Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Dec:152:105252.
doi: 10.1016/j.exphem.2025.105252. Epub 2025 Sep 17.

A 30-gene classifier distinguishes low-risk MDS HSPCs from healthy HSPCs

Affiliations

A 30-gene classifier distinguishes low-risk MDS HSPCs from healthy HSPCs

Pawan Bhat et al. Exp Hematol. 2025 Dec.

Abstract

Myelodysplastic syndromes (MDS) are a group of malignant clonal disorders that are characterized by functional impairment of hematopoiesis, morphologic dysplasia, and genetic heterogeneity. While less likely to transform to acute leukemia, lower-risk MDS (LR-MDS) include patients with IPSS-M moderate low risk, low risk, and very low risk patients and have a limited median survival of 3 to 10 years. Further, there is growing interest in discovering translational targets of LR-MDS pathophysiology. Clonal populations within the hematopoietic stem and progenitor (HSPC) to myeloid differentiation spectrum are widely considered to be a major contributor to MDS pathophysiology. A granular assessment of cell-type and lineage-specific states that contribute to LR-MDS pathophysiology remains to be elucidated. Here, we leverage single-cell transcriptomics to characterize cell states across the HSPC-myeloid differentiation landscape in LR-MDS. We develop a 30-gene score to classify LR-MDS HSPCs and identify novel molecular features of LR-MDS. The genes in our score suggest dysfunction in vesicular trafficking, which we further resolve across the myeloid differentiation axis. The gene products of vesicular trafficking-related pathways may be suitable translational targets for LR-MDS.

PubMed Disclaimer

Conflict of interest statement

Declaration of competing interest PBF is supported by a Mark Foundation for Cancer Research Endeavor Award, the Gabrielle's Angel Foundation for Cancer Research, a Novartis Global Scholar Award, NIDDK R56 (R56DK138826), and a VA MERIT Award (I01BX005991). AGB is on the scientific advisory board of NashBio. RSW is supported by the Mark Foundation for Cancer Research Endeavor Award, the Edward P. Evans Foundation, and the O'Neal Comprehensive Cancer Center. RL is supported by the Mark Foundation for Cancer Research Endeavor Award, NIH/NCI (R01CA259480), American Cancer Society (RSG-22-036-01-DMC), and the Gabrielle’s Angel Foundation for Cancer Research. MRS reports consultancy or membership on a board or advisory committee for BMS, CTI, Forma, Geron, GSK, Karyopharm, Rigel, Ryvu, Taiho, and Treadwell; patents and royalties from Boehringer Ingelheim and Empath Biosciences; research funding from ALX Oncology, Astex, Incyte, Takeda, and TG Therapeutics; and equity in Empath Biosciences, Karyopharm, and Ryvu. The other authors do not have any conflicts of interest to declare in relation to this work.

Figures

Figure 1.
Figure 1.. Feature selection for the classification of LR-MDS HSPCs.
(A) Integrated UMAP of LR-MDS (n = 16) and HD (n = 8) transcriptomes. (B) GSEA results with selected pathways (FDR-adjusted p-value < 0.05). (C) Schematic of score development. (D) Network plot of hub genes and co-expressed genes (edge correlation > 0.4). (E) VPS13B and ERC1 expression in single cells, comparing MDS and HD. (F) Validation of MDSPrimID classifier on bulk RNA-seq data from an independent cohort. The validation cohort included LR-MDS samples, classified as IPSS-R Low or Int-1 (n = 26), and normal samples (n = 23). (G) Differential gene expression of CD34+ cells comparing LR-MDS and HD samples from the validation cohort. Genes from MDSPrimID that were significantly upregulated or downregulated are labeled (Log 2-Fold Change > 0.5 or < −0.5, FDR-adjusted p-value < 0.05).
Figure 2.
Figure 2.. Resolving MDSPrimID by myeloid differentiation states.
(A) Integrated UMAP colored by HSPC and myeloid metaclusters (top) and pseudotime ordering (bottom). Gray dots represent cells not assigned to HSPC or myeloid metaclusters. (B) Gene expression resolved by pseudotime of differentially expressed MDSPrim genes between MDS and HD (Log 2-Fold Change > 1 or < −1, FDR-adjusted p-value < 0.05). Hierarchical clustering reveals two MDS clusters (MDS1 and MDS2) and one HD cluster. (C,D) Gene-wise expression plots of single cells ordered by pseudotime, organized by transcriptional regulation or vesicular trafficking.

References

    1. Steensma DP Myelodysplastic Syndromes: Diagnosis and Treatment. Mayo Clin. Proc 90, 969–983 (2015). - PubMed
    1. Bernard E et al. Molecular International Prognostic Scoring System for Myelodysplastic Syndromes. NEJM Evid. 1, EVIDoa2200008 (2022).
    1. Sekeres MA & Taylor J Diagnosis and Treatment of Myelodysplastic Syndromes: A Review. JAMA 328, 872–880 (2022). - PubMed
    1. Trowbridge JJ & Starczynowski DT Innate immune pathways and inflammation in hematopoietic aging, clonal hematopoiesis, and MDS. J. Exp. Med 218, e20201544 (2021). - PMC - PubMed
    1. Guess T et al. Distinct Patterns of Clonal Evolution Drive Myelodysplastic Syndrome Progression to Secondary Acute Myeloid Leukemia. Blood Cancer Discov. 3, 316–329 (2022). - PMC - PubMed