Taq DNA polymerases, RNAiso Plus, and First Strand cDNA Synthesis Kit were obtained from TaKaRa (Dalian, China). LPS from Escherichia coli 0111:B4 and peptidoglycan (PGN) and LTA from Staphylococcus aureus were obtained from Sigma (St. Louis, MO, USA). Trehalose from Saccharomyces cerevisiae was purchased from Amresco (Solon, OH, USA).

Mud crabs (~150 g each) purchased from an aquaculture farm in Chongming Country (Shanghai, China) were acclimated in 400 L tanks with aerated seawater for a week (mud crabs are not endangered or protected species, and no special permissions are required). Healthy crabs were randomly selected to investigate the tissue distribution and expression profiles of SpBark. For immune stimulation, Staphylococcus aureus and Vibrio parahemolyticus were cultured overnight in Luria–Bertani medium, collected by centrifugation, and re-suspended in PBS (10 mM Na₂HPO4, 1.8 mM KH₂PO4, 140 mM NaCl, and 2.7 mM KCl; pH 7.3). After washing twice with PBS, the resultant suspensions were used as bacterial inocula. Subsequently, 200 μL of each inoculum (2 × 10⁷ CFU in PBS) was injected into the base of the right fifth leg of each crab. The corresponding control was challenged with the same volume of PBS. At each time point after injection (0, 2, 6, 12, 24, 48, and 72 h), hemolymph was drawn from the base of the legs by using a syringe preloaded with ice-cold anticoagulant buffer (0.45 M NaCl, 0.1 M glucose, 30 mM trisodium citrate, 26 mM citric acid, and 10 mM ethylenediaminetetraacetic acid; pH 4.6) (34). The hemocyte pellets were collected for RNA extraction after centrifugation at 800 × g for 10 min at 4°C. In addition, other tissues, including the heart, gills, hepatopancreas, stomach, intestine, connective tissue, and muscle, of healthy crabs were dissected; washed with sterile PBS; and collected for RNA extraction. For each tissue sample, at least three crabs were used to eliminate individual differences. Another two batches of previously isolated RNA samples were used to eliminate the differences among batches. All animal experimental procedures were conducted in accordance with the National Institutes of Health's Guide for the Care and Use of Laboratory Animals.

The total RNA from hemocytes and other collected tissues was isolated with RNAiso Plus, and DNase I (Promega, USA) was added into the extracted total RNA to remove the contaminated DNA. The first stand cDNA was synthesized using the First Strand cDNA Synthesis Kit (TaKaRa, Dalian) in accordance with the manufacturer's instructions.

The original cDNA sequence of SpBark was obtained through high-throughput transcriptome sequencing with an RNA mixture isolated from the hemocytes, gill, and hepatopancreas, and it was further confirmed by PCR with five pairs of primers (Table 1). The PCR was performed under the following parameters: 95°C for 5 min; 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 60 s; and finally 72°C for 10 min. The DNA products were separated by agarose gel electrophoresis, subcloned into a pMD-19T vector, and sequenced via a commercial company (Sangon, China).

The identities of SpBark with other Bark-like proteins were revealed using the online Basic Local Alignment Search Tool Program (BLASTP) (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The domain architecture was predicted using simple modular architecture research tool (SMART) (http://smart.embl-heidelberg.de). Multiple alignment was generated through the ClustalX 2.0 program (http://www.ebi.ac.uk/tools/clustalw2) and GENEDOC software. The putative amino acid sequence was deduced and predicted on http://web.expasy.org/translate/. The signal peptide was predicted with SignalP (http://www.cbs.dtu.dk/services/SignalP/). The pI and molecular weight (MW) were calculated on http://web.expasy.org/compute_pi/. MEGA 7.0 was used to construct a neighbor-joining tree, and the bootstrap of 1,000 was selected to assess reliability.

qRT-PCR was conducted to investigate the relative expression level of SpBark in a real-time thermal cycler Quantstudio 6 Flex (ABI, USA) according to a previous protocol (35). A pair of primers (SpBarkRF and SpBarkRR) was designed to generate an amplicon of SpBark. Another pair of primers (18SRF and 18SRR) was synthesized to produce a 121-bp fragment of 18S rRNA as reference. qRT-PCR was programed as follows: an initial denaturation at 94°C for 3 min; 40 cycles of 94°C for 10 s and 60°C for 1 min; and finally a melting curve analysis from 65 to 95°C. The total volume of reaction mixture was 20 μL (10 μL of 2 × SYBR Premix Ex Taq, 2 μL of cDNA, and 4 μL of each primer). All tests were performed thrice with individual templates. The algorithm of 2^−ΔCT was used to investigate tissue distribution of SpBark. To analyze the time-course expression profiles, the formula of 2^−ΔΔCT was used to normalize the data in two steps. First, the expression of SpBark was normalized to the reference gene (18S rRNA) in the same sample. Second, the relative expression of SpBark in the experimental sample was normalized to that in the control sample. Unpaired Student's t-test was used to analyze significant differences (^*P < 0.05; ^**P < 0.01).

To explore the potential immune function of SpBark, four predicted functional domains (SRCRD1, SRCRD2, SRCRD3, and CTLD) were overexpressed using an E. coli expression system. On the basis of the SpBark cDNA sequence, four primer pairs (SRCRD1EF and SRCRD1RF, SRCRD2EF and SRCRD2ER, SRCRD3EF and SRCRD3ER, and CTLDEF and CTLDER) were designed to produce DNA fragments encoding SRCRD1, SRCRD2, SRCRD3, and CTLD, respectively (Table 1). These DNA fragments were digested and then inserted into the pET32a expression vectors. The constructed plasmids were transformed into competent E. coli cells for overexpression. Recombinant proteins were induced by adding isopropyl-β-d-thiogalactoside (IPTG) to a final concentration of 0.5 mM at 37°C for 6 h and purified using High Affinity Ni-NTA Resin (GenScript, Nanjing) in accordance with the manufacturer's instructions.

Nine kinds of microorganisms, including four Gram-negative bacteria (Vibrio harveyi, V. parahemolyticus, V. alginolyticus, and E. coli), three Gram-positive bacteria (S. aureus, Bacillus subtilis, and B. megaterium), and two fungi (Candida albicans and Pichia pastoris), were used to investigate the microorganism-binding activities of SRCRD1, SRCRD2, SRCRD3, and CTLD. The experiment was performed following a protocol used in our previous study (36). In brief, the microorganisms (1 × 10⁸ CFU) were incubated in 1 mL of each recombinant protein (200 μg/mL) with gentle rotation for 30 min at room temperature. The microorganisms were pelleted, washed four times with 1 mL of TBS (150 mM NaCl, 10 mM Tris–HCl, pH 7.5), and eluted with 200 μL of 7% SDS solution with mild agitation for 5 min at room temperature. The microorganism pellets were rewashed thrice with 1 mL TBS. Finally, the elution and pellet of each microorganism were sampled and analyzed on 12.5% SDS-PAGE. The recombinant proteins (SRCRD1, SRCRD2, SRCRD3, and CTLD) were sampled as the positive controls. The binding activity to microorganisms was determined by Western blot. After the proteins were transferred onto a PVDF membrane, the membrane was blocked with 4% bovine serum albumin (BSA) in TBS and incubated with peroxidase-conjugated monoclonal antibody against His-tag (GenScript, Nanjing). The target proteins were visualized with an ECL Western blot reagent kit.

Enzyme-linked immunosorbent assay (ELISA) was performed to test the binding abilities of four domain proteins (SRCRD1, SRCRD2, SRCRD3, and CTLD) to microbial polysaccharides and acetylated LDL (ac-LDL) following a previously described protocol with slight modifications (37). Microbial polysaccharides, including PGN and LTA from S. aureus, LPS from E. coli 0111:B4, and β-glucan from Laminaria digitata, together with human ac-LDL were used in this study. Each well of the microtiter plate was incubated with 100 μL of ac-LDL (50 μg/mL) at 4°C overnight or microbial polysaccharide (20 μg/mL) at 37°C until the plate came to desiccation. Wells incubated with 100 μL of distilled water were employed as blank control. Each well was blocked with BSA (1 mg/mL, 200 μL) at 37°C for 2 h and washed four times with TBST (0.05% Tween 20 in TBS). Subsequently, a series of diluted recombinant protein SRCRD1, SRCRD2, SRCRD3, CTLD, or TRX (0–5 μM in TBS containing 0.1 mg/mL BSA) was added to polysaccharide-coated plates. TRX was used as negative control. The serially diluted recombinant proteins supplemented with 5 mM of CaCl₂ were added to ac-LDL-coated plates. After incubation at room temperature for 3 h, the plates were rinsed with TBST four times. Each well was then incubated with 100 μL of peroxidase-conjugated mouse monoclonal anti-His antibody (1:5,000 dilution in TBS with 0.1 mg/mL BSA) at 37°C for 2 h. The color reaction was developed with 0.01% 3,3′,5,5′-tetramethylbenzidine (Sigma, USA) and stopped by adding 2 M H₂SO₄. The absorbance was read at 405 nm on a Spark 10 M microplate reader (Tecan, Switzerland). The assay was conducted in triplicate. For a competitive binding assay, purified recombinant CTLD was diluted to a final concentration of 0.25 μM and pre-incubated with different LPS concentrations (0.0005, 0.0025, 0.005, 0.025, 0.05, 0.25, 0.5, and 2.5 μM) by gentle shaking at room temperature for 2 h (38). Subsequently, the pre-incubated CTLD protein (100 μL per well) was added to the plates coated with LPS. The following steps were the same as those in ELISA above. The competitive binding assay was repeated three times.

The microorganisms used in microorganism-binding assay were applied for agglutination assay following a method described by (36). Microorganisms were harvested at the mid-logarithmic growth phase by centrifugation at 5,000 × g for 5 min, re-suspended in TBS after washing three times, and finally adjusted to 2 × 10⁸ cells/mL (for bacteria) or 2 × 10⁷ cells/mL (for fungi). The microorganism suspensions were mixed with equal volume (100 μL) of diluted recombinant protein (SRCRD1, SRCRD2, SRCRD3, CTLD, or TRX) in TBS at the protein concentration range of 0.48–15.2 μM with or without 10 mM CaCl₂ at 28°C for 30 min. TRX tag protein served as the negative control. Agglutination was observed under a light microscope.

A partial SpBark DNA fragment, which covered the SpBark CTLD, was amplified by PCR with specific primers linked to the T7 promoter (Table 1) and used as the template to generate dsSpBark (SpBark dsRNA) with T7 RiboMAX™ Express RNAi System (Promega, USA). The dsGFP (EGFP dsRNA) was synthesized as control with primers listed in Table 1. To obtain better RNA interference efficiency, the crabs with an average weight of ~25 g were selected and divided into two groups (four crabs in each group). Each crab in the experimental group was injected with 25 μg of dsSpBark into the base of the fourth leg. The crabs in the control group were injected with an equal amount of dsGFP. A second dsRNA injection was performed 24 h later in the same way. Hemocytes were collected from each group at 40 and 48 h after the first dsRNA injection. RNAi efficiency was assessed by qRT-PCR using the total RNAs from hemocytes with a pair of primers located outside the above DNA fragment (Table 1). Unpaired t-test was used to analyze significant differences (^*P < 0.05; ^**P < 0.01).

V. parahemolyticus was incubated with recombinant CTLD protein and then injected into crabs to determine whether coating bacteria with CTLD could influence bacterial clearance. Approximately 500 μL of CTLD protein in PBS (400 μg/mL) was mixed with an equal volume of V. parahemolyticus (1 × 10⁸ CFU) with gentle rotation at 28°C for 60 min. The same number of V. parahemolyticus incubated with TRX tag protein (400 μg/mL) and equal volume of PBS were also prepared to serve as controls. Mud crabs were divided into three groups (four crabs in each group). After incubation, the mixtures (200 μL) were injected into crabs. At each time point post-injection (5, 15, and 30 min), hemolymph (200 μL) was collected from crabs and mixed with an equal volume of anticoagulant buffer. After serial dilution with PBS, the samples (100 μL) were plated onto the LB plates. These plates were then incubated at 37°C overnight, and the number of bacterial clones on the plates was counted, which was used to determine the number of residual bacteria in hemolymph. After validating that SpBark expression could be silenced by injection of dsSpBark, we investigated whether SpBark knockdown could affect bacterial clearance. Each crab was injected with 200 μL of V. parahemolyticus suspension (2 × 10⁷ cells) at 40 h after injection with dsSpBark or dsGFP. After mock PBS injection, the crabs were treated with an equal amount of bacteria in the same way. The number of residual bacteria in hemolymph was calculated using the method described above. The experiments were performed three times, and the results presented the mean of three individual experiments. Unpaired t-test was used to analyze significant differences (^*P < 0.05; ^**P < 0.01).

The labeling of V. parahemolyticus with fluorescein isothiocyanate isomer I (FITC, Sigma) was conducted as described earlier (32). Overnight cultures of V. parahemolyticus were washed twice with PBS, re-suspended in carbonate buffered saline (pH 9.5) containing 0.2 mg/mL FITC, and then incubated at 25°C for 1 h. Subsequently, the FITC-labeled bacteria were rewashed four times to remove dissociated FITC and then re-suspended in PBS (~1 × 10⁹ CFU/mL). Approximately 200 μL of FITC-labeled bacteria was injected into the mud crabs at 40 h after injections with dsSpBark and dsGFP. Hemolymph was collected at 30 min after bacterial injection using a syringe preloaded with anticoagulant containing 4% paraformaldehyde. After incubation at 4°C for 15 min, the hemocytes were centrifuged and re-suspended in anticoagulant. Trypan blue (2 mg/mL, Sigma) was added into the hemocyte suspension and incubated for 5 min to quench the fluorescence of non-phagocytized bacteria. After washing with PBS twice, the fixed hemocytes were dropped onto poly-L-lysine-coated glass slides, and the phagocytized bacteria were evidently detected under a fluorescence microscope (Nikon Eclipse Ti-E, Japan). A total of 600 cells (each group) were counted to determine the phagocytic percentage, which was defined here as [hemocytes ingesting bacteria/all cells observed or tested] × 100%. The phagocytosis assay was conducted three times. The data were subjected to Student's t-test, and differences with P < 0.05 were considered statistically significant.

The full-length cDNA sequence of SpBark is 10,053 bp long, including a 24-bp 5′ untranslated region (UTR), a 1,437-bp 3′ UTR containing a poly(A) tail, and an 8,592-bp open reading frame encoding a polypeptide of 2,863 amino acids (GenBank Accession No. MH595537) (Figure 1S). The deduced protein had an 18-amino-acid signal peptide at the N-terminus and a transmembrane region close to the C-terminus. Its mature peptide had an estimated MW of 320.8 kDa and a theoretical pI of 5.58. Three putative SRCR domains, including SRCRD1 (Leu¹⁰⁶ to Asp²¹⁰), SRCRD2 (Val⁹⁷⁶ to Tyr¹⁰⁸²), and SRCRD3 (Val¹⁷⁹⁷ to Gly¹⁹¹⁴), and a following CTLD (Cys²⁵¹² to Glu²⁶⁰⁸) were found in this deduced protein (Figure 1A). The pIs of SRCRD1, SRCRD2, SRCRD3, and CTLD were 5.12, 4.58, 5.48, and 4.95, respectively.

The BLASTP search analysis demonstrated that SpBark shared 71–73% identity with four Bark beetle-like proteins from Penaeus vannamei (four isoforms: ROT62695, XP_027233243, XP_027233244, and XP_027233245). SpBark also shared 59% identity with Hyalella azteca Bark-like protein (XP_018022137) but no more than 42% identity with other putative Bark-like proteins. The alignment of three SRCR domains revealed that six cysteine residues in these SRCR domains, which might be responsible for forming three disulfide bonds to stabilize the dimensional structure of SRCR domains, were well-conserved even though these three amino acid sequences shared a low identity of 29.4% (Figure 1B). Another two cysteine residues at the N-terminus of SRCRD3 were also found, suggesting that SRCRD3 may have a different structure from SRCRD1 and SRCRD2. Alignment of the CTLDs of lectins from mud crab, shrimp, human, and amphioxus showed that four cysteine residues (Cys49, Cys127, Cys142, and Cys150), which can form two disulfide bonds to stabilize the structure of classic CTLDs, were well-conserved in SpBark CTLD. Two additional cysteine residues, which may form another disulfide bond, were found at the N-terminus of each domain. Moreover, some remarkable amino acid motifs, such as EPN (Glu-Pro-Asn), QPD (Gln-Pro-Asp), and WND (Trp-Asn-Asp), responsible for carbohydrate binding were not found in SpBark CTLD. These results suggested that SpBark CTLD was unique with different sequence information from the classic CTLDs (Figure 1C).

On the basis of BLASTP results, a phylogenetic tree was constructed using Bark-like proteins to analyze the evolutionary relationships among them. In this tree, Bark-like proteins were divided into two large clusters (Figure 1D). SpBark together with other crustacean Bark-like proteins was grouped into the crustacean cluster, and the other Bark-like proteins were clustered into the insect group. SpBark and four shrimp Bark-like proteins formed a meaningful subcluster with the node value of 99, indicating that Bark-like proteins may widely exist in decapod crustaceans possessing special biological function different from the ones in insects.

qRT-PCR analysis revealed that SpBark was highly expressed in the hepatopancreas, hemocytes, gills, intestine, and stomach, and it was detected in the heart and muscle at a low expression level (Figure 2A). The temporal expression profiles of SpBark in hemocytes after bacterial challenges were further investigated to explore whether this gene participated in immune defense. SpBark was remarkably increased 2–6 h after challenge with V. parahemolyticus and reached the highest level (~7 fold increase) at 6 h post-injection. With regard to the challenge with S. aureus, significant induction (~2 fold increase) was observed at 12 h post-injection (Figure 2B). These results suggested that SpBark was involved in antibacterial responses of mud crabs.

The recombinant proteins of the four structural domains of SpBark were overexpressed in soluble form and then purified by Ni-NTA His Bind Resin. The predicted MWs of recombinant SRCRD1, SRCRD2, SRCRD3, and CTLD (each fused with a ~16 kDa Trx tag) were approximately 27.8, 28.0, 29.7, and 27.4 kDa, respectively. The position of each purified protein is roughly in agreement with the size of the corresponding recombinant protein (Figure 3A).

The microbial cell-binding activities of the four structural domain proteins were tested by Western blot. The recombinant protein was detected in pellets representing that it possessed strong binding activity, and the one in elution meant that it exhibited weak binding toward microorganisms. On the basis of these criteria, CTLD exhibited strong binding activity to all tested microorganisms, including four Gram-negative bacteria (V. harveyi, V. parahemolyticus, V. alginolyticus, and E. coli), three Gram-positive bacteria (S. aureus, B. subtilis, and B. megaterium), and two fungi (C. albicans and P. pastoris). By contrast, SRCRD2 and SRCRD3 only exhibited weak binding activities to different kinds of microorganisms, and SRCRD1 possessed no evident binding activity to microorganisms (Figure 3B). These results implied that SpBark CTLD was the major functional domain involved in bacterial recognition.

The binding activities of SpBark SRCRD1, SRCRD2, SRCRD3, and CTLD to ac-LDL and several kinds of polysaccharides were investigated by performing ELISA. In the ac-LDL binding assay, only three of the four domain proteins (SRCRD1, SRCRD2, and CTLD) exhibited ac-LDL-binding activities in varying degrees (Figure 3E). Among them, SpBark CTLD possessed the strongest binding activity, and SRCRD1 displayed more potent binding activity than SRCRD2. Given that the binding activity to mLDL is a marked property of scavenge receptors, SpBark possessing ac-LDL binding activity suggested that it might function as a scavenger receptor in mud crab. In addition, SRCRD1, SRCRD3, and CTLD exhibited notable LPS-binding activities, but none of the four domain proteins displayed apparent binding activities to the other tested polysaccharides, such as LTA, PGN, and trehalose; this finding indicated that the microorganism-binding activity exhibited by CTLD might be through binding to LPS or other untested components on the cell surface (Figures 3C,F–H). This study also revealed that CTLD displayed a much stronger LPS-binding activity than SRCRD1 and SRCRD3. The specific LPS-binding activity exhibited by CTLD was further determined by a competitive binding assay. The results demonstrated that the binding ability of CTLD to immobilized LPS on the microplates was significantly reduced by pre-incubation of CTLD with LPS, and the binding activity to immobilized LPS became weak with increasing amount of LPS to pre-incubate CTLD (Figure 3D). These results suggested that the SpBark binding activity to Gram-negative bacteria was mainly through binding LPS with its CTLD.

The agglutination activities of SRCRD1, SRCRD2, SRCRD3, and CTLD were investigated to explore whether they were involved in bacterial agglutination. Results demonstrated that only SpBark CTLD of these four proteins exhibited apparent agglutination activity in the presence of Ca²⁺ (Figure 4). The agglutination activities of CTLD to different microorganisms greatly varied. Table 2 shows that the minimal agglutinating concentration (MAC) of CTLD to Gram-positive bacteria S. aureus and B. megaterium was the same as that to fungus P. pastoris, which was lower than the MACs to Gram-negative bacteria. This finding indicates that CTLD possessed stronger agglutinating abilities to Gram-negative bacteria than the other tested microorganisms. SpBark CTLD could agglutinate bacteria, suggesting that it can function like many common CTLDs acting as a binding site of PRR.

Given that SpBark CTLD exhibited specific binding abilities to LPS and Gram-negative bacteria, we investigated whether this protein facilitated bacterial clearance in crabs through its binding activity. V. parahemolyticus pre-incubated with CTLD, Trx protein, or PBS was injected into healthy crabs. The number of bacteria in vivo was significantly decreased at 15 min post-injection by pre-incubating V. parahemolyticus with CTLD compared with that treated with Trx or PBS; however, no significant differences in bacterial numbers were found at 5 min post-injection among these three groups. At 30 min after the injection of bacteria treated with CTLD, an extremely significant decrease in the number of bacteria (P < 0.01) in crabs was observed compared with that in control crabs (Figure 5). These results revealed that pre-incubation of V. parahemolyticus with CTLD would facilitate bacterial clearance in vivo.

To investigate the in vivo function of SpBark, RNAi of SpBark and bacterial clearance assays were performed. qRT-PCR analysis indicated that the transcripts of SpBark in hemocytes significantly declined at 40 and 48 h after the first injection of dsSpBark (Figure 6A). This result indicated that the injection of SpBark dsRNA into crabs could remarkably suppress SpBark expression. After SpBark knockdown, V. parahemolyticus was injected into crabs, and the residual number in hemolymph was counted to assess the bacterial clearance ability. As shown in Figure 6B, the residual number of V. parahemolyticus in vivo significantly increased at 15 and 30 min post-injection compared with that in dsGFP or PBS group. This result suggested that the bacterial clearance ability in hemolymph of SpBark-silenced crabs was severely impaired.

To reveal the mechanism by which SpBark facilitates bacterial clearance, bacterial phagocytosis assay was conducted. After SpBark was silenced, FITC-labeled bacteria were injected into crabs. Hemocytes were collected to determine the phagocytosis rate. The results showed that the phagocytic rate of hemocytes in SpBark-silenced crabs was significantly lower than that in control crabs: the phagocytosis rate was decreased by ~45% by knockdown of SpBark (from 9.7 to 5.3%) (Figures 6C,D). This result suggests that SpBark modulated hemocyte phagocytosis of V. parahemolyticus.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.