Perspective Mitotic Recombination: Why? When? How? Where?

Matthew C. LaFave1,2, Jeff Sekelsky1,2,3* 1 Curriculum in Genetics and Molecular Biology, University of North Carolina, Chapel Hill, North Carolina, United States of America, 2 Cell and Molecular Biology Training Program, University of North Carolina, Chapel Hill, North Carolina, United States of America, 3 Department of Biology, University of North Carolina, Chapel Hill, North Carolina, United States of America

DNA damage repair, loss of heterozy- RCO. Since there is also evidence that suggests that a considerable fraction of the gosity, and chromosome rearrangement single-stranded nicks and gaps are recom- breaks that results in RCOs occur before are important aspects of genome stability, binogenic [5], it is likely that several types replication. and all are tied to mitotic recombination. of DNA lesions may be important for Despite the importance of mitotic recom- spontaneous mitotic recombination events. How? bination, the most basic questions about In addition, some recombinogenic agents this process remain poorly understood. (such as ultraviolet radiation) are thought What is the molecular mechanism by This is in part because mitotic recombi- to produce nicks that result in DSBs when which mitotic recombination is accom- nation, in contrast to meiotic recombina- the nicked DNA is replicated [6]. Thus, plished? The results presented by Lee et al. tion, is rare on a per cell division basis [1]. the question of why becomes tied up with suggest that RCOs with different gene A number of systems have been devised to the question of when. conversion tract lengths may be produced detect or select for mitotic recombination. by different mechanisms. Short tracts may In this issue of PLoS Genetics, Lee et al. [2] When? result from a DSB repair pathway involv- describe a novel system that represents a ing heteroduplex formation followed by major step forward in the study of At what point in the cell cycle does mismatch repair [1]. A heteroduplex is a spontaneous mitotic recombination events. mitotic recombination occur? Whereas region of DNA composed of strands that Their studies have given us new insights meiotic recombination occurs during mei- are derived from two different chromo- into the why, when, how, and where of osis, most mitotic recombination probably somes. Polymorphisms between the two mitotic recombination. does not occur during , but during chromosomes will result in mismatches, Mitotic recombination was first de- interphase. Analysis of gene conversion and repair of these mismatches can result scribed by Stern in his classic Drosophila tracts associated with RCOs provides clues in gene conversion. Although this mecha- experiments [3]. For Stern, ‘‘recombina- about when during interphase mitotic nism has been proven to be important for tion’’ referred only to reciprocal crossovers recombination takes place. Gene conver- meiotic gene conversion, in which the (RCOs) (Figure 1A). A severe limitation of sion is a nonreciprocal exchange of genetic conversion tracts are usually 1–2 kb long, most RCO assays is that only one of the information. Normal gene conversion evidence that it can produce the very long two reciprocal products can be recovered. between homologous chromosomes pro- conversion tracts (average of 12 kb) ob- Barbera and Petes [4] devised a clever duces a 3:1 ratio of alleles (Figure 1C); served by Lee et al. is lacking. method to recover both products of RCOs however, Lee et al. also detected 4:0 and Lee et al. argue that the very large in . They used this 3:1/4:0 hybrid tracts (Figure 1D). Lee et conversion tracts (some up to 100 kb in method to measure rates of spontaneous al. argue that a 4:0 tract most likely results length) may reflect a different process, gap and induced mitotic recombination. Lee et when a break occurs prior to DNA repair [7]. If a DSB is processed to form a al. have brought increased power to this replication, but repair takes place after gap, or if two DSBs occur on the same assay by performing it in diploids with replication. As depicted in Figure 8 of Lee chromatid, the potential for extensive ,0.5% heterology between the sequences et al., replication of a broken chromatid heteroduplex formation is eliminated. of homologous chromosomes. This design results in sister chromatids that are both Instead, the entire gap is filled using the allowed mapping of RCOs at high resolu- broken at the same position. Since both homologous chromosome as a template, tion, and also allowed study of another are broken, the homologous chromosome resulting in a long patch of gene conver- aspect of recombination—gene conversion must be used as a repair template. If both sion. If these long tracts are indeed (Figure 1C and 1D). Their analysis led to broken chromatids repair off the homolo- produced by gap repair, it raises the several key findings that provide unique gous chromosome, a 4:0 or 4:0/3:1 hybrid question of whether the RCOs associated and sometimes surprising insights into is produced, depending on whether or not with these tracts arise from a double- questions about mitotic recombination. both tracts are identical. The high fre- (DHJ) intermediate. quency of 4:0 and 3:1/4:0 hybrid tracts Meiotic crossovers in S. cerevisiae involve a Why? Why does mitotic recombination, which Citation: LaFave MC, Sekelsky J (2009) Mitotic Recombination: Why? When? How? Where? PLoS Genet 5(3): can be detrimental, occur? Answering this e1000411. doi:10.1371/journal.pgen.1000411 question begins with asking what initiates Editor: Gregory P. Copenhaver, The University of North Carolina at Chapel Hill, United States of America the process. Lee et al. suggest that most Published March 13, 2009 spontaneous RCOs are initiated by DNA Copyright: ß 2009 LaFave, Sekelsky. This is an open-access article distributed under the terms of the double-strand breaks (DSBs). Recombina- Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any tional repair of a DSB requires a template; medium, provided the original author and source are credited. when the homologous chromosome serves Competing Interests: The authors have declared that no competing interests exist. that role, it provides the opportunity for an * E-mail: [email protected]

PLoS Genetics | www.plosgenetics.org 1 March 2009 | Volume 5 | Issue 3 | e1000411 Figure 1. Reciprocal crossovers and gene conversion. (A) An RCO is depicted between chromatids of two homologous chromosomes. One segregation pattern results in daughter cells that have become homozygous for the sequence distal to the crossover site. (B–D) A close-up view of the region outlined by the dotted box, showing different gene conversion tract configurations detectable using markers a through d. (B) No conversion tract, either because there was no gene conversion or the tract was too small to be detected with the markers available. All markers are still present in a 2:2 ratio. (C) A typical gene conversion event produces a tract that alters some of the markers (b and c) to a 3:1 ratio. Note that conversion tracts can only be detected if both reciprocal products (i.e., both daughter cells) are recovered and analyzed, as done by Lee et al. (D) Lee et al. observed some tracts that were wholly or partially 4:0. In the example shown here, marker b has segregated 4:0, but marker c has segregated 3:1; this is therefore a 4:0/3:1 hybrid gene conversion tract. doi:10.1371/journal.pgen.1000411.g001

DHJ intermediate [8], but it is unknown Lee et al. allows high-resolution molecular assay used by Lee et al., such as focusing whether this structure can be produced mapping of hotspots for mitotic recombi- on other regions of the genome or across a long gap. nation. They found that sites of RCOs, incorporating DNA repair mutants, will and therefore the initial sites of spontane- surely aid in future studies. Where? ous damage, were nonrandomly distribut- Many questions concerning mitotic ed. Furthermore, the authors uncovered recombination remain to be answered. Are there hotspots for mitotic recombi- evidence for the existence of one region Perhaps the most basic of these is what nation as there are for meiotic recombi- with elevated RCOs. This is exceptionally makes certain regions more prone to nation? It is believed that some sites are interesting because it represents a region mitotic breakage and recombination than hotter for DSB formation than others. prone to spontaneous rather than induced others? The approach of Lee et al. can Common fragile sites (CFSs), regions of DSBs. The fact that such a hotspot could address this question—and others—in an the genome prone to chromosomal DSBs, be detected by examining only 1% of the exciting new way by focusing on regions are a normal feature of mammalian genome makes this discovery more in- prone to spontaneous damage, as opposed chromosomes, and analogous regions have triguing still. It will be exciting to see if been identified in yeast [9]. Most studies of to induced damage. These and other CFSs have relied on the use of replication other such hotspots for spontaneous dam- studies of mitotic recombination, a process inhibitors to increase the frequency of age and mitotic recombination exist, and both fundamental and far reaching, prom- breaks, followed by cytological detection what role they play in genome stability. ise to continue to provide interesting [10]. In contrast, the approach taken by Taking advantage of the flexibility of the insights into causes of genome instability.

References 1. Paques F, Haber JE (1999) Multiple pathways of overs in Saccharomyces cerevisiae. Proc Natl Acad 8. Schwacha A, Kleckner N (1995) Identification of double-strand break-induced recombination in Sci U S A 103: 12819–12824. double Holliday junctions as intermediates in Saccharomyces cerevisiae. Microbiol Mol Biol Rev 5. Lettier G, Feng Q, de Mayolo AA, Erdeniz N, meiotic recombination. Cell 83: 783–791. 63: 349–404. Reid RJD, et al. (2006) The role of DNA double- 9. Lemoine FJ, Degtyareva NP, Lobachev K, 2. Lee PS, Greenwell PW, Dominska M, Gawel M, strand breaks in spontaneous homologous recom- Petes TD (2005) Chromosomal translocations in Hamilton M, et al. (2009) A fine-structure map of bination in S. cerevisiae. PLoS Genet 2(11): e194. yeast induced by low levels of DNA polymerase: a spontaneous mitotic crossovers in the yeast doi:10.1371/journal.pgen.0020194. model for chromosome fragile sites. Cell 120: Saccharomyces cerevisiae. PLoS Genet 5(3): 6. Galli A, Schiestl RH (1999) Cell division 587–598. e1000410. doi:10.1371/journal.pgen.1000410. transforms mutagenic lesions into deletion-re- 10. Glover TW, Berger C, Coyle J, Echo B (1984) 3. Stern C (1936) Somatic crossing over and combinogenic lesions in yeast cells. Mutat Res DNA polymerase alpha inhibition by aphidicolin segregation in . Genetics 429: 13–26. induces gaps and breaks at common fragile sites 21: 625–730. 7. Szostak JW, Orr-Weaver T, Rothstein RJ, in human chromosomes. Hum Genet 67: 4. Barbera MA, Petes TD (2006) Selection and Stahl FW (1983) The double-strand break repair 136–142. analysis of spontaneous reciprocal mitotic cross- model for recombination. Cell 33: 25–35.

PLoS Genetics | www.plosgenetics.org 2 March 2009 | Volume 5 | Issue 3 | e1000411