Single-cell transcriptomics reveals phenotypic and functional alterations of erythroid progenitor cells in tuberculosis patients.

Publication date: Mar 01, 2026

Anemia is a common complication of tuberculosis (TB), closely associated with impaired host immune function and poor prognosis; however, its impact on the dynamic alterations of erythroid progenitor cells (EPCs) remains largely unexplored. In this study, we integrated and analyzed single-cell transcriptomic data from peripheral blood mononuclear cells (PBMC) of healthy controls, latent tuberculosis (LTB), TB, and disseminated tuberculosis (DTB) cohorts, as well as 18F-FDG-labeled TB lung tissues and adjacent uninvolved regions. At single-cell resolution, we identified ten EPCs clusters within PBMC, among which clusters 2-6 represented DTB-specific subpopulations (EPC_DTB). These EPC_DTB exhibited metabolic reprogramming (including glycolysis, arginine, and lipid metabolism) accompanied by upregulation of IFITM3/JUN and ribosomal genes, consistent with enhanced differentiation. EPC_DTB and EPC_TB interacted with macrophages, dendritic cells, and monocytes through the ANNEXIN and GALECTIN signaling pathways, while these immune cells reciprocally regulated EPCs subsets via the MIF signaling axis. In lung tissues, EPCs (12 clusters) displayed disease-specific characteristics, with clusters 0, 3, and 6 uniquely present in TB lesions. The EPC_High subpopulation within inflammatory foci showed significant upregulation of GZMA, IL32, CCL5, and CHI3L1, whereas EPC_Low communicated with immune cells, epithelial cells, and stromal cells through the MDK-NCL signaling pathway. Moreover, we established an LTB predictive model based on EPC_LTB signature genes in PBMC (GRIK3, S100B, ZDHHC19, and LRRN3). Our study uncovers the heterogeneity of EPCs across different TB stages and tissue compartments, delineates their bidirectional crosstalk with immune cells, and establishes a link among anemia, immune regulation, and EPCs biology, thereby proposing potential diagnostic biomarkers and therapeutic targets.

Semantics

Original Article