The Zhikong scallops (Chlamys farreri) used in this study were collected from a local farm in Yantai, Shandong province, China. The studies involving animals were reviewed and approved by the respective Animal Research and Ethics Committees of Ludong University. The animals had shell height of 52–58 mm. The scallops were cultured in filtered and aerated seawater at 19 ± 1°C for one week before processing.

The genome of the Zhikong scallop has been sequenced (25), and based on IKK gene sequence alignment, we found several IKK homologous genes in the C. farreri genome. In this study, we focused on one IKK gene, and the forward and reverse primers, CfIKK3-F and CfIKK3-R, respectively, were designed for PCR amplification of its open reading frame (ORF) ( Table 1 ). The PCR products were purified using a SanPrep column DNA gel extraction kit (Sangon Biotech, Shanghai, China) and ligated into the pEASY-T1 cloning vector (TransGen Biotech, Beijing, China). Positive colony clones were sequenced to obtain the complete coding sequence for this gene. The coding amino acid sequences were obtained using an online translation tool (http://web.expasy.org/translate/). The CDD/SPARCLE database (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and simple modular architecture research tool (SMART, http://smart.embl-heidelberg.de/) were employed to predict the protein domains of CfIKK3. The IKK family protein sequences of the different species were acquired from the NCBI database (http://www.ncbi.nlm.nih.gov/guide/proteins/). Multiple sequence alignments were performed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/). Phylogenetic analysis was performed based on the IKK protein phylogenetic tree, which was constructed using the neighbor-joining algorithm in MEGA software (v.5.05) (http://www.megasoftware.net).

After acclimation, the 160 scallops were randomly divided into four groups of 40 scallops each. Each group of scallop was injected intramuscularly with any of 50 μL phosphate-buffered saline (PBS; 0.14 M NaCl, 3 mM KCl, 10 mM Na₂HPO₄, and 1.5 mM KH₂PO₄, pH 7.4), 50 μL lipopolysaccharide (LPS; 10 mg/mL in PBS; Sigma, USA), 50 μL peptidoglycan (PGN; 10 mg/mL in PBS; Sigma), or 50 μL polyinosinic–polycytidylic acid (poly(I:C); 1.0 mg/mL in PBS; In vivoGen, USA). The gills of three animals from each group were randomly sampled at 0, 3, 6, 12, 24, 48, and 72 h after PBS, LPS, PGN, and poly(I:C) stimulation. In addition, nine types of tissues (gills, hemocytes, feet, striated muscle, smooth muscle, hepatopancreas, gonad, mantle, and kidney) were collected from three untreated scallops to study the tissue distribution of CfIKK3 mRNA.

A Tiangen RNA extraction kit (Tiangen, China) was used to isolate total RNA from the scallop tissues. Reverse transcription PCR was performed using a PrimeScript RT reagent kit with gDNA eraser (Takara, Japan). qRT-PCR was performed using SYBR Green 2× Master Mix (Takara) and an ABI 7500 real-time thermal cycler (Applied Biosystems, USA). Primers CfIKK3-qRT-F/R were used to amplify the target gene fragments; qRT-PCR was programmed as follows: 6 min at 95°C, followed by 40 cycles of 10 s at 95°C and 30 s at 60°C. The specificity of the PCR products was analyzed using a melting curve. Relative gene expression levels were evaluated using the 2^–ΔΔCt method (26). The housekeeping EF-1α gene (GenBank accession number DT716075.1) was used as the internal control gene. The CfIKK3 expression levels following PAMP challenge were normalized to the expression in PBS-treated scallops. Differences were considered significant if the p-value was less than 0.05.

Currently, no mollusk cell line can successfully transfect protein-expression plasmids. Human embryonic kidney (HEK) 293T cells were used for protein expression. The cell lines present in this study were obtained from the American Type Culture Collection. Cells were grown in Dulbecco’s modified Eagle’s medium (Sigma, USA) supplemented with 10% fetal bovine serum (Gibco, USA) and 1× penicillin–streptomycin solution (Sangon, China). HEK293T cells were grown at 37°C in 5% CO₂ and subcultured every 2–3 days.

For the expression of Myc- and FLAG-tagged proteins, PCR fragments encoding full-length or truncated mutants were amplified using Phusion High-Fidelity DNA polymerase (Thermo Fisher Scientific, USA). Gene-specific primers used for amplification are listed in Table 1 . For Myc-tagged protein expression, plasmids pCMV-Myc (Clontech, USA) or pcDNA3.0-6×Myc (constructed in our lab, based on pcDNA3.0) were linearized with EcoRI (New England Biolabs, USA) or XhoI (New England Biolabs), respectively. For FLAG-tagged protein expression, the plasmid pCMS-EGFP-FLAG (constructed in our laboratory, based on pCMS-EGFP) was digested with XhoI (New England Biolabs). The purified PCR products were then inserted into the digested plasmid using an OK Clon DNA ligation kit (Accurate Biotechnology (Hunan) Co., Ltd., China).

The constructed protein expression plasmid was transfected into HEK293T cells using lipofectamine 3000 reagent (Invitrogen) according to the manufacturer’s instructions.

HEK293T cells were grown to an appropriate density in 10 cm dishes (Corning, USA). Afterwards, the cells were co-transfected with 5 μg Myc-tagged protein expression plasmid and 5 μg FLAG-tagged plasmid (pCMS-FLAG as a control). Twenty-four hours after transfection, the cells were harvested and lysed using a cell lysis buffer (Beyotime, China). Next, 20 μL of cell lysate was isolated as an input sample to detect protein expression. An appropriate amount of anti-FLAG M2 magnetic beads (Sigma-Aldrich) was added to the remaining lysate samples. The mixture was incubated at 4°C for > 2 h with gentle shaking, after which the beads were rinsed 3 times with cell lysis buffer. Twenty microliters of input samples and rinsed magnetic beads were then mixed with 2× protein sodium dodecyl sulfate-polyacrylamide gel electrophoresis loading buffer (Takara) at 100°C under denaturation for 5 min. Western blot assays were employed for target protein examination using an anti-Myc antibody (Abbkine Scientific Co., Ltd., Wuhan, China) and an anti-FLAG antibody (Sigma).

Proteins were extracted from HEK293T cells using RIPA lysis buffer (Beyotime, China), separated on a 10% SDS-PAGE gel, and transferred to PVDF membranes (Roche). After blocking with 3% nonfat milk in TBST (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, and 0.05% Tween-20) for 2 h, the membrane was incubated with a primary antibody (anti-Myc or FLAG antibody in TBST; 1:1000 dilution) for 3 h. After three washes in TBST, horseradish peroxidase-labeled goat anti-mouse IgG (Abbkine Scientific Co., Ltd., diluted 1:1000 in TBST) was used as the secondary antibody and incubated for 2 h at room temperature. Unbound IgG was then washed away, and the signals were visualized with Meilunbio^® fg super sensitive ECL luminescence reagent (Meilun Biotechnology Co., Ltd., Dalian, China), according to the manufacturer’s instructions. Protein bands were obtained using a ProteinSimple FluorChem E ultrasensitive, fully automated imaging analysis system (Bio-Techne, USA).

HEK293T cells were used to overexpress the CfIKK3 protein in order to study the activation of reporter genes such as NF-κB and interferon-stimulated response element (ISRE) using dual-luciferase reporter (DLR) experiments. HEK293T cells were cultured in 24-well plates and grown to approximately 50% density. The cells were then co-transfected with CfIKK3 (or its truncated expression plasmid), NF-κB reporter plasmid or ISRE reporter plasmid (Beyotime, China), and Renilla pRL-TK internal control vector (Promega, USA). Cells were harvested 24 h after transfection, and the transcriptional activity of the reporter genes was assayed using the Firefly & Renillalight luciferase reporter assay kit (Meilun Biotechnology Co., Ltd., Dalian, China). Firefly luciferase activity was normalized to Renilla luciferase activity. To ensure reliability, experiments in each group were performed in triplicate.

Based on the C. farreri genome (25), the primers CfIKK3-F and CfIKK3-R were designed to amplify the ORF of the scallop IKK gene. The complete coding sequence was confirmed and submitted to the GenBank database (GenBank accession number OP764589). Because this is the third IKK gene sequence we obtained from the C. farreri scallop genome, we named it CfIKK3. The polypeptide encoded by CfIKK3 is 773 amino acids long. The N-terminus of CfIKK3 protein is characterized by a typical protein kinase domain like that of other IKK family proteins. Two more domains, namely, a ubiquitin-like domain (ULD) and coiled-coil domain 1 (CCD1) were also present in IKK3 ( Figure 1 ).

The CfIKK3 protein has a similar domain composition to vertebrate TBK1/IKKϵ family proteins, all of which have a kinase domain (KD), ULD, and CCD1 ( Figure 2A ). Multiple sequence alignment of the KD sequences in these proteins showed that this domain is highly conserved ( Figure 2B ). As a characteristic domain of the IKK protein, we speculated that this scallop IKK protein has a conserved function. Based on the results of the sequence alignment, we constructed a phylogenetic tree of the IKK proteins of different species ( Figure 3 ). The figure clearly shows a bifurcation of IKK proteins into the IKKα/IKKβ family and the TBK1/IKKϵ family. Interestingly, CfIKK3 belongs to the TBK1/IKKϵ family and was first grouped with the TBK1/IKKϵ protein of other invertebrates, such as the Pacific oysters Crassostea gigas and Litopenaeus vannamei.

The expression pattern of CfIKK3 was verified by qRT-PCR. Tissue-specific expression of this gene is shown in Figure 4A . The CfIKK3 gene was detected in all tested scallop tissues, with the highest expression level in gill tissues. The expression level of this gene in blood cells was relatively lower. Simultaneously, to reveal the function of CfIKK3 in host innate immunity, we studied the temporal expression patterns of CfIKK3 under bacterial or virus-associated PAMPs attack. qRT-PCR results showed that CfIKK3 mRNA expression levels showed a significant inducible expression pattern after stimulation with bacterial-associated PAMPs (LPS, PGN) and virus-associated PAMP (poly(I:C)) ( Figures 4B–D ). For example, after LPS challenge, the expression of the CfIKK3 gene was significantly upregulated at 6 h after stimulation; after PGN challenge, the expression of the CfIKK3 gene was significantly increased at 72 h after stimulation; and after poly(I:C) challenge, the expression of the CfIKK3 gene was elevated at 6 h and 72 h after stimulation.

TLR family proteins are a key class of PPRs that play key roles in cellular innate immunity against bacteria and viruses. TLRs can recruit the adaptor protein MyD88 for immune signaling (27, 28). In the present study, we investigated the potential relationship between CfIKK3 and the adaptor protein MyD88 (GenBank accession number DQ249918.1). Notably, co-IP and immunoblotting results showed that CfIKK3 can directly interact with CfMyD88 ( Figure 5 ). Additionally, truncated mutants of protein expression plasmids were constructed to explore the key protein domains that mediate the interaction of these two proteins ( Figures 5A, C ). As shown in Figure 5B , in addition to the wild-type CfIKK3, the CfIKK3-P1 mutant (KD only) and the CfIKK3-P2 mutant (KD + ULD) could also combine with CfMyD88, suggesting a critical role for the KD domain of CfIKK3 in mediating the interaction. Based on this result, we further investigated the binding of truncated CfMyD88 protein to the KD of CfIKK3. As shown in Figure 5D , in addition to the wild-type CfMyD88, the CfMyD88-P1 mutant (DEATH domain only), CfMyD88-P2 mutant (DEATH domain, the peptide length is slightly longer than that of CfMyD88-P1), and CfMyD88-P3 mutant (DEATH domain + TIR domain) could also interact with CfIKK3-P1, suggesting a critical role for the DEATH domain of CfMyD88 in mediating this interaction.

In addition to TLR family proteins, RLR family proteins constitute another class of key PPRs (29, 30). RLR proteins are usually localized in the cytoplasm and transduce signals by recruiting adaptor protein MAVS (31, 32). We investigated the relationship between CfIKK3 and the adaptor protein MAVS (GenBank accession number OM069740, Figure 6 ). Co-ip and western blotting results did not indicate an interaction between the full-length CfIKK3 and CfMAVS. However, CfIKK3-P1, CfIKK3-P2, and CfIKK3-P3 (CCD1 only) could directly bind to CfMAVS, suggesting that both KD and CCD1 of CfIKK3 may be involved in the interaction with MAVS ( Figure 6B ). Furthermore, we constructed a truncated CfMAVS ( Figure 6C ) and studied the interaction of each truncated MAVS protein with CfIKK3-P1. The results showed that CfMAVS-P1 (CARD only), CfMAVS-P2 (CARD + DEATH domain), and CfMAVS-P3 (full-length CfMAVS without TM domain) could directly bind to CfIKK3-P1, suggesting that the CARD of CfMAVS might be sufficient for binding to the KD of CfIKK3 ( Figure 6D ). At the same time, we also studied the interaction of each truncate protein of CfMAVS with CfIKK3-P3. The results showed that both CfMAVS-P2 and P3 could directly bind to CfIKK3-P3, while CfMAVS-P1 did not, implying that the DEATH domain of CfMAVS mediates binding to the CCD1 domain of CfIKK3 ( Figure 6E ).

In this study, co-IP assays were conducted to determine the binding relationship between CfIKK3 and our previously cloned IKK family proteins (CfIKK1 and CfIKK2). The results showed that CfIKK3 did not substantially bind to CfIKK1 (results not shown), but CfIKK3 had obvious self-association ( Figure 7A ). Furthermore, the regions mediating homodimer formation were verified using constructed truncations. Interestingly, the results showed that CfIKK3-FL-flag, CfIKK3-P1-flag, CfIKK3-P2-flag, and CfIKK3-P3-flag interacted with CfIKK3-P1-myc (KD only) ( Figure 7B ). The results also showed that CfIKK3-FL-flag, CfIKK3-P1-flag, CfIKK3-P2-flag, and CfIKK3-P3-flag interacted with CfIKK3-P3-myc ( Figure 7C ).

In addition, co-ip results showed that CfIKK3 might be interacting with CfIKK2, another TBK1/IKKϵ family member of C. farreri. CfIKK3-P1-flag and CfIKK3-P2-flag bound to CfIKK2-myc, suggesting that the KD of CfIKK3 might be sufficient for binding to CfIKK2 ( Figure 8 ).

DLR gene assays were employed to detect the possible activation of CfIKK3 on target reporter genes (NF-κB or ISRE). The results showed that the overexpression of CfIKK3 in HEK293T cells did not activate the ISRE reporter gene (data not shown). However, overexpression of CfIKK3 can significantly activate the NF-κB signaling pathway, and this activation has a significant dose-dependent effect. From the results of the activation of NF-κB by each truncated mutant protein, KD + ULD appeared to have a significant dose-dependent activation effect on NF-κB reporter genes ( Figure 9 ).