CHIP is one of the most common age-related conditions in medicine. It arises when a single hematopoietic stem cell acquires a mutation in genes such as DNMT3A, TET2 or ASXL1, giving its descendants a growth advantage so that an expanding clone of genetically distinct blood cells gradually takes over blood production. The condition is called indeterminate because it causes no obvious disease on its own, yet large population studies have shown that people with CHIP face substantially elevated risks of hematologic malignancies, including acute myeloid leukemia and myelodysplastic syndromes, as well as cardiovascular disease. What has remained murky is precisely how the bone marrow itself changes when a clone begins its silent expansion.

To answer that question, the Toronto team assembled a cohort of more than forty individuals undergoing hip surgery, roughly half of whom carried CHIP-associated mutations and half of whom did not. Because the patients were otherwise comparable in age, body mass index and blood counts, the researchers could isolate the effect of clonal hematopoiesis on the marrow niche without the confounding influence of overt blood disease. From femoral head bone cores obtained during surgery, they freshly isolated both hematopoietic cells and the non-hematopoietic stromal cells that form the structural scaffolding of the marrow, then subjected these populations to single-cell RNA sequencing at unprecedented depth.

The resulting single-cell atlas, encompassing tens of thousands of cells, delivered a striking surprise. The hematopoietic compartment itself looked remarkably normal: the proportions of stem and progenitor cells, mature immune populations, monocyte subsets and T cells were largely preserved between CHIP carriers and controls, and functional assays of colony formation showed no gross defect in the blood cells’ output. Instead, the most dramatic differences appeared in the stromal compartment. Fibroblasts, the connective tissue cells that weave the collagen framework of the marrow, were markedly expanded in CHIP samples and had shifted into transcriptional states that closely resemble cancer-associated fibroblasts, the notorious stromal cells that sculpt and support solid tumors.

Beyond the expansion of fibroblasts, the team detected a coordinated activation of extracellular matrix programs across the mesenchymal and vascular compartments. Cells upregulated collagen genes including COL3A1 and COL6A3, along with broader pathways governing matrix remodeling. Histological staining of the biopsy tissue confirmed the molecular signal: reticulin and Masson’s trichrome stains revealed increased collagen deposition in CHIP marrow, and staining for alpha-smooth muscle actin, a marker of activated myofibroblasts, showed enlarged fibroblast-rich zones. In other words, the marrow of CHIP carriers displayed the early, prefibrotic hallmarks that in full-blown disease characterize myelofibrosis, the scarring disorder seen in advanced bone marrow cancers.

Perhaps the most visually compelling results came from spatial transcriptomics, a technology that maps gene expression onto intact tissue sections rather than dissociated cells. Using the Xenium platform, the researchers discovered that the fibroblast expansion was not diffuse but organized into discrete, fiber-enriched regions that they describe as inflammatory hubs. These spatially confined zones were populated by active fibroblasts and CXCL12-positive stromal cells, macrophages and lymphoid cells, and pathway analysis showed they were strongly enriched for proinflammatory signaling, most notably transforming growth factor-beta and tumor necrosis factor pathways. TGF-beta is the master driver of fibrosis throughout the body, while TNF is a key inflammatory cytokine previously implicated in giving mutant stem cells a selective advantage.

The spatial analysis also revealed altered cellular choreography within these hubs. Multiplex immunohistochemistry showed that monocytes and T cells were preferentially located near alpha-SMA-positive fibroblasts in CHIP marrow, and computational colocalization analysis confirmed significant attraction between these immune populations and the activated stromal cells. Receptor-ligand interaction modeling across cell types uncovered rewired communication networks, including altered signaling through LIF, VCAM1 and the adhesion molecule ADGRE2, suggesting that the mutant hematopoietic clone and its remodeled niche are engaged in an ongoing molecular dialogue that neither population could sustain alone.

To probe the functional consequences of this crosstalk, the team turned to laboratory experiments in which myeloid cells with DNMT3A or TET2 knockdown were cultured alongside primary fibroblasts. These co-culture assays demonstrated that fibroblast contact influenced the behavior of the mutant myeloid cells, including their adhesion properties and expression of molecules such as CD18 and CD312, providing mechanistic support for the idea that the fibro-inflammatory niche actively participates in clonal dynamics rather than merely observing them. While the authors are careful to note that these in vitro systems cannot fully recapitulate the complexity of human marrow, they establish a plausible link between the observed stromal remodeling and the fitness advantages that mutant clones enjoy in an inflammatory environment.

The conceptual implications of the study are considerable. For years, research on CHIP has focused on cell-intrinsic mechanisms: how mutations in DNMT3A and TET2 reprogram stem cell responses to inflammation, allowing mutant cells to tolerate or even exploit inflammatory cytokines that harm their wild-type competitors. The new work adds a crucial cell-extrinsic dimension, showing that CHIP is not simply a story about mutant cells outcompeting normal ones in an unchanged environment, but about the environment itself being transformed into a prefibrotic, inflammation-soaked scaffolding that may feed forward to accelerate malignant progression. This reframing aligns CHIP with a broader principle in cancer biology, where tumor-associated stromal remodeling is increasingly recognized as an active accomplice rather than a passive bystander.

There are also potential clinical dividends. If fibro-inflammatory hubs are indeed a hallmark of CHIP, they could serve as biomarkers for identifying which carriers are most likely to progress to overt malignancy, complementing existing risk models based on clone size and mutation type. More ambitiously, the druggability of stromal and inflammatory pathways raises the prospect of interventions aimed at the niche rather than the clone, an approach that has shown promise in other fibrotic diseases. The researchers have deposited their entire dataset in public repositories, including the CELLxGENE platform, allowing the global community to interrogate this newly mapped prefibrotic landscape. As the population ages and CHIP becomes an ever more common incidental finding, understanding the scarred terrain on which mutant clones march may prove as important as understanding the mutations themselves.

Subject of Research: Fibro-inflammatory remodeling of the bone marrow microenvironment in clonal hematopoiesis of indeterminate potential

Article Title: Prefibrotic bone marrow microenvironment is a hallmark of clonal hematopoiesis

Article References: Aguilar-Navarro, A. G., Edun, G., Gower, M., Kant, J., Li, X., Yang, D., Nader, M., Fernandez, M., Jahangiri, S., Tsao, E., Joshi, P., Kossinna, P., Caloren, L. C., Davey, J. R., Zywiel, M. G., Suderman, R. P., Kapoor, A. I., Akens, M. K., Changoor, A., … Tikhonova, A. N. (2026). Prefibrotic bone marrow microenvironment is a hallmark of clonal hematopoiesis. Nature Immunology. https://doi.org/10.1038/s41590-026-02668-3

Image Credits: AI Generated

DOI: 10.1038/s41590-026-02668-3

Keywords: clonal hematopoiesis, CHIP, bone marrow microenvironment, fibroblasts, fibrosis, inflammation, single-cell RNA sequencing, spatial transcriptomics, TGF-beta, TNF, hematopoietic stem cells, leukemia risk