How does the cell count the number of ectopic copies of a gene in the premeiotic inactivation process acting in Ascobolus immersus?

Repeated genes, artificially introduced in Ascobolus immersus by integrative transformation, are frequently inactivated during the sexual phase. Inactivation is observed in about 50% of meioses if duplicated genes are at ectopic chromosomal locations, and in 90% of meioses if genes are tandemly repeated. Inactivation is associated with extensive methylation of the cytosine residues of the duplicated sequences and is induced in the still haploid nuclei of the dikaryotic cell which will undergo karyogamy and subsequent meiosis. Only repeated sequences become methylated. This raises the intriguing question of how the premeiotic inactivation machinery is informed that a nucleus contains multiple copies of a gene. By using in crosses recombinant strains of A. immersus in which either one, two or three genetically independent copies of the exogenous amdS gene from Aspergillus nidulans had been introduced, we could follow the premeiotic inactivation of each one of the ectopic amdS copies. This led us to propose that a prerequisite for inactivation is a premeiotic pairing of repeated sequences and that each copy can undergo successive cycles of pairing. In fact, once methylated, a copy can pair with a still unmethylated copy, so that an uneven number of copies can be subject to inactivation.