Genome editing of FTR42 improves zebrafish survival against virus infection by enhancing IFN immunity

Summary

The development of CRISPR-Cas9 technology introduces an efficient tool for precise engineering of fish genomes. With a short reproduction cycle, zebrafish infection mode can be referenced as antiviral breeding researches in aquaculture fish. Previously we identified a crucian carp-specific gene ftrca1 as an inhibitor of interferon response in vitro. Here, we demonstrate that genome editing of zebrafish ftr42, a homolog of ftrca1, generates a zebrafish mutant (ftr42^lof/lof) with an improved resistance to SVCV infection. Zebrafish ftr42 acts as a virus-induced E3 ligase and downregulates IFN antiviral response by facilitating TBK1 protein degradation and also IRF7 mRNA decay. Genome editing results in loss of function of zebrafish ftr42, which enables zebrafish to have enhanced interferon response, thus improving zebrafish survival against virus infection. Our results suggest that fine-tuning fish IFN innate immunity through genome editing of negative regulators can genetically improve viral resistance in fish.

Graphical abstract

Highlights

• •FTR42 loss-of-function zebrafish show improved resistance to virus infection
• •Zebrafish FTR42 facilitates TBK1 protein degradation to inhibit IFN response
• •Zebrafish FTR42 facilitates IRF7 mRNA decay to inhibit IFN response

Biological sciences; Gene process; Molecular biology

Introduction

Viral diseases cause severe losses in fish farming that has contributed the most to global protein supply and food security for the past two decades.^1,2 The living water environment of fish limits a wide use of vaccines and drugs, which are not very successful to prevent viral diseases in pisciculture. Selectively breeding fish variants with resistance to viral diseases is highly desirable.^2,3 Genome editing is thought as an effective tool to expedite genetic improvement in aquaculture; however, its potential in fish breeding for resistance to viral diseases has not yet been broadly exploited.² Given a long reproduction cycle for farmed fish species, a zebrafish (Danio rerio) susceptibility model with SVCV (spring viremia of carp virus) is beneficial to provide a reference for this goal. Notedly, SVCV is a highly pathogenic pathogen for some important farmed cyprinids, such as common carp (Cyprinus carpio) and grass carp (Ctenopharyngodon idella).⁴

In vertebrates, interferon (IFN) immunity is the first line of host defense against viral infection.^5,6,7 Virus infection can be rapidly recognized by host pattern recognition receptors (PRR) including retinoic acid-inducible gene-I (RIG-1)-like receptors (RLRs), thus recruiting adaptor proteins to activate protein kinase TANK-binding kinase 1 (TBK1) and downstream IFN regulatory factor (IRF) 3 and IRF7, and subsequently inducing the expression of type I IFNs.⁸ The produced IFNs in turn upregulate the expression of hundreds of IFN-stimulated genes (ISGs) with antiviral effects, defending host against viral infection.⁹ These results suggest that the IFN immunity confers a fast and broad-spectrum disease resistance, thus representing a potential source for breeding resistance. Despite the necessity of cellular IFN response, unregulated IFN expression is pathogenic and often fatal in mammals,⁶ and in fish, an IFN-transgenic medaka strain yields an increased susceptibility to viral infection.¹⁰ However, comparison analyses of different clones of Gibel carp (Carassius auratus gibelio) reveal the relevance of a high virus resistance to a high IFN level,^11,12,13 indicating that appropriately promoting IFN response might be beneficial to largely improve fish resistance to viral infection, although there is a lack of useful genes for genome editing.

During viral infection, host IFN immunity is fine-tuned synergistically by various kinds of positive and negative regulators.⁸ Genome editing of zebrafish LGP2 creates a loss-of-function mutant with severely impaired zebrafish survival against viral infection,¹⁴ implying that genome editing of molecules with negative regulatory functions might be a potential strategy for genetic improvement of disease resistance trait in fish.^15,16 Recently, genome editing of negative regulators has been used in crops to create complete or partial loss-of function alleles, and this strategy has been successfully used to rapidly create novel crop variants with multi-pathogen resistance.^17,18 However, it is unclear which genes are suitable for genome editing to improve fish resistance to viral infection.

Tripartite motif (TRIM) proteins encode a class of E3 ubiquitin ligases,^19,20 and roughly half of human TRIM members contribute to regulating the innate immune response.²¹ In comparison with over 80 TRIM proteins in human, there are more than 200 in zebrafish due to an unprecedented gene expansion, which forms a finTRIM (fish novel TRIM, FTR) subfamily exclusively in teleost fish.^22,23,24 Interestingly, most finTRIM genes are transcriptionally upregulated by IFNs although their functions are largely unknown.²⁵ We previously identified a crucian carp-specific finTRIM gene, named ftrca1 (finTRIM C. auratus 1),²⁶ and found that during virus infection, FTRCA1 downregulates fish IFN response by selectively attenuating TBK1 protein and IRF7 mRNA expression,¹⁵ implying that FTRCA1 functions as a negative regulator of fish IFN immunity in vitro.

Zebrafish ftr42 is a homologous gene of ftrca1.²⁶ In the current study, we used the CRSIPR/Cas9 technology to generate a zebrafish mutant (ftr42^lof/lof), with a modified ftr42 allele being loss-of-protein expression and loss-of-function. The ftr42^lof/lof zebrafish are normal in visible phenotypes and importantly, they showed a stronger resistance to SVCV infection than wild type (WT) zebrafish. Mechanistically, FTR42 function-deficiency enhanced IFN immunity in zebrafish following SVCV infection. Considering that the antiviral mechanisms in fish are generally conserved, our results suggest that fine-tuning of fish IFN innate immunity by genome editing of a negative regulator is able to genetically improve the viral resistance trait in aquaculture fish.

Results

FTR42 deficiency benefits zebrafish survival against SVCV infection

Previous reports showed that FTRCA1, a species-specific inhibitor of fish IFN immunity, is phylogenetically close to a subset of zebrafish finTRIM members including FTR42.²⁶ Zebrafish FTR42 has a similar protein size to FTRCA1, with a similar protein structure composed of an N-terminal RING finger domain, two B-boxes, and a coiled-coil domain immediately preceded by a C-terminal PRY-SPRY domain (Figure S1A). Like ftrca1,²⁶ zebrafish ftr42 was transcriptionally induced by SVCV infection in tissues (Figure S1B), showing an inducible expression pattern similar to zebrafish Ifnφ1 gene in liver, kidney, gill and spleen (Figure S1C). Sequence analysis of a 5′ flanking DNA reveals two putative ISRE motifs (Figure S1D), a characteristic of ISG promoters.^26,27 This DNA sequence-driven luciferase reporter plasmid (FTR42pro-luc) was activated significantly by SVCV infection, poly(I:C) incubation (analogous to extracellular dsRNA), poly(I:C) transfection (analogous to intracellular dsRNA) (Figure S1E), or transfection with IRF3, IRF7 or zebrafish IFNφ1 (Figure S1F). These results indicated that zebrafish ftr42 is an ISG like ftrca1.

Using the CRISPR/Cas9 technology to target 5′ end of ftr42’s ORF in exon 1 followed by homozygous screening (Figure 1A), we obtained a zebrafish mutant line (ftr42^lof/lof), with one base mutation immediately preceded by one base insertion at the Cas9 targeting site (Figure 1B). During over two-year successive breeding in our lab, this mutant showed a normal visible phenotype in growth, reproduction and feeding. We compared viral susceptibility between mutant zebrafish and wild-type (WT) zebrafish. After zebrafish larvae (4 dpf) were immersed with SVCV (5×10⁶ TCID₅₀/mL) for two days, only 6 deaths were seen in ftr42^lof/lof larvae but 36 deaths in WT larvae (n = 50) (Figure 1C). Zebrafish larvae started to die at 12 h post infection (5×10⁶ TCID₅₀/mL), and at 72 h post infection, no WT larvae survived, but nearly half of ftr42^lof/lof larvae were still alive (Figure 1D). Similar assays with zebrafish adults (3-months old) showed that, SVCV injection (1×10⁸ TCID₅₀/mL) caused a more severe hemorrhage in lower jaw and peritoneal regions in WT fish than in ftr42^lof/lof fish (Figure 1E). At 7 days post injection, half of ftr42^lof/lof adults were alive, in contrast to no survivals in WT adults (Figure 1F). Consistently, WT larvae had a high-level mRNA expression of SVCV genes over ftr42^lof/lof larvae (Figure 1G), which was also detected for WT zebrafish tissues versus ftr42^lof/lof zebrafish tissues (Figure 1H). These results indicated that FTR42 deficiency is beneficial to zebrafish resistance against SVCV infection.

Figure 1.

ftr42-deficient zebrafish show increased resistance to SVCV infection

(A and B) Scheme of the genomic structure shows CRISPR/Cas9-targeted site in exon 1 of zebrafish FTR42, causing a reading frameshift due to 1-bp insertion and 1-bp mutation (A), which was verified by sequencing (B).

(C) Representative images showed WT and ftr42^lof/lof zebrafish larvae (4 dpf, n = 50 per group) immersed with SVCV (5×10⁶ TCID₅₀/mL) for 48h. Dead larvae showed obvious body curling.

(D) Survival ratios of WT and ftr42^lof/lof zebrafish larvae (n = 100 per group) were counted at indicated time points after immersion challenge with SVCV (5×10⁶ TCID₅₀/mL).

(E) Representative images showed the hemorrhagic comparison of WT and ftr42^lof/lof zebrafish adults (90 dpf) that were i.p. injected for 48h with SVCV (1×10⁸ TCID₅₀/mL), 20 μL per fish.

(F) Survival ratios of WT and ftr42^lof/lof zebrafish adults (n = 30 per group) were counted at indicated time points after injection challenge with SVCV (1×10⁸ TCID₅₀/mL).

(G and H) FTR42 deficiency impaired SVCV replication in zebrafish larvae (G) and adults (H) following SVCV infection. WT and ftr42^lof/lof zebrafish larvae (4 dpf) or adults (90 dpf) were challenged with SVCV as in (D) and (F), respectively. At indicated times, SVCV genes, including nucleoprotein (N), RNA polymerase (L), glycoprotein (G), matrix protein (M) and phosphoprotein (P) were detected by RT-qPCR.

Zebrafish survival curves were analyzed by Log-Rank test. Data were shown as mean ± SD (N = 3). p values were calculated using Student’s t test. ∗∗p < 0.01, ∗∗∗p < 0.001.

See also Figure S1 and Table S1.

FTR42-deficient zebrafish show enhanced IFN immunity

Since FTR42 was induced by virus infection and IFN, we hypothesized that FTR42 deficiency made a change in zebrafish IFN antiviral immunity. As expected, SVCV infection yielded a more robust and a time-dependent mRNA induction of ifnφ1 and ISGs (mxb and pkz) (Figure 2A), and of ftr42 (Figure 2B), in ftr42^lof/lof larvae than in WT larvae. Similarly, SVCV-infected tissues from ftr42^lof/lof adults displayed a high mRNA expression of ifnφ1 and ftr42 mRNA (Figures 2C and 2D). These results indicated that ftr42^lof/lof zebrafish show a stronger IFN immunity than WT zebrafish following SVCV infection.

Figure 2.

FTR42-deficient zebrafish show an enhanced IFN immunity

(A and B) FTR42-deficient zebrafish larvae showed an enhanced IFN immunity. WT and ftr42^lof/lof larvae (4 dpf) were immersed with SVCV (5×10⁶ TCID₅₀/mL) for indicated time points, followed by RT-qPCR analyses of ifnφ1, mxb and pkz (A) and ftr42 (B).

(C and D) FTR42-deficient zebrafish adults showed an enhanced IFN immunity. WT and ftr42^loflof adults (90 dpf) were i.p. injected with SVCV (1×10⁸ TCID₅₀/mL) for indicated time points, followed by RT-qPCR analyses of ifnφ1 (C) and ftr42 (D) in various tissues.

(E) GFP was continuously expressed in zebrafish embryos by microinjection. A GFP plasmid (60 pg/nL) was microinjected into one-cell-stage zebrafish embryos (1 nL per embryo), followed by fluorescence microscopy at indicated time points. Scale bar: 0.5 mm.

(F) IFNφ1 promoter was activated in zebrafish embryos by poly(I:C) or TBK1. One-cell-stage embryos were microinjected with DrIFNφ1pro-luc (50 pg/nL) and pRL-TK (2.5 pg/nL), together with poly(I:C) (50 pg/nL) or TBK1 (50 pg/nL), in a total volume of 1 nL per embryo. At the indicated time points, embryos were collected for luciferase assays.

(G and H) IFNφ1 promoter activation was inhibited in zebrafish embryos by FTR42. One-cell-stage WT embryos were microinjected as in (F) with DrIFNφ1pro-luc, poly(I:C) and FTR42 (50 pg/nL), together with pRL-TK (2.5 pg/nL) (G). Or WT and ftr42^lof/lof embryos were injected with DrIFNφ1pro-luc, poly(I:C) and pRL-TK, respectively (H). 60 h later, luciferase assays were performed.

Data were shown as mean ± SD (N = 3). p values were calculated using Student’s t test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

See also Figure S2 and Table S1.

It was surprising that ftr42^lof/lof zebrafish had a normal transcription of FTR42 (termed FTR42-Cas9 thereafter). We next analyzed the potential function of FTR42-wt (derived from wild-type zebrafish). Overexpression of FTR42-wt in fish cells followed by SVCV infection yielded more severe cytopathogenic effects (CPEs) and a higher level of viral titers than overexpression of empty vector (EV) (Figure S2A). Similarly, FTR42-wt suppressed mRNA expression of cellular IFN and ISGs in fish cells when transfected with poly(I:C) (Figure S2B), and also attenuated luciferase activity of zebrafish IFNφ1-promoter driven luciferase plasmid (IFNφ1pro-luc), by poly(I:C) and virus infection (SVCV and GCRV) (Figures S2C and S2D), indicating that FTR42-wt promotes viral replication likely by downregulating fish IFN response.

Compared to FTR42-wt mRNA encoding a 539-aa-protein, FTR42-Cas9 mRNA was predicated to have two possible ORFs by bioinformatic analyses (Figure S2E). To determine a possibility of whether FTR42-Cas9 mRNA could be translated into proteins, we constructed two expression plasmids by insertion of an HA tag into both predicated ORFs (Figure S2F); however, no proteins were detected by overexpression assays (Figure S2G). Further luciferase assays revealed an inhibitory potential of FTR42-wt, but not of FTR42-Cas9 (Figure S2H). Therefore, FTR42-Cas9 is a loss-of-function allele although it is transcriptionally induced in ftr42^lof/lof zebrafish following viral infection.

Subsequently, the IFN immunity between WT zebrafish and ftr42^lof/lof zebrafish was compared. Microinjection of a GFP reporter plasmid into one-cell-stage zebrafish embryos yielded a continuous and a robust expression of GFP from 12 to 72 h, indicating that zebrafish embryos could be used to do overexpression assays (Figure 2E). Actually, microinjection of poly(I:C) or TBK1 into WT embryos time-dependently induced luciferase activity of IFNφ1pro-luc (Figure 2F). In another assays, poly(I:C)-induced IFN promoter activation was abrogated in WT embryos by simultaneous microinjection of FTR42 (Figure 2G), and notedly, poly(I:C) stimulated a more significant luciferase activity in ftr42^lof/lof embryos than in WT embryos (Figure 2H). These results indicated that FTR42 deficiency enhances IFN immunity in zebrafish.

FTR42 attenuates TBK1 protein level to impair IFN response

How FTR42 downregulated IFN response was determined. Compared to overexpression of empty vector as control, overexpression of FTR42 inhibited IFNφ1 promoter activation by MAVS, STING, TBK1 and IRF7 but not by IRF3 (Figure 3A). Therefore, we initially focused on the effect of FTR42 on TBK1. It showed that, overexpression of FTR42 in EPC cells significantly attenuated TBK1-induced mRNA expression of cellular ifn and ISGs (mx and viperin) (Figure 3B), but did not have any effects on mRNA expression of the transfected TBK1 plasmids (Figure 3C), all of which were replicated by dose-dependent assays (Figure 3D). Western blotting showed that FTR42 decreased TBK1 expression at protein level, in a time-dependent manner (Figure 3E), and in a dose-dependent manner (Figure 3F). These results indicated that FTR42 downregulates TBK1-mediated fish IFN response by targeting TBK1 at protein level but not at mRNA level.

Figure 3.

FTR42 impairs IFN response by attenuating TBK1 protein level in vitro

(A) Overexpression of FTR42 inhibited IFN promoter activation by MAVS, MITA, TBK1 and IRF7 but not by IRF3. CAB cells seeded in 48-well plates overnight were transfected with DrIFNφ1pro-luc, each of signaling molecules (MAVS, MITA, TBK1, IRF3 and IRF7) or empty vector as control (100 ng each), together with FTR42 at increasing amounts. 10 ng pRL-TK was included as internal control. 30 h later, cells were harvested for luciferase assays.

(B–D) FTR42 downregulated TBK1-trigged IFN response. EPC cells seeded in 6-well plates were transfected with the indicated plasmids as in (A), and were then collected at the indicated time points (B and C) or at 30 h post transfection (D), for RT-qPCR analyses of mRNA expression of cellular ifn and ISGs (B and D), and mRNA expression from the transfected FTR42-vt and TBK1-vt (C and D).

(E and F) FTR42 attenuated TBK1 protein levels in vitro. EPC cells seeded in 3.5 cm² dishes overnight were transfected with TBK1, FTR42 or pcDNA3.1 as control (1 μg each), and at different time points, the cells were harvested for western blotting analyses (E). Or the cells were transfected for 30 h with TBK1 and FTR42 at the indicated doses, followed by western blotting (F).

Data were shown as mean ± SD (N = 3). p values were calculated using Student’s t test. ∗∗∗p < 0.001.

See also Figure S3 and Table S1.

To investigate the interaction between FTR42 and TBK1 proteins, fluorescent confocal microscopy illuminated that either FTR42 or TBK1 was expressed in cytoplasm when overexpressed individually, and they were co-localized when overexpressed together (Figure S3A). Using the truncates of FTR42 and TBK1 (Figure S3B), Co-IP assays showed that two domain-deleted truncates FTR42, ΔRING and ΔSPRY, did not interact with TBK1 (Figure S3C), but any a domain of TBK1 could bind to FTR42 (Figure S3D).

FTR42 targets TBK1 for protein degradation dependent on its E3 ligase activity

TRIM proteins generally have E3 ligase activity dependent on their RING domains.²⁵ We made eight single-site mutants of FTR42 and one multiple-site mutant, E3-mut (Figure 4A), because these amino acids are conserved in the RING domains of TRIM proteins (Figure S4A), and are essential for E3 ligase activity in FTRCA1²⁶ and human TRIM40,²⁸ respectively. Overexpression of His-ubiquitin and FTR42-Flag followed by affinity purification of His-ubiquitin using Ni²⁺-NTA resin showed that the full-length FTR42 was ubiquitinated more significantly than ΔRING or E3-mut, indicating that an intact FTR42’s RING domain was essential for its E3 ligase activity (Figure 4B). Consistently, ΔRING or E3-mut failed to suppress TBK1-triggered IFN promoter activation (Figure 4C). Similar results were seen for eight single-site mutants of FTR42 (Figure S4B).

Figure 4.

FTR42 inhibits TBK1-triggered IFN response dependent on its E3 ligase activity

(A) Schematic diagrams of full-length FTR42, ΔRING and E3-mut.

(B) FTR42 but not ΔRING or E3-mut possessed the E3 ligase activity. HEK293T cells seeded in 10 cm² dishes were transfected with His-Ub, together with FTR42-Flag or ΔRING-Flag or E3-mut-Flag (5 μg each). 30 h later, the cells were lysed and incubated with Ni²-NTA overnight, followed by western blotting to determine the ubiquitination of FTR42 and two mutants.

(C) FTR42-ΔRING or E3-mut failed to inhibit TBK1-triggered IFNφ1 promoter activation. CAB cells seeded in 48-well plates overnight were transfected with DrIFNφ1pro-luc, TBK1 or empty vector (100 ng each), together with FTR42 or ΔRING or E3-mut at increasing amounts. 30 h later, the cells were harvested for luciferase assays.

(D) FTR42-ΔRING or E3-mut failed to attenuate TBK1 protein level in vitro. CAB cells seeded in 3.5 cm² dishes were co-transfected with the indicated plasmids (1 μg each) for 30 h, followed by western blotting assays with tag-specific Abs.

(E and F) FTR42-triggered TBK1 degradation was blocked by MG132 but not by NH₄Cl or chloroquine. CO cells seeded in 12-well plates were co-transfected with indicated plasmids for 24 h, followed by treatment with MG132, NH₄Cl, or chloroquine for additional 6 h.

p values were calculated using Student’s t test. ∗∗∗p < 0.001; ns, not significant.

See also Figure S4.

Western blotting showed that ΔRING or E3-mut failed to facilitate TBK1 protein degradation (Figure 4D), indicating that the E3 ligase activity is essential for FTR42-targeted TBK1 protein degradation. Subsequently, incubation of MG132 (an inhibitor of ubiquitin proteasome-dependent degradation pathway) resulted in a significant accumulation of TBK1 protein in the presence of FTR42 (Figure 4E), but either NH₄Cl or chloroquine (two inhibitors of autophagy-lysosome–dependent degradation pathway) had not a similar effect (Figure 4F). These results together suggest that FTR42 promotes TBK1 protein degradation through the proteasome-dependent degradation pathway and thus downregulates IFN response.

FTR42-deficient zebrafish show enhanced TBK1 protein expression and improved TBK1-mediated IFN immunity

Consistent with the aforementioned results showing no effects of FTR42 on mRNA expression of TBK1 in vitro (Figure 3C), time course analyses of SVCV-infected ftr42^lof/lof zebrafish tissues revealed a similar mRNA expression of tbk1 but an enhanced one of ifnφ1, compared to WT zebrafish (Figure 5A). Nearly a constant level of tbk1 mRNA was also detected in WT larvae and ftr42^lof/lof larvae following SVCV infection (Figure 5B). These results indicated that zebrafish tbk1 mRNA is constitutively expressed and is not impacted by viral infection or FTR42 deficiency. On the contrary, ftr42^lof/lof larvae exhibited an enhanced TBK1 protein level over WT larvae (Figure 5C), which was also seen in SVCV-infected gill from ftr42^lof/lof adults (Figure 5D).

Figure 5.

FTR42 impairs IFN response by attenuating TBK1 protein level in vivo

(A and B) FTR42-dificient zebrafish showed an enhanced expression of ifnφ1 mRNA over WT zebrafish but a constant tbk1 mRNA expression in response to SVCV infection. Zebrafish adults (90 dpf, A) and zebrafish larvae (4 dpf, B) were challenged with SVCV. At the indicated time points, tissues or larvae were sampled for RT-qPCR analyses of cellular ifnφ1 and tbk1 mRNA.

(C and D) FTR42-dificient zebrafish showed a higher level of TBK1 protein than WT zebrafish upon SVCV infection. Zebrafish larvae (C) and adults (D) were infected with SVCV. At the indicated time points, larvae or gill were sampled for western blotting analyses of endogenous TBK1 protein.

(E) TBK1 activated IFN promoter more significantly in ftr42^lof/lof embryos than in WT embryos. One-cell-stage embryos from WT and FTR42-dificient zebrafish were microinjected with TBK1 (50 pg), DrIFNφ1pro-luc (50 pg) and pRL-TK (2.5 pg), in a volume of 1 nL per embryo. At 48 h or 60 h post injection, the embryos were collected for luciferase assays.

(F) TBK1 induced transcriptional expression of ifnφ1 and mxb more significantly in ftr42^lof/lof embryos than in WT embryos. One-cell-stage embryos were microinjected with TBK1(60 pg), in a volume of 1 nL per embryo. 48 h later, ifnφ1 and tbk1 mRNA was detected by RT-qPCR.

(G) A constant expression of tbk1 mRNA was detected in ftr42^lof/lof and WT embryos after microinjection. The same samples in (F) were used to detect mRNA expression from the injected TBK1 plasmid by RT-qPCR.

Data were shown as mean ± SD (N = 3). p values were calculated using Student’s t test. ∗∗p < 0.01, ∗∗∗p < 0.001; ns, not significant.

The effect of FTR42 on TBK1-mediated IFN response was further determined in WT and ftr42^lof/lof embryos. Microinjection of TBK1 could stimulate IFN promoter activation in WT embryos, which was better in ftr42^lof/lof embryos, particularly at 60 h post injection (Figure 5E). Under the same conditions, endogenous ifnφ1 and mxb gene transcription was significantly induced by microinjection of TBK1, with a stronger induction in ftr42^lof/lof embryos than in WT embryos (Figure 5F). Notedly, similar levels of tbk1 mRNA expression were detected in WT embryos and ftr42^lof/lof embryos (Figure 5G). These results indicated that FTR42 impairs IFN response by attenuating TBK1 protein level in vivo.

FTR42 facilitates IRF7 mRNA decay to downregulate IFN response in vitro

We further determined how FTR42 downregulated IRF7-mediated IFN response. Overexpression assays showed that IRF7 protein was decreased by FTR42 in a time-dependent manner (Figure 6A), and in a dose-dependent manner (Figure 6B). However, FTR42 did not target IRF7 for protein degradation, because there was no change in FTR42-triggered IRF7 protein attenuation when protein degradation pathways were blocked by MG132 and NH₄Cl or chloroquine, respectively (Figure 6C). Contrarily, irf7 mRNA expression was time-dependently reduced by FTR42 (Figure 6D), and IRF7-mediated transcription of cellular ifn and ISGs were reduced concomitantly (Figure 6E). Similar results were replicated in a dose-dependent assay (Figure 6F).

Figure 6.

FTR42 facilitates IRF7 mRNA decay to downregulate IFN response in vitro

(A–C) FTR42 attenuated IRF7 protein levels independently of protein degradation. CAB cells seeded in 3.5 cm² dishes overnight were transfected with IRF7, together with FTR42 (1 μg ecah) for different time points (A) or with FTR42 at increasing amounts (B) for 30 h; or CAB cells seeded in 12-well plates were transfected for 24 h with FTR42 and IRF7 (0.5 μg each), followed by treatment with MG132, NH₄Cl, or chloroquine for additional 6 h (C).

(D–F) FTR42 downregulated IRF7-trigged IFN response by attenuating IRF7 transcription. CAB cells seeded in 3.5 cm² dishes were transfected with IRF7 and FTR42 (1 μg each) for different time points (D and E), or CAB cells seeded in 6-well plates were transfected for 24 h with IRF7 (0.5 μg) and FTR42 at increasing amounts (F), followed by RT-qPCR detection of mRNA expression from the transfected FTR42 and IRF7 plasmid (D and F), or from cellular ifn, mx and viperin (E and F).

(G) FTR42-mediated reduction of IRF7 protein was blocked by CHX. CAB cells seeded in 12-well plates overnight were transfected for 24 h with IRF7 (0.5 μg) and FTR42 at increasing doses (0, 0.2, 0.4, 0.8 μg), followed by addition of CHX or DMSO as control. Another 6 h later, the cells were collected for western blotting (left panel), followed by densitometric quantification of IRF7 protein expression normalized to β-actin (right panel).

(H) FTR42-mediated reduction of IRF7 mRNA was not impaired by ActD. CAB cells seeded in 12-well plates overnight were transfected as (F). 24 h later, ActD (1 μg/mL) was added to the plates for 4 h or 7 h, followed by RT-qPCR detection of mRNA from the transfected IRF7 plasmid.

Data were shown as mean ± SD (N = 3). p values were calculated using Student’s t test. ∗∗p < 0.01, ∗∗∗p < 0.001.

See also Figure S5 and Table S1.

These results indicated that FTR42 downregulated fish IFN response by attenuating IRF7 expression at mRNA level but not at protein level. To reinforce this notion, we evaluated FTR42-directed IRF7 expression in the presence of cycloheximide (CHX) or actinomycin D (ActD). As expected, IRF7 protein reduction was abolished by the protein synthesis inhibitor CHX (Figure 6G), but irf7 mRNA reduction was not impaired by the transcription inhibitor ActD (Figure 6H). Therefore, FTR42 attenuates IRF7 expression by facilitating irf7 mRNA decay.

Subsequent assays were performed to determinate whether the E3 ligase activity of FTR42 contributed to downregulating IRF7-mediated IFN response. Both zebrafish IFNφ1 and IFNφ3 promoters were used, because IRF7 is mainly responsible for IFNφ3 expression and minorly for IFNφ1 expression.²⁹ Whereas the full-length FTR42 displayed a potential to inhibit IRF7-triggered fish IFN promoter activation (Figure S5A), this inhibition was not seen for E3-mut or ΔRING (Figure S5B). Either E3-mut or ΔRING failed to reduce IRF7 protein expression (Figures S5C and S5D). Therefore, the E3 ligase activity is essential for FTR42 to downregulate IRF7-mediated fish IFN response.

FTR42-deficient zebrafish show enhanced IRF7 mRNA expression and improved IRF7-mediated IFN response

Consistent with the result that FTR42 reduced irf7 mRNA expression in vitro (Figure 6), RT-qPCR analyses of SVCV-infected tissues from zebrafish adults showed an increased mRNA expression of irf7 and ifnφ1 in ftr42^lof/lof fish, compared to that in WT zebrafish (Figure 7A), which was similarly detected by comparison of ftr42^lof/lof larvae and WT larvae (Figure 7B). Consistently, SVCV infection resulted in an increased IRF7 protein level in ftr42^lof/lof zebrafish larvae (Figure 7C), and also in gill from ftr42^lof/lof zebrafish adults (Figure 7D).

Figure 7.

FTR42 facilitates IRF7 mRNA decay to downregulate IFN response in vivo

(A and B) FTR42-dificient zebrafish showed enhanced mRNA expression of irf7 and ifnφ1 over WT zebrafish upon SVCV infection. Zebrafish adults (90 dpf, A) and zebrafish larvae (4 dpf, B) were challenged with SVCV. At the indicated time points, tissues or larvae were sampled for RT-qPCR analyses of cellular ifnφ1 and irf7.

(C and D) FTR42-dificient zebrafish showed a higher level of IRF7 protein than WT zebrafish upon SVCV infection. Zebrafish larvae (C) and adults (D) were infected with SVCV. At the indicated time points, larvae or gill were sampled for western blotting analyses of endogenous IRF7 protein.

(E) IRF7 stimulated IFN promoter activation more significantly in ftr42^lof/lof embryos than in WT embryos. One-cell-stage embryos from WT and FTR42-dificient zebrafish were microinjected with IRF7 (50 pg), DrIFNφ1pro-luc (50 pg) and pRL-TK (2.5 pg), in a volume of 1 nL per embryo. At the indicated time points, the embryos were collected for luciferase assays.

(F) IRF7 induced transcriptional expression of ifnφ1 and mxb more significantly in ftr42^lof/lof embryos than in WT embryos. One-cell-stage embryos were microinjected with IRF7 (60 pg), in a volume of 1 nL per embryo. 48 h later, ifnφ1 and mxb mRNA was detected by RT-qPCR.

(G) FTR42-dificient zebrafish expressed a higher level of IRF7 mRNA than WT zebrafish after microinjection. The same samples in (F) were used to detect irf7 mRNA from the microinjected IRF7 plasmid by RT-qPCR.

Data were shown as mean ± SD (N = 3). p values were calculated using Student’s t test. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001.

Subsequently, microinjection of IRF7 into WT embryos significantly stimulated IFN promoter activation at 60 h post injection, with more enhanced stimulation when microinjection was done in ftr42^lof/lof embryos (Figure 7E). Whereas ifnφ1 and mxb were transcriptionally induced in WT embryos by microinjection of IRF7, a more robust induction was detected in ftr42^lof/lof embryos (Figure 7F). Under the same conditions, ftr42^lof/lof embryos had a high level of irf7 mRNA over WT embryo (Figure 7G). These results together indicated that FTR42 deficiency enhances IRF7-mediated IFN response by attenuating irf7 mRNA decay in vivo.

FTR42 restores the susceptibility of ftr42^lof/lof zebrafish to SVCV infection

To further determine the relevance of FTR42 to virus resistance in zebrafish, we attempted to restore the expression of FTR42 in ftr42^lof/lof zebrafish by microinjecting FTR42 plasmids into one-cell-stage embryos. WT embryos were parallelly microinjected as control. Three days later, the injected embryos developed into larvae, followed by immersion challenge with SVCV. As shown in Figure 8A, compared to mock injection (with empty vector, EV), FTR42 injection led to a decreased survival in ftr42^lof/lof larvae (survivals: 35:9, n = 50), and in WT larvae (survivals: 25:11, n = 50). However, mock injection or E3-mut injection nearly did not change the improved virus resistance of ftr42^lof/lof larvae compared to WT larvae (survivals: 35:25 for EV injection versus 43:29 for E3-mut injection, n = 50). Daily counts of survivals showed that at 2 days post infection, all ftr42^lof/lof larvae were dead after FTR42 injection, while 54% and 76% survived after E3-mut injection or mock injection (Figure 8B). These results indicated that microinjection of FTR42 but not E3-mut can reduce the viral susceptibility of ftr42^lof/lof zebrafish to a level similar to WT zebrafish.

Figure 8.

FTR42 but not E3-mut reduces the susceptibility of ftr42^lof/lof zebrafish to SVCV infection

(A) Representative images of WT and ftr42^lof/lof zebrafish larvae (4 dpf, n = 50 per group) after SVCV immersion challenge (5×10⁶ TCID₅₀/mL) for 36 h. zebrafish larvae were raised from embryos that had been microinjected at one cell stage with FTR42 or E3-mut or empty vector (60 pg each) in a volume of 1 nL.

(B) Survival ratios were calculated by counting deaths of WT and ftr42^lof/lof zebrafish larvae (n = 50 per group), which were developed from embryos through microinjection and subsequently immersed with SVCV as in (A).

(C–E) Transcriptional expression of ifnφ1 and mxb (C), irf7 and tbk1 (D), and five SVCV genes (E) were compared between WT and ftr42^lof/lof zebrafish larvae, which were developed from embryos through microinjection and subsequently immersed for 48 h with SVCV as in (A).

Zebrafish survival curves were analyzed by Log-Rank test. Data were shown as mean ± SD (N = 3). p values were calculated using two-way ANOVA analysis. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; ns, not significant.

We next determined the change in IFN immunity under the condition as in Figure 8B. Compared to mock injection, FTR42 injection reduced mRNA expression of ifnφ1 and mxb in ftr42^lof/lof larvae back to a level similar to WT larvae with mock injection, but E3-mut injection did not (Figure 8C). Similarly, injection with FTR42 rather than E3-mut reduced irf7 mRNA expression to a level equivalent to WT larvae (Figure 8D, left panel). However, nearly similar levels of tbk1 mRNA were detected regardless of what was injected (Figure 8D, right panel). Consistently, mock-injected WT larvae and FTR42-injected ftr42^lof/lof larvae displayed a higher mRNA expression of five SVCV genes than mock- or E3-mut-injected ftr42^lof/lof larvae (Figure 8E). These results suggest that FTR42 is the causative gene contributing to zebrafish virus resistance, because FTR42 deficiency can improve zebrafish survival against viral infection by enhancing host IFN response.

Discussion

It is well documented that zebrafish do not develop fully functional adaptive immunity until 4–6 weeks post fertilization.^14,30,31,32 In the current study, zebrafish larvae (<10 days post fertilization) are challenged with SVCV infection, indicating that the consequent mortality and survival are ascribed to the innate immunity in zebrafish. Compared to WT zebrafish, both ftr42^lof/lof zebrafish larvae and adults show a better survival against viral infection and also a stronger IFN immunity, which is further verified by genetical complementation assays showing that, FTR42 microinjection can return the improved viral resistance and enhanced IFN immunity of ftr42^lof/lof zebrafish to a level equivalent to WT zebrafish. Given that FTR42 is a negative regulator of fish IFN response, our results suggest that genome editing of FTR42 has significantly improved the virus resistance in zebrafish, as the resulting FTR42 deficiency leads to a robust IFN immune response toward viral infection.

FTR42 is an E3 ligase belonging to a fish-specific TRIM subfamily, termed finTRIM subfamily, which largely originates from lineage-specific, genus-specific and even species-specific gene expansion.^22,23,24,26 The ongoing and rapid expansion of finTRIM subfamily is speculated to help fish resist virus infection.^23,24,26 This is consistent with emerging evidences that have related the pivotal function of TRIM proteins to the host innate antiviral response.^19,21,33 Actually, more than one-third of TRIM genes are upregulated by IFNs in human.^20,34 In fish, most finTRIM genes are initially identified as ISGs.²⁵ FTRCA1 is a crucian carp-specific finTRIM gene, which is phylogenetically homologous to five zebrafish genes, including FTR41, FTR42, FTR58, FTR57, and FTR94.²⁶ Although FTRCA1 and any one zebrafish gene do not possess a “one to one” orthologous relationship, FTR42 downregulates IFN response by facilitating degradation of TBK1 protein and IRF7 mRNA in zebrafish, similar to FTRCA1 in crucian carp.^15,26 These results indicate that both FTR42 and FTRCA1 exert a conserved regulation of the IFN immune response in different fish species.

FTR42 and FTRCA1 genes exhibit common virus- and IFN-inducible features, implying that they might act as a brake to progressively restrict cellular IFN production in a dose-dependent manner, with a full effect when the invading virus is cleared. This feature is essential for homeostatic regulation of host IFN response to largely avoid cell damages by an excessive IFN production.⁶ That is, cellular IFN production must return to the basal level in normal physical conditions. Unfortunately, during the evolution of virus-cell interactions, the suppression of innate immune signaling by TRIM proteins may be usurped by invading viruses to facilitate viral replication.²⁰ This might be a possible interpretation why loss-of-function of FTR42 is beneficial to zebrafish survival against viral infection.

The vertebrate IFN is initially identified as a broad-spectrum antiviral cytokine by inducing the expression of ISG sets with antiviral effects.^5,35 Now it has become clear that IFNs also induce an epigenomic signature by activating various sets of ISGs, thus priming immune cells and training innate immune memory.⁹ However, IFN genes seem not to be ideal targets for resistance breeding in fish. In mammals, IFN transgenic mice are male sterility due to a continuous and excess IFN signaling,^36,37,38 although displaying a significantly enhanced viral resistance in some cases.³⁹ By contrast, IFN transgenic medaka (Oryzias latipes) exhibits an increased susceptibility to virus infection and an impaired IFN signaling likely by a prolonged exposure to IFN at high levels,¹⁰ a desensitization effect as found in human.⁴⁰ These results indicate that, to appropriately promote host IFN response, just toward virus infection but not in normal conditions, might be essential for host survival against virus infection. Apparently, zebrafish FTR42 is induced time-dependently during viral infection as an inhibitor of host IFN response; therefore, FTR42 deficiency confers a rapider and stronger IFN immunity in zebrafish just in response to virus infection, which is more conductive to zebrafish clearing virus infection. Besides the improved resistance to virus infection, ftr42^lof/lof zebrafish are viable, without visible changes in growth, feeding and reproduction during over two-year successive breeding in our lab. These promising observations indicate that the homologous genes of zebrafish FTR42 might be ideal targets for genome editing to improve the viral disease-resistant trait in farmed fish species.

Recently, some negative regulators of fish IFN immunity, similar to FTR42 in the present study, have been identified,^41,42,43 including TRIM and HERC genes.^15,16,26,44 These negative regulators might be particularly compatible with genome editing in fish breeding for resistance to viral diseases, if they do not largely impair other economic traits in fish. It is believed that the genome-editing technology can efficiently create diverse alleles with complete or partial loss-of-function, which has been recently utilized in rice,¹⁸ and also in model fish.^45,46 The ftr42^lof/lof zebrafish generate a novel FTR42 allele (called FTR42-Cas9). Interestingly, FTR42-Cas9 is transcriptionally induced by SVCV infection, but is not translated into a function protein in ftr42^lof/lof zebrafish. A similar phenomenon is observed in ISG15-deficient patients.⁴⁷ These patients have one-base mutation in exon 2 of ISG15, resulting in a normal transcription but loss-of-expression of ISG15 protein due to generating a premature stop codon. However, the cells derived from these patients exhibit a broad-spectrum resistance to viral infection in vitro, as a result of enhanced IFN immunity.⁴⁷ We do not know whether there are differences in survival rates against viral infection between zebrafish mutations with a complete or a partial FTR42’s deficiency. Given that the magnitude of cellular IFN response is tightly related to a complete/partial genome edition within a targeted gene, and that appropriate production of cellular IFN is essential for the improved resistance to viral infection in fish, selective screening of diversified CRISPR/Cas9-created alleles would be helpful and also be easy to rapidly obtain the fish variants with preferred resistance traits.

Limitations of the study

Due to lack of more aquatic viruses to infect zebrafish, the broad-spectrum disease resistance of ftr42^lof/lof zebrafish needs further verification, also including resistance to bacterial infection. Our results in the current study reveal that zebrafish FTR42 downregulates IFN response by targeting TBK1 and IRF7 for protein and mRNA degradation, indicating that FTR42 is a multifaceted inhibitor to target different signaling factors for shaping fish IFN immunity. This raises a question of whether other signal factors, besides TBK1 and IRF7, are targeted by FTR42 to fine-tune the IFN response. As a broad-spectrum cytokine, fish IFNs also contribute to survival and functionality of fish B cells,⁴⁸ implying a possibility that, besides the enhanced innate IFN immunity, a promoted adaptive immunity during viral infection might be activated in ftr42^lof/lof zebrafish adults. Despite these interesting issues unknown, our results demonstrate that properly increasing fish IFN immunity by genome editing of negative regulators is a good strategy to genetically improve virus resistance in aquaculture fish.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rabbit monoclonal anti-HA	Cell Signaling Technology	Cat# 3724; RRID: AB_1549585
Rabbit monoclonal anti-Flag	Cell Signaling Technology	Cat# 14793; RRID: AB_2572291
Rabbit polyclonal anti-β-actin	Cell Signaling Technology	Cat# 4967; RRID: AB_330288
Rabbit polyclonal anti-DrTBK1	This paper	N/A
Rabbit polyclonal anti-CaIRF7	Gong et al.14	N/A
Rabbit polyclonal anti-CaSTAT1	This paper	N/A
Bacterial and virus strains
Spring viraemia of carp virus (SVCV)	Wu et al.26	N/A
Glass carp reovirus (GCRV)	Li et al.16	N/A
Chemicals, peptides, and recombinant proteins
MG132	MedChemExpress	HY-13259; CAS: 133407-82-6
NH4Cl	MedChemExpress	HY-Y1269; CAS: 12125-02-9
Cycloheximide (CHX)	MedChemExpress	HY-12320; CAS: 66-81-9
Actinomycin D (ActD)	MedChemExpress	HY-17559; CAS: 50-76-0
Chloroquine (CQ)	Cell Signaling Technology	Cat# 14774; CAS: 50-63-5
PEI	Sigma-Aldrich	919012; CAS: 49553-93-7
poly(I:C)	Sigma-Aldrich	Cat# P0913; CAS: 42424-50-0
Critical commercial assays
Hifair® T7 High Yield RNA Synthesis Kit	YEASEN	10623ES60
TrueCut™ Cas9 Protein v2	Invitrogen	A36499
Dual-Luciferase® Reporter Assay System 10-Pack	Promega	E1960
SteadyPure Universal RNA Extraction Kit	Accurate Biotechnology	AG21017
MonScript™ RTIII Super Mix with dsDNase	Monad	MR05201
Hieff® qPCR SYBR Green Master Mix (No Rox)	YEASEN	Cat# 11201ES08
Deposited data
Raw and analyzed data	This paper	Lead Contact, Yibing Zhang (ybzhang@ihb.ac.cn)
Experimental models: Cell lines
HEK293T	ATCC	CRL-3216
Epithelioma papulosum cyprini cells (EPC)	ATCC	CRL-2872
Ovary cells of grass carp (CO)	Kept in IHB, CAS	N/A
Crucian carp (C. auratus L.) blastulae embryonic cells (CAB)	Kept in IHB, CAS	N/A
Experimental models: Organisms/strains
Zebrafish (Danio rerio) strain AB	China Zebrafish Resource Center	N/A
Oligonucleotides
see Table S1
Recombinant DNA
pcDNA3.1(+)	Invitrogen	Cat# V79020
Software and algorithms
GraphPad Prism 9.0.0	N/A	https://www.graphpad.com/
ZEN	N/A	https://www.zeiss.com.cn/
Figdraw	N/A	https://www.figdraw.com/static/index.html#/

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Yibing Zhang (ybzhang@ihb.ac.cn).

Materials availability

The ftr42-deficient zebrafish mutant line and plasmids generated in this study are available from the lead contact without restriction.

Data and code availability

• •Raw and analyzed data reported in this paper will be shared by the lead contact upon request.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Experimental model and study participant details

Cell lines and virus strains

Epithelioma papulosum cyprini (EPC) cells and Human embryonic kidney (HEK) 293T cells are from ATCC. Grass carp C. idellus ovary (CO) cells, and crucian carp (C. auratus) blastulae embryonic (CAB) cells are kept in our Institute of Hydrobiology, Chinese Academy of Sciences. All cell lines were tested free of mycoplasma contamination. HEK293T cells were cultured at 37°C in Dulbecco’s Modified Eagle’s medium (DMEM; HyClone) supplemented with 10% fetal bovine serum (FBS; ExCell Bio) and 1% Penicillin-Streptomycin (Life Technologies). CAB cells, EPC cells and CO cells were cultured at 28°C in medium 199 (HyClone) supplemented with 10% FBS and 1% Penicillin-Streptomycin (Life Technologies).

Spring viraemia of carp virus (SVCV), a negative sense single-stranded RNA virus, and glass carp reovirus (GCRV), a double stranded RNA virus, were all propagated in EPC cells. When 80% cytopathic effects (CPEs) appeared, the cells were frozen and thawed three times, followed by collection of the supernatants at -80°C. Virus titers were determined by a 50% tissue culture-infective dose (TCID₅₀) assay.¹⁴

Experimental models

AB strain wild zebrafish (Danio rerio) were provided by the China Zebrafish Resource Center (CZRC), and were raised, maintained and bred according to the standard protocols.¹⁴ All zebrafish were randomly grouped according with principle of randomization. All studies were approved by the Animal Care and Use Committee of Institute of Hydrobiology, Chinese Academy of Sciences.

Method details

Plasmids

The open reading frames (ORF) of zebrafish FTR42 (GenBank: XM_690458.8) was cloned into EcoR V site of pcDNA3.1(+) vector (Invitrogen) which had a Flag or a HA coding sequence in the Not I site. FTR42 mutants including E3-mut, and truncates including ΔRING, were derived from these FTR42 plasmids. The ORFs of zebrafish RIG-I (GenBank: NM_001306095.1), MDA5 (GenBank: NM_022168.4), MAVS (GenBank: NM_001080584.2), MITA (GenBank: NM_001080584.2), TBK1 (GenBank: NM_001044748.2), IRF3 (GenBank: NM_001044748.2), and IRF7 (GenBank: NM_200677.2) were subcloned into EcoR I/BamH I site of pcDNA3.1(+) vector. Zebrafish TBK1-Flag, TBK1-HA and IRF7-HA were constructed by inserting the corresponding ORFs into EcoR V site of pcDNA3.1(+) vector. FTR42pro-luc was made by insertion of a 5’-flanking sequence (-425 to +75) of FTR42 into pGL3-Basic (Promega). Zebrafish IFNφ1pro-luc was made by insertion of a 5’-flanking sequence (-586 to +38) of zebrafish ifnφ1 gene into the Xho I/Hand III site of pGL3-Basic. And DrIFNφ3pro-luc was made by insertion of a 5’-flanking sequence (-1448 to -8) of zebrafish ifnφ3 gene into the Kpn I/Xho I site of pGL3-Basic. For subcellular localization, the ORFs of zebrafish FTR42 and TBK1 were inserted into pEGFP-N3 or pCS2-mCherry vectors (Clontech), respectively. The primers used were listed in Table S1.

Antibodies and reagents

Anti-FLAG (1:5000, catalog no. 14793S), anti-HA (1:5000, catalog no. 3724S), and anti-β-actin (1:5000, catalog number: 4967S) were purchased from Cell Signaling Technology (Danvers, MA) and ABclonal Technology (Wuhan, China), respectively. Anti-TBK1 Ab was generated by immunization of rabbits with a purified peptide corresponding to 360–550 aa of zebrafish TBK1. Anti-CaIRF7 and anti-CaSTAT1 Abs were generated by immunization of rabbits with purified full length peptides of crucian carp IRF7 and STAT1.^14,49

MG132, NH₄Cl, Cycloheximide (CHX) and Actinomycin D (ActD) were purchased from MedChemExpress (Monmouth Junction, NJ). Chloroquine was from Cell Signaling Technology (Danvers, MA). Polyinosinic-polycytidylic acid [poly(I:C)] was from Sigma-Aldrich (St. Louis, MO).

CRISPR/Cas9 editing in zebrafish

Single-guide RNA (sgRNA) was designed against a sequence in exon 1 of zebrafish ftr42 by a CRISPR design tool, termed CCTop-CRISPR/Cas9 target online predictor (https://cctop.cos.uni-heidelberg.de:8043/). PUC9-gRNA vector was used to amplify the ftr42 sgRNA template with following primers: forward 5’-GTAATACGACTCACTATAGGATTCCAGACACTGCAGACGTTTTAGAGCTAGAAATAGC-3’ and reverse 5’-AAAAGCACCGACTCGGTGCC-3’. SgRNA was synthesized using T7 High Yield RNA Synthesis Kit (Yeasen).

One-cell-stage zebrafish embryos were microinjected with a mixture of 1×Cas9 protein (Invitrogen) and 0.15 ng sgRNA, in a total volume of 1 nL per embryo. At 24 h post microinjection, total DNA was extracted from embryos to amplify the genomic DNA fragments including CRISPR/Cas9-targeted sites with primers listed in Table S1, followed by sequencing to determine whether sgRNA was conductive. If so, the embryos were raised up until maturity as F0 zebrafish. Further breeding of a mature F0 with a WT zebrafish (strain AB) generated F1 zebrafish with heterozygous mutation (ftr42^+/lof). F1 male and F1 female, with the same phenotype, were next bred to obtain F2 zebrafish with homozygous mutation (ftr42^lof/lof). Up to now, the ftr42^lof/lof zebrafish are viable with normal growth, feeding and reproduction after over two-year succussive breeding in our lab.

Viral infection in zebrafish

According to our previous reports,^14,50 zebrafish larvae (4 dpf) were transferred to petri dishes filled with SVCV solution, where SVCV was diluted in egg water to 5×10⁶ TCID₅₀/mL. 30 larvae at a time point were collected for RT-qPCR and 50 for western blotting. Zebrafish adults (90 dpf) were intraperitoneal (i.p.) injected with SVCV solution (1×10⁸ TCID₅₀/mL), 20 μL each. 3 zebrafish at a time point were collected for tissue sampling, or just for taking a picture, after anesthetization with tricaine methane sulfonate (MS-222). All operations were performed at 28°C. Death curves were made by counting the deaths of zebrafish larvae or adults every few hours, until one group of fish was completely dead. Notedly, the dead fish must be removed immediately to keep the water clean.

Microinjection in zebrafish

Typically, one-cell-stage embryos after fertilization were immediately microinjected with injection solution, 1 nL per embryo. For luciferase assays, pRL-TK, promoter-driven luciferase reporter plasmid and expression construct were mixed at a ratio of 1:20:20 as injection solution, and at a time point, 15 embryos were collected to detect luciferase activities. For RT-qPCR assays, 30 embryos were collected at each time point.

Cell transfection and luciferase assays

Cell transfection was performed according to our previous reports.^15,50 Typically, cells seeded overnight were transfected with indicated plasmids which were diluted with Opti-MEM and mixed with polyethylenimine (PEI, molecular mass 25000 Da, Sigma-Aldrich), at a ratio of 1:3 (plasmids (μg)/PEI (μl)) for cultured fish cells and 1:5 for cultured HEK293T cells.

For luciferase assays, cells were transfected with indicated plasmids at a ratio of 1:10:10 (pRL-TK, promoter-driven luciferase reporter plasmid, expression construct). Empty vector was added to make a consistent amount of transfected plasmid in each sample. Luciferase activities were measured by a Junior LB9509 luminometer (Berthold, Pforzheim, Germany), and were expressed as a value relative to the amounts of Renilla luciferase activity in the same sample according to the Dual-Luciferase reporter assay system (Promega, Madison, WI). All results were shown as a representative of more than three independent experiments, each performed in triplicate.

Real-time PCR

Total RNAs were extracted using Universal RNA Extraction Kit (Accurate Biology, Hunan, China) according to manufacturer’s protocol. First-strand cDNA was synthesized using oligo(dT)₂₀VN and reverse transcriptase MonScript 5× RTIII super mix (Monad Biotech, Suzhou, China). RT-qPCR was performed with HiEff quantitative PCR SYBR Green master mix (Yeasen Biotechnology, Shanghai, China) on the CFX96 real-time system (Bio-Rad). The relative genes expression was normalized to β-actin. All experiments were independently repeated at least three times, with samples in triplicate. The primers for qRT-PCR assays are listed in Table S1.

Western blotting

Cell samples were boiled directly with 1×SDS loading buffer for 10 min. For zebrafish tissue samples, 150 μL NP-40 lysis buffer were added to the samples, followed by ground with homogenizer, and boiling for 10 min after 5×SDS loading buffer was added. After high-speed centrifugation to remove tissue fragments, supernatants were ultrasonically broken to remove the genome in the samples. All samples were separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membrane (Millipore). These membranes were blocked with milk and incubated with the indicated Abs, and then imaged the blots by using an ImageQuant LAS 4000 system (GE Healthcare).

Fluorescence microscopy

HEK293T cells seeded on cover glasses in 6-well plates overnight were transfected with indicated plasmids. At 30 h post transfection, cells were washed twice with ice-cold PBS and fixed by 4% (v/v) paraformaldehyde (PFA) for 30 min. Samples were washed three times to remove PFA, incubated with 0.2% triton X-100 for 15 min, washed again with PBS, and strained with DAPI (Beyotime) for 15 min, followed by observation with a confocal microscope (ZEN Blue lite confocal system: objectives, ×340; analysis software, ZEN 2.3 [Blue edition]).

Quantification and statistical analysis

All tests above were performed by three independent biological repeats unless otherwise indicated. All quantitative data are expressed as mean ± SD (N=3). The results were analyzed and graphed using the GraphPad Prism 9 software. And the Fluorescence microscopy experiment results were analyzed by ZEN 2.3. The differences between groups were calculated by Student’s t-test to compare two groups or by two-way ANOVA analysis to compare three groups. Zebrafish survival curves were analyzed by Log-Rank test. All of the statistical details of experiments can be found in the figure legends. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; ns, not significant.

Published: March 12, 2024