According to the type of DNA damage, cells have different systems for detecting and repairing DNA lesions. Some types of DNA damage may be minor, resulting simply in an altered base, for example. Others, such as DNA cross-linking, may be more problematic and block DNA replication (the replication fork stalls) or block transcription (the RNA polymerase stalls).
Different molecular sensors identify different types of DNA damage, triggering an appropriate DNA repair pathway. If the DNA lesion is substantial and initial repair is not effective, cell cycle arrest may be triggered. The arrest may be temporary (we consider this later in the context of cancer) or be more permanent. In other cases, as often hap pens in lymphocytes, apoptosis may be triggered.
The DNA repair process occasionally involves simple reversal of the molecular steps that cause DNA damage. This, however, is rare in human cells. Examples include the use of O-6-methylguanine DNA methyltransferase to reverse methylation of guanine at the O6 position, and the use of DNA ligase to repair broken phosphodiester bonds (DNA nicks).
Normally, DNA repair pathways do not directly reverse the damage process. Instead, according to the type of DNA lesion, one of several alternative DNA repair pathways is used. Most of the time, the repair needs to be made to one DNA strand only, but some times there is a need to repair both strands, as in the case of interstrand cross-linking (see Figure 1D, part [ii]) and double-strand DNA breaks (see Figure 1A, part [iii]).
Fig1. Four classes of chemical damage to DNA. (A) DNA strand breakage. A single strand may be broken by simple cleavage of a phosphodiester bond (i) or by a more complex single-strand break (ii) where the ends are damaged and sometimes one or more nucleotides are deleted. Double-strand DNA breaks (iii) occur when both strands are broken at sites that are in very close proximity. (B) Base deletion. Hydrolysis cleaves the covalent N-glycosidic bond (shown in deep blue) connecting a base to its sugar. (C) Base modification. Altered bonding or added chemical groups are shown in red. Examples are: 8-oxoguanine (i), which base-pairs to adenine and so induces mutations; thymidine glycol (ii), which is not a mutagen but blocks DNA polymerase; and a DNA adduct (iii) formed by covalent bonding, in this case bonding of an aromatic hydrocarbon, benzo(a)pyrene, to N7 of a guanine residue. (D) Base cross-linking. This involves formation of new covalent bonds linking two bases that may be on the same DNA strand (an intrastrand cross-link) or on complementary DNA strands (an interstrand cross-link). The former includes cyclobutane pyrimidine dimers: linked carbon atoms 4 and 5 on adjacent pyrimidines on a DNA strand (i). This is the most prevalent form of damage incurred by exposure to ultraviolet light from the sun. The anticancer agent cisplatin, (NH3)2 PtCl2 , causes interstrand cross-links by covalently bonding the N7 nitrogen atoms of guanines on opposite strands (ii).
Errors in DNA replication and chemical damage to DNA are a constant throughout life. Inevitably, however, some mistakes are made in repairing DNA and there are also inherent weaknesses in detecting some base changes (as described below). Inefficiency in detecting and repairing DNA damage is an important contributor to generating mutation. We consider the major DNA repair mechanisms in the next two sections—see Table 1 for a road map of how individual types of DNA changes are repaired.
Table1. FREQUENT TYPES OF DNA DAMAGE/ALTERATION AND HOW THEY ARE REPAIRED IN HUMAN CELLS
Repair of DNA damage or altered sequence on a single DNA strand
DNA damage or an error in DNA replication mostly results in one strand having a DNA lesion or a wrongly inserted base but leaves the complementary DNA strand unaffected at that location. In that case, the undamaged complementary strand may be used as a template to direct accurate repair. Different repair mechanisms are available as listed below.
Base-excision repair (BER)
This pathway is specifically aimed at lesions where a single base has either been modified or excised by hydrolysis to leave an abasic site (~20,000 such events occur in each nucleated cell every day). To replace a modified base by the correct one, a specific DNA glycosylase cleaves the sugar–base bond to delete the base, producing an abasic site. For all abasic sites, the residual sugar–phosphate residue is removed using a dedicated endonuclease and phosphodiesterase. The gap is filled using a DNA polymerase (to insert the correct nucleotide) and DNA ligase (Figure 2A). Note that the same DNA repair machinery occasionally makes a more substantial long-patch repair that involves replacing more than a single base.
Fig2. Base-excision and nucleotide-excision repair. (A) Base-excision repair. This pathway repairs modified bases and abasic sites produced by depurination or depyrimidination. Modified bases are first removed by a DNA glycosylase that cleaves the N-glycosidic bond that connects the base to the sugar, producing an abasic site. The cell has a range of different DNA glycosylases that are specific for common modified bases, such as 8-oxoguanine DNA glycosylase and uracil DNA glycosylase (as shown here). Abasic sites are usually repaired by excising the remaining sugar–phosphate residue (using a phosphodiesterase and specialized endonuclease) then inserting the correct nucleotide to match the undamaged complementary DNA strand. (B) Nucleotide-excision repair. After identification of a bulky DNA lesion, this type of repair involves opening-out the double helix containing the lesion over a considerable distance (using a helicase to unwind the DNA). Subsequently, an excision nuclease makes cuts on either side of the lesion on the damaged DNA strand, generating an oligonucleotide of about 30 nucleotides containing the damaged site, which is then degraded. The resulting gap is repaired by DNA synthesis, using the undamaged strand as a template, and sealed using a DNA ligase.
Single-strand break repair
Single-strand breaks (SSBs)—also called DNA nicks—are very common. They are eas ily reversed by DNA ligase but oxidative attack can cause deoxyribose residues to dis integrate, producing more complex strand breakage. A type of base-excision repair is employed: strand breaks are rapidly detected and briefly bound by a sensor molecule, poly(ADP-ribose), that initiates repair by attracting suitable repair proteins to the site. The 3′ or 5′ termini of most SSBs are damaged and need to be restored. The gap is then filled using a DNA polymerase and DNA ligase.
Nucleotide-excision repair (NER)
This mechanism allows repair of bulky, helix-distorting DNA lesions. After the lesion is detected, the damaged site is opened out and the DNA is cleaved some distance away on either side of the lesion, generating an oligonucleotide of about 30 nucleotides containing the damaged site that is discarded. Resynthesis of DNA is carried out using the opposite strand as a template (Figure 2B). The priority is to rapidly repair bulky lesions that block actively transcribed regions of DNA. A specialized subpathway, transcription-coupled repair, initiates this type of repair after detection of RNA polymerases that have stalled at the damaged site. Otherwise, an alternative global, genome nucleotide-excision repair pathway is used.
Base mismatch repair
This mechanism corrects errors in DNA replication. The mismatch repair (MMR) com ponents work closely with the DNA replication machinery. In human cells, three types of protein dimer carry out most of the repairs (Figure 3A). Two of them—hMutSα and hMutSβ—are needed to identify base mismatches. The former identifies base–base mismatches but can also handle mismatching due to single nucleotide insertions or deletions; hMutS can spot base mismatching for different sizes of very short insertions or deletions (which frequently occur at short tandem repeats as a result of replication slippage, the tendency for DNA polymerase to stutter or skip forward at tandem repeats).
Fig3. Mismatch repair for correcting replication errors. (A) Major classes of MutS or MutL dimers in human mismatch repair. (B) Mechanism of 5′-directed mismatch repair in eukaryotic cells. Replication errors on a newly synthesized strand result in base mismatches that can be recognized by a MutS–MutL complex. The MutS component works as a clamp that can slide along the DNA, allowing it to scan for a base–base mismatch (using MutSα, shown here as subunits MSH2 and MSH6) or an unpaired insertion/deletion loop (often using MutSβ). MutLα, which has an endonuclease function, can form a ternary complex with MutS and DNA. After the newly replicated DNA has been identified (by having a pre-existing nick in the DNA), PCNA (proliferating cell nuclear antigen) and RFC (replication factor C) are loaded onto the newly replicated DNA, where they help trigger the endonuclease function of PMS2 to make a new nick close to the replication error. EXO1 exonuclease is recruited to excise the sequence containing the replication error, making a gapped DNA. The resulting stretch of single-stranded DNA (stabilized by binding the RPA protein) is used as a template for the re-synthesis of the correct sequence using high-fidelity DNA polymerase δ (POL δ), followed by sealing with DNA ligase I. (Adapted from Geng H & Hsieh P [2013] In: DNA Alterations in Lynch Syndrome: Advances in Molecular Diagnosis and Genetic Counseling [M Vogelsang, ed.]. With permission from Springer Science and Business Media. Copyright © 2013.)
The MMR machinery cannot simply repair one of the two strands at random: there has to be a way of distinguishing the original (correct) strand from the newly replicated strand with the incorrect sequence that needs to be repaired. Before being repaired by DNA ligase, nicks (single-strand breaks) are common on a freshly replicated DNA strand, and in human (and other eukaryotic) cells the strand distinction is achieved by identifying a nearby nick on the newly replicated DNA strand. Then hMutLα cleaves the newly replicated strand close to the mismatch and recruits an exonuclease to excise a short stretch of DNA containing the replication error so that the DNA can be re-synthesized and repaired (Figure 3B). Errors in base mismatch repair are important in some cancers.
Repair of DNA lesions that affect both DNA strands
Double-strand DNA breaks (DSBs) are normally rare in cells. They do occur naturally, however, and are necessary for specialized DNA rearrangements in B and T cells that maximize immunoglobulin and T-cell receptor diversity (as described below).
DSBs also occur by accident, as a result of chemical attack on DNA by endogenous or externally-induced reactive oxygen species (but at much lower frequencies than single-strand breaks). In these cases, DNA repair is required, but can sometimes be difficult to perform. For example, when the two complementary DNA strands are broken simultaneously at sites sufficiently close to each other, neither base pairing nor chroma tin structure may be sufficient to hold the two broken ends opposite each other. The DNA termini will often have sustained base damage and the two broken ends are liable to become physically dissociated from each other, making alignment difficult.
Unrepaired DSBs are highly dangerous to cells. The break can lead to inactivation of a critically important gene, and the broken ends are liable to recombine with other DNA molecules, causing chromosome rearrangements that may be harmful or lethal to the cell. Cells respond to DSBs in different ways. Two major DNA repair mechanisms can be deployed to repair a DSB, as listed below, but if repair is incomplete, apoptosis is likely to be triggered.
• Homologous recombination (HR)-mediated DNA repair. This highly accurate repair mechanism requires a homologous intact DNA strand to be available to act as a template strand. Normally, therefore, it operates after DNA replication (and before mitosis), using a DNA strand from the undamaged sister chromatid as a template to guide repair (Figure 4). It is important in early embryogenesis, when many cells are rapidly proliferating, and in the repair of proliferating cells; it occurs after the DNA has replicated, notably in S phase.
• Nonhomologous end-joining (NHEJ). No template strand is needed here: the bro ken ends are simply fused together quickly. Specific proteins bind to the exposed DNA ends and recruit a special DNA ligase, DNA ligase IV, to rejoin the broken ends. Unlike HR-mediated DNA repair, NHEJ is, in principle, always available to cells, but it is most important for the repair of differentiated cells and of proliferating cells in G1 phase before the DNA has replicated.
Fig6. Homologous recombination-mediated repair of double-strand DNA breaks. The double-strand break (DSB) in the chromatid at top is repaired using as a template the undamaged DNA strands in the sister chromatid (note, to make the mechanism easier to represent, the upper chromatid is shown in an unusual format; the 3′ → 5′ strand is placed above the 5′ → 3′ strand). The first step is to cut back the 5′ ends at the double-strand break to leave protruding single-strand regions with 3′ ends. Following strand invasion, each of the single-strand regions forms a duplex with an undamaged complementary DNA strand from the sister chromatid, which acts as a template for new DNA synthesis (newly synthesized DNA copied from the sister chromatid DNA is highlighted in yellow). Following DNA synthesis, the ends are sealed using DNA ligase. The repair is highly accurate because for both broken DNA strands, undamaged sister chromatid DNA strands act as a template to direct incorporation of the correct nucleotides during DNA synthesis.
Repair of DNA interstrand cross-links
Cross-linking can occur between bases on complementary strands of a double helix, either as a result of endogenous metabolites or through exogenously supplied chemicals, notably many anticancer drugs (see the example of cisplatin in Figure 1D, part [ii]). Interstrand cross-links seem to be repaired using a combination of nucleotide-excision repair, translesion synthesis, homologous recombination, and a complex of multiple different protein subunits that are encoded by genes mutated in Fanconi anemia (Fanconi anemia DNA repair pathway).
Undetected DNA damage, DNA damage tolerance, and translesion synthesis
DNA damage may sometimes go undetected. For example, cytosines that occur within the dinucleotide CpG are highly mutable as a result of inefficient DNA repair. In vertebrates, the dinucleotide CpG is a frequent target for DNA methylation, converting the cytosine to 5-methylcytosine (5-meC). Deamination of cytosine residues normally produces uracil, which is efficiently recognized as a foreign base in DNA, but deamination of 5-methylcytosine produces thymine that may go undetected as an altered base (Figure 5). As a result, C→T substitutions are the most frequent type of single nucleotide change in our DNA.
Fig5. Why C→T transitions are so common in human DNA. Deamination of cytosine is a common event in our cells and normally produces uracil, a base that is usually found in RNA, not DNA. A dedicated enzyme, uracil DNA glycosylase, recognizes uracil residues in our DNA and removes them as part of the base-excision DNA repair pathway (see Figure 2A). However, as in the DNA of other vertebrates, many of our cytosines are methylated at the 5′ carbon. Deamination of 5′-methylcytosine produces thymine, a base normally found in DNA. Although a stable C-G base pair has been replaced by a T-G base mismatch, the base mismatch may often escape detection by the base mismatch repair system (which focuses on DNA replication events). At the subsequent round of DNA replication, the thymine will form a T-A base pair, effectively producing a C→T mutation.
Sometimes, DNA lesions may be identified but not repaired before DNA replica tion (damage tolerance). For example, DNA lesions that block replication may be bypassed rather than repaired, and nonclassic DNA polymerases are required to resume DNA synthesis past the damaged site (translesion synthesis). Subsequently, the gap in the daughter strand opposite the lesion is filled-in, and later on the lesion can be repaired using the daughter strand as a template in nucleotide-excision repair. The nonclassic DNA polymerases used in translesion synthesis exhibit a low fidelity in DNA replication (Table 2). They have a higher success in incorporating bases opposite a damaged site, but they are prone to error by occasionally inserting the wrong base. As a result, replication forks are preserved, but quite often at the cost of mutagenesis.
Table2. NONCLASSIC DNA-DEPENDENT MAMMALIAN DNA POLYMERASES
