The Schizosaccharomyces pombe strain EF1 (CBS 356) is haploid, prototrophic, respiratory competent, and of homothallic mating type. From restriction enzyme analysis the length of the mitochondrial genome is 17.3 kilobase pairs, which is in good agreement with the value of 17.1 kilobase pairs determined by electron microscopy. The mitochondrial genome of strain EF1 is thus about 2.3 kilobase pairs shorter than that of strain ade7-50h- (about 19.4 kilobase pairs). A restriction map was constructed for 11 enzymes: For most, but not all of them, the pattern is nearly identical to that of strain ade7-50h-. The genes for the large ribosomal RNA, the subunits 1, 2, and 3 of cytochrome c oxidase, subunits 6 and 9 of ATP synthetase, and cytochrome b were localized by hybridization with mitochondrial DNA probes from Saccharomyces cerevisiae. The gene order was found to be the same in both yeast strains. From Southern hybridization of strain ade7-50h- with nick-translated mitochondrial DNA from strain EF1 it is evident that strain EF1 does not possess the intron, which is present in the cytochrome b gene of Schizosaccharomyces pombe strain ade7-50h-. Crosses between strain ade7-50h- and EF1 demonstrate that both the nuclear and the mitochondrial genomes are able to recombine. The mitochondrial genomes of 2 out of 30 independently isolated hybrids between the two strains are described as the result of recombination between the two parental mitochondrial genomes.

We report evidence for random drift of mitochondrial allele frequencies in zygote clones of Saccharomyces cerevisiae and Schizosaccharomyces pombe. Monofactorial and bifactorial crosses were done, using strains resistant or sensitive to erythromycin (alleles ER, ES), oligomycin (OR, OS), or diuron (DR, DS). The frequencies of resistant and sensitive cells (and thus the frequencies of the resistant and sensitive alleles) were determined for each of a number of clones of diploid cells arising from individual zygotes. Allele frequencies were extremely variable among these zygote clones; some clones were "uniparental," with mitochondrial alleles from only one parent present. These observations suggest random drift of the allele frequencies in the population of mitochondrial genes within an individual zygote and its diploid progeny. Drift would cease when all the cells in a clone become homoplasmic, due to segregation of the mitochondrial genomes during vegetative cell divisions. To test this, we delayed cell division (and hence segregation) for varying times by starving zygotes in order to give drift more time to operate. As predicted, delaying cell division resulted in an increase in the variance of allele frequencies among the zygote clones and an increase in the proportion of uniparental zygote clones. The changes in form of the allele frequency distributions resembled those seen during random drift in finite Mendelian populations. In bifactorial crosses, genotypes as well as individual alleles were fixed or lost in some zygote clones. However, the mean recombination frequency for a large number of clones did not increase when cell division was delayed. Several possible molecular mechanisms for intracellular random drift are discussed.

No abstract