On the cellular origins of multiple myeloma

Abstract

Abstract Introduction A long-standing question in multiple myeloma (MM) is the nature of its stem cell origin and whether transformation involves compartments prior to terminal plasma cell (PC) differentiation. Of particular interest is the potential involvement of pre-germinal center (GC) naïve B cells or even earlier stages such as hematopoietic stem cells (HSCs) or B-cell precursors. Advances in sequencing and lineage-resolved cell sorting allow re-evaluation of this question, with particular relevance to the origins of relapse following deep MRD-negative responses to therapies targeting mature PCs. The extent and pattern of somatic hypermutation (SHM) in immunoglobulin (Ig) genes can provide insight into timing and nature of GC exposure during clonal evolution. We analyzed Ig gene rearrangements from whole-genome sequencing (WGS) with compartment-specific sorting to define hematopoietic relationships and pinpoint the cellular origin of MM transformation. Methods We performed WGS on 353 matched tumor-normal pairs from 294 MGUS, SMM, NDMM, and RR MM patients (median coverage: 64× tumor, 34× normal). Ig loci were profiled using IgCaller, which utilizes split-read and discordant-pair evidence to reconstruct V(D)J junctions, detect abnormal class-switch recombination (CSR) events, and call chromosomal translocations. Variant calling was performed via the harmonized Myeloma Genome Pipeline 1000. Our workflow, shmCaller (https://github.com/pblaney/shmCaller), was applied to sorted compartments to quantify and track SHM fingerprinting across lineages. Thirteen patients underwent flow sorting into hematopoietic and B-cell subsets: HSCs, multipotent progenitors (MPPs), lymphoid-primed MPPs, pre/pro-B cells, naïve B cells, memory B cells, normal plasma cells, and malignant plasma cells (MPCs), with an immunophenotyping panel (CD34, CD38, CD10, CD19, CD45RA, CD90, CD27, CD56, CD117 and CD138) and each fraction underwent 100x whole-exome sequencing (WES) with a uniform pipeline. Results Productive V(D)J rearrangements were detected in 82% of cases: IGH+IGK (35%), IGH+IGL (15%), IGK only (18%), IGL only (8%), and IGH only (6%). V, D, and J gene usage (e.g., IGHV3-30, IGKV1-5, IGLV1-40) showed no major differences across MGUS, SMM, and MM. Canonical translocations were found in 43% of cases: t(11;14) (21%), t(4;14) (12%), t(8;14) and t(6;14) (3% each), t(14;16) and t(14;20) (2% each), with breakpoints mapped to IG constant and VDJ regions. Novel non-canonical rearrangements were detected: IGH to 11q13.4 (13%), 19p13.3 (PIP5K1C, 9%), 19p13.11 (2%), and IGK to 20q11.21 upstream of BCL2A1 (14%). Intraclonal SHM analysis of IGHV sequences revealed variable mutation burdens across MGUS, SMM, and MM. Productively rearranged IGHV cases exhibited diverse SHM patterns. Several cases showed additional mutations within dominant Ig clones, consistent with ongoing or reinitiated SHM, potentially driven by GC reentry or persistent chromatin accessibility. We also performed low-input WES across all flow sorted compartments. Preliminary analysis revealed cases with mutational burden restricted to MPCs and others with shared mutations in bone marrow-confined pre-GC populations. One SMM case harbored a pathogenic TP53 L137Q mutation across HSCs, MPPs, and MPCs, consistent with possible early acquisition in a precursor selectively contributing to PC differentiation. These findings suggest that initiation of transformation events may arise at multiple stages of hematopoietic development, informing models of clonal evolution, disease persistence, and potential relevance of clonal hematopoiesis of indeterminate potential to MM pathogenesis. Additional high-depth sorting and sequencing are underway and will be integrated into the final analysis to refine the temporal and lineage context of MM transformation. Conclusion Our integrated analysis of Ig rearrangements, SHM patterns, and lineage-resolved sequencing supports a model in which multiple myeloma originates from clonal events across hematopoietic development, from early progenitors to differentiated B-cell compartments. While malignant transformation likely occurs later within the plasma cell lineage, early driver mutations in stem-like cells and ongoing SHM activity underscore the complexity of clonal evolution. These findings highlight the need to define lineage-specific vulnerabilities and temporal dynamics throughout the myeloma disease course.