Studies on DNA Replication, Repair, and Mutagenesis in Eukaryotic and Prokaryotic Cells

Roger Woodgate, PhD, Head, Section on DNA Replication, Repair and Mutagenesis
Ekaterina Chumakov, PhD, Staff Scientist
Alexandra Vaisman, PhD, Senior Research Fellow
Mary McLenigan, BS, Chemist
John McDonald, PhD, Biologist
Elena Curti, PhD, Visiting Fellow
Kiyonobu Karata, PhD, Visiting Fellow
Wojciech Kuban, PhD, Visiting Fellow
Eiji Ohashi, PhD, Visiting Fellow
Tara Howard, BSc, Predoctoral Fellow
Catherine Theisen, Student Fellow

Under optimal conditions, the fidelity of DNA replication is extremely high. Indeed, it is estimated that, on average, only one error occurs for every 10 billion bases replicated. However, given that living organisms are continually subjected to a variety of endogenous and exogenous DNA-damaging agents, optimal conditions rarely prevail in vivo. While all organisms have evolved elaborate repair pathways to deal with such damage, the pathways rarely operate with 100 percent efficiency. Thus, the persisting DNA lesions are replicated, but with much lower fidelity than is undamaged DNA. Our aim is to understand the molecular mechanisms by which mutations are introduced into damaged DNA. The process, commonly referred to as translesion synthesis (TLS) or translesion replication (TR), is facilitated by one or more members of the Y-family of DNA polymerases that are conserved from bacteria to humans. Based on phylogenetic relationships, Y-family polymerases may be broadly classified into five subfamilies; DinB-like (polIV/pol kappa-like) proteins are ubiquitous and found in all domains of life; in contrast, the Rev1-like, Rad30A (pol eta)-like, and Rad30B (pol iota)-like polymerases are found only in eukaryotes and the UmuC (polV)-like polymerases only in prokaryotes. We continue to investigate TLS in all three domains of life: bacteria, archaea, and eukaryotes.

Mutagenesis in prokaryotes

Curti, Howard, Karata, McDonald, Theisen, Vaisman, Woodgate

Escherichia coli possesses five known DNA polymerases (pols); given their multiplicity, the pols require exquisite regulation of their access to genomic DNA. The cell’s main replicase is the pol III holoenzyme while pol I is responsible for the maturation of Okazaki fragments and for filling gaps generated during nucleotide excision repair. As part of the cell’s global SOS response to DNA damage, pol II, pol IV, and pol V are significantly upregulated; under these conditions, they may alter the fidelity of DNA replication by potentially interfering with the ability of pol I and pol III to complete their cellular functions. To gain insight into the interplay among E. coli’s five DNA polymerases, we exploited the fact that each polymerase leaves a distinct genetic fingerprint when copying DNA. We analyzed the spectrum of rpoB mutants generated in a set of 13 isogenic strains carrying a mutation in one or more of E. coli’s chromosomally encoded DNA polymerases as well as in three related strains that moderately overexpressed polV-like polymerases. The key to our study was that all strains carried a defect in mismatch repair, thereby allowing us to monitor polymerase-specific misincorporation events rather than assaying mutations that escape repair. Indeed, it is well known that mismatch repair largely protects an organism against transition mutations and has little effect on transversions; therefore, it was no surprise that, in our rpoB assays, the spectrum of mutations was dominated by AT→GC and CG→TA transitions at four well-defined hot spots. The mutations were not dependent on pol II, pol IV, or pol V, as we recovered a considerable number of mutants at the hot spots in the absence of each polymerase. However, given that the magnitude of mutagenesis at certain positions varied considerably in the polymerase-deficient strains, it appears that all three polymerases can modulate the extent of mutagenesis at the hot spots. In contrast, in a strain lacking polA, we observed a striking reduction in transition mutagenesis, suggesting that such mutagenic events are largely dependent on pol I. The observation that pol I is responsible for most transition mutagenesis is remarkable in that pol I possesses both 3′→5′ and 5′→3′ exonuclease activities and is generally thought to be an accurate polymerase, with misincorporations occurring with a frequency of less than 1 in a million bases replicated.

Our study also revealed that transversion mutations are largely dependent on the combined actions of pol IV and pol V and that the cellular levels of pol V limit transversion mutagenesis. Furthermore and in contrast to pol I, pol IV, and pol V, which promote mutagenesis, pol II appears to play a largely anti-mutagenic role by suppressing mutations at certain hot spots.

Until now, it has tacitly been assumed that spontaneous mutations occurring in the E. coli genome arise largely during genome duplication performed by the cell’s replicase pol III. However, our observations led us to conclude that, in fact, considerable switching occurs among E. coli’s five DNA polymerases in the absence of exogenous DNA damage and that such interplay modulates the extent of spontaneous mutagenesis occurring on the E. coli chromosome.

Curti E, McDonald JP, Mead S, Woodgate R. DNA polymerase switching: effects on spontaneous mutagenesis in Escherichia coli. Mol Microbiol 2008, in press.
Mead S, Vaisman A, Valjavec-Gratian M, Karata K, Vandewiele D, Woodgate R. Characterization of polVR391: a Y-family polymerase encoded by rumA′B from the IncJ conjugative transposon, R391. Mol Microbiol 2007;63:797-810.

Characterization of archaeal Dpo4

McDonald, Vaisman, Woodgate; in collaboration with Bell

Orthologues of E. coli pol IV are found in all domains of life, including archaea. The best-characterized archaeal orthologue is Dpo4, which originates from the hyperthermophilic crenarchaeota Sulfolobus solfataricus. We previously showed that Dpo4 is a low-fidelity polymerase; it can facilitate the bypass of a number of lesions that block cellular replicases. Interestingly, the crystal structure of Dpo4 revealed that, although the enzyme possesses structural domains similar to replicases, its active site is large and solvent-exposed, explaining its ability to accommodate bulky adducts in that site. While Dpo4 is moderately processive, the polymerase possesses a putative “PIP-box,” which is thought to facilitate an interaction between the polymerases and PCNA (proliferating cell nuclear antigen), the cell’s sliding-clamp. In a recent collaborative study with Stephen Bell, we assayed the effect of PCNA on the processivity of Dpo4 when replicating an undamaged DNA template and investigated the mechanism whereby circular PCNA is loaded onto DNA.

Unlike the case of eukaryotic PCNA, which is composed of a homotrimer of PCNA protomers, previous studies on S. solfataricus PCNA demonstrated that the sliding clamp in fact consists of a heterotrimer of three discrete PCNA subunits (PCNA1, PCNA2, and PCNA3). We first confirmed that the processivity of Dpo4 is greatly stimulated in the presence of heterotrimeric PCNA1/2/3. Given that changing the two phenylalanines to alanines in the consensus PIP-box of Dpo4 abrogated the polymerase’s ability to be stimulated by PCNA, we concluded that the increased processivity occurred through a physical interaction between PCNA and Dpo4. Interestingly, yeast two-hybrid analyses revealed that Dpo4 specifically interacts with the PCNA1 subunit and shows little to no affinity for PCNA2 or PCNA3 subunits.

We next investigated the mechanism of clamp loading. While molecular dynamics simulations suggest that only a single interface within PCNA may be opened during loading, we have yet to test biochemically the number of inter-subunit interfaces—one, two, or even three—that are opened in PCNA during the loading process in order to facilitate the entry of DNA during loading. Accordingly, we exploited the unique asymmetry of Sulfolobus solfataricus PCNA. Our strategy was to fuse PCNA subunits covalently by using linker peptide sequences to join the N-terminus of one PCNA subunit to the C-terminus of its neighbor, thereby covalently sealing inter-subunit interfaces. Employing such constructs in clamp-loading assays allowed us to determine the number and identity of interfaces important for loading. All combinations of PCNA1, PCNA2, and PCNA3, which reconstituted the heterotrimeric PCNA, stimulated Dpo4 activity. With the dimer fusion constructs, we could detect replication factor C (RFC)–dependent stimulation in reactions containing the PCNA1/2 fusion plus PCNA3 and when the PCNA2/3 fusion plus PCNA1 was present. Significantly, the PCNA3/1 fusion plus PCNA2 could not support RFC-dependent stimulation of Dpo4 activity. Our data therefore indicate that productive loading of PCNA requires the opening of only a single inter-subunit interface—specifically, that between PCNA3 and PCNA1.

d’Abbadie M, Hofreiter M, Vaisman A, Loakes D, Gasparutto D, Cadet J, Woodgate R, Pääbo S, Holliger P. Molecular breeding of polymerases for amplification of ancient DNA. Nat Biotechnol 2007;25:939-943.
Dionne I, Brown NJ, Woodgate R, Bell SD. On the mechanism of loading the PCNA sliding clamp by RFC. Mol Microbiol 2008;68:216-222.

Characterization of eukaryotic Y-family polymerases

Chumakov, Kuban, McDonald, Ohashi, Woodgate; in collaboration with Yang, Wilson

DNA is subject to a variety of chemical modifications that alter its structure. Such alterations can block basic cellular functions such as transcription and/or replication and can lead to cell death, mutagenesis, and, in higher eukaryotes, cancer. One such modification is DNA methylation, which can be caused by endogenous chemicals, products of metabolism, environmental exposure, or treatment with several cancer chemotherapeutics. Not surprisingly, cells have developed several evolutionarily conserved mechanisms for repairing or tolerating this type of DNA damage, including base excision repair (BER), nucleotide excision repair (NER), recombination, and TLS. The major products in DNA exposed to S_N2-methylating agents are N7-methylguanine and N3-methyladenine (3MeA). 3MeA accounts for approximately 20 percent of the base damage formed by S_N2-methylating agents and is considered the major cytotoxic lesion produced by such chemicals. However, it has been extremely difficult to prove that 3MeA directly blocks replication, as 3MeA’s half-life in vitro is estimated at between 12 and 24 hours, thereby precluding biochemical analysis. To circumvent these problems, we synthesized a stable 3-deaza analogue of the nucleoside 3-methyl-2′-deoxyadenosine (3dMeA) that we incorporated into synthetic oligonucleotides and used as templates for DNA replication in vitro. As expected, the 3dMeA lesion blocked both human DNA polymerases alpha and delta. In contrast, human polymerases eta, iota, and kappa as well as S. cerevisiae pol eta were able to bypass the lesion, albeit with varying efficiencies and accuracy. To confirm the physiological relevance of our findings, we demonstrated that, in S. cerevisiae lacking 3MeA repair, human pol eta, pol iota, and especially pol kappa are capable of restoring methyl methanesulfonate (MMS) resistance to the normally MMS-sensitive strain.

In a collaborative study with Samuel Wilson, we investigated the role of DNA polymerase iota in BER, which plays a major role in maintaining genomic integrity. For example, the BER pathway predominantly repairs alkylated DNA bases in higher eukaryotes. During single-nucleotide BER, a damage-specific glycosylase removes an alkylated base, and apurinic/apyrimidinic endonuclease cleaves the resulting abasic site. The outcome is a nick with a 5′-dRP group that can be removed by the dRP lyase activity of pol beta; the same polymerase then fills the single-nucleotide gap. Finally, a DNA ligase seals the nick, thereby completing repair.

Mouse embryonic fibroblasts (MEF) deficient in pol beta demonstrate increased sensitivity to alkylating agents such as MMS, yet cellular extracts from pol beta–null cells evidence some residual BER activity, suggesting that other polymerases could substitute for pol beta in BER. Previous studies indicated that reversal of MMS hypersensitivity requires the dRP lyase activity of pol beta; therefore, the polymerase that substitutes for pol beta in BER must also possess dRP lyase activity. Interestingly, several years ago, we demonstrated that human pol iota has intrinsic dRP lyase activity and, accordingly, hypothesized that pol iota might function as a back-up BER enzyme in pol beta–null cells. To test this hypothesis directly, we determined the MMS sensitivity of MEFs deficient in pol beta and pol iota. We found that MEFs doubly deficient in pol beta and pol iota were more sensitive to the killing effects of MMS than MEFs deficient in pol beta alone, suggesting that pol iota may indeed participate in the BER of MMS-induced lesions.

However, the pol beta–deficient MEFs used in our study had different origins. To minimize the effect of genetic background variability, we expressed mouse pol iota in the pol beta–deficient and pol iota–deficient cells at a level equal to or greater than wild-type pol iota. Unfortunately, analysis of growth inhibition revealed that expression of pol iota did not affect the cells’ resistance to the alkylating agents MMS or N-methyl-N′-nitro-N-nitrosoguanidine. Thus, we concluded that the difference in the MMS sensitivity of the pol iota–deficient and –proficient cell lines is unlikely to result from the difference in pol iota expression but rather is attributable to uncharacterized differences between the two cell lines. As a consequence, we concluded that, although pol iota possesses dRP lyase activity, expression of the polymerase has negligible impact on the alkylation damage sensitivity of cells lacking pol beta.

Frank EG, Woodgate R. Increased catalytic activity and altered fidelity of human DNA polymerase iota in the presence of manganese. J Biol Chem 2007;282:24689-24696.
Plosky BS, Frank EG, Berry DA, Vennall GP, McDonald JP, Woodgate R. Eukaryotic Y-family polymerases bypass a 3-methyl-2′-deoxyadenosine analog in vitro and methyl methanesulfonate-induced DNA damage in vivo. Nucleic Acids Res 2008;36:2152-2162.
Poltoratsky V, Horton JK, Prasad R, Beard WA, Woodgate R, Wilson SH. Negligible impact of pol iota expression on the alkylation sensitivity of pol beta-deficient mouse fibroblast cells. DNA Repair 2008;7:830-833.
Yang W, Woodgate R. What a difference a decade makes: insights into translesion DNA synthesis. Proc Natl Acad Sci USA 2007;104:15591-15598.

Role of DNA polymerase iota in humans with the xeroderma pigmentosum variant (XP-V) syndrome

McLenigan, Vaisman, Woodgate; in collaboration with Kraemer

Xeroderma pigmentosum variant (XPV) patients have normal DNA excision repair yet are predisposed to develop sunlight-induced cancer. They exhibit a 20-fold higher-than-normal frequency of UV-induced mutations with a highly unusual spectrum. The primary defect in XP-V cells is a lack of functional pol eta, which normally inserts adenine nucleotides opposite photoproducts involving thymine. Both the high frequency and striking difference in spectrum of UV-induced mutations in XP-V cells strongly suggest that, in the absence of pol eta, translesion replication of UV-induced DNA lesions is catalyzed by an alternate DNA polymerase that is much more error-prone than pol eta. Indeed, we previously presented data suggesting that pol iota may be responsible for the hypermutability observed in patients with XP-V.

As part of a collaborative study with Kenneth Kraemer, we surveyed the molecular basis of XP-V worldwide by measuring levels of pol eta protein in skin fibroblasts from putative XP-V patients (age 8–66 years) from 10 families in North America, Turkey, Israel, Germany, and Korea. DNA sequencing identified 10 distinct POLH mutations, suggesting that the mutations arose independently. They included two splicing mutations, one nonsense mutation, five frameshift mutations (three deletion and two insertion), and two mis-sense mutations. Nine of the mutations were located in the catalytic domain of the polymerase and one in the C-terminus of the protein involved in mediating protein-protein interactions between pol eta and its partners.

We also examined the level of pol eta protein in 16 cell strains from other UV-sensitive, cancer-prone patients. As expected, the cells exhibited normal post–UV cell survival, host cell reactivation, or DNA repair. However, unlike true XP-V cells, the cells had normal levels of pol eta protein, and we were unable to detect any mutations in POLH. We therefore considered the possibility that these clinical phenotypes may have arisen because of mutations in another polymerase involved in the translesion synthesis of UV photoproducts. An obvious candidate was pol iota, given that our previous studies had demonstrated that it interacts physically with pol eta and that both polymerases co-localize to sites of UV-induced DNA damage. We therefore measured the level of pol iota protein in the cells of interest and sequenced the entire POLI gene. However, all cells exhibited normal levels of pol iota protein and no mutations in POLI. The molecular defect causing the phenotype resembling XP-V in these cells thus remains to be determined.

Inui H, Oh KS, Nadem C, Ueda T, Khan SG, Metin A, Gozukara E, Emmert S, Slor H, Busch DB, Baker CC, DiGiovanna JJ, Tamura D, Seitz CS, Gratchev A, Wu WH, Chung KY, Chung HJ, Azizi E, Woodgate R, Schneider TD, Kraemer KH. Xeroderma pigmentosum-variant patients from America, Europe, and Asia. J Invest Dermatol 2008;128:2055-2068.
Wang Y, Woodgate R, McManus TP , Mead S, McCormick JJ, Maher VM. Evidence that in xeroderma pigmentosum variant cells, which lack DNA polymerase eta, DNA polymerase iota causes the very high frequency and unique spectrum of UV-induced mutations. Cancer Res 2007;67:3018-3026.

Collaborators

Stephen D. Bell, PhD, Sir William Dunn School of Pathology, University of Oxford, Oxford, UK
Kenneth H. Kraemer, MD, Basic Research Laboratory, Center for Cancer Research, NCI, Bethesda, MD
Samuel H. Wilson, MD, Laboratory of Structural Biology, NIEHS, Research Triangle Park, NC
Wei Yang, PhD, Laboratory of Molecular Biology, NIDDK, Bethesda, MD

For further information, contact woodgate@helix.nih.gov or visit http://sdrrm.nichd.nih.gov