DNA repair by a Rad22-Mus81-dependent pathway that is independent of Rhp51

Abstract

In budding yeast most Rad51-dependent and -independent recombination depends on Rad52. In contrast, its homologue in fission yeast, Rad22, was assumed to play a less critical role possibly due to functional redundancy with another Rad52-like protein Rti1. We show here that this is not the case. Rad22 like Rad52 plays a central role in recombination being required for both Rhp51-dependent and -independent events. Having established this we proceed to investigate the involvement of the Mus81-Eme1 endonuclease in these pathways. Mus81 plays a relatively minor role in the Rhp51-dependent repair of DNA damage induced by ultraviolet light. In contrast Mus81 has a key role in the Rad22-dependent (Rhp51-independent) repair of damage induced by camptothecin, hydroxyurea and methyl-methanesulfonate. Furthermore, spontaneous intrachromosomal recombination that gives rise to deletion recombinants is impaired in a mus81 mutant. From these data we propose that a Rad22-Mus81-dependent (Rhp51-independent) pathway is an important mechanism for the repair of DNA damage in fission yeast. Consistent with this we show that in vitro Rad22 can promote strand invasion to form a D-loop that can be cleaved by Mus81.

Figure 1

Hypothetical model showing potential roles that Mus81–Eme1 might have in processing stalled and broken replication forks. Nascent strands are shown as red lines with arrowheads at 3′ ends. The replication fork block is indicated by a solid circle, and the position of Mus81–Eme1 cleavage sites by the labelled arrows.

Figure 2

Spot assay comparing the genotoxin sensitivity of different rad22 mutant strains. The strains used were MCW1221 (wild-type), MCW1285 (new rad22⁻), MCW1222 [rad22⁻; (22)], MCW4 [rad22⁻; (40)] and MCW1501 [rad22⁻; (41)]. Cultures were serially diluted and spotted onto YES agar as described in Materials and Methods. The neat spot ‘1’ represents 10⁵ cells plated.

Figure 3

Rhp51-dependent and -independent recombination and DNA repair. (A) Spot assay comparing the sensitivity of wild-type (MCW1221), rhp51⁻ (MCW1088), rad22⁻ (MCW1285), and rad22⁻ rhp51⁻ (MCW1335) strains to UV, HU, MMS and CPT. (B) Survival curves showing the UV and γ-ray sensitivity of the same strains as in ‘A’. (C) Schematic of the recombination substrate and recombinant products. Solid and open circles represent the ade6-L469 and ade6-M375 mutations, respectively. (D) Comparison of the spontaneous recombinant frequencies of wild-type (MCW429), rhp51⁻ (MCW1224), rhp55⁻ (MCW431), rhp54⁻ (MCW1225), rad22⁻ (MCW1494), rhp51⁻ rad22⁻ (MCW1038), rhp55⁻ rad22⁻ (MCW1227), rhp51⁻ rhp55⁻ (MCW1228) and rhp54⁻ rhp55⁻ (MCW1229) strains. Error bars represent standard deviations about the mean.

Figure 4

Mus81 functions in both Rhp51-dependent and -independent repair pathways depending on the type of DNA damage. (A) Spot assay showing the relative sensitivities to UV, HU, MMS and CPT of wild-type (MCW45), mus81⁻ (MCW745), rhp51⁻ (MCW3) and mus81⁻ rhp51⁻ (MCW892) strains (left panel), and wild-type (MCW1221), mus81⁻ (MCW1502), rad22⁻ (MCW1285) and mus81⁻ rad22⁻ (MCW1337) strains (right panel). (B) Survival curves showing the sensitivity of wild-type (MCW1221), mus81⁻ (MCW1502), rhp51⁻ (MCW1088), rad22⁻ (MCW1285), mus81⁻ rhp51⁻ (MCW1235) and mus81⁻ rad22⁻ (MCW1337) strains to UV. (C) Survival curves showing the CPT sensitivity of wild-type (MCW42), mus81⁻ (MCW745), rhp51⁻ (MCW3) and mus81⁻ rhp51⁻ (MCW891) strains (left panel), and wild-type (MCW1221), mus81⁻ (MCW1502), rad22⁻ (MCW1285) and mus81⁻ rad22⁻ (MCW1337) strains (right panel). (D) Spot assay showing the relative growth and genotoxin sensitivity of rhp51⁻ rhp54⁻ (MCW1497), mus81⁻ rhp51⁻ (MCW1235), mus81⁻ rhp54⁻ (MCW1498) and mus81⁻ rhp51⁻ rhp54⁻ (MCW1499) strains after 3 and 5 days growth as indicated. (E) Spot assay showing the relative sensitivities to UV, HU, MMS and CPT of MCW1335 (rad22⁻ rhp51⁻), MCW1337 (rad22⁻ mus81⁻), MCW1235 (rhp51⁻ mus81) and MCW1478 (rad22⁻ rhp51⁻ mus81⁻).

Figure 5

Effect of mus81⁻ on the frequency and type of recombinants in different mutant backgrounds. The strains used were MCW429 (wild-type), MCW988 (mus81⁻), MCW431 (rhp55⁻), MCW992 (rhp55⁻ mus81⁻), MCW1494 (rad22⁻), MCW1500 (rad22⁻ mus81⁻), MCW1227 (rad22⁻ rhp55⁻) and MCW1039 (rad22⁻ rhp55⁻ mus81⁻).

Figure 6

Rad22-promoted D-loop formation and D-loop cleavage by Mus81. (A) Schematic of D-loop formation. The asterisk indicates the position of the 5′-³²P-end-label on the partial duplex. (B) Rad22-promoted D-loop formation. Reactions are described in Materials and Methods and contained Rad22 (6 nM, lanes b and e; 12 nM, lanes c and f; and 24 nM, lanes d and g) and 10 mM MgCl₂. Rad22 was pre-incubated with partial duplex or φX174 DNA as indicated. (C) Dependence on homology for D-loop formation. Reactions are described in Materials and Methods. (D) Effect of Mus81 on Rad22-promoted φX174-based D-loop formation. Reactions contained 12 nM Rad22 and 14 nM Mus81–Eme1 as indicated. (E) Cleavage of purified φX174-based D-loops by Mus81–Eme1. Reactions contained 7 nM (lane c) or 14 nM (lanes d and e) Mus81–Eme1 and 10 mM MgCl₂ as indicated. The D-loop (lane a) was dissociated by heat treatment at 96°C for 2 min.