Single-cell profiling reveals MDSC-driven immunosuppressive remodeling of the tumor microenvironment mediates ruxolitinib resistance in myelofibrosis Free

Introduction Myelofibrosis (MF) is a myeloproliferative neoplasm characterized by clonal hematopoietic stem cell proliferation, bone marrow fibrosis, and extramedullary hematopoiesis. Despite ruxolitinib's frontline role in MF, ∼50% of patients develop resistance with limited salvage options. Uncovering ruxolitinib resistance mechanisms is thus imperative.

Methods Single-cell RNA sequencing (scRNA-seq) was performed on bone marrow samples from 6 MF patients (4 ruxolitinib-resistant, 2 ruxolitinib-sensitive controls). Cell types were annotated using marker genes, and differential expression/pathway analysis was conducted. Pseudotime analysis was used to reconstruct differentiation trajectories and cell-cell communication networks were inferred to map microenvironmental crosstalk. Flow cytometric validation was systematically performed on an independent cohort to validate scRNA-seq-predicted MDSC expansion.

Results Our integrative analysis identified 12 clusters and 7 major cell types in scRNA-seq data, with significant expansion of myeloid cells and concurrent reduction in T cells in ruxolitinib-resistant versus sensitive MF patients. Differential expression analysis revealed upregulation of S100A8/S100A9 genes (log2FC>1, p<0.001) in myeloid cells, and KEGG pathway enrichment demonstrated compensatory activation of PI3K-Akt, MAPK, Rap1 signaling. Subclustering of the expanded myeloid lineage uncovered dominant accumulation of MDSC subsets (including M-MDSC and G-MDSC) accounting for 49.99% and 17.69% of the population. Pseudotime trajectory analysis further indicated myeloid differentiation arrest at the MDSC stage in ruxolitinib-resistant patients, with trajectory-dependent genes S100A/S100A9 and PADI4 showing progressive dysregulation along the pseudotime axis (by Monocle2). CellChat analysis revealed fundamentally reprogrammed interaction networks in ruxolitinib resistance: physiological crosstalk between naïve to CD8⁺T cells dominated in sensitive patients, whereas resistant specimens exhibited pathologically enhanced signaling from M-MDSC and G-MDSC to CD8⁺Tcm. Ligand-receptor analysis within specific cellular niches revealed resistin as a resistance-specific pathway. This rewired crosstalk potently sustains inflammation amplification and T-cell suppression in refractory MF.

Conclusion Our single-cell analysis establishes MDSC expansion as a hallmark of ruxolitinib resistance in myelofibrosis, characterized by S100A8/S100A9 overexpression and activation of bypass signaling pathways (PI3K/AKT, MAPK, Raf1). MDSCs drive a self-reinforcing immunosuppressive niche that suppresses CD8⁺ T-cell cytotoxicity through resistin signaling. These insights nominate co-targeting strategies (e.g., S100A8/A9 inhibitors or resistin antibody) to overcome resistance.

Acknowledgement: This research was funded by the Zhejiang Provincial Health High-level Innovative Talent Project (2022-2026).

*Correspondence to:

Jian Huang, M.D., Ph.D., Department of Hematology, The First Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou 310003, Zhejiang, China. E-mail:househuang@zju.edu.cn