The hypothesis that chromosome size affects the rate and distribution of meiotic crossovers in budding yeast was tested. Map distance and interference were measured in the same genetic intervals present on either small (340 and 508 kb) or large (917 and 1085 kb) chromosomes, respectively. No differences were observed.
DURING meiosis, crossovers between homologous chromosomes, in combination with sister-chromatid cohesion, create physical connections that promote the accurate segregation of chromosomes to opposite poles at the first meiotic division (Bascom-Slack et al. 1997; Petronczki et al. 2003). Genetic interference is a phenomenon by which a crossover in one interval decreases the probability that additional crossovers will occur nearby. The distribution of crossovers by interference has been postulated to ensure that every pair of homologs receives at least one. The degree of interference and the number of crossovers per meiosis varies between organisms. For example, in nematodes interference is strong, such that each chromosome pair undergoes only a single crossover (Hillers and Villeneuve 2003). In contrast, in fission yeast there is no interference, but because there is a large number of crossovers and only three chromosomes, the probability that each chromosome will receive a crossover is high (Munz 1994). In budding yeast, there are ∼90 crossovers per meiosis (Mortimer et al. 1991). There are 16 pairs of homologs and these chromosomes vary in size from 230 to 1530 kb (Saccharomyces Genome Database). Mutations that abolish interference increase nondisjunction of small chromosomes preferentially, consistent with the idea that crossovers are distributed to ensure that all chromosomes receive at least one (Sym and Roeder 1994). Studies in the literature using bisected and translocated chromosomes suggest that chromosome size is important in determining the rate of recombination (in centimorgans per kilobase) as well as the degree of interference (Kaback et al. 1992, 1999). We sought to test this hypothesis by comparing map distance and interference values obtained from the same genetic intervals present on either small or large chromosomes. Contrary to what has been previously reported, we saw no effect of chromosome size on either of these parameters. We conclude, therefore, that differences in the rate of recombination and interference observed between chromosomes in budding yeast, at least for some intervals, are a function of their DNA sequence as opposed to the size of the chromosome.
To allow direct comparison between the same DNA sequences on different-size chromosomes, ectopic recombination was used to make a reciprocal translocation that exchanges the right arm of chromosome VII for the last ∼8 kb of the right arm of chromosome III and vice versa, using the approach described in Borde et al. (2000)(Figure 1, A and B). Recombination between ade2 truncation alleles on chromosomes III and VII generated Ade+ cells that were screened by polymerase chain reaction (PCR) and Southern blot analysis to confirm the presence of the translocation (Figure 1B). Three intervals on chromosomes III and VII were then analyzed on either a short chromosome (340 and 508 kb, respectively) or a long chromosome (917 and 1085, respectively). The isogenic diploids, NH598 (which contains the native chromosomes) and NH607 (which contains the translocation chromosomes), were sporulated and dissected. Over 1000 tetrads were analyzed for each diploid. The map distance of each interval was unaffected by the change in chromosome size (Table 1). Gene conversion was also unaffected (data not shown). To monitor interference, the ratio of observed nonparental ditypes (or NPDs, which arise from four-strand double crossovers) to those expected assuming no interference, was calculated. An NPD ratio of 1 indicates no interference, while an NPD ratio of 0 indicates complete interference (Snow 1979). Although differences in NPD ratio were observed, these differences are not likely to be significant as the standard deviations of all the intervals overlap, with the exception of the LEU2-THR4 interval (where at least 48 NPDs were scored for each diploid; Table 1). In the LEU2-THR4 interval, the NPD ratio is increased in the translocation diploid, in contradiction to what was expected from the literature. The prediction was that moving this interval from a small chromosome to a large chromosome should increase the amount of interference (i.e., decrease the NPD ratio).
Our experiments unambiguously demonstrate that changes in chromosome size, by either an increase of 2.6-fold or a decrease of 2.1-fold, have no effect on the rate of recombination or the degree of interference for specific intervals on chromosomes III and VII. Why do these results differ from those in the literature? One possible explanation is that intervals on different chromosomes are differentially sensitive to changes in chromosome size. In the work by Kaback et al. (1992)(1999), recombination was measured using translocations between chromosomes I and II, while this work used translocations between chromosomes III and VII. Alternatively, there may be genetic differences between the SK1 strains used in this study and the non-SK1 strains used by the Kaback lab that are responsible. Whatever the reason, the main finding of this article is that the phenomenon of chromosome size affecting recombination rates and interference is not a general one.
Different intervals in our experiment clearly exhibited different rates of recombination as measured by centimorgans per kilobase. The highest rates were observed for the chromosome III intervals MAT-THR4 and HIS4-LEU2, irrespective of chromosome size (Table 2). Interestingly, chromosome III also contained the interval that exhibited the lowest rate of recombination. Not surprisingly, the rate of recombination correlates well with the frequency of double-strand breaks (DSBs) formed in or near these intervals (Baudat and Nicolas 1997; Gerton et al. 2000; Borde et al. 2004). Whereas chromosome III has well-defined hotspots for DSB formation, the breaks on chromosome VII are more evenly distributed (Gerton et al. 2000; Borde et al. 2004). Therefore, differences in recombination rates between chromosomes are likely due to DNA sequences that lead to different chromosomal contexts that influence the frequency of DSB formation, rather than to chromosome size.
Many thanks go to Neil Hunter and Michael Lichten for strains, plasmids, and ideas. Rebecca Cole assisted in the tetrad dissection. We are grateful to Bernadette Holdener for use of her CHEF gel apparatus. This work was supported by National Institutes of Health grant GM-50717.
↵ 1 Present address: USDA, ARS, NAA, Plum Island Animal Disease Center, P.O. Box 848, Greenport, NY 11944.
Communicating editor: M. Lichten
- Received July 15, 2004.
- Accepted August 24, 2004.
- Genetics Society of America