Anecdotal , Historical and Critical Commentaries on Genetics Edited by James F . Crow and William F . Dove The Holliday Junction on Its Thirtieth Anniversary

I N the 1940s and 50s, the apparent lack of reciprocality in the production of bacteriophage recombinants led to the ascendancy of copychoice schemes for recombination in those little creatures (STURTEVANT, cited in HERSHEY and ROTMAN 1949). The subsequent descrip tion of gene conversion (non-4:4 meiotic segregation) in Neurospora (MITCHELL 1955a,b) was viewed by some as an illustration of the applicability of such schemes to meiotic exchange ( e . g . , FREESE 1957). However, to ROBIN HOLLIDAY, copychoice was not a decent explanation for meiotic conversion. He found the potential for a merent explanation in the embryonic field of DNA repair. The trick was to make recombination involve a heteroduplex, in which a marked segment of one chain carries information from one chromosome while the corresponding segment of the other chain carries information from its homolog. [The word “strand” is used by classical geneticists to denote a chromosome or a chromatid. It is used by those lacking a classical education to refer to a polynucleotide chain as defined by WATSON and CRICK (1953). We’ll avoid confusion by not using it at all.] Such locally heteroduplex products of recombination had been hypothesized to account for heterozygous particles of phage T4 ( 1954). Then, enzymes analogous to enzymes proposed to repair UV damage could recognize violations of Watson-Crick pairing at the marked site and operate on the heteroduplex, removing a bit from one chain or the other. Deviations from 4 4 segregation would (or, at least, could) result. Failure of the hypothetical mis match correction enzymes to operate on a given heteroduplex site would result in meiotic products that would segregate alleles in the first post-meiotic mitosis. The demonstrated occurrence in some fungi of such post-meiotic segregations (PMS) fully justified the assumption of heteroduplexes in meiotic recombination. In fungi, about half of the tetrads manifesting either deviations from a 4:4 ratio (conversion) or 4:4 tetrads with PMS at a given site in two of the four haploid products are reciprocally recombined for markers flanking that site (they are usually tetratype for those markers). The crossover typically involves the chromatid that is