A novel intracellular signaling pathway elicited by DM9CP-6 regulates immune responses in oysters

Abstract

DM9 domain containing protein (DM9CP) is a recently discovered novel pattern recognition receptor (PRR) with the ability to recognize various microbes, but its role as a cytosolic PRR to track the invading microbes and trigger signaling pathway is still not clear. In the present study, a DM9CP (designated as CgDM9CP-6) and an intracellular regulatory molecule, 14-3-3 (designated as Cg14-3-3ε) were identified from oyster Crassostrea gigas, which contained two tandem DM9 repeats and a 14-3-3ε domain, respectively. CgDM9CP-6 and Cg14-3-3ε were higher expressed in haemocytes, and their mRNA expression levels increased significantly after Vibrio splendidus stimulation. CgDM9CP-6 could bind various polysaccharides (LPS, PGN, MAN, and D-mannose) and microbes (Staphylococcus aureus, Micrococcus luteus, V. splendidus, Escherichia coli, Yarrowia lipolytica, and Pichia pastoris) in vitro, and it was observed to be colocalized with the FITC-labeled V. splendidus, E. coli and Y. lipolytica in haemocytes in vivo. The pull-down, surface plasmon resonance (SPR) and Co-Immunoprecipitation (Co-IP) assays all demonstrated that Cg14-3-3ε was able to interact with CgDM9CP-6 in vitro or in vivo. After the expression of CgDM9CP-6 and Cg14-3-3ε was knocked down separately by RNAi, the nuclear translocation of CgRel in haemocytes was inhibited, and the mRNA expressions of interleukin17-3 (CgIL17-3), CgIL17-6, CgLysozyme and CgBigDef1 in haemocytes all decreased significantly after the oysters were stimulated with V. splendidus. The results collectively indicated that CgDM9CP-6 could function as an intracellular PRR to be associated with Cg14-3-3ε to trigger the NF-κB pathway, which eventually regulated the immune responses including the expressions of inflammatory cytokines and antimicrobial molecules in oysters.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-025-02389-4.

Introduction

Pattern recognition receptors (PRRs) are a class of receptors to recognize pathogen-associated molecular patterns (PAMPs) and trigger the inducible defense responses against invading pathogens [1–3]. They are normally divided into membrane receptors, such as Toll-like receptors (TLRs) and C-type lectin receptors (CLRs) and cytoplasmic receptors, such as nucleotide-binding oligomerization domain-like receptors (NLRs) and retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) [4, 5]. DM9 domain containing proteins (DM9CP)s are recently documented to function as extracellular PRRs to bind various PAMPs and microbes [6–8], and then mediate various immune responses, such as phagocytosis [6], agglutination [6, 9], encapsulation [8], and inflammation [7].

DM9CPs are PRRs to bind various PAMPs and microbes and present in lower vertebrates (fish) [10, 11] and invertebrates (Arthropoda, Mollusca, and Platyhelminthes) [6, 7, 12–17]. DM9CP was firstly identified from Drosophila melanogaster [12] and in mosquito Anopheles gambiae. Among which, Pfs47 receptor protein from mosquito with four tandem DM9 domains could bind the surface protein Pfs47 of parasite [16]. DM9CPs had also been identified in crab Eriocheir sinensis and oyster Crassostrea gigas [6–9, 18–21], respectively. Most of them were demonstrated to function as extracellular PRRs to bind various PAMPs and microbes in cell-free haemolymph [6–9, 18, 20–22].

Different PRRs have specific subcellular localizations to recognize different PAMPs or microbes and activate different signaling pathways. For example, TLRs and CLRs on the cell surface recognize various PAMPs or microbes and activate the nuclear factor κB (NF-κB), activator protein 1 (AP-1), cAMP response element-binding protein (CREB), and nuclear factor of activated T cells (NFAT) [23, 24]. NLR family apoptosis inhibitory protein (NAIP) located inside the cell can directly recognize bacterial flagella and trigger the Caspase-mediated pyroptosis [25]. The cytoplasmic RLR can recognize the intermediate double-stranded RNA (dsRNA) and activate the interferon regulatory factor (IRF) signaling [26]. CgDM9CP-5 from C. gigas was reported to be secreted in cell-free haemolymph and could activate CgIntegrin-mediated mitogen-activated protein kinase (MAPK) signals after its binding to different PAMPs [7]. In our previous study, most oyster DM9CPs were found to be highly expressed in haemocytes and some of them were also located in haemocyte cytoplasm [6–9, 18, 19]. It is speculated that the cytosolic DM9CPs may function as intracellular PRRs to sense different pathogens and initiate immune response.

14-3-3s are a very important class of intracellular regulatory molecules to regulate a wide variety of signals transduction pathways by promoting the cytoplasmic export of pre-phosphorylated substrates [27–30]. They are thought to alter the stability, activity and/or subcellular localization of the target protein via protein-protein interaction within the cell to mediate signaling pathways [27, 31–33]. Many 14-3-3 targets have already been identified, such as Raf-1 [34], the cell cycle regulator Cdc25 [35], histone deacetylases [36], and Nuclear factor kappa B (NF-κB)/P65 [37]. Among which, NF-κB homolog has been identified in fish, shrimp, and oyster. In fish, NF-κB pathway could promote the expression of interleukin-1β (IL-1β) and β-Defensin1 [38]. In shrimp and oyster, NF-κB pathway induced the expressions of various cytokines and/or antibacterial peptides [7, 20, 39, 40]. For example, in C. gigas, CgRel could translocate into nucleus to regulate the expressions of CgIL17s and CgBigDef1 after V. splendidus stimulation [41].

The Pacific oyster C. gigas is important economic Mollusca in the aquaculture industry [42]. In the past decades, the frequent outbreak of disease has caused serious mortality of oysters, which threatens the development of aquaculture industry in China. The recent studies have reported that DM9CPs function as PRRs and involve in the antibacterial responses [6–9, 18]. In the present study, the intracellular CgDM9CP-6 was identified from C. gigas with the objectives to (1) determine its activity to recognize and trace the invading microbes in haemocytes, (2) identify its interaction with Cg14-3-3ε to trigger NF-κB pathway, and (3) confirm its role in the regulation of immune responses of inflammatory cytokines and antimicrobial effectors in oysters.

Materials and methods

Animals

Pacific oysters (C. gigas, shell length 12–16 cm) collected from a local farm in Dalian, Liaoning, China, were cultured in aerated seawater at 15 ± 2 °C for 2 weeks before processing. Female mice were purchased from Dalian Institute of Drug Control. All the experiments were performed following the animal ethics guidelines approved by the Ethics Committee of Dalian Ocean University.

cDNA cloning and sequence analysis of CgDM9CP-6 and Cg14-3-3ε

The full-length cDNA fragments of CgDM9CP-6 and Cg14-3-3ε were amplified by PCR with primers (Table S1) according to the sequence information of CgDM9CP-6 (LOC105327353) and Cg14-3-3ε (LOC105321521) from the National Center for Biotechnology Information (NCBI) database (https://www.ncbi.nlm.nih.gov/).

DM9CPs and 14-3-3εs were searched from the genomes of different species in metazoa to construct the multiple sequence alignments with Clustal X v2.0 program. Phylogenetic trees were constructed by employing Molecular Evolutionary Genetics Analysis software with the neighbor-joining (NJ) method. Simple Modular Architecture Research Tool (SMART) was employed to predict the conserved domains. The promoter region was analyzed using TRANSFAC for potential transcriptional factor binding sites (http://www.generegulation.com/pub/programs.html).

Sample collection and immune stimulation

One hundred and forty-four oysters were randomly divided into two groups (stimulation group and control group). The two groups received an injection with 100 µL V. splendidus (2 × 10⁸ CFU mL⁻¹) or PBS, respectively. To alleviate the stress and injuries of the animals, oysters were pre-cooled on ice before experiments. Nine individuals were randomly sampled from each group for the haemolymph collection at 0, 3, 6, 12, 24, 48 and 72 h after V. splendidus and PBS stimulation, respectively. The haemolymph collected from three oysters was pooled together as one sample and there were three parallels at each time point. The haemocytes were harvested by centrifugation at 800 × g, 4 °C for 10 min. The haemolymph, labial palp, mantle, adductor muscle, gill, gonad and hepatopancreas were collected from other nine oysters for analyzing the tissue distribution of CgDM9CP-6 and Cg14-3-3ε. All these samples were stored at −80 °C for subsequent RNA extraction.

Structural prediction of CgDM9CP-6, Cg14-3-3ε and CgRel protein using AlphaFold3

The three-dimensional structures of CgDM9CP-6, Cg14-3-3ε and CgRel protein were predicted using AlphaFold3 server (https://alphafoldserver.com/). All predictions were carried out using default parameters [43]. The reliability of the predicted models was assessed based on the predicted Local Distance Difference Test (pLDDT) scores, with structures scoring above 70 considered reliable [44]. The predicted models were visualized and refined using PyMOL (http://pymol.sourceforge.net/) to highlight key structural features and domains [45]. These final models were then employed in further structural and functional analyses.

Molecular docking

The molecular structure of LPS (Compound CID: 375581774) was obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) [46]. The most stable conformation of LPS was determined using Chem3D software version 22.0.0 [47]. Molecular flexible docking between LPS and CgDM9CP-6 was conducted using AutoDock Vina [48]. The docking results were visualized using PyMOL.

Reverse transcription quantitative PCR (RT-qPCR) analysis of target genes

RT-qPCR was performed to analyze the mRNA expressions of target genes with specific primers (Table S1). The fragments of elongation factor (CgEF, NM_001305313) amplified with primers CgEF-RT-F/-R (Table S1) were employed as the internal reference. RT-qPCR was programmed at 95 °C for 10 min, followed by 40 cycles at 95 °C for 10 s and 60 °C for 45 s. The final product was analyzed via melting analysis from 65 °C to 95 °C. The relative mRNA expressions levels of target genes were analyzed by 2^−△△Ct method [49].

Recombinant expression and purification of CgDM9CP-6 and Cg14-3-3ε

The recombinant proteins of CgDM9CP-6 and Cg14-3-3ε were expressed and purified as previously described with some modification [50]. The coding sequences of the two genes were amplified by the primers of CgDM9CP-6-ORF-F/-R and Cg14-3-3ε-ORF-F/-R (Table S1). The PCR products of CgDM9CP-6 were digested with Nde I and Hind III, gel-purified, and ligated into pET-30a plasmid to construct the recombinant plasmid pET-30a-CgDM9CP-6. The PCR products of Cg14-3-3ε were digested with BamH I and Xho I, gel-purified and ligated into the pGEX-4T-1 plasmid to construct the recombinant plasmid pGEX-4T-1-Cg14-3-3ε. Trx-tag from pET-32a and GST-tag from pGEX-4T-1 plasmid were used as control. The recombinant plasmids pET-30a-CgDM9CP-6 and pGEX-4T-1-Cg14-3-3ε were transformed into E. coli Transetta (DE3) (TransGen Biotech, Beijing, China), respectively. The positive strain E. coli was verified by DNA sequencing. The recombinant proteins of CgDM9CP-6 (rCgDM9CP-6) and Cg14-3-3ε (rCg14-3-3ε) were separately purified by affinity chromatography using His-Bind resin and GST-Bind resin following the manufacturer’s instructions and stored at −80 °C for subsequent experiments. The recombinant protein of CgRel (rCgRel) was prepared as previously reported [51].

Western blotting assay of CgDM9CP-6 and Cg14-3-3ε

The specificity of antibody was determined by western blotting assay. The rCgDM9CP-6 and rCg14-3-3ε were injected into six-weeks old mouse to acquire polyclonal antibody as the method described previously [50]. The haemocyte lysates were separated by SDS-PAGE and transferred onto a 0.22 μm pore nitrocellulose membrane. The anti-CgDM9CP-6 and anti-Cg14-3-3ε were used as primary antibody and horseradish peroxidase (HRP)-labeled goat anti-mouse IgG (Beyotime, Shanghai, China) was used as second antibody. The membrane was incubated in Western lighting ECL substrate system (Thermo Fisher Scientific, Waltham, MA) in dark for 1 min, and the immune-blotted protein bands were visualized using chemiluminescent imaging system (Amersham Imager 600, GE Healthcare Life Sciences, Marlborough, MA). The antisera of CgRel were prepared as previously reported [51].

PAMP binding assay of rCgDM9CP-6

The PAMP binding activity of rCgDM9CP-6 was determined by enzyme-linked immunosorbent assay (ELISA) [52]. Briefly, PAMPs including LPS from E. coli 0111: B4 (Sigma-Aldrich), PGN from S. aureus (Sigma-Aldrich), mannan from Saccharomyces cerevisiae (Sigma-Aldrich) and D-(+)-mannose from wood (Sigma-Aldrich) were coated on wells of a 96-well microtiter plate (20 µg well⁻¹) at 37 °C for 1 h. One hundred microlitres rCgDM9CP-6 (1.56, 3.13, 6.25, 12.50, 25.00, 50.00 and 100.00 µg mL⁻¹) was added into the wells of 96-well plates and incubated at 37 °C for 1 h. The same concentration of rTrx was used as control. After washed three times with TBST, the anti-6 × His Tag mouse monoclonal antibody (1:1000 dilution) was used as primary antibody and HRP-labeled goat anti-mouse IgG (1:1000 dilution; Beyotime, Shanghai, China) was as the second antibody. The absorbance at 405 nm was measured by Infinite M1000 PRO (Tecan, Sweden). The reading of rCgDM9CP-6 group was employed as sample group (P), and the reading of rTrx group and TBS group were used as negative group (N) and blank group (B), respectively. Samples with P (sample) - B (blank)/N (negative) - B (blank) > 2.1 were considered as positive. Each experiment was repeated in triplicates and the results were given in terms of the mean of three individual measurements ± S.D. (N = 3).

Microbial binding assay of rCgDM9CP-6

Bacteria V. splendidus were isolated and preserved in our laboratory [53]. E. coli Transetta (DE3) was purchased from TransGen Biotech (TransGen Biotech, Beijing, China). Staphylococcus aureus, Micrococcus luteus, Y. lipolytica and P. pastoris were obtained from the Microbial Culture Collection Center (Beijing, China). Microbial binding activities of rCgDM9CP-6 towards Gram-positive bacteria (Micrococcus luteus, Staphylococcus aureus), Gram-negative bacteria (V. splendidus, E. coli) and fungi (Y. lipolytica, Pichia pastoris) were examined by western blotting assay. The microbes (1 × 10⁹ CFU mL⁻¹) were suspended in TBS, and incubated with rCgDM9CP-6 (100 µL, 1 mg mL⁻¹) under slight rotation at 4℃ overnight. After rinsed by TBS for five times, the mixture containing microbes and binding proteins was then analyzed by western blotting assay with anti-His as primary antibody and HRP-labeled goat anti-mouse IgG (Beyotime, Shanghai, China) as second antibody.

Bio-layer interferometry (BLI) assay

The BLI assay was conducted on Octet RED96 System (ForteBio, Fremont CA) as previous description [7]. Briefly, the BLI assay was performed on Octet RED96 System (ForteBio) at 29 °C with PBST as running buffer. The proteins of rCgDM9CP-6, rCgRel and rTrx were biotinylated using EZ-Link Sulfo-NHS-LC-Biotinylation Kit (Genemore, Suzhou, China) for 30 min. The streptavidin biosensor was transferred into PBS with Tween 20 buffer containing rCg14-3-3ε at the indicated concentration (17.6, 11.7, 0.88, 0.59 µM) (for association) or the analyte-free buffer (for dissociation). The binding constant was determined using software provided by Fortebio (Data Analysis 8.5). The association and dissociation data were used to calculate equilibrium K_D values based on a 1:1 binding model.

Pull-down assay of CgDM9CP-6, Cg14-3-3ε, and CgRel

The Pull-down assay was performed to analyze the interaction between rCgDM9CP-6 (His tag) and rCg14-3-3ε (GST tag), rCg14-3-3ε (GST tag) and rCgRel (His tag). The purified rCgDM9CP-6 (1 mg mL⁻¹, 100 µL) was incubated with His resin (200 µL) at 4℃ for 2 h, which was then incubated with the purified rCg14-3-3ε (1 mg mL⁻¹, 100 µL) at 4℃ for 2 h. The reaction mixture was washed with PBS three times, and the collected His resin samples were analyzed by using 12% SDS-PAGE. The interaction between rCg14-3-3ε and rCgRel protein was performed similarly as above. As controls, rTrx (His tag) was used in pull-down assays with rCg14-3-3ε (GST tag), while rGST (GST tag) was used in pull-down assays with rCgDM9CP-6 (His tag) and rCgRel (His tag), respectively.

Co-Immunoprecipitation (Co-IP) assay of Cg14-3-3ε, CgDM9CP-6, and CgRel

The Co-IP assay was performed to examine the interaction between Cg14-3-3ε and CgDM9CP-6, Cg14-3-3ε and CgRel in vivo. Protein A resin (Sangon, Shanghai, China) (200 µL) was mixed with 1 mL PBS in EP tubes. After washed with PBS for three times by centrifugation at 3,000 rpm, 4℃ for 5 min, anti-Cg14-3-3ε, anti-CgDM9CP-6 or anti-CgRel (20 µL) were individually added into the tubes and incubated at 4℃ for 1 h. Then the haemocyte lysates were added into the tubes and incubated at 4℃ for 2 h, respectively. After washing with PBS for three times, the resulting pellet (bound protein, Ab, and protein A) was added into the electrophoresis sample buffer and denatured for 5 min. The protein samples were analyzed by SDS-PAGE and western blotting assay using anti-Cg14-3-3ε, anti-CgDM9CP-6 and anti-CgRel, respectively.

Immunocytochemistry assay of CgDM9CP-6 and Cg14-3-3ε

Immunocytochemistry assay was employed to observe the subcellular localization of CgDM9CP-6 and Cg14-3-3ε in haemocytes. The haemocytes were collected from the sinus of three oysters at 2 h after V. splendidus stimulation and dropped onto the slides coated poly-L-lysine. The following steps were according to the previous report [50] with anti-CgDM9CP-6 or anti-Cg14-3-3ε (diluted 1:500 in 3% BSA) as primary antibody and Alexa Fluer 488-labeled and goat anti-mouse IgG (diluted 1:1000 in 3% BSA) as second antibody. The 2-(4-Amidinophenyl)−6-indolecarbamidine dihydrochloride (DAPI) (Solarbio life sciences, China, diluted 1:1000 in BSA) was used to stain the haemocyte nucleus. The prepared slides were washed again with PBST three times and mounted in buffered glycerin (50%) for observation by fluorescence microscope (Axio Imager A2, ZEISS).

Microbes including E. coli, V. splendidus and Y. lipolytica were fixed with 4% paraformaldehyde (PFA) and then mixed with 1 mg mL⁻¹ FITC (Sigma, USA) in 0.1 M NaHCO₃ (pH 9.0) buffer with continuous gentle stirring at 4 °C overnight. The FITC-labeled microbes were washed with PBS for five times to eliminate free FITC molecules. These microbes were incubated with haemocytes at room temperature for 1 h. The haemocytes containing FITC-labeled microbes were then incubated with anti-CgDM9CP-6 as the primary antibody and Alexa Fluor 647-labeled Goat Anti-Mouse IgG as secondary antibody. 4’,6-Diamidino-2-phenylindole dihydrochloride (DAPI) was used to stain nucleus. Fluorescence signals were observed under inversion fluorescence microscope (Axio Imager A2, ZEISS) to analyze the co-localization of CgDM9CP-6 and FITC-labeled microbes.

RNA interference (RNAi) of target genes in vivo

The fragments of CgDM9CP-6 and Cg14-3-3ε amplified with primers CgDM9CP-6-Fi/-Ri, Cg14-3-3ε-Fi/-Ri (Table S1) were used as templates for dsRNA synthesis, respectively. The EGFP fragments amplified from pEGFP vector with primers EGFP-Fi/-Ri (Table S1) were used as control. The dsRNAs were synthesized according to the method described in previous report [54]. Thirty-six oysters were employed for the RNAi experiment, and they were randomly divided into four groups (Blank, EGFP-RNAi, CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi groups). The oyster in each group received an injection of 100 µL dsRNAs (1 µg µL⁻¹) specific for EGFP, CgDM9CP-6 and Cg14-3-3ε, respectively. To enhance the RNAi efficiency, a second injection was conducted at 12 h after the first injection. The untreated oysters were employed as Blank group. Each group contained nine oysters, and they were divided into three parallels. The total RNA was extracted from haemocytes and assessed by RT-qPCR with specific primers (Table S1) to evaluate their RNAi efficiency.

In CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi groups, the mRNA transcripts of CgIL17-3, CgIL17-6, CgLysozyme and CgBigDef1 in haemocytes were analyzed by RT-qPCR with specific primers (Table S1) at 12 h after V. splendidus stimulation. The Nuclear Protein Extraction Kit (Solarbio, China) was used to obtain the nuclear and cytoplasm protein of haemocytes. And the subcellular location of CgRel was analyzed by immunocytochemistry and western blotting assay.

Chromatin immunoprecipitation (ChIP) assay of CgRel

ChIP assay was performed to examine the activity of CgRel to bind the NF-κB binding sites in the promoter regions of CgIL17-3, CgIL17-6 and CgLysozyme by using a ChIP Assay Kit (Beyotime, China; P2078) according to the manufacturer’s instruction. The collected haemocytes were resuspended in the modified Leibovitz L-15 medium and formaldehyde (1% final concentration) was added into the haemocyte suspension to crosslink the DNA and protein. The cross-linked fragmented DNA was precleared with protein A agarose/salmon sperm DNA and then was incubated with anti-CgRel at 4℃ overnight. The cross-linked DNA was de-crosslinked with 200 mM sodium chloride at 65℃ for 4 h and the proteins were removed by treatment with proteinase K. The resultant DNA was extracted using the phenol/chloroform/isoamyl alcohol method. The immunoprecipitates were analyzed by using RT-PCR with primers specific for the fragments containing CgRel binding sites (Table S1).

Dual-luciferase reporter assays

The plasmid construction and dual-luciferase reporter assays were performed as previously described with some modifications [55]. Briefly, CgRel cDNA fragments were amplified and inserted into pcDNA3.1⁽⁺⁾ vector (Miaoling, China) at Nde I and Xba I restriction sites. The promoter sequences of CgIL17-3, CgIL17-6 and CgLysozyme were amplified using their specific primers (Table S1) and inserted into pGL3-basci vector (Miaoling, China) at Kpn I and Hind III restriction sites. All recombinant plasmids were transformed into trans5α (DE3) cells (TransGen, China) and extracted using EndoFree Plasmid Max Kit (Tiangen, China).

The HEK293T cells were cultured as previous described [55] and divided into nine experimental groups: pGL-control (transfected with pGL3-control), pGL3-basic (pGL3-basic co-transfected with pcDNA3.1⁽⁺⁾ or pcDNA3.1-CgRel), pGL3-CgIL17-3 promotor (pGL3-CgIL17-3 promotor co-transfected with pcDNA3.1⁽⁺⁾ or pcDNA3.1-CgRel), pGL3-CgIL17-6 promotor (pGL3-CgIL17-6 promotor co-transfected with pcDNA3.1⁽⁺⁾ or pcDNA3.1-CgRel) and pGL3-CgLysozyme promotor (pGL3-CgLysozyme promotor co-transfected with pcDNA3.1⁽⁺⁾ or pcDNA3.1-CgRel). The group of pGL-control was employed as positive group, pGL3-basic group served as blank group, and pGL3-CgIL17-3, -CgIL17-6 and -CgLysozyme promotor group was used to confirm the binding activity of CgRel. In all groups, the pRL-TK renilla luciferase plasmid (Miaoling, China) was co-transfected as an internal control using Lipofectamine TM3000 (Invitrogen, USA). Luciferase activity was measured at 48 h post-transfection using dual-luciferase reporter assay system (Promega, USA).

Statistical analysis

All the data were expressed as mean ± S.D. and analyzed by Statistical Package for Social Sciences (SPSS) 18.0. The significant differences among groups were tested by one-way analysis of variance (ANOVA). The comparison between two independent samples was analyzed by t-test. Statistically significant differences were designated at p < 0.05 (N = 3).

Results

Sequence characteristics of CgDM9CP-6 and Cg14-3-3ε

The well-annotated genomes from the species in metazoan subphyla/phyla were screened to search DM9CPs, they were found to be presented in Vertebrates (only limited in teleost and lamprey), Urochordata, Arthropoda, Mollusca, Annelida, Platyhelminthes, and Coelenterata (Fig. 1A). There were two branches in the evolutionary tree of DM9CPs (Vertebrate branch and Invertebrate branch). Seven DM9CPs from oyster (C. gigas) were clustered to be a sister branch to that from anemone (Actinia tenebrosa) and coral (Dendronephthya gigantea) (Fig. 1A). The amino acid sequences of DM9CPs in different species were not conserved (Fig. S1A). There were one to four DM9 domains identified in DM9CPs, such as one DM9 domain in zebrafish, lamprey and ascidian DM9CP, two tandem DM9 domains in mosquito, crab, oyster, polychaete, trematode, anemone and coral DM9CP, as well as four tandem DM9 domains in fruit fly DM9CP (Fig. 1A). In addition, there also existed other domains in some DM9CPs, such as suppressor of cytokine signaling (SOCS) domain in ascidian DM9CP, Farnesoic acid O-methyl transferase (FAMeT) domain in crab DM9CP, and Aerolysin (as a cytolytic toxin exported by Aeromonas hydrophila) domain in teleost and lamprey DM9CP.

Fig. 1.

Phylogenetic tree analysis and structural domains of DM9CPs and 14-3-3εs. A Phylogenetic tree analysis and structural domains of CgDM9CP-6 and DM9CPs from other species. Fg, Fasciola gigantica; Sj, Schistosoma japonicum; Dg, Dimorphilus gyrociliatus; Ci, Ciona intestinalis; Es, Eriocheir sinensis; Ac, Anopheles cruzii; Ag, Anopheles gambiae; Dm, Drosophila melanogaster; Dm, Drosophila miranda; Dg, Dendronephthya gigantea; At, Actinia tenebrosa; Cg, Crassostrea gigas; Dr, Danio rerio; Lr, Lethenteron reissneri; Tn, Thalassophryne nattereri. B Phylogenetic tree analysis and structural domains of Cg14-3-3ε and 14-3-3εs from other species. Hs, Homo sapiens; Mm, Mus musculus; Dr, Danio rerio; Xl, Xenopus laevis; Bb, Bufo bufo; Ag, Anopheles gambiae; Dm, Drosophila melanogaster; Cg, Crassostrea gigas; Cv, C. virginica; Ob, Octopus bimaculoides. Red box: Signal peptide; DM9 with gray box: the unconservative domain

The amino acid sequences of 14-3-3εs were relatively conserved (Fig. S1B), and there were two distinct branches (Vertebrate branch and Invertebrate branch) in the phylogeny tree. 14-3-3εs from Mollusca were clustered and dropped into the invertebrate branch (Fig. 1B) and they all had the conserved 14-3-3ε domain like those from other reported species (Fig. 1B).

The mRNA spatiotemporal expressions of CgDM9CP-6 and Cg14-3-3ε

qPCR was performed to analyze the mRNA expression patterns of CgDM9CP-6 and Cg14-3-3ε. The mRNA transcripts of CgDM9CP-6 were detected in all the tested tissues, including hepatopancreas, gonad, gill, adductor muscle, mantle, labial palp and haemolymph with the higher expression level in labial palp and haemolymph, which were 5.28-fold and 7.51-fold higher (p < 0.05) than that in hepatopancreas (Fig. 2A). The mRNA expression levels of Cg14-3-3ε were higher in gill and haemolymph, which were 23.79 and 21.43-fold (p < 0.05) of that in mantle (Fig. 2B). After V. splendidus stimulation, the mRNA expression levels of CgDM9CP-6 and Cg14-3-3ε in haemocytes were examined by RT-qPCR. The mRNA expression levels of CgDM9CP-6 increased significantly at 3 h (1.40-fold, p < 0.05), 6 h (1.60-fold, p < 0.05), 12 h (3.24-fold, p < 0.01), and 48 h (1.93-fold, p < 0.05) (Fig. 2C). Similarly, the mRNA expression levels of Cg14-3-3ε also increased significantly at 3 h (3.39-fold, p < 0.01), 6 h (2.30-fold, p < 0.05), and 12 h (2.78-fold, p < 0.01) (Fig. 2D).

Fig. 2.

Tissue distribution and expression pattern of CgDM9CP-6 and Cg14-3-3ε. A-B The mRNA expressions of CgDM9CP-6 and Cg14-3-3ε in different tissues analyzed by qPCR. C-D The mRNA expressions of CgDM9CP-6 and Cg14-3-3ε in haemocytes at 0, 3, 6, 12, 24, 48 and 72 h after V. splendidus stimulation by qPCR. Each value was shown as mean ± S.D. (N = 3). Different letters: p < 0.05 (one-way ANOVA). Asterisks: *p < 0.05, **p < 0.01 (t-test)

The activity of CgDM9CP-6 to bind different PAMPs and microbes

The recombinant plasmids of pET-30a-CgDM9CP-6 and pGEX-4T-1-Cg14-3-3ε were transformed into E. coli transetta (DE3), respectively. The recombinant proteins were purified by affinity chromatography using His-Bind resin and GST-Bind resin, and there were distinct bands detected in nitrocellulose membrane. The predicted molecular weight of CgDM9CP-6 was 16.4 kDa. rCgDM9CP-6 containing N-6×His, N-S, N-Thrombin, N-EK and C-6×His in the expression vector pET-30a (+) was about 16.4 kDa + 6.0 kDa = 22.4 kDa (Fig. 3A). The predicted molecular weight of Cg14-3-3ε was about 30.1 kDa. rCg14-3-3εcontaining GST-tagged in the expression vector pET-GST is about 30.1 kDa + 25 kDa = 55.1 kDa (Fig. 3C). The purified rCgDM9CP-6 and rCg14-3-3ε were used for preparing polyclonal antibodies. Distinct single bands with similar weight as CgDM9CP-6 and Cg14-3-3ε were observed by western blotting assay with haemocyte lysates, indicating the high specificity and efficiency of the polyclonal antibodies (Fig. 3B and D). The activity of CgDM9CP-6 to bind LPS, PGN, MAN and D-mannose was determined by ELISA. rCgDM9CP-6 displayed high affinity to MAN (P/N = 8.39) and D-mannose (P/N = 7.52), low affinity to LPS (P/N = 4.18) and PGN (P/N = 2.86). The P/N values of rCgDM9CP-6 to bind LPS, PGN, MAN and D-mannose increased from 0.20 to 4.18, 0.22 to 2.86, 0.79 to 8.39 and 0.12 to 7.52 with the increase of rCgDM9CP-6 concentration from 3.13 to 100.00 µg mL⁻¹ (Fig. 3E). As a control, no binding affinity was observed in the group of rTrx (data not shown). To elucidate the potential binding mode of LPS with CgDM9CP-6, the putative sites of DM9CP-6 to bind LPS were mapped onto the predicted CgDM9CP-6 structure, and AutoDock Vina was used to perform the flexible docking. The LPS was found to form extensive hydrogen bonds with residues Ala104, Glu137, Arg103, Gly113, Gly42, Ala19, Ile30, Phe29, Ile17, Asp66 and Leu68 in CgDM9CP-6 (Fig. S3A).

Fig. 3.

Recombinant protein and polyclonal antibody of CgDM9CP-6 and Cg14-3-3ε, PAMP and microbial binding activity of rCgDM9CP-6 as well as the co-localization of CgDM9CP-6 with microbes. A SDS-PAGE analysis of rCgDM9CP-6. Lane M: standard protein molecular weight marker; Lane 1: negative control (without induction); Lane 2: induced rCgDM9CP-6; Lane 3: purified rCgDM9CP-6. B Western blotting assay with CgDM9CP-6 antibody in haemocytes. Lane M: standard protein molecular weight marker; Lane 1: native CgDM9CP-6 protein in haemocytes by using CgDM9CP-6 antibody. C SDS-PAGE analysis of rCg14-3-3ε. Lane M: standard protein molecular weight marker; Lane 1: negative control (without induction); Lane 2: induced rCg14-3-3ε; Lane 3: purified rCg14-3-3ε. D Western blotting assay with Cg14-3-3ε antibody in haemocytes. Lane M: standard protein molecular weight marker; Lane 1: native Cg14-3-3ε protein in haemocytes by using Cg14-3-3ε antibody. E The PAMP binding affinity of CgDM9CP-6 by ELISA assay. F The microbial binding activity of rCgDM9CP-6 by Western blotting assay. G The co-localization between CgDM9CP-6 and microbes in haemocytes by Confocal analysis. Haemocytes were incubated with FITC-labeled E. coli, V. splendidus, Y. lipolytica, and S. aureus

There were bands observed after rCgDM9CP-6 was incubated with Gram-positive bacteria (M. luteus, S. aureus), Gram-negative bacteria (V. splendidus, E. coli), and fungi (Y. lipolytica, P. pastoris), while no band was observed in the rTrx group (Fig. 3F). The microbes stained by FITC were observed in green, while anti-CgDM9CP-6 combined with Alexa Fluor^® 647-conjugated antibody was observed in red (Fig. 3G). After haemocytes were incubated with the FITC-labeled microbes for 1 h and then added anti-CgDM9CP-6 as the primary antibody and Alexa Fluor 647-labeled Goat Anti-Mouse IgG as the second antibody. And the green signals of FITC-labeled microbes (V. splendidus, E. coli, and Y. lipolytica) were co-localized with the red signals of CgDM9CP-6 in haemocytes (Fig. 3G). The haemocyte ratios containing co-localization of CgDM9CP-6 and V. splendidius, E. coli and Y. lipolytical were 37.5%, 26.1% and 49.8%, respectively.

The interactions between Cg14-3-3ε and CgDM9CP-6, Cg14-3-3ε and CgRel

The interactions between rCg14-3-3ε and rCgDM9CP-6, and rCg14-3-3ε and rCgRel were analyzed by pull-down assay, BLI assay and Co-IP assay, respectively. There were two distinct bands (rCgDM9CP-6 and rCg14-3-3ε) observed in the elute liquid after the pull-down assay (Fig. 4A and B). The activity of rCg14-3-3ε to bind rCgDM9CP-6 was found to be of dose dependent. With the increase of rCg14-3-3ε concentration from 0.59 to 17.6 µM, the activities to bind rCgDM9CP-6 also increased. The affinity K_D of rCgDM9CP-6 to bind rCg14-3-3ε was 3.81 × 10⁻⁶ M (Fig. 4H). The bands of CgDM9CP-6 and Cg14-3-3ε were observed after the haemocyte lysates were coimmunoprecipitated by CgDM9CP-6 and Cg14-3-3ε, respectively (Fig. 4K). Similarly, two distinct bands (rCg14-3-3ε and rCgRel) were also observed in the elute liquid after the pull-down assay (Fig. 4C and D). In the control group, two proteins (rCgDM9CP-6 and rCgRel) were used as bait proteins to pull down rGST, and only the bands of rCgDM9CP-6 and rCgRel were observed in the elute liquid, respectively (Fig. 4E and G). When rCg14-3-3ε was used as the bait protein to pull down rTrx, only the band of rCg14-3-3ε was observed in the elute liquid (Fig. 4F). The activity of rCg14-3-3ε to bind rCgRel was found to be of dose dependent. With the increase of rCg14-3-3ε concentration from 0.59 to 17.6 µM, the activities to bind rCgRel also increased. The affinity K_D of rCg14-3-3ε to bind rCgRel was 8.26 × 10⁻⁷ M (Fig. 4I). There was no binding ability observed in the control group, when rTrx was reacted with the biotin-labeled rCg14-3-3ε (Fig. 4J). Similarly, the bands of Cg14-3-3ε and CgRel were observed after the haemocyte lysates were coimmunoprecipitated by CgRel and Cg14-3-3ε, respectively (Fig. 4L).

Fig. 4.

The interaction of rCgDM9CP-6, rCg14-3-3ε, and rCgRel. A Pull down by rCgDM9CP-6 (His). Lane 1, purified rCgDM9CP-6 (His); lane 2, purified rCg14-3-3ε (GST); lane 3, washed liquid; lane 4, eluted liquid. B Pull down by rCg14-3-3ε (GST). Lane 1, purified rCg14-3-3ε (GST); lane 2, purified rCgDM9CP-6 (His); lane 3, washed liquid; lane 4, eluted liquid. C Pull down by rCg14-3-3ε (GST). Lane 1, purified rCg14-3-3ε (GST); lane 2, purified rCgRel (His); lane 3, washed liquid; lane 4, eluted liquid. D Pull down by rCgRel (His). Lane 1, purified rCgRel (His); lane 2, purified rCg14-3-3ε (GST); lane 3, washed liquid; lane 4, eluted liquid. E Pull down by rGST. Lane 1, purified rCgDM9CP-6 (His); lane 2, purified rGST; lane 3, washed liquid; lane 4, eluted liquid. F Pull down by rTrx-His. Lane 1, purified rCg14-3-3ε (GST); lane 2, purified rTrx-His; lane 3, washed liquid; lane 4, eluted liquid. G Pull down by rGST. Lane 1, purified rCgRel (His); lane 2, purified rGST; lane 3, washed liquid; lane 4, eluted liquid. H The binding activity of biotin-labeled rCgDM9CP-6 (His-Tag) to rCg14-3-3ε (GST-Tag). I The binding activity of biotin-labeled rCgRel (His-Tag) to rCg14-3-3ε (GST-Tag). J The binding activity of biotin-labeled rTrx-His (His-Tag) to rCg14-3-3ε (GST-Tag). K The interaction between CgDM9CP-6 with Cg14-3-3ε in vivo by Co-IP assay. L The interaction between Cg14-3-3ε with CgRel in vivo by Co-IP assay

The three-dimensional structures were predicted using AlphaFold3 to analyze the potential binding sites of Cg14-3-3ε to bind CgDM9CP-6 and CgRel. Six potential binding sites (Asp226, Lys215, Tyr131, Asn43, Glu15, and Lys215) of Cg14-3-3ε could interact with (Arg18, Trp80, Asn73, Lys143, Lys143, and Asp79) of CgDM9CP-6, respectively (Fig. S3B). Similarly, thirteen potential binding sites (Lys153-Tyr563, Asn43-Thr460, Gln16-Lys372, Glu15-Arg373, Arg61-Asn447, Glu241-Lys441, Lys50-Thr454, Asn176-Leu451, Asn176-Asp452, Arg130-Asp452, Asn227-Arg449, Asp226-Arg449, and Asp226-Arg397) were revealed between Cg14-3-3ε and CgRel (Fig. S3C). They all could form stable interactions through hydrogen bonds and other non-covalent interactions.

Subcellular location of CgDM9CP-6 and Cg14-3-3ε in haemocytes

Immunocytochemistry assay was used to observe the subcellular location of CgDM9CP-6 and Cg14-3-3ε in haemocytes. Anti-CgDM9CP-6 and anti-Cg14-3-3ε combined with Alexa Fluor^® 488-conjugated antibody were all observed in green signals. The nucleus stained by DAPI was observed in blue signals. The positive green signals of CgDM9CP-6 and Cg14-3-3ε were observed in the cytoplasm of haemocytes in PBS or V. splendidus group (Fig. 5A and B).

Fig. 5.

CgRel nuclear translocation in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters, respectively. A-B Subcellular location of CgDM9CP-6 and Cg14-3-3ε in haemocytes. C CgRel nuclear translocation in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters after V. splendidus stimulation. D The statistical analysis of co-localization between CgRel and nucleus in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters after V. splendidus stimulation using Image-Pro Plus software. E RNAi efficiencies in CgDM9CP-6-RNAi oysters. F RNAi efficiencies in Cg14-3-3ε-RNAi oysters. G The CgRel distribution in nucleus and cytoplasm of haemocytes in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters after V. splendidus stimulation, respectively. Each value was shown as mean ± S.D. (N = 3). Asterisks: *p < 0.05 (t-test)

The translocation of CgRel in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters after V. splendidus stimulation

The expressions of CgDM9CP-6 and Cg14-3-3ε were knocked done by RNAi to analyze their regulation on the nuclear translocation of CgRel. The mRNA expression levels of CgDM9CP-6 and Cg14-3-3ε decreased significantly in haemocytes of CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters (0.28-fold and 0.22-fold compared with EGFP-RNAi oysters, p < 0.05), respectively (Fig. 6A and B). The band intensities of CgDM9CP-6 and Cg14-3-3ε were reduced in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters, respectively (Fig. 5E and F). Anti-CgRel combined with Alexa Fluor^® 488-conjugated antibody was observed in green signals and the nucleus stained by DAPI was observed in blue. In CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters, the green signals of CgRel in nucleus were reduced after V. splendidus stimulation (Fig. 5C). The co-localization values between CgRel and nucleus also decreased significantly in haemocytes, which were 0.44-fold (p < 0.05) and 0.47-fold (p < 0.05) of that in EGFP-RNAi oysters (Fig. 5D). The band intensities of CgRel were enhanced in haemocyte cytoplasm and they were inhibited in nucleus of CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters after V. splendidus stimulation, respectively, compared with that in EGFP-RNAi oysters (Fig. 5G; Fig. S2A, 2B).

Fig. 6.

The mRNA expressions of CgIL17s, Cglysozyme and CgBigDef1 in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters, respectively. A The mRNA expressions of CgDM9CP-6, CgIL-17s and Cglysozyme in CgDM9CP-6-RNAi oysters after V. splendidus stimulation by qPCR. B The mRNA expressions of Cg14-3-3ε, CgIL-17s and Cglysozyme in Cg14-3-3ε-RNAi oysters after V. splendidus stimulation by qPCR. C Analysis of the promoters of CgIL-17-3, CgIL-17-6, and Cglysozyme. The 5’ untranslated regions were obtained and analyzed using the online AliBaba2.1 tool (http://gene-regulation.com/). D The binding activity of CgRel to CgIL-17-3, CgIL-17-6 and CgLysozyme promoter regions. The haemocytes were collected as a pool for the ChIP assay. The immunoprecipitates were analyzed using RT-PCR with primers specific for the fragments containing the CgRel binding sites. E The binding activities of CgRel to CgIL-17-3, CgIL-17-6 and CgLysozyme promoter regions in CgDM9CP-6-RNAi or Cg14-3-3ε-RNAi oysters. The number showed relative ratio of the band intensity (anti-Rel/input) of each panel. F Dual luciferase report assay was used to analyze the transcription activity of CgRel to CgIL17-3, CgIL17-6 and CgLysozyme promoter regions in HEK293T cells. The pGL3-control was used as positive control. G-H The mRNA expressions of CgBigDef1 in CgDM9CP-6-RNAi or Cg14-3-3ε-RNAi oysters after V. splendidus stimulation. Each value was shown as mean ± S.D. (N = 3). Asterisks: *p < 0.05, **p < 0.01 (t-test)

The mRNA expressions of inflammatory cytokines and antimicrobial effectors in CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters after V. splendidus stimulation

The regulation of CgDM9CP-6 and Cg14-3-3ε on the mRNA expressions of inflammatory cytokines and antimicrobial effectors was further investigated after the expressions CgDM9CP-6 and Cg14-3-3ε were knocked down by RNAi. In CgDM9CP-6-RNAi oysters, the mRNA expression levels of CgIL17-3 (0.28-fold, p < 0.01), CgIL17-6 (0.18-fold, p < 0.01), CgLysozyme (0.19-fold, p < 0.05) and CgBigDef1 (0.08-fold, p < 0.01) in haemocytes decreased significantly at 12 h after V. splendidus stimulation, compared with that of EGFP-RNAi oysters, respectively (Fig. 6A and G). In Cg14-3-3ε-RNAi oysters, the mRNA expression levels of CgIL17-3 (0.20-fold, p < 0.01), CgIL17-6 (0.25-fold, p < 0.01), CgLysozyme (0.15-fold, p < 0.01) and CgBigDef1 (0.12-fold, p < 0.01) decreased significantly after V. splendidus stimulation, compared with that of EGFP-RNAi oysters, respectively (Fig. 6B and H).

NF-κB/Rel responsive elements were identified in the upstream sequences of CgIL17-3, CgIL17-6 and CgLysozyme (Fig. 6C). ChIP assay was performed to determine whether CgRel could bind the fragments containing the responsive elements of CgIL17-3, CgIL17-6 and CgLysozyme. The positive signals of NF-κB/Rel responsive elements from CgIL17-3, CgIL17-6 and CgLysozyme were observed from the immunoprecipitates when CgRel antibody was incubated with haemocytes as the pool for ChIP (Fig. 6D). In CgDM9CP-6-RNAi and Cg14-3-3ε-RNAi oysters, the positive signals of NF-κB/Rel responsive elements from CgIL17-3, CgIL17-6 and CgLysozyme were reduced in haemocytes (Fig. 6E), indicating that CgDM9CP-6 and Cg14-3-3ε regulated the CgRel-dependent transcription of CgIL17-3, CgIL17-6, and CgLysozyme.

The dual-luciferase reporter assay was engaged to examine the transcription activity of CgRel on CgIL17-3, CgIL17-6 and CgLysozyme promotors in HEK293T cells. The positive control group (pGL-control) presented strong relative luciferase activity about 10.9. As control, the elative luciferase activities of pGL3-basic + pcDNA3.1⁽⁺⁾ and pGL3-basic + pcDNA3.1/CgRel group were about 1.55 and 1.49, respectively. Compared with their respective pcDNA3.1⁽⁺⁾ control groups, the relative luciferase activities of pGL3-CgIL17-3 + pcDNA3.1-CgRel, pGL3-CgIL17-6 + pcDNA3.1-CgRel and pGL3-CgLysozyme + pcDNA3.1-CgRel increased significantly, which were 1.94-fold 1.87-fold, 2.13-fold compared with that in pcDNA3.1/CgRel group, respectively, p < 0.01 (Fig. 6F).

Discussion

DM9CP is recently reported to function as PRR to trigger various immune responses [6–9, 18]. Like the lectins in invertebrates [56, 57], DM9CPs act as extracellular PRRs to bind various microbes [6–8] and mediate phagocytosis [6], agglutination [6, 9], and encapsulation [8]. Recently, CgDM9CP-5 in oyster cell-free haemolymph was reported to interact with membrane CgIntegrin to activate MAPK signals after it recognized and bound V. splendidus [7]. In the present study, a novel cytosolic CgDM9CP6 was characterized to bind various PAMPs and microbes and interact with Cg14-3-3ε to promote CgRel translocation into haemocyte nucleus, which eventually induced the mRNA expressions of IL-17s and lysozyme to defend against microbial invasion.

In invertebrates, DM9CPs with one or more DM9 domains are widely distributed in various tissues and display the activity to bind various bacteria [6–9, 16, 18, 22]. In mosquito, Pfs47 receptor protein had four DM9 domains [16], and in crab and oyster, DM9CPs normally have two DM9 domains [7, 22]. In the present study, DM9CPs were only in teleost and lamprey and they except for containing DM9 domain also had an Aerolysin domain, and the Aerolysin domain is as a cytolytic toxin exported by Aeromonas hydrophila, indicating that Aerolysin domain was inserted into the DM9CPs in fish and lamprey. In invertebrates, DM9CPs were presented in multiple phyla, such as Urochordata, Arthropoda, Mollusca, Annelida, Platyhelminthes, and Coelenterata. While their amino acid sequences were not conserved in different species, suggesting that they were not evolutionarily conserved. According to the evolutionary tree of DM9CPs, there were two branches (Vertebrate branch and Invertebrate branch). The DM9CPs from oyster were close to those from anemone and coral, and they all had two DM9 repeat domains. While the domains in DM9CPs from different species of phyla were not always the same, indicating that DM9 domains were lost, or re-organized and other domains were inserted into the DM9CPs during evolution. The results also suggested that their functions might be different and more complex. DM9CPs are mainly distributed in different immune tissues and their mRNA expressions are induced after immune stimulation [6–9, 16, 18, 22]. In the present study, CgDM9CP-6 mRNA was found to be highly expressed in haemocytes, similarly to CgDM9CP-1/4 identified previously [6, 18]. DM9CP is documented to function as PRR to bind different PAMPs and microbes [6–9, 16, 18]. In mosquito, Pfs7Rec was able to bind Pfs7 of P. falciparum [16]. In crab and oyster, EsDM9CP and CgDM9CPs exhibited binding affinity to different PAMPs, and Gram-positive bacteria, Gram-negative bacteria, and fungi [6–9, 18, 22]. In the present study, rCgDM9CP-6 was also found to bind multiple PAMPs (LPS, PGN, MAN, and D-mannose), and various microbes such as Gram-positive bacteria (M. luteus, S. aureus), Gram-negative bacteria (V. splendidus, E. coli), and fungi (Y. lipolytica, P. pastoris). In our previous study, most DM9CPs could secrete into cell-free hemolymph, which could function as extracellular PRRs to sense microbes and then bind with membrane receptor to trigger intracellular immune signals. While in the present study, it was interesting that CgDM9CP-6 could only distribute in hemocytes, while not cell-free hemolymph. The intracellular CgDM9CP-6 co-localized with various microbes (V. splendidus, E. coli, and Y. lipolytica) suggesting most pathogens that were engulfed by oyster hemocytes might occur evasion in hemocytes and could be sensed again by intracellular PRRs. The results indicated that CgDM9CP-6 was able to function as an intracellular PRR to recognize the invading microbes in haemocytes.

After recognizing microbes and PAMPs, PRRs often bind with intracellular proteins to mediate signal activation and cascade amplification. Most DM9CPs were secreted into the cell-free haemolymph and some of them could be detected in haemocytes [6–9, 18, 20]. The extracellular DM9CP upon recognizing PAMPs normally binds to the membrane receptor to activate downstream immune signaling. For example, in C. gigas, CgDM9CP-5 as secreted PRR upon recognizing microbes could interact with the membrane receptor CgIntegrin to activate the MAPK signaling pathway in haemocytes [7]. In the present study, CgDM9CP-6 was mainly located in haemocyte cytoplasm, and it was co-localizated with V. splendidus in haemocytes. 14-3-3ε mainly exists in the cellular cytoplasm and functions as a scaffolding protein in the signaling pathway [27, 31–33]. In the present study, Cg14-3-3ε was identified from C. gigas with a conserved 14-3-3ε domain, indicating that Cg14-3-3ε might have the same function as 14-3-3ε. Apart from this, Cg14-3-3ε mRNA was highly expressed in haemocytes and its protein was mainly distributed in the cytoplasm, which both were similar as that of CgDM9CP-6. Also, Cg14-3-3ε mRNA expression increased significantly in haemocytes after V. splendidus stimulation. The results indicated that Cg14-3-3ε and CgDM9CP-6 might have some relationship in defending against microbial invasion. Further study demonstrated that CgDM9CP-6 locating in haemocyte cytoplasm interacted with Cg14-3-3ε. The molecular docking analyses were performed to investigate the binding sites and potential competitive relationships between these interactions. The binding sites of CgDM9CP-6 with LPS and Cg14-3-3ε were distinct and non-overlapping. The results demonstrated that CgDM9CP-6 in haemocyte cytoplasm was able to interact with Cg14-3-3ε upon recognizing intracellular V. splendidus.

14-3-3ε plays an important role in mediating the NF-κB signaling pathway to induce the expressions of immune effectors. For example, in human hepatocellular carcinoma, 14-3-3ε enhanced NF-κB activation and promoted nuclear translocation of NF-κB [58]. In mouse, both global and myeloid-specific deletion of 14-3-3ε induced NF-κB activation resulting in exacerbated inflammatory arthritis [32]. While, in other species, there are still no reports about the relation between 14-3-3ε and NF-κB. In the present study, Cg14-3-3ε was able to interact with NF-κB/Rel to induce NF-κB/Rel nuclear translocation in haemocytes. The direct relation between 14-3-3ε and NF-κB/Rel was proposed, which was helpful for understanding the function of 14-3-3ε in mediating the NF-κB signaling pathway in molluscs. As NF-κB/CgRel normally was combined with CgIκB (IκB-Rel complex) located in the cytoplasm and Cg14-3-3ε might promote the release of CgRel from the IκB-Rel complex. Of course, the detailed mechanism should be confirmed in the future study. 14-3-3 supports the production of several cytokines, such as IL-13, IFN-γ, and IL-17 A in human peripheral blood mononuclear cells (PBMC) and CD8⁺ T cells [59, 60]. In Drosophila melanogaster, 14-3-3ε was required for the Rab11-positive vesicle function, which in turn enabled antimicrobial peptide secretion [61]. In the present study, CgDM9CP-6 and Cg14-3-3ε could regulate the mRNA expressions of CgIL17-3, CgIL17-6, CgLysozyme and CgBigDef1 in haemocytes after V. splendidus stimulation. And NF-κB/Rel could bind to the promoter regions of CgIL17-3, CgIL17-6, and CgLysozyme. IL17 has the function of inducing antimicrobial peptides (AMPs). Also, CgNF-κB/Rel was reported to regulate the mRNA expressions of CgIL17-3 and CgIL17-6 [51]. The results collectively indicated that there was an intracellular DM9CP-6/14-3-3ε/NF-κB signaling pathway, which regulated the mRNA expressions of IL17s and AMPs in oyster haemocytes.

In conclusion, CgDM9CP-6 could function as an intracellular PRR to recognize different microbes, which then interacted with scaffolding protein Cg14-3-3ε to recruit CgNF-κB/Rel. The recruited CgNF-κB/Rel eventually translocated into haemocyte nucleus to induce the mRNA expressions of CgIL17-3, CgIL17-6, CgLysozyme, and CgBigDef1 (Fig. 7). These results demonstrated that the intracellular DM9CP-6-14-3-3ε-NF-κB/Rel pathway regulated the expressions of IL17s, lysozyme, and BigDef1 in haemocytes of oysters, which provided reference for studying the intracellular PPR-mediated antibacterial immune signaling pathway. The present study clarified that the intracellular PRR-mediated antibacterial immune pathway in oysters could monitor and promptly eliminate pathogens that escaped into the cells. This discovery contributed to understand the interaction mechanism between intracellular pathogens and hosts in bivalves and provide a theoretical basis for the prevention and control of bivalve pathogens.

Fig. 7.

The intracellular DM9CP-6-14-3-3ε-Rel signaling pathway in regulating the mRNA expressions of CgIL17s, Cglysozyme and CgBigDef1 in oysters

Authors’ contributions

Conceptualization: J.S., and L.S.; methodology: J.S., and Y.L.; formal analysis: J.S., and Y.L.; investigation: Y.L., and Y.L.; resources: J.S., L.W., and L.S.; writing – original draft: J.S., and Y.L.; writing – review and editing: J.S., L.W., and L.S.; visualization: J.S., and Y.L.; supervision: J.S., L.W., and L.S.; project administration: J.S., L.W., and L.S.; funding acquisition: J.S., and L.S.

Funding

This research was supported by grants from National Natural Science Foundation of China (32222086, 32230110), the fund for China Agriculture Research System (CARS-49) and Outstanding Talents and Innovative Teams of Agricultural Scientific Research in MARA, the innovation team of Aquaculture Environment Safety from Liaoning Province (LT202009), and Liaoning Revitalization Talents Program (XLYC2203087).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.

Conflict of interest

No potential conflict of interest was reported by the author(s).