Functional Peroral Infectivity Complex of White Spot Syndrome Virus of Shrimp

ABSTRACT

White spot syndrome virus (WSSV) is a major cause of disease in shrimp cultures worldwide. The infection process of this large circular double-stranded DNA virus has been well studied, but its entry mechanism remains controversial. The major virion envelope protein VP28 has been implicated in oral and systemic viral infection in shrimp. However, genetic analysis of viral DNA has shown the presence of a few genes related to proteins of per os infectivity factor (PIF) complex in baculoviruses. This complex is essential for the entry of baculoviruses, large terrestrial circular DNA viruses, into the midgut epithelial cells of insect larvae. In this study, we aimed to determine whether a PIF complex exists in WSSV, the components of this complex, whether it functions as an oral infectivity complex in shrimp, and the biochemical properties that contribute to its function in a marine environment. The results revealed a WSSV PIF complex (~720 kDa) comprising at least eight proteins, four of which were not identified as PIF homologs: WSV134, VP124 (WSV216), WSSV021, and WSV136. WSV134 is suggested to be a PIF4 homolog due to predicted structural similarity and amino acid sequence identity. The WSSV PIF complex is resistant to alkali, proteolysis, and high salt, properties that are important for maintaining infectivity in aquatic environments. Oral infection can be neutralized by PIF-specific antibodies but not by VP28-specific antibodies. These results indicate that the WSSV PIF complex is critical for WSSV entry into shrimp; the complex’s evolutionary significance is also discussed.

IMPORTANCE White spot disease, caused by the white spot syndrome virus (WSSV), is a major scourge in cultured shrimp production facilities worldwide. This disease is only effectively controlled by sanitation. Intervention strategies are urgently needed but are limited by a lack of appropriate targets. Our identification of a per os infectivity factor (PIF) complex, which is pivotal for the entry of WSSV into shrimp, could provide new targets for antibody- or dsRNA-based intervention strategies. In addition, the presence of a PIF complex with at least eight components in WSSV, which is ancestrally related to the PIF complex of invertebrate baculoviruses, suggests that this complex is structurally and functionally conserved in disparate virus taxa.

INTRODUCTION

White spot syndrome virus (WSSV) is a major scourge in shrimp cultivation worldwide, and causes huge economic losses in the shrimp industry. The virus was discovered in the early 1990s in Southeast Asia and has since spread worldwide (1). WSSV has a very wide host range among crustaceans, and even occurs in annelids (2, 3). WSSV infects various tissues of both ectodermal and mesodermal origins. The major target of the virus in crustaceans is the hepatopancreas (4). As a result of the infection, the animals lose appetite, become discolored and lethargic, and serve as prey for other crustaceans and natural enemies. The disease quickly spreads in ponds and other environments, and shrimp die in about a week, often showing white spots on the carapace (5, 6).

WSSV is the only member (species) of the Nimaviridae family (7). The enveloped virions contain a double-stranded (ds) circular DNA molecule of approximately 300 kilobase pairs (kb), depending on the isolate. The viral DNA encodes approximately 180 open reading frames (ORFs), many of which have unknown functions. Virions contain over 40 virus-encoded proteins, of which VP19, VP24, VP26, and VP28 are the major proteins and VP664 is the largest viral protein (664 kDa) (8 to 12). The last protein is a nucleocapsid protein. Viral DNA also contains multiple regions of homologous repeats, a characteristic shared with other large circular dsDNA viruses of invertebrates, such as nudiviruses, baculoviruses, hytrosaviruses, and bracoviruses (13).

Recently, endogenous nimavirus sequences were identified and were found to be integrated into the genomes of several crustaceans (14, 15). These include Marsupenaeus japonicas endogenous nimavirus (MjENV), Penaeus monodon endogenous nimavirus (PmENV), Metapenaeus ensis nimavirus (MeENV), Sesarmops intermedium endogenous nimavirus (SiENV), Hemigrapsus takanoi endogenous nimavirus (HtENV), Metopaulias depressus WSSV-like virus, and Litopenaeus vannamei endogenous nimavirus (LvENV). The genomes of these endogenous viruses, if present, range from ~200 kb to 279 kb. Chionoecetes opilio bacilliform virus (CoBV), which causes milky hemolymph syndrome in the snow crab Chionoecetes opilio, is phylogenetically related to WSSV but has not yet been officially assigned to the Nimaviridae family. Twenty-eight conserved genes (also called core genes) and 11 ancestral genes (which are conserved in most, but not all, genomes) were identified in the genomes of these viruses (14). In this study, these viruses are referred to as potential members of the Nimaviridae family.

The entry process of WSSV into shrimp at the cellular and molecular levels is not understood. The virus enters per os to infect the stomach of the shrimp and/or via the gills or nephrocomplex (16). One of the major virion envelope proteins, VP28, has been implicated in oral entry (17), but the mechanism and receptors involved have not been elucidated. VP28 has also been implied in systemic infection by WSSV, but this is controversial (18 to 20). Vaccines based on various WSSV proteins (VP26 and VP28) have been developed but are only partially effective (increased survival) (17, 21).

Genetic analysis of the WSSV genome revealed the presence of at least five homologs of the per os infectivity factor (PIF) genes of baculoviruses (PIF0 [22] [VP53B, WSV115], PIF1 [23] [VP187, WSV209], PIF2 [24] [VP110, WSV035], PIF3 [25] [VP39A, WSV306], and PIF5 [26] [VP150, WSV011]) (14, 27), suggesting that WSSV also encodes a PIF complex. There are other nuclear arthropod large DNA viruses (NALDV), such as bracoviruses, hytrosaviruses, and nudiviruses, that also harbor PIF homologs (14, 28 to 32), but the function of PIFs and the PIF complex has been mainly characterized in baculoviruses and bracoviruses (33, 34). In Autographa californica multinucleopolyhedrovirus (AcMNPV), the most studied baculovirus, 10 PIF proteins have been identified. Of these proteins, 9 (except PIF5) form a stable, alkali- and protease-resistant complex, which is essential for survival in the midgut and entry into the midgut epithelial cells of susceptible insect larvae (35, 36). In the case of baculoviruses, the PIF complex is present only in occlusion body-derived virions (ODVs), which are responsible for oral infection within the midgut, but not in budded virions (BVs), which are responsible for systemic infection inside the larval body. The presence of five baculovirus PIF homolog genes in the WSSV genome may signal the presence of a similar PIF complex on the surface of virions, but whether it has the same function, if it exists, in WSSV infection remains to be elucidated.

In this study, we characterized a PIF complex from WSSV virions, identified the protein components and their genes, and showed the functional significance of this complex for WSSV infection of shrimp. We conclude that a PIF complex consisting of at least eight proteins is involved in the entry of WSSV into shrimp. The fact that WSSV is ancestral to other large invertebrate dsDNA viruses shows the evolutionary significance of the PIF complex in viral infection by these viruses in aquatic and terrestrial invertebrates.

RESULTS

WSSV PIFs are envelope proteins.

Previously, five WSSV PIF homologs (PIF0, PIF1, PIF2, PIF3, and PIF5) were identified based on gene homology with their counterparts in baculoviruses (14). To verify whether these proteins are virion envelope proteins, WSSV virions were purified and the fractions of virion envelope and nucleocapsids were separated, as described in Materials and Methods. The quality of the purified virions and nucleocapsid fraction was checked by electron microscopy. As shown in Fig. 1A, the intact virions and naked nucleocapsid were clearly observed, indicating the virus purification and fractionation of envelop and nucleocapsid were successful. Polyclonal antibodies against the five WSSV PIF homologs were generated, and Western blotting showed that the majority of these proteins were present in the envelope fraction (E) and were not (or only weakly) associated with the nucleocapsid (NC) (Fig. 1B). The antibody against VP28, the major envelope protein of WSSV, showed VP28 mainly in the envelope fraction of purified virions, indicating the successful separation of the envelope and nucleocapsid. The open reading frame of pif0 encodes 968 amino acids (aa), resulting in a full-length molecular size of 108 kDa. However, a polyclonal antibody against the N terminus of PIF0 (13 to 631 aa) reacted with an ~56- kDa band on an SDS-PAGE gel (immunoblot). We speculate the possible presence of cleavage of the PIF0 protein, similar to the case for baculovirus PIF0 (37). The observed molecular size of PIF1, PIF2, and PIF3 on the immunoblot of virion or envelope fractions was as predicted, which reacted to the bands of ~174, 108, and 47 kDa, respectively (Fig. 1B). For PIF5, the antibody detected a band of predicated size (~144 kDa), but also some other bands, including a major band of ~100 kDa. These possible degraded forms of PIF5 were also reported previously (8).

FIG 1.

White spot syndrome virus (WSSV) per os infectivity (PIF) homologs are envelope proteins. (A) Transmission electron microscopy observation of purified WSSV and nucleocapsids. The scale bar is 200 nm. (B) Western blot analysis of WSSV PIF homolog proteins. The total WSSV virion (virion), envelope fraction (E), and nucleocapsid (NC) were analyzed with SDS-PAGE and probed with individual antibody. The antibody against major envelope protein VP28 was used as positive control for envelope proteins. The expected signal is indicated with an arrow.

WSSV PIF homologs form a multiprotein complex.

We next analyzed the envelope fraction using blue native PAGE (BN-PAGE), a method for isolating native membrane protein complexes (38). Western blotting after BN-PAGE detected a single ~720-kDa band with specific antibodies against each of the four WSSV PIF homologs, but not with the PIF5 antibody, which probed two bands near 140 kDa (Fig. 2A). Therefore, as is the case for baculoviruses, the homologous PIF proteins of WSSV also very likely form a protein complex that is 720 kDa in size and comprises PIF0-3, but not PIF5. Anti-VP28 identified multiple bands, likely representing multimers of VP28, but not the 720-kDa band (Fig. 2A).

FIG 2.

Identification of the WSSV PIF complex. (A) BN-PAGE followed by Western blot using antibodies against known WSSV PIF homologs. The WSSV envelope fraction was analyzed with BN-PAGE and probed with antibodies against each PIF and VP28 (positive control). The PIF complex signal is indicated with an arrow. The VP28 protein complexes were also indicated. (B) Verification of antibodies against candidate proteins in the WSSV PIF complex. Polyclonal antibodies were used for Western blot analysis to detect each protein from samples of viriron, envelope (E), and nucleocapsid (NC) fraction. (C) Identification of novel components of the WSSV PIF complex. The WSSV PIF complex isolated from BN-PAGE in panel A was analyzed by Western analysis after BN-PAGE with individual antibodies. Antibodies against PIF1 and PIF2 were used as positive control for the complex. The four additional components identified within the complex are labeled in red.

Novel components of the WSSV PIF complex.

The estimated accumulated size of the four PIFs (PIF0-3) is approximately 436 kDa, which is significantly less than the observed 720 kDa, suggesting that additional components were associated with the complex or multimers of PIF0-3. To identify other candidates, the 720-kDa band was sliced from BN-PAGE and subjected to liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis. In total, 37 viral proteins were identified, including PIF0-3 (Table 1), and their respective ORFs were identified. We then attempted to generate antibodies against candidate proteins with a relatively high abundance in the complex band and successfully obtained 20 specific antibodies, including those used in Fig. 1B (Table 1). Protein fractions from the nucleocapsid and envelope were analyzed by Western blotting using these antibodies, and all these candidate PIF proteins were membrane associated, although a few (VP26, WSV073, and WSV077) were also detected in the NC fraction (Fig. 2B). Western blotting analyses after BN-PAGE showed that, in addition to the known WSSV PIF homologs, only antibodies against VP124, WSV021, WSV134, and WSV136 reacted with the complex band (Fig. 2C), suggesting that these four proteins are associated with the WSSV PIF complex. This suggests that the 720-kDa complex may comprise at least eight viral proteins: PIF0, PIF1, PIF2, PIF3, VP124, WSV021, WSV134, and WSV136. The new pif genes are all ancestral genes of Nimaviridae (14), implying that the putative PIF complex may have existed in the ancestor of this viral family and is important in the viral life cycle.

TABLE 1.

Viral proteins identified in the WSSV 720 kDa band by LC-MS/MS^a

Proteinb	mol wt (kDa)	Test 1		Test 2		Test 3
		No. of unique peptides	Peak area (mAU·min)	No. of unique peptides	Peak area (mAU·min)	No. of unique peptides	Peak area (mAU·min)
WSV386 (VP12B)	6.76	1	2.2E + 10	1	2.3E + 07	1	2.4E + 10
WSV311 (VP26)	22.15	14	9.5E + 09	11	4.2E + 09	11	5.9E + 09
WSV002 (VP24)	23.13	16	2.5E + 09	8	7.8E + 08	9	1.3E + 09
WSV216 (VP124)	131.88	70	1.3E + 09	50	4.8E + 08	62	8.5E + 08
WSV209 (VP187, PIF1)	174.14	79	1.2E + 09	58	4.3E + 08	65	7.5E + 08
WSV442 (VP95)	89.37	42	1.1E + 09	38	5.0E + 08	39	7.0E + 08
WSV115 (VP53B, PIF0)	108.14	42	1.1E + 09	31	3.4E + 08	31	6.0E + 08
WSV421 (VP28)	22.13	7	1.1E + 09	7	2.9E + 08	7	5.0E + 08
WSV035 (VP110, PIF2)	108.30	62	9.1E + 08	50	3.2E + 08	54	6.7E + 08
WSV325 (VP56)	51.04	29	7.6E + 08	18	2.6E + 08	24	5.1E + 08
WSV134	24.98	11	7.4E + 08	8	2.0E + 08	9	3.3E + 08
WSV198 (VP32)	31.37	13	6.8E + 08	10	1.6E + 08	11	3.2E + 08
WSV021	22.78	4	6.8E + 08	5	2.3E + 08	5	4.1E + 08
WSV293a (VP14)	10.94	10	6.5E + 08	9	2.3E + 08	11	3.8E + 08
WSV136	14.54	11	5.8E + 08	6	7.6E + 07	8	1.7E + 08
WSV306 (VP39A, PIF3)	47.44	16	2.1E + 08	4	3.2E + 07	10	5.6E + 07
WSV432	11.48	6	2.1E + 08	1	1.0E + 08	3	1.3E + 08
WSV271 (VP136)	134.52	34	1.1E + 08	24	4.6E + 07	32	5.9E + 07
WSV340 (VP31)	29.83	12	9.0E + 07	4	3.7E + 07	6	4.8E + 07
WSV011 (VP150, PIF5)	143.82	15	8.9E + 07	9	3.5E + 07	15	5.3E + 07
WSV339 (VP39)	32.01	12	7.1E + 07	6	3.0E + 07	7	4.6E + 07
WSV073	47.22	8	6.6E + 07	7	1.8E + 07	7	3.2E + 07
WSV321 (VP16)	13.12	4	5.9E + 07	3	1.9E + 07	4	4.3E + 07
WSV077	33.23	7	5.5E + 07	5	1.9E + 07	8	3.6E + 07
WSV327 (VP90)	96.06	28	5.3E + 07	11	2.1E + 07	21	3.1E + 07
WSV285	123.33	27	4.8E + 07	10	2.1E + 07	20	3.0E + 07
WSV338 (VP62)	48.19	5	4.6E + 07	7	1.8E + 07	7	2.6E + 07
WSV006	32.73	3	4.5E + 07	2	2.2E + 07	3	1.9E + 07
WSV076	31.53	11	3.7E + 07	7	1.8E + 07	9	2.9E + 07
WSV254	31.54	8	2.9E + 07	3	9.6E + 06	6	1.4E + 07
WSV414 (VP19)	13.23	2	2.8E + 07	1	6.7E + 06	2	1.3E + 07
WSV256 (VP52B)	43.14	9	2.5E + 07	3	8.8E + 06	3	1.5E + 07
WSV259 (VP38)	35.44	5	2.5E + 07	3	5.9E + 06	3	1.2E + 07
WSV143	259.05	12	2.2E + 07	6	8.0E + 06	5	1.1E + 07
WSV181	43.21	7	1.8E + 07	4	1.3E + 07	4	1.8E + 07
WSV269	56.30	6	1.3E + 07	4	5.1E + 06	4	8.9E + 06
WSV390	35.81	5	8.5E + 06	1	5.5E + 06	2	4.0E + 06

^aThe BN-PAGE isolation and LC/MS-MS were performed three times, and the viral proteins identified in all three tests are shown in the table. The number of unique peptides and ion peak area for each protein are listed; the proteins are ordered according to the peak area of test 1 from high to low.

^bProteins with antibodies successfully obtained are in bold and underlined.

Structure prediction suggests WSV134 as a PIF4 homolog.

The four newly identified viral proteins (VP124, WSV021, WSV134, and WSV136) of the WSSV PIF complex showed no obvious homology to any of the proteins in the databases using BLAST. We then used AlphaFold2 (39) to predict the folding structures of these four proteins along with PIF homologs from other NALDV. Interestingly, we found that WSV134 has a predicted domain that is highly similar to that of PIF4 homologs from baculoviruses, nudiviruses, bracoviruses, and Apis mellifera filamentous virus (AmFV) (Fig. 3A). This domain comprises seven antiparallel β-sheets (β1 to β7) surrounded by two to three α helixes. β6 is linked to the loop between β4 and β5 by a predicted conserved disulfide bond, so that β7 folds back antiparallel with β1 (Fig. 3A). There is no similar domain in the Worldwide Protein Data Bank search by an online Dali server (http://ekhidna2.biocenter.helsinki.fi/dali/); therefore, we assigned this domain as “PIF4 domain.” The topology of the PIF4 domain was highly similar among viruses from different families (Fig. 3A). The predicted secondary structure of PIF4 is also well conserved among NALDVs, and the two cysteines that potentially form a disulfide bond are 100% conserved (Fig. 3B). Based on the above data, we propose that WSV134 is a PIF4 homolog. Similar analyses were conducted for VP124, WSV021, and WSV136; however, no convincing predictions could be made for these proteins.

FIG 3.

Structural prediction and bioinformatics analyses indicate WSV134 as a PIF4 homolog. (A) Predicted structure of PIF4 homologs from five representative invertebrate large DNA viruses. The PIF4 homolog structures were predicted by AlphaFold2 and visualized with PyMOL, and the PIF4 domain is shown in red and is framed (upper panel). The secondary structure topology of the PIF4 domain is illustrated in the lower panel with α helix in red, β sheet in blue, and a red broken line indicating the disulfide bond. (B) Secondary structure and sequence alignment of the PIF4 domain from representative invertebrate large DNA families. Sequence alignment was performed by ClustalW and exported by an ESPript 3.0 on-line server. AcMNPV PIF4 secondary structure is shown on top, and PIF4 domain containing sequence alignment is listed below. Conserved and semiconserved residues are presented in white font against a red background or red font framed in a blue box. Virus abbreviations and PIF4 sequence accession numbers are the following: AcMNPV, Autographa californica multinucleopolyhedrovirus, NP_054126.1; HearNPV, Helicoverpa amigera nucleopolyhedrovirus, NP_075154.1; CpGV, Cydia pomonella granulovirus, NP_148873.1; NeleNPV, Neodiprion lecontei nucleopolyhedrovirus, YP_025257.1; CuniNPV, Culex nigripalpus nucleopolyhedrovirus, NP_203393.1; GbNV, Gryllus bimaculatus nudivirus, YP_001111354.1; Hz2NV, Helicoverpa zea nudivirus-2, YP_004956787.1; OrNV, Oryctes rhinoceros nudivirus, YP_002321344.1; PmNV, Penaeus monodon NV, YP_009051934.1; CcBV, Cotesia congregata bracovirus, CAR31579.1; CiBV, Chelonus inanitus bracovirus, CAR40196.1; MmBV, Microplitis mediator bracovirus, AYH52229.1; AmFV, Apis mellifera filamentous virus, YP_009165908.1; WSSV, white spot syndrome virus, YP_009220517.1; CoBV, Chionoecetes opilio bacilliform virus, GAV93170.1; MjENV, Marsupenaeus japonicus endogenous nimavirus, GBG35418.1; PmENV, Penaeus monodon endogenous nimavirus, GBG35529.1; MeENV, Metapenaeus ensis endogenous nimavirus, GBG35448.1; LvENV, Litopenaeus vannamei endogenous nimavirus, ASM378908v1.

Stability of WSSV PIF complex under different conditions.

Unlike baculovirus, where the primary infection is initiated in the alkaline midgut of insects, the infection of WSSV in the aquatic environment appears to be neutral, but in higher salt conditions. Baculovirus PIF complexes are resistant to alkaline and proteolytic decay. This makes investigating the possible resistance of the PIF complex of WSSV to proteolytic decay under neutral and alkaline conditions compelling. To this end, the envelope proteins of WSSV were treated with proteinase K (40 ng/mL) at neutral and alkaline pH for 0 to 30 min, and the degradation of the PIF complex was investigated. In BN-PAGE, the ~720-kDa complex band could still be detected by a PIF1 antibody under different pH conditions within 10 min digestion, and a subcomplex containing PIF1 was retained even after 30 min digestion; however, VP28 multimers tend to be degraded under proteinase treatment and could hardly be detected after alkaline proteolytic digestion (Fig. 4A). Without proteinase K, PIF complex and VP28 multimers were not degraded under pH 7.0 or 11.0 even after 30 min incubation (data not shown). These results suggest that the WSSV PIF complex is resistant to proteolytic attack under both neutral and alkaline conditions.

FIG 4.

PIF complex is tolerant to variant physiochemical environments. (A) Stability of WSSV PIF complex under different pH and protease treatments. The WSSV envelope fraction was exposed to neutral (pH 7) or alkaline environment (pH 11) with proteinase K (PK, 40 ng/mL). After incubation for 0, 5, 10, and 30 min at 37°C, the proteinase was inactivated by adding proteinase inhibitor, and the sample was analyzed for protein complex with blue native PAGE following Western blot with antibodies against PIF1 or VP28. (B) Salt tolerance of PIF complex from WSSV and AcMNPV. Virion envelope fraction was treated with 0, 150, and 300 mM NaCl individually, and further incubated with or without proteinase K at 37°C for 5 min. Proteinase inhibitors were added to stop the digestion. Samples were analyzed for protein complex with PIF1 or VP28 antibodies. The WSSV and AcMNPV PIF complex signal is indicated with an arrow, and the VP28 associated complex with a half bracket.

Considering the salty and brackish water of shrimp habitats, determining the salt tolerance of the WSSV PIF complex is important. The salinity of seawater used for shrimp culture in this study is 17 to 20‰, i.e. 290 to 340 mM. The virus envelope fraction was treated with three different dosages of sodium chloride (NaCl: 0, 150, and 300 mM) with or without proteinase K digestion, followed by BN-PAGE and complex detection by Western blotting using PIF1 or VP28 antibodies. The 720-kDa complex band could be clearly detected under these salt concentrations, although the amount appeared to be slightly reduced at 300 mM upon addition of proteinase K. The signal of the VP28 multimeric complex was noticeably reduced after proteinase digestion in a saline environment (150 to 300 mM NaCl) (Fig. 4B). These data suggest that the WSSV PIF complex is likely to be stable in salty seawater, which might be compatible with its function in initiating viral infection in shrimp.

Based on these findings, the question of the similarity in the salt tolerance of the baculovirus PIF complex arose. To this end, the AcMNPV ODV envelope was fractionated and subjected to different salt concentrations with or without proteinase K treatment. The antibody against AcMNPV PIF1 probed the ~500-kDa baculovirus PIF complex even after treatment with 300 mM salt and proteinase K (Fig. 4B, right panel). Taken together, the WSSV PIF complex is both alkaline and salt stable, and this physiochemical property appears to be conserved during evolution.

WSSV PIFs are involved in host oral infection.

Next, we sought to determine whether the WSSV PIF complex, similar to the baculovirus PIF complex, plays an important role in entry into shrimp. To address this issue, we employed an antibody neutralization method. The purified WSSV virions were first incubated with individual polyclonal antibodies against PIF0, PIF1, PIF2, or PIF3 and then subjected to Litopenaeus vannamei per os inoculation as described in Materials and Methods. At 36 h post infection (p.i.), the viral load in the hemolymph or gills was analyzed by qPCR. As shown in Fig. 5A, the antibody against PIF1 or PIF3 reduced the viral load both in the gill (left panel) and hemolymph (right panel), implying that PIFs as part of the complex play a role in the oral infection of shrimp. To determine whether the neutralization result was specific, a series of different dilutions of PIF1 and PIF3 antibodies were used for neutralization, and the results showed a dose-dependent effect of PIF1 and PIF3 antibodies on the inhibition of virus replication (Fig. 5B). These data confirm that PIFs play a role in virus entry per os.

FIG 5.

Neutralization by PIF1 or PIF3 antibody reduces the oral infectivity of WSSV. (A) Freshly purified WSSV virions were incubated with control serum (CS) or antiserum against PIFs or VP28 (dilution 1:5) at 25°C for 1 h, then delivered into shrimp per os, 4 to 6 shrimp each group. At 36 h p.i., shrimp gill and hemolymph were harvested for DNA extraction, and WSSV viral genomes were quantified by qPCR. The column represents the mean value ± SD; comparison of mean values between serum-treated WSSV group and WSSV group was analyzed by unpaired two-tailed t test. *, P < 0.05; ns, no significant difference. (B) Dose-dependent inhibition of PIF1 and PIF3 antibody on virus load. A serial dilution (1:2 to 1:16) of PIF1 and PIF3 antiserum or control serum (CS, dilution 1:2) was incubated with freshly purified WSSV virions; at 36 h p.i., the virus load from gills and hemolymph was quantified by qPCR. Quantification is shown in mean ± SD (n = 4 to 6). *, P < 0.05 (unpaired two-tailed t test); ns, no significant difference.

DISCUSSION

In this study, we sought to uncover a candidate multiprotein complex of WSSV that might function in viral entry, especially in gut infection per os. Through isolation by BN-PAGE, protein identification by LC/MS-MS, and complex verification by Western blotting with specific antibodies, we found that, apart from PIF0-3, at least four additional proteins (VP124, WSV021, WSV134, and WSV136) are components of the WSSV PIF complex. A previous study analyzed the WSSV envelope protein complex using 2-D blue native/SDS-PAGE and identified an 832-kDa protein complex comprising VP53B (PIF0), VP187 (PIF1), VP110 (PIF2), and VP124 (40). Considering the overlapping protein content and the same order of magnitude in the size of both protein complexes, we believe that the WSSV PIF complex identified in this study is similar to the complex reported by Li et al. (40).

The WSSV PIF complex appears to be a large protein complex in the virion envelope according to BN-PAGE staining (Fig. 2A) (40), and contains at least eight proteins: PIF0 (VP53B, WSV115), PIF1 (VP187, WSV209), PIF2 (VP110, WSV035), PIF3 (VP39A, WSV306), PIF4 (WSV134), VP124 (WSV216), WSV021, and WSV136 (Fig. 6A). Similar to baculovirus (35), a PIF5 homolog protein (VP150, WSV011) is not a component of the WSSV PIF complex (Fig. 2B). The putative size of the complex (~720 kDa) was only measured by BN-PAGE and should be confirmed using other techniques. However, the size of the complex appeared to be significantly larger than the accumulated size of the eight components (PIFs) (~632 kDa). This may suggest the presence of unknown PIFs in the PIF complex or multimers of PIFs in the complex. Interestingly, the nine PIF-related genes, including PIF complex components and PIF5, are among the 11 ancestral genes in nimaviral genomes (14), suggesting that PIFs are associated with an ancient entry mechanism of invertebrate DNA viruses.

FIG 6.

Proposed PIF5 and PIF complex model in WSSV and NALDVs. (A) WSSV infects shrimp host via PIF proteins in its envelope. Among these, eight proteins (PIF0, PIF1, PIF2, PIF3, PIF4, VP124, WSV021, and WSV136) form a PIF complex (dash framed); one protein (PIF5) is distantly related. The topology of the proteins was predicted by TMHMM (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0). (B) PIF5 and PIF complex are common features among NALDVs. A maximum likelihood phylogenetic tree was built based on the concatenated aligned protein sequences of PIF0-3 of different NALDVs. Number on the nodes indicates maximum likelihood nonparametric bootstrap supports (500 replicates). The images of representative hosts, the viral family, and the number of PIF homologs are indicated on the top of the tree, and the proposed ancestral PIF complex and PIF5 are placed at the root of tree.

Although oral infection is one of the major routes for the natural spread of WSSV, the detailed molecular entry mechanism remains largely elusive. Two major viral envelope proteins, VP24 and VP28, may play a role in oral infection. VP24 was found to bind chitin and a polymeric immunoglobulin receptor (pIgR)-like protein, thus facilitating viral binding in the digestive tract (41) and subsequent viral internalization (42). VP28 was reported to interact with various host factors, as well reviewed by Verbruggen et al. (43). Oral delivery of recombinant VP28 protein can inhibit WSSV infection (44); however, its mechanism is unclear. We used antibodies against WSSV PIFs and VP28 to treat purified WSSV virions, and subsequently tested for oral infectivity. Antibodies against PIF1 and PIF3 showed a dose-dependent inhibitory effect on WSSV replication in vivo, but the antibody against VP28 had no effect (Fig. 5). Therefore, our data indicate that WSSV PIF1 and PIF3 are likely involved in oral infection. WSSV PIF1 (VP187, WSV209) binds to β-integrin, which is generally distributed in various cell types (45). This suggests that PIF1 may serve as a binding protein in oral infections. These results suggest that the WSSV PIF complex is functional during WSSV infection.

Given the various habitats of their invertebrate hosts, the PIF homolog-containing viruses experience very different physiological conditions, such as high salt and neutral pH conditions in the case of WSSV and crustaceans (46, 47), whereas baculoviruses have an alkaline and proteolytic environment in insect guts. The mechanism through which PIF sustains various environmental conditions has raised our interest. The integrity of the baculoviral PIF complex is essential for its function, and prevents proteinase digestion of virions released in insect midguts (35, 36). Surprisingly, the PIF complex from WSSV was stable under both neutral and alkaline pH conditions (Fig. 4A). We believe that the wide pH stability of the PIF complex has a selective advantage for viruses evolving from aquatic to terrestrial environments. In addition, the WSSV PIF complex is salt tolerant, as is the baculovirus PIF complex (Fig. 4B). It is hypothesized that the ancestral PIF complex already harbored the physiological properties of general tolerance to wide ranges of pH and salinity. In contrast, the major envelope protein, VP28, and its associated complex were less stable in saline environments (Fig. 4B) and were more sensitive to protease K digestion in both neutral and alkaline conditions (Fig. 4A). Our data suggest that VP28 tends to be degraded in the salty and digestive guts of shrimp and question its intrinsic role in WSSV oral infection (44). VP28 may somehow protect the PIF complex and make it available for interaction with receptors after degradation. Collectively, our data strongly suggest that the PIF complex mediates the oral infection of WSSV in shrimp (Fig. 6A).

The multicomponent functional PIF complex found in the WSSV expands our insight into PIF evolution. Different numbers of PIF homologs have been identified in NALDVs. Viruses from Hytrosaviridae that infect dipteran hosts, and AmFV, which infects honey bees, contain six PIF homologs, including PIF0-5 (14, 28 to 30, 48, 49). Bracoviruses, which have mutualistic symbioses with wasps and viruses from Nudiviridae that infect invertebrate hosts from Coleoptera, Diptera, and Lepidoptera, contain eight PIF homologs (PIF0 to 6 and PIF8) (27, 31, 32). Baculoviruses that infect insects of Lepidoptera, Hymenoptera, and Diptera contain 10 PIFs (PIF0-9). The PIF5 and PIF complex-mediated entry mechanism appears to be a common feature of these NALDVs (Fig. 6B). Previously, sequence homology failed to identify WSV134 as a PIF4 homolog, but structural homology now identifies it; it is referred to as WSSV PIF4. The other putative PIFs (VP124, WSV021, and WSV136) may have evolved further away from a common ancestor, implying that other unidentified PIF homologs may exist in distinctly related viruses such as AmFV and hytrosaviruses.

The PIFs-mediated entry mechanism is unknown and may be relatively more complicated, as most viruses utilize one or two proteins for host cell entry. We came up with a few possible points to describe the property of such a conserved entry mechanism. (i) The entry is mediated by multiple proteins, including PIF5 and the multicomponent conserved PIF complex; PIF5 and PIF complex may function at different entry steps, and work in concert to initiate a successful virus infection. (ii) The stability of PIF5 and PIF complex under the proteolytic environment is important for their function. We recently resolved the 3D structure of AcMNPV PIF5 and found that the stability of the protein under alkaline proteolytic conditions is critical for its function (50). Likewise, we showed that the PIF complex is stable in a variety of physiochemical environments, making it a heritable trait driving the evolution of NALDVs over time, space, and environmental conditions. (iii) PIF0 appears to be cleaved, as shown in WSSV (Fig. 1B) and in baculoviruses (37). As many viral fusion proteins are cleaved to expose fusion peptide to mediate virus-host membrane fusion during entry, whether PIF0 is the fusion protein has not been determined. We believe that elucidation of the PIF complex-mediated virus entry mechanism will increase insight into NALDV entry and identify novel targets for disease control of WSSV in shrimp farming.

MATERIALS AND METHODS

Production and purification of WSSV for mass spectrometry.

WSSV was produced in Procambarus clarkii as previously reported (51). Briefly, a sample of the WSSV stock was injected into crayfish P. clarkii. After 7 days, the hemolymph of the dying crayfish was freshly collected and subjected to sucrose gradient (30 to 60% wt/vol) ultracentrifugation (Beckman Coulter SW28 rotor, 20,000 rpm for 1 h at 4°C) for virus purification. The virus band was collected at the interface of 40 to 50% sucrose, sedimented at 20,000 rpm for 45 min, suspended in 50 mM Tris-HCl (pH 7.4), and stored at −80°C until use. To separate the envelope and nucleocapsid, WSSV virions were treated with 1% NP40 at 4°C for 30 min, followed by centrifugation at 20,000 × g for 30 min. The supernatant was collected as the envelope fraction, and the nucleocapsid-containing pellet was resuspended. Samples of purified virions and nucleocapsids were loaded onto a copper grid and negatively stained with 2% (wt/vol) phosphotungstic acid for examination using transmission electron microscopy (HITACHI H-7000FA). The purified virions were used for identification of the PIF complex and LC/MS-MS protein analysis.

Antibodies.

The WSSV proteins (in bold and underlined in Table 1) were expressed using the pET28a vector and purified from E. coli BL21. Monospecific polyclonal antibodies were obtained after the immunization of rabbits. Ac PIF1 antibody was generated previously (52).

BN-PAGE and LC-MS/MS analysis.

The virion envelope fraction was extracted and analyzed by BN-PAGE followed by LC-MS/MS as previously described (35). Briefly, the virion suspension was treated with 1% NP-40 detergent, and the envelope fraction was analyzed using BN-PAGE, according to the manufacturer’s instructions (Thermo Fisher Scientific). Guided by PIF0-3 antibody probing, the visible PIF complex band from Coomassie blue staining was sliced out, digested in-gel, and further analyzed by LC-MS/MS.XX Proteins were identified by searching against the WSSV-CN database (accession number AF332093) of NCBI.

Stability assays of PIF complex under different conditions.

To test the alkaline sensitivity of the PIF complex, the WSSV envelope fraction was exposed to pH 7.0 or pH 11.0 by adding 1/4 volume of 4 × PBS (pH 7.0 or pH 11.0 adjusted), followed by proteinase K (40 ng/mL) digestion at 37°C for 0, 5, 10, and 30 min. The digestion was stopped by adding soybean trypsin inhibitor (SBTI) (Sigma-Aldrich, St. Louis, MO, USA) and cOmplete Protease Inhibitor Cocktail (Roche, Basel, Switzerland). For the salt tolerance assay, the AcMNPV ODV envelope fraction was prepared as described previously (35), and the envelope fraction of WSSV or AcMNPV ODV was treated with 0, 150, and 300 mM NaCl (pH 7.0). The sample was further incubated with or without proteinase K at 37°C for 5 min, and digestion was terminated as described above. After all treatments, the samples were adjusted using a 4 × BN-PAGE sample buffer and analyzed using BN-PAGE. The WSSV PIF complex, VP28, and AcMNPV PIF complex were probed using antibodies against WSSV PIF1, VP28, and Ac PIF1, respectively.

Antibody neutralization assay for WSSV per os infection.

For the per os infection assay, the strain WSSV-CN03 (53) was used and produced in Cherax quadricarinatus and purified from dying crayfish tissues as previously described (54). The purified WSSV particles were stored in TMN buffer (50 mM Tris-HCl, 5 mM MgCl₂, 150 mM NaCl, pH 7.5) at 4°C and used within 1 week. Virus concentration was measured by spectrophotometry (55), C (virions/μL) = f OD₆₀₀ = 3.34 × 10⁸ × OD₆₀₀.

Whiteleg shrimp (Litopenaeus vannamei) with an average weight of 10 g and body length of 12 cm were maintained in seawater with 1.7 to 2% salinity at 25°C, supplemented with an air pump. WSSV (1 × 10⁸ virions/20 μL/shrimp) was delivered into the lumen of the esophagus using a flexible silicone tube, as described previously (41). For the antibody-neutralizing assay, the virions were preincubated with antiserum against individual PIFs or VP28 at 25°C for 1 h. Nonimmune rabbit serum was used as the negative control. Four to six shrimp were used in each group.

Hemolymph (10 μL) and gills (10 mg) from shrimp isolated at 36 h p.i. were used for total DNA extraction. The viral load in each sample was measured by qPCR using the WSSV fluorescent quantitative PCR detection kit (Xiamen Lulong Biotech Co., Ltd., China). The amplification reactions were performed as follows: denaturation at 95°C for 2 min, followed by 40 cycles at 94°C for 10 s and 60°C for 30 s. A t test was performed to calculate the P value, and P < 0.05 was considered significant.