Fig. 7: Assessment of vascular alterations in human CHIP, MDS, and a CHIP Dnmt3a mouse model.

A Violin plots of angiogenesis signature scores across stromal cell types comparing Control, CHIP, and MDS. B Violin plots of representative genes within the angiogenesis signature across the stromal cell types comparing Control, CHIP, and MDS. C Immunofluorescence of BM sections showing UEA1 (green, all vasculature) and CD105 (red, sinusoids) and DAPI (blue) across Control, CHIP, and MDS. Asterisks are highlighting autofluorescence from erythrocytes. Scale bars: 25 µm. D, E Quantification of CD105⁺ sinusoids (D) and CD105^-UEA1⁺ arterioles/arteries (E) across conditions (n = 3/condition). Medians with 95% confidence intervals are shown. Statistical test: one-way ANOVA, Tukey’s test. F Scatter plot depicting the nearest distance of the vasculature (in µm) in Control, CHIP, and MDS samples. Shortened distances in MDS indicate disrupted vasculature organization. Medians with 95% confidence intervals are shown. Statistical test: one-way ANOVA with FDR correction. G Schematic of the CHIP Dnmt3a-mutant mouse study groups: femurs were collected from young control (24 weeks), young Dnmt3a^R878H (24 weeks), and old control (68 weeks), n = 3/group. Created in BioRender. Prummel, K. (https://BioRender.com/pqyu7b8). H Giemsa staining from young control, young Dnmt3a^R878H, and old control mouse femurs, illustrating reduced cellularity in young Dnmt3a^R878H and old mice (n = 3/group). Scale bars: 200 µm. I Bar plot showing the cellularity (BM/adipocyte ratio) across the groups (n = 3/group). Both young Dnmt3a^R878H and old control mice display significantly reduced cellularity compared to young controls. Means with SD are shown. Statistical test: one-way ANOVA, Tukey’s test. J Immunofluorescence of EMCN (red) and DAPI (blue) in femur sections from young control, young Dnmt3a^R878H, and old control mice. Scale bars: 100 µm. K Quantification of the total sinusoid area in the femur (n = 3/condition). Means with SD are shown. Statistical test: two-way ANOVA, Tukey’s test. L Quantification of the mean sinusoidal area (sinusoid size) in the femur (n = 3/condition). Means with SD are shown. Statistical test: one-way ANOVA with FDR correction. M Schematic illustration summarizing BM niche remodeling in CHIP and LR-MDS. A, B Statistical test: two-sided Wilcoxon rank-sum test with Benjamini–Hochberg procedure. Created in BioRender. Guezguez, B. (https://BioRender.com/8tl5dtf). Source data are provided in the Source Data file and Supplementary Data 7. n.s. not significant.