The Base Excision Repair Pathway

Molecular Mechanisms and Role in Disease Development and Therapeutic Design

David M Wilson III

• Publisher: WSPC
• Year: 2016
• ISBN:
  9789814719742
• Topic: Ciencias biológicas
• Subtopic: Biología molecular

-->

This book will serve as the preeminent text book on the topic of "base excision repair", a key DNA repair pathway that protects cells from most spontaneous forms of DNA damage, including oxidative lesions that arise both in the nuclear and mitochondrial genomes. The book, which includes contributions from many of the world leaders in the field, provides a detailed description of the molecular mechanisms of base excision repair, as well as its emerging relationship to epigenetic regulation, the aging process and human disease, such as cancer susceptibility, immunological defects and neurological disorders. The book will also cover the state-of-the-art technologies being developed to assess base excision repair capacity among individuals in the population, in addition to the strategies being employed to target base excision repair as part of therapeutic paradigms to eradicate disease, namely cancer.

This book represents one of the most extensive efforts to date to cover the topic of "base excision repair". It includes chapters by many of the most established investigators in the field, from all over the world.

--> Contents:

• Genomic Uracil — Dangers and Benefits in Processing (Hans E Krokan, Geir Slupphaug and Bodil Kavli)
• The GO Repair Pathway: OGG1 and MUTYH (Amelia H Manlove, Nicole N Nuñez and Sheila S David)
• The DNA Glycosylases that Recognize and Remove Free Radical-damaged Pyrimidines (Andrea J Lee and Susan S Wallace)
• Alkyladenine DNA Glycosylases (Carrie M Margulies and Leona D Samson)
• The Abasic Endonuclease APE1: Much More Than a DNA Repair Enzyme (Mattia Poletto, Lisa Lirussi, Giulia Antoniali and Gianluca Tell)
• Central Steps in Mammalian BER and Regulation by PARP1 (Rajendra Prasad, Julie K Horton, Yuan Liu and Samuel H Wilson)
• Processing Strand Break Termini in the DNA Single-Strand Break Repair Pathway (Zahra Shire, Bingcheng Jiang, Ismail Abdou and Michael Weinfeld)
• The PARP-XRCC1 Axis in Base Excision Repair (Vincent Pennaneach and Janet Hall)
• DNA Ligases in Base Excision Repair (Alan E Tomkinson and Ishtiaque Rashid)
• Base Excision Repair and Epigenetic Regulation (Alain R Weber, Primo Schär and David Schuermann)
• Base Excision Repair in the Immune System (Carol E Schrader and Jeroen E J Guikema)
• Base Excision Repair in the Etiology of Lupus and Cancer (Khadijeh Alnajjar, Sreerupa Ray, Allison N Schechter, Alireza G Senejani, Stephany Robison, Lindsey A Criswell, Julie Dragon and Joann B Sweasy)
• Base Excision Repair in Trinucleotide Repeat Expansion Disorders (Margherita Bignami and Filomena Mazzei)
• Neurodegeneration Caused by Accumulation of an Oxidized Base Lesion, 8-oxoguanine, in Nuclear and Mitochondrial DNA: From Animal Models to Human Diseases (Yusaku Nakabeppu)
• Assessing BER Capacity in the Human Population (Isaac Alexander Chaim and Zachary David Nagel)
• Prognostic and Predictive Significance of Base Excision Repair in Human Cancers (Reem Ali and Srinivasan Madhusudan)
• Enzymes in the Base Excision Repair Pathway as Targets for Small Molecule Mediated Therapeutics (Eric F Johnson, Amanda K McCullough and R Stephen Lloyd)
• Mitochondrial Base Excision Repair (Mateus P Mori, Daniela Tathiana Soltys and Nadja C de Souza Pinto)
• Base Excision Repair in Aging (Ricardo Gredilla, Vilhelm A Bohr and Tinna Stevnsner)

-->
--> Readership: Researchers, academics, scientists and clinicians, particularly for radiological and medical oncologists, epidemiologists, cancer prevention specialists, and graduate students in cancer biology, who are interested in DNA repair mechanisms and related susceptibility to cancer. -->
Keywords:DNA Repair;DNA Damage;Cancer;Immunodeficiency;Neurodegeneration;Disease Susceptibility;Treatment;InhibitorReview:0

Chapter 1

Genomic Uracil — Dangers and Benefits in Processing

Hans E. Krokan*^,‡, Geir Slupphaug*^,† and Bodil Kavli*

*Department of Cancer Research and Molecular Medicine,
Norwegian University of Science and Technology,
NO-7491 Trondheim, Norway

^†PROMEC Core Facility for Proteomics and Metabolomics,
NTNU and the Central Norway Regional Health Authority,
Norwegian University of Science and Technology,
NO-7491 Trondheim, Norway
^‡[email protected]

All cells normally have thymine rather than uracil in their DNA, although low levels of uracil occur in genomes due to deamination of cytosine and incorporation of dUMP. These uracils are usually removed by base excision repair (BER) initiated by a uracil DNA glycosylase. However, the uracil-containing DNA bacteriophage PBS2 can survive in Bacillus subtilis because it encodes a peptide inhibitor (Ugi) of uracil DNA glycosylase (Friedberg et al., 1975). Furthermore, bacterial and mammalian DNA polymerases use dUTP and dTTP with essentially equal efficiency (Bessman et al., 1958; Wist et al., 1978), and the structure of DNA containing uracil in U:A pairs is equal to that containing T:A pairs (Langridge and Marmur, 1964). Thus, there is no obvious chemical reason why uracil could not work as the carrier of genetic information, except for the danger of cytosine deamination to the mutagenic precursor uracil. However, bacteria carrying mutations that strongly increase uracil incorporation into the genome are growth arrested (el-Hajj et al., 1992). Thus, cells have adapted to a life with thymine in the genome and do not tolerate large amounts of U:A pairs.

Spontaneous deamination of DNA-cytosine to uracil (Lindahl and Nyberg, 1974) and incorporation of dUMP instead of dTMP (Tye et al., 1977; Wist et al., 1978) were reported approximately four decades ago. More recently, it has been established that genomic uracil may also result from targeted and untargeted enzymatic deamination of cytosine in DNA by activation-induced cytosine deaminase (AID) or other members of the APOBEC-family. These relatively uncommon events are in most cells efficiently corrected by BER initiated by a uracil DNA glycosylase. In mammalian cells, four different uracil DNA glycosylases are known, including nuclear UNG2 and mitochondrial UNG1 (both encoded by the UNG gene), single-strand-selective monofunctional uracil DNA glycosylase (SMUG1), thymine DNA glycosylase (TDG) and methyl-CpG binding domain protein 4 (MBD4). These enzymes have different biochemical properties and expression patterns and have only in part overlapping functions (Kavli et al., 2007; Krokan and Bjoras, 2013; Krokan et al., 2002; Krokan et al., 2014; Lindahl, 2013; Olinski et al., 2010).

Incorporated dUMP may be the most abundant source of genomic uracil. The resulting U:A pairs are not miscoding, but may cause mutations by translesion synthesis (TLS) at abasic sites (AP sites) generated by uracil DNA glycosylase base excision (Auerbach et al., 2005) or from errors generated during the BER process (Akbari et al., 2009; Bennett et al., 2001). Cytosine deamination, resulting in U:G mismatches (Lindahl, 1974, 1993; Lindahl and Nyberg, 1974), is more intriguing, since copying of U during replication will invariably result in the generation of a U:A pair, and eventually a T:A pair, in place of the original C:G pair. Repair of U:G mismatches is therefore essential to avoid mutations. Furthermore, 5-methylcytosine (5-meC) in CpG sequences, a post-replicative cytosine modification that regulates gene expression, is also subject to deamination, resulting in potentially mutagenic G:T mismatches (Coulondre et al., 1978; Duncan and Miller, 1980). The rate of spontaneous deamination of 5-meC in double-stranded DNA (dsDNA) is approximately twice as high as that of cytosine (Shen et al., 1994). In addition, deaminated 5-meC residues (G:T pairs) are apparently less efficiently repaired, relative to G:U pairs, as indicated by several-fold higher occurrence of mutations in CpG contexts (Campbell et al., 2012; Duncan and Miller, 1980; Meng et al., 2015; Rideout et al., 1990). Notably, C:G to T:A transition mutations are the most common mutations in human cancer (Kandoth et al., 2013), as well as in mice (Stuart et al., 2000), and most likely a significant fraction of these arise from deamination of cytosine and 5-meC. In general, cytosine deamination is more than two orders of magnitude higher in nucleosides, nucleotides and single-stranded DNA (ssDNA) as compared to double-stranded DNA (dsDNA) (Lindahl, 1993; Shapiro, 1980). Since it is not known how much of the genomic DNA is on average in a single-stranded form, the number of spontaneous cytosine deaminations per day is not exactly known, as discussed previously (Kavli et al., 2007).

The exocyclic amino groups of genomic adenine and guanine are also subject to deamination, resulting in hypoxanthine and xanthine, respectively. However, the rate of spontaneous adenine deamination in DNA is some two orders of magnitude lower than that of cytosine (Lindahl, 1993; Shapiro, 1980). The rate of spontaneous deamination of guanine in DNA to xanthine is apparently not known, but assumed to be low. However, nitrosative deamination of adenine and guanine in DNA, as well as cytosine, may be significant and medically important, particularly during inflammation. Here, nitric oxide (NO^.) and nitrosative derivatives such as nitrous anhydride and peroxynitrite are formed in substantial amounts (Dedon and Tannenbaum, 2004). Deamination by peroxynitrite is apparently restricted to guanine (Burney et al., 1999).

Genomic cytosine is rather unique among DNA bases due to its chemical instability, as well as its many enzymatic modifications of great significance to DNA repair, gene regulation, innate immunity, adaptive immunity and cancer development, as reviewed recently (Harris and Dudley, 2015; Hashimoto et al., 2015; Knisbacher et al., 2015; Krokan et al., 2014). The present Chapter focuses on the generation and processing of genomic uracil in mammalian cells and its biological significance.

1.Sources of Uracil in DNA

1.1.Incorporation of dUMP — a major source of genomic uracil

The ability of dUTP, as well as 5-bromo dUTP, to replace dTTP in enzymatic synthesis of DNA by a DNA polymerase was first demonstrated in the laboratory of Arthur Kornberg. In fact, Escherichia coli polymerase I (Bessman et al., 1958) and mammalian DNA polymerases (Wardle et al., 2008; Wist et al., 1978) use dUTP and dTTP with equal efficiency. dUTP is a normal intermediate in the biosynthesis of dTTP, but is usually kept at a low level, presumably mainly due to an efficient dUTPase that is present in both prokaryotes (Bertani et al., 1961) and eukaryotes (Wist et al., 1978). Interestingly, archae use a complex strategy to keep uracil out of the genome. Their DNA polymerase incorporates dUMP less efficiently than dTMP and may form protein complexes with dUTPase (Hogrefe et al., 2002; Slupphaug et al., 1993) and even with archae uracil DNA glycosylase (Connolly et al., 2003). Furthermore, archae DNA polymerases may use a “reading head pocket” in the enzyme to terminate 1–4 bases ahead of template dUMP (Connolly et al., 2003; Fogg et al., 2002). Thus, archae have several mechanisms to prevent dUMP-incorporation and to handle dUMP-residues in the template.

In bacteria and eukaryotes, it is reasonable to assume that the dUTP:dTTP ratio in a given cell correlates with dUMP incorporation. This ratio varies between tumor cell lines, but is in the range of 0.01–0.05 (Grogan et al., 2011; Horowitz et al., 1997; Studebaker et al., 2005; Traut, 1994; Webley et al., 2000; Wilson et al., 2012), or even lower (Weil et al., 2013). This would suggest that the levels of incorporation of dUMP would be in the range of 2–10 × 10⁷ in a single round of replication of a double set of human chromosomes containing 30% T. Presumably, most of these uracils are rapidly removed by post-replicative BER (Nilsen et al., 2000; Otterlei et al., 1999; Wist et al., 1978), since the genomic level of uracil is 4–5 orders of magnitude lower than predicted (Galashevskaya et al., 2013; Rona et al., 2015).

Remarkably, activated peripheral blood mononuclear cells (PBMCs) have a ratio of dUTP:dTTP of ~1, whereas macrophages have ratios as high as ~60. Since PBMCs and macrophages are differentiated non-proliferating cells, replicative incorporation of dUMP in nuclear DNA is not an issue. Rather, the high level of dUTP is required as part of a defense mechanism, in which dUMP is incorporated by reverse transcriptase into proviral-HIV1 (human immunodeficiency virus) DNA or other retroviruses infecting the cell. Subsequently, uracilated proviral DNA can be degraded in a process initiated by UNG2, thereby preventing integration (Kennedy et al., 2011; Weil et al., 2013).

1.1.1.Consequences of dUMP-incorporation in DNA

In mammalian nuclei, incorporated dUMP is very rapidly removed by BER largely initiated by UNG (Nilsen et al., 2000; Otterlei et al., 1999; Wist et al., 1978). Persistent uracil in DNA may affect gene expression, since some transcription factors require T in the binding site and do not bind to a U-containing recognition sequence (Verri et al., 1990). AP sites resulting from removal of uracil in U:A pairs also negatively affect in vitro gene expression in a plasmid-based system, and this decreased transcription was relieved by knockdown of UNG1/2, but not SMUG1. Gene expression in a U:G construct in an otherwise identical plasmid was similarly affected, but unexpectedly, neither knockdown of UNG1/2, SMUG1 or TDG influenced gene expression. This result may indicate that processing of U:A is different from that of U:G (Luhnsdorf et al., 2014). AP sites generated through removal of uracil in U:A base pairs are mutagenic through at least two mechanisms; TLS and errors made during BER. Interestingly, in yeast, mutagenesis is linked to excision of incorporated dump and active transcription. Mutations largely occur at hot spots in a process that requires Ung1 to generate AP sites and a TLS polymerase (Polζ or Rev1) to introduce complex mutations (Kim and Jinks-Robertson, 2009). These results suggest that transcription-dependent genome instability may at least in part depend on dUMP incorporation and the subsequent generation of AP sites. This is quite intriguing since Ung contributes to mutagenesis and one would expect that mutagenesis by such a mechanism would be selected against. Possibly, dUMP incorporation may have so far not identified physiological advantages.

In yeast, in vivo, mutations from AP sites opposite adenine are most frequently caused by Rev1-dependent dCMP-insertions, leading to AT>CG transversions (Auerbach et al., 2005). Rev1 and Polζ are also involved in TLS when the AP site is opposite guanine (Auerbach et al., 2005; Chan et al., 2013). REV1 and POLζ are involved in TLS of AP sites in human cells as well (Weerasooriya et al., 2014). In vitro BER of a plasmid DNA with one distinct U:A lesion by extracts of embryonic mouse cells resulted in an error frequency of approximately 6 × 10⁻⁴ (Bennett et al., 2001). In a similar study, the error frequency during in vitro BER of U:A pairs in human cell extracts was approximately 1.2 × 10⁻⁴ and 3.6 × 10⁻⁴ in extracts from proliferating and non-proliferating cells, respectively. A targeted one-nucleotide deletion apparently caused by DNA POLβ, a key protein for the repair synthesis reaction during BER (see Chapter 6), was the most common mutation (Akbari et al., 2009). It is notable that the frequency of incorporation of dUMP estimated from dUTP:dTTP ratios (2–10 × 10⁷ in one round of replication) and the reported error frequencies in BER of U:A alone (~1.2 × 10⁻⁴) would suggest a mutation frequency of 2–8 × 10³ per round of replication in mammalian cells. Mutations from TLS over AP sites would add to this number. Mutation frequencies from AP sites were found to be even higher (0.6–2.6%) after transfection of an AP site-containing vector into monkey cells and depended upon the base in the opposite strand (Gentil et al., 1992). These estimated mutation rates are orders of magnitude higher than those observed, since the total intergeneration mutation rate from all sources measured by whole-genome sequencing of human cells was found to be only ~1.1 × 10⁻⁸ per position (Roach et al., 2010). This corresponds to less than 100 mutations per diploid genome per sexual generation and less than 1 mutation per cell generation. Earlier reports reached similar mutation rates using other methods (Drake et al., 1998; Nachman and Crowell, 2000). In conclusion, as far as mutations from processing of U:A is concerned, either measured BER errors are too high, or dUMP-incorporation estimated from dUTP:dTTP ratios is too high.

In light of ...