Identification of a Novel Nonstructural Protein ... - Journal of Virology

JOURNAL OF VIROLOGY, Nov. 2006, p. 10419–10427 0022-538X/06/$08.00⫹0 doi:10.1128/JVI.00698-06 Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Vol. 80, No. 21

Identification of a Novel Nonstructural Protein, VP9, from White Spot Syndrome Virus: Its Structure Reveals a Ferredoxin Fold with Specific Metal Binding Sites䌤 Yang Liu,1 Jinlu Wu,1 Jianxing Song,1,2 J. Sivaraman,1 and Choy L. Hew1* Department of Biological Sciences1 and Department of Biochemistry,2 National University of Singapore, Singapore Received 6 April 2006/Accepted 14 August 2006

White spot syndrome virus (WSSV) is a major pathogen in shrimp aquaculture. VP9, a full-length protein of WSSV, encoded by open reading frame wsv230, was identified for the first time in the infected Penaeus monodon shrimp tissues, gill, and stomach as a novel, nonstructural protein by Western blotting, mass spectrometry, and immunoelectron microscopy. Real-time reverse transcription-PCR demonstrated that the transcription of VP9 started from the early to the late stage of WSSV infection as a major mRNA species. The structure of full-length VP9 was determined by both X-ray and nuclear magnetic resonance (NMR) techniques. It is the first structure to be reported for WSSV proteins. The crystal structure of VP9 revealed a ferredoxin fold with divalent metal ion binding sites. Cadmium sulfate was found to be essential for crystallization. The Cd2ⴙ ions were bound between the monomer interfaces of the homodimer. Various divalent metal ions have been titrated against VP9, and their interactions were analyzed using NMR spectroscopy. The titration data indicated that VP9 binds with both Zn2ⴙ and Cd2ⴙ. VP9 adopts a similar fold as the DNA binding domain of the papillomavirus E2 protein. Based on our present investigations, we hypothesize that VP9 might be involved in the transcriptional regulation of WSSV, a function similar to that of the E2 protein during papillomavirus infection of the host cells. Shrimp aquaculture has become an important industry worldwide during the last few decades. Intensive cultivation and massive worldwide trade of the shrimp and other aquaculture industries have led to the emergence and spread of numerous viral pathogens. Out of these viral pathogens, shrimp white spot syndrome virus (WSSV) is the most devastating to crustacean species. To date, there is no effective treatment available for WSSV infection. WSSV belongs to a new virus family, Nimaviridae, under a new genus, Whispovirus, which shares low sequence homology with other DNA viruses (16, 24). WSSV is an enveloped virus with a 305-kb double-stranded circular DNA genome. Its genome was completely sequenced, with 180 predicted open reading frames (ORFs) (27). Most of these WSSV genes bear poor sequence homology with other known proteins, and thus the function of these genes cannot be predicted. However, a large amount of work has been carried out on the identification and characterization of WSSV structural proteins, including envelope and other capsid proteins. To date, a total of 39 structural proteins are known from WSSV (12, 23). A structure-based functional study is a promising approach to elucidate the function of such WSSV proteins. Besides the envelope and capsid proteins, nonstructural proteins are required for replication of the viral genome, production of the virus particle, and inhibition of the host cell functions. These proteins are therefore potential candidates for drug design and the development of vaccines.

* Corresponding author. Mailing address: Department of Biological Sciences, National University of Singapore, 14 Science Drive 4, Singapore, 117543. Phone: 65-68742692. Fax: 65-67795671. E-mail: dbshewcl @nus.edu.sg. 䌤 Published ahead of print on 6 September 2006.

Here we report, for the first time, the identification of a novel, nonstructural WSSV protein, VP9. We have demonstrated its abundance both at mRNA and protein levels. In addition, the presence of VP9 in WSSV-infected tissues detected by Western blotting and mass spectrometry was in agreement with immunoelectron microscopy results. Furthermore, in order to shed light on its presumptive function, both the X-ray and nuclear magnetic resonance (NMR) structures of VP9 were determined. In addition, we have studied the metal binding properties of VP9, using NMR titration and X-ray diffraction of Cd2⫹-bound crystals of VP9. Based on these studies, we propose that VP9 might recognize DNA in a manner similar to its structural homolog, the papillomavirus-1 E2 protein, and thus possibly acts as a transcriptional regulator of WSSV. MATERIALS AND METHODS Real-time quantitative RT-PCR and statistical analysis. Penaeus monodon shrimps were sampled at different time points after viral infection, and total RNA was extracted from the shrimp gills by TRIzol extraction. Five micrograms of total RNA isolates was transcribed into cDNA using a SuperScript III First Strand Synthesis System (Invitrogen). WSSV delayed early gene dnapol (4) and late gene vp28 (25) were used as controls, and shrimp ␤-actin was used as an internal control. Real-time reverse transcription-PCR (RT-PCR) conditions were optimized by adjusting the annealing temperature. PCR was performed using a Light Cycler (Roche). For each specimen, reactions were performed in duplicate. Subsequently, a Tukey’s multiple comparison post test, following oneway analysis of variance at a 95% confidence level, was performed using Prism, version 4.00, for Windows (GraphPad Software). Cloning, expression, purification, and production of polyclonal antibody. Using specific primers 5⬘ CGCGCGCATATGGCCACCTTCCAGACTGAC 3⬘ and 3⬘ CGCGGATCCTTATTCTGTTGTTGGCAC 5⬘, gene vp9 was PCR amplified and ligated with pET15b through NdeI/BamHI restriction enzyme sites. The plasmid was transformed into Escherichia coli BL-21 cells (DE3; Novagen). The overexpressed VP9 protein was purified by a nickel-nitrilotriacetic acid column (QIAGEN) followed by on-column thrombin cleavage (Sigma). His

10419

10420

LIU ET AL.

J. VIROL.

FIG. 1. Transcriptional analysis of WSSV by real-time RT-PCR. In this study, relative quantification was calculated using ⌬CT, where ⌬CT refers the difference between the cycle thresholds (CTs) of the target genes and the housekeeping gene ␤-actin. ⌬CT for vp9 (wsv230) is colored in blue, dnapol (wsv514) in black, and vp28 (wsv421) in red.

tag-removed VP9 was further purified by ion exchange chromatography (Amersham) followed by gel filtration chromatography using a Superdex 30 column (Amersham). Fractions containing the native VP9 were pooled and concentrated to 20 mg ml⫺1 using Amicon (Millipore). In addition, 15N- and 15N/13C-labeled VP9 were prepared in M9 medium with the addition of (15NH4)2SO4 for 15N labeling and (15NH4)2SO4 and [13C]glucose for 15N/13C double labeling, respectively. All samples used for NMR studies were dissolved in a buffer containing 20 mM sodium phosphate, pH 6.8, and Sigma cocktail proteinase inhibitor. In addition, anti-VP9 polyclonal antibodies were raised from the New Zealand White female rabbit, and the serum was purified by agarose A (Roche). Western blot and immunoelectron microscopy analyses. Virus inocula were prepared (26) and injected intramuscularly into healthy crayfish (Cherax quadricarinatus). After 4 to 5 days, WSSV virions were purified by sucrose gradient centrifugation as described previously (11). Purified WSSV was further separated into envelope and nucleocapsid fractions (28). Gills and stomachs from moribund crayfish were homogenized separately in a lysis buffer (1% NP-40; 150 mM NaCl, 50 mM Tris, pH 8.0, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride). Ten micrograms of each sample was loaded onto a 10 to 20% gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis ready gel (Bio-Rad) followed by Western blot analysis. For Western blotting, three antibodies were used, including anti-␤-actin, anti-VP28, and anti-VP9. In addition, immunoelectron microscopy was carried to determine the location of VP9 on virions (28); for immunoelectron microscopy, three antibodies used included anti-VP28, antiVP664, and anti-VP9. To reconfirm the authenticity of VP9, total tissue lysates were separated by performing a 10 to 20% gradient gel in parallel with the gel for Western blotting. A protein band at approximately 9 kDa was found and analyzed with in-gel tryptic digestion matrix-assisted laser desorption ionization– time of flight mass spectrometry analysis as described previously (21). NMR experiments and NMR structure calculation. 1H, 15N, and 13C NMR data were acquired at 298 K on an 800 MHz Avance spectrometer (Bruker) or on a 500 MHz Avance spectrometer (Bruker) equipped with both an actively shielded cryoprobe and pulse field gradient units. The NMR experiments included 15N-edited heteronuclear single quantum coherence spectroscopy (HSQC), HNCACB, and CBCA (CO)NH for backbone assignments and threedimensional total correlation spectroscopy (3D-TOCSY), three-dimensional nuclear overhauser enhancement spectroscopy (3D-NOESY), and HCCH-TOCSY for side chain assignments. All NMR data were processed with NMRPipe (7) and analyzed with NMRView (13). Structure calculation was performed with the program CYANA (9). Distance constraints were obtained from 1H/1H NOEs derived from 15N-NOESY and 13C-NOESY spectra. Hydrogen bond restraints were derived from HSQC-based hydrogen-deuterium exchange experiments. The phi and psi angle constraints were generated from the TALOS program (5). Ten conformers with the lowest final values of the DYANA target function were chosen to represent the most probable structures. NMR titration experiments. To study the interaction between VP9 and various divalent metals, two-dimensional 1H-15N HSQC spectra of the 15N-labeled VP9 were measured at a concentration of 0.1 mM in the absence and presence of divalent metals, such as Zn2⫹, Cd2⫹, Ca2⫹, and Mg2⫹. The final ratio of the VP9 to metal was 1:4. The perturbed residues were assigned by superimposing the two 15 N-HSQC spectra in the absence and the presence of different metals. Crystallization and data collection. Crystals of VP9 were obtained by using the hanging drop vapor diffusion method. Initial crystallization conditions were es-

tablished by Hampton Research screens (HR screen II) and were further optimized. Best crystals were obtained with a reservoir solution consisting of 2 M sodium acetate, 100 mM morpholinepropanesulfonic acid, pH 6.3, 25 mM cadmium sulfate, and 3% glycerol. Crystals grew to approximate sizes of ⬃0.2 by 0.2 by 0.1 mm3 over 7 days. Prior to mounting, crystals were briefly soaked for ⬃20 s in a cryoprotectant solution consisting of a mixture of 50:50 mineral oil and paraffin oil, picked up in a nylon loop, and flash cooled at 100 K in a nitrogen gas cold stream (Oxford Cryosystems, Oxford, United Kingdom). Synchrotron data were collected at X12C beam-line, National Synchrotron Light Source, Brookhaven National Laboratory. Our aim was to collect a complete sulfur single wavelength anomalous diffraction (SAD) data set. A total of 1,080 frames (360° three times) were collected at 1.7-Å wavelength with an oscillation of 1.0° using a charge-coupled device detector to 2.3-Å resolution. Diffraction data were processed using the program HKL2000 (17). Crystals belonged to the space group P212121, with a ⫽ 73.23 Å, b ⫽ 76.97 Å, c ⫽ 78.24 Å, with four molecules forming an asymmetric unit. Subsequently, a high-resolution data set was collected up to 2.2 Å for phase extension and refinement. Structure solution and refinement. For phase determination, the resolution range of 2.6 to 20.0 Å was chosen. During phasing trials, a strong anomalous contribution from Cd2⫹ was identified with appropriate f⬘ and f⬙. Assignment as Cd2⫹ was consistent with the high concentration of CdSO4 that was essential for crystallization. This interpretation explains why SAD phasing around the S absorption edge was unsuccessful. Initial phase calculations were carried out with SOLVE (22). Subsequent heavy-atom refinement and density modification was carried out using SHARP (6). The resulting phase gave an overall figure of merit of 0.69. The starting electron density map was further improved by phase extension up to 2.35 Å using the program wARP, which was able to trace the mainchain atoms up to ⬃38% of the model. The remaining parts of the model were built manually using the program O (14). Further cycles of model building alternating with refinement using the program CNS (2) resulted in a final model with an R factor of 0.225- (Rfree ⫽ 0.275) to 2.35-Å resolution with no ␴ cutoff used during refinement. Noncrystallographic symmetry (NCS) restraints were used only during the initial stage of refinement. The VP9 model comprises 316 residues, 8 Cd2⫹ ions, and 125 water molecules in the asymmetric unit. The first well-ordered residue was Ala2. The last two residues and the first three residues of the linker region could not be traced in the electron density map and were not modeled. Validation of the model was done using PROCHECK (15). PDB coordinates. The NMR structures and X-ray structure Protein Data Bank (PDB) coordinates of VP9 were deposited in the PDB under numbers 2GJI and 2GJ2, respectively.

RESULTS AND DISCUSSION Transcriptional analysis. The transcriptional analysis of VP9 was carried out with real-time quantitative RT-PCR using RNAs from shrimp tissues before infection (0 h) and at 2, 4, 6, 10, 12, 24, and 72 h after WSSV challenge (Fig. 1). The vp9 transcript was detected from 2 h postinfection (hpi), while dnapol and vp28 (WSSV late gene) was detected only from 10 hpi. These results indicated that vp9 is an early gene which

VOL. 80, 2006

Nsp VP9 FROM WSSV AND ITS STRUCTURE DETERMINATION

FIG. 2. Localization analysis of VP9 by Western blotting. (A) WSSV virions; (B) noninfected stomach; (C) noninfected gill; (D) WSSV-infected stomach; (E) WSSV-infected gills; (F) envelope fraction; (G) nucleocapsid fraction. Also shown are localization analyses by immunoelectron microscopy using (H) anti-VP28 (envelop protein) as a positive control, (I) anti-VP664 (nucleocapsid) as a positive control, and (J) anti-VP9, respectively.

10421

might code for a nonstructural protein. The level of vp9 transcript was found to be nearly identical to dnapol or vp28 after 10 hpi up to 72 hpi. Statistical analysis showed no significant difference in the transcript levels in comparisons of vp9 to dnapol and vp28 (except at 24 hpi), respectively. However, at 24 hpi, there was significant increase of vp28 relative to vp9 (P ⬎ 0.05). This might be due to the large amount of viral particles being assembled at this point of time postinfection. The transcriptional profile from real time RT-PCR agreed well with Western blot, mass spectrometric (MS), and immunoelectron microscopy results. This experimental evidence strongly supports that VP9 is a full-length nonstructural WSSV protein with a high abundance both at mRNA and protein levels. VP9 as a nonstructural protein. Western blot analysis was used to establish VP9 as a nonstructural protein. This study revealed that the anti-VP9 antibody recognized a protein band of 9 kDa not from WSSV virion or noninfected tissues (Fig. 2A, B, C) but only from the WSSV-infected stomach (Fig. 2D) and gills (Fig. 2E). Further, WSSV virion was separated into envelop and nucleocapsid fractions; however, anti-VP9 failed to recognize VP9 from both fractions (Fig. 2F and G). This band was analyzed by mass spectrometry, and its identity was confirmed as VP9 by N-terminal sequencing. The observation that VP9 was detected only from the WSSV-infected tissues but not from purified virions was consistent with our immunogold labeling experiment results, which showed no gold labeling signal by transmission electron microscopy (Fig. 2H, I, and J). As all protein samples were equally loaded and the size of VP9 (9 kDa) is nearly one-third that of VP28 (28 kDa), it was suggested that the expression level of VP9 protein relative to

FIG. 3. Crystal structure of VP9. Two dimers of one asymmetry unit with four cadmium ions (blue spheres) are shown. VP9 is shown as ribbons with the ␣-helix in red, the ␤-sheet in cyan, and the random coil and loop in gray.

10422

LIU ET AL.

J. VIROL. TABLE 1. Data collection and refinement statistics

Data set

Refinement

Phasing

Cell parameters Space group

a ⫽ 74.13 Å; b ⫽ 78.21 Å; c ⫽ 78.98 Å P212121

a ⫽ 73.23 Å; b ⫽ 76.97 Å; c ⫽ 78.24 Å P212121

Data collection Resolution range Wavelength Observed hkl Unique hkl Completeness Overall (I/␴I) Rsym a

2.2–50.0 Å 1.1 Å 249473 43491 99.1% (2.30–2.2; 95.4%) 15.6 6.3%

2.3–50.0 Å 1.7 Å 214496 38474 97.4% (2.35–2.3; 78.4%) 17.5 5.7%

Refinement Resolution range 关I⬎␴(I)兴 Rworkb (no. of reflections) Rfreec (no. of reflections) rmsd bond lengths rmsd bond angles

2.35–45.0 Å 0.225 (34,301) 0.275 (1,821) 0.012 Å 1.7°

Avg B factors Main chain atoms Side chains atoms

45.349 Å2 50.347 Å2

Ramachandran plot Most favored regions Additional allowed regions Generously allowed regions Disallowed regions

93.2% 6.5% 0.4% 0.0%

Rsym ⫽ ⌺/Ii ⫺ 具I典 ⌺/Ii, where Ii is the intensity of the ith measurement and 具I典 is the mean intensity for that reflection. Rwork ⫽ ⌺/Ii Fobs ⫺ Fcalc /⌺ Fobs , where Fcalc and Fobs are the calculated and observed structure factor amplitudes, respectively. c Rfree was determined with the same equation as that for Rwork, but 6% of the total reflections were chosen at random and omitted from refinement. a b

that of VP28 (one of the most abundant viral structural proteins) was nearly 1 to 1. These results confirmed that VP9 is a full-length nonstructural WSSV protein with a high abundance at protein level. Structural studies. The three-dimensional structures of VP9 were independently determined both by X-ray crystallography and NMR spectroscopy. The root mean square deviation (rmsd) between X-ray and NMR models was ⬃1.194 Å for 74 C␣ atoms (Phe4 to Thr76), which indicated a good agreement between the two structures. It is worth mentioning that the NMR structure was determined in conditions free of metal ions, whereas the X-ray structure was determined with Cd2⫹ ions. The observed metal ion was located on the monomer interface of the dimer and also on the surface of the molecule. This suggested that the metal ion was not essential for the folding of VP9. The structure of VP9 revealed a ferredoxin fold, a well-known nucleotide recognition fold. The following structural description is based on the crystal structure. X-ray. The crystal structure of VP9 from WSSV was determined by the SAD method from synchrotron data and refined to a final R factor of 0.225 (Rfree ⫽ 0.275) at 2.35-Å resolution (Fig. 3). The VP9 model was refined with good stereochemical parameters (Table 1). The asymmetric unit consists of four molecules comprising 79 residues each from Ala2 to Thr80 and a total of 125 water molecules. The monomer of VP9 molecule adopts a mixed ␣/␤ fold with overall dimensions of 32 Å by 25 Å by 20 Å. A total of six ␤-strands and two ␣-helices are found per molecule. The anti-parallel ␤-strands ␤31␤42␤21␤62 assemble into a single ␤-sheet (␤-sheet I), which forms one

face of the molecule. The ␤-sheet II is rather small, consisting of only two ␤-strands, ␤11␤52. The two ␣-helices ␣1 and ␣2 as well as ␤-sheet II are packed on the same side of ␤-sheet I. In the crystal, two dimers of one asymmetric unit are related by a twofold noncrystallographic symmetry (NCS) approximately parallel to the b axis. Of the eight Cd2⫹ ions of an asymmetric unit, five of them formed an almost perfect tetrahedral coordination sphere, whereas the remaining three had five coordinating atoms. In every case at least one water molecule was involved in the coordination. In the coordination sphere, the distance between the Cd2⫹ and the coordinating oxygen or sulfur atoms ranged from 1.9 to 2.3 Å. Cd2⫹ ion coordination was found at the monomer interface of the dimer between the side chains of Asp9, Cys46 of one monomer, and Glu31 of the second monomer of the dimer. These intermonomer coordination bonds maintained the rigid architecture of the VP9 dimer. Three Cd2⫹ ions were found to be in the dimerization region of the VP9 dimer. In addition, hydrogen bonding and extensive hydrophobic interactions participated in the stabilization of the dimer. There are five intermonomer hydrogen bonding contacts (⬍3.2 Å). Interactions of the monomers within the dimer mainly comprised ␤-sheet I, ␣2helix, and its connecting loop from each monomer. NMR. To assess whether the dimerization observed in crystal structure is relevant in solution, we have conducted extensive characterization by gel filtration, dynamic light scattering (DLS), analytic ultracentrifuge (AUC), and NMR relaxation measurements. The gel filtration and DLS results indicated that the apparent molecular mass was ⬃16 kDa, while analytic

VOL. 80, 2006


FIG. 4. NMR structure of VP9. Shown are ten lowest energy structures of VP9 superimposed as ribbons with ␣-helix in red, ␤-sheet in cyan, and random coil and loop in gray.

10423

may represent a critical mechanism to facilitate the conformational switch of VP9 required for the proposed DNA binding. Sequence and structural homology. Blast searches revealed that VP9 has sequence homologies only with a few other WSSV proteins of unknown function. It exhibits a maximum identity of 43% with wsv234 and a minimum of 31% with wsv231 from WSSV. Searches for structurally similar proteins within the PDB were performed with the program DALI (10). Significant structural similarities were found with several nucleotide binding and metal transport/binding proteins. The highest structural similarity was observed between VP9 and bovine papillomavirus-1 E2 DNA-binding protein (PDB code 2BOP), yielding an rmsd of 3.0 Å for 63 C␣ atoms, with 17% sequence identity. This was followed by a metal transporting ATPase (PDB code, 1MWY; rmsd of 2.4 Å for 59 C␣ atoms, with 14.6% identity) and Atx1 metallochaperone (PDB code 1CC8; rmsd of 2.6 Å for 62 C␣ atoms; 13.4% identity). These structurally homologous proteins were superimposed with O program, and their conformational similarities and differences were examined. Except for ␣2 of VP9 and other small local structural differences, all the secondary structural elements were comparable and superimposable. ␣2 of VP9 was in a completely different disposition. The observed structural differences could be partly responsible for its functional specificity in WSSV.

TABLE 2. NMR structural statistics d

ultracentrifuge and NMR relaxation studies estimated the apparent molecular mass to be ⬃11 to 12 kDa. Moreover, only one set of HSQC peaks was observed for VP9 residues. These results imply that in solution, VP9 protein exists in fast exchange equilibrium between monomer and symmetric dimer. Since in solution VP9 undergoes a fast exchange between monomer and symmetric dimmer, the solution structure of VP9 was also determined by heteronuclear, multidimensional NMR spectroscopy. Ten conformers with the lowest target function values were chosen to represent the most probable structures from 100 randomized starting models (Fig. 4). Assignment of backbone and side chain resonances was accomplished by a combination of double- and triple-resonance experiments. Briefly, backbone assignments were complete except for two proline residues. More than 92% of side chains were assigned. The final NMR structure of the full-length VP9 was calculated and refined with the CYANA program. This calculation was based on 1,572 manual and autoassigned distance restraints (548 intraresidue, 349 sequential, 212 medium range, and 463 long range) and 74 backbone dihedral-angle restraints (Table 2). The Ramachandran map indicated that the majority of the residues (84.9%) had angular averages in energetically most favorable regions, 15.1% in additional allowed regions, and none in the generously allowed or disallowed regions. Based on our results, the dimeric X-ray structure would be biologically relevant, although the crystal packing force may shift the monomer-dimer equilibrium, thus favoring the formation of the observed dimeric structure. We propose that in fact the fast exchange between monomer and dimer in solution

Restraints used in structure calculation

Value

No. of distance constraints Total .............................................................................1,572 Intraresidue ( i ⫺ j ⫽ 0).......................................548 Sequential ( i ⫺ j ⫽ 1)..........................................349 Medium range (1 ⬍ i ⫺ j ⬍ 5) ..........................212 Long range ( i ⫺ j ⬎5) .........................................463 No. of torsion angle constraints ...................................74 ⌽ ...................................................................................37 ⌿...................................................................................37 No. of hydrogen bond restraints ..................................24 Residual CYANA target function value .....................0.45 Å2 Residual NOE distance constraint violationsa (no. ⬎ 0.1 Å) ......................................................1 Residual dihedral angle constraint violationsa (no. ⬎ 5.0°) .........................................................0 Geometric statistics Avg backbone atom rmsd to meanb.........................0.37 Å ⫾ 0.07 Å Avg heavy atom rmsd to meanb ...............................0.87 Å ⫾ 0.13 Å Ramachandran analysisc Residues in most favored regions ............................84.9% Residues in additional allowed regions ...................15.1% Residues in generously regions.................................0.0% Residues in disallowed regions .................................0.0% a

Constraints violated in more than five structures. rmsd for residues 2 to 80. Residues 1 to 82. d Ten structures with the lowest CYANA target function were selected out of 100 structures, calculated using CYANA from randomized staring structures. b c

10424

LIU ET AL.

J. VIROL.

FIG. 5. (a) Simulated annealing Fo-Fc omit map in the dimerization region of VP9. All three cadmium ions (green) and all atoms within 3.5 Å of cadmium ions were omitted prior to refinement and map calculation. The map was contoured at a level of 2.5 ␴. This figure was prepared using the program Bobscript. (b) Stick representation for the cadmium coordination sphere. A yellow dashed line indicates the coordination bond. Cd2⫹ is in green and a water molecule is in red. Asp9, Cys46, and Glu31 from one monomer are shown in cyan, and the other monomer in magenta.

Metal binding sites. In the electron density map of the native protein, there were eight strong peaks corresponding to metal ions (Fig. 5a). Based on the coordination, we have interpreted these peaks as Cd2⫹ ions. It is worth mentioning here that the Cd2⫹ ions were essential for crystallization; these metal ions must have been acquired during the crystallization process. The presence of tightly bound divalent metal ions has been previously reported for the structural homologs of VP9 (1, 18).

It is noteworthy to observe that during crystallization, VP9 was unable to crystallize in the absence of Cd2⫹, and even a reduction of the Cd2⫹ concentration led to the destabilization of crystals, strongly indicating the crucial role of Cd2⫹ ions to form intermolecular contacts for crystal formation (Fig. 5b). Various divalent metal ions are integral parts of several viral proteins and are necessary for their survival and pathogenesis (3). These bound metal ions are also required for nucleocapsid protein-transactivation response-RNA interactions. Zinc is the most common divalent metal ion that influences a variety of viral infections (3). Proteins from viruses such as human immunodeficiency virus type 1, hepatitis C, hepatitis B, herpes simplex, pox, rubella, influenza, corona, human papilloma, Ebola, picorn, and rotavirus essentially bind with Zn2⫹ and carry out the host infection (3). In VP9, the observed divalent metal ion coordinates with Asp, Glu, and Cys. Since there is no histidine and only one cysteine is present in the sequence of VP9, the observed coordinating side chains could be considered an unconventional binding site for Zn2⫹. Recently, Shi et al. (19) reported a similar unconventional Zn2⫹ binding site with His, Asp, and Glu. VP9 has identical coordinating side chains, except that histidine is replaced by cysteine. In the crystal, substitution of Cd2⫹ ions for the natural Zn2⫹ binding sites has been observed. Similar substitutions from the crystallization buffers have been previously reported (19). In VP9, all the natural Zn2⫹ binding sites might be taken up by the Cd2⫹ ions present in the crystallization solution. To further verify

VOL. 80, 2006


10425

FIG. 6. Dual 1H-15N HSQC spectra of VP9 in the absence (red) and presence (blue) of Zn-sulfate (a) and Cd-sulfate (b). Shown is a superimposition of the 1H-15N HSQC spectra of the free form of VP9 in red and in the complex with zinc sulfate/cadmium sulfate in blue at a ratio of about 1:4 at pH 6.72 and 298 K. Forty-two perturbed peaks that either disappeared or underwent chemical shifts refer to Asn56, Tyr43, Glu31, Glu72, Arg19, Met44, Leu12, Gln74, Val45, Gly50, Gly57, Leu11, Leu47, Lys35, Leu55, Glu21, Gly14, Thr6, Ile59, Ser36, Val13, Arg63, Asp40, Asp15, Met76, Thr52, Leu48, Cys46, Val78, Ala2, Thr80, Thr81, Ile77, Phe4, Asp7, Ala32, Phe10, Asp9, Leu64, Gln5, Glu61, and Leu62. Of 42 perturbed peaks, Asp9, Phe10, Glu31, Val45, Cys46, Leu47, and Leu48 were perturbed the most due to close locations to the metal binding sites.

this fact, the NMR titration experiments with various divalent metal ions were carried out. NMR metal titration. The binding interactions of VP9 with the four most common metal ions, Zn2⫹, Cd2⫹, Mg2⫹, and Ca2⫹, were studied. The results indicated that only Zn2⫹ and Cd2⫹ were able to bind with VP9. Therefore, we have monitored the binding interaction between 15N-labeled VP9 and

zinc/cadmium. Figure 6a shows the binding profiles of VP9 with Zn2⫹. A detailed analysis of the HSQC titration resulted in the identification of 42 perturbed peaks that either disappeared or underwent chemical shifts. Peaks disappeared because of NMR resonance broadening, indicating that binding led to a significant increase in the conformational exchange on the microsecond-millisecond time scale. The pattern for cad-

10426

LIU ET AL.

J. VIROL.

demonstrated, our studies, including the transcriptional, Western blot, MS, and immunoelectron microscopy results, have identified VP9 as a novel, nonstructural, and abundant protein in the WSSV-infected host tissues. The structure of VP9 is the first structure to be reported for WSSV proteins. The X-rayand NMR-based structural studies revealed that VP9 possesses a DNA recognition fold with specific metal binding sites. A possible natural metal cofactor could be zinc ions (20). Taken together, we speculate that VP9 might be involved in the transcriptional regulation of WSSV, similar to its most structurally homologous counterpart, the E2 protein in papillomavirus-1. Our findings have identified a new candidate protein suitable for further studies towards drug and vaccine development against WSSV infections. ACKNOWLEDGMENTS

FIG. 7. Proposed DNA binding model. Superimposition of E2 monomer (with DNA molecules) on VP9 monomer. DNA is shown in stick representation (blue), E2 in ribbon representation (red), and VP9 in ribbon representation (yellow). ␣1 of E2 and VP9 is highlighted in magenta and cyan, respectively, as shown in the box.

mium was similar to that for zinc (Fig. 6b). However, there is no detectable interaction between VP9 with either magnesium or calcium (data not shown) compared to zinc/cadmium. Our data suggested that VP9 binds to both zinc and cadmium. Functional implications. The dominant transcriptional regulator of the papillomavirus-1 E2 protein, which shares the closest structure homology to VP9, binds to its DNA target through a dimeric arrangement of E2. The E2/DNA complex has been crystallized (PDB code, 2BOP). The bound DNA is severely but smoothly bent over the E2 structure through the interactions between the major grooves and a pair of symmetrically disposed ␣-helices of the E2 dimer. The superposition of the VP9 and papillomavirus-1 E2 bound with the DNA fragment reveals a possible DNA binding mode of VP9. The specific DNA sequence for recognition by VP9 has not yet been established. A possible DNA recognition region is located at ␣1 (Thr17-Thr26) and the ␤-turn (Ser36-Asp40). The helix ␣1 is highly conserved among all structural homologs of VP9. In the superimposed model, the side chains from Arg19 and Lys25 from ␣1 are facing the DNA. Figure 7 shows the DNA binding model for VP9. It shows only the monomer of VP9 superimposed with the monomer of the E2 DNA binding domain. In the E2 crystal structure, the DNA fragment binds with the dimer. However, in the case of VP9, the superimposed DNA fragment has to undergo a conformational change to engage with both monomers of the dimer. Similar conformational changes upon DNA binding were previously documented for several DNA protein complexes (8). Although the exact cellular function of VP9 has yet to be

We acknowledge X12C beam-line, Brookhaven National Laboratory, NSLS, for the data collection, and we thank Anand Saxena for help during data collection, Lim Daina for the technical support of real-time RT-PCR work, Fan Jingsong for the NMR data collection, Shashikant Joshi and Lin Zhi for the valuable discussions, and C. F. Lo for providing VP28 and VP664 antibodies. We thank Li Shaowei, National Institute of Diagnostics and Vaccine Development in Infectious Diseases, Xiamen University, China, for conducting analytical ultracentrifuge experiments. We also acknowledge the facilities provided by the Protein and Proteomic Center (PPC), Department of Biological Sciences, National University of Singapore. We are grateful for the funding support by MOE (Ministry of Education of Singapore) to Choy L. Hew. REFERENCES 1. Banci, L., I. Bertini, S. Ciofi-Baffoni, L. A. Finney, C. E. Outten, and T. V. OⴕHalloran. 2002. A new zinc-protein coordination site in intracellular metal trafficking: solution structure of the apo and Zn (II) forms of ZntA (46-118) J. Mol. Biol. 323:883–897. 2. Brunger, A. T., P. D. Adams, G. M. Clore, W. L. DeLano, P. Gros, R. W. Grosse-Kunstleve, J. S. Jiang, J. Kuszewski, M. Nilges, N. S. Pannu, R. J. Read, L. M. Rice, T. Simonson, and G. L. Warren. 1998. Crystallography and NMR system. A new software suite for macromolecular structure determination. Acta Crystallogr. Sect. D 54:905–921. 3. Chaturvedi, U. C., and R. Shrivastava. 2005. Interaction of viral proteins with metal ions: role in maintaining the structure and functions of viruses. FEMS Immunol. Med. Microbiol. 43:105–114. 4. Chen, L. L., H. C. Wang, C. J. Huang, S. E. Peng, Y. G. Chen, S. J. Lin, W. Y. Chen, C. F. Dai, H. T. Yu, C. H. Wang, C. F. Lo, and G. H. Kou. 2002. Transcriptional analysis of the DNA polymerase gene of shrimp white spot syndrome virus. Virology 301:136–147. 5. Cornilescu, G., F. Delaglio, and A. Bax. 1999. Protein backbone angle restraints from searching a database for chemical shift and sequence homology. J. Biomol. NMR 13:289–302. 6. de La Fortelleqq, E., and G. Bricogne. 1997. Maximum-likelihood heavy-atom parameter refinement for multiple isomorphous replacement and multiwavelength anomalous diffraction methods. Methods Enzymol. 276:472–494. 7. Delaglio, F., S. Grzesiek, G. W. Vuister, G. Zhu, J. Pfeifer, and A. Bax. 1995. NMRPipe: a multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR 6:277–293. 8. Hegde, R. S., S. R. Grossman, L. A. Laimins, and P. B. Sigler. 1992. Crystal structure at 1.7 A of the bovine papillomavirus-1 E2 DNA-binding domain bound to its DNA target. Nature 359:505–512. 9. Herrmann, T., P. Guntert, and K. Wuthrich. 2002. Protein NMR structure determination with automated NOE assignment using the new software CANDID and the torsion angle dynamics algorithm DYANA. J. Mol. Biol. 319:209–227. 10. Holm, L., and C. Sander. 1995. Dali: a network tool for protein structure comparison. Trends Biochem. Sci. 20:478–480. 11. Huang, C., L. Zhang, J. Zhang, L. Xiao, Q. Wu, D. Chen, and J. K. Li. 2001. Purification and characterization of white spot syndrome virus (WSSV) produced in an alternate host: crayfish, Cambarus clarkii. Virus Res. 76:115– 125. 12. Huang, C., X. Zhang, Q. Lin, X. Xu, Z. Hu, and C. L. Hew. 2002. Proteomic analysis of shrimp white spot syndrome viral proteins and characterization of a novel envelope protein VP466. Mol. Cell Proteomics 1:223–231. 13. Johnson, B. A. 2004. Using NMRView to visualize and analyze the NMR spectra of macromolecules. Methods Mol. Biol. 278:313–352.

VOL. 80, 2006


14. Jones, T. A., J.-Y. Zou, S. W. Cowan, and M. Kjeldgaard. 1991. Improved methods for building models in electron density maps and the location of errors in these models. Acta Crystallogr. Sect. A 47:110–119. 15. Laskowski, R. A., M. W. MacArthur, D. S. Moss, and J. M. Thornton. 1993. PROCHECK: a program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 26:283–291. 16. Mayo, M. A. 2002. A summary of taxonomic changes recently approved by ICTV. Arch. Virol. 147:1655–1663. 17. Otwinowski, Z., and W. Minor. 1997. Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol. 276:307–326. 18. Rosenzweig, A. C., D. L. Huffman, M. Y. Hou, A. K. Wernimont, R. A. Pufahl, and T. V. OⴕHalloran. 1999. Crystal structure of the Atx1 metallochaperone protein at 1.02 A resolution. Structure Fold Des. 7:605–617. 19. Shi, W., C. Zhan, A. Ignatov, B. A. Manjasetty, N. Marinkovic, M. Sullivan, R. Huang, and M. R. Chance. 2005. Metalloproteomics: high-throughput structural and functional annotation of proteins in structural genomics. Structure 13:1473–1486. 20. Sivaraman, J., R. S. Myers, L. Boju, T. Sulea, M. Cygler, V. Jo Davisson, and J. D. Schrag. 2005. Crystal structure of Methanobacterium thermoautotrophicum phosphoribosyl-AMP cyclohydrolase HisI. Biochemistry 44:10071– 10080. 21. Song, W. J., Q. W. Qin, J. Qiu, C. H. Huang, F. Wang, and C. L. Hew. 2004.

22. 23.

24. 25. 26. 27. 28.

10427

Functional genomics analysis of Singapore group iridovirus: complete sequence determination and proteomic analysis. J. Virol. 78:12576–12590. Terwilliger, T. C. 2002. Automated structure solution, density modification and model building. Acta Crystallogr. D Biol. Crystallogr. 58:1937–1940. Tsai, J. M., H. C. Wang, J. H. Leu, H. H. Hsiao, A. H. Wang, G. H. Kou, and C. F. Lo. 2004. Genomic and proteomic analysis of thirty-nine structural proteins of shrimp white spot syndrome virus. J. Virol. 78:11360– 11370. van Hulten, M. C., J. Witteveldt, S. Peters, N. Kloosterboer, R. Tarchini, M. Fiers, H. Sandbrink, R. K. Lankhorst, and J. M. Vlak. 2001. The white spot syndrome virus DNA genome sequence. Virology 286:7–22. van Hulten, M. C. W., M. Westernberg, S. D. Goodall, and J. M. Vlak. 2000. Identification of two major virion protein genes of white spot syndrome virus of shrimp. Virology 266:227–236. Wu, J., M. Arimoto, T. Nishizawa, and K. Muroga. 2002. Preparation of an inoculum of white spot syndrome virus for challenge tests in Penaeu japonicus. Fish Pathol. 37:65–69. Yang, F., J. He, X. Lin, Q. Li, D. Pan, X. Zhang, and X. Xu. 2001. Complete genome sequence of the shrimp white spot bacilliform virus. J. Virol. 75: 11811–11820. Zhang, X., C. Huang, X. Tang, Y. Zhuang, and C. L. Hew. 2004. Identification of structural proteins from shrimp white spot syndrome virus (WSSV) by 2DE-MS. Proteins 55:229–235.