DNA binding residues in the RQC domain of Werner protein are critical for its catalytic activities

WRN expression and purification

For over a decade, purification of WRN helicase has been performed in our laboratory using a cumbersome and time-consuming procedure [9]. Recently, a new approach was devised, resulting in a reduction of the number of chromatographic steps, increased purity of the final product, and in overall simplification of the entire process. The new protocol uses Invitrogen's Bac-to-Bac® baculovirus expression system with a modified pFastBac1 vector. The modified vector, pFastBac1-InteinCBDAla, contains intein/chitin binding domain (InteinCBD) from the pTXB1 plasmid (IMPACT™ Kit, New England Biolabs) inserted downstream from the multiple cloning site. Additionally, SmaI restriction site and Alanine codon were incorporated in front of the InteinCBD to allow for cloning of blunt-ended inserts, and to improve efficiency of the intein self-cleavage, respectively (described in Experimental Procedures).

The WRN gene was PCR-amplified using primers incorporating hexahistidine (6xHis) and DYKDDDDK (FLAG tag) at the 5' and 3' termini, respectively. The PCR product was inserted between NotI and SmaI sites of the pFastBac1-InteinCBDAla plasmid, forming the 6xHis-WRN-FLAG/pFastBac1-InteinCBDAla construct (Figure 2A), which drives expression of WRN protein with N-terminal 6xHis and C-terminal InteinCBD tags. Next, the construct was used with the Bac-to-Bac® Baculovirus Expression System (Invitrogen) to generate recombinant baculoviruses, which were then used to express the dual-tagged WRN protein in High Five™ insect cells. The purification scheme consisted of two chromatographic steps. First, the cell lysate was applied to nickel-charged HisTrap column, followed by a chitin column. Chitin bead-bound 6xHis-WRN-FLAG-InteinCBD fusion protein was incubated overnight in the presence of thiols to induce the self-cleavage activity of the intein. Post-cleavage, the InteinCBD domain remained bound to the chitin resin, while released 6xHis-WRN-FLAG protein was eluted, concentrated and stored at -80°C, as described in Experimental Procedures (Figure 2B, 2C). The presence of two affinity tags at opposite termini simplified the purification scheme, allowed for elimination of most proteolytic fragments, and increased recovery of the full length WRN protein. The FLAG tag was not employed during purification, and was placed at the C-terminus of WRN to be utilized in future experiments.

Effect of the mutations on DNA unwinding activity

The recently solved crystal structure of the WRN RQC-DNA duplex complex demonstrated that the amino acid residues in the α2-α3 loop and the β-hairpin motifs of the RQC domain are responsible for the DNA binding [8]. Thus, we hypothesized that the RQC domain may be important for WRN's DNA unwinding activity. For this purpose, two mutant proteins with single amino acid substitutions, R993A and F1037A, as well as a double mutant R993A/F1037A, have been isolated and biochemically characterized. The sites were chosen because mutations in Arg-993 and Phe-1037 displayed most significant reductions in DNA binding affinity [8]. First, the WRN wild type and mutant proteins were purified and their purity estimated to be > 90% by SDS-PAGE (Figure 2D). We performed heat denaturation experiment to evaluate folding of purified proteins. SYPRO Orange fluorescent dye (SO), which increases its fluorescence emission intensity upon binding with hydrophobic residues of the unfolded protein, can be used for monitoring of protein unfolding, and thereby indirectly protein structure [10, 11]. Here, we used the SO dye to monitor heat denaturation curves, and to determine the melting temperatures of the wild type and mutant WRN proteins. As shown in Figure 2E, WRN wild type, R993A, F1037A and R993A/F1037A gave similar heat denaturation profiles, suggesting that all proteins were folded similarly. The melting temperatures (Tmobs) of these proteins were around 45 °C, as summarized in Table 1, suggesting that stabilities of the mutant proteins were comparable to that of the wild type protein.

Since we had confirmed that the structural properties of the wild type and mutant proteins were similar, we performed helicase assays on forked duplex substrates using these recombinant proteins. All mutants, WRN R993A, WRN F1037A, and WRN R993A/F1037A, showed significantly decreased DNA unwinding activity on forked duplex (Figure 3A, 3B). It is noted that DNA unwinding by these mutants was detected when the protein concentration for the assay was increased excessively (see Supplementary Figure S1). We then examined DNA binding abilities of these mutant proteins by performing EMSA. R993A, F1037A and R993A/F1037A did not show any DNA binding affinity, similarly to the results demonstrated in a previous study where the WRN RQC domain fragment was used with the corresponding mutations (Figure 3C) [8]. We also determined whether the mutant proteins possessed ATPase activity, and found that they did not exhibit ATP hydrolysis in the presence of either ssDNA or dsDNA (Figure 3D).

WRN helicase is able to resolve a variety of DNA structures, including forks, flaps, D-loops, bubbles, HJ and G4 DNAs, all of which are thought to represent intermediates in DNA replication and recombination processes [3]. Therefore, we further examined the DNA unwinding activity of the WRN variants on HJ, G4, D-loop and bubble DNA substrates. As shown in Figure 4, WRN R993A, F1037A and R993A/F1037A could not resolve any of these DNA structures. These results strongly support the hypothesis that the β-hairpin structure and the unique DNA binding mode of RQC play integral roles in the specialized activities of WRN, and this may also be the case for other human RECQ helicases with conserved RQC structure.

Effect of the mutations on WRN exonuclease and DNA annealing activities

Since WRN helicase is known to exhibit other enzymatic activities, such as 3'→5' exonuclease and DNA strand annealing, we tested these activities using the WRN mutant proteins. First, we examined exonuclease activity using a 5' overhang DNA substrate. Surprisingly, neither R993A nor F1037A mutant proteins exhibited exonuclease activity (Figure 5A), suggesting that the RQC domain is not only responsible for the helicase activity, but is also involved in regulating the exonuclease activity. WRN exonuclease is known to be stimulated by Ku70/80 heterodimeric protein [12-14]. Therefore, we expected that the interaction between Ku protein and WRN mutants would stabilize the WRN mutant-DNA substrate complex structure, thereby improving their exonuclease activity. As shown in Figure 5B, however, Ku heterodimer was not able to stimulate exonuclease activity of any mutant protein (Figure 5B: lanes 7-15), while, as reported previously, it stimulated the exonuclease activity of wild type WRN (Figure 5B: lanes 4-6). We also examined the DNA strand annealing activity of the WRN mutants. As shown in Figure 5C, R993A, F1037A and R993A/F1037A exhibited significantly lower activity than the wild type (< 15% of wild type), suggesting the involvement of the RQC domain in strand annealing activity as well. We have previously demonstrated using truncated proteins that the region between the RQC and HRDC of WRN is required for ssDNA annealing [15]. However, our results in this study suggest that the RQC domain itself may also contribute to the strand annealing activity of the full length protein.

Assessment of mutant protein folding

Since our biochemical experiments showed that mutations in the RQC domain cause loss of all enzymatic activities of WRN, we then asked whether the mutant proteins were folded properly. An earlier gel filtration study suggested that WRN protein exists as a trimer form in solution [16]. However, electron microscopy studies suggested that WRN is likely to be a dimer in solution while it forms higher oligomers such as a tetramers in the presence of DNA substrate [17]. These findings led us to hypothesize that the oligomeric status of WRN protein, which changes in the presence of the DNA substrate, may be important for its enzymatic functions. Thus, if the mutant proteins were folded properly, they might form higher oligomeric structures with the wild type protein when mixed in the presence of DNA substrates. The hetero-oligomer could thereby facilitate DNA unwinding, but at a lower level of activity than the oligomer consisting of wild type only. Based on this notion, we performed helicase assays with mixed wild type and mutant proteins. The representative gel image and summarized data are shown in Figure 6. When 2 nM wild type WRN alone (lane 2) was compared with 2 nM wild type with 4 nM R993A mutant (lane 9), the helicase activity increased an additional 10 to 15 %. This is more than additive effect, suggesting that the R993A mutant protein was able to stimulate the helicase unwinding activity of WRN wild type. Similar results were observed for F1037A and R993A/F1037A (Figure 6A, 6B: lane 12, lane 15). The above results, taken together with the heat denaturation experiment (Figure 2E), indicate that mutant proteins, R993A, F1037A and R993A/F1037A are most likely folded properly.

Mutations disrupt the functional interaction with NEIL1 Several studies have revealed that the RQC domain of WRN is important not only for DNA binding but also for protein-protein interactions, including FEN1 [18], NEIL1 [19], PARP1 [20], p97/VCP [21] and TRF2 [22]. We have previously reported that WRN containing a K577M mutation, which inactivates the helicase activity, retained the ability to stimulate NEIL1's incision activity [19]. Therefore we chose to investigate how the WRN mutations introduced in this study affected the functional interaction with NEIL1. We performed the NEIL1 incision assays using a 5-OHU-containing bubble substrate. The results showed that the wild type WRN stimulated NEIL1's incision activity in a concentration dependent manner (Figure 7A: lanes 3 to 5, Figure 7B), whereas the mutant proteins could not stimulate NEIL1's activity (Figure 7A: lanes 6 to 14, Figure 7B).

Plasmid construction

pFastBac1 plasmid (Invitrogen) was PCR-linearized using primers: pFastBac1-F, 5'-GTCGAGAAGTACTAGAGGATC-3', and pFastBac1-R2, 5'-GGAATTAAGCTTGGTAC CGCATG-3'. The PCR product's 5' blunt end was phosphorylated using T4 Polynucleotide Kinase (PNK, New England Biolabs), and the 3' end digested with HindIII (New England Biolabs). Mycobacterium xenopi GyrA intein/chitin binding domain fusion was PCR-amplified from pTXB1 plasmid (New England Biolabs) with primers: InteinCBD-SmaI-Ala-F, 5'-GGAATTAAGCTTCCCGGGGCCTGCATCACGGGAGATGCAC-3', and InteinCBD-Stop-R, 5'-TCATTG AAGCTGCCACAAGGCA-3', which incorporated SmaI restriction site (underlined) and Alanine codon (italicized). The PCR product was digested with HindIII, phosphorylated using PNK and ligated into pFastBac1, resulting in the formation of pFastBac1-InteinCBDAla plasmid.

WRN gene was PCR amplified from YEp195SpGAL-WRN plasmid [39] using primers: hWRN-NotI-6xHis-F, 5'-GGAATTGCGGCCGCATGTCTCATCACCATC ACCATCACAGTGAAAAAAAATTGGAAACAAC-3' and WRN-FLAG-blns-R, 5'-CTTATCGTCGTCATC CTTGTAATCACTAAAAAGACCTCCCCTTTTC-3', adding sequences coding hexahistidine (6xHis tag), and DYKDDDDK (FLAG tag) to the 5' and 3' termini, respectively. The PCR product was digested with NotI, phosphorylated with PNK, and ligated into NotI/SmaI digested pFastBac1-InteinCBDAla vector, producing 6xHis-WRN-FLAG/pFastBac1-InteinCBDAla construct. The sequence of final construct was confirmed by direct sequencing. All DNA oligonucleotides were purchased form Integrated DNA Technologies (Coralville, IA).

Arginine 993 and Phenylalanine 1037 were substituted into Alanine by site-directed mutagenesis. The mutagenic primers were designed such that the codons for Arg (CGT) and Phe (TTT) are changed to those for Ala (GCG, GCT), respectively. We also constructed the double mutants, R993A/F0137A. The nucleotide sequences were confirmed by sequencing.

Generation of recombinant baculoviruses

Recombinant baculoviruses were generated using Bac-to-Bac® Baculovirus Expression System (Invitrogen). Briefly, 6xHis-WRN-FLAG/pFastBac1-InteinCBDAla construct was transformed into DH10Bac™ E. coli strain (Invitrogen), and recombinant bacmids purified from large white colonies. The insertion of the 6xHis-WRN-FLAG/InteinCBDAla gene was verified by PCR. Next, bacmids were transfected into Sf9 insect cells using Cellfectin II® reagent (Invitrogen). The recombinant viruses were subsequently amplified in Sf9 cells grown at 27 °C in TNM-FH medium (Mediatech) supplemented with 10% fetal bovine serum (Sigma-Aldrich).

Protein expression and purification

High Five™ insect cells (Invitrogen) were maintained in suspension with aeration in ESF 921 medium (Expression Systems, Woodland, CA) at 27 °C. The cells were infected with recombinant baculovirus at multiplicity of infection (MOI) of 10, grown for 48 hours and harvested by centrifugation. All purification steps were carried out at 4 °C. Insect cell pellets were sonicated in lysis buffer (50 mM Tris HCl, pH 8.0, 300 mM NaCl, 10% glycerol, 25 mM Imidazole, 1 mM PMSF, 1 tablet/50 ml Complete EDTA-free Protease Inhibitor Cocktail, Roche), spun 30 min. at 37,000g, and filtered through 45 μm membrane. Clarified lysate was loaded onto HisTrap FF 5 ml column equilibrated in Buffer A (50 mM Tris HCl, pH 8.0, 300 mM NaCl, 10% glycerol) supplemented with 25 mM Imidazole, using ÄKTApurifier instrument (GE Healthcare). The column was washed with 10 column volumes (CV) of Buffer A + 25 mM Imidazole, and the protein eluted with 10 CV of Buffer A + 250 mM Imidazole. Peak fractions containing protein of interest were pooled, and loaded 3 times by gravity flow onto 3 mL chitin column (New England Biolabs) equilibrated with Buffer A. The column was washed with additional 20 CV of Buffer A, followed by 3 CV of Buffer A + 100 mM 2-Mercaptoethanol. The flow was stopped and the intein self-cleavage reaction allowed to proceed for 18 hrs. at 4°C. Cleaved protein was eluted with 5 CV of Buffer A + 100 mM 2-Mercaptoethanol, peak fractions were pooled and concentrated using Amicon® Ultra-4 centrifugal filter unit with 50 kDa cutoff (Millipore). Purified concentrated protein was diluted 1:1 with cold 100% glycerol, aliquoted and stored at -80 °C. Protein concentration was determined by Bradford method (Bio-rad), and protein purity was analyzed on SDS-PAGE. Purifications of all three mutant proteins were performed as described above for wild type protein.

Recombinant protein of Ku70/80 heterodimer was purified as described previously [12]. Recombinant NEIL1 protein was obtained previously [40].

Heat denaturation analysis

To determine the protein stability, thermofluor assay was carried out using an iQ cycler instrument (Bio-rad). Protein stability measurements were performed in buffer containing 50 mM HEPES-NaOH, pH 8.0, 60 mM NaCl, 1 mM DTT and 15 % (v/v) glycerol. The protein was used at 0.1 μg/μl concentration (2 μg in 20 μl reaction) and SYPRO Orange (Invitrogen) was used at 1x to 20x concentration range (1:250 to 1:5000 dilution of 5000x stock). Heat denaturing curves were observed within the temperature range of 20 °C to 90 °C, with approximately 2 °C/min rate of temperature change, collecting data every 0.5 °C. Obtained data was normalized by the van't Hoffs analysis [41].

Helicase assay

Helicase unwinding assays were performed as described previously [42]. Reactions (20 μl) contained 0.5 nM substrate, 2.5 mM ATP and the indicated concentrations of WRN wild type and mutants in 40 mM Tris-HCl pH 8.0, 50 mM NaCl, 5 mM MgCl2, 100 μg/ml BSA, 2 mM ATP, 1 mM DTT. Reactions were carried out at 37 °C for 30 min, and terminated by the addition of 10 μl of SDS stop solution (2% SDS, 50 mM EDTA, 30% glycerol, 0.1% bromophenol blue, 0.1% xylene cyanol). Products were separated on 8% non-denaturing polyacrylamide gel. Radiolabeled DNA was visualized using Typhoon phosphorimager, (Typhoon 9400, GE Healthcare) and quantified using ImageQuant software (Molecular Dynamics). The D-loops were prepared and characterized as described previously [43]. G4 DNA substrate was prepared and helicase reaction was performed as described previously [44, 45].

Substrate binding analysis

Electro mobility shift assays (EMSA) were performed to analyze the binding ability as described previously [46]. Binding reactions (20 μl) were conducted in standard binding buffer (40 mM Tris-HCl, pH 8.0, 1 mM EDTA, 20 mM NaCl, 8% glycerol, and 20 μg/ml BSA). Protein and DNA substrate concentrations are indicated in Figure legends. Reactions were incubated for 30 min on ice and directly loaded onto 5% non-denaturing polyacrylamide gels. Products were visualized using Typhoon phosphorImager (GE Healthcare) and quantified using ImageQuant software (Molecular Dynamics).

ATPase assay

ATPase assays were performed as described previously [4]. Buffers for standard ATPase reactions contained 20 mM HEPES-NaOH, pH 8.0, 0.05 mM ATP, 40 μg/ml BSA, 1 mM DTT. ATPase reactions employed WRN and 12.5 μCi of [γ-32P]ATP. Reactions were incubated for 1 h at 30 °C with 150 ng of nucleic acid substrates (M13 mp18 ssDNA for ssDNA and pUC19 for dsDNA), and stopped by the addition of 5 μl of 0.5 M EDTA. ATP hydrolysis was analyzed by polyethyleneimine thin-layer chromatography using 1 M formic acid, 0.8 M LiCl mobile phase. The results were analyzed using Typhoon phosphorimager (GE Healthcare) and ImageQuant software (Molecular Dynamics).

Exonuclease assay

Exonuclease assays were performed as described previously [12, 14]. Reactions (10 μl) were performed in buffer (40 mM Tris-HCl, pH 8.0, 4 mM MgCl2, 5 mM DTT, 2 mM ATP, and 0.1 mg/ml BSA) containing DNA substrate (0.5 nM) and WRN and/or Ku70/80 as indicated. Samples were incubated at 37 °C for 15 min. Reactions were terminated by addition of equal volume of formamide stop dye (80% formamide, 0.5×TBE, 0.1% bromophenol blue, and 0.1% xylene cyanol). Products were heat-denatured for 5 min at 95 °C, loaded on 14% denaturing polyacrylamide gels, visualized using Typhoon phosphorImager (GE Healthcare), and quantified using ImageQuant software (Molecular Dynamics).

Single strand DNA annealing assay

DNA strand annealing assays were performed as described previously [15, 47]. C80 and G80 oligonucleotides were used. Reactions (20 μL) were carried out with 0.5 nM of each oligonucleotide, one of which was 5'-32P-end-labeled, in 20 mM Tris-HCl, pH 7.5, 2 mM MgCl2, 40 μg/mL BSA and 1 mM DTT at 37 °C for 15 min. Concentrations of WRN proteins used in the reactions were indicated in Figure legends. Reactions were stopped by the addition of stop buffer (50 mM EDTA, 1% SDS and 50% glycerol) and immediately loaded onto 8 % polyacrylamide gel. Radiolabeled DNA was detected using Typhoon Imager (GE Healthcare), and percentage of annealed oligonucleotide was quantified using ImageQuant software (Molecular Dynamics).

NEIL1 incision assay

NEIL1 incision assays were performed as described previously [19, 40]. NEIL1 substrates used were 51-mer oligonucleotides (purchased from Midland Co.) with the damaged base (5-OHU) at residue 26. The damage-containing oligonucleotide was labeled with [γ-32P]ATP and T4 polynucleotide kinase (New England Biolabs), and annealed to the complementary oligonucleotide containing a C opposite the lesion. DNA strand cleavage at the site of the lesion due to the intrinsic AP-lyase activity of NEIL1 after base excision was assayed by incubation of 32P-labeled substrate with NEIL1 (1 nM) alone or together with WRN wild type or mutant proteins (1-4 nM) as indicated. Reactions (10 μl) were performed in a buffer containing (40 mM Tris, pH 8.0, 4 mM MgCl2, 5 mM DTT, 0.1 μg/μl BSA) at 37 °C for 15 min. After stopping the reaction with 70% formamide/30 mM NaOH, the products were analyzed by denaturing gel electrophoresis. Radiolabeled products were visualized using Typhoon phosphorImager (GE Healthcare) and quantified using ImageQuant software (Molecular Dynamics).