Replication stress-induced Exo1 phosphorylation is mediated by Rad53/Pph3 and Exo1 nuclear localization is controlled by 14-3-3 proteins
Mechanisms controlling DNA resection at sites of damage and affecting genome stability have been the subject of deep investigation, though their complexity is not yet fully understood. Specifically, the regulatory role of post-translational modifications in the localization, stability and function of DNA repair proteins is an important aspect of such complexity.
Here, we took advantage of the superior resolution of phosphorylated proteins provided by Phos-Tag technology to study pathways controlling the reversible phosphorylation of yeast Exo1, an exonuclease involved in a number of DNA repair pathways. We report that Rad53, a checkpoint kinase downstream of Mec1, is responsible for Exo1 phosphorylation in response to DNA replication stress and we demonstrate a role for the type-2A protein phosphatase Pph3 in the dephosphorylation of both Rad53 and Exo1 during checkpoint recovery. Fluorescence microscopy studies showed that Rad53-dependent phosphorylation is not required for the recruitment or the release of Exo1 from the nucleus, whereas 14-3-3 proteins are necessary for Exo1 nuclear translocation.
By shedding light on the mechanism of Exo1 control, these data underscore the importance of post-translational modifications and protein interactions in the regulation of DNA end resection.
Keywords14-3-3 Budding yeast DNA replication Exonuclease-1 Phosphorylation Pph3 Rad53 Sub-cellular localization
The fidelity of DNA replication is essential to maintain a stable genome. Errors occurring during replication facilitate the development of cancer [1, 2]. In budding yeast, DNA replication starts at defined sequences or origins that are distributed throughout chromosomes and where proteins of the origin recognition complex (ORC) bind upon mitotic exit [3, 4]. Timely recruitment of additional components leads to the formation of fully competent replisomes that, upon firing, move bi-directionally away from origins . DNA damage represents a physical impediment to replication causing fork stall and collapse, eventually resulting in chromosome breaks and genome rearrangements . To prevent this, a replication checkpoint has evolved as surveillance mechanism that controls components of the replisome , thus allowing to coordinate cell cycle arrest with DNA repair.
Exo1 is a DNA repair nuclease of the Rad2 family originally identified in the fission yeast S. pombe where its activity was shown to increase ~ fivefold during meiosis, suggesting a role for Exo1 in homologous recombination . Similar observations were made in Saccharomyces cerevisiae and D. melanogaster [9, 10]. Budding yeast Exo1 was also shown to participate in the processing of DNA ends at double-strand breaks (DSB) , in mitotic recombination  as well as in end-resection at uncapped telomeres . A role in mutation avoidance and mismatch correction was described for S. pombe Exo1  and later confirmed in S. cerevisiae, demonstrating physical and genetic interaction between yeast Exo1 and the DNA mismatch repair (MMR) proteins Msh2  and Mlh1 . Both S. cerevisiae and human EXO1 were shown to participate in the process of nucleotide excision repair (NER) after UV irradiation [16, 17]. Additionally, human EXO1 plays an important role in the repair of DSB by HR where, in a two-step resection mechanism, it was shown to carry out the extensive resection that is necessary to generate recombinogenic structures [18, 19]. Studies conducted in S. cerevisiae showed redundancy between Exo1 and Rad27 in processing Okazaki fragments during DNA replication  and the recruitment of yeast Exo1 to stalled replication forks was shown to contribute to fork stability by counteracting fork reversal .
Biochemically, EXO1 catalyzes the removal of mononucleotides from the 5′-end of the DNA duplex, showing a strong preference for blunt-ended, 5′-recessed termini and DNA nicks. EXO1 can process single-stranded DNA, though less efficiently than double-stranded DNA [22, 23, 24], and displays 5′-ssDNA-flap-specific endonuclease activity but does not possess endonuclease activity at bubble-like structures . The mechanism by which EXO1 acts on DNA was investigated at the molecular level in a study where the catalytic domain of EXO1 was co-crystallized with a 5′ recessed-end substrate. The data showed that, in analogy to other FEN nucleases, EXO1 first splays apart the DNA duplex inducing a sharp bend proximal to the cleavage site and then frays two nucleotides at the 5′ end, facilitating access of the catalytic site to the scissile phosphodiester bond .
A number of laboratories, including ours, have reported that both yeast and human EXO1 are regulated by post-translational modifications such as phosphorylation, ubiquitylation and sumoylation [26, 27, 28, 29, 30, 31, 32]. A proteome wide study aimed at identifying in vivo targets of checkpoint kinases reported S372 as site of phosphorylation in yeast Exo1 . A study addressing the effect of telomere dysfunction on DNA resection described four additional sites of phosphorylation in yeast Exo1 that appeared to exert a negative effect on Exo1 activity .
In the present study, we employed Pho-Tag gel technology to resolve phosphorylated Exo1 in response to stalled replication or DNA damage and identified components of the Mec1–Rad53–Dun1 pathway as well as the phosphatase Pph3 in the dynamic control of yeast Exo1 phosphorylation. Furthermore, we explored the role of 14-3-3 proteins and Rad53 on the subcellular localization of Exo1 during checkpoint activation and recovery.
Exo1 phosphorylation upon replication stress is Rad53-dependent and Dun1-independent
These data unequivocally indicate that Rad53 is the component of the Mec1 pathway responsible for Exo1 phosphorylation in response to DNA replication stress.
Pph3 controls Rad53 and Exo1 dephosphorylation during recovery
The established lack of viability of glc7-null cells  prompted us to take advantage of a catalytic mutant of the phosphatase (glc7-E101Q), which displays defects in glucose metabolism but normal cell cycle progression and chromosome segregation . The data showed that the pattern of Exo1 or Rad53 mobility during recovery from HU in the glc7-E101Q background was not altered in comparison to control cells (Fig. 2a), ruling out a contribution from Glc7 in the dephosphorylation of these proteins during checkpoint recovery.
Taken together, these data indicate that Pph3 is the major phosphatase involved in Rad53 and Exo1 dephosphorylation during recovery from HU.
14-3-3 proteins control Exo1 localization
To examine possible effects of 14-3-3 proteins on the localization of Exo1-GFP we employed the bmh1-280 bmh2Δ double mutant strain, which is characterized by normal cell cycle progression in unperturbed conditions, but shows recovery defects in response to HU . The mutant Bmh1-280 protein carries a single point mutation (E136 > G) in helix αE, affecting a residue close to the amino acids that form salt bridges and hydrogen bonds with the ligand . In bmh1-280 bmh2Δ cells (hereafter 14-3-3-deficient cells), the intense green fluorescence signal of Exo1-GFP remained diffused in response to treatment with HU, failing to shuttle into the nucleus (Fig. 3).
These data suggest that 14-3-3 proteins are necessary to facilitate Exo1 nuclear translocation, whereas the Rad53/Pph3 control of Exo1 phosphorylation does not seem to play a role in the export of Exo1 from the nucleus.
Phosphorylation of yeast Exo1 upon checkpoint activation was so far investigated by means of cumbersome mass spectrometry and the effect of Exo1 phosphorylation on DNA resection was indirectly assessed from patterns of cell viability [31, 33]. We have previously shown the power of Phos-Tag technology, which allows resolving a number of phosphorylated Exo1 isoforms induced in response to stalled replication . In the present study, combining Phos-Tag SDS-PAGE resolution of proteins and Western blot analysis (Fig. 1a and Additional file 1: Fig. S1), we confirm and extend these observations, demonstrating that Exo1 undergoes phosphorylation in response to DNA damage by an alkylating agent as well as by an agent that stalls DNA replication. With regard to the protein kinase cascade signaling to Exo1, we previously compared the pattern of Exo1 mobility in wild type and mec1/sld1Δ cells, concluding that retarded Exo1 migration in Phos-Tag gels depended on Mec1 activity . Extending those observations, we now provide evidence for a similar retardation of migration in rad53-K227A cells, but not in dun1Δ cells, challenged with a replication inhibitor (Fig. 1b). This indicates that Exo1 phosphorylation in response to HU is predominantly Rad53-dependent.
The major Ser–Thr protein phosphatase activities in the cells, namely those associated with the type-1 and type-2A enzymes Glc7 and Pph3, respectively, have been shown to play a major role during recovery from DNA replication stress or MMS-induced damage through dephosphorylation of γH2AX and Rad53 [34, 36, 37]. We demonstrate that absence of Pph3 caused an evident retardation of Exo1 dephosphorylation during recovery from HU and also contributed to delay dephosphorylation of Rad53 (Fig. 2a). Checkpoint switch-off by caffeine in pph3Δ cells showed that dephosphorylation of both Rad53 and Exo1 was delayed (Fig. 2c), indicating that Pph3 may act directly or indirectly on Exo1.
We have previously shown that 14-3-3 proteins bind both human and yeast Exo1, with the distinction that the former establishes a constitutive interaction, whereas the latter binds 14-3-3 only upon stalled replication . Hence, checkpoint-dependent phosphorylation of yeast Exo1 likely creates the conditions for interaction with 14-3-3 proteins, whereas in human EXO1 suitable sites are likely phosphorylated in a checkpoint-independent manner. In human EXO1, at least six phosphorylation sites potentially responsible for this protein–protein interaction were identified . Functional analysis with phospho-null mutants indicated that they have no role in nuclear import of EXO1 . A subsequent study identified the domain 508–750 of EXO1, which comprises a number of previously identified sites of phosphorylation, as the region responsible for interaction with 14-3-3 . The authors showed that 14-3-3 protein binding through this region causes inhibition of EXO1-mediated DNA resection in a cell-free system but not in a biochemical resection assay with purified components, indicating the participation of yet unidentified components in this process . Regardless of this, the study revealed a functional role for the interaction of 14-3-3 proteins with EXO1. Such modulatory effect of 14-3-3 proteins on the human enzyme is in agreement with our previous findings in budding yeast, where we showed that lack of functional Bmh1 and Bmh2 proteins (the yeast homologues of human 14-3-3 proteins) unleashes Exo1 activity, causing pathological resection of DNA . Using wild type and mutant strains, in the present study we discovered an additional role of yeast 14-3-3 proteins, namely an effect on the dynamic shuttling of Exo1 through cell compartments. We provide visual evidence that yeast Exo1 is recruited to nuclei during regular transition through S-phase (Additional file 1: Fig. S3) as well as in response to replication stress (Fig. 3), in a manner that depends on 14-3-3 proteins. Release of Exo1 from nuclei of wild type cells occurs during the recovery phase (Fig. 3). In Rad53-deficient cells recovering from replication stress, the release of Exo1 from nuclei was impaired (Fig. 3), a fact that can be correlated to failed phosphorylation by kinase-defective Rad53. However, in pph3Δ cells where wild type Rad53 remained partially phosphorylated during initial phases of recovery (Fig. 2a), hence possibly able to phosphorylate Exo1, and when Exo1 itself was hyper-phosphorylated, we did not observe preferential exclusion of Exo1 from the nucleus (Fig. 4), indicating that the Rad53/Pph3 control of Exo1 phosphorylation does not likely play a role in nuclear export of Exo1.
Interestingly, in the absence of functional 14-3-3 proteins, we observed that Exo1 remained distributed in all cell compartments, regardless of the presence of HU (Fig. 3), with the nuclear sub-population of Exo1 likely being the one responsible for the reported over-resection of DNA . This suggests that nuclear localization of Exo1 requires interaction with 14-3-3 proteins. Since it was previously reported that yeast 14-3-3 proteins bind to the checkpoint kinase Rad53, influencing its DNA damage-dependent functions , and we showed that yeast 14-3-3 proteins bind Exo1 upon replication stress , it is likely that 14-3-3 proteins coordinate phosphorylation of Exo1 by Rad53.
This work demonstrates that the dynamic phosphorylation of yeast Exo1 in response to stalled replication is under the control of the checkpoint kinase Rad53. During recovery from HU the phosphatase Pph3 results to be the major controller of Rad53 and Exo1 phosphorylation. Furthermore, this study extended previous observations on the interaction between Exo1 and 14-3-3 proteins showing that the latter contribute to shuttle Exo1 into the nucleus under conditions of normal DNA replication or upon stress caused by low nucleotide availability.
As a whole, this study sheds further light on the control of a key component of the DNA resection machinery.
Materials and methods
The antibodies used in this study were: mouse monoclonal anti-Myc (9E10, sc-40), mouse monoclonal (sc-74427) and goat polyclonal (sc-6749) anti-Rad53, rat monoclonal anti-α-tubulin (sc-53030) from Santa Cruz Biotechnology. Hydroxyurea was purchased from Bio-Basic Canada Inc., whereas α1-mating factor and all other reagents were from Sigma.
Saccharomyces cerevisiae strains
List of S. cerevisiae strains used in this study
MATa leu2-3,112 trp1-1 can1-100 ura3-1 ade2-1 his3-11,15 [phi+]
MATa rad53-K227A::KANMX4 exo1Δ::HIS3
MATa EXO1-Myc::KANMX4::exo1 dun1Δ::URA3
MATa EXO1-Myc::KANMX4::exo1 pph3Δ::URA3
MATa glc7-E101Q::KANMX4 EXO1-Myc::NAT1::exo1
MATa bmh2Δ::NAT1 bmh1Δ::HIS3::bmh1-280::LEU2
MATa bmh2Δ::NAT1 bmh1Δ::HIS3::bmh1-280::LEU2 exo1::URA3 rad53-K227A::KANMX4
MATa bmh2Δ::NAT1 bmh1Δ::HIS3::bmh1-280::LEU2 EXO1-GFP::KITRP1-1::exo1
MATa rad53-K227A::KANMX4 EXO1-GFP::KITRP1-1::exo1
MATa rad53-K227A::KANMX4 bmh2Δ::NAT1 bmh1Δ::HIS3::bmh1-280::LEU2 EXO1-GFP::KITRP1-1::exo1
MATa EXO1-GFP::KITRP1-1::exo1 pph3Δ::URA3
Log-phase cells were treated with 5 µg/ml α-factor for 120 min, released in fresh YPD containing 50 µg/ml pronase and harvested in a time-course fashion.
Wild-type and mutant strains were grown exponentially. Serial dilutions (1:10) were spotted on YPD plates containing different HU concentrations and grown for 3 days before scoring.
Protein extraction and Western Blotting
Western blot analysis of yeast proteins was carried out upon TCA extraction . To visualize Exo1, an optimized Phos-tag system (50–150 µM Phos-tag reagent) was employed according to . Proteins were transferred to PVDF (0.45 µm pore size, Machery-Nagel) overnight at 4 °C applying constant voltage (30 V). Immunoblot analysis was performed as previously described  and proteins were visualized using the FUSION SOLO® chemiluminescence imaging system (Vilber).
Flow cytometric analysis
Flow cytometric analysis was performed as described .
Yeast cells were fixed in 4% formaldehyde and mounted with Vectashield® (containing DAPI). Cells were imaged with an Olympus 1X71 fluorescence microscope. An oil immersion 100× objective was used. DAPI was rendered in red as false color.
NRC designed, performed and analyzed experiments; GDG performed experiments; SF designed and analyzed experiments and wrote the manuscript. All authors read and approved the final manuscript.
We would like to thank Dr. J. Oullet and Prof. Y. Barral, ETH Zurich, for tools and assistance with fluorescence microscopy as well as Farah M’hmedi for technical support. We are also indebted to Dr. K. Engels, Prof. A. Pellicioli and members of SF laboratory for helpful suggestions.
The authors declare that they have no competing interests.
Availability of data and materials
All data generated or analyzed during this study are included in this published article.
Consent for publication
All authors agree with publishing this study.
Ethics approval and consent to participate
This work was supported by grants from the Swiss Foundation for Fight against Cancer and from the University of Zurich Research Funds.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 43.Andersen SD, Keijzers G, Rampakakis E, Engels K, Luhn P, El-Shemerly M, Nielsen FC, Du Y, May A, Bohr VA, et al. 14-3-3 checkpoint regulatory proteins interact specifically with DNA repair protein human exonuclease 1 (hEXO1) via a semi-conserved motif. DNA Repair (Amst). 2012;11(3):267–77.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.