Gale, Molly

Abstract Abstract 3527 Increasing evidence compiled from our lab and several others is illuminating the importance of epigenetic mechanisms in the pathogenesis of multiple myeloma (MM). The biochemical modifications that govern epigenetics are DNA methylation, and post-translational modifications of histone proteins on arginine and lysine residues. Modifications are generally associated with activation or repression of gene transcription depending upon the specific position of the modification. Histone methylation, catalyzed by histone methyltransferases (HMT) and histone demethylases (HDMT) are crucial for the proper programming of the genome during development. Deregulation of the methylation machinery can alter chromatin configuration and disrupt normal transcriptional programs, both features of cancer cells. Misregulated methylation may result from several genetic alterations in chromatin-modifying enzymes, which can include mutations, over-expression or chromosomal abnormalities. Indeed, genomic alterations of MLL, NSD1, CREBBP(CBP) MMSET and UTX, all encoding proteins involved in post-translational histone modifications, have been implicated in MM. The purpose of our study was to interrogate existing genomics datasets to identify additional HMT or HDMT genes that may be important in the pathogenesis of MM and, thus could become potential therapeutic targets. To identify novel HMT or HDMT in MM we mined existing expression, copy number and whole genome sequencing data generated as part of the Multiple Myeloma Research Consortium Genomics Initiative. Collectively, the datasets converged on numerous alterations involving histone methylation on lysine 9 (H3K9). One of these genes, a HDMT called JMJD1C, was over-expressed in approximately 15% of MM samples examined. We validated over-expression of JMJD1C expression by RT-PCR in over 50 clinical MM samples and 10 human myeloma cell lines (HMCLs). We also confirmed over-expression of JMJD1C by immunohistochemistry (IHC) of a tissue microarray (TMA) consisting of over 60 MM samples. Next we examined the levels of JMJD1C expression and H3K9 methylation by IHC, RT-PCR and western blot to determine the extent to which JMJD1C expression is associated with a decrease in H3K9 methylation (or vice versa). We demonstrated this negative correlation in 6/10 HMCL including OCIMY7 and KMS11 and in over 60% of clinical samples examined on the TMA. Furthermore, we demonstrated that H3K9 methylation, a repressive mark associated with heterochromatin, is low in a large percentage of samples. In addition, we generated double knockout isogenic cell line pairs using zinc finger nuclease technology to evaluate the direct role of JMJD1C on H3K9 methylation, cell function and chromatin and expression level changes using ChIP-seq and RNA-seq analyses. In summary our preliminary results demonstrate that JMJD1C is highly expressed in a significant percentage of MM patients. Our results demonstrate a noticeable inverse relationship of JMJD1C expression to H3K9. This reflects the importance of JMJD1C in modifying the H3K9 repressive mark and suggests that loss of repression due to JMJD1C-mediated histone demethylation disrupts the chromatin machinery and leads to neoplastic processes. Further studies are ongoing to realize JMJD1C as a potential and novel therapeutic target in MM. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 1285 Introduction: Recent studies have now demonstrated that both genic and global hypomethylation characterizes the multiple myeloma (MM) epigenome. We previously conducted a study to measure changes in DNA methylation at approximately 1500 CpG (GoldenGate Methylation BeadArray, Illumina) loci in 193 MM samples used in the Multiple Myeloma Research Consortium Genomics Initiative. The purpose of this study was to determine the extent to which DNA methylation changes in MM correlate to changes at the gene expression level. Methods: To evaluate this correlation we performed three independent analyses. In the first approach we analyzed matching gene expression and DNA methylation data using a Pearson correlation to assess the linear relationship between the datasets. In the second approach methylation and gene expression data from 193 matching samples were quantile-normalized and standardized across genes and samples. Methylation data was then discretized into three groups (methylated, unchanged, unmethylated) according to the mean and standard deviation of the values for each probe on the array. We then computed a t-test to compare the gene expression differences for each gene represented on the methylation array between samples in the methylated and unmethylated groups. We also performed a class enrichment analysis to assess DNA methylation/gene expression correlations that might be associated with hyperdiploidy or molecular subtype. In the third approach we performed RT-PCR and methylation pyrosequencing to validate results in a subset of samples used for the analysis and in an independent cohort comprised of 50 MM samples. Results: By Pearson correlation, only 31 (2.1%) CpG loci (corresponding to 24 unique genes) had a gene expression probe with a statistically significant correlation. Using the discretization approach, we identified 382 loci (25.3%, 309 unique genes) with methylation-expression correlations. Although we identified more correlations by discretizing the data, methylated and unmethylated samples were represented by an average of only 30 and 27 samples per group. We identified correlations where the directionality of methylation and expression values were either in the opposite (negative) or same direction (positive). Among the genes identified, the tumor suppressor genes CDKN2A and DLC1 interestingly demonstrated a positive correlation (methylated and expressed) and the IGF1R and IL17RB genes were negatively correlated (unmethylated and expressed). A class enrichment analysis revealed that samples with DLC1, CDKN2A, IGF1R or IL17RB methylation were associated with hyperdiploid MM. Conversely, samples that were unmethylation for these genes were generally non-hyperdiploid. These findings were validated by RT-PCR and methylation pyrosequencing. Conclusion: Overall, our findings demonstrated, albeit in a limited number of genes, that DNA methylation changes are weakly associated to gene expression. These data suggest that CpG methylation may have other functional consequences such as predisposing the genome to global gene transcriptional changes or chromosomal instability. While future studies are needed to determine the exact role of DNA methylation, we identified a number of genes regulated by an epigenetic mechanism with important clinical and biological implications to MM and warrant further study. Disclosures: No relevant conflicts of interest to declare.