H19 influenza A virus exhibits species-specific MHC class II receptor usage

Abstract

Avian influenza A virus (IAV) surveillance in Northern California, USA, revealed unique IAV hemagglutinin (HA) genome sequences in cloacal swabs from lesser scaups. We found two closely related HA sequences in the same duck species in 2010 and 2013. Phylogenetic analyses suggest that both sequences belong to the recently discovered H19 subtype, which thus far has remained uncharacterized. We demonstrate that H19 does not bind the canonical IAV receptor sialic acid (Sia). Instead, H19 binds to the major histocompatibility complex class II (MHC class II), which facilitates viral entry. Unlike the broad MHC class II specificity of H17 and H18 from bat IAV, H19 exhibits a species-specific MHC class II usage that suggests a limited host range and zoonotic potential. Using cell lines overexpressing MHC class II, we rescued recombinant H19 IAV. We solved the H19 crystal structure and identified residues within the putative Sia receptor binding site (RBS) that impede Sia-dependent entry.

Keywords: HA structure; MHC class II; avian influenza virus surveillance; entry receptor; glycan array; hemagglutinin subtype H19; host range; influenza A virus; receptor binding site; recombinant H19 influenza A virus.

Figure 1.. Lesser Scaups in North California harbor IAV of the phylogenetically divergent H19 subtype.

A) Phylogenetic analysis of the A/lesser scaup/California/3087/2010 (*) and A/lesser scaup/California/1742/2013 (**) HA amino acid sequences (purple) and representative sequences from all IAV subtypes was performed by the Maximum Likelihood method. Black triangle represents H19 A/common pochard/Kazakhstan/Kz52/2008. H9 clade members including H8, H9, H12 and H19 are highlighted in gray. Scale bar, estimated amino acid substitutions per site. B) Evolutionary divergence between indicated HA subtypes was calculated from H19 (n=3), H8 (n=122), H9 (n=352) and H12 (n=156) amino acid sequences and plotted as inter-subtype difference. C) Crystal structure of the A/lesser scaup/California/3087/2010 HA trimer (PDB code 8VCC) is shown as a ribbon representation. HA1 and HA2 subunits are highlighted in purple and blue, respectively. N-inked glycans are shown as yellow carbons. D) Crystal structure of the putative RBS of the A/lesser scaup/California/3087/2010 HA is shown. Structural conserved RBS elements 130-loop, 190-helix and 220-loop are indicated. Conserved and non-conserved RBS residues (H3 numbering) among H19 and other HA subtypes in A are highlighted in green and magenta, respectively. E) Surface representation of H19 the A/lesser scaup/California/3087/2010 putative RBS in comparison to H9 A/swine/Hong Kong/9/1998 (PDB code 1jsh, brown), H1 A/California/04/2009 (PBD code 3UBN, cyan) and H18 A/flat-faced bat/Peru/033/20210 (PDB code 4k3x, pink). Conserved and non-conserved RBS residues are highlighted as in D. C-E) Structural representations of HA proteins were generated using PyMOL.

Figure 2.. H19 does not bind the canonical IAV receptor Sia.

A) Hemagglutination assays with recombinant HA proteins (rHA) from A/lesser scaup/California/3087/2010 (rH19 CA/10), A/Hong Kong/1/1968 (rH3 HK/68) and A/Viet Nam/1203/2004 (rH5 VN/04) were performed with turkey or chicken red blood cells (RBC). Serial dilutions were performed starting at 0.25 μg/μl of rHA. B) Vero E6 cells were transfected with indicated expression plasmids. At 24 h post transfection, cells were treated with TPCK-treated trypsin and fusion was induced by incubation at low pH=5. Cells were recovered for 4 h in supplemented DMEM, fixed with 3.7% formaldehyde and subjected to immunohistochemistry. Syncytia formation was visualized by HA staining (magenta). Nuclei were stained with DAPI (blue). Panels on the right show zoom images of the insets on the left. Scale bars, 150 μm. Representative images from n=3 independent experiments are shown. C) Sia binding of rH19 CA/10 was tested on a glycan microarray containing non-sialylated glycans (1–3, gray), α2,3- (4–6, black) or α2,6- (7–9, white) linked Neu5Ac, α2,3- (10–12, orange) or α2,6- (13–15, yellow) linked Neu5Gc, and glycolipids containing α2,3- and α2,8-linked Neu5Ac (A-M, cyan). rH5 VN/04 and rH3 CH/13 served as positive controls. Mean RFU ± s.d. from n=4 technical repeats are shown.

Figure 3.. HLA-DR orthologs from duck, swan, bat, and mouse function as entry receptors for H19.

A) HEK293T cells were transfected with expression plasmids encoding HLA-DR, -DQ, - DP and HLA-DR orthologs from S. scrofa (SLA-DR), G. gallus (B-L), E. fuscus (Ef-DR), P. alecto (Pa-DR), M. lucifugus (Ml-DR), A. fuligula (Af-DR), Mus musculus (H2-E) or the HLA-DQ ortholog in Mus musculus (H2-A), or mCherry as negative control. At 48 h post transfection, cells were inoculated with the indicated BlaM1-VLPs and entry positive cells were quantified by flow cytometry. B) HEK293T cells were transfected with expression plasmids encoding HLA-DR orthologs from A. fuligula (Af-DR), C. atratus (Ca-DR), A. platyrhynchos (Ap-DR), S. hondurensis (Sh-DR), A. jamaicensis (Aj-DR), Z. californianus (Zc-DR), M. p. furo (Mpf-DR). At 48 h post transfections cells were inoculated with the indicated BlaM1-VLPs and entry positive cells were analyzed as in A. C) H2-E surface levels from MDCK cells overexpressing H2-E (LV-H2-E) and control cells (LV-ctrl) were measured by flow cytometry. D) Cells in C were inoculated with the indicated BlaM1-VLPs and entry positive cells were quantified by flow cytometry as in A. E) Cells in C were inoculated with VSV-pseudotypes expressing EGFP and firefly luciferase. At 24 h p.i. cells were fixed; cell nuclei were stained with DAPI and EGFP-positive cells were analyzed by fluorescence microscopy. Scale bars, 300 μm. F) Entry of VSV-pseudotypes in E was determined by luciferase signals. Relative light units (RLU) are plotted. Dotted line indicates the detection limit. G) MDCK LV-H2-E were pre-treated with anti-H2-E or anti-6xHis antibody and inoculated with VSV-H19 in the presence of the indicated antibodies. At 24 h p.i., luciferase signals were measured and normalized to untreated samples. Statistical significance was determined by unpaired, t-test (two-tailed). *P ≤ 0.05, **P ≤ 0.01. H) Vero E6 cells were co-transfected with expression plasmids encoding H19 (A/LS/Cal/10) and HLA-DR, H2-E or with the empty vector (EV) control. At 24 h post transfection, cells were treated with TPCK-treated trypsin and fusion was induced by incubation at low pH=5. Cells were recovered for 4 h in supplemented DMEM, fixed with 3.7% formaldehyde and subjected to immunohistochemistry. Syncytia formation was visualized by staining for H19 (magenta) and HLA-DR or H2-E (green). Nuclei were stained with DAPI (blue). Bottom panels show zoom images of the insets above. Scale bars, 150 μm. I) Cells in C) were incubated with the indicated rHAs for 1 h at 4°C. Binding of rHAs was detected using Alexa-647-labelled anti-6xHis antibody and analyzed by flow cytometry. Histograms (left) and its quantification (right) show rHA binding. A, B, D, F, G, I) Data are means ± s.d. from n≥2 independent experiments. C, E, H, I) Representative histograms or images from n=3 independent experiments are shown.

Figure 4.. Recombinant H19N6 and H19N6stop IAV replicate in MDCK cells expressing the mouse H2-E.

A) Recombinant H19N6 IAV was rescued by transfection of HEK293T and MDCK LV-H2-E co-cultures with rescue plasmids encoding PA, PB1, PB2, HA, NP, M, and NS of A/lesser scaup/California/3087/2010 and the NA of A/duck/Minnesota/104/1974 (H4N6). MDCK LV-H2-E or control cells were inoculated with rescue supernatants and overlaid with agar. At 48 h p.i. plaques were visualized by crystal violet staining. B) Indicated cell lines were inoculated with H19N6 or H1N1 (A/PR/8/34) at an MOI of 0.1 or 0.01, respectively. Virus titers at the indicated times p.i. were determined by plaque assay. Data are means ± s.d. from n=3 independent experiments. C) Rescue plasmid encoding N6 A/duck/Minnesota/104/1974 with premature stop codon was generated. Amino acids (aa) of full length N6 and N6stop are indicated. Recombinant H19N6, H19N6stop and chimeric viruses thereof with the internal genes of A/PR/8/34 were rescued and their NA activities were measured. Values are shown as fold change over background. Data are means ± s.d. from n=2 independent experiments. D) MDCK LV-H2-E cells were inoculated with the indicated viruses at an MOI of 0.1. Virus titers at the indicated times p.i. were determined by plaque assay. Data are means ± s.d. from n=3 independent experiments. Statistical significance was determined by unpaired, t-test (two-tailed). **P ≤ 0.01, ****P ≤ 0.0001.

Figure 5:. Amino acid substitutions at positions 134, 137 and 226 in the H19 putative RBS restore Sia-dependent entry.

A) Sia was modelled into the H19 (PDB code 8VCC) putative RBS by alignment with the Sia complexed H9 structure (PDB code 1jsh) using PyMOL. Conserved (Y98, S136, W153, T155, H183, E190, L194, Y195, G225, G228) and non-conserved (T134, I226) key RBS residues are highlighted in green and magenta, respectively. Sia is shown as carbon structure in yellow. 130-loop residue P137 is highlighted in blue. Potential clashes of Sia with T134 and P137 are indicated with arrows. B) Luciferase-encoding VSV-pseudotypes carrying wild type or mutant H19 (A/LS/Cal/10) and N1 (A/WSN/33) were generated. MDCK LV-ctrl and LV-H2-E cells were inoculated with VSV-pseudotypes carrying the indicated H19 single mutants. At 24 h p.i., entry of VSV-pseudotypes was determined by luciferase signals. Relative light units (RLU) are plotted. Dotted line indicates the detection limit. C) VSV-pseudotypes carrying the indicated H19 double mutants were generated and tested as in B. D) VSV-pseudotypes carrying the indicated H19 triple mutants were generated and tested as in B. E, F) MDCK LV-ctrl cells in E or MDCK LV-H2E in F were treated with bacterial sialidase (NA) prior to inoculation with VSV-pseudotypes carrying the indicated H19 mutations. B-E) Data are means ± s.d. from n=3 independent experiments. Statistical significance was determined by unpaired, t-test (two-tailed). **P ≤ 0.01, ****P ≤ 0.0001.

Figure 6:. Amino acids at positions 40, 43, 44, 61, 62 and 64 in the MHC-II α1-domain confer species-specific H19-entry.

A) HEK293T cells were transfected with expression plasmids encoding the illustrated wild type or chimeric MHC-II. At 48 h post transfection cells were inoculated with indicated BlaM1-VLPs and entry positive cells were quantified by flow cytometry. B) Amino acid sequence alignment of the HLA-DR (human), H2-E (mouse), Af-DR (tufted duck), Ap-DR (mallard), Ca-DR (black swan), Pa-DR (black flying fox) α1-domains is shown. Non-conserved amino acids between HLA-DR (non-functional) and H2-E, Af-DR, Ap-DR, Ca-DR, Pa-DR (functional) are highlighted in red. Patches of surface exposed non-conserved amino acids are indicated. C) Surface representation of HLA-DR (PDB code 3pdo). α- and β-chain are colored white and gray, respectively. Non-conserved amino acids from B) are colored red. Amino acids involved in peptide binding are colored orange. D) HEK293T cells were transfected with expression plasmids encoding HLA-DR, H2-E, or the indicated HLA-DR α1-domain mutants. HLA-DR α1-domain mutants were co-transfected with the wild type β-chain of HLA-DR. At 48 h post transfection, entry of indicated BlaM1-VLPs was determined as in A. E) Surface levels of wild type and mutant HLA-DR in D were determined by flow cytometry. Empty vector transfected cells are shown in black; HLA-DR wild type or mutants are shown in shades of gray form dark to light gray: wild type, N40L/Q43K/S44R, M61I/A62E/K64S, G74A/R75K, T104K/Y109N. F) Entry of indicated BlaM1-VLPs was tested on HEK293T cells expressing HLA-DR, H2-E, or the indicated HLA-DR mutants as in A. Statistical significance was tested by unpaired t-test (two-sided) to compare corresponding entry levels in HLA-DR transfected cells to cells transfected with H2-E or the indicated HLA-DRA mutants. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001. G) Surface levels of wild type and mutant HLA-DR in F were determined by flow cytometry. Empty vector transfected cells are shown in black; HLA-DR wild type or mutants are shown in shades of gray form dark to light gray: wild type, N40L/Q43K/S44RM61I/A62E/K64S, M61I/A62E/K64S/G74A/R75K, N40L/Q43K/S44R/G74A/R75K, N40L/Q43K/S44R/T104K/Y109N, M61I/A62E/K64S/T104K/Y109N, G74A/R75K/T104K/Y109N. H). Entry of indicated BlaM1-VLPswas determined in HEK293T cells transfected with wild type Af-DR or the indicated Af-DRA mutants as in A. A, D, F, H) Data are means ± s.d. from n=3 independent experiments. E, G) Representative histograms from n=3 independent experiments are shown.