NEM inhibitor

Characterization of the endonuclease activity of the replication‑associated protein of beak and feather disease virus

Jui‑Kai Chen1 · Chiaolong Hsiao2 · Jian‑Shin Wu1 · Shin‑Yi Lin2 · Chi‑Young Wang1,3

Abstract

Beak and feather disease virus (BFDV) belongs to the family Circoviridae. A rolling-circle replication strategy based on a replication-associated protein (Rep) has been proposed for BFDV. The Rep gene of BFDV was expressed and purified, and it was shown to cleave short oligonucleotides containing the conserved nonanucleotide sequence found in the replication origin of circoviruses. This endonuclease activity was most efficient in the presence of the divalent metal ions Mg2+ and Mn2+. Rep proteins containing mutation in the ATPase/GTPase motifs and the 14FTLNN18, 61KKRLS65, 89YCSK92, and 170GKS172 motifs lacked endonuclease activity. The endonuclease activity was not affected by ATPase inhibitors, with the exception of N-ethylmaleimide (NEM), or by GTPase inhibitors, but it was decreased by treatment with the endonuclease inhibitor L-742001. Both the ATPase and GTPase activities were decreased by site-directed mutagenesis and deletion of the ATPase/GTPase and endonuclease motifs. The Rep protein was able to bind a double-stranded DNA fragment of P36 (dsP36) containing the stem-loop structure of the replication origin of BFDV. All of the Rep mutant proteins showed reduced ability to bind this fragment, suggesting that all the ATPase/GTPase and endonuclease motifs are involved in the binding. Other than NEM, all ATPase, GTPase, and endonuclease inhibitors inhibited the binding of the Rep protein to the dsP36 fragment. This is the first report describing the endonuclease activity of the Rep protein of BFDV.

Introduction

Beak and feather disease virus (BFDV) is a member of the family Circoviridae and has a single-stranded circular DNA (ss cDNA) genome [1]. It is a non-enveloped icosahedral virus with a diameter of 15 nm [2]. It causes beak and feather disease (PBFD) in various species of psittacine birds, which manifests as successive feather dystrophy and malformation of the beaks and claws. Occasionally, immunodepletion of the thymus and bursa of Fabricius can lead to chronic infec- tion followed by a higher probability of a secondary infec- tion, which subsequently results in death [3]. This disease has been documented in different wild and captive popula- tions of psittacine birds across America, Australia, Europe, and Asia, including Taiwan [4, 5]. Thus, understanding the molecular characteristics of BFDV and finding a reliable strategy for controlling the virus have become important issues in Taiwan.
The BFDV genome is 2000 nucleotides long and encodes two major viral proteins, the replication-associated protein (Rep) and the capsid protein (Cap), which are encoded on the sense and antisense strand, respectively [6]. Numerous ss cDNA viruses, including members of the families Circoviridae and Geminiviridae, have been found to employ a rolling-cir- cle replication (RCR) mechanism [7]. The Rep proteins of these viruses, which are multifunctional single-subunit pro- teins, are required for viral DNA replication and several other biochemical functions. These proteins (1) bind and cleave a nonanucleotide sequence located at the apex of a hairpin-loop structure containing the origin of replication through their endonuclease activity, (2) supply the energy required for DNA replication through their ATPase activity, and (3) unwind the double-stranded (ds) intermediates for DNA strand elongation during replication through their helicase activity [8]. Bind- ing and hydrolysis of ATP via the ATPase activity of the Rep protein has been observed with porcine circovirus (PCV) and tomato yellow leaf curl virus (TYLCV) [8]. Similar findings have also been reported for BFDV, and its Rep protein exhibits both ATPase and GTPase activity, which requires through the Walker A motif, the Walker B motif, and a novel 193GYDG196 motif located at its C-terminus [9, 10]. The replication of ss cDNA viruses is initiated by cleavage of the phosphodiester bond between the T and A nucleotides of the conserved nona- nucleotide sequence at the hairpin of the replicative interme- diate. Once the DNA is cleaved, cellular DNA polymerase uses the free 3’-OH as a primer for subsequent DNA syn- thesis, while the Rep protein binds the phosphate at the 5′ terminal end [11]. The endonuclease activity, including both the cleavage and covalent binding steps, is performed by the N-terminus of the Rep protein in PCV and geminiviruses [12, 13]. The amino acid sequences at the N-terminal ends of the Rep proteins of several viruses with endonuclease activity have been compared and shown to contain conserved motifs, includ- ing the FTLN motif of PCV and the FLTY motif of TYLCV, which are involved in the binding of the Rep protein to ssDNA. The YxxK motif of the Rep protein of PCV includes a tyros- ine residue that is necessary for covalent binding to cleaved viral DNA [12, 14]. These motifs use metal ions as cofactors for their endonuclease activity, and in the case of PCV, the cleavage of the ssDNA has been shown to be most efficient in the presence of Mg2+ or Mn2+ [11]. To our knowledge, the endonuclease activity of the Rep protein in BFDV has not yet been studied, and the protein motifs involved in this activity remained to be characterized.
In this study, we investigated the endonuclease activity of the Rep protein of BFDV and identified motifs involved in this activity. The dsDNA region containing the replication ori- gin bound by the Rep protein was identified, and the ATPase/ GTPase and endonuclease motifs were found to be involved in DNA binding.

Materials and methods

Construction of a T7 SUMO expression vector

The Rep gene of BFDV, which was amplified from a BFDV strain isolated in Taiwan (GenBank accession no. KC980909) and then cloned into the plasmid pET32b con- taining a thioredoxin (Trx) tag, was obtained from a previ- ous study, and the plasmid clone served as the backbone for subsequent constructs [9]. The Rep gene of BFDV was amplified from the recombinant plasmid using the primers SUMO-Rep F and SUMO-Rep R (Supplementary Table 1). The PCR conditions were 94 °C for 2 min followed by 25 cycles at 94 °C for 15 s, 55 °C for 15 s, and 72 °C for 3.5 min, and then a final extension step at 72 °C for 10 min. Competent cells (Lucigen Corporation, USA) were trans- formed with 100 μg of the PCR product and 25 ng of pETite vector. The correct clones were selected and verified by DNA sequencing. Plasmids were prepared and used to trans- form Escherichia coli BL21 (DE3) cells.

Site‑directed mutagenesis

Site-directed mutagenesis was performed to construct Rep ATPase/GTPase point mutants, the Rep endonucle- ase deletion mutants, and Rep endonuclease point mutants (Fig. 1). The primers used for mutant construction are listed in Supplementary Table 1, and their modified nucleotides are underlined. The Rep Δ165-172 (Rep-box-6-del), Rep Δ193-196 (Rep-box-7-del), and Rep Δ198-203 (Rep-box- 8-del) constructs were described in a previous report [9]. Constructs encoding Rep Δ51-55 (Rep-box-2-del) and Rep Y55A (Rep-box-2-mut) proteins were constructed using the primers listed in Supplementary Table 1. The PCR steps were performed as per the instructions of the QuickChange Lighting Site-Directed Mutagenesis Kit (Agilent Technolo- gies, USA): 95 °C for 2 min; 18 cycles at 95 °C for 20 s, 60 °C for 10 s, and 68 °C for 3 min and 24 s; and a final extension at 68 °C for 5 min. The PCR products were cloned into pET32b after Dpn I digestion. The correct clones were verified by DNA sequencing. Plasmids were purified and used to transform E. coli BL21 (DE3) cells.

Expression and purification of recombinant proteins

Protein expression in transformed E. coli BL21 (DE3) cells was induced by treatment with 1.0 mM isopropyl- β-D-1-thiogalactopyranoside (IPTG) at 20 °C for 4 h. The bacteria were pelleted, suspended in binding buffer (250 mM NaCl, 20 mM Tris-HCl, and 10 mM imidazole, pH 7.5) containing 1 M phenylmethanesulfonyl fluoride, and digested with lysozyme at a concentration of 0.5 mg/ mL. After sonication and centrifugation, the supernatant was incubated overnight with Ni Sepharose beads (GE, USA) at 4 °C. The bound beads were washed sequentially with washing buffer I (250 mM NaCl, 20 mM Tris-HCl, and 20 mM imidazole, pH 7.5), washing buffer II (250 mM NaCl, 20 mM Tris-HCl, and 40 mM imidazole, pH 7.5), and washing buffer III (250 mM NaCl, 20 mM Tris-HCl, and 80 mM imidazole, pH 7.5) and subsequently eluted with elution buffer (250 mM NaCl, 20 mM Tris-HCl, and 500 mM imidazole, pH 7.5). The eluted proteins were dialyzed overnight against PBS buffer at 4 °C and then concentrated for further use.

Removal of SUMO tags from the expressed Rep protein

The purified SUMO fusion protein was incubated over- night with the SUMO express protease (Lucigen Cor- poration, USA) and 2 mM fresh dithiothreitol at room temperature. The separation procedures were followed according to the instructions of the Capturem His-Tagged Purification Miniprep Kit (Clontech Laboratories, USA). Briefly, 400 μl xTractor buffer was used to equilibrate the column. The digested protein was loaded into the column and subsequently centrifuged at 11,000×g for 1 min. Three hundred microliters of wash buffer I and II containing (20 mM and 40 mM imidazole, respectively) was loaded onto the column, and the sample was centrifuged again at 11,000×g for 1 min. Finally, 300 μl elution buffer was loaded onto the column, and the sample was centrifuged again at 11,000×g for 1 min. The flowthrough obtained after each centrifugation step was subjected to SDS-PAGE analysis.

Western blotting

Proteins were resolved by SDS-PAGE and transferred to a PVDF membrane. The membrane was blocked with 5% skim milk in TBST buffer (150 mM NaCl, 20 mM Tris, and 0.1% Tween 20, pH 7.5) and then treated with TBST buffer containing an anti-Rep monoclonal antibody at 4 °C overnight. After washing, the bound monoclonal antibodies were probed with horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibody (Santa Cruz, USA). The membrane was developed using an enzyme-linked chemilu- minescence system (ECL Pro, Perkin Elmer, USA).

ATPase and GTPase activity analysis

A colorimetric ATP or GTP assay kit based on the forma- tion of a green color when malachite green reagent reacts with free phosphate (Innova Biosciences, United Kingdom) was used according to the manufacturer’s instructions. One hundred microliters of Tris-HCl (pH 7.5) was used to dilute 200 nmol of the purified Rep protein, which was further incubated with 0.5 mM ATP at 25 °C for 30 min. An enzyme-linked immunosorbent assay was used to detect the phosphomolybdate-malachite green complex based on the absorbance value at a wavelength of 260 nm. The background value derived from the non-enzymatic release of γ-phosphate in the absence of the Rep protein and sub- strates was determined and subsequently subtracted from the enzymatic activity of each group. To set up a standard curve for the measurement of the phosphate concentration (μM) and the rate of release of phosphate (pmol/min/μg) as an indicator of ATPase or GTPase activity, different con- centrations of inorganic acids (50, 45, 40, 35, 30, 25, 20, 15, 10, 5, and 2.5 μM) were tested in parallel in each assay. The effect of the endonuclease inhibitor L-742001 (Sigma, USA) was assessed at different concentrations (0.6, 0.9, 1.2, and 1.5 mM). This inhibitor was incubated with the Rep protein for 60 min prior to the addition of substrate. For ATPase or GTPase activity measurement, data from triplicate experi- ments are presented as the mean (%) ± standard deviation (SD). Statistical analysis was performed using one-way ANOVA and the Student’s t-test. A significant difference was defined as p < 0.05. Endonuclease activity assays Cleavage assays included 3 μg of each purified Rep protein dissolved in 50 mM Tris (pH 7.4), 0.1 M NaCl, and 2.5 mM MgCl2 or 2.5 mM MnCl2 with or without 0.25 mM ssDNA oligonucleotides (P10, 5′-TAGTATT^ACC-3′; P12, 5′-TAG TATT^ACCCC-3′). These oligonucleotides contain the con- served nonamer sequence located in the genome of BFDV, in which ^ indicates the scissile phosphodiester bond and the underlined letters denote the conserved nonanucleotides. Samples were incubated at 37 °C for 1 h and then subjected to 12% SDS-PAGE analysis. The effect of different com- binations of metal ions, including 2.5 mM NaCl, MgCl2, MnCl2, CaCl2, and ZnCl2, on the endonuclease activity of the Rep protein was studied. Thirty mM EDTA was added prior to incubation with P10 or P12 oligonucleotides in the presence of NaCl, MgCl2, and MnCl2 to assess its effect. In addition, the nucleotidyl transfer activity of the Rep protein- P10 and Rep protein-P12 adducts was assessed using 0.25 mM oligonucleotide P15 (5′-GCGGCGGTTAGTATT^-3′), which includes residues −1 to −15 located at the 5′ end of the scissile bond and serves as a preformed acceptor. This oligonucleotide was added to the cleavage mixture, which was then incubated at 37 °C for 1 h in order to examine whether the free Rep protein was released after P15 oligo- nucleotides were joined to the linked oligonucleotides of the adducts. All of the covalent protein-DNA adducts were isolated on an SDS-PAGE gel and visualized by staining with Coomassie blue. The effects of the ATPase inhibitors sodium orthovanadate (NaVO4), sodium azide (NaN3), and N-ethylmaleimide (NEM), the GTPase inhibitors dynasore and linoleic acid, and the endonuclease inhibitor L-742001 on the endonuclease activity of the Rep protein were tested. All inhibitors (Sigma, USA) were incubated with the Rep protein for 60 min prior to the addition of P10 or P12 oligo- nucleotides to the reaction mixture. ImageMaster software (Amersham, USA) was used to perform the densitometric analysis on SDS-PAGE gels. Electrophoretic mobility shift assay (EMSA) Three dsDNA molecules corresponding to sequences pre- sent in the BFDV genome, including dsP16 (5′-17GGGCAC CGGGGCACTG32-3′), dsP26 (5′-7CCGCCGCCTGGGGCA CCGGGGCACTG32-3′), and dsP36 (5′-7CCGCCGCCT GGGGCACCGGGGCACTGCAGCCAT39-3′) were used as probes in the EMSA assay. The locations of these sequences in the BFDV genome are shown in Supplementary Fig. 2B. One ssDNA was synthesized with a biotin label at the 5′ terminal end for generating dsDNA probes. The sense and antisense oligonucleotides with or without the biotin label were heated together at 95 °C for 10 min and subsequently cooled down to room temperature. An EMSA was conducted using a LightShift chemiluminescence system (Thermo Fisher Scientific, USA). Twenty microliters of binding buffer (2.5% glycerol, 5 mM MgCl2, 0.05% NP-40, and 50 ng of poly dI-dC per µl), 50 ng of biotin-labeled probe, and 1.5 μg of Rep protein were incubated at room temperature for 5 min. In the competition group, 150 ng of unlabeled probe was pre-incubated with samples at room temperature for 20 min prior to the addition of 50 ng of biotin-labeled probe. The complexes were resolved on a 6% native polyacrylamide gel and then transferred to a nylon membrane. After UV cross-linking at 120 mJ/cm2, the membrane was blocked and hybridized with HRP-conjugate at room temperature for 15 min. After washing, the membrane was developed using an enzyme-linked chemiluminescence system. The Rep protein was pre-treated with ATPase, GTPase, and endonuclease inhibitors for 60 min before the addition of dsP36 oligo- nucleotides. ImageMaster software (Amersham, USA) was used to perform densitometric analysis on the blots. Data from triplicate experiments are presented as mean values (%). 3D modeling of the functional motifs of the Rep protein The 3D structure of the Rep protein of BFDV was modeled using the SWISS-MODEL web server [15–17]. A homol- ogy modeling approach was used to construct a 3D model of the BFDV by using Blast and HHBlits to search template libraries, which included SMTL version 2017-11-09 and PDB release 2017-11-03 [18, 19]. The template selected for homology modeling of the Rep protein endonuclease motif was the replication initiation protein of PCV (PDB entry: 2HW0), and the helicase from papillomavirus in complex with its enhancer E2 (PDB entry: 1TUE) was selected as the template for the ATPase/GTPase motif. To merge the endo- nuclease motif and the ATPase/GTPase motif of the Rep protein, the MDA (multidomain assembler) approach was used with the UCSF Chimera package [20, 21]. The ATP- binding site of the Rep protein was modeled as described [22]. Results Expression and purification of BFDV Rep and its mutants In a previous study, the Walker A, Walker B, and 193GYDG196 motifs were found to be involved in the ATPase/ GTPase activity of the BFDV Rep protein [9]. Therefore, Rep proteins carrying a point mutation in each of these motifs, namely Rep D195A (Rep-box-7-mut), Rep D202A (Rep-box-8-mut), and Rep K171A (Rep-box-6-mut), respectively, were separately expressed and purified. Their molecular weights were 53.5 kDa, 53.5 kDa, and 53.49 kDa, respectively, as determined by Western blot using anti-Rep monoclonal antibodies (Fig. 2A, B, and C). Moreover, to identify motifs associated with the endonuclease activity, Rep proteins, containing deletions, namely Rep Δ14-18 (Rep-box-1-del), Rep Δ61-65 (Rep-box-3-del), and Rep Δ89-92 (Rep-box-4-del), or point mutations, namely Rep N17N18A (Rep-box-1-mut), Rep K61K62A (Rep-box-3- mut), and Rep Y89A (Rep-box-4-mut), were produced in a soluble form, purified, and analyzed by SDS-PAGE and Western blotting using anti-Rep monoclonal antibodies (Fig. 2D-I). The molecular weights of the Rep Δ14-18 (Rep- box-1-del), Rep Δ61-65 (Rep-box-3-del), Rep Δ89-92 (Rep- box-4-del), Rep N17N18A (Rep-box-1-mut), Rep K61K62A (Rep-box-3-mut), and Rep Y89A (Rep-box-4-mut) proteins were 52.96 kDa, 52.93 kDa, 53.06 kDa, 53.46 kDa, 53.43kDa, and 53.45 kDa, respectively. Mass spectrometry was also used to validate the authenticity of the expressed pro- teins (data not shown). An additional deletion mutant, Rep Δ51-55 (Rep-box-2-del), and a point mutant, Rep Y55A (Rep-box-2-mut), were also expressed, purified, and vali- dated. The molecular weights of the Rep Δ51-55 (Rep-box- 2-del) and Rep Y55A (Rep-box-2-mut) proteins were 52.95 kDa and 53.45 kDa, respectively (data not shown). A soluble Rep protein with an N-terminal SUMO tag was expressed and purified as a 47 kDa protein (Supplementary Fig. 1A). After digestion with SUMO protease and subsequent purifi- cation, an untagged Rep protein with a molecular weight of 32 kDa was obtained (Supplementary Fig. 1B). The identity of this protein was confirmed by Western blotting using anti- Rep monoclonal antibodies (Supplementary Fig. 1C). Cleavage and binding of the oligonucleotides P10 and P12 by the BFDV Rep protein The cleavage and covalent binding activities of a Rep pro- tein with a Trx tag were tested. Both the P10 and P12 oligonucleotides were cleaved and bound by these Rep proteins within the conserved seven-nucleotide sequence TAGTATT, as indicated by slower migration of Rep pro- tein-DNA adducts in an SDS-PAGE gel than that of the Rep protein alone (Fig. 3A). The identity of the free Rep protein and the Rep protein-DNA adducts was confirmed by Western blotting using an anti-Rep monoclonal anti- body (data not shown). For the tagged-Rep protein, on average, 52.93% and 51.55% of the Rep protein was pre- sent in adducts with P10 and P12, respectively. To rule out any effect of the Trx tag on the endonuclease activity, an untagged Rep protein prepared using a SUMO expression system was incubated with the oligonucleotides P10 and P12. Similarly, the untagged Rep protein was catalytically active, and 50.1% and 53.13% of the Rep protein, on aver- age, formed complexes with P10 and P12, respectively (Supplementary Fig. 2A). This demonstrated that the tagged Rep proteins exhibited the same level of activity as the untagged ones. Covalent adducts containing the tri- nucleotide ACC and the pentanucleotide ACCCC in P10 and P12, respectively, exhibited nucleotidyl transfer activ- ity. Increased amounts of Rep protein were released from the Rep-P10 and Rep-P12 adducts when the concentration of oligonucleotide P15 was increased from 10 to 20 mM (Fig. 3B and C). Effect of divalent metal ions on endonuclease activity The cleavage of and covalent binding to oligonucleotides P10 and P12 by the Rep protein were only observed in reac- tion mixtures containing 0.1 M NaCl. For oligonucleotide P10, the mean amount of the DNA-Rep protein adduct increased 16.4% and 20.45% compared to NaCl only after the addition of 2.5 mM MgCl2 and MnCl2, respectively. For oligonucleotide P12, the mean amount of the DNA-Rep protein adduct increased 11.66% and 15.66% compared to NaCl only after the addition of 2.5 mM MgCl2 and MnCl2, respectively. The cleavage was not enhanced any further by the simultaneous presence of NaCl, MgCl2, and MnCl2, and this activity was partially inhibited by the addition of 30 mM EDTA prior to incubation with P10 or P12. Moreover, ZnCl2 and CaCl2 did not affect the cleavage activity of Rep proteins in the presence of NaCl. The relative potency of divalent metal ions for endonuclease activity was as follows: Mn2+ > Mg2+ > Mg2+ + Mn2+ > Zn2+ > Ca2+. Thus, Mn2+ and Mg2+ were the most efficient divalent metal ions for the endonuclease activity (Fig. 4).

Lack of endonuclease activity of Rep mutant proteins

The endonuclease activity of Rep mutant proteins, includ- ing Rep Δ165-172 (Rep-box-6-del), Rep Δ193-196 (Rep- box-7-del), Rep Δ198-203 (Rep-box-8-del), Rep K171A (Rep-box-6-mut), Rep D195A (Rep-box-7-mut), and Rep D202A (Rep-box-8-mut), was examined. The results indi- cated that deletion or point mutations in any of the ATPase/ GTPase motifs abrogated the cleavage and covalent binding activities of the respective Rep protein and prevented for- mation of an adduct with oligonucleotide P10. When these ATPase/GTPase mutant proteins were tested using oligonu- cleotide P12, no endonuclease activity was detected. The mutants Rep Δ14-18 (Rep-box-1-del), Rep Δ61-65 (Rep- box-3-del), Rep Δ89-92 (Rep-box-4-del), Rep Δ170-172 (Rep-box-5-del), Rep N17N18A (Rep-box-1-mut), Rep K61K62A (Rep-box-3-mut), Rep Y89A (Rep-box-4-mut), and Rep K171A (Rep-box-5-mut) proteins lost all their abil- ity to cleave and covalently bind oligonucleotide P10 and to cleave oligonucleotide P12 (Fig. 5A and B). As expected, the endonuclease activity of the mutants Rep Δ51-55 (Rep- box-2-del) and Rep Y55A (Rep-box-2-mut) was reduced with both P10 and P12 (Supplementary Fig. 3A).

Inhibition of the endonuclease activity of the Rep protein by NEM and inhibition of the endonuclease, ATPase, and GTPase activity of the Rep protein by L‑742001

A previously conducted study indicated that the Rep pro- tein of BFDV has dual ATPase/GTPase activity that is sig- nificantly inhibited by ATPase inhibitors such as NaVO4, NEM, and NaN3 as well as by the GTPase inhibitors such as dynasore and linoleic acid [9]. In this study, when the Rep protein was treated with NaVO4 or NaN3 for 60 min, its endonuclease activity was not affected. The cleavage of P10 and P12 oligonucleotides and the formation of DNA- Rep protein adducts were observed. However, no DNA- protein adduct was seen when the ATPase inhibitor NEM was tested (Fig. 6A). The endonuclease activity of the Rep protein for P10 and P12 oligonucleotides was not affected by either dynasore or linoleic acid (Fig. 6B). When the Rep protein was treated with 0.6, 0.9, or 1.2 mM L-742001, the amount of Rep protein-DNA adduct observed decreased in a dose-dependent manner. The endonuclease activity of the Rep protein on oligonucleotides P10 was completely inhibited when 1.5 mM L-742001 was used (Fig. 6C). The cleavage and covalent binding to P12 was also inhibited by L-742001 in a dose-dependent manner. The complete inhi- bition of endonuclease activity of the Rep protein for P12 oligonucleotides was observed at a concentration of 1.5 mM (Fig. 6D). When Rep proteins alone were treated with 0.6, 0.9, 1.2, and 1.5 mM L-742001 for 60 min, the rela- tive levels of ATPase activity exhibited by these proteins were 100% ± 2.06%, 64% ± 2.34%, 54.9% ± 1.17%, 46% ± 2.01%, and 36.95% ± 0.44%, respectively (all p < 0.05). The relative GTPase activity levels for Rep proteins alone when treated with 0.6, 0.9, 1.2, and 1.5 mM L-742001 for 60 min were 56.41% ± 3.18%, 52.11% ± 3.89%, 36.81% ± 3.73%, and 18.31 ± 1.29% (all p < 0.05) (Fig. 6E). Loss of ATPase/GTPase activity of the mutant Rep proteins except for Rep Δ51‑55 (Rep‑box‑2‑del) and Rep Y55A (Rep‑box‑2‑mut) The relative levels of the ATPase activities of the Rep protein, Rep D195A (Rep-box-7-mut) protein, and Rep D202A (Rep-box-8-mut) protein were 100% ± 1.59%, 49.49% ± 2.47%, and 33.32% ± 4.56%, respectively (all p < 0.05) (Fig. 7A). The relative levels of GTPase activi- ties of the Rep protein, Rep D195A (Rep-box-7-mut) protein, and Rep D202A (Rep-box-8-mut) protein were 100% ± 1.24%, 36.56% ± 2.5%, and 26.85% ± 3.35%, respectively (all p < 0.05) (Fig. 7B). The relative levels of ATPase and GTPase activities compared with those of the Rep protein from a previous study were 49.97% ± 1.58% and 55.45% ± 1.27%, respectively (p < 0.05) [9]. The relative levels of the ATPase activities of the Rep protein, Rep Δ14-18 protein (Rep-box-1-del), Rep Reduced binding of BFDV dsDNA by mutant Rep proteins, and blockage of DNA binding by inhibitors The Rep protein containing a Trx tag did not bind dsP16 or dsP26 DNA, but it bound dsP36 DNA (data not shown). To exclude the possibility that the binding between the Rep pro- tein and dsP36 DNA was caused by the Trx tag, an untagged Rep protein was tested for its ability to bind dsP36 DNA. A shifted band corresponding to a dsDNA-protein com- plex was observed, and this complex was no longer visible after the addition of unlabeled dsP36 DNA (Supplementary Fig. 2C). All of the Rep mutants carrying a deletion or point mutation in the ATPase/GTPase motifs exhibited decreased dsP36 DNA binding ability. The relative binding levels of the Rep protein, Rep Δ165-172 (Rep-box-6-del) protein, Rep Δ193-196 (Rep-box-7-del) protein, and Rep Δ198-203 (Rep-box-8-del) were 100%, 78.97%, 76.55%, and 66.82%, respectively (Fig. 8A). The relative binding levels of the Rep protein, Rep K171A (Rep-box-5-mut) protein, Rep D195A protein (Rep-box-7-mut), and Rep D202A (Rep-box-8-mut) protein were 100%, 76.88%, 84.75%, and 59.21%, respec- tively (Fig. 8B). The DNA binding abilities of most of the Rep proteins with deleted or point mutants at the predicted endonuclease motifs were also reduced. The relative bind- ing levels of the Rep protein, Rep Δ14-18 (Rep-box-1-del) protein, Rep Δ61-65 (Rep-box-3-del) protein, Rep Δ89-92 (Rep-box-4-del) protein, and Rep Δ170-172 (Rep-box-5- del) protein were 100%, 87.63%, 104.72%, 97.22%, and 90.36%, respectively (Fig. 8C). The relative binding levels of the Rep protein, Rep N17N18A (Rep-box-1-mut) protein, Rep K61K62A (Rep-box-3-mut) protein, and Rep Y89A (Rep-box-4-mut) protein were 100%, 95.3%, 86.23%, and 86.34%, respectively (Fig. 8D). In contrast, the Rep Δ51-55 (Rep-box-2-del) protein and Rep Y55A (Rep-box-2-mut) protein showed higher levels of binding to dsP36 DNA (Supplementary Fig. 3D). The relative binding levels of the Rep protein, Rep Δ51-55 (Rep-box-2-del) protein, and Rep Y55A (Rep-box-2-mut) protein were 100%, 110.31%, and 99.08%, respectively. When the Rep protein was pretreated with NaVO4, NaN3, dynasore or linoleic acid for 60 min, the binding of the Rep protein to dsP36 DNA was completely inhibited. However, a small amount of Rep protein continued to bind dsP36 DNA after pretreatment with NEM for 60 min (Fig. 9A). Next, when the Rep protein was pretreated with L-742001 for 60 min, the binding of the Rep protein to dsP36 DNA was significantly inhibited (Fig. 9B). Association of the ATPase/GTPase motif with the endonuclease motif in the Rep protein structural model A 3D model of the Rep protein suggested that the endo- nuclease motif is intimately associated with the ATPase/ GTPase motif in three dimensions through helix-loop and sheet-loop interactions (Fig. 10A). The boxes 6, 7, and 8, located in the N-terminal motif, form an interaction surface that appears to stabilize the conformation of the ATP bind- ing site in the C-terminal motif (Fig. 10B). The modeled ATP binding site was located in loop252-260 and loop165-172. The mutations in boxes 3, 5, and 6 were close to the dsDNA binding site, while those in boxes 1, 4, 7, and 8 were dis- tant from this site. It is noteworthy that the dsDNA binding site residue K61 of the Rep protein, positioned in the major groove of the dsDNA, did not form electrostatic interactions with phosphate oxygen of the dsDNA. This residue is in a loop region connected by a beta-strand with amino acid residues in box 2 (Fig. 10C). Discussion To facilitate biochemical studies and improve the solubil- ity of expressed proteins, fusion proteins with His or GST tags have been used to examine the endonuclease activity of the Rep proteins of ssDNA virus such as PCV. Since this strategy was proven to be effective in a previous study, a BFDV Rep protein carrying a Trx tag was used in this study [7, 9, 11]. Furthermore, a Rep protein with a SUMO tag was expressed and purified after the removal of the SUMO tag using SUMO protease [23]. The unmodified Rep protein and the Rep protein tagged with Trx yielded similar results, indicating that the biological activity of the expressed Rep protein was not affected by the presence of the tag. Since the SUMO system involves a cumbersome procedure and resulted in much lower yields of the Rep protein, the pre- ferred methodology adopted in this study was tagging the Rep protein with a Trx tag. Acidianus two-tailed virus (ATV) contains a chaperone protein p892, which belongs to the AAA + ATPase super- family, carrying both the Walker A and Walker B motifs. The ATPase motifs are located at the N-terminus of the p892 protein and facilitate the binding of viral dsDNA [24]. In contrast, the ATPase/GTPase motifs of BFDV are located at the C-terminus of the Rep protein, and these motifs were not involved in the binding of viral dsDNA containing the replication origin. The endonuclease motifs of the Rep pro- tein of BFDV and the p892 protein of ATV are located at the N-terminus and C-terminus, respectively. The endonuclease motif of the p892 protein of ATV belongs to the PDD/ExK family and is similar to those found in type II restriction endonucleases but differs from the endonuclease motifs in the Rep protein of BFDV [24]. Although 0.1 M NaCl had been used as the basic ingre- dient of the cleavage buffer in a previous study, different divalent metal ions were also added to the buffer to exam- ine their potential [11]. It is believed that divalent metal ions played an important role in the cleavage activities of endonucleases. A range of divalent ions, including Mn2+, Mg2+, Ca2+, Cu2+, Ni2+, and Zn2+, were found to effectively stimulate the endonuclease activity of the relaxase TrwC in an in vitro assay [25]. This finding corroborates our finding that Mg2+ and Mn2+ were the strongest cofactors for endo- nuclease activity of the Rep protein of BFDV. Although the coordination geometry of the protein-metal ion complex is similar for Mg2+ and Mn2+, Mg2+ is the more plentiful and accessible ion under physiological conditions. The cleavage activity of the NS1 protein of mouse minute virus and the U94 protein of human herpesvirus (HHV) have also been shown to be enhanced by Mg2+ [26, 27]. The endonuclease activity of PCV Rep protein is enhanced by addition of Mn2+ or Mg2+ [11]. This is due to neutralization of repulsive elec- trostatic forces between negatively charged amino acid side chains and nucleic acids as well as the ability of the specific coordination geometry of the protein-Mn2+/Mg2+ complex to promote endonuclease reactivity. The intergenic region of BFDV includes a stem-loop structure comprising inverted repeats of 5′-CAGGCG GCGG-3′ flanking the 5′-TAGTATTAC-3′ nonanucleotide at its apex. The sequence of the nonanucleotide of PCV is identical to that of BFDV, but the inverted repeat of PCV, 5′-AAGTGCGCTG-3′, is different. This suggests that although both viruses belong to the genus Circovirus, differ- ent species may be infected by these viruses due to the dif- ferences in their genomic structures. Both PCV and BFDV had the same cleavage site in the nonanucleotide, which is located between the seventh and the eighth nucleotides. The presence of at least three copies of the CCCTAA motif has been shown to be required for the binding of the U94 pro- tein of HHV to dsDNA [27]. Similar sequence specificity for binding of the Rep proteins of PCV and BFDV to viral DNA has been observed with respect to the iterons of the CGGCAG and GGGGCACC/T motifs, respectively. At least two copies of these iterons are necessary for binding [7, 28]. The GGGGCACC/T iteron of BFDV is located next to the stem-loop structure, in a position similar to that of the CGG CAG iteron of PCV, and this site is regarded as the minimal binding site for the Rep protein [7]. Despite the presence of two copies of the GGGGCACC/T iteron in both dsP16 and dsP26 DNA, only dsP36 DNA can be bound by the BFDV Rep protein. In addition to the iterons, the right arm of the stem-loop structure, which is present in dsP36 DNA, but not dsP16 or dsP26, appears to be necessary to stabilize the dsDNA-Rep protein complex [11]. Similar to PCV, several conserved motifs, including the 14FTLNN18 motif, 51HLQGY55 motif, 89YCSK92 motif, and 170GKS172 motif, are found in the Rep protein of BFDV. The FTLNN, HLQGY, and YCSK motifs were associated with the endonuclease activity of PCV and BFDV. When the YCSK motif of the PCV Rep protein was mutated, its ATPase activity and ability to bind the viral dsDNA remained unaffected. Only the endonuclease activity of the Rep protein of PCV was abrogated [7]. As discussed above, the 3D model of the Rep protein shows that the N-terminal endonuclease motifs stabilize the conformation of the C-ter- minal ATPase motifs. Therefore, deletions in the 14FTLNN18 motif, 61KKRLS65 motif, and 89YCSK92 motif would be pre- dicted to affect the stability of the ATP binding site, result- ing in decreased ATPase/GTPase activity. On the other hand, the 51HLQGY55 motif, which is located immediately beneath the dsDNA binding site, presumably introduces conforma- tional tension at the ATP binding site. This tension might be released by deletions or substitutions in the 51HLQGY55 motif, leading to a more ideal ATP binding site for catalysis. Moreover, when the lysine residues in the 165GPPGCGKS172 motif, 61KKRLS65 motif, and 170GKS172 motif, which are close to the dsDNA binding site, were mutated to alanine, the dsDNA binding ability of the Rep protein was lost due to a lack of electrostatic interactions. Recent studies have indi- cated that, upon dsDNA binding, the structural conformation in a remote region of the Rep protein of PCV might change and thereby affect the dsDNA binding affinity [11]. In addi- tion, such structural changes have been shown to involve the catalytic site. Since the mutations in the 193GYDG196 motif, 14FTLNN18 motif, and 89YCSK92 motif were in a remote region and involved the catalytic site, it would be reason- able to conclude that these mutants had an impact on the dsDNA binding affinity of the Rep protein. Normally, the dsDNA binding site residue K61 of the Rep protein, which does not form electrostatic interactions with the dsDNA, is connected by a beta-strand to the 51HLQGY55 motif. Upon deletion of the 51HLQGY55 motif, the residue K61 may alter its position relative to the dsDNA backbone, leading to bet- ter dsDNA binding. Each RCR cycle ends with the cleavage of a nonanucleotide and re-joining of the new ssDNA carried out by the Rep protein, and this cleavage of the phosphodiester bond does not require ATP. Therefore, ATPase inhibitors, such as NaVO4 and NaN3, used at concentrations that are inhibitory for ATPase activity, do not affect the endonuclease activity. However, the endonuclease activity of the Rep protein was inhibited by a thiol-conjugating agent, NEM. A study has shown that NEM can irreversibly modify cysteine residues, which might cause a change in protein structure and a loss of activity [29]. NEM has been used to inhibit the RNase H activity of the reverse transcriptase of human immuno- deficiency virus 1 by blocking its endonuclease activity [30]. When the DNA-binding protein (DBP) of Autographa californica multiple nucleopolyhedrovirus was treated with NEM, although the binding capacity of DBP for dsDNA was not completely inhibited, the ability of DBP to unwind and renature the replicative intermediate was weakened. This is similar to our observations with the Rep protein of BFDV, in which the ability of the protein to bind dsP36 DNA was partially inhibited by NEM treatment [31]. The RNA-dependent RNA polymerase of influenza virus comprises the PA, PB1, and PB2 subunits. The N-terminal domain of the PA subunit, which belongs to the PDD/ExK family, possesses endonuclease activ- ity that can be inhibited by L-742001, as was the case with BFDV. This endonuclease inhibitor has been used as an antiviral inhibitor for bunyaviruses, arenaviruses, and orthomyxoviruses. Since L-742001 chelates divalent metal ions through its diketo acid moiety to inhibit the endonuclease activity of PA of influenza virus, it interacts efficiently with divalent-cation-binding proteins, including the Rep protein of BFDV [32, 33]. Most importantly, to the best of our knowledge, this is the first report showing that L-742001 inhibits the endonuclease activity of the Rep protein of a circovirus. References 1. Rahaus M, Wolff MH (2003) Psittacine beak and feather disease: a first survey of the distribution of beak and feather disease virus inside the population of captive psittacine birds in Germany. J Vet Med B Infect Dis Vet Public Health 50:368–371 2. Bassami MR, Berryman D, Wilcox GE, Raidal SR (1998) Psit- tacine beak and feather disease virus nucleotide sequence analysis and its relationship to porcine circovirus, plant circovirus, and chicken anemia virus. Virology 249:453–459 3. Tood D (2000) Circovirus: immunosuppressive threats to avian species: a review. Avian Pathol 29:373–394 4. Varsani A, Regnard GL, Bragg R, Hitzeroth II, Rybicki EP (2011) Global NEM inhibitor genetic diversity and geographical and host-species dis- tribution of beak and featherdisease virus isolates. J Gen Virol 92:752–767
5. Huang SW, Chiang YC, Chin CY, Tang PC, Liu PC, Wang CY (2016) The phylogenetic and recombinational analysis of beak and feather disease virus Taiwan isolates. Arch Virol 161:2969–2988
6. Hsu CM, Ko CY, Tsai HJ (2006) Detection and sequence analysis of avian polyomavirus and psittacine beak and feather disease virus from psittacine birds in Taiwan. Avian Dis 50:348–353
7. Steinfeldt T, Finsterbusch T, Mankertz A (2007) Functional analy- sis of cis- and trans-acting replication factors of porcine circovirus type 1. J Virol 81:5696–5704
8. Clerot D, Bernardi F (2006) DNA helicase activity is associated with the replication initiator protein Rep of tomato yellow leaf curl geminivirus. J Virol 80:11322–11330
9. Huang SW, Liu HP, Chen JK, Shien YW, Wong ML, Wang CY (2016) Dual ATPase and GTPase activity of the replication-asso- ciated protein (Rep) of beak and feather disease virus. Virus Res 213:149–161
10. Orozco BM, Miller AB, Settlage SB, Hanley-Bowdoin L (1997) Functional domains of a geminivirus replication protein. J Biol Chem 272:9840–9846
11. Vega-Rocha S, Byeon IL, Gronenborn B, Gronenborn AM, Cam- pos-Olivas R (2007) Solution structure, divalent metal and DNA binding of the endonuclease domain from the replication initiation protein from porcine circovirus 2. J Mol Biol 367:473–487
12. Laufs J, Schumacher S, Geisler N, Jupin I, Gronenborn B (1995) Identification of the nicking tyrosine of geminivirus Rep protein. FEBS Lett 377:258–262
13. Steinfeldt T, Finsterbusch T, Mankertz A (2001) Rep and Rep’ protein of porcine circovirus type 1 bind to the origin of replica- tion in vitro. Virology 291:152–160
14. Campos-Olivas R, Louis JM, Clérot D, Gronenborn B, Gronen- born AM (2002) The structure of a replication initiator unites diverse aspects of nucleic acid metabolism. Proc Natl Acad Sci USA 99:10310–10315
15. Arnold K, Bordoli L, Kopp J, Schwede T (2006) The SWISS- MODEL workspace: a web-based environment for protein struc- ture homology modelling. Bioinformatics 22:195–201
16. Benkert P, Biasini M, Schwede T (2011) Toward the estimation of the absolute quality of individual protein structure models. Bioin- formatics 27:343–350
17. Biasini M, Bienert S, Waterhouse A, Arnold K, Studer G, Schmidt T, Klefer F, Gallo Cassarino T, Bertoni M, Bordoli L, Schwede T (2014) SWISS-MODEL: modelling protein tertiary and quater- nary structure using evolutionary information. Nucleic Acids Res 42:W252–W258
18. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25:3389–3402
19. Remmert M, Biegert A, Hauser A, Soding J (2011) HHblits: light- ning-fast iterative protein sequence searching by HMM–HMM alignment. Nat Methods 9:173–175
20. Hertig S, Goddard TD, Johnson GT, Ferrin TE (2015) Multid- omain assembler (MDA) generates models of large multidomain proteins. Biophys J 108:2097–2102
21. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE (2004) UCSF Chimera—a visualiza- tion system for exploratory research and analysis. J Comput Chem 25:1605–1612
22. Abbate EA, Berger JM, Botchan MR (2004) The X-ray struc- ture of the papillomavirus helicase in complex with its molecular matchmaker E2. Genes Dev 18:1981–1996
23. Cherney BW, Chaudhry B, Bhatia K, Butt TR, Smulson M (1991) Expression and mutagenesis of human poly(ADP-ribose) poly- merase as a ubiquitin fusion protein from Escherichia coli. Bio- chemistry 30:10420–10427
24. Erdmann S, Scheele U, Garrett RA (2011) AAA ATPase p529 of Acidianus two-tailed virus ATV and host receptor recognition. Virology 421:61–66
25. Boer R, Russi S, Guasch A, Lucas M, Blanco AG, Perez-Luque R, Coll M, de la Cruz F (2006) Unveiling the molecular mechanism of a conjugative relaxase: the structure of TrwC complexed with a 27-mer DNA comprising the recognition hairpin and the cleavage site. J Mol Biol 358:857–869
26. Tewary SK, Liang L, Lin Z, Lynn A, Cotmore SF, Tattersall P, Zhao H, Tang L (2015) Structures of minute virus of mice rep- lication initiator protein N-terminal domain: insights into DNA nicking and origin binding. Virology 476:61–71
27. Trempe F, Gravel A, Dubuc I, Wallaschek N, Collin V, Wal- laschek N, Collin V, Gilbert-Girard S, Morissette G, Kaufer BB, Flamand L (2015) Characterization of human herpesvirus 6A/B U94 as ATPase, helicase, exonuclease and DNA-binding proteins. Nucleic Acids Res 43:6084–6098
28. Londono A, Riego-Ruiz L, Arguello-Astorga GR (2010) DNA- binding specificity determinants of replication proteins encoded by eukaryotic ssDNA viruses are adjacent to widely separated RCR conserved motifs. Arch Virol 155:1033–1046
29. Su D, Delaplane S, Luo M, Rempel DL, Vu B, Kelley MR, Gross ML, Georgiadis MM (2011) Interaction of APE1 with a redox inhibitor: Evidence for an alternate conformation of the enzyme. Biochemistry 50:82–92
30. Zhan X, Tan CK, Scott WA, Mian AM, Downey KM, So AG (1994) Catalytic distinct conformation of the ribonuclease H of HIV-1 reverse transcriptase by substrate cleavage patterns and inhibition by azidothymidylate and N-ethylmaleimide. Biochem- istry 33:1366–1372
31. Mikhailov VS, Vanarsdall AL, Rohrmann GF (2008) Isolation and characterization of the DNA-binding protein (DBP) of the Autographa californica multiple nucleopolyhedrovirus. Virology 370:415–429
32. DuBios RM, Slavish PJ, Baughman BM, Yun MK, Bao J, Webby RJ, Webb TR, White SW (2012) Structural and biochemical basis for development of influenza virus inhibitors targeting the PA endonuclease. PLoS Pathog 8:e1002830
33. Yan Z, Zhang L, Fu H, Wang Z, Lin J (2014) Design of the influenza virus inhibitors targeting the PA endonuclease using 3D-QSAR modeling, side-chain hopping, and docking. Bioorg Med Chem Lett 24:539–547

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.