Shotgun Identification of the Structural Proteome of Shrimp White Spot Syndrome Virus and iTRAQ Differentiation of Envelope and Nucleocapsid Subproteomes *S

White spot syndrome virus (WSSV) is a major pathogen that causes severe mortality and economic losses to shrimp cultivation worldwide. The genome of WSSV contains a 305-kb double-stranded circular DNA, which encodes 181 predicted ORFs. Previous gel-based proteomics studies on WSSV have identified 38 structural proteins. In this study, we applied shotgun proteomics using off-line coupling of an LC system with MALDI-TOF/TOF MS/MS as a complementary and comprehensive approach to investigate the WSSV proteome. This approach led to the identification of 45 viral proteins; 13 of them are reported for the first time. Seven viral proteins were found to have acetylated N termini. RT-PCR confirmed the mRNA expression of these 13 newly identified viral proteins. Furthermore iTRAQ (isobaric tags for relative and absolute quantification), a quantitative proteomics strategy, was used to distinguish envelope proteins and nucleocapsid proteins of WSSV. Based on iTRAQ ratios, we successfully identified 23 envelope proteins and six nucleocapsid proteins. Our results validated 15 structural proteins with previously known localization in the virion. Furthermore the localization of an additional 12 envelope proteins and two nucleocapsid proteins was determined. We demonstrated that iTRAQ is an effective approach for high throughput viral protein localization determination. Altogether WSSV is assembled by at least 58 structural proteins, including 13 proteins newly identified by shotgun proteomics and one identified by iTRAQ. The localization of 42 structural proteins was determined; 33 are envelope proteins, and nine are nucleocapsid proteins. A comprehensive identification of WSSV structural proteins and their localization should facilitate the studies of its assembly and mechanism of infection.

Shrimp white spot syndrome virus (WSSV), 1 which belongs to Nimaviridae (Whispovirus) family, is an enveloped, double-stranded DNA virus (1,2). It can cause up to 100% mortality in shrimp within 7-10 days, resulting in huge economic losses to the shrimp farming industry (3,4). This virus infects shrimp and other freshwater and marine crustaceans including crabs and crayfish (5). Genomes of three WSSV isolates originating from China (GenBank TM accession number AF332093), Taiwan (GenBank accession number AF440570), and Thailand (GenBank accession number AF369029) have been completely sequenced (6 -9). WSSV originating from China contains a 305-kb double-stranded circular DNA, which encompasses 181 putative ORFs with 50 or more amino acids (9). Most of these ORFs are of unknown function because no homologues to known genes can be found in public databases.
Viral structural proteins have critical functions not only in viral structure and assembly but also in the early stages of infection, cell adhesion, signal transduction, and evasion of the rapidly deployed antiviral defenses of the host. Previously 18 structural proteins from WSSV were identified by using one-dimensional (1D) SDS-PAGE and MALDI-TOF or ESI Q-TOF mass spectrometers (10). Recently 33 WSSV structural proteins resolved by 1D SDS-PAGE were identified using the on-line LC-ESI Q-TOF mass spectrometer, increasing the number of structural proteins identified to 38 by these two proteomics studies (11). Our previous study on Singapore grouper iridovirus suggested that the 1D gel-based approach and the LC-based shotgun approach are equally effective and complementary to each other (12). A number of novel viral proteins were detected by using the shotgun approach. In an effort to achieve a better understanding of the structural proteome of WSSV, shotgun proteomics, which involves direct digestion of total proteins to complex peptide mixtures, followed by the automated identification of the peptides by LC-MS/MS was initiated. In total, 45 viral structural proteins were identified from the purified WSSV, including 32 proteins previously identified and 13 proteins reported for the first time.
Determining the localization of structural proteins in the virion is important to elucidate their roles in both virus assembly and infection. Western blot analysis and immunogold electron microscopy (IEM) are two conventional approaches to localize the viral proteins. For WSSV, IEM has been used to detect 13 envelope proteins and one nucleocapsid protein (10,(13)(14)(15)(16)(17)(18)(19)(20)(21)(22). Recently a more systematic study on WSSV has led to the differentiation of seven envelope proteins, five tegument proteins, and four nucleocapsid proteins by Western blot analysis and two additional nucleocapsid proteins by MS (23). To date, the localization of 27 ORFs in the virion has been determined among the known structural proteins (6,24). In this study, we applied a complementary proteomics approach to examine the localization of structural proteins in the virion by iTRAQ (isobaric tags for relative and absolute quantification). iTRAQ is a newly developed LC-based quantitative proteomics approach, which allows for comparison of up to four different samples simultaneously (25). It has been successfully applied to measure the enrichment of organelle proteins (26) as an alternative approach to the localization of organelle proteins by isotope tagging using cleavable ICAT (27) and protein correlation profiling (28). Determining the viral protein localization by iTRAQ is based on the principle that the enrichment of envelope proteins in the detergent-solubilized fraction and nucleocapsid proteins in the pellet fraction can be quantified by the reporter ions, whereas their protein identities can be determined by other MS/MS fragment ions. Using this approach, we identified 23 envelope proteins and six nucleocapsid proteins among which 12 envelope proteins and two nucleocapsid proteins are reported for the first time. Our results demonstrated that iTRAQ is a powerful approach for rapid protein localization and that it can also be applied to study other enveloped viruses. A better understanding of WSSV structural proteins and the localization in the virion will shed more light on virus assembly, its infection pathway, and the discovery of antiviral drugs.

EXPERIMENTAL PROCEDURES
Proliferation and Isolation of WSSV Virions-WSSV used in this study originated from WSSV-infected Penaeus chinensis (China isolate). Virus inoculums were prepared from the hemolymph of infected red claw crayfish, Cherax quadricarinatus, as described previously (29). After centrifugation at 1,500 ϫ g for 10 min, the supernatant was filtered with a 0.45-m filter and injected intramuscularly into healthy crayfish. After 4 -6 days, hemolymph extracted from moribund crayfish was centrifuged at 2,000 ϫ g for 10 min. The supernatant was layered on top of a 30 -60% (w/v) stepwise sucrose gradient and centrifuged at 53,000 ϫ g for 1 h at 4°C. The virus band was collected and then mixed with TN buffer (20 mM Tris-HCl, 400 mM NaCl, pH 7.4) and repelleted at 53,000 ϫ g for 1 h at 4°C. The resulting pellet was washed with TN buffer to remove sucrose and then resuspended in TN buffer. The purified virus samples were negatively stained with 2% phosphotungstic acid and examined under the transmission electron microscope to check the purity and quantity.
Shotgun LC-MALDI MS Analysis of WSSV Structural Proteins-Viral protein extraction, in-solution digestion, and LC separation of tryptic peptides were carried out following procedures described previously (12). Briefly 5 volumes of 50 mM Tris-HCl with 0.1% SDS, pH 8.5, were used to extract proteins from purified WSSV virions. The extracted proteins were reduced with triscarboxyethylphosphine, alkylated with iodoacetamide, and then digested with sequencing grade porcine trypsin (Promega, Madison, WI). The digested peptide mixture was separated using an Ultimate LC system (Dionex-LC Packings, Sunnyvale, CA) equipped with a Probot MALDI spotting device. Approximately 10 g of peptide mixture were captured by a 0.3 ϫ 1-mm trap column (3-m C 18 PepMap, 100 Å, Dionex-LC Packings) and separated by a 0.075 ϫ 150-mm analytical column (3-m C 18 PepMap, 100 Å, Dionex-LC Packings) at a flow rate of 0.4 l/min. The mobile phases A and B were 2% ACN, 0.05% TFA and 80% ACN, 0.04% TFA, respectively. The LC gradients used were 0 -20% B in 10 min, then 20 -60% B over 3 h, 60 -100% B in 1 min, and kept at 100% B for 5 min. The LC fractions were mixed with MALDI matrix solution (7 mg/ml ␣-cyano-4-hydroxycinnamic acid and 130 g/ml ammonium citrate in 75% ACN) before spotting onto MALDI target plates.
An ABI 4700 Proteomics Analyzer MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA) was used to analyze the samples. The instrument was controlled by 4000 Series Explorer version 3.0. For MS analyses, typically 1,000 subspectra were accumulated. Peaks were detected with the minimum signal to noise ratio (s/n) set to 15, and the peaks were deisotoped. MS/MS analyses were carried out using nitrogen at a collision energy of 1 kV and a collision gas pressure of ϳ9.0 ϫ 10 Ϫ7 torr. Two thousand to 10,000 subspectra were combined for each spectrum using stop conditions based on the quality of the data. The spectra were smoothed using the Savitsky-Golay method with points across peak (full-width half-maximum) set to 3, and polynomial order was set to 4. The peaks were deisotoped, and only the peaks with s/n Ն10 were picked. GPS Explorer software version 3.0 (Applied Biosystems) was used to create and search files with the MASCOT search engine version 2.0 (Matrix Science, Boston, MA) to identify viral proteins. A database (64,335 entries) containing all predicted ORFs from three WSSV isolates (2,013 entries) together with the International Protein Index human database version 3.16 (www.ebi.ac.uk/IPI/IPIhelp.html, 62,322 entries) was used to minimize false positive identifications. The search was restricted to tryptic peptides, and one missing cleavage was allowed. Cysteine carbamidomethylation, N-terminal acetylation and pyroglutamation (Glu or Gln), and methionine oxidation were selected as variable modifications. Precursor error tolerance and MS/MS fragment error tolerance were set to 60 ppm and 0.4 Da, respectively. Only the top ranked peptide matches were taken into consideration for protein identification. For peptide matches with an expect value Ͼ0.05, the MS/MS spectra were further validated manually.
Bioinformatics Analysis of WSSV Structural Proteins-To characterize the previously unknown WSSV structural proteins, the homology analysis was achieved by searching InterProScan (30,31). Putative signal sequences and transmembrane domains were predicted by dense alignment surface (32). PSORT was used for the prediction of protein cellular localization in cells based on the Swiss-Prot data.
Isolation of Total RNA and RT-PCR-WSSV-infected crayfish gills were treated with RNAlater (Qiagen, Hilden, Germany). Total RNA was isolated from these tissues using an RNeasy minikit (Qiagen). To remove any residual DNA, RNA solution was treated with DNA-free kit (Ambion, Austin, TX) following the protocol described. Gene-specific primers were used to amplify the target genes by the One-Step RT-PCR kit (Qiagen) (primer pairs are provided in Supplemental Data I). All procedures were performed according to the manufacturer's instruction. Briefly cDNA was reverse transcribed at 50°C for 45 min. The PCR amplification segment was started with an initial heating step at 95°C for 15 min to activate HotStarTaq DNA polymerase and simultaneously inactivate reverse transcriptase. After activation of DNA polymerase, PCR amplification reactions were performed (40 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 2 min). A final extension step was carried out at 72°C for 7 min. RT-PCR products were resolved on 1.2% agarose gels. The genes of two ORFs (wsv143 and wsv161) were divided into several short fragments (less than 1,600 bp) to get complete coverage of the entire coding region. For controls, purified RNA was added after inactivating reverse transcriptase to exclude the possibility of genomic DNA contamination.
Separation of Viral Envelope Proteins and Nucleocapsid Proteins-A previously validated separation procedure described by Tsai et al. (23) was adopted with modification. Briefly purified virus was treated with buffer A containing 20 mM Tris-HCl, 5 mM EDTA-Na 2 , 1% Triton X-100, 0.5 M NaCl, 1ϫ protease inhibitor mixture (Roche Diagnostics Asia Pacific Pte. Ltd.), pH 7.4, at 4°C for 30 min. Then the virus was divided into two equal portions. One portion was set aside as a control for total viral proteins, whereas the other portion was separated into two fractions, supernatant and pellet, by centrifugation at 200,000 ϫ g for 1 h at 4°C. The pellet fraction was washed one more time with buffer A and centrifuged again for 1 h. The pellet was resuspended in an equal volume of buffer A as the envelope fraction.
Western Blot Analysis of Envelope and Nucleocapsid Fractions-Virion-associated proteins from each fraction were resolved by 12% SDS-PAGE and processed for Western blot analysis. The nitrocellulose membrane was blocked with 5% nonfat milk in 1ϫ TBST (20 mM Tris base, 137 mM NaCl, 0.1% Tween 20, pH 7.6) at room temperature and then subjected to Western blotting. Primary antibodies were used with the following concentrations: rabbit anti-VP28 (wsv421) polyclonal antibody (1:2,000) and mouse anti-VP24 (wsv002) and anti-VP466 (wsv308) polyclonal antibody (1:2,000). The secondary antibodies, antirabbit or anti-mouse horseradish peroxidase-conjugated antiserum, were diluted 1:5,000 (GE Healthcare). Pierce SuperSignal West Pico chemiluminescent substrate (Pierce) was used according to the manufacturer's protocol, and the protein bands were visualized using Lumi-Film chemiluminescent detection film (Roche Applied Science).
iTRAQ Labeling and Two-dimensional (2D) LC-MALDI MS to Determine Viral Protein Localization-One hundred micrograms of total viral proteins and the equivalent envelope and nucleocapsid fractions (separated from 100 g of total viral proteins) were processed using a 2D Clean-Up kit (GE Healthcare) and resuspended in the dissolution buffer (0.5 M triethylammonium bicarbonate, pH 8.5, containing 0.1% SDS). The samples were then reduced, and cysteines were blocked according to the protocol of the iTRAQ kit (Applied Biosystems). Ten microliters of a 1 g/l trypsin (Applied Biosystems) solution were added, and the samples were digested at 37°C overnight. The samples were vacuum-dried, reconstituted with 30 l of dissolution buffer, and labeled with iTRAQ tags as follows: total viral proteins, iTRAQ 114 reagent; the envelope fraction, iTRAQ 115 reagent; and the nucleocapsid fraction, iTRAQ 116 reagent. The labeled samples were then pooled and purified using a strong cation exchange (SCX) column (Applied Biosystems), and the bound peptides were eluted with 5% NH 4 OH in 30% methanol.
MS analysis was performed as described above, and the MS/MS analysis settings were the same as those for the shotgun analysis except that the collision gas pressure was changed to ϳ2 ϫ 10 Ϫ6 torr. For the precursor ions with s/n Ն100, 6,000 shots were combined for each spectrum. For the precursors with s/n between 50 and 100, 10,000 shots were acquired. The peak processing and detection parameters were the same as those for the shotgun analysis described above. GPS Explorer software version 3.5 (Applied Biosystems) using the MASCOT search engine (version 2.1, Matrix Science) was used for peptide and protein identifications and iTRAQ quantification. The database used was the same as mentioned above and restricted to tryptic peptides. Cysteine methanethiolation, N-terminal iTRAQ labeling, and iTRAQ-labeled lysine were selected as fixed modifications, and methionine oxidation was selected as a variable modification. One missing cleavage was allowed. Precursor error tolerance and MS/MS fragment error tolerance were set to 120 ppm and 0.4 Da, respectively. Maximum peptide rank was set to 1, and minimum ion score confidence interval (C.I.; %) for peptide was set to 0. For proteins with low ion scores (Յ30), the MS/MS spectra were manually inspected. Antibody Preparation and Western Blot Analysis of wsv432-The full-length gene of wsv432 was PCR-amplified and inserted into a modified pET vector containing a C-terminal His 6 tag. After sequencing, the construct was transformed into Escherichia coli strain BL21 Star (DE3) (Invitrogen), and the protein was expressed after isopropyl ␤-D-thiogalactopyranoside induction at 18°C. The recombinant protein was purified using a nickel-nitrilotriacetic acid column, and its identity was confirmed by MS. The antibody was prepared by Bam Biotech Co., Ltd. (Xiamen, Fujian, China) using the purified fusion protein to immunize the rabbits. Proteins from the virion, the envelope, and the nucleocapsid were resolved by SDS-PAGE and subjected to Western blot analysis as described above. The anti-wsv432 antibody was diluted 1:1,000.
Localization of wsv432 in the Virion by IEM-The purified virus was treated with 0.1% Tween 20 at room temperature for 30 s. After washing with 0.2 M phosphate buffer, pH 7.3, to remove the detergent, the virus suspension was absorbed on carbon-coated nickel grids. Rabbit anti-wsv432 antibody was used to recognize wsv432 in the viral particles, whereas preimmune rabbit serum was included in parallel as a negative control. The second antibody was goat antirabbit IgG conjugated with 15-nm gold particles (Electron Microscopy Sciences, Hatfield, PA). The subsequent immunogold labeling was carried out according to Leu et al. (33). The specimens were examined under a transmission electron microscope.

Identification of Virion-associated Proteins by Shotgun Proteomics-
The purity of isolated WSSV virions using sucrose gradient separation was confirmed by electron microscopy FIG. 2. RT-PCR amplification of the cDNAs of 13 newly identified structural proteins of WSSV. For each protein, the left lane is the genomic DNA contamination control, and the right lane is its corresponding cDNA product. All protein coding sequences were completely amplified including wsv143 and wsv161. For wsv143, five RT-PCR products (wsv143-1 to -5) were amplified to cover the entire coding region; whereas for wsv161, two RT-PCR products (wsv161-1 and -2) were produced. Shotgun Proteomics and Subproteomes of WSSV (Fig. 1A). Most of purified virions were intact. The structural proteome of WSSV was analyzed by shotgun proteomics using off-line LC-MALDI workflow. In total, 45 viral proteins were detected with amino acid sequences that matched WSSV ORFs (Table I). Except for wsv198 and wsv419, all other identified proteins contained at least one top ranked peptide match with MASCOT expect value Ͻ0.05 (Supplemental Data II). wsv198 was included as a true identification as it had three top ranked peptides with MASCOT scores of 26, 24, and 15 (Supplemental Data II). Although wsv419 only contained one top ranked peptide with a MASCOT score of 28, the MS/MS spectrum was manually inspected and was considered as a reliable assignment. One protein was identified in WSSV (China isolate); however, the predicted ORF can only be found in the Taiwan (wssv349) and Thailand (ORF144) isolates, not in the China isolate. Thus, we named it wssv349. Although a comprehensive study on the WSSV proteome was carried out by the gel-based approach earlier, our shotgun method successfully identified an additional 13 structural proteins in WSSV. Potential functions of these 13 ORFs were predicted by InterProScan. It was not surprising to find that putative functions could only be assigned for two of these proteins because most of the WSSV ORFs are functionally unknown. wsv143 is homologous to SOX transcription factors, whereas wsv161 has five sequence regions that are homologous to prichextensn, which represents a signature for proline-rich extensins of plant cell wall proteins.
FIG. 3. Western blot analysis of total proteins, envelope proteins, and nucleocapsid proteins. The three antibodies used were anti-VP466, -VP28, and -VP24, respectively. The total protein concentrations were 0.3 g for the left panel and 12 g for the right panel.

FIG. 4. The iTRAQ labeling workflow and 2D LC-MS for the localization of structural proteins in WSSV.
The virus envelopes were separated from the nucleocapsids after detergent treatment. To quantitatively distinguish envelope proteins from nucleocapsid proteins, total viral proteins, proteins from the envelope fraction, and the nucleocapsid fraction were reduced, cysteine-blocked, and digested with trypsin, respectively. The tryptic peptides were labeled with 114, 115, and 116 iTRAQ reagents, respectively. Then the combined peptide mixture was cleaned up using SCX chromatography. The eluate was further separated by 2D LC coupling SCX with reverse-phase chromatography, and the column effluent was added with MALDI matrix solution and spotted on MALDI target plates, which were later analyzed by an ABI 4700 Proteomics Analyzer MALDI-TOF/TOF mass spectrometer.

Post-translational Modification of WSSV Proteins-From
the MS/MS data, seven identified proteins contained acetylated N termini (Table II and Supplemental Data III). wsv134, wsv198, and wsv360 were acetylated at the N-terminal methionine residues. wsv131, wsv289, wsv414, and wsv432 were acetylated at the second amino acids, indicating that the N-terminal methionines of these proteins were removed after translation.
RT-PCR Confirmation of the Gene Expression-To confirm the expression of the genes of those 13 newly identified viral proteins by shotgun proteomics, RT-PCR was conducted to detect the existence of mRNAs for these ORFs. Our results showed that all these genes were expressed (Fig. 2). The sizes of the RT-PCR products of each gene are provided in Supplemental Data I.

Separation of WSSV Envelope Proteins and Nucleocapsid
Proteins-Viral structural proteins were separated into envelope and nucleocapsid fractions after Triton X-100 and 0.5 M NaCl treatment (Fig. 1B) (23). In the present investigation, we could not distinguish the tegument proteins from envelope proteins, so we classified viral structural proteins as envelope proteins (including tegument proteins) and nucleocapsid proteins. Western blotting was performed using antibodies against VP28, VP24, and VP466 to check the separation of viral proteins (Fig. 3). Our results confirmed that VP466 is a nucleocapsid protein (23). As expected, envelope proteins and nucleocapsid proteins were enriched in their respective fractions after separation.
Localization of Structural Proteins in WSSV by iTRAQ-A quantitative proteomics experiment was carried out using iTRAQ reagents and 2D LC-MALDI MS to localize various structural proteins in WSSV. The workflow of this technique is outlined in Fig. 4. After separation, total viral proteins, envelope proteins, and nucleocapsid proteins were labeled with iTRAQ reagents 114, 115, and 116, respectively. Because of their enrichment through the separation process, we expected that envelope proteins would have 115/114 ratios higher than 116/114 ratios, whereas nucleocapsid proteins would instead have lower 115/114 ratios compared with 116/ 114 ratios. Fig. 5 shows the representative MS/MS spectra of iTRAQ ratios from envelope protein wsv009 and nucleocapsid protein wsv289 determined in this study (for the rest of the spectra, see Supplemental Data IV). Based on the iTRAQ reporter ion ratios, we observed that 23 structural proteins are envelope proteins and six are nucleocapsid proteins (Table  III). These included seven proteins with best ion scores Յ30 (wsv198, wsv216, wsv230, wsv238, wsv242, wsv256, and wsv289). Because these proteins have been confidently identified as viral structural proteins by the shotgun proteomics study and the peptides matched were all top ranked in the iTRAQ study, these protein assignments should be reliable. The MS/MS spectra were manually inspected for further validation (Supplemental Data IV). Using the iTRAQ approach, we confirmed 11 envelope proteins and four nucleocapsid proteins that were identified previously, among which VP24 and VP26 are envelope proteins, whereas VP466 is a nucleocapsid protein (23,34). Moreover we identified an additional 12 envelope proteins and two nucleocapsid proteins, demonstrating the effectiveness of this approach in the study of subproteomes of viruses. Five newly identified structural proteins by shotgun proteomics and 23 known proteins were further verified in the iTRAQ study. One envelope protein, wsv295, was identified as a viral structural protein for the first time in our iTRAQ study. Expression and Localization of the Structural Protein wsv432 in WSSV-Recombinant wsv432, a protein newly identified by shotgun proteomics, was expressed in E. coli and purified (Fig. 6A). The MS analysis confirmed the identity of this protein (data not shown). Western blot analysis indicated that this protein was present on the virus envelope, not the nucleocapsid (Fig. 6B). IEM was carried out to further validate the presence of wsv432 in WSSV. To verify the localization of wsv432 on virus particles, the virus was treated with Tween 20 to partially separate the envelope from the nucleocapsid. Subsequent immunogold labeling demonstrated that gold particles were localized on the envelopes but not on the naked nucleocapsids, whereas no gold particles were observed on the negative control (Fig. 7). The result revealed that wsv432 is an envelope protein; this is consistent with our Western blot result. DISCUSSION For the past few years, gel-based techniques have been successfully used to study the WSSV proteome, and 38 viral structural proteins were identified (10,11). To further characterize the protein components of this virus, shotgun proteomics was applied, and this has led to the identification of 45 structural proteins in our study, including 13 proteins that have not been identified before. The reliability of our findings was verified by ORF-specific RT-PCR for those newly identified proteins. Surprisingly wsv230, which was defined as nonstructural protein VP9 (wsv230) in a previous study (35), was detected in the virion by both shotgun and iTRAQ approaches. The postulation that wsv230 is a non-structural protein was based on the Western blot and IEM results. However, the inability to detect wsv230 by antibody-based approaches might be due to the low abundance of this protein in the virion. Both Western blot analysis and IEM confirmed that wsv432 identified by shotgun proteomics, not by 1D gel MS, is an envelope protein. Moreover this protein contains an RGD motif for cell adhesion. It was reported that the threonine residue right after RGD is important for its interaction with integrin (36). As for wsv432, the threonine residue is replaced with a serine, implying that this protein might also have cell adhesion activity important for virus entry.
A genome-wide transcription profile of WSSV ORFs using DNA microarrays to understand the regulation of WSSV gene expression has been reported (37)(38)(39). For the 13 previously unknown proteins, five of these genes started expression at 2 h postinfection (hpi), one started expression at 12 hpi, and six started expression at 24 hpi based on the DNA microarray data (Table I) (11,37).
Fifty-five WSSV structural proteins identified by three proteomics studies are summarized in Fig. 8. It is noted that 17 proteins were uniquely identified by the LC-based approach, 10 were uniquely identified by the gel-based approach, and 28 were identified by both, confirming the complementary nature of these two approaches. Among the 17 ORFs identified by shotgun proteomics, four structural proteins, including two envelope proteins, VP187 (wsv209) and VP124 (wsv216), and two nucleocapsid proteins, VP160B (wsv037) and VP160A (wsv289), have been identified by gel-based MS analyses in three recent publications (19,23,40). However, four structural proteins, VP1684 (wsv001), VP357 (wsv129), VP184 (part of wsv303), and VP448 (wsv526), were only identified from 1D gel by MALDI-TOF peptide mass fingerprinting and could not be confirmed by MS/MS data, suggesting that these protein identifications are of low reliability and that further evidence is needed to support these findings (10). Because only visible bands were excised from the gel, it is possible that those low abundance proteins might not be included for the identification. Therefore, the LC-based approach has an advantage in identifying low abundance proteins such as wsv230. Moreover several highly basic proteins, such as wsv021 (pI 9.35) and wsv324 (pI 9.3), were also identified by the LC method. Altogether a total of 58 structural proteins have been identified in WSSV, including 55 proteins identified by shotgun and previous proteomics studies, one envelope protein, wsv295, identified by our iTRAQ study, and two nucleocapsid proteins, VP35 (wsv493) and VP15 (wsv214), reported previously (6,11).
Traditionally Western blot analysis and IEM have been used to determine the localization of viral proteins. These approaches require a substantial amount of work such as the generation and characterization of the antibodies. In addition, some of these results were controversial. For example, VP466 FIG. 8. A summary of WSSV structural proteins identified by proteomics studies. WSSV structural proteins identified in the present study and two earlier reports are summarized. *, ORFs identified in other literature as described under "Discussion" (19,23,40). Underlined ORFs were also verified by the iTRAQ study. **, altogether 58 structural proteins were identified, including 55 proteins identified by these three studies, wsv295 identified by iTRAQ, and VP35 and VP15 reported previously. was determined by IEM as an envelope protein, whereas Western blot analysis suggested that it is a nucleocapsid protein (10,23). The precise localization of proteins in WSSV by non-quantitative MS analysis has been hindered by the difficulties in the successful separation of envelope proteins and nucleocapsid proteins. Western blot analysis could detect the traces of envelope proteins VP28 and VP24 in the nucleocapsid fraction. In the present study, quantitative MS analysis was introduced to discriminate viral envelope proteins and nucleocapsid proteins according to their distributions and hence localization. As we expected, two distinct types of iTRAQ reporter ion spectra were achieved for enve-lope proteins and nucleocapsid proteins, respectively (Fig. 5). However, some iTRAQ ratios (116/114) of nucleocapsid proteins were only slightly higher than that of 115/114 because of the presence of solubilized nucleocapsid proteins in the envelope fraction. This is most likely due to partial dissociation of nucleocapsid proteins caused by osmotic shock during the high salt treatment and the separation process (23,41). As a verification of our approach, the localization results of structural proteins in the virion based on iTRAQ ratios are in good agreement with the previous investigations, including 11 envelope proteins and four nucleocapsid proteins (19,23,40). Seven of the newly characterized proteins contain predicted transmembrane domains, supporting their classification as envelope proteins (Table IV). Our data demonstrated the feasibility of accurate localization of viral structural proteins using the iTRAQ approach even though contaminant-free fractions could not be obtained. The ability to distinguish envelope and nucleocapsid proteins without the need of complete separation represents a significant advance in the localization of viral structural proteins. wsv143 and wsv271 were two newly identified nucleocapsid proteins. DNA microarray data demonstrated that wsv143 is an early gene, which is consistent with its potential function as a transcription factor (37). Elucidation of the function of this protein should be useful to uncover the mechanism of virus replication. Although both wsv143 and wsv271 contain a KGD or RGD motif for cell adhesion, it is less likely that they may have such a function as they are localized in the nucleocapsid. Moreover we also found that all nine nucleocapsid proteins are predicted to target to the nucleus by PSORT search and do not contain any transmembrane domains predicted by dense alignment surface, implying that these two criteria are required for most nucleocapsid proteins. All known localizations of structural proteins of WSSV are summarized in Table IV. In the present investigation, more comprehensive proteomics analyses of the WSSV components and their localization in the virion were achieved using shotgun and iTRAQ technologies. Our results demonstrated that iTRAQ is an effective MS-based approach to distinguish the envelope and nucleocapsid proteins. This approach should also be applicable to study the protein localization of other enveloped viruses. In summary, WSSV is composed of at least 58 structural proteins, including 13 viral proteins identified by shotgun proteomics and one identified by iTRAQ in this study. Among these structural proteins, the localization of 33 envelope proteins and nine nucleocapsid proteins was determined, including 12 envelope proteins and two nucleocapsid proteins identified by our iTRAQ study. Our findings could provide new information to investigate the molecular mechanisms of virus assembly and virus entry.