Large yellow croakers (18.0 ± 1.5 cm in length and 60.0 ± 15.0 g in weight) were purchased from Ningde Fufa Fishing Co., Ltd. The fish were kept in a 25°C inflatable seawater tank and fed with commercial feed for two weeks before being used for subsequent experiments as our previous reports (10, 27). The fish used in the present study were conducted according to the guidelines of the Animal Administration and Ethics Committee of Jimei University (Permit No. 2021-4).

HEK 293T cells were purchased from the Chinese Typical Culture Preservation Center and cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS, Invitrogen-Gibco), 100 U/mL penicillin (P) and streptomycin (S). The cells were cultured in a 5% CO₂ incubator at 37°C (10, 27). Large yellow croaker muscle cells (LYCMS) were maintained in L15 medium (Boster, USA) supplemented with 10% fetal bovine serum and 100 U/mL penicillin and streptomycin at 25°C. The transfection of the plasmids into the cells mentioned above was performed by using Lipofectamine^® 3000 (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions.

To understand the tissue distribution of TRAF4 in large yellow croakers, six healthy fish were randomly selected and anesthetized with 0.01% eugenol, then organs/tissues including gill, liver, spleen, head kidney, trunk kidney, intestine, heart, brain, skin, and muscle were dissected and stored in RNAlater, and blood was collected and stored in Trizol for subsequent total RNA extraction.

To study the expression patterns of TRAF4 in large yellow croakers under different immune stimulations, the fish were divided into five groups (with each group contained 50 fish) and injected with different pathogen-associated molecular patterns (PAMPs). In brief, each fish in challenge groups was intraperitoneally injected with 100 μL suspension of Pseudomonas plecoglossicida (5×10⁵ CFU/mL), polyinosinic-poly-cytidylic acid potassium salt (poly I:C) (P9582, Sigma, 1 mg/mL), lipopolysaccharides (LPS) (L3024, Sigma, from Escherichia coli O111:B4, 0.5 mg/mL), and peptidoglycan (PGN) (69554, from Bacillus subtilis, Sigma, 1 mg/mL), respectively. The control group was injected with 100 µL PBS solution (27). Six fish were randomly selected from each group at 6, 12, and 24 h post-injection (hpi) and anaesthetized as described above, different organs/tissues such as gill, intestine, spleen, head kidney, and blood were collected for RNA extraction experiment.

Total RNA was extracted from the above tissue/organ samples using Eastep™ Super Total RNA Extraction Kit (Promega) according to the manufacturer’s instructions. The quality and concentration of RNA were determined by agarose gel electrophoresis and OD260/280 analysis. 1 μg total RNA was taken from each sample and digested with RNase free DNase I (Thermo Scientific™), then RevertAid First Strand cDNA Synthesis Kit (#K1622, Thermo Scientific™) was used to reverse transcribe the RNA into cDNA, followed by storing in a refrigerator at -80°C for subsequent gene cloning or quantitative real-time PCR analysis (qRT-PCR).

Based on TRAF4 transcriptome data of large yellow croaker (GenBank Accession No. XM_010749366.3), the full-length ORF of large yellow croaker TRAF4 was amplified by using specific primers. The amplification conditions were as follows: 95°C denatured 3 min, 35 cycles of denaturation at 95°C, annealing at 56°C for 30 s, extension at 72°C for 2 min. The confirmed ORF sequence of TRAF4 was inserted into pTurboGFP-N vector to construct GFP fluorescent tracer plasmid and pcDNA3.1/myc-His (−) A vector for subsequent eukaryotic overexpression analysis, which were verified by sequencing analysis, respectively. Primers used in the present study are shown in Table 1 .

Protein prediction was carried out using the software at ExPASy Bioinformatics Resource Portal (http://www.expasy.org/proteomics), with conserved domain structures detected by using the Conserved Domain Database (CDD) on NCBI (http://www.ncbi.nlm.nih.gov/cdd) (28). Vertebrate TRAF4 protein sequences were searched using the Basic Local Alignment Search Tool (BLAST) analysis tool of the National Center for Biotechnology Information website (NCBI, http://www.ncbi.nlm.nih.gov/blast). Amino acid multiple alignments were performed using the Clustal X program and the results were edited using the GeneDoc program. The phylogenetic analysis was conducted using the neighbor-joining method in MEGA 7 (29).

The above cDNA was used as a template, and the volume of qPCR amplification was 10 µL, including 2×SYBR Green Master Mix 5 µL, forward primer 0.5 µL (4 mM), reverse primer 0.5 µL (4 mM), cDNA 1 µL (100 ng/µL), and nuclease-free water 3 µL.

qRT-PCR was performed using Go Taq^® qPCR Master Mix (Promega, Madison, USA) on the Roche LightCycler^®480 II quantitative real-time detection system (Roche, Switzerland). The PCR amplification conditions were as follows: 95°C for 5 min, followed by 40 cycles at 95°C for 20 s, 58°C for 15 s, and 72°C for 15 s, with a final extension at 72°C for 10 min, and then the reaction temperature was increased from 65°C to 95°C at a rate of 0.5°C per second, and the melting curve was analyzed. All the reactions were carried out in a 384-well plate and their mean values were recorded. The relative expression of the target gene was normalized to the expression of β-actin, which has been proved to be a reliable reference gene and widely used for gene expression analysis in large yellow croaker (10, 27), and then analyzed using a comparative Ct method (2^-ΔΔCt) (30). All primers used for qRT-PCR were listed in Table 1 .

To determine the effect of large yellow croaker TRAF4 on IRF3, IRF7, and type I IFN promoter activation, HEK 293T cells were transiently transfected with 100 ng pGL4-IRF3-pro (Chinese invention patent number: ZL201710457836.8), pGL4-IRF7-pro (Chinese invention patent number: ZL201710457820.7) or pGL4-IFN1-pro (Chinese invention patent application number: 201710456729.3), 10 ng of pRL-TK (Promega, Madison, WI) together with increasing concentrations of pcDNA3.1-TRAF4 (10, 50, 100 or 200 ng) or pcDNA3.1 empty vector with the same concentrations for luciferase activity assays.

To determine the effect of different domains of TRAF4 on IRF3, IRF7, and type I IFN promoter activation, HEK 293T cells were transiently transfected with 100 ng pGL4-IRF3-pro, pGL4-IRF7-pro or pGL4-IFN1-pro, 10 ng pRL-TK together with 100 ng full-length TRAF4 and its truncated forms including TRAF4-ΔRING (residues 63-470), TRAF4-ΔRING/zinc1 (residues 161-470), TRAF4-MATH (residues 273-470), and TRAF4-ΔMATH (residues 1-303) for luciferase activity assays.

To investigate the association between large yellow croaker TRAF4 and TRIF or TRAF6, HEK 293T cells were seeded overnight in a 24-well plate with 1 × 10⁵ cells per well. The next day, the cells were transiently transfected with 100 ng pGL4-IRF3-pro or pGL4-IRF7-pro, 10 ng of pRL-TK together with 100 ng pcDNA3.1-TRAF4 and pcDNA3.1-TRIF or pcDNA3.1-TRAF6 alone or in combination of two, with the total amount of transfected plasmids balanced with the pcDNA3.1 empty vector.

At 24 hours post-transfection (hpt), the transfected cells described above were collected and lysed for subsequent luciferase activity assays using Dual-Luciferase Reporter System (Promega), with luciferase activity measured by a Promega GloMax^® 20/20 luminometer according to the manufacturer’s instructions. All data were normalized to Renilla luciferase activity and expressed as fold relative to the control group.

To detect the subcellular localization of TRAF4 and truncated forms of TRAF4 in large yellow croaker, HEK 293T cells were cultured on sterilized coverslips in 6-well plates overnight (2 × 10⁵ cells per well). On the second day, 5 µg plasmids of pTurbo-TRAF4-GFP, pTurbo-TRAF4-ΔRING-GFP, pTurbo-TRAF4-ΔRING/zinc1-GFP, pTurbo-TRAF4-MATH-GFP, pTurbo-TRAF4-ΔMATH-GFP, and pTurboGFP-N (vector control) were transfected into HEK 293T cells with Lipofectamine^® 3000, respectively. At 24 hpt, the transfected cells growing on the coverslips were washed with PBS, fixed with 4% paraformaldehyde at room temperature for 10 min, and then permeated with Triton X-100. The cells on the coverslips were washed with PBS and sealed with one drop of the mounting medium (VECTASHIELDR Hard Set™ Mounting Medium with DAPI, Vector Laboratories, CA). Finally, the cells were examined and photographed under a confocal microscope (Laika, Germany).

To determine the effects of large yellow croaker TRAF4 with TRIF and TRAF6 on the induction of downstream antiviral and inflammation-related genes expression, LYCMS cells were seeded on 6-well plates at a density of 2 × 10⁵ cells per well, and then cells were co-transfected with 2.5 μg pcDNA3.1-TRAF4 together with 2.5 μg pcDNA3.1-TRIF or pcDNA3.1-TRAF6 alone or in a combination of two, with the total amount of transfected plasmids balanced with the pcDNA3.1 empty vector. At 48 hpt, the cells were collected for total RNA extraction and cDNA synthesis. Subsequently, qRT-PCR was performed to detect the mRNA expression levels of antiviral and inflammation-related genes, including IRF3, IRF7, ISG15, ISG56, Mx, RSAD2, IL-1β, and TNF-α. All primers used for qRT-PCR are shown in Table 1 .

Statistical analysis was conducted using one-way analysis of variance (ANOVA), followed by Duncan’s multiple range test using SPSS version 25, with different superscripts indicating statistically different (P < 0.05), * P < 0.05, and ** P < 0.01 considered statistically significant, respectively.

To study the role of TRAF4 in host immunity of large yellow croaker, the cDNA of TRAF4 was cloned from the spleen of the fish and named as Lc-TRAF4 (GenBank accession No. ON246166). The identified ORF of Lc-TRAF4 consists of 1,413 nucleotides, encoding a protein of 470 amino acids (aa). Based on the analysis of conserved domains by using NCBI CDD, it was found that Lc-TRAF4 contains a RING finger domain (15-59 aa), two zinc finger domains (102-156 aa, and 210-269 aa), and a MATH domain (308-463 aa) ( Figure S1 ).

The amino acid sequence of Lc-TRAF4 was compared with TRAF4 of other vertebrates, including grouper (E. coioides), medaka (Oryzias latipes), rainbow trout (Oncorhynchus mykiss), fugu (T. rubripes), zebrafish (Danio rerio), Atlantic salmon (Salmo salar), channel catfish (Ictalurus punetaus), African clawed frog (Xenopus laevis), mouse (Mus musculus), and human (Homo sapiens) ( Figure S1 ), which revealed a relatively conservative RING finger domain, zinc finger domains, and MATH domain in the species that mentioned. Additionally, the amino acid sequence of Lc-TRAF4 had a similarity of 99% with the TRAF4 of grouper, medaka, rainbow trout, and fugu, 98% with zebrafish, and 97% with channel catfish. Compared with TRAF4 of amphibians and mammals, the similarity level decreased, with a similarity of 92% to African clawed frog and 90% to mouse and human ( Table 2 ), indicating a high degree of sequence conservation across vertebrate TRAF4 proteins.

To further understand the phylogenic relationship of TRAF4 in vertebrates, the phylogenetic tree was constructed based on the amino acid sequences of Lc-TRAF4 and other vertebrates, including teleosts, amphibians, birds, and mammals. The results showed that these TRAF4 orthologs were divided into two groups, with mammalian TRAF4 gathering in one clade, whereas teleost, bird, and amphibian TRAF4 gathering in another clade. Additionally, Lc-TRAF4 was clustered into fish TRAF4 family and was most closely related to TRAF4 of sea bream (Sparus aurata) ( Figure 1 ).

Based on the analysis of genomic sequences of Lc-TRAF4 and other vertebrate TRAF4 genes in NCBI database, the genomic structures of large yellow croaker, fugu, zebrafish, African clawed frog, duck (Anas platyrhynchos), mouse, and human were analyzed. The results showed that the length of TRAF4 genes have a range of 4,560 bp to 39,956 bp from teleosts to mammals, with the shortest found in duck and the longest found in African clawed frog, respectively. In addition, the exon-intron organization of TRAF4 in teleosts, amphibians, birds, and mammals was similar, including 7 exons and 6 introns ( Figure 2 ). All exons of vertebrate TRAF4 were very conserved, which were 143, 52, 105, 162, 162, 156, and 633 bp in length, respectively, whereas the sizes of introns were varied considerably in the detected species ( Figure 2 ).

The results of qRT-PCR analysis showed that Lc-TRAF4 was ubiquitously expressed in all detected organs/tissues, including the gill, spleen, head kidney, trunk kidney, liver, heart, brain, skin, muscle, intestine, and blood, with the highest expression level found in the brain, and the lowest observed in the heart ( Figure 3 ).

To further understand the role of Lc-TRAF4 in the host immune response, the expression patterns of Lc-TRAF4 in mucosal immune tissues (gill and intestine), peripheral immune organ/tissue (spleen and head kidney), and peripheral blood under various PAMPs stimulations were also determined. The results showed that the expression levels of Lc-TRAF4 were significantly induced under poly I:C stimulation in the gill, intestine, spleen, head kidney, and peripheral blood, with a 1.9-fold increase at 6 hpi in gill ( Figure 4A ), 2.4- and 2.1-fold, a 3.1- and 4.1-fold, and a 2.5- and 3.5-fold increase at 6 and 12 hpi in the intestine, spleen, and head kidney ( Figures 4B–D ), and a 2.6-fold increase at 12 hpi in peripheral blood ( Figure 4E ), respectively. The expression levels of Lc-TRAF4 were also significantly up-regulated upon LPS challenge, which was a 2.4-, 12.6-, and 4.7-fold increase at 6 hpi in the gill, intestine, and spleen ( Figures 4A–C ), a 2.0-fold increase at 12 hpi in the head kidney ( Figure 4D ), and a 20.2- and 10.1-fold increase at 6 and 12 hpi in peripheral blood ( Figure 4E ), respectively. In addition, the expression levels of Lc-TRAF4 were also significantly induced under PGN stimulation, with a 2.7- and 5.6-fold increase at 6 hpi in the gill and intestine ( Figures 4A,B ), a 2.1-, 2.3-, and 3.0-fold increase at 6, 12, and 24 hpi in the spleen ( Figure 4C ), a 2.7-fold increase at 12 hpi in the head kidney ( Figure 4D ), and a 5.3- and 2.2-fold increase at 6 and 12 hpi in peripheral blood ( Figure 4E ), respectively. Additionally, Lc-TRAF4 was also significantly induced in response to P. plecoglossicida infection, with a 2.9- and 1.9-fold, a 2.1- and 4.3-fold increase at 6 and 12 hpi in gill and peripheral blood ( Figures 4A, E ), a 1.6-fold increase at 6 hpi in the intestine ( Figure 4B ), a 3.4- and 2.5-fold increase at 12 and 24 hpi in the spleen ( Figure 4C ), and a 2.4-fold increase at 12 hpi in the head kidney ( Figure 4D ), respectively. Nevertheless, the expression levels of Lc-TRAF4 were significantly down-regulated at 24 hpi in the head kidney and peripheral blood under LPS and PGN stimulation ( Figures 4D, E ), respectively.

To investigate whether Lc-TRAF4 plays a role in IRF3, IRF7, and type I IFN promoter activation, expression plasmids pcDNA3.1-TRAF4 or pcDNA3.1 empty vector together with IRF3, IRF7, or IFN1 promoter luciferase reporter plasmids were co-transfected into HEK 293T cells and then detected by luciferase assays. The results showed the overexpression of Lc-TRAF4 could significantly induce the activation of IRF3, IRF7, and IFN1 promoters, which provoking a typical plasmid dose-dependent manner. With the transfection amount going higher, the induction level increased, being up to 2.0-, 3.8-, and 2.2-fold relative to the control at a plasmid dose of 200 ng in IRF3, IRF7, and IFN1 promoter activation, respectively ( Figures 5A–C ).

To determine the role of various domains of Lc-TRAF4 in IRF3, IRF7 or IFN1 promoter activation, four truncated forms of domains deletion expression vectors were constructed, including TRAF4-ΔRING (deletion of the RING finger domain), TRAF4-ΔRING/zinc1 (deletion of the RING finger and the first zinc finger domain), TRAF4-MATH (only containing the MATH domain), and TRAF4-ΔMATH (deletion of the MATH domain) ( Figure 5D ). Based on dual-luciferase reporter assays, it was found that the truncated forms of Lc-TRAF4, except for the only containing MATH domain mutant, can also effectively induce IRF3, IRF7, and IFN1 promoter activation. However, the promoter activation levels induced by truncated forms of Lc-TRAF4, including TRAF4-ΔRING, TRAF4-ΔRING/zinc 1, and TRAF4-ΔMATH were significantly lower compared to the full-length form of TRAF4 ( Figures 5E–G ). It is thus suggested that the RING finger and zinc finger domains of Lc-TRAF4 may function in the activation of IRF3, IRF7, and IFN1 signaling.

In order to understand the subcellular localization of Lc-TRAF4, the full-length ORF of Lc-TRAF4 was subcloned into pTurboGFP-N vector and then transfected into HEK 293T cells, followed by examining under a confocal microscope at 24 hpt. The results showed that the pTurbo-TRAF4-GFP fusion protein was distributed in the cytoplasm, with bright green spots identified around the nucleus. To determine the function of various domains of Lc-TRAF4 in the subcellular distribution, four truncated forms of domain deletion expression vectors, including TRAF4-ΔRING, TRAF4-ΔRING/zinc1, TRAF4-MATH, and TRAF4-ΔMATH were also transfected into HEK 293T cells and assayed. It was revealed that TRAF4-ΔRING, TRAF4-ΔRING/zinc1, TRAF4-MATH, and TRAF4-ΔMATH had nearly the same subcellular distribution with the full length of Lc-TRAF4, which were also located in the cytoplasm with some aggregates around the nucleus. However, the control cells transfected with an empty pTurboGFP-N vector showed a global cytosolic localization as well as the nucleus ( Figure 6 ).

As we have mentioned above, Lc-TRAF4 overexpression was sufficient to activate the promoter activity of IRF3 and IRF7, which are involved in the host antiviral signaling pathway. It is thus important to reveal the molecules that associated with Lc-TRAF4 in such signaling. To address this question, expressing plasmids including Lc-TRAF4, Lc-TRIF, and Lc-TRAF6 were co-transfected with IRF3 or IRF7 promoter reporter plasmids into HEK 293T cells and then detected by dual-luciferase reporting system. The results showed that co-expression of Lc-TRAF4 with Lc-TRIF could induce higher activation of IRF3 and IRF7 promoter compared with the only overexpression of Lc-TRAF4 or Lc-TRIF ( Figures 7A, B ). Additionally, co-transfection of Lc-TRAF4 with Lc-TRAF6 could also remarkably up-regulated the level of IRF3 and IRF7 promoter activation in comparison with the only transfection of Lc-TRAF4 or Lc-TRAF6 ( Figures 7A, B ). These results indicated that Lc-TRAF4 is associated with Lc-TRIF and Lc-TRAF6 in the activation of IRF3 and IRF7 signaling.

To further confirm the association of Lc-TRAF4 with Lc-TRIF and Lc-TRAF6 in the host immune-related signaling pathway, qRT-PCR analysis was also performed to reveal the association of Lc-TRAF4 with Lc-TRIF and Lc-TRAF6 on the regulation of antiviral as well as inflammation-related genes expression, including IRF3, IRF7, ISG15, ISG56, Mx, RSAD2, IL-1β, and TNF-α. The results showed that the mRNA expression levels of some immune-related genes described above were significantly induced under the overexpression of Lc-TRAF4 alone, including IRF3, IRF7, ISG15, ISG56, RSAD2, IL-1β, and TNF-α, whereas the expression of Mx was not affected ( Figures 8A–H ). Interestingly, co-expression of Lc-TRAF4 with Lc-TRIF and Lc-TRAF6 resulted in significantly higher expression levels of the detected immune-related genes compared with the only overexpression of Lc-TRAF4, Lc-TRIF or Lc-TRAF6 ( Figures 8A–H ), indicating that Lc-TRAF4 strongly associate with Lc-TRIF and Lc-TRAF6 in in induction of host antiviral and inflammatory signaling.