P-236: Integrated coding and non-coding transcriptomic atlas reveals how malignant plasma cells shape the bone marrow microenvironment in multiple myeloma

Introduction: Multiple myeloma (MM) is a highly heterogeneous cancer, posing challenges for prognosis and treatment. The bone marrow microenvironment (BMME) has emerged as a significant factor driving MM progression and response to therapy, with dysregulated immune populations significantly shaping the BMME. We propose that these dysregulated populations correlate with specific malignant plasma cell phenotypes, collectively influencing patient outcomes. Given the emerging role of noncoding RNAs (ncRNAs) in regulating cellular phenotypes, we constructed the first combined coding and non-coding transcriptomic atlas of the MM-BMME to elucidate how distinct malignant cell phenotypes alter the immune landscape.

Methods: We performed single-cell RNA-seq on 481 bone marrow samples from MM patients enrolled in the MMRF CoMMpass study. The sequencing data was aligned to the GRCh38 genome expanded with noncoding transcripts from LncBook2.0 and GENECODEv34, followed by QC, batch correction, clustering, and supervised analysis.

Results: Approximately 1.9 million cells were captured from immune, plasma, and stromal compartments, with ncRNAs comprising 46% of represented genes: 22,102 long ncRNAs (lncRNAs) and 3,112 small ncRNAs. Major immune compartments (B lymphoid, myeloid, and NK and T lymphoid) were further divided into 70 subclusters, with ncRNAs representing 24-38 of the top 100 and 10-24% of the total differentially expressed genes within each compartment, highlighting the ability of ncRNAs to differentiate immune subtypes. Thirty of these subclusters were differentially abundant (P < 0.05) between patients with different specific cytogenetic abnormalities, suggesting that malignant plasma cell genotypes shape the immune microenvironment. Most notably, among patients with 1q21 gain and 17p53 deletion, an increase in TGFβ-stimulated monocytes and a decrease in cytotoxic CD8+ T and CD56dim NK cells, respectively, were observed.
Examining the malignant plasma cell profiles, we identified 14 ncRNAs differentially expressed across cytogenetic abnormalities, of which 6 lncRNAs were associated with poor overall survival. Combined high expression of these lncRNAs in plasma cells was linked to poor survival (P < 0.05, HR = 3.7). Stratifying patients based on enrichment of these outcome-associated lncRNAs revealed changes in immune composition, including an increase in CD4+ Tregs (P < 0.05), further supporting our hypothesis that malignant cell phenotypes shape the BMME. Lastly, by constructing regulatory networks, we identify how key ncRNAs putatively modulate the phenotypes of plasma and immune cells.


Conclusions: Integrating coding and noncoding RNAs, we present the first high-resolution transcriptomic atlas of the MM-BMME, revealing distinct immune subpopulations associated with specific malignant plasma cell phenotypes. Collectively, this work highlights a central role for ncRNAs in shaping the MM-BMME, highlighting their potential as therapeutic targets and prognostic markers for MM.