Differential requirement for DICER1 activity during the development of mitral and tricuspid valves

Abstract

Mitral and tricuspid valves are essential for unidirectional blood flow in the heart. They are derived from similar cell sources, and yet congenital dysplasia affecting both valves is clinically rare, suggesting the presence of differential regulatory mechanisms underlying their development. Here, we specifically inactivated Dicer1 in the endocardium during cardiogenesis and found that Dicer1 deletion caused congenital mitral valve stenosis and regurgitation, whereas it had no impact on other valves. We showed that hyperplastic mitral valves were caused by abnormal condensation and extracellular matrix (ECM) remodeling. Our single-cell RNA sequencing analysis revealed impaired maturation of mesenchymal cells and abnormal expression of ECM genes in mutant mitral valves. Furthermore, expression of a set of miRNAs that target ECM genes was significantly lower in tricuspid valves compared to mitral valves, consistent with the idea that the miRNAs are differentially required for mitral and tricuspid valve development. We thus reveal miRNA-mediated gene regulation as a novel molecular mechanism that differentially regulates mitral and tricuspid valve development, thereby enhancing our understanding of the non-association of inborn mitral and tricuspid dysplasia observed clinically.

Keywords: Conditional knockout; Heart development; MicroRNA; Valvulogenesis.

Fig. 1.

Endocardial-deletion of Dicer1 leads to mitral valve stenosis and regurgitation. (A) The percentage of living mutant embryos (Nfac1-Cre;Dicer1^loxp/loxp) from the cross between male Nfac1-Cre;Dicer1^loxp/+ and female Dicer1^loxp/loxp mice. Data were calculated from 10 litters of mice at each stage. (B) Whole-mount examination of P1 hearts. The left atrium of the mutant heart was dilated with blood trapped inside. Images representative of ten experiments. (C) Morphological examination of valves at P1. The arrow shows the reduced size of the orifice of the mutant mitral valve. (D) Quantification of leaflet area from different valves using ImageJ. Total areas were added together. The area of control leaflets was set at 100%. (E) Quantification of anterior and posterior leaflets area of mitral valves from control (n=8) and mutant hearts (n=5). Error bars show standard error (s.e.m.). (F,G) Mitral valve flow measurement through echocardiography color Doppler. The white arrow indicates regurgitation. Summary data of mitral valve flow velocity from WT (n=5) and mutant (n=6) hearts in shown in G. The box represents the 25–75th percentiles, and the median is indicated. The whiskers show the 5–95th percentiles. **P<0.01; ***P<0.001; ns, not significant (unpaired, two tailed Student's t-test); la, left atrium; ra, right atrium; v, ventricle; AL, anterior leaflet; PL, posterior leaflet.

Fig. 2.

Endocardial deletion of Dicer1 leads to defects in condensation and ECM remodeling in mitral valves. (A) Mitral valves of control (ctrl) and mutant (mut) hearts at E17.5 and E18.5. (B) Quantification of areas of mitral valves (n=5). (C) Sections of control and mutant hearts at E18.5 covering the mitral valve area were stained with antibodies against cardiac myosin heavy chain (MHC, mf20, red) and Ki67 (green). Total nuclei were visualized with DAPI staining (blue). (D) Quantification of proportion of Ki67-postive nuclei from images as in C (n=4). (E) Sections of control and mutant hearts at E18.5 were stained with antibodies against cardiac MHC (mf20, red) and cleaved CASP3 (green). Images representative of four experiments. (F,G) Cardiac sections were stained with an mf20 antibody (red) and DAPI (DAPI) as shown in F. The boxed areas indicate the base (B), middle (M) and distal (D) positions of mitral valves. At least five sections of each heart from four hearts were measured and results are shown in G. We did not observe a difference between anterior and posterior leaflets, and thus data from both leaflets were combined. (H) Collagen gel contraction assay using cells isolated from mitral valves at E18.5. Data were averaged from four independent cultures. At 48 h post incubation, the areas of gels without cells were set at 100%, and the relative area (against the gel without cells) was then calculated. Data in B,D,G,H is shown as mean±s.e.m. *P<0.05; **P<0.01; ****P<0.0001; ns, not significant (unpaired, two tailed Student's t-test).

Fig. 3.

scRNA-Seq analysis of mitral valves of E17.5 hearts. (A) The mitral valves from E17.5 hearts were isolated and were then subjected to scRNA-Seq analysis. Clusters corresponding to cardiomyocytes, epicardial cells, endocardial cells, endothelial cells and valve mesenchymal cells are indicated. (B) The volcano chart shows differentially expressed genes with the x-axis showing the fold of alteration and the y-axis showing the Benjamini–Hochberg adjusted P value. The dotted lines show the threshold (adjusted P<0.05, expression altered by at least 20%). 10 ECM genes are indicated. (C) Trajectory and pseudotime analysis of valve mesenchymal cells. The mesenchymal cells were further divided into five subclusters using Monocle 3.0. The pseudotime is indicated with the color and the trajectory pathways are shown with black lines in the UMAP chart. (D) The percentage of cells in each subcluster.

Fig. 4.

Col1a1, Col3a1 and Tnxb are targets of a group of miRNAs in mitral valves. (A) Cardiac sections (E18.5) were stained with antibodies as indicated. The charts in the right panels show the quantification of immunofluorescence intensity normalized against area measured using ImageJ. The intensity of control samples is set to 1.0. Data are shown as mean±s.e.m. (n=4). (B) Luciferase reporter analysis was performed using various reporter constructs harboring the 3′UTR of indicated genes in tsA58-AVM cells. The value of cells transfected with scrambled miRNA was set at 1.0. (C) RNA samples were purified from cells isolated from wild type tricuspid and mitral valves. qRT-PCR was then performed to examine expression of various miRNAs. The expression level of miRNAs was normalized against Snord65. Data in B,C are shown as mean±s.e.m. (n=4). L, left side, mitral valve; R, right side, tricuspid valve. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001; ns, not significant (unpaired, two tailed Student's t-test).