DNA Excision Repair Assays

David Mu; Aziz Sancar    Department of Biochemistry and Biophysics, University of North Carolina School of Medicine, Chapel Hill, North Carolina, 27599

DNA lesions at specific sites in the genome can cause mutation or induce recombination, and may result in other DNA rearrangement reactions. These changes can ultimately lead to cancers. It has been estimated that 70–80% of cancers are caused by endogenous or exogenous agents that damage DNA (1). Similarly, many drugs used in cancer chemotherapy are DNA-damaging chemicals. Some patients and tumors are not responsive to these drugs, whereas others, after an initial favorable response, become refractory. It has been suggested that elevated DNA repair activity contributes to drug resistance (2). Thus, it is of clinical and scientific significance to understand the molecular mechanisms that repair DNA.

Of all the DNA repair mechanisms, nucleotide excision repair is probably the most important in view of the wide variety of DNA lesions that can be acted on by excision repair. In nucleotide excision repair, the damage is removed from DNA in the form of 12–13 nucleotides (prokaryotes) or 24–32 nucleotides (eukaryotes), by dual incisions of the damaged strand through an ATP-dependent multisubunit enzyme system we refer to as excision nuclease (or excinuclease). Defective nucleotide excision repair gives rise to an autosomal recessive hereditary disorder called xeroderma pigmentosum (XP) (3). From cell fusion studies, this disease was found to be genetically heterogeneous and classified into complementation groups A through G (4–5). Proteins defined by these seven complementation groups are a part of the excision nuclease, which is the operational definition for dual incision activity that requires all seven subunits (6).

In recent years, important advances in excision repair have significantly increased our understanding of DNA repair. Detailed accounts of these advances have been documented (5, 7, 7a). It is the purpose of this article to review the various tools, i.e., repair assays, used to study excision repair. Some of these assays have been in use for many years and others have been developed recently and have been instrumental in the rapid progress in the enzymology/molecular biology of excision repair.

All of the repair assays are broadly classified into two categories, in vitro and in vivo, although in some cases the line between them is blurry. We discuss the theoretical principles of the various assays, their specific use, and their advantages and disadvantages. Because this review is not intended to be a laboratory manual, no attempt is made to describe the technical details of various assays.

I In Vitro Assays

A Nicking/Incision Assay

This assay measures the damage-dependent incisions of DNA. The earliest and still widely used version of this method (endonuclease-sensitive site assay) measures the average size distribution of DNA in alkaline sucrose gradients, following treatment with T4 endonuclease V, which incises at the sites of pyrimidine dimers (8). A popular version of the nicking assay is based on the conversion of covalently closed circular, supercoiled plasmid DNA into a nicked, relaxed form. The conversion is commonly monitored by three methods: alkaline sucrose gradient (9), nitrocellulose filter binding (10), and agarose gel electrophoresis (11). Although this nicking assay can be carried out with relative ease, it does not detect incisions at the nucleotide level, nor can it distinguish a repair endonuclease such as T4 endonuclease V from a repair excision nuclease such as the Uvr(A)BC excinuclease of Escherichia Coli (12). To circumvent this problem, linear DNA fragments, containing damage randomly distributed throughout the DNA or at a specific position, are labeled only at either the 5′ or the 3′ end and subjected to the action of repair proteins. The incised products are then analyzed using denaturing acrylamide gels to visualize the precise incision sites. Although linear DNAs containing damaged nucleotides at random sites [obtained by exposing the DNA to irradiation or to various model carcinogens such as psoralen, 2-(N-acetoxyacetylamino)fluorene (AAAF), (+)anti-benzo(a)pyrene-7,8-dihydro-diol-9,10-epoxide (BPDE), or cisplatin] have been used as substrates (13), uniquely modified DNAs are a better choice because they offer more defined analysis of the excision reaction, such as the order of incision, and provide more unambiguous data.

B Excision Assay

As shown in Fig. 1, nucleotide excision repair is generally considered a two-stage event: damage-guided dual incision (excision) and repair synthesis. Excision assay refers to the method of detecting the damage-carrying oligonucleotide as a result of the first stage. At least three isotopic labeling schemes have been put forth to detect the excised, lesion-containing oligomer: (1) A radiolabel is incorporated in the vicinity of the lesion in a synthetic substrate, such that the released oligomer carries the label and can be resolved on a sequencing gel (14). (2) The substrate is not radiolabeled. However, following the repair reaction, the excised oligomer is radiolabeled by deoxynucleotidyl terminal transferase before separating on a sequencing gel (13, 15). (3) The substrate is not radiolabeled. Following excision, the products are separated on a sequencing gel and the excised fragment is located by Southern hybridization (16).

Method 1, employing an internally isotopically labeled DNA substrate, is commonly used in the authors’ laboratory because it is superior to others in terms of simplicity, sensitivity, and specificity. Furthermore, this is the only assay that allows one to carry out rigorous quantitative analysis of the excision nuclease in cell-free extracts or reconstituted systems. When the products of the excision nuclease are examined by denaturing polyacrylamide gels, radiolabeled oligomers containing the lesion will appear as a result of the two nicks, 5′ and 3′ to the damage. In essence, excision assay is identical to incision assay in terms of the experimental procedure. The only difference lies in the radiolabel position in the substrate DNA. The incision assay requires terminally labeled DNA whereas the excision assay employs a substrate that is internally radiolabeled at a phosphodiester bond in the vicinity (5′ or 3′) of the lesion, such that the dual incision will release the damage and the radiolabel in the same fragment, which can be analyzed using denaturing electrophoresis.

The nature of the damage used to synthesize the substrate DNA for either incision or excision assay generally does not affect the assay because the excision nuclease practically excises any type of damage (7). The choice is normally governed by the availability of the lesion in the precursor form ready for phosphoramidite chemistry so that the particular damage can be incorporated into an oligonucleotide by a commercial oligonucleotide synthesizer (17). Subsequently, this lesion-carrying oligomer is assembled into a longer, double-stranded DNA through annealing and ligation with other oligomers (18, 19). Cholesterol-DNA (19), biotin-DNA adducts (J. Reardon, personal communication), and UV photoproducts such as cis-syn-cyclobutane thymine dimer (17) are among those that are routinely incorporated into oligomers using phosphoramidite chemistry.

The other method of making a short oligonucleotide containing a defined lesion is to damage the oligomer of a special nucleotide sequence that contains a site hypersensitive to a particular compound (20). Following the reaction, the desired oligomer product is isolated by gel electrophoresis or high-performance liquid chromatography. A good example is platinated oligonucleotides, which are usually generated by reacting DNA with the anticancer drug, cisplatin, forming an intrastrand adduct, usually at (GpG), (ApG), and to a lesser extent (GpTpG) sites (21). A second consideration regarding substrate preparation is the substrate length. For the bacterial excision nuclease, a DNA fragment as short as 40 nucleotides is sufficient for excision to take place (22), whereas the minimal length for the human enzyme is 100 nucleotides (23).

In addition to the linear substrate, covalently closed circular plasmid DNA of several Kilobases is also utilized as internally labeled substrate, despite the fact that it is more laborious to make such a substrate. In fact, the excision assay for the human excision nuclease was originally...