Epigenetic changes play important roles in carcinogenesis and influence initial steps in neoplastic transformation by altering genome stability and regulating gene expression. To characterize epigenomic changes during the transformation of normal plasma cells to myeloma, we modified the HELP assay to work with small numbers of purified primary marrow plasma cells. The nano-HELP assay was used to analyze the methylome of CD138+ cells from 56 subjects representing premalignant (MGUS), early and advanced stages of myeloma as well as healthy controls. Plasma cells from premalignant and early stages of myeloma were characterized by striking, widespread hypomethylation. Gene specific hypermethylation was seen to occur in the advanced stages and cell lines representative of relapsed cases were found to be sensitive to decitabine. Aberrant demethylation in MGUS occurred primarily in CpG islands while differentially methylated loci in cases of myeloma occured predominantly outside of CpG islands and affected distinct sets of gene pathways, demonstrating qualitative epigenetic differences between premalignant and malignant stages. Examination of the methylation machinery revealed that the methyltransferase, DNMT3A, was aberrantly hypermethylated and underexpressed, but not mutated in myeloma. DNMT3A underexpression was also associated with adverse overall survival in a large cohort of patients, providing insights into genesis of hypomethylation in myeloma. These results demonstrate widespread, stage specific epigenetic changes during myelomagenesis and suggest that early demethylation can be a potential contributor to genome instability seen in myeloma. We also identify DNMT3A expression as a novel prognostic biomarker and suggest that relapsed cases can be therapeutically targeted by hypomethylating agents.

Purpose Even though recent studies have shown that genetic changes at enhancers can influence carcinogenesis, most methylomic studies have focused on changes at promoters. We used renal cell carcinoma (RCC), an incurable malignancy associated with mutations in epigenetic regulators, as a model to study genome wide patterns of DNA methylation at a high resolution. Experimental Design and Results Analysis of 1.3 million CpGs by the HELP assay in RCC and healthy microdissected renal tubular controls demonstrated that the RCC samples were characterized by widespread hypermethylation that preferentially affected gene bodies. Aberrant methylation was particularly enriched in kidney specific enhancer regions associated with H3K4Me1 marks. Various important underexpressed genes such as SMAD6 were associated with aberrantly methylated, intronic enhancers and these changes were validated in an independent cohort. MOTIF analysis of aberrantly hypermethylated regions revealed enrichment for binding sites of AP2alpha, AHR, HAIRY, ARNT and HIF-1 transcription factors, reflecting contributions of dysregulated hypoxia signaling pathways in RCC. The functional importance of this aberrant hypermethylation was demonstrated by selective sensitivity of RCC cells to low levels of decitabine. Most importantly, methylation of enhancers was predictive of adverse prognosis in 405 cases of RCC in multivariate analysis. Additionally, parallel copy number analysis from MspI representations demonstrated novel cnvs that were validated in independent cohort of patients. Conclusions Our study is the first high resolution methylome analysis of RCC; demonstrates that many kidney specific enhancers are targeted by aberrant hypermethylation and reveals the prognostic importance of these epigenetic changes in an independent cohort.

Alterations in DNA methylation are seen in cancers and have also been examined in clear cell renal cell carcinoma (ccRCC). Numerous tumor suppressor genes have been reported to be partially or completely silenced due to hypermethylation of their promoters in single-locus studies, and the use of hypomethylating agents has been shown to restore the expression of many of these genes in vitro. In particular, members of the Wnt and TGF-beta pathways, pro-apoptotic genes such as APAF-1 and negative cell-cycle regulators such as KILLIN have been shown to be epigenetically silenced in numerous studies in ccRCC. Recently, TCGA analysis of a large cohort of ccRCC samples demonstrated that aberrant hypermethylation correlated with the stage and grade in kidney cancer. Our genome-wide studies also revealed aberrant widespread hypermethylation that affected regulatory regions of the kidney genome in ccRCC. We also observed that aberrant enhancer hypermethylation was predictive of adverse prognosis in ccRCC. Recent discovery of mutations affecting epigenetic regulators reinforces the importance of these changes in the pathophysiology of ccRCC and points to the potential of epigenetic modulators in the treatment of this malignancy.

Edited by Alex Toker

784 Gene expression is a tightly regulated process and is influenced by aberrant epigenetic changes that can lead to carcinogenesis. We used the HELP (HpaII tiny fragment Enrichment by Ligation-mediated PCR) assay to perform an unbiased genome-wide analysis of DNA methylation in 11 MGUS, 16 newly diagnosed myeloma, 17 relapsed myeloma, and 8 healthy control samples. The HELP assay uses differential methylation specific restriction digestion by HpaII and MspI followed by amplification, 2 color labeling, and co-hybridization to quantitatively determine individual promoter methylation of 25,626 loci. The methylome analysis was performed using CD 138+ sorted bone marrow plasma cells in all cases. We observed extensive DNA methylation changes in myeloma that were seen even in MGUS cases when compared to normal plasma cells. Unsupervised clustering of all samples showed that MGUS samples were distinct but epigenetically closer to normal plasma cells than to cases of newly diagnosed or relapsed Myeloma. MGUS cases were characterized predominantly by aberrant hypomethylation, with 456 hypomethylated probes versus 130 hypermethylated probes (cutoff criteria were defined as a difference of means > 1.5 and significance of this difference with p < 0.001, Figure 1) that affected pathways regulated by NF-kb transcription factor. Untreated newly diagnosed myeloma samples were also predominantly hypomethylated (461 hypomethylated probes, 83 hypermethylated probes) when compared to controls. In addition to NF-kB, the MAP kinase and PI3 kinase regulated pathways were affected by hypomethylated genes. In contrast, cases of relapsed myeloma showed predominantly hypermethylated loci (221 hypomethylated, 560 hypermethylated probes) when compared to controls and involved the TNF and retinoblastoma pathways. A large number of important genes including the tumor suppressors CDKN2A and CDKN2B were aberrantly hypermethylated in this cohort. Figure 1: Volcano plot of differentially methylated genes, plotting difference of means against the −log(10) of the respective p-value. The cutoff for the definition for significant differentially methylated probes was set at a difference of means >1.5 or < −1.5 and a p-value < 0.001. Hypomethylated probes are depicted in green, hypermethylated probes in red. The respective numbers are documented in the same color. A) comparison of normal plasma cells with MGUS. B) comparison of normal plasma cells with newly diagnosed MM. C) comparison of normal plasma cell with relapsed myeloma Figure 1:. Volcano plot of differentially methylated genes, plotting difference of means against the −log(10) of the respective p-value. The cutoff for the definition for significant differentially methylated probes was set at a difference of means >1.5 or < −1.5 and a p-value < 0.001. Hypomethylated probes are depicted in green, hypermethylated probes in red. The respective numbers are documented in the same color. A) comparison of normal plasma cells with MGUS. B) comparison of normal plasma cells with newly diagnosed MM. C) comparison of normal plasma cell with relapsed myeloma Analysis of our differentially methylated targets using the Molecular Signatures Database (MSigDB, Tamayo, et al. 2005, PNAS 102, 15545–15550), showed significant overlap for hypomethylated genes in MGUS and new multiple myeloma (MM). These gene sets contain genes with promoter regions around transcription start site containing motifs which match annotation for transcription factors SP1 and TCF3 as well as enrichment for genes with gene ontology annotations for plasma membrane proteins. Hypermethylated genes in new MM show significant overlap with genes in the neighborhood of RAD23A, CTBP1, G22P1 and SMC1, suggesting impairment of DNA repair, as well as enrichment for genes associated with apoptosis and programmed cell death. In conclusion, genome-wide DNA methylation analysis is able to clearly differentiate between normal bone marrow plasma cells, MGUS as well as new MM and relapsed MM cells. Correlation of significantly differentially methylated genes with published gene sets reveals enrichment for genes involved in DNA repair, cell-cell signaling, cell death, apoptosis and cell cycle regulation. Disclosures: Mehta: Celgene: Consultancy, Speakers Bureau; Takeda/Millennium: Speakers Bureau; Onyx: Research Funding. Singhal:Takeda/Millennium: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Onyx: Research Funding; Celgene: Speakers Bureau.