D(

SCD

a

ARRAA

KBRAG

1

a

h0

Virus Research 213 (2016) 149–161

Contents lists available at ScienceDirect

Virus Research

j ourna l h o mepa ge: www.elsev ier .com/ locate /v i rusres

ual ATPase and GTPase activity of the replication-associated proteinRep) of beak and feather disease virus

hr-Wei Huang, Hao-Ping Liu, Jui-Kai Chen, Yao-Wen Shien, Min-Liang Wong,hi-Young Wang ∗

epartment of Veterinary Medicine, National Chung Hsing University, 250 Road Kuo Kuang, Taichung 402, Taiwan

r t i c l e i n f o

rticle history:eceived 12 June 2015eceived in revised form 1 December 2015ccepted 1 December 2015vailable online 3 December 2015

eywords:eak and feather disease viruseplication-associated proteinTPaseTPase

a b s t r a c t

Background: Psittacine beak and feather disease affects parrots resulting in an immunosuppressive dis-ease that is often characterized by an abnormal shape and growth of the animal’s beak, feathers, andclaws. Beak and feather disease virus (BFDV) is a single-stranded circular DNA virus and is classified asa member of the Circoviridae family. Two major open reading frames (ORFs) are known to encode thereplication-associated (Rep) protein and the capsid-associated (Cap) protein.Methods: The Rep and Cap genes of BFDV were fused with tags and then expressed and purified, respec-tively. Both the ATPase and GTPase activities of the recombinant Rep protein are measured. The substrateand ion preference, the optimal conditions, the effects of ATPase and GTPase inhibitors and the presenceof Cap protein on the ATPase and GTPase activity of the Rep protein are examined. Finally, the effects ofthe Walker A motif, the Walker B motif, and a novel GYDG motif of the Rep protein on the ATPase andGTPase activities are studied by various mutants.Results: The recombinant Rep protein could display ATPase activity and GTPase activity. The Repprotein was able to hydrolyze both deoxyribonucleotides and ribonucleotides. Among nucleosidetriphosphates and deoxynucleoside triphosphates, the substrate preference orders were found to beATP > GTP > CTP ≥ UTP and dATP > dGTP > dTTP > dCTP, respectively. Both the ATPase and GTPase activityof the BFDV Rep protein required magnesium ions and the presence of calcium ions significantly inhib-ited the ATPase and GTPase activity of the Rep protein. The optimal temperatures for ATPase activity andGTPase activity were both 56 ◦C, while their optimal pH values were both pH 7.5. Both the ATPase activ-ity and GTPase activity of the BFDV Rep protein were significantly down-regulated by polynucleotidesand the dsDNA of 36 bp (located in origin of replication) of BFDV genome. The ATPase activity of theBFDV Rep protein was found to be more sensitive to sodium azide than sodium orthovanadate andN-ethylmaleimide. Linoleic acid more strongly inhibited the GTPase activity of the Rep protein thandynasore. This suggested the Rep protein of BFDV should be classified as an F-type ATPase and polyun-saturated fatty acid-sensitive GTPase. In the presence of 10 ng of the Cap protein, the ATPase activityand GTPase activity of the BFDV Rep protein were significantly increased. Furthermore, the BFDV Repprotein contains the Walker A motif, the Walker B motif and a novel GYDG motif. Both the ATPase activity

and the GTPase activity of various deletion and site-directed mutants of the Rep protein affecting thesemotifs were significantly reduced, suggesting all the three motifs contribute to the ATPase and GTPaseactivities; specifically. In addition, the ATPase activity and GTPase activity of the deletion mutants of theRep protein were reversed in the presence of the Cap protein. This is the first example of dual ATPase andGTPase activity being reported in the Rep protein of BFDV.

. Introduction

The genome of beak and feather disease virus (BFDV) consists of single-stranded circular DNA (ss cDNA) and the virus belongs to

∗ Corresponding author. Fax: +886 4 22862073.E-mail address: cyoungwang@dragon.nchu.edu.tw (C.-Y. Wang).

ttp://dx.doi.org/10.1016/j.virusres.2015.12.001168-1702/© 2015 Elsevier B.V. All rights reserved.

© 2015 Elsevier B.V. All rights reserved.

the Circoviridae family (Rahaus and Wolff, 2003). The virus con-sists of a non-enveloped capsid with a diameter of between 14and 16 nm (Bassami et al., 1998). Interestingly, both the virion andgenome size of BFDV makes it the most compact and smallest viral

pathogen that affects parrots. The virus leads to successive mal-formation, fracture, and necrosis of the bird’s feathers, beak, andclaws. In addition to the cracked outer layers of the beaks and claws,the birds are prone to secondary infections because they suffer

150 S.-W. Huang et al. / Virus Research 213 (2016) 149–161

Table 1Primers used for the construction of the Cap and Rep proteins and the Rep mutant proteins.

Primer name Sequences (5′-3′) Nucleotide positiona

Cap F primer CGGGTCGACTCTAACTGTGCATGCGCTAAAT 1239–1259Cap R primer CCCCTCGAGAGAAGTACTGGGATTGTTTGG 1955–1972Rep F outer primer TTAGTCGACATGCCGTCCAAGGAGGGTTC 132–151Rep R outer primer AGGCTCGAGATAATTGATGGGGTGGG 982–998Rep-GST F primer GTGGATCCATGCCGTCCAAGGAGGGT 132–148Rep-GST R primer GACCCGGGTCAATAATTGATGGGGTG 984–998Motif I deletion, Rep F inner GGTTGACGTAATCTACAGATGGGCCAATGAGC 608–623/648–663Motif I deletion, Rep R inner GCTCATTGGCCCATCTGTAGATTACGTCAACC 608–623/648–663Rep-I2F primer CCACCGGGGTGTGGCGCAAGTAGATGGGCCAA 627–658Rep-I2R primer TTGGCCCATCTACTTGCGCCACACCCCGGTGG 627–658Motif II deletion, Rep F inner GAATGGTGGGACGAGGATGTCGTCGTCTT 696–708/721–736Motif II deletion, Rep R inner GCTCATTGGCCCATCTGTAGATTACGTCAACC 696–708/721–736GYDG motif deletion, Rep F inner TATGACGGGGAGGATTGGCTACCGTATTGC 711–725/753–767

AATCC

ftafwatta

n(fct(etbcs

CiRbGi1ohTaovaSAtBcRAwashm

GYDG motif deletion, Rep R inner GCAATACGGTAGCC

a Nucleotide position number was based on GenBank accession no. KC510146.

rom immunosuppression that is caused by lymphocytic deple-ion of the thymus and bursa of Fabricius. Since its first report in

psittacine species from Australia in 1970, this disease has beenound in both captive and wild populations of psittacine around theorld, including Southern Australia, Europe, Africa, the Americas,

nd Asia (Varsani et al., 2011). An epidemiology study has revealedhat this disease was transmitted from Old World psittacine specieso New World psittacine species via international bird movementnd trade (Bert et al., 2005).

With a genome size that ranges from 1992 nucleotides to 2018ucleotides, the virus encodes three major open reading framesORFs), namely V1, V2, and V3 on the sense strand together withour further ORFs, C1, C2, C3, and C4, which are found on theomplementary antisense strand (Raue et al., 2004). Among them,he replication-associated (Rep) protein and the capsid-associatedCap) protein are encoded by ORF V1 and ORF C1, respectively (Hsut al., 2006). In Taiwan, the positive rate for BFDV among asymp-omatic birds is 24% and the major source of BFDV in Taiwan haseen linked to the import of exotic birds (Su et al., 2007). In suchircumstances, both surveillance for the presence of BFDV and thetudy of BFDV have become important issues.

The Rep protein of several ss c-DNA animal viruses, such as theircoviridae, Geminiviridae and Nanoviridae, are involved in the

nitiation of rolling-circle replication (Steinfeldt et al., 2007). Theep protein of porcine circovirus (PCV) has been shown to haveoth ATP binding and ATP hydrolysis activity through a conservedKS motif. If this GKS motif is mutated, the replication of PCV is

nhibited due to there being lower ATPase activity (Warrener et al.,993). The down-regulation or abolishment of the ATPase activityf the Rep protein of tomato yellow leaf curl virus (TYLCV) and ofepatitis C virus (HCV) have been shown to inhibit the replication ofYLCV and HCV (Desbiez et al., 1995; Tate et al., 1989). The ATPasectivity has also been found to be associated with the NS4B proteinf classical swine fever virus (CSFV), the VP4 protein of bluetongueirus (BTV), the A32L protein of orf virus, and the �A protein ofvian reovirus (ARV) (Gladue et al., 2011; Ramadevi and Roy, 1998;u et al., 2007). Most of those proteins include at least two typicalTPase motifs, one of which is the Walker A motif (G/AXXXXGK/ST)

hat is involved in the ATP binding, while the other is the Walker motif (hhhhDD/E), which affects the chelation of the enzyme’sofactor, magnesium ions. Both of these motifs are present in theep protein of BFDV (Walker et al., 1982). This contrasts with the32L protein of orf virus, where there are a motif III and a motif IV,hich allow the formation of a beta-stranded structure with a polar

mino acid at the C-terminus and a hydrophobic beta-strandedtructure at the N-terminus, respectively; these structural featuresave been shown to be associated with the ATPase activity. No suchotifs are present in the NS4B protein of CSFV or the Rep protein

TCCCCGTCATA 711–725/753–767

of BFDV. Moreover, in a number of studies GTPase activity has beenfound to be associated with proteins exhibiting the ATPase activity,but with varying levels of activity. For example, the A32L proteinof orf virus, the �A protein of ARV, and the NS4B protein of CSFVhave ATPase activity levels that are much higher than their GTPaseactivity levels (Gladue et al., 2011; Lin et al., 2011; Su et al., 2007).However, the GTPase activity of the VP4 protein of BTV has beenshown to be higher than its ATPase activity (Ramadevi and Roy,1998). In fact, cell signalling, viral replication, and membrane fusionhave all been shown to be related to the hydrolysis of ATP and GTPfor the NS4B protein of HCV and the 2C protein of poliovirus (Einavet al., 2004; Pfister and Wimmer, 1999). In addition to ATP and GTPbeing the preferred substrates of the VP4 protein of BTV and the�A protein of ARV, the presence of nucleoside triphosphate hydro-lase activity has also been demonstrated for UTP and CTP in theseviruses but at lower activity levels (Ramadevi and Roy, 1998; Suet al., 2007).

The aim of this study is to examine the ATPase and GTPase activ-ity of the Rep protein of BFDV and to characterize its biochemicalfeatures and enzymatic kinetics. The effects of deletions and muta-tion on the BFDV Rep protein in terms of ATPase activity and GTPaseactivity were explored. The results indicated that the BFDV Rep pro-tein has comparable levels of ATPase and GTPase activity and thatthe Walker A motif, the Walker B motif, and the GYDG motif are allnecessary for both ATPase activity and GTPase activity to be present.

2. Materials and methods

2.1. Plasmid construction and verification

The Rep and Cap genes of BFDV (GenBank accession No.KC510146) were cloned into pET32b. The Rep gene was cloned intopGEX. To create mutants, the Rep gene was used as a template fora PCR-driven strategy (Heckman and Pease, 2007). All primers forthe construction of the clones are listed in Table 1. To clone theRep or Cap gene, the amplified products from the first round PCRwere digested followed by cloning. First round PCR products wereused as templates for the second round PCR using outer flankingprimers. The first round of PCR involved 94 ◦C for 30 min followedby 32 cycles of 94 ◦C for 30 s, 55 ◦C for 30 s, and 72 ◦C for 60 s. Thesecond round of PCR involved 94 ◦C for 30 min followed by 36 cyclesof 94 ◦C for 30 s, 55 ◦C for 30 s, and 72 ◦C for 1 min 30 s. Both Sal I andXho I were used to digest the PCR products, which were cloned intopET32b or pGEX. After sequence verification, plasmids containing

the Cap gene, the Rep gene, the Rep gene with the Walker A motifdeleted, the Rep gene with the Walker B motif deleted, and the Repgene with the GYDG motif deleted were respectively named as Cap,Rep, Rep-GST, Rep-box-1-del, Rep-box-2-del, and Rep-box-3-del.

S.-W. Huang et al. / Virus Research 213 (2016) 149–161 151

F racteo Rep-bt

2

iCptTLUc2Dc

2

(i�r1(sFiwi(awfsEd

2S

wwE1BMa

ig. 1. Diagram of the Rep protein of beak and feather disease virus showing the chaf Rep proteins are labelled: Rep, Rep-box-1-del, Rep-box-2-del, Rep-box-3-del, andhat have been deleted or mutated in various mutants.

.2. Site-directed mutagenesis

Site-direct mutagenesis of the Walker A motif (K171A)nvolved using the mutated primers Rep I-2F primer: 5′-CACCGGGGTGTGGCGCAAGTAGATGGGCCAA-3′ and Rep I-2Rrimer: 5′-TTGGCCCATCTACTTGCGCCACACCCCGGTGG-3′ for PCR;he modified nucleotides are identified in Table 1 by underlining.he PCR program followed the protocol of the QuickChangeighting Site-Directed Mutagenesis Kit (Agilent Technologies,SA). Briefly, PCR involved 95 ◦C for 2 min followed by 18 cyclesonsisting of 95 ◦C for 20 s, 60 ◦C for 10 s, and 68 ◦C for 3 min and4 s with a final extension of 68 ◦C for 5 min. After digestion withpn I, the products were cloned into pET32b and their sequenceonfirmed. This clone was named Rep-box-1-mut.

.3. Expression of the recombinant proteins and their purification

The Escherichia coli BL21 strain (Invitrogen) or Top10 strainThermoFisher) transformed with the expression vectors wasnduced at 20 ◦C for 8 h or 30 ◦C for 4 h with 1.0 mM isopropyl--d-1-thiogalactopyranoside (IPTG). The pelleted bacteria were

esuspended in binding buffer (250 mM NaCl, 20 mM Tris–HCl, and0 mM imidazole, pH 7.5) plus 1 M phenylmethanesulfonyl fluoridePMSF). Samples were digested with 0.5 mg/mL lysozyme and thenonicated. Each supernatant was purified using Ni Sepharose 6 Fastlow (GE Healthcare). The nickel-charged sepharose beads werencubated with the supernatant at 25 ◦C for 10 min and then washed

ith washing buffer (500 mM NaCl, 50 mM Tris–HCl, and 40 mMmidazole, pH 7.5). Bound protein was eluted by elution buffer500 mM NaCl, 50 mM Tris–HCl, and 400 mM imidazole, pH 7.5)nd dialyzed. The GST-fusion proteins were purified by incubationith glutathione-coupled sepharose beads (GE Healthcare) at 4 ◦C

or 2 h. The GST-fusion proteins were immobilized on the beads andtored in PBST buffer (1 × phosphate-buffered saline with 2 mMDTA, 0.1% �-mercaptoethanol, 0.2 mM PMSF, and 5 mM benzami-ine) at 4 ◦C prior to use.

.4. Immunoprecipitation, GST-pull down assay,DS-polyacrylamide gel electrophoresis, and Western blotting

The liver and spleen samples from BFDV-positive grey parrotsere solubilized in PBS by sonication. Ten �g of the supernatantere diluted in RIPA lysis buffer (50 mM Tris, 150 mM NaCl, 10 mM

DTA, 0.1% SDS, 1% NP-40, 1 mM phenylmethylsulfonyl fluoride,

0 �g/ml aprotinin, pH 7.5). The lysates were incubated with anti-FDV Cap or anti-BFDV Rep monoclonal antibodies provided by theolecular Biology Lab of National Chung Hsing University, Taiwan

nd then with 75 �l of 50% agarose-coupled goat anti-mouse IgG

ristic motifs associated with ATPases and GTPases. Both the full length and mutantsox-1-mut. The one-letter symbols represent the amino acid sequences of the motifs

(Sigma, USA). DF-1 cells were used as negative control. The pelletedbeads were washed with RIPA lysis buffer. To assess the interac-tion of the Cap and Rep proteins, tissue lysates were incubatedwith anti-BFDV Cap monoclonal antibody and then immunopre-cipitated with agarose-coupled anti-mouse IgG. The purified orprecipitated proteins were separated by the 10% SDS-PAGE andthen stained with Coomassie Brilliant Blue. The purified proteinswere dissolved in 5% formic acid and subjected into the electro-spray ionization tandem mass spectrometric analysis. The identityof the BFDV Cap and Rep proteins were confirmed by the Mascotprogram. For the GST-pulldown assay, 4 �g of Cap or thioredoxin(Trx) proteins were incubated with 5 �g of immobilized GST orGST-Rep proteins in Nonidet P-40 (NP-40) buffer (1% NP-40, 20 mMTris–HCl pH 7.5, 150 mM NaCl, 1 mM Na3VO4, 5 mM EDTA pH 8.0,10% glycerol, 10 mg/ml leupeptin, 10 mg/ml aprotinin, 1 mM PMSF).After an overnight incubation, the bound proteins were washedand separated by 10% SDS-PAGE. For Western blotting, proteinswere transferred, blocked with TBST (150 mM NaCl, 20 mM Tris, and0.1% Tween 20, pH 7.5) containing 7.5% skimmed milk, and incu-bated with anti-His (Genetex, USA), anti-BFDV Cap or anti-BFDVRep monoclonal antibodies or the transferred proteins on the mem-brane were stained with Ponceau S. After washing, the membranewas probed with a horseradish peroxidase (HRP)-conjugated goatanti-mouse antibody (Jackson ImmunoResearch, USA). An enzyme-linked chemiluminescence system (ECL Pro, Perkin Elmer, USA) wasused for development.

2.5. Measurement of ATPase activity and GTPase activity

The release of free �-phosphate (Pi) from ATP or GTP by theRep protein was quantified using a colorimetric ATPase assay kit(Innova Biosciences, United Kingdom) followed the manufacturer’sinstructions. Two hundred nmoles of protein were dissolved in100 �L Tris–HCl (pH 7.5) and incubated with 0.5 mM ATP at 25 ◦C for30 min. After termination, the absorbance at OD260, which was pro-portional to the amount of phosphomolybdate malachite complexpresent, was measured using an enzyme-linked immunosorbentassay. The value of a blank assay that assessed the non-enzymaticrelease of Pi contamination in the absence of proteins and sub-strates was then subtracted from the enzymatic activity. TheATPase and GTPase activities of Trx were measured as negativecontrols. A set of inorganic acids (50, 45, 40, 35, 30, 25, 20, 15,10, 5, and 2.5 �M) was run in parallel to provide a standard curve,which was then used to calculate the total released phosphate

(�M) and the release rate of phosphate (pmol/min/�g) repre-senting the levels of ATPase or GTPase activity. To obtain a timecourse, the reaction was incubated for 30, 60, 90, and 120 min.For the optimal pH, the reaction was incubated in buffers at pH

152 S.-W. Huang et al. / Virus Research 213 (2016) 149–161

Fig. 2. The pET32b plasmids carrying Rep and Cap genes express the Rep and Cap proteins, respectively. Expression and purification of the Trx, Rep (Rep, Rep-box-1-del,Rep-box-2-del, Rep-box3-del, and Rep-box-1-mut), and Cap proteins were carried out and the proteins were separated by SDS-PAGE. Both crude bacterial lysates without (−)or with (+) IPTG-induction are shown (A). The purified Trx, Rep, Rep-box-1-del, Rep-box-2-del, Rep-box3-del, Rep-box-1-mut, and Cap proteins were resolved by SDS-PAGE(B). The purified proteins were examined using anti-His, anti-Rep protein, and anti-Cap protein monoclonal antibodies by Western blotting (C). The reactivity of anti-Rep andanti-Cap monoclonal antibodies against native Rep and Cap proteins was confirmed by the immunoprecipitation and Western blotting (D). The ATPase and GTPase activitiesof the Rep protein of beak and feather disease virus and the Trx protein (E). The free phosphate (�M) released from ATP and GTP by using 50 nM, 100 nM, and 200 nM proteinsin assay. The means ± SD from three independent experiments are presented.

S.-W. Huang et al. / Virus Research 213 (2016) 149–161 153

Fig. 3. The effect of divalent cations (MgCl2, MnCl2, ZnCl2, CaCl2, and MgCl2 plus MnCl2) on the ATPase (A) and GTPase (B) activities of the Rep protein. The ATPase and GTPaseactivities in 2.5 mM MgCl2 were both defined as 100%. The effect of temperature (25 ◦C, 37 ◦C, 56 ◦C, and 80 ◦C) on the ATPase (C) and GTPase (D) activities of the Rep protein.T pH (5T resena

5nNafioGM

he ATPase and GTPase activities at 25 ◦C were both defined as 100%. The effect ofhe ATPase and GTPase activities at pH 5.5 were both defined as 100%. Data are repre identified by *p < 0.05.

.5, pH 6.5, pH 7.5, and pH 8.5. The effects of ATPase inhibitors,amely sodium azide (NaN3), sodium orthovanadate (NaVO4), and-ethylmaleimide (NEM), and GTPase inhibitors, namely dynasorend linoleic acid, were assessed; all inhibitors were purchasedrom Sigma–Aldrich, USA. Specifically, the Rep protein was pre-ncubated with inhibitors for 15 min and 30 min before the addition

f the substrate. To assess the effect of divalent ions, ATPase andTPase activity levels were measured in the presence of 2.5 mMg2+, Mn2+, Zn2+, and Ca2+. To calculate the Vmax and the Michaelis

.5, 6.5, 7.5, and 8.5) on the ATPase (E) and GTPase (F) activities of the Rep protein.ted as the means ± SD from three independent experiments. Significant difference

constant (KM) values, the velocity of the enzymatic reactions wasmeasured in the presence of 25, 50, 100, 200, 400, 800, 1200, and1600 �M ATP or GTP. A Linweaver–Burk plot using reciprocal val-ues for the substrate concentrations and velocities was obtainedand the KM and Vmax values were calculated by the least squareestimates method. To assess substrate specificity, 0.5 mM ATP, GTP,

CTP, UTP, dATP, dGTP, dCTP, and dTTP were used. To measure dose-dependent effects, ATPase or GTPase activity was determined whenthe Rep protein was present at 50, 100, and 200 nmoles. In order

154 S.-W. Huang et al. / Virus Research 213 (2016) 149–161

Fig. 4. The free phosphates (�M) released from ATP and GTP using 200 nM Rep and 200 nM Trx proteins measured at 30 min, 60 min, 90 min, and 120 min after the starto TPase1 calcuc SD fro

tppmomp00P

da

2

NTw

f incubation (A). A Lineweaver–Burk plot of enzyme kinetics of the ATPase and G200, and 1600 �M) were used as substrates. Least square estimates were used tooncentrations are shown in the inset (B). The results are presented as the means ±

o assess the effect of the presence of the Cap protein, the Caprotein (10, 25, 50, and 100 ng) was pre-incubated with the Reprotein for 30 min prior to the ATPase and GTPase activity measure-ents; this was repeated using 100 ng Cap protein in the presence

f mutant Rep proteins. The effect of polynucleotides was deter-ined by pre-incubating the Rep protein for 30 min with 0.25 mM

oly(dA), 0.25 mM poly(dG), 0.25 mM poly(dC), 0.25 mM poly(dT),.25 mM poly(dA) plus poly(dT), 0.25 mM poly(dC) plus poly(dG),.25 mM P10 (5′-TAGTATTACC-3′) single-stranded DNA, 0.25 mM12 (5′-TAGTATTACCCG-3′) single-stranded DNA, or 0.25 mM P36

(5′-CCGCCGCCTGGGGCACCGGGGCACTGCAGCCATTGG-3′)ouble-stranded DNA. The P10, P12, and P36 DNA sequences arell located in origin of replication of BFDV.

.6. Enzyme-linked immunosorbent assays (ELISA)

The Cap protein (250 ng) in TBS buffer (25 mM Tris–HCl, 150 mMaCl, pH 7.5) was used to coat ELISA plates. After washing withBST buffer and blocking with 5% bovine serum albumin, the platesere washed three times with TBST and then different amounts

of the Rep protein. Various amounts of ATP and GTP (25, 50, 100, 200, 400, 800,late the Km and Vmax values. The effects of velocity (nmole/min/mg) and substratem three independent experiments.

(10 ng, 12.5 ng, 25 ng, 50 ng, 100 ng, and 150 ng) of Rep proteinwere added in the presence of micrococcal nuclease (40 units/ml),which was followed by incubation for 30 min. After washing threetimes, monoclonal antibody against the BFDV Rep protein wasadded and the mixture incubated for 60 min. For the negative con-trol, the Trx protein (250 ng) was coated onto the ELISA platesand then challenged with the Rep proteins. After washing threetimes, horseradish peroxidase (HRP)-conjugated goat anti-mouseantibody (Jackson ImmunoResearch, USA) was added as the sec-ondary antibody. After final washing, the chromogenic substratewas added and the absorbance values were quantified at 450 nmby spectrophotometer.

2.7. Statistical analysis

The ATPase and GTPase activity levels were determined

as the mean (%) ± standard deviation (SD) value measured aspmol/min/�g from triplicated experiments. The relative activityor absorbance value at 450 nm was determined using the mean.Finally, one-way ANOVA and the Student’s t-test were used for sta-

Resea

ta

3

3BR

hipbwp5cwtamtrvtapfa

t5nileaiGfaRt(

3A

pw9iw9AtettCt

5

S.-W. Huang et al. / Virus

istical analysis. The presence of a significant difference was defineds p < 0.05.

. Results

.1. Expression and purification of the Cap and Rep proteins ofFDV and the dose-dependent ATPase and GTPase activities of theep protein

The Walker A motif, the Walker B motif, and the GYDG motifave been characterized as motifs associated with ATPase activ-

ty and magnesium binding; all three motifs are found in the Reprotein of BFDV. Accordingly, the following proteins, Trx, Rep, Rep-ox-1-del, Rep-box-2-del, Rep-box-3-del, Rep-box-1-mut, and Capere expressed and purified (Fig. 1). The expected sizes of theseroteins in soluble form were found to be 20 kDa, 53 kDa, 53 kDa,3 kDa, 53 kDa, 53 kDa, and 49 kDa, respectively, and these wereonfirmed by SDS-PAGE (Fig. 2A and B). Since Trx and a 6-His tagere separately added to the N-termini and C-termini of these pro-

eins, respectively, the purified proteins were also checked usingnti-His antibody and Western blotting (data not shown). Theass spectrometry confirmed that the identity of recombinant pro-

eins with 49 kDa and 53 kDa was Rep and Cap proteins of BFDV,espectively. The authenticity of the recombinant proteins was alsoerified by Western blotting using monoclonal antibodies againsthe Rep and Cap proteins of BFDV (Fig. 2C). Besides, the reactivity ofnti-Rep and anti-Cap monoclonal antibodies to native Rep and Caproteins was confirmed by the immunoprecipitation using proteinsrom the BFDV-infected bird. The molecular weights of native Repnd Cap proteins were 33 kDa and 29 kDa, respectively (Fig. 2D).

Free phosphate (�M) release from the substrate ATP increasedo 30.41 ± 4.85, 47.34 ± 1.52, and 75.95 ± 1.79, respectively, when0, 100, and 200 nM of the Rep protein were present. However,o change in the release of free phosphate (�M) was detected

n the presence of the Trx, with release remaining at backgroundevel. The amounts of free phosphate (�M) produced in the pres-nce of 50, 100, and 200 nM the Trx were 0.12 ± 0.12, 0.72 ± 0.12,nd 0.5 ± 0.17, respectively. However, the dose-dependent effect ofncreased amounts of the Rep protein was less obvious when theTPase activity was measured. The free phosphate (�M) released

rom the substrate GTP was increased to 28.25 ± 5.17, 44.57 ± 3.16,nd 69.35 ± 3.26 in the presence of 50, 100, and 200 nM of theep protein, respectively. The background levels of GTPase activi-ies were 0.01 ± 0.02, 1.16 ± 0.02, and 1.06 ± 0.09 �M, respectivelyFig. 2E).

.2. The effects of divalent ions, temperature, and pH on theTPase and GTPase activities of the Rep protein of BFDV

The relative ATPase activities of the Rep proteins in theresence of MgCl2, MnCl2, ZnCl2, CaCl2, and MgCl2 + MnCl2ere 100 ± 8.59%, 56.99 ± 7.47%, 48.79 ± 5.35%, 31.32 ± 5.35%, and

9.74 ± 9.59%, respectively (Fig. 3A). The relative GTPase activ-ties of the Rep proteins in the presence of the same ions

ere 100 ± 4.25%, 46.77 ± 6.5%, 45.07 ± 5.35%, 19.17 ± 6.84%, and6.32 ± 6.84%, respectively (Fig. 3B). These results suggest that theTPase and GTPase activities of the Rep are largely depending on

he presence of magnesium ions. Furthermore, ZnCl2 is able toffectively inhibit the ATPase activity and the GTPase activity ofhe Rep protein. Similar to magnesium ions, calcium ions are ableo activate the ATPase activity of the Rep protein. However, 2.5 mM

aCl2 would seem to significantly suppress the GTPase activity ofhe Rep protein.

The relative ATPase activities of the Rep protein at 25 ◦C, 37 ◦C,6 ◦C, and 80 ◦C were 100 ± 5.63%, 137.32 ± 9.27%, 145.07 ± 6.49%,

rch 213 (2016) 149–161 155

and 23.12 ± 5.35%, respectively (Fig. 3C). Furthermore, the relativeGTPase activities were 100 ± 6.5%, 119.39 ± 7.47%, 152.49 ± 8.86%,and 50.04 ± 5.35% at the same temperatures, respectively (Fig. 3D).The ATPase activity and GTPase activity of the Rep were almostabolished at 80 ◦C.

The relative ATPase activities of the Rep protein were 100 ± 8.6%,131.77 ± 5.35%, 134.81 ± 7.47%, and 132.78 ± 6.84% at pH 5.5, pH6.5, pH 7.5, and pH 8.5, respectively (Fig. 3E), thus the optimal pHwas 7.5. Similarly, the relative GTPase activities of the Rep protein atthe same pH values were 100 ± 8.86%, 131.21 ± 4.25%, 136.8 ± 6.5%,and 135.54 ± 6.38%, respectively (Fig. 3F), which means the optimalpH for GTPase activity was pH 7.5. Thus, the GTPase activity of theRep is the same under neutral conditions with ATPase activity ofthe Rep.

3.3. The time course and enzymatic kinetics of the ATPase andGTPase activities of the Rep protein of BFDV

Next we explored the time-dependence of free phosphates (�M)release from ATP in the presence of 200 nM Rep protein; thisincreased giving values of 31.85 ± 1.74, 65.86 ± 2.09, 83.73 ± 0.53,and 86.01 ± 2.16 after 30, 60, 90 and 120 min. There were nochanges in the release of free phosphate (�M) as the various con-centrations of Trx with the activity remaining at background level.The free phosphates (�M) produced by 200 nM of Trx after 30,60, 90, and 120 min were 1.16 ± 0.06, 1.17 ± 0.03, 1.16 ± 0.04, and1.09 ± 0.04, respectively.

The time-dependent effect was less obvious when the GTPaseactivity of the Rep protein was explored. Using the same timeperiods, the free phosphate (�M) released from the substrateGTP by 200 nM of the Rep protein were increased to 31.21 ± 0.57,63.83 ± 1.08, 83.68 ± 1.5, and 88.13 ± 2.21, respectively, whilethe free phosphate (�M) released by 200 nM of the Trx were1.39 ± 0.05, 1.81 ± 0.1, 1.86 ± 0.15, and 1.8, respectively (Fig. 4A).These findings indicate that both the ATPase activity and theGTPase activity of the Rep protein approached a plateau phaseafter 90 min when 200 nM of the Rep protein is used. To ensurethat ATP and GTP hydrolysis does followed a linear relationshipbetween speed and substrate concentration, 50 nmoles of the Repprotein were used. A linear relationship was found to exist withina range of 25 �M and 1200 �M of ATP/GTP and the initial veloc-ity (nmole/min/mg) between 25 �M and 400 �M for the ATPaseactivity of the Rep was higher than for the GTPase activity of theRep. This is demonstrated in the inset of Fig. 3B. Least square esti-mation was used to determine the Vmax and the Km and straightlines were obtained using a Lineweaver–Burk plot. The Km andVmax values for the ATPase activity and for the GTPase activity were79.25 ± 2.51 nmole/min/mg and 2.33 ± 0.01 �M and 65.37 ± 0.92nmole/min/mg and 2.36 ± 0.02 �M, respectively (Fig. 4B). The bind-ing affinity of the Rep protein for GTP compared that for ATPdecreased only by 1.01-fold, which suggests that, in terms of bind-ing affinity, the ATPase activity of the Rep protein of BFDV is verysimilar to that of the protein’s GTPase activity.

3.4. The nucleotide substrate specificity of the Rep proteins ofBFDV and the effect of polynucleotides on the ATPase and GTPaseactivities of the Rep protein of BFDV

The substrate specificity of the Rep protein of BFDV wasdetermined using various nucleotides. The relative activities ofthe Rep protein when hydrolysing ATP, UTP, GTP, CTP, dATP, dTTP,dGTP, and dCTP were 100 ± 8.05%, 23.59 ± 5.35%, 71.27 ± 4.43%,

26.64 ± 4.43%, 145.28 ± 4.25%, 43.18 ± 3.25%, 60.38 ± 2.13%, and20.76 ± 2.46%, respectively (Fig. 5A). Significant differences wereobserved with respect to all NTPase activities of the Rep protein(p < 0.05). The efficiency of hydrolysis of the ribonucleotide ATP was

156 S.-W. Huang et al. / Virus Resea

Fig. 5. The substrate specificity of the NTPase activities of the Rep protein. The rela-tive activities are presented as the means ± SD from three independent experimentsaod

hwHatbdaahs

nd the activity with ATP was defined as 100% (A). The effect of polynucleotidesn the ATPase activity (B) and GTPase activity (C) of the Rep protein. Significantifferences are indicated by *p < 0.05.

igher than that of GTP, which is similar to the difference observedhen the same deoxyribonucleotides (dATP and dGTP) are used.owever, the efficiency of hydrolysis for different ribonucleotidesnd their deoxynucleotide counterparts varied. This suggestedhat the Rep protein of BFDV should be regarded as preferentiallyeing an ATPase and a GTPase. The effects of single-stranded andouble-stranded deoxyribopolynucleotides on the ATPase activity

nd GTPase activity were tested. The relative levels of ATPasectivities of the Rep protein in the presence of single-strandedomopolymers poly(dA), poly(dT), poly(dG), and poly(dC),ingle-stranded P10 sequence, single-stranded P12 sequence,

rch 213 (2016) 149–161

double-stranded poly(dA-dT), poly (dG-dC), and double-strandedP36 sequence were 100 ± 8.85%, 64.78 ± 8.51%, 87.51 ± 6.84%,77.97 ± 8.05%, 78.83 ± 7.67%, 99.65 ± 7.67%, 98.09 ± 6.38%,75.19 ± 7.37%, 73.28 ± 6.84%, and 68.08 ± 6.49%, respectively(Fig. 5B). A significant difference (p < 0.05) were observed for thesingle-stranded homopolymers poly(dA), homopolymers poly(dT),homopolymers poly(dG), homopolymers poly(dC), the double-stranded poly (dA-dT), the double-stranded poly (dG-dC), and thedouble-stranded P36 sequence. The relative levels of the GTPaseactivities of the Rep protein in the presence of the same substrateswere 99.99 ± 2.03%, 72.27 ± 3.5%, 73.91 ± 2.12%, 47.33 ± 1.96%,65.83 ± 4.32%, 82.22 ± 4.55%, 74.47 ± 1.78%, 77.68 ± 2.83%,41.79 ± 1.58%, and 60.92 ± 3.92%, respectively (Fig. 5C). Thusexcept the double-stranded P10 and P12 sequences, both theATPase and GTPase activities were found to be lower when theenzyme acted on deoxyribopolynucleotides.

3.5. The effect of inhibitors on the ATPase and GTPase activities ofthe Rep protein of BFDV

The relative ATPase activities of the Rep protein alone and afterthe treatment with 0.1 mM, 1 mM and 10 mM NaN3 for 10 min were100 ± 4.42%, 49.65 ± 5.35%, 45.89 ± 6.14%, and 42.5 ± 8.86%, respec-tively (Fig. 6A). The relative levels of the ATPase activities alone andafter the treatment with 0.1 mM, 1 mM, and 10 mM NaN3 for 30 minwere 100 ± 5.63%, 46.18 ± 5.35%, 42.16 ± 9.27%, and 39.52 ± 7.47%,respectively (Fig. 6B). Thus significant differences in the ATPase inthe presence of NaN3 (p < 0.05) were observed at 0.1 mM, 1 mM, and10 mM at both 10 min and 30 min. The relative ATPase activities ofthe Rep protein were 100 ± 7.7%, 73.55 ± 5.63%, 66.93 ± 5.63%, and68.4 ± 3.24% alone and after treatment with 0.01 mM, 0.1 mM, and1 mM NaVO4 for 10 min, respectively (Fig. 6A). The relative ATPaseactivities of the Rep protein alone and treated with the same con-centrations of NaVO4 for 30 min were 100 ± 5.35%, 72.8% ± 5.35%,65.86 ± 5.35%, and 67.01 ± 7.47%, respectively (Fig. 6B). Thus sig-nificant differences (p < 0.05) were seen at 0.01 mM, 0.1 mM, and1 mM for both 10 min and 30 min. The relative ATPase activities ofthe Rep protein alone and treated with 0.1 mM, 1 mM, and 10 mMNEM for 10 min were 100 ± 4.91%, 89.93 ± 6.5%, 88.06 ± 7.47%,and 78.41 ± 8.86%, respectively (Fig. 6A). The ATPase activities ofthe Rep protein alone and after treatment with the same con-centrations of NEM for 30 min were 100 ± 8.05%, 86.55 ± 6.84%,70.74 ± 9.27%, and 59.32 ± 8.05%, respectively (Fig. 6B). Thus sig-nificant differences (p < 0.05) were found at 0.1 mM, 1 mM and10 mM for 10 min treatment and 0.1 mM, 1 mM, and 10 mM for30 min treatment. The relative GTPase activities alone and aftertreatment with 5 nM, 50 nM, and 500 nM dynasore for 10 min were100 ± 6.49%, 88.64 ± 3.25%, 91 ± 4.43%, and 89.35 ± 6.84%, respec-tively (Fig. 6C). The relative levels of GTPase activities of the Repprotein alone and after treatment with 5 nM, 50 nM, and 500 nMdynasore after 30 min treatment were 100 ± 6.14%, 84.65 ± 5.35%,75.02 ± 5.35%, and 64.67% ± 8.05%, respectively (Fig. 6D). Signifi-cant differences were observed at 5 nM, 50 nM, and 500 nM for10 min and 5 nM, 50 nM, and 500 nM for 30 min (all p < 0.05).This suggests that a prolonged incubation can increase the effec-tiveness of dynasore on the Rep protein. The relative GTPaseactivities of the Rep protein alone and when treated with 10 �M,100 �M, and 1000 �M linoleic acid were 100 ± 8.86%, 85.1 ± 6.14%,70.08 ± 6.38%, and 70.59% ± 7.47%, respectively, after 10 min treat-ment (Fig. 6C). The relative GTPase activities of the Rep proteinalone and when treated with 10 �M, 100 �M, and 1000 �M linoleic

acid for 30 min were 100 ± 7.67%, 49.47 ± 5.35%, 29.16 ± 6.84%, and30.11 ± 8.05%, respectively (Fig. 6D). Thus significant differenceswere observed at 10 �M, 100 �M, and 1000 �M after 10 min treat-ment and at 10 �M, 100 �M and 1000 �M after 30 min treatment.

S.-W. Huang et al. / Virus Research 213 (2016) 149–161 157

Fig. 6. The effects of inhibitors on the ATPase activity and GTPase activity of the Rep protein. The Rep proteins were treated with sodium azide (NaN3) (0.1, 1, and 10 mM),s (0.1,

d d 60 me 00%.

Ts

3B

fiwttawa1rr5FtCaGctp

odium orthovanadate (NaVO4) (0.01, 0.1, and 1 mM), or N-ethylmaleimide (NEM)ynasore (5, 50, and 500 nM) or linoleic acid (10, 100, and 1000 �M) for 10 min (C) anxperiments. The ATPase and GTPase activities without inhibitors were defined as 1

hus, the inhibitory effect of linoleic acid was greater than dyna-ore.

.6. Interaction between the Cap protein and the Rep protein ofFDV

The interaction between the native Cap and Rep proteinsrom the tissue of BFDV-infected birds was verified by the co-munoprecipitation. The native Rep protein, which has a molecular

eight of 33 kDa, was detected by Western blot and this confirmedhe interaction between the Cap and Rep proteins (Supplemen-al Fig. 1). Immobilized Cap protein was challenged with differentmounts of the Rep proteins and binding of the Rep proteinas found to increase in a dose-dependent manner. The mean

bsorbance values were 0.15, 0.16, 0.17, 0.27, 0.49, and 0.49 for0 ng, 12.5 ng, 25 ng, 50 ng, 100 ng, and 150 ng of the Rep proteins,espectively. For the negative control, the mean absorbance valuesemained 0.08, 0.08, 0.1, 0.1, 0.12, and 0.14 for 10 ng, 12.5 ng, 25 ng,0 ng, 100 ng, and 150 ng of the Trx proteins, respectively (Fig. 7A).or the GST-pull down assay, the Cap protein was precipitated withhe GST-Rep protein rather than the GST protein. The precipitatedap protein was detected by both anti-His and anti-Cap monoclonalntibodies. The Trx protein was not precipitated by the GST and the

ST-Rep proteins (Fig. 7B). This excluded the possibility of the asso-iation of the Rep and Cap protein via the fusion tags and indicatedhe existence of an in vitro association between the Rep and Caproteins.

1, and 10 mM) for 10 min (A) and 60 min (B). The Rep proteins were treated within (D). The relative activities are shown as the means ± SD from three independent

Significant difference are indicated by *p < 0.05.

3.7. The effect of the Cap protein of BFDV on the ATPase andGTPase activities of the Rep protein of BFDV

The relative ATPase activities of 100 ng Cap protein and 100 ngRep protein were 6.61 ± 1.09% and 98.89 ± 1.04%, respectively.The relative ATPase activities of Rep protein in the presence of10 ng, 25 ng, 50 ng, and 100 ng Cap proteins were 107.85 ± 1.8%,106.74 ± 2.28%, 121.07 ± 1.09%, and 125.48 ± 1.95%, respectively.This increase suggests that, while there is no ATPase activity asso-ciated with the Cap protein and that the ATPase activity of theRep protein is up-regulated by the presence of Cap protein in adose-dependent manner and that significant differences could beobserved in the presence of only 10 ng of Cap protein (all p < 0.05;Fig. 7C).

The relative GTPase activities of 100 ng Cap protein and 100 ngRep protein were 2.42 ± 2.78% and 96.06 ± 3.41%, respectively. Therelative GTPase activities of 100 ng Rep protein in the presence of10 ng, 25 ng, 50 ng, and 100 ng of Cap protein were 109.24 ± 1.31%,115 ± 8.71%, 124.24 ± 2.15%, and 133.79 ± 5.01%, respectively. Thusthe GTPase activity of the Rep protein are also up-regulated by theCap protein in a dose-dependent manner and a significant differ-ence was observed in the presence of only 10 ng of Cap protein (allp < 0.05; Fig. 7D). Interestingly, the up-regulation of GTPase activitywas more profound than that of ATPase activity when the Cap pro-tein is present and this is supported by the relative ATPase activity

levels of the Rep protein compared to GTPase activity levels of theRep protein in the presence of 100 ng Cap protein (125.48 ± 1.95%vs. 133.79 ± 5.01%).

158 S.-W. Huang et al. / Virus Research 213 (2016) 149–161

Fig. 7. The interaction between Rep and Cap proteins. Cap or Trx protein (250 ng) coated on a plate was reacted with different amounts of the Rep protein (25, 50, 100, 150,200, and 250 ng). The absorbance values indicate the binding of the two proteins and are presented as means from three independent experiments (A). Cap or Trx proteinwas individually incubated with immobilized GST or GST-Rep protein. The bound protein was analyzed by Western blotting using anti-His and anti-Cap antibodies. Theamounts of GST or GST-Rep proteins present during the pull-down assay were assessed by Ponceau S staining (B). The effects on ATPase activity and GTPase activity of theRep protein (100 ng) when the Cap protein (10, 25, 50, and 100 ng) is present. The ATPase activity and GTPase activity of the Rep protein alone (100 ng) were defined as1 ented

i

3aa

25rlmlgo1r1(aest

00%. The relative levels of the ATPase activity (C) and GTPase activity (D) are presndicated by *p < 0.05.

.8. The ATPase and GTPase activity of the mutant Rep proteinsnd the effect of the Cap protein on the ATPase and GTPasectivity of the mutant Rep proteins

The relative ATPase activities of Rep, Rep-box-1-del, Rep-box--del, Rep-box-3-del, and Rep-box-1-mut were 105.39 ± 5.56%,3.99 ± 0.85%, 58.09 ± 3.19%, 51.72 ± 2.33%, and 49.97 ± 1.58%,espectively. The Walker A-del and Walker B-del mutants bothead to significant reductions in the ATPase activity. Further-

ore, the ATPase activity of the AYDG-del mutants of Rep proteinost about 42% of the wild-type enzyme’s activity, which sug-ests that magnesium ions are necessary for the ATPase activityf the Rep protein. The relative ATPase activities of Rep-box--del, Rep-box-2-del, Rep-box-3-del, and Rep-box-1-mut wereeversed to 112.69 ± 2.71%, 105.39 ± 8.42%, 105.62 ± 4.45%, and04.58 ± 5.49%, respectively, in the presence of the Cap proteinFig. 8A). The relative ATPase activities of mutant Rep proteinsre able to be reverted to the levels of the Rep protein alone. As

xpected, all relative activities of the mutant Rep proteins wereignificantly lower than that of the Rep protein in the presence ofhe Cap protein.

as the means ± SD from three independent experiments. Significant difference are

Similarly, the relative GTPase activities of the Rep, Rep-box-1-del, Rep-box-2-del, Rep-box-3-del, and Rep-box-1-mut were102.47 ± 2.46%, 60.27 ± 2.52%, 57.34 ± 1.84%, 53.64 ± 1.25%, and55.45 ± 1.27%, respectively (Fig. 8B), which suggests that GTPaseactivity is also affected when magnesium ion cannot bind. Finally,the lysine (K) residue of the Walker A motif (G/AXXXXGK/ST) wasfound to be important to both the ATPase activity and the GTPaseactivity of the BFDV Rep protein. The relative ATPase activitiesof Rep-box-1-del, Rep-box-2-del, Rep-box-3-del, and Rep-box-1-mut were respectively reversed to 119.59 ± 3.14%, 116.92 ± 3.2%,116.97 ± 3.48%, and 120.59 ± 4.6%, respectively, in the presence ofthe Cap protein. The relative GTPase activities of mutant Rep pro-teins can reverse. Noteworthy, the relative GTPase activities ofRep-box-1-del and Rep-box-1-mut returned to a similar level tothat of the Rep protein in the presence of the Cap protein (Fig. 8B).

4. Discussion

In order to obtain a high yield of the target proteins, bet-ter protein purification, good solubility of the expressed proteins,and correct post-translation modification of the expressed pro-

S.-W. Huang et al. / Virus Research 213 (2016) 149–161 159

Fig. 8. The relative levels of the ATPase activity (A) and GTPase activity (B) of the Rep, Rep-box-1-del, Rep-box-2-del, Rep-box-3-del, and Rep-box-1-mut proteins alone andi ree indw red wR resen

tilfNwSwtu(

insnaepadmatitmGitin1

abitHAWtTt2

n the presence of the Cap protein. Results are presented as the means ± SD from there both defined as 100%. Significant differences are indicated by *p < 0.05 compaep-box-2-del, Rep-box-3-del, and Rep-box-1-mut proteins alone vs. those in the p

eins, the proteins of interest were expressed as tagged proteinsn bacteria (Patterson et al., 2013; Sarker et al., 2015). No glycosy-ation site was founded on the Cap protein based on the predictionrom the http://www.cbs.dtu.dk/services/NetNGlyc website. The-terminal thioredoxin tag and a lower induction temperatureere used in this study to preserve adequate biological activity.

uch a system helps to purify the capsid proteins of circovirus,hich is notoriously difficult to purify in soluble forms. In addi-

ion to the Rep protein of BFDV, this tagging system has also beensed to analyse the ATPase activity of the A32L protein of orf virusLin et al., 2010).

If we examine the Rep protein of PCV, this protein’s ATPase activ-ty is dependent on the presence of Mg2+ ions as a co-factor and isot affected by the presence of either single-stranded or double-tranded DNA (Steinfeldt et al., 2007). The hydrolysis of ATP isecessary to allow the replication of PCV because the separationnd unwinding of the origin of replication of DNA would need thenergy. These characteristics are very similar to those of the Reprotein of BFDV. However, an inhibition of the ATPase and GTPasectivity of the Rep protein of BFDV were observed when eitherouble-stranded DNA or single-stranded DNA was present. Severalechanisms allowing such DNA-independent ATPase and GTPase

ctivity have been proposed. Firstly, this may be an intrinsic charac-eristic of the Rep protein. Secondly, because large amounts of Mg2+

ons are sequestered by single-stranded and double-stranded DNA,his will result in there being less cofactor available when these

olecules are present, which in turn will inhibit the ATPase andTPase activity levels. Thirdly, other host proteins are likely to be

mportant when the above activities are involved in binding DNA tohe Rep protein (Bideshi and Federici, 2000). Since the ATPase activ-ty and GTPase activity of the Rep protein of BFDV are inhibited byucleic acids, the enzyme is not likely to be a helicase (Stauber et al.,997).

Similar to geminiviruses, the activation of the ATPase activitynd GTPase activity of the Rep protein of BFDV occurs through theinding of ATP and GTP to their appropriate domains. This bind-

ng serves as a conformation switch and is involved in a signalransduction cascade initiated by heterotrimetric G proteins anda-ras p21-like GTPases (Lin et al., 1992). In such a system, theTPase activity and GTPase activity, which are controlled by thealker A and Walker B motifs, respectively, are able to determine

he immune response of the host and the virulence of the virus.herefore, mutations of the Walker A motif are known to result in

he production of non-infectious progeny of CSFV (Gladue et al.,011).

ependent experiments. The ATPase activity and GTPase activity of the Rep proteinith the Rep protein alone and **p < 0.05 for the activities of the Rep, Rep-box-1-del,ce of the Cap protein.

The GXXXXGK, DXXG, and NKXD domains are known to beassociated with GTP-binding proteins and are usually separatedby 40 amino acids and then by another 40 amino acids. However,the GXXXXGK and GXXG domains of the Rep protein of BFDV areonly separated by 32 amino acids. Similar to �/� tubulin, a GTP-binding protein, it exhibits an atypical spacing rule of less than40 amino acids (Dever et al., 1987). When the A32L protein of orfvirus and the NS4B protein of CSFV were examined, the ATPaseconsensus sequence with mismatches can be used as the commonphosphate-binding sequence for ATP and GTP (Gladue et al., 2011;Lin et al., 2011). However, the dissociation constant (Kd) value ofsuch a sequence within a GTP-binding protein is higher than that ofa ATP-binding protein suggesting that the affinity for ATP is higherthan that of GTP. Interestingly, the Rep protein of BFDV is the onlyone that exhibits similar levels of ATPase and GTPase activity. Itis possible that, unlike larger viruses, the Rep protein of BFDV hasto be multifunctional (Thompson et al., 2009). Since the GXXXXGKsequence belongs to the consensus domains of both ATPase andGTPase, it is reasonable that there is both the ATPase activity andGTPase activity of BFDV Rep protein are reduced in the Rep-box-1-del mutant protein. Similar to the r�A of ARV, the lysine (K) residueof the Walker A motif (GXXXXGK) is involved in the binding of the�-phosphates and �-phosphates of NTPs. If it was mutated intoan alanine (A) residue, both the ATPase and GTPase activities areknown to be abrogated (Su et al., 2007). The GTPase activity of Rep-box-1-del protein is close to that of the Rep protein with the WalkerA motif mutation implying that the hydrophobic residue at the 2ndposition near the glycine residue (G) plays a significant role in theGTPase activity. A negatively charge amino acid, aspartic acid (D),located at the 1st or/and 3rd positions of the DYDG of the Rep pro-tein of BFDV is able to bind magnesium ions. This is similar to for thesituation with the AYDG motif of the A32L protein of orf virus (Linet al., 2011). Since the GTPase activity of the Rep protein with motifI, II, and III deletions are profoundly reduced, all of the identifiedmotifs would seem to participate in ATPase and GTPase activity.

The optimal pH values for the ATPase and GTPase activities ofBFDV are both 7.5, which implies that the two activities are simul-taneously active at the same pH value. A significant inhibition ofthe ATPase activity of BFDV was observed at 80 ◦C suggesting theRep protein is a thermo-sensitive protein. The optimal activity ofthe ATPase of the Rep protein and of the r�A protein of ARV occurin the presence of Mg2+ ions and the presence of Zn2+ ions sig-nificantly inhibits both the ATPase and GTPase activities (Su et al.,

2007). While the ATPase activity of the r�A protein of ARV is notaffected by Mn2+ ions, the ATPase and GTPase activity of the Repprotein of BFDV are inhibited by Mn2+ ions suggesting that the

1 Resea

eno2RVV(Gewc(A1h5em

aTptciBmprabbmttCRtpt

uRgocibLc1tait

C

A

m2

60 S.-W. Huang et al. / Virus

ffect of different ions on the ATPase activity is determined by theature of protein. The Vmax and Km values of the ATPase activityf the Rep protein of BFDV are 79.25 ± 2.51 nmole/min/mg and.33 �M, respectively. The Vmax value of the ATPase activity of theep protein of BFDV is close to that of the ATPase activity of theP4 protein of BTV but has a higher Km value than that of theP4 protein of BTV; these are lower than those found in bacteria

Ramadevi and Roy, 1998; Sarin et al., 2001). The binding affinity ofTP is 1.17-fold higher than that of ATP. No difference in hydrolysisfficiency between purine nucleotides and pyrimidine nucleotidesas found. Similar to the NS4B protein of HCV, the hydrolysis effi-

iency of the Rep protein of BFDV for dATP is higher than for ATPThompson et al., 2009). Numerous viral proteins have significantTPase and GTPase activity levels (Bisallion et al., 1997; Kuo et al.,996; Ramadevi and Roy, 1998). For BTV, viral NTPase is able toydrolyze NTPs to provide energy and to cleave GTP into GMP for′-end capping of RNA (Ramadevi and Roy, 1998). Other than as annergy supply, the hydrolysis of dNTP by the Rep protein of BFDVay be associated with the synthesis of viral DNA.The Rep protein of BFDV exhibiting both ATPase and GTPase

ctivities aking it similar to the Ras GTPase (Thompson et al., 2009).his protein can trigger a phosphoinositol 3-kinase (PI3K)-Aktathway leading to an increase in intracellular calcium concentra-ion. This is related to viral infection and results in an increase inalcium levels. Similar to the NS3 protein of the JE virus, there is annhibition of the ATPase and GTPase activity of the Rep protein ofFDV (Kuo et al., 1996). The GTPase activity of the Rep protein isuch more sensitive to the inhibitory effect of calcium than the NS3

rotein. The GTPase of the Rep protein of BFDV may serve as a sur-ogate enzyme with respect to GTP hydrolysis that does not require

calcium-dependent phosphorylated intermediate. An interactionetween the Rep and Cap proteins of BFDV was already identifiedy co-immunoprecipitation and co-localization under the confocalicroscopy suggests that the Cap protein of BFDV is able to facili-

ate the nuclear movement of the Rep protein (Heath et al., 2006);his is consistent with the in vitro interaction between the Rep andap proteins of BFDV. Both the ATPase and GTPase activity of theep protein of BFDV are up-regulated in the presence of Cap pro-ein, which suggests that the interaction between the Rep and Caproteins may lead to conformational and/or functional changes inhe Rep protein.

ATPases are usually classified into F-type, P-type, or V-typesing their sensitivity to various inhibitors (Sarin et al., 2001). Theep protein of BFDV is most sensitive to sodium azide (F-type) sug-esting that it is closest to being F-type ATPase. Although sodiumrthovanadate (P-type) can usually bring about conformationalhanges in an ATPase, resulting in a reduced affinity for calciumons, this inhibitor had no effect on the BFDV Rep protein, proba-ly because the enzyme is calcium-independent (Tate et al., 1985).inoleic acid can modulate the G protein activity by altering theonformation of the cellular Na+ channel (Mukhopadhyay et al.,997). The GTPase activity of the Rep protein of BFDV is more sensi-ive to inhibition by linoleic acid than dynasore. Dynasore is able toct as a non-competitive GTPase inhibitor with respect to the bind-ng of dynamins to lipid vesicles (Macia et al., 2006). This suggestshat the replication of circovirus may involve lipid vesicles.

onflict of interest

None

cknowledgments

The authors thank Dr. Ralph Kirby for critical review of theanuscript. Funding: This work was supported by the grant 102-

313-B-005-011 from the National Science Council, Taiwan.

rch 213 (2016) 149–161

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.virusres.2015.12.001.

References

Bassami, M.R., Berryman, D., Wilcox, G.E., Raidal, S.R., 1998. Psittacine beak andfeather disease virus nucleotide sequence analysis and its relationship toporcine circovirus, plant circoviruses, and chicken anaemia virus. Virology 249(2), 453–459.

Bert, E., Tomassone, L., Peccati, C., Navarrete, M.G., Sola, S.C., 2005. Detection ofbeak and feather virus (BFDV) and avian polyomavirus (APV) DNA in psittacinebirds in Italy. J. Vet. Med. B Infect. Dis. Vet. Public Health 52 (2), 64–68.

Bideshi, D.K., Federici, B.A., 2000. DNA-independent ATPase activity of theTrichoplusia ni granulovirus DNA helicase. J. Gen. Virol. 81 (6), 1601–1604.

Bisallion, M., Bergeron, J., Lemay, G., 1997. Characterization of the nucleotidetriphosphate phosphohydrolase and helicase activities of the reovirus �1protein. J. Biol. Chem. 272 (29), 18298–18303.

Desbiez, C., David, C., Mettouchi, A., Laufs, J., Gronenborn, B., 1995. Rep protein oftomato yellow leaf curl geminivirus has an ATPase activity required for viralDNA replication. Proc. Natl. Acad. Sci. U. S. A. 92 (12), 5640–5644.

Dever, T.E., Glynias, M.J., Merrick, W.C., 1987. GTP-binding domain: threeconsensus elements with distinct spacing. Proc. Natl. Acad. Sci. U. S. A. 84 (7),1814–1818.

Einav, S., Elazar, M., Danieli, T., Glenn, J.S., 2004. A nucleotide binding motif inhepatitis C virus (HCV) NS4B mediates HCV RNA replication. J. Virol. 78 (20),11288–11295.

Gladue, D.P., Gavrilov, B.K., Holinka, L.G., Fernandez-Sainz, I.J., Vepkhvadze, N.G.,Rogers, K., O’Donnell, V., Risatti, G.R., Borca, M.V., 2011. Identification of anNTPase motif in classical swine fever virus NS4B protein. Virology 411 (1),41–49.

Heath, L., Williamson, A., Rybicki, E.P., 2006. The capsid protein of beak and featherdisease virus binds to the viral DNA and is responsible for transporting thereplication-associated protein into the nucleus. J. Virol. 80 (14), 7219–7225.

Heckman, K.L., Pease, L.R., 2007. Gene splicing and mutagenesis by PCR-drivenoverlap extension. Nat. Protoc. 2 (4), 924–932.

Hsu, C.M., Ko, C.Y., Tsai, H.J., 2006. Detection and sequence analysis of avianpolyomavirus and psittacine beak and feather disease virus from psittacinebirds in Taiwan. Avian Dis. 50 (3), 348–353.

Kuo, M.D., Chin, C., Hsu, S.L., Shiao, J.Y., Wang, T.M., Lin, J.H., 1996. Characterizationof the NTPase activity of Japanese encephalitis virus NS3 protein. J. Gen. Virol.77 (9), 2077–2084.

Lin, F.Y., Chan, K.W., Wang, H.C., Hsu, W.L., Wong, M.L., 2010. Functional expressionof the recombinant ATPase of orf virus. Arch. Virol. 155 (10), 1701–1705.

Lin, F.Y., Chan, K.W., Wang, C.Y., Wong, M.L., Hsu, W.L., 2011. Purification andfunctional motifs of the recombinant ATPase of orf virus. Protein Expr. Purif. 79(2), 210–216.

Lin, J.J., Phillips, A.M., Hearst, J.E., Sancar, A., 1992. Active site of (A)BC excinucleaseII. Binding, bending, and catalysis mutants of UvrB reveal a direct role in 3′ andan indirect role in 5′ incision. J. Biol. Chem. 267 (25), 17693–17700.

Macia, E., Ehrlich, M., Massol, R., Boucrot, E., Brunner, C., Kirchhausen, T., 2006.Dynasore, a cell-permeable inhibitor of dynamin. Dev. Cell 10 (6), 839–850.

Mukhopadhyay, S., Ramminger, S.J., McLaughlin, M., Gambling, L., Olver, R.E.,Kemp, P.J., 1997. Direct modulation of G-proteins by polyunsaturated fattyacids: a novel eicosanoid-independent regulatory mechanism in thedeveloping lung. Biochem. J. 326 (3), 725–730.

Patterson, E.I., Swarbrick, C.M., Roman, N., Forwood, J.K., Raidal, S.R., 2013.Differential expression of two isolates of beak and feather disease virus capsidprotein in Escherichia. coli. J. Virol. Methods 189 (1), 118–124.

Pfister, T., Wimmer, E., 1999. Characterization of the nucleoside triphosphataseactivity of poliovirus protein 2C reveals a mechanism by which guanidineinhibits poliovirus replication. J. Biol. Chem. 274 (11), 6992–7001.

Ramadevi, N., Roy, P., 1998. Bluetongue virus core protein VP4 has nucleosidetriphosphate phosphohydrolase activity. J. Gen. Virol. 79 (10), 2475–2480.

Rahaus, M., Wolff, M.H., 2003. Psittacine beak and feather disease: a first survey ofthe distribution of beak and feather disease virus inside the population ofcaptive psittacine birds in Germany. J. Vet. Med. B Infect. Dis. Vet. Public Health50 (8), 368–371.

Raue, R., Johne, R., Crosta, L., Burkle, M., Gerlach, H., Muller, H., 2004. Nucleotidesequence analysis of a C1 gene fragment of psittacine beak and feather diseasevirus amplified by real-time polymerase chain reaction indicates a possibleexistence of genotypes. Avian Pathol. 33 (1), 41–50.

Sarin, J., Aggarwal, S., Chaba, R., Varshney, G.C., Chakraborti, P.K., 2001. B-subunitof phosphate-specific transporter from Mycobacterium tuberculosis is athermostable ATPase. J. Biol. Chem. 276 (48), 44590–44597.

Sarker, S., Ghorashi, S.A., Swarbrick, C.M., Khandokar, Y.B., Himiari, Z., Forwood,

J.K., Raidal, S.R., 2015. An efficient approach for recombinant expression andpurification of the viral capsid protein from beak and feather disease virus(BFDV) in Escherichia coli. J. Virol. Methods 215–216, 1–8.

Stauber, N., Costas, J.M., Sutton, G., Monastyrskaya, K., Roy, P., 1997. Bluetonguevirus VP6 protein binds ATP and exhibits an RNA-dependent ATPase function

Resea

S

S

T

T

other ATP-requiring enzymes and a common nucleotide binding fold. EMBO J.1 (8), 945–951.

Warrener, P., Tamura, J.K., Collett, M.S., 1993. RNA-stimulated NTPase activityassociated with yellow fever virus NS3 protein expressed in bacteria. J. Virol.67 (2), 989–996.

S.-W. Huang et al. / Virus

and a helicase activity that catalyze the unwinding of double-stranded ofdouble-stranded RNA substrates. J. Virol. 71 (10), 7220–7226.

teinfeldt, T., Finsterbusch, T., Mankertz, A., 2007. Functional analysis of cis- andtrans-acting replication factors of porcine circovirus type 1. J. Virol. 81 (11),5696–5704.

u, Y.P., Shien, J.H., Liu, H.J., Yin, H.S., Lee, L.H., 2007. Avian reovirus core protein �Aexpressed in E. coli possesses both NTPase and RTPase activities. J. Gen. Virol.88 (6), 1797–1805.

ate, C.A., Bick, R.J., Blaylock, S.L., Youker, K.A., Scherer, N.M., Entman, M.L., 1989.Nucleotide specificity of canine cardiac sarcoplasmic reticulum. Differentialalteration of enzyme properties by detergent treatment. J. Biol. Chem. 264 (14),

7809–7813.

hompson, A.A., Zou, A., Yan, J., Duggal, R., Hao, W., Molina, D., Cronin, C.N., Wells,P.A., 2009. Biochemical characterization of recombinant hepatitis C virusnonstructural protein 4B: evidence for ATP/GTP hydrolysis and adenylatekinase activity. Biochemistry 48 (5), 906–916.

rch 213 (2016) 149–161 161

Varsani, A., Regnard, G.L., Bragg, R., Hitzeroth, I.I., Rybicki, E.P., 2011. Global geneticdiversity and geographical and host-species distribution of beak and featherdisease virus isolates. J. Gen. Virol. 92 (4), 752–767.

Walker, J.E., Saraste, M., Runswick, M.J., Gay, N.J., 1982. Distantly related sequencesin the alpha-subunits and beta-subunits of ATP synthase, myosin, kinases and