Multiple Myeloma Evolution Is Characterized By Dynamic Epigenetic Landscapes

The biological mechanisms driving MM progression and multi-drug resistance remain unclear. To better characterize the molecular mechanisms driving malignant transformation and disease progression, we generated a new database comprising 1,016 bone marrow biopsies from patients across the whole MM spectrum. These samples were clinically, genetically (844 RNA-Seq, 870 WES, 7 scRNA-Seq), and epigenetically (11 scATAC-Seq, 8 CUT&Tag) characterized. While we have observed increased mutation frequency in MM driver genes in later disease stages, no specific mutations statistically correlated with malignant transformation, nor with the development of refractory disease. Enrichment analysis of transcriptional data identified biological pathways and underlying molecular events associated with MM progression and therapy resistance: cell adhesion, inflammatory cytokines, TME-dependence and hematopoietic cell identity were under-expressed in NDMM vs. premalignant disease; whereas cell cycle, energy metabolism, DNA repair, and protein/RNA synthesis/degradation were overexpressed in LRMM compared to NDMM. Importantly, this transcriptional dysregulation was anecdotally confirmed in patients with multiple sequential samples. To identify the main biological pathways responsible for the inter-sample transcriptomic variability of our cohort, we performed a principal component analysis (PCA) using single sample gene set enrichment analysis (ssGSEA) NES of Cancer Hallmarks. The first component (PC1) was associated with aforementioned hallmarks differentially expressed in NDMM and LRMM samples. This suggested PC1 as a possible prognostic biomarker, which was confirmed by Cox Proportional Hazard models, associated with shorter TTP in SMOL, PFS in NDMM, and OS in active MM. To elucidate the mechanisms driving such aberrant transcriptional programs, we applied a robust dimensionality reduction technique (t-SNE) and a clustering method (fuzzy c-means) to RNA-Seq data to identify clusters of co-expressing genes, thus characterizing MM transcriptomic topology. In total, 16,738 genes were grouped in 500 gene clusters putatively controlled by shared transcriptional programs. Enrichment analysis showed that gene clusters under-expressed during malignant transformation were enriched for H3K27me3, whereas gene clusters overexpressed in refractory disease were enriched for H3K27ac. We validated these findings using scATAC-Seq in 11 MM samples (1 MGUS, 3 SMOL, 4 NDMM, 1 ERMM and 2 LRMM), which confirmed that a greater decrease in chromatin accessibility occurred among the most under-expressed gene clusters during malignant transformation. scATAC-Seq data also revealed the differential accessibility of B cell master regulators' binding motifs across disease progression (e.g., SPI1 in MGUS, FOS in SMOL, the IRF and POU family in NDMM, RUNX2 and RFX5 in LRMM) and suggested MM pioneer TFs by correlating the accessibility of TF's genes to their targets' expression (e.g., SPI1, BACH2, RELA, IRF4, among others). CUT&Tag data from 4 NDMM and 4 LRMM samples confirmed that the strongest H3K27ac signal was found in the topological region overexpressed in refractory disease, with a positive correlation between H3K27ac signal and gene expression. Super enhancer (SE) analysis demonstrated that, while genes associated with active SEs showed higher level of expression than genes subjected to regular enhancers (RE) in all samples, LRMM presented increased expression of genes under both SE/RE control compared to their counterparts in NDMM. These observations, in conjunction with an increased expression of transcription regulation-associated DNA-binding proteins (DBPs; e.g., CDK9, CTCF, MED1, YY1), suggests that the aberrant gene expression in LRMM may be due to differential expression of these DBPs.

In conclusion, MM progression, characterized by transcriptional dysregulation in response to TME and therapeutic stress, is epigenetically driven by DBPs whose activity is altered by genetic, cytogenetic, and epigenetic events. Future studies will determine the role of such masterregulators in MM biology and drug resistance, as well as allow us to propose patient-specific strategies for therapy re-sensitization.

Kulkarni:M2GEN: Current Employment. Zhang:M2GEN: Current Employment. Hampton:M2GEN: Current Employment. Baz:Janssen: Membership on an entity's Board of Directors or advisory committees, Research Funding; GSK: Honoraria, Membership on an entity's Board of Directors or advisory committees; Pfizer: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees; Shattuck labs: Membership on an entity's Board of Directors or advisory committees; genentech: Membership on an entity's Board of Directors or advisory committees; BMS: Membership on an entity's Board of Directors or advisory committees, Research Funding; karyopharm: Research Funding; celgene: Consultancy, Honoraria; Sanofi: Consultancy, Honoraria; Merck: Research Funding. Shain:Bristol Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Amgen: Speakers Bureau; Takeda: Honoraria, Speakers Bureau; Sanofi: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Janssen: Honoraria, Speakers Bureau; Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Adaptive: Honoraria; AbbVie: Research Funding; GlaxoSmithKline: Speakers Bureau; Karyopharm: Research Funding, Speakers Bureau.