Introduction Multiple myeloma (MM) is a fatal malignancy of plasma cells and is preceded by a precursor stage, smoldering myeloma (SMM). There is a 50% risk of progression from SMM to overt MM within five years of diagnosis. Most SMM patients do not receive treatment until progression to overt MM and early intervention is beyond the current standard of care. Large sequencing studies have revealed that SMM patients with similar genetic backgrounds experience varying rates of disease progression, overall survival, and treatment response. These findings indicate an unmet need to better define the molecular requirements of disease progression. We hypothesize that altered chromatin states in pre-myeloma cells act as priming events for clonal expansion and enable the transition from pre-myeloma conditions to MM. Characterization of the pre-myeloma epigenome at single-cell resolution may offer insights into the dynamic reprogramming that occurs during myeloma development and provide new biomarkers of disease progression. Methods Bone marrow (BM) aspirates were sampled from 26 patients and healthy donors at different stages of the disease. BM samples were enriched for plasma cells using CD138 microbeads and subjected to single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq, 10x Genomics). This cohort comprised five healthy donors, seven low risk (LR), five intermediate risk (IR), and three high risk (HR) SMM patients as well as six newly diagnosed (ND) MM patients based on the 2/20/20 risk stratification model. Batch effects, data denoising, and differential accessibility testing were performed using deep generative modeling via PeakVI. CytoTRACE was used to model cell plasticity and HOMER was used to infer enriched transcription factor footprints in differential accessibility regions (DARs). g:Profiler was used for gene ontology analyses. Results We analyzed >81,000 scATAC-seq profiles. To contextualize tumor cell emergence from normal plasma cells (NPCs), we supplemented our dataset with ~1,000 cells from a public dataset of relapsed refractory MM (RRMM) patients and healthy donors (Fredeet al., Nat Cell Biol, 2021). We identified ~17,000 LRSMM vs. NDMM DARs, ~24,000 HRSMM vs. LRSMM DARs and ~5,000 HRSMM vs. IRSMM DARs. Comparing cells from SMM and MM donors with NPCs revealed few or zero significant DARs, suggesting that intrasample heterogeneity must be properly resolved to accurately describe detailed molecular annotations of each cell group. We thus performed unsupervised clustering of all NPC, SMM, and MM cells to resolve 11 distinct clusters. We defined three NDMM clusters, three NPC/LRSMM clusters, and individual RRMM, NDMM/ HRSMM, HRSMM, IRSMM/LRSMM, NDMM/LRSMM clusters. Comparison of these clusters revealed ~13,000 LRSMM vs. NPCs DARs and ~1,600 IRSMM/LRSMM vs. NPC DARs. Gene ontology analysis of genes neighboring DARs (<50 kb) displayed signatures associated with the P38-MAPK, JNK, and WNT signaling pathways as well as differentiation and development, consistent with previous observations of de-differentiation and increased epigenomic plasticity in MM. BATF, POU3F3, SMAD2, IRF4, JUNB, MYC, FOSL1, and SP1 binding sites were enriched in DARs. We ordered these cell groups along an axis of disease progression, and work to build a deep, generative time-resolved model of MM emergence from pre-myeloma is ongoing. Conclusions We report the largest single-cell chromatin profiling experiment on primary MM cells, focusing on phenotypically early traits of SMM. We characterize distinct sub-groups of SMM patient BM cells and their corresponding regions of accessible chromatin. We are expanding this study through sample-matched WGS to delineate independent genetic and epigenomic contributions to MM progression. scATAC-seq analyses of the matched CD138-negative immune cell fractions are underway alongside analyses of single-cell transcriptome data for both CD138 positive and negative fractions. We anticipate that a deeper understanding of the chromatin states associated with the establishment of overt myeloma will lead to novel therapeutic targets and offer improved predictive prognostics for MM.
Read full abstract