Tid1/Rdh54 promotes dissociation of Dmc1 from nonrecombinogenic sites on meiotic chromatin

Abstract

The meiosis-specific recombinase Dmc1 plays a critical role in DNA strand exchange in budding yeast. Tid1/Rdh54, a member of the Swi2/Snf2 family of DNA translocases, has been shown to stimulate Dmc1-dependent recombination. Tid1and its budding yeast paralog Rad54 have a variety of biochemical activities that may contribute to their biological function. Here we demonstrate that Dmc1 can associate with chromatin in the absence of DNA double-strand breaks (DSBs), and Tid1 suppresses this association. Chromatin immunoprecipitation experiments indicate that an activity shared by Tid1 and Rad54 is required for normal assembly of Dmc1 at DSB sites in preparation for recombination. These results lead to a model in which the ATP hydrolysis-dependent DNA translocase activity of Tid1 acts to promote dissociation of Dmc1 from nonreombinogenic sites on chromatin, with Rad54 being able to substitute for this function in the absence of Tid1. The tendency of Dmc1 to form unproductive interactions with chromatin is proposed to be a consequence of the mechanism of strand exchange. The results raise the possibility that ATP hydrolysis-dependent disruption of nonproductive recombinase-DNA interactions is a feature shared with other homologous recombination systems.

Figure 1.

TID1 blocks Dmc1 focus formation in the absence of DSBs. (A) Examples of typical representatives of Dmc1 staining patterns. (Left) Wild -type nucleus displaying brightly staining Dmc1 foci. (Second from left) Background staining with anti-Dmc1 as observed in a dmc1Δ nucleus. (Second from right) spo11Δ nucleus displaying “faint” Dmc1 staining. (Right) spo11Δ tid1Δ nucleus showing the type of Dmc1 staining pattern scored as “patchy.” Note the large number of faintly staining foci combined with a predominant staining structure. (B) Fifty unselected nuclei from wild-type, tid1Δ, spo11Δ, spo11Δ tid1Δ, and dmc1Δ strains were scored with respect to having a Dmc1 staining pattern that was “bright,” “faint,” “patchy,” or “none” as described in the text.

Figure 2.

Dmc1 foci in spo11Δ tid1Δ do not result from elevated Dmc1 levels, SPO11-independent chromosome breaks, or SPO11-independent ssDNA tracts. (A) Western blot analysis of Dmc1 protein levels from wild-type, tid1Δ, spo11Δ, and spo11Δ tid1Δ compared with Arp7 loading control. (B) Analysis of the integrity of meiotic chromosomes during a meiotic time course from wild-type, tid1Δ, spo11Δ, and spo11Δ tid1Δ strains by CHEF gel electrophoresis (per strain from left to right: 0, 3.5, 4, 6, and 8 h in meiosis). Note that in wild type, formation of DSBs at 3.5 h is evident, followed by restoration of chromatin integrity at 8 h. spo11Δ and spo11Δ tid1Δ mutants do not show loss of chromosome integrity. (C) Three representative spread nuclei prepared from cells collected 4 h after transfer to sporulation medium. The relevant genotype is indicated. Dmc1 foci are shown in green, and RPA foci are shown in red.

Figure 3.

Dmc1 ChIP. The average hotspot:coldspot ratios corrected to input DNA are shown for each strain (n = 3). (A) Preferential association of Dmc1 at the HIS4::LEU2 hotspot depends on SPO11. Dmc1 enrichment was not detected in control experiments without cross-linking and without antibody (data not shown). Hotspot and coldspot signal is not detectable in strains lacking DMC1. Dmc1 is not enriched at hotspots in spo11Δ tid1Δ mutants, but twofold more DNA is immunoprecipitated at both the hotspot and the coldspot after 4 h in sporulation medium vs. the spo11Δ single mutant. (B) Dmc1 does not preferentially associate with DSBs at HIS4 ::LEU2 in a tid1Δ rad54Δ double mutant. “WT 20%” represents a mixture of cells consisting of 20% wild-type cells from 4 h in meiosis with 80% wild-type cells from a mitotic culture. This mixture was designed to mimic the low efficiency of meiotic induction in the tid1Δ rad54Δ mutant in order to demonstrate that the ChIP assay is capable of detecting preferential hotspot association, even when meiotic induction is inefficient. (C) Dmc1 Western blot showing accuracy of mixing in the “WT 20%” sample compared with tid1Δ rad54Δ.

Figure 4.

Tid1 ATPase activity is required for dissociation of Dmc1 from dsDNA. Distribution of Dmc1 focus counts from spo11Δ and spo11Δ tid1 mutants after 4 h of meiosis. Focus counts from 50 unselected nuclei are plotted in ascending order.

Figure 5.

Model for Tid1 activity. Tid1 has an ATP-dependent function that opposes Dmc1–dsDNA interaction. This can occur at two steps during meiosis. First, Dmc1 is removed from nonspecific sites on duplex DNA prior to DSB formation, creating a pool of Dmc1 available for binding when DSBs do form. After DSB formation and strand exchange, Tid1 then removes Dmc1 from the duplex product of strand exchange to allow recombination to be completed.