GSK3β Plays a Negative Role During White Spot Syndrome Virus (WSSV) Infection by Regulating NF-κB Activity in Shrimp Litopenaeus vannamei

Glycogen synthase kinase-3 (GSK3), a cytoplasmic serine/threonine-protein kinase involved in a large number of key cellular processes, is a little-known signaling molecule in virus study. In this study, a GSK3 protein which was highly similar to GSK3β homologs from other species in Litopenaeus vannamei (designated as LvGSK3β) was obtained. LvGSK3β was expressed constitutively in the healthy L. vannamei, at the highest level in the intestine and the lowest level in the eyestalk. White spot syndrome virus (WSSV) reduced LvGSK3β expression was in immune tissues including the hemocyte, intestine, gill and hepatopancreas. The inhibition of LvGSK3β resulted in significantly higher survival rates of L. vannamei during WSSV infection than the control group, and significantly lower WSSV viral loads in LvGSK3β-inhibited L. vannamei were observed. Knockdown of LvGSK3β by RNAi resulted in increases in the expression of LvDorsal and several NF-κB driven antimicrobial peptide (AMP) genes (including ALF, PEN and crustin), but a decrease in LvCactus expression. Accordingly, overexpression of LvGSK3β could reduce the promoter activity of LvDorsal and several AMPs, while the promoter activity of LvCactus was increased. Electrophoretic mobility shift assays (EMSA) showed that LvDorsal could bind to the promoter of LvGSK3β. The interaction between LvGSK3β and LvDorsal or LvCactus was confirmed using co-immunoprecipitation (Co-IP) assays. In addition, the expression of LvGSK3β was dramatically reduced by knockdown of LvDorsal. In summary, the results presented in this study indicated that LvGSK3β had a negative effect on L. vannamei by mediating a feedback regulation of the NF-κB pathway when it is infected by WSSV.

Glycogen synthase kinase-3 (GSK3), a cytoplasmic serine/threonine-protein kinase involved in a large number of key cellular processes, is a little-known signaling molecule in virus study. In this study, a GSK3 protein which was highly similar to GSK3b homologs from other species in Litopenaeus vannamei (designated as LvGSK3b) was obtained. LvGSK3b was expressed constitutively in the healthy L. vannamei, at the highest level in the intestine and the lowest level in the eyestalk. White spot syndrome virus (WSSV) reduced LvGSK3b expression was in immune tissues including the hemocyte, intestine, gill and hepatopancreas. The inhibition of LvGSK3b resulted in significantly higher survival rates of L. vannamei during WSSV infection than the control group, and significantly lower WSSV viral loads in LvGSK3b-inhibited L. vannamei were observed. Knockdown of LvGSK3b by RNAi resulted in increases in the expression of LvDorsal and several NF-kB driven antimicrobial peptide (AMP) genes (including ALF, PEN and crustin), but a decrease in LvCactus expression. Accordingly, overexpression of LvGSK3b could reduce the promoter activity of LvDorsal and several AMPs, while the promoter activity of LvCactus was increased. Electrophoretic mobility shift assays (EMSA) showed that LvDorsal could bind to the promoter of LvGSK3b. The interaction between LvGSK3b and LvDorsal or LvCactus was confirmed using co-immunoprecipitation (Co-IP) assays. In addition, the expression of LvGSK3b was dramatically reduced by knockdown of LvDorsal. In summary, the results presented in this study indicated that LvGSK3b had a negative effect on L. vannamei by mediating a feedback regulation of the NF-kB pathway when it is infected by WSSV.

INTRODUCTION
Shrimp is one of the main internationally traded aquatic products, which has important economic value. As an important aquaculture shrimp species, the produce of Litopenaeus vannamei accounts for nearly two-thirds of the total output of shrimp worldwild (1). In recent years, emerging diseases caused by various pathogens have been the main menace for sustainable development of the shrimp industry worldwide. Among these pathogens, white spot syndrome virus (WSSV) has always been a viral agent of greatest concern, although other viruses such as Taura syndrome virus (TSV), infectious hypodermal and hematopoietic necrosis virus (IHHNV) and decapod iridescent virus 1 (DIV1) were reported to cause losses in farmed shrimp (2)(3)(4)(5). As an invertebrate, the innate immunity is the first defense line against viral infections in shrimp (6). Although a considerable number of scientific studies have focused on this for several years, knowledge of the host-virus interaction is quite insufficient. Thus, it is in urgent need to investigate the mechanisms of immune defence for disease controling in shrimp culture.
Glycogen synthase kinase 3 (GSK3), a multifunctional serine/ threonine-protein kinase, initially identified as a key regulator of glycogen metabolism and insulin signaling, has been verified as being involved in various physiological and pathological processes, such as protein synthesis, signal transduction, cell proliferation and differentiation, immune response, inflammation and tumorigenesis, and so on (7,8). GSK3 is highly conserved in species from the amoeba, Dictyostelium discoideum, to humans. In vertebrates, there are two isoforms of GSK3 (GSK3a and GSK3b) known, which share a highly conserved catalytic domain but differ at both termini (9). GSK3a and GSK3b perform different functions in many biological processes. GSK3a is mainly involved in glycogen metabolism, while GSK3b functions as a convergence point for multiple signaling pathways, including nuclear factor kB (NF-kB), Wnt/ b-catenin, phosphatidylinositide 3-kinase/protein kinase B (PI3K/Akt), Hedgehog and STAT signaling pathways (10)(11)(12). Recent studies have unearthed GSK3b's role in viral infections through its alteration of viral entry, replication and egress. GSK3b is known as an AKT substrate which plays a crucial role in influenza viral entry (13,14). GSK3b silence could decrease the hepatitis C virus (HCV) replication and the production of infectious particle while GSK3b overexpression enhanced HCV replication (15). Studies in vitro displayed that specific inhibitors (small molecules) of GSK3b significantly decreased Tat-dependent replication of human immunodeficiency virus 1 (HIV1) (16). In addition, GSK3b inhibitors could effectively prevent the assembly and release of HCV because of the key role in HCV virion assembly and release GSK3b played, through inhibiting apolipoprotein synthesis (17). But overall, studying on the function of GSK3b and its related signaling pathways in the research field of virology is still in its initial stage, especially in invertebrates.
GSK3b has been demonstrated to play a role in the interaction between host and virus in shrimp, but studies on its specific regulation mechanism during the anti-viral immune reaction in shrimp is required, as GSK3b is a cross point in several signal pathways (18). An important pro-inflammatory pathway promoted by GSK3b is the activation of NF-kB (19). GSK3b plays a vital role in gene transcription by the way of regulating the promoter-specific recruitment of NF-kB (20). Through restraining the transcriptional activation of NF-kB, GSK3b inhibitors could lessen the inflammatory cytokine production of the TLR pathway (21). Consistently, activation of the PI3K/Akt/GSK3b signaling pathway reduced the NF-kB nuclear translocation and the pathway acts as the influencing factor in inflammatory protection (22). NF-kB plays an essential role in the antiviral immunity and is an obstacle to the survival of virus (23). The Toll and immune deficiency (IMD) pathways are Two NF-kB related signaling pathways which are considered as the main regulatory factors of the anti-viral immune response in shrimp (24). Building on the current research, the mechanism behind GSK3b mediation of NF-kB during WSSV infection in L. vannamei was studied in this paper. The results showed that L. vannamei GSK3b (LvGSK3b) has a negative effect for the shrimp during WSSV infection by regulating NF-kB activity, which is beneficial for the study on the host-virus interactional mechanisms and may offer potential strategies for disease prevention in shrimp industry.

Shrimp and WSSV
L. vannamei (approximately 4-6 g each) were purchased from a local shrimp farm in Zhanjiang City, Guangdong Province, China. As our previous study (25) described, L. vannamei were cultured in a recirculating water tank system for over 7 days before the experiments with water salinity and temperature maintained at 27‰ salinity at 25-27°C. The shrimp were fed with a commercial shrimp pellet diet (Haid Group, Guangzhou, China) twice daily. At the beginning of all experimental treatments, shrimp (5% of the total) were analyzed and ensured to be free of WSSV via PCR method performed according to published standard operating procedures. The WSSV (Chinese strain, AF332093) was extracted from WSSV-infected shrimp muscle tissue and stored at −80°C. Before injection, the muscle tissue was homogenised and prepared as a WSSV inoculum with approximately 1×10 5 copies/µl, as previously described (26).

Cloning the cDNA and Genome of LvGSK3b
Total RNA and the genomic DNA were extracted from the shrimp tissues using the RNeasy Mini kit (Qiagen, Hilden, Germany) and TIANGEN Marine Animals DNA Kit (TIANGEN, Guangzhou, China), respectively. First-strand cDNA synthesis was performed using a cDNA Synthesis Kit (Takara, Dalian, China) following the manufacturer's instructions. The partial cDNA sequence of LvGSK3b was obtained from transcriptomic sequencing of L. vannamei (27) and its full-length cDNA sequence was cloned by a rapid amplification of cDNA ends (RACE) PCR using the SMARTer ™ RACE cDNA Amplification kit (Clontech, Japan), following the user manual and the specific primers used are listed in Table 1. The genomic DNA sequences of LvGSK3b were obtained via PCR, according to a previously published method (28) and were based on the genome sequences of Penaeus vannamei breed Kehai No.1 (GenBank No. NW_020869560). The PCR products were cloned into the pMD-20 vector (Takara, Japan) and sequenced. The gene sequences obtained in this study have been deposited in the NCBI GenBank (http://www. ncbi.nlm.nih.gov/genbank/).

Genomic Walking
The 5' flanking regulatory regions of LvGSK3b were cloned by genome walking PCR amplification via the GenomeWalker Universal Kit (Clontech, Dalian, China), according to the methods of a previous paper (29). Two pairs of primers, AP1/LvGSK3b-R1 and AP2/LvGSK3b-R2, were used to perform the primary and nested genome walking PCR amplification. The PCR products were cloned into the pMD-20T vector (Takara, Dalian, China) and sequenced. Primers are listed in Table 1.

Bioinformatics Analyses
The protein domain topologies and the genome sequences were analysed using the SMART program (http://smart.emblheidelberg.de/) and Splign program (https://www.ncbi.nlm.nih. gov/sutils/splign/splign.cgi). Protein sequences of GSK3b homologues from other species were selected from the NCBI databases. Multiple sequence alignments were performed using the ClustalX 2.0 program (http://www.ebi.ac.uk/tools/clustalw2) and visualized using GeneDoc. The similarities among the GSK3b proteins were calculated in GeneDoc. The neighborjoining phylogenetic trees were constructed based on the amino acid sequences using MEGA 5.0 (http://www.megasoft-ware.net/ download_form), applying the Poisson distribution substitution

Tissue Distribution, Immune Challenge, and Gene Expression Analyses
Eleven tissues, including the hemocytes, hepatopancreas, gill, heart, stomach, pyloric caecum, nerve, epithelium, eyestalk, intestine and muscle, were obtained from 10 healthy L. vannamei for RNA extraction. For the challenge experiments, 60 healthy L. vannamei divided into two groups (n = 30) were intramuscularly injected with WSSV (10 6 copies/g) in the third abdominal segment. L. vannamei injected with PBS were used as controls. At 0, 4, 8, 12, 24, 48, and 72 h post-injection (hpi), three animals from each group were randomly sampled for hemocyte, hepatopancreas, gill and intestine collection. Total RNA extraction, reverse transcription and real-time PCR (qPCR) analysis were performed as detailed in previous research (30). The qPCR was performed in a Roche Light Cycler480 thermal cycler (Roche Applied Science, Germany). Elongation factor 1a (EF1a, GenBank accession No. GU136229) was used as the internal control. The expression level of LvGSK3b was calculated using the Livak(2 −DDCt ) method after normalization to LvEF1a. All samples were tested in triplicate. Primer sequences are listed in Table 1.

RNA Interference
As described in our previous studies (31,32), the doublestranded RNAs (dsRNAs) specific to LvGSK3b, LvDorsal and GFP (as control) were synthesized using an in vitro transcription method, by using the T7 RiboMAX ™ Express RNAi System (Promega, USA), and the quality and amount of dsRNA were checked after annealing via gel electrophoresis. For the RNA interference efficiency determination, healthy L. vannamei were injected with 50 µl PBS containing 5 µg LvGSK3b, LvDorsal or GFP dsRNA. At 24 and 48 hpi, the hemocytes from each group were sampled for qPCR to detect the RNAi efficiency, the effects of LvGSK3b on the expression of LvDorsal, LvCactus and nine antimicrobial peptides (AMP) genes and the effects of LvDorsal on the expression of LvGSK3b in L. vannamei. Primer sequences are listed in Table 1.

WSSV Challenge Experiments in LvGSK3b-Inhibited L. vannamei
LvGSK3b was inhibited both on the enzyme activity and mRNA expression levels by injecting with the specific inhibitor and dsRNA. Specific inhibitor of LvGSK3b, named TWS-119 (Selleck Chemicals, USA) was used with the dose refer to the Inhibitory concentration 50 (IC50) (33). At 4, 8 and 12 hpi, the hemocytes from each group were sampled for the detection of inhibition efficiency of LvGSK3b activities, using the GSK3b Kinase Activity Quantitative Detection Kit (Genmed, China). The expression of LvGSK3b was inhibited by RNAi as described above in Section RNA Interference. For the mortality test, L. vannamei were injected with 50 µl PBS containing or not containing 10 6 WSSV after 8 h of TWS-119 injection or 48 h of dsRNA injection. Cumulative mortality was recorded every 4 h and the differences between groups were analysed using the Mantel-Cox (log-rank c 2 test) method with the software GraphPad Prism. To investigate the genome copies of WSSV, total DNA was extracted from muscle at 48, 72 and 96 hpi and absolute real-time PCR was performed using primers WSSV32678-F/WSSV32753-R and a Taq Man fluorogenic probe, as in a previous study (34).

Dual-Luciferase Reporter Assays
The gene-specific primers were designed to amplify the ORF of LvGSK3b without introducing a stop codon. The amplified product was cloned into a pAc5.1/V5-His A vector (Invitrogen) to allow subsequent expression of V5-tagged LvGSK3b proteins. All of the vectors with reporter genes used in this paper, including Dorsal, Cactus and AMPs in shrimp or Drosophila, were obtained from our previous studies (30,35).
Since no permanent shrimp cell line was available, Drosophila Schneider 2 (S2) cells were cultured for promoter activity analysis at 28°C in Schneider's Insect Medium (Sigma, USA) containing 10% fetal bovine serum (Gibco, USA) for the in vitro experiments. After 24 h of culturing in a 96-well plate, the S2 cells of each well were transfected with 0.05 mg of firefly luciferase reporter-gene plasmids, 0.01 mg of pRL-TK renilla luciferase plasmid (Promega, as an internal control) and 0.05 mg of (otherwise indicated) either pAc-LvGSK3b-V5 plasmids or pAc5.1-Basic plasmids (control). Each experiment was repeated in 6 wells. 48 hours post-transfection, the activity of the luciferases was detected in order to calculate the relative ratios of firefly and renilla luciferase activity using the Dual-Glo ® Luciferase Assay System kit (Promega, USA), according to the manufacturer's instructions.

Electrophoretic Mobility Shift Assay
An electrophoretic mobility shift assay (EMSA) was performed using a Light Shift Chemiluminescent EMSA kit (Thermo Fisher Scientific, Waltham, MA, USA) according to a previously published method (36). Briefly, the biotin-labeled or unbiotinlabeled probes were designed using the Dorsal binding motif sequence. The mutant probe was designed by deleting the Dorsal-binding motif sequence. All of the probes were synthesized by Life Technologies (Shanghai, China) and the sequences are listed in Table 1. RHD domain of Dorsal (Dorsal-RHD) was cloned into the modified pGEX-4T plasmid to get rDorsal-RHD-GST. GST tag of rDorsal-RHD-GST was then removed using rPorcine Enterokinase from Glutathione Resin Kit. The purification productions of rDorsal-RHD-GST and rDorsal-RHD were analyzed using SDS-PAGE and stained with Coomassie blue. Purified rDorsal-RHD protein (10 mg) was incubated with 20 fmol of the probes for the binding reactions. The reactions were separated on a 5% native PAGE gel, transferred to positively charged nylon membranes (Roche, Germany) and cross-linked by UV light. Then, the biotin-labeled DNA on the membrane was detected by chemiluminescence and developed on X-ray film, followed by enhanced chemiluminescence (ECL) visualisation (Tanon, Shanghai, China).

Co-Immunoprecipitation Assays
Co-immunoprecipitation (Co-IP) assays were performed as described in the previous study (36). To investigate the interaction of LvGSK3b with LvDorsal or LvCactus, pAc-LvGSK3b-GFP or pAc5.1-GFP (control) were co-transfected with individual plasmids along with pAc-LvDorsal-HA or pAc-LvCactus-HA into the S2 cells. Forty eight hours post transfection, cells were harvested and washed with ice-cold PBS three times, and then lysed in IP Lysis Buffer with a protease inhibitor cocktail (Sigma, USA). Both Co-IP and reciprocal Co-IP experiments were carried out using anti-HA affinity gel (Sigma, USA) and samples were subjected to SDS-PAGE assay. The precipitated protein was examined using western-blots, with rabbit anti-GFP antibody as the primary antibody, and alkaline phosphatase-conjugated goat anti-rabbit as the secondary antibody (Sigma, USA). A standardized aliquot (5%) of each total input cell lysates was also examined as control.

Statistical Analyses
The relative gene expression data were statistically analysed and all data were presented as mean ± SD. Student's t-tests were used to make comparisons among groups. For the mortality analysis, the data were statistically analysed using GraphPad Prism (Graphpad, San Diego, CA, USA) to obtain Kaplan-Meier plots (log-rank c 2 test).

Bioinformatics Analyses of LvGSK3b
As shown in Figure 1A,  Figure 1B). Analysis of the genome sequences showed that all of the exon-intron boundaries in LvMyD88 conform to the consensus GT/AG rule for splicing (37). SMART analysis revealed that the LvGSK3b protein contained one protein kinase domain named STKc_GSK3 between residues 54 and 338 ( Figure 1C). Multiple sequence alignment indicated that LvGSK3b shares a high similarity of more than 80% with GSK3b proteins from invertebrates to vertebrates, with the lowest homology (80%) to GSK3b from Aplysia californica and the highest homology (100%) to GSK3b from Penaeus monodon (Figure 2A), which is also a shrimp species. The neighbour-joining (NJ) phylogenetic tree showed that GSK3a and GSK3b were closely related to their own homologues from other species. All of the GSK3b proteins used in this study were separated into five clades including the crustacea, insecta, arachnids, molluscs and vertebrates groups ( Figure 2B). GSK3b from L. vannamei, P. monodon, Daphnia pulex and Eriocheir sinensis were clustered together in the crustacea group. L. vannamei GSK3b and P. monodon GSK3b showed the closest evolutionary relationship. As shown in Figure 3A, the promoter region of LvGSK3b (1,169 bp in length) contained two potential Dorsal binding sites (−700 to −711 bp and −962 to −973 bp) ( Figure 3B).

Expression Profiles of LvGSK3b in Healthy and WSSV Challenged L. vannamei
LvGSK3b was detected in all of the examined tissues ( Figure 4). The relative expression levels of LvGSK3b in the other tissues were normalised to that in eyestalk, which was the lowest and set as the baseline (1.0). LvGSK3b was expressed highest in the intestine, in which the expression level of LvGSK3b was 14.07fold over the baseline. As shown in Figure 4, LvGSK3b expression in the gill, muscle, nerve, heart, hepatopancreas and hemocyte were all nearly 5.00-fold over the baseline and not significantly different from each other.
Using the four important immune tissues, including hemocytes, hepatopancreas, intestine and gill, LvGSK3b expression after WSSV challenge was investigated using qPCR ( Figure 5). The expression of LvGSK3b was reduced to different degrees by WSSV challenge in the haemocytes, intestine and gill but not in the hepatopancreas. In the hepatopancreas there was no significant difference in LvGSK3b expression between the infection group and the control group at all of the sampled timepoints ( Figure 5B). In the hemocyte, LvGSK3b expression was significantly down-regulated starting at 4 hpi, reaching a low-point at 12 hpi at 25.21% of the control. At 72 hpi, LvGSK3b expression recovered to a level not perceptibly different from the control ( Figure 5A). In the intestine, the expression of LvGSK3b in the infection group was notably lower than the control group at 4, 8 and 72 hpi, with the lowest level at 8 hpi at 21.00% of the control ( Figure 5C). When compared with the control group, LvGSK3b expression in the gill of the infection group was significantly decreased at 12, 24 and 48 hpi by 72.02%, 54.78% and 65.16%, respectively ( Figure 5D).

Function of LvGSK3b in WSSV Infection
Refer to IC50 of TWS-119 in animals (33), a dose of 5 µg/g TWS-119 per shrimp was chosen for the LvGSK3b-inhibition test. As Figure 6A shows, the enzyme activity of LvGSK3b was significantly reduced at 4, 8 and 12 hpi, with the lowest level at 8 hpi, which was the timepoint chosen for the following WSSV challenge experiment. In the TWS-119 + WSSV group, the survival rates of L. vannamei were notably higher than that of the PBS + WSSV group starting at 56 hpi ( Figure 6B). At 124 hpi, the survival rate of L. vannamei in the TWS-119 + WSSV group was 43.60% but the L. vannamei in the PBS+WSSV group had all died. The final survival rate of L. vannamei at 144 hpi was 10.30% for the TWS-119 + WSSV group. Consistently, the WSSV genome copies in the muscle of the TWS-119 injected L. vannamei at 48, 72 and 96 hpi were significantly less than those in the PBS injected L. vannamei ( Figure 6C).    RNAi was also performed to investigate the role of LvGSK3b during WSSV infection. qPCR was used to check the silencing efficiency of LvGSK3b. As shown in Figure 6D, at 48 h after injection of LvGSK3b dsRNA, the expression levels of LvGSK3b at 24 and 48 hpi were significantly down-regulated to~0.43-fold and~0.21-fold of the GFP dsRNA injection group, respectively. Similarly as for the results when LvGSK3b were inhibited by TWS-119, the survival rates of L. vannamei in the LvGSK3b dsRNA + WSSV group were also significantly higher than that of the GFP dsRNA + WSSV group, and the final survival rate of L. vannamei at 144 hpi was 53.85 and 14.29% for the TWS-119 + WSSV group and GFP dsRNA + WSSV group, respectively ( Figure  6E). Furthermore, the viral loads of the LvGSK3b dsRNA + WSSV group were significantly lower than those of the GFP dsRNA + WSSV control group, with a~15.52-fold,~40.78-fold and~7.06fold decrease at 48, 72 and 96 h, respectively ( Figure 6F).

Influence of LvGSK3b on Expression of NF-kB Related Genes in L. vannamei
NF-kB has important roles in the regulation of AMP genes in shrimp (2). Considering the fact that the promoter of LvGSK3b contained two potential Dorsal binding sites, the expression of NF-kB related genes (Dorsal, Cactus and AMP genes) in LvGSK3b dsRNA injected shrimp were investigated to primarily explore the underlying mechanism of LvGSK3b during WSSV infection. When compared with the control GFP dsRNA, the LvGSK3b dsRNA significantly increased LvDorsal , gill (C) and intestine (D) from WSSV or PBS challenged L. vannamei. qPCR was performed in triplicate for each sample. Expression values were normalized to those of LvEF1a using the Livak (2 −DDCt ) method and the data were provided as the means ± SD of triplicate assays. The statistical significance was calculated using Student's t-test (*p < 0.05, **p < 0.01). expression but decreased LvCactus expression in the haemocyte of L. vannamei. Consistent with this, the expression of several AMP genes, including three antilipopolysaccharide factors (LvALF1, LvALF2, LvALF3), two penaeidins (LvPEN3, LvPEN4) and one crustin (LvCRU1) were significantly upregulated at 48 h after injection with LvGSK3b dsRNA ( Figure 7A).

Influence of LvDorsal on Expression of LvGSK3b in L. vannamei
The expression of LvGSK3b in LvDorsal dsRNA injected shrimp was also investigated. When LvDorsal was knocked down by dsRNA injection, the expression of LvGSK3b in the hemocyte of L. vannamei was significantly higher than the control GFP dsRNA injected group ( Figure 7B).

Regulatory Effects of LvGSK3b on the Promoter Activities of NF-kB Related Genes
Previous results in vivo showed that LvGSK3b could regulate the expression of NF-kB related genes, so the effects of LvGSK3b on the promoter activities of NF-kB related genes were further explored in vitro using dual-luciferase reporter assays. As shown in Figure 8,

Interaction Between LvDorsal and the LvGSK3b Promoter
To further investigate the interaction between LvDorsal and the LvGSK3b promoter, EMSA was performed using the purified 6His-tagged RHD domain of Dorsal protein (rDorsal-RHD) expressed in E. coli cells. As shown in Figure 9, rLvDorsal-RHD, but not the control rTrx, effectively retarded the mobility of the bio-labeled probes 1 and 2. The DNA/protein complex was markedly reduced by the competitive unlabeled probe 1 at the level of 10×, 50× and 100×. Contrarily, the DNA/protein complex was only faintly reduced by the competitive 10× and 50× unlabeled probe 2 but markedly reduced by the competitive 100× unlabeled probe 2. These results indicate the specificity of the interaction between rLvDorsal-RHD and the two probes.

Interaction of LvGSK3b With LvDorsal and LvCactus
To explore whether the interaction of LvGSK3b with LvDorsal or LvCactus occurs in L. vannamei, Co-IP analyses were carried out by using the S2 cell line. GFP-tagged LvGSK3b or GFP (as a control) were co-expressed with HA-tagged LvDorsal or HAtagged LvCactus, respectively. The results showed that both LvDorsal-HA and LvCactus-HA co-precipitated with GFPtagged LvGSK3b but not the control GFP ( Figures 10A, B), indicating an interaction of LvGSK3b with LvDorsal and LvCactus.

DISCUSSION
Previous studies on GSK3, concentrating on vertebrates, have shown that it is a multifunctional enzyme involved in glycogen metabolism, insulin signaling, inflammatory response and innate A B Drosophila S2 cells were transfected with the protein expression vectors (LvGSK3b and the pAc5.1 empty vector as a control), the reporter gene plasmid, and the pRL-TK Renilla luciferase plasmid (as an internal control). After 48 h, the cells were harvested for measurement of luciferase activity using the dualluciferase reporter assay system (Promega). The bars indicated the mean ± SD of the luciferase activity (n = 6). The statistical significance was calculated using Student's t-test (*p < 0.05, **p < 0.01).

A B
FIGURE 10 | LvGSK3b interacted with LvDorsal (A) and LvCactus (B). Co-immunoprecipitation (Co-IP) assays showed that the GFP-tagged LvGSK3b but not the control GFP protein can be co-precipitated by HA-tagged LvDorsal or LvCactus. Input: western-blotting analysis of the input cell lysates (5%) before immunoprecipitation. immunity (38). However, the function of GSK3 in invertebrates has rarely been reported. In the present study, a GSK3b homolog was cloned from L. vannamei (LvGSK3b) and was found to be strongly downregulated in vivo after WSSV infection, while the inhibition of LvGSK3b significantly increased the survival rate of L. vannamei in response to WSSV infection. The effects of inhibition and overexpression of LvGSK3b on NF-kB related genes were the opposite. In addition, interaction between LvDorsal and the LvGSK3b promoter, and the interaction of LvGSK3b with LvDorsal/LvCactus were revealed. All of the results indicated that LvGSK3b has a negative effect on the shrimp by negatively regulating NF-kB activity in L. vannamei when it is infected by WSSV. In vertebrates, GSK3 comprises two homologues, GSK3a and GSK3b, encoded by two similar genes, which share similar Serine/Threonine kinase domains but differ substantially in their termini (39). Unique to the GSK3a is an N-terminal domain consisting of 63 residues that is glycine-rich. Resembling the previously identified GSK3b, L. vannamei GSK3b has Ser9 and Tyr216 phosphorylation sites, a STKc_GSK3 domain but no glycine-rich extension at the Nterminus, and the NJ phylogenetic tree showed that LvGSK3b belonged to the cluster of GSK3b homologues from other species, which suggested that LvGSK3b was a member of the b isoform of the GSK3 family in L. vannamei. Based on the information on NCBI, most of the GSK3 genes identified were GSK3b in invertebrates including P. monodon, D. pulex and E. sinensis, and GSK3a was only reported in Drosophila species. In fact, only GSK3b could be found in L. vannamei through scanning the transcriptome and genomic data published on NCBI. The reason why GSK3b but not GSK3a is common in invertebrates has not yet been revealed.
In E. sinensis, GSK3b regulated hemocyte phagocytosis in the immune response (40). A previous study reported that LvGSK3b played an important role in shrimp immunomodulation, and shrimp might promote the apoptosis to restrain WSSV infection by inhibition of LvGSK3b (18). Similarly, it was also found that LvGSK3b expression decreased with WSSV challenge, and LvGSK3b inhibition by injection with a specific inhibitor or RNAi reduced the WSSV loads in L. vannamei in this study. More intuitively, our study showed that the survival rates of LvGSK3b inhibited L. vannamei after the WSSV challenge was greater. All of the results further verified the close relationship between LvGSK3b and WSSV infection, implying the positive function of LvGSK3b on WSSV replication. To our knowledge, this study was a rare report about the role GSK3b plays during viral infection in invertebrates. Comparatively speaking, many more studies about the effect of GSK3b on viral replication have been performed on vertebrates. In humans, there was a significant drop in GSK3b expression in response to Japanese encephalitis virus (JEV) infection (41), and the replication of HCV (15), HIV-1 (16), varicella zoster virus (VZV) (42) and severe acute respiratory syndrome coronavirus (SARS-CoV) (43) was decreased by silencing GSK3b. In addition, GSK3b functioned in inhibiting porcine epidemic diarrhoea virus (PEDV) replication in vitro (44). Thus, it can be seen that the effect of GSK3b on viral replication is consistent in vertebrates and invertebrates, indicating the functional conservatism of GSK3b in different animal species.
As an intersection of multiple cell signaling pathways such as Wnt/b-catenin, NF-kB and PI3K/AKT, the function of GSK3b in proliferation, differentiation, apoptosis, and immune responses could not be ignored (10)(11)(12). This raises difficulties for clarifying the role of GSK3b and the associated regulatory mechanisms regarding virology. Nevertheless, great progress has been made in humans. For example, the downregulation of GSK3b resulted in a long viral persistence of JEV by enhancing the stability of cyclin D1 (39). For the herpes simplex virus type 1 (HSV-1) -induced Kaposi's sarcoma-associated herpesvirus (KSHV) reactivation, the activation of the MAPK and the PTEN/PI3K/AKT/GSK3b pathways was required (45). VZV benefited its own survival and replication through activating AKT signaling that induces the phosphorylation of GSK3b, which in turn blocks Bad's proapoptotic pathway (40). In addition, the decrease of GSK3b expression can also inhibit replication of the influenza virus by positively regulating the virus-induced host cell apoptosis (46). Similar results were presented in the functional study of GSK3b in shrimp. LvGSK3b play an important role in WSSV clearance by mediating apoptosis (18). Consistently, this study found that the expression of LvGSK3b was inhibited in L. vannamei upon WSSV challenge and the mortality rates of LvGSK3b-inhibited L. vannamei in response to WSSV infection were significantly reduced when compares with the control group. Besides, LvGSK3b expression has a great impact on the expression and transcriptional regulation of LvDorsal, LvCactus and AMP genes, and LvGSK3b interacted with LvDorsal and LvCactus. Furthermore, knockdown of LvDorsal significantly reduced the expression of LvGSK3b. All these results implied that LvGSK3b plays a role during virus infection by regulating NF-kB activity in L. vannamei and there was a feedback regulatory loop existed. Briefly speaking, the silence of LvGSK3b by dsRNA resulted in upregulation of LvDorsal and downregulation of LvCactus. The downregulation of LvCactus could indirectly increase the expression of LvDorsal, followed by the increased expression of AMP genes which acted as important effectors of host defense against WSSV (24). At the same time, the increase of LvDorsal brought about the downregulation of LvGSK3b could enhance the effect. When suffer WSSV infection, the inhibition of LvGSK3b expression was also strengthened. All the series of reactions led to the final result was that the survival rates of LvGSK3b-inhibited L. vannamei significantly increased when suffer WSSV infection. To our knowledge, although the role of the GSK3b or NF-kB related signaling pathways in viral infection and the control of NF-kB signaling activity by GSKb has been confirmed separately (18)(19)(20)(21)(22)(23)37), no previous studies have explored the functional mechanisms of GSKb in viral infection from the perspective of the regulation of NF-kB; therefore, this study is the first such report.
In conclusion, this study revealed that LvGSK3b has a negative effect on L. vannamei by negatively regulating NF-kB activity when it is infected by WSSV. The results detail the immune mechanisms and provide a theoretical basis for enhancing the immunity and improving the disease resistance of shrimp, which lays the foundation for the establishment of effective prevention and control of WSS.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
The experiments were conceived and designed by SZ and LS. SZ, L-LZ, CH, HY, and SY collected samples and performed the experiments. MD, SZ, and LS analyzed the data and completed the writing of the paper. All authors contributed to the article and approved the submitted version.