Genome-wide association mapping and targeted loss of function studies identify Shroom3 as a driver of hyperpolyploidy and ventricular dilation

Figures

• 

  Fig. 1.. Cardiomyocyte (CM) ploidy is variable across the rat strains of the Hybrid Rat Diversity Panel (HRDP). CMs were isolated from weanling (3 to 4-wk-old) rats using Langendorff perfusion with collagenase II to digest the heart into a single-cell preparation. (A) Representative image of CMs labeled with ploidy class. Percent of MNDCMs (B) and hyperpolyploid CMs (C) across 75 strains of the HRDP (N = 3 to 7). Colored bars represent strains that were selected for further studies to assess baseline cardiac physiology. (D) Pearson correlation between MNDCMs and hyperpolyploid CMs.

• 

  Fig. 2.. Frequency of hyperpolyploid CM correlates with echocardiographic measurements of contractility and ventricular lumen size. 8-wk-old female rats were subjected to echocardiography. (A) Example images in systole (s) and diastole (d) for two divergent strains (M520 and HXB23). (B) Left ventricular ejection fraction (%LVEF) across 13 strains of the HRDP (N = 9 to 20) animals per strain. Spearman correlations between frequency of hyperpolyploid to LVEF (C), LV systolic area (D), LV diastolic area (E), and LV mass (F); calculated from long axis B-mode. Images collected for assessment of CM ploidy were used to assess CM dimensions. P-values and r-coefficients indicated on graphs.

• 

  Fig. 3.. Manhattan plots for QTL mapping for frequency of MNDCMs and hyperpolyploid CMs. Frequency of MNDCMs (A) and hyperpolyploid CMs (B) were subjected to QTL mapping using genome-wide efficient mixed-model analysis. Significant (red line) and suggestive (blue line) loci are defined by permutation-based phenotype-specific genome-wide thresholds of 0.05 and 0.63, respectively. (C) Schematic of prioritization strategy for identification of candidate genes associated with frequency of hyperpolyploid CMs. (D) UMAP of annotated clusters identified from single-nucleus RNA sequencing of ventricular tissue in BN-Lx, F344, and M520 (n = 2 hearts per strain). (E–G). Expression of cytoskeleton-related candidate genes Shroom3 (E), Inppl1 (F), and Ptpn21 (G), identified through prioritization of Hyperpolyploid CM loci. Labels indicate the percentage of CMs with positive expression of each gene. (H and I) Association between %hyperpolyploid CM and specific Shroom3 variants. P-values were calculated using a Student t test with Welch’s correction.

• 

  Fig. 4.. Validation of Shroom3 in CM-specific knockout mice. CM ploidy and baseline cardiac function were assessed using 6-wk-old Shroom3 cKO and Myh6-Cre+ control mice. (A) Frequency of each CM population accounting for both nucleation and nuclear ploidy. P-values calculated by MANOVA with Tukey HSD pairwise comparisons. (B) Representative images from echocardiography (M-mode). Echocardiography was performed to obtain %LVEF (C) and left ventricular systolic (D) and diastolic (E) area; calculated from long axis B-mode and (F) left ventricular mass corrected calculated from short axis M-mode. Colored data points indicate females (red) or males (blue). P-values calculated by a Student t test.

• 

  Fig. 5.. Shroom3 variants G1073S and R977W disrupt actin binding. (A) Schematic of SHROOM3 protein with PDZ, ASD1, and ASD2 domains. SNVs found in the HRDP strains that are predicted to be damaging are labeled in their approximate locations across the Top. (B) Representative blots following transfection of HEK293T cells, coimmunoprecipitation for FLAG antibody, and blotting against FLAG, ROCK1, and β-Actin. (C and D) Quantifications for β-Actin and ROCK1 normalized to FLAG. P-values calculated from one-way ANOVA with Tukey’s multiple comparisons test. (E) Coimmunoprecipitation was also performed following Ad-Shroom3-1073G, Ad-Shroom3-1073S, or Ad-GFP infected NRVCMs. Representative blots against FLAG, β-Actin, α-Cardiac Actin. (F and G) Quantifications for β-Actin and α-Cardiac Actin normalized to FLAG. P-values calculated from a Student t test. (H) Heatmap of differentially expressed genes (DEGs) in Shroom3 1073G–infected vs. GFP-infected NRVCMs and Shroom3 1073S infected vs. GFP infected NRVCMs. DEGs defined by |log2(Fold Change)| > 1 and FDR < 5%. Gene clusters are annotated with enriched terms from DAVID (I) Top 10 terms from gene set enrichment analysis (GSEA) with biased expression…