Meiotic gene conversion in yeast tetrads and the theory of recombination.

ONVENTIONAL recombination theory has held that genetic recombination occurs after DNA replication, when each chromosome is composed of two chromatids; that crossing over involves a full-chromatid interaction between nonsister chromatids, and yields reciprocal products from breakage reunion events; and that either no crossing over occurs between sister chromatids, or if it occurs it is unrelated to homologous exchanges. Experience with this theory has shown that in a chromosome arm the occurrence of one exchange interferes with another exchange in an adjacent region. This has been manifested by detecting significantly fewer double crossovers than would be predicted on the assumption that individual exchanges occur independently of each other. Furthermore, among the multiple exchanges that do occur no evidence for chromatid interference has been found; two-, three-, and four-strand double exchanges are typically observed to occur in the expected 1:2: 1 ratio (PERKINS 1962). This theory has had the consequence that multiply recombined progeny derived from random sperm or random spores have been interpreted in terms of multiple exchanges between the genetic markers, even when the reciprocal products themselves have not been available for analysis. However, recombination need not be a reciprocal process. The initial report by MITCHELL (1955) of nonreciprocal recombination (gene conversion) was followed by others (reviewed by WHITEHOUSE and HASTINGS 1965) from which the following has become clear. (1 ) Recombination within a cistron may be nonrecip-