Multifactorial origins of heart and gut defects in nipbl-deficient zebrafish, a model of Cornelia de Lange Syndrome.

Cornelia de Lange Syndrome (CdLS) is the founding member of a class of multi-organ system birth defect syndromes termed cohesinopathies, named for the chromatin-associated protein complex cohesin, which mediates sister chromatid cohesion. Most cases of CdLS are caused by haploinsufficiency for Nipped-B-like (Nipbl), a highly conserved protein that facilitates cohesin loading. Consistent with recent evidence implicating cohesin and Nipbl in transcriptional regulation, both CdLS cell lines and tissues of Nipbl-deficient mice show changes in the expression of hundreds of genes. Nearly all such changes are modest, however--usually less than 1.5-fold--raising the intriguing possibility that, in CdLS, severe developmental defects result from the collective action of many otherwise innocuous perturbations. As a step toward testing this hypothesis, we developed a model of nipbl-deficiency in zebrafish, an organism in which we can quantitatively investigate the combinatorial effects of gene expression changes. After characterizing the structure and embryonic expression of the two zebrafish nipbl genes, we showed that morpholino knockdown of these genes produces a spectrum of specific heart and gut/visceral organ defects with similarities to those in CdLS. Analysis of nipbl morphants further revealed that, as early as gastrulation, expression of genes involved in endodermal differentiation (sox32, sox17, foxa2, and gata5) and left-right patterning (spaw, lefty2, and dnah9) is altered. Experimental manipulation of the levels of several such genes--using RNA injection or morpholino knockdown--implicated both additive and synergistic interactions in causing observed developmental defects. These findings support the view that birth defects in CdLS arise from collective effects of quantitative changes in gene expression. Interestingly, both the phenotypes and gene expression changes in nipbl morphants differed from those in mutants or morphants for genes encoding cohesin subunits, suggesting that the transcriptional functions of Nipbl cannot be ascribed simply to its role in cohesin loading