Enzyme studies on the products of mitotic gene conversion in Saccharomyces cerevisiae.

Abstract

1. The accuracy of genetic information transfer by mitotic gene conversion between homologous chromosomes has been investigated using a diploid yeast strain heteroallelic at the locus is1. 2. The mutants defective at the locus is1 required isoleucine for growth and lacked threonine dehydratase activity. Threonine dehydratase in Saccharomyces cerevisiae catalyses the initial step in the isoleucine biosynthetic pathway, transformation of threonine to α-ketobutyrate. This enzyme was characterized in the crude extract using the following criteria: Specific activity, stability, inactivation by dilution, pH dependence, Michaelis constants, stabilization by the cofactor pyridoxalphosphate, feed back inhibition by isoleucine, activation by valine and the ratio serine over threonine dehydratase activities. 3. Altogether 23 prototrophic revertants were analysed. These were derived by gene conversion from two different inactive mutant alleles combined in a heteroallelic diploid. All convertants produced threonine dehydratase that was indistinguishable from wild type enzyme as judged by the above mentioned criteria. In contrast to these conditions found with conversional revertants, mutational revertants derived from the two alleles produced threonine dehydratase at a barely detectable level which did not allow for detailed analysis. 4. The results were interpreted to mean that gene conversion is mediated by an accurate process of information transfer between homologous chromosomes as it is predicted by the hybrid DNA model of gene conversion. 5. Mitotic gene conversion is discussed as a mechanism generating intra-individual variegation. This variegation is considered to be strictly directed in contrast to undirected variegation caused by mutation. Moreover, starting with a given pair of non-identical alleles certain conversional derivatives are expected to occur obligatorlily during ontogenesis and differentiation. Such a situation could be realized in the antibody forming systems of vertebrates where on individual synthesises many thousands of closely related but still different proteins. It is unlikely that all this variability is being coded for by the same number of nonallelic structural genes.