In addition to the increasingly well understood repertoire of mutations in coding sequences of oncogenes and
tumor suppressor genes, genome-wide sequencing of human cancers has revealed abundant mutations in
non-coding sequences. Some of these non-coding mutations are found in regulatory sequence elements
(promoters, enhancers, insulators) that determine gene expression. However, even when found in such
sequences, most of these mutations are expected to be neutral, i.e., passengers, with only a small percentage
having a functional impact. Here we seek to address the challenge of identifying oncogenic non-coding
mutations, focusing on a common and difficult to treat human hematologic malignancy, multiple myeloma
(MM). Our approach entails combined genetic-epigenetic mapping, focusing on allele-specific DNA methylation
(ASM), followed by functional assays. We will perform whole-genome sequencing (WGS) and whole-genome
bisulfite sequencing (WGBS; methyl-seq) in 30 MM cases, representing two of the most common molecular
subtypes of this cancer, each paired with non-neoplastic peripheral blood B cells from the same patients. This
procedure will identify differentially methylated regions with ASM (ASM DMRs) that have arisen de novo in the
tumor cells due to somatic mutations that destroy or create transcription factor binding sites (TFBS) in
promoter/enhancer elements or CTCF sites in insulator elements. These findings will nominate candidate
functional non-coding mutations, which have declared their functional activity by conferring the local physical
asymmetry, namely ASM, between mutant and wild-type alleles. We will vet these mutations for functional
effects by creating them in MM cell lines using CRISPR-Cas9 mutagenesis, followed by epigenetic and
biological assays on the isogenic pairs of wild-type and mutant cell lines. Using our mapping approach in a
small pilot series of paired samples from MM patients, we have already identified candidate regulatory
mutations, including a point mutation in a putative enhancer element of the TEAD1 gene, which we have
functionally validated as producing loss of methylation of its flanking CpGs. Lastly, we will test for recurrence of
mutations in the same regulatory sequences in several hundred MM cases in our tumor banks, apply
bioinformatic approaches to determine whether the de novo ASM-associated mutations, even if not highly
recurrent, preferentially participate in biological pathways that drive MM progression, and study sequential
clinical samples to determine whether additional non-coding regulatory mutations in these same pathways
arise during MM disease progression. Our overall goals are to develop and validate a highly practical and
generalizable approach for identifying functional non-coding mutations in human cancers, and to use the
specific data from this project to identify new targets for treatment of MM.