Inhibition of DYRK1A, via histone modification, promotes cardiomyocyte cell cycle activation and cardiac repair after myocardial infarction

Figure 1. Loss of DYRK1A promotes cardiomyocyte cell cycle reentry and improves cardiac repair following MI in adult hearts. (a) DYRK1A expression in mouse heart during development and post-MI in adult mouse. (a1) Immunoblotting was used to detect DYRK1A expression in postnatal day 1 (P1) and adult mouse hearts (n=6 mice per group). Representative immunoblots (left) and quantification of protein levels (right) are shown. ⁎⁎⁎P<0.001 versus P1 group. (a2) Adult mice were subjected to MI or sham-operation, and protein expression of DYRK1A in the heart was determined at day 7 post-MI (n=6 mice per group). Representative immunoblots (left) and quantification of protein levels (right) are shown. ⁎⁎⁎P<0.001 versus sham group. (b) Schematic diagram of cardiomyocyte-specific deletion of DYRK1A (DYRK1A CKO). DYRK1Aflox/flox mice were crossed with α-MHCMerCreMer mice to generate tamoxifen-inducible DYRK1A CKO mice. (c) Overview of the experimental setup. After the intraperitoneal injection (IP) of tamoxifen every other day thrice, control (αMHCMerCreMer) and α-MHCMerCreMer/DYRK1Aflox/flox mice (DYRK1A CKO) were subjected to MI. Echocardiography was performed every 7 days until removal of the heart,…

Figure 2. DYRK1A knockdown induces transcriptional reprogramming of networks that govern cardiomyocyte cell cycle activation. (a) Schematic workflow of transcriptome analysis. (b) Heatmap profile of RNA-sequencing (RNA-seq) analysis to show differentially expressed genes in primary neonatal cardiomyocytes transfected with scramble (si-NC) or DYRK1A siRNA (si-DYRK1A) (n=3 samples per group). (c) Gene ontology (GO) analysis of upregulated genes in si-DYRK1A cardiomyocytes (relative to si-NC group) show multiple, significantly enriched GO terms relevant to cell proliferation. (d) Heatmap profile of upregulated cell cycle regulators (relative to si-NC group). (e–f) Validation of selected cell cycle regulators from RNA-seq by qRT-PCR and immunoblotting (n=4 samples per group in e; n=6 samples per group in f). Representative immunoblots (left) and quantification of protein levels (right) are shown in f. (⁎⁎P<0.01, ⁎⁎⁎P<0.001) versus si-NC group. (g) The dynamic bias of differentially expressed genes in si-NC and si-DYRK1A cardiomyocytes. Gene set enrichment analysis was used to analyze clusters of genes that regulate cell cycle. For e–f, all bars express mean ± SD and data are analyzed using the…

Figure 3. DYRK1A knockdown increases deposition of H3K27ac and H3K4me3 on promoters of cell cycle genes in cardiomyocytes. (a) Changes of histone modifications induced by DYRK1A knockdown in vitro. Immunoblotting was performed to show global changes of histone modifications, using proteins extracted from cardiomyocytes in primary culture transfected with scramble (si-NC) or DYRK1A siRNA (si-DYRK1A) for 48 h (n=6 samples per group). Representative immunoblots (left) and quantification of protein level (right) are shown. ⁎⁎⁎P<0.001 versus si-NC group. (b) Schematic workflow of H3K4me3 and H3K27ac ChIP-seq. (c) Distribution of H3K4me3 or H3K27ac ChIP-seq peaks in chromosomes. The positions of H3K4me3 and H3K27ac enrichment were aligned, according to chromosome position in the outermost circle. (d) Pie chart of the genomic distribution of H3K4me3 or H3K27ac binding peaks, including promoters, 5′ UTR, 3′ UTR, exons, introns, and intergenic regions. (e) ChIP-seq density heatmaps in cardiomyocytes transfected with si-NC or si-DYRK1A. Signals were ranked by H3K4me3 and H3K27ac read intensity within ±2k bp of peak from the transcription start sites. (f) Visualization of H3K4me3 or H3K27ac ChIP-seq…

Figure 4. Inhibition of DYRK1A decreases KAT6A and WDR82 phosphorylation to induce H3K27ac and H3K4me3 modification in cardiomyocytes. (a) Schematic workflow of immunoprecipitation (IP)-mass spectrometry to identify binding partners of DYRK1A in cardiomyocytes. (b) Selection strategy of candidate effectors for DYRK1A in regulating H3K4me3 and H3K27ac enrichments from mass spectrometry profiling. (c) Mass spectrogram profile of WDR82 (upper figure) and KAT6A (lower figure). MS2 spectra of WDR 82 and KAT6A peptides are shown, and the intensity of peptide fragments is plotted against the mass-to-charge ratio (m/z). (d) Validation of the interaction of DYRK1A with WDR82 or KAT6A, and determination of DYRK1A domain responsible for the binding. The binding of HA-tagged full-length DYRK1A (lane 4), or its deletion mutants (N-terminal domain deletion, lane 5; kinase catalytic domain deletion, lane 6; C-terminal domain deletion, lane 7) with FLAG-tagged WDR82 or KAT6A was examined by co-transfection/co-IP experiments. Lane 1, control without transfection; lane 2, control with only DYRK1A transfection. Lane 3 control with only WDR82 or KAT6A transfection. TF: transfection. IP: immunoprecipitation.…

Figure 5. Pharmacological inhibition of DYRK1A promotes cardiac repair after MI in adult hearts. (a) Schematic diagram of the experimental design. Harmine was dissolved in drinking water at a concentration of 0.05 mg/ml. After MI or sham surgery, echocardiography was performed every 7 days until the removal of the hearts. Proliferation and histological analyses were performed at 7 and 35 days, respectively. (b–d) Cardiomyocyte cell cycle activity induced by harmine in vivo. Immunofluorescence staining (left) and quantification (right) of Ki67- (b), pH3- (c), and Aurkb- (d) positive cardiomyocytes in the peri-infarct area in vehicle- or harmine-treated mice. Ki67-, pH3-, and Aurkb-positive cardiomyocytes are indicated by arrows. Scale bar=40 μm in b and c; scale bar=20 μm in d. CMs: cardiomyocytes (>10000 cardiomyocytes from five mice per group). ⁎⁎⁎P<0.001 versus vehicle group. (e–h) Echocardiographic analysis of cardiac function from vehicle- or harmine-treated adult mice. Representative M-mode tracing images (day 35 post-MI) and quantitative analysis are shown. LVEF: left ventricular ejection fraction; LVFS: left ventricular fractional shortening; LVIDs: left ventricular internal…

Figure 6. Graphic abstract. DYRK1A phosphorylates WDR82 and KAT6A to reduce deposition of H3K4me3 and H3K27ac on the promoters of cell cycle genes, thereby limiting their transcriptional activation and inhibiting cardiomyocyte cell cycle activity. Either pharmacological inhibition or genetic deletion of DYRK1A promotes cardiomyocyte cell cycle activation and cardiac repair in adult heart post-MI.