Full text loading...
Chapter 9 : Mechanism of 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 discusses the molecular mechanism of nucleotide excision repair (NER) in eukaryotes, with emphasis on the reaction mechanism in mammalian cells and in the budding yeast, Saccharomyces cerevisiae. The NER mechanism is highly conserved in eukaryotes, and most components and features of the reaction mechanism are very similar in these two organisms. A cell-free system that reflects NER in mammalian cells was developed in the late 1980s and was followed by experimental approaches that measure repair synthesis or the excision of damage-containing oligonucleotide fragments in extracts from mammalian cells and yeast. These techniques have served to identify and track proteins required for NER and have provided specific assays for the purification of NER proteins. The results of genetic studies and the biochemical systems have facilitated the reconstitution of the mammalian and yeast NER machinery with purified protein components and DNA molecules containing single lesions placed at specific sites. To set the stage for the discussion to follow, it is useful to first summarize the reconstitution results. The incision step of NER in the yeast S.cerevisiae has been reconstituted with UV-irradiated DNA and a set of proteins comprising Rad14, Rad4-Rad23, RPA, TFIIH, and the nucleases Rad2 and Rad1-Rad10. The chapter also discusses mechanism of assembly and action of the NER machinery, modulation and regulation of NER in eukaryotes, and evolution of the eukaryotic NER system.
Schematic model for the NER pathway in mammalian cells. A lesion causing some disruption of the duplex DNA structure (represented by the circle) is bound by XPC-RAD23B, in an initial step termed distortion recognition. In the second step, a preincision complex is formed surrounding the lesion in an ATPdependent reaction involving TFIIH, XPA, and RPA. If a site of damage is located in this step, dual incision takes place on the damaged strand by XPG on the 3’ side of the adduct and by ERCC1-XPF on the 5’ side. This incision releases a fragment of about 27 nt, and the resulting gap is filled by a DNA polymerase (pol) holoenzyme and sealed by DNA ligase.
NER synthesis in vitro using fractionated cell extracts. Extracts from CHO cells of different rodent cell genetic complementation groups were fractionated on a phosphocellulose column to produce a CFII fraction and mixed with purified RPA and PCNA proteins in buffer containing deoxynucleotides and [32P]dATP. The reaction mixture included two plasmids of slightly different size, one UV irradiated (+ UV) and one unirradiated (— UV). DNA isolated from the reaction mixtures was linearized with a restriction enzyme and separated by agarose gel electrophoresis. (Top) Ethidium bromide-stained agarose gel. (Bottom) Auto-radiograph. Fractionated cell extracts were from CHO cells defective in NER complementation groups 1, 2, 3, or 4 as indicated (defective genes ERCC1, XPD [ERCC2], XPB [ERCC3], or XPF [ERCC4], or mixtures of extracts from two different groups as indicated). (Adapted from reference 16 .)
Electron micrographs showing repair synthesis patches in plasmid DNA. The plasmid DNA was irradiated with UV light and then incubated with human cell extracts and buffer that included biotinylated dUTP instead of dTTP. The plasmid DNA was purified and prepared for electron microscopy. Grids were soaked in the presence of streptavidin conjugated to 10-nm-diameter collodial-gold particles. The plasmids are 1 μm in circumference. On the left is a dark-field view, and on the right is a bright-field view. (Adapted from reference 237 with permission of Elsevier.)
Dual-incision pattern for NER in human cells. At a T-T CPD in DNA, the human NER system incises the damaged strand at the 5th or 6th phosphodiester bond on the 3’ side and at the 22nd, 23rd, or 24th phosphodiester bond on the 5’ side. This generates an oligonucleotide 26 to 29 nt long, containing the damage ( 101 ).
Detection of NER incisions in eukaryotes at lesions placed at specific sites in DNA. (A) Detection of specific incision events by using plasmids containing an internal 32P label near a single adduct. (Left) An oligonucleotide containing a single adduct (closed circle) is labeled with 32P, incorporated into closed-circular duplex DNA, and purified. Dual incisions (as shown by the arrows) release products ca. 24 to 30 nt long that can be (right) separated by gel electrophoresis. In this example, DNA containing an AAF-guanine adduct was used, positioned within a 5-nt unpaired “bubble structure” (B5 AAF+). A control DNA substrate contained only the bubble structure, without the AAF adduct (B5 AAF—). The DNA samples were incubated with whole-cell extract from XPC-defective cells, supplemented with the indicated amounts of purified XPC-RAD23B protein. Molecular mass markers are shown at the left (sizes in nucleotides). (B) Detection of incision events by the end-labeling method. (Left) An unlabeled oligonucleotide containing a single adduct (closed circle) is incorporated into closed-circular duplex DNA and purified. Dual incisions (as shown by the arrows) release oligonucleotide products ca. 24 to 30 nt long. The oligonucleotides produced by NER are annealed to a template with an extension of four G residues and labeled with [32P]dCTP by extension with DNA polymerase. (Right) The products, separated by gel electrophoresis, are in the range of 27 to 35 nt long, 4 nt longer than the original excised products. In this example, the adduct was a cisplatin adduct, incubated for the indicated times with a CFII fraction from HeLa cells and purified RPA. Positions of molecular mass markers are shown at the left (nucleotides). (Panel A adapted from reference 229 . Panel B adapted from reference 219 .)
NER by fractionated extracts from human cells. (A) Fractionation of human cell extracts as shown to yield CFIA (containing RPA), CFIB (containing PCNA), and CFII (containing XP proteins and other factors required for NER). (B) Repair synthesis in UV-irradiated plasmid DNA is observed by autoradiography only when all three fractions (CFIA, CFIB, and CFII) are mixed or when fraction CFIA is replaced with purified RPA. (Top) Ethidium bromide-stained agarose gel; (bottom) autoradiograph. Similar experiments show that fraction CFIB can be replaced by purified PCNA. (Adapted from reference 218 .)
Reconstitution of the NER dual-incision reaction with proteins from human cells. (A) Stained gels of purified human NER incision factors. Recombinant XPA, RPA, and ERCC1-XPF proteins were purified after expression in E. coli. Recombinant XPC-RAD23B and XPG proteins were purified after expression in insect cells. TFIIH was purified from HeLa cells. Numbers at left of gels are in kilodaltons. (B) Dual incision at DNA damage with the purified proteins shown in panel A. A plasmid containing a single 1,3 (dGpTpG) cisplatin intrastrand adduct was incubated with XPA, RPA, XPC-RAD23B, XPG, ERCC1-XPF, and/or TFIIH, and the products of dual incision were detected as in Fig. 9–5B . Reaction mixtures with all six factors are labeled “complete,” and repair factors were individually omitted as indicated in the other lanes. (Adapted from reference 5 .)
Reconstitution of the NER incision reaction with proteins from S. cerevisiae. (A) Stained gels of purified yeast NER factors. Recombinant proteins as indicated were expressed in yeast, and are shown with molecular mass markers (lane M). (B) Reconstitution of yeast NER incision on UV-irradiated closed-circular DNA. Incision of the plasmid causes it to migrate as a nicked form on the ethidium bromide-stained agarose gel, as shown. Plasmids were either UV irradiated or not, as indicated, and incubated either with all factors from panel A (factors omitted, None) or with omission of the indicated NER factor. (Adapted from reference 75 .)
TFIIH subunits in human cells and in the budding yeast S. cerevisiae. There are 10 protein subunits of TFIIH in each organism. Each subunit is shown opposite its ortholog; for example, human XPD and S. cerevisiae Rad3 are orthologs. “Subcomplexes” of TFIIH are described in the literature, including five subunit associations in each organism that are resistant to multiple purification procedures, and other groupings as indicated.
Ring structures formed by TFIIH from S. cerevisiae and human cells. (A) View of the three-dimensional reconstruction of core TFIIH at 18-A resolution. The shading identifies regions corresponding to Rad3 (top), Ssl1 (middle), and Tfb1, Tfb2, and Tfb3 (bottom). (B) Overall shape of human TFIIH as derived from electron microscopy. The positions of subunits CDK7 (cdk7), XPD, XPB, and GTF2H2 (p44) are inferred by immunolabeling experiments and are indicated by arrows. The positions of GTF2H1 (p62) and GTF2H4 (p52) are not assigned on the model. (Panel A adapted from reference 27 . Panel B adapted from reference 215 .)
Model for the promoter melting function of TFIIH in RNA Pol II transcription. DNA near the transcription initiation site is tightly wrapped around TFIIH. Here, the XPB subunit of TFIIH is shown beginning to unwind or “melt” the DNA in preparation for transcription initiation. (Adapted from reference 38 .)
Diagram showing the extent of helix opening during NER around a single 1,3-d(GpTpG)-cisplatin adduct (Pt). T residues sensitive to treatment with a combination of potassium permanganate and hot piperidine are indicated in gold. The arrows indicate the experimentally determined sites of 3’ and 5’ incision at this adduct. (Adapted from reference 33 .)
The DNA helicase activity of Rad3 protein is inhibited by the presence of base damage in the DNA strand to which it binds and translocates in the 5’ → 3’ direction (gold). When the base damage is placed on the opposite strand (grey), no inhibition is observed. (Adapted from reference 166 .)
Both a distortion and a chemical modification of DNA are necessary to create a good NER substrate. At the top is a diagram of an unmodified backbone with a normal deoxynucleotide and a backbone containing a selenophenyl adduct at the C4’ position of the deoxyribose. This adduct does not in itself markedly distort the DNA structure. The gels (below) show the result of measuring NER capacity by monitoring the release of dual-incision products (similar to the assay in Fig. 9–5 ) from a 147-nt substrate containing the indicated alteration. Neither the nondistorting selenophenyl adduct nor a distorting bubble without an adduct is repaired (first three lanes), but an adduct combining these two features is recognized and repaired by NER, as shown in the last two lanes. (Adapted from reference 90 .)
(A) Diagram of some of the interactions between NER core factors in mammalian cells. Established interactions confirmed independently in several laboratories are indicated. (B) Diagram of some of the interactions between NER core factors in the yeast S. cerevisiae. Established interactions confirmed independently in several laboratories are indicated. Interactions with individual subunits of TFIIH are not shown in detail.
Order of events leading to dual incision by NER factors in mammalian cells. After the events shown here, the gap resulting from release of a fragment containing the damage is filled in by a DNA polymerase holoenzyme, as indicated in Fig. 9–1 .
Principle of a photobleaching technique for quantification of the immobilized fraction of a molecule in living cells. (A) A theoretical pattern showing a confocal plane (gold disk) before photobleaching, where fluorescently labeled molecules (small gold spheres) are homogeneously distributed. (B) A laser beam (cones) is focused in the center of the nucleus during spot bleaching. In this case, a bleach time of 4 s was used, during which time mobile molecules diffuse through the laser cone area. (C) Theoretical fluorescence patterns showing the result of 100% immobile, 0% immobile, or 33% immobile fluorescent molecules. (Adapted from reference 99 .)
Individual mobility of GFP-tagged NER proteins demonstrated by fluorescence recovery after photobleaching. A strip of the nucleus is subjected to a 0.1-s pulse of photobleaching, and then the relative fluorescence within the strip is monitored as a function of time. Different GFP-tagged NER factors show different rates of recovery of fluorescence within the strip. (Adapted from data provided courtesy of W. Vermeulen.)
Localized irradiation to visualize assembly of NER factors at damaged sites. (Top) Strategy for irradiation of cells in distinct areas of cell nuclei by placing a UV-blocking filter containing 3-μm pores over human fibroblast cells attached to a coverslip. (Bottom) Schematic showing results of irradiating cells with UV light at 30 J/m2 (or no UV as a control), fixing the cells 15 min later, and staining using antibodies (Ab) against XPC protein or XPA protein. Both XPC and XPA proteins are concentrated in irradiated areas in normal cells (first row). In XP-A cells, XPC protein becomes concentrated in irradiated areas (second row). In XP-C cells, XPA protein does not become concentrated in irradiated areas (third row). (Data from reference 250 .)
The PCNA heterotrimer interacts with a variety of proteins containing the consensus motif QXX(Hp)XX(Ar)(Ar) (in which Hp represents a hydrophobic residue, Ar represents an aromatic residue, and X represents any residue) at a binding pocket located between the two domains of each subunit of the PCNA trimer. Here, the crystal structure of PCNA complexed to a peptide derived from the cell cycle inhibitor p21 (gold) ( 74 ) shows the location of the binding site for protein-protein interactions.
(A) The ubiquitin-proteasome pathway uses elaborate enzymatic machinery to degrade proteins. The crystal structure of the 20S core particle of the S. cerevisiae proteasome shows that its 28 subunits are arranged as four stacked rings ( 73 ). Substrates can enter through either end of the core particle to access catalytic sites located in the central cavities of the complex. (B) The complete 26S proteasome consists of the 20S core catalytic particle and a 19S regulatory particle. Electron microscopy studies show that the 19S assembly binds on either end of the 20S core. The 19S assembly consists of a “lid” and a “base” region that is in contact with the core. Multiple Rpn proteins (shown with the prefix n in the figure) are components of the lid. The base is composed of a hexameric ring of AAA ATPases (Rpt1 to Rpt6, designated t1 to t6 in the figure) and the proteins Rpn1 and Rpn2. Rad23 and Rpn10 associate with the base and promote binding by substrates containing polyubiquitin chains ( 249 ). (Panel B adapted from Glickman et al. [ 70a ] with permission of Elsevier.)