Full text loading...
Chapter 8 : Nucleotide Excision Repair in Eukaryotes
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
This chapter considers the cell biology and genetics of nucleotide excision repair (NER) in eukaryotes and some biochemical properties of NER gene products. It presents experimental demonstration of NER in eukaryotic cells, and rate of removal of cyclobutane pyrimidine dimmers (CPD) and (6-4) photoproducts in yeast cells and mammalian cells. The discovery that the hereditary disease xeroderma pigmentosum (XP) affects NER in human cells established a powerful genetic framework for exploring molecular aspects of the pathogenesis of this disease and the mechanism of NER in eukaryotes. It discusses the molecular cloning of individual yeast genes for NER as well as their mammalian orthologs and considers biochemical information gleaned from the characterization of individual genes and their gene products. The majority of the genes and proteins considered in this chapter are required for or involved in steps associated with base damage recognition and bimodal incision of damaged DNA during NER in eukaryotes.
Repair synthesis of DNA during NER can be visualized by autoradiography. Cultured human HeLa cells were either unirradiated (top) or exposed to UV irradiation at 10 J/m2 (bottom) and were then labeled for 2 h with [3H]thymidine. Following autoradiography, cells in the S phase of the cell cycle show intense labeling of their nuclei ( 63 ). The remaining unirradiated cells show very low background labeling except for the occasional cell just entering or leaving S phase (top). Non-S-phase cells exposed to irradiation (bottom) show many autoradiographic grains, reflecting repair synthesis (so-called unscheduled DNA synthesis). (Adapted from reference 63 with permission.)
Complementation of the UV-radiation sensitivity of immortalized XP-A cells. Shown is the survival after UV radiation as measured by colony-forming ability in repair-proficient human fibroblasts WI38VA13 (black), parental XP-A cells (grey), and XPA gene transfectants (gold). (Adapted from reference 403 .)
Ligation-mediated PCR to detect locations of CPD. (A and B) Duplex DNA containing a CPD (triangle) is cleaved with T4 denV enzyme and denatured (A), and a gene-specific primer P1 is annealed and extended with DNA polymerase to the end of the fragment (B). (C) Linker DNA (black rectangles) is ligated to the ends of the DNA. (D) PCR is then performed with a second gene-specific primer P2 and a primer annealing to the linker. (E) The DNA products are separated on a sequencing gel and transferred to a membrane. A strand generated with gene-specific primer P3 is used as a hybridization probe to detect the PCR products.
Detection of CPD and (6–4)PP in UV-irradiated DNA at the trinucleotide sequence TC-X. CPD are converted into DSB with a 5’ phosphate group by cleavage with T4 denV and by photolyase treatment to create ligatable ends. The resulting DNA break positions can be detected by ligation-mediated PCR (see Fig. 8–3 ). (6–4)PP and their Dewar isomers are converted into strand breaks with 5’ phosphate groups by cleavage in hot alkaline buffer. Note than an amplification product derived from a (6–4)PP is 1 nucleotide shorter than the product derived from a CPD at the same dipyrimidine sequence. Only one strand of the DNA duplex is shown. (Adapted from reference 300 .)
A procedure for the isolation and end labeling of restriction fragments containing a gene, illustrated here for an RsaI restriction fragment from the MFA2 gene. (Top) Probes 1 and 2 are complementary to 3’ ends of the transcribed strand (TS) and the nontranscribed strand (NTS), respectively, of the MFA2-containing RsaI fragment. Each probe has a hexa(dT) overhang, connected to six further bases with a biotinylated residue at the 5’ end. RsaI-restricted DNA is denatured and annealed to probe 1 or 2 to isolate the MFA2 TS or NTS. (Bottom) These annealed fragments are separated from genomic DNA by using streptavidin-coated magnetic beads which bind the biotin. The beads, and consequently the bound sequences, are retained in the tube by placing a magnet at the side of the tube, while nonbound genomic DNA is washed away. The MFA2 fragments are then end labeled by incorporating six radiolabeled dATP molecules opposite the hexa(dT) sequence with DNA polymerase. These labeled fragments are eluted from the beads. To detect UV radiation-induced CPD in a fragment (represented by a triangle on the TS in the figure), the DNA is incised with T4 denV enzyme and the fragments are separated according to size on a DNA sequencing gel (see Fig. 10–18 for an example). (Adapted from reference 407 .)
Kinetics of the removal of (6–4)PP (A) and CPD (B) from the DNA of UV-irradiated CHO cells. (Adapted from reference 273 .)
Defective incision of UV-irradiated DNA in a rad3 mutant of S. cerevisiae in vivo. After UV irradiation, the wild-type (RAD) strain (A) and a rad16 mutant strain (C) generate incisions in DNA that shift the sedimentation position of DNA toward the top of the alkaline sucrose gradients. These incisions persist at 36°C because of the presence of the temperature-sensitive cdc9 mutation that prevents completion of NER by preventing DNA ligation. At 25°C DNA ligase is active and NER can be completed. This allows the formation of higher-molecular-weight DNA in wild-type and rad16 cells, which sediments faster in the gradients. No strand breaks are detected in the DNA of the rad3 mutant (B), even at 36°C, indicating that this mutant is defective in the incision of UV-irradiated DNA ( 118 ).
DNA damage-specific incision in UV-irradiated normal and XP fibroblasts. The figure shows the relative retention of DNA on filters as measured by the alkaline elution technique. Normal human fibroblasts generate transient strand breaks in their DNA during postirradiation incubation. This results in a reduced retention of the DNA on the filters following denaturation of the DNA in alkali because of the smaller size of the DNA fragments. The gradual return to higher-molecular-weight DNA associated with the completion of NER is reflected by increased filter retention. Incision of DNA in XP cells is reduced or absent as measured by this technique ( 109 ).
Detection of NER incisions by the alkaline unwinding technique. Confluent monolayers of a human fibroblast cell line were exposed to incident UV-C radiation doses of 0, 5, 10, or 20 J/m2 and then incubated at 37°C for the times indicated. At each time point, cells were lysed and exposed to alkaline conditions (0.03 M NaOH, plus 0.15 NaCl at 0°C). NER incisions present in the cells at the time of lysis serve as sites of unwinding to form single-stranded DNA that is monitored to calculate the number of breaks. The number of breaks at any time reflects ongoing NER. (Adapted from reference 94 .)
Comparison of the kinetics of repair synthesis in human fibroblasts and the loss of ESS in DNA sensitive to the T4 denV enzyme, which nicks at CPD in DNA. Cells for the repair synthesis assay were UV irradiated with a dose of 10 J/m2 and for the ESS assay with a dose of 5 J/m2.
(A) Deletion of the RAD10 gene of S. cerevisiae results in more sensitivity to killing by UV radiation than that of a rad10-1 point mutant, to a level comparable to that sustained by a rad3 mutant ( 453 ). (B) Deletion of the RAD7 gene results in less sensitivity to killing by UV radiation than deletion of the RAD10 gene does ( 299 ). (Note the different UV dose scales in panels A and B.)
A mutant strain of S. pombe with a deletion of the rad13+ gene (the S. pombe ortholog of S. cerevisiae RAD2) is only marginally sensitive to UV radiation compared with an NERdefective strain (compare the UV radiation dose to that in Fig. 8–11 ). A plasmid carrying the cloned rad13+ gene corrects the UV radiation sensitivity of the S. pombe rad13 mutant. (Adapted from reference 55 .)
Functional domains of human XPA protein. NLS, nuclear localization signal. (Adapted from reference 197 .)
Binding of human XPA protein to DNA with altered base pairing. XPA protein was incubated together with 19-mer DNA duplexes, and the protein-DNA interaction monitored by an electrophoretic mobility shift assay. One strand of each substrate was labeled with 32P at its 5’ end. The positions of bound XPA-DNA complexes (B) and free DNA fragments (F) are indicated in the autoradiographs shown. (A) DNA fragments modified by site-directed AAF or benzo[a]pyrene (B[a]P) carcinogen-DNA adducts. (B) DNA fragments contained a C4’-pivaloyl adduct (denoted by *) in the center. (Adapted from reference 53 .)
Domain organization of single-strand-DNA-binding (SSB) proteins. (A) Domain organization of SSB proteins in eukaryotes, bacteria, and archaea. Eukaryotes have a heterotrinieric SSB protein (RPA) with six OB-folds (gold boxes), four of which participate in DNA binding (A, B, C, and D). The third DNA-binding domain (labeled C) of RPA1 has a zinc-binding domain insertion (yellow box). Bacteria have a homotetrameric SSB protein with a single OB-fold in each polypeptide. The C-terminal third of E. coli SSB is a highly flexible “tail” (grey box) with an acidic terminus (black) that does not participate in DNA binding but plays a role in protein-protein interactions. Crenarchaeal SSB has a similar organization. (B) The OB-fold ( 269 ) is a five-strand, barrel-shaped structure that binds to ssDNA. Two views of an OB-fold are shown here. Most SSBs contain one or several tandem OB-fold domains. A variant of the OB-fold binds to dsDNA and is found in DNA ligases (see Fig. 6–38). (Panel A adapted from reference 192 .)
The crystal structure of two OB-fold domains from the large subunit (RPA1) of human RPA shows how basic residues from the strands and connecting loops of the OB barrel contact the DNA backbone and aromatic residues stack against the bases of ssDNA ( 33 ) (see also Fig. 8–15 ).
Interaction domains in Rad1 and Rad10 proteins. The regions between amino acids 90 and 210 of S. cerevisiae Rad10 and amino acids 809 to 997 of Rad1 are required for specific interaction between the two polypeptides.
Rad1-Rad10 endonuclease cleaves partially duplex molecules. A bubble substrate (left) was formed by annealing two 90-mer oligonucleotides. These were of complementary sequence except for the central 30 unpaired nucleotides, which had 30 T residues on one strand and 30 C residues on the other (as in Fig. 8–27 ). One strand was labeled with 32P on the 5’ end, as indicated by an asterisk. A mixture of Rad1 and Rad10 proteins was incubated with the bubble, and the products were separated on a denaturing polyacrylamide gel and subjected to autoradiography (right). Rad1-Rad10 cleaves the labeled strand on the 5’ side of the bubble at the position shown, whereas neither Rad1 nor Rad10 alone shows cleavage activity. (Adapted from reference 76 .)
Structure-specific nuclease activity of ERCC1-XPF. A 46-mer stem-loop structure (top), labeled at the 5’ end with 32P, is incised by purified ERCC1-XPF protein at the positions shown by the arrows within the duplex near the 5’ junction with the loop. Cleavage sites were analyzed by electrophoresis on a denaturing gel (bottom). (Adapted from reference 205 .)
Model for the role of Rad1-Rad10 endonuclease in recombinational repair by the single-strand annealing pathway. Following the introduction of a DSB between two homologous sequences, the broken DNA ends are resected by a 5’ → 3’ exonuclease. Complementary overlapping 3’ ends then anneal, and the structure-specific nuclease activity of Rad1-Rad10 trims off the 3’-ended unpaired single strands. The DNA is then repaired by resynthesis and ligation, resulting in a deletion product in which sequences between two direct repeats have been lost. (Adapted from reference 76 .)
Domain structure of XPF and ERCC1. XPF consists of two conserved areas (grey) separated by a less highly conserved region in the middle. The N-terminal area includes homology to SF2-family helicase domains I to III and a predicted leucine zipper (L-zipper) region. In the C-terminal region there is an area of similarity to ERCC1. Part of this region is highly similar to a region near the C terminus of archaeal helicases. In ERCC1, the first ca. 100 amino acids are dispensable and the remainder of the protein (light gold) has similarity to XPF. Some known mutations in CHO ERCC1 mutant cells and human XPF mutant cells are indicated. (Adapted from reference 120 .)
Structural organization of XPF-like nucleases. In euryarchaea (top), as exemplified by Pyrococcus Hef, the nuclease is formed by homodimerization of two molecules, each containing a helicase domain, a nuclease domain, and a domain with two HhH motifs. Independent dimerization occurs between the HhH domains (dark gold) and the nuclease domains (grey) of each monomer. In the crenarchaea (middle), as exemplified by the enzyme from Sulfolobus, the nuclease is a homodimer of a shorter protein containing only nuclease and HhH domains. This enzyme is assisted in anchoring to the DNA by crenarchaeal PCNA. The eukaryotic nuclease (bottom) is a heterodimer of two related subunits, as exemplified by human XPF-ERCC1. The larger XPF subunit contains the active nuclease domain but has disrupted helicase motifs and one disrupted C-terminal HhH domain. XPF is dimerized with a smaller ERCC1 subunit having two HhH domains and a disrupted nuclease domain. (Adapted from reference 283 .)
Highly conserved residues in the nuclease domain of XPF and Mus81 homologs. The alignment shows amino acids 712 to 724 of human XPF (Hs XPF) with the homologous regions of the XPF orthologs from mouse (Mm), Drosophila (Dm), S. pombe (Sp), and S. cerevisiae (Sc). The middle group of aligned sequences consists of putative helicase/nucleases from the euryarchaea Archaeoglobus fulgidus (Af), Methanococcus jannaschii (Mj), Methanothermobacter thermoautotrophicus (Mt), Pyrococcus abyssi (Pa), and the crenarcheaon S. solfataricus (Ss). The bottom group of sequences consists of eukaryotic Mus81 homologs. Results of alanine substitution mutations for the residues indicated in gold are shown in Fig. 8–24 . (Adapted from reference 93 .)
Nuclease activity of wild-type and mutant ERCC1-XPF. A stem-loop substrate labeled with 32P (as in Fig. 8–19 ) was incised by wild-type ERCC1-XPF (WT) or mutant ERCC1-XPF in the presence of 0.4 mM MnCl2 (A) or 2 mM MgCl2 (B). An autoradiograph is shown of the 46-mer substrate and 9- to 10-mer products separated on a denaturing polyacrylamide gel. Highly conserved Asp (D), Glu (E), Arg (R), and Lys (K) residues were changed to Ala (a) as indicated. Residues 714, 715, 716, and 720 (bold) are depicted in Fig. 8–23 . (Adapted from reference 93 .)
Structure of a thermostable homodimer of an XPF/Mus81 family enzyme from A. pernix, bound to dsDNA. The DNA in the crystal structure (shown horizontally as a stick model representation) represents part of a 3’ flap substrate. A second DNA duplex is modelled as part of the same 3’ flap substrate and is shown vertically as a backbone ribbon. The archaeal enzyme has a catalytic nuclease domain and a DNA-binding domain separated by a flexible linker. The two subunits of the homodimer are shown in two shades of gold, with twofold axes of symmetry indicated by black rods in the figure. The DNA-binding and nuclease domains become coupled on binding a DNA substrate, and the DNA is likely to be bent as part of the recognition of a double-strand DNA-single strand DNA discontinuity. (Adapted from reference 278 a.)
Proposed scheme for substrate recognition by dimeric nucleases such as the archaeal XPF/Mus81 homologs and the mammalian heterodimeric ERCC1-XPF nuclease, based on the crystal structure shown in Fig. 8–25 . In this scheme, one subunit (B) binds to the DNA substrate and presents it for cleavage by the catalytic subunit (A). For ERCC1-XPF, the catalytically inactive “nuclease” domain of the ERCC1 subunit could stabilize the fold of the XPF nuclease and/or interact with the DNA. The HhH domains of ERCC1 and XPF are a dimerization interface, and one or both HhH subunits might also bind to DNA, as seen in the crystal structure of the archaeal XPF/Mus81 homolog ( Fig. 8–25 ). Although some XPF/Mus81 homologs lack the N-terminal helicaselike domain of eukaryotic XPF ( Fig. 8–21 ), it is likely that this domain also participates in binding to DNA. The modified DNA strand labeled X that is subject to cleavage by the XPF/Mus81 nucleases may be present in a bubble substrate, a single-stranded 3’ flap, or one strand of a four-way (Holliday) junction. Different enzymes from the XPF/Mus81 family cleave these DNA structures with different efficiencies.
XPG nuclease cleaves on the 3’ side of bubble structures. Two 90-mer oligonucleotides were annealed to form a bubble structure. There is 30 bp of paired DNA on each side of the bubble; the bubble itself arises because the central part of the structure comprises 30 T residues on one strand and 30 C residues on the other. The top strand was labeled with 32Patthe 5’ end and treated with XPG, generating a 60-mer, as shown in the autoradiograph of a denaturing polyacrylamide gel. Precise size markers were generated by Maxam-Gilbert DNA-sequencing reactions (T + C and G + A). For the lanes at the right, the bubble was treated with two concentrations of S1 nuclease to nick in the single-stranded area. (Adapted from reference 95 .)
Highly conserved residues in the XPG and FEN1 nuclease family. The upper part of the figure represents the conserved N-terminal (N) and internal (I) regions of some members of the structure-specific nuclease family. The total number of amino acids in each protein is indicated. Within each region, some of the most highly conserved amino acid residues are indicated, with identical residues in grey. Human XPG nuclease is inactivated if either of the highly conserved residues Asp77 or Glu791 (gold) is changed to Ala. In XP patients XP124LO and XP125LO, the causative mutation is a change of the highly conserved Ala792 (gold) to Val. (Adapted from reference 68 .)
Protein sequence similarity between yeast Rad4 and human and Drosophila XPC. The diagram shows protein sequence similarity between the S. cerevisiae Rad4 protein and its orthologs human XPC and Drosophila Mus210. The most highly conserved region of about 370 amino acid residues is in the C-terminal part of the proteins (indicated in gold). Less highly conserved regions are present in the N-terminal part of the proteins (grey). The percentages of amino acid sequence identity and similarity are shown for these regions.
Human XPC protein is associated with RAD23 proteins and centrin 2. A protein complex purified from HeLa cells, containing XPC complementing activity, contains XPC protein tightly bound to RAD23B or RAD23A as well as to the small centrin 2 (CEN2) protein. Positions of migration of the proteins on a silver-stained gel are shown. The asterisks indicate products derived from proteolysis of the large XPC subunit. (Adapted from reference 9 .)
Functional domains in two human homologs of S. cerevisiae Rad23, RAD23A and RAD23B. The proteins have 58% identity and 67% similarity. The ubiquitin-like (UbL) and ubiquitin-associated (UbA) domains are shown, as well as a region that binds XPC protein (XPC-BD). The 398-amino-acid (aa) S. cerevisiae Rad23 protein has a similar domain organization and an overall 30% identity and 42% similarity to human RAD23B.
Preferential binding of XPC-RAD23B protein to UVirradiated DNA. 32P-labeled 136-bp duplex DNA (1 ng) was unirradiated or UV irradiated with the indicated doses and incubated with XPC-RAD23B protein and 100 ng nonspecific poly(dI-dC) competitor DNA. (Adapted from reference 27 .)
Binding of XPC-RAD23B to DNA containing small bubbles. The figure shows an autoradiograph of the products of a gel mobility shift assay using human XPC-RAD23B protein and 32P-labeled probes (about 55 bp long) containing duplex DNA, a region of three unpaired bases, or a region of five unpaired bases. (Adapted from reference 385 .)
Partial amino acid sequence alignment of proteins sharing helicase consensus motifs. Two of the seven well-conserved helicase motifs ( Fig. 8–37 ), designated I and II, are shown here. Motif I, also known as a Walker type A sequence, contains a GK(T/S) sequence that is important in binding ATP. Motif II contains a DExH sequence important in binding a divalent cation, usually Mg2+ S. cerev., S. cerevisiae; Drosoph., Drosophila.
(A) Rad3 DNA helicase activity can be demonstrated by measuring the displacement of a radiolabeled oligonucleotide annealed to complementary circular ssDNA. (B and C) Rad3 protein displaces such oligonucleotides with a strict 5’ → 3’ polarity with respect to the strand to which it is bound (the single-stranded circle) and with a pronounced pH dependence. Panel B represents an autoradiograph following gel electrophoresis, showing the displacement of the radiolabeled oligonucleotide. In lane 1, no Rad3 protein was added. Lanes 5 to 8 represent reactions at pH 5.3, 5.6, 5.9 and 6.2, respectively. Lane 2 is a control in which the substrate was heat denatured. The pH profile is shown graphically in panel C. The ATPase activity of Rad3 has the same pH requirements. (Adapted from reference 389 .)
Yeast strains with mutations in the RAD3 gene that encode polypeptides carrying amino acid substitutions in conserved helicase motifs are defective in NER, as evidenced by a marked sensitivity to killing by UV radiation ( 376 ). The rad3-20 mutant allele encodes a polypeptide in which Gly47 in the highly conserved GKT helicase motif (see Fig. 8–34 ) is replaced with Asp; the rad3-21 allele encodes a polypeptide in which Lys48 in the GKT helicase motif is replaced with Glu; the rad3-1 allele encodes a polypeptide in which Glu236 in the DExH helicase motif II (see Fig. 8–34 ) is replaced with Lys, and the rad3-24 allele encodes a polypeptide in which Gly604 (helicase motif VI [ Fig. 8–37 ]) is replaced with Arg. (Adapted from reference 376 .)
The predicted Rad16 polypeptide of S. cerevisiae contains consensus helicase family motifs (I to VI). These show extensive amino acid sequence identity to other members of the Swi/Snf family of proteins shown here. Amino acids shown in gold are identical in all or nearly all 10 polypeptides. Among the remaining amino acids, most of the differences reflect conservative substitutions. The conserved motifs are not contiguous in any of the proteins but are separated by protein sequences of different lengths (not shown here).
Electrophoretic mobility shift measurements using a radiolabeled, UV-irradiated oligonucleotide and extracts from human cells. Extracts from repair-proficient cells and cells representing different XP complementation groups cause a shift of the mobility of the DNA in the gel. Only XP complementation group E cells lack activity for a major binding factor, designated DDB. (Adapted from reference 59 .)
DDB1 as an intrinsic component of an E3 ubiquitin ligase with a wide range of substrates. DDB1 associates with other E3 ubiquitin ligase components including the cullin CUL4A and the ring finger protein RBX1 (which associates with an E2 enzyme in the ubiquitin pathway). Different partners can interact with DDB1, each of which is a “substrate adaptor” for a target of ubiquitination. The variable partners of DDB1 include DDB2 (with targets affecting the NER pathway), CSA (with partners affecting TC-NER), and V proteins of paramyxoviruses (with STAT targets affecting the interferon signaling pathway). (Provided by V. RapicOtrin and A.S. Levine.)