Mapping

Several tools are available for mapping and positional cloning in the zebrafish. These include a dense map of sequence-specific length polymorphisms (SSLP), an increasing number of expressed sequence tags (ESTs) mapped on genetic and radiation hybrid panels, and a variety of genomic libraries. Furthermore, the zebrafish genome has been sequenced in its entirety and a high degree of synteny to the human and pufferfish (Fugu rubripes) genomes has become apparent. These reagents make the zebrafish an extremely versatile genetic system for the generation of mutant phenotypes and cloning of the corresponding genes.

For mapping of our mutants we employ the following strategy:
DNA from 40 mutant and 40 wildtype embryos is distributed into 4 pools of 20 each. Using a collection of 239 CA markers (approximately 10 per chromosome), we perform bulk segregant analysis (see Fig. 1A,B). This can be done by hand (956 PCR reactions or ten 96-well plates) or by using a Tecan robot, if available. Linkage is presumed, when both mutant pools show only one amplicon, while the two wildtype pools show an additional (presumably wildtype) amplicon (see Fig.1B). Linkage is then confirmed by testing the linked markers on a panel of 88 mutants with 8 wildtype controls (Fig.1C). The number of recombinants identified allows a rough estimate of the tested marker from the mutant locus (See Fig. 1D). The distance in the case depicted in Figure 1C is: 1/87 = 0.01149 i.e. approximately 1.15cM (one cM in zebrafish is on average approximately 600kB). Markers North and South of the mutant locus are established (“flanking markers”) and the closest ones are subsequently used to test a panel of 1,500 mutants (fine mapping). This number of mutants should allow for a recombinant resolution of 30-40kB, well within the limits of a single BAC.

Mapping mutations can also be done by EP. This methods uses gynogenetic diploid animals and tests mutant and wildtype pools (see above) with centromeric CA markers. The rationale behind this method is that there is no recombination between the mutant locus and the centromeric marker on either of the duplicated chromosomes (otherwise the individual would be a heterozygote). Therefore, if the mutant pool shows one amplicon only, and the wildtypes another one (this happens with a centromeric mutation, see Figure 2, CA1), or an additional amplicon, the mutation is linked to the chromosome corresponding to the centromeric marker giving this pattern. Figure 2 also shows an unlinked marker (CA2). One advantage of this method is that the distance of the mutant locus from the centromere can be calculated by counting the number of mutant and the total number of embryos: distance in cM = 50 (1 - (2 x mutant number/total number of embryos)). This facilitates the search for linked CA markers. Cave: EP mapping can only be done in the first mapcross generation!

If no CA markers are available that are close enough to the region of the mutant locus to start a walk, single strand conformational polymorphism (SSCP) can be used to identify recombinants which help to orient the walk (Fig. 3). In this example, the wildtype (WT) genotype is characterized by three additional SSCP complexes (1-3), which identifies one mutant (mut) individual as a recombinant (rec). Alternatively, PCR products for a given genomic region (200-300bp) from wildtype, mutant and heterozygous individuals can be resolved by melting curve analysis (MCA), based on the difference in melting temperature. temperature conferred by the base change (SNP or actual mutation). Figure 4 shows an example of the capacity of MCA to distinguish individuals from a p53 +/- incross.