition from centromeric satellite regions (40). In HCCs and urothelial carcinomas, a precancerous to a malignant state. Carcinogenesis, 28, 2434–2442. DNA hypomethylation of these regions is correlated with copy number 17. Arai,E. et al. (2008) Genetic clustering of clear cell renal cell carcinoma based alterations on chromosomes 1 (41) and 9 (42), respectively, where satel- on array-comparative genomic hybridization: its association with DNA meth- ylation alteration and patient outcome. Clin. Cancer Res., 14, 5531–5539. lite regions are plentiful. Correlations between the clustering based on 18. Arai,E. et  al. (2006) Regional DNA hypermethylation and DNA methyl- Infinium assay and copy number alterations should be further examined. transferase (DNMT) 1 protein overexpression in both renal tumors and cor- Taken together, the data suggest that in CIMP-positive clear cell RCCs responding nontumorous renal tissues. Int. J. Cancer, 119, 288–296. belonging to Cluster B, DNA hypermethylation of distinct CpG islands 19. Arai,E. et al. (2009) Genome-wide DNA methylation profiles in both pre- participates even in the very early and precancerous stages. Such DNA cancerous conditions and clear cell renal cell carcinomas are correlated with methylation alterations occurring in the precancerous stages may induce malignant potential and patient outcome. Carcinogenesis, 30, 214–221. more aggressive tumor phenotypes and poorer patient outcome in Clus- 20. Arai,E. et  al. (2011) Genome-wide DNA methylation profiles in renal ter B. On the other hand, in the other pathway of renal carcinogenesis tumors of various histological subtypes and non-tumorous renal tissues. leading to clear cell RCCs in Cluster A, DNA hypomethylation may be a Pathobiology, 78, 1–9. 21. Bibikova,M. et al. (2009) Genome-wide DNA methylation profiling using later event (Table I) than DNA hypermethylation on CpG islands. We are Infinium® assay. Epigenomics, 1, 177–200. now performing exome, transcriptome and proteome analyses of RCCs 22. Eble,J.N. et  al. (2004) Renal cell carcinoma. World Health Organization belonging to both clusters. Such multilayer omics analyses may identify Classification of Tumours. Pathology and Genetics. Tumours of the Uri- the upstream genetic events inducing DNA methylation profiles and key nary System and Male Genital Organs. IARC Press, Lyon, pp. 10–43. signal pathways that characterize Clusters A and B. 23. Fuhrman,S.A. et al. (1982) Prognostic significance of morphologic param- eters in renal cell carcinoma. Am. J. Surg. Pathol., 6, 655–663. 24. Sobin,L.H. et al. (2002) International Union Against Cancer (UICC). TNM Supplementary material Classification of Malignant Tumors. 6th edn. Wiley-Liss, New York, pp. 193–195. Supplementary Figures S1–S5 and Tables S1–S5 can be found at 25. Kanai,T. et al. (1987) Pathology of small hepatocellular carcinoma. A pro- http://carcin.oxfordjournals.org/. posal for a new gross classification. Cancer, 60, 810–819. 26. Sambrook,J. et al. (2001) Molecular Cloning: A Laboratory Manual. 3rd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 6.14–6.15. Funding 27. Breiman,L. (2001) Random forests. Mach. Learn., 45, 5–32. 28. Morris,M.R. et  al. (2010) Epigenetics of renal cell carcinoma: the path Program for Promotion of Fundamental Studies in Health Sciences towards new diagnostics and therapeutics. Genome Med., 2, 59. of the National Institute of Biomedical Innovation (NiBio); Grant-in- 29. McRonald,F.E. et al. (2009) CpG methylation profiling in VHL related and Aid for the Third Term Comprehensive 10-Year Strategy for Cancer VHL unrelated renal cell carcinoma. Mol. Cancer, 8, 31. Control from the Ministry of Health, Labor and Welfare of Japan; 30. Issa,J.P. (2004) CpG island methylator phenotype in cancer. Nat. Rev. Can- National Cancer Center Research and Development Fund; Grants-in- cer, 4, 988–993. Aid for Scientific Research (B) and for Young Scientists (B) from the 31. Toyota,M. et al. (1999) CpG island methylator phenotype in colorectal can- Japan Society for the Promotion of Science (JSPS). cer. Proc. Natl. Acad. Sci. U.S.A., 96, 8681–8686. 32. Shen,L., et al. (2007) Integrated genetic and epigenetic analysis identifies three different subclasses of colon cancer. Proc. Natl. Acad. Sci. U.S.A., Acknowledgements 104, 18654–18659. 33. Toyota,M. et al. (1999) Aberrant methylation in gastric cancer associated Conflict of Interest Statement: None declared. with the CpG island methylator phenotype. Cancer Res., 59, 5438–5442. 34. Nagashio,R. et al. (2011) Carcinogenetic risk estimation based on quantifi- cation of DNA methylation levels in liver tissue at the precancerous stage. References Int. J. Cancer, 129, 1170–1179. 1. Ljungberg,B. et al. (2011) The epidemiology of renal cell carcinoma. Eur. 35. Gotoh,M. et al. (2011) Diagnosis and prognostication of ductal adenocar- Urol., 60, 615–621. cinomas of the pancreas based on genome-wide DNA methylation profil- 2. Arai,E. et  al. (2011) Genetic and epigenetic alterations during renal car- ing by bacterial artificial chromosome array-based methylated CpG island cinogenesis. Int. J. Clin. Exp. Pathol., 4, 58–73. amplification. J. Biomed. Biotechnol., 2011, 780–836. 3. Baldewijns,M.M. et  al. (2010) VHL and HIF signalling in renal cell car- 36. Nishiyama,N. et  al. (2010) Genome-wide DNA methylation profiles in cinogenesis. J. Pathol., 221, 125–138. urothelial carcinomas and urothelia at the precancerous stage. Cancer Sci., 4. Barrett,I.P. (2010) Cancer genome analysis informatics. Methods Mol. 101, 231–240. Biol., 628, 75–102. 37. Arai,E. et al. (2009) Genome-wide DNA methylation profiles in liver tissue 5. Zhang,J. et al. (2011) International Cancer Genome Consortium Data Portal- at the precancerous stage and in hepatocellular carcinoma. Int. J. Cancer, -a one-stop shop for cancer genomics data. Database (Oxford), 2011, bar026. 125, 2854–2862. 6. Brannon,A.R. et al. (2010) Renal cell carcinoma: where will the state-of- 38. Abe,M. et al. (2008) Identification of genes targeted by CpG island meth- the-art lead us? Curr. Oncol. Rep., 12, 193–201. ylator phenotype in neuroblastomas, and their possible integrative involve- 7. Dalgliesh,G.L. et  al. (2010) Systematic sequencing of renal carcinoma ment in poor prognosis. Oncology, 74, 50–60. reveals inactivation of histone modifying genes. Nature, 463, 360–363. 39. Eden,A. et  al. (2003) Chromosomal instability and tumors promoted by 8. van Haaften,G. et al. (2009) Somatic mutations of the histone H3K27 dem- DNA hypomethylation. Science, 300, 455. ethylase gene UTX in human cancer. Nat. Genet., 41, 521–523. 40. Hansen,R.S. et  al. (1999) The DNMT3B DNA methyltransferase gene is 9. Varela,I. et  al. (2011) Exome sequencing identifies frequent mutation of mutated in the ICF immunodeficiency syndrome. Proc. Natl. Acad. Sci. the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature, 469, U.S.A., 96, 14412–14417. 539–542. 41. Wong,N. et al. (2001) Hypomethylation of chromosome 1 heterochromatin 10. Baylin,S.B. et  al. (2011) A decade of exploring the cancer epigenome - DNA correlates with q-arm copy gain in human hepatocellular carcinoma. biological and translational implications. Nat. Rev. Cancer, 11, 726–734. Am. J. Pathol. 59, 465–471. 11. Boumber,Y. et al. (2011) Epigenetics in cancer: what’s the future? Oncol- 42. Nakagawa,T. et al. (2005) DNA hypomethylation on pericentromeric satel- ogy (Williston Park), 25, 220–226, 228. lite regions significantly correlates with loss of heterozygosity on chromo- 12. Jones,P.A. et al. (2007) The epigenomics of cancer. Cell, 128, 683–692. some 9 in urothelial carcinomas. J. Urol., 173, 243–246. 13. Kanai,Y. (2010) Genome-wide DNA methylation profiles in precancerous conditions and cancers. Cancer Sci., 101, 36–45. Received December 11, 2012; revised April 17, 2012; accepted May 11, 2012 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Carcinogenesis Pubmed Central

Single-CpG-resolution methylome analysis identifies clinicopathologically aggressive CpG island methylator phenotype clear cell renal cell carcinomas

Loading next page...
 
/lp/pubmed-central/single-cpg-resolution-methylome-analysis-identifies-Jh0n8cVQhy

References (80)

Publisher
Pubmed Central
Copyright
© The Author 2012. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0143-3334
eISSN
1460-2180
DOI
10.1093/carcin/bgs177
Publisher site
See Article on Publisher Site

Abstract

Carcinogenesis vol.33 no.8 pp.1487–1493, 2012 doi:10.1093/carcin/bgs177 Advance Access Publication May 18, 2012 Single-CpG-resolution methylome analysis identifies clinicopathologically aggressive CpG island methylator phenotype clear cell renal cell carcinomas 1 2 1 efforts (6). Such efforts have revealed that renal carcinogenesis Eri Arai , Suenori Chiku , Taisuke Mori , 1 3 involves inactivation of histone-modifying genes, such as SETD2 (7), Masahiro Gotoh , Tohru Nakagawa , 3 1, a histone H3 lysine 36 methyltransferase, JARIDIC (KDM5C (7)), Hiroyuki Fujimoto and Yae Kanai * a histone H3 lysine 4 demethylase, and UTX (KMD6A (8)), a histone Division of Molecular Pathology, National Cancer Center Research Institute, H3 lysine 27 demethylase, as well as the SWI/SNF chromatin-remod- Tokyo 104-0045, Japan, Science Solutions Division, Mizuho Information eling complex gene, PBRM1 (9). Non-synonymous mutations of the and Research Institute, Inc., Tokyo 101-8443, Japan and Department of NF2 gene and truncating mutations of the MLL2 gene have also been Urology, National Cancer Center Hospital, Tokyo 104-0045, Japan reported (7). However, such gene mutations cannot fully explain the *To whom correspondence should be addressed. Tel: +81 3 3542 2511; clinicopathological diversity of clear cell RCCs. Fax: +81 3 3248 2463; Not only genetic, but also epigenetic events appear to accumulate Email: ykanai@ncc.go.jp during carcinogenesis, and both types of event reflect the clinicopatho- To clarify the significance of DNA methylation alterations dur - logical diversity of cancers in various organs in association with each ing renal carcinogenesis, methylome analysis using single-CpG- other (10–12). DNA methylation alterations are one of the most con- resolution Infinium array was performed on 29 normal renal sistent epigenetic changes in human cancers (13–16). In fact, on the cortex tissue (C) samples, 107 non-cancerous renal cortex tis- basis of methylation-specific PCR (MSP), combined bisulfite restric- sue (N) samples obtained from patients with clear cell renal cell tion enzyme analysis (17,18) and bacterial artificial chromosome array- carcinomas (RCCs) and 109 tumorous tissue (T) samples. DNA based methylated CpG island amplification (BAMCA (19,20)), we have methylation levels at 4830 CpG sites were already altered in N suggested that non-cancerous renal cortex tissue obtained from patients samples compared with C samples. Unsupervised hierarchical with RCCs is already at the precancerous stage associated with DNA clustering analysis based on DNA methylation levels at the 801 methylation alterations, even though no remarkable histological changes CpG sites, where DNA methylation alterations had occurred in are evident and there is no association with chronic inflammation or N samples and were inherited by and strengthened in T samples, persistent infection with viruses or other pathogenic microorganisms. clustered clear cell RCCs into Cluster A (n = 90) and Cluster B Genome-wide analysis using BAMCA revealed that DNA methylation (n  =  14). Clinicopathologically aggressive tumors were accumu- status in non-cancerous renal cortex tissue at the precancerous stage lated in Cluster B, and the cancer-free and overall survival rates was basically inherited by the corresponding clear cell RCC in indi- of patients in this cluster were significantly lower than those of vidual patients (19). DNA methylation alterations at the precancerous patients in Cluster A. Clear cell RCCs in Cluster B were char- stage may confer further susceptibility to genetic and epigenetic alter- acterized by accumulation of DNA hypermethylation on CpG ations and generate more malignant clear cell RCCs (2,13). However, islands and considered to be CpG island methylator phenotype in our previous studies using BAMCA, the resolution and the number (CIMP)-positive cancers. DNA hypermethylation of the CpG sites of probes were limited. Therefore, further analysis is needed to clarify on the FAM150A, GRM6, ZNF540, ZFP42, ZNF154, RIMS4, the significance of DNA methylation alterations in renal carcinogenesis. PCDHAC1, KHDRBS2, ASCL2, KCNQ1, PRAC, WNT3A, Recently, methylome analysis using the Infinium array has made TRH, FAM78A, ZNF671, SLC13A5 and NKX6-2 genes became it possible to interrogate 27 000 highly informative CpG sites at sin- hallmarks of CIMP in RCCs. On the other hand, Cluster A was gle-CpG resolution (21). In order to clarify the significance of DNA characterized by genome-wide DNA hypomethylation. These methylation alterations during renal carcinogenesis, we used the data indicated that DNA methylation alterations at precancerous Infinium BeadChip system to perform genome-wide DNA methyla- stages may determine tumor aggressiveness and patient outcome. tion analysis of 29 samples of normal renal cortex tissue (C) obtained Accumulation of DNA hypermethylation on CpG islands and from patients without any primary renal tumors, 107 samples of non- genome-wide DNA hypomethylation may each underlie distinct cancerous renal cortex tissue (N) from patients with clear cell RCCs pathways of renal carcinogenesis. and 109 samples of tissue from the tumors (T) themselves. Correla- tions between the genome-wide DNA methylation profiles and clin- icopathological parameters were then examined. Introduction Materials and methods Clear cell renal cell carcinoma (RCC) is the most common histological Patients and tissue samples subtype of adult kidney cancer and frequently affects working-age adults in midlife. In general, RCCs at an early stage are curable by The 109 T samples and corresponding 107 N samples showing no remarkable histological changes were obtained from materials that had been surgically nephrectomy. However, some RCCs relapse and metastasize to dis- resected from 110 patients with primary clear cell RCCs. These patients did tant organs, even if the resection has been considered complete (1). not receive preoperative treatment and underwent nephrectomy at the National Such clinicopathological diversity may be attributable to distinct path- Cancer Center Hospital, Tokyo, Japan. There were 79 men and 31 women with a ways of renal carcinogenesis (2). It is well known that clear cell RCCs mean (±SD) age of 62.8 ± 10.3 years (range 36–85 years). Histological diagnosis are characterized by inactivation of the Von Hippel–Lindau tumor- was made in accordance with the World Health Organization classification (22) suppressor gene (3). In addition, systematic resequencing and exome (Supplementary Figure S1, available at Carcinogenesis Online). All the tumors analysis of RCCs are now being performed by The Cancer Genome were graded on the basis of criteria described previously (23) and classified Atlas (4), The Cancer Genome Project (5) and other international according to the pathological Tumor-Node-Metastasis (TNM) classification (24). The criteria for macroscopic configuration of RCC (17–19) followed those Abbreviations: BAMCA, bacterial artificial chromosome array-based meth- established for hepatocellular carcinoma (HCC): type 3 (contiguous multinodular ylated CpG island amplification; C, normal renal cortex tissue obtained from type) HCCs show poorer histological differentiation and a higher incidence of patients without any primary renal tumor; CIMP, CpG island methylator phe- intrahepatic metastasis than type 1 (single nodular type) and type 2 (single notype; HCC, hepatocellular carcinoma; N, non-cancerous renal cortex tissue nodular type with extranodular growth) HCCs (25). The presence or absence obtained from patients with clear cell renal cell carcinomas; NCBI, National of vascular involvement was examined microscopically on slides stained with Center for Biotechnology Information; RCC, renal cell carcinoma; T, tumor- hematoxylin–eosin and elastica van Gieson. The presence or absence of tumor ous tissue; TNM, Tumor-Node-Metastasis. thrombi in the main trunk of the renal vein was examined macroscopically. © The Author 2012. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons. org/licenses/by-nc/3.0), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. E.Arai et al. RCC is usually enclosed by a fibrous capsule and well demarcated, and hardly pyrosequencing method (Supplementary Figure S2, available at Car- ever contains fibrous stroma between cancer cells. Therefore, we were able to cinogenesis Online). With regard to the well-known methylation- obtain cancer cells from surgical specimens, avoiding contamination with both silencing Von Hippel–Lindau tumor-suppressor gene (probe Target ID: non-cancerous epithelial cells and stromal cells. cg22782492), DNA hypermethylation (Δβ > 0.1) was detected in 12 T–N For comparison, 29 samples of normal renal cortex tissue (C1–C29) were (12%) of 104 patients, for whom both N and T samples were assayed obtained from materials that had been surgically resected from 29 patients in the same experimental batch. This incidence corresponded to that in without any primary renal tumor. These patients included 18 men and previous reports (28,29). Taken together, the data confirmed the reliabil- 11 women with a mean (±SD) age of 61.4 ± 10.8 years (range 31–81 years). Of ity of the present Infinium assay. these patients, 22 had undergone nephroureterectomy for urothelial carcinomas of the renal pelvis and ureter, 6 had undergone nephrectomy with resection of retro- Although precancerous conditions in the kidney have been rarely peritoneal sarcoma around the kidney, and the remaining 1 had undergone para- described, our previous study suggested that N samples are already at aortic lymph node dissection for metastatic germ cell tumor, which resulted in precancerous stages, from the viewpoint of altered DNA methylation, simultaneous nephrectomy because it was not possible to preserve the renal artery. despite the absence of any remarkable histological changes and the All patients included in this study provided written informed consent, and lack of association with chronic inflammation and persistent infection the study was approved by the Ethics Committee of the National Cancer Cen- with viruses or other pathogenic microorganisms (17–20). (a) In ter, Tokyo, Japan. fact, the logistic model adjusted by sex, age and experimental batch Infinium assay revealed that DNA methylation levels on 4830 probes were already High-molecular-weight DNA from fresh frozen tissue samples was extracted altered in N samples compared with those in C samples (FDR, using phenol–chloroform, followed by dialysis (26). Five-hundred-nanogram q = 0.01, Table I). (b) In order to reveal DNA methylation alterations aliquots of DNA were subjected to bisulfite conversion using an EZ DNA inherited by clear cell RCCs themselves, ordered differences of DNA TM Methylation-Gold Kit (Zymo Research, Irvine, CA). Subsequently DNA methylation level from C to N and then to T samples were examined methylation status at 27 578 CpG loci was examined at single-CpG resolution by the cumulative logit model adjusted by sex, age and experimental using the Infinium HumanMethylation27 Bead Array (Illumina, San Diego, batch. Ordered differences from C to N and then to T samples were CA). This array contains CpG sites located within the proximal promoter observed on 11  089 probes (FDR, q  =  0.01, Table I). (c) In order regions of the transcription start sites of 14 475 consensus coding sequences in to reveal the cancer-prone DNA methylation alterations, differences the National Center for Biotechnology Information Database. On average, two assays were selected per gene, and from 3 to 20 CpG sites for more than 200 in DNA methylation levels between 104 paired samples of N and cancer-related and imprinted genes. Forty control probes were employed for T assayed in the same experimental batch were examined using the each array; these included staining, hybridization, extension, bisulfite conver - Wilcoxon matched pairs test. Significant differences between N and sion and negative controls. An Evo robot (Tecan, Switzerland) was used for the corresponding clear cell RCCs themselves were observed on automated sample processing. Whole-genome amplification was performed 10 870 probes (FDR, q = 0.01, Table I). using the Infinium Assay Kit (Illumina (21)). After hybridization, the specifi- DNA hypermethylation frequently occurred at the very early stages cally hybridized DNA was fluorescence labeled by a single-base extension of renal carcinogenesis [(a) in Table I], whereas DNA hypometh- reaction and detected using a BeadScan reader (Illumina) in accordance with ylation was also observed during progression to established cancers the manufacturer’s protocols. The data were then assembled using GenomeS- [(b) and (c) in Table I]. Eight hundred and one probes satisfied all of tudio methylation software (Illumina). At each CpG site, the ratio of the fluo- rescent signal was measured using a methylated probe relative to the sum of the the above criteria (a)–(c) (Table I): DNA methylation alterations on methylated and unmethylated probes, i.e. the so-called β-value, which ranges these 801 probes (Supplementary Table S2, available at Carcinogen- from 0.00 to 1.00, reflecting the methylation level of an individual CpG site. esis Online) were already evident in N samples, and were inherited by and strengthened in T samples. Statistics In the Infinium assay, the call proportions (P-values for detection of signals above the background <0.01) for 32 probes (shown in Supplementary Table S1, available at Carcinogenesis Online) in all of the tissue samples exam- ined were less than 90%. Since such a low proportion may be attributable to Table I. DNA methylation alterations during renal carcinogenesis polymorphism at the probe CpG sites, these 32 probes were excluded from the present assay. In addition, all CpG sites on chromosomes X and Y were The number of probes showing DNA hypermethylation and DNA excluded, to avoid any gender-specific methylation bias, leaving a final total hypomethylation of 26 454 autosomal CpG sites. (a) The probes on which DNA methylation levels were altered in samples of Infinium probes showing significant differences in DNA methylation lev- non-cancerous renal cortex tissue (N) obtained from patients with clear els between the 29 C and 107 N samples were identified by a logistic model cell RCCs relative to those in samples of normal renal cortex tissue (C) adjusted by sex, age and experimental batch. Ordered differences from 29 obtained from patients without any primary renal tumor. (Logistic model C to 107 N and then to 109 T samples themselves were examined by the adjusted by sex, age and experimental batch; FDR, q = 0.01.) cumulative logit model adjusted by sex, age and experimental batch. Dif- DNA hypermethylation (β > β ) N C ferences of DNA methylation status between 104 paired samples of N and DNA hypomethylation (β < β ) the corresponding T obtained from a single patient and assayed in the same N C Total 4830 experimental batch were examined by Wilcoxon matched pairs test. A false (b) The probes on which DNA methylation levels showed ordered differences discovery rate (FDR) of q = 0.01 was considered significant. Unsupervised from C to N, and then to tumorous tissue (T) samples. hierarchical clustering (Euclidan distance, Ward method) based on DNA (Cumulative logit model adjusted by sex, age and experimental batch; FDR, methylation levels (Δβ ) was performed in patients with clear cell RCCs. T–N q = 0.01.) Correlations between clusters of patients and clinicopathological parameters DNA hypermethylation (β < β < β , C N T were examined using Wilcoxon rank sum test and Fisher’s exact test. Survival . . β < β =. β or β =. β < β ) curves of patients belonging to each cluster were calculated by the Kaplan– C N T C N T DNA hypomethylation (β > β > β , Meier method, and the differences were compared by the log-rank test. The C N T . . number of Infinium assay probes showing DNA hyper- or hypomethylation β > β =. β or β =. β > β ) C N T C N T in each cluster and the average DNA methylation levels (Δβ ) of each clus- Total 11 089 T–N ter were examined using Wilcoxon rank sum test at a significance level of (c) The probes on which DNA methylation levels differed between T and the P < 0.05. The CpG sites discriminating the clusters were identified by Fish- corresponding N samples (Wilcoxon matched pairs test; FDR, q = 0.01) er’s exact test and random forest analysis (27). 5408 DNA hypermethylation (Δβ > 0) T–N DNA hypomethylation (Δβ < 0) 5462 T–N Total 10 870 Results Among the 4589 probes, 2675 showed DNA hypermethylation in T samples DNA methylation alterations during renal carcinogenesis than in C samples (β > β ; FDR, q = 0.01). T C First, DNA methylation levels of representative CpG sites based on Among the 241 probes, 126 showed DNA hypomethylation in T samples the Infinium assay were clearly verified using a highly quantitative than in C samples (β < β ; FDR, q = 0.01). T C 1488 Single-CpG-resolution methylome analysis of RCCs Fig. 1. Unsupervised hierarchical clustering using DNA methylation levels (Δβ ) on the 801 probes in 104 patients with clear cell RCCs. The 801 probes T–N satisfied all of the criteria (a), (b) and (c) in ‘DNA methylation alterations during renal carcinogenesis’ in Results and Table I. On the 801 probes, DNA methylation alterations occurred at the precancerous stages and were inherited by and strengthened in clear cell RCCs themselves. (A) 104 patients with clear cell RCCs were hierarchically clustered into Cluster A (n = 90) and Cluster B (n = 14). The DNA methylation levels (Δβ ) are shown in the color range maps. T–N −6 The cluster trees for patients and probes are shown at the top and left of the panel, respectively. (B) The cancer-free (P = 3.59 × 10 ) survival rates of Stage I–III −2 patients in Cluster B were significantly lower (log-rank test) than those of patients in Cluster A. Overall (P = 1.32 × 10 ) survival rates of all patients in Cluster B were significantly lower (log-rank test) than those of patients in Cluster A. 1489 E.Arai et al. Epigenetic clustering of clear cell RCCs DNA methylation profiles of clear cell RCCs belonging to each cluster Unsupervised hierarchical clustering using DNA methylation levels The distribution of DNA methylation levels (Δβ ) in all 26  454 T–N (Δβ ) on the above 801 probes, on which DNA methylation altera- probes for 104 clear cell RCCs belonging to Cluster A or B is sum- T–N tions occurred at the precancerous stages and may continuously par- marized along chromosomes in Figure 2A. Clear cell RCCs belonging ticipate in renal carcinogenesis, subclustered 104 patients with clear to Cluster B clearly showed accumulation of DNA hypermethylation cell RCCs, of whom both N and T samples were assayed in the same (Δβ > 0.1) relative to DNA hypomethylation, whereas clear cell T–N experimental batch, into Cluster A (n  =  90) and Cluster B (n  =  14, RCCs belonging to Cluster A showed greater DNA hypomethylation Figure 1A). The clinicopathological parameters of clear cell RCCs (Δβ < −0.1) relative to DNA hypermethylation (Figure 2A). T–N belonging to Clusters A and B are summarized in Table II. (The The proportions of the probes showing the various degrees of DNA number of samples for each TNM stage is also described in Supple- hypermethylation in T samples compared with the corresponding N mentary Table S3, available at Carcinogenesis Online.) Epigenetic samples (Δβ > 0.1, 0.2, 0.3, 0.4 or 0.5) for all 26 454 probes, and T–N clustering of clear cell RCCs was dependent on neither age nor sex of the proportions of the probes showing various degrees of DNA hypo- the patients (Table II). Clear cell RCCs belonging to Cluster B had a methylation in T samples compared with the corresponding N samples larger diameter, more frequent macroscopically evident extranodular (Δβ < −0.1, −0.2, −0.3, −0.4 or −0.5) for all 26 454 probes in clear T–N (type 2) or multinodular (type 3) growth, vascular involvement, renal cell RCCs belonging to Clusters A and B are summarized in Figure vein tumor thrombi, infiltrating growth, tumor necrosis and renal pel- 2B. Although the probes showing prominent DNA hypomethylation vis invasion, and also had higher histological grades and pathologi- (Δβ < −0.5) were accumulated slightly more in Cluster B than in T–N cal TNM stages than those in Cluster A (Table II). Figure 1B shows Cluster A, the incidence of DNA hypomethylation in Clusters A and B the Kaplan–Meier survival curves of patients belonging to Clusters did not reach a statistically significant difference (Δβ < −0.1, −0.2, T–N A and B. The period covered ranged from 42 to 4024  days (mean, −0.3 or −0.4, Figure 2B). On the other hand, the probes showing DNA 1821 days). The cancer-free and overall survival rates of patients in hypermethylation were markedly accumulated in Cluster B relative to Cluster B were significantly lower than those of patients in Cluster A Cluster A, regardless of the degree of DNA hypermethylation (Δβ T–N −6 −2 (P = 3.59 × 10 and P = 1.32 × 10 , respectively, Figure 1B). > 0.1, 0.2, 0.3, 0.4 or 0.5, Figure 2B). These data indicate that clear cell RCCs belonging to Cluster B are characterized by accumulation of DNA hypermethylation. Table II. Correlation between the subclassification of patients with clear The top 61 probes (including the 60th and 61st, which showed cell RRCs based on DNA methylation profiles and the clinicopathological equivalent P-values) on which DNA methylation levels (Δβ ) dif- parameters T–N −6 fered markedly between Clusters A and B (P < 1.056  × 10 , Wil- Clinicopathological parameters Cluster A Cluster B P coxon rank sum test) are listed in Supplementary Table S4, available (n = 90) (n = 14) at Carcinogenesis Online. Although only 19 246 probes (72.8%) out −2 b of the total of 26  454 were located within CpG islands, 60 (98.4%) Age 62.08 ± 67.36 ± 8.36 × 10 10.08 11.06 of the top 61 probes located within CpG islands showed DNA hyper- −1 c Sex Male 63 11 5.47 × 10 methylation in clear cell RCCs belonging to Cluster B (Δβ > 0.097, T–N Female 27 3 Supplementary Table S4, available at Carcinogenesis Online): only −4 b Tumor 5.10 ± 8.75 ± 1.07 × 10 1 probe among the top 61 was located within a non-CpG island and diameter (cm) 3.19 2.85 showed DNA hypomethylation (Δβ   =  −0.425 ± 0.096 in Cluster −4 c T–N Macroscopic Type 1 37 1 6.29 × 10 B). Taken together, the data indicated that Cluster B is significantly configuration Type 2 29 2 correlated with clinicopathological phenotype and characterized by Type 3 24 11 −6 c frequent DNA hypermethylation on CpG islands. Such characteristics Predominant G1 47 1 8.33 × 10 of clear cell RCCs in Cluster B are similar to those of CpG island histological G2 35 4 grades G3 7 7 methylator phenotype (CIMP)-positive cancers (30,31) in other well- G4 1 2 studied organs, such as those of the colon (32) and stomach (33). In −4 c Highest G1 8 0 5.67 × 10 other words, our single-CpG-resolution methylome analysis identi- histological G2 43 1 fied, for the first time, CIMP-positive clear cell RCCs as Cluster B. grades G3 24 4 G4 15 9 −4 c Vascular Negative 54 1 2.45 × 10 Hallmark CpG sites of CIMP-positive clear cell RCCs involvement Positive 36 13 Scattergrams of DNA methylation levels (β values) in T and the corre- −3 c Renal vein Negative 69 5 3.38 × 10 sponding N samples from representative patients with clear cell RCCs tumor thrombi Positive 21 9 −4 c belonging to Clusters A and B (Supplementary Figure S3, available at Predominant Expansive 84 7 1.86 × 10 Carcinogenesis Online) indicated that probes for which DNA meth- growth pattern Infiltrative 6 7 −3 c ylation levels were low in the N samples and for which the degree of Most aggressive Expansive 57 4 2.06 × 10 growth pattern Infiltrative 33 10 DNA hypermethylation in T samples relative to the corresponding N −6 c Tumor necrosis Negative 71 2 4.86 × 10 samples was prominent (marked by red circles in panels E to H in Positive 19 12 Supplementary Figure S3, available at Carcinogenesis Online) were −2 c Invasion to Negative 83 10 3.98 × 10 obvious only in Cluster B, and not in Cluster A. renal pelvis Positive 7 4 Therefore, in order to discriminate clear cell RCCs belonging to −5 c Pathological Stage I 50 0 5.41 × 10 Cluster B from those belonging to Cluster A, we first focused on the TNM stage Stage II 1 1 probes for which the average β value in all N samples was less than Stage III 23 9 0.2 and the incidence of more than 0.4Δβ was markedly higher in Stage IV 16 4 T–N −6 Cluster B relative to Cluster A (P < 1.98 × 10 , Fisher’s exact test). The number of samples in each TNM stage was described in Supplementary Among such probes, 16 (the FAM150A, GRM6, ZNF540, ZFP42, Table S3, available at Carcinogenesis Online. ZNF154, RIMS4, PCDHAC1, KHDRBS2, ASCL2, KCNQ1, PRAC, P-values of <0.05 are in italics. ZNF154, WNT3A, TRH, FAM78A and ZNF671 genes) showed Wilcoxon rank sum test. >0.4Δβ in 6 (42.8%) or more RCCs among the 14 belonging to T–N Fisher’s exact test. Cluster B, but only in 2 (2.2%) or fewer RCCs among the 90 belong- If the tumor showed heterogeneity, findings in the predominant area were ing to Cluster A (Table IIIA). DNA methylation levels (Δβ ) on the described. T–N 16 CpG sites differed completely between Clusters A and B (Supple- If the tumor showed heterogeneity, the most aggressive features of the tumor were described. mentary Figure S4, available at Carcinogenesis Online). In addition, 1490 Single-CpG-resolution methylome analysis of RCCs random forest analysis (27) (Supplementary Figure S5, available at DNA methylation status in N samples were significantly altered relative Carcinogenesis Online) using the 869 probes on which DNA meth- to those in C samples. The single-CpG-resolution analysis revealed that ylation levels (Δβ ) differed significantly between Clusters A and the DNA methylation status of 4830 CpG sites was actually altered T–N B [FDR (q  =  0.01)] identified the top four probes that were able to at the precancerous stage in comparison to normal renal cortex tissue discriminate Cluster B from Cluster A in Table IIIB. Two probes were samples. In addition, it was revealed that alterations at the precancerous shared by Tables IIIA and IIIB. Thus, CpG sites on the 18 probes stages tended to involve DNA hypermethylation [Table I(a)]. Among can be considered as hallmarks of CIMP-positive clear cell RCCs, i.e. the 801 probes we selected, DNA methylation alterations occurred at clear cell RCCs belonging to Cluster B. the precancerous stage and were inherited by, and strengthened in, clear cell RCCs themselves, indicating that DNA methylation alterations on the 801 probes may participate continuously in renal carcinogenesis Discussion from the precancerous stage until cancers have become established. The DNA methylation profiles of these 801 probes clustered clear cell RCCs Here, we have reported the results of methylome analysis of 245 renal into clinicopathologically valid subclusters: clear cell RCCs belonging tissue samples at single-CpG resolution. To our knowledge, no study to Cluster B showed clinicopathological parameters reflecting tumor involving Infinium analysis of such a large number of renal tissue aggressiveness, and patients with Cluster B tumors showed a poorer samples has been reported to date. We have been focusing on DNA outcome. Quantitative reverse transcription–PCR analysis indicated that methylation alterations at the precancerous stage: our previous studies DNA hypermethylation may result in significantly reduced expression using methylation-specific PCR, combined bisulfite restriction enzyme of representative genes listed in Tables IIIA and IIIB and Supplementary analysis and bacterial artificial chromosome arrays suggested that N Table S4, available at Carcinogenesis Online (Supplementary Table S5, samples are already at the precancerous stage associated with DNA available at Carcinogenesis Online). These findings suggest that DNA methylation alterations (17–20). First, we identified the probes on which Fig. 2. (A) Distribution of DNA methylation levels (Δβ ) in all 26 454 probes in 104 clear cell RCCs belonging to Cluster A or B. The DNA methylation T–N levels are shown in the color range maps. Clear cell RCCs belonging to Cluster A are skewed toward DNA hypomethylation (Δβ < −0.1, cold color) relative T–N to DNA hypermethylation (warm color). Clear cell RCCs belonging to Cluster B clearly showed accumulation of DNA hypermethylation (Δβ > 0.1, warm T–N color) relative to DNA hypomethylation (cold color). (B) The proportions of the probes showing the various degrees of DNA hypermethylation, when the tumor tissue (T) sample was compared with the corresponding non-cancerous renal cortex tissue (N) sample (Δβ > 0.1, 0.2, 0.3, 0.4 or 0.5, warm color), to all probes, T–N and the proportions of the probes showing the various degrees of DNA hypomethylation, when the T sample was compared with the corresponding N sample (Δβ < −0.1, −0.2, −0.3, −0.4 or −0.5, cold color), to all probes in Clusters A and B. Bar, standard deviation. The probes showing DNA hypermethylation T–N were markedly accumulated in Cluster B relative to Cluster A, regardless of the degree of DNA hypermethylation (Δβ > 0.1, 0.2, 0.3, 0.4 or 0.5). The probes T–N showing prominent DNA hypomethylation (Δβ < −0.5) were slightly accumulated in Cluster B compared with Cluster A. These data indicated that clear cell T–N RCCs belonging to Cluster B are mainly characterized by accumulation of DNA hypermethylation. 1491 E.Arai et al. Table IIIA. CpG sites as hallmarks of the CpG island methylator phenotype of clear cell RRCs a b c d f Target ID Chr Position CpG island Gene symbol P The number of tumors whose Δβ > 0.4 (%) T–N Cluster A (n = 90) Cluster B (n = 14) −12 cg17162024 8 53,478,454 Y FAM150A 2 (2.2) 12 (85.7) 4.60 × 10 −11 cg14859460 5 178,422,244 Y GRM6 0 (0) 10 (71.4) 3.84 × 10 −8 cg03975694 19 38,042,472 Y ZNF540 2 (2.2) 9 (64.3) 3.64 × 10 −8 cg06274159 4 188,916,867 Y ZFP42 1 (1.1) 8 (57.1) 9.91 × 10 −8 cg08668790 19 58,220,662 Y ZNF154 1 (1.1) 8 (57.1) 9.91 × 10 −7 cg19332710 20 43,438,865 Y RIMS4 2 (2.2) 8 (57.1) 4.68 × 10 −6 cg12629325 5 140,306,458 Y PCDHAC1 2 (2.2) 7 (50) 5.10 × 10 −6 cg18239753 6 62,995,963 Y KHDRBS2 2 (2.2) 7 (50) 5.10 × 10 −6 cg06263495 11 2,292,004 Y ASCL2 2 (2.2) 7 (50) 5.10 × 10 −6 cg17575811 11 2,466,409 Y KCNQ1 1 (1.1) 7 (50) 1.21 × 10 −6 cg12374721 17 46,799,640 Y PRAC 2 (2.2) 7 (50) 5.10 × 10 −7 cg21790626 19 58,220,494 Y ZNF154 0 (0) 7 (50) 1.62 × 10 −6 cg01322134 1 228,194,448 Y WNT3A 0 (0) 6 (42.9) 1.98 × 10 −6 cg01009664 3 129,693,613 Y TRH 0 (0) 6 (42.9) 1.98 × 10 −6 cg12998491 9 134,152,531 Y FAM78A 0 (0) 6 (42.9) 1.98 × 10 −6 cg19246110 19 58,238,928 Y ZNF671 0 (0) 6 (42.9) 1.98 × 10 Probe ID for the Infinium HumanMethylation27 Bead Array. Chromosome. National Center for Biotechnology Information (NCBI) Database (Genome Build 37). Y means CpG island. The probes satisfied the following criteria: (i) the average β value for all samples of non-cancerous renal cortex tissue (N) was <0.2, e e (ii) >0.4Δβ was observed in six or more clear cell RCCs (≥42.9%) in Cluster B, whereas >0.4Δβ in two or fewer clear cell RCCs (≤2.2%) in Cluster A and T–N T–N −6 f (iii) the incidence of >0.4Δβ was markedly higher in Cluster B than in Cluster A (P < 1.98 × 10 , Fisher’s exact test ). T–N Table IIIB. CpG sites as hallmarks of the CpG island methylator phenotype of clear cell RRCs a b c d e Target ID Chr Position CpG island Gene symbol P Δβ (mean ± SD) T–N Cluster A (n = 90) Cluster B (n = 14) f −7 cg17162024 8 53,478,454 Y FAM150A 0.126 ± 0.120 0.499 ± 0.184 3.40 × 10 −7 cg22040627 17 6,617,030 Y SLC13A5 0.045 ± 0.072 0.283 ± 0.103 2.64 × 10 f −7 cg14859460 5 178,422,244 Y GRM6 0.077 ± 0.105 0.434 ± 0.184 1.10 × 10 −7 cg09260089 10 134,599,860 Y NKX6-2 0.078 ± 0.083 0.372 ± 0.150 2.26 × 10 Probe ID of the Infinium HumanMethylation27 Bead Array. Chromosome. NCBI Database (Genome Build 37). Y means CpG island. Top four probes capable of discriminating Cluster B from Cluster A identified by random forest analysis (Supplementary Figure S5, available at Carcinogenesis Online) using the 869 probes on which the DNA methylation levels (Δβ ) were differed significantly between Clusters A and B (Wilcoxon rank sum test). T–N The FAM150A and GRM6 genes were shared by Tables IIIA and IIIB. methylation alterations occurring at the precancerous stage determine CpG islands. Since Cluster B was significantly associated with both both the aggressiveness of RCCs and the outcome of affected patients frequent DNA hypermethylation on CpG islands and distinct clinico- through alterations of gene expression levels. pathological phenotypes of clear cell RCCs, RCCs belonging to Clus- Unsupervised hierarchical clustering based on our previous study ter B can be recognized as CIMP-positive clear cell RCCs on the basis using bacterial artificial chromosome arrays also clustered clear cell of the definition of well-studied CIMP-positive cancers (30,31) such RCCs into clinicopathologically valid subclusters: 14% of examined as colorectal cancer (32) and stomach cancer (33), although Morris RCCs belonged to a subcluster showing clinicopathological param- et al. (28) previously considered that the relevance of the CIMP-pos- eters reflecting tumor aggressiveness and poorer patient outcome itive phenotype to RCCs had not yet been clearly defined. Although (19). DNA methylation profiles in N samples based on BAMCA data McRonald et al. (29) suggested that a subset of RCCs might display were also inherited by the corresponding clear cell RCC developing CIMP based on findings indicating that the distribution of the number in the same patient. In this study, 14% of the clear cell RCCs sub- of methylated CpGs in individual tumors differed from the expected jected to Infinium analysis belonged to Cluster B. BAMCA is suitable Poisson distribution, they did not identify distinct CpG sites that could for detecting DNA methylation alterations occurring in a coordinated become hallmarks for CIMP in the kidney. It has been suggested that, manner on individual large regions of chromosomes (34–37), whereas in order to identify CIMP-positive cancers in specific organs, marker 27 000 Infinium array is suitable for detecting DNA methylation alter - CpG sites or genes that are specific to each organ or histological type ations on promoter regions of specific genes. Different methodologies of tumor should be used (38), rather than classical CIMP marker identified similar clinicopathologically valid subclusters of RCCs, genes (30,31) that were originally identified in colorectal cancers. indicating that such clustering based on DNA methylation profiles is The present single-CpG-resolution analysis identified such hallmark not accidental but reproducible, and may reflect the distinct epigenetic CpG sites for the first time. Using the 18 CpG sites in Tables IIIA and pathway of renal carcinogenesis. IIIB, CIMP-positive RCCs or RCCs equivalent to those in the present In contrast to Cluster A, which appeared to be characterized by Cluster B could be reproducibly identified. These 18 CpG sites may accumulation of DNA hypomethylation (Figure 2A), Cluster B was be useful for further clarifying the molecular basis of the epigenetic clearly characterized by accumulations of DNA hypermethylation on pathway of renal carcinogenesis. 1492 Single-CpG-resolution methylome analysis of RCCs 14. Arai,E. et al. (2010) DNA methylation profiles in precancerous tissue and DNA methylation alterations are known to result in chromosomal cancers: carcinogenetic risk estimation and prognostication based on DNA instability through chromatin configuration changes (39). In fact, ger - methylation status. Epigenomics, 2, 467–481. mline mutations of the de novo DNA methyltransferase DNMT3B 15. Kanai,Y. (2008) Alterations of DNA methylation and clinicopathological gene have been reported in patients with immunodeficiency, centro- diversity of human cancers. Pathol. Int., 58, 544–558. meric instability and facial anomalies (ICF) syndrome, a rare recessive 16. Kanai,Y. et al. (2007) Alterations of DNA methylation associated with abnor- autosomal disorder characterized by DNA hypomethylation of peri- malities of DNA methyltransferases in human cancers during transition from centromeric satellite regions (40). In HCCs and urothelial carcinomas, a precancerous to a malignant state. Carcinogenesis, 28, 2434–2442. DNA hypomethylation of these regions is correlated with copy number 17. Arai,E. et al. (2008) Genetic clustering of clear cell renal cell carcinoma based alterations on chromosomes 1 (41) and 9 (42), respectively, where satel- on array-comparative genomic hybridization: its association with DNA meth- ylation alteration and patient outcome. Clin. Cancer Res., 14, 5531–5539. lite regions are plentiful. Correlations between the clustering based on 18. Arai,E. et  al. (2006) Regional DNA hypermethylation and DNA methyl- Infinium assay and copy number alterations should be further examined. transferase (DNMT) 1 protein overexpression in both renal tumors and cor- Taken together, the data suggest that in CIMP-positive clear cell RCCs responding nontumorous renal tissues. Int. J. Cancer, 119, 288–296. belonging to Cluster B, DNA hypermethylation of distinct CpG islands 19. Arai,E. et al. (2009) Genome-wide DNA methylation profiles in both pre- participates even in the very early and precancerous stages. Such DNA cancerous conditions and clear cell renal cell carcinomas are correlated with methylation alterations occurring in the precancerous stages may induce malignant potential and patient outcome. Carcinogenesis, 30, 214–221. more aggressive tumor phenotypes and poorer patient outcome in Clus- 20. Arai,E. et  al. (2011) Genome-wide DNA methylation profiles in renal ter B. On the other hand, in the other pathway of renal carcinogenesis tumors of various histological subtypes and non-tumorous renal tissues. leading to clear cell RCCs in Cluster A, DNA hypomethylation may be a Pathobiology, 78, 1–9. 21. Bibikova,M. et al. (2009) Genome-wide DNA methylation profiling using later event (Table I) than DNA hypermethylation on CpG islands. We are Infinium® assay. Epigenomics, 1, 177–200. now performing exome, transcriptome and proteome analyses of RCCs 22. Eble,J.N. et  al. (2004) Renal cell carcinoma. World Health Organization belonging to both clusters. Such multilayer omics analyses may identify Classification of Tumours. Pathology and Genetics. Tumours of the Uri- the upstream genetic events inducing DNA methylation profiles and key nary System and Male Genital Organs. IARC Press, Lyon, pp. 10–43. signal pathways that characterize Clusters A and B. 23. Fuhrman,S.A. et al. (1982) Prognostic significance of morphologic param- eters in renal cell carcinoma. Am. J. Surg. Pathol., 6, 655–663. 24. Sobin,L.H. et al. (2002) International Union Against Cancer (UICC). TNM Supplementary material Classification of Malignant Tumors. 6th edn. Wiley-Liss, New York, pp. 193–195. Supplementary Figures S1–S5 and Tables S1–S5 can be found at 25. Kanai,T. et al. (1987) Pathology of small hepatocellular carcinoma. A pro- http://carcin.oxfordjournals.org/. posal for a new gross classification. Cancer, 60, 810–819. 26. Sambrook,J. et al. (2001) Molecular Cloning: A Laboratory Manual. 3rd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pp. 6.14–6.15. Funding 27. Breiman,L. (2001) Random forests. Mach. Learn., 45, 5–32. 28. Morris,M.R. et  al. (2010) Epigenetics of renal cell carcinoma: the path Program for Promotion of Fundamental Studies in Health Sciences towards new diagnostics and therapeutics. Genome Med., 2, 59. of the National Institute of Biomedical Innovation (NiBio); Grant-in- 29. McRonald,F.E. et al. (2009) CpG methylation profiling in VHL related and Aid for the Third Term Comprehensive 10-Year Strategy for Cancer VHL unrelated renal cell carcinoma. Mol. Cancer, 8, 31. Control from the Ministry of Health, Labor and Welfare of Japan; 30. Issa,J.P. (2004) CpG island methylator phenotype in cancer. Nat. Rev. Can- National Cancer Center Research and Development Fund; Grants-in- cer, 4, 988–993. Aid for Scientific Research (B) and for Young Scientists (B) from the 31. Toyota,M. et al. (1999) CpG island methylator phenotype in colorectal can- Japan Society for the Promotion of Science (JSPS). cer. Proc. Natl. Acad. Sci. U.S.A., 96, 8681–8686. 32. Shen,L., et al. (2007) Integrated genetic and epigenetic analysis identifies three different subclasses of colon cancer. Proc. Natl. Acad. Sci. U.S.A., Acknowledgements 104, 18654–18659. 33. Toyota,M. et al. (1999) Aberrant methylation in gastric cancer associated Conflict of Interest Statement: None declared. with the CpG island methylator phenotype. Cancer Res., 59, 5438–5442. 34. Nagashio,R. et al. (2011) Carcinogenetic risk estimation based on quantifi- cation of DNA methylation levels in liver tissue at the precancerous stage. References Int. J. Cancer, 129, 1170–1179. 1. Ljungberg,B. et al. (2011) The epidemiology of renal cell carcinoma. Eur. 35. Gotoh,M. et al. (2011) Diagnosis and prognostication of ductal adenocar- Urol., 60, 615–621. cinomas of the pancreas based on genome-wide DNA methylation profil- 2. Arai,E. et  al. (2011) Genetic and epigenetic alterations during renal car- ing by bacterial artificial chromosome array-based methylated CpG island cinogenesis. Int. J. Clin. Exp. Pathol., 4, 58–73. amplification. J. Biomed. Biotechnol., 2011, 780–836. 3. Baldewijns,M.M. et  al. (2010) VHL and HIF signalling in renal cell car- 36. Nishiyama,N. et  al. (2010) Genome-wide DNA methylation profiles in cinogenesis. J. Pathol., 221, 125–138. urothelial carcinomas and urothelia at the precancerous stage. Cancer Sci., 4. Barrett,I.P. (2010) Cancer genome analysis informatics. Methods Mol. 101, 231–240. Biol., 628, 75–102. 37. Arai,E. et al. (2009) Genome-wide DNA methylation profiles in liver tissue 5. Zhang,J. et al. (2011) International Cancer Genome Consortium Data Portal- at the precancerous stage and in hepatocellular carcinoma. Int. J. Cancer, -a one-stop shop for cancer genomics data. Database (Oxford), 2011, bar026. 125, 2854–2862. 6. Brannon,A.R. et al. (2010) Renal cell carcinoma: where will the state-of- 38. Abe,M. et al. (2008) Identification of genes targeted by CpG island meth- the-art lead us? Curr. Oncol. Rep., 12, 193–201. ylator phenotype in neuroblastomas, and their possible integrative involve- 7. Dalgliesh,G.L. et  al. (2010) Systematic sequencing of renal carcinoma ment in poor prognosis. Oncology, 74, 50–60. reveals inactivation of histone modifying genes. Nature, 463, 360–363. 39. Eden,A. et  al. (2003) Chromosomal instability and tumors promoted by 8. van Haaften,G. et al. (2009) Somatic mutations of the histone H3K27 dem- DNA hypomethylation. Science, 300, 455. ethylase gene UTX in human cancer. Nat. Genet., 41, 521–523. 40. Hansen,R.S. et  al. (1999) The DNMT3B DNA methyltransferase gene is 9. Varela,I. et  al. (2011) Exome sequencing identifies frequent mutation of mutated in the ICF immunodeficiency syndrome. Proc. Natl. Acad. Sci. the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature, 469, U.S.A., 96, 14412–14417. 539–542. 41. Wong,N. et al. (2001) Hypomethylation of chromosome 1 heterochromatin 10. Baylin,S.B. et  al. (2011) A decade of exploring the cancer epigenome - DNA correlates with q-arm copy gain in human hepatocellular carcinoma. biological and translational implications. Nat. Rev. Cancer, 11, 726–734. Am. J. Pathol. 59, 465–471. 11. Boumber,Y. et al. (2011) Epigenetics in cancer: what’s the future? Oncol- 42. Nakagawa,T. et al. (2005) DNA hypomethylation on pericentromeric satel- ogy (Williston Park), 25, 220–226, 228. lite regions significantly correlates with loss of heterozygosity on chromo- 12. Jones,P.A. et al. (2007) The epigenomics of cancer. Cell, 128, 683–692. some 9 in urothelial carcinomas. J. Urol., 173, 243–246. 13. Kanai,Y. (2010) Genome-wide DNA methylation profiles in precancerous conditions and cancers. Cancer Sci., 101, 36–45. Received December 11, 2012; revised April 17, 2012; accepted May 11, 2012

Journal

CarcinogenesisPubmed Central

Published: May 18, 2012

There are no references for this article.