Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ɛ, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae

ARTICLE IN PRESS

G Model MUT-11211; No. of Pages 9

Mutation Research xxx (2012) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Mutation Research/Fundamental and Molecular Mechanisms of Mutagenesis journal homepage: www.elsevier.com/locate/molmut Community address: www.elsevier.com/locate/mutres

Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ␧, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae Joanna Kraszewska 1 , Marta Garbacz 1 , Piotr Jonczyk, Iwona J. Fijalkowska ∗∗ , Malgorzata Jaszczur ∗ Institute of Biochemistry and Biophysics, Polish Academy of Sciences, Pawinskiego 5A, 02-106 Warsaw, Poland

a r t i c l e

i n f o

Article history: Received 24 February 2012 Received in revised form 5 June 2012 Accepted 7 June 2012 Available online xxx Keywords: DNA polymerase epsilon (Pol ␧) DNA polymerase zeta (Pol ␨) Error-prone DNA polymerase Spontaneous mutagenesis DNA replication fidelity

a b s t r a c t The Saccharomyces cerevisiae DNA polymerase epsilon holoenzyme (Pol ␧ HE) is composed of four subunits: Pol2p, Dpb2p, Dpb3p and Dpb4p. The biological functions of Pol2p, the catalytic subunit of Pol ␧, are subject of active investigation, while the role of the other three, noncatalytic subunits, is not well defined. We showed previously that mutations in Dpb2p, a noncatalytic but essential subunit of Pol ␧ HE, influence the fidelity of DNA replication in yeast cells. The strength of the mutator phenotype due to the different dpb2 alleles was inversely proportional to the strength of protein–protein interactions between Pol2p and the mutated forms of Dpb2p. To understand better the mechanisms of the contribution of Dpb2p to the controlling of the level of spontaneous mutagenesis we undertook here a further genetic analysis of the mutator phenotype observed in dpb2 mutants. We demonstrate that the presence of mutated forms of Dpb2p in the cell not only influences the intrinsic fidelity of Pol ␧ but also facilitates more frequent participation of error-prone DNA polymerase zeta (Pol ␨) in DNA replication. The obtained results suggest that the structural integrity of Pol ␧ HE is a crucial contributor to accurate chromosomal DNA replication and, when compromised, favors participation of error prone DNA Pol ␨ in this process. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Experimental data available to date have shown that the fidelity of chromosomal DNA replication is maintained by three highly conserved processes: base selection conducted by DNA polymerases, the 3 → 5 exonucleolytic proofreading activity of the DNA polymerases and the postreplicative mismatch repair system (MMR). The combination of these three processes leads to an average mutation rate in the range of 10−10 per base pair replicated [1]. For years genetic and biochemical studies of the mechanisms determining the fidelity of DNA replication have been mainly concentrated on the MMR system and on the catalytic subunits of prokaryotic and eukaryotic replicative DNA polymerase holoenzymes. However, recently the contributions of at least two other elements that may influence the final fidelity of DNA replication and the level of spontaneous mutagenesis are being intensely investigated. These are noncatalytic subunits of

∗ Corresponding author. Present address: Department of Biological Sciences, Molecular and Computational Biology Section, University of Southern California, 1050 Childs Way, RRI 113 University Park, Los Angeles, CA 90089-2910, United States. Tel.: +1 213 740 5191; fax: +1 213 821 1138. ∗∗ Corresponding author. Tel.: +48 22 592 1113; fax: +48 22 592 2190. E-mail addresses: [email protected] (I.J. Fijalkowska), [email protected] (M. Jaszczur). 1 These authors contributed equally to this work.

DNA polymerase holoenzymes and translesion synthesis (TLS) DNA polymerases. Despite the recent progress in the understanding of the structure of multisubunit replicative polymerase holoenzymes, the exact physiological role of their accessory subunits in eukaryotic cells remains unknown. Several genetic studies have suggested that noncatalytic subunits of DNA polymerases may influence the fidelity of DNA replication. DNA polymerase III holoenzyme (Pol III HE), the major replicative DNA polymerase in Escherichia coli, is a large complex of 17 subunits [2,3]. It has been shown that some mutations in the dnaX gene which encodes the ␶ subunit of Pol III HE lead to a mutator phenotype [4–6]. Also in Saccharomyces cerevisiae the accessory subunits of the main replicases have been demonstrated to affect the fidelity of DNA replication. Three DNA polymerase holoenzymes are involved in chromosomal DNA replication in S. cerevisiae [7]. DNA polymerase alpha holoenzyme (Pol ␣ HE) comprises of four subunits – Pol1p, Pol12p, Pri1p and Pri2p, DNA polymerase epsilon holoenzyme (Pol ␧ HE) is also a heterotetramer composed of Pol2p, Dpb2p, Dpb3p and Dpb4p, and DNA polymerase delta holoenzyme (Pol ␦ HE) is a heterotrimer of Pol3p, Pol31p and Pol32p. Mutations in the PRI1 and PRI2 genes, encoding the respective subunits of Pol ␣ HE increase the spontaneous mutation rate [8]. Moreover, S. cerevisiae strains lacking Pol32p, a nonessential subunit of Pol ␦ HE, exhibit an increased frequency of deletions of sequences flanked by short direct repeats [9]. Genetic data strongly suggest that the accessory subunits of S. cerevisiae Pol ␧ HE, Dpb2p, Dpb3p and Dpb4p, contribute to the

0027-5107/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.mrfmmm.2012.06.002

Please cite this article in press as: J. Kraszewska, et al., Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ␧, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae, Mutat. Res.: Fundam. Mol. Mech. Mutagen. (2012), http://dx.doi.org/10.1016/j.mrfmmm.2012.06.002

G Model MUT-11211; No. of Pages 9 2

ARTICLE IN PRESS J. Kraszewska et al. / Mutation Research xxx (2012) xxx–xxx

high fidelity of DNA replication. Earlier work has demonstrated that a dpb3 mutant displays a mutator phenotype [10,11]. Also Aksenonova et al. [12] have shown that deletion of the DPB3 and DPB4 genes encoding two nonessential subunits of Pol ␧ HE elevates spontaneous frameshift and base substitution rates in vivo. Recently, we have characterized several temperature-sensitive dpb2 alleles. Some of them caused a very strong mutator effect [13,14]. We also found that an impaired interaction between the catalytic Pol2 subunit and Dpb2p was responsible for the observed mutator phenotype in S. cerevisiae cells carrying mutated alleles of the DPB2 gene. Moreover, we observed that the strength of the mutator phenotype for the different dpb2 alleles was inversely proportional to the strength of the protein–protein interaction between Pol2p and the mutated forms of Dpb2p. Those results suggested that the structural integrity of Pol ␧ HE is an important contributor to the accurate chromosomal DNA replication. Dpb2p, by stabilizing the holoenzyme, could prevent DNA replication errors indirectly by at least two mechanisms: by influencing the nucleotide insertion fidelity of Pol ␧ HE and/or by affecting the integrity and stability of replisom. Which in consequence could influence the processing of mismatched primer terminus by participation of low-fidelity TLS DNA polymerases, such as DNA polymerase zeta (Pol ␨). DNA Pol ␨ is an error-prone enzyme belonging to so called TLS polymerases [15–18]. Prokaryotic and eukaryotic cells contain several such DNA polymerases, e.g., Pol IV and Pol V in E. coli or Pol ␩ (eta), Pol ␨ and Rev1 in S. cerevisiae. Pol ␨ from S. cerevisiae consists of two subunits: Rev3p, the catalytic subunit, and Rev7p, an accessory protein which enhances the catalytic activity of the polymerase [19]. Rev3p is a member of DNA polymerase family B. It lacks a 3 → 5 exonuclease activity. Genetic studies in S. cerevisiae have shown that Pol ␨ is required for mutagenesis induced by UV light and by other exogenous genotoxic agents [19,20]. Despite being considered a TLS polymerase Pol ␨ alone has in fact a very limited ability to bypass DNA lesions due to its low efficiency of inserting nucleotides opposite a lesion site. However, Pol ␨ is very proficient in extending from nucleotides inserted opposite a lesion by other DNA polymerases and in extending primers containing a mismatched terminal nucleotide on undamaged DNA [20–25]. The mismatch extension activity of Pol ␨ is stimulated by the deoxycytidyl transferase Rev1 protein [26] and by the DNA polymerase processivity factor – proliferating cell nuclear antigen (PCNA) [11,27]. The mechanisms by which TLS polymerases, such as Pol ␨, access undamaged DNA at the replication fork are recently becoming of interest. In prokaryotes, in actively dividing bacteria, lack of Pol IV does not significantly affect the level of spontaneous mutations, indicating that Pol IV has only a limited access to the replication fork [28,29]. In contrast, in dividing yeast cells Pol ␨ is responsible for 30–60% of spontaneous mutations [30–35]. Recently, it has been shown that in E. coli and in yeast strains carrying mutations in replicative DNA polymerases [5,11,12,36,37] the involvement of Pol IV and Pol ␨ in spontaneous mutagenesis is increased, respectively. Here we show that the presence of mutations in Dpb2 subunit increases the involvement of Pol ␨ in DNA replication in S. cerevisiae cells. This conclusion is based on the results demonstrating that (i) in the dpb2 mutants the spontaneous mutagenesis is partially Pol ␨-dependent, (ii) the spectrum of mutations observed in the dpb2 mutants confirms the involvement of Pol ␨ in replication, (iii) in the pol30K164R background, where the Pol ␨ participation in DNA replication is reduced, the observed dpb2-dependent mutagenesis is decreased. Interestingly, even when the activity of Pol ␨ is abolished in the dpb2 mutant strains the observed mutator effects are still significantly elevated comparing to DPB2 and DPB2 rev3 strains. These results indicate that other factors, besides Pol ␨, for example decreased Pol ␧ insertion fidelity, are responsible for

the dpb2 mutator effect. The obtained results suggest that Dpb2p affects at least two important fidelity factors: participation of Pol ␨ in the replication fork, and fidelity of Pol ␧ HE. 2. Materials and methods 2.1. Media and growth conditions S. cerevisiae strains and their genotypes are listed in Table 1. Strains were grown in standard media [41]. Yeast strains were grown in nonselective YPD medium (yeast complete medium), containing 1% yeast extract, 1% peptone and 2% glucose. SD minimal medium (0.67% yeast nitrogen base without amino acids, 3% glucose) supplemented with appropriate l-amino acids and nucleotides was used for yeast transformation and mutagenesis assays. The frequency of forward mutations at the CAN1 locus was determined by using SD plates additionally supplemented with lcanavanine (60 mg/l), an analog of arginine. Propagation of plasmids was performed in E. coli strains DH5␣ or MG7. 2.2. Construction of the cassette for disruption of chromosomal DPB2 The cassette for disruption of the DPB2 gene was isolated from BY4743 dpb2::kanMX4/DPB2 yeast cells (EUROSCARF accession no. Y25590) by the gaprepair method using EcoRI/StuI-linearized pKF106 [13]. Homologous recombination took place in both the promoter and terminator sequences of DPB2 giving two kinds of plasmids depending on which allele was rescued. The plasmid containing the PDPB2 -kanMX4-TDPB2 cassette was named pKF111, while one bearing the wild-type gene was named pKF112. The alleles obtained were confirmed by sequencing. 2.3. Construction of the cassette for disruption of chromosomal lys2-801 To construct the disruption cassette lys2::hph, the hph gene was PCR-amplified with primers: 5 CACTCGCGACACCTTATGTATCATACAC3 (NruI site underlined) and 5 GTCTCGCGACAATTAATACGACTCACT3 (NruI site underlined) and pAG32 as the template (EUROSCARF, the plasmid is part of the “DEL-MARKER-SET” accession no. P30106). The obtained PCR product (1879 bp) was NruI-digested and ligated with NruI-digested pRS317 [42]. The clone bearing the hph gene in the orientation opposite to that of the LYS2 sequence was chosen, giving the pJK20 plasmid. 2.4. Construction of the pJK19 vector used for plasmid shuffling Construction of the pJK19 vector was a few steps procedure. First, the 1995-bp Ecl136II fragment of pMG29 (DPB2 under heterologous promoter PMET25 and terminator TCYC1 ) was subcloned into Ecl136II-linearized pRS317 [42]. Then using primers: 5 ACAGCAGCAATCCAGCAGTATGAGTATCAC3 and 5 AAGCGGCCGCAATACGACTCACTATAGG3 (NotI site underlined) the 1734-bp fragment of pMG29 was PCR-amplified. Finally, the PCR product (confirmed by DNA sequencing) was NotI/EcoNI-digested and cloned into NotI/EcoNI-digested vector in the first step, yielding pJK19 (LYS2 DPB2). 2.5. Introduction of the DPB2 alleles into the POL30 mutants Strains are haploid derivatives of wild-type DF5 [43] and were modified by H.D. Urlich to MAT˛ his3-200 leu2-3,2-112 lys2-801 trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30wt::LEU2 and MAT˛ his3-200 leu2-3,2-112 lys2-801 trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30(K164R)::LEU2. To introduce the DPB2 alleles with the plasmid shuffle technique into POL30 and pol30K164R strains the LYS2 marker was used. To prevent reversion of the chromosomal lys2-801 allele to the wild-type allele, the chromosomal lys2-801 allele was disrupted by a lys2::hph cassette obtained from digestion pJK20 with PflMI/StuI. Both strains mentioned above (POL30 and pol30K164R) were transformed with the obtained 4211-bp PflMI/StuI fragment carrying lys2::hph cassette, incubated at 30 ◦ C for up to 5 days on selective hygromycin B (300 ␮g/1 ml YPD) plates and then duplicated onto SD-Lys− plates to confirm their Lys− phenotype. Transformants were incubated at 30 ◦ C for up to 5 days on selective hygromycin B (300 ␮g/1 ml YPD) plates. Obtained colonies were duplicated onto SD-Lys− plates to confirm their Lys− phenotype. To confirm this integration genetically, genomic DNA of the created lys2::hph POL30 and lys2::hph pol30K164R strains was isolated and used as a template in PCR reactions to amplify the LYS2 locus with primers 5 ACCGATGGCTGTGTAGAAGT3 and 5 GGTCTGGATAGAGAAGTTGG3 . The presence of a 1215-bp PCR product confirmed integration of the lys2::hph cassette into the chromosomal lys2-801 allele. Next, in a few step procedure the chromosomal wild-type DPB2 genes were disrupted and the DPB2 alleles were introduced on centromeric plasmids to the obtained strains. DPB2 gene is essential so the first step of the disruption of chromosomal DPB2 was transformation of pJK19 (LYS2 DPB2) to the lys2::hph strains and transformants incubation at 30 ◦ C for up to 3 days on plates selective for lysine prototrophy. Then the PDPB2 -kanMX4-TDPB2 cassette was excised from pKF111 as a 3.1-kb SacII/Bsp120I fragment and used for yeast transformation. Transformants were incubated at 30 ◦ C for up to 5 days on selective G418 (200 ␮g/ml YPD) plates. To finally confirm the integration of the PDPB2 -kanMX4TDPB2 cassette into the chromosomal DPB2 locus, genomic DNA of the constructed

Please cite this article in press as: J. Kraszewska, et al., Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ␧, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae, Mutat. Res.: Fundam. Mol. Mech. Mutagen. (2012), http://dx.doi.org/10.1016/j.mrfmmm.2012.06.002

G Model MUT-11211; No. of Pages 9

ARTICLE IN PRESS J. Kraszewska et al. / Mutation Research xxx (2012) xxx–xxx

3

Table 1 Yeast strains used in this study. Strain

Genotype

Source

CD138 SC696

MATa, his7, leu2, lys1, ura3, ogg1::TRP1 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 DPB2::CaURA3 ogg1::TRP1 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-100::CaURA3 ogg1::TRP1 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-103::CaURA3 ogg1::TRP1 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2- GG2899-2900 CAN1 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2- GG2899-2900 CAN1 pol2-4 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 DPB2::CaURA3

[38] This study

SC708 SC709 I(-2)I-7B-YUNI300 I(-2)I-7B-YUNI300 pol2-4 SC91

SC94

MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 DPB2::CaURA3

SC99

MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-101::CaURA3 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 dpb2-101::CaURA3 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-100::CaURA3 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 dpb2-100::CaURA3 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-103::CaURA3 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 dpb2-103::CaURA3 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 DPB2::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 DPB2::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-101::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 dpb2-101::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-100::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 dpb2-100::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 dpb2-103::CaURA3 rev3::LEU2 MATa trp1-289 his7-2 leu2-::kanMX4 ura3- ade2-1 lys2-GG2899-2900 CAN1 pol2-4 dpb2-103::CaURA3 rev3::LEU2 MAT˛ his3-200 leu2-3,2-112 lys2-801 trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30wt::LEU2 MAT˛ his3-200 leu2-3,2-112 lys2-801 trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30(K164R)::LEU2 MAT˛ his3-200 leu2-3,2-112 lys2::hph trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30wt::LEU2 MAT˛ his3-200 leu2-3,2-112 lys2::hph trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30(K164R)::LEU2 MAT˛ his3-200 leu2-3,2-112 lys2::hph trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30wt::LEU2 dpb2::kanMX4 [pGJ2 (DPB2 HIS3)] MAT˛ his3-200 leu2-3,2-112 lys2::hph trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30wt::LEU2 dpb2::kanMX4 [pMJ103 (dpb2-103, HIS3)] MAT˛ his3-200 leu2-3,2-112 lys2::hph trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30(K164R)::LEU2 dpb2::kanMX4 [pGJ2 (DPB2 HIS3)] MAT˛ his3-200 leu2-3,2-112 lys2::hph trp1-1 (am) ura3-52 pol30::URA3 YIp128-P30-POL30(K164R)::LEU2 dpb2::kanMX4 [pMJ103 (dpb2-103, HIS3)] MAT˛ his31 leu20 lys20 ura30

SC101 SC146 SC148 SC152 SC154 SC377 SC378 SC379 SC380 SC381 SC382 SC383 SC384

SC705 SC706 SC731 SC732 SC733 SC734 BY4742

dpb2::kanMX4 strains was isolated and used as a template in PCR reactions with pairs of primers: 5 GAATACTGGCTTACCGAG3 with 5 CGTATGTGAATGCTGGTC3 and 5 CACCGACTGCAACAGA3 with 5 GTCAAGACTGTCAAGGAG3 . The presence of the 1083-bp and 798-bp products, respectively, indicated that the PDPB2 -kanMX4-TDPB2 cassette had replaced the chromosomal DPB2 gene. We also excluded the presence of WT DPB2 allele in constructed strain by PCR reaction with primers: DPB2 UP: 5 CACCGACTGCAACAGA3 and DPB2 DOWN: 5 GAATACTGGCTTACCGAG3 . The 3628-bp PCR product was detected only in strains carrying DPB2 allele. Introduction of the DPB2 alleles into the obtained lys2::hph dpb2::kanMX4 POL30 [pJK19 (LYS2 DPB2)] and lys2::hph dpb2::kanMX4 pol30K164R [pJK19 (LYS2 DPB2)] strains was performed with the plasmid shuffle technique with aminoadipate (␣AA) [44]. First, the strains were transformed with plasmids pGJ2 (pRS313 HIS3 DPB2) or pMJ103 (pRS313 HIS3 dpb2-103) [13] and selected for histidine prototrophy on plates

This study This study [39] Pavlov Y. [13] derivative of I(-2)I-7BYUNI300 [13] derivative of I(-2)I-7BYUNI300 pol2-4 [13] [13] [13] [13] [13] [13] This study This study This study This study This study This study This study This study Ulrich H. D. Ulrich H. D. This study This study This study This study This study This study [40]

incubated at 23 ◦ C for up to 5 days. Then the His+ transformants were isolated and toothpicked twice at 23 ◦ C onto plates additionally containing lysine and ␣AA (a selective agent against Lys+ cells) to remove pJK19 bearing the wild-type LYS2 and DPB2 genes. Obtained colonies were duplicated onto new SD-Lys− plates to confirm Lys− phenotype. Additionally, the temperature-sensitive phenotype was confirmed for strains with the dpb2-103 allele. 2.6. Construction of the cassette for disruption of chromosomal REV3 To delete the REV3 gene encoding the catalytic subunit of Pol ␨, plasmids carrying the rev3::CaURA3 and rev3::LEU2 disruption cassettes were constructed. First, 5 and 3 REV3-flanking sequences (upstream of the ATG codon and downstream of the TAA codon) were PCR-amplified and cloned into pRS314 [45]. BY4742 chromosomal

Please cite this article in press as: J. Kraszewska, et al., Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ␧, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae, Mutat. Res.: Fundam. Mol. Mech. Mutagen. (2012), http://dx.doi.org/10.1016/j.mrfmmm.2012.06.002

G Model MUT-11211; No. of Pages 9

ARTICLE IN PRESS J. Kraszewska et al. / Mutation Research xxx (2012) xxx–xxx

4

DNA was used as a template for both PCR reactions [40]. The 5 flanking region was amplified with primers: 5 CAGGTCGACATGCTCTCCATACCGCTATTG3 (SalI site underlined) and 5 CCGCTGCAGCATTTCCAAACATTTTATTTCGCC3 (PstI site underlined). The obtained PCR product (1159 bp) was SalI/PstI-digested and ligated with SalI/PstI-digested pRS314, giving the pUPrev3 plasmid. The 3 REV3-flanking region was amplified using primers 5 CTGCCCGGGTAATCTAGACACAGATATGTCTCGC3 (SmaI site underlined) and 5 CTTGAGCTCGACATAACACCGCAGGAAGAA3 (SacI site underlined). The resulting PCR product (962 bp) was SmaI/SacI-digested and cloned into SmaI/SacI-digested pUPrev3. The resulting plasmid, named pUPDWrev3, contained 5 and 3 REV3-flanking regions. Next, a DNA fragment containing the CaURA3 gene (URA3 from Candida albicans) was cloned between the 5 and 3 REV3-flanking regions. The CaURA3-containing fragment was PCR-amplified using primers (PstI site underlined) and 5 CGACTGCAGAGAACGTGGACTCCAAC3 5 GATCCCGGGCCCAACAGTTGCGCAG3 (SmaI site underlined) and pKF117 as a template [13]. The PCR product (1798 bp) was PstI/SmaI-digested and cloned into PstI/SmaI-digested pUPDWrev3. The resulting plasmid, named pGJ100, contains the rev3::CaURA3 cassette for replacement of the chromosomal REV3 gene. To construct the rev3::LEU2 disruption cassette, DNA fragment containing the LEU2 gene was PCR-amplified using primers 5 CGGCTGCAGTCGGTGATGACGGTGAAAAC3 (PstI site underlined) and 5 ATTCCCGGGAAAAACTTGATTAGGGTGATGG3 (SmaI site underlined) and pRS315 (LEU2) as a template. The resulting PCR product (2722 bp) was used to replace the CaURA3 marker in pGJ100. pGJ100 was digested with PstI/SmaI and the fragment containing the CaURA3 gene was replaced with PstI/SmaI-digested fragment containing the LEU2 gene. The resulting plasmid, named pGJ101, contains the rev3::LEU2 disruption cassette. All PCR products were confirmed by DNA sequencing. To isolate the rev3::CaURA3 (3922 bp) and rev3::LEU2 (4843 bp) disruption cassettes from pGJ100 and pGJ101, respectively, the plasmids were digested with SalI/SacI and appropriate DNA fragments containing the cassettes were isolated.

nucleotides. Colonies were usually taken from two to three independent isolates of each strain. The cultures were grown at 23 ◦ C to stationary phase. Next, yeast cells were collected by centrifugation, washed and resuspended in water. Undiluted suspensions were plated on selective plates supplemented with l-canavanine whereas appropriate dilutions were plated on non-selective plates. All plates were incubated for 7–10 days at 23 ◦ C and colonies were counted. The frequency of forward mutations was measured at the CAN1 locus. Each experiment was repeated at least 3 times. Tables show results from the representative experiments. Mutant frequency was determined by dividing the mutant count by the total cell count. The median value of mutation rate and the 95% confidence limits were used to compare spontaneous mutagenesis in different strains. The mutation rates were calculated as described previously [13,47] and the p-value was determined using Statistica 6.0 software (nonparametric Mann–Whitney criterion applied to the mutant yield distributions of 10–20 independent cultures for the two compared strains based on the result of individual experiment [48]). 2.10. CanR mutation spectra Cultures from single independent colonies of each strain were grown at 23 ◦ C to stationary phase on SD minimal medium supplemented with appropriate l-amino acids and nucleotides. Each culture was plated on lcanavanine-supplemented plates and incubated at 23 ◦ C. The frequency of forward mutations at the CAN1 locus was determined as above. Then a single colony was chosen randomly from each plate. Next, chromosomal DNA was isolated from each chosen CanR colony (as described by Gietz and Woods [46]) and the CAN1 locus was amplified using primers 5 AAGAGTGGTTGCGAACAGAG3 and 5 GGAGCAAGATTGTTGTGGTG3 . Sequencing reactions were performed using primers 5 ATATTTGACAGGGAACAAGT3 , 5 GATGGCTCTTGGAACGGA3 , 5 TCGTTACTGCTGCATTTG3 and 5 CTGATGTGCGAGATTGAG3 , 5 CAAAGGTTTTGCCACATATC3 . Mutation distribution and occurrences in strains were compared using chi-square analysis.

2.7. Integration into the REV3 locus To replace the chromosomal REV3 gene with the rev3::CaURA3 or rev3::LEU2 disruption cassettes, the dpb2 temperature sensitive strains and DPB2 control strains were transformed with the mentioned cassettes (as described by Gietz and Woods [46]). Transformants were selected for uracil or leucine prototrophy on plates incubated at 23 ◦ C for up to 10 days. The colonies were replicated onto new SD-Ura or SD-Leu plates. The ability to grow without uracil or leucine suggested that the rev3::CaURA3 or rev3::LEU2 cassette, respectively, was integrated into the REV3 locus replacing the REV3 gene. To finally confirm the integration of the rev3::CaURA3 cassette into the REV3 locus, genomic DNA of the constructed dpb2 rev3::CaURA3 and DPB2 rev3::CaURA3 strains was isolated and used as a template in PCR reactions with primers 5 GCAAGTGTAGCGGTCACG3 and 5 GTGCGCGGATATAAAGGAG3 . These primers recognize a sequence within the CaURA3-containing fragment from the rev3::CaURA3 disruption cassette and a sequence downstream of the REV3 gene. The presence of the 1223-bp PCR product indicated that the rev3::CaURA3 cassette had replaced the REV3 gene and the resulting strains were rev3. Similarly, integration of the rev3::LEU2 cassette into the REV3 locus was confirmed using primers 5 AGATAGGGTTGAGTGTT3 and 5 GTGCGCGGATATAAAGGAG3 . The presence of the 1234-bp product indicated that the rev3::LEU2 cassette had replaced the chromosomal REV3 gene. To confirm the presence or lack of WT REV3 in the constructed strains the following primers were used: Rev3-up: 5 GATAAGTATTCACTAACACC3 and Rev3-down: 5 CTTAGAGGATACGAAGATTC3 . The 4912-bp PCR product was detected only in strains carrying REV3 allele. 2.8. Disruption of the chromosomal OGG1 gene To disrupt the OGG1 gene, an ogg1::TRP1 cassette was used. To obtain this cassette a fragment of chromosomal DNA from strain CD138 [38] containing ogg1::TRP1 was PCR-amplified using primers 5 CACCAGTTTTCTCGCGG3 and 5 CTTTCTCCACAAGGCAT3 . To replace the chromosomal OGG1 gene, the dpb2 strains and the DPB2 control strain derivative of I(-2)I-7B-YUNI300 [13] were transformed with the isolated ogg1::TRP1 cassette (as described by Gietz and Woods [46]). Transformants were incubated at 23 ◦ C for up to 10 days on plates selective for tryptophan prototrophy. Then, the obtained colonies were replicated onto new SD-Trp− plates. The ability to grow on medium without tryptophan indicated that the ogg1::TRP1 cassette had replaced the OGG1 gene. To confirm this integration, genomic DNA of the created dpb2 ogg1::TRP1 and DPB2 ogg1::TRP1 strains was isolated and used as a template in PCR reactions to amplify the OGG1 locus with the above mentioned primers. The 275-bp PCR product indicated the presence of WT OGG1 while the presence of the single 976-bp PCR product indicated that replacement of the chromosomal OGG1 gene with the ogg1::TRP1 cassette. 2.9. Measurement of spontaneous mutation frequency and calculation of mutation rates To measure the spontaneous reversion rates of forward mutation to canavanine resistance (CanR), 10–20 yeast cultures of particular strain were started from single colonies in liquid SD medium supplemented with required amino acids and

3. Results 3.1. The dpb2 mutator effect is partially dependent on Pol ␨ activity We have previously shown that yeast strains carrying various mutated alleles of dpb2 express a mutator phenotype [13,14]. Our further analysis has revealed that errors arising in the dpb2 mutants are replication errors corrected by both the mismatch repair and the 3 → 5 exonuclease proofreading activity. The dpb2 mutations impaired the interaction between the catalytic Pol2p and the mutated Dpb2 subunits of Pol ␧ and cause temperature-sensitivity. Interestingly, we observed a negative correlation between the strength of the dpb2 mutator phenotype and the relative strength of the Dpb2p–Pol2p interaction. We hypothesized that impaired interactions between Pol ␧ HE subunits could destabilize the entire holoenzyme at the replication fork and thus could allow other DNA polymerases to participate in DNA replication. To check if the dpb2-dependent mutator effect is entirely or partially Pol ␨-dependent we inactivated Pol ␨ in the strains carrying three previously characterized dpb2 mutant alleles (dpb2-100, dpb2-101, dpb2-103) [13]. The dpb2 alleles were integrated into the chromosome of strain I(-2)I-7B-YUNI300 and an isogenic strain with a REV3 deletion (rev3) and assayed for forward mutation rate at the CAN1 locus (Table 2). Because any mutation that inactivates the arginine permease encoded by CAN1 results in the CanR phenotype, this assay allows measuring of a wide spectrum of mutations like any of the possible base substitutions, frameshifts and other classes of mutations [49,50]. As shown in Table 2, the mutation rate at the CAN1 locus was two-fold lower in the rev3 strain than in the wild-type one, reflecting the spontaneous antimutator phenotype of the rev3 mutation. This indicates that ∼50% of spontaneous mutations is probably due to the participation of Pol ␨ in the error-prone bypass of the endogenous DNA lesions present at physiological level and from error-prone copying of undamaged DNA [11,33,35]. The mutation rates in the dpb2 mutants were elevated 10-fold, 3-fold and 8-fold for dpb2-100, dpb-101 and dpb2-103, respectively, comparing to the DPB2 strain (Table 2). Interestingly, deletion of the REV3 gene in strains carrying the dpb2

Please cite this article in press as: J. Kraszewska, et al., Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ␧, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae, Mutat. Res.: Fundam. Mol. Mech. Mutagen. (2012), http://dx.doi.org/10.1016/j.mrfmmm.2012.06.002

ARTICLE IN PRESS

G Model MUT-11211; No. of Pages 9

J. Kraszewska et al. / Mutation Research xxx (2012) xxx–xxx Table 2 Effect of REV3 disruption on mutation rates in haploid strains with different DPB2 alleles. Straina

dpb2-100 strain, 6 × 10−7 in the dpb2-101 strain and 18 × 10−7 in dpb2-103 strain (Table 2). Therefore, the dpb2 strains are 3–10-fold mutators even when Pol ␨ is inactive.

Mutation rates CanR × 10−7

POL2 DPB2 POL2 DPB2 rev3 POL2 dpb2-100 POL2 dpb2-100 rev3 POL2 dpb2-101 POL2 dpb2-101 rev3 POL2 dpb2-103 POL2 dpb2-103 rev3 pol2-4 DPB2 pol2-4 DPB2 rev3 pol2-4 dpb2-100 pol2-4 dpb2-100 rev3 pol2-4 dpb2-101 pol2-4 dpb2-101 rev3 pol2-4 dpb2-103 pol2-4 dpb2-103 rev3

5

4 (3–6)b 2 (1–2) 41 (36–49) 21 (18–24) 12 (11–14) 8 (6–11) 32 (27–37) 20 (16–23) 21 (17–27) 24 (18–33) 305 (247–364) 119 (102–159) 141 (133–153) 67 (56–81) 196 (134–279) 108 (86–153)

Fold increasec

p-valued

1 0.5 10 5 3 2 8 5 5 5 76 30 35 17 49 27

0.001033 0.000931 0.021685 0.002437 0.713192 0.001460 0.002415 0.021685

a The dpb2 alleles were integrated into the DPB2 chromosomal locus of I(-2)I7B-YUNI300 [13]. b 95% confidence limits are shown in parentheses. c The fold increase – mutation rate in dpb2 strains divided by mutation rate in wild-type strain (DPB2). d The p-value for REV3 vs. isogenic rev3 strain calculated as described in Section 2.

mutator alleles partially decreased the rate of CanR spontaneous mutagenesis. The REV3 deletion in the dpb2-100 mutated strain entailed 2-fold decrease in CanR mutagenesis rates from the level of 41 × 10−7 to 21 × 10−7 (Table 2). In the case of dpb2-103 rev3 we observed decrease in CanR mutagenesis rate from 32 × 10−7 to 20 × 10−7 (Table 2). Interestingly, the strongest effects of the rev3 were observed for those strains carrying the Dpb2p mutated variants (Dpb2p-100 and Dpb2p-103) which had significantly impaired interaction with Pol2p [13]. To calculate the rate of mutations that are specifically dependent on Pol ␨ we subtracted the mutation rate observed in Pol ␨-deficient strain from the mutation rate seen in the presence of Pol ␨. The observed reduction in the rate of mutagenesis in the strains with inactivated Pol ␨ is 2 × 10−7 (4 × 10−7 minus 2 × 10−7 ) in the DPB2 strain but 20 × 10−7 in dpb2-100 strain (Table 2). To rule out the possibility that the contribution of Pol ␨ to mutagenesis in the dpb2 strains varied depending on their genetic background, we tested the Pol ␨-dependence not only in the I(2)I-7B-YUNI300 background but also in BY4743. Similar results were obtained in both backgrounds (data not shown). Taken together, our results suggest that in the dpb2 mutator strains Pol ␨ participates in DNA replication more often than it does in the DPB2 strain. It is possible that the increased contribution of Pol ␨ in the dpb2-dependent mutagenesis reflects the phenomenon described as defective-replisome-induced mutagenesis (DRIM). DRIM represents Pol ␨-dependent error-prone DNA synthesis and/or extension of mismatches generated by the defective replisome [11,35,37,51,52]. However, even when the activity of Pol ␨ is abolished in the dpb2 mutant strains, they still show significantly elevated mutation rates relative to that seen in the isogenic DPB2 strain (DPB2 rev3 versus dpb2 rev3, p-value < 0.05). This indicates that other factors besides Pol ␨, for example a decreased Pol ␧ insertion fidelity, are responsible for the dpb2 mutator effect. To calculate the rate of mutations attributable to the mutator forms of Pol ␧ HE we subtracted the mutation rate observed in the DPB2 rev3 strain from the mutation rates seen in the dpb2 rev3 strains. The observed increase in the rate of mutagenesis is 19 × 10−7 (21 × 10−7 − 2 × 10−7 ) in the

3.2. Mutated Dpb2p variants influence DNA Pol ␧ insertion fidelity Deletion of the REV3 gene decreases the mutation rates by 30–50% in the dpb2 mutants when frequency of CanR mutants measured. Interestingly, in the dpb2 rev3 strains the frequency of CanR mutants is increased by 3–10-fold comparing to the DPB2 rev3 strain (Table 2). This means that a significant portion of mutations are not due to Pol ␨ participation in DNA replication. Therefore, the question arises about the nature of the remaining 50–70% of mutations. To check whether these remaining mutations arise as DNA Pol ␧ replication errors we attempted to increase the frequency of mutations produced by this polymerase. We used the pol2-4 allele that inactivates the intrinsic 3 → 5 exonuclease activity of DNA Pol ␧ [53,54]. Mutations that abolish the 3 → 5 exonuclease proofreading activity lead to the mutator phenotype by increasing the level of uncorrected replication errors. As shown in Table 2, strains carrying the pol2-4 mutation have 5-fold elevated rates of CanR mutagenesis. Previously, and in this paper, we have shown that mutator effects of the dpb2 and pol2-4 mutations are multiplicative (the error rate in the double pol2-4 dpb2 mutant is higher than the sum of error rates in the two single mutants), which suggests that errors arising in dpb2 strains are corrected by the proofreading activity of Pol ␧ [13]. Thus the pol2-4 mutation alone increases the frequency of CanR mutagenesis 5-fold, dpb2-100 alone 10-fold, while the double pol2-4 dpb2-100 strain exhibits 76-fold increased rate of CanR formation relative to the POL2 DPB2 strain (see Table 2). A similar multiplicative increase of CanR mutagenesis was observed for the other two pol2-4 dpb2 strains tested. To determine if the dpb2 mutations influence the Pol ␧ insertion fidelity, we increased the frequency of Pol ␧-dependent mutations by introducing the pol2-4 allele into the dpb2 mutants and then we abolished the Pol ␨ activity (by deleting the REV3 gene). The pol2-4 dpb2 rev3 triple mutants showed a partial decrease of mutagenesis toward the CanR phenotype comparing to the pol2-4 dpb2 double mutants, similarly as it was observed in the POL2 background. Additionally, the triple mutants exhibited a multiplicative mutator effect relative to the dpb2 rev3 and pol2-4 rev3 double mutants. For example, for the pol2-4 dpb2-100 rev3 strain we observed a 2.6-fold decrease in the rate of CanR mutagenesis, to the level of 119 × 10−7 , relative to the 305 × 10−7 observed for pol2-4 dpb2-100. Similar partial Pol ␨-dependence was observed for the pol2-4 dpb2-101 and pol2-4 dpb2-103 strains. The obtained results suggest that Dpb2p influences at least two important fidelity factors: the participation of the error-prone Pol ␨ in the replication fork, and the insertion fidelity of Pol ␧ HE. 3.3. pol30-K164R mutation weakens the spontaneous dpb2-dependent mutator phenotype The DNA processivity clamp PCNA is an important regulatory target in the recruitment of Pol ␨ to the replication fork [55–58]. It has been demonstrated that both monoubiquitination and sumoylation of PCNA at Lys164 is important for the Pol ␨-dependent increase of spontaneous mutagenesis in S. cerevisiae strains carrying mutations in catalytic subunits of Pol ␧ or Pol ␦ [11]. To check whether the observed partial Pol ␨-dependence of the mutator phenotype of the dpb2 mutants is connected with modification of PCNA at Lys164, we replaced the wild-type POL30 gene encoding PCNA with a mutant pol30-K164R allele in the dpb2-103 genetic background. The dpb2-103 allele was chosen as a moderate mutator.

Please cite this article in press as: J. Kraszewska, et al., Defect of Dpb2p, a noncatalytic subunit of DNA polymerase ␧, promotes error prone replication of undamaged chromosomal DNA in Saccharomyces cerevisiae, Mutat. Res.: Fundam. Mol. Mech. Mutagen. (2012), http://dx.doi.org/10.1016/j.mrfmmm.2012.06.002

G Model MUT-11211; No. of Pages 9

ARTICLE IN PRESS J. Kraszewska et al. / Mutation Research xxx (2012) xxx–xxx

6

Table 3 Effect of pol30-K164R mutation in haploid DPB2 and dpb2-103 strains. Relevant genotypea

Mutation rates CanR × 10−7

DPB2 DPB2 pol30-K164R dpb2-103 dpb2-103 pol30-K164R

8 (7–9)b 8 (7–8) 18 (17–22)* d 12 (11–14)* d

Table 4 Spontaneous mutagenesis rate in CAN1 locus. Types of mutations

Fold increasec 1 1 2.25 1.5

a Strains (SC731, SC732, SC733, SC734, Table 1) with DPB2 or dpb2-103 allele present on centromeric plasmid (pGJ2 or pMJ103 [13]). b 95% confidence limits are shown in parentheses. c The fold increase – mutation rate in DPB2 pol30-K164R or dpb2-103 strains divided by mutation rate in wild type strain (DPB2). d Statistically significant differences are indicated by asterisk. The p-value is