Selective advantage of mutant stem cells in clonal hematopoiesis occurs by attenuating the deleterious effects of inflammation and aging

Abstract

Clonal hematopoiesis (CH) arises when hematopoietic stem cells (HSC) acquire mutations in genes, includingDNMT3AandTET2, conferring a competitive advantage through a mechanism that remains unclear. To gain insight into how CH mutations enable gradual clonal expansion, we used single-cell multi-omics with high-fidelity genotyping on CH bone marrow samples. Most of the selective advantage of mutant cells occurs within HSCs.DNMT3AandTET2-mutant clones expand further in early progenitors, whileTET2mutations accelerate myeloid maturation in a dose-dependent manner. Unexpectedly, both mutant and non-mutant HSCs from CH samples are enriched for inflammatory and aging transcriptomic signatures, compared to HSC from non-CH samples, revealing a non-cell autonomous mechanism. However,DNMT3AandTET2-mutant HSCs have an attenuated inflammatory response relative to wild-type HSCs within the same sample. Our data support a model whereby CH clones are gradually selected because they are more resistant to the deleterious impact of inflammation and aging.

Identification of a human hematopoietic stem cell subset that retains memory of inflammatory stress

“…6I, and Table S13). Notably, downregulation of inflammatory genes IL1R1 and ID2 within CH MUT HSC2/HSC-iM was found, in line with observed downregulation of TNFα via NFkB signaling within this population (72). These results are similar to those from the in vivo inflammation-recovery model wherein molecular changes following inflammatory treatment occurred predominantly within the HSC-iM population (Fig.…”

“…By establishing an inflammation-recovery xenotransplantation model able to recapitulate HSC states found in aging and CH, we have identified clues towards answering these important questions. Critically, the HSC-iM subset identified from our xenograft model is remarkably similar, at a molecular level, to an HSC subset found in adult BM from our separate CH study (72). We found that DNMT3A and TET2 mutations have both cell-autonomous and non-cellautonomous effects on HSCs from CH donors.…”

“…As "inflammaging" -chronic, low-grade inflammation -develops with physiological aging, we asked whether the HSC-iM signature accumulates in human HSC across the human lifespan (63). To this end, we analyzed a 10x scMultiome bone marrow CD34+ HSPC data set comprised of two older-aged (OA; 70-80yr), two middleaged (MA; 50-60yr), and two young adult (YA; 20-30yr) donors that we generated (64). We sought to minimize donor-specific effects by supplementing our analysis with additional published scRNA-seq datasets of human HSPCs from OA, MA, and YA donors: Ainciburu et al comparing three OA (60-80yr) against three YA (20-30yr) donors (65), Zhang et al comparing three OA (60-80yr) against two YA (20-30yr) donors (66), and published data sets within BoneMarrowMap comparing 14 MA (40-60yr) against 19 YA (18-40yr) donors (67)(68)(69)(70).…”

“…To understand whether inflammatory memory in human HSC is associated with CH, we investigated TARGET-seq+ data from our CH study of nine DNMT3A or TET2-mutated CH donors and four age-matched non-CH controls (Fig. 5E) (64). TARGET-seq+ enables simultaneous single-cell profiling of transcriptome, genotype, and FACS immunophenotype, enabling comparisons of mutant (CH MUT ) and wildtype (CH WT ) HSCs in CH with non-CH HSCs from age-matched controls (non-CH).…”

“…Given the strength of enrichment patterns for HSC-iM in the context of CH, our CH data set presented an opportunity to ask if HSC-I and HSC-iM could be identified in an unbiased manner from human adults where functionally validated immunophenotypic markers (73) were used to isolate HSPC. In our separate CH study (64), we had performed unsupervised feature selection and clustering using the Self-Assembling Manifolds (SAM) tool on primitive HSPC, which revealed three clusters of human HSC within the cohort, termed HSC1, HSC2, and HSC3 (Fig. 6A).…”