Involvement of Molecular Chaperonins in Nucleotide Excision Repair

UvrA is one of the key Escherichia coli proteins involved in removing DNA damage during the process of nucleotide excision repair. The relatively low concentrations (nanomolar) of the protein in the normal cells raise the potential questions about its stability in vivo under both normal and stress conditions. In vitro, UvrA at low concentrations is shown to be stabilized to heat inactivation by E. colimolecular chaperones DnaK or the combination of DnaK, DnaJ, and GrpE. These chaperone proteins allow sub-nanomolar concentrations of UvrA to load UvrB through >10 cycles of incision. Guanidine hydrochloride-denatured UvrA was reactivated by DnaK, DnaJ, and GrpE to as much as 50% of the native protein activity. Co-immunoprecipitation assays showed that DnaK bound denatured UvrA in the absence of ATP. UV survival studies of a DnaK-deficient strain indicated an 80-fold increased sensitivity to 100 J/m2 of ultraviolet light (254 nm) as compared with an isogenic wild-type strain. Global repair analysis indicated a reduction in the extent of pyrimidine dimer and 6–4 photoproduct removal in the DnaK-deficient cells. These results suggest that molecular chaperonins participate in nucleotide excision repair by maintaining repair proteins in their properly folded state.

The UvrA protein is an essential subunit of UvrABC endonuclease (1) and plays an important role in DNA damage recognition and nuclease assembly (2). Binding of UvrA to damaged DNA is specific for a local conformational change at the site of the damaged base (2)(3)(4). The primary mechanistic function of UvrA in the UvrABC system is to interact with UvrB to form a heterotrimer (UvrA 2 B), which helps to locate the site of DNA damage. After dissociation from the UvrA 2 B-DNA complex, UvrA is believed to load more UvrB proteins onto the site of DNA damage (5).
High concentrations of UvrA have been found to be inhibitory to the incision reaction in vitro (6 -8) and can confer UV sensitivity in vivo (9). This phenomenon is due either to competition between UvrA 2 and the UvrA 2 B complex for binding to the damaged DNA, a lack of UvrA 2 dissociation from the UvrA 2 B-DNA pre-incision complex, or to the formation of nonproductive dead end complexes (which could be caused by partially denatured or improperly folded proteins). Within the cell, the relative concentrations and the interactions between UvrA and UvrB must be well controlled to allow UvrA to participate in multiple rounds of incision. One group of proteins which often help to facilitate proper protein-protein interactions are the heat shock or chaperone proteins.
Here we report interactions of UvrA with heat shock proteins of the DnaK, DnaJ, and GrpE family. Our results indicate the following: 1) UvrA is heat-labile; 2) UvrA is stabilized by DnaK or the combination of DnaK, DnaJ, and GrpE under normal in vitro conditions; and 3) UvrA in the presence of DnaK, DnaJ, and GrpE undergoes multiple cycles of UvrB loading, leading to increased incision efficiency; 4) denatured UvrA can be refolded and reactivated by DnaK, DnaJ, and GrpE; 5) a DnaK mutant strain shows decreased survival after UV and a lower extent of photoproduct removal.

EXPERIMENTAL PROCEDURES
UvrA, UvrB, and UvrC Purification-UvrA was purified from E. coli strain MH1 UvrA containing the overproducing plasmid, pSST10 (graciously supplied by L. Grossman, Johns Hopkins University), in which the uvrA gene is under the control of the heat-inducible PL promoter. UvrB and UvrC were overproduced from E. coli strain CH296 containing plasmids pUNC211 and pDR3274, respectively (graciously supplied by A. Sancar, University of North Carolina). All three proteins were purified to homogeneity as described previously (32 Heat Shock Proteins-All heat shock proteins used in this study were purchased from StressGen Biotechnologies. BPDE-damaged DNA-The 50-base pair oligonucleotides containing a single BPDE adduct, (ϩ)-cis-anti-BPDE, were prepared as described previously (2). The phosphorylated BPDE-11-mers (30 pmol) (generously supplied by Dr. Nick Geacintov, Department of Chemistry, New York University) were ligated with equal moles of 20-mer (5Ј-terminally labeled with 32 P) and phosphorylated 19-mer in the presence of the complementary strand 50-mer and T4 DNA ligase in a 30-l solution containing 50 mM Tris-HCl, pH 7.8, 10 mM MgCl 2 , 10 mM DTT, 1 mM ATP, and 50 g/ml BSA. The ligation was carried out at 16°C for 12 h. After ligation, products were purified and then reannealed with the non-damaged 50-mer complementary strand and then purified on an 8% polyacrylamide native gel. The double-stranded character and homogeneity of the 50-base pair substrates were examined by a restriction assay (2) and analyzed on a 12% polyacrylamide sequencing gel under denaturing conditions. Thermal Stabilization Assay-UvrA (0.25 or 1.0 nM) was incubated in the presence or absence of DnaK or the combination of DnaK, DnaJ, and GrpE at 37°C for discrete periods of time or at various concentrations, as indicated in the figure legends, in 10 l of UvrABC buffer (50 mM Tris-HCl, pH 7.5, 50 mM KCl, 10 mM MgCl 2 , 5 mM DTT, and 1 mM ATP). After heat inactivation, the 5Ј-terminally labeled BPDE-DNA substrate (2 nM) with UvrB (200 nM), UvrC (50 nM), and more ATP (1 mM) were mixed into the reaction to initialize the UvrABC-mediated incision of the adduct at 37°C for 60 min. The reaction was stopped by addition of EDTA to 25 mM, followed by heating at 90°C for 3 min. The incision products were then analyzed in a 10% (w/v) polyacrylamide sequencing gel under denaturing conditions with TBE buffer (50 mM Tris, 50 mM boric acid, and 1 mM EDTA, pH 8).
Stabilization during Incision-UvrA (0.25 or 1.0 nM), UvrB (200 nM), and UvrC (50 nM) were first mixed in the absence or presence of heat shock proteins (DnaK, DnaJ, and GrpE 0.9/0.01/0.2 M) in 60 l of UvrABC buffer (50 mM Tris-HCl, pH 7.5, 50 mM KCl, 10 mM MgCl 2 , 5 mM DTT, and 1 mM ATP). Then the 5Ј-terminally labeled BPDE-DNA substrate was added immediately to initiate the incision at 37°C. After the indicated periods of incision, a 10-l aliquot was withdrawn and heated at 90°C for 3 min to stop the reaction. The incised products were analyzed in a 10% (w/v) polyacrylamide sequencing gel under denaturing conditions. UV Survival-The E. coli strains used in this study are as follows: MC4100, wild-type cells; BB1553, a ⌬dnaK52 sidB1 mutant; GW8306, BB1553 cells containing pJM41 plasmid encoding wild-type DnaK (pBR322-P lac dnaK ϩ ); and GW8304, BB1553 cells harboring the pBR322 plasmid (all strains were generously provided by G. C. Walker, Massachusetts Institute of Technology) (33,51). The sidB1 mutation allows stable propagation of cells in which dnak has been deleted. SidB1 is a mutation in rpoH, the gene for 32. The BB1553 cells have a similar growth rate to that of wild-type cells, MC4100 (33). MC4100 and BB1553 cells were grown in LB. GW8304 and GW8306 cells were grown in LB containing 100 g/ml ampicillin, 60 g/ml kanamycin, and 0.5 mM IPTG. Cells were grown to saturation overnight at 30°C and then diluted by about 200-fold into the same medium, followed by continuous growth at the same temperature to an optical density of 0.4 at 600 nm. Cells were pelleted, washed, and resuspended to the original density in 0.9% NaCl solution. Irradiation of the cell suspensions was performed with a germicidal lamp (254 nm) at 100 microwatts for various periods to damage the cells. The irradiated cells were then appropriately diluted with 0.9% NaCl solution on ice, plated on LB agar (for MC4100 and BB1553 cells) or LB agar containing 100 g/ml ampicillin, 60 g/ml kanamycin, and 0.5 mM IPTG (for GW8304 and GW8306 cells), and incubated at 30°C in the dark. Plates containing 100 -800 colonies were counted to calculate the survival.
Cell Growth and DNA Preparation for DNA Repair Assay-Cells were grown at 30°C in Davis medium supplemented with 0.4% glucose, 0.2% casamino acids, and 1 g/ml vitamin B1. Cultures were grown to mid log phase (approximately 3 ϫ 10 8 cells/ml) as measured by A 600 . Cells were collected by filtration on 0.45-mm Millipore filters, washed with prewarmed Davis medium, and resuspended in Davis medium. Cells were UV-irradiated in a glass Petri dish on a rotating platform under a germicidal lamp with a dose of 40 J/m 2 ultraviolet-light (254 nm) and placed in a flask containing growth supplements. Samples of the culture were removed at various times and mixed with an equal volume of ice-cold NET (100 mM NaCl, 10 mM Tris, pH 8.0, 10 mM EDTA) buffer. Cells were pelleted by centrifugation at 4°C, resuspended in TE, pH 8.0, and were lysed by addition of 1 mg/ml lysozyme and 100 mg/ml RNase A and incubation for 15 min at 37°C. Proteinase K (100 mg/ml) and Sarkosyl (0.5%) were then added, and the mixture was incubated at 50°C for 1 h. The DNA was extracted with phenol/ chloroform and precipitated with 2.5 M ammonium acetate and 2 volumes of 95% ethanol. Purified DNA was resuspended in TE, pH 8.0. The DNA was lightly sonicated using a Branson sonifier, and the concentration was determined by fluorometry using Hoechst 33258.
Immunoassay for Global Nucleotide Excision Repair-The repair of cyclobutane dimers (CPD) and 6 -4 photoproducts (6 -4 PP) in each sonicated DNA sample was determined using an immunoassay. Following denaturation by boiling, 25 ng (CPD) or 250 ng (6 -4PP) of each DNA sample was loaded in triplicate onto a Hybond N ϩ membrane using a slot blot apparatus. The membrane was incubated for 2 h in the presence of mouse antibodies against either CPDs (TDM-2) or 6 -4 PPs (64M-2) diluted 1:2000 in phosphate-buffered saline (antibodies were a generous gift of Dr. Toshio Mori). Horseradish peroxidase-conjugated secondary antibodies were used at a dilution of 1:5000 and detected using the enhanced chemiluminescence detection system (Amersham Pharmacia Biotech) and subsequent analysis on phosphor screens (Bio-Rad).
Reactivation Assay-UvrA protein (126 nM) was inactivated by incubation in 21 l of 6 M guanidine HCl at 30°C for 30 min. The denatured protein (0.5 l) was then diluted 125-fold into the refolding buffer (50 mM Tris-HCl, pH 7.5, 50 mM KCl, 10 mM MgCl 2 , 5 mM DTT, 1 mM ATP), with or without the heat shock proteins DnaK, DnaJ, and GrpE (1.0/ 0.03/0.64 M) and subsequently incubated at 37°C for the periods indicated. After the incubation, aliquots of the refolding mixture were added with UvrB (200 nM), UvrC (50 nM), more ATP (1 mM), and the 5Ј-terminally labeled BPDE-DNA substrate (2 nM) to start the incision reactions. The incision was carried out at 37°C for 60 min. The incised products were then analyzed in the polyacrylamide sequencing gel under denaturing conditions.
Immunoprecipitation-UvrA (0.25-5 nM) was first incubated with 145 nM DnaK (StressGen Biotechnologies) at 37°C for 30 min in 20 l of IM buffer (50 mM Tris-HCl, pH7.5, 50 mM KCl, 10 mM MgCl 2 , and 5 mM DTT). Murine monoclonal anti-DnaK (StressGen Biotechnologies) antibody (1 g) was added to the mixture, which was then incubated with gentle shaking at 4°C for 1 h. After the incubation, the sample was mixed with 12 g of swollen protein A-Sepharose 4B fast flow beads (Sigma) and continuously incubated with gentle shaking at 4°C for another hour. The sample was centrifuged at 12,000 ϫ g at 4°C for 5 min, and the supernatants were removed. The Immunobeads were washed with IM buffer at 4°C three times. The beads were mixed with 20 l of SDS-polyacrylamide gel electrophoresis sample buffer and heated at 90°C for 5 min. The mixture was loaded onto an 8% (w/v) SDS-polyacrylamide gel electrophoresed for separation. The gel was blotted to a Hybond nitrocellulose membrane (Amersham Pharmacia Biotech) using a Hoefer electrotransfer unit in a buffer containing 10 mM CAPS, pH 11, and 10% MeOH. The membrane was then treated with rabbit anti-UvrA polyclonal following the ECL Western blotting procedures (Amersham Pharmacia Biotech) to identify the UvrA protein bands.

RESULTS
Stabilization of UvrA by Hsp-Models of the action mechanism of the UvrABC system indicate that UvrA is capable of loading multiple UvrB molecules at damaged sites (1). Whereas UvrA has been shown to turn over on randomly damaged plasmids (5), we have been unable to demonstrate its turnover on defined DNA substrates containing a site-specific BPDE-N 2guanine adduct (2). Since UvrB loading takes several minutes, we reasoned that UvrA at dilute concentrations might be heatlabile, and molecular chaperone proteins might help stabilize UvrA. Fig. 1 shows the thermal stability of UvrA at 37°C in the absence and presence of all three heat shock proteins (DnaK, DnaJ, and GrpE). The thermal stability of UvrA was assessed by the incision activity of UvrABC system on a DNA substrate containing a single site-specific BPDE adduct. To characterize UvrA specifically under conditions where other components were in excess and would not significantly affect the results, a limited concentration of UvrA was used in our experiments, relative to the DNA substrate and the UvrB and UvrC subunits. As shown in Fig. 1, in the absence of the heat shock proteins, UvrA quickly lost its activity at 37°C, while the presence of the heat shock proteins greatly protected UvrA from inactivation. Since the lost enzymatic activity was not restored after subsequent incubation with DnaK, DnaJ, and GrpE (data not shown), it appears that the heat leads to the irreversible aggregation of UvrA.
The effects of heat shock proteins on the UvrA stabilization were further examined as shown in Fig. 2. UvrA was incubated at 37°C for 30 min in the presence of increasing concentrations of heat shock proteins, and UvrA stability was assessed by incision activity as before. In the absence of heat shock proteins, UvrA failed to withstand the physiological temperature of 37°C for 30 min and lost its ability to support incision. Acetylated BSA provided a little, if any, stabilization of UvrA. In contrast, UvrA activity, as measured UvrABC incision efficiency, dramatically increased with increasing concentrations of the heat shock proteins (Fig. 2, a and b). These results suggest a specific interaction of the heat shock proteins with UvrA. Although DnaJ and GrpE work together to help DnaKmediated refolding of polypeptides (12,34), we observed no significant differences in the protection of UvrA from denaturing when DnaJ and GrpE were added with DnaK. It appears that DnaK uses its binding/releasing/rebinding power, driven by ATP binding (35)(36)(37)(38)(39)(40)(41)(42), to stabilize UvrA.
The effects of DnaK on the thermal stability of UvrA were further confirmed by a gel mobility shift assay for the binding of UvrA to the BPDE-DNA substrate under the same experimental conditions. As shown in Fig. 2c, heat shock proteins helped to stabilize binding of UvrA to the BPDE substrate.
The effects of adding heat shock proteins during the incision are shown in Fig. 3. Although the excess of UvrB and UvrC preserves some UvrA activity in the absence of DnaK, DnaJ, and GrpE, much higher incision efficiency has been achieved in the presence of the heat shock proteins, especially for the lower UvrA concentration. There is even a significant difference in the initial rates of incision. Furthermore, as shown in Fig. 3, the majority of BPDE-DNA substrate molecules (20 pmol) has been incised even by 1.25 pmol of UvrA (in dimer) in 1 h, at a stoichiometric ratio of DNA/UvrA 2 of 16:1. In this case, a single UvrA dimer molecule has loaded at least 10 UvrB molecules. From these experiments, it is clear that UvrA turned over efficiently during the incision. These results are consistent with a previous report (5) of those with higher concentrations of UvrA and a damage-containing plasmid but with the absence of heat shock proteins.
Reactivation of UvrA by Hsps-One of the most important characteristics of heat shock proteins is their ability to refold denatured protein substrates into their native structures and to allow restoration of their enzymatic activities. To explore the relationship of UvrA with E. coli heat shock proteins of the Hsp70 family, we conducted UvrA reactivation experiments. To avoid aggregation among the unfolded polypeptides, 6 M guanidine hydrochloride was used instead of thermal denaturation. After the inactivation of UvrA, the denatured protein  in panel b). An aliquot from the incubation mixture was then incubated with ATP (1 mM), UvrB (200 nM), UvrC (50 nM), and 5Ј-labeled BPDE-DNA substrate (2 nM) at 37°C for 60 min to assay incision of the substrate by UvrABC nuclease (panels a and b). UvrA (0.25 nM) was also incubated with 5Ј-labeled BPDE-DNA substrate (2 nM) in the UvrABC buffer in the presence of increasing concentrations of DnaK at 37°C for 60 min to assay binding of UvrA to the substrate (filled triangles in panel c). The UvrA-DNA complex was analyzed on a 3% native polyacrylamide gel. Data in panels b and c are the means of at least three independent experiments. was diluted 1:125 into refolding buffer containing DnaK, DnaJ, and GrpE to allow restoration of the activity. To determine the background activity, the denatured UvrA was also diluted into the same buffer without heat shock proteins. As shown in Fig.  4, the heat shock proteins were able to restore about 50% of the native incision activity of UvrA within 10 min. As a control, BSA or DnaK, DnaJ, and GrpE in the absence of ATP showed no or little power to reactivate UvrA (Fig. 4b). Individually, the member of the Hsp70 family showed some ability to restore UvrA activity, but the combination of all three provided the highest level of reactivation of unfolded UvrA.
Immunoprecipitation of UvrA-DnaK Complex-The reactivation of UvrA by DnaK, DnaJ, and GrpE strongly suggests a direct interaction of the heat shock proteins with UvrA. We therefore investigated the potential binding between denatured UvrA and DnaK in an immunoprecipitation assay that has been widely used for study of protein-protein interactions (43). In this assay (see "Experimental Procedures"), native UvrA was heated at 37°C in the presence of DnaK in a buffer containing no ATP. Then DnaK antibody was added and allowed to bind to DnaK as well as to any potential UvrA-DnaK complex. Treatment with Immunobeads, which recognize the antibody, followed by centrifugation, allows the separation of the bead-bound proteins from the proteins that are free in solution. The immunoprecipitated samples were analyzed by SDS-polyacrylamide gel electrophoresis and probed with UvrA antibody (␣UvrA) on Western blots. As shown in Fig. 5, the identified bands indicate that DnaK acting as molecular chaperone does bind to UvrA in this assay. This binding may act to prevent UvrA from denaturing and aggregating (Fig. 5).
UV Sensitivity of a DnaK-deficient Strain-The results from the in vitro studies suggested a potential role for heat shock proteins in nucleotide excision repair in vivo. To help confirm this idea, we conducted in vivo experiments to study the UV sensitivity of DnaK-deficient cells as compared with wild-type cells. Fig. 6 demonstrates the survival of wild-type cells MC4100 and BB1553 mutant (⌬dnaK52 sidB1) (33) versus UV dose. In comparison to wild-type cells, BB1553 cells displayed significantly more sensitivity to UV irradiation at various doses and were about 80-fold more sensitive at 100 J/m 2 . As a control, BB1553 mutant cells carrying the plasmid, pJM41, which encodes DnaK under an IPTG-inducible lac promoter showed similar UV resistance as wild-type cells (Fig. 6). These results demonstrate that the dnaK deletion, but the sidB1 mutation, conferred UV sensitivity. Since both wild-type and mutant cells grew at the similar rate at 30°C (33), one of the possible explanations for these results is that the DnaK deletion mutant cells may have a limited capacity for nucleotide excision repair, due to the reduction of UvrA activity. This hypothesis was tested below.
Repair Kinetics of UV Photoproducts in DnaK-deficient Cells-The increased UV sensitivity exhibited by BB1553 cells could result from a number of different factors including lack of DNA replication resumption or slower kinetics of DNA repair. In order to investigate the latter, BB1553 and MC4100 were irradiated with 40 J/m 2 of ultraviolet light (254 nm), and the frequency of pyrimidine dimers and 6 -4 photoproducts was determined by DNA slot blot analysis using specific monoclonal antibodies to each photoproduct. The results collected in Fig. 7 indicate that BB1553 cells, DnaK-deficient, had slower rates of repair of 6 -4 photoproducts and lower extents of repair for both types of DNA photoproducts. These results are entirely consistent with the in vitro results indicating DnaK is neces- The K, J, and E indicate DnaK, DnaJ, and GrpE, respectively. After the incubation, aliquots of the mixture were added to UvrB (200 nM), UvrC (50 nM), more ATP (1 mM), and the 5Ј-terminally labeled BPDE-DNA substrate (2 nM) to start the incision reaction (37°C, for 60 min). The activity was determined relative to a control sample treated in the same way but without heat shock proteins. The control was the sample in which the substrate was directly incised by UvrABC nuclease without any pre-incision treatment. Data are representative of three independent experiments. sary for the turnover of UvrA and that without the action of DnaK in vivo there is a limited number of active UvrA molecules that would lead to lower extents of repair.

DISCUSSION
UvrA is a critical protein in E. coli nucleotide excision repair because of its direct role in damage recognition and in guiding UvrB to the damaged site. In this report we sought to understand why UvrA could not act catalytically during incision reactions on defined substrates containing a site-specific lesion. In the course of these studies we found the following. 1) Heat shock proteins allowed UvrA to participate in multiple rounds of UvrB loading and to act catalytically in the incision reaction.
2) DnaK or the combination of DnaK, DnaJ, and GrpE helped to protect UvrA from thermal denaturation and irreversible inactivation. 3) DnaK interacted directly with denatured UvrA forming a protein-protein complex. 4) Denatured but not aggregated UvrA could be reactivated by DnaK, DnaJ, and GrpE to up to ϳ50% of the native protein activity. 5) UV survival and repair kinetics in a DnaK-deficient cells indicate an important in vivo role for DnaK in nucleotide excision repair, presumably by its interaction with UvrA.
The dnaK, dnaJ, and grpE genes of E. coli were first identified by mapping mutations that were unable to support bacteriophage replication (44). Subsequent analysis indicated that these genes encode heat shock proteins that are ubiquitous throughout nature (11,12). These proteins function as molecular chaperones to maintain proteins in the properly folded state. For example, Walker and co-workers (45) have shown that GroE heat shock protein is absolutely required for SOS-induced mutagenesis, presumably through its stabilization of UmuC.
Whereas earlier studies indicated that UV light and naladixic acid were able to induce DnaK and GroEL in an SOSindependent manner (46), no enhanced survival functions have been attributed to the induction of these proteins. Experiments presented in Fig. 6 clearly show that a DnaK-deletion strain, One g of murine monoclonal anti-DnaK (as indicated with ϩ for ␣DnaK) or UvrA antibody (as indicated with ϩ for ␣UvrA) was then added to the mixture followed by incubation at 4°C for 1 h. The sample was then mixed with 12 g of swollen protein A-Sepharose beads and continually incubated at 4°C for another hour. The Immunobeads were precipitated and washed at least three times with the same buffer at 4°C. The bead-bound proteins were resolved on an 8% (w/v) SDS-polyacrylamide gel, which was then blotted to a nitrocellulose membrane. The UvrA in the membrane was identified with UvrA antibody in the ECL Western blotting assay. BB1553, is significantly more UV-sensitive than an isogenic wild-type strain. Furthermore repair kinetics of 6 -4 PP and CPD in these two strains (Fig. 7) indicate that loss of DnaK activity leads to a decrease in the overall extent of repair for both CPD and 6 -4 PP and a slower rate of repair for 6 -4 PP.
These in vivo findings are entirely consistent with our in vitro results that indicate that the addition of DnaK, DnaJ, and GrpE to UvrA leads to increased rate and extent of incision of a defined substrate containing a single BPDE-N 2 guanine adduct (Fig. 3). UvrA is extremely heat-labile when in dilute concentrations, with a t1 ⁄2 of less than 5 min at 37°C (Figs. 1  and 2). However, the addition of DnaK either singularly or with DnaJ and GrpE greatly stabilized UvrA to heat denaturation and led to multiple loadings of UvrB.
The data presented in this work demonstrate that DnaK, DnaJ, and GrpE directly interact as molecular chaperones with UvrA to keep UvrA in its native state. It is believed that exposed hydrophobic side chains of denatured proteins are the primary determinants for DnaK binding (47,48). The analysis of crystal structure of DnaK complexed with a peptide at its substrate-binding site has indicated that the binding of DnaK was centered to its substrate on hydrophobic residues that become completely buried in a deep pocket on DnaK (42). This suggests that DnaK binds to unfolded proteins that expose hydrophobic residues on their surface. In our experiments, UvrA was denatured by heating at 37°C, and UvrA may have aggregated in the absence of chaperones, resulting in irreversible loss of UvrA enzymatic activity. Incubation of the UvrA aggregate with DnaK, DnaJ, and GrpE failed to recover any UvrA activity (data not shown). In the presence of DnaK, non-native UvrA was protected from aggregation or misfolding most likely by the binding of DnaK to the UvrA.
UvrA after denaturation by guanidine hydrochloride was reactivated up to about 50% of the original activity by DnaK or the combination of DnaK, DnaJ, and GrpE in the presence of ATP. The difference between the heat and guanidine hydrochloride treatments is that the former results in aggregation of UvrA and the latter leads to the formation of unfolded UvrA without aggregation. Our results indicated that although there were no significant differences between DnaK or the combination of DnaK, DnaJ, and GrpE in heat stabilization of UvrA, the combination of all three proteins was evidently more efficient in reactivating unfolded UvrA (Fig. 4). A similar observation was reported for the interactions of DnaK or DnaK, DnaJ, and GrpE with RNA polymerase by Skowyra et al. (18) and Ziemienowicz et al. (49). The difference in reactivation may be attributed to a higher efficiency for DnaK working together with coenzymes of DnaJ and GrpE. DnaJ has been reported to stimulate the ATP hydrolysis of DnaK to form ADP-DnaK intermediate that binds peptide substrate tightly (39,50). Then GrpE dissociates the ADP from DnaK, resulting in the release of substrate (36,38).
In conclusion, data presented here clearly show that DnaK stabilizes UvrA from heat inactivation and allows UvrA to act catalytically during in vitro incision reactions. Measurements of in vivo repair kinetics and UV survival of a DnaK-deficient E. coli strain clearly indicate a role for DnaK in nucleotide excision repair. Future studies will be directed toward understanding whether DnaK and other E. coli heat shock proteins can help promote the release of UvrA from the UvrAB-DNA complex and whether these chaperone proteins act on other components of the nucleotide excision repair system. It will be of interest to determine whether heat shock proteins help aid in the proper assembly/disassembly, and thus efficiency, of multi-subunit protein machines involved in eucaryotic nucleotide excision repair.