Over the past couple of decades, zebrafish (Danio rerio) has become established as one of the foremost model systems for studying vertebrate development. High fecundity, the external development of an optically clear embryo and the genetic tractability of the zebrafish facilitate the systematic investigation of a wide variety of developmental processes, in great part through forward genetic screens. For such screens, genetically related fish (most often siblings) are bred together, and phenotypes arising in the progeny are assessed. The phenotypic and molecular analysis of zebrafish mutants identified in these screens elucidates gene function and provides an entry point for the dissection of the molecular pathways underlying vertebrate development.
A novel mutant from our pilot screen showing altered gene expression in the eye.
In the 1990’s, the Driever and Nüsslein-Volhard groups isolated several thousand mutations in hundreds of essential genes by screening for randomly induced mutations that cause visible morphological defects in embryos (Driever et al., 1996; Haffter et al., 1996). More recently, genetic screens using more complex screening techniques (e.g., in situ hybridization, immunochemistry, behavioural analysis and confocal microscopy) or involving more complex genetics (e.g., enhancer screens and maternal effect screens) are being carried out to isolate mutations that lead to more subtle and specific phenotypes.
In collaboration with several other research groups, we are undertaking a new genetic screen that incorporates an enhancer screen for mutations that lead to increased Wnt signalling. As we describe below, the screen is a little more complicated than most others that have been performed in fish as we do the screen in a line of fish carrying a mutation in a gene encoding a key player in the Wnt signal transduction cascade.
The intention of the screen is to identify mutations that:
- interact with the Wnt signalling pathway (Rodrigo Young, Wilson lab)
- disrupt brain asymmetry and/or laterality (Heather Stickney, asymmetry research pages)
- perturb eye morphogenesis (Gaia Gestri and Florencia Cavodeassi eye development research pages)
- influence liver or pancreas development (Elke Ober’s lab)
- affect neural crest and jaw development (Christiana Ruhrberg’s webpage ; Quentin’s webpage?)
- disturb lateral line development Miguel Allende’s lab
Background to why we planned this screen
A key aim for this screen is to identify mutations that lead to enhanced Wnt signaling in the developing embryo. The Wnt/Beta-catenin/Tcf pathway regulates development, regeneration, stem cell behavior and disease progression. In modulating these diverse processes, it remains unclear how the context dependent outcome of signalling is determined, which genes are critical to this process and whether their identity varies in different contexts. Our focus is to identify genes that modulate the action and output of the Wnt signalling pathway during brain and eye development. To do this, we have developed a forward genetic enhancer screen on a hdl/tcf3a mutant background, which is sensitised to small variations in Wnt signaling levels.
Large-scale synthetic enhancer screens have been extremely successful in revealing the distributed nature and complexity of genetic networks in Drosophila and C.elegans. For logistical reasons (notably the challenge of finding phenotypes in only 1/16th of progeny), the approach is not widely used in vertebrates. This is problematic as the complexity of the vertebrate genome means that studying combinations of mutations is even more important for understanding networks of gene action than it is in invertebrates. Indeed, studies of cancer and congenital conditions in humans attest that phenotypes arising from multiple gene mutations are the norm, not the exception.
The zebrafish is an ideal vertebrate model in which to perform enhancer screens as large clutch sizes facilitate identification of double mutant phenotypes. We hypothesis that because Wnt-signalling is the common denominator in a wide variety of biological processes, molecules other than the core components of the signal transduction cascade, including those that cross-talk from other signalling pathways modulate pathway output and define context dependent activity.
In designing the screen, we selected a line of fish homozygous for a mutation in the Tcf3a transcriptional effector of the Wnt pathway. Zebrafish lacking maternal and zygotic headless(hdl)/tcf3a display a loss of forebrain and eyes caused by overactivation of Wnt signalling. Due to redundancy with Tcf3b, embryos lacking only zygotic Tcf3a (zhdl) are viable with no overt phenotypes. However, zhdl mutants are sensitized to perturbations in Wnt signalling. Thus in the zhdl background, the absence of Tcf3a makes the system less robust and more vulnerable to perturbation. zhdl is therefore an excellent background on which to screen for enhancer mutations in any of the genes in the signalling networks by which the Wnt pathway regulates brain development.
Scheme of mutagenesis indicating the allelic composition of hdl/tcf3a and a putative new mutation at each step of the screen.
To conduct the enhancer screen, we induced mutations in homozygous tcf3a mutant male fish and crossed these to homozygous mutant females (Fig.2). To prevent all the fry dying, we rescued the progeny by expression of exogenous wildtype tcf3a RNA. This resulted in an F1 generation that are homozygous the headless/tcf3a mutation and heterozygous for novel mutations. We crossed these fish to a mapping line to generate F2 families. The incorporation of a mapping cross at this stage speeds up the identification of the mutated genes following identification of new phenotypes in the next generation. The screen itself is conducted in the F3 generation and for enhancers of Wnt signalling, we are looking for phenotypes present in homozygous double mutants – 1 in 16 of the progeny of sibling crosses. For the pilot screen, we analysed about 100 mutagenised genomes.
Eyeless phenotype enhancer mutations. A) wildtype. B) tcf3a+/-;enha36/a36 double mutant C) tcf3a+/-;enhg11/g11 double mutant.
For the enhancer screen, we focus on Wnt-dependent phenotypes affecting the forebrain and eyes and have already found quite a few eyeless synthetic enhancer mutations (Fig. 3,4) – this is very encouraging. In the multitude of other screens conducted throughout the community, there are only a handful of mutations known to perturb specification of the eye and our screen has already found a handful more! This gives us a lot of confidence that we will identify many new mutations that simply could not be detected on a wild-type background. We anticipate that the project will begin to elucidate key nodes in the gene networks that impact upon Wnt signalling during brain development.
The a12 enhancer mutation. A) Wildtype B) tcf3a+/-;enha12/a12 embryo with apoptotic tectum. C) Genotyped tcf3a+/-; enha12/a12 embryo with additional eye phenotype.
The screen is, of course, not limited to only finding mutations that interact with the headless/tcf3a mutation and with collaborating labs, we contemporaneously screened for many other phenotypes. For instance, we have several mutations affecting brain asymmetry including several that generate symmetric "double-left" phenotypes and others where both body and brain asymmetries are disturbed (Fig. 5).
Asymmetry phenotypes. A) Sibling B) laterality mutant with reversed expression of lov in the habenulae (arrow) and sepB in the liver (arrowhead) C) mutant displaying bilaterally symmetric habenulae and a normally positioned liver.
A critical aspect of the screen is to be able to quickly map and clone the mutations we identify. This is again proceeding very well: several mutations are now cloned and map locations have been identified for most of our other favourites from the pilot screen.
The Wellcome Trust have generously provided us with funds to undertake this project.