Single-Cell Profiling of ~1.7 Million Plasma Cells across >250 Individuals Reveals Transcriptional Signatures of Progression and Clonal Expansion in Multiple Myeloma and Its Precursor Stages

Introduction

Multiple myeloma (MM) is a plasma cell malignancy frequently preceded by precursor conditions, monoclonal gammopathy of undetermined significance (MGUS) and smoldering multiple myeloma (SMM). Patients with MGUS and SMM exhibit variable risk of progression to MM that cannot be fully explained by differences in genetic alterations. Characterizing both transcriptional and clonotype features in healthy and malignant plasma cells at a single-cell level may help to identify molecular mechanisms of disease progression and improve prognostication. To address this, we performed single-cell RNA and B cell receptor sequencing on bone marrow (BM) and peripheral blood (PB) CD138-positive samples collected from >250 patients with various disease stages and healthy donors, yielding ~1.7 million plasma cell transcriptional and clonal profiles. Here, we present a comprehensive investigation of their transcriptional programs including clonal expansion during tumorigenesis.

Methods

We performed single-cell RNA and V(D)J sequencing on 342 samples (252 BM, 90 PB) from 39 patients with MGUS, 132 patients with SMM, 51 patients with MM, and 41 healthy donors (263 individuals in total). CellRanger v6.0.1. was used to compute gene expression count matrices and identify clonotype information per single cell. Low-quality droplets (with ≥15% mitochondrial expression, or ≤400 genes covered) and potential multiplets (with ≥5000 total genes detected, or ≥50,000 total UMIs) were filtered out. Malignant plasma cells were annotated based on the matched V(D)J clonotype information. Transcriptional signatures were extracted by SignatureAnalyzer v0.0.9 at the single-cell level, and clonotype features were analyzed by Dandelion v0.3.7.

Results

Overall, we sequenced 1,681,085 plasma cells, including 1,256,207 malignant and 424,878 normal plasma cells with the clonotype information. Malignant plasma cells from the same patient formed a unique cluster, and these malignant clusters across different patients with the same IgH translocation were closer to each other. We derived multiple gene expression signatures that capture various programs related to dysregulated plasma cell biology as well as tumor stage by comparing healthy and malignant plasma cells between healthy donors and patients across various disease risk groups. Investigating the signature activities across disease stages at a single-cell level, we observed that: (i) the upregulation of CD47, which promotes immune evasion, was established in SMM compared to MGUS; (ii) immune regulation signature was upregulated in plasma cells and was predictive of survival (P-value < 0.0001; e.g., IFI6, a negative regulator of innate immune responses by modulating RIG-I activation); (iii) signatures related to bone biology, hypoxia, metabolic activity, and cytomobility were associated with tumor stage; and (iv) normal plasma cells are unexpectedly heterogeneous, with a clear distinction between circulating and bone marrow-resident normal plasma cells. We also found an MGUS-specific signature enriched in both malignant and normal cells, which could be useful for prognosis of disease progression. After integrating copy-number variations and B-cell receptor clonotypes per cell, we interestingly identified 4 SMM patients that had two distinct tumors with different cytogenetics. These cases may shed light on what transcriptional and genomic factors contributed to the risk, evolution and progression to MM.

Conclusion

We generated the largest clonotype-resolved single-cell cohort of patients with MM and its precursor stages (MGUS and SMM), and characterized novel transcriptional signatures related to the transformation of normal to malignant plasma cells and circulating and bone marrow-resident plasma cells among disease stages. Our findings in this dataset could be a useful resource to (i) score plasma cells at a single-cell level for the assessment of disease stages, and (ii) nominate novel targets or prioritize target validation for the treatment or prevention of MM.