One complication of the breeding scheme for diploid organisms outlined in Figure 3-2.2a in the previous section is that, in accordance with Mendel's Laws (see Chapter 1), only one in two of the progeny of the F1 mutant heterozygote will inherit the mutant allele. This means that in intercrosses between randomly selected male and female siblings in the F2 generation, only one in four (1/2 × 1/2) crosses will be between individuals that are both heterozygous for the newly induced mutant allele. Thus multiple different pairs of F2 siblings must be mated to ensure that at least one pair will generate homozygous progeny. The probability, P, that a cross will be established between two parents that are both heterozygous for the same gene is given by the equation P = (1 - 0.75ⁿ), where n is the number of crosses made. For example, to be 90% sure of detecting a mutation, eight different pair matings must be made among the F2 progeny of the mutagenized individual.

This limitation could be avoided if one could follow the transmission of the mutagenized chromosomes directly down the generations. Then, individuals heterozygous for a mutagenized chromosome could be identified directly in the F2 generation. In principle, this can be accomplished if the chromosome in the mutagenized parent carries a preexisting dominant mutation that gives a visible phenotype, such as altered pigmentation. This would allow the transmission of that chromosome to be followed from one generation to the next simply by following the inheritance of the abnormal pigmentation trait (Figure 3-3.1). Only one marked chromosome at a time can be followed in each set of screens, so a full mutant screen of the whole genome requires a separate set of breeding experiments for each chromosome.

However, this general strategy makes the major assumption that no recombination of the marked chromosome occurs during meiosis. If it does, the mutagenized chromosome will not be transmitted intact from one generation to the next and the marker mutation is highly likely to become separated from any newly induced mutation on the original chromosome (see Figure 3-3.1). To prevent this happening, special chromosomes that suppress recombination are used in some organisms, especially in Drosophila. These chromosomes – known as balancer chromosomes – are typically highly rearranged due to multiple inversions, a structure that suppresses recombination by inhibiting pairing at meiosis. Moreover, any recombination event that does occur is likely to generate an abnormal chromosome with deleted and duplicated regions, resulting in an inviable gamete. By using a balancer chromosome that also carries a dominant genetic marker, so that progeny carrying both the mutagenized chromosome and the balancer are easily identifiable, the transmission of the intact mutagenized chromosomes can easily be followed through successive generations, removing the need to set up random matings of multiple pairs in the F2 generation (Figure 3-3.2). Another useful feature of Drosophila is the absence of recombination in males; some types of genetic analysis can be simplified by picking only males for future breeding.

Up to now we have discussed the general principles of mutant screening, which apply whether one is looking for mutants in general or for mutations affecting a particular process. In practice, the latter type of screen is much more commonly carried out. In the ideal screen, mutant individuals should be detectable by simple non-invasive inspection of their morphology or behavior: in many cases, this can be achieved by taking advantage of some indirect consequence, or surrogate, of the primary defect of interest. For example, in the case of the yeast wee1 mutation mentioned in section 3-0, the objective of the original experiment was to identify mutations affecting the entry of cells into mitosis during the cell-division cycle. Direct detection of such mutations would require a comparison of the doubling times of mutant and wild-type cell populations or the direct visualization of cytokinesis in individual cells, neither of which can easily be achieved on the scale necessary to detect rare events. However, premature entry into mitosis causes cells to divide before they have grown sufficiently, thereby giving rise to abnormally small daughter cells. Thus by using cell size as a surrogate phenotype – one that can be readily scored in thousands of individual cells – it is possible to select for the desired mutations, which can then be investigated in more detail.

The same principle is exemplified by the isolation of the sevenless mutation in the fruit fly Drosophila. This mutation identifies a gene specifically required for the differentiation of a single photoreceptor cell in each of the clusters of photoreceptors that make up the compound eye of the fly (Figure 3-3.3). Direct detection of such a subtle defect would necessitate the detailed analysis of histological sections of eyes from individual flies. However, the R7 cell is specifically required for flies to detect UV light, to which they are preferentially attracted. Thus it is possible instead simply to screen for mutant flies that fail to walk towards a UV light source, thereby selecting mutants in which the differentiation of the R7 cell is compromised.

Of course, it may not always be possible, or indeed necessary, to devise such indirect assays for gene function. The advent of sophisticated molecular technology has made it possible to screen directly for defects in specific tissues and we will discuss these techniques in detail in Chapter 5.

As discussed in section 3-2, sampling the same newly induced mutation many times can be avoided by ensuring that it arose in a mature gamete rather than in a progenitor or stem cell. However, preexisting mutations that are present in the genetic background of mutagenized individuals will be transmitted multiple times to their F1 progeny. This potential complication can be avoided by rendering the chromosomes of the mutagenized individuals isogenic. An isogenic chromosome is one in which both alleles at every locus are identical on both copies of the chromosome. Thus a phenotypically normal isogenic individual by definition lacks any preexisting mutant alleles.

Making an individual isogenic for all its chromosomes is not a trivial task but can be approached by repeated inbreeding of individuals over successive generations. In organisms where balancer chromosomes are available, isogenizing an individual chromosome is relatively straightforward: males and females heterozygous for a genetically identical chromosome over a balancer chromosome are derived from a single individual and interbred. Because the balancer chromosome suppresses recombination, the same unrecombined chromosome will be inherited from both parents, generating animals homozygous for the intact chromosome. This process can be repeated until all the chromosomes have been isogenized.

Using isogenic chromosomes in mutant screens is also important when it comes to the molecular analysis of mutations, as it eliminates molecular polymorphisms in the DNA that can confound the identification of mutational change. Consider for instance, a neutral polymorphism in DNA sequence present at low frequency in the mutagenized population and that happens by chance to reside in the gene inactivated by the newly induced mutation. Because the polymorphism is neutral it will show no phenotype in the parental population and its existence will be quite unsuspected. Individuals carrying both a new mutation and the low-frequency variant will show a mutant phenotype, but when the mutant allele is sequenced and compared to the majority wild-type allele, it will be impossible to tell without further genetic experiments which sequence difference is the polymorphism and which the new mutation.

balancer chromosome: a chromosome with multiple inversions and which therefore cannot undergo recombination with its homolog.

isogenic chromosome: in a diploid organism, a chromosome in which both alleles at every locus are identical on both copies.