Topological N-glycosylation and site-specific N-glycan sulfation of influenza proteins in the highly expressed H1N1 candidate vaccines

Yi-Min She,Terry D Cyr,Aaron Farnsworth,Xuguang Li

doi:10.1038/s41598-017-10714-2

Abstract

The outbreak of a pandemic influenza H1N1 in 2009 required the rapid generation of high-yielding vaccines against the A/California/7/2009 virus, which were achieved by either addition or deletion of a glycosylation site in the influenza proteins hemagglutinin and neuraminidase. In this report, we have systematically evaluated the glycan composition, structural distribution and topology of glycosylation for two high-yield candidate reassortant vaccines (NIBRG-121xp and NYMC-X181A) by combining various enzymatic digestions with high performance liquid chromatography and multiple-stage mass spectrometry. Proteomic data analyses of the full-length protein sequences determined 9 N-glycosylation sites of hemagglutinin, and defined 6 N-glycosylation sites and the glycan structures of low abundance neuraminidase, which were occupied by high-mannose, hybrid and complex-type N-glycans. A total of ~300 glycopeptides were analyzed and manually validated by tandem mass spectrometry. The specific N-glycan structure and topological location of these N-glycans are highly correlated to the spatial protein structure and the residential ligand binding. Interestingly, sulfation, fucosylation and bisecting N-acetylglucosamine of N-glycans were also reliably identified at the specific glycosylation sites of the two influenza proteins that may serve a crucial role in regulating the protein structure and increasing the protein abundance of the influenza virus reassortants.

Highlights

Influenza viruses are among the most virulent pathogens, causing severe respiratory illness and human pandemic deaths worldwide
To thoroughly examine the glycosylation states and the structural consequences caused by mutations of HA and NA in the high-yielding pandemic H1N1 candidate reassortant vaccines, we utilized high-resolution LTQ-FT and Orbitrap mass spectrometry-based proteomics approaches to identify the N-glycan structure of intact glycopeptides of the protein digests by low-energy collision-induced dissociation (CID) followed by multi-stage tandem mass spectrometry (MS3)
Sequence alignment of influenza proteins derived from two candidate vaccines of NIBRG-121xp and New York Medical College (NYMC)-X181 revealed the amino acid identities of 99.12% in HA and 99.79% in NA (Supplementary Information Figs S1 and S2)

Summary

Introduction

Influenza viruses are among the most virulent pathogens, causing severe respiratory illness and human pandemic deaths worldwide. Most interesting of these are the complex glycans of influenza glycoproteins usually featuring a large diversity of bi-, tri- and tetra-antennary structures with additional modifications of fucosylated, sialylated, sulfated groups on the carbohydrate side chains. To thoroughly examine the glycosylation states and the structural consequences caused by mutations of HA and NA in the high-yielding pandemic H1N1 candidate reassortant vaccines, we utilized high-resolution LTQ-FT and Orbitrap mass spectrometry-based proteomics approaches to identify the N-glycan structure of intact glycopeptides of the protein digests by low-energy collision-induced dissociation (CID) followed by multi-stage tandem mass spectrometry (MS3). We quantitatively analyzed the candidate vaccines of NIBRG-121xp and NYMC-X179A both derived from the influenza virus strain A/California/7/2009 by different manufacturers, and evaluated the antigenic stability of the HA and NA proteins[3, 5]. Despite the considerable progress in the determination of 3D crystal structures and post-translational modifications of influenza proteins[20,21,22,23,24,25,26,27,28,29,30], much remains to be understood about the structure-function relationship between the specific compositions of N-glycans and the biosynthesis of influenza viruses

Methods

Results

Conclusion