Inherited resilience to clonal hematopoiesis by modifying stem cell RNA regulation

Abstract

Somatic mutations that increase hematopoietic stem cell (HSC) fitness drive their expansion in clonal hematopoiesis (CH) and predispose to blood cancers. Although CH frequently occurs with aging, it rarely progresses to overt malignancy. Population variation in the growth rate and potential of mutant clones suggests the presence of genetic factors protecting against CH, but these remain largely undefined. Here, we identify a non-coding regulatory variant, rs17834140-T, that significantly protects against CH and myeloid malignancies by downregulating HSC-selective expression and function of the RNA-binding protein MSI2. By modeling variant effects and mapping MSI2 binding targets, we uncover an RNA network that maintains human HSCs and influences CH risk. Importantly, rs17834140-T is associated with slower CH expansion rates in humans, and stem cell MSI2 levels modify ASXL1-mutant HSC clonal dominance in experimental models. These findings leverage natural resilience to highlight a key role for post-transcriptional regulation in human HSCs, and offer genetic evidence supporting inhibition of MSI2 or its downstream targets as rational strategies for blood cancer prevention.

Figure 1:. Inherited resilience to CHIP and myeloid malignancy at the 17q22 locus.

(A) Approach for meta-analysis of clonal hematopoiesis of indeterminate potential (CHIP) across UK Biobank (UKB), Geisinger Health Study (GHS) and All of Us (AoU) cohorts. (B) Waterfall plot showing meta-analysis odds ratio effect sizes (95% CI) for the most significant variant at 24 CHIP-associated loci. (C) Protective effect of rs80093687 for CHIP by genotype (top), by driver mutation (middle), or for myeloid malignancies (MyMs) (bottom). AML = acute myeloid leukemia; MPN = myeloproliferative neoplasm; MDS = myelodysplastic syndrome. (D) Associations between rs80093687 and blood cell traits. PLT CRIT = plateletcrit; MPV = mean platelet volume; EO = eosinophil; PLT = platelet; WBC = white blood cell; NEUTRO = neutrophil, MONO = monocyte; RETIC = reticulocyte; LYMPH = lymphocyte; RBC = red blood cell; # = count.

Figure 2:. Regulatory variant rs17834140-T protects from CHIP through HSC-selective downregulation of MSI2.

(A) LocusZoom plot identifying rs17834140 at the 17q22 CHIP-resilience haplotype (R²=1.0 with rs80093687) as a putative causal regulatory variant, with tracks for chromatin accessibility (ATAC) and transcription factor binding (ChIP-seq) in primary human HSPCs. (B) Normalized chromatin accessibility of the rs17834140-harboring regulatory element in 19 hematopoietic cell types, indicating HSC-selective regulatory activity. (C) H3K27ac-HiChIP track showing physical contact between rs17834140 and the MSI2 promoter in human HSPCs. (D) Reduced MSI2 mRNA expression in CD34⁺CD45RA⁻CD90⁺ HSC-enriched cells, 3 days after CRISPR interference targeting the rs17834140-harboring regulatory element. N=3 donors. (E) Reporter assay showing enhancer activity of rs17834140 wild-type (C) or CHIP-resilience (T) alleles in CD34⁺CD45RA⁻CD90⁺ HSC-enriched cells. N=3 donors. (F) rs17834140 is predicted to disrupt a conserved GATA2 binding motif, a factor known to bind at this site. (G) Allele-specific ATAC-sequencing reads at rs17834140 in HSPCs showing reduced chromatin accessibility with the resilience (T) allele. N=3 donors. (H) Editing strategy modeling rs17834140-T effect with ~100bp (ENH-1) or ~300bp (ENH-2) microdeletions, or MSI2-knockout (MSI2-KO). (I) Reduced MSI2 mRNA expression in CD34⁺CD45RA⁻CD90⁺ HSC-enriched cells, 3 days after enhancer microdeletions. N=4 donors. (J) Flow cytometry showing reduced intracellular MSI2 protein abundance in CD34⁺CD45RA⁻CD90⁺ gated HSC-enriched cells, 3 days after editing. N=3 donors.

Figure 3:. Genetic variation-driven loss of MSI2 enhancer reduces human HSC fitness.

(A) Outline of HSC assays used to assess CHIP-resilience variant effect. (B) Representative flow cytometry gating for LT-HSC (CD34⁺CD45RA⁻CD90⁺CD133⁺EPCR⁺ITGA3⁺) immunophenotyping, six days after editing adult CD34⁺ HSPCs. (C-D) Proportion and total numbers of phenotypic LT-HSCs on day 8 of serum-free culture (6 days post-editing) of adult CD34⁺ HSPCs, relative to AAVS1-edited controls. N=4 donors. (E-F) Proportion and total numbers of phenotypic LT-HSCs on day 14 of cytokine-free culture (11 days post-editing) of cord blood CD34⁺ HSPCs, relative to AAVS1-edited controls. N=3–5 donors. (G) Total human cells achieving long-term engraftment in bone marrow (BM) of NBSGW mice, 16-weeks after transplantation with AAVS1, ENH-1 or MSI2-KO edited cord blood HSPCs. N=12 mice. (H) Total human CD45⁺ engrafted cells in BM expressing lineage markers (CD34⁺ HSPCs, CD19⁺ B-cells, CD33⁺ myeloid cells). N=12 mice. (I) Number of colonies formed by BM-selected human CD34⁺ cells cultured in methylcellulose, 16 weeks post-transplantation. N=12 mice.

Figure 4:. Modified RNA regulation underlying CHIP-resilience in human HSCs.

(A) Overview of approach to determine MSI2-regulated RNA network in human HSCs. (B) MSI2-HyperTRIBE edit sites are enriched in 3’ UTR of target mRNAs with the UAG MSI2-recognition motif. (C) Volcano plot of differentially expressed genes (DEGs) between ENH-1 vs. AAVS1-edited HSCs. (D) Gene set enrichment analysis (GSEA) of ENH-1 vs. AAVS1 edited HSCs. (E) CHIP-resilient HSCs (ENH-1 vs. AAVS1) show negative enrichment for CD34⁺ MSI2-HyperTRIBE mRNA targets. (F) Identification of an RNA network that is MSI2-bound and functionally downregulated in CHIP-resilient HSCs. (G) Ribo-seq profiling of human HSCs with GSEA showing CHIP-resilience network mRNAs are enriched for ribosome protected fragments (RPFs). (H) TET2-mutant HSC/MPPs show positive enrichment for the MSI2-DOWN CHIP-resilience network, relative to their wild-type counterparts from the same donors. (I) Overall survival of pediatric acute myeloid leukemia (AML) patients in the TARGET-AML cohort, stratified by expression of either MSI2 or the MSI2-DOWN CHIP-resilience network. Log-rank test.

Figure 5:. Stem cell MSI2 levels modify clonal dominance of ASXL1-mutant HSCs.

(A) Approach to assess associations between rs17834140-T and longitudinal CHIP progression in a cohort with serial blood samples ~6 years apart. (B) Growth rate [% VAF/year] of CHIP-mutations stratified by C/C (N=481) or C/T (N=32) genotype at rs17834140. (C) Odds of CHIP transience (>2% VAF at baseline, <2% VAF at follow-up) by rs17834140-T for driver mutations. (D) Dual modeling of germline CHIP-resilience (ENH-1 microdeletion, MSI2-knockout, AAVS1 control) with ASXL1-editing in primary adult CD34⁺ HSPCs, quantifying LT-HSCs 7 days post-editing. N=3 donors. (E) RNA-seq of CD34⁺CD45RA⁻CD90⁺ HSC-enriched cells showing that ASXL1 mutations enrich for an activated HSC signature, which is attenuated by MSI2 loss (ENH-1 microdeletion or MSI2-knockout). (F) Workflow of murine model evaluating MSI2 and Asxl1 cooperativity. (G) Quantification of Lin⁻Sca-1⁺c-Kit⁺ (LSK) cells, 1-month post-doxycycline treatment in murine model. N=10–17 mice per condition. (H) Peripheral red blood cell (RBC) and white blood cell (WBC) counts 8-months post-doxycycline in murine model. N=10–16 mice per condition. (I) Representative blood smears from bone marrow and peripheral blood, showing myelodysplastic syndrome (MDS) features (hyposegmented Pelger-Huët-like neutrophils and binucleate erythroid precursors) in Asxl1^f/f+MSI2^DOX mice. (J) Model of how MSI2 levels in HSCs regulate stem cell pool size (grey) and clonal advantage of somatic CHIP driver mutations (red).