Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a curative therapy for a variety of malignant and non-malignant hematologic diseases. However, its success is often limited by graft-versus-host disease (GVHD), a major complication in which donor T cells recognize recipient antigens as foreign, leading to widespread tissue damage. Regulatory T cells (Tregs) are essential for controlling alloimmune responses and promoting immune tolerance. Adoptive transfer of Tregs has emerged as a promising strategy to reduce GVHD without compromising graft-versus-tumor (GVT) effects. Clinical trials using defined ratios of donor Tregs and conventional T cells (Tcons) in allo-HSCT have demonstrated improved GVHD-free, relapse-free survival (GRFS). Despite these advances, the molecular characteristics of Tregs that mediate immune tolerance in the human bone marrow (BM)—a critical site of hematopoietic recovery—remain poorly understood. To characterize transcriptional and chromatin accessibility differences between BM and peripheral blood (PB) Tregs, we analyzed publicly available single-cell RNA sequencing (scRNA-seq) and single-cell ATAC sequencing (scATAC-seq) datasets from bone marrow mononuclear cells (BMMCs) and peripheral blood mononuclear cells (PBMCs). We performed shared nearest neighbor clustering, pseudobulk differential expression analysis, and gene ontology enrichment to identify transcriptional and epigenetic programs enriched in BM Tregs. From the scRNA-seq analysis, we identified differential expression of surface markers such as CXCR4 in BM Tregs. The scATAC-seq analysis further revealed differential accessibility at loci associated with FOXP3-related transcriptional regulators. To validate and extend these findings at the protein level, we developed a mass cytometry (CyTOF) panel informed by our transcriptomic and epigenomic results. Using this panel, we profiled BMMCs and PBMCs from healthy adults (n = 8) and identified distinct surface markers and transcriptional regulators that differentiate Treg populations across compartments. Notably, an average of 16.2% of FOXP3+ cells in peripheral blood expressed CXCR4, compared to 37.4% in bone marrow. Given that most adult allo-HSCT procedures utilize mobilized peripheral blood rather than bone marrow as the graft source, we investigated whether the phenotypic differences between BM and PB Tregs could also be detected in mobilized apheresis products. Using CyTOF, we analyzed Tregs from donor peripheral blood collected under two mobilization regimens: granulocyte colony-stimulating factor (G-CSF) alone, and G-CSF combined with plerixafor (n = 8). Several markers differentially expressed between BM and PB Tregs were also differentially expressed between the G-CSF and G-CSF + plerixafor conditions. In particular, we found that an average of 12.15% of FOXP3+ cells in G-CSF–treated donors expressed CXCR4, compared to 37.4% in donors treated with G-CSF + plerixafor (n=8). To determine the functional relevance of these markers, we examined their expression in expanded Tregs and performed gene-editing experiments. Tregs from healthy PBMCs were expanded ex vivo using IL-2 and anti-CD3/CD28 tetramers, and CyTOF analysis confirmed expression of several key markers, including CXCR4, which was expressed on an average of 65% of expanded Tregs (n = 4). Using CRISPR-Cas9, we knocked out individual markers in expanded Tregs and characterized the edited cells by CyTOF. Loss of specific markers altered the expression of other proteins, suggesting regulatory interactions. For example, CXCR4 knockout reduced FOXP3 expression from an average of 67.25% in mock-edited cells to 38.68% in knockout cells (n = 4). We further conducted bulk ATAC-seq and RNA-seq on edited and unedited Tregs, revealing changes in chromatin accessibility and gene expression. Suppression assays using conventional T cells and CRISPR-edited Tregs showed that marker deletion also affected Treg suppressive function. Together, our work identifies distinct transcriptional, epigenetic, and phenotypic features of human BM Tregs and demonstrates that specific markers contribute to their suppressive function. These findings provide mechanistic insight into Treg biology in the context of allo-HSCT and offer strategies to enhance Treg-based cell therapies for GVHD prevention.

This content is only available as a PDF.
Sign in via your Institution