Chakrabarti M, Al-Sammarraie N, Gebere MG, Bhattacharya A, Chopra S, Johnson J, Peña EA, Eberth JF, Poelmann RE, Gittenberger-de Groot AC, Azhar M.J Cardiovasc Dev Dis. 2020 May 24;7(2):E19. doi: 10.3390/jcdd7020019.PMID: 32456345 Free article.
Transforming growth factor beta3 (TGFB3) gene mutations in patients of arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD1) and Loeys-Dietz syndrome-5 (LDS5)/Rienhoff syndrome are associated with cardiomyopathy, cardiac arrhythmia, cardiac fibrosis, cleft palate, aortic aneurysms, and valvular heart disease. Although the developing heart of embryos express Tgfb3, its overarching role remains unclear in cardiovascular development and disease. We used histological, immunohistochemical, and molecular analyses of Tgfb3-/- fetuses and compared them to wildtype littermate controls. The cardiovascular phenotypes were diverse with approximately two thirds of the Tgfb3-/- fetuses having one or more cardiovascular malformations, including abnormal ventricular myocardium (particularly of the right ventricle), outflow tract septal and alignment defects, abnormal aortic and pulmonary trunk walls, and thickening of semilunar and/or atrioventricular valves. Ventricular septal defects (VSD) including the perimembranous VSDs were observed in Tgfb3-/- fetuses with myocardial defects often accompanied by the muscular type VSD. In vitro studies using TGFβ3-deficient fibroblasts in 3-D collagen lattice formation assays indicated that TGFβ3 was required for collagen matrix reorganization. Biochemical studies indicated the ‘paradoxically’ increased activation of canonical (SMAD-dependent) and noncanonical (MAP kinase-dependent) pathways. TGFβ3 is required for cardiovascular development to maintain a balance of canonical and noncanonical TGFβ signaling pathways.
Figure 1 Tgfb3 deletion leads to an impaired myocardium in mouse embryos. (A–C), Cross- sections of E15.5-16.5 fetuses showing cardiac muscle actin (clone HHF35) immunohistochemistry for wildtype (A) and two different TGFβ3-deficient fetuses (B,C). TGFβ3-deficient fetuses developed non-uniform right ventricular myocardium with heterogeneous regions of thicker (B, left arrow) and thinner (B, bottom arrow) myocardium and spongy (B, right arrow) and thicker (C, right arrow) left ventricular myocardium compared to the wildtype littermate fetus (A), Other mutant fetus had a muscular ventricular septal defect (VSD) (C, asterisk) (n = 6). (D,E), Representative images revealed Tgfb3 expression (green-blue punctate dots, arrows) in the compact and trabecular myocardium of the right (D) and left ventricles (E) from wildtype E12.5 embryo. Scale bars = 200 µm for (A–C), 50 µm for (D,E). Abbreviations: rv, right ventricle; lv, left ventricle.
Figure 2 Tgfb3 knockout mouse fetuses exhibit outflow tract septal and alignment defects. (A–D), Hematoxylin and eosin staining of hearts for E15.5-16.5 wildtype (A) and two different Tgfb3−/− fetuses (B–D). Tgfb3−/− fetuses develop both double-outlet right ventricle (B, arrow) and perimembranous ventricular septal defect (C,D; arrows). Scale bars = 200 µm for (A–D). Abbreviations: rv, right ventricle; lv, left ventricle.
Figure 3 Loss of Tgfb3 in mice results in the outflow tract cushion and vascular wall defects. (A–E), Cross-sections of E15.5–16.5 fetuses showing cardiac muscle actin (clone HHF35) immunohistochemistry for wildtype controls (A,C) and TGFβ3-deficient fetuses (B,D,E). Tgfb3−/− fetuses develop dysmorphic pulmonary (B) and aortic valve (D,E), and abnormal ascending aortic and pulmonary trunk walls (B) morphology. Tgfb3−/− fetuses also demonstrating hypoplastic outlet septum (E). (F) Representative image of a wildtype embryo (E11.5) using RNAscope in situ hybridization reveals Tgfb3 expression (brown punctate dots) in the vascular wall (arrowhead) and outflow tract septum (arrow). (G,H), Apoptosis (E13.5) using TUNEL staining (brown colored nuclei) in outflow tract septum mesenchyme. Compared to wildtype controls (G), Tgfb3−/− OFT septum have a reduced number of apoptotic cells (H). (I) Quantification of fraction of cells undergoing apoptosis. Mean ± SEM of % average apoptosis from at least 4 sections for each sample was used for comparison. Quantification was predominantly done in the area of outflow tract septum marked by a dotted circle. Reduced apoptosis in Tgfb3−/− hearts was noted as compared to wildtype embryos (** p = 0.004, Student’s t test; p = 0.07, Nonparametric (Mann Whitney test)). (J–L), Cell proliferation (E13.5) using phospho-histone H3 (Ser10) immunohistochemistry. Mean ± SEM of average pHistoneH3+ cells/section from at least 4 sections for each sample was used for comparison. Quantification was mainly restricted to the region around the fibrous outflow tract septum (L). Increased cell proliferation in Tgfb3−/− hearts (K, arrows) was observed as compared to wildtype (J) embryos (* p = 0.04, Student’s t test; p > 0.05, Nonparametric (Mann Whitney test)). Scale bars = 100 µm for (A–E,F–H,J,K). Abbreviations: rv, right ventricle; av, aortic valve; pv, pulmonary valve; ao, ascending aorta; ots, outlet septum, rvot, right ventricular OFT.
Figure 4 Tgfb3 knockout fetuses exhibit atrioventricular valve thickening. (A,B) H&E stained sections of wildtype (A) and Tgfb3−/− fetus (E14.5-15.5) showing mitral valve and tricuspid valve thickening (B, arrowheads). Note that the ventricular myocardium is thin and abnormal with muscular ventricular septal defect in Tgfb3−/− (B, arrow). (C,D), Cardiac muscle actin (clone HHF35) immunohistochemistry of cross-sections of E15.5-16.5 wildtype (C) and Tgfb3−/− (D) fetuses. Tgfb3−/− fetus develop thickened mitral valves (D, arrow). Scale bars = 200 µm for (A,B), 50 µm for (C,D). Abbreviations: rv, right ventricle; lv, left ventricle; tv, tricuspid valve; mv, mitral valve.
Figure 5 Collagen lattice formation assay. (A–D), dissecting microscope images of fibroblast-laden collagen gels after 5-days using three independent fibroblast cell lines (n = 3) from wildtype (A,C) or Tgfb3−/− embryos (B,D) in the presence of low (0.1 ng/mL) (A,B) and high (1 ng/mL) (C,D) doses of exogenous recombinant TGFB1. TGFβ3-deficient whole mouse embryonic fibroblasts treated with the low dose of exogenous TGFB1 have decreased contractility compared to the wildtype fibroblasts while higher doses rescue contractility and lattice formation (E). p-Values are indicated in the histogram. Numerical data are presented as scatter dot-plots with the box denoting the mean; error bars identify the SEM.
Figure 6 TGFβ3 is required for TGFβ-SMAD cardiovascular pathway activation. (A–C), Western blot analysis of fetal hearts and the accompanied ascending aorta (E18.5) shows protein levels of the phosphorylated SMADs (pSMAD2, pSMAD3, pSMAD1/5) and total SMADs (SMAD2, SMAD3, SMAD1/5). (D) A common and independent β-actin blot (C, bottom) was used for normalizing data from pSMAD2, pSMAD3, and pSMAD1/5 blots. There was no difference in the β-actin levels between any cardiovascular tissues taken from wildtype and Tgfb3−/− mice. Densitometric quantification of phosphorylated proteins after normalization to β-actin and total non-phosphorylated proteins is shown on the right (A–C). E, Representative western blots of pooled samples (3 hearts and aortas/sample) from wildtype, Tgfb3+/−, and Tgfb3−/− fetuses (E14.5) for pSMAD2, SMAD2, and β-actin. Each western blot was repeated three times with similar results. Western blots and densitometric quantification show the levels of pSMAD2 (60 kDa) and SMAD2 (60 kDa) (A,D), pSMAD3 (52 kDa) and SMAD3 (52 kDa) (B), pSMAD1/5 (60 kDa) and SMAD1/5 (60 kDa) (C), and β-actin (42 kDa) (A–D). Note that levels of pSMAD2, pSMAD3, and pSMAD1/5 are increased in individual samples taken from TGFβ3-deficient mice (A–C). The levels of pSMAD2 were also higher in pooled samples of both Tgfb3+/− and Tgfb3−/− fetuses compared to pooled sample from wildtype fetuses. Tgfb3−/− had slightly higher levels of pSMAD2 than the Tgfb3+/− fetuses. Data points excluded from statistical analysis (Student’s t test) are indicated by grey symbols (A–C). Individual densitometric values from each pooled samples were plotted (E). All western blots (A–C) with individual samples were done in triplicate. Thus, all data points represent an average values of three independent blots (A–C). p-values are shown in the figure. Numerical data from multiple individual samples are presented as scatter dot-plots with boxes denoting the mean; error bars indicate the SEM.