biostudies-arrayexpressDavid JohnMus musculushttps://www.ebi.ac.uk/biostudies/studies/E-MTAB-17569Clonal haematopoiesis (CH), the expansion of blood-cell clones carrying acquired mutations, is an independent risk factor for cardiovascular disease. Although mutations in DNMT3A and TET2 are well studied, the cardiovascular consequences of KDM6A mutations remain unclear. KDM6A is an X-linked histone demethylase that is commonly mutated in patients with heart failure. Here, we perform multi-omics profiling and functional characterisation of mouse models and patient-derived data to show that haematopoietic KDM6A loss impairs cardiac recovery after myocardial infarction. KDM6A deficiency increases myeloid-cell recruitment to the injured heart and reprogrammes monocytes, macrophages and neutrophils towards inflammatory, migratory and glycolytic states. Patients with heart failure and KDM6A-driven CH show similar pro-inflammatory monocyte signatures. KDM6A-silenced macrophages promote cardiomyocyte hypertrophy and cardiac fibroblast activation. Importantly, IL-1β blockade post-MI rescued the adverse cardiac phenotype of haematopoietic KDM6A loss. These findings identify KDM6A-driven CH as a driver of immune dysregulation and IL-1β signalling as a potential therapeutic target in in CH-associated heart failure.biostudies-arrayexpressNucleic Acid Extraction - Subsequently, red blood cells were lysed with 1x RBC lysis buffer (Biolegend) for 5 minutes. Following washing, cells were stained with CD45 (Biolegend, clone 30-F11) and 7-AAD (BD) and sorting of CD45high, 7-ADDNeg cells were performed for downstream single-cell analyses. Sorting was performed with an Aria Fusion (BD). CD45+ cells were centrifuged (300g, 5min, 4oC) and resuspended in pre-chilled lysis solution (Tris-HCl pH 7.4 10mM, NaCl 10mM, MgCl2 3mM, Tween-20 0.1%, Nonidet P40 Substitute 0.1%, Digitonin 0.01%, 1% BSA) for 5 minutes on ice. Nuclei were then washed with a 1ml wash solution (Tris-HCl pH 7.4 10mM, NaCl 10mM, MgCl2 3mM, Tween-20 0.1%, 1% BSA) and centrifuged (500g, 5min, 4oC). VSequencing - The nuclei suspensions were counted manually and diluted according to manufacturer’s protocol to obtain 10.000 single cell data points per sample. Each sample was run separately on a lane in Chromium controller with Chromium Next GEM Single Cell ATAC Reagent Kits v1.1 (10xGenomics). Single-cell ATACseq library preparation was done using standard protocol. Sequencing was done on Nextseq2000.Library Construction - iable nuclei were stained with 7-ADD and 7-ADDhigh nuclei were FACS sorted before proceeding to downstream single-cell ATAC library preparation.Sample Collection - Mice were sacrificed using cervical dislocation under isofluorane anaesthesia. Hearts were perfused via the left ventricle with cold PBS and excised before being dissected and enzymatically digested (in RPMI 1640 media; 450 U/ ml collagenase I, 125 U/ml collagenase XI, 60 U/ml DNase I, and 60 U/ml hyaluronidase, 30 minutes, 37oC degrees). Following digestion, mechanical dissociation was carried out (gentleMACS™) and cells were filtered through a 40-μm nylon mesh to obtain a single cell suspension. After obtaining the cardiac single cell suspension a tissue debris removal was used according to manufacturer’s instructions (Miltenyi).OrganizationMINSEQE ScoreAssays and DataProcessed DataMAGE-TAB FilesData Transformation - Downstream analyzed using peak vs cells matrices were analyzed using Signac (1.14.0) in R studio with R 4.4.2. Briefly, data were normalized via frequency-inverse document frequency (TF-IDF) normalization. Top variable features were selected for dimensional reduction with singular value decomposition (SVD) on the TD-IDF matrix for latent semantic indexing (LSI). The first LSI component was removed from downstream analysis as it was correlated with sequencing depth (visualized with the DepthCor() function in Signac). Cell clusters were then visualized in a low-dimensional space with UMAP. A gene activity matrix was computed by summing the fragments intersecting the gene body and promoter region to visualize canonical markers and interpret cell clusters, and subsequently log-normalized. To additionally help with the deconvolution of cell clusters, we used our unpaired single-cell RNA sequencing data to identify correlation patterns between the gene activity matrix and scRNA-seq data. This was achieved by identifying a set of anchors between the RNA and ATAC data followed by a transfer of cell labels to the ATAC data. Gene activity was then visualized with CoveragePlot() to ensure validity of cell type assignment. Cluster specific peaks or within each cluster for Kdm6aWT vs Kdm6aΔ-Haem comparison were generated using the “FindMarkers” function. ATAC-peaks were linked to genes for each cell cluster with the gABC-scoring function from STARE (v.1.0.4). The unified ATAC peaks were used as candidate enhancer and the counts per million (CPM) in each peak of each cell cluster used as enhancer activity. Scanpy was used for file access56. As gene annotation, the M27 version from GENCODE was used. The window size for STARE was set to 5MB, the score cutoff to 0.02 and as contact the estimate based on distance was chosen as follows: “STARE_ABCpp -a gencode.vM27.annotation.gtf.gz -b <unified ATAC peaks> -n 5+ -w 5000000 -f False -o <output_dir>”. Processing involving genomic coordinates made use of pybedtools (0.8.1). For analysis of the signal coverage, the bam-files for each cell cluster were created with the help of the bam-merge-master functionality from STITCHIT, and subsequently converted to bigwig files with bamCoverage from deeptools (v3.5.1) (normalize with CPM and bin size of 20). The coverage plots were then generated with the plotHeatmap function from deeptools. For determining the location of regions with respect to genes, each base pair was assigned to a location and the percentage given in relation to the summed base pairs of all peaks. Promoters were defined as ±200 bp around all annotated TSS of all genes, and introns were defined as parts of the gene body that were not labelled as anything else. Colours for the locations were taken from the colorcet Python package based on Glasbey et al63.Volcano plots were generated using EnhancedVolcano in R (v.1.24.0). Upset plots were generated with UpSetR (v1.4.0).Sequence Alignment - Raw reads were aligned against the mouse genome (mm39, Ensembl 104 release) and counted by StarSolo. Peak calling was done by MACS and the parameters used were --nomodel --shift -100 --extsize 200 --broad. Preprocessed counts were further analyzed using Scanpy. Basic cell quality control was conducted by the following criteria: a.) mean fragment length = 300, b.) 7.8 < total counts per cell (log) < 12, c.) Reads in peaks > 0.2, d.) Reads in Chr.M < 0.1.MetabolomicsUnknownTranscriptomicsGenomicsProteomicsIllumina NovaSeq 6000RNA-seq of coding RNA from single cellsMus musculusDavid JohnfalseHaematopoietic loss of KDM6A impairs cardiac recovery in heart failure via epigenetic reprogramming of myeloid cells (scATAC-SEQ)Clonal haematopoiesis (CH), the expansion of blood-cell clones carrying acquired mutations, is an independent risk factor for cardiovascular disease. Although mutations in DNMT3A and TET2 are well studied, the cardiovascular consequences of KDM6A mutations remain unclear. KDM6A is an X-linked histone demethylase that is commonly mutated in patients with heart failure. Here, we perform multi-omics profiling and functional characterisation of mouse models and patient-derived data to show that haematopoietic KDM6A loss impairs cardiac recovery after myocardial infarction. KDM6A deficiency increases myeloid-cell recruitment to the injured heart and reprogrammes monocytes, macrophages and neutrophils towards inflammatory, migratory and glycolytic states. Patients with heart failure and KDM6A-driven CH show similar pro-inflammatory monocyte signatures. KDM6A-silenced macrophages promote cardiomyocyte hypertrophy and cardiac fibroblast activation. Importantly, IL-1β blockade post-MI rescued the adverse cardiac phenotype of haematopoietic KDM6A loss. These findings identify KDM6A-driven CH as a driver of immune dysregulation and IL-1β signalling as a potential therapeutic target in in CH-associated heart failure.2026-09-04T00:00:00Z2026-09-04T16:31:29.343Z2026-09-04T12:20:32.555ZE-MTAB-17569ERP204952EFO_0002944EFO_0004170EFO_0005684EFO_0004917EFO_0005518EFO_0003816EFO_0004184