Figure 1. (A) SMART analysis of DDX41s. Coiled coil domain, DEAD domain, HELIC domain, and ZnF_C2HC domain were labeled in the sequence. (B) Distributions of scDDX41 expression in different tissues of mandarin fish. The expression levels of scDDX41 were detected by RT-qPCR. (C) Phylogenetic tree of DDX41 proteins from various species. A phylogenetic tree was constructed using the Neighbor-Joining method in MEGA v10.0, with 1000 bootstrap replications. The bootstrap values were indicated at the nodes of the tree. (D) Expression levels of scDDX41 in cells treated with poly(I:C) at indicated times. (E) Expression levels of scDDX41 in cells treated with poly (dA:dT) at indicated times. (F) Expression levels of scDDX41 in cells infected with MRV at indicated times. The β-actin gene served as internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance was indicated by asterisks, with ** referring to p < 0.01. ns, non-significant.

Figure 1. (A) SMART analysis of DDX41s. Coiled coil domain, DEAD domain, HELIC domain, and ZnF_C2HC domain were labeled in the sequence. (B) Distributions of scDDX41 expression in different tissues of mandarin fish. The expression levels of scDDX41 were detected by RT-qPCR. (C) Phylogenetic tree of DDX41 proteins from various species. A phylogenetic tree was constructed using the Neighbor-Joining method in MEGA v10.0, with 1000 bootstrap replications. The bootstrap values were indicated at the nodes of the tree. (D) Expression levels of scDDX41 in cells treated with poly(I:C) at indicated times. (E) Expression levels of scDDX41 in cells treated with poly (dA:dT) at indicated times. (F) Expression levels of scDDX41 in cells infected with MRV at indicated times. The β-actin gene served as internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance was indicated by asterisks, with ** referring to p < 0.01. ns, non-significant.

Figure 2. scDDX41 induced the activities of IFN-β-luc and NF-κB-luc promoters. (A) Cells were transfected with 0.2, 0.4, 0.6, or 0.8 μg of scDDX41 expression plasmid or an empty vector together with IFN-β-luc (0.4 μg/well) and pRL-TK (0.04 μg/well) plasmids. Luciferase assays were performed 36 h after the transfection. (B) Cells were transfected with 0.2, 0.4, 0.6, or 0.8 μg of scDDX41 expression plasmid or an empty vector together with NF-κB-luc (0.4 μg/well) and pRL-TK (0.05 μg/well). Luciferase assays were performed 36 h after the transfection. (C) Schematic of full-length and scDDX41 mutants with the DEAD domain, HELIC domain, and residue numbers as indicated. Various scDDX41 fragments were inserted into the C-terminus of pCMV-myc. (D,E) DEAD and HELIC domains of scDDX41 for IFN and NF-κB activation. Cells were transfected with 0.4 μg/well of various expression plasmids of scDDX41, scDDX41 mutants, or empty vector together with the reporter plasmid 0.04 μg/well pRL-TK as well as 0.4 μg/well of IFN-β-luc or NF-κB-luc plasmid. Luciferase assays were performed 36 h after the transfection. All luciferase assays were repeated at least three times, and data are means ±SD (n = 3) from single representative experiments. * p < 0.05, ** p < 0.01 between normal cells and stimulated cells.

Figure 2. scDDX41 induced the activities of IFN-β-luc and NF-κB-luc promoters. (A) Cells were transfected with 0.2, 0.4, 0.6, or 0.8 μg of scDDX41 expression plasmid or an empty vector together with IFN-β-luc (0.4 μg/well) and pRL-TK (0.04 μg/well) plasmids. Luciferase assays were performed 36 h after the transfection. (B) Cells were transfected with 0.2, 0.4, 0.6, or 0.8 μg of scDDX41 expression plasmid or an empty vector together with NF-κB-luc (0.4 μg/well) and pRL-TK (0.05 μg/well). Luciferase assays were performed 36 h after the transfection. (C) Schematic of full-length and scDDX41 mutants with the DEAD domain, HELIC domain, and residue numbers as indicated. Various scDDX41 fragments were inserted into the C-terminus of pCMV-myc. (D,E) DEAD and HELIC domains of scDDX41 for IFN and NF-κB activation. Cells were transfected with 0.4 μg/well of various expression plasmids of scDDX41, scDDX41 mutants, or empty vector together with the reporter plasmid 0.04 μg/well pRL-TK as well as 0.4 μg/well of IFN-β-luc or NF-κB-luc plasmid. Luciferase assays were performed 36 h after the transfection. All luciferase assays were repeated at least three times, and data are means ±SD (n = 3) from single representative experiments. * p < 0.05, ** p < 0.01 between normal cells and stimulated cells.

Figure 3. Overexpression of scDDX41 induces the expression of IFN-I, ISGs, and inflammatory cytokines. After transfection with scDDX41-myc or pCMV-myc at 24 h, cells were harvested and the expression levels of scDDX41 (A), scIFN-h (B), scMx (C), scISG15 (D), scViperin (E) and scTNF-α (F) genes were detected. Overexpression of scHELIC and scDEAD induced the expression of scMx (G) and scISG15 (H) in MFF-1 cells. The β-actin gene served as the internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to p < 0.01.

Figure 3. Overexpression of scDDX41 induces the expression of IFN-I, ISGs, and inflammatory cytokines. After transfection with scDDX41-myc or pCMV-myc at 24 h, cells were harvested and the expression levels of scDDX41 (A), scIFN-h (B), scMx (C), scISG15 (D), scViperin (E) and scTNF-α (F) genes were detected. Overexpression of scHELIC and scDEAD induced the expression of scMx (G) and scISG15 (H) in MFF-1 cells. The β-actin gene served as the internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to p < 0.01.

Figure 4. Identification of the interaction between scDDX41 and scSTING by Co-IP assay. (A) scDDX41 interacted with scSTING. Cells were transfected with the indicated plasmids. At 36 h post-transfection, the cell lysates were precipitated with an anti-flag or anti-myc mAb in conjunction with protein G-Sepharose beads and detected by WB analysis using anti-myc or anti-flag mAbs. The expression of the transfected proteins was analyzed by immunoblotting with anti-myc and anti-flag mAbs. (B) DEAD and HELIC domains of scDDX41 interacted with scSTING. Cells were co-transfected with DEAD-myc, HELIC-YFP, and scSTING-flag or empty vector. Immunoprecipitation assays with anti-flag antibody (IP: Flag) and Western blot analysis were performed with anti-Flag, anti-myc, or anti-YFP antibodies. (C) DEAD domain interacted with scSTING-NTD and scSTING-CTD. Cells were co-transfected with myc-DEAD and flag-scSTING-NTD, flag-scSTING-CTD, or empty vector. Immunoprecipitation assays with anti-flag antibody (IP: Flag) and WB analysis were performed with anti-Flag or anti-myc antibodies. (D) HELIC domain interacted with scSTING-NTD but not with scSTING-CTD. Cells were co-transfected with HELIC-YFP and flag-scSTING-NTD, flag-scSTING-CTD, or empty vector. Immunoprecipitation assays with anti-flag antibody (IP: Flag) and WB analysis were performed with anti-flag or anti-YFP antibodies.

Figure 4. Identification of the interaction between scDDX41 and scSTING by Co-IP assay. (A) scDDX41 interacted with scSTING. Cells were transfected with the indicated plasmids. At 36 h post-transfection, the cell lysates were precipitated with an anti-flag or anti-myc mAb in conjunction with protein G-Sepharose beads and detected by WB analysis using anti-myc or anti-flag mAbs. The expression of the transfected proteins was analyzed by immunoblotting with anti-myc and anti-flag mAbs. (B) DEAD and HELIC domains of scDDX41 interacted with scSTING. Cells were co-transfected with DEAD-myc, HELIC-YFP, and scSTING-flag or empty vector. Immunoprecipitation assays with anti-flag antibody (IP: Flag) and Western blot analysis were performed with anti-Flag, anti-myc, or anti-YFP antibodies. (C) DEAD domain interacted with scSTING-NTD and scSTING-CTD. Cells were co-transfected with myc-DEAD and flag-scSTING-NTD, flag-scSTING-CTD, or empty vector. Immunoprecipitation assays with anti-flag antibody (IP: Flag) and WB analysis were performed with anti-Flag or anti-myc antibodies. (D) HELIC domain interacted with scSTING-NTD but not with scSTING-CTD. Cells were co-transfected with HELIC-YFP and flag-scSTING-NTD, flag-scSTING-CTD, or empty vector. Immunoprecipitation assays with anti-flag antibody (IP: Flag) and WB analysis were performed with anti-flag or anti-YFP antibodies.

Figure 5. Involvement of scDDX41 in scSTING-mediated IFN expression, and scDDX41 recognizes dsDNA through the DEAD domain. (A) Cells transfected with an IFN-β-luc 400 ng/well and TK (40 ng/well) plus 400 ng/well) of the expression vectors for pCMV-myc, pCMV-myc and scDDX41-myc, pCMV-myc and scSTING-myc, scDDX41-myc, and scSTING-myc. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to p < 0.01. scDDX41 enhanced the IFN-I response induced by STING. (B–E) Cells seeded in 6-well plates were transfected or co-transfected with scDDX41 (400 ng/well), and scSTING (400 ng/well). The cells transfected with pCMV-myc acted as negative control. The expression levels of interferon signaling molecules including scIFN-h, scMx, scISG15, and scViperin were examined using RT-qPCR. The β-actin gene served as the internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to p < 0.01. (F) scDDX41 recognizes ISD through the DEAD domain using EMSA with biotin-labeled (Bio-) or unlabeled (Unbio-) probes ISD. The black lines indicate where parts of the image were joined. (G) Immunoblot analysis of the immunoprecipitated purified myc-tagged hsDDX41, scDDX41, or scDDX41ΔDEAD recombinant proteins incubated individually with biotinylated ISD and probed with anti-myc antibodies.

Figure 5. Involvement of scDDX41 in scSTING-mediated IFN expression, and scDDX41 recognizes dsDNA through the DEAD domain. (A) Cells transfected with an IFN-β-luc 400 ng/well and TK (40 ng/well) plus 400 ng/well) of the expression vectors for pCMV-myc, pCMV-myc and scDDX41-myc, pCMV-myc and scSTING-myc, scDDX41-myc, and scSTING-myc. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to p < 0.01. scDDX41 enhanced the IFN-I response induced by STING. (B–E) Cells seeded in 6-well plates were transfected or co-transfected with scDDX41 (400 ng/well), and scSTING (400 ng/well). The cells transfected with pCMV-myc acted as negative control. The expression levels of interferon signaling molecules including scIFN-h, scMx, scISG15, and scViperin were examined using RT-qPCR. The β-actin gene served as the internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to p < 0.01. (F) scDDX41 recognizes ISD through the DEAD domain using EMSA with biotin-labeled (Bio-) or unlabeled (Unbio-) probes ISD. The black lines indicate where parts of the image were joined. (G) Immunoblot analysis of the immunoprecipitated purified myc-tagged hsDDX41, scDDX41, or scDDX41ΔDEAD recombinant proteins incubated individually with biotinylated ISD and probed with anti-myc antibodies.

Figure 6. Overexpression of scDDX41 attenuates MRV infection. After transfection with scDDX41-myc or pCMV-myc at 24 h, cells were infected with MRV and harvested for RT-qPCR independently at the indicated time points. (A) Expression levels of scDDX41 in cells infected with MRV at indicated times. (B–D) Expression levels of mcp, ICP-18, and DNA pol genes in cells infected with MRV at indicated times. (E–I) Expression levels of scIFN-h, scMx, scISG15, scViperin, and scTNF-α genes in cells infected with MRV at indicated times. (J) Correlation of viral load in samples measured by TaqMan qPCR. (K) The titer of virus infection was measured on a 96-well cell culture plate via the finite dilution method. β-actin gene served as the internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to pD. rerio DDX41 might obstruct the DDX41-STING signaling pathways [11]. Although the DDX41 of fish is highly homologous with the DDX41 from other species, there are differences between the STING of fish and those of other species [17,22]. The N-terminal TM domain of fish STING presents diversity; for example, grass carp STING has three TM structure domains [23], while grouper STING contains four TM domains [24], and the scSTING contains five TM domains [17]. The TM domains of STING in fishes may lead to different mechanisms from those of mammals due to their high degree of unconserved sequences. The signal activation of STING depends on the IRF3 and NFkB. Different from the mammalian STING, the zebrafish STING can significantly stimulate a downstream NF⁃κB signal but the IRF3⁃IFN signal is weaker, which is due to the difference of the C terminal of STING [22,25]. Those findings indicated the difference in the activation of the DDX41-STING pathway between mammals and teleost fish. Interestingly, the DEAD domain of scDDX41 could directly interact with not only scSTING-NTD but also scSTING-CTD, and the HELIC domain of scDDX41 could also directly interact with scSTING-NTD. Hence, the DEAD and HELIC domains of scDDX41 contributed to the interaction of scDDX41 with scSTING to induce STING-dependent IFN-I and inflammatory immune responses. As such, the activation of the DDX41-STING pathway might differ between mammals and teleost fish. The detailed mechanisms of the DDX41 HELIC domains in synergistically regulating the STING-mediated innate immune response in fish should be further studied. Figure 6. Overexpression of scDDX41 attenuates MRV infection. After transfection with scDDX41-myc or pCMV-myc at 24 h, cells were infected with MRV and harvested for RT-qPCR independently at the indicated time points. (A) Expression levels of scDDX41 in cells infected with MRV at indicated times. (B–D) Expression levels of mcp, ICP-18, and DNA pol genes in cells infected with MRV at indicated times. (E–I) Expression levels of scIFN-h, scMx, scISG15, scViperin, and scTNF-α genes in cells infected with MRV at indicated times. (J) Correlation of viral load in samples measured by TaqMan qPCR. (K) The titer of virus infection was measured on a 96-well cell culture plate via the finite dilution method. β-actin gene served as the internal control to calibrate the cDNA template for all samples. Vertical bars represent ±SD (n = 3). Statistical significance is indicated by asterisks, with ** referring to pD. rerio DDX41 might obstruct the DDX41-STING signaling pathways [11]. Although the DDX41 of fish is highly homologous with the DDX41 from other species, there are differences between the STING of fish and those of other species [17,22]. The N-terminal TM domain of fish STING presents diversity; for example, grass carp STING has three TM structure domains [23], while grouper STING contains four TM domains [24], and the scSTING contains five TM domains [17]. The TM domains of STING in fishes may lead to different mechanisms from those of mammals due to their high degree of unconserved sequences. The signal activation of STING depends on the IRF3 and NFkB. Different from the mammalian STING, the zebrafish STING can significantly stimulate a downstream NF⁃κB signal but the IRF3⁃IFN signal is weaker, which is due to the difference of the C terminal of STING [22,25]. Those findings indicated the difference in the activation of the DDX41-STING pathway between mammals and teleost fish. Interestingly, the DEAD domain of scDDX41 could directly interact with not only scSTING-NTD but also scSTING-CTD, and the HELIC domain of scDDX41 could also directly interact with scSTING-NTD. Hence, the DEAD and HELIC domains of scDDX41 contributed to the interaction of scDDX41 with scSTING to induce STING-dependent IFN-I and inflammatory immune responses. As such, the activation of the DDX41-STING pathway might differ between mammals and teleost fish. The detailed mechanisms of the DDX41 HELIC domains in synergistically regulating the STING-mediated innate immune response in fish should be further studied.

Table 1. Primers used for cloning.

Table 1. Primers used for cloning.

NamesSequence (5′–3′)5′ RACE-FCTAATAGCACTCACTATAGGGCAAGCAGTGGTATCAACGCAGAGT5′ RACE-RTGAGATGCTGATGCTGGTCAAGGAG3′ RACE-FTGATGGATCTTAAAGCCCTGC3′ RACE-RACTCTGCGTTGATACCACTGCTTGCCCTATAGTGAGTGCTATTAGscDDX41-FCGGAATTCCGGAGACCGACAATCGACCCscDDX41-RGGGGTACCTTAGAAGTCCATTGAGCTATGAGCscDEAD-FCGGAATTCCGCCACCAGCAATTCTAAAAGGscDEAD-RGGGGTACCTTACATCTTGGCCTCCTCTTTGscHELIC-FCGGAATTCCGCTTTTATTTGCTGAGAAGAAGGscHELIC-RGGGGTACCTTAATTAATGAACGTAGTGGCscHELIC-YFP-FCGGAATTCCGATGCTTTTATTTGCTGAGAAGAAGscHELIC-YFP-RGGGGTACCCCATTAATGAACGTAGTGGCISD-FTACAGATCTACTAGTGATCTATGACTGATCTGTACATGATCTACA ISD-RTGTAGATCATGTACAGATCAGTCATAGATCACTAGTAGATCTGTA

Table 2. Primers used for RT-qPCR.

Table 2. Primers used for RT-qPCR.

Gene NamesPrimersSequences (5′–3′)Primer EfficiencyscDDX41Forward
ReverseACGATTATGTTCCGTACATTCCAGTCAA
TCCTCATCCCTCTGCTCCTCCC0.99MRV mcpForward
ReverseGTCACCCTGCCTTACAACGAAA
CACGATGGGCTTGACTTCTCC0.96MRV ICP-18Forward
ReverseAGTTTGACGCCAGCTTTCACG
TGCCATACCGTCGCACTCG0.98MRV DNA polymeraseForward
ReverseGGCGTCAAGTGCCATCAG
CAGGCGAGACAGTCTTCCAAT0.98IFN-hForward
ReverseCGCTCTGCTGTGATTGGC
GGGACTCCACCTCTGCCTTT0.99scMXForward
ReverseGGATTCTGACATCGGGAGCAA
GTGCAGTAGACTCATGCTGT0.98scISG15Forward
ReverseCGACGAGACTGTGAGCGACTTC
CATTCATCATCTCCCTGCCTTGGT0.99scViperinForward
ReverseCCAAGAGGGGCCTCAAACTT
CTGACACTTGGGAGCTGGAG0.95scTNF-αForward
ReverseAGCCAGGCATCGTTCAGAGTCT
CTGTCCTCCTGAGCGGTGTCTT0.96β-actinForward
ReverseCCCTCTGAACCCCAAAGCCA
CAGCCTGGATGGCAACGTACA0.96