MDA5 cleavage by the Leader protease of foot-and-mouth disease virus reveals its pleiotropic effect against the host antiviral response.

The RIG-I-like receptor (RLR) melanoma differentiation-associated gene 5 (MDA5) plays a key role in triggering innate antiviral response during infection by RNA viruses. MDA5 activation leads to transcription induction of type-I interferon (IFN) and proinflammatory cytokines. MDA5 has also been associated with autoimmune and autoinflammatory diseases by dysfunctional activation of innate immune response in the absence of infection. Here, we show how foot-and-mouth disease virus (FMDV) counteracts the specific antiviral effect exerted by MDA5 targeting the protein for cleavage by the viral Leader protease (Lpro). MDA5 overexpression had an inhibitory effect on FMDV infection in IFN-competent cells. Remarkably, immunostimulatory viral RNA co-immunoprecipitated with MDA5 in infected cells. Moreover, specific cleavage of MDA5 by Lpro was detected in co-transfected cells, as well as during the course of FMDV infection. A significant reduction in IFN induction associated with MDA5 cleavage was detected by comparison with a non-cleavable MDA5 mutant protein with preserved antiviral activity. The Lpro cleavage site in MDA5 was identified as the RGRAR sequence in the conserved helicase motif VI, coinciding with that recently reported for Lpro in LGP2, another member of the RLRs family involved in antiviral defenses. Interestingly, specific mutations within the MDA5 Lpro target sequence have been associated with immune disease in mice and humans. Our results reveal a pleiotropic strategy for immune evasion based on a viral protease targeting phylogenetically conserved domains of immune sensors. Identification of viral strategies aimed to disrupt MDA5 functionality may also contribute to develop new treatment tools for MDA5-related disorders.

RNA immunization can protect mice against foot-and-mouth disease virus.

In previous work we have reported the immunization of swine using in vitro-transcribed foot-and-mouth disease virus (FMDV) RNA. With the aim of testing whether RNA-induced immunization can mediate protection against viral infection, a group of Swiss adult mice was inoculated with FMDV infectious transcripts. In most inoculated animals viral RNA was detected in serum at 48-72h postinoculation. A group of the RNA-inoculated mice (11 out of 19) developed significant titers of neutralizing antibodies against FMDV. Among those animals that were successfully challenged with infectious virus (15 out of 19), three out of the eight animals immunized upon RNA inoculation were protected, as infectious virus could not be isolated from sera but specific anti-FMDV antibodies could be readily detected. These results suggest the potential of the inoculation of genetically engineered FMDV RNA for virulence and protection assays in the murine model and allow to explore the suitability of RNA-based FMDV vaccination in natural host animals.

Tolerance to mutations in the foot-and-mouth disease virus integrin-binding RGD region is different in cultured cells and in vivo and depends on the capsid sequence context.

Engineered RNAs carrying substitutions in the integrin receptor-binding Arg-Gly-Asp (RGD) region of foot-and-mouth disease virus (FMDV) were constructed (aa 141-147 of VP1 capsid protein) and their infectivity was assayed in cultured cells and suckling mice. The effect of these changes was studied in the capsid proteins of two FMDVs, C-S8c1, which enters cells through integrins, and 213hs(-), a derivative highly adapted to cell culture whose ability to infect cells using the glycosaminoglycan heparan sulfate (HS) as receptor, acquired by multiple passage on BHK-21 cells, has been abolished. The capsid sequence context determined infectivity in cultured cells and directed the selection of additional replacements in structural proteins. Interestingly, a viral population derived from a C-S8c1/L144A mutant, carrying only three substitutions in the capsid, was able to expand tropism to wild-type (wt) and mutant (mt) glycosaminoglycan-deficient CHO cells. In contrast, the 213hs(-) capsid tolerated all substitutions analysed with no additional mutations, and the viruses recovered maintained the ability of the 213hs(-) parental virus to infect wt and mt CHO cells. Viruses derived from C-S8c1 with atypical RGD regions were virulent and transmissible for mice with no other changes in the capsid. Substitution of Asp143 for Ala in the C-S8c1 capsid eliminated infectivity in cultured cells and mice. Co-inoculation with a neutralizing monoclonal antibody directed against the type C FMDV RGD region abolished infectivity of C-S8c1 virus on suckling mice, suggesting that FMDV can infect mice using integrins. Sequence requirements imposed for viral entry in vitro and in vivo are discussed.

The 3' end of the foot-and-mouth disease virus genome establishes two distinct long-range RNA-RNA interactions with the 5' end region.

The untranslated regions (UTRs) of the foot-and-mouth disease virus (FMDV) genome contain multiple functional elements. In the 5' UTR, the internal ribosome entry site (IRES) element governs cap-independent translation initiation, whereas the S region is presumably involved in RNA replication. The 3' UTR, composed of two stem-loops and a poly(A) tract, is required for viral infectivity and stimulates IRES activity. Here, it was found that the 3' end established two distinct strand-specific, long-range RNA-RNA interactions, one with the S region and another with the IRES element. These interactions were not observed with the 3' UTR of a different picornavirus. Several results indicated that different 3' UTR motifs participated in IRES or S region interactions. Firstly, a high-order structure adopted by both the entire IRES and the 3' UTR was essential for RNA interaction. In contrast, the S region interacted with each of the stem-loops. Secondly, S-3' UTR interaction but not IRES-3' UTR interaction was dependent on a poly(A)-dependent conformation. However, no other complexes were observed in mixtures containing the three transcripts, suggesting that these regions did not interact simultaneously with the 3' UTR probe. Cellular proteins have been found to bind the S region and one of these also binds to the 3' UTR in a competitive manner. Our data suggest that 5'-3'-end bridging through both direct RNA-RNA contacts and RNA-protein interactions may play an essential role in the FMDV replication cycle.