Multi-omics reveals immune features in immune and non-immune cells, an IFN-γ/IFN-α-B2M positive feedback loop, and targeted metabolic therapy in multiple myeloma

Abstract Introduction: Multiple myeloma (MM) remains incurable despite ongoing advancements in treatment, with the majority of patients eventually relapsing. Although advances in proteasome inhibitors, immunomodulatory drugs, and hematopoietic cell transplantation have improved survival, long-term outcomes remain suboptimal. High-throughput single-cell RNA sequencing (scRNA-seq), bulk RNA-seq, and single-cell metabolic profiling enable unprecedented resolution of tumor microenvironment (TME) heterogeneity and metabolic reprogramming. Here, we integrate the largest multi-omics dataset to date , aiming to: (1) identify novel immune and non-immune cell subpopulations driving immune evasion and disease progression; (2) elucidate key signaling circuits—particularly an IFN-γ/IFN-α–B2M positive feedback loop; and (3) discover targeted metabolic therapies. Methods: We analyzed scRNA-seq data from 113 individuals (healthy controls and MM patients across disease stages; 111,640 cells) alongside bulk RNA-seq from the MMRF cohort and GEO. Cell clustering, annotation, pseudo-time trajectories, ligand-receptor interactions (iTALK), transcription factor networks, and pathway enrichment (GSEA/GSVA) were performed in R. Single-cell metabolic activity was quantified using the scMetabolism R package. CIBERSORTx deconvolution of bulk RNA-seq enabled survival-linked abundance estimates for each cluster. Key subpopulations were validated by multiplex immunofluorescence. In vitro drug screening on MM cells evaluated four metabolic inhibitors-galloflavin (LDHA), CB-839 (GLS1), NAC (OXPHOS), EGCG (multi-target)-and two combination regimens (galloflavin+NAC, CB-839+EGCG) via CCK-8 proliferation assays, flow cytometry, Western blot, and CalcuSyn synergy analysis. A retrospective clinical cohort assessed serum IFN-γ, IL-6, CRP, β2-microglobulin, ferritin, iron, transferrin, and TSAT levels. Results: We defined 23 plasma cell clusters and diverse non-plasma subsets (including HSCs, erythroid progenitors, erythroblasts, macrophages, T and B cell subsets, and CAFs). Relapsed cases were enriched for exhausted T cells, γδ T cells, fibroblast progenitor cells (FPCs), and myeloid-erythroid progenitor cells (MEPCs), all of which correlated with inferior prognosis. Two novel bone marrow subpopulations-ISG15⁺ B cells and LYZ⁺KIT⁺ erythroid progenitors (MEPCs)-expressed immune checkpoint molecules, were survival-associated, and participated in an IFN-γ/IFN-α-B2M positive feedback loop. Cell–cell communication analysis revealed predominant B2M–TFRC signaling and upregulation of KRAS, IL6-JAK-STAT3, MYC, IFN-γ, and IFN-α pathways. IFN-γ from effector memory CD8⁺ T cells and IFN-α from non-classical monocytes established the IFN-γ/IFN-α–B2M feedback circuit via ISGF3-mediated upregulation of B2M and TFRC, sustaining downstream STAT, MYC, and anti-apoptotic signaling. Chronic, sustained, and excessive activation of the IFN-α signaling pathway may also contribute to T cell exhaustion. Metabolic profiling showed malignant plasma cells exhibited increased glycolysis, oxidative phosphorylation (OXPHOS), and glutamine metabolism, with overexpression of LDHA, GLS1, and related enzymes. All four agents suppressed MM cell proliferation; the CB-839+EGCG combination demonstrated strong synergy (CI <1), induced G2/M arrest, enhanced apoptosis (↑Bax/Bcl-2 ratio), and markedly inhibited proliferation. Clinical specimens confirmed elevated IFN-γ, IL-6, CRP, β2-microglobulin, and ferritin, alongside reduced serum iron, transferrin, and TSAT in MM patients. Conclusion: This multi-omics study delineates the complex cellular architecture of the MM TME, uncovers a novel IFN-γ/IFN-α–B2M positive feedback loop central to immune evasion and disease progression, and identifies CB-839+EGCG as a promising metabolic therapy. Our atlas provides a comprehensive resource for elucidating MM pathogenesis and tailoring targeted immunometabolic interventions.