Abstract Introduction: Clonal hematopoiesis (CH) is common with aging and associated with an increased risk for atherosclerosis (AS) and other cardiovascular diseases (CVD), although molecular mechanisms underlying atherogenesis in CH are only beginning to emerge. TET2 is among the most commonly mutated genes in CH and TET2 mutant CH (TET2 CH) increases AS in mouse models (Jaiswal et al, NEJM, 2017). While Tet2 mutant macrophages induce inflammatory cytokine genes and secrete cytokines linked to atherogenesis in mice, how this occurs is unclear. High Mobility Group A1 (HMGA1) chromatin regulators activate transcriptional networks involved in clonal expansion and inflammatory signals in diverse settings, including myeloproliferative neoplasms, acute leukemias (Li et al, Blood 2022), and solid tumors (Luo et al, JCI 2025, Chia et al, JCI 2023). Moreover, variants in the HMGA1 locus increase the risk for CH. We therefore hypothesized that HMGA1 drives clonal expansion, inflammatory signaling, and atheroma formation in the setting of TET2 CH. Methods: To investigate HMGA1 in TET2 CH AS, we used an established mouse model of AS with biallelic deletion of the low-density lipoprotein receptor gene (Ldlr-/-). Ldlr-/- mice underwent bone marrow (BM) transplantation from donors with hematopoietic-specific (vav), biallelic Tet2 mutation and intact Hmga1 or Hmga1 deficiency (heterozygous or homozygous). Following a high fat diet for 18 weeks, we compared atherogenesis, macrophage infiltration, and serum cytokines. Serum lipid profiles were assessed in all mice. To define mechanisms underlying Hmga1 in Tet2 CH, we analyzed BM-derived c-Kit+ cells via single cell RNA sequencing (scRNAseq) followed by gene set enrichment analysis (GSEA). Results:We discovered a striking decrease in atheromatous plaque area of the descending aorta (-50%, p0.005) inLdlr-/- mice with Tet2 CH and Hmga1 heterozygous deficiency compared to those with Tet2 CH and intact Hmga1. Decreases were more pronounced in mice with Tet2 CH and Hmga1 homozygous deficiency (-65%, p0.001).Macrophage infiltration was markedly decreased in the kidney, spleen, and lungs of Tet2 CH mice with Hmga1haploinsufficiency. Within the aortic root, Hmga1 deficiency also decreased plaque sizes (-43.5%, p0.001) and macrophage infiltration (-36.6%, p0.0001). Serum cytokine analysis revealed a significant decrease in inflammatory cytokines from Tet2 CH mice with Hmga1 haploinsufficiency, including CXCL1, CXCL2, IL-1α, IL-6, and TNFa, all of which are linked to AS in humans. We also found decreases in other serum factors implicated in CVD, such as insulin-like growth factor binding proteins (IGFBPs 1, 2, 3, 5, & 6), E-selectin, and P-selectin. By contrast, serum lipid levels were similar across all genotypes. scRNAseq revealed that intact Hmga1 in Tet2 CH expands the mutant HSC population andgranulocyte/macrophage progenitors (GMPs) compared to Tet2 CH with Hmga1 haploinsufficiency. To elucidate mechanisms underlying Hmga1, we performed GSEA which revealed that Hmga1 activates multiple pathways governing inflammation, including: inflammatory response, TNFa via NFkB signaling, and interferon alpha signaling networks. Intriguingly, Hmga1 also up-regulates gene sets involved in endothelial cell activation and leukocyte trafficking. Preliminary genome wide association studies (GWAS) of human cohorts show that HMGA1 variants are associated with an increased risk for CH and CVD, including AS. Conclusions: Together, our studies reveal Hmga1 as a novel driver of atherogenesis in Tet2-mediated CH. Hmga1 expands the Tet2 mutant HSC compartment and increases pro-inflammatory granulocyte-monocyte progenitors in the bone marrow, resulting in downstream increases in macrophage infiltration, circulating inflammatory cytokines, and atheroma formation in a mouse model of AS and TET2 CH. Strikingly, loss of just a single Hmga1 allele within the hematopoietic compartment is sufficient to decrease atherogenesis and circulating levels of inflammatory cytokines. Mechanistically, Hmga1 induces inflammatory gene networks including those that activate endothelial cells. Finally, HMGA1 gene variants are associated with CH and AS in GWAS, underscoring the relevance of our studies to humans. Further studies to dissect actionable mechanisms mediated by HMGA1 have the potential to reveal new therapeutic interventions to prevent AS and associated CVD.
Thompson et al. (Mon,) studied this question.