RESUMEN
FOG-2 is a multi-zinc finger protein that binds the transcriptional activator GATA4 and modulates GATA4-mediated regulation of target genes during heart development. Our previous work has demonstrated that the Nucleosome Remodeling and Deacetylase (NuRD) complex physically interacts with FOG-2 and is necessary for FOG-2 mediated repression of GATA4 activity in vitro. However, the relevance of this interaction for FOG-2 function in vivo has remained unclear. In this report, we demonstrate the importance of FOG-2/NuRD interaction through the generation and characterization of mice homozygous for a mutation in FOG-2 that disrupts NuRD binding (FOG-2(R3K5A)). These mice exhibit a perinatal lethality and have multiple cardiac malformations, including ventricular and atrial septal defects and a thin ventricular myocardium. To investigate the etiology of the thin myocardium, we measured the rate of cardiomyocyte proliferation in wild-type and FOG-2(R3K5A) developing hearts. We found cardiomyocyte proliferation was reduced by 31±8% in FOG-2(R3K5A) mice. Gene expression analysis indicated that the cell cycle inhibitor Cdkn1a (p21(cip1)) is up-regulated 2.0±0.2-fold in FOG-2(R3K5A) hearts. In addition, we demonstrate that FOG-2 can directly repress the activity of the Cdkn1a gene promoter, suggesting a model by which FOG-2/NuRD promotes ventricular wall thickening by repression of this cell cycle inhibitor. Consistent with this notion, the genetic ablation of Cdkn1a in FOG-2(R3K5A) mice leads to an improvement in left ventricular function and a partial rescue of left ventricular wall thickness. Taken together, our results define a novel mechanism in which FOG-2/NuRD interaction is required for cardiomyocyte proliferation by directly down-regulating the cell cycle inhibitor Cdkn1a during heart development.