Presence of two RIG-I-like receptors, MDA5 and LGP2, and their dsRNA binding capacity in a perciform fish, the snakehead Channa argus

Highlights

• Two RLRs, MDA5 and LGP2, were identified in the snakehead (C. argus) with conserved domain structure.

• RIG-I was not found in the fish, nor in other species of fish in the Acanthomorphata, Clupeiformes and Polypteriformes.

• MDA5, LGP2 and MAVS in snakehead were constitutively and inductively expressed.

• MDA5 in snakehead had poly(I:C) binding capacity and can interact with MAVS.

Abstract

Fish retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) are critical RNA sensors in cytoplasm and are involved in antiviral innate immunity. However, some species of fish lack RIG-I gene, and the function of RLR members in RIG-I-absent fish is poorly understood. In the present study, MDA5, LGP2 and MAVS genes were identified in commercially important snakehead Channa argus. But, RIG-I gene was not found in this fish, and a systematic analysis of RLRs in available genome database of fish indicated the absence of RIG-I in the Acanthomorphata, Clupeiformes and Polypteriformes, suggesting that loss events of RIG-I gene may have occurred independently three times in the evolutionary history of fish. The MDA5, LGP2 and MAVS in snakehead have conserved protein domains and genomic location based on sequence, phylogenetic and syntenic analyses. These genes are constitutively expressed in healthy fish and can be induced by polyinosinic and polycytidylic acid (poly(I:C)) stimulation in vitro. It is further revealed that the snakehead MDA5 and LGP2 have binding capacity with dsRNA, such as poly(I:C), and MDA5 can interact with MAVS, implying the antiviral function of MDA5 in the RIG-I-absent fish.

Introduction

Retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) are cytosolic RNA sensors involved in antiviral immunity (Akira et al., 2006; Kawai and Akira, 2006). Mammalian RLR family contains three members, RIG-I (also known as DEAD box polypeptide 58, DDX58), melanoma differentiation-associated gene 5 (MDA5, or interferon induced with helicase C domain 1, IFIH1), and laboratory of genetics and physiology 2 (LGP2, or DExH box polypeptide 58, DHX58), which all belong to DExD/H box RNA helicases (Loo and Gale, 2011; Weber and Weber, 2014). Indeed, a central DExD/H box RNA helicase domain with a C-terminal domain is found in the three RLRs; however, only RIG-I and MDA5 have the N-terminal caspase activation and recruitment domains (CARDs) (Loo and Gale, 2011; Mukherjee et al., 2014).

The three mammal RLR genes are widely expressed in most tissues with relatively low levels, but their expression can be strongly induced by interferons (IFNs) and viruses (Loo and Gale, 2011). In fact, RLR proteins are able to detect a variety of viruses, including members from the Paramyxoviridae, Rhabdoviridae, Orthomyxoviridae and Picornaviridae, as well as the synthetic dsRNA, polyinosinic and polycytidylic acid (poly(I:C)), although the recognition of ligands by RIG-I and MDA5 exhibits certain degrees of difference, such as RNA length and structure (Loo and Gale, 2011; Yoneyama et al., 2015). Activated RLRs, except LGP2, trigger IFN expression through binding to the adaptor molecule, mitochondrial antiviral signalling adaptor (MAVS, also known as IPS-1, VISA, or Cardif), which provides the docking sites for IFN regulatory factors (IRFs) and TANK binding kinase 1 (TBK1) and is associated with phosphorylation of IRFs (Liu and Cao, 2016; Yoneyama et al., 2015). LGP2 is unable to interact directly with MAVS due to the absence of CARD, but is involved in antiviral immunity through the regulation of RIG-I and MDA5 signalling (Liu and Cao, 2016; Loo and Gale, 2011).

To date, RLRs have been identified in a range of fish species, including zebrafish (Danio rerio), grass carp (Ctenopharyngodon idella), rainbow trout (Oncorhynchus mykiss), Japanese flounder (Paralichthys olivaceus), channel catfish (Ictalurus punctatus), large yellow croaker (Larimichthys crocea), Atlantic cod (Gadus morhua) etc. (Chen et al., 2017). Like in mammals, RLRs in fish are predicted to contain similar functional domains, and it has been reported that both fish RIG-I and MDA5 are associated with MAVS to induce type I IFN expression, and thus are involved in antiviral immune response, but LGP2 may show functional duality in having positive or negative regulatory roles in antiviral immunity (Chang and Zhang, 2017; Chen et al., 2017; Lazarte et al., 2019; Li et al., 2017). Surprisingly, RIG-I seems to be lost in some teleost species, such as large yellow croaker, Nile tilapia (Oreochromis niloticus), mandarin fish (Siniperca chuatsi), torafugu (Takifugu rubripes), medaka (Oryzias latipes), and some other species (Chen et al., 2017; Gao et al., 2018; Gu et al., 2019; Shen et al., 2016; Zou et al., 2009). Although the sequence and expression pattern of MDA5 and LGP2 genes have been reported in these ‘RIG-I lack’ species (Gao et al., 2018; Gu et al., 2019; Shen et al., 2016), the ligand recognition of MDA5 and LGP2 and their interaction with MAVS are less investigated (Lazarte et al., 2019).

The snakehead (Channa argus) with commercial importance in aquaculture belongs to the Perciformes and is cultured widely in some Asian countries (Liu and Li, 2012; Liu et al., 2015; Xu et al., 2017). The understanding of its immune system may contribute to the development of immune-related strategies for healthy management of snakehead aquaculture, and may also provide knowledge for understanding the function of MDA5 and LGP2 in perciform fish which lack RIG-I. In this study, two RLR members, MDA5 and LGP2, as well as the adaptor, MAVS, were cloned in C. argus, and in silico analyses were performed to reveal sequence, phylogenetic and syntenic characteristics of MDA5, LGP2 and MAVS genes. The expression profiles of these genes were investigated in vivo and in vitro. The binding capacity of MDA5 and LGP2 was examined against the synthetic dsRNA, polyinosinic and polycytidylic acid (poly(I:C)), and MAVS was found to associate with MDA5 but not LGP2 in signalling. Furthermore, a deep data mining of teleost RLR genes was conducted on available genomic sequences of fish, which should contribute to the understanding of evolutionary composition of RLRs in fish.