Abstract Multiple myeloma (MM) and its aggressive subtype plasma cell leukemia (PCL) are transcriptionally-driven hematological malignancies, heavily dependent on enhancer activity. Enhancers are distal regulatory elements at which transcription factors (TFs) bind, recruiting coactivators that promote chromatin remodeling to stimulate target gene expression via physical proximity of enhancer and promoter. MM cells show high expression of essential TFs, including IRF4, MYC, PRDM1 and IKZF1/3, which function by binding to and activating oncogenic enhancers. A detailed understanding of chromatin regulation in MM patients requires the comparison of histone modifications, chromatin accessibility and TF/cofactor binding distribution within the same cells, which has so far been unachievable. Patient analysis has mainly relied on ATAC-seq and RNA-seq, with limited inference of TF binding and chromatin state, owing to the difficulty in obtaining the large numbers of MM cells required for more detailed characterization. Many observations of enhancer function in MM therefore come from cell lines, which only offer an approximation of MM physiology. To overcome the limitations of in vitro models and primary cell availability, we employed TOPmentation (TF-OPtimized ChIPmentation) to generate ChIP-seq-quality data for CD138+ cells from MM and PCL patients. This technique uses as few as 100,000 cells for histone modifications, and 250,000 cells for TFs, allowing us to profile up to 15 features per patient. In addition, we used the base-pair resolution chromosome conformation capture technique micro-capture-C (MCC) to visualize enhancer-promoter interactions in a PCL patient sample. This allowed us to dissect large oncogenic enhancers, identifying the key TF binding sites within each enhancer that directly contact the gene promoter. Using this dual approach, for the first time we have generated integrated maps of TF occupancy, active and repressive histone modifications, and chromatin interactions in MM patient samples. We characterized the epigenetic landscapes of two t(11;14) and two t(4;14) patients, finding broadly similar chromatin profiles in each subtype. Key MM enhancers were retained at oncogenes including IRF4, MYC,PRDM1 and IKZF1/3, indicating a convergence of gene regulation in MM originating from distinct initiating genetic events. At many enhancers, we observed multiple TFs co-binding at the same sites, indicating cooperative activation of target genes. By combining TOPmentation with MCC, we found that many of these sites interact with target gene promoters, directly implicating them in gene regulation. In MM, patients typically display mutually exclusive upregulation of CCND1 or CCND2. As previously established, we found strong activation at the CCND1 locus by IGH translocation in t(11;14) patient samples, whereas in t(4;14) patients CCND2 was upregulated via an upstream super-enhancer. Surprisingly, in each subtype the silent CCND gene was not actively repressed, but rather existed in a poised, bivalent state marked by active H3K4me3 and repressive H3K27me3 modifications. This argues that intricate regulation of CCND1/2 is required to maintain optimal levels of expression for tumor growth. Comparison of patient and cell line data showed broad conservation of epigenetic features. However, we also observed examples of distinct gene regulation, for example at the PRDM1 locus. In several MM cell lines (including KMS12, H929 and MM1S), transcription of PRDM1 initiates from a distal promoter. However, this locus was inactive in the patient samples analyzed, and instead an alternate, proximal promoter was favored. This study for the first time reports the genomic binding distribution of TFs and histone modifications in MM patient cells, demonstrating the importance of comprehensive epigenomic analysis to capture the complexity in MM gene regulation. This emphasizes the need for validation of cell line models to provide a more physiologically relevant understanding of enhancer-driven oncogene regulation. As we expand this patient-derived dataset, it will provide a valuable resource to better understand MM biology and inform future strategies for precision epigenetic therapy.
Alexia Katsarou (Mon,) studied this question.