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

SUMMARY

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 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-II), which facilitates viral entry. Unlike the broad MHC-II specificity of H17 and H18 from bat IAV, H19 exhibits a species-specific MHC-II usage that suggests a limited host range and zoonotic potential. Using cell lines overexpressing MHC-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.

Graphical Abstract

eTOC Blurb

Karakus et al. uncovers H19 influenza A virus genomes in North Californian ducks. Instead of using the conventional influenza virus receptor sialic acid, H19 uses MHC class II proteins to enter host cells. MHC class II receptor usage of H19 is restricted to certain species suggesting a limited zoonotic potential.

INTRODUCTION

Influenza A virus (IAV) is a zoonotic pathogen with a broad host range including diverse mammalian and avian species. The natural IAV reservoir resides in waterbirds and shorebirds, from where IAV can occasionally spill over into marine mammals, domesticated animals such as pigs and poultry, or humans ¹. Such zoonotic events can introduce IAV subtypes into the human population to which humans lack pre-existing immunity and carry the potential to cause devastating pandemics ². Based on genetic and antigenic differences of the surface glycoproteins hemagglutinin (HA) and neuraminidase (NA), IAV is categorized into different subtypes. To date, 18 different HA and 11 different NA subtypes have been described. While all conventional HA and NA subtypes are found in wild birds ^3,4, the most recently discovered subtypes H17N10 and H18N11 were exclusively detected in South American fruit bats ^5,6.

Infection of host cells is initiated by the engagement of host cell receptors via the influenza virus’ HA. The HA also carries the fusion machinery, which triggers fusion between viral and endosomal membranes, delivering the viral genome into the target cells’ cytoplasm ⁷. Conventional HA subtypes (H1-H16) bind to Sia receptors, which are terminal sugar moieties on host cell glycoproteins and glycolipids. The configuration of the Sia linkage to glycoconjugates confers species specificity and represents a fundamental determinant in shaping the IAV host range. Avian-adapted and human-adapted IAV have binding preference to α2,3- and α2,6-linked Sia, respectively ^8–11. In contrast, H17 and H18 lack the ability to bind to the canonical IAV receptor or other glycans^6,12,13. Instead, H17 and H18 use the major histocompatibility complex class II (MHC-II) from multiple species including humans, pigs, chickens, and bats as entry receptors ¹⁴. Several evolutionary conserved amino acids within the α-chain of the MHC-II heterodimer have been proposed to potentially interact with H18 and to be critical for H18N11 infection ¹⁵. However, the role of MHC-II as IAV receptor remains poorly understood.

Due to its pandemic potential and the constant threat IAV poses to poultry and livestock industries, IAV surveillance in wild birds is of global importance. Despite being key for risk assessment of potentially dangerous IAV strains circulating in wild animals, virus surveillance improves our understanding of IAV ecology and diversity ¹⁶. In the past two decades, such studies identified three IAV subtypes including H16 ¹⁷, H17 ⁵ and H18 ⁶. Recently, Fereidouni et al. ¹⁸ reported genome sequences of a previously undescribed IAV HA subtype, designated H19, in a common pochard (Aythya ferina) in Kazakhstan (A/common pochard/Kazakhstan/Kz52/2008). However, an infectious virus was not isolated from the IAV positive duck sample, impeding further characterization of the H19 subtype ¹⁸. Independently, during routine avian IAV surveillance in 2010 and 2013 in Northern California, USA we detected genome sequences of two avian IAVs with distinct HA sequences in lesser scaups (Aythya affinis). To our surprise, their sequence similarity with HA A/common pochard/Kazakhstan/Kz52/2008 suggests that both belong to the H19 subtype. As a fundamental molecular determinant of the IAV host range and zoonotic potential, our study aimed at examining the previously unknown receptor specificity of H19.

RESULTS

Lesser scaups in North California harbor IAV of the phylogenetically divergent H19 subtype

Two IAV positive cloacal swabs were collected from lesser scaups (A. affinis) during routine avian IAV surveillance at hunter check stations in North California, USA. Samples named UCD10-3087 and UCD13-1742 were collected at the Sacramento National Wildlife Refuge, California, USA, on January 12, 2010, and at the Grizzly Island Wildlife Refuge, California, USA, on November 2, 2013, respectively. IAV genome sequences in samples UCD10-3087 and UCD13-1742 were designated A/lesser scaup/California/3087/2010 and A/lesser scaup/California/1742/2013, respectively. Genome sequences encoding the internal viral genes PA, PB1, PB2, NP, M, NS from the 2010 sample and NP, NS, M from the 2013 sample mapped to contemporary avian IAV strains (Table S1). Only partial or no NA genomes sequences were recovered from A/lesser scaup/California/3087/2010 or A/lesser scaup/California/1742/2013, respectively. The partial NA sequence of A/lesser scaup/California/3087/2010 was closely related to A/duck/Minnesota/104/1974 (H4N6) (Table S1) but contained a large internal deletion of 1040 bp corresponding to the NA head domain. Also, it lacked the first 140 nucleotides encoding the cytoplasmic tail, transmembrane domain, and parts of the stalk domain (Figure S1) ¹⁹. For A/lesser scaup/California/1742/2013, no full-length genome sequences of the polymerase subunits (PB2, PB1 and PA) were recovered either. To determine the subtype of the detected HA genome sequences, we selected 71 HA sequences representing all HA subtypes reported to date (Table S2) and performed a phylogenetic analysis (Figure 1A and Figure S2A). Our analysis revealed that HA sequences from A/lesser scaup/California/3087/2010 (OR611720) and A/lesser scaup/California/1742/2013 (OR611723) form a distinct cluster within the H9 clade, indicating a separate HA subtype (Figure 1A). Both HA sequences are closely related and share 99.05 % nucleotide sequence identity and 98.75 % amino acid sequence identity. We also included the recently published HA A/common pochard/Kazakhstan/Kz52/2008, which was proposed as a new candidate HA subtype, designated H19 ¹⁸. Nucleotide and amino acid sequence identities between HA A/lesser scaup/California/3087/2010 and HA A/common pochard/Kazakhstan/Kz52/2008 of 83.6 % and 91.27 %, respectively, suggest that both belong to the same H19 subtype. Indeed, the inter-subtype amino acid distances between HA A/lesser scaup/California/3087/2010 and HA subtypes within the H9 clade are similar to those between H8, H9 and H12 (Figure 1B). This further indicates that HA of A/lesser scaup/California/3087/2010 and A/lesser scaup/California/1742/2013 together with A/common pochard/Kazakhstan/Kz52/2008 belong to a distinct subtype. Further analysis of the intra-subtype differences within the H9 clade, revealed that H19 sequences are more divergent from one another than sequences within H8 or H12 subtypes, but slightly less divergent compared to sequences of the H9 subtype (Figure S2B).

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.

To characterize the structure of a prototypic H19, we determined the crystal structure of the A/lesser scaup/California/3087/2010 HA ectodomain at a resolution of 2.4 Å (Figure 1C, Table S3). The overall structure of the H19 trimer highly resembles other HA structures characterized by a membrane-distal globular head domain and a membrane-proximal stalk domain containing the monobasic cleavage site ²⁰. The latter allows for proteolytic cleavage of the HA0 precursor into the disulfide-linked subunits HA1 and HA2 (Figure S3), representing the fusion competent active conformation ²⁰. Analyses of the putative receptor binding site (RBS) of H19 revealed that structural features such as the 130-loop, 190-helix, and the 220-loop ²⁰ are highly conserved (Figure 1D). We next analyzed the key RBS residues at positions 98, 134, 136, 153, 155, 183, 190, 194, 195, 225, 226, and 228 (H3 numbering) between H19 and other HA subtypes. We first determined the most abundant amino acids at these positions among conventional HA subtypes from Figure 1A (Figure S4, Table S4). While most of the key RBS residues are conserved between H19 and conventional HA subtypes, there are two changes at positions 134 and 226. Instead of glycine (Gly, G) at position 134, which is highly conserved across H1-H16, H19 contains threonine (Thr, T). Furthermore, position 226, which together with 228 determines avian- and human-type Sia receptor binding ²¹, contains an isoleucine (Ile, I) instead of glutamine (Gln, Q) or leucine (Leu, L). We next compared the structure of the putative RBS of H19 to H1, H9 and H18 (Figure 1E). The putative RBS of H18, where only four (W153, H183, E190 and Y195) of the key RBS residues are conserved, adopts a shallow and wide conformation unable to accommodate Sia⁶. Even though structural differences between the putative RBS of H19 and conventional HA (H1 and H9) are subtle, compared to H9 and H1 the putative RBS of H19 adopts a slightly more closed conformation.

H19 does not bind the canonical IAV receptor Sia

To study the receptor specificity of H19, we performed a hemagglutination assay, which exploits the ability of the IAV HA to bind Sia receptors on red blood cells (RBC) thereby cross-linking and preventing RBC from pelleting ²². We generated and tested recombinant H19 (A/lesser scaup/California/3087/2010) ectodomains. As controls we used recombinant H3 (A/Hong Kong/1/1968 or A/Switzerland/9715293/2013) and H5 (A/Viet Nam/1203/2004), representing human-type and avian-type HAs, respectively. To our surprise, recombinant H19 failed to hemagglutinate chicken or turkey RBC (Figure 2A). We next tested the ability of H19 to induce syncytia after low-pH exposure, which requires the HA to interact with host cell receptors, such as Sias, present on neighboring cells. Unlike H1 (A/WSN/33), H19 (A/lesser scaup/California/3087/2010) did not induce syncytia in transfected Vero E6 cells (Figure 2B). To investigate the receptor usage of H19 more comprehensively, we used glycan microarrays containing synthetic glycans with α2,3- and α2,6-linked Sias (Neu5Ac or Neu5Gc), and glycolipids with α2,3- and α2,8-linked Sias (Neu5Ac) ^23,24. The same approach has previously identified a sialylated ganglioside receptor for an influenza B-like virus from spiny eel ²⁵. As expected, recombinant H3 and H5 controls specifically bound to N-glycans containing α2,6-linked Neu5Ac (positions 5, 6) and N-glycans (7–9) or glycolipids (A, D, I, K) containing α2,3-linked Neu5Ac, respectively (Figure 2C). However, we did not detect specific binding of recombinant H19 to any of the glycans presented on the microarray (Figure 2C). Altogether, this strongly suggests that H19 does not bind to Sia.

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.

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

The absence of Sia binding prompted us to test if H19, like bat IAV H17 and H18 can use MHC-II as host cell receptor. We inoculated human embryonic kidney (HEK) 293T cells transiently expressing MHC-II from multiple species with virus-like particles (VLP) composed of the IAV matrix protein M1 fused to β-lactamase (BlaM1) ²⁶ pseudotyped with H19 (A/lesser scaup/California/3087/2010), H19 (A/lesser scaup/California/1742/2013), or H18 (A/bat/Peru/33/2013) as control. While all transfected MHC-II constructs rendered HEK293T cells susceptible to H18-BlaM1 VLP entry as previously shown ¹⁴, only HLA-DR orthologs from black flying fox (Pteropus alecto, Pa-DR), mouse (H2-E) and tufted duck (A. fuligula, Af-DR) rendered HEK293T cells susceptible to H19-BlaM1 VLPs (Figure 3A). Notably, tufted ducks (A. fuligula) belong to the same genus as lesser scaups (A. affinis). We expanded our MHC-II panel to investigate the potential host range of H19 and tested HLA-DR orthologs from black swan (Cygnus atratus, Ca-DR), mallard (Anas platyrhynchos, Ap-DR), Honduran yellow-shouldered bat (Sturnira hondurensis, Sh-DR), Jamaican fruit bat (Artibeus jamaicensis, Aj-DR), sea lion (Zalophus californianus, Zc-DR), and ferret (Mustela putorius furo, Mpf-DR). In addition to tufted duck, HLA-DR orthologs from black swan and mallard rendered HEK293T cells susceptible to H19-BlaM1 VLP entry (Figure 3B). We next generated MDCK cells overexpressing H2-E (LV-H2-E) and corresponding control cells (LV-ctrl) by lentiviral transduction. H2-E surface levels were confirmed by flow cytometry (Figure 3C). We inoculated MDCK LV-ctrl and LV-H2-E cells with H1N1-, H18- or H19-pseudotyped BlaM1 VLPs (Figure 3D) or vesicular stomatitis virus (VSV) encoding EGFP and firefly luciferase (Figure 3E and 3F). H2-E expression rendered MDCK cells susceptible to H19-VLP entry (Figure 3C–F). Increasing amounts of a monoclonal antibody against the α-chain of H2-E significantly inhibited VSV-H19 entry on MDCK LV-H2-E in a dose-dependent manner, while a control antibody did not have an effect (Figure 3G). To probe for interaction between H2-E and H19, we performed a syncytia formation assay in Vero E6 cells. When co-expressed with H2-E, but not with HLA-DR, H19 induced syncytia in Vero E6 cells upon a low-pH pulse (Figure 3H, Figure S5A). In contrast, H1 induced syncytia independent of MHC-II, and H18 only did so when H2-E or HLA-DR was co-expressed (Figure S5B). Consistently, we detected binding of recombinant H19 on MDCK LV-H2-E but not on control cells (Figure 3I). Taken together, our results suggest that MHC-II from distinct species function as entry receptors for H19.

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.

Recombinant H19N6 and H19N6stop IAV replicate in MDCK cells expressing H2-E

To further study a prototypic H19 IAV, we used reverse genetics to rescue an infectious virus from recombinant DNA plasmids encoding the HA segment and the internal segments of A/lesser scaup/California/3087/2010. Because a complete NA sequence of A/lesser scaup/California/3087/2010 was not recovered from the original sample, we used the NA of A/duck/Minnesota/104/1974 (H4N6), being the most closely related to the partial NA sequences recovered from the lesser scaup samples. We successfully rescued a recombinant H19N6 IAV using co-cultures of HEK293T and MDCK LV-H2-E (Figure 4A). Plaque formation was only detected on monolayers of MDCK LV-H2-E but not on control cells (Figure 4A). Growth kinetics of recombinant H19N6 IAV revealed that only MDCK cells expressing H2-E but not the control cell line supported H19N6 replication (Figure 4B). As expected, no difference was observed for our control H1N1 (A/PR/8/34) (Figure 4B). Since H19 does not bind Sia, we hypothesized that sialidase function of NA could be dispensable for H19 IAV replication, as described for bat IAV ²⁷. We introduced a premature stop codon into the N6 coding sequence, generating a truncated N6 (N6stop) lacking the enzymatically active head domain ²⁸ (Figure 4C). We rescued two pairs of isogenic H19 IAV, termed H19N6, H19N6stop, PR8-H19N6 and PR8-H19N6stop, carrying either wild type or truncated N6 A/duck/Minnesota/104/1974 with the internal genes of A/lesser scaup/California/3087/2010 or A/PR/8/34. We confirmed the absence of detectable sialidase activity in viruses carrying the truncated N6 (Figure 4C). To evaluate the impact of absent sialidase activity on H19 IAV replication, we inoculated MDCK LV-H2-E with the generated viruses at a low multiplicity of infection (MOI) of 0.1. Both viruses carrying the non-functional N6 efficiently replicated in MDCK LV-H2-E despite a modest attenuation compared to their counterparts with the wild type N6 (Figure 4D). Notably, this attenuation was more pronounced for the NA-deficient virus carrying the internal genes of A/PR/8/34. We conclude that NA function is not essential but might enhance H19 IAV replication in culture. Moreover, to exclude the possibility that reduced multivalency affected recombinant H19 to bind to Sias on glycan arrays in Figure 2C, we tested H19N6 virions i.e. PR8-H19N6. Indeed, we did not detect any specific binding of PR8-H19N6 to the presented glycans (Figure S6).

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.

Amino acid changes at positions 134, 137 and 226 in the putative RBS of H19 restore Sia-dependent entry

Since H19 did not mediate infection in the absence of MHC-II expression due to its inability to bind Sias, we aimed at assessing the number of amino acid changes in the H19 putative RBS to acquire Sia binding. We first modelled Sia into the putative RBS of H19 by structurally aligning the crystal structure of H19 A/lesser scaup/California/3087/2010 (PDB code 8VCC) and H9 A/swine/Hong Kong/9/1998 complexed with α2,3-linked Sia (PDB code 1jsh) (Figure 5A). We identified that residues T134 and P137 in the 130-loop potentially clash with the modelled Sia (Figure 5A). To identify mutations that restore Sia binding of H19, we mutated residues T134, P137 and I226. As a proxy for Sia mediated entry, we analyzed infectivity of luciferase-encoding VSV-pseudotypes carrying H19 mutants in MDCK control cells. We first generated single mutations T134 to G134, which is highly conserved across HA subtypes (Table S4), I226 to Q226 or L226, which together with G228 are determinants of avian- or human-type Sia receptor specificity, respectively ²¹. Furthermore, we mutated P137 to A137 and R137, representing amino acids predominantly present in H1 (Table S4) and in the most closely related HA sequence A/duck/Hong Kong/Y439/1997 (H9N2), respectively. To generate VSV-pseudotypes carrying H19 mutants that potentially bind Sia, we co-expressed N1. None of the H19 single mutants infected MDCK LV-ctrl, while infection of MDCK LV-H2-E was comparable (Figure 5B). Also, most of the H19 double mutants did not facilitate entry in MDCK LV-ctrl while retaining robust entry levels in MDCK LV-H2-E (Figure 5C). For the double mutant carrying mutations T134G and P137R, we detected low entry levels in control cells, indicating that Sia binding was potentially restored to a certain degree (Figure 5C). We next generated H19 mutants carrying combinations of mutations in all three positions. VSV-pseudotypes carrying H19 mutants T134G/P137A/I226Q, T134G/P137R/I226Q or T134G/P137A/I226L but not T134G/P137R/I226L showed robust entry levels in control cells comparable to H1N1-pseudotypes (Figure 5D). We suggest that three simultaneous mutations at positions 134 (G), 137 (A or R) and 226 (Q) are sufficient to restore Sia binding of H19. To validate these results, we treated MDCK LV-ctrl with bacterial sialidase to cleave Sia prior to inoculation with VSV-pseudotypes. Sialidase treatment strongly reduced surface levels of α2,3- and α2,6-Sias as measured by lectin staining (Figure S7A). Indeed, treatment of MDCK LV-ctrl with sialidase significantly reduced entry levels of the three H19 mutants with restored Sia-dependent entry (Figure 5E, Figure S7B). Notably, entry of pseudotypes carrying wild type H19 on MDCK LV-ctrl was not affected (Figure S7B). To test if the H19 mutations also affected MHC-II-mediated entry, we tested the effect of sialidase treatment on VSV-pseudotypes carrying the H19 mutants in MDCK LV-H2-E. While entry of H1N1-pseudotypes was strongly inhibited, entry of H19 mutants remained at similar levels suggesting that the mutations that restored Sia-dependent entry did not abolish MHC-II mediated entry (Figure 5F and Figure S7C). Similar to bat IAV entry and infection in culture ^29–31, pretreatment with sialidase enhanced entry of VSV-pseudotypes carrying the wild type H19 in MDCK LV-H2-E.

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.

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

To investigate which MHC-II amino acids confer species specificity, we generated chimeras of MHC-II that can or cannot facilitate H19-entry. To avoid potential incompatibilities within chimeras, we selected the human MHC-II HLA-DR and its mouse orthologue H2-E, which share amino acid sequence homology of 75.98% and 75.76% between α-chains and β-chains, respectively. We generated chimeric α- and β-chains by swapping α1-, α2-, β1-, and β2-domains between HLA-DR and H2-E (Figure 6A, Figure S8A–C). Combinations of chimeric and wild type α- and β-chains of HLA-DR and H2-E were expressed in HEK293T cells and surface levels were measured by flow cytometry (Figure S8D). For multiple combinations, we could not detect MHC-II surface expression (Figure S8D), potentially due to incompatibilities between the protein subunits. We continued our mapping attempts by selecting five MHC-II chimeras named ch1, ch2, ch3, ch4 and ch5, (Figure 6A) which showed robust surface expression levels (Figure S8D) and tested their ability to facilitate H19- and H18-entry(Figure 6A). While chimera ch4 showed reduced H18-entry, in cells transfected with ch1, ch2, ch3 and ch5, H18-entry levels were comparable to H2-E or HLA-DR transfected cells, indicating that ch1, ch2, ch3 and ch5 are functional entry receptors for H18-VLPs. However, only ch2 and ch5, but not ch1 and ch3, rendered HEK293T cells susceptible to H19-entry (Figure 6A). Even though we cannot exclude some contribution of the β1-domain our results suggest that the MHC-II α- but not the β-chain is sufficient to facilitate species-specific H19-entry. To identify amino acids in the MHC-II α-chain, that determine species specificity, we compared the membrane distal α1-domain of HLA-DR orthologs that represent functional (H2-E (mouse), Af-DR (tufted duck), Ap-DR (mallard), Ca-DR (black swan), Pa-DR (black flying fox)) and non-functional (HLA-DR (human)) entry receptors for H19 (Figure 6B, Figure 3A, B). We identified four groups of amino acids within the HLA-DR α1-domain that were distinct from H2-E and other functional MHC-II (Figure 6B and C). Amino acids N40/Q43/S44, M61/A62/K64, G74/R75 and T104/Y109 in HLA-DR potentially preventing H19-entry are located as four patches on the α1-domain (Figure 6C). We mutated amino acid patches shown in Figure 6C individually to amino acids present in H2-E. HEK293T cells were transfected with the HLA-DR α-chains carrying the mutations N40L/Q43K/S44R, M61I/A62E/K64S, G74A/R75K or T104K/Y109N with the wild type HLA-DR β-chain and tested for recovery of H19-entry (Figure 6D). While all mutants retained their ability to facilitate H18-entry (Figure 6D) and surface expression (Figure 6E), only the M61I/A62E/K64S mutant partially restored H19-entry (Figure 6D). We next combined mutations from two amino acid patches and tested if they could recover H19-entry. Indeed, the combination of the mutations N40L/Q43K/S44R and M61I/A62E/K64S largely restored H19-entry to comparable levels obtained for H18-entry (Figure 6F). HLA-DR α-chain mutants carrying the mutations M61I/A62E/K64S and G74A/R75K or M61I/A62E/K64S and T104K/Y109N only partially restored H19-entry, while the mutations N40L/Q43K/S44R and G74A/R75K, T104K/Y109N and N40L/Q43K/S44R or T104K/Y109N and G74A/R75K did not facilitate H19-entry (Figure 6F). Similar to single patch mutants, expression levels of the HLA-DR mutants in Figure 6F were slightly reduced (Figure 6G) and likely account for the minor reductions in entry levels compared to HLA-DR wild type. We concluded that amino acids at positions 40, 43, 44, 61, 62 and 64 in the MHC-II α1-domain potentially confer species-specific H19-entry. By introducing the reciprocal mutations from HLA-DR into the H2-E α1-domain we could confirm the requirement of both amino acid patches to facilitate H19-entry (Figure S8E). Next, we tested if the mapped amino acids were also important for the duck MHC-II Af-DR to facilitate H19-entry. When we introduced the corresponding amino acids from HLA-DR into the Af-DR α1-domain, mutations at positions 65, 66 and 68 (corresponding to residues 61, 62 and 64 in HLA-DR), but not 40, 43 and 44, abolished H19-entry while only minimally affecting H18-entry (Figure 6H). This suggests that the residues at positions 65, 66 and 68 in Af-DR are critical to confer species-specific H19-entry.

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 ⁴⁵are 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.

DISCUSSION

Like H17 and H18 ^5,6, H19 represents yet another HA subtype that exhibits MHC-II receptor specificity instead of Sia. While our binding studies using glycan arrays that display α2,3- and α2,6-linked Sias (Neu5Ac or Neu5Gc), and glycolipids containing α2,3- and α2,8-linked Sias ²³ suggest that H19 does not bind those Sias, we have not tested other Sia versions. One of the major Sia modification, 9-O-acetyl, which is bound by the influenza C virus hemagglutinin-esterase-fusion surface glycoprotein ³², would be disfavored in binding the pocket of an IAV HA due to the absence of the 9-OH group, which directly interacts with the conserved Y98 ³³. This residue is also conserved in H19, hence binding to 9-O-acetylated Sias would be similarly unfavored by its putative RBS. Other Sia modifications, such as 4-O-acetyl and N-glycolyl at the C5 are not displayed in the duck intestine ³⁴ and avian species ³⁵, respectively. Given that H19 was found in cloacal samples from ducks, we speculate that H19 is unlikely to bind such Sia versions either. The discovery of two closely related H19 sequences in lesser scaups in 2010 and 2013 suggests that this IAV subtype is circulating among lesser scaups and possibly other aquatic bird species in Northern California, USA. The recent report of a duck sample in Kazakhstan from 2008, where closely related H19 sequences were detected ¹⁸, indicates a potential global circulation of H19 IAVs in wild ducks. However, more IAV surveillance is needed to learn about its prevalence in the wild. Traditionally IAV HA and NA subtypes are classified based on antigenic properties determined by double radial immunodiffusion assay ^17,36. In contrast, the classification of the IAV subtypes H17N10 and H18N11 from bats ^5,6, as well as H19, ¹⁸ is based on nucleotide sequence divergence. While our study supports H19 as a distinct IAV subtype by sequence inference, antigenic classification remains to be confirmed.

Unlike H17 and H18, which can use classical human MHC-II and homologs from various species ¹⁴, MHC-II usage of H19 is limited to distinct species. This includes avian species like black swan, tufted duck, and mallard, and mammalian species such as black flying fox and mouse. The discovery of H19 in diving ducks by us and other groups ¹⁸, suggests that H19 has adapted to use HLA-DR homologs from aquatic birds. Phylogenetically, H19 is distantly related to H17 and H18, suggesting that MHC-II receptor usage has evolved independently. However, it might also be possible that Sia and MHC-II dual receptor usage was present in a common ancestor for all group I IAV, and that Sia binding was lost independently for bat IAV and avian H19 IAV. This scenario is supported by our observations that human and avian H2N2 IAV can have dual receptor specificity and use both Sia and MHC-II receptors (in press).

While it has been challenging to biochemically confirm the interaction between MHC-II and H17 or H18, in silico docking experiments proposed a model of the putative interaction interface between H18 and HLA-DR ¹⁵. We identified six polymorphic amino acids at positions 40, 43, 44, 61, 62 and 64 in the membrane-distal α1-domain of MHC-II that determine species-specific H19-entry, and these positions coincide with the putative interaction site identified for H18 and HLA-DR ¹⁵. Our mapping approach between the human HLA-DR and its mouse ortholog H2-E revealed that all six residues were required to facilitate H19-entry. For Af-DR, the HLA-DR ortholog from tufted ducks, which represents the MHC-II receptor potentially encountered by H19 in its natural host, we found that mutation of the residues 65, 66 and 68 (corresponding to residues 61, 62 and 64 in HLA-DR) was sufficient to abolish H19-entry. This suggests that the residues 65, 66 and 68 on the α1-domain of Af-DR are critical to facilitate H19-entry. We speculate that binding affinities between H19 and Af-DR or other functional MHC-II such as H2-E might differ. Depending on the strength of this interaction, which might involve other residues we have not mapped, mutation of either amino acid patch comprising residues 40, 43, 44 or 61, 62, 64 is sufficient to disrupt or restore the interaction between H19 and MHC-II of a given species.

Previous efforts to propagate infectious H19 IAV from the duck sample in Kazakhstan using conventional methods were unsuccessful, thereby hindering further characterization of this newly identified IAV subtype ¹⁸. Despite failed attempts to recover infectious virus from both lesser scaup samples using conventional approaches, creating MDCK cells overexpressing H2-E allowed us to rescue recombinant H19 IAV. Even though full-length sequences of the NA segments from both lesser scaup samples remain unidentified, the NA segment of A/duck/Minnesota/104/1974 (H4N6) was compatible with the remaining seven segments from A/lesser scaup/California/3087/2010. The genome sequences of H19 IAV found in 2010 and 2013 harbor internal gene segments that share over 95 % sequence identity with those of avian IAV of the H3, H4, H5 or H9 subtypes. In contrast to bat IAV, where packaging incompatibilities with conventional IAV have been demonstrated ^37,38, our experiments show that the HA segment of A/lesser scaup/California/3087/2010 was compatible with the internal genes of A/PR/8/34 (H1N1). This suggests that reassortment of H19 IAV with other conventional IAV is possible. A common feature of the bat IAV subtypes H17N10 and H18N10, which do not bind Sia receptors, is that they encode a full-length NA-like segment without sialidase function ^6,39,40. In contrast, no full-length NA genomes were detected by us and others ¹⁸ with the H19 genome sequences. However, we speculate that the partial NA sequence with an internal truncation we detected in one of the duck samples might suggest that H19 IAV carry a non-functional NA. IAV containing NA segments with internal deletions have been isolated from patients treated with neuraminidase inhibitors ^41,42. Despite lacking enzymatic activity, such NA segments maintain conserved RNA replication and packaging signals required for viral particle formation. The successful rescue of a recombinant H19 IAV carrying a non-functional NA suggests that it does not require neuraminidase functions for replication but likely require a complete set of genome segments for efficient packaging. Interestingly, even though NA function was not required it was beneficial for replication and entry. It was previously described that sialidase treatment enhances H17N10 and H18N11 infection ^29–31. Likewise, H19-entry was enhanced in tissue culture upon treatment with bacterial sialidase. This could explain the observed attenuation of recombinant H19 IAV without a functional NA. Regardless of the NA and internal gene segments, replication of recombinant H19 IAV was limited to MDCK cells overexpressing MHC-II. In future, such cell lines combined with conventional isolation methods could help discover unidentified IAV with MHC-II receptor specificity. Detection of H19 sequences in cloacal swabs of ducks, indicates that H19 IAV replicates in their intestinal tract and is shed through the feces of infected animals. As reported for low pathogenic avian IAVs, ⁴³ H19 IAV may transmit via the oral-fecal route among ducks. For bat IAV H18N11, infection studies in Neotropical Jamaican fruit bats have demonstrated ²⁷ that virus replication occurs in the follicle-associated epithelium (FAE) of jejunal Peyer's patches, a part of the gut-associated lymphoid tissue that harbors lymphoid cells, such as dendritic cells and macrophages, which typically express MHC-II. Apart from hematopoietic cells, MHC-II expression has been detected in epithelial cells of the mouse intestine at steady state ⁴⁴. Assuming IAV subtypes that use MHC-II receptors have a similar tissue tropism, we speculate that H19 IAV may also replicate in the FAE of gut-associated lymphoid tissues in ducks. However, future studies are needed to determine the cell and organ tropism of H19 IAV in their natural hosts. The use of mice having H2-E might also facilitate understanding the cell tropism of H19 IAV.

As a prototypic IAV H19 glycoprotein we solved the crystal structure of the A/lesser scaup/California/3087/2010 HA. Despite the overall conservation of key RBS residues and architecture, we suggest that the slightly closed conformation hinders binding of H19 to Sia. We identified two amino acids at positions 134 and 137 within the 130-loop that potentially interfere with Sia binding. Remarkably, we restored Sia-dependent H19-entry when we introduced simultaneous mutations at those two positions and position 226. Notably, position 226, along with position 228, determines the specificity of the HA to bind to α2,3- or α2,6-linked Sia receptors ²¹. Overall, this suggests that only three amino acid substitutions are sufficient to switch the receptor specificity of H19 and carries the potential to expand its host range.

In summary, our study demonstrates that H19, a previously uncharacterized HA subtype, exhibits MHC-II receptor specificity instead of Sia binding. This highlights the existence of IAV subtypes beyond the bat IAV subtypes that rely on MHC-II as their receptor. While bat IAVs exhibit broad MHC-II usage, H19 displays selective MHC-II usage, primarily within aquatic birds. Similar to the role of the Sia linkage in determining species-specificity for conventional IAVs, polymorphic amino acids in the MHC-II α1-domain may determine the host range of IAV utilizing MHC-II receptors. The evolutionary driving forces for IAV to adopt MHC-II receptor usage remain a fascinating subject for future research.

STAR METHODS

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and regents may be directed to and will be fulfilled by the lead contact, Dr. Adolfo García-Sastre (adolfo.garcia-sastre@mssm.edu).

Materials availability

All unique reagents generated in this study are available from the lead contact without restriction.

Data and code availability

• IAV genome sequences have been deposited at the National Center for Biotechnology Information (NCBI) and are publicly available as of the date of publication. Accession numbers are listed in the key resources table.

• Electron density map of the A/lesser scaup/California/3087/2010 HA protein has been deposited in the Protein Data Bank (PDB) and is publicly available as of the date of publication. The accession code is listed in the key resources table.

• This paper does not report original code.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Fluorescein isothiocyanate (FITC)-labelled anti-H2-Eα antibody clone M5/114.15.2	Abcam	Catalog no. ab93561, RRID:AB_1056401 2
Allophycocyanin (APC)-labelled anti-human HLA-DR antibody clone L243	Biolegend	Catalog no. 307610, RRID:AB_314687
Phycoerythrin (PE)-labelled anti-human HLA-DR, DP, DQ Antibody clone Tu39	Biolegend	Catalog no. 361716, RRID:AB_2750318
pan-HA antibody CR9114	Dreyfus, C. et al., 46	N/A
Goat anti-Human IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 647	ThermoFisher Scientific	Catalog no. A21445, RRID:AB_2535862
anti-human HLA-DR antibody clone L243	Biolegend	Catalog no. 307602, RRID:AB_314680
Goat anti-Mouse IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 488	ThermoFisher Scientific	Catalog no. A11029, RRID:AB_2534088
anti-mouse I-A/I-E Antibody	Biolegend	Catalog no. 107602, RRID:AB_313317
Donkey anti-Rat IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 488	ThermoFisher Scientific	Catalog no. A11029
anti-VSV-G antibody	Kerafast	Catalog no. EB0010, RRID:AB_2811223
anti-6xHis antibody	ThermoFisher Scientific	Catalog no. MA1-21315, RRID:AB_557403
mouse anti-his Alexa 647	ThermoFisher Scientific	Catalog no. MA1-21315-A647, RRID:AB_2610647
human anti streptag	IBA	Catalog no. 2-1507-001, RRID:AB_513133
goat anti-human Alexa-555	ThermoFisher Scientific	Catalog no. A-21433, RRID:AB_2535854
goat-anti-human Alexa 647	ThermoFisher Scientific	Catalog no. A-21445, RRID:AB_2535862
Bacterial and virus strains
A/PR/8/34 (H1N1)	This paper	N/A
H19N6	This paper	N/A
H19N6stop	This paper	N/A
H19N6 (6+2)	This paper	N/A
H19N6stop (6+2)	This paper	N/A
Recombinant vaccinia virus expressing bacteriophage T7 RNA polymerase (MVA-T7)	Whitt et al. 47	N/A
Chemicals, peptides, and recombinant proteins
LIVE/DEAD™ Fixable Near-IR Dead Cell Stain Kit	ThermoFisher Scientific	Catalog no. L10119
L-1-tosylamido-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin	Sigma-Aldrich	Catalog no. T-8802
Trypsin inhibitor from soybean	ThermoFisher Scientific	Catalog no. 17075029
Diethylaminoethyl (DEAE)-dextran hydrochloride	Sigma-Aldrich	Catalog no. D9885
ViaFect	Promega	Catalog no. E4982
Lipofectamine™ 2000 Transfection Reagent	ThermoFisher Scientific	Catalog no. 11668019
2-(N-Morpholino)ethanesulphonic acid (MES)	ThermoFisher Scientific	Catalog no. AAJ62081AK
4’,6-diamidino-2-phenylindole (DAPI)	ThermoFisher Scientific	Catalog no. D1306
Neuraminidase from Vibrio cholerae	Sigma-Aldrich	Catalog no. N6514-1UN
Neuraminidase from Vibrio cholerae	Sigma-Aldrich	Catalog no. 11080725001
Bovine Serum Albumin V-standard grade	ThermoFisher Scientific	Catalog no. 50-753-3053
UltraPure™ 0.5M EDTA, pH 8.0	ThermoFisher Scientific	Catalog no. 15575020
Biotinylated Sambucus Nigra Lectin	Vector Labs	Catalog no. B-1305-2
Biotinylated Maackia Amurensis Lectin II	Vector Labs	Catalog no. B-1265-1
FITC-labelled streptavidin	ThermoFisher Scientific	Catalog no. SA1001
FuGENE® HD transfection reagent	Promega	Catalog no. E2311
Critical commercial assays
Cell staining buffer	Biolegend	Catalog no. 420201
MagMAX-96 AI/ND Viral RNA Isolation kit	ThermoFisher Scientific	Catalog no. AM1835
Superscript III one-step RT-PCR system	ThermoFisher Scientific	Catalog no. 12574018
QIAquick Gel Extraction kit	Qiagen	Catalog no. 28704
TOPO™ TA Cloning™ Kit for Sequencing	ThermoFisher Scientific	Catalog no. K4575J10
QIAprep Spin Miniprep Kit	Qiagen	Catalog no. 27104
LiveBLAzer™ FRET-B/G Loading Kit with CCF2-AM	ThermoFisher Scientific	Catalog no. K1032
ONE-Glo Luciferase Assay System	Promega	Catalog no. E6120
NA-Star™ Influenza Neuraminidase Inhibitor Resistance Detection Kit	ThermoFisher Scientific	Catalog no. 374422
In-Fusion Snap Assembly cloning	Takara	Catalog no. 638948
Deposited data
Crystal structure of H19 influenza A virus hemagglutinin from A/lesser scaup/California/3087/2010	PDB	8VCC
Influenza A virus (A/Lesser Scaup/CA/3087/2010(H19)) segment 4 hemagglutinin (HA) gene, complete cds	NCBI	OR611720
Influenza A virus (A/Lesser Scaup/CA/1742/2013(H19)) segment 4 hemagglutinin (HA) gene, complete cds	NCBI	OR611723
Influenza A virus (A/Lesser Scaup/CA/1742/2013(H19)) segment 8 nuclear export protein (NEP) and nonstructural protein 1 (NS1) genes, complete cds	NCBI	OR602902
Influenza A virus (A/lesser scaup/CA/1742/2013(H19)) segment 7 matrix protein 2 (M2) and matrix protein 1 (M1) genes, complete cds	NCBI	OR479890
Influenza A virus (A/lesser scaup/CA/1742/2013(H19)) segment 5 nucleocapsid protein (NP) gene, complete cds	NCBI	OR479766
Experimental models: Cell lines
Human embryonic kidney (HEK) cells 293T	ATCC	Catalog no. CRL-3216; RRID:CVCL_0063
Madin-Darby canine kidney (MDCK) cells	ATCC	Catalog no. CCL-34,
Vero (clone E6) cells from African green monkey kidney cells	ATCC	Catalog no. CRL-1586, RRID:CVCL_0574
Baby hamster kidney (BHK-21) cells	ATCC	Catalog no. CCL-10, RRID:CVCL_1915
Expi293F	ThermoFisher Scientific	Catalog no. A14527
Expi293F GNTI-	ThermoFisher Scientific	Catalog no. A39240
HEK293S GnTI(-)	ATCC	Catalog no. CRL-3022
MDCK LV-H2-E	This paper	N/A
MDCK LV-ctrl	This paper	N/A
Oligonucleotides
MBTuni-12: 5’-ACGCGTGATCAGCAAAAGCAGG-3’	Zhou, B. & Wentworth et al. 48	N/A
MBTuni-13: 5’-ACGCGTGATCAGTAGAAACAAGG-3’	Zhou, B. & Wentworth et al. 48	N/A
M13 forward primer: 5’-TGTAAAACGACGGCCAGT-3’	Ghedin et al. 49	N/A
M13 reverse primer: 5’-CAGGAAACAGCTATGACC-3'	Ghedin et al. 49	N/A
pDZ SapI forward primer (PA, PB1, PB2): 5’-CCGAAGTTGGGGGGGAGCGAAAGCAGG-3’	This paper	N/A
pDZ SapI forward primer (HA, NP, M, NA, NS): 5’-CCGAAGTTGGGGGGGAGCGAAAGCAGG-3’	This paper	N/A
pDZ SapI reverse primer (all segments): 5’-GGCCGCCGGGTTATTAGTAGAAACAAGG-3’	This paper	N/A
N6-G290A-frw: 5’-GTCCTTAGATAGAATGTGCTATGAGTTTACTTCACACAAGG-3’	This paper	N/A
N6-G290A-rev: 5’-CCTTGTGTGAAGTAAACTCATAGCACATTCTATCTAAGGAC-3’	This paper	N/A
rH19 A/lesser scaup/California/3087/2010 frw: 5’-TTTGGCAAAGAATTCGCCACCATGTGGAAACTCG-3’	This paper	N/A
rH19 A/lesser scaup/California/3087/2010 rev: 5’-TGGGACCAAGCGGCCGCTGATCTTGTAAACTGTA-3’	This paper	N/A
rH5 A/Viet Nam/1203/2004 (HALo) frw 5’-TTTGGCAAAGAATTCGCCACCATGGAGAAAATAGTG-3’	This paper	N/A
rH5 A/Viet Nam/1203/2004 (HALo) rev 5’-TGGGACCAAGCGGCCGCTTATTTGGTAAATTCC-3’	This paper	N/A
rH3 A/Hong Kong/1/1968 frw: 5’-TTTGGCAAAGAATTCGCCACCATGAAGACCATCATTG-3’	This paper	N/A
rH3 A/Hong Kong/1/1968 rev: 5’-TGGGACCAAGCGGCCGCTGTCTTTGTATCCAGAC-3’	This paper	N/A
See table S5 for the remaining primers used in this study
Recombinant DNA
pLVX-IRES-puro	Clontech	Catalog no. 632183
pLVX-IRES-neo	Clontech	Catalog no. 632181
pLVX-H2-EA-IRES-puro	This paper	N/A
pLVX-H2-EB-IRES-neo	This paper	N/A
pCAGGS-H19 A/lesser scaup/California/3087/2010	This paper	N/A
pCAGGS-H19 A/lesser scaup/California/1742/2013	This paper	N/A
pCAGGS-A/WSN/33-H1	Karakus et al. 14.	N/A
pCAGGS-A/WSN/33-BlaM1	Karakus et al. 14.	N/A
pCAGGS-A/bat/Guat/164/09 (H17N10)-BlaM1	Karakus et al. 14.	N/A
pCAGGS-A/bat/Peru/033/10-H18	Karakus et al. 14.	N/A
pmCherry-C1	Clontech	Catalog no. 632524
pHLA-DRA	Karakus et al. 14	N/A
pHLA-DRB	Karakus et al. 14	N/A
pHLA-DQA	Karakus et al. 14	N/A
pHLA-DQB	Karakus et al. 14	N/A
pHLA-DPA	Karakus et al. 14	N/A
pHLA-DPB	Karakus et al. 14	N/A
pH2-EA	Karakus et al. 14	N/A
pH2-EB	Karakus et al. 14	N/A
pSLA-DRA	Karakus et al. 14	N/A
pSLA-DRB	Karakus et al. 14	N/A
pB-LA	Karakus et al. 14	N/A
pB-LB	Karakus et al. 14	N/A
pEf-DRA	Karakus et al. 14	N/A
pEf-DRB	Karakus et al. 14	N/A
pPa-DRA	Karakus et al. 14	N/A
pPa-DRB	Karakus et al. 14	N/A
pMl-DRA	Karakus et al. 14	N/A
pMl-DRB	Karakus et al. 14	N/A
pAf-DRA	This paper	N/A
pAf-DRB	This paper	N/A
pAp-DRA	This paper	N/A
pAp-DRB	This paper	N/A
pCa-DRA	This paper	N/A
pCa-DRB	This paper	N/A
pSh-DRA	This paper	N/A
pSh-DRB	This paper	N/A
pAj-DRA	This paper	N/A
pAj-DRB	This paper	N/A
pZc-DRA	This paper	N/A
pZc-DRB	This paper	N/A
pMpf-DRA	This paper	N/A
pMpf-DRB	This paper	N/A
pChA1	This paper	N/A
pChA2	This paper	N/A
pChB1	This paper	N/A
pChB2	This paper	N/A
pHLA-DRA mutants	This paper	N/A
pH2-EA mutants	This paper	N/A
pAf-DRA mutants	This paper	N/A
pCAGGS-H19 mutants	This paper	N/A
pVSV∆G-EGFP-firefly luciferase	This paper	N/A
Set of Helper Plasmids (VSV-N, VSV-P, VSV-L, VSV-G)	Kerafast	Catalog no. EH1012
pDZ-PA A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-PB1 A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-PB2 A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-HA A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-NP A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-M A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-NS A/lesser scaup/California/3087/2010	This paper	N/A
pDZ-NA A/duck/Minnesota/104/1974(H4N6)	This paper	N/A
pDZ-NAstop A/duck/Minnesota/104/1974(H4N6)	This paper	N/A
pCAGGS-rH19 A/lesser scaup/California/3087/2010	This paper	N/A
pCAGGS-rH5 A/Viet Nam/1203/2004 (HALo)	This paper	N/A
pCAGGS-rH3 A/Hong Kong/1/1968	This paper	N/A
pCAGGS-rH3 A/Hong Kong/1/1968	This paper	N/A
pCD5-rH3 A/Switzerland/9715293/2013	This paper	N/A
Software and algorithms
FlowJo	BD	https://www.flowjo.com/solutions/flowjo
MEGA	Open source	https://www.megasoftware.net/
CodonCode Aligner	CodonCode Corporation	https://www.codoncode.com/aligner/
PyMOL	Schrödinger	https://pymol.org/
GraphPad Prism v10	Graphpad software	https://www.graphpad.com/
Other
KingFisher Magnetic Particle Processor	ThermoFisher Scientific	Catalog no. A31508
Ion Torrent PGM next-generation sequencing system	ThermoFisher Scientific	Catalog no. A25511
Beckman Coulter Gallios	Beckman Coulter	RRID:SCR_01963 9
BD FACS Canto II flow cytometer	BD	N/A
EVOS M5000	ThermoFisher Scientific	Catalog no. AMF5000, RRID:SCR_02365 0
BioTek Synergy Neo2	Agilent Technologies	N/A
Superose 6 Increase 10/300 GL column	Cytiva	Catalog no. 29091596
Oryx Nano crystallization robot	Douglas Instruments	N/A
Innoscan 710	Innopsys	N/A

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Cells

Human embryonic kidney cells (HEK) 293T, Madin-Darby canine kidney (MDCK), Vero (clone E6) cells from African green monkey kidney, and baby hamster kidney (BHK-21) cells were purchased from the American Type Culture Collection (ATCC). All cell lines were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with fetal bovine serum (FBS 10% v/v, Corning) and Pen-Strep (100 U/mL penicillin and 100 μg/mL streptomycin, Corning). Expi293F (ThermoFisher, catalog no. A14527) and Expi293F GNTI- (ThermoFisher, catalog no. A39240) cells were cultured in suspension in Expi293 Expression medium (ThermoFisher, catalog no. A1435102). To generate MDCK cells stably expressing H2-E (MDCK LV-H2-E) or control cells (MDCK LV-ctrl), cDNAs encoding H2-Ea (alpha chain, NM_010381) or H2Eb1 (beta chain, NM_010382.2) were synthesized (ThermoFisher Scientific) and cloned into pLVX-IRES-puro (Clontech catalog no. 632183) or pLVX-IRES-neo (Clontech, catalog no. 632181) using XhoI and NotI restriction sites. To generate lentiviruses encoding pLVX-IRES-H2-EA-puro, pLVX-IRES-H2-EB-neo, and corresponding empty vector controls, HEK293T cells were co-transfected with pMD2.G, pCMVdR8.91 and the respective lentiviral constructs. At 48 h post transfection, supernatants were harvested. Cell debris were removed from supernatants by centrifugation at 3000 rpm for 5 min and filtration through a 0.45 μm filter. MDCK cells were co-transduced with pLVX-H2-EA-IRES-puro- and pLVX-H2-EB-IRES-neo-encoding or corresponding control lentiviruses in the presence of 8 μg/mL polybrene. At 48 h post transduction cells were selected with 2.5 μg/mL puromycin (InvivoGen, catalog no. ant-pr-1) and 300 μg/mL G418 (InvivoGen, catalog no. ant-gn-1). H2-E surface expression was confirmed by flow cytometry.

Viruses

Influenza virus strain A/PR/8/34 (H1N1) grown and titrated by plaque assay on MDCK cells. H19N6 and H19N6stop were grown and titrated by plaque assay on MDCK LV-H2-E cells.

METHOD DETAILS

IAV surveillance and sequencing

Cloacal swabs were collected from lesser scaups at hunter check stations during routine avian influenza virus (AIV) surveillance. Sample UCD10-3087 was collected at the Sacramento National Wildlife Refuge on January 12, 2010, and UCD13-1742 was collected at the Grizzly Island Wildlife Refuge on November 2, 2013. Samples were collected using a sterile rayon swab and stored in viral transport media then brought to the laboratory and frozen at −80 °C. Initially the samples were screened for AIV by RT-PCR using established protocols ⁵⁰. Briefly, RNA was extracted using MagMAX-96 AI/ND Viral RNA Isolation kit (ThermoFisher Scientific, catalog no. AM1835), on a KingFisher Magnetic Particle Processor (ThermoFisher Scientific, Waltham, MA). The RT-PCR was set up to screen for the AIV matrix gene. The matrix gene cycle threshold (Ct) PCR value for sample UCD10-3087 was 31.5 and for UCD13-1742 it was 32.6. Samples with a Ct value <45 is considered positive. Each RNA extract was tested in one replicate, the RT-PCR included a known IAV isolate as positive control and viral transport medium (VTM) diluent as a negative control. Virus isolation was attempted by inoculation into embryonated chicken eggs (Charles River, CT), embryonated mallard eggs (Metzer Farms, CA) and MDCK cells. Each specimen was inoculated into eggs and cells twice using a described protocol ⁵¹. Egg samples were screened by RT-PCR for IAV RNA after each isolation attempt and cells were examined for cytopathic effects. However, no virus was isolated from those samples. For full genome sequencing, viral RNA was extracted from the original cloacal swab samples using the MagMax-96 AI/ND Viral RNA kit (ThermoFisher Scientific, catalog no. AM1835) on a KingFisher Magnetic Particle Processor (Thermo Scientific, MA) and served as a template in the multi segment RT-PCR method ⁵². Briefly, using the Superscript III one-step RT-PCR system (ThermoFisher Scientific, catalog no. 12574018) and the two primers MBTuni-12 (5’-ACGCGTGATCAGCAAAAGCAGG-3’) and MBTuni-13 (5’-ACGCGTGATCAGTAGAAACAAGG-3’) to amplify all genome segments. The resulting cDNA amplicons were run on an agarose gel, processed using the QIAquick Gel Extraction kit (Qiagen, CA) then directly cloned using the TOPO TA kit (Invitrogen, CA) following kit directions ⁴⁸. Sets of eight recombinant plasmid pellets were purified using the QIAprep Spin Miniprep Kit (Qiagen, CA), eluted with water and digested with EcoRI. Digests were analyzed on a gel and submitted for Sanger sequencing using the M13 forward (F) and reverse (R) primers ⁴⁹, (F primer: 5’-TGTAAAACGACGGCCAGT-3’; R primer: 5’-CAGGAAACAGCTATGACC-3') to the UC Davis UCDNA Sequencing Facility. Sequences were analyzed using MEGA6 and CodonCode Aligner. For each viral gene segment except for the NA segment, where a positive cDNA amplicon was obtained, at least two clones were sequenced. No polymorphisms were identified. Genome sequences from A/lesser scaup/California/3087/2010 and A/lesser scaup/California/1742/2013 were deposited on GenBank at NCBI and are summarized in Table S1. The partial coding sequence of the NA segment from A/lesser scaup/California/3087/2010 in Figure S1 was obtained by sequencing of amplicons directly after RT-PCR using the Ion Torrent PGM next-generation sequencing system (ThermoFisher Scientific).

Phylogenetic analysis

71 IAV HA sequences representing all HA subtypes (H1-H19) reported to date were selected from the National Center for Biotechnology Information (NCBI) data base covering sequences from a broad range of HA-NA combinations, times of isolation, hosts, and geographical locations (Table S2). Amino acid sequences were aligned with MUSCLE and a phylogenetic tree was constructed by the Maximum Likelihood method, using the MEGA11 ⁵³ software with the default parameters and 1000 bootstrap replications. To determine the evolutionary divergence between (inter-subtype distance) and among (intra-subtype distance) HA subtypes of the H9 clade, amino acid sequences from H19 (n=3), H8 (n=122), H9 (n=352) and H12 (n=156) were analyzed by calculating the number of amino acid substitutions per site by averaging over all sequence pairs between groups. Analyses were conducted using the Poisson correction model ⁵⁴. This analysis involved 633 amino acid sequences. All ambiguous positions were removed for each sequence pair (pairwise deletion option). There was a total of 222 positions in the final dataset. Evolutionary analyses were conducted in MEGA11 ⁵³.

H2-E and HLA-DR surface staining

After detaching with 0.25% trypsin-ethylenediamine tetraacetic acid (EDTA) (Corning), cells were resuspended in DMEM supplemented with 10% FBS and Pen-Strep (100 U/mL penicillin and 100 μg/mL streptomycin, Corning). After two washes by centrifugation at 1200 rpm for 3 min and resuspension with cell staining buffer (Biolegend, catalog no. 420201), the cells were incubated for 1 h at 4 °C with fluorescein isothiocyanate (FITC)-labelled anti-H2-Eα antibody clone M5/114.15.2 (Abcam, catalog no. ab93561), allophycocyanin (APC)-labelled anti-human HLA-DR antibody clone L243 (Biolegend, catalog no. 307610) or phycoerythrin (PE)-labelled anti-human HLA-DR, DP, DQ Antibody clone Tu39 (Biolegend, catalog no. 361716) diluted 1:50 in cell staining buffer. Dead cells were excluded from analysis using the LIVE/DEAD^™ Fixable Near-IR Dead Cell Stain Kit (ThermoFisher Scientific, catalog no. L10119) according to the manufacturer’s instruction. After three more washes with cell staining buffer as described above, FITC or APC signals from 5’000–10’000 live cells were measured using a Beckman Coulter Gallios or BD FACS Canto II flow cytometer and analyzed using FlowJo v10 software.

IAV-based virus-like particle (BlaM1-VLP) production and inoculation

cDNAs encoding HA from A/lesser scaup/California/3087/2010 (OR611720) and A/lesser scaup/California/1742/2013 (OR611723) were synthesized (ThermoFisher Scientific, Azenta Life Sciences) and cloned into pCAGGS using restriction sites NotI and XhoI. To generate pseudotyped BlaM1 VLPs ²⁶, HEK293T cells were co-transfected in 6-well plates with 2 μg of pCAGGS-A/WSN/33-BlaM1 and 1 μg of pCAGGS-A/lesser scaup/California/3087/2010-H19 or pCAGGS-A/lesser scaup/California/1742/2013-H19, or with 2 μg of pCAGGS-A/bat/Guat/164/09 (H17N10)-BlaM1 and 1 μg of pCAGGS-A/bat/Peru/033/10-H18 ¹⁴. For transfection, ViaFect (Promega) was used at a DNA to ViaFect ratio of 1:3. 6 h post transfection, supernatants were replaced with Opti-MEM containing Pen-Strep (100 U/mL penicillin and 100 μg/mL streptomycin, Corning). After harvesting VLPs 72 h post transfection, they were treated with 5 μg/mL L-1-tosylamido-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (Sigma-Aldrich, catalog no. T-8802) for 20 min at 37 °C for HA cleavage followed by trypsin inactivation with 10 μg/mL trypsin inhibitor from soybean (ThermoFisher Scientific, catalog no. 17075029) for 20 min at 37 °C. For infection, cells were washed with phosphate-buffered saline (PBS) and incubated with 200 μL of BlaM1 VLPs in the presence of 0.1 μg/mL diethylaminoethyl (DEAE)-dextran hydrochloride and 2% FBS for 4 h at 37 °C. Cells were collected by trypsinization and incubated with the fluorogenic β-lactamase substrate CCF2-AM (ThermoFisher Scientific, catalog no. K1032) for 1 h at 37 °C. Samples were analyzed on a Beckman Coulter Gallios or BD FACS Canto II flow cytometer. Dead cells were excluded using LIVE/DEAD^™ Fixable Near-IR Dead Cell Stain Kit (ThermoFisher Scientific). Entry positive cells were quantified by gating on events with cleaved CCF2-AM using FlowJo v10 software.

Syncytia formation assay

Vero E6 cells were seeded onto poly-L-lysine-coated 24-well plates at 150’000 cells per well. The next day, cells were co-transfected with expression plasmids encoding alpha and beta chains of either HLA-DR or H2-E ¹⁴, or pCAGGS-empty vector, together with either pCAGGS-A/WSN/33-H1, -A/bat/Peru/033/10-H18 or -A/lesser scaup/California/3087/2010-H19. Lipofectamine2000 (ThermoFisher) was used as transfection reagent at a DNA: Lipofectamine2000 ratio of 1:2. The next day, cells were washed with PBS and treated with 1 μg/mL TPCK-treated trypsin for 10 min at 37 °C before being exposed to low pH (50 mM 2-(N-Morpholino)ethanesulphonic acid (MES), 150 mM NaCl, pH 5.2) for 20 min at 37 °C. Cells were washed twice with PBS and subsequently incubated with DMEM supplemented with 10% FBS and Pen-Strep (100 U/mL penicillin and 100 μg/mL streptomycin, Corning) for 4 h at 37 °. Cells were fixed with 3.7% (v/v) paraformaldehyde (PFA) for 15 min and stained for HA with the pan-HA antibody CR9114 ⁴⁶ and Goat anti-Human IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor^™ 647 (ThermoFisher, catalog no. A21445), for HLA-DR with anti-human HLA-DR antibody clone L243 (Biolegend, catalog no. 307602) and Goat anti-Mouse IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor^™ 488 (ThermoFisher Scientific, catalog no. A11029), for H2-E with anti-mouse I-A/I-E Antibody (Biolegend, catalog no. 107602) and Donkey anti-Rat IgG (H+L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor^™ 488 (ThermoFisher Scientific, catalog no. A11029). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). Pictures were acquired on an EVOS M5000 (ThermoFisher Scientific) fluorescence microscope.

Ectopic expression of MHC-II in HEK293T cells

Plasmids encoding alpha- or beta-chains of HLA-DR, -DQ, -DP, HLA-DR orthologues from pigs (Sus scrofa, SLA-DR), chicken (Gallus gallus, B-L) and bats (Eptesicus fuscus, Ef-DR; Pteropus alecto, Pa-DR; Myotis lucifugus, Ml-DR) were previously described ¹⁴. cDNA sequences of the following MHC-II genes were synthesized (ThermoFisher Scientific) and cloned into pmCherry-C1 (Clontech, catalog no. 632524) using the restriction sites EcoRI and NheI, replacing the mCherry-encoding sequence: A. fuligula (tufted duck) HLA class II histocompatibility antigen, DR alpha chain-like (AfDRA, XM_032205226.1), A. fuligula class II histocompatibility antigen, B-L beta chain-like (AfDRB, XM_032205134.1), A. platyrhynchos (mallard) HLA class II histocompatibility antigen, DR alpha chain (ApDRA, NM_001310349.1), A. platyrhynchos (mallard) class II histocompatibility antigen, B-L beta chain (ApDRB, NM_001310815.1), C. atratus (black swan) HLA class II histocompatibility antigen, DR alpha chain (CaDRA, XM_035572397.1), C. atratus (black swan) class II histocompatibility antigen, B-L beta chain (CaDRB, XM_035572399.1), S. hondurensis (Honduran yellow-shouldered bat) HLA class II histocompatibility antigen, DR alpha chain (Sh-DRA, XM_037063365.1), S. hondurensis (Honduran yellow-shouldered bat) DLA class II histocompatibility antigen, DR-1 beta chain-like, transcript variant X1 (Sh-DRB, XM_037063363.1), A. jamaicensis (Jamaican fruit bat) HLA class II histocompatibility antigen, DR alpha chain-like (Aj-DRA, XM_037155244.1), A. jamaicensis (Jamaican fruit bat) DLA class II histocompatibility antigen, DR-1 beta chain-like (Aj-DRB, XM_037155243.1), Z. californianus (sea lion) HLA class II histocompatibility antigen, DR alpha chain (Zc-DRA, XM_035728695.1), Z. californianus (sea lion) MHC-II antigen (Zc-DRB, AY491464.1), M. p. furo (ferret) HLA class II histocompatibility antigen, DR alpha chain (Mpf-DRA, XM_045074660.1), M. p. furo (ferret) DLA class II histocompatibility antigen, DR-1 beta chain (Mpf-DRB, XM_045074661.1). For transfection, HEK293T cells were seeded onto poly-L-lysine-coated 24-well plates and transfected with 1 μg of pmCherry-C1 as control, or co-transfected with plasmids encoding alpha- and beta-chains of the above-mentioned MHC-II proteins using Lipofectamine2000 (ThermoFischer Scientific) at a DNA to Lipofectamine2000 ratio of 1:2. At 48h post transfection, cells were inoculated with BlaM1-VLPs and entry-positive cells were determined as described above.

Cloning of chimeric H2-E/HLA-DR and HLA-DRA mutants

Expression plasmids encoding chimeric alpha chains (chA1, chA2), chimeric beta chains (chB1, chB2), or HLA-DR alpha chain mutants were cloned into pmCherry-C1 (Clontech, catalog no. 632524) by In-Fusion Snap Assembly cloning (Takara, catalog no. 638948) using the primers indicated in Table S5. Expression plasmids encoding alpha and beta chains of HLA-DR or H2-E were used to amplify corresponding alpha-1, alpha-2, beta-1, and beta-2 domains containing 15 nucleotides upstream or downstream overlapping with the to-be-combined alpha or beta chain from the other species. All chimeric constructs were cloned to contain the signal peptides and the cytoplasmic tails from the alpha or beta chain of HLA-DR. HLA-DR, H2-E and Af-DR alpha chain mutants were cloned with the primers indicated in Table S5. Plasmids encoding mutant, chimeric or wildtype versions of HLA-DRA, HLA-DRB, H2-EA and H2-EB were co-transfected into HEK293T cells as described above. At 48 h post transfection, cells were analyzed for surface expression of HLA-DR or H2-E by flow cytometry or inoculated with BlaM1-VLPs as described above.

VSV∆G/G*/ H1N1*/ H18*/ H19*-EGFP-luciferase generation and infection

To generate single-cycle VSV-pseudotypes expressing EGFP and firefly luciferase, the VSV-G ORF in pVSV1XN-EGFP ⁵⁵ was replaced by that of firefly luciferase using restriction enzymes MluI and NheI. Recombinant rVSV∆G/G*-EGFP-luciferase was recovered as previously described ⁴⁷. Briefly, BHK-21 cells were inoculated with recombinant vaccinia virus expressing bacteriophage T7 RNA polymerase (MVA-T7), and subsequently co-transfected with T7-expression plasmids encoding VSV-N, -P, -L, -G (Kerafast, catalog no. EH1012) and the antigenomic VSV∆G genome encoding EGFP and firefly luciferase. At 48 h post transfection, supernatants containing rVSV∆G/G*-EGFP-luciferase were collected and titrated on BHK-21 cells by counting EGFP-positive cells at 24 h post infection (p.i.). rVSV∆G/G*/ H1N1*/ H18*/ H19*-EGFP-luciferase generation and infection

To amplify primary recovered rVSV∆G/G*-EGFP-luciferase, HEK293T cells were seeded onto 6-well plates and transfected with 2 μg of pCAGGS-VSV-G using ViaFect (Promega) at a DNA: ViaFect ratio of 1:3. The next day, transfected cells were inoculated at an MOI of 0.1 for 90 min at 37 °C. Inoculum was then replaced with 2 mL of OptiMEM and incubated at 37 °C for another 48 h. Once, more than 90% of the cells were EGFP positive, the supernatant containing amplified rVSV∆G/G*-EGFP-luciferase were collected, titrated, and used as a working stock to generate rVSV∆G-EGFP-luciferase pseudotyped with H1N1 (A/WSN/33), H18 (A/bat/Peru/2010) or H19 (A/lesser scaup/California/3087/2010).

To generate VSV∆G/H1N1*/ H18*/ H19*-EGFP-luciferase, HEK293T cells were seeded onto 6-well plates and co-transfected with 1 μg of pCAGGS-A/WSN/33-H1 and 1 μg of pCAGGS-A/WSN/33-N1 per well, or transfected with 2 μg of pCAGGS-A/bat/Peru/033/10-H18 or 2 μg of pCAGGS-A/lesser scaup/California/3087/2010-H19 using ViaFect (Promega) at a DNA: ViaFect ratio of 1:3. The next day, transfected cells were inoculated with rVSV∆G/G*-EGFP-luciferase at an MOI of 3 for 90 min at 37 °C. The inoculum was then replaced with 2 mL of OptiMEM containing 1 μg/mL anti-VSV-G antibody (Kerafast, catalog no. EB0010) and incubated at 37 °C. At 24 h p.i. supernatants were collected and clarified by centrifugation at 3000 rpm for 5 min. Pseudotypes were treated with TPCK-treated trypsin (Sigma-Aldrich, catalog no. T-8802) for 20 min at 37 °C for HA cleavage followed by trypsin inactivation with 10 μg/mL trypsin inhibitor from soybean (ThermoFisher Scientific, catalog no. 17075029) for 20 min at 37 °C.

For infection with VSV∆G/H1N1*/ H18*/ H19*-EGFP-luciferase, MDCK LV-ctrl and LV-H2-E cells were inoculated in 96-well plates with 50 μl of the corresponding VSV-pseudotypes in the presence of 0.1 μg/mL DEAE-dextran hydrochloride and 2% FBS for 90 min at 37 °C. The inoculum was replaced with supplemented DMEM, and cells were incubated at 37 °C. At 24 h p.i., Cells were fixed with 3.7% (v/v) paraformaldehyde (PFA) for 15 min and nuclei were stained with DAPI. EGFP-positive cells were analyzed using an EVOS M5000 (ThermoFisher Scientifc). Luciferase signals from unfixed samples were measured using the ONE-Glo Luciferase Assay System (Promega) using a plate reader (BioTek, Synergy Neo2). For sialidase treatment, the cells were incubated with 200mU/mL of bacterial neuraminidase (Sigma-Aldrich, catalog no. N6514-1UN or catalog no. 11080725001) for 90 min at 37 °C prior to infection as described above.

Lectin staining

MDCK LV-ctrl or LV-H2-E cells were seeded on 24-well plates at 100’000 cells per well. The next day cells were washed once with PBS and treated with 200mU/mL bacterial neuraminidase (Sigma-Aldrich, catalog no. N6514-1UN or catalog no. 11080725001) diluted in non-supplemented DMEM for 90 min at 37 °C. After detachment of cells with 0.25% trypsin-EDTA (Corning), cells were resuspended in supplemented DMEM and centrifuged at 1200 rpm for 5 min at 4 °C. Prior to staining, cell pellets were washed with Dulbecco's phosphate-buffered saline with Calcium and Magnesium (DPBS, ThermoFisher Scientific, catalog no. 14040117) supplemented with 2% BSA (ThermoFisher Scientific, catalog no. 50-753-3053) and 1mM EDTA (ThermoFisher, catalog no. 15575020) by resuspension and centrifugation at 1200 rpm for 5 min. The same buffer was used throughout the lectin staining protocol. Cells were incubated with biotinylated Sambucus Nigra Lectin (SNA, Vector Labs, catalog no. B-1305-2) or biotinylated Maackia Amurensis Lectin II (MALII, catalog no. B-1265-1) at 2 μg/mL or 1 μg/mL, respectively, for 1 h at 4 °C. After washing the cells twice by centrifugation and resuspension, cells were incubated with FITC-labelled streptavidin (ThermoFisher Scientific, catalog no. SA1001) diluted 1:500 for 1 h at 4 °C. LIVE/DEAD^™ Fixable Near-IR Dead Cell Stain Kit (ThermoFisher Scientific) was used to stain dead cells. After cells were washed three times, FITC levels were measured by flow cytometry using a BD FACS Canto II instrument. 10’000 live cells were acquired and median fluorescence intensities (MFI) from FITC signals were quantified using FlowJo v10 software. MFIs from FITC signals in sialidase treated samples were normalized to those from untreated samples.

Blocking assay with anti-H2-E antibody

MDCK LV-H2-E cells were incubated with increasing concentrations of anti-mouse I-A/I-E antibody (Biolegend, catalog no. 107602) or anti-6xHis antibody (ThermoFisher Scientific, catalog no. MA1-21315) for 1 h at 4 °C prior to infection. Cells were inoculated with rVSV∆G/H19*-EGFP-luciferase in the presence of increasing concentrations of either anti-H2-E or anti-6xHis antibodies for 90 min at 37 °C. Cells were washed once with PBS and incubated in supplemented DMEM at 37 °C. At 24 h p.i., luminescence signals as a proxy for VLP entry were measured using the ONE-Glo Luciferase Assay System (Promega).

Virus rescues

cDNAs encoding PB2 (CY157045.1), PB1 (CY157044.1), PA (CY157043.1), HA (OR611720), NP (CY157041.1), M (CY157040.1) and NS (CY157042.1) segments from A/lesser scaup/California/3087/2010 (CY180462.1), and NA from A/duck/Minnesota/104/1974(H4N6) were synthesized (Azenta Life Sciences) and cloned by In-Fusion Snap Assembly cloning (Takara, catalog no. 638948) into the rescue vector pDZ ⁵⁶, which was previously linearized with the restriction enzyme SapI. PA, PB1 and PB2 segments were cloned using forward primer 5’-CCGAAGTTGGGGGGGAGCGAAAGCAGG-3’ and reverse primer 5’-GGCCGCCGGGTTATTAGTAGAAACAAGG-3’, while HA, NP, M, NA and NS segments were cloned using forward primer 5’-CCGAAGTTGGGGGGGAGCAAAAGCAGG-3’ and the same reverse primer as above. Italicized and underlined letters indicate nucleotides complementary to the overhangs of the SapI-digested pDZ-vector and the conserved IAV segments non-coding region (NCR), respectively. The N6 segment with the premature stop codon (pDZ-N6stop) was cloned by In-Fusion Snap Assembly cloning with the primers as above and mutagenesis primers (N6-G290A-frw: 5’- GTCCTTAGATAGAATGTGCTATGAGTTTACTTCACACAAGG-3’, N6-G290A-rev: 5’- CCTTGTGTGAAGTAAACTCATAGCACATTCTATCTAAGGAC-3’), leading to truncated N6 protein of 96 amino acids. H19N6, H19N6stop with the internal segments from A/lesser scaup/California/3087/2010 and the corresponding 6+2 viruses containing the internal segments from A/Puerto Rico/8/34 (H1N1) were essentially rescued as previously described ⁵⁷. Briefly, HEK293T cells were transfected with eight rescue plasmids encoding the individual IAV genomic segments with FuGENE^® HD transfection reagent (Promega, catalog no. E2311) at a DNA to transfection reagent ratio of 1:3. 24 h post transfection MDCK LV-H2-E cells were co-cultured with the transfected HEK293T cells in the presence of 0.5 μg/mL TPCK-treated trypsin for another 48 h, until cytopathic effects were observed. Newly generated viruses were collected, plaque-purified on MDCK LV-H2-E cells and sequence-verified.

Virus inoculation

For virus growth curves, 120‘000 cells were seeded in 24-well plates and inoculated with the indicated viruses at the indicated MOI, which is defined as the ratio between infectious virions and the number of inoculated cells, in PBS supplemented with 0.3% bovine serum albumin (BSA), 1 mM Ca²⁺/Mg²⁺, Pen-Strep (100 U/mL penicillin and 100 μg/mL streptomycin, Corning). After incubation for 1 h at 37 °C, virus inoculum was replaced with OptiMEM (ThemoFisher) with 0.5 μg/mL TPCK-treated trypsin (Sigma-Aldrich, catalog no. T-8802). At the indicated times post infection (p.i), virus titer in the supernatant was determined by standard plaque assay on MDCK LV-H2-E cells.

NA-activity assay

The NA-Star^™ Influenza Neuraminidase Inhibitor Resistance Detection Kit (ThermoFisher Scientific, catalog no. 374422) was used to confirm the absence of NA activity in the recombinant H19N6stop and H19N6stop (6+2) viruses as per the manufacturer’s instruction. Briefly, viruses were serially diluted 2-fold in assay buffer in duplicates in a 96-well plate and incubated for 15 min at RT. 10 μL of NA-Star substrate was added to each well containing 50 μL of diluted virus and incubated for 20 min at RT. After the addition of 60 μL of NA-Star Accelerator per well, luminescence signals were measured using a plate reader (BioTek, Synergy Neo2).

Cloning of H19 mutants

H19 mutants were cloned (Clontech) by In-Fusion Snap Assembly cloning (Takara, catalog no. 638948) using pCAGGS-A/lesser scaup/California/1742/2013-H19 with the primers indicated in Table S5 into pCAGGS previously digested with NotI and XhoI. Once single mutants were generated, H19 mutants with multiple mutations were generated by sequentially introducing additional mutations. Expression plasmids encoding H19 mutants were co-transfected with pCAGGS-A/WSN/33-N1 to generate VSV-pseudotypes encoding luciferase and EGFP as described above.

Expression and purification of recombinant HA proteins

Ectodomains from A/lesser scaup/California/3087/2010 H19 (Met 1-Ileu 525), A/Viet Nam/1203/2004 (HALo) H5 (Met 1-Ileu 528) and A/Hong Kong/1/1968 H3 (Met 1-Ileu 567) were cloned by In-Fusion Snap Assembly cloning (Takara, catalog no. 638948) into pCAGGS containing a T4 trimerization domain with an upstream thrombin cleavage site and a downstream 6xHis-tag ⁵⁸ previously digested with restriction enzymes EcoRI and NotI using primers rH19 frw 5’-TTTGGCAAAGAATTCGCCACCATGTGGAAACTCG-3’, rH19 rev 5’-TGGGACCAAGCGGCCGCTGATCTTGTAAACTGTA-3’, rH5 frw 5’-TTTGGCAAAGAATTCGCCACCATGGAGAAAATAGTG-3’, rH5 rev 5’-TGGGACCAAGCGGCCGCTTATTTGGTAAATTCC-3’, rH3 frw 5’-TTTGGCAAAGAATTCGCCACCATGAAGACCATCATTG-3’ and rH3 rev 5’-TGGGACCAAGCGGCCGCTGTCTTTGTATCCAGAC-3’. Italicized and underlined letters indicate nucleotides complementary to the overhangs of the digested vector and the cds of the corresponding HA, respectively. Recombinant HA proteins (rHA) were produced in Expi293F cells (ThermoFisher, catalog no. A14527) or Expi293F GnTI-cells (ThermoFisher, catalog no. A39240) and purified using nickel-nitrilotriacetic acid (Ni-NTA) columns ⁵⁹. To verify the correct size and susceptibility to trypsin digestion, 5 μg of rHA were treated with 2 μg/mL of TPCK-treated trypsin for 2 min at 37 °C or left untreated and analyzed on a reducing sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) by Coomassie staining. The open reading frame of H3 A/Switzerland/9715293/2013 was codon optimized by Genscript and cloned into the pCD5 expression vector in frame with a GCN4 trimerization motif (KQIEDKIEEIESKQKKIENEIARIKK), a superfolder GFP and the Twin-Strep-tag (WSHPQFEKGGGSGGGSWSHPQFEK); IBA, Germany) ^23,60,61. The trimeric HAs were expressed in HEK293S GnTI(-) cells and purified with sepharose strep-tactin beads (IBA Life Sciences, Germany) according to the manufacturer’s instructions.

Hemagglutination assay

25 μg of rHA were serially diluted 2-fold in PBS and incubated with 0.5% chicken or turkey red blood cells (RBC) in a total volume of 100 μL/well for 30 min at 4 °C. Hemagglutination of RBC was visually assessed.

Glycan array

Glycan array binding analysis of the rHAs was carried out as described here ^23,24. Briefly, recombinant hexahistidine-tagged HA was precomplexed with a mouse anti-his Alexa 647 (ThermoFisher Scientific, catalog no. MA1-21315-A647) or human anti streptag (IBA, catalog no. 2-1507-001), and goat-anti-mouse Alexa 647 antibodies (Invitrogen, catalog no. A21235) or goat-anti-human Alexa 647 (ThermoFisher Scientific, catalog no. A-21445). This was done in 50 μl PBS-T (phosphate-buffered saline with 0.1% Tween-20) in a 4:2:1 molar ratio, incubated for 15 min on ice, and the applied on the array for 90 min. Following multiple washes with PBS-T, PBS, and deionized water the arrays were scanned to detect HA binding using an Innoscan 710 (Innopsys). Mean relative fluorescence unit (RFU) and standard deviation values were imported into Prism 7.0 and the corresponding graph was generated. Viruses were incubated overnight at 4°C in the presence of 200 nm Oseltamivir and stained using a human CR9114 anti-HA-stem antibody and finally detected with goat anti-human (ThermoFisher Scientific, catalog no. A-21433) Alexa-555 conjugated antibodies.

Binding assay

MDCK LV-H2-E or LV-ctrl were detached with 0.25% trypsin-EDTA (Corning) and resuspended in DMEM supplemented with 10% FBS and Pen-Strep (100 U/mL penicillin and 100 μg/mL streptomycin, Corning). Cells were then washed once with FACS buffer (Biolegend, catalog no. 420201) and incubated with 5 μg of rHA for 1 h at 4 °C. Cells were washed twice by resuspension with FACS buffer and centrifugation at 1200 rpm for 3 min. To stain for cell-bound rHA, cells were incubated with mouse anti-his Alexa 647 (ThermoFisher Scientific, MA1-21315-A647) for 30 min at 4 °C. LIVE/DEAD^™ Fixable Near-IR Dead Cell Stain Kit (ThermoFisher Scientific) was used to stain dead cells. After two more washes with FACS buffer, samples were analyzed on a Beckman Coulter Gallios Flow Cytometer. Binding of rHA was quantified by gating on Alexa-647-positive cells using FlowJo v10 software.

Crystallization and structure determination of H19

Prior to crystallization, purified H19 ectodomain was treated with thrombin to remove the foldon domain and 6xHis-tag. The cleaved trimeric H19 ectodomain was then passed through a Ni-NTA column to remove the cleaved 6xHis-tag and foldon domain. The flowthrough fraction, containing the cleaved H19 ectodomain, was further purified by size exclusion chromatography using a Superose 6 Increase 10/300 GL column (Cytiva, catalog no. 29091596) in a buffer containing 20 mM Tris pH 8.0 and 100 mM NaCl. Crystallization trials were conducted at a protein concentration of 15 mg/mL. Initial screens were conducted with an Oryx Nano crystallization robot (Douglas Instruments) at 20 °C in a sitting drop format, with 0.25 μl of protein mixed with an equal volume of reservoir solution. Initial crystals were observed in solutions containing 15% polyethylene glycol (PEG) 20,000 and 0.1 M 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) pH 7.0 after 2 days. The crystals were further optimized by varying the concentration of PEG 20,000 and the pH of the buffer in a hanging drop format, with 1 μl of protein mixed with 1 μl of reservoir solution. The crystals were cryoprotected in a stepwise manner using a reservoir solution containing 5–25% glycerol and then flash-cooled in liquid nitrogen. X-ray diffraction data were collected at the National Synchrotron Light Source (NSLS)-II 17-ID-2 beamline at the Brookhaven National Laboratory (BNL) under cryogenic conditions. The diffraction data were processed using the autoPROC pipeline available at the NSLS-II 17-ID beamline. The structure was solved by molecular replacement with Phaser-MR using the structure from A/black-headed gull/Sweden/2/99(H16N3) (PDB code 4F23) as a search model. Subsequent iterative manual building and refinement were performed using Coot and Phenix, respectively. All molecular graphics figures were prepared using PyMOL (Schrödinger). The atomic coordinate and structure factor has been deposited in the Protein Data Bank under the accession code 8VCC.

QUANTIFICATION AND STATISTICAL ANALYSIS

Quantifications and statistics were performed as described in the figure legends and methods section. Statistical significance was determined using GraphPad Prism software. Unless indicated differently, data was collected from n=3 biologically independent experiments and is shown as mean ± SD.

Highlights.

• Influenza A virus H19 subtype does not bind the canonical receptor sialic acid.

• H19 utilizes MHC class II proteins from distinct species for host cell entry.

• Mutating the putative canonical RBS of H19 restores sialic acid-dependency.

• Residues in the MHC class II α-chain confer species-specific entry of H19.