Human Stem Cells Retain Inflammatory Memory That Shapes Lifelong Blood Health

Inflammation is a fundamental driver of biological aging and blood malignancy, yet how human haematopoietic stem cells (HSCs) — the long-lived progenitors that sustain lifelong blood production — adapt to repeated inflammatory insults has remained poorly understood. This landmark study in Nature addresses that gap by introducing a new experimental framework and identifying a functionally and molecularly distinct HSC state with lasting consequences for health.

The research team developed xenograft inflammation-recovery models in which human HSCs were exposed to inflammatory stimuli and then analyzed at single-cell resolution using multiomics (combined transcriptomics and epigenomics). This approach revealed two transcriptionally and epigenetically distinct HSC subsets. One, termed HSC inflammatory memory (HSC-iM), uniquely retained a molecular imprint of prior inflammatory exposure even after apparent recovery, demonstrating that HSCs possess a form of non-genetic cellular memory.

Functionally, HSC-iM cells displayed enhanced quiescence — a state of reduced cycling activity — and produced less haematopoietic output compared to non-iM HSCs. This restrained behavior may be a protective adaptation to preserve the HSC pool after inflammatory depletion, but it simultaneously limits regenerative capacity. The HSC-iM molecular signature was validated across multiple physiological settings: it was enriched in HSCs from COVID-19 recovery patients, individuals with sickle cell disease, aged donors, and those with clonal haematopoiesis, confirming that the xenograft model faithfully recapitulates in vivo biology.

The study also explored how clonal haematopoiesis (CH) mutations — somatic mutations that accumulate with age and confer a competitive advantage to HSC clones — intersect with HSC-iM. Notably, CH mutations within HSC-iM cells attenuated the quiescence phenotype by promoting HSC activation and differentiation, suggesting these mutations may partly escape the suppressive effects of inflammatory memory, potentially explaining why CH clones expand preferentially with age and repeated inflammation.

Perhaps most clinically significant, the HSC-iM transcriptional program was transmitted to differentiated immune progeny in both xenograft and physiological settings, meaning the inflammatory memory extends beyond stem cells to influence mature immune cell function. Population cohort analyses then demonstrated that enrichment of the HSC-iM gene program in circulating blood cells was associated with a significantly elevated all-cause mortality risk score, directly linking this stem cell state to heterogeneous health outcomes across human lifespans. Together, these findings reframe HSCs not as passive reservoirs but as active recorders of inflammatory history, with far-reaching implications for aging biology, immunology, and hematologic disease.