MMeetthhooddss iinn MMoolleeccuullaarr BBiioollooggyy TTMM VOLUME 160 NNuucclleeaassee MMeetthhooddss aanndd PPrroottooccoollss EEddiitteedd bbyy CCaatthheerriinnee HH.. SScchheeiinn HHUUMMAANNAA PPRREESSSS Processivity of DNA Repair Enzymes 3 1 Processivity of DNA Repair Enzymes R. Stephen Lloyd 1. Introduction 1.1. General Considerations of Processivity DNA repair enzymes monitor a host’s genome for structural aberrations caused by exogenous damage (ionizing or UV radiation, chemical exposure) or spontaneous dam- age (deamination, oxidation, or base loss). Lack of repair at such sites can result in replication and transcription blockage, and potentially error-prone replication bypass that could eventually lead to cell transformation. Several human diseases, such as xero- derma pigmentosum, Cockayne’s syndrome, trichothiodystrophy, and human heredi- tary nonpolyposis colon cancer, are associated with inefficient DNA repair mechanisms (reviewed in ref. 1). As these lesions occur infrequently in DNA, enzymes must dis- criminate between normal, intact DNA and damaged DNA in circumstances where the former predominates. One of the mechanisms used by DNA repair enzymes to locate damaged sites has been described as a processive search mechanism (reviewed in refs.2and3). In this context, “processive” does not relate to the activity of enzymes (DNA or RNA poly- merases or exonucleases) that create their substrate through a processive activity. Here, “processive” means that multiple catalytic events occur on a defined piece of DNA (whether a plasmid or a chromatin domain) prior to the enzyme’s macroscopic diffu- sion from the DNA to which it was initially bound. After it has been released from that DNA, the enzyme may become bound to the same or another DNA molecule, upon which multiple catalytic events can again take place. The factors that effect processivity have been extensively reviewed for several pro- teins(2–5). These studies indicate that the major attractive force between the proteins and DNA are electrostatic, but the energy of net movement between the DNA and protein comes from simple Brownian motion. Some enzymes hydrolyze ATP to track unidirectionally along DNA, but these are not discussed in this chapter. From:Methods in Molecular Biology, vol. 160: Nuclease Methods and Protocols Edited by: C. H. Schein © Humana Press Inc., Totowa, NJ 3 4 Lloyd 1.2. Discovery of In Vitro Processive Nicking Activity of DNA Glycosylase/Abasic (AP)-Site Lyase on DNA Containing Cyclobutane Pyrimidine Dimers T4 endonuclease V (now referred to as T4-pdg for pyrimidine dimer glycosylase) is an enzyme that initiates repair at UV light-induced cyclobutane pyrimidine dimers in double-stranded DNA (reviewed in refs.3and 6). The mechanism of action involves: 1. Recognition and binding to the dimer; 2. Flipping the nucleotide that is opposite the 5' pyrimidine of the dimer to an extrahelical position on the enzyme; 3. Scission at the N-C-1 glycosyl bond; and 4. Phosphodiester-bond scission via a (cid:96)-elimination reaction (AP lyase activity). The processive nicking activity of T4-pdg was originally observed by Dr. M. L. Dodson: when limiting concentrations of T4-pdg were added to heavily UV irradiated supercoiled (form I) plasmids, double-strand breaks appeared in the plasmid popula- tion (linear form III DNA) prior to all form I DNA being converted to either nicked circles (form II DNA) or form III DNA. The early appearance of plasmids containing double-stranded breaks indicated that T4-pdg incised most, if not all, dimers on one DNA molecule prior to enzyme dissociation (7). Form III linear DNA is formed by the incision at two dimers in complementary strands in close proximity. Processive nick- ing activity can be most easily measured by damaging a plasmid at multiple sites and monitoring for the accumulation of form III DNA as a function of the remaining form I DNA (Fig. 1)(7,8). This assay is described in Subheading 3.1.1.As the interactions are primarily electrostatic, increases in salt concentrations decrease processive nicking, i.e., encourage the enzyme to release its substrate before completing the reaction (Fig. 2). Processive nicking reactions can also be readily distinguished from a random accu- mulation of breaks within a population of plasmids by measuring the number average molecular weight of denatured DNAs by either denaturing agarose gel electrophoresis or velocity sedimentation in alkaline sucrose gradients following enzymatic treatment. These techniques will be described in Subheadings 3.1.2. and 3.1.3., respectively. 1.3. Discovery of Processive Plasmid Repair Within Intact Cells The processivity of enzymes, such as T4-pdg, as described above, is not merely an in vitro artifact of well-controlled biochemical enzymology, but rather is operative in intact, living cells (9,10). Dr. Elliott Gruskin pioneered the concepts and methodolo- gies to obtain evidence for the processivity of T4-pdg and the Escherichia colinucle- otide-excision repair system. His experimental procedure (described in Method 4) was to irradiate cells harboring plasmids in order to introduce 5 or 10 pyrimidine dimers per plasmid within E. coli, that contained limiting concentrations of T4-pdg, and then mea- sure the kinetics of the accumulation of fully repaired plasmid molecules. This experi- mental design assumed that the rate-limiting step in the complete repair of a dimer site was the recognition of the lesion and incision at that site. The subsequent steps of AP endonucleolytic cleavage, polymerization, and ligation were presumed to be rapid. If T4-pdg initiated repair at all dimer sites within a subset of plasmids, then that subset of plasmids should be completely free of dimers prior to any repair in the remaining plas- mids. Thus, if repair occurs processively, the kinetics of the accumulation of damage- Processivity of DNA Repair Enzymes 5 Fig. 1. (A)Time course analysis of T4 endonuclease V-nicking of form I DNA containing 10 or 25 dimers per molecule. Endonuclease V at 0.5 ng/µL was reacted in 25 mMNaCl with form I DNA containing 10 and 25 dimers per molecule. The three topological forms of DNA are as follows: form I DNA, supercoiled covalently closed circular DNA ((cid:1), 10 dimers per molecule; (cid:2), 25 dimers per molecule); form II DNA, nicked or open circular DNA ((cid:3), 10 dimers per molecule; (cid:4), 25 dimers per molecule); and form III DNA, monomer length linear DNA ((cid:5), 10 dimers per molecule; (cid:6), 25 dimers per molecule). (B) Change in the rate of disappearance of form I DNA as a function of the average number of dimers per molecule. The negative natural logarithm of the mass fraction of form I DNA was calculated for each time point taken in the time course endonuclease V-nicking assay (A) on DNA containing 10 ((cid:1)), or 25 ((cid:2)) dimers per molecule. free plasmids will be linear with repair time, and the rate of accumulation of fully repaired DNAs will be inversely proportional to the UV dose. If these assumptions were valid, the repair rate of plasmids containing five dimers should be twice as fast as for DNAs con- taining 10 dimers on average. In contrast, if repair occurred randomly within a cell, then the kinetics of “distributive” repair would be characterized by a significant time lag to the accumulation of fully repaired plasmid molecules (seeFig. 3). Intracellular plasmid repair kinetics initiated by T4-pdg displayed all the predicted characteristics of processive nicking at dimers and a processive completion of the base- excision repair pathway (9). Additionally, repair initiated by photolyase, an enzyme known to have low affinity for nontarget DNA, and thus predicted not to repair dimers processively, appeared to be distributive (10). For photolyase, fully repaired plasmid 6 Lloyd Fig. 2. Modulation of the processive nicking activity of T4 endonuclease V on dimer-con- taining form I DNA through changes in the NaCl concentration of the reaction.Time course reactions were analyzed in which form I DNA containing 25 dimers per molecule were incu- bated with endonuclease V at 0.17 ng/µL in various NaCl concentrations. (A)Loss of form I DNA at the following NaCl concentrations: (cid:1), 0 mM;(cid:2), 25 mM;(cid:7), 50 mM; and (cid:1), 100 mM. (B)Accumulation of form II DNA at the following NaCl concentrations: (cid:3), 0 mM;(cid:4), 25 mM; (cid:8), 50 mM; and (cid:3), 100 mM.(C)Accumulation of form III DNA at the following NaCl concen- trations:(cid:5), 0 mM;(cid:6), 25 mM;(cid:9), 50 mM; and (cid:5), 100 mM. Processivity of DNA Repair Enzymes 7 Fig. 3. In vivo quantitation of accumulation of endonuclease V-resistant form I DNA. In vivo kinetic analyses of endonuclease V-initiated excision repair were performed on uvrA– recA–denV+E. coligrown at 40°C in Luria broth containing 100 µg/mL ampicillin, 0.8% glu- cose, and 5 mCi/mL [3H]thymidine.E. coliwas UV-irradiated and incubated at 37°C. Plasmid DNA was isolated at the specified time points and treated in vitro with purified endonuclease V. Reaction products were resolved by electrophoresis through 1.2% agarose gels. Accumula- tion of endonuclease V-resistant form I DNA was determined by liquid scintillation spectros- copy of dissolved agarose slices containing form I, II, and III DNAs at each time point. (cid:1), uvrA–recA–denV+, UV dose of 450 J/m2;(cid:2),uvrA–recA–denV+, UV dose of 900 J/m2. molecules accumulated only after several hours time lag, but then accumulated expo- nentially. Repair initiated by the nucleotide excision-repair system, UvrABC, showed characteristics of limited processivity(10). 1.4. Biological Significance of Processive Repair Since electrostatic interactions are the primary attractive forces between the repair enzymes and DNA, it was hypothesized that neutralization of basic residues on the surface of T4-pdg that binds DNA could decrease in vitro and in vivo processivity of the mutant enzyme, without necessarily diminishing catalytic activity. Dr. Diane Dowd was first in testing this hypothesis by creating a series of site-directed mutants of T4-pdg that retained full catalytic activity, but were compromised with respect to their processivity(11–13). Since no crystal structure was known for T4-pdg at the time of these studies, the choice of which sites to mutate was based on molecular modeling analyses. These predictions utilized circular dichroism data and turn prediction com- puter programs of all (cid:95)-helical proteins to identify probable helix nucleation sites. Helix wheel diagrams suggested (cid:95)-helices with a series of positively charged residue along one face of the helix, and these sites were mutated. These studies revealed a good correlation between in vitro and in vivo processivity assays, and showed that those mutants with diminished processivity had significantly reduced survival following UV challenge as measured by colony-forming ability. Specifically, some mutant enzymes were created that incised heavily irradiated plasmids by a distributive mechanism, even at low salt concentrations (Fig. 4). Mutant enzymes that displayed this loss in processive nicking activity (Fig. 5) were also unable to enhance UV survival in DNA 8 Lloyd Fig. 4. Analysis of T4 endonuclease V-nicking of form I DNA containing 20–25 dimers/ molecule.Cellular lysates containing endonuclease V were added to 1.0 µg of UV-irradiated [3H] pBR322 in 20 µL 10 mMTris-HCl (pH 8.0), 1 mMEDTA, and 10 mMKCl. Solutions were incubated at 37°C for 30 min. Shown is a representative graph; measurements were repro- ducible to within 10%. (cid:4), endonuclease V (wild-type); (cid:1), Gln-26; (cid:5), Gln-33; (cid:6), Gln-26,33. repair-deficient cells (Fig. 6). These data were the first that demonstrated the biologi- cal significance of nontarget DNA binding. Attempts to enhance the nontarget DNA binding of T4-pdg by converting neutral amino-acid side chains to basic ones were successful in increasing nontarget DNA binding, but these mutants were unable to enhance UV survival above that of the wild type enzyme and in some cases could depress survivals(14–16). 2. Materials 2.1. Buffers and Solutions 1. 10 mM Tris-HCl, pH 7.5, 1 mM EDTA. 2. 1 M NaCl. Processivity of DNA Repair Enzymes 9 Fig. 5. Analysis of in vivo plasmid repair.Cells expressing endonuclease V wild type, Gln-26, Gln-33, or Gln-26,33 were harvested, resuspended in 43 mM Na HPO , 20 mM KH PO , 9 mM 2 4 2 4 NaCl, and 19 mMNH Cl. They were irradiated at 900 J/m2at 0°C, mixed with LB media, and 4 incubated at 37°C for varying times in order to initiate repair. Plasmid DNA was extracted, incu- bated with endonuclease V for 1 h at 37°C, and analyzed for endonuclease V-resistant DNA. These data were the average of two experiments. (cid:4), endonuclease V (wild-type); (cid:1), Gln-26; (cid:6), Gln-33; (cid:5), Gln-26,33. 3. 10 mMTris-HCl (pH 7.5), 1 mMEDTA, 100 mg/mL bovine serum albumin (BSA), vary- ing [NaCl]. 4. Reaction stop buffers: a. For neutral agarose gels: 10 mMTris-HCl (pH 7.5), 1 mMEDTA, 20% (w/v) sucrose, 2% (w/v) SDS, 0.2% (w/v) Bromophenol blue. b. For denaturing agarose gels: same as item 4a, except also contains 30 mM NaOH. c. For velocity sedimentation through alkaline sucrose gradients: same as item 4aexcept also contains 100 mM NaOH. 5. 1X TAE. 6. 100 µg/mL ethidium bromide; store in brown bottle. 7. 30 mM NaOH, 1 mM EDTA. 8. Luria broth: 10 g bacto-tryptone, 5 g bacto-yeast extract, 10 g NaCl/L deionized H O pH 2 to 7.0 with 5 g NaOH sterilized by autoclaving. 10 Lloyd Fig. 6. Colony-forming ability of UV-irradiated repair-deficient E. coli containing denV+,denV–, or mutant denV constructs.Shown is the average of three measurements. (cid:2), AB2480 with pGX2608-denV–; (cid:4), denV+; (cid:1), denV Gln-26; (cid:6), denV Gln-33; (cid:5), denV Gln-26,33. 9. 2X SSC = 17.53 g NaCl, 8.82 g sodium citrate, per 1 L deionized H O; pH to 7.0 with 2 NaOH. 10. 5% alkaline sucrose = 5% (w/v) sucrose. 11. 20% alkaline sucrose = 20% (w/v) sucrose. 12. Buffered salt solution for irradiation of intact E. coli cells: 43 mM Na HPO , 22 mM 2 4 KH PO , 9 mM NaCl, 19 mM NH Cl. 2 4 4 13. Cell lysis suspension buffer: 50 mMTris-HCl (pH 8.0), 50 mMEDTA, 8% (w/v) sucrose, 5% (v/v) Triton X-100. 14. 10 mg/mL lysozyme in H O (Sigma Chemical). 2 15. 2 propanol. 16. Buffer-saturated phenol. 17. Chloroform/isoamyl alcohol (24:1). 18. 2.5 M sodium acetate. 19. 95% ethanol. 20. 10 mMTris-HCl (pH 8.0), 0.2 mMEDTA, 100 mMNaCl, 5% (v/v) glycerol, 2 mg/mL BSA. 21. 14C or 3H-labeled thymidine. 22. Scintillation fluid. 23. E. coliuvrA–recA– cells (i.e., AB2480). 2.2. Equipment 1. Short-wave (254 nm) UV lamp (Spectroline [Westbury, NY] model EF-16 spectronics). 2. Short-wave (254 nm) UV monitor (International Light, Newbury Port, MA) radiometer/ photometer. 2.3. Enzyme 1. T4-pdg is commercially available through Pharmingen (San Diego, CA). Processivity of DNA Repair Enzymes 11 3. Methods 3.1. In Vitro Plasmid-Nicking Assay 3.1.1. Analyses by Native Agarose Gel Electrophoresis 3.1.1.1. PREPARATIONOF UV-IRRADIATED DNA 1. Dialyze covalently closed circular DNA (form I DNA) into 10 mMTris-HCl (pH 7.5), 1 mM EDTA. 2. Dilute DNA to 0.1 µg/µL in 10 mMTris-HCl (pH 7.5), 1 mMEDTA; aliquot-sufficient DNA for the complete experiment, plus an additional 10% to serve as a master mixture. 3. Prewarm short-wave UV lamps at least 15 min prior to irradiation of DNA. 4. Adjust UV lamp height so that at the height where the DNA is to be exposed, the UV meter reads 100 µW/cm2. 5. Using continuous stirring, expose DNA to UV light (100 µW/cm2) for varying times in which a 4 kb plasmid will accumulate 1 cyclobutane pyrimidine dimer per plasmid DNA mol in 10 s (7). 6. Remove 50-µL aliquots of the irradiated DNA since for each experimental variable condi- tion (different salt concentrations; enzyme concentrations, and so on), at least five time points will be needed, each containing 1 µg DNA. 7. Add an equal volume of 10 mMTris-HCl (pH 7.5), 1 mMEDTA, 100 µg/mL BSA; this buffer will also contain varying NaCl concentrations to adjust the final salt between 0.01 and 0.2 M. 3.1.1.2. INCISION KINETICS 1. Prewarm the 100-µL master mixture to the desired temperature for at least 5 min. 2. Dilute pure T4-pdg in 10 mMTris-HCl (pH 7.5), 1 mMEDTA, 100 µg/mL BSA (nano- gram quantities of T4-pdg are generally sufficient to incise microgram quantities of DNA for 15 min at 37°C). 3. Remove a 20-µL DNA aliquot and add it to 20 µL of the reaction stop buffer (10 mMTris- HCl, pH 7.5, 1 mM EDTA, 20% [w/v] sucrose, 2% [w/v] SDS, 0.2% [w/v] bromophenol blue) for a no enzyme control. 4. Add 5 µL T4-pdg to the remaining 80-µL master reaction mixture, mix, and return to appropriate temperature. 5. Remove 21-µL aliquots at selected times and add to an equal volume of reaction stop buffer. 3.1.1.3. DATA ANALYSIS 1. Separate the form I, II, and III DNAs by electrophoresis (4 h at 100 V constant) through a 1% horizontal agarose gel, and run in 1X TAE buffer. 2. Stain agarose gel for at least 2 h in 1X TAE supplemented with 0.5 µg/mL ethidium bro- mide; avoid prolonged exposure to visible or UV light by covering staining container with aluminum foil. 3. Place fully stained gel on a short-wave UV light box and capture the image, using a cam- era attached to an imaging system, such as Appligene Imager, in which the data can be analyzed and quantitated as a TIFF file. 4. Prior to quantitating the percentiles of forms I, II, and III DNAs, the raw data values obtained for form I DNA should be multiplied by 1.42, a correction factor that normalizes for the reduced binding of ethidium bromide by covalently closed circular DNA.