Hemato-vascular specification requires arnt1 and arnt2 genes in zebrafish embryos.

During embryonic development, a subset of cells in the mesoderm germ layer are specified as hemato-vascular progenitor cells, which then differentiate into endothelial cells and hematopoietic stem and progenitor cells. In zebrafish, the transcription factor npas4l (cloche) is required for the specification of hemato-vascular progenitor cells. However, it is unclear whether npas4l is the sole factor at the top of the hemato-vascular specification cascade. Here, we show that arnt1 and arnt2 genes are required for hemato-vascular specification. We found that arnt1;arnt2 double mutant zebrafish embryos, but not arnt1 or arnt2 single mutants, lack blood cells and most endothelial cells. arnt1/2 mutants have reduced or absent expression of etsrp and tal1, the earliest known endothelial and hematopoietic transcription factor genes. We found that Npas4l binds both Arnt1 and Arnt2 proteins in vitro, consistent with the idea that PAS domain-containing bHLH transcription factors act in a multimeric complex to regulate gene expression. Our results demonstrate that npas4l, arnt1 and arnt2 act together to regulate endothelial and hematopoietic cell fate, where each gene is necessary, but not sufficient, to drive hemato-vascular specification.

Mechanistic toxicology in light of genetic compensation.

Mechanistic toxicology seeks to identify the molecular and cellular mechanisms by which toxicants exert their deleterious effects. One powerful approach is to generate mutations in genes that respond to a particular toxicant, and then test how such mutations change the effects of the toxicant. CRISPR is a rapid and versatile approach to generate mutations in cultured cells and in animal models. Many studies use CRISPR to generate short insertions or deletions in a target gene and then assume that the resulting mutation, such as a premature termination codon, causes a loss of functional protein. However, recent studies demonstrate that this assumption is flawed. Cells can compensate for short insertion and deletion mutations, leading toxicologists to draw erroneous conclusions from mutant studies. In this review, we will discuss mechanisms by which a mutation in one gene may be rescued by compensatory activity. We will discuss how CRISPR insertion and deletion mutations are susceptible to compensation by transcriptional adaptation, alternative splicing, and rescue by maternally derived gene products. We will review evidence that measuring levels of messenger RNA transcribed from a mutated gene is an unreliable indicator of the severity of the mutation. Finally, we provide guidelines for using CRISPR to generate mutations that avoid compensation.