Meiotic Recombination in Yeast: Coronation of the Double-Strand-Break Repair Model

ciprocal exchanges (conversions) that are the result of Meiosis reduces the chromosome complement of a euDNA lost from one parent in the neighborhood of the karyote from diploidy to haploidy. It is typically marked DSB site and then compensated for by synthesis temby crossing over, the reciprocal exchange of segments plated on the uncut homologue, as called for by the of DNA between homologous chromosomes. In many DSBR model (Figure 1). creatures, including yeast, the orderly distribution of Second, in the DSBR model, the two duplexes particichromosomes through meiosis is dependent upon this pating in the joint molecule are held together by a pair crossing over. Crossing over certainly contributes to of Holliday junctions (Figure 1c). In a Holliday junction, genetic variability as well, and that may be an important strands of the same polarity are swapped between dufactor in its maintenance. Molecular models for genetic plexes. Schwacha and Kleckner showed that their joint exchange became possible only with the elucidation of molecules can be dissociated into duplexes, some the double-stranded structure of DNA by Watson and crossed over for the distant markers and some not, Crick in 1953. Many of the daunting data on meiotic by the in vitro action of an enzyme, Holliday junction recombination were rationalized by Robin Holliday’s resolvase, isolated from E. coli. The specificity of this germinal proposal of 1964. enzyme makes it likely that the joint molecules are, inIn Holliday’s proposal, the initial event in meiotic redeed, held together by Holliday junctions. Furthermore, combination was presumed to be nicks in one of the the DSBR model proposed that alternate patterns of two strands of each of the participating DNA duplexes. resolution of these junctions could give either crossover In that of Meselson and Radding, which addressed some or noncrossover products (Figure 1d). It follows from shortcomings in Holliday’s model, the initiating event the parental nature of the flanking marker arrangements was a nick in one strand of one of the participants. In in the joint molecules that there is an even number of 1983, Szostak et al. proposed that meiotic recombinasuch junctions and that the same two strands are tion was initiated by the breakage of both strands of swapped at each junction, as called for by the model one of the two participating duplexes, as did Resnick (Figure 1c). The simple view is that there are exactly (1976) a few years earlier.Subsequent experimental suptwo such junctions in each joint molecule, exactly as port for this radical proposal of an apparently foolhardy postulated by the model. mechanism has established the Double-Strand-Break The paper of Schwacha and Kleckner puts the DSBR Repair (DSBR) model for recombination (Figure 1) as model on more secure footing than has been enjoyed the dominant paradigm. This support included direct, by any preceding model for meiotic recombination. At physical detection of meiosis-specific double-strand an important level, the problemof meiotic recombination breaks, the correlation of these breaks with initiation in yeast appears to have been solved. However, having sites for recombination, and the demonstration of resecsolved the problem by vindicating the model, the aution of the 59-ended strands on each side of a break to thors, in the Discussion in the same paper, challenge a create 39-ended overhangs that are about 600 bases feature of that model. long. However, transition structures between the stage The model (Figure 1c) predicts segments of biparental of resected ends and completed recombinant molecules DNA (heteroduplex DNA) in the joint molecule on the remained undetected. two sides of the original DSB. If the two parents have In 1994, two groups (Schwacha and Kleckner, 1994; appropriately marked DNA, the heteroduplex nature of Collins and Newlon, 1994) described “joint molecules” these segments should be detectable by restriction enisolated from meiotic cells. Each joint molecule conzyme analysis designed to detect the predicted mistained four continuous single strands, two from each matches. The Discussion in the paper by Schwacha and parent, none of which was recombinant for markers that Kleckner (1995) focuses primarily on their failure to find flanked the initiation site. These features of the joint heteroduplex DNA in the joint molecules described in molecules were precisely those of the recombination their 1994 paper — no mismatches were found in douintermediate of the 1983 DSBR proposal. Nevertheless, ble-stranded DNA isolated by melting apart joint molethose very features weretaken by someof theauthors as cules, even though the method used did detect missuggesting that the participating duplexes were merely matches arising later, when fully formed recombinant “kissing” and had not yet gotten down to the serimolecules appeared. They reconcile these data by proous business of genetic exchange. Recently, however, posing that heteroduplexes arise not at the time of joint Schwacha and Kleckner (1995) have demonstrated that molecule formation but during their resolution. This is the joint molecules are truly recombinational intermediin conflict with the details of the DSBR model as preates and that they have two additional key properties sented in Figure 1. Schwacha and Kleckner support their demanded of the bimolecular intermediate of the DSBR position by citing work of Goyen and Lichten (1993) that