To identify IRF genes, the transcriptome and whole genome sequence databases of the C. farreri were searched using the available IRF protein sequences from representative invertebrates and vertebrates, including Homo sapiens, Mus musculus, Gallus gallus, Xenopus tropicalis, Danio rerio, C. gigas, P. fucata, Mytilus galloprovincialis, Hyriopsis cumingii, Pecten maximus, Lottia gigantea, Biomphalaria glabrata, and Elysia chlorotica. These IRF proteins from representative species were retrieved from NCBI (http://www.ncbi.nlm.nih.gov), Ensembl (http://useast.ensembl.org), MolluscDB (http://mgbase.qnlm.ac/), and OysterBase (http://www.oysterdb.com/) databases. TBLASTN was used to obtain the initial pool of IRFs transcriptome sequences from the Zhikong scallop, and then, BLASTN was performed to verify the cDNA sequences by comparing the transcriptome sequences with the whole genome sequences. The candidate CfIRFs sequences were submitted to the ORF Finder program (https://www.ncbi.nlm.nih.gov/orffinder/) to predict the open reading frame (ORF), and the ORFs were translated into amino acid sequences. The translated sequences were submitted to the SMART program (http://smart.embl-heidelberg.de/) for identification of the signal peptide and other conserved domains. The putative isoelectric point (pI) and molecular weight were computed using the Compute pl/Mw (http://web.expasy.org/compute_pi/). The subcellular localization and nuclear localization signals (NLSs) were predicted through the online prediction website (http://www.csbio.sjtu.edu.cn/bioinf/euk-multi-2/, https://www.genscript.com/wolf-psort.html/, https://sunflower.kuicr.kyoto-u.ac.jp/~smatsuda/slplocal.html, http://nls-mapper.iab.keio.ac.jp/cgi-bin/NLS_Mapper_form.cgi). The protein structures of all the identified IRF proteins were drawn with IBS1.0.3 software. Multiple alignment analysis of CfIRFs were performed with the ClustalW multiple alignment programs (http://www.ebi.ac.uk/clustalw/).

The IRF proteins from other vertebrates and invertebrates listed in the Section 2.1 were used for phylogenetic analysis together with the Zhikong scallop IRFs. The amino acid sequences of IRF proteins from these species were retrieved from the NCBI and Ensembl Genome Browser. Protein sequences were aligned using the ClustalW method in the MEGA-X software (Sudhir et al., 2018). Phylogenetic relationships of IRF amino acid sequences were estimated using maximum-likelihood (ML) analyses with FastTree 2.0.0. FastTree accounts for variable rates of evolution across sites by assigning each site to one of 20 categories, with the rates geometrically spaced from 0.05 to 20. FastTree sets each site to its most likely category by using a Bayesian approach with a gamma prior. Branch supports evaluated 10,000 pseudo-replicates of the ultrafast bootstrap procedure (Thi et al., 2017). Whole amino acid sequences were used in the phylogenetic analyses. This analysis involved a total of 59 amino acids across 13 species. The accession numbers of 59 IRFs are listed in Supplementary Table 1 .

For expressional analysis, the RPKM (reads per kilo per million reads) value of each IRF gene was retrieved from the published RNASeq datasets of various developmental stages and adult tissues of Zhikong scallop (Li et al., 2017). To visualize the expression patterns of IRF genes in Zhikong scallop, the expressional heatmaps were shown via heatmap package under the R environment and the statistical analysis of the data was performed with edgeR package under R environment using the F-test. Differences were considered significant at p < 0.05. For examining the correlation relationship of IRF genes, a regression analysis was performed.

Two-year-old healthy Zhikong scallops were collected from artificial scallop-rearing substrates installed in Xunshan Fishery Group Co., Rongcheng (Shandong Province, China). All the procedures involved in the handling and the treatment of scallops during this study were approved by the Ocean University of China Institutional Animal Care and Use Committee (OUC-IACUC) prior to the initiation of the study. The scallops were acclimated in the laboratory at ambient seawater temperature for 1 week prior to the experiments, which is within the optimum temperature range for their survival.

Gram-negative (V. anguillarum) bacteria were used to challenge scallops in our study (Zhi et al., 2011). V. anguillarum was cultured in liquid 2216E broth (5 g/L of Tryptone, 1 g/L of yeast extract, and 0.1 g/L of C6H5Fe·5H2O, pH = 7.6) at 28°C and harvested by centrifugation at 2000×g for 5 min, as described by Kong et al. The pellet was suspended in filtered seawater and was adjusted to 1×10⁷ CFU/ml in seawater, respectively (Zhi et al., 2011; Ragab et al., 2014).

A total of 75 individuals were randomly and equally divided into five groups. At 0 h, 5 h, 24 h, 48 h, and 72 h post-infection, 5 individuals were randomly collected from each group. The 0 h group was employed as the control group, and other groups were used as experimental groups. The hemolymph samples were collected from adductor muscles using a syringe and were immediately centrifuged at 800×g, 4°C for 10 min to harvest the hemocytes (Gao et al., 2007). The extracted hemocyte sample was immediately frozen in liquid nitrogen and then subsequently frozen at −80°C before processing.

Total RNA was isolated following the method described by Hu et al. (2010), and then was digested with DNase I (TaKaRa, Shiga, Japan). A Nanovue Plus spectrophotometer (GE Healthcare, NJ, USA) was used to assess the concentration and purity of RNA; RNA integrity was determined by agarose gel electrophoresis. First-strand cDNA was synthesized using Moloney murine leukemia virus (MMLV) reverse transcriptase (Thermo, USA) following the manufacturer’s protocol. All of the cDNA products were diluted to 5 ng/ml for use as the template in real-time PCR.

Real-time PCR was conducted using the SsoFast™ EvaGreen^® Supermix on a Light Cycler 480 Real-time PCR System (Roche Di-agnostics, Mannheim, Germany). The running program was as follows: 50°C for 2 min, 94°C for 10 min, and 40 cycles at 94°C for 15 s and at 62°C for 1 min. Cytochrome B (CB), DEAD-βox RNA helicase (HELI), and EF1-A gene were designated as internal reference genes for the normalization of gene expression in healthy adults and test subjects during the real-time PCR experiment, respectively (Li et al., 2010; Feng et al., 2013). All the primers used in the real-time PCR were designed using Primer Premier 5.0 and are listed in Table 1 .

Data from the real-time PCR were analyzed using the Relative Expression Software Tool (REST) version 2009 (Pfaffl et al., 2002); gene expression is shown as the fold change. For the experimental groups, the control group (0 h) was used for normalization. The statistical analysis of the data was performed with SPSS (version 16.0) software using the independent t-test. Differences were considered to be significant at p < 0.05.

The full length of ORFs of three CfIRF genes was amplified from C. farreri cDNA, using primers listed in Table 1 . For CfIRF1 and CfIRF2, the PCR products were ligated and subcloned into pEGFP-N1 (Clontech, USA) by way of overlap extension PCR to construct recombinant plasmids pEGFP-CfIRF1 and pEGFP-CfIRF2, while for the CfIRF1-like, PCR products were digested with KpnI and SmaI, and ligated and subcloned into pEGFP-N1 vector digested by the corresponding restriction enzymes to construct recombinant plasmids pEGFP-CfIRF1-like. The constructed recombinant plasmids were subsequently verified by DNA sequencing.

In scallops, as well as in other marine bivalves, there are no mature cell lines. Thus, we choose HEK293T cells to perform our experiment. HEK293 cells were maintained in Modified Eagle Medium (MEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Invitrogen, USA) and antibiotics (100 mg/L streptomycin and 10⁵ U/L penicillin, Gibco) at 37°C in a humidified incubator under 5% CO₂. For DNA transfection, cells were seeded and allowed to grow to more than 70% confluence, and then plasmids were transfected by using the Lipofectamine 3000 Reagent (Invitrogen, USA) following the manufacturer’s recommendations. HEK293T cells were transiently co-transfected with 0.8 μg of expression plasmid and 1 μl of Lipofectamine 3000 in each well in a 24-well plate. All assays were performed with three independent transfections.

At 48 h post-transfection, HEK293 cells were washed with PBS twice and fixed with paraformaldehyde for 15 min. Then, the cells were washed three times with PBS and were stained with DAPI (Sigma, USA) to mark the nucleus followed by washing three times. Immunofluorescence was visualized and captured with confocal microscopy (Nikon, Japan).

The full length of ORFs of three CfIRF genes was amplified from C. farreri cDNA, using primers listed in Table 1 . For CfIRF1 and CfIRF2, the PCR products were ligated and subcloned into pcDNA3.1 V5/H vector (Invitrogen, USA) by way of overlap extension PCR to construct recombinant plasmids pcDNA3.1-IRF1 and pcDNA3.1-IRF2, while for the CfIRF1-like, the pcDNA3.1-IRF1-like was constructed using the same method as mentioned above but digested with NotI and KpnI. The constructed recombinant plasmids were subsequently verified by DNA sequencing. For reporter plasmids, pISRE-Luc (ClonTech, USA) was used, and pRL-TK renilla luciferase plasmid (Progema, USA) and pGL3-basic vector (Progema, USA) were used as an internal control and blank group, respectively. EndoFree Plasmid MiKit (OMEGA, USA) was used for the transfection of the plasmids according to the manufacturer’s instruction.

The cell culture assays were performed according to Section 2.6. For dual-luciferase reporter assays, HEK293T cells were transiently co-transfected with 0.2 mg of expression plasmid, 0.5 mg of reporter gene plasmid, 0.01 mg of pRL-TK renilla luciferase plasmid, and 0.1 μl of Lipofectamine 3000 in each well in a 24-well plate. The luciferase reporter vector pGL3-basic was used as a blank group. All assays were performed with three independent transfections. At 48 h post-transfection, HEK293 cells were washed with PBS twice and lysed. Firefly and renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega, USA) according to the manufacturer’s instruction. Cell lysate (20 μl) was transferred to a 1.5-ml EP tube and 100 μl of luciferase assay reagent II and 100 μl of Stop & Glo^® Reagent were added in sequence, then firefly and renilla luciferase activities were measured, respectively.

Four IRF genes, CfIRF1, CfIRF1-like, CfIRF2, and CfIRF8, were identified from the genome of Zhikong scallop. The basic information (total length, ORF length, number of exons, amino acids length, theoretical pI, and weight of protein) of these IRF members were summarized in Table 2 . The ORFs of CfIRF1, CfIRF1-like, CfIRF2, and CfIRF8 were respectively 1,107, 5,913, 1,053, and 1,311 bp, encoding 368, 1,970, 350, and 436 amino acids. The predicted molecular weights of these four genes ranged from 40.05 to 212.51 KD, with the predicted pI values from 4.97 to 6.96 ( Table 2 ). The genomic structural analysis showed that the length as well as the exon–intron pattern of these four CfIRF genes varied greatly. The longest CfIRF1 gene was 28,913 bp with 9 exons, and the shortest CfIRF2 gene was 6,565 bp with 10 exons. Moreover, CfIRF1-like and CfIRF8 had 9 and 8 exons, respectively ( Figure 1 ).

All four CfIRF proteins had a single well-conserved DBD domain (113 aa in length), which was found to be helix-turn-helix at the N-terminal ( Figure 2 ). Furthermore, CfIRF8 was predicted to contain an IRF-associated domain (IAD) at the C-terminus, and CfIRF1-like possessed seven C2H2-type (classical) zinc fingers (ZnF_C2H2) at the C-terminal ( Figure 2 ). Multiple sequence alignments of CfIRFs showed that they share high similarity within the DBD domain ( Figure 3 ). Moreover, five conserved tryptophan (W) residues were revealed, in which the W⁷¹ residue was found mutated as Y⁷¹ in CfIRF1 protein.

To confirm the identification of the CfIRFs, a phylogenetic tree was constructed through the ML method using 59 IRF proteins across 13 species. Four IRF subfamilies (IRF-1, -3, -4, and -5) were identified ( Supplementary Figure 1 ) and all CfIRFs were clustered into its own clade. The result showed that CfIRF1, CfIRF1-like, and CfIRF2 were clustered into the IRF-1 subfamily. CfIRF8 was firstly grouped together with IRF8-like from Peten maximus and IRF8 from C. gigas, and they were clustered into the IRF-4 subfamily.

The RPKM data were used to analyze the expression patterns of four CfIRF genes during eleven developmental stages ( Figure 4A , Supplementary Figure 2A ). Their expression during early development (from zygote to trochophore larvae) was quite low (<10 RPKM). After D-shaped larvae formation, CfIRF1-like and CfIRF8 remained at a low expression (<5 RPKM) while expression of CfIRF1 and CfIRF2 was obviously elevated. Although CfIRF2 showed higher expression level than CfIRF1 (>3-fold), their expression pattern was similar, both of which showing the highest expression in creeping larvae and juvenile. In adult tissues, we found that expression of CfIRF1 and CfIRF2 was obviously higher than CfIRF1-like and CfIRF8 ( Figure 4C , Supplementary Figure 2B ), consistent with their expression tendency during development. In general, CfIRF1-like and CfIRF8 remained low expression in all the tissues (<25 RPKM), while CfIRF1 was dominantly expressed in hepatopancreas and hemolymph (>160 RPKM), and the highest expression of CfIRF2 was observed in gill and hepatopancreas (>220 RPKM). To illustrate the role of CfIRF1 and CfIRF2 in development stages and adult tissues, we performed correlation analysis between the expression of CfIRF1 and CfIRF2 with E2/T ratio. According to the results, CfIRF1 has a significantly positive correlation with CfIRF2 expression both in different developmental stages (r = 0.86, p < 0.001) ( Figure 4B ) and in adult tissues (r = 0.65, p < 0.05) ( Figure 4D ).

To examine the immune responses of CfIRFs to V. anguillarum challenge (Qiu et al., 2007; Costa et al., 2009), their expression level was investigated at four time points (5 h, 24 h, 48 h, and 72 h) after infection ( Figure 5 ). Overall, three CfIRFs from the IRF-1 subfamily (CfIRF1, CfIRF1-like, and CfIRF2) were significantly upregulated. Notably, CfIRF2 showed the most sensitive responses, which showed significant upregulation after 24-h infection (>13-fold, p < 0.05) and sustained at a significantly higher expression level till 72 h. Moreover, the expressions of CfIRF1 and CfIRF1-like were significantly upregulated at 72 h post-infection (>12-fold, p < 0.05), while CfIRF8 only showed mild upregulated tendency after 24-h infection (4- to 8-fold).

Subcellular localization prediction showed that CfIRF1, CfIRF1-like, and CfIRF2 all possess nucleus localization ( Table 3 ). Furthermore, ¹⁶³RSRRRKKPCVKKE¹⁷⁵ in CfIRF1 was predicted as a nucleus localization signal with a high score (score: 9.5), further suggesting that CfIRF1 was a nuclear-localized protein. HEK293T cells were transfected with plasmids encoding pEGFP-tagged CfIRF1, CfIRF1-like, or CfIRF2 to investigate their subcellular localization. The immunofluorescence image analysis showed that recombinant CfIRF1-GFP protein was located in the nucleus, consisting of its subcellular localization prediction and the NLS prediction. Unlike CfIRF1, the fluorescent signal of CfIRF1-like and CfIRF2 recombinant proteins was distributed in both nuclei and cytoplasm ( Figure 6 ). Thus, the divergence of CfIRFs by subcellular localization tentatively suggested that these three CfIRFs may be involved with different cellular functions.

To analyze the transcription activities of CfIRF1, CfIRF1-like, and CfIRF2, dual-luciferase reporter assays were performed in HEK293T cells. As control, pcDNA3.1-IRFs were co-transfected with pGL3-basic, and it did not show any effect on the ISRE reporter. By using pISRE-Luc, after co-transfection with pcDNA3.1-IRF1, pcDNA3.1-IRF1-like, and pcDNA3.1-IRF2, the luciferase activity of the ISRE reporter was significantly upregulated ( Figure 7 ), suggesting their obvious transcriptional activation on interferon-stimulated response element. These results indicated that these CfIRFs could activate the expression of ISRE luciferase reporter genes, suggesting that scallop IRFs can specifically activate interferon signaling.