2 Dna Damage and Repair
Introduction
Preserving the integrity of the genome is of paramount importance to any organism. However, our DNA is under continual assault from both endogenous and exogenous damaging agents. These include the products of normal metabolism (reactive oxygen species [ROS]), ultraviolet (UV) and ionizing radiation, and genotoxic compounds such as those found in cigarettes. Inefficient or inadequate DNA repair can have dire consequences, which can lead to cell death, affect essential metabolic processes such as gene transcription, or result in the creation of a mutator phenotype.1 Decreased efficiency of DNA repair pathways have been linked with cancer predisposition in humans. This is clearly illustrated by studies that show that mutations in DNA repair pathways are associated with cancer-predisposing syndromes in humans. Furthermore, the loss of the ability to effectively repair damaged DNA leads to an increase in genomic instability and can contribute to tumorigenesis.2 DNA repair pathways are highly conserved throughout evolution, with many of the initial discoveries in the field having been made in Escherichia coli, but are still highly applicable to human cells. The purpose of this chapter is to give an overview of the five main repair pathways (BER, NER, MMR, NHEJ, and HR), and to highlight how mutations in these pathways affect tumor formation and response to therapy (Fig. 2-1).
DNA Damage Response Pathways
There are numerous mechanisms for the repair of damaged DNA, which are determined in part by the nature of the damage to which they respond and the presence of functioning repair pathways. For example, in the absence of functional BER, a single-strand break (SSB) can form a DNA double-strand break (DSB), which will then be processed by HR.3 In virtually all DNA repair pathways, DNA damage is detected by sensors that create a signal that is further transduced and amplified to initiate the repair pathway. The most critical DNA repair pathways are those that remove damaged bases that result as a consequence of cellular metabolism or genotoxic insult. BER, NER, and MMR pathways are involved in removing damaged bases and are discussed in the following text.
Base Excision Repair
The BER pathway has a central role to play in preserving genomic integrity. Alterations in BER signaling are predicted to predispose humans to cancer and have been associated with colorectal adenomas and colon cancer. In contrast to other DNA repair pathways discussed in this chapter, only recently has an association between loss of BER and a human disease syndrome been identified. Patients affected by this syndrome, termed E. coli mutY human homolog (MYH)–associated polyposis (MAP), develop large numbers of colonic polyps (in the tens to hundreds range) by the age of 40, and nearly 50% present with colon cancer.4 The essential role of this pathway is supported by the finding that mice nullizygous for the genes involved were either lethal, indicating an absolute requirement for the gene product, or showed little to no phenotype because of a high degree of redundancy within the pathway. For example, loss of APE1, LIG1, LIG3, XRCC1, FEN1, and POLB all result in embryonic lethality, whereas deletion of NTH1, OGG1, UNG, AAAG, and MUTYH result in no discernible developmental defects.5 The absolute requirement for BER stems from its ability to repair DNA damage from both exogenous and endogenous sources. The majority of damage to the base portion of each nucleotide is the result of the products of normal metabolism that oxidize or alkylate DNA, such as ROS. Included in these lesions is the formation of 8-oxo-7,8-dihdroguanine (8-oxoG), which has mutagenic properties and also blocks DNA replication and transcription. BER is also used by the cell to handle the high rate of depurination (loss of purines from the DNA), and it has been estimated that a human being loses approximately a trillion guanines from his or her DNA every hour. In addition, BER is also used to remove a large number of cytosines that become spontaneously deaminated to form uracil. The BER mechanism has been extensively characterized in both E. coli and mammals, and includes five stages, shown schematically in Fig. 2-2. In brief, there is an initial recognition of the damaged base by DNA N-glycosylases. Different glycosylases recognize specific lesions, although there is considerable redundancy between them. The damaged base is then removed to give an apurinic/apyrimidinic (AP) site, which is processed by an AP-specific endonuclease (monofunctional glycosylase) or with an additional AP-lyase (bifunctional glycosylase), which leaves a single-strand interruption. These gaps are then filled by a DNA polymerase either with a single nucleotide (short patch) or polynucleotides (longer repair patch) before a ligation occurs. During BER, an SSB is formed as an intermediate. These SSBs are also generated by exogenous agents such as ionizing radiation and other oxidizing compounds. In response to ionizing radiation, BER effectively repairs the damage (SSB) although some additional enzymatic activities are thought to be required, such as PARP1. The control of BER has been recently described and indicates that BER proteins that are not actively involved in repair are ubiquitinated and degraded by the proteosome.6
Nucleotide Excision Repair
NER was discovered in the 1960s through elegant studies on the effects of UV irradiation on DNA synthesis and repair replication in bacteria. Since then it has been characterized extensively in mammals and has been described as the principle repair pathway for the removal of bulky adducts induced by UV radiation or other environmental carcinogens.7–9 In contrast to the BER pathway, damage detected by the NER proteins is not done in a specific manner because there is no equivalent to DNA N-glycosylases involved in BER, described previously. It is hard to imagine that a specific glycosylase could evolve for each type of DNA damage, especially in an environment in which exogenous damaging agents are numerous and constantly attack epithelial cells in the skin. Consequently, NER has been described as a more flexible version of BER. The major lesions recognized by NER are bulky adducts that result from intrastrand crosslinks, UV-induced cyclobutane pyrimidine dimers (CPDs) and 6-4 photoproducts. When these lesions are not removed, they distort the DNA helix and in doing so prevent DNA replication and transcription, which can result in the employment of the error-prone translesion synthesis repair pathway. The process of NER can be subdivided into two pathways: global genome repair (GG-NER) or global genomic repair (GGR), and transcription coupled repair-nucleotide excision (TC-NER) or transcription-coupled repair (TCR).10 The process of GG-NER is genome wide (i.e., lesions can be removed from anywhere). In contrast, TC-NER only removes lesions in DNA strands of actively transcribed genes. When a DNA strand that is being actively transcribed becomes damaged, the ribonucleic acid (RNA) polymerase can block access to the site of damage and hence prevent DNA repair. TC-NER has evolved to prevent this by effectively removing the RNA polymerase from the site of damage to allow the repair proteins access. The mechanism of GG-NER and TC-NER differ only in the detection of the lesion; the subsequent pathway is the same. In general NER involves the recognition of the lesion, by proteins specific to GG-NER or TC-NER, followed by incision of the DNA strand near the damage, removal of the affected stretch of DNA, repair replication using the complementary strand as a template, and, finally, ligation to seal the 3′ end of the repair patch with the parental DNA (Fig. 2-3).11,12
Three human syndromes that occur as a result of mutation of the NER pathways have been identified: xeroderma pigmentosum (XP), Cockayne syndrome (CS) and trichothiodystrophy (TTD).13 Many of the principle genes involved in NER are called XP, for example XPA, XPB, and so on. This is because they were identified in individuals suffering from XP.14 XP results from mutation in any one of seven complementation groups: XPA, XPB (ERCC3), XPC, XPD (ERCC2), XPE (DDB2), XPF and XPG. Interestingly XP patients with mutations in XPC or XPE are deficient in GG-NER, whereas the remaining are deficient in both GG-NER and TC-NER. XP presents early in life as extreme sensitivity to the sun; freckles appear at a young age (1 year) even after short exposure to the sun, and significant pigmentation abnormalities are common. The eyes of XP patients are photosensitive and often exhibit eye damage as a result of constant UV-induced conjunctivitis and keratitis. This extreme sensitivity to UV translates to a 1 in 4000-fold increased risk of skin cancer in XP individuals, often appearing as early as age 10.15 These malignancies are usually squamous and basal cell carcinomas, but melanomas have also been reported. Neurologic abnormalities also frequently occur, including peripheral neuropathy, sensorineural deafness, and progressive mental retardation. Similar to XP patients, those with CS are also very sensitive to UV, although they do not exhibit the increased risk of skin cancer. Instead CS patients have distinct developmental abnormalities, which include growth retardation (dwarfism), progressive cognitive impairment, skeletal abnormalities, severe mental retardation, cataracts, and retinitis pigmentosa. CS patients tend to die younger than the age of 20 (the average life span is 12 years) as a result of infectious or renal complications rather than cancer. The genetic determinants of CS are mutations in either the CSA or CSB genes, both of which are essential for recognition of DNA damage and initiation of the TC pathway of NER. CS individuals remain proficient in GG-NER. Finally, mutation of the XPB or XPD genes results in TTD. Again extreme sensitivity to the sun is a characteristic in some of these patients (approximately 50%), although they possess a low incidence of developing skin cancer. TTD individuals are similar to CS in presentation but have the additional symptom of brittle nails and hair; this is a result of reduced sulfur content in the component proteins. The differences and similarities between these three syndromes have proved intriguing, in that they all show UV sensitivity but only XP individuals have an increased risk of skin cancer. This has been attributed to the determination that loss of GG-NER is what predisposes individuals to skin cancer. In contrast, the developmental abnormalities associated with these syndromes (retardation, dwarfism, etc.) are the result of a loss of TC-NER. Given the essential role of NER in the removal of UV-induced damage it is not surprising that XP individuals succumb to skin cancer, but an increased risk of other malignancies would also be expected and this has not been noted. There have been reports documenting modest alteration in NER genes as a result of polymorphisms that contribute to the risk of solid tumors.
Mismatch Repair
During normal replication and recombination, mistakes can be made that result in the insertion of an incorrect nucleotide leading to a mismatch. These mismatches also occur as a result of base modifications. If mismatches escape surveillance of the proofreading capabilities of DNA polymerases, there is a permanent change in sequence or mutation. The MMR pathway functions to identify such mistakes and correct them before they are propagated further. MMR also functions to remove small loops (insertion–deletion loops [IDL]) in the DNA that arise as a consequence of spontaneous slippage-dependent misalignment between primer and template DNA strands. These occur particularly in highly repetitive microsatellite regions that, in the absence of effective MMR, show a large increase in the frequency of spontaneously occurring mutations (microsatellite instability [MSI]).16 The process of MMR can be subdivided into four components: First, the mismatch must be identified by sensors that transduce a signal; second, MMR factors are recruited; third, the newly synthesized strand harboring the mismatch is identified and the incorrect or altered nucleotides are excised; and in the fourth stage, resynthesis and ligation of the excised tract occurs. MMR was first characterized in E. coli, which encodes MUT genes; homologues of these gene products have now been identified and extensively characterized in both yeast and humans.17 See Fig. 2-4 for a schematic representation and an indication of the critical gene products. The significance of MMR was highlighted when it became apparent in the 1990s that hereditary nonpolyposis colorectal cancer (HNPCC) and sporadic MSI-positive colon cancers were caused by defects in MMR. HNPCC, also known as Lynch syndrome, affects 1 in 1000 people and is characterized by early-onset colorectal cancers that have a high level of MSI. Individuals with HNPCC have an 80% risk of colorectal cancer in their lifetime, as well as an increased risk of other malignancies such as endometrial (50% risk), ovarian, and gastric malignancies. In 70% to 80% of the germline mutations found in HNPCC, the mutations occur in the MLH1 or MLH2 genes. In 10% of cases there is a mutation in MSH6, and, in rare cases, PMS1, PMS2, MLH3, and EXO1.18 Epigenetic silencing of the MMR genes has also been noted and shown to increase the mutation rate. In very rare cases, both copies of an MMR gene, such as MLH1, MSH2, MSH6, or PMS2 have been found mutated (homozygous germline mutations).19 These individuals have a reduced life span and succumb to malignancy of hematologic or brain origin.
The MMR proteins have been described as having a role to play in cell cycle checkpoints and apoptosis. Specifically, both hMutSα- and hMutLα-deficient cells exhibit defective S-phase checkpoint responses.20 The proposed model behind this observation is that MMR proteins act as sensors of DNA damage and then recruit the checkpoint proteins and hence activate these signaling pathways. This is supported by the findings that hMutSα and hMutLα physically interact with ataxia-telangiectasia mutated (ATM), ataxia-telangiectasia Rad3-related (ATR), ATR-interacting protein (ATRIP), c-Abl, and p73 in cells exposed to DNA-damage-inducing agents. Also supportive of this model is the wealth of data indicating that cells that become defective in MMR gain resistance to cytotoxic agents. Interestingly, patients with MSI-positive tumors have a greater chance of survival than those that do not, although whether this is as a result in differences in tumorigenesis or therapeutic response is unclear.
