Nitric oxide synthase-3 promotes embryonic development of atrioventricular valves

Abstract

Nitric oxide synthase-3 (NOS3) has recently been shown to promote endothelial-to-mesenchymal transition (EndMT) in the developing atrioventricular (AV) canal. The present study was aimed to investigate the role of NOS3 in embryonic development of AV valves. We hypothesized that NOS3 promotes embryonic development of AV valves via EndMT. To test this hypothesis, morphological and functional analysis of AV valves were performed in wild-type (WT) and NOS3(-/-) mice at postnatal day 0. Our data show that the overall size and length of mitral and tricuspid valves were decreased in NOS3(-/-) compared with WT mice. Echocardiographic assessment showed significant regurgitation of mitral and tricuspid valves during systole in NOS3(-/-) mice. These phenotypes were all rescued by cardiac specific NOS3 overexpression. To assess EndMT, immunostaining of Snail1 was performed in the embryonic heart. Both total mesenchymal and Snail1(+) cells in the AV cushion were decreased in NOS3(-/-) compared with WT mice at E10.5 and E12.5, which was completely restored by cardiac specific NOS3 overexpression. In cultured embryonic hearts, NOS3 promoted transforming growth factor (TGFβ), bone morphogenetic protein (BMP2) and Snail1expression through cGMP. Furthermore, mesenchymal cell formation and migration from cultured AV cushion explants were decreased in the NOS3(-/-) compared with WT mice. We conclude that NOS3 promotes AV valve formation during embryonic heart development and deficiency in NOS3 results in AV valve insufficiency.

Figure 1. Malformation of mitral and tricuspid valves in NOS3^−/− mice, which are rescued by cardiomyocyte-specific NOS3 overexpression (NOS3^Tg;NOS3^−/−).

Representative tissue sections with Masson’s trichrome staining of mitral (A–D) and tricuspid (G–J) valves in WT, NOS3^−/−, NOS3^Tg and NOS3^Tg;NOS3^−/− mice at P0. (E and F) Quantification of mitral valve size. (K and L) Quantification of tricuspid valve size. Anterior and posterior leaflets represent the valve leaflet closer to the septum and ventricular free wall, respectively. (M) Valve size was measured at the proximal (hinge) and distal aspects of the leaflets. White arrows represent proximal and distal measurements. Black arrow points to the hinge of the valve. Scale bar = 60 µm. Data are mean ± SEM from 5–8 mice per group. *P<0.05 vs. WT (NOS3^+/+), ^† P<0.05 vs. NOS3^−/− mice.

Figure 2. Regurgitation of mitral valve in NOS3^−/− mice at P0.

Backflow of mitral valves was determined by color (A) and pulsed-wave (B) Doppler echocardiography. Backflow during systole is indicated by arrows. WT mice had no backflow in mitral valves. However, NOS3^−/− mice showed significant mitral valve backflow. (C–D) Quantification of mitral valve regurgitation. Significant regurgitation was observed in NOS3^−/− mice, which were rescued by cardiomyocyte-specific NOS3 overexpression. Data are mean ± SEM from 6–7 mice per group. *P<0.01 vs. all other groups. +/+, WT; −/−, knockout; Tg, transgenic. Tg;−/− and Tg;+/+ indicate NOS^Tg;NOS3^−/− and NOS3^Tg, respectively.

Figure 3. Regurgitation of tricuspid valve in NOS3^−/− mice at P0.

Backflow of tricuspid valves was determined by color (A) and pulsed-wave (B) Doppler echocardiography. Backflow during systole is indicated by arrows. WT mice had minor backflow in tricuspid valves. However, NOS3^−/− mice showed marked tricuspid valve backflow. (C–D) Quantification of tricuspid valve regurgitation. Significant regurgitation was observed in NOS3^−/− mice, which were rescued by cardiomyocyte-specific NOS3 overexpression. Data are mean ± SEM from 6–7 mice per group. *P<0.01 vs. WT (NOS3^+/+), ^† P<0.01 vs. NOS3^−/−. +/+, WT; −/−, knockout; Tg, transgenic. Tg;−/− and Tg;+/+ indicate NOS^Tg;NOS3^−/− and NOS3^Tg, respectively.

Figure 4. Cardiac function determined using echocardiography at P0.

(A–D) Both left and right ventricular ejection fraction and fractional shortening were significantly decreased in the NOS3^−/− mice compared with WT controls, which were rescued by cardiomyocyte-specific NOS3 overexpression. (E–F) E/A ratio showed no significant difference between all groups. (G–H) Left and right atrial size was significantly increased in NOS3^−/− mice compared with WT controls, which was rescued by cardiomyocyte-specific NOS3 overexpression. Data are mean ± SEM from 6–7 mice per group. *P<0.05 vs. WT (NOS3^+/+), ^† P<0.05 vs. NOS3^−/−. +/+, WT; −/−, knockout; Tg, transgenic. Tg;−/− and Tg;+/+ indicate NOS^Tg;NOS3^−/− and NOS3^Tg, respectively.

Figure 5. EndMT is impaired in the AV cushion of NOS3^−/− hearts at E12.5.

(A) 3D reconstructions (Left and middle columns, Red for Snail1⁺ cells; Grey for cushion) and representative Snail1 immunostaining (Right column, Brown staining indicates Snail1⁺ cells). Scale bar = 50 µm. (B) Quantification of Snail⁺ mesenchymal cell volume from 3D reconstructed in E12.5 AV cushions. (C) Total number of mesenchymal cells (Snail positive and negative cells) in E12.5 cushions was quantified. (D) Quantification of endocardial cushion size at E12.5 from 3D reconstructions. (E) Quantification of Snail1⁺ endocardial cushion mesenchymal cells at E10.5 from 3D reconstructions. (F) Quantification of total endocardial cushion mesenchymal cells at E10.5. (G) Quantification of endocardial cushion size at E10.5 from 3D reconstructions. N = 3–4 mice per group. *P<0.05 vs. WT(NOS3^+/+) or WT control. ^† P<0.05 vs. NOS3^−/−.

Figure 6. cGMP regulates expression of Snail1, TGFβ and BMP2 in E12.5 hearts.

(A) cGMP levels in P0 hearts. (B–D) E12.5 ex vivo heart cultures. Cultured WT and NOS3^−/− hearts were treated with ODQ (100 µM), a soluble guanylate cyclase inhibitor, or 8-Br-cGMP (2 mM), a cGMP analog for 6 hrs. Snail1, TGFβ and BMP2 mRNA levels were determined by real-time RT-PCR. (E) E12.5 ex vivo heart cultures from WT and NOS3^−/− mice were treated with recombinant TGFβ protein (10 ng/ml). TGFβ mRNA levels were determined by real-time RT-PCR. N = 6 hearts per group for C–G. *P<0.05 vs. WT(NOS3^+/+) or WT control. ^† P<0.05 vs. NOS3^−/− or NOS3^−/− control.

Figure 7. Atrioventricular (AV) endocardial cushion explant culture from E10.5 NOS3^−/− and WT mice.

(A) Spindle shaped mesenchymal cells (arrows) were formed after 24 hours of culture. Lower panel is an enlargement of the boxed area. (B) The total number of mesenchymal cells per mm of explant cushion tissue was significantly decreased in NOS3^−/− compared to WT explants. Data are mean ± SEM from 6 explants per group. *P<0.05 vs. WT(NOS3^+/+). Black scale bar = 100 µm, white scale bar = 10 µm. (C) Proposed signaling pathway by which NOS3 promotes endocardial EndMT and AV valve development.