Nucleic Acids Research

Competition between DSB hotspots does not involve nucleosome rearrangement

To date, three types of alterations have been found to affect the frequency of DSBs at a site without changing its underlying sequence. Full activity of a DSB hotspot in the HIS4 promoter depends on the function of the Rap1, Bas1 and Grf2 transcription factors (6,18). Mutants that block binding of these proteins to the HIS4 promoter both eliminate DSBs and cause significant changes to chromatin structure at this site (6,19). Second, inversions and deletions that allow transcription from an upstream gene to impinge upon the ARG4 DSB site cause a loss of DSBs from the ARG4 promoter (15); these rearrangements are also associated with qualitative changes in chromatin structure at the DSB site. In contrast, competition between DSB sites, in which a strong DSB site is found to reduce the frequency of breaks at nearby sites (9-11), appears to occur without significantly altering chromatin structure at the affected site, as assayed with DNase I (9,20). In the current study, we found that the positions of nucleosomes in the insert-borne arg4 gene, as revealed by MNase digestion (21), are not modified by competitive inactivation. In particular, the insert-borne arg4 DSB site displayed similar patterns of MNase-hypersensitive sites in both full-length and [Delta]L-[Delta]R his4::arg4 strains, despite the >10-fold difference in DSB frequencies seen in the two constructs. These results indicate that competition between recombination hotspots is not mediated by an extensive remodeling of chromatin and that sequences within the affected DSB sites remain accessible to exogenous factors.

Competition results in a loss of meiotic induction of MNase sensitivity at affected DSB sites

This and other studies (4,12) have shown that the MNase hypersensitivity of strong DSB sites increases markedly during meiosis. This is true of the DSB sites at ARG4, CYS3 (4) and YCR47C-ARE1 (this study) and is also true of break sites within the his4::ARG4 constructs at the junction of unrelated yeast and/or pBR322 sequences. On the other hand, there is little meiotic induction of breaks or MNase hypersensitivity in the arg4 DSB site of the full-length his4::ARG4 construct (Figs 3 and 5); deletion of pBR322 sequences restores both DSBs and MNase sensitivity induction to this site. Thus, competitive inactivation of the insert-borne ARG4 site suppresses meiotic induction of both DSBs and MNase hypersensitivity, despite the fact that the underlying sequence and basal chromatin structure of this site are unaffected by the presence or absence of flanking pBR322 sequences. This result, together with the observation that induction of MNase hypersensitivity at DSB sites depends on protein functions required for DSB formation (e.g. Mre11) (12,13), lends strong support to the suggestion that the induction of MNase hypersensitivity signals a critical step in DSB formation.

What is the nature of this step? The requirement for genes necessary for DSB formation in the meiotic induction of MNase sensitivity leads us to consider the possibility that this induction reflects the binding of a multi-protein complex in preparation for DSB formation. Alternatively, the meiotic increase in MNase sensitivity might be due to activities that actively remodel chromatin or that modify nucleosomes in a process analogous to that seen in transcriptional activation (22-24), so as to make DSB sites more available to break-forming proteins. The loss of this induction under conditions of competitive inactivation might reflect the absence of either the DSB-forming complex or of the above-mentioned chromatin-modifying activities. It should be noted that the premeiotic and meiotic levels of MNase hypersensitivity at the inactivated ARG4 DSB site (in the full-length his4::arg4 insert) (Fig. 3B) was very similar to the case observed in a mre11[Delta] mutant (12). Thus, the former possibility is probably more likely.

This suggestion, in turn, raises a second question: what prevents such factors from acting on the full-length insert-borne arg4 DSB site? This site resides in a region of open chromatin, one that displays nuclease hypersensitivity in premeiotic cells. It should be pointed out that the his4::arg4 insert reduces the frequency of meiotic recombination in the LEU2 gene, located 17 kb from HIS4 (9); it also suppresses DSBs in an ~60 kb region around the insertion site (T.-C.Wu and M.Lichten, unpublished data). Thus, if competition between DSB sites is due to an absence of critical factors at affected DSB sites, the distribution of these factors may be controlled at the level of chromosomal or nuclear domains. Elucidation of the mechanism of DSB competition should reveal important aspects of meiotic chromosome and/or nuclear structure.

Chromatin structure at the junction of yeast and foreign DNA and at the breakpoints of chromosomal rearrangements

Chromatin at the junctions of pBR322 and yeast sequences in the full-length his4::ARG4 insert is hypersensitive to both DNase I (9) and MNase (present data) and thus forms preferential sites for meiotic DSBs. Others have shown that insertions of bacterial DNA or of yeast telomere sequences can create nuclease-hypersensitive sites that are hotspots for meiotic recombination (6,18,25). Such unusually accessible regions are not specific to bacterial sequences, since open sites are also created at the junction between his4 and URA3 yeast sequences in the full-length insert and at the junction between URA3 and ARG4 sequences in the [Delta]L-[Delta]R insert. We do not know how such discontinuities in chromosomal context create nuclease-hypersensitive sites in chromatin. One possibility is that nucleosomes are displaced at the break point by a collision between phased nucleosomes directed by different chromosomal sequences. Another is that these joints disrupt an arrangement of sequences, established over the course of evolution of the genome, that direct the assembly of nucleosomes into regular arrays.

Irrespective of the mechanism, the formation of nuclease-hypersensitive sites at the junction of juxtaposed sequences may act, over time, to preserve the normal arrangement of the genome. If rearrangements create strong meiotic DSB sites (6,9,20), these novel junctions could lead to chromosome loss due to unrepaired breaks. They might also serve as recipients of genetic information during meiotic gene conversion (26,27). This, in turn, should lead to the preferential loss of these sites over time, as they are replaced by normal chromosomal sequences (28). In the natural yeast populations, such juxtapositions may be brought about by unequal sister chromatid exchange, chromosome breakage and rejoining and insertion of movable genetic elements. All are large scale chromosomal rearrangements that are expected to be harmful in most cases. The formation of open chromatin at these novel junctions, in creating meiotic DSB sites, would allow meiotic recombination to participate in the removal or reduction of such chromosomal rearrangements from the population (29).

ACKNOWLEDGEMENTS

We thank K. Kobayashi and S. Okamoto for their assistance with these experiments. This work was supported by a research grant from the Human Frontier Science Program (RG493/95), a grant for the ‘Biodesign Research Program’ from RIKEN, CREST of the JST (Japan Science and Technology) and grants from the Ministry of Education, Science and Culture, Japan.