Dual control of Yen1 nuclease activity and cellular localization by Cdk and Cdc14 prevents genome instability

Abstract

The careful orchestration of cellular events such as DNA replication, repair, and segregation is essential for equal distribution of the duplicated genome into two daughter cells. To ensure that persistent recombination intermediates are resolved prior to cell division, the Yen1 Holliday junction resolvase is activated at anaphase. Here, we show that the master cell-cycle regulators, cyclin-dependent kinase (Cdk) and Cdc14 phosphatase, control the actions of Yen1. During S phase, Cdk-mediated phosphorylation of Yen1 promotes its nuclear exclusion and inhibits catalytic activity by reducing the efficiency of DNA binding. Later in the cell cycle, at anaphase, Cdc14 drives Yen1 dephosphorylation, leading to its nuclear relocalization and enzymatic activation. Using a constitutively activated form of Yen1, we show that uncontrolled Yen1 activity is detrimental to the cell: spatial and temporal restriction of Yen1 protects against genotoxic stress and, by avoiding competition with the noncrossover-promoting repair pathways, prevents loss of heterozygosity.

Graphical abstract

Figure 1

Yen1^ON Is a Constitutively Active Version of Yen1 (A) Schematic representation of Yen1, indicating the 25 phosphoresidues identified by mass spectrometry (above) and the nine predicted Cdk consensus target sites (below). Cdk sites are highlighted in red. (B) Analysis of the nuclease activity of Yen1^ON. Soluble lysates were prepared from asynchronous cultures of strains in which YEN1-myc9, nuclease-dead YEN1^EEAA-myc9, and YEN1^ON-myc9 alleles replaced endogenous YEN1. Following α-myc immunoaffinity purification, the IPs were analyzed by western blotting and for nuclease activity using ³²P-labeled Holliday junction DNA. Upper panel: western blots of the lysate and IP, with Pgk1 as a loading control. Lower panel: resolution assays. (−), untreated substrate. (C) Yen1^ON overexpression causes increased MMS sensitivity. Ten-fold dilutions of wild-type strain BY4741 transformed with centromeric plasmids carrying different YEN1 alleles (see Figure S1) were plated on SC-URA with or without MMS and imaged after 3 days. (D) Yen1^ON activity escapes cell-cycle control. The activity profiles of Yen1-myc9, Yen1^EEAA-myc9 and Yen1^ON-myc9 were analyzed as in (B) following their immunoprecipitation from synchronized cultures. As, asynchronous culture; G1, α-factor treated culture; S, S and G2 cells 25 min after α-factor release; M, G2/M cells 60 min after α-factor release. See also Figure S1.

Figure 2

Premature Activation of Yen1 Causes Increased DNA Damage Sensitivity and Loss of Heterozygosity (A) Expression of Yen1^ON at physiological levels causes MMS sensitivity. Assays were carried out as described in Figure 1C, using strains in which endogenous YEN1 was replaced by the indicated YEN1 alleles. Cells were plated on YPD containing the indicated amount of drugs. (B) YEN1^ON suppresses the DNA damage sensitivity of mus81Δ mutants. As in (A), but employing strains deleted for MUS81. (C) Mild suppression of the MMS sensitivity of sgs1Δ mutants by YEN1^ON. (D) YEN1^ON suppresses the synthetic lethality of mus81Δ sgs1Δ double mutants. Diploid strains homozygous for the indicated YEN1 alleles and heterozygous for MUS81/mus81Δ and sgs1Δ/SGS1 were sporulated and analyzed by tetrad dissection. Images were taken after 3 days incubation at 30°C. The expected position of each mus81Δ sgs1Δ colony is indicated (red circles). (E) YEN1^ON promotes crossover formation and loss of heterozygosity during mitotic recombination. The percentage of different recombination events in YEN1 and YEN1^ON strains is indicated. CO, crossovers; NCO, noncrossovers; BIR, break-induced replication. LOH indicates the total loss of heterozygosity (the sum of CO and BIR events). p values were calculated using a chi-square test with Pearson’s correction. See also Table S1.

Figure 3

Biochemical Activation and Subcellular Localization Are Independent Layers of Yen1 Regulation (A) Nuclear localization of Yen1 throughout the cell cycle. Yen1-myc18 at different stages of the cell cycle was analyzed by immunofluorescence. Spindle morphology and DNA were visualized using anti-tubulin antibodies and DAPI. The contours of the cells, determined from differential interference contrast (DIC) images, are depicted by a dotted line. (B) As in (A), but with Yen1^ON. (C) Subcellular localization of Yen1 phosphomutants. Nuclear enrichment (as opposed to diffuse pan-cellular staining) of the indicated Yen1 mutants was analyzed as in (A) and quantified (200 cells per condition). The schematic indicates the mutations at the Cdk sites in each allele. Protein domains are colored as in Figure 1A. (D) Analysis of the biochemical activities of the Yen1 mutants used in (C). Myc9-tagged proteins were immunoprecipitated from S phase-arrested cells, and their nuclease activities were assayed. (E) Suppression of the mus81Δ phenotype depends on both the biochemical activation and nuclear localization of Yen1. mus81Δ strains carrying the indicated YEN1 alleles were analyzed as in Figure 2B.

Figure 4

Cdk and Cdc14 Phosphatase Regulate the Nuclease Activity and Subcellular Localization of Yen1 (A) Schematic representation of the experimental conditions employed in the Cdk inhibition experiments (B and C). (B) Cdk inhibition in S phase cells increases Yen1 activity. cdc28-as1 cells carrying galactose-inducible YEN1-myc9 were treated with HU. To one-half of the culture, galactose was added before DMSO (lane 1) or the Cdc28-as1 kinase inhibitor 1NM-PP1 (lane 2). The other half was split, and galactose was added after DMSO (lane 3) or 1NM-PP1 (lane 4). Yen1-myc9 was immunoprecipitated and assayed for HJ resolution. (C) Yen1^ON activity is refractory to Cdk inhibition. As in (B), except that YEN1^ON-myc9 was expressed after addition of DMSO (3) or 1NM-PP1 (4). The amounts of immunoprecipitated protein were reduced compared with (B) in order to limit DNA cleavage. (D) cdc14-1 mutants fail to activate Yen1 at restrictive temperature. Yen1-myc9 IPs from the indicated strains (grown at 25°C and then shifted to 37°C for 2 hr) were tested for nuclease activity. (E) Yen1^ON activity is refractory to Cdc14 inactivation. As in (D), but employing cdc14-1 strains carrying either YEN1-myc9 or YEN1^ON-myc9. (F) Cdc14 is required for nuclear enrichment of Yen1 at anaphase. The immunofluorescence analysis of Yen1-myc18 or Yen1^ON-myc18 during anaphase in CDC14 or cdc14-1 strains at the restrictive temperature was carried out as in Figure 3. The percentage of the cells displaying the most representative type of Yen1 distribution (nuclear or cellular) is shown. The contours of the cells, determined from DIC images, are depicted by a dotted line.

Figure 5

Direct Activation of Yen1 by Cdc14 Phosphatase (A) Analysis of Yen1 activity in TAB6-1 (CDC14^P116L) mutants. Yen1-myc9 IPs from CDC14 and TAB6-1 strains treated with HU (enriched at the G1/S phase transition) were tested for nuclease activity. (B) Overexpression of Cdc14 in cycling cells results in Yen1 activation. Cycling cultures of the indicated strains were supplemented with 2% galactose for 2 hr to induce Cdc14 expression. As controls, similar cultures were left uninduced. Lysates and IPs were analyzed as described in Figure 1A. (C) Dephosphorylation and activation of Yen1 by ectopic overexpression of Cdc14 in HU-treated cells. The indicated strains were blocked at G1/S and Cdc14 expression was induced by adding 2% galactose for 2 hr. As, asynchronous culture. (D) Physical interactions between Yen1 and Cdc14. P_GAL1-CDC14 mutants were induced by addition of 0.5% galactose for 90 min. For each culture, lysates were prepared and split into two, and proteins were immunoprecipitated with either α-FLAG or α-myc. Proteins were analyzed by western blotting.

Figure 6

Phosphorylation of Yen1 Modulates DNA Binding (A) FTH-tagged Yen1, Yen1^EEAA, and Yen1^ON were purified from asynchronous cells and analyzed by SDS-PAGE and Coomassie staining (1 μg each). (B) Yen1^ON exhibits elevated nuclease activity. FTH-tagged Yen1 (Y), Yen1^EEAA (E), and Yen1^ON (O) (10 nM) were analyzed for nuclease activity using a static Holliday junction X0, nicked HJ X0, mobile HJ X26, replication fork, 5′-flap (upper panel); 3′-flap, splayed arm, double-stranded DNA (dsDNA) (60 bp), oligo dT₆₀, and dsDNA (120 bp). (−): no protein. (C) SDS-PAGE of purified Yen1^P and Yen1^λ (1 μg each). Yen1-FTH was expressed in cdc14-1 cells arrested at 37°C and purified without (Yen1^P) or with (Yen1^λ) λ-phosphatase treatment. (D) Phosphorylated Yen1 exhibits reduced nuclease activity. Upper panels: increasing amounts of Yen1^λ or Yen1^P (0, 0.25, 0.5, 1, 2, 4, 8, 16 nM) were incubated with 5′-flap DNA. Lower panels: Yen1^λ or Yen1^P (0, 1, 2, 4, 8, 16, 32, 64 nM) with HJ X0. (E) Yen1^P exhibits reduced ability to bind DNA. DNA binding assays were carried out using dsDNA, 5′-flap, and X0 HJ DNA, with either Yen1^λ (upper panel) or Yen1^P (lower panel). Proteins were used at 0, 10, 20, 40, 80, 160, and 320 nM.

Figure 7

Regulatory Control of Yen1 by Cdk and Cdc14 (A) Cdk and Cdc14 control the phosphorylation status of Yen1. In S phase, Cdk phosphorylates Yen1 to (1) promote its nuclear exclusion by impairing its NLS function and, potentially, enhance Msn5-dependent export, and (2) reduce its DNA binding affinity, avoiding the cleavage of S phase-specific DNA structures such as replication or early recombination intermediates. At anaphase, Cdc14 dephosphorylates Yen1 to reinstate a fully functional NLS and potentially block Msn5-mediated export, leading to its nuclear accumulation. Additionally, the removal of phosphate groups increases the DNA binding affinity and catalytic activity of Yen1, allowing it to resolve HR intermediates that have persisted until anaphase. Red circles (P) depict phosphate groups. (B) Yen1^ON cannot be turned off or shuttled to the cytoplasm by Cdk. This may lead to the unscheduled and detrimental cleavage of replication or early recombination intermediates. In mutants that accumulate HR intermediates, the constitutive activation of Yen1^ON provides an alternative way to process these potentially toxic DNA structures.