Full text loading...
Chapter 6 : Base Excision Repair
Category: Microbial Genetics and Molecular Biology
Ebook: Choose a downloadable PDF or ePub file. Chapter is a downloadable PDF file. File must be downloaded within 48 hours of purchase
Base excision repair (BER) is probably the most frequently used DNA repair mode in nature. In this pathway, the excision of base damage is initiated by the action of a class of DNA repair enzymes called DNA glycosylases. The initial enzymatic event during BER forms sites in DNA without a base, called apurinic, apyrimidinic, or abasic (AP) sites. The removal of AP sites is initiated by a second class of BER enzymes called apurinic/apyrimidinic (AP) endonucleases, which specifically recognize these sites in duplex DNA. This chapter considers individual biochemical events during BER, where an altered base is removed by a DNA glycosylase and the resulting AP site is cleaved by an AP endonuclease. Next, it discusses the steps required to complete the repair process. First, a normal DNA base is incorporated by a DNA polymerase to replace the excised damaged residue, and then the integrity of the polynucleotide chain is restored by a DNA ligase. Photoincision by an AP endonuclease always takes place 5’ to an AP site, leaving 3’-OH and 5’-deoxyribosephosphate (dRp) termini. For the removal of 5’-dRp termini, gap-filling, and ligation, different BER strategies and enzymes are employed by higher eukaryotes, lower eukaryotes, and bacteria. Further, the chapter talks about sequential interactions between proteins in BER, and introduces effect of chromatin structure on DNA repair processes.
Key Concept Ranking
- Nuclear Magnetic Resonance Spectroscopy
Pathways of BER is a multistep process that corrects endogenous damage to DNA caused by hydrolysis, ROS, and other metabolites. It is catalyzed by a lesion-specific DNA glycosylase, an AP site-specific endonuclease, another endonuclease or 2’-deoxyribose phosphodiesterase, DNA polymerase, and DNA ligase. Initially, the damaged base is excised from DNA by cleavage of the N-glycosyl bond between the base and a deoxyribose sugar, and then the remaining abasic nucleotide (AP site) is excised and replaced by repair synthesis. In the long-patch repair option (shown as the branch point on the right), strand displacement synthesis by DNA polymerase creates a 5’ flap that is cleaved by the flap endonuclease FEN1 to create a ligatable nick. Some DNA glycosylases have an associated lyase activity that cleaves the DNA backbone 3’ of the damaged nucleotide (left branch of pathway). The resulting 3’ modified nucleotide is removed by a phosphodiesterase before repair synthesis and ligation to complete the repair. (Adapted from M. D. Wyatt, J. M. Allan, A. Y. Lau, T. E. Ellenberger, and L. D. Sanson, Bioessays 21:668-676, 1999.)
Simple DNA glycosylases catalyze the hydrolysis of N-glycosyl bonds by using general acid/base chemistry (see, e.g., references 97 , 120 , and 284 ). In the simplest example, an activated water molecule displaces the damaged base to generate an AP site and the free base. A general base (typically an aspartate or glutamate residue) abstracts a proton to activate the water nucleophile for attack of the anomeric C1 carbon of 2’-deoxyribose. A general acid catalyst (depicted as A:H) can accelerate the reaction by protonating the base to make it a better leaving group.
Structure of an AP site and cleavage by AP endonuclease and AP lyase activities. In the center is a depiction of an abasic site (AP site) in one strand of DNA. Cleavage on the 5’ side of the AP site by an AP endonuclease (left) results in a 3’ OH terminus and a 5’ dRp residue. Cleavage on the 3’ side of an AP site by an AP lyase activity (associated with some DNA glycosylases) results in a 3’ end with a 3’ unsaturated aldehydic end and a 5’-phosphorylated end. (Adapted from reference 39 .)
The HhH is a double-stranded-DNA binding motif that was first identified in the E. coli Nth DNA glycosylase ( 203 ). Subsequent crystal structures of the E. coli 3-meA DNA glycosylase ( 204 , 465 ) revealed an unexpected structural similarity to Nth (EndoIII), including the conservation of the HhH motif. The conserved core of HhH DNA glycosylases consists of two α-helical domains flanking the enzyme active site, as revealed by the crystal structures shown here ( 46 , 98 , 102 , 140 , 203 , 204 , 258 , 465 ). Aside from the HhH motif and a few key residues such as a conserved aspartic acid ( Fig. 6–2 ), the residues lining the substrate-binding pocket (circled) are quite variable, reflecting the different specificities of these HhH enzymes.
The phyletic distribution of 234 HhH DNA glycosylases from all three domains of life indicates that these widely distributed enzymes can be classified into six major families by protein sequence similarity. The families reflect predicted specificities for different types of DNA damage. It is notable that the number of HhH enzymes per genome varies widely, perhaps reflecting alternative DNA repair pathways or variation in the amount of DNA damage sustained by different organisms. (Adapted from reference 76 .)
Several different families of UDG proteins have evolved, removing uracil from DNA in different sequence contexts ( 310 ). Representative structures from three of the UDG families (shown here) have a common protein fold, reflecting their shared catalytic properties. The structural differences in the regions flanking the conserved core no doubt relate to the distinctive substrate specificities of mismatch-specific UDGs and members of other families that excise uracil without regard to sequence context.
The uracil-binding pocket of human UNG makes favorable electrostatic interactions with the hydrogen-bonding groups of the base that contribute to catalytic specificity ( 260 ). Thymine is excluded because its bulky C5 methyl group would clash with the side chain of Tyr147. Purine bases are also too large to fit in the substrate-binding pocket of UNG.
Partial amino acid sequence alignment for enzymes in five families of UDG enzymes. Portions of the active-site motifs A and B of five UDG families are shown. Conserved residues within each family are shaded. Highly conserved residues among all five UDG families are darkly shaded. In motif A, the position of a conserved aspartate (D) or asparagine (N) involved in catalysis in families 1, 2, and 3 is shown by an asterisk (*). A conserved aromatic residue involved in the stacking interaction with uracil is also indicated by an asterisk. The enzymes indicated are from the eukaryotes H. sapiens (Hs), M. musculus (Mm), X. laevis (Xl), S. cerevisiae (Sc), and S. pombe (Sp); from the eubacteria E. coli (Ec), M. tuberculosis (Mt), T. maritima (Tm), and T. thermophilus (Tt); and from the archaea P. aerophilum (Pa) and A. fulgidis (Af). (Adapted from reference 63 .)
A protein inhibitor (UgI) encoded by a B. subtilis bacteriophage protects the uracil-containing DNA of the phage from “repair” by host UNG. The small, acidic UgI protein binds to UNG enzymes from a variety of organisms. The crystal structure of human UNG complexed to UGI shows that the UgI inhibitor occupies the DNA-binding surface of the enzyme, acting as a tight-binding mimic of the DNA substrate ( 259 ).
(A) The enzyme 3-meA-DNA glycosylase (Gly) I (TagA) catalyzes the selective excision of free 3-meA from DNA. (B) The enzyme 3-meA-DNA glycosylase II (AlkA) additionally catalyzes the excision of 7-methylguanine (7-meG) and 3-methylguanine (3-meG).
The positively charged, alkylated purines that are excised from DNA by E. coli AlkA (3-methyladenine DNA glycosylase II) are chemically unstable in comparison to unmodified A and G. Consequently, these methylpurines are good leaving groups, prone to spontaneous hydrolysis of the N-glycosyl bond and release from DNA (solid circles; [compare Fig. 6–2 ]). The active site of AlkA exhibits little selectivity, accelerating the glycosylase reaction equally well (several hundredfold) in complexes with DNA containing unmodified or methylated purines (open circles). The greater chemical reactivity of alkylated purine bases is sufficient to explain how an unselective enzyme can function selectively in vivo ( 28 ). (Adapted from reference 286 .)
Activity of 3-meA DNA glycosylases I and II following gel filtration of extracts of unadapted (A) and adapted (B) E. coli cells. Enzyme I is sensitive to product inhibition by free 3-meA. Hence, addition of this base shows that most of the activity present in unadapted cells represents 3-meA-DNA glycosylase I (TagA). In adapted cells, 3-meA-DNA glycosylase II activity (AlkA) elutes earlier from the gel filtration and can be readily distinguished because it is resistant to inhibition by 3-meA.
The alkylated N3 positions of adenine and guanine and the O2 position of cytosine and thymine (shown in black shading) occupy the minor groove of the DNA helix. Other sites of alkylation in bases, such as O6 of guanine, O4 of thymine, and N7 of guanine (shown in gold) occupy the major groove.
The crystal structure of AlkA in complex with a DNA-based inhibitor shows that the enzyme engages the minor groove to bend the DNA and flip the target nucleotide into a pocket in the enzyme active site ( 164 ). The signature HhH motif (grey) contacts the DNA backbone and positions the substrate over the active site.
The structure of human MPG is unrelated to that of E. coli AlkA ( Fig. 6–14 ), despite the similar functions of these enzymes. Like AlkA, MPG catalyzes base flipping by engaging the minor groove of a DNA substrate. In complex with DNA containing 1,N6 ethenoadenine (edA), MPG flips the alkylpurine substrate into an aromatic binding pocket ( 208 ). A bound water molecule is positioned for the backside attack of the N-glycosyl bond.
The glycosylase reaction catalyzed by AlkA, MPG, OGG1, and other purine-specific DNA glycosylases can be accelerated by protonation of the base to make it a better leaving group ( 120 , 284 ). As the N-glycosyl bond is broken, the increased electronegativity of the base is offset by protonation. Modified abasic sites in DNA that mimic the transition state of this reaction serve as tight-binding inhibitors of many different DNA glycosylases ( 164 , 181 , 362 ). (Adapted from reference 284 .)
The E. coli FPG DNA glycosylase processes a structurally diverse group of oxidized and fragmented purine and pyrimidine substrates that are generated by exposure to reactive oxygen and ionizing radiation. FPG substrates include 7,8-dihydro-8-oxoguanosine (8-oxoG), 5-hydroxycytosine (5-OH-C), 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine (FaPy-G), and 4,6-diamino-5-N-methylformamidopyrimidine (FaPy-A).
A crystal structure of the unliganded T. thermophilus Fpg (MutM protein) ( 399 ) first revealed a bilobed protein that is structurally unrelated to the human 8-oxoG DNA glycosylase OGG1 ( Fig. 6–21 ). Crystal structures of Lactococcus lactis ( 368 ), Bacillus stearothermophilus ( 121 ), and E. coli ( 132 ) Fpg proteins bound to DNA show that Fpg proteins strongly bend the DNA to expose and position the target nucleotide for nucleophilic attack by the imino nitrogen of an N-terminal proline residue. A C-terminal zinc-binding domain makes additional DNA contacts that position the enzyme on its DNA substrate ( 475 ).
Enzymatic systems for protection of cells from the mutagenic effects of 7,8-dihydro-8-oxoguanine (8-hydroxyguanine). (A) The predominant tautomeric form of the adduct, 8-oxoG, is shown. (B) The adduct can be excised from DNA by Fpg (MutM) glycosylase, and subsequent completion of BER can restore the original G-C base pair (right). If the 8-oxoG adduct is not removed prior to DNA replication, DNA synthesis may retain the 8-oxoG base pair (left), allowing a subsequent opportunity for BER. Alternatively, an 8-oxoG:A base pair may form during DNA replication (center). The misincorporated A can be removed by the MutY DNA glycosylase and replaced by C, yielding a further opportunity for removal of the 8-oxoG lesion by Fpg. (C) ROS can additionally lead to the formation of 8-oxo-dGTP. This deoxynucleoside triphosphate (dNTP) is a substrate for MutT protein, which removes it from the dNTP pool by hydrolysis (top right). If replication occurred with some 8-oxo-dGTP in the dNTP pool, replication would be largely accurate because T is preferentially inserted opposite A (top left). However, in the event that 8-oxo-dGMP is incorporated into DNA during replication, it may be either correctly base-paired opposite C (bottom left) or mispaired opposite A (center). In the latter event, excision of the mispaired A by the MutY DNA glycosylase can result in an A·T → C·G transversion mutation (bottom right).
Point mutations in human MYH are associated with an increased frequency of sequence transversions and a predilection to colorectal cancer. Two clinically important sites of these mutations, corresponding to E. coli MutY residues Gly260 and Tyr88, contact DNA in the crystal structure of E. coli MutY complexed to DNA ( 119 ).
The crystal structure of human OGG1 bound to DNA ( 46 ) shows how the enzyme recognizes both bases of an 8-oxoG-C base pair to initiate base excision repair and restoration of a G·C base pair. If A is misinserted during replication opposite 8-oxoG, the removal of 8-oxoG by OGG1 would promote G-to-T transversions ( Fig. 6–19 ). For error-free repair, the 8-oxoG-A mismatch glycosylase MYH removes the mispaired A so that the 8-oxoG-C pairing can be restored and correctly processed by OGG1.
Crystal structures of E. coli Nth DNA glycosylase ( 203 ) first revealed the HhH double-stranded-DNA-binding motif that is a hallmark of a large superfamily of DNA glycosylases ( 273 ). The Fe4S4 metal-binding domain of Nth plays a structural role and contains DNA-binding residues ( 407 ).
Illustration of the way in which the M. luteus PD-DNA glycosylase activity was discovered. For simplicity, only one DNA strand is shown. If DNA radiolabeled at its 5’ end and containing a CPD is cleaved 5’ to the dimer by a conventional endonuclease (left), DNA fragments of a particular size are expected in a sequencing gel. However, the fragments observed following incubation with the M. luteus enzyme migrate in sequencing gels as if they are approximately 1 nucleotide larger, suggesting the cleavage mechanism on the right.
Relative sensitivity of bacteriophages T4, T2, T6, and T5 to UV radiation. Phage T4 is significantly more resistant than the other phages.
When phage T2 is exposed to modest doses of UV radiation and introduced into E. coli in the presence of heavily UV irradiated (and hence functionally inactivated) phage T4, phage T2 shows enhanced survival relative to that observed in the absence of the coinfection. A polypeptide encoded by the phage T4 v gene contributes to the survival of irradiated phage T2. This effect is called v gene reactivation.
Enzyme-mediated cleavage of just one of the N-glycosyl bonds in a CPD (symbolized as T < >T) by a PD-DNA glycosylase results in the formation of free thymine following monomerization by photoreversal or the use of DNA photolyase. Measurement of the free thymine provides an assay for the PDDNA glycosylase activity.
The amount of tritium released as free thymine from [3H] thymine-labeled UV-irradiated DNA incubated with the phage T4 PD-DNA glycosylase is half that lost from the DNA as thymine-containing CPD (note the different scales on the y axes). This result indicates that thymine-containing CPD are the probable source of the free thymine and that only one of the N-glycosyl bonds in the dimer is cleaved.
T4 denV cleaves the N-glycosyl bond of the 5’ T of a cis-syn CPD (cf. Fig. 6–26 ) and then catalyzes a β-elimination reaction that cleaves DNA on the 5’ side of the CPD. The crystal structure of T4 denV ( 428 ) shows that a relatively simple protein fold supports both of these catalytic activities. The enzyme flips the adenine (Ade) opposite the 5’ T of the CPD to gain access to the lesion.
The active site of T4 denV includes a cluster of basic residues (Arg3, Arg22, and Arg26) that bind to DNA and a catalytically essential residue Glu23 ( 266 ), which presumably activates a water molecule for attack of the N-glycosyl bond of the 5’ T of a CPD.
Stereo diagram of the active-site residues in T4 denV that stabilize the adenine (Ade) opposite the 5’ T of a CPD in a flipped-out orientation, providing access to the N-glycosyl bond that is cleaved by the T4 denV glycosylase.
Use of [32P]DNA to distinguish between the cleavage of phosphodiester bonds by AP endonucleases and AP lyases. An AP site is shown attached to DNA with a 32P-labeled phosphate (gold) 5’ to the AP site. Cleavage by an AP endonuclease generates a labeled 5’-terminal deoxyribose-5’-phosphate residue (5’-dRp). The radiolabel can be released by alkali-catalyzed β-elimination, yielding free 4-hydroxy-5-[32P]phospho-2-pentenal (ddR5p), which, on further δ-elimination, yields free radiolabeled phosphate. Cleavage by an AP lyase (left) leaves the radiolabel as part of the 3’- terminal structure, where it is resistant to release from the DNA by alkali treatment. However, further treatment of this substrate with an AP endonuclease releases the label as free ddR5p. (Adapted from reference 216 .)
The protein folds of the AP site-specific endonucleases APEX1 (APE1) ( 262 ) and XthA (exonuclease III) ( 263 ) closely resemble the structure of DNase I ( 205 ). All three enzymes are metal-dependent nucleases with a mixed α,β-fold that engages DNA with a series of binding loops. APEX1 strongly bends the DNA and flips the abasic nucleotide into the active site for cleavage of the DNA backbone 5’ to the AP site.
Nfo (endonuclease IV) cleaves DNA 5’ of AP sites to enable repair synthesis by a DNA polymerase. A crystal structure of T. maritima Nfo reveals a TIM barrel-like fold of the enzyme that is adapted for metal binding in the active site and interactions with the DNA substrate of the nuclease reaction ( 169 ). Nfo kinks the bound DNA at a ca. 90° angle to position the AP site nucleotide adjacent to three bound Zn2+ ions (indicated as Zn1, Zn2, and Zn3).
The active site of Nfo harbors three Zn2+ metal ions (indicated as spheres) that coordinate the scissile phosphate at the site of DNA cleavage. The positively charged metal ions can act as a Lewis acid to stabilize a water-derived hydroxide ion for attack of the phosphodiester backbone of DNA and to counteract the developing negative charge on the DNA during the cleavage reaction.
Paraquat induction of Nfo activity. (A) AP endonuclease activity is induced in extracts of wild-type and nth (endonuclease III) mutant strains but not in nfo (endonuclease IV) mutant strains. (B) Cells transfected with a plasmid carrying the cloned nfo+ gene show induction of AP endonuclease activity, while strains that overexpress the nth+ gene do not. (Adapted from reference 54 .)
Reconstitution of BER in vitro with purified human proteins. In this example, the DNA substrate was a 41-mer oligonucleotide annealed to its complementary strand, with a U•G base pair at position 21 (or a C•G base pair as a control). The U-containing strand was labeled with 32P at the 5’ end. Substrate DNA was incubated with the human enzymes indicated in buffer including deoxynucleotide precursors and ATP. The reaction products were analyzed by autoradiography after separation in a denaturing polyacrylamide gel. A combination of the uracil-DNA glycosylase UNG and the AP endonuclease APEX1 cleaved the chain to yield a 20-mer labeled oligonucleotide (lane 6). Inclusion of Pol p in the reaction mixture increased the product length to 21 nucleotides by addition of a single nucleotide, occasionally adding several nucleotides (lane 7). Addition of DNA ligase III completed repair, generating a 41-mer (lane 8). (Adapted from reference 200 .)
The chemical mechanism of enzymatic ligation of DNA was worked out in the 1960s, shortly after the first DNA ligases were isolated ( 214 ). The reaction comprises three chemical steps, in which an adenylate group is transferred from NAD or ATP cofactor to a lysine side chain (step 1) and then to the 5’ phosphate end of a DNA substrate (step 2), activating the DNA for attack by the 3’ OH end and displacement of the adenylate (Ado-phosphate) to seal the nick (step 3). All three steps of the reaction require divalent metal ions. mRNA-capping enzymes that add GMP to the 5 ‘ phosphate of mRNAs catalyze a reaction that is analogous to steps 1 and 2 of the ligation reaction ( 377 ), using a similar protein fold ( Fig. 6–39 ). The E in this figure indicates DNA ligase enzyme, and B is a nucleobase.
Bacteriophage T7 DNA ligase was the first ligase to be crystallized ( 398 ). The enzyme consists of an adenylation domain (AdD) and an OB-fold domain (OBD), an arrangement that is universally conserved in DNA and RNA ligases and mRNAcapping enzymes. The AdD contains many of the highly conserved active site residues, and the OB-fold functions in binding to double-stranded DNA and assisting in the adenylation of the enzyme during step 1 (cf. Fig. 6–36).
The bacteriophage T7 DNA ligase (Lig) ( 398 ), Chlorella virus DNA ligase ( 290 ), and Chlorella mRNA-capping enzyme ( 145 ) have similar, two-domain folds catalyzing common reactions (cf. Fig. 6–37 ). Each enzyme consists of an adenylation/guanylation domain that binds nucleotide cofactor and an OB-fold domain that contributes binding interactions with the substrate.
Multidomain bacterial DNA ligases are NAD + -dependent enzymes. Their overall organization is typified by the crystal structure of DNA ligase from Thermusfiliformis ( 213 ). In addition to the AdD and OBD, which are present in all ligases ( Fig. 6–39 ), bacterial DNA ligases have a zinc-binding domain (ZnB) and a C-terminal α-helical domain, termed the HhH domain, that contributes strongly to DNA-binding activity ( 178 ).
Three mammalian DNA ligases have been well characterized. Ligase 1 functions in the joining of Okazaki fragments during replication, Ligase 3 (which exists in two isoforms, α and β) participates in BER, and Ligase 4 functions in nonhomologous DNA end joining and the rearrangement of immunoglobulin genes. All three ligases share a conserved central region that contains most of the conserved sequence motifs found in all DNA ligases, including a “conserved peptide” (consv.) motif with residues required for step 1 of ligation ( Fig. 6–37 ). Unique regions flanking the conserved catalytic core participate in protein interactions that give rise to the specific biological function(s) of each DNA ligase. aa, amino acids. (Adapted from reference 419 .)
The crystal structure of human DNA ligase I in complex with a nicked, 5’-adenylated DNA substrate shows that the enzyme encircles the DNA ( 307 ). The DBD of ligase I is located immediately N terminal to the catalytic core (cf. Fig. 6–43 ). The DBD contacts the minor groove on either side of the nick and stabilizes the DNA substrate in an underwound conformation that exposes the ends of the nick to the enzyme active site ( Fig. 6–44 ). (B) View of the exposed surface of DNA ligase I showing how the ends of the nicked DNA are buried in the enzyme complex, where two Mg2+ ions are bound near the ends of the nicked DNA.
The N-terminal DBD of ligase I (cf. Fig. 6–42 ) is conserved in mammalian ligases III and IV ( 239 ), suggesting that all three ligases bind DNA substrates in a similar manner. The shape of the DBD somewhat resembles the HhH domain of bacterial DNA ligases ( Fig. 6–40 ). The similar double-stranded-DNA- binding activity of the HhH domain (cf. Fig. 6–40 ) could indicate that bacterial ligases also encircle their DNA substrates.
The 5’-adenylated DNA substrate in complex with human DNA ligase I is underwound and sharply kinked in the region of the nick ( 307 ). The DNA in the enzyme complex adopts an A-form helix adjacent to the 3’ OH end of the nick and a B-form helix adjacent to the 5’ phosphate end. DNA ligases strongly discriminate against RNA strands on the 5’ phosphate side of the nick, consistent with a requirement to form a B-form helix in order to intimately contact the enzyme. An RNA:DNA hetereoduplex would preferentially form an A-form helix that would not complement the DNA-binding surface of ligase I in this region. In contrast, an RNA strand is well tolerated on the 3’ OH side of the nick during the ligation reaction, consistent with the A-form DNA conformation in this region of the ligase-DNA complex. Thus, the A-form-to-B-form transition in DNA structure across the active site of DNA ligase I explains the strong preference for DNA ligation.
Nonconservative amino acid substitutions at Glu566 and Arg771 of DNA ligase I resulted from mutations in LIG1 of a patient with severe immunodeficiency ( 15 , 443 ). The affected residues are located near the active site and DNA-binding surface, respectively, of human DNA ligase I.
Model for AP endonuclease- and PNKP-dependent BER pathways in mammalian cells. Three BER subpathways (I, II, and III) defined by the type and reaction mechanism of DNA glycosylases are shown. Monofunctional glycosylases (M) generate AP sites, which are cleaved by APEX1 to leave a 5’-deoxyribosephosphate terminus. It is removed by Pol p, producing a single-nucleotide gap for nucleotide addition (pathway II). When NTH1 or OGG1 carry out β-elimination, APEX1 can remove the resulting 3’ dRp residue to generate a single-nucleotide gap with a 3’ OH (pathway I). With a NEIL enzyme as the initial glycosylase, a 3’ phosphate terminus is generated, which is then removed by PNKP (pathway III). (Adapted from reference 451 .)
PARP enzymes and poly(ADP) ribosylation. (A) On binding of PARP1 to DNA breaks, its catalytic activity is stimulated to add branched poly(ADP) ribose chains to nuclear protein acceptors. NAD+ is used as the source of poly(ADP) ribose, releasing nicotinamide (Nam) in the reaction. The adenine (Ade), ribose (Rib), and phosphate (P) groups in NAD+ are indicated. The major nuclear substrate for this reaction is the PARP1 enzyme itself. The resulting chains can be broken down by another enzyme, poly(ADP) ribose glycohydrolase. (B) Domain structure of two nuclear PARP enzymes from human cells. PARP1 includes a DNA-binding domain containing Zn fingers near the N terminus, a nuclear localization signal (NLS), an automodification domain to which poly(ADP) ribose is added and which contains a BRCT motif, and a catalytic domain near the C terminus. PARP2 contains a DNA binding domain and catalytic domain but no automodification domain. (Adapted from reference 171 .)
Sequential protein handoffs during BER. The singlenucleotide replacement pathway for BER is shown. The example shown is for repair of a T residue arising when 5-methylcytosine (meC) in a CpG sequence in a mammalian genome (A) is deaminated (B). TDG glycosylase removes the thymine and recruits the APEX1 endonuclease (C). APEX1 cleaves the chain on the 5’ side of the abasic site and recruits Pol p; TDG dissociates (D). Pol β releases the remnant 5’-deoxyribosephosphate (dRp), inserts a C residue, and recruits the LIG3-XRCC1 complex (E). LIG3 seals the nick as Pol β dissociates (F). The LIG3-XRCC1 complex is liberated (G). To restore the DNA to its original methylation state, a DNA methyltransferase would need to act on the newly synthesized C residue. (Adapted from reference 227 .)