Break-induced replication requires all essential DNA replication factors except those specific for pre-RC assembly

Abstract

Break-induced replication (BIR) is an efficient homologous recombination (HR) pathway employed to repair a DNA double-strand break (DSB) when homology is restricted to one end. All three major replicative DNA polymerases are required for BIR, including the otherwise nonessential Pol32 subunit. Here we show that BIR requires the replicative DNA helicase (Cdc45, the GINS, and Mcm2-7 proteins) as well as Cdt1. In contrast, both subunits of origin recognition complex (ORC) and Cdc6, which are required to create a prereplication complex (pre-RC), are dispensable. The Cdc7 kinase, required for both initiation of DNA replication and post-replication repair (PRR), is also required for BIR. Ubiquitination and sumoylation of the DNA processivity clamp PCNA play modest roles; in contrast, PCNA alleles that suppress pol32Delta's cold sensitivity fail to suppress its role in BIR, and are by themselves dominant inhibitors of BIR. These results suggest that origin-independent BIR involves cross-talk between normal DNA replication factors and PRR.

Figure 1.

Experimental system and model of BIR. (A) An HPHMX-marked HO cut site (gray bar) is integrated into the CAN1 gene on Ch V, deleting the 3′ end portion of the gene; the remaining sequences are represented as CA. The AN1 donor sharing 1157-bp homology with CAN1 is integrated into Ch XI. At the start, PCR primers P1 and P2, which are used to monitor the initiation of new DNA synthesis, will not amplify because they sit on different chromosomes. (B) Following HO expression, a DSB is generated and the broken chromosome ends are resected by 5′-to-3′ exonucleases to generate ssDNA. (C) The Rad51 recombinase binds to the 3′-ended single strand and mediates the homology search and strand invasion of the donor template to form a D-loop. (D) The BIR replication machinery is recruited to the recombination intermediate, and the leading and lagging strand coordinately initiate new DNA synthesis. The initiation of new DNA synthesis is detected by PCR when the P2 sequences are covalently linked to the P1 sequences. (E) BIR DNA synthesis extends to the end of the template chromosome, resulting in a nonreciprocal translocation in which the CAN1 gene is restored, while the entire donor chromosome arm is duplicated and sequences distal to the original HO cut site are lost.

Figure 2.

The Mcm2–7 helicase is required for BIR DNA synthesis. (A–D) Appearance of BIR repair product, as monitored by PCR, in wild-type (WT), mcm4-td, cdc45-td, and psf2-td cells arrested in nocodazole at either 25°C or 37°C. Error bars represent ±SEM for wild type, mcm4-td, and cdc45-td, and the error range for two independent measurements of psf2-td.

Figure 3.

Cdt1, Dpb11, and Sld3 are required for BIR DNA synthesis. (A–E) Appearance of BIR repair product, as monitored by PCR, in cdc6-1, orc6-td, cdt1-td, dpb11-1, and sld3-7td cells arrested in nocodazole at either 25°C or 37°C. Error bars represent ±SEM for cdc6-1, orc6-td, cdt1-td, and sld3-7td, and the error range for two independent measurements of dpb11-1. (F) Efficiency of BIR in wild type (WT) and dpb11-1 as measured by viability following a DSB at the permissive temperature of 25°C. Error bars represent ±SEM.

Figure 4.

Mcm10 and Cdc7 are required for the initiation of BIR DNA synthesis, and Ctf4 is important for BIR. (A) Appearance of BIR repair product, as monitored by PCR, in mcm10-td cells arrested in nocodazole at either 25°C or 37°C. Error bars represent the error range for two independent measurements. (B) Efficiency of BIR in wild-type (WT), ctf4Δ, ctf18Δ, chl1Δ, mrc1Δ, and tof1Δ cells, as measured by viability following a DSB. Error bars represent ±SEM. (C) Appearance of BIR repair product, as monitored by PCR, in cdc7-4 cells arrested in nocodazole at either 25°C or 37°C. Error bars represent the error range for two independent measurements. (D) Efficiency of BIR in wild-type (WT), cdc7-4, mcm4Δ74-174, and cdc7-4 mcm4Δ74-174 cells, as measured by viability following a DSB at the permissive temperature of 25°C or restrictive temperature of 37°C. Error bars represent ±SEM.

Figure 5.

Rad18-mediated ubiquitination and Siz1-mediated sumoylation of PCNA are important for BIR but not GC. (A) Efficiency of BIR in wild-type (WT), pol30K164R, rad18Δ, siz1Δ, pol30K164R rad18Δ, pol30K164R siz1Δ, rad18Δ siz1Δ, pol30K127R, K164R, mgs1Δ, ubc13Δ, rad30Δ rev3Δ, and srs2Δ cells, as measured by viability following a DSB. Error bars represent ±SEM. (B) Appearance of BIR repair product, as monitored by PCR, in wild-type (WT) and srs2Δ cycling cells. Error bars represent the error range for two independent measurements. (C) The experimental system to study GC at the CAN1 locus (disrupted by URA3 creating a 376-bp gap) on Ch V. The homologous sequences that serve as the donor for repair by GC are found on Ch XI. (D) Efficiency of GC in wild-type (WT), rad51Δ, pol32Δ, rad18Δ, and siz1Δ cells, as measured by viability following a DSB. Error bars represent ±SEM.

Figure 6.

Novel PCNA alleles are required for BIR. (A) PCNA mutant alleles pol30-89 and pol30-92 suppress pol32Δ cold sensitivity. Tenfold serial dilutions of the indicated strains incubated for 6 d at 18°C or 3 d at 30°C. (B) PCNA mutant alleles pol30-89 and pol30-92 are defective in BIR. Efficiency of BIR in wild-type (WT), pol32Δ, pol30-89, pol30-89 pol32Δ, pol30-92, and pol30-92 pol32Δ cells, as measured by viability following a DSB. Error bars represent ±SEM. (C) PCNA mutant alleles pol30-89 and pol30-92 are dominant-negative in BIR. Efficiency of BIR in wild-type (WT) (POL30), pol30-89 POL30, and pol30-92 POL30 cells, as measured by viability following a DSB. Error bars represent ±SEM. (D) PCNA mutant alleles pol30-89 and pol30-92 are not required for GC. Efficiency of GC in wild-type (WT) (pol30-0), pol30-89, and pol30-92 cells, as measured by viability following a DSB. Error bars represent ±SEM.