Melanoma differentiation-associated gene 5 in zebrafish provoking higher interferon-promoter activity through signalling enhancing of its shorter splicing variant

Melanoma differentiation-associated gene 5 (MDA5) is one of the three members in the retinoic acid-inducible gene I-like receptor (RLR) family, which are cytoplasmic pathogen recognition receptors recognizing intracellular viruses. In the present study, MDA5 and its spliced shorter forms, named as MDA5a and MDA5b, were identified in zebrafish. MDA5a and MDA5b can be up-regulated in cell lines following the infection of a negative ssRNA virus, the spring viraemia of carp virus (SVCV), and an intracellular Gram-negative bacterial pathogen Edwardsiella tarda, implying that the RLR may also be able to sense elements released from bacteria. The over-expression of MDA5a and MDA5b in fish cells resulted in significant induction of type I interferon promoter activity and enabled the protection of transfected cells against SVCV infection. Furthermore, the shorter spliced form, MDA5b when co-transfected with MDA5a or mitochondrial antiviral signalling protein (MAVS), induced a significantly higher level of interferon promoter activity, indicating that MDA5b may function as an enhancer in the interaction between MDA5 and MAVS.


Introduction
The innate immune defence against pathogenic invasion is mediated through the recognition of conserved pathogen-associated molecular patterns (PAMPs) by pattern recognition receptors (PRRs). Toll-like receptors, nucleotide-binding oligomerization domain (NOD)-like receptors and retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) are three major families of PRRs in vertebrates. [1][2][3] The RLR family contains three members, RIG-I, melanoma differentiation-associated gene 5 (MDA5), and laboratory of genetics and physiology 2 (LGP2), which share some common and homologous core structural domains, such as caspase activation and recruitment domain (CARD) at the N terminus (with the exception of LGP2, which has no CARD), DExD/H box helicase domain (DEXDc), helicase C-terminal domain (HELICc), regulatory domain (RD; also referred to as repressor domain or C-terminal domain) at the C terminus. [3][4][5] RIG-I signals through its tandem CARDs as a multimeric complex, and MDA5 may also signal in a homocomplex, while LGP2, through its RD at the C terminus, plays a role in inhibiting RIG-I, but not MDA5, signalling despite being able to form a complex with either RIG-I or MDA5. 6 Both RIG-I and MDA5 initiate the host immune response through interaction of CARDs between each of the PRRs and mitochondrial antiviral signalling protein (MAVS; also known as Cardif, IPS-1 or VISA), 7,8 leading to the phosphorylation of downstream signalling molecules, interferon regulatory factors 3 and 7, and to the subsequent translocation of the molecules into nucleus for inducing the expression of interferon-a/b (IFN-a/b) and IFN-stimulated genes. 1,7,9 Among the RLR family, RIG-I was initially reported to sense viral dsRNA, 10  recognition of viral dsRNA. 5,11,12 The recognition of dsRNA by RIG-I or MDA5 is related to different features of ligands, such as 5′-triphosphate constituents, 13 and the length of dsRNA molecules. 14 In mammals, RIG-I and MDA5 may both functionally sense a diverse range of negative-stranded or positive-stranded RNA viruses. [15][16][17] Recently, it was reported that RIG-I and MDA5 may also sense bacterial RNA or DNA released from the pathogenic bacterium Listeria monocytogenes to trigger the production of IFN-b and interleukin-1b. 18 In teleost fish, the homologue of human MDA5 has been characterized functionally only in rainbow trout, Oncorhynchus mykiss, which was responsive to the infection of RNA viruses, such as viral haemorrhagic septicaemia virus and salmon alphavirus. 19 In the present study, the MDA5, and its shorter splicing variants, named as MDA5a and MDA5b, were identified in zebrafish, Danio rerio. The expression of the two variants was then examined in cell lines following infection with spring viraemia of carp virus (SVCV), a negative ssRNA virus as well as Edwardsiella tarda, an intracellular Gram-negative pathogenic bacterium. The role of MDA5a and MDA5b in inducing IFN promoter, and in inducing an antiviral state was also examined in a fish cell line. The possible interaction between the two variants, MDA5a and MDA5b, and between the variants and MAVS, the adaptor molecule triggered by MDA5 and RIG-I, was also examined in cell lines. The present study therefore provides the first report on the existence of MDA5 spliced variants and their functions in fish and probably in any other vertebrates.
Transfection of plasmids in ZF4 cells was performed using the Amaxa Nucleofector II transfection system (Lonza, Cologne, Germany) under Program T20, whereas in EPC cells and HEK 293T cells Lipofectamine 2000 (Invitrogen) was used. All of the procedures were performed according to the manufacturers' instructions. All experiments involving cell lines and animals complied with the local and institutional guidelines for the experimental use of animals.

Virus and bacterium used for infection
The negative ssRNA virus SVCV was propagated in EPC cells at 25°, with titres determined by plaque assay under methylcellulose overlay (0Á5% in DMEM).
Edwardsiella tarda PPD130/91, an intracellular Gramnegative bacterial pathogen, 22 was grown in tryptic soy broth (BD Biosciences, Sparks, MD) liquid culture overnight at 28°. Quantification of the bacterial culture was performed by spectrophotometry and plating dilution on tryptic soy agar (BD Biosciences). SVCV and E. tarda were both confirmed as infective in zebrafish and ZF4 cells. 22,23 .
RNA extraction, molecular cloning and plasmid construction Total RNA was extracted using Trizol Reagent (Gibco) from organ/tissue samples, including kidney, spleen and intestine of zebrafish adults stimulated for 24 hr with the injection of 100 ng poly I:C, and the RNA was treated with RNase-free DNase I (Fermentas Life Sciences, Vilnius, Lithuania) and RiboLock TM RNase inhibitor (Fermentas Life Sciences) according to the manufacturer's protocol. The DNase I-treated RNA was used for the synthesis of first-strand cDNA for 5′ RACE and 3′ RACE using GeneRacer TM Kit (Invitrogen). Based on the partial sequence of zebrafish MDA5 (XM_689032.3), nested primers zfMDA5-5-R1, zfMDA5-5-R2, zfMDA5-3-F1 and zfMDA5-3-F2 (Table 1) were designed and used in 5′ RACE or 3′ RACE using a GeneRacer TM Kit (Invitrogen) according to the manufacturer's instructions. For 5′ RACE, the annealing temperature of the first round and nested PCR was 56°, whereas it was 60°and 58°, respectively, for 3′ RACE. The PCR products were cloned into the pMD18-T vector (Takara, Dalian, China) for sequencing by Sangon Biotech Co., Ltd (Shanghai, China).
For eukaryotic expression, the open reading frame (ORF) of zebrafish MDA5a and MDA5b was cloned and inserted into the pcDNA3.1/myc-His(-)A vector (Invitrogen). For subcellular localization, the entire ORF of MDA5a and MDA5b was also inserted into the pTur-boGFP-N vector (Evrogen, Moscow, Russia). The ORF of zebrafish MAVS was sub-cloned into pcDNA3.1/myc-His (-)A vector (Invitrogen) and p3xFLAG-CMV TM -14 expression vector (Sigma Aldrich, St Louis, MO), respectively. All plasmid constructs were verified by sequencing analysis. The primers with the restricted enzyme cutting sites are listed in Table 1.

Sequence analysis
Protein prediction was performed using software at the ExPASy Bioinformatics Resource Portal (http://web.expasy.org/proteomics). Analysis of conserved domains was performed using NCBI CDD. 24 The zebrafish MDA5 cDNA sequences obtained were searched against zebrafish genome (Zv9) for further identification. The intron/exon structures of genomic sequences were determined by alignment of the full-length cDNA sequence to genomic sequence using BLAST.

Quantitative real-time PCR
For SVCV or E. tarda infection, ZF4 cells were passaged into six-well plates for 24 hr at a concentration of 5 9 10 6 cells per well. Cells were then infected with SVCV or E. tarda at a multiplicity of infection (MOI) of 5 or 10, respectively, and 6 and 24 hr post infection (hpi) cells were collected for RNA extraction as described above. The first-strand cDNAs were synthesized by Superscript reverse transcriptase (Fermentas Life Sciences) and oligo(dT) primer using RNase-free DNase I-treated RNA. Quantitative real-time PCR was performed using iQ TM SYBR Green Supermix (Bio-Rad, Singapore) on a Bio-Rad CFX96 Real-Time System under the following conditions: 3 min at 95°, followed by 40 cycles of 15 seconds at 94°, 15 seconds at 58°and 30 seconds at 72°. All reactions were performed in triplicate in a 96-well plate, with the mean value recorded. The relative expression level of target genes was normalized to the expression of glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Fold changes were calculated by comparison with the corresponding controls using the comparative Ct method (2 ÀDDCt ). 25 Three experiments were conducted for statistical analysis using paired Student's t-test, with the probability level of P < 0Á05 set as statistical significance. All primers used for quantitative real-time PCR are listed in Table 1, with one pair of primers named as qGAPDH-F/ R from a previous report. 26 Fluorescence microscopy and luciferase activity assay For subcellular localization of MDA5a and MDA5b, ZF4 cells were cultured in fresh flasks for 24 hr, and 5 lg of each of the following constructs, pTurbo-MDA5a-GFP, pTurbo-MDA5b-GFP and pTurboGFP-N (vector control), were transfected into 5 9 10 6 cells using Amaxa Nucleofector II transfection system (Lonza) under Pro- GeneRacer TM 5′ Primer, GeneRacer TM 5′ Nested Primer, GeneRacer TM 3′ Primer and GeneRacer TM 3′ Nested Primer are from the GeneRacer TM Kit (Invitrogen); and qGAPDH-F and qGAPDH-R were reported in a previous study. 26 gram T20. Cells were then transferred into 25-cm 2 fresh flasks for 24 hr of culture at 28°, and transfected cells were stained with Hoechst 33342 (Sigma-Aldrich) before being examined under a fluorescence microscope (Zeiss, Oberkochen, Germany). Subsequently, cells were collected to confirm MDA5a-GFP and MDA5b-GFP fusion proteins by Western blotting, with Anti-TurboGFP antibody (Evrogen, CAT. #AB513) diluted at 1 : 5000 as the primary antibody. The bands were detected by using Immobilon TM Western Chemiluminescent HRP Substrate (Millipore Corporation, Billerica, MA) and ECL Western blot system (LAS-4000 mini; Fujifilm, Tokyo, Japan) according to the manufacturers' instructions. For luciferase activity assay, EPC cells seeded overnight in 24-well plates at 2 9 10 5 cells per well were transiently transfected with 100 ng IFNpro-Luc, a zebrafish type I IFN promoter-driving luciferase plasmid 27 and 10 ng pRL-TK (Promega, Madison, WI), together with pcDNA3.1-MDA5a, pcDNA3.1-MDA5b or pcDNA3.1 (vector control) at various concentrations (10, 50, 100 or 200 ng) using Lipofectamine 2000 (Invitrogen). Twenty-four hours post-transfection, cells were harvested and lysed using Dual-Luciferase Reporter System (Promega), with luciferase activity measured on a Junior LB9509 luminometer (Berthold, Pforzheim, Germany). Data were normalized to the Renilla Luciferase activity, and expressed as the fold change relative to control group transfected with empty vector pcDNA3.1. Cell lysates were examined by Western blotting analysis.
To determine the possible association between MDA5a, MDA5b and MAVS, EPC cells seeded overnight in 24well plates were transiently transfected with 100 ng each of the plasmids pcDNA3.1-MDA5a, pcDNA3.1-MDA5b, or pcDNA3.1-MAVS, or with combinations of any two, together with luciferase reporter plasmids (100 ng IFNpro-Luc and 10 ng pRL-TK). The pcDNA3.1 empty vector was used to balance the total volume of transfected plasmids. Twenty-four hours post-transfection, cells were harvested and luciferase activity was measured as described above.

Antiviral effect
EPC cells seeded in 24-well plates at a concentration of 1 9 10 6 cells per well were transfected with 1 lg pcDNA3.1-MDA5a, pcDNA3.1-MDA5b or pcDNA3.1 empty vector (control), respectively. Twenty-four hours post-transfection, transfected cells were washed and infected with SVCV at an MOI of 1, 0Á1 or 0Á01, respectively, and 48 hpi, the culture supernatants were collected for the determination of virus titres by standard plaque assay. Cell monolayers were then fixed in 10% paraformaldehyde for 1 hr, before being stained with 0Á5% crystal violet for the observation of cytopathic effect.

Antibodies, co-immunoprecipitation and Western blotting
Rabbit polyclonal antibody against zebrafish MDA5 was generated by a commercial supplier (GenScript, Nanjing, China) using standard procedures with synthetic peptide, VRNRPPEPDEEAEHC of the MDA5. Mouse monoclonal antibody against Flag was purchased from Sigma.
For co-immunoprecipitation experiments, HEK 293T cells seeded in six-well plates were co-transfected with p3xFLAG-MAVS and pcDNA3.1-MDA5a, or with p3xFLAG-MAVS and pcDNA3.1-MDA5b, and with p3xFLAG-MAVS and pcDNA3.1 as control. Twenty-four hours post-transfection, cells were collected and lysed in 500 ll extraction buffer as described above. Cellular debris was removed by centrifugation at 12000 g for 10 min at 4°, and supernatants were transferred to fresh tubes and incubated overnight at 4°with anti-Flag resin conjugate (FLAG Tagged Protein Immunoprecipitation Kit; Sigma). The resin was washed six times with ice-cold wash solution as described by the manufacturer, and eluted with 40 ll 2 9 SDS sample buffer by boiling for 10 min at 95°. The precipitates were detected by Western blotting with anti-MDA5 and anti-Flag antibodies.
For Western blotting, equal amounts of protein samples were separated on 8% or 12% SDS-PAGE gels and then electrophoretically transferred to a polyvinylidene difluoride membrane (Millipore Corporation). Afterwards, the membrane was blocked with PBS containing 5% nonfat dry milk for 1 hr at room temperature, before three washes, each for 5 min, in PBS containing 0Á1% Tween 20 (PBST). The membrane was incubated with primary antibody in PBS containing 2% milk for 2 hr at room temperature, and 1 hr incubation with secondary antibody also at room temperature, with three washes in PBST following each incubation. The membrane was stained with Immobilon TM Western Chemiluminescent HRP Substrate (Millipore Corporation) and examined under the ECL Western blot system LAS-4000 mini; Fujifilm).

Sequence analysis of MDA5 transcripts in zebrafish
Using primers to amplify the entire ORF of MDA5 in zebrafish, two transcripts named as MDA5a (GenBank accession no. JX462556) and MDA5b (JX462557), were identified, which were 3 398, and 3 027 bp encoding 997 and 685 amino acids, respectively. Using NCBI CDD (http://www.ncbi.nlm.nih.gov/Structure/cdd/docs/ cdd_search.html), it was found that MDA5a and MDA5b contained conserved domains, such as two CARDs, and a DEXDc domain, with the former having an HELICc  domain and an RD domain that were absent in MDA5b (Fig. 1a).
Searching the zebrafish genome, MDA5a and MDA5b were found to be encoded by a single gene, which was located on zebrafish chromosome 9. The MDA5a gene consists of 16 exons, with 15 intervening introns, spanning approximately 69 kb of genomic sequence, while MDA5b lacks partial exons 11 and 13 and the entire exon 12, so causing a premature stop in exon 11 with the transcription of a predicted 685 amoino acids, being a shorter form of MDA5 protein, named as MDA5b (Fig. 1b).

Localization of MDA5a and MDA5b in ZF4 cells
To determine the subcellular localization of MDA5a and MDA5b, their ORFs were inserted into the pTurboGFP-N expression vector for transient transfection in ZF4 cells. Fluorescent microscopy revealed that both pTurbo-MDA5a-GFP and pTurbo-MDA5b-GFP fusion proteins had an entire cytosolic distribution except in the nucleus, whereas control cells transfected with pTurboGFP-N empty vector showed a cytosolic as well as nucleic distribution (Fig. 2a). The presence of pTurbo-MDA5a-GFP and pTurbo-MDA5b-GFP in the cytosol was also confirmed using a polyclonal antibody against the GFP protein by Western blotting analysis (Fig. 2b).

Expression of MDA5a, MDA5b, MAVS and Mx in ZF4 cells infected with SVCV and E. tarda
MDA5a and MDA5b were both significantly up-regulated in ZF4 cells in response to SVCV or to E. tarda infection in the 24-hr experimental period, with only one exception that MDA5b had a statistically non-significant increase at 6 hr following SVCV infection (Fig. 3). MDA5a had in general a higher level of expression, when compared with MDA5b, and its significant increase following SVCV or E. tarda infection was respectively above threefold and   fourfold at 6 hpi, and all were about sevenfold at 24 hpi (Fig. 3). MDA5b increased about 1Á5-fold at 6 hr following SVCV or E. tarda infection, and accordingly about threefold and twofold at 24 hpi (Fig. 3). MAVS and Mx (myxovirus resistance A, GenBank accession no. NM_182942.4) were also up-regulated at mRNA level in ZF4 cells following SVCV or E. tarda infection (Fig. 3). MAVS was significantly up-regulated only at the early stage of infection, i.e. at 6 hpi, with the expression being above threefolds higher following SVCV or E. tarda infection; but its increased expression was not significant at 24 hpi. In addition, significantly increased expression of the antiviral protein Mx, which is known to be up-regulated in virus infection, was also observed in ZF4 cells following either SVCV or E. tarda infection, with a rather higher expression level (about 14-fold) observed at 6 hr following E. tarda infection (Fig. 3).

Luciferase assay on the activation of IFN promoter by MDA5a and MDA5b
Subsequently, luciferase assay was used to determine whether MDA5a and MDA5b induced the activation of IFN, with the constructed expression plasmids, pcDNA3.1-MDA5a and pcDNA3.1-MDA5b co-transfected with IFN reporter plasmid in EPC cells. The over-expression of MDA5a provoked a strong activation of IFN promoter, by up to 42-fold against the control when the amount of each plasmid was 10 ng, and up to 106-fold when the amount was 100 ng. MDA5b also induced the activation of IFN promoter, although the induced level was not as high as that for MDA5a, with fivefold when the plasmid amount was 10 ng and 34-fold for 100 ng (Fig. 4a). Additionally, the protein of the transfected EPC cells was also examined by Western blotting analysis using polyclonal antibody against zebrafish MDA5, which confirmed the successful over-expression of MDA5a and MDA5b in EPC cells after transfection (Fig. 4b).

Interaction of MDA5 variants and between the variants and MAVS in IFN signalling
To reveal possible interactive roles between MDA5a and MDA5b in IFN signalling, plasmids expressing separately MDA5a or MDA5b were co-transfected into EPC cells with the IFN reporter plasmid. The luciferase assay showed that the co-transfection of MDA5a with MDA5b induced significantly higher levels of IFN promoter activ-  ity, when compared with separate over-expression of MDA5a or MDA5b (Fig. 6a). Furthermore, the co-expression of MDA5b with MAVS induced a significantly higher level of IFN activation in comparison with the only transfection of MDA5b, while co-expression of MDA5a with MAVS showed no significant impact on the IFN promoter activity when compared with the activation level in the transfection of MDA5a alone (Fig. 6b). These results suggested that the splicing variant MDA5b may act as a positive regulator of MDA5a-mediated IFN activation.
Finally, the interaction of each variant, MDA5a and MDA5b, with MAVS was determined by co-immunoprecipitation with Tag or specific antibodies. In HEK 293T cells co-transfected with p3xFLAG-MAVS and pcDNA3.1-MDA5a, anti-Flag antibody-immunoprecipitated protein complex was recognized by anti-MDA5 antibody (Fig. 7). Moreover, in cells co-transfected with p3xFLAG-MAVS and pcDNA3.1-MDA5b, co-immunoprecipitation and Western blotting analysis also showed that the MDA5b interacted with MAVS (Fig. 7). These data suggested that both MDA5a and MDA5b could form a complex with MAVS during the signal pathway.

Discussion
In this study, MDA5 and its shorter splicing form have been identified for the first time in a teleost fish, the zebrafish, representing also the first report of MDA5 variants in vertebrates. MDA5a and MDA5b can all activate the expression of type I IFN, as demonstrated using zebrafish type I promoter-driving luciferase plasmid in a fish cell line, and when transfected, they can all protect cells against the infection of negative ssRNA virus, SVCV. Intriguingly, it is found that the MDA5a and MDA5b can all interact with their adaptor molecule, MAVS, and that the shorter form, MDA5b, functions also as an enhancer of the MDA5 signalling in the activation of IFN expression.
Research in mammals has shown that MDA5 is able to immunologically recognize positive-stranded ssRNA viruses, such as picornavirus, West Nile virus and Dengue virus, and dsRNA viruses, such as reoviruses and synthetic analogue of dsRNA virus, poly I:C. 12,15,[28][29][30] The only research on MDA5 in other classes of vertebrates was carried out in rainbow trout, and it is revealed that the expression of MDA5 can be significantly induced in cell lines following the infection of negative-stranded, or positive-stranded, ssRNA viruses, such as viral haemorrhagic septicaemia virus and salmon alphavirus, respectively, and that its over-expression can protect cells against viral haemorrhagic septicaemia virus infection. 19 In the present study, MDA5 and its shorter splicing form, MDA5b, can be significantly induced at mRNA level in cell lines following the infection of negative-stranded ssRNA virus, SVCV, and can also protect cells against SVCV infection. Taken together, it is likely that MDA5 may be able to immunologically recognize RNA viruses, ranging from ssRNA and dsRNA viruses to negative-and positive-stranded RNA viruses. Additionally, Luthra et al. 31 reported that mammalian MDA5 sense viral mRNA from parainfluenza 5 virus in association with RNase L, which cleaves viral mRNA into small RNA, and such RNA products were suggested to have cleavage marker or processing signature and could be recognized by MDA5. Pichlmair et al. 32 also suggested that higher-order RNA structures of dsRNA and ssRNA contributed to the induction of MDA5-mediated IFN activation. Hence, the recognition of extensive viral RNA by MDA5 requires some conserved structures or signals that distinguish viral from host cellular RNA. However, exact features of the  In addition to viral RNA recognition for the production of type I IFN, 15,33 cytosolic RLRs, RIG-I, MDA5 and STING were found to be able to recognize nucleic acids RNA/DNA secreted from infected live bacteria, L. monocytogenes with the subsequent triggering of IFN-b production. 18 Listeria monocytogenes is an intracellular Gram-positive pathogen, and the secreted RNA/DNA may be mediated through SecA2, which is one of the two secretion pathways present in some Grampositive pathogenic bacteria, and contributes to the secretion of specific virulence factors, 34 as the mutant bacteria lacking SecA2 elicited less RIG-I and MDA5-dependent IFN-b production, although the exact feature of the RNA/DNA is unknown. 18 The E. tarda, used in the present study is also intracellular, but Gram-negative, and possesses secretion systems, with types III and VI secretion systems as the main tools for the secretion of virulence factors; 35,36 but it is not yet known that nucleic acids, or small RNA/DNA, are secreted through the systems, which may be of significance for further study. However, it can be at least tentatively concluded that the cytosolic RLR, MDA5, in mammals and in teleost fish can recognize elements from intracellular Gram-positive or Gram-negative bacteria.
More interestingly, a spliced shorter form, MDA5b, was found in zebrafish in addition to the presence of the mammalian homologue, MDA5. It is revealed that MDA5b resulted from the alternative splicing of the same gene as MDA5, and its expression can also be up-regulated by SVCV and E. tarda infection; but MDA5b lacks the HELICc and RD, which are present in MDA5. It has been demonstrated that RD in RIG-I and LGP2 plays a role in keeping the molecules in an inactive conformation, and upon ligand binding, the RD unmasks CARDs for signalling activity. 6 However, such a role was not found in MDA5, and it the exact role of RD in MDA5 is unknown. 6,37 In comparison of the structural domains of MDA5a and MDA5b, it is suggested that the N-terminal CARDs and DEXDc domains of MDA5 are sufficient to  induce an antiviral response, which is consistent with the finding in RIG-I that the over-expression of the N-terminal CARDs was only able to protect cells against virus infection, 38 although a lower level of IFN induction was observed for MDA5b. On the other hand, when cotransfected with MAVS or MDA5a, MDA5b induced much more significantly the level of IFN promoter activity. It is therefore hypothesized that in addition to its IFN-inducing role, the splicing shorter variant, MDA5b, which lacks HELICc and RD, may function as a positive regulator in MDA5-mediated induction of type I IFN production, being conserved in sensing PAMPs but diversified in exemplifying the IFN production and immune response. Alternative splicing has been reported as a major source for producing functional diversity of proteins, 39 and several cases have been reported for other intracellular sensors of PAMPs, or even RLRs. In mammals, a short isoform of NOD2 was identified, which served as an endogenous inhibitor of NOD2 signalling pathways. 40,41 Three splicing variants of MAVS were reported in mammals and showed different roles in RIG-I/MAVS signalling, in which one splicing variant MAVS 1a functions as an inhibitor of RIG-I/MAVS-mediated nuclear factor-jB and IFN-b activation, whereas the other, MAVS 1b, activates IFN-b promoter but not nuclear factor-jB pathway. 42 In teleost fish, LGP2 was found to have a variant, named as LGP2b, which played an inhibitory role in LGP2-mediated antiviral signalling. 19 However, the exact mechanism involved in the amplification of IFN promoter activity by the co-transfection of MDA5b with MDA5a or MAVS requires further investigation.
In conclusion, MDA5 and its shorter splicing variant, named as MDA5a and MDA5b, respectively, were identified in zebrafish. Both MDA5a and MDA5b can be induced at mRNA level in cells infected with ssRNA virus, SVCV and intracellular Gram-negative bacteria E. tarda, and the RLRs can also trigger type I IFN promoter activity and an antiviral state in cell lines. Interestingly, MDA5b when co-transfected with MAVS or MDA5a, induced a significantly higher level of IFN promoter activity, although it exhibited a relatively lower level of effect on the promoter activity. It is therefore considered that MDA5b may function as an enhancer in the interaction between MDA5 and MAVS, but the interactive mechanism remains to be further investigated.