N-linked glycosylation sites in G protein of infectious hematopoietic necrosis virus (IHNV) affect its virulence and immunogenicity in rainbow trout

Infectious hematopoietic necrosis virus (IHNV) causes infectious hematopoietic necrosis in salmonid fish, resulting in substantial economic losses to the aquaculture industry worldwide. The G protein, which harbors the major antigenic determinants of IHNV, is an envelope glycoprotein that plays an important role in both pathogenicity and immunogenicity of IHNV. Previous studies have demonstrated that changes to viral glycosylation sites may affect replication and immunogenicity, but little is known about the specific contributions of G protein glycosylation to IHNV replication and pathogenicity. In this study, we predicted four N-linked glycosylation sites at position 56, 379, 401, and 438 Asp (N) in G protein, and using a reverse genetics system developed in our laboratory, constructed nine recombinant viruses with single, triple, or quadruple glycosylation site disruptions using alanine substitutions in the following combinations: rIHNV-N56A, rIHNV-N379A, rIHNV-N401A, rIHNV-N438A, rIHNV-N56A-N379A-N401A, rIHNV-N56A-N379A-N438A, rIHNV-N56A-N401A-N438A, rIHNV-N379A-N401A-N438A, and rIHNV-N56A-N379A-N401A-N438A. Our results confirmed that all four asparagines are sites of N-linked glycosylation, and Western blot confirmed that mutation of each predicted N-glycosylation sited impaired glycosylation. Among the nine recombinant IHNVs, replication levels decreased significantly in vitro and in vivo in the triple and quadruple mutants that combined mutation of asparagines 401 and 438, indicating the importance of glycosylation at these sites for efficient replication. Moreover, juvenile rainbow trout mortality after challenge by each of the nine mutants showed that, while eight mutants suffered almost 100% cumulative mortality over 30 days, the mutant with a single alanine substitution at position 438 resulted in cumulative mortality of less than 50% over 30 days. This mutant also elicited specific anti-IHNV IgM production earlier than other mutants, suggesting that glycosylation of asparagine 438 may be important for viral immune escape. In conclusion, our study reveals the effect of G protein glycosylation on the pathogenicity and immunogenicity of IHNV and provides a foundation for developing a live-attenuated vaccine.