The Role of Dbf4-Dependent Protein Kinase in DNA Polymerase z-Dependent Mutagenesis in Saccharomyces cerevisiae

The yeast Dbf4-dependent kinase (DDK) (composed of Dbf4 and Cdc7 subunits) is an essential, conserved Ser/Thr protein kinase that regulates multiple processes in the cell, including DNA replication, recombination and induced mutagenesis. Only DDK substrates important for replication and recombination have been identified. Consequently, the mechanism by which DDK regulates mutagenesis is unknown. The yeast mcm5-bob1 mutation that bypasses DDK’s essential role in DNA replication was used here to examine whether loss of DDK affects spontaneous as well as induced mutagenesis. Using the sensitive lys2DA746 frameshift reversion assay, we show DDK is required to generate “complex” spontaneous mutations, which are a hallmark of the Polz translesion synthesis DNA polymerase. DDK co-immunoprecipitated with the Rev7 regulatory, but not with the Rev3 polymerase subunit of Polz. Conversely, Rev7 bound mainly to the Cdc7 kinase subunit and not to Dbf4. The Rev7 subunit of Polz may be regulated by DDK phosphorylation as immunoprecipitates of yeast Cdc7 and also recombinant Xenopus DDK phosphorylated GST-Rev7 in vitro. In addition to promoting Polzdependent mutagenesis, DDK was also important for generating Polz-independent large deletions that revert the lys2DA746 allele. The decrease in large deletions observed in the absence of DDK likely results from an increase in the rate of replication fork restart after an encounter with spontaneous DNA damage. Finally, nonepistatic, additive/synergistic UV sensitivity was observed in cdc7D pol32D and cdc7D pol30-K127R,K164R double mutants, suggesting that DDK may regulate Rev7 protein during postreplication “gap filling” rather than during “polymerase switching” by ubiquitinated and sumoylated modified Pol30 (PCNA) and Pol32. KNOWLEDGE of how mutations are produced is important for understanding genetic variability in populations and the process of evolution. Cells have evolved sophisticated molecular mechanisms for maintaining the integrity of the genome. Most DNA repair mechanisms have high fidelity and prevent mutations by removing damaged DNA and replacing the resulting gap using the undamaged, complementary strand as a template (Waters et al. 2009; Boiteux and Jinks-Robertson 2013). If not repaired, some lesions block the replicative DNA polymerases (Pols a, d, and e) during S phase and require bypass by an alternative errorfree or error-prone mechanism. Error-free bypass usually involves a template switch to the undamaged sister chromatid, while error-prone bypass uses translesion synthesis (TLS) DNA polymerases to synthesize new DNA directly across the lesion. There are at least 15 identified TLS polymerases in human cells but only three in the yeast Saccharomyces cerevisiae: Polz, Polh, and Rev1, which are encoded by the REV3-REV7, RAD30, and REV1 genes, respectively (Boiteux and Jinks-Robertson 2013). TLS polymerases have low fidelity and processivity on undamaged DNA templates and lack an associated exonuclease proofreading activity. The misregulation or loss of TLS polymerases is known to result in a number of human diseases, underscoring the importance of this process. Patients suffering from a variant form of xeroderma pigmentosum, for example, lack Polh and have an increased risk for skin cancer (Kennedy and D’Andrea 2006). In Fanconi anemia, another high-risk familial Copyright © 2014 by the Genetics Society of America doi: 10.1534/genetics.114.165308 Manuscript received April 14, 2014; accepted for publication May 23, 2014; published Early Online May 28, 2014. These authors contributed equally to this work. Present address: Gevo, 345 Inverness Dr. S., Bldg. C, Ste. 310, Englewood, CO 80112. Present address: Department of Biology, 7300 Reinhardt Circle, Reinhardt University, Waleska, GA 30183. Corresponding author: Department of Biochemistry and Molecular Genetics, University of Colorado, 12801 E. 17th Ave, MS8101, Aurora, CO 80045. E-mail: [email protected] Genetics, Vol. 197, 1111–1122 August 2014 1111 cancer syndrome, patients are TLS defective due to a failure to recruit Rev1, a nontemplate directed deoxycytidyl terminal transferase, to the site of DNA damage (Kim et al. 2012). In early studies by Kilbey (Njagi and Kilbey 1982) and subsequent studies by our laboratory, induced mutagenesis in S. cerevisiae was reduced in hypomorphic cdc7 mutants including an inactive cdc7 kinase-dead (KD) mutant (Hollingsworth et al. 1992; Ostroff and Sclafani 1995; Pessoa-Brandão and Sclafani 2004). Dbf4-dependent kinase (DDK) is conserved from yeast to humans and is composed of the Cdc7 kinase catalytic and Dbf4 regulatory subunit. Dbf4 is low in G1 phase of the cell cycle and required for kinase activity (Sclafani and Holzen 2007). DDK is a Ser/Thr protein kinase that regulates the initiation of DNA replication by phosphorylating one or more members of the Mcm2–7 DNA helicase (Sheu and Stillman 2006; Randell et al. 2010; Sheu and Stillman 2010; Heller et al. 2011; reviewed in Tanaka and Araki 2013). Either mcm5 (mcm5-bob1) (Hardy et al. 1997) or mcm4 (mcm4D74-174) (Sheu and Stillman 2010) mutations can bypass the requirement for DDK during replication. However, cdc7D mcm5-bob1, dbf4D mcm5-bob1, or cdc7D dbf4D mcm5-bob1 strains are defective in other DDK functions, including the regulation of meiotic recombination (Matos et al. 2008; Wan et al. 2008) and resistance to DNA damaging agents such as UV and MMS (Pessoa-Brandão and Sclafani 2004), suggesting the kinase interacts with and phosphorylates additional substrates that promote mutagenesis and meiotic recombination (Sclafani 2000). In this report, we investigated the role of DDK in mutagenesis and demonstrate DDK is required for spontaneous as well as induced Polz-dependent mutagenesis. Based on biochemical analyses, we suggest DDK regulates mutagenesis by direct interaction with and phosphorylation of the Rev7 subunit of Polz. Our results are important, given the conservation of DDK’s role in TLS in human cells (Day et al. 2010; Yamada et al. 2013), and are significant with regard to human disease because overexpression of yeast DDK increases mutagenesis (Sclafani et al. 1988; Sclafani and Jackson 1994), human DDK overexpression is observed in many types of cancer cells (Hess et al. 1998; Nambiar et al. 2007; Bonte et al. 2008; Clarke et al. 2009; Kulkarni et al. 2009), and DDK is a therapeutic target in cancer patients (Rodriguez-Acebes et al. 2010). Materials and Methods Yeast strains, media, and plasmids Yeast strains and plasmids used are listed in Table 1 and Table 2, respectively. Yeast strains were grown in yeast extract/ peptone/dextrose (YPD) with 2% glucose or in synthetic defined minimal media supplemented with appropriate amino acids and 2% glucose (Pessoa-Brandão and Sclafani 2004). Galactose induction of tagged genes for subsequent immunoprecipitation was as described (Oshiro et al. 1999). Briefly, exponentially growing cells using raffinose as carbon source were induced by the addition galactose to 2% for 2 hr. Because pSF2 has a copper-inducible promoter, pSF2-containing cells were induced by the addition of 0.1 mM CuSO4 for 2 hr. Genetic methods for yeast strain construction, tetrad analysis, and transformation were as previously described (Pessoa-Brandão and Sclafani 2004). Gene deletions marked with the kanMX cassette were constructed by PCR amplification of an appropriate disruption cassette, transformation, and subsequent selection with G418 (Winzeler et al. 1999). The pol30-K127,K164R allele was introduced at the POL30 locus by two-step allele replacement using XbaI-digested plasmid 721 and verified by sequencing genomic DNA. The mcm5-bob1-2 allele contains a CT-to-TC mutation that converts codon 83 from proline to leucine and creates a DdeI site. This allele was introduced into strain SJR1177 by two-step allele replacement using MluI-digested pRAS490, which contains mcm5-bob1-2 in pRS306 (Pessoa-Brandão and Sclafani 2004; Dohrmann and Sclafani 2006). The cdc7D::HIS3 allele was introduced into RSY1183 by one-step allele replacement using a 3-kb MluI–EcoRI fragment from pRS277-cdc7D::HIS3 (Jackson et al. 1993). Presence of the cdc7D::HIS3 allele was verified by restriction digest and by loss of cdc7 genetic complementation. Strain YSS13 contains Myc-tagged Rev7 and was obtained from Marco Muzi-Falconi (University of Milan, Milan, Italy) (Sabbioneda et al. 2005). pLPB60 encodes GST-tagged Rev7 and was constructed by insertion of REV7 into the NcoI/BamHI sites of the pYES263 expression vector (Melcher 2000). Plasmid pLPB46, which inserts a FLAG (Sigma-Aldrich) epitope tag immediately after the initiating methionine of Pol30, was constructed by PCR-overlap mutagenesis (Ho et al. 1989). Plasmid pBL211 was used as template for PCR using outside primers M13Fwd (59-TGTAAAACGACGGCCAGT-39) and M13Rev (59-TCACACAGGAAACAGCTATGAC-39), complementary to the plasmid backbone, and internal primers Pol30-FlagFwd (59-ATGgattacaaggatgacgacgataagTTAGAA GCAAAATTTGAAGAAGC-39) and Pol30-FlagRev (59-cttatcg tcgtcatccttgtaatcCATTTTTTCTCTCTTTTGACTGCG-39; lowercase letters indicate the FLAG epitope tag). The resulting PCR fragment, which includes the FLAG-Pol30 construct and the POL30 promoter, was then digested with XhoI and BamHI and cloned into pRS426. The FLAG-Pol30 gene was functional as plasmid pLPB46 complemented a lethal pol30D::kanMX4 allele. All tagged constructs used in this study (Table 1 and Table 2) have been shown to function normally by genetic complementation: Cdc7-myc9 (Oshiro et al. 1999), HA-Dbf4 (Oshiro et al. 1999), HA-Cdc7 (Hardy and Pautz 1996), GST-Mcm2 (Lei et al. 1997), GST-Rev1 and GST-Rev3 (Nelson et al. 1996), and REV7-myc13 (Sabbioneda et al. 2005). GST-Rev7 produced in this study was tested by complementation of a rev7D::KanMX4 null mutant.