neuraminidase

Abstract

The invention relates to a polypeptide comprising a neuraminidase head group, therapeutic uses thereof, and methods of producing said polypeptide.

Description

[0001] NEURAMINIDASE

[0002] Field of the Invention

[0003] The present invention relates to a polypeptide comprising a neuraminidase head group with improved properties. The invention also relates to the therapeutic uses of the polypeptide, for example, as a vaccine. The invention also relates to methods for producing the neuraminidase head group.

[0004] Background to the Invention

[0005] Influenza neuraminidase (NA) is an important target for protective antibodies and therapeutic drugs against influenza viruses. Currently, influenza subunit vaccines focus on haemagglutinin and contain minimal amounts of NA, despite emerging evidence from multiple studies indicating NA's importance as an independent protective antigen. However, recombinant NA protein expression is often bedevilled by low yield and poor stability, with substantial variation in NA protein yields across different virus strains. We found that NA yield varied from 0.1 to 30 mg / L in different expression systems found in the literature (Figure 8). Non-mammalian expression systems, such as Pichia pastoris, and insect cells that have been used previously, have their own disadvantages ranging from instability, post-translational modifications, limited scalability for complex proteins, and lack of human-like glycosylation.

[0006] Purified NA has been safely administered to humans as a vaccine candidate and NAI antibodies to NA have been shown to be an independent correlate of protection against influenza infection. Enhancing current vaccines by addition of NA proteins could substantially boost their effectiveness, particularly among high-risk elderly, immunocompromised groups, and potential pandemic influenza viruses. A major challenge has been the production of NA at optimal yields. It is an aim of the invention to produce stable and immunogenic NA protein at high yields to address this critical issue.

[0007] Summary of the Invention

[0008] The present inventors surprisingly found that it is possible to combine the antigenic features of a poorly expressed NA (e.g. from a circulating influenza) with the expression and stability properties of a high expressing NA by grafting the L01 and L23 loops from the former onto the structural “scaffold” of the latter. The present invention therefore provides hybrid proteins with the desired antigenic properties that have greatly improved expression. These hybrids assemble into stable tetramers, retain enzymatic activity, and are inhibited by standard sialic acid analogue inhibitor drugs. The loop-grafting concept is straightforward and may be broadly applicable to all NA subtypes.

[0009] The present inventors surprisingly found that the structures formed by the grafted loops in the context of their new scaffold were recognized by a collection of human mAbs raised to the donors of the NA surface loops, and crystal structures showed the loops of donor and recipient to be essentially superposable. Mice immunised with the NA hybrids elicited inhibitory antibodies to the NA loop donor.

[0010] The invention is thus based on three simple principles: 1) the majority of protective epitopes for the antibody response are present in the L01 and L23 loops on the top surface of the NA head as defined by Varghese and Colman 1983, 2) the stability of the NA tetramer depends mainly on interactions between monomers encoded in the remainder of the NA sequence (the “scaffold”), and 3) the structure of the loops will be maintained when grafted from one scaffold to another. Loop grafting should be useful for NA-based vaccine design and manufacturing, as well as drug and antibody screening.

[0011] Thus, the present invention provides a polypeptide comprising the head region of a first neuraminidase, wherein the L01 and L23 loops are substituted with the L01 and L23 loops of a second neuraminidase.

[0012] The invention also provides a tetramer comprising four polypeptides that each comprises a neuraminidase head region, wherein at least one of the polypeptides is a polypeptide of the invention.

[0013] The invention further provides a polynucleotide encoding the polypeptide of the invention, or one or more polynucleotides encoding the tetramer of the invention.

[0014] The invention also provides a particle comprising the polynucleotide of the invention, wherein the polynucleotide is an RNA molecule, and a pharmaceutically acceptable excipient or carrier.

[0015] The invention also provides a virus comprising the polypeptide or the tetramer of the invention, or a virus-like particle (VLP) comprising the polypeptide or the tetramer of the invention. The invention further provides a pharmaceutical composition comprising the polypeptide, the tetramer, the polynucleotide, the particle, the virus, or the VLP of the invention.

[0016] The invention also provides a method of producing a chimeric protein, the method comprising aligning the amino acid sequence of the head region of a first neuraminidase and a second neuraminidase to identify the L01 and L23 loops, and substituting the L01 and L23 loops of the first neuraminidase with the L01 and L23 loops of the second neuraminidase.

[0017] Brief Description of the Figures

[0018] Fig. 1: Surface antigenic loops transfer method to design NA-hybrid protein, a) NA Anatomy. Each monomer of the tetrameric NA head is composed of a six -bladed propeller-like structure. Each blade unit consists of a P-sheet composed of four P-strands connected by loops arranged in a W-shape. The loop preceding P-strand 1 is termed LoopOl and the loop connecting P-strands 2 and 3 is termed Loop23. The twelve 01 and 23 loops in each monomer point up and surround and contribute to the enzyme active site, b) Antigenic surface formation. The key antigenic surface on the top of each NA monomer is formed by the twelve 01 and 23 loops from the six P-sheet units and the C-terminal domain. The twelve loops and the C-terminal domain in one monomer are shaded to show the top surface surrounding the active site. The remaining four Beta strands and the Loops 12 and 34 we refer to as the “scaffold”. Oseltamivir is shown in the sialic acid receptor binding site. A calcium ion at its binding site is shown as a circle, c) Hybrid NA design. Loops 01 and 23 and the C-terminal domain, and the scaffold are shown separately. Oseltamivir and the binding residues are included for reference. The surface antigenic loops 01 and 23 of an NA of interest were transferred on to the scaffold of a high- expressing NA candidate to improve the protein yield and stability. Figures were generated with PDB 4B7J using UCSF ChimeraX.

[0019] Fig 2: NA protein sequence alignment highlighting surface loops and active site, a) Protein sequence alignment of the scaffold of N1 from H5N1 A / mute swan / England / 053054 / 2021 (mS), and loop donors H1N1 A / Califomia / 7 / 2009 (Nl / 09) and A / Wisconsin / 588 / 2019 (Nl / 19) neuraminidases. The N2 sequence H2N2 A / Tokyo / 3 / 1967 from Varghese et. al (1983, 1991), that was used to define the loops (PDB 1NN2) is also included. mSNl is used as a reference sequence and identical residues are shown as dots. The sequence conservation is shown by bars. The numbering here is based on N1 numbering as used in the annual Crick Reports (https: / / www.crick.ac.uk / sites / default / files / 2022- 10 / Crick%20report%20Sep2022%20for%20SH2023_to%20post.pdf). Loops 01 and 23 that form the top antigenic surface are highlighted. Loops annotation is based on Varghese et al. (see Table 1). Residues that form the catalytic site (8 residues) and framework of the enzymatic cavity (11 residues) are annotated with bars. These residues are highly conserved between NA subtypes and the majority are part of the surface loops - 7 / 8 catalytic residues and 8 / 11 scaffold residues. Two scaffold residues are at the edge of loop B2L01. The figure was generated using Geneious Prime, b) Number of amino acid differences in the loops and scaffolds of loop donors Nl / 09 and Nl / 19 and the mS N1 recipient are shown. Twelve dissimilar residues within Loops 01 and 23 of Nl / 09 were transferred to mSNl scaffold to form Nl / 09 hybrid. Similarly, 16 dissimilar residues within Loops 01 and 23 of Nl / 19 were transferred to mS N1 scaffold to form Nl / 19 hybrid.

[0020] Fig 3: Characteristics of NA hybrid proteins including epitope specificity. a) Characteristics of NA proteins. NA proteins were expressed in a transient mammalian ExpiCHO expression system. In principle NA proteins could be expressed in any viable expression system for establishing the donor and recipient of the L01 and L23 Loops. The optimal scaffold donor may be different in different expression systems. Gene constructs including affinity purification tags, SpyTag and an artificial tetramerisation domain tetrabrachion at N-terminus were cloned in the pcDNA3.1- vector. Proteins were purified using the 6His tag and Nickel-sepharose HisTrap purification columns. See Fig. 9 for constructs design and SDS-PAGE of purified proteins. Nl / 09 hybrid means residues of loops 01 and 23 of Nl / 2009 grafted to mSNl (H5N1 / 2021) scaffold. Nl / 19 hybrid means loops 01 and 23 of Nl / 2019 grafted to mSNl (H5N1 / 2021) scaffold. The expression yield of NA proteins and their hybrid forms is included in the 2ndcolumn. Nl / 19 protein didn’t express in ExpiCHO cells in two instances. Size-exclusion chromatography graphs are rd shown in the 3 column. Elution volume of 10-14 mL indicates tetrameric form of the protein. MUNANA and ELLA activity of the NA proteins are in 4* and 5* columns and the nanoDSF thermal melting temperature is in final column. The sharp narrow peak and the higher melting temperature indicate higher protein stability, b) Epitope specificity has been transferred with the loops. Human monoclonal antibodies, previously published and some new, were titrated for ELISA binding of NA proteins. Areas under the curve were ranked after normalisation with one of the strongest binding mAb (see Fig. 8) ‘+++’ denotes >70% binding, ‘++’ 40-70% binding, ‘+’ 10-40%, and <10% as a non-binder. Loops cross-reactive mAbs AG7C, AF9C, Z2B3 and 1G01, all defined by crystal structures, show full binding to all NA proteins. Seven mAbs (NmAb) do not bind mSNl but bind to Nl / 09, Nl / 19 and their hybrid forms. NmAb-03 is specific to Nl / 19 surface Loops. mAb CD6 is a scaffold dependent mAb that shows binding to Nl / 19 hybrid protein but does not bind the Nl / 19 protein on the cell surface. mAb Z2C2 is specific for mSNl and Nl / 09, hence did not bind Nl / 19 or its hybrid form.

[0021] Fig 4: The fold of the scaffold neuraminidase is strictly conserved in the loops- grafted hybrid proteins, a, X-ray crystal structures of the protomers of H5N1 / 21 (mSNl), Nl / 09Loops-mS Scaffold (Nl / 09 hybrid), and Nl / 19Loops-mS Scaffold (Nl / 19 hybrid) are shown in cartoon representation. In the mSNl protein, six L01 and six L23 loops are labelled. The C-terminal domain (CTD) and the calcium ion are labelled. Eight highly conserved residues in the catalytic site (R118, D151, R152, R225, E277, R293, R368, and Y402) surrounding oseltamivir are shown as sticks. Grafted L01 and L23 residues in the Nl / 09 and Nl / 19 hybrid proteins are shown as sticks. Residues that are the same in Nl / 09 and Nl / 19 hybrids are labelled, b, Nl / 09 and Nl / 19 hybrid structures in putty tube representation with colour and radius reflecting local RMS deviation values (generally <0.5 A) throughout the structures, with modestly higher values in the inherently flexible B1L23 and B6L23 loops, c, Structures in surface representation displaying the active site cavities. The msNl structures revealed heterogeneity in the active site conformation. Within a tetramer, one monomer showed a wider cavity (i) and another showed a narrower cavity (ii) due to different trajectories of the B1L23 loop (150-loop). The Y344N in Nl / 09 and Nl / 19 hybrids, compared to msNl, caused a minor, local widening of the rim of the cavity

[0022] Fig 5: NA-hybrid proteins are immunogenic and provide in vivo protection against virus challenge, a) Immunogenicity of NA hybrid proteins. BALB / c mice (n=6 / group) were immunised with 0.5 pg NA coupled to the mi3 virus-like particles (NA- VLP) adjuvanted with 1 : 1 vol / vol Addavax™ (squalene-based oil-in-water nano- emulsion). Intramuscular immunisations were done twice at the interval of three weeks and sera were harvested three weeks post booster dose to assess the antibody response. Neuraminidase activity inhibition (NAI) IC50 titres measured using fetuin-based Enzyme- Linked Lectin Assay (ELLA) are shown as a separate dot for each mouse. mAb AG7C was used as a positive control. Pooled sera to unconjugated mi3 VLP was negative control and showed no inhibition at 1 :40 dilution (not included here). Geometric mean with 95% confidence interval is shown, b) Protection against virus challenge. DBA / 2 mice (n=6 / group) were immunised as above and were challenged with intranasally administered 200 LD50 of H1N1 / 2009 (X-179A) virus. Weights were monitored for two weeks. Loss of >20% initial weight was considered an endpoint. mAb AG7C (10 mg / Kg prophylaxis) was used as a positive control. Empty VLP pooled sera was a negative control and mice reached the endpoint within day 5-7 post virus infection. The ELLA NA inhibition graphs of these pooled sera are shown in Fig 9. Figures were made using GraphPad Prism vlO. Kruskal-Wallis test was used for statistical analysis, ns: non-significant (p-value>0.05), ** means p-value<0.005. Kaplan-Meier survival analysis was done with Logrank Mantel -Cox test for comparison.

[0023] Fig 6: Loop grafting between two distant N1 NAs: H5N1 A / mute swan / England / 053054 / 2021 (mS) and H1N1 A / PR / 8 / 1934 (PR8). a) Number of amino acid differences between mSNl and PR8 N1 and their loop grafted hybrids are shown. Eighteen dissimilar residues (5% of total head region residues) within Loops 01 and 23 were grafted to make the hybrid proteins, b) NA proteins were expressed in a transient mammalian ExpiCHO expression system. The expression yield of NA proteins and their hybrid forms is included in the 2ndcolumn. Size-exclusion chromatography graphs are in the 3rdcolumn. Elution volume of 10-14 mL indicates tetrameric form of the protein and 14-15 ml indicates trimeric or dimeric nature of the protein. MUNANA and ELLA activity of the NA proteins are in 4thcolumn and the nanoDSF thermal melting temperature is in the final column. The sharp narrow peak and the higher melting temperature indicate the higher protein stability, c) Epitope specificity has been transferred with Loops with a few exceptions. Antibodies, previously published and some new, were titrated for ELISA binding of NA proteins. Area under curve was ranked after normalisation with one of the strongest binding mAb (see Fig 7a) ‘+++’ denotes >70% binding, ‘++’ 40-70% binding, ‘+’ 10-40%, and <10% as a non-binder. Loops cross-reactive mAbs recognised both mSNl and PR8N1 and their loops grafted proteins. PR8N1 Loops specificity is shown by NmAb-03. mAb CD6 is a major scaffold dependent mAb and showed binding to PR8Loops-mS but not the PR8 and mSLoops-PR8 (see Fig. 7b). Similarly, NmAb-20 is a mSNl Loops + scaffold dependent mAb. mSNl Loops specificity is shown by NmAb-02 and Nm Ab-13.

[0024] Fig 7: NA hybrid proteins elicited loop-specific NA inhibiting antibodies and provided loop-specific in vivo protection against virus challenge. a,b,c, Immunogenicity of NA hybrid proteins. BALB / c mice (n=6 / group) were immunised with 0.5 pg NA coupled on to the mi3 virus-like particles (NA-VLP) adjuvanted with 1 : 1 vol / vol Addavax™ (squalene-based oil-in-water nano-emulsion). Intramuscular immunisations were done twice at the interval of three weeks and sera were harvested three weeks post booster dose to assess the antibody response. Neuraminidase activity inhibition (NAI) IC50 titres measured using fetuin-based Enzyme-Linked Lectin Assay (ELLA) are shown as a separate dot for each mouse. mAb AG7C was used as a positive control. Empty VLP pool sera were negative controls and showed no inhibition at 1 :40 dilution (not included here). Geometric mean with 95% confidence interval is shown. d,e,f,g, Protection against virus challenge. BALB / c mice (n=6 / group) were immunised as above. Pooled sera antibodies were assessed in ELLA assay before virus challenge (e) and IC50 values shown are sera reciprocal dilution. Mice were challenged with intranasally administered 1000 LD50 of PR8 virus (Cambridge strain, 104TCID50). Weights were monitored for two weeks. Loss of >20% initial weight was considered an endpoint. mAb AG7C (10 mg / Kg prophylaxis) was used as a positive control. Empty VLP pool sera were negative controls and mice reached the endpoint by day 6 post virus infection. Importantly, immunogens with PR8 Loops protected 100% mice from virus challenge and mS Loops did not. Mean and standard deviations are shown. Figures were made using GraphPad Prism vlO. Kruskal-Wallis test was used for statistical analysis, ns: non-significant (p- value>0.05), *** denotes p-value<0.0005. Kaplan-Meier survival analysis was used with Logrank Mantel-Cox test for comparison.

[0025] Fig 8: Literature study of recombinant NA expression in various cell expression systems.

[0026] Fig 9: a) Gene constructs for NA proteins. All the constructs are the same except for Nl / 09 which contains extra residues 69-81 in NA head and has no Strep II tag. b) NA and hybrid proteins were expressed as tetramers. Purified proteins on SDS-PAGE with Coomassie staining, in reducing condition and cross-linked before loading on the gel. 2 pg of NA protein was incubated with 6 mM BS3 cross-linking reagent for 30 min and the reaction was stopped by adding IM Tris-HCl pH 8.0. Proteins were separated on reducing 4-12% Bis-Tris SDS-PAGE and lx MES SDS buffer, c) Inhibition of enzyme activity of NA hybrid proteins by small molecule inhibitors and mAbs. Oseltamivir and Zanamivir inhibit the function of mSNl, Nl / 09 hybrid, and Nl / 19 hybrid proteins suggesting the active site is intact and functional. mAb 1G01 inhibited mSNl and Nl / 09 hybrid suggesting the cross-reactive epitopes recognised by 1G01 have been preserved. 1G01 fails to inhibit recognise Nl / 19 due to substitution at N222K (see figure 2 alignment) and this has been confirmed by the loss of inhibition of NA activity of Nl / 19 hybrid, d) Coupling of NA to mi3 virus-like particles (VLP) using SpyCatcher technology. SDS-PAGE showing the NA+mi3, NA and mi3. Two pg mi3 with SpyCatcher003 tag covalently linked to 5 pg NA (molar ratio 1 :0.4). These NA-VLPs were used for mouse immunisations.

[0027] Fig 10: Recognition of Nl / 09 NA protein expressed as tetramers with VASP or tetrabrachion tetramerisation domains. Note that CD6 did not bind to N1 / 09-VASP and but did bind to N1 / 09-TB relatively weakly compared to mAbs AG7C and 1G01 that (unlike CD6) bind within a single monomer.

[0028] Fig 11: Binding titration of mAbs against recombinant soluble proteins or NA expressed on the surface of virus infected cells a) Twenty-five anti-Nl mAbs were titrated for binding against mSNl, Nl / 09 and Nl / 19 and their loop transferred hybrid variants, b) Binding titration of eighteen mAbs that bind either PR8 N1 or mSNl on wild type sequence proteins and their loop-grafted hybrid proteins. Area under curve (AUC) was calculated and normalised against one of the strongest binders to rank the order.

[0029] Fig 12. Pooled sera from immunised mice (n=6) were titrated in ELLA assay. Naive sera and sera from mice immunised with empty VLP were used as negative controls. mAb AG7C was used as a positive assay control. H1N1 / 2009 virus = X-179A (A / Califomia / 07 / 2009); H1N1 / 2021 = A / Sydney / 5 / 2021 (Differs from Nl / 19 only by V453M in the C-terminal domain), mSNl = NA from H5N1 A / mute swan / England / 053054 / 2021. Figure 13: The selected N8 sequences chosen for expression in ExpiCHO cells without tetramerisation domain. Numbers on branches = substitutions per site.

[0030] Figure 14: Characterization of N8 neuraminidase (NA) proteins. The NA proteins were expressed using a transient ExpiCHO mammalian expression system. Gene constructs were cloned into the pcDNA3.1- vector, with construct design features summarized in the second column (6H = 6xHis tag; ST = SpyTag; TB = tetrabrachion tetramerization domain). Proteins were purified through the 6xHis affinity tag using Nickel-Sepharose HisTrap columns. The putative oligomeric state of each protein was evaluated by size-exclusion chromatography (SEC), in which 30 pg of protein were applied to an Agilent AdvanceBio SEC 300 A column. An elution time of ~10 min corresponded to the tetrameric form. Enzymatic activity was measured using ELLA and MUNANA for which ECso and Area under curve (AUC) values of titration curves are included. Protein stability was assessed by nanoDSF thermal melting analysis. A sharp, narrow SEC peak along with a higher melting temperature indicated high protein stability. ELLA = Enzyme-linked lectin assay; MUNANA = 20-(4-methylumbelliferyl)-a-D-N- acetylneuraminic acid assay.

[0031] Figure 15: ELLA inhibition of NA activity by mAbs. (A) and commercial NA inhibiting drugs (B). IC50 of inhibition titration curves ae shown: in ng / ml for mAbs and nM for drugs. msNl (clade 2.3.4.4b H5N1 NA protein) is included as a control.

[0032] Figure 16: NA-VLP Immunised mouse sera were tested for inhibition of NA proteins using ELLA. Anti NA cross-reactive mAb 1G01 is an assay control. Geometric mean of five biological replicates and 95% confidence internal are shown. Data were analysed by the non-parametric Kruskal -Wallis test followed by Dunn's multiple comparison tests in GraphPad Prism. * indicates p < 0.05; ** indicates p < 0.005; ns indicates not significant.

[0033] Figure 17: A,B) Sequence alignment of chN8, N8Henan, and loop-grafted N8Henan hybrids. A) chN8 is shown as the reference sequence with N8 numbering. B) msNl is shown as the reference sequence with msNl numbering. The top loops and the underside loop B5L34 are indicated. PDB 4WA3 is included since its structure was used as a guide for loop annotations forN8 NA. The loop annotation is labelled as v2.1 which differs slightly from the previous msNl annotations. C) Number of amino acid differences between the scaffold donor (chN8) and the loop donor (N8 Henan) is shown. The difference between N8Henan hybrid-1 and hybrid-2 is that, in hybrid-1, loop B5L34 is additionally grafted onto the chN8 scaffold along with loops L01 and L23. Furthermore, the sequence boundaries differ: hybrid- 1 spans residues 83-470, whereas hybrid-2 spans residues 80-470.

[0034] Figure 18: A,B) Sequence alignment of X-31 N2 NA, N2 / 2021 NA and loop- grafted NA hybrids. A) X-31 NA is shown as the reference sequence with N2 numbering. B) msNl is shown as the reference sequence with msNl numbering. The loop annotation is labelled as v3 which differs slightly from previous msNl annotations. C) Number of amino acid differences between the scaffold donor and the loop donors are shown.

[0035] Figure 19: N2 protein yields in ExpiCHO expression system.

[0036] Figure 20: Alignment of msNl with PDB 1NN2 highlighting differences in loop annotations. The msNl vl loop annotation is used as the reference. Loop definitions from Varghese and Colman (1983, 1991) are also shown for comparison.

[0037] Brief Description of the Sequence Listing

[0038] SEQ ID NO: 1 is a H7 HA signal sequence

[0039] SEQ ID NO: 2 is a IgKappa signal sequence

[0040] SEQ ID NO: 3 is a Strep Tag II

[0041] SEQ ID NO: 4 is a SpyTag

[0042] SEQ ID NO: 5 is a 6His purification tag

[0043] SEQ ID NO: 6 is a tetrabrachion tetramerisation domain

[0044] SEQ ID NO: 7 is an mSNl (A / mute swan / England / 053054 / 2021) head region sequence SEQ ID NO: 8 is a H7 HA signal sequence, Strep Tag II, linker, 6His purification tag, linker, SpyTag, linker, tetrabrachion tetramerisation domain and linker

[0045] SEQ ID NO: 9 is an Nl / 09 (A / California / 07 / 2009) head region sequence

[0046] SEQ ID NO: 10 is an IgKappa signal sequence, 6His purification tag, linker, SpyTag, linker, tetrabrachion tetramerisation domain, and linker

[0047] SEQ ID NO: 11 is an Nl / 19 (A / Wisconsin / 588 / 2019) head region sequence

[0048] SEQ ID NO: 12 is an Nl / 09 hybrid (Nl / 09 Loops - mSNl Scaffold) head region sequence SEQ ID NO: 13 is an Nl / 19 hybrid (Nl / 19 Loops - mSNl Scaffold) head region sequence SEQ ID NO: 14 is a PR8 N1 (A / PR / 8 / 1934) head region sequence SEQ ID NO: 15 is an mSLoops-PR8 Frame (mSNl Loops - PR8 Scaffold) head region sequence

[0049] SEQ ID NO: 16 is a PR8Loops-mS Frame (PR8 N1 Loops - mSNl Scaffold) head region sequence

[0050] SEQ ID NO: 17 is an mSNl B1L01 loop

[0051] SEQ ID NO: 18 is an mSNl B1L23 loop

[0052] SEQ ID NO: 19 is SVA, an mSNl B2L01 loop

[0053] SEQ ID NO: 20 is an mSNl B2L23 loop

[0054] SEQ ID NO: 21 is an mSNl B3L01 loop

[0055] SEQ ID NO: 22 is an mSNl B3L23 loop

[0056] SEQ ID NO: 23 is an mSNl B4L01 loop

[0057] SEQ ID NO: 24 is an mSNl B4L23 loop

[0058] SEQ ID NO: 25 is an mSNl B5L01 loop

[0059] SEQ ID NO: 26 is an mSNl B5L23 loop

[0060] SEQ ID NO: 27 is an mSNl B6L01 loop

[0061] SEQ ID NO: 28 is an mSNl B6L23 loop

[0062] SEQ ID NO: 29 is an Nl / 09 B1L23 loop

[0063] SEQ ID NO: 30 is an Nl / 09 B4L01 loop

[0064] SEQ ID NO: 31 is an Nl / 09 B5L01 loop

[0065] SEQ ID NO: 32 is an Nl / 09 B5L23 loop

[0066] SEQ ID NO: 33 is an Nl / 09 B6L01 loop

[0067] SEQ ID NO: 34 is an Nl / 09 B6L23 loop

[0068] SEQ ID NO: 35 is an Nl / 19 B3L01 loop

[0069] SEQ ID NO: 36 is an Nl / 19 B3L23 loop

[0070] SEQ ID NO: 37 is an Nl / 19 B5L01 loop

[0071] SEQ ID NO: 38 is an Nl / 19 B5L23 loop

[0072] SEQ ID NO: 39 is a PR9 N1 B1L23 loop

[0073] SEQ ID NO: 40 is a PR9 N1 B3L01 loop

[0074] SEQ ID NO: 41 is a PR9 N1 B3L23 loop

[0075] SEQ ID NO: 42 is a PR9 N1 B4L01 loop

[0076] SEQ ID NO: 43 is a PR9 N1 B5L01 loop

[0077] SEQ ID NO: 44 is a PR9 N1 B5L23 loop SEQ ID NO: 45 is a PR9 N1 B6L01 loop

[0078] SEQ ID NO: 46 is a PR9 N1 B6L23 loop

[0079] SEQ ID NO: 47 is the C-terminal domain of an mSNl (A / mute swan / England / 053054 / 2021) head region sequence

[0080] SEQ ID NO: 48 is a H7 HA signal sequence, 6His purification tag, linker, SpyTag and linker

[0081] SEQ ID NO: 49 is a head region sequence of chN8

[0082] (H5N8|A / chicken / England / 030720 / 2020|EPI_ISL_626652|83-470)

[0083] SEQ ID NO: 50 is a head region sequence of chN8

[0084] (H5N8|A / chicken / England / 030720 / 2020|EPI_ISL_626652|80-470)

[0085] SEQ ID NO: 51 is a head region sequence of N8Henan TB (H3N8|A / Henan / 4-14 / 2022|

[0086] EPI_ISL_12277128|83-470)

[0087] SEQ ID NO: 52 is a H7 HA signal sequence, a strep tag II sequence, a linker, 6His purification tag, linker, SpyTag and linker

[0088] SEQ ID NO: 53 is a N8Henan hybrid-1 (83-470) head region sequence

[0089] SEQ ID NO: 54 is a N8Henan hybrid-2 (80-470) head region sequence

[0090] SEQ ID NO: 55 is an chN8 B1L01 loop

[0091] SEQ ID NO: 56 is an chN8 B1L23 loop

[0092] SEQ ID NO: 57 is an chN8 B2L01 loop

[0093] SEQ ID NO: 58 is an chN8 B2L23 loop

[0094] SEQ ID NO: 59 is an chN8 B3L01 loop

[0095] SEQ ID NO: 60 is an chN8 B3L23 loop

[0096] SEQ ID NO: 61 is an chN8 B4L01 loop

[0097] SEQ ID NO: 62 is an chN8 B4L23 loop

[0098] SEQ ID NO: 63 is an chN8 B5L01 loop

[0099] SEQ ID NO: 64 is an chN8 B5L23 loop

[0100] SEQ ID NO: 65 is an chN8 B6L01 loop

[0101] SEQ ID NO: 66 is an chN8 B6L23 loop

[0102] SEQ ID NO: 67 is an chN8 B5L34 loop

[0103] SEQ ID NO: 68 is an chN8 c-terminal domain

[0104] SEQ ID NO: 69 is an N8Henan B1L01 loop

[0105] SEQ ID NO: 70 is an N8Henan B1L23 loop SEQ ID NO: 71 is an N8Henan B2L01 loop

[0106] SEQ ID NO: 72 is an N8Henan B2L23 loop

[0107] SEQ ID NO: 73 is an N8Henan B3L01 loop

[0108] SEQ ID NO: 74 is an N8Henan B3L23 loop

[0109] SEQ ID NO: 75 is an N8Henan B4L01 loop

[0110] SEQ ID NO: 76 is an N8Henan B4L23 loop

[0111] SEQ ID NO: 77 is an N8Henan B5L01 loop

[0112] SEQ ID NO: 78 is an N8Henan B5L23 loop

[0113] SEQ ID NO: 79 is an N8Henan B6L01 loop

[0114] SEQ ID NO: 80 is an N8Henan B6L23 loop

[0115] SEQ ID NO: 81 is an N8Henan B5L34 loop

[0116] SEQ ID NO: 82 is an N8Henan c-terminal domain

[0117] SEQ ID NO: 83 is a TEV Protease Site

[0118] SEQ ID NO: 84 is an avitag

[0119] SEQ ID NO: 85 is a H7 HA signal sequence, linker, avitag, linker, 6His purification tag, linker, SpyTag, linker, TEV protease site, tetrabrachion tetramerisation domain and linker SEQ ID NO: 86 is a head region sequence of X-31 NA (H3N2 / 1968) 82-469

[0120] SEQ ID NO: 87 is a head region sequence of N2 / 2021 (H3N2 A / Darwin / 6 / 2021) 82-469

[0121] SEQ ID NO: 88 is a head region sequence of X-31 hybrid (X-31Loops-N2 / 21 scaffold)

[0122] SEQ ID NO: 89 is a head region sequence of N2 / 2021 hybrid (N2 / 21 -X-31 scaffold)

[0123] SEQ ID NO: 90 is an X-31 NA B1L01 loop

[0124] SEQ ID NO: 91 is an X-31 NA B1L23 loop

[0125] SEQ ID NO: 92 is an X-31 NA B2L01 loop

[0126] SEQ ID NO: 93 is an X-31 NA B2L23 loop

[0127] SEQ ID NO: 94 is an X-31 NA B3L01 loop

[0128] SEQ ID NO: 95 is an X-31 NA B3L23 loop

[0129] SEQ ID NO: 96 is an X-31 NA B4L01 loop

[0130] SEQ ID NO: 97 is an X-31 NA B4L23 loop

[0131] SEQ ID NO: 98 is an X-31 NA B5L01 loop

[0132] SEQ ID NO: 99 is an X-31 NA B5L23 loop

[0133] SEQ ID NO: 100 is an X-31 NA B6L01 loop

[0134] SEQ ID NO: 101 is an X-31 NA B6L23 loop SEQ ID NO: 102 is an X-31 NA c-terminal domain

[0135] SEQ ID NO: 103 is an N2 / 2021 B1L01 loop

[0136] SEQ ID NO: 104 is an N2 / 2021 B1L23 loop

[0137] SEQ ID NO: 105 is an N2 / 2021 B2L01 loop

[0138] SEQ ID NO: 106 is an N2 / 2021 B2L23 loop

[0139] SEQ ID NO: 107 is an N2 / 2021 B3L01 loop

[0140] SEQ ID NO: 108 is an N2 / 2021 B3L23 loop

[0141] SEQ ID NO: 109 is an N2 / 2021 B4L01 loop

[0142] SEQ ID NO: 110 is an N2 / 2021 B4L23 loop

[0143] SEQ ID NO: 111 is an N2 / 2021 B5L01 loop

[0144] SEQ ID NO: 112 is an N2 / 2021 B5L23 loop

[0145] SEQ ID NO: 113 is an N2 / 2021 B6L01 loop

[0146] SEQ ID NO: 114 is an N2 / 2021 B6L23 loop

[0147] SEQ ID NO: 115 is an N2 / 2021 c-terminal domain

[0148] Detailed Description of the Invention

[0149] General definitions

[0150] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as is commonly understood by a person skilled in the art to which this invention belongs.

[0151] In general, the term “ comprising" is intended to mean including but not limited to. For example, the phrase “a polypeptide comprising a head region" should be interpreted to mean that the polypeptide has a head region, but the polypeptide may comprise further elements. In some embodiments of the invention, the word “comprising" may be replaced with the phrase “consisting of . The term “consisting of is intended to be limiting. For example, the phrase “a polypeptide consisting of a head region" should be interpreted to mean that the polypeptide has a head region and contains no further components.

[0152] In some embodiments of the invention, the word “comprising” may be replaced with the phrase “consisting essentially of’ . The term “consisting essentially of” means that specific further components can be present, namely those not materially affecting the essential characteristics of the subject matter. The singular forms “a”, “an”, and “the” include plural referents unless the context clear dictates otherwise. Thus, for example, reference to “an amino acid” includes one or more instances or versions of such amino acids.

[0153] The terms “protein” and “polypeptide” are used interchangeably herein, and are intended to refer to a polymeric chain of amino acids of any length.

[0154] The terms “nucleic acid molecule”, “polynucleotide” , “nucleotide sequence” and “nucleic acid sequence” are intended to refer to a polymeric chain of any length of nucleotides, including deoxyribonucleotides, ribonucleotides, or analogues thereof. For example, the nucleic acid molecule, polynucleotide or nucleotide sequence may comprise DNA (deoxyribonucleotides) or RNA (ribonucleotides). The nucleic acid molecule, polynucleotide or nucleotide sequence may consist of DNA. The nucleic acid molecule, polynucleotide or nucleotide sequence may be mRNA. Since the nucleic acid molecule, polynucleotide or nucleotide sequence may comprise RNA or DNA, all references to T (thymine) nucleotides may be replaced with U (uracil).

[0155] For the purpose of this invention, in order to determine the percent identity of two sequences (such as two polynucleotide or two polypeptide sequences), the sequences are aligned for optimal comparison purposes (e.g., gaps can be introduced in a first sequence for optimal alignment with a second sequence). The nucleotide or amino acid residues at each position are then compared. When a position in the first sequence is occupied by the same amino acid as the corresponding position in the second sequence, then the amino acids are identical at that position. The percent identity between the two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = number of identical positions / total number of positions in the reference sequence x 100).

[0156] Typically, the sequence comparison is carried out over the length of the reference sequence. For example, if the user wished to determine whether a given (“test”) sequence has at least 80% identity to SEQ ID NO: 1, SEQ ID NO: 1 would be the reference sequence. To assess whether a sequence has at least 80% identity to SEQ ID NO: 1 (an example of a reference sequence), the skilled person would carry out an alignment over the length of SEQ ID NO: 1, and identify how many positions in the test sequence were identical to those of SEQ ID NO: 1. If at least 80% of the positions are identical, the test sequence is at least 80% identical to SEQ ID NO: 1. If the sequence is shorter than SEQ ID NO: 1, the gaps or missing positions should be considered to be non-identical positions. The skilled person is aware of different computer programs that are available to perform an alignment between two sequences. An alignment between two sequences can be accomplished using a mathematical algorithm. For example, an alignment may be performed using the Needleman and Wunsch algorithm (Needleman and Wunsch, 1970, J Mol Biol.;48(3):443- 53) which aligns the sequences optimally over the entire length). Sequences of substantially different lengths may alternatively be aligned using a local alignment algorithm (e.g.15 Smith and Waterman algorithm (Smith and Waterman, 1981, J Theor Biol. ;91 (2):379-80) or Altschul algorithm (Altschul SF et al., 1997, Nucleic Acids Res.;25(17):3389-402; Altschul SF et al., 2005, Bioinformatics.;21(8): 1451-6). Those skilled in the art can determine appropriate parameters for measuring alignment, including any algorithms needed to achieve maximal alignment over the full length of the sequences being compared.

[0157] An amino acid sequence “ corresponding to" specified positions of a specified SEQ ID NO may be the amino acid sequence at the specified positions of the particular SEQ ID NO recited. For example, an amino acid sequence that corresponds to an mSNl B1L01 loop (i.e. positions 26-38 of SEQ ID NO: 7) may be the sequence at the 26-38th amino acids in the sequence. Alternatively, an amino acid sequence “corresponding to" positions X-Y of a specified SEQ ID NO may be an amino acid sequence which is not the X-Y111amino acids, but rather the amino acid sequence that best aligns to positions X-Y in the specified SEQ ID NO. It is within the capabilities of the person skilled in the art to determine which amino acids in an alternative neuraminidase amino acid sequence “correspond to" the specified positions in the specified SEQ ID NO (e.g. SEQ ID NO: 7). For example, the person skilled in the art may perform a sequence alignment of the alternative neuraminidase amino acid sequence with the specified SEQ ID NO using a suitable alignment algorithm such as that of Needleman and Wunsch described above, and determine which region of the alternative neuraminidase amino acid sequence best aligns to the specified positions in the specified SEQ ID NO. For example, the skilled person is able to align the alternative neuraminidase amino acid sequence with SEQ ID NO: 7 and determine which amino acid sequence best aligns, and therefore corresponds to, e.g. positions X-Y or 26-38 of SEQ ID NO: 7. In some cases, the sequence alignment of the alternative neuraminidase amino acid sequence with the specified SEQ ID NO may comprise increasing the weighting of the alignment of the sequence outside of the loops when compared to the alignment within the loops. This may be useful where amino acid mutations (substitutions, deletions or additions) occur at the boundaries of the loops.

[0158] All publications, patents and patent applications cited herein, whether supra or infra, are hereby incorporated by reference in their entirety.

[0159] Polypeptide

[0160] Neuraminidase (NA) is a mushroom shaped type II homotetrameric protein. The polypeptide chain of the NA monomeric head (head region) folds into six, topologically identical, four- stranded, antiparallel P-sheets which are arranged like the blades of a propeller. Each of the six P-sheet blades, numbered Bl -B6, is composed of four antiparallel P-strands in a W-shape numbered S1-S4. The P-strands in the “W” are connected at the top and bottom by loops (Fig. la). The loop between the fourth strand of the preceding beta sheet and the first strand of the following sheet (L01), and the loop between the second and third strands (L23) form most of the top surface and the active site of the NA and were delineated in the first crystal structure of the N2 NA (Varghese, J. N., Laver, W. G. & Colman, P. M. Structure of the influenza virus glycoprotein antigen neuraminidase at 2.9 A resolution. Nature 303, 35-40 (1983); Colman, P. M., Varghese, J. N. & Laver, W. G. Structure of the catalytic and antigenic sites in influenza virus neuraminidase. Nature 303, 41-44 (1983)) and in subsequent structures (reviewed in Air, G. M. Influenza neuraminidase. Influenza and other Respiratory Viruses 6, 245-256 (2012)). As there are six P-sheets in each NA monomer, there are six L01 and six L23 loops that together form most of the top surface of each identical NA monomer with a final contribution from the C-terminal domain of ~17 amino acids (Fig. la).

[0161] The L01 and L23 loops encircle the enzymatic cavity that constitutes the sialic acid binding site. NA cleaves off host sialic acids, aiding the release of newly formed virions from infected cells. While the loop residues contributing to the active site of NA are completely conserved, the remaining sequence of these loops can be highly variable. Many antibodies that inhibit enzyme activity and are protective in vivo bind to these loops. Multiple studies have also shown that the majority of, but not all, monoclonal antibodies (mAbs) to NA select resistant viruses with amino acid replacements in these loops. Analysis of the evolution of NAs has shown that the great majority of sequence change over time occurs at the surface exposed residues of NA, which include the residues of the L01 and L23 loops.

[0162] The L12 loop is the loop between the first and second strands of a beta sheet, and the L34 loop is the loop between the third and fourth strands of a beta sheet. In some cases, in particular for the group I neuraminidase N8, the B5L34 loop may form part of the immunogenic surface of the neuraminidase.

[0163] In addition to the head region, wild type neuraminidase comprises a stalk region (otherwise referred to as a ‘stem’ region), a transmembrane region and a cytoplasmic region).

[0164] The invention provides a polypeptide comprising the head region of a first neuraminidase. The L01 and L23 loops of the first neuraminidase are substituted with the L01 and L23 loops of a second neuraminidase. The resulting polypeptide comprises a neuraminidase head region that is a hybrid / chimera of two neuraminidase proteins, comprising a scaffold of the first neuraminidase on which the L01 and L23 loops of the second neuraminidase are added. The top surface exposed region of the resulting polypeptides mirrors that of the second neuraminidase, whilst benefitting from the expression and stability characteristics of the scaffold from the first neuraminidase.

[0165] The head region (otherwise known as the head group that contains the catalytic domain) is the sequence comprising a six-bladed propeller structure, with each blade comprising four anti-parallel P-sheets, and a C-terminal domain. The head region is the sequence that is C-terminal of the stalk of the wild-type neuraminidase. The head region typically comprises exosialidase activity. In some embodiments, the head region comprises an amino acid sequence starting at a position corresponding to any of amino acid positions 69 to 95 of the neuraminidase of H5N1 A / mute swan / England / 053054 / 2021 and ending at the C-terminus of the neuraminidase. In some embodiments, the head region comprises an amino acid sequence corresponding to amino acid positions 82 to 469 of the neuraminidase of H5N1 A / mute swan / England / 053054 / 2021. In some embodiments, the head region comprises an amino acid sequence corresponding to SEQ ID NO: 7. In some embodiments, the head region comprises an amino acid sequence corresponding to amino acid positions 69 to 469 of However, H1N1 A / Califomia / 07 / 2009 NA (Nl / 09). In some embodiments, the head region comprises an amino acid sequence corresponding to SEQ ID NO: 9. For example, the head region of the first neuraminidase may comprise SEQ ID NO: 7 or a sequence which corresponds to the amino acid sequence of SEQ ID NO: 7. For example, the head region of the first neuraminidase comprising a sequence which corresponds to the amino acid sequence of SEQ ID NO: 7 may comprise SEQ ID NO: 86 or 87. In some embodiments, the head region comprises an amino acid sequence corresponding to amino acid positions 80 to 480 or 83 to 470 of H5N8|A / chicken / England / 030720 / 2020|EPI_ISL_626652. In some embodiments, the head region comprises an amino acid sequence corresponding to SEQ ID NO: 49. In some embodiments, the head region comprises an amino acid sequence corresponding to SEQ ID NO: 50. For example, in some embodiments, the head region of the first neuraminidase may comprise SEQ ID NO: 49 or 50.

[0166] As used herein, the term “ substituted" is intended to mean that the amino acid sequences that are substituted (e.g. the L01 and L23 loops of the head region) are identical to that of the second neuraminidase, whilst the remaining residues are identical to that of first neuraminidase. The residues that are identical to that of the first neuraminidase may be referred to as a ‘scaffold’ or ‘framework’. For example, the invention also provides a polypeptide comprising a neuraminidase head region, comprising the head region scaffold of a first neuraminidase and the L01 and L23 loops of a second neuraminidase. In some cases, one or more of the L01 and L23 loops are identical in the first and second neuraminidase. In some embodiments and as discussed herein, further amino acids sequences of the first neuraminidase are substituted with those of the second neuraminidase. In some embodiments, provided herein is a polypeptide comprising the head region of a first neuraminidase, wherein the head region comprises a substitution of the L01 and L23 loops with the L01 and L23 loops of a second neuraminidase.

[0167] The skilled person will understand that any neuraminidase can be used as the first or second neuraminidase in the polypeptides of the invention. The first and second neuraminidases are different proteins. The first and second neuraminidases comprise different L01 and L23 loop sequences. For example, the second neuraminidase comprises at least one amino acid substitution, addition or deletion in the L01 and L23 loops when compared to the first neuraminidase. In other words, eleven of the twelve loops making up the L01 and L23 loops of the neuraminidase may be identical in the first and second neuraminidases, whilst the remaining loop comprises at least one amino acid substitution, addition or deletion in the second neuraminidase when compared to the first neuraminidase. In some embodiments, the second neuraminidase comprises at least two, at least three, at least four, at least five, at least 10 or at least 15 amino acid substitutions, additions or deletions in the L01 and L23 loops when compared to the first neuraminidase.

[0168] In some embodiments, the first neuraminidase is capable of providing a good yield in an expression system. The first neuraminidase may have a greater yield than the second neuraminidase. The head region of the polypeptide of the invention may express as well as, or better than, the head region of the first neuraminidase. In some embodiments, the head region of the polypeptide of the invention has a greater yield than the second neuraminidase. In some embodiments, the polypeptide of the invention has the same (+ / - 10%) yield or a greater yield than the first neuraminidase. In some embodiments, the first neuraminidase is a polypeptide that yields at least 10 mg / L, at least 20 mg / L, at least 30 mg / L, at least 40 mg / L, at least 50 mg / L, at least 60 mg / L, at least 70 mg / L, at least 80 mg / L, at least 90 mg / L, at least 100 mg / L, at least 150 mg / L, at least 200 mg / L, at least 250 mg / L, or at least 300 mg / L. In some embodiments, the first neuraminidase is a polypeptide that yields at least 50mg / L. In some embodiments, the first neuraminidase has a yield that is greater than the yield of the second neuraminidase. In some embodiments, the neuraminidase has a yield that is at least 2 times, at least 3 times, at least 4 times, at least 5 times or at least 10 times greater than the yield of the second neuraminidase. In some embodiments, the first neuraminidase has a yield that is at least 10 mg / L, at least 20 mg / L, at least 50 mg / L, at least 100 mg / L or at least 200 mg / L greater than the yield of the second neuraminidase. In some embodiments, the head region of the polypeptide of the invention has a yield that is greater than the yield of the second neuraminidase. In some embodiments, the head region of the polypeptide of the invention has a yield that is at least 2 times, at least 3 times, at least 4 times, at least 5 times or at least 10 times greater than the yield of the second neuraminidase. In some embodiments, the head region of the polypeptide of the invention has a yield that is at least 10 mg / L, at least 20 mg / L, at least 50 mg / L, at least 100 mg / L or at least 200 mg / L greater than the yield of the second neuraminidase. In some embodiments, the head region of the polypeptide of the invention yields at least 10 mg / L, at least 20 mg / L, at least 30 mg / L, at least 40 mg / L, at least 50 mg / L, at least 60 mg / L, at least 70 mg / L, at least 80 mg / L, at least 90 mg / L, at least 100 mg / L, at least 150 mg / L, at least 200 mg / L, at least 250 mg / L, or at least 300 mg / L. In some embodiments, the head region of the polypeptide of the invention yields at least 50mg / L.

[0169] The yield is typically calculated as the total yield of a construct comprising SEQ ID NO: 8 and the head region of the neuraminidase and encoded in a pcDNA3.1 vector, expressed in ExpiCHO cells for 8 days at 37°C, 8% CO2. The construct used in the calculation comprises, from N-terminus to C-terminus, SEQ ID NO: 8 and the head region of the first neuraminidase. In some embodiments, the yield may be calculated as the total yield of a construct comprising SEQ ID NO: 48 and the head region of the neuraminidase and encoded in a pcDNA3.1 vector, expressed in ExpiCHO cells for 8 days at 37°C, 8% CO2. The construct used in the calculation comprises, from N-terminus to C-terminus, SEQ ID NO: 48 and the head region of the first neuraminidase. The skilled person will recognise that any appropriate expression system may be used to express the construct, and that the parameters provided herein are used to define conditions under which a first neuraminidase is considered to express well. In some embodiments, the yield may be further calculated following harvesting of the ExpiCHO culture supernatant followed by Ni-NTA-sepharose affinity chromatography purification.

[0170] In some embodiments, the yield may be calculated based on a construct comprising the head region of the neuraminidase and encoded in a pcDNA3.1 vector, expressed in ExpiCHO cells for 8 days at 32°C, 5% CO2.

[0171] In some embodiments, the first neuraminidase has good stability. Stability may be measured by melting temperature (Tm). The first neuraminidase may have greater stability than the second neuraminidase. The head region of the polypeptide of the invention may have the same or better stability than the head region of the first neuraminidase. In some embodiments, the head region of the polypeptide of the invention has greater stability than the second neuraminidase. In some embodiments, the polypeptide of the invention has the same (+ / - 10%) stability or a greater stability than the first neuraminidase. The polypeptide / neuraminidase is typically capable of forming (in other words, forms) a tetramer. Without wishing to be bound by theory, it is believed that the scaffold of the first neuraminidase comprises the majority of the residues responsible for inter-monomer interactions in the tetramer, and so a stable first neuraminidase typically results in a stable head region in the polypeptide of the invention. In some embodiments, the head region of the first neuraminidase has a Tmof at least 55 °C, at least 56 °C, at least 57 °C, at least 58 °C, at least 59 °C, at least 60 °C, at least 61 °C, at least 62 °C, at least 63 °C, at least 64 °C or at least 65 °C. In some embodiments, the head region of the first neuraminidase has a Tmof at least 60 °C. In some embodiments, the head region of the first neuraminidase has a Tmthat is greater than the Tmof the head region of the second neuraminidase. In some embodiments, the head region of the polypeptide of the invention has a Tmthat is greater than the Tmof the head region of the second neuraminidase. For example, the Tmof the head region of the polypeptide of the invention may be at least 5 °C, at least 6 °C, at least 7 °C, at least 8 °C, at least 9 °C or at least 10 °C greater than the Tmof the head region of the second neuraminidase. The Tmis typically calculated based on a construct comprising SEQ ID NO: 8 and the head region of the neuraminidase in DPBS buffer with calcium, wherein the Tmis measured by nano differential scanning fluorimetry. Other methods of measuring Tmare known in the art. The construct used in the calculation comprises, from N-terminus to C-terminus, SEQ ID NO: 8 and the head region of the first neuraminidase. In some embodiments, the Tmis calculated based on a construct comprising SEQ ID NO: 48 and the head region of the neuraminidase in DPBS buffer with calcium, wherein the Tmis measured by nano differential scanning fluorimetry. The construct used in the calculation comprises, from N-terminus to C-terminus, SEQ ID NO: 48 and the head region of the first neuraminidase.

[0172] In some embodiments, the Tmmay be calculated based on a construct comprising the head region of the neuraminidase and encoded in a pcDNA3.1 vector, expressed in ExpiCHO cells for 8 days at 32°C, 5% CO2.

[0173] In some embodiments, the first neuraminidase is derived from an influenza virus. The influenza may be influenza A, influenza B, influenza C or influenza D. In some embodiments, the influenza is influenza A or influenza B. In some embodiments, the influenza is influenza A. In some embodiments, the first neuraminidase is an Nl, N2, N3, N4, N5, N6, N7, N8 or N9 neuraminidase. In some embodiments, the first neuraminidase is an Nl, N2, N4, N5 or N8 neuraminidase. In some embodiments, the first neuraminidase is an Nl, N2 or N8 neuraminidase. In some embodiments, the first neuraminidase is an Nl or N2 neuraminidase.

[0174] In some embodiments, the first neuraminidase is a group 1 neuraminidase, i.e. Nl, N4, N5 or N8. In some embodiments, the first neuraminidase is an Nl or an N8 neuraminidase. In some embodiments, the first neuraminidase is an Nl neuraminidase. In some embodiments, the first neuraminidase is the neuraminidase of H5N1 A / mute swan / England / 053054 / 2021, or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of H5N1 A / mute swan / England / 053054 / 2021. In some embodiments, the first neuraminidase is the neuraminidase of H1N1 A / PR / 8 / 1934 (PR8), or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of H1N1 A / PR / 8 / 1934 (PR8). In some embodiments, the first neuraminidase is an N8 neuraminidase. In some embodiments, the first neuraminidase is the neuraminidase of H5N8|A / chicken / England / 030720 / 2020, or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of H5N8|A / chicken / England / 030720 / 2020.

[0175] In some embodiments, the first neuraminidase is a group 2 neuraminidase, i.e. N2, N3, N6, N7 or N9. In some embodiments, the first neuraminidase is an N2 neuraminidase. In some embodiments, the first neuraminidase is the neuraminidase of X-31 NA (H3N2 / 1968), or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of X-31 NA (H3N2 / 1968). In some embodiments, the first neuraminidase is the neuraminidase of N2 / 2021 (H3N2 A / Darwin / 6 / 2021), or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of N2 / 2021 (H3N2 A / Darwin / 6 / 2021).

[0176] In some embodiments, the first neuraminidase comprises a head region comprising the amino acid sequence of SEQ ID NO: 7, 14, 49, 50, 86 or 87. In some embodiments, the first neuraminidase comprises a head region comprising the amino acid sequence of SEQ ID NO: 7 or 14. In some embodiments, the first neuraminidase comprises a head region comprising the amino acid sequence of SEQ ID NO: 49 or 50. In some embodiments, the first neuraminidase comprises a head region comprising the amino acid sequence of SEQ ID NO: 86 or 87.

[0177] In some embodiments, the first neuraminidase and the second neuraminidase are of the same neuraminidase subgroup. For example, the first neuraminidase may be an N1 neuraminidase and the second neuraminidase may be an N1 neuraminidase. In other embodiments, the first neuraminidase and the second neuraminidase are of different neuraminidase subgroups. For example, the first neuraminidase may both be group 1 neuraminidases but of different subtype, such as an N1 and an N8 neuraminidase for the first and second neuraminidases, respectively. In some embodiments, the first neuraminidase and the second neuraminidase are of the same neuraminidase group. The first neuraminidase and the second neuraminidase may be group I neuraminidases. For example, the first neuraminidase may be an N1 neuraminidase and the second neuraminidase may be an N8 neuraminidase. The first neuraminidase and the second neuraminidase may be group II neuraminidases.

[0178] In some embodiments, the second neuraminidase is an Nl, N2, N3, N4, N5, N6, N7, N8 or N9 neuraminidase. In some embodiments, the second neuraminidase is an Nl, N2, N4, N5 or N8 neuraminidase. In some embodiments, the second neuraminidase is an Nl, N2 or N8 neuraminidase. In some embodiments, the second neuraminidase is a group

[0179] I neuraminidase. In some embodiments, the second neuraminidase is Nl or N8. In some embodiments, the second neuraminidase is a group II neuraminidase. In some embodiments, the second neuraminidase is an Nl or N2 neuraminidase.

[0180] In some embodiments, the second neuraminidase is an Nl neuraminidase.

[0181] In some embodiments, the second neuraminidase is the neuraminidase of a circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals. The circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals may be a strain identified by the WHO Global Influenza Programme (http s : / / www. who . int / tool s / flunet) .

[0182] In some embodiments, the C-terminal domain (CTD) of the head region of the first neuraminidase is substituted with the CTD of the second neuraminidase. In some embodiments, the CTD is the C-terminal 17 amino acids of the neuraminidase. In some embodiments, the CTD corresponds to the C-terminal 17 amino acids of SEQ ID NO: 7. In some embodiments, the CTD corresponds to SEQ ID NO: 47. In some embodiments, the CTD is the sequence corresponding to residues 372-388 of SEQ ID NO: 7. In embodiments wherein the CTD is the sequence corresponding to residues 372-388 of SEQ ID NO: 7, the first and / or second neuraminidase may be group I neuraminidase, such as an Nl or N8 neuraminidase. In some embodiments, the CTD is the C-terminal 11 amino acids of the neuraminidase. In some embodiments, the CTD corresponds to the C-terminal

[0183] I I amino acids of SEQ ID NO: 7. The CTD is preferably the sequence corresponding to residues 378-388 of SEQ ID NO: 7. In embodiments wherein the CTD is the sequence corresponding to residues 378-388 of SEQ ID NO: 7, the first and / or second neuraminidase may be group II neuraminidase, such as an N2 or N8 neuraminidase. The amino acid position at the C-terminal end of the CTD is the C-terminal amino acid of the neuraminidase sequence. In some embodiments, the amino acid position of the N-terminal end of the CTD may be varied to take into account factors discussed herein. In some embodiments, the N-terminal side of the CTD may be extended or reduced by up to four amino acids upstream or downstream, such as by up to two amino acids upstream or downstream or by up to 1 amino acid upstream or downstream. For example, in some embodiments, the CTD may be the amino acid sequence corresponding to residues 372 ± 4 to 388 of SEQ ID NO: 7, such as 372 ± 2 to 388 or 372 ± 1 to 388 of SEQ ID NO: 7. In some embodiments, the CTD may be the amino acid sequence corresponding to residues 378 ± 4 to 388 of SEQ ID NO: 7, such as 378 ± 2 to 388 or 378 ± 1 to 388 of SEQ ID NO: 7. In some embodiments, the CTD is the sequence following B6L34, i.e. the sequence C- terminal of the loop connecting P-sheets 3 and 4 in blade 6 of the neuraminidase head region. In some embodiments, the CTD is the sequence that is C-terminal of P-sheet 4 in blade 6 of the neuraminidase head region. In some embodiments, the CTD is the solvent exposed sequence that is C-terminal of B6L34. In some embodiments, the CTD is the solvent exposed sequence that is C-terminal of -sheet 4 in blade 6 of the neuraminidase head region.

[0184] In some embodiments, the polypeptide further comprises a neuraminidase stalk domain, a neuraminidase transmembrane domain and / or a neuraminidase cytoplasmic domain. In some embodiments, the polypeptide further comprises a neuraminidase stalk domain, a neuraminidase transmembrane domain and a neuraminidase cytoplasmic domain. The neuraminidase stalk domain, a neuraminidase transmembrane domain and / or a neuraminidase cytoplasmic domain is typically of the same neuraminidase group as the first neuraminidase, such as of the same neuraminidase subtype as the first neuraminidase. In some embodiments, the neuraminidase stalk domain, a neuraminidase transmembrane domain and / or a neuraminidase cytoplasmic domain is of the same neuraminidase group as the first and second neuraminidase, such as the same neuraminidase subtype as the first and second neuraminidase. In some embodiments, the polypeptide comprises in a N- terminus to C-terminus direction: an optional neuraminidase cytoplasmic domain, an optional neuraminidase transmembrane domain, an optional neuraminidase stalk domain and the neuraminidase head group. In some embodiments, the polypeptide further comprises a tetramerisation domain. The tetramerisation domain is typically N-terminal of the neuraminidase head group. In other words, the polypeptide may comprise in a N-terminus to C-terminus direction a tetramerisation domain and a neuraminidase head group. Any suitable tetramerisation domain may be used in the invention. For example, the tetramerisation domain may be a vasodilator-stimulated phosphoprotein (VASP) tetramerisation domain, a tetrabrachion tetramerisation domain, a measles phosphoprotein tetramerisation domain, a Sendai virus phosphoprotein tetramerisation domain, GCN4-pLI tetramerisation domain, or an Arabidopsis thaliana transcription factor tetramerisation domain. In some embodiments, the tetramerisation domain is a tetrabrachion tetramerisation domain.

[0185] L01, L12, L23, and L34 loops

[0186] The polypeptide of the invention comprises L01 and L23 loops of the second neuraminidase. The neuraminidase head region is made up of six P-sheet blades, numbered B1-B6 in the direction from N-terminus to C-terminus, each composed of four anti -parallel P-strands numbered 1-4 in the direction from N-terminus to C-terminus. The L01 loop is the loop between the fourth strand of the preceding beta sheet and the first strand of the following sheet. The L23 loop is the loop between the second and third strands of a beta sheet. The L12 loop is the loop between the first and second strands of a beta sheet. The L34 loop is the loop between the third and fourth strands of a beta sheet. The loops are typically referred to using nomenclature BXLYZ, where x is the number of the blade, and LYZ is the identity of the loop within said blade.

[0187] The loops may be identified by any suitable means. Typically, the same means for identifying a loop is used for both the first neuraminidase and the second neuraminidase. Various factors may be considered when identifying a loop, such as sequence alignment with an annotated reference sequence, solvent accessibility in the three-dimensional structure (based on a crystal structure or a predicted structure), evidence for evolutionary selection, viral resistance mutations selected with mAbs in vitro and defined crystal structures of bound antibodies. An example is provided in the Nl / 19 hybrid design, residues N200S, immediately C-terminal of the identified B2L23 loop could have been considered part of the loop. Despite the N200S substitution not being included in the hybrid protein, the hybrid retained the epitope specificity of the Nl / 19 NA loop donor (see Examples). The same principles apply to the identified of the C-terminal domain.

[0188] Table 1 provides the sequences and positions of the L01 and L23 loops in mSNl, Nl / 09, Nl / 19 and PR8 N1 neuraminidases. Accordingly, in some embodiments, a loop of a neuraminidase, for example a B1L01 loop, is an amino acid sequence which corresponds to the amino acid sequence of one of the mSNl loops listed in Table 1, for example the mSNl B1L01 loop. In some embodiments, the L01 and L23 loops are the amino acid sequences corresponding to residues 26-38, 54-75, 95-97, 115-118, 138-147, 163-170, 188-197, 212-215, 233-266, 283-288, 314-318 and 349-356 of SEQ ID NO: 7. This numbering is otherwise referred to as ‘vl’ or ‘version 1’. In some embodiments, the first and / or second neuraminidase is a group 1 neuraminidase. In some embodiments, the first and / or second neuraminidase is a Nl neuraminidase.

[0189] In some embodiments, the amino acid positions of the ends of the loops may be varied to take into account the factors discussed herein. In some embodiments, the N- terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 26 ± 4 to 38 ± 4, 54 ± 4 to 75 ± 4, 95 ± 4 to 97 ± 4, 115 ± 4 to 118 ± 4, 138 ± 4 to 147 ± 4, 163 ± 4 to 170 ± 4, 188 ± 4 to 197 ± 4, 212 ± 4 to 215 ± 4, 233 ± 4 to 266 ± 4, 283 ± 4 to 288 ± 4, 314 ± 4 to 318 ± 4 and 349 ± 4 to 356 ± 4 of SEQ ID NO: 7. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream, and / or the C- terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 26 ± 2 to 38 ± 2, 54 ± 2 to 75 ± 2, 95 ± 2 to 97 ± 2, 115 ± 2 to 118 ± 2, 138 ± 2 to 147 ± 2, 163 ± 2 to 170 ± 2, 188 ± 2 to 197 ± 2, 212 ± 2 to 215 ± 2, 233 ± 2 to 266 ± 2, 283 ± 2 to 288 ± 2, 314 ± 2 to 318 ± 2 and 349 ± 2 to 356 ± 2 of SEQ ID NO: 7. For example, the B1L01 loop corresponding to residues 26 ± 2 to 38 ± 2 of SEQ ID NO: 7 may refer to a loop corresponding to residues 24 to 36, 25 to 36, 26 to 36, 27 to 36, 28 to 36, 24 to 37, 25 to 37, 26 to 37, 27 to 37, 28 to 37, 24 to 38, 25 to 38, 26 to 38, 27 to 38, 28 to 38, 24 to 39, 25 to 39, 26 to 39, 27 to 39, 28 to 39, 24 to 40, 25 to 40, 26 to 40, 27 to 40 or 28 to 40. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream, and / or the C- terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream. In some embodiments, the L01 and L23 loops may be the amino acid sequences corresponding to residues 26 ± 1 to 38 ± 1, 54 ± 1 to 75 ± 1, 95 ± 1 to 97 ± 1, 115 ± 1 to 118 ± 1, 138 ± 1 to 147 ± 1, 163 ± 1 to 170 ± 1, 188 ± 1 to 197 ± 1, 212 ± 1 to 215 ± 1, 233 ± 1 to 266 ± 1, 283 ± 1 to 288 ± 1, 314 ± 1 to 318 ± 1 and 349 ± 1 to 356 ± 1 of SEQ ID NO: 7. As used herein, the term “upstream” refers to an amino acid on the N- terminus side of the specified amino acid position, and the term “downstream” refers to an amino acid on the C-terminus side of the specified amino acid position.

[0190] In some embodiments, the L01 and L23 loops are the amino acid sequences corresponding to residues 26-38, 54-75, 95-97, 115-118, 138-147, 163-170, 188-197, 212- 215, 233-266, 283-288, 314-318 and 347-356 of SEQ ID NO: 7. In some embodiments, the amino acid positions of the ends of the loops may be varied to take into account the factors discussed herein. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream, and / or the C- terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 26 ± 4 to 38 ± 4, 54 ± 4 to 75 ± 4, 95 ± 4 to 97 ± 4, 115 ± 4 to 118 ± 4, 138 ± 4 to 147 ± 4, 163 ± 4 to 170 ± 4, 188 ± 4 to 197 ± 4, 212 ± 4 to 215 ± 4, 233 ± 4 to 266 ± 4, 283 ± 4 to 288 ± 4, 314 ± 4 to 318 ± 4 and 347 ± 4 to 356 ± 4 of SEQ ID NO: 7. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 26 ± 2 to 38 ± 2, 54 ± 2 to 75 ± 2, 95 ± 2 to 97 ± 2, 115 ± 2 to 118 ± 2, 138 ± 2 to 147 ± 2, 163 ± 2 to 170 ± 2, 188 ± 2 to 197 ± 2, 212 ± 2 to 215 ± 2, 233 ± 2 to 266 ± 2, 283 ± 2 to 288 ± 2, 314 ± 2 to 318 ± 2 and 347 ± 2 to 356 ± 2 of SEQ ID NO: 7. For example, the B1L01 loop corresponding to residues 26 ± 2 to 38 ± 2 of SEQ ID NO: 7 may refer to a loop corresponding to residues 24 to 36, 25 to 36, 26 to 36, 27 to 36, 28 to 36, 24 to 37, 25 to 37, 26 to 37, 27 to 37, 28 to 37, 24 to 38, 25 to 38, 26 to 38, 27 to 38, 28 to 38, 24 to 39, 25 to 39, 26 to 39, 27 to 39, 28 to 39, 24 to 40, 25 to 40, 26 to 40, 27 to 40 or 28 to 40. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream. In some embodiments, the L01 and L23 loops may be the amino acid sequences corresponding to residues 26 ± 1 to 38 ± 1, 54 ± 1 to 75 ± 1, 95 ± 1 to 97 ± 1, 115 ± 1 to 118 ± 1, 138 ± 1 to 147 ± 1, 163 ± 1 to 170 ± 1, 188 ± 1 to 197 ± 1, 212 ± 1 to 215 ± 1, 233 ± 1 to 266 ± 1, 283 ± 1 to 288 ± 1, 314 ± 1 to 318 ± 1 and 347 ± 1 to 356 ± 1 of SEQ ID NO: 7. As used herein, the term ^upstream" refers to an amino acid on the N-terminus side of the specified amino acid position, and the term ^dow nstream" refers to an amino acid on the C-terminus side of the specified amino acid position.

[0191] In some embodiments, a loop of a neuraminidase, for example a B1L01 loop, is an amino acid sequence which corresponds to the amino acid sequence of one of the chN8 loops listed in Table 3, for example the chN8 B1L01 loop. In some embodiments, the L01 and L23 loops are the amino acid sequences corresponding to residues 23-35, 58-75, 95- 99, 115-120, 136-147, 163-171, 187-193, 212-219, 236-263, 280-288, 314-320 and 347- 356 of SEQ ID NO: 7. This numbering is otherwise referred to as ‘v2.1’, ‘v2.1+’, ‘version 2’ or ‘version 2.1’. In some embodiments, the first and / or second neuraminidase is a group 1 neuraminidase. In some embodiments, the first and / or second neuraminidase is an N8 neuraminidase.

[0192] In some embodiments, the amino acid positions of the ends of the loops may be varied to take into account the factors discussed herein. In some embodiments, the N- terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 23 ± 4 to 35 ± 4, 58 ± 4 to 75 ± 4, 95 ± 4 to 99 ± 4, 115 ± 4 to 120 ± 4, 136 ± 4 to 147 ± 4, 163 ± 4 to 171 ± 4, 187 ± 4 to 193 ± 4, 212 ± 4 to 219 ± 4, 236 ± 4 to 263 ± 4, 280 ± 4 to 288 ± 4, 314 ± 4 to 320 ± 4 and 347 ± 4 to 356 ± 4 of SEQ ID NO: 7. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream, and / or the C- terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 23 ± 2 to 35 ± 2, 58 ± 2 to 75 ± 2, 95 ± 2 to 99 ± 2, 115 ± 2 to 120 ± 2, 136 ± 2 to 147 ± 2, 163 ± 2 to 171 ± 2, 187 ± 2 to 193 ± 2, 212 ± 2 to 219 ± 2, 236 ± 2 to 263 ± 2, 280 ± 2 to 288 ± 2, 314 ± 2 to 320 ± 2 and 347 ± 2 to 356 ± 2 of SEQ ID NO: 7. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream. In some embodiments, the L01 and L23 loops may be the amino acid sequences corresponding to residues 23 ± 1 to 35 ± 1, 58 ± 1 to 75 ± 1, 95 ± 1 to 99 ± 1, 115 ± 1 to 120 ± 1, 136 ± 1 to 147 ± 1, 163 ± 1 to 171 ± 1, 187 ± 1 to 193 ± 1, 212 ± 1 to 219 ± 1, 236 ± I to 263 ± l, 280 ± 1 to 288 ± 1, 314 ± 1 to 320 ± 1 and 347 ± 1 to 356 ± 1 of SEQ ID NO: 7. As used herein, the term “upstream” refers to an amino acid on the N-terminus side of the specified amino acid position, and the term “downstream” refers to an amino acid on the C-terminus side of the specified amino acid position.

[0193] In some embodiments, a loop of a neuraminidase, for example a B1L01 loop, is an amino acid sequence which corresponds to the amino acid sequence of one of the X-31 NA loops listed in Table 3, for example the X-31 NA B1L01 loop. In some embodiments, the L01 and L23 loops are the amino acid sequences corresponding to residues 23-35, 56-75, 95-99, 115-120, 138-147, 163-171, 188-197, 212-219, 233-267, 281-289, 314-320 and 348-356 of SEQ ID NO: 7. This numbering is otherwise referred to as ‘v3’ or ‘version 3’. In some embodiments, the first and / or second neuraminidase is a group 2 neuraminidase. In some embodiments, the first and / or second neuraminidase is an N2 neuraminidase.

[0194] The amino acid positions of the ends of the loops may be varied to take into account the factors discussed herein. In preferred embodiments, the N-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 23 ± 4 to 35 ± 4, 56 ± 4 to 75 ± 4, 95 ± 4 to 99 ± 4, 115 ± 4 to 120 ± 4, 138 ± 4 to 147 ± 4, 163 ± 4 to 171 ± 4, 188 ± 4 to 197 ± 4, 212 ± 4 to 219 ± 4, 233 ± 4 to 267 ± 4, 281 ± 4 to 289 ± 4, 314 ± 4 to 320 ± 4 and 348 ± 4 to 356 ± 4 of SEQ ID NO: 7. For example, in some embodiments, the N-terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to two amino acids upstream or downstream. For example, the L01 and L23 loops may be the amino acid sequences corresponding to residues 23 ± 2 to 35 ± 2, 56 ± 2 to 75 ± 2, 95 ± 2 to 99 ± 2, 115 ± 2 to 120 ± 2, 138 ± 2 to 147 ± 2, 163 ± 2 to 171 ± 2, 188 ± 2 to 197 ± 2, 212 ± 2 to 219 ± 2, 233 ± 2 to 267 ± 2, 281 ± 2 to 289 ± 2, 314 ± 2 to 320 ± 2 and 348 ± 2 to 356 ± 2 of SEQ ID NO: 7. For example, in some embodiments, the N-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to one amino acid upstream or downstream. In some embodiments, the L01 and L23 loops may be the amino acid sequences corresponding to residues 23 ± 1 to 35 ± 1, 56 ± 1 to 75 ± 1, 95 ± 1 to 99 ± 1, 115 ± 1 to 120 ± 1, 138 ± 1 to 147 ± 1, 163 ± 1 to 171 ± 1, 188 ± 1 to 197 ± 1, 212 ± 1 to 219 ± 1, 233 ± 1 to 267 ± 1, 281 ± 1 to 289 ± 1, 314 ± 1 to 320 ± 1 and 348 ± 1 to 356 ± 1 of SEQ ID NO: 7. As used herein, the term “upstream” refers to an amino acid on the N-terminus side of the specified amino acid position, and the term “downstream” refers to an amino acid on the C-terminus side of the specified amino acid position.

[0195] In some embodiments, a polypeptide of the invention comprises the head region of a first neuraminidase, wherein one or more loops, e.g. L01 loops, are substituted with the corresponding loops of a second neuraminidase. The skilled person can determine whether a polypeptide comprising a head region of a first neuraminidase comprises a loop substitution by aligning the polypeptide to a reference polypeptide, locating the amino acid sequence in the polypeptide which corresponds (i.e. aligns) to the given loop in the reference polypeptide and determining if the amino acid sequence at the positions corresponding to the loop is the same as the amino acid sequence in the reference polypeptide, wherein the reference polypeptide is a corresponding native polypeptide. If the amino acid sequence is not the same, the polypeptide comprises a loop substitution at that position. For example, to determine whether a polypeptide comprising an mSNl neuraminidase head region has a B1L01 loop substitution, the skilled person would align the polypeptide to the mSNl neuraminidase head region of SEQ ID NO: 7 and determine whether the amino acid sequence at the positions corresponding to positions 26-38 of SEQ ID NO: 7 is RIGSKGDVFVIRE. If the amino acid sequence at those positions is other than RIGSKGDVFVIRE, the polypeptide comprising an mSNl neuraminidase head region has a B1L01 loop substitution. For example, if the amino acid sequence at those positions is the amino acid sequence of the B1L01 loop of a second neuraminidase (e.g. an Nl / 09 neuraminidase), the polypeptide comprising an mSNl neuraminidase head region has a B1L01 loop substitution with a second neuraminidase (e.g. an Nl / 09 neuraminidase). Table 1 : Loop annotations, N1 numbering is based off the mSNl head region as set out in in SEQ ID NO: 7. Residues in loops that differ from mSNl sequence are marked in bold and underline Table 3: Loop annotations, numbering is based off the mSNl head region as set out in in SEQ ID NO: 7. Exemplary sequences for each numbering scheme (v2.1 or v3) provided.

[0196] The invention provides a polypeptide comprising the head region of a first neuraminidase, wherein at least one of the L01 and L23 loops are substituted with the corresponding L01 and L23 loops of a second neuraminidase. In some embodiments, one or more loops selected from B3L01, B3L23, B4L01, B5L01, B5L23, B6L01 and B6L23 of the head region of the first neuraminidase are substituted with loops B3L01, B3L23, B4L01, B5L01, B5L23, B6L01 and B6L23 of the head region of the second neuraminidase. In some embodiments, loops B3L01, B3L23, B4L01, B5L01, B5L23, B6L01 and B6L23 of the head region of the first neuraminidase are substituted with loops B3L01, B3L23, B4L01, B5L01, B5L23, B6L01 and B6L23. The nomenclature B6L23 is understood as referring to loop 23 of blade 6 of a neuraminidase head region. It has been identified that loops B3L01, B3L23, B4L01, B5L01, B5L23, B6L01 and B6L23 are immunodominant, i.e. the key antibody response is directed to residues within these seven loops. In some embodiments, the polypeptide comprises the head region of a first neuraminidase, wherein the L01 and L23 loops are substituted with the L01 and L23 loops of a second neuraminidase.

[0197] In some embodiments, the B5L34 loop of the first neuraminidase is substituted with the B5L34 loop of the second neuraminidase. In other words, the B01 loops, the B23 loops and the B5L34 loop of the first neuraminidase are substituted with the B01 loops, the B23 loops and the B5L34 loop of the first neuraminidase. For certain neuraminidases, such as the N8 neuraminidase, this loop has been identified as contributing to the immunogenicity of the neuraminidase. In some embodiments, the B5L34 loop is the amino acid sequence corresponding to residues 299-307 of SEQ ID NO: 7. In some embodiments, the amino acid positions of the ends of the loops may be varied to take into account the factors discussed herein. In some embodiments, the N-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream, and / or the C-terminal side of the loops may be extended or reduced by up to four amino acids upstream or downstream, such as by up to two amino acids upstream or downstream, such or by up to 1 amino acid upstream or downstream. For example, the B5L34 loop may be the amino acid sequences corresponding to residues 299 ± 4 to 307 ± 4, 299 ± 2 to 307 ± 2, or 299 ± 1 to 307 ± 1 of SEQ ID NO: 7. As used herein, the term "upstr am" refers to an amino acid on the N-terminus side of the specified amino acid position, and the term "downstream" refers to an amino acid on the C-terminus side of the specified amino acid position. Accordingly, in some embodiments, the first and / or second neuraminidase is a group 1 neuraminidase. In some embodiments, the first and / or second neuraminidase is an N8 neuraminidase. The CTD of the first neuraminidase may be further substituted with the CTD of the second neuraminidase, as described herein.

[0198] In some embodiments, one or more amino acid residue of the L01 and L23 loops, and CTD that are involved in inter-monomer interactions are further substituted with the corresponding amino acid residue of the first neuraminidase. Exemplary amino acids are identified in Ellis et al. In some embodiments, one or more residue corresponding to amino acid position 96 or 196 of SEQ ID NO: 7 are further substituted with the amino acid of the first neuraminidase corresponding to position 96 or 115 of the first neuraminidase. In some embodiments, the CTD is the CTD of the first neuraminidase.

[0199] In some embodiments, substitution is achieved by modifying the nucleic acid sequence encoding the head region of the first neuraminidase, for example by modifying the nucleic acid sequence encoding the L01 and L23 loops of the first neuraminidase. In some embodiments, substitution is achieved by replacing the nucleic acids encoding the L01 and / or L23 loops in a polynucleotide encoding the head region of the first neuraminidase with nucleic acids encoding the L01 and / or L23 loops of a second neuraminidase. In some embodiments, the polynucleotide is a DNA molecule. In some embodiments, the polynucleotide is an RNA molecule. In some embodiments, the polynucleotide is codon optimised. The polynucleotide may be codon optimised depending on the expression system to be used. For example, the polynucleotide may be codon optimised for insect, yeast, plant, or animal expression. The polynucleotide may be codon optimised for expression in mammals, such as in humans or Cricetuleus griseus.

[0200] In some embodiments, the L12 and L34 loops of the first neuraminidase are substituted with the L12 and L34 loops of a second neuraminidase. Exemplary definitions of the L12 and L34 loops are provided in Varghese et al (Nature 303, 35-40 (1983)). A skilled person would be able to determine the amino acid positions corresponding to the L12 and L34 loops defined in Varghese et al using the methods described herein.

[0201] Tetramer

[0202] Wild-type neuraminidase forms a tetramer of four identical monomers. Some neuraminidase head region monomers are able to form stable tetramers without the stalk and transmembrane domains of neuraminidase. Polypeptides comprising the head region of a neuraminidase and a tetramerisation domain are typically able to form stable tetramers. For example, the polypeptide may comprise a vasodilator-stimulated phosphoprotein (VASP) tetramerisation domain; a tetrabrachion tetramerisation domain; a neuraminidase stalk domain, transmembrane domain and cytoplasmic domain; a measles phosphoprotein tetramerisation domain; a Sendai virus phosphoprotein tetramerisation domain; or an Arabidopsis thaliana transcription factor tetramerisation domain. In some embodiments, the polypeptide comprises a tetrabrachion tetramerisation domain.

[0203] In some embodiments, the head region of the first neuraminidase is capable of forming (in other words, forms) a tetramer that has a Tmof at least 55 °C, at least 56 °C, at least 57 °C, at least 58 °C, at least 59 °C, at least 60 °C, at least 61 °C, at least 62 °C, at least 63 °C, at least 64 °C or at least 65 °C. In some embodiments, the head region of the first neuraminidase is capable of forming (in other words, forms) a tetramer that has a Tm of at least 60 °C. For the purposes of calculating the Tm, the tetramer is a homotetramer. In some embodiments, a tetramer of the head region of the first neuraminidase has a Tmthat is greater than the Tmof a tetramer of the head region of the second neuraminidase. In some embodiments, a tetramer of the head region of the polypeptide of the invention has a Tmthat is greater than the Tmof a tetramer of the head region of the second neuraminidase. For example, the Tmof the tetramer of the head region of the polypeptide of the invention may be at least 5 °C, at least 6 °C, at least 7 °C, at least 8 °C, at least 9 °C or at least 10 °C greater than the Tmof a tetramer of the head region of the second neuraminidase. The Tmis typically calculated based on a construct comprising SEQ ID NO: 8 and the head region of the first neuraminidase in DPBS buffer with calcium, wherein the Tmis measured by nano differential scanning fluorimetry. The Tmis typically the melting temperature of the disassociation of the tetramer, i.e. to lower multimers or monomers. Other methods of measuring Tmare known in the art. The construct used in the calculation comprises, from N-terminus to C-terminus, SEQ ID NO: 8 and the head region of the first neuraminidase. The skilled person will understand that any tetramerisation domain or purification tags may be used in the expression of polypeptides of the invention. In some embodiments, the Tm may be calculated based on a construct comprising SEQ ID NO: 48 and the head region of the first neuraminidase in DPBS buffer with calcium, wherein the Tmis measured by nano differential scanning fluorimetry. The construct used in the calculation may comprise, from N-terminus to C-terminus, SEQ ID NO: 48 and the head region of the first neuraminidase.

[0204] The present invention provides a tetramer comprising four polypeptides that each comprise a neuraminidase head region, wherein at least one of the polypeptides is a polypeptide according to the invention. In some embodiments the tetramer comprises two or more polypeptides of the invention. In some embodiments, the tetramer comprises three or more polypeptides of the invention. In some embodiments, all four polypeptides are polypeptides according to the invention. The majority of antibodies generated in response to a neuraminidase bind within a single monomer, as opposed to having an epitope across two or more monomers of the tetramer. Furthermore, it is understood that the majority of the stability achieved by the polypeptide of the invention is based on the scaffold of the first neuraminidase. Accordingly, in embodiments wherein the tetramer comprises two or more polypeptides according to the invention, the polypeptides according to the invention may be different. For example, the different polypeptides according to the invention may comprise L01 and L23 loops of different second neuraminidases. The different polypeptides according to the invention may comprise the same first neuraminidase scaffold. In some embodiments, the polypeptides according to the invention in the tetramer are identical. In some embodiments, the tetramer is a homotetramer, i.e. comprises four identical polypeptides according to the invention.

[0205] Polynucleotide

[0206] The invention also provides a polynucleotide encoding the polypeptide of the invention or one or more polynucleotides encoding the tetramer of the invention. The polynucleotide may be a DNA polynucleotide or an RNA polynucleotide. The DNA polynucleotide may be a single-stranded or double-stranded DNA polynucleotide. The RNA polynucleotide may be a single-stranded or double-stranded DNA polynucleotide.

[0207] DNA polynucleotides typically comprise a combination of adenosine (A), guanosine (G), cytidine (C) and thymidine (T) nucleotides. In an RNA polynucleotide, the T nucleotides are typically replaced by uridine (U), which retains the ability to base pair with A. RNA polynucleotides therefore typically comprise a combination of A, C, G and U nucleotides. Other naturally-occurring, non-naturally occurring, modified and / or synthetic nucleotides may also be present in the polynucleotides described herein, particularly nucleotides that may improve stability, transcription or translation of the polynucleotide described herein. For example, the polynucleotide may include inosine (I), 5 ’methylcytidine (5meC), N6-methyladenosine, pseudouridine, N1 -methylpseudouridine, deoxyuridine (dU), abasic nucleotides, threose nucleotides (TNA), glycerol nucleotides (GNA), locked nucleotides (LNA) and peptide nucleotides (PNA). In some cases, then polynucleotide may comprise 5’methyl cytidine (5meC), N6-methyladenosine, pseudouridine and / or N1 -methylpseudouridine.

[0208] In some embodiments, the polynucleotide is codon optimised. The polynucleotide may be codon optimised depending on the expression system to be used. For example, the polynucleotide may be codon optimised for insect, yeast or animal expression. The polynucleotide may be codon optimised for expression in mammals, such as in humans or Cricetuleus griseus.

[0209] 31 The invention also provides a vector comprising a polynucleotide or one or more polynucleotides according to the invention. In some embodiments, the vector is a transposon, a plasmid, a virus or a phage vector.

[0210] In some embodiments, the vector is a particle comprising the RNA molecule and a pharmaceutically acceptable excipient or carrier. In some embodiments, the particle is a viral particle, virus-like particle, or lipid nanoparticle. In some embodiments, the RNA molecule is an mRNA molecule. In some embodiments, the mRNA molecule is for use as an mRNA therapeutic. In some embodiments, the viral particle is an adeno-associated virus (AAV), for example, for use in gene therapy.

[0211] In some embodiments, the present invention provides a virus comprising a polynucleotide of the invention. In some embodiments, the present invention provides a virus-like particle (VLP) comprising a polynucleotide of the invention. In some embodiments, the virus or VLP further comprises a polynucleotide encoding an influenza hemagglutinin. In some embodiments, the influenza hemagglutinin is the influenza hemagglutinin of a circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals. The circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals may be a strain identified by the WHO Global Influenza Programme (http s : / / www. who . int / tool s / flunet) .

[0212] The invention also provides a cell comprising the polypeptide, tetramer, polynucleotide or vector according to the invention. In some embodiments, the cell expresses the polypeptide or tetramer according to the invention. In some embodiments, the cell expresses a soluble form of the polypeptide or the tetramer. In some embodiments, the cell expresses a membrane-bound form of the polypeptide of the polypeptide or the tetramer. In some embodiments, the cell is a eukaryotic cell. In some embodiments, the cell is an insect cell, a yeast cell or a mammalian cell. In some embodiments, the cell is a human cell. In some embodiments, the cell is a non-human mammalian cell, such as a mouse, rat, hamster, cat, dog, pig, goat, sheep, horse, cow, camel or non-human primate cell. In some embodiments, the cell is a sf9 or sf21 cell. In some embodiments, the cell is a S. cerevisiae or a P. pastoris cell. In some embodiments, the cell is a HEK293 cell, for example a cell derived from the HEK293 cell line (e.g. a HEK293F or Expi293F cell). In some embodiments, the cell is a Chinese hamster ovary (CHO) cell, for example a cell derived from a CHO cell (e.g. an ExpiCHO cell). In some embodiments, the cell is a chicken cell. In some embodiments, the invention provides a chicken egg comprising the polypeptide, tetramer, polynucleotide or vector according to the invention, for instance, a viral vector comprising the polynucleotide of the invention.

[0213] Other products and compositions

[0214] In some embodiments, the present invention provides a virus comprising a polypeptide or a tetramer of the invention. In some embodiments, the present invention provides a virus-like particle (VLP) comprising a polypeptide or a tetramer of the invention.

[0215] In some embodiments, the virus is an influenza virus. In some embodiments, the virus further comprises an influenza hemagglutinin. In some embodiments, the influenza hemagglutinin is the influenza hemagglutinin of a circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals. The circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals may be a strain identified by the WHO Global Influenza Programme (https: / / www.who.int / tools / flunet). In some embodiments, the virus is a live-attenuated virus, such as a live-attenuated influenza virus. In some embodiments, the virus is an inactivated virus, such as an inactivated influenza virus. In some embodiments, the virus is a live virus, i.e. an infectious virus.

[0216] Any VLP may be used in the invention. The VLP may be an mi3 VLP, a hepatitis B core (HBc) and surface antigen (HBsAg) VLP, a bacteriophage QP VLP, an influenza VLP, a Newcastle disease virus VLP, a tobacco mosaic virus VLP, a hepatitis E virus VLP or a ferritin-based nanoparticle. VLPs are discussed in Mohsen & Bachmann (Cell Mol Immunol. 2022 Sep;19(9):993-1011). A VLP typically comprises multiple copies of the polypeptide or the tetramer of the invention. In some embodiments, the VLP further comprises an influenza hemagglutinin. In some embodiments, the influenza hemagglutinin is the influenza hemagglutinin of a circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals. The circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals may be a strain identified by the WHO Global Influenza Programme (https: / / www.who.int / tools / flunet). The polypeptide or tetramer may be conjugated to the VLP through any suitable means, such as via pairs of binding tags as described herein.

[0217] In some embodiments, the VLP is an mi3 VLP. The mi3 VLP may be a SpyCatcher-mi3 VLP, and the tetramer may comprise a SpyTag capable of forming an isopeptide bond with the SpyCatcher on the VLP. The mi3 VLP may be a SpyTag-mi3 VLP, and the tetramer may comprise a SpyCatcher capable of forming an isopeptide bond with the SpyTag on the VLP. Any SpyCatcher / SpyTag binding pair may be used in the invention. In some embodiments, the SpyCatcher / SpyTag binding pair is selected from the group consisting of SpyCatcher / SpyTag, SpyCatcher / SpyTag002, SpyCatcher / SpyTag003, SpyCatcher002 / SpyTag, SpyCatcher002 / SpyTag002, SpyCatcher002 / SpyTag003, SpyCatcher003 / SpyTag, SpyCatcher003 / SpyTag002, SpyCatcher003 / SpyTag003, SpyTag / Ktag, SnoopCatcher / SnoopTag, SnoopCatcher / SnoopTagJr, DogCatcher / DogTag, SnoopTagJr / DogTag, and Pilin-C / IsopepTag.

[0218] The invention also provides a composition comprising a polypeptide, tetramer, polynucleotide, particle, virus or VLP as described herein.

[0219] The composition may be an immunogenic composition. The immunogenic composition is capable of eliciting antibodies when administered to a host cell. The antibodies may be capable of binding to the influenza strain from which the second neuraminidase is derived. In some embodiments, the antibodies may be capable of neutralising the infection and / or virulence of the influenza strain from which the second neuraminidase is derived, as demonstrated within the examples of the application. The immunogenic composition may be capable of eliciting an immune response, e.g. a protective immune response, such as a cell-mediated and / or an antibody response, against the influenza strain from which the second neuraminidase is derived, such as Nl / 09 influenza.

[0220] The composition (e.g. immunogenic composition) of the invention may comprise an adjuvant. Any suitable adjuvant which enhances effectiveness of the composition may be used with the invention. The adjuvant may be an adsorbent. The adjuvant may be an adsorbent that does not enhance immunogenicity of the composition. The adjuvant may be an aluminium adjuvant, such as aluminium hydroxide, ALHYDROGEL®, aluminium phosphate, potassium aluminium sulphate and / or alum. A useful adjuvant for use with the invention may be oil-in-water emulsion formulations, such as for example MF59™ (containing 5% Squalene, 0.5% Tween 80, and 0.5% Span 85 formulated into submicron particles), SAF (containing 10% Squalane, 0.4% Tween 80, 5% pluronic-blocked polymer L121, and thr-MDP either microfluidized into a submicron emulsion or vortexed to generate a larger particle size emulsion), and Ribi™ adjuvant system (RAS) (containing 2% Squalene, 0.2% Tween 80, and one or more bacterial cell wall components from the group consisting of monophosphorylipid A (MPL), trehalose dimycolate (TDM), and cell wall skeleton (CW S), e.g. MPL+CWS (Detox™)). A useful adjuvant for use with the invention may be saponin adjuvants, such as Stimulon™ or particles generated therefrom such as ISCOMs (immunostimulating complexes). Other useful adjuvants include Complete Freund's Adjuvant (CFA) and Incomplete Freund's Adjuvant (IF A), cytokines, such as interleukins (eg. IL-1, IL-2, IL-4, IL-5, IL-6, IL-7, IL-12, etc.), interferons (eg. gamma interferon), macrophage colony stimulating factor (M-CSF), tumor necrosis factor (TNF), alum, etc; and other substances that act as immunostimulating agents to enhance the effectiveness of the composition.

[0221] The composition (e.g. immunogenic composition) may be a pharmaceutical composition. The composition may further comprise a pharmaceutically acceptable excipient, carrier or diluent. The components of the pharmaceutical compositions are capable of being co-mingled with the polypeptide, tetramer, polynucleotide, particle, virus or VLP of the present invention, and with each other, in a manner such that there is no interaction which would substantially impair the desired pharmaceutical efficacy. Preferred pharmaceutical compositions are sterile and pyrogen free.

[0222] In some embodiments, the composition further comprises an additional therapeutic agent. In some embodiments, the additional therapeutic agent is a vaccine. In some embodiments, the additional therapeutic agent is an immunomodulator.

[0223] The composition of the invention may comprise one or more pharmaceutically acceptable salts. A "pharmaceutically acceptable salt" refers to a salt that retains the desired biological activity of the parent compound and does not impart any undesired toxicological effects. Examples of such salts include acid addition salts and base addition salts. As used herein, a pharmaceutically acceptable salt is a salt with a pharmaceutically acceptable acid or base. Pharmaceutically acceptable acids include both inorganic acids such as hydrochloric, sulphuric, phosphoric, diphosphoric, hydrobromic or nitric acid and organic acids such as oxalic, citric, fumaric, maleic, malic, ascorbic, succinic, tartaric, benzoic, acetic, methanesulphonic, ethanesulphonic, benzenesulphonic or p- toluenesulphonic acid. Pharmaceutically acceptable bases include alkali metal (e.g. sodium or potassium) and alkali earth metal (e.g. calcium or magnesium) hydroxides and organic bases such as alkyl amines, aralkyl amines and heterocyclic amines.

[0224] The composition of the invention may comprise a pharmaceutically-acceptable carrier. The pharmaceutically-acceptable carrier includes any carrier that does not itself induce the production of antibodies harmful to the individual receiving the composition Suitable carriers are typically large, slowly metabolized macromolecules such as proteins, polysaccharides, polylactic acids, polyglycolic acids, polymeric amino acids, amino acid copolymers, lipid aggregates (such as oil droplets or liposomes), and inactive virus particles. Such carriers are well known to those of ordinary skill in the art. Additionally, these carriers may function as immunostimulating agents, which may also be referred to as adjuvants.

[0225] Typically, the compositions of the invention are prepared as injectables, either as liquid solutions or suspensions, solid forms suitable for solution in, or suspension in, liquid vehicles prior to injection may also be prepared. The compositions may be lyophilized. Liquid formulations allow the compositions to be administered directly from their packaged form, without the need for reconstitution in an aqueous medium. In the embodiment where the composition is in a lyophilized form and requires reconstitution, the composition can be provided in the form of a kit which can comprise two vials, or can comprise one ready-filled syringe and one vial, with the contents of the syringe being used to reconstitute the contents of the vial prior to administration to a subject.

[0226] The composition of the invention may be provided as a kit comprising instructions to enable the kit to be used in the methods and medical uses described herein or details regarding which subjects the method may be used for.

[0227] Compositions can be presented in vials, or they can be presented in ready-filled syringes. The syringes can be supplied with or without needles. A syringe will include a single dose of the composition, whereas a vial can include a single dose or multiple doses as described herein. Kits can include a measured dose for administration to a subject. Therapy

[0228] The polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention are useful in therapy. Methods are disclosed herein for inducing an immune response to influenza in a subject using any of the disclosed compositions. The immune response can be a protective immune response. Accordingly, provided herein is a polypeptide, tetramer, polynucleotide, particle, virus, VLP or a composition as described herein, for use in medicine. The invention further relates to the use of a polypeptide, tetramer, polynucleotide, particle, virus, VLP or a composition described herein in a method for treatment of the human or animal body by therapy. The invention further relates to the use of a polypeptide, tetramer, polynucleotide, particle, virus, VLP or a composition described herein in the manufacture of a medicament.

[0229] As explained above, the polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions provided herein the invention are useful in treating or preventing influenza infection or an influenza related disorder. Hence, the invention also provides a method of treating or preventing influenza infection or an influenza related disorder in a subject, wherein the method comprises administering to said subject an effective amount of a polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition as described herein. Further provided is use of a polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition as described herein for the manufacture of a medicament for use in treating or preventing influenza infection or an influenza related disorder.

[0230] The polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition of the invention may be administered to a subject in need thereof in order to prevent the onset or reoccurrence of one or more symptoms of the influenza infection. This is prophylaxis. Hence, the invention also provides a vaccine against influenza, such as Nl / 09 influenza. The vaccine may have a vaccine efficacy against influenza of at least 10%, e.g. >20%, >30%, >40%, >50%, >60%, >70%, >80%, >85%, >90%, or more.

[0231] The polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition of the invention may be used in combination with other agents. For example, the polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition of the invention may be used in combination with vaccines against other influenza viruses.

[0232] The immunogenic composition may be co-immunised with one or more further vaccines. The different vaccines can be administered either separately or as a combination. For separate administration, the vaccines will typically be administered at different sites e.g. one vaccine to the left upper arm, and a second vaccine to the right upper arm. Although the vaccines are administered separately, they may be administered at substantially the same time (e.g. during the same medical consultation or visit to a healthcare professional or vaccination centre), such as within 1 hour of each other. For administration for the multiple vaccines as a combination, a combination vaccine i.e. a single composition in which the different immunogens are admixed, may be used. Administration may involve a single dose schedule, but will usually involve a multiple dose schedule, such as in a prime boost protocol. An initial dose and an additional dose can be administered within days, weeks, or months of each other. Suitable intervals between priming doses can be routinely determined e.g. between 4-16 weeks, such as one month or two months. The initial administration of the mixture can be followed by booster immunization of the same of different mixture, with at least one booster, such as two boosters. The method can include administering 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10 doses.

[0233] The immunogenic composition of the invention may be administered parenterally, e.g. by injection, either subcutaneously, intramuscularly, or transdermally. Additional formulations suitable for other modes of administration include oral and pulmonary formulations, suppositories, and transdermal applications.

[0234] The immunogenic composition of the invention may be administered in a variety of dosage forms. Thus, it can be administered orally, for example as tablets, troches, lozenges, aqueous or oily suspensions, dispersible powders or granules.

[0235] The subject to be treated is a mammal, in particular a human. However, it may be non-human. Preferred non-human animals include, primates, such as marmosets or monkeys, commercially farmed animals, such as horses, cows, sheep or pigs, and pets, such as dogs, cats, mice, rats, guinea pigs, ferrets, gerbils or hamsters.

[0236] The subject can be any animal that is capable of being infected by a bacterium as described in more detail herein.

[0237] For prophylactic treatment, the subject is typically asymptomatic.

[0238] For example, the subject who is to be immunized is a human being, who may be any age e.g. 0-12 months old, 1-5 years old, 5-18 years old, 18-55 years old, or more than 55 years old. In one embodiment, the subject who is immunized is an adolescent (e.g. 12- 18 years old) or an adult (18 years or older). The subject may be an adolescent or adult who has been immunized against influenza previously, or immunized against another strain of influenza, and who receives a booster dose of an immunogenic composition according to the invention to protect against the influenza strain from which the second neuraminidase is derived.

[0239] The subject who is to be immunized may be at increased risk of infection with influenza (e.g. at increased risk relative to the average risk in the general population). A prophylactically effective amount of the agent or composition may be administered to a subject. The dose may be determined according to various parameters, especially according to the compound used; the age, weight and condition of the subject to be treated; the route of administration; and the required regimen. Again, a physician will be able to determine the required route of administration and dosage for any particular subject. For example, the immunogenic composition may comprise any suitable amount of polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition per unit dose. Suitable amounts of the polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition may be from 0.1 to 200 pg per unit dose, particularly 10 pg, 20 pg, 25 pg, 50 pg or 100 pg. Per unit dose, aqueous immunogenic compositions of the invention may comprise a total concentration of polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition of less than 200pg / ml, less than lOOpg / ml or less, 80pg / ml or less, 50pg / ml or less, 25pg / ml or less, 20pg / ml or less, 15pg / ml or less, lOpg / ml or less. Per unit dose, aqueous immunogenic compositions of the invention may comprise a total concentration of polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition of from 5pg / ml to 200pg / ml, from 5pg / ml to lOOpg / ml, from lOpg / ml to lOOpg / ml, from lOpg / ml to 80pg / ml, from lOpg / ml to 50pg / ml, 25pg / ml to 50pg / ml. Per unit dose, immunogenic compositions of the invention may comprise a total concentration of polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition of more than lOOpg / ml, more than 80pg / ml, more than 50pg / ml, more than 25pg / ml, more than 20pg / ml, more than 15pg / ml or more than lOpg / ml.

[0240] Further applications

[0241] The polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention may be used as diagnostic reagents. Hence, the invention also provides the use of a polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition according to the invention in the manufacture of a diagnostic reagent for detecting the presence of influenza, such as Nl / 09 influenza. For example, the polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention can be used in immunoassays to detect levels of antibodies to influenza, e.g. within biological samples, including for example, blood or serum samples. The polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention may be used to raise antibodies against neuraminidase, such as Nl / 09 neuraminidase. Hence, the invention also provides the use of a polypeptide, tetramer, polynucleotide, particle, virus, VLP or composition according to the invention in the manufacture of antibodies raised against neuraminidase (such as Nl / 09 neuraminidase). The anti-neuraminidase antibodies can be used in immunoassays to detect antigen levels, e.g. within biological samples, including for example, blood or serum samples.

[0242] Appropriate protocols for the immunoassays described herein can be determined by the person skilled in the art. For example, the immunoassay may be based, for example, upon competition, direct reaction, or sandwich type assays. The polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention or the antibodies raised from polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention may be labelled e.g. using a probe, and the probe may be, for example, fluorescent, chemiluminescent, radioactive, or dye molecules. Appropriate assays which amplify the signals from the probe are known to the person skilled in the art; examples of which are assays which utilize biotin and avidin, and enzyme-labelled and mediated immunoassays, such as ELISA assays.

[0243] Kits suitable for immunodiagnosis are also provided, wherein the kit comprises the appropriate labelled reagents, including the polypeptides, tetramers, polynucleotides, particles, viruses, VLPs and compositions of the invention, in suitable containers, along with the remaining reagents and materials (for example, suitable buffers, salt solutions, etc.) required for the conduct of the assay, as well as suitable set of assay instructions.

[0244] Method

[0245] The present invention provides a method of producing a chimeric protein. The method comprises aligning the amino acid sequence of the head region of a first neuraminidase and a second neuraminidase to identify the L01 and L23 loops. The method further comprises substituting the L01 and L23 loops of the first neuraminidase with the L01 and L23 loops of the second neuraminidase.

[0246] In another embodiment, the method may comprise aligning the amino acid sequence of the head region of a first neuraminidase with a reference sequence, and the head region of second neuraminidase with a reference sequence, which the reference sequence is annotated with loops, to thereby identify the L01 and L23 loops. The method may further comprises substituting the L01 and L23 loops of the first neuraminidase with the L01 and L23 loops of the second neuraminidase.

[0247] The skilled person will understand that they are many methods and algorithms suitable for alignment, such as those described herein. The L01 and L23 loops of the first neuraminidase will typically be annotated because the first neuraminidase is to be used as a platform / scaffold on which loops of a circulating influenza strain are added. However, in some embodiments, the aligning step comprises aligning the amino acid sequence of the head region of a first and second neuraminidase to the head region of one or more reference neuraminidase head regions to thereby identify the L01 and L23 loops. The reference neuraminidase head region may vary depending on the use in the claim. In some embodiments, the reference neuraminidase may be of the same neuraminidase group or subtype as the first or second neuraminidase. For example, where the first and second neuraminidase are N1 neuraminidases, the reference neuraminidase may be an N1 neuraminidase. In some embodiments, the reference neuraminidase is the N2 neuraminidase annotated in Varghese et al (Nature 303, 35-40 (1983)). In some embodiments, alignment may be performed by aligning the amino acid sequences of the neuraminidases. Any suitable sequence alignment may be used, such as those described herein. In some embodiments, structural alignment may be performed by aligning the structures of the neuraminidase proteins. In some embodiments, alignment may be performed by predicting the structure of the first or second neuraminidase based on the structure of a reference neuraminidase. Typically, the L01 and L23 loops are identified in the aligning step by determining the amino acid residues of the first and / or second neuraminidase that correspond to the L01 and L23 loops in an annotated reference sequence, for example, by sequence-based or structure-based alignment.

[0248] The substituting step defines the desired result, i.e. a chimeric protein comprising the L01 and L23 loops of the second neuraminidase and the remaining sequence of the head region (e.g. the scaffold) from the first neuraminidase. This may be achieved by any means known to the skilled person. In some embodiments, the L01 and L23 loops of the first neuraminidase are substituted with the L01 and L23 loops of the second neuraminidase. In some embodiments, the scaffold of the second neuraminidase is substituted with the scaffold of the first neuraminidase. The resulting chimeric protein is the same. In some embodiments, substitution comprises replacing the entirety of the sequences to the substituted. In some embodiments, substitution comprises only replacing the sequences that in the L01 and L23 loops that differ between the first and second neuraminidases.

[0249] Typically, the substitution step is performed at the polynucleotide level. In some embodiments, the substitution step comprises substituting the nucleic acid sequences encoding the L01 and L23 loops of the first neuraminidase and with the nucleic acid sequences encoding the L01 and L23 loops of the second neuraminidase. This results in a polynucleotide encoding a chimeric protein comprising the head region scaffold of the first neuraminidase and the L01 and L23 loops of the second neuraminidase. In some embodiments, the method further comprises expressing the polynucleotide to produce the chimeric protein. Any suitable expression system may be used, such as the cells described herein. In some embodiments, the chimeric protein is a polypeptide of the invention.

[0250] In some embodiments, the method further comprises identifying the B5L34 loop of the first neuraminidase and the second neuraminidase, and substituting the B5L34 loop of the first neuraminidase with the B5L34 loop of the second neuraminidase. In some embodiments, the method further comprises identifying the CTD of the first neuraminidase and the second neuraminidase, and substituting the CTD of the first neuraminidase with the CTD of the second neuraminidase. The B5L34 loop and / or the CTD may be aligned and / or identified as described herein, e.g. as for the L01 and L23 loops.

[0251] In some cases, the method further comprises codon optimising the nucleic acid sequence encoding the first neuraminidase, the second neuraminidase or the chimeric protein. The nucleic acid sequences may be codon optimised depending on the expression system to be used. For example, the nucleic acid sequences may be codon optimised for insect, yeast or animal expression. The nucleic acid sequences may be codon optimised for expression in mammals, such as in humans or Cricetuleus griseus. In some embodiments, the invention also provides a chimeric protein produced by the method.

[0252] The present method is pragmatic and easily applied to any given NA subtype. This simple technique improved expression of stable N 1 NA proteins and could form a starting point for further improvement by structural analysis. The present method is useful for NA- based vaccine design and manufacturing, as well as drug and antibody screening.

[0253] The following examples illustrate the invention

[0254] Examples

[0255] EXAMPLE 1

[0256] Design of loop-grafted NA hybrid proteins

[0257] NA proteins were expressed in a transient mammalian ExpiCHO expression system using a gene construct comprising the N-terminus signal sequence, purification tags, SpyTag, a tetrabrachion tetramerization domain (SEQ ID NO: 8) and the NA head ectodomain. The protein was purified from clarified supernatants and analysed using sizeexclusion chromatography, SDS-PAGE with BS3 cross linking, thermal unfolding, and enzymatic assays. We found that H5N1 A / mute swan / England / 053054 / 2021 NA (mSNl) expressed at -380 mg / L as a tetramer with minimal aggregation, exhibiting high enzymatic activity and a high melting temperature (Fig. 3 a). However, H1N1 A / California / 07 / 2009 NA (Nl / 09) expressed at -15 mg / L with a tendency to form aggregates, and H1N1 A / Wisconsin / 588 / 2019 NA (Nl / 19) had a variable expression from undetectable to low (performed twice).

[0258] Using the PDB 1NN2 (H2N2 A / Tokyo / 3 / 1967) structure and the top surface loop annotations by Varghese et al. (Varghese, J. N., Laver, W. G. & Colman, P. M. Structure of the influenza virus glycoprotein antigen neuraminidase at 2.9 A resolution. Nature 303, 35-40 (1983)) as a guide, we annotated the twelve L01 and L23 loops of the N1 neuraminidases (Fig. 1 a,b, Fig. 2, Table 1). As described by Varghese et al., the top surface is formed by the loops interconnecting the four anti-parallel P-strands arranged in a "W" topology (Fig. la). The loop preceding P-strand 1, which connects with the fourth strand of the preceding P-sheet is termed LoopOl, and the loop connecting P-strands 2 and 3 is termed Loop23. The twelve 01 and 23 loops in each monomer form the top antigenic surface, and surround and contribute to the enzyme active site. We transferred 12 dissimilar residues within the loops of Nl / 09 and 16 dissimilar residues of Nl / 19 to the H5N1 / 2021 scaffold (mSNl) to create Nl / 09 and Nl / 19 hybrids (Fig. 2a, b). From now on, Nl / 09Loops-mS Scaffold and Nl / 19Loops-mS Scaffold will be referred to as the Nl / 09 hybrid and Nl / 19 hybrid, respectively. The C-terminal domain (CTD) of the final ~17 amino acids of the N1 protein also forms part of the top surface. However, in these N1 NAs the CTD was conserved.

[0259] Nl / 09 and Nl / 19 hybrid proteins expressed at higher level as stable tetramers

[0260] There was a substantial increase in protein yield for both Nl / 09 and Nl / 19 hybrid proteins (Fig. 3a). These hybrids expressed as tetramers with negligible aggregation, as demonstrated by size-exclusion chromatography (Fig. 3 a) and cross-linked tetrameric complex of -250 kDa resolved on SDS-PAGE (Fig. 9). The hybrids were enzymatically active in both small substrate MUNANA (489 Da) and large substrate fetuin (49 kDa) based assays. The Nl / 09 hybrid showed an 8°C improvement in its melting temperature Tmto 64.4°C compared to the Nl / 09 protein (56.4°C), while the Tmof the Nl / 19 hybrid (72.4°C) surpassed the H5N1 mSNl scaffold donor (68.7°C). The 4.3°C difference in melting temperature between the mSNl protein (68.7°C) and the Nl / 09 hybrid (64.4°C) may indicate that the sequence variation in the loops also contributes to stability of the tetramer.

[0261] Binding by mAb CD6 (Ellis, D. et al. Structure-based design of stabilized recombinant influenza neuraminidase tetramers. Nat Commun 13, 1825 (2022); Wan, H. et al. Structural characterization of a protective epitope spanning A(HlNl)pdmO9 influenza virus neuraminidase monomers. Nature Communications 6, 1-10 (2015)) was also informative as CD6 binds across two protomers of the tetrameric NA and makes contacts with both loop and scaffold residues. It has previously been shown that binding of CD6 is dependent on the formation of compact closed conformation tetramers, whereas a more open conformation N1 2009 tetramer failed to bind CD6.

[0262] Further, we compared the binding of CD6 to N1 / 09-VASP (with human vasodilator stimulated phosphoprotein tetramerisation domain) and N1 / 09-TB [with tetrabrachion (TB) tetramerisation domain] proteins expressed in ExpiCHO cells and found that CD6 recognised only the latter (Fig. 10). CD6 bound better to Nl / 09 infected cells than to the recombinant N1 / 09-TB protein (Fig. 3b, Fig. 11). This aligns with Ellis et al.'s findings that the 2009 N1 protein (with a VASP tetramerisation domain) predominantly formed open tetramers that failed to bind CD6. Our recombinant N 1 / 09 protein (with TB domain) may have formed a mixture of open and closed tetramers, also suggested by the appearance of aggregates on size-exclusion chromatography and the relatively low Tmof 56.4°C (Fig.

[0263] 3a). In contrast, CD6 bound strongly to the Nl / 09 hybrid protein, which had a higher Tmof 64.4°C. CD6 also bound strongly to mSNl.

[0264] The CD6 epitope was lost on seasonal H1N1 viruses isolated after 2015 (unpublished observation) so CD6 bound to cells infected with Nl / 09 but not Nl / 19 virus, but CD6 bound to the Nl / 19 hybrid protein strongly, suggesting that CD6's binding to the Nl / 19 hybrid depended on the scaffold residues donated by the msNl protein, and that the Nl / 19 hybrid protein had formed a compact tetramer compatible with the Nl / 19 hybrid's higher Tmof 72.4°C (Figure 3a).

[0265] Nl / 09 and Nl / 19 hybrid proteins retained epitope specificity and enzyme activity

[0266] Next, we assessed the NA hybrid proteins for the epitopes recognised by monoclonal antibodies isolated from humans infected with influenza virus or vaccinated with inactivated seasonal influenza vaccine. A comprehensive analysis was done using a collection of twenty-five human mAbs, isolated by us from vaccinated or infected individuals or synthesized in the laboratory based on published sequences (Fig. I la). Results from the binding titration of fifteen mAbs on mSNl and hybrid proteins and NA on the virus-infected cell membrane are depicted in Fig. 3b. These included a set of five broadly reactive mAbs, including four, 1G01, Z2B3, AG7C and with crystal structures showing binding to the L01 and L23 loops, and nine strain specific mAbs that distinguished between loop donor and recipient. Cross-reactive mAbs showed full binding to all proteins - mSNl, loop-donors, and hybrid proteins. Seven mAbs which did not bind the mSNl protein but bound Nl / 09, also demonstrated binding to the Nl / 09 hybrid protein, showing their specificity for the transferred loops. Similarly, NmAb-03, displaying specific binding to Nl / 19 NA on infected cells, exclusively bound to the Nl / 19 hybrid protein. By contrast, mAb Z2C2 which is specific for mSNl and Nl / 09, bound neither Nl / 19-infected cells nor the Nl / 19 hybrid protein. These results showed that for a wide collection of cross reactive and specific mAbs, isolated from influenza infected or vaccinated humans, binding was retained on the hybrid proteins. Together these results suggested that the tertiary structure of the loops had been retained after grafting onto the msNl scaffold.

[0267] NA inhibiting drugs, oseltamivir and zanamivir, inhibited the NA activity of the hybrid proteins (Fig 9). Similarly, mAb 1G01, a known catalytic site targeting mAb also inhibited the activity of mSNl and Nl / 09 hybrid but not Nl / 19 hybrid. This result was expected since the mAb 1G01 lost its inhibitory activity on 2019 H1N1 seasonal influenza11due to the substitution N222K (N1 numbering) located within the loop B3L01 (unpublished observation, Fig. 3b and Fig. 9). While enzyme inhibition was lost, detectable binding of 1G01 was retained to Nl / 19 cells and Nl / 19 hybrid proteins (Fig. 3b, Fig. I la). msNl, Nl / 09 hybrid and Nl / 19 hybrid crystal structures are nearly identical

[0268] To assess whether loop grafting influenced the local or overall structure of the neuraminidase, we determined high resolution crystal structures of msNl, the Nl / 09 hybrid, and the Nl / 19 hybrid (Fig. 4, Table 2). Overlay analysis revealed that Nl / 09 hybrid and Nl / 19 hybrid were nearly identical in structure to msNl, with overall rootmean-square (RMS) deviations of <0.25 A across equivalent Ca atoms upon overlay (Fig. 4b). This is consistent with previous findings, where N1 neuraminidases within a subtype typically exhibit an RMS deviation between 0.2-0.4 A (Xu, X., Zhu, X., Dwek, R. A., Stevens, J. & Wilson, I. A. Structural Characterization of the 1918 Influenza Virus H1N1 Neuraminidase. Journal of Virology 82, 10493-10501 (2008); Sun, X. et al. Structure of Influenza Virus N7: the Last Piece of the Neuraminidase “Jigsaw” Puzzle. J Virol 88, 9197-9207 (2014)). All segments of the scaffold region in both NA hybrids superposed extremely well onto msNl, with RMS deviation values of less than 0.5 A between the most distantly aligned Ca residue pairs. Minor deviations were observed in the B1L23 (150- loop) and B6L23 (430-loop) regions bordering the active-site cavity, with the RMS deviation distances of up to ~1.0 A between aligned Ca residue pairs. The remaining loops showed RMS deviation values of less than 0.5 A.

[0269] While the scaffold is structurally conserved across the three resolved structures, each possess unique active site conformations, which are primarily influenced by the L01 and L23 loops surrounding the active site cavity. Indeed, analysis of msNl, Nl / 09 hybrid, and Nl / 19 hybrid structures revealed some variability in active site conformations, with the greatest variability existing between the cavities of two msNl molecules within the asymmetric unit of the crystal, reflecting different trajectories of the B1L23 (150-loop) (Fig. 4c). Additionally, we observed that the Y344N substitution in the Nl / 09 and Nl / 19 hybrids slightly widened the enzymatic cavity compared to msN 1.

[0270] The structural and monoclonal antibody binding data thus showed that the twelve or sixteen substitutions in loops matching the 2009 and 2019 seasonal Nls grafted onto the avian N1 scaffold did not affect the protein fold. Indeed, the NA hybrid proteins exhibited a high degree of structural conservation, consistent with all previously published crystal structures of N1 neuraminidase proteins.

[0271] Table 2 Crystallographic data collection and refinement statistics.

[0272] DATA COLLECTION mSNl Nl / 09 hybrid Nl / 19 hybrid

[0273] Beamline DLS 124 DLS 124 DLS 124 Wavelength (A) 0.6199 0.6199 0.6199 Space Group 4 4 C 2 2 2i Cell Dimensions a, b, c (A) 92.2, 92.2, 147.9 92.1, 92.1, 146.8 115.5, 234.4,

[0274] 115.7 a, P, Y (°) 90, 90, 90 90, 90, 90 90, 90, 90

[0275] Resolution range (A) 147.90-1.98 [2.01- 41.18-1.94 [1.97- 82.31-1.78 [1.98- 1.98] 1.94] 1.78]

[0276] Rmerge 0.766 [2.784] 0.792 [3.109] 0. 367 [1.705] l / o (I) 2.8 [0.7] 4.1 [0.7] 5.4 [1.6] CC1 / 2 0.976 [0.342] 0.983 [0.282] 0.988 [0.564] Completeness (%) 100 [99.8] 99.9 [98.3] 94.2 [65.2]*

[0277] Multiplicity 14.2 [14.1] 13.9 [10.9] 13.4 [12.9]

[0278] REFINEMENT

[0279] Resolution (A) 92.17-1.98 41.18 -1.94 64.69-1.78

[0280] No. reflections 83,874 89,971 58,510

[0281] Rwork / Rfree 0.211 / 0.238 0.183 / 0.213 0.233 / 0.267

[0282] No. atoms protein 6,296 6,279 6,164 ligands 127 102 116 solvent 619 695 474 Average B-factors protein 29.3 24.2 18.23 ligand 65.4 66.3 60.12 solvent 35.8 33.6 25.0

[0283] Ramachandran (%) favoured 96.21 96.32 96.22 allowed 3.55 3.68 3.58 outlier 0.24 0 0.25

[0284] RMS bond lengths (A) 0.003 0.013 0.004 bond angles (°) 0.67 1.16 0.71

[0285] * *Ellipsoidal completeness as determined by STARANISO The values between brackets are for the highest-resolution shell

[0286] Nl / 09 and Nl / 19 hybrid proteins are immunogenic and elicit NA enzyme inhibiting (NAI) antibody responses

[0287] We have previously shown that recombinant NA presented on a mi3 particle is strongly immunogenic at doses as low as 0.1 pg (Rahikainen, R. et al. Overcoming Symmetry Mismatch in Vaccine Nanoassembly through Spontaneous Amidation. Angewandte Chemie - International Edition 60, 321-330 (2021)), in contrast to 10-100 fold higher doses of free NA protein often described in the literature. Our recombinant NA proteins were coupled onto mi3 virus-like particles (NA-VLP) via the covalent bond between the SpyTag on the NA protein and SpyCatcher003 sequence on the mi3 protein (Fig 9d). Mice were immunised with 0.5 pg NA-VLP adjuvanted with 1 : 1 vol / vol Addavax™ (squalene-based oil-in-water nano-emulsion). Intramuscular immunisations were administered twice at three-week intervals, and sera were harvested three weeks postbooster dose to assess the antibody response. Neuraminidase activity inhibition (NAI) IC50 titres were measured using a fetuin-based Enzyme-Linked Lectin Assay (ELLA).

[0288] All NA proteins were immunogenic in mice, generating binding and NA inhibiting antibodies (NAI) to themselves (Fig. 5a). mSNl elicited NAI antibody titres against itself that cross inhibited viral Nl / 09 NA. However, Nl / 09 protein did not induce a cross- reactive response against mSNl, although the Nl / 09 hybrid did elicit antibodies that inhibited mSNl. This difference may have been due to improved structural integrity of the Nl / 09 hybrid, or possibly due to induction of inhibitory antibodies to the mSNl scaffold. The mSNl and the Nl / 19 hybrid both induced strong inhibitory antibodies to themselves, but the sera to mSNl failed to cross-inhibit the enzyme activity of Nl / 19-like virus (Fig. 12b, c) and the Nl / 19 hybrid protein did not generate a cross-inhibitory NAI antibody response to mSNl. So, in this case the common mSNl scaffold did not generate cross-inhibitory antibodies. Nl / 19 had four significant additional amino acid substitutions (see discussion below) compared to Nl / 09 in the transferred L01 and L23 loops which may have been responsible for the loss of cross reactivity between 2019 N1 with mSNl. However, the Nl / 19 hybrid did generate antibodies that cross-inhibited Nl / 09 virus, so all four immunogens mSNl, Nl / 09, Nl / 09 hybrid and, Nl / 19 hybrid induced antibodies that could inhibit 2009 viral neuraminidase.

[0289] Murine Challenge Studies

[0290] We challenged immunised DBA / 2 mice with a lethal dose of H1N1 / 2009 (X-179A) virus, and monitored weights for at least 14 days. Mice immunised with empty VLP lost >20% initial weight and hence were culled within 5-7 days post-infection. Vaccination with all the tested NA proteins mSNl, Nl / 09, Nl / 09 hybrid and Nl / 19 hybrid protected mice from severe weight loss on challenge with X-179A. These results mirrored several studies in the literature which showed that immunisation with the 2009 N1 could provide at least partial protection in mice and ferrets to the avian H5N1 challenge (Easterbrook, J. D. et al. Virology 432, 39-44 (2012); Sandbulte, M. R. et al. PLoS Medicine 4, 0265-0272 (2007); Rockman, S. et al. J Virol 87, 3053-3061 (2013)), and is compatible with the well characterised human mAbs that cross-inhibit broadly within the N1 subtype.

[0291] In this case all four immunogens induced inhibitory antibody to 2009 viral NA, resulting in protection from challenge with H1N1 / 2009 virus.

[0292] Although there is broad cross reactivity of antisera within the N1 subtype, it is not absolute as demonstrated above with mSNl and the Nl / 19 hybrid which differed by 16 amino acids in the transferred loops, compared to 12 differences for the Nl / 09 hybrid. We noted that the Cambridge strain of H1N1 A / PR / 8 / 1934 and mSNl differed by 18 residues in the L01 and L23 loops which may be sufficient to prevent cross-inhibition. We then exchanged the L01 and L23 loops between A / PR / 8 / 34 and mSNl to look for correlation between antibody cross-reactivity and protection with the source of the L01 and L23 loops. Loop transfer between two distant N1 NAs: H5N1 A / mute swan / England / 053054 / 2021 (mS) and H1N1 A / PR / 8 / 1934 Cambridge Strain (PR8)

[0293] We exchanged the L01 and L23 top surface loops between mS and PR8 which differ by 18 residues (Fig. 6a). Unlike the Nl / 09 and Nl / 19 proteins, PR8 N1 expressed well (-105 mg / L; Tm 55.8°C) as tetramers with minimal aggregation (Refer to SDS-PAGE in Fig. 9). PR81oops-mS (comprising Loops 01 and 23 from PR8 combined with the mSNl scaffold) expressed at a nearly equal yield (108 mg / L; Tm 58.1°C) and showed minimal aggregation (Fig. 6b). Conversely, msLoops-PR8 (consisting of Loops 01 and 23 from mSNl combined with the PR8 scaffold) gave a lower yield at (-22 mg / L; Tm 64.3°C) but also showed a single peak in melting temperature and assembled into a tetramer with minimal aggregation (Fig. 6b).

[0294] Epitope specificity between loop-exchanged mS and PR8 hybrid proteins

[0295] We assessed the epitope specificity of these proteins, following the methods applied for Nl / 09 and Nl / 19 hybrids. We titrated 17 mAbs for binding to the original and loops- exchanged NA proteins, with results for representative 11 mAbs shown in Fig. 1 lb. Ten cross-reactive mAbs showed full binding to all proteins - PR8 N1 and mSNl and their loops-exchanged variants. Among specific mAbs, NmAb-03 specifically bound to PR8 N1 and PR8Loops-mS but did not bind to mSNl or mSLoops-PR8. Similarly, NmAb-02 and - 13 are mS Loops-specific and did not bind to PR8 N1 or PR8Loops-mS. CD6 is a scaffolddependent mSNl binder and did not bind to PR8 N1 NA. CD6 maintained its binding to mSNl scaffold and to PR8Loops-mS but not to mSLoops-PR8, indicating its dependence on scaffold residues from the H5N1 donor. mAbs NmAb-20 and Z2C2 were exceptions. NmAb-20 bound mSNl and mSLoops-PR8 suggesting that it was mSNl loops specific. However, in addition d gained binding to PR8Loops-mS. mAb Z2C2 bound to both PR8 and mSNl, but lost binding to PR8Loops-mS hybrid, implying that this epitope had been lost in the hybrid protein. These two exceptions out of 17 mAbs studied in detail suggest that the transferred loops may show a limited amount of difference to either donor, perhaps at the margins of the loops. mS and PR8 loop-exchanged hybrid proteins elicited loop-specific NA inhibiting antibodies and provided loop-specific protection in vivo against virus challenge

[0296] BALB / c mice were immunised with 0.5 pg NA-VLP adjuvanted with AddaVax™ (Fig. 7a), as previously described for Nl / 09 in vivo experiments. mSNl and mSLoops-PR8 generated equivalent NAI sera titres towards mSNl, whereas PR8 N1 and PR8Loops-mS elicited no detectable titres to mSNl (Fig. 7b). Similarly, PR8 and PR8Loops-mS Scaffold generated equivalent titres against PR8 virus (Fig. 7c). Interestingly, mSLoops-PR8 Scaffold elicited a low NAI titre (albeit ~10-fold lower compared to PR8 Nl; p=0.0009) against PR8 virus, implying that in this combination some inhibitory antibody may have been generated either to the PR8 scaffold or to conserved epitopes in the L01 and L23 loops.

[0297] In an independent experiment, we challenged immunised mice with a lethal dose of PR8 virus (Cambridge strain) (Fig. 7d). Weight-survival curves showed that 6 / 6 mice immunised with PR8 NA and PR8Loops-mSScaffold NA survived without weight loss, whereas all mice vaccinated with the mSLoops-PR8Scaffold reached the endpoint, similar to the negative control group immunised with empty VLP (Fig. 7f,g). For mice immunised with mSNl NA, 2 / 6 mice survived but with significant weight loss. These experiments demonstrated that survival matched pre-exposure to vaccine presenting the PR8 L01 and L23 loops. The serology of vaccinated mice showed the same pattern (Fig. 7e). PR8 and PR8Loops-mS Scaffold generated inhibitory titres to PR8 virus NA, but mSNl and m SLoop s-PR8 Scaffold elicited no detectable NAI titres to PR8 virus (<160) in this experiment. The majority of the NAI serum titres were generated against the loops and this matched protection in this pair of NAs that differed by 18 residues in the L01 and L23 loops.

[0298] Discussion

[0299] Influenza neuraminidase protein is an attractive target for influenza vaccines and therapeutics, yet has been beset by low yields and stability, and variations due to strain differences. We have devised a loop grafting technique to improve the yield and stability NA protein in vitro. Our Loop-grafting concept is straightforward and may be broadly applicable within other NA subtypes. The role of the tetramerisation domain

[0300] In preliminary experiments we compared the VASP and Tetrabrachion (TB) tetramerisation domains for expression and immunisation with the 2009 N1 protein. We found that the VASP domain supported the formation of a NA tetramer but the tetrameric protein did not bind the CD6 antibody that binds across two monomers (Figure 10) (confirming results of Ellis et al. Nat Commun 13, 1825 (2022)). By contrast the Nl / 09- TB did bind CD6 and protected mice against matched viral challenge, as shown here. This result suggested that, while the VASP-linked 2009 N1 does form tetramers as defined by size-exclusion and cross-linking, the protein may not be in an optimal state for vaccination. We therefore have used the tetrabrachion domain to form tetramers throughout this report.

[0301] The basis for stable tetramer formation

[0302] The hybrid proteins we produced all appeared to form stable tetramers as defined by SEC, BS3 crosslinking and melting temperature. They were all active in ELLA (large substrate) and MUNANA (small substrate) enzyme activity assays, and were inhibited by standard small molecule inhibitors. We tested a wide range of virus specific and cross- reactive human mAbs and with few exceptions the antibodies bound to the hybrid proteins. In addition, the CD6 antibody, that binds across two monomers and only to fully formed tetramers, bound to our hybrid proteins held together by the TB tetramerization domain. Finally, the crystal structures of our hybrid N1 proteins confirmed that the grafted loops had retained their expected conformations. Together these data suggested that the hybrid proteins had folded correctly and combined the expression and stability characteristics of the “scaffold” donor with the antigenic properties of the loop donor.

[0303] The distribution of epitopes on Neuraminidase

[0304] Varghese, Laver, and Colman described the structure of L01 and L23 loops on the top surface of N2 NA and showed that these loops contribute largely to the active site, notably including seven of eight conserved residues in the catalytic site and eight of eleven conserved residues that support the site (Varghese et al. Nature 303, 35-40 (1983)). Also, in the L01 and L23 loops surrounding the active site they noted multiple variable residues that vary seasonally associated with antigenic drift, and that can be selected for viral resistance by monoclonal antibodies in vitro. Resistance mutations selected for by three N2 murine mAbs and N9 mAbs were within loops B3L01, B5L01 and B5L23. Over the last forty years many similar studies have been done with murine and human monoclonal antibodies to various NAs. We have collected a set that identifies thirty-one sites of amino acid substitution selected by monoclonal antibodies in independent experiments. Twenty- five of these are located in Loops 01 and 23, with three additional sites in immediate neighbour positions to these loops (90% altogether). Three additional sites of selection can be assigned to the underside of the NA head at the interface of stalk and head (position 88), B4 Loop 12 (position 285), B4 Loop34 (position 309). In addition, a recent study of human sera identified position 386 on B5 loop 34 as responsible for an antigenic change between Nl / 1977 and Nl / 1986 (Daulagala, P. et al. mBio 14, e00084-23 (2023)). In addition a recent study of natural seasonal evolution of N2 neuraminidase identified the ten most significant amino acid positions contributing to antigenic variation (Catani et al. eLife 2023;12:RP90782. DOI: https: / / doi.org / 10.7554 / eLife.90782). Nine of these changes are in loops L01 and L23, confirming the original observations of Colman et al (Nature, 303, 41- 43, 1983).

[0305] Crystal structures of bound Fab fragments give further information on the footprints of protective antibodies. The majority describe binding to the L01 and L23 loops, with recent examples also confirming binding to the side (CD6, NA-22) and underside of NA. The majority of antibodies defined by crystallography bind within the surface of a single monomer, with recent structural evidence that a few mAbs binds across two monomers. Finally, several structures of antibodies that bind within the active sites of a broad range of NAs have been described, all of which contact the L01 and L23 loops. From these data we suggest that most antibodies generated by NA that are likely to be protective bind to the L01 and L23 loops, while a minority bind to epitopes on the underneath and side of the NA head that are not included within the L01 and L23 loops.

[0306] Definition of Loops

[0307] To align the top surface loops in N1 proteins, we used loop annotations from the N2 NA structure resolved to 2.9 A by Varghese et al. as a reference.

[0308] Some residues at the margins of loops are contact residues for some mAbs (For example, 1G01 : residue 134 preceding Bl L23, 1G01 and Z2B3: residues 179-180 following B2L01, and NC41 and NClO: residue 363 preceding B5L23; Nl / 2021 numbering), and in a few cases were selected as escape variants by antibodies or antisera. An element of judgement is therefore reasonable when selecting surface residues to graft. In addition the C-terminal domain (CTD) contributes to surface residues and may be suitable for grafting, since residues within this section of NA have been selected by mAbs in vitro and evolve over time. However, the CTD of N1 was conserved in all of our examples.

[0309] Vaccination with hybrid NAs

[0310] Our NA vaccination strategy is based on our earlier evidence that linking tetrameric NAs to the mi3 vaccine-like particle (valency of sixty) via a SpyTag / SpyCatcher covalent linkage results in enhanced immunogenicity and dose sparing, with doses as low as 0.1 pg of NA protein able to induce NAI antibody (Rahikainen, R. et al. Overcoming Symmetry Mismatch in Vaccine Nanoassembly through Spontaneous Amidation. Angewandte Chemie - International Edition 60, 321-330 (2021)). This compares well to the higher doses of pure protein used in the majority of studies of NA immunity (Zhang, X. & Ross, T. M. Anti -neuraminidase immunity in the combat against influenza. Expert Review of Vaccines 23, 474-484 (2024)). In the present experiments we opted for two doses of 0.5 pg of mi3 linked NA that gave full protection in preliminary experiments-(not shown).

[0311] In our first set of grafting experiments between Nl / 2009 and mSNl (H5N1 2021) we found that while protein expression was greatly improved, antisera from animals were cross reactive for NA inhibition between the scaffold H5N1 donor, the hybrid (with 12 amino acid replacements in the L01 and L23 loops) and the 2009 Loop donor. In addition, all immunised animals were protected from challenge with 2009 H1N1 virus. These results were consistent with evidence that within the N1 subtype there is broad cross-reactivity and that immunisation with 2009 N1 provides at least partial protection against an avian H5N1. Cross-reactivity could have been for epitopes anywhere within the NA structure.

[0312] The sequence difference in the loops between 2019 N1 donor and the H5N1 recipient was greater (16 amino acids), sufficient to provide an antigenic distance that prevented cross inhibition by antisera from animals immunised with these two NAs. The NA inhibitory activity of the antisera was predominantly specific for the L01 and L23 loops. Nl / 19 had four significant additional amino acid substitutions compared to Nl / 09 in the transferred L01 and L23 loops: N222K (B3L01), N244D (B3L23), N270K (B4L01) and K432E (B6L23). All four of these residues have been selected for resistance by inhibitory mAbs in vitro, or form contacts with mAbs demonstrated in crystal structures, so these residues are likely to have been responsible for the loss of cross reactivity between Nl / 2019 and mSNl.

[0313] The lack of cross inhibition between sera raised to the avian mSNl and the seasonal Nl / 19 showed that cross reactivity of antisera to the N1 NAs is not absolute.

[0314] We found that the NA of mouse-virulent H1N1 virus A / PR / 8 / 1934 (Cambridge strain) differed by 18 amino acids in the L01 and L23 loops from the mS H5N1 scaffold donor. We therefore prepared two further hybrids in which the loops were exchanged between these two NAs. The relevant hybrids induced NA inhibitory sera that were largely loop specific, and protection against a high dose challenge with A / PR / 8 / 34 was now dependent on matched L01 and L23 loops. In this case antibodies to the scaffold could not have been protective, and protection correlated with the NA inhibition activity induced.

[0315] Methods

[0316] Cell lines and viruses

[0317] ExpiCHO cells were used for the production of NA proteins and monoclonal antibodies. They were handled according to the manufacturer's protocol. MDCK- SIAT1 (Madin-Darby Canine Kidney cells stably transfected with human a 2,6- sialyltransferase, SIAT1) cells were used for the production of viruses, and for virus infection for epitope specificity assays. Cells were maintained in DIO medium [Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10% (v / v) foetal calf serum (Sigma- Aldrich, F9665), 2 mM glutamine, 100 U / mL penicillin and 100 pg / mL streptomycin], with incubation in a humidified 5% CO2 37 °C incubator. DMEM supplemented with 2 mM glutamine, 10 mM HEPES, 0.1% BSA, 100 U / mL penicillin and 100 pg / mL streptomycin was used for virus growth and virus dilution. The medium is referred to as virus growth medium (VGM). All cell lines were tested to be mycoplasma- free.

[0318] Viruses were obtained from the Worldwide Influenza Centre, Crick Institute, London UK. Viruses were propagated by infecting a monolayer of MDCK-SIAT1 with ~0.1 MOI (multiplicity of infection) virus for 1 h before replacing with VGM containing 0.5-1 pg / mL TPCK -treated trypsin. Virus supernatant was harvested after 48 h. Monoclonal antibodies

[0319] Antibodies AG7C, AF9C, Z2C2, and Z2B3 were previously published. mAb 1G01 was produced in the lab using the antibody sequence obtained from PDB 6Q23. Genes were synthesized by GeneArt (Life Technologies), cloned into antibody expressing vectors, and expressed in ExpiCHO cells. Similarly, mAb CD6 was synthesized using the sequence obtained from PDB 4QNP. The mAb is chimeric since the mouse variable region is cloned with the human constant regions.

[0320] NmAbs and 24- 1C are anti -NA humans mAbs provided by our collaborator Prof. Kuan Ying Huang (National University Taiwan), which we produced in ExpiCHO expression system using their expression plasmids (see Fig. 11). The study protocol and informed consent of human monoclonal antibody isolation were approved by the ethics committee at the National Taiwan University Hospital and Chang Gung Memorial Hospital and were carried out in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Written informed consent was received from each adult prior to inclusion in the study.

[0321] Neuraminidase protein expression and purification

[0322] Gene constructs (except for the Nl / 09 protein (SEQ ID NO: 10)) were designed to have H7 HA signal sequence (MNTQILVFALIAIIPTNADKI), Strep -tag II (SAWSHPQFEK), G3S linker, 6xHis, S3GSG linker, SpyTag (AHIVMVDAYKPTK), G2SG4S linker, and tetrabrachion tetramerisation domain from Staphylothermus marinus (Streltsov, V. A., Schmidta, P. M. & Breschkina, J. L. M. K. Structure of an influenza a virus n9 neuraminidase with a tetrabrachion-domain stalk. Acta Crystallographica Section F: Structural Biology Communications 75, 89-97 (2019)), GGSGTG linker (SEQ ID NO: 8), and finally an NA ectodomain at the C-terminus (SEQ ID NOs: 7, 9, 11-16). Sequences were synthesized as human-codon-optimized cDNAs by GeneArt (Life Technologies) and cloned into pcDNA3.1 / - plasmid for transfection. Proteins were expressed in ExpiCHO expression system (Thermo Fisher). Briefly, ExpiCHO cells were cultured in a humidified Multitron Cell incubator (Infers HT) at 37 °C with 8% (v / v) CO2, rotating at 125 rpm, for at least 2 passages before transient transfection. The Max -titre protocol from the manufacturer was used and proteins were harvested on day 8 / 9 post-transfection. Culture supernatants were clarified by centrifugation at 3000 g for 10 min at room temperature, followed by filtration through a 0.22 pm filter (RapidFlow Nalgene). Proteins were purified from cell culture supernatants using Ni-NTA-sepharose (prepacked HisTrap HP column from Cytiva) using the AKTA Pure purification system. The binding buffer was 20 mM Sodium Phosphate (NfeHPCU), 0.5 M NaCl, 20 mM Imidazole, pH 7.4, and the elution buffer was 20 mM Sodium Phosphate (ISfeHPCU), 0.5 M NaCl, 500 mM Imidazole, pH 7.4. The eluate was then buffer exchanged to Dulbecco's phosphate buffered saline (DPBS) with Calcium and Magnesium (Gibco™ 14040133) using 7K MWCO (molecular weight cut off) Zebaspin desalting columns (Thermo Fisher). Protein aliquots were stored in -80 °C for long-term storage and at 4 °C for short-term use and storage.

[0323] Size-exclusion chromatography

[0324] The putative oligomeric state of NA proteins was assessed via size-exclusion chromatography (SEC). Proteins were run on Superdex 200 Increase 10 / 300 GL column (Cytiva) equilibrated with PBS at a flow rate of 0.5 mL / min. Graphs were plotted using GraphPad Prism software.

[0325] NA enzymatic activity assays

[0326] The enzymatic activity of NA proteins was measured using ELLA (Enzyme-linked lectin assay; large substrate fetuin Mw: 49 kDa) and MUNANA (small fluorescent substrate 20-(4-methylumbelliferyl)-a-D-N-acetylneuraminic acid Mw:489 Da) assays.

[0327] ELLA: ELLA assays were performed. In brief, serially-diluted heat-inactivated sera were incubated with pre-titrated virus (NA source) or recombinant NA protein for 1 h. The dilution medium was DMEM (Gibco) supplemented with 2 mM glutamine, lOmM HEPES, 0.1% (w / v) BSA, 100 U / mL penicillin and 100 pg / mL streptomycin. The virus or NA concentration was pre-determined by titrating to measure the lowest concentration at top plateau, ensuring at least a 5-fold signal-to-noise ratio. The mixture was transferred to a Nunc Immunoassay ELISA plate (Thermo Fisher, 439454) pre-coated overnight with 25 pg / mL fetuin (Sigma-Aldrich, F3385). The plate was incubated for 18-20 h at 37 °C with 5% (v / v) CO2 in a tissue culture incubator. The NA enzymatic activity was detected by adding HRP-conjugated peanut agglutinin (Sigma-Aldrich, L7759) at 1 pg / mL after washing the plates 3 times with PBS and then developing with 50 pL TMB substrate (SeraCare). The enzymatic reaction was stopped after 5-10 min using 50 pL 1 M H2SO4 and the absorbance was read at 450 nm using a CLARIOstar plate reader (BMG Labtech).

[0328] MUNANA: The protocol from the WHO Worldwide Influenza Centre, Crick Institute was adapted. The sialidase activity of NA proteins or viruses were measured by incubation of serially-diluted NA (50 uL) in 32.5 mM 2-(N-morpholino)ethanesulfonic acid (MES) buffer pH 6.5 containing 4 mM CaCh with 50 uL of 100 pM MUNANA substrate. The reaction mix was incubated for 1 h at 37 °C and stopped by adding 50 uL stop solution (0.1 M glycine, 25% ethanol, pH 10.7). Changes in fluorescence was measured at 460 nm (excitation 355 nm) using a CLARIOstar. The NA concentration giving fluorescence intensity of 40-60K units (~ 100-fold signal from background) was used in NA inhibition assays. NA inhibition titres of anti-sera or mAbs were determined by incubation of serially diluted sera / mAbs with NA / viruses for 1 h before the addition of MUNANA substrate.

[0329] % NA activity was calculated as {(X — Min) / (Max — Min)} * 100 where X = measurement value, Min = buffer only, Max = NA or virus alone. The 50% inhibiting titre / concentration (IC50) was determined using non-linear regression curve fit on GraphPad Prism. Kruskal-Wallis test was used for statistical analysis. All graphs were plotted using GraphPad Prism software. nanoDSF thermal unfolding

[0330] The thermal stability and unfolding of NA proteins were measured using the nanoDSF (differential scanning fluorimetry) technique on a Prometheus Panta instrument (Nanotemper). NanoDSF monitors changes in the intrinsic tryptophan or tyrosine fluorescence resulting from alterations in the 3D structure of proteins as a function of temperature. Capillaries were filled with 10 pL of NA protein at 0.4 mg / mL, placed into the sample holder and the temperature was increased from 20 to 90°C at a ramp rate of l°C / min, with one fluorescence measurement taken per 0.1°C. The fluorescence intensity ratio (Emsso nm / Em33o mn) and its first derivative were calculated using the manufacturer’s software. Two measurements were taken for each experiment, and the means are shown. The experiment was repeated to ensure reproducibility. Epitope specificity analysis

[0331] A collection of monoclonal antibodies developed in-house and published in the literature was used to determine their binding specificity to NA proteins. Antibodies were serially diluted 5-fold from 10,000 to 1 ng / mL. mAbs were tested for binding to recombinant protein or NA expressed on the surface of MDCK-SIAT1 cells infected overnight with the virus. Antibody binding was detected using a secondary goat-anti human HRP antibody. The area under the curve was measured for each mAb’s titration curve and normalised as a percentage of relative binding to one of the highest binding mAb for comparison. The binding experiments were repeated, and the data from a single experiment are included. Data analysis and graphs generation was performed using GraphPad Prism. mi3 VLP conjugations

[0332] SpyCatcher003-mi3 virus-like particles (Lot No. 108-23-001-007) produced in Bacillus sublilis. were kindly provided by Ingenza Ltd. NA and VLP conjugations were done in DPBS with Calcium and Magnesium (Gibco™ 14040133) at various molar ratios. The conjugation was analysed using reduced SDS-PAGE. The conjugates were confirmed to have enzymatic activities and binding of mAbs binding equivalent to the unconjugated proteins.

[0333] Mice immunisation and virus challenge

[0334] Animal experiments were conducted in compliance with the UK Animals (Scientific Procedures) Act Project License (PP9362617), following the principles of the 3Rs (Replacement, Reduction and Refinement). Female BALB / c OlaHsd or DBA / 2 OlaHsd mice (7-8 weeks old at the start of the experiment) were obtained from Envigo and housed in individually vented cages in a specialised unit for infectious diseases. The mice were housed in accordance with the UK Home Office ethical and welfare guidelines, fed standard chow, and had access to water ad libitum. Mice were anesthetized with isofhirane (Abbott) and immunised via intramuscular injections with two doses of 0.5 pg of NA- VLP immunogens adjuvanted with AddaVax™ 1 : 1 vol / vol, administered 3 weeks apart. Serum samples were collected 3 weeks after the booster dose: intermediate samples were obtained via tail vein and terminal samples via cardiac puncture of euthanised mice. Whole blood collected in microtainer SST tubes (BD) was allowed to clot at room temperature for 1-2 h before centrifugation at 10,000 g for 10 min. The clarified sera were transferred to fresh tubes and stored at -20 °C.

[0335] For virus challenge experiments, 50 uL virus (passaged and titrated in the laboratory) was administered intranasally to anesthetised mice. Virus dose was 104TCID50 (200 LD50) of X-179A (H1N1 A / Califomia / 07 / 2009) in DBA / 2 mice and 104TCID50 (1000 LD50) of H1N1 A / PR / 8 / 1934 in BALB / c mice. The weight and clinical signs of mice were monitored regularly over a study period of 2-3 weeks until they either reached their endpoint (<80% of their initial weight) or recovered to their initial weight. Mice reaching the endpoint were humanely euthanised. Kaplan-Meier survival analysis with the Logrank Mantel-Cox test was used for comparison.

[0336] Sequence Alignment

[0337] Neuraminidase sequences were obtained from PDB files where referred, or from Global Initiative on Sharing All Influenza Data (GISAID). Amino acid sequences of the neuraminidase were aligned using Muscle alignment in Geneious Prime, and all alignment figures were generated using Geneious Prime.

[0338] Crystallization, data collection and structure determination

[0339] In each well of 96-well sitting drop plates (Greiner), 100 nL of purified neuraminidase was combined with 100 nL of precipitant. The concentrations used for crystallisation were 4.6 mg / mL for mSNl, 5 mg / mL for Nl / 09 hybrid and 4.5 mg / mL for Nl / 19 hybrid. For crystal formation, the mixtures were equilibrated against 90 pL of precipitant at 20.5 °C.

[0340] Crystals of msNl were obtained using the Morpheus B2 precipitant [0.09M halogens (NaFl, NaBr, Nal), 0.1 M imidazole / MES pH 6.5, 30% ethylene glycol / PEG 8K; Hampton Research], The Nl / 09 hybrid crystallized using precipitant H4 of the Pentaerytritol screen (25% pentaerytritol ethoxylate 797, 0.1 M MES, 0.05 M MgC12, pH 6.5; Jena Bioscience) and the Nl / 19 hybrid using PGA-screen condition G2 (5% PGA-LM, 30% PEG550 MME, 0.1 M Tris pH 7.8; Molecular Dimensions). Glycerol was added to 25% (v / v) for cryoprotection of the Nl / 19 hybrid crystals. Diamond beamline 124 (Harwell, UK) was used for diffraction data collection at 100K. Data processing used the Xia2 programme suite and, for the Nl / 19 hybrid, autoPROC and STARANISO. The molecular replacement program PHASER was used for structure elucidation. Model building was done using COOT and model refinement using Phenix. Data-collection and refinement statistics are provided in Supplementary Table 1. Figures were prepared using UCSF ChimeraX.

[0341] EXAMPLE 2

[0342] N8 neuraminidase as a scaffold for loop grafting to produce high-yield selfassembling tetramers

[0343] Introduction

[0344] Human infections with avian influenza A(H3N8) and A(H5N8) have been reported following direct transmission from poultry in China (2022) and Russia (2020), respectively. The N8 neuraminidase carried by these viruses is novel to humans, leaving the population immunologically naive to this antigen. Although seasonal H3N2 viruses are endemic in humans, sera from individuals exposed to or vaccinated against H3N2 failed to cross-neutralize avian H3N8 viruses. Moreover, a human-isolated H3N8 virus was transmissible between ferrets via respiratory droplets.

[0345] Results

[0346] An N8-based vaccine was pursued and expressed the N8 neuraminidase from A / chicken / England / 030720 / 2020 | EPI_ISL_626652 (H5N8) in ExpiCHO cells. Unlike other NAs, this N8 NA (hereafter referred to as “chN8”) self-tetramerized when expressed without a tetramerization domain. The protein yielded 43 mg / L in the ExpiCHO system, exhibited high activity in both MUNANA and ELLA assays, and was recognized by the catalytic-site-targeting, broadly reactive mAbs 1G01 and Z2B3.

[0347] To determine whether self-tetramerization is a common feature of N8 neuraminidases, we expressed five additional N8 NAs spanning diverse hosts, clades, and years (Figure 13):

[0348] 1. A / duck / Ukraine / 1 / 1963 H3N8 (Sequence identical to PDB 2HT5)

[0349] 2. A / harbor seal / Massachusetts / 1 / 2011 H3N8 (Sequence differs from PDB 4WA3 by 3 residues) 3. A / gyrafalcon / Washington / 41088-6 / 2014 H5N8 (Sequence identical to PDB 5HUN)

[0350] 4. A / Equine / Maryland / 1 / 22 H3N8

[0351] 5. A / Henan / 4- 14 / 2022 H3N8

[0352] The 5 NAs were selected from a phylogenetic analysis of N8 neuraminidase sequences. Among 7,120 N8 sequences deposited in GISAID (Global Initiative on Sharing All Influenza Data) as of 3 January 2022, ~70 representative sequences were selected to cover a broad range of isolation years, host species, and geographic origins. Multiple sequence alignment was performed using MUSCLE, and a neighbor-joining phylogenetic tree was generated from the alignment. Five sequences representing a broad selection were selected for expression in the lab.

[0353] None of the five N8 NAs expressed at higher concentration, nor they showed NA activity like chN8. For example, the N8Henan TDminus (lacking a transmembrane domain) protein (Figure 14) showed poor quality, as indicated by size-exclusion chromatography (SEC) and protein melting curves, both of which displayed broad, diffuse peaks rather than sharp, well-defined ones.

[0354] N8 hybrids using chN8 as a scaffold chN8 TDminus protein showed good activity on ELLA and MUN ANA (Figure 14). However, it showed unusual biophysical characteristics. SEC revealed a broad retention peak spanning 10-15 minutes (corresponding to -200-50 kDa), suggesting the presence of heterogeneous species ranging from tetramers to monomers. Consistent with this, protein melting analyses showed two distinct Tmvalues (43.6 °C and 72.0 °C), further indicating structural heterogeneity.

[0355] The N8Henan neuraminidase (A / Henan / 4- 14 / 2022 H3N8) containing a tetrabrachion (TB) tetramerization domain (N8Henan TB) was expressed poorly (<1 mg / L; Figure 14) and was therefore selected as the donor for loop grafting onto the chN8 scaffold. For the L01 and L23 loop annotations, slight modifications in the loop regions were made compared to previously published loop-grafting method based on mSNl (Example 1).

[0356] This new loop annotation is referred to as v2.1+. Loop regions were shortened or extended depending on whether the residues were predicted to be exposed or buried, as determined by NetSurfP2.0. Additional considerations were based on whether residues at loop ends had been reported as part of mAb or antigenic epitopes. Twelve L01 and L23 top loops were grafted from N8Henan and additionally included the B5L34 loop, which was predicted as a B-cell epitope by BepiPred (cutoff set to higher confidence, top 20%) (Figure 17). The resulting loop-grafted construct, N8Henan hybrid-1, was expressed as TDminus and yielded 533 mg / L. Despite the high expression level, the protein precipitated at concentrations above ~10 mg / mL in PBS (pH 7.4), an issue resolved by diluting to <5 mg / mL and buffer exchanging into TBS (pH 8.0). Hybrid-1 displayed a clean ~50 kDa band on reduced SDS-PAGE, a melting temperature of 67.5 °C, and an SEC profile consistent with a mixture of tetrameric and lower-order oligomeric species (Figure 14). Although hybrid- 1 was used in immunization studies, it did not yield protein crystals.

[0357] To improve crystallization, the variation within the B5L34 loop sequence (VQN) was reverted to TQT, matching the parental chN8 scaffold. Construct design was also refined in N8Henan hybrid-2 by extending the coding region (residues 80-470 versus 83- 470 in hybrid-1) and simplifying the expression cassette to retain only a 6His tag, removing Strep-tagll, SpyTag, and additional linkers. These optimizations resulted in markedly improved expression (1450 mg / L), increased thermostability (Tm= 68.3 °C), and a homogeneous tetrameric SEC profile. Importantly, hybrid-2 crystallized successfully, yielding crystals that diffracted to 1.5 A, although the precise construct feature enabling crystallization remains unclear.

[0358] N8 NA proteins are inhibited by NA inhibiting mAbs and drugs

[0359] To confirm that produced N8 proteins were in proper structural conformation, inhibition by NA-inhibiting mAbs and known commercial drugs was done using ELLA. mAbs FNI9 and 1G01, the subtype cross-reactive mAbs targeting NA catalytic site were used. Among drugs, Oseltamivir, Zanamivir and Peramivir were used. N8Henan TB and N8Henan hybrids-1 and 2 were inhibited in a similar patter by both mAbs and drugs. N8Henan hybrid-2 was inhibited with lower IC50 compared to N8Henan hybrid- 1 (Figure 15).

[0360] N8Henan and the hybrid NA produced similar in vivo antibody response chN8 TDminus, N8 Henan TB, and the N8 Henan hybrid were each conjugated to mi3 virus-like particles (VLPs) and used to immunize groups of mice, as previously described in Example 1. Immunized sera (n = 5 per group) were assayed for inhibition of N8 NA proteins (Figure 16). All groups elicited NA-inhibiting antibody responses against each other. chN8 TDminus and the N8Henan TB showed cross-reactive responses to each other. The N8Henan hybrid- 1 elicited a statistically higher response against both N8Henan TB and itself compared to chN8 TDminus.

[0361] Importantly, antisera from N8Henan TB and hybrid-1 groups produced similar inhibition profiles (statistically indistinguishable, p=>0.99 for inhibition of N8Henan TB and p=0.69 for inhibition of N8Henan hybrid-1) against all NA proteins tested, indicating a high degree of antigenic similarity between these two constructs.

[0362] Discussion

[0363] The selected neuraminidase (chN8) naturally may form a self-assembling tetramer without requiring an external tetramerization domain, a property not universal among N8 subtypes. chN8 serves as a scaffold donor, improving the yield and stability of other low- expressing N8 variants.

[0364] The B5L34 loop can be successfully grafted while preserving both protein yield and structural conformation. This represents a second demonstration of loop grafting functionality in a subtype beyond N1 neuraminidase.

[0365] EXAMPLE 3

[0366] Feasibility of Loop Grafting in N2 Neuraminidase

[0367] Loop grafting was performed in a further Group II NA candidate. Two distant N2 neuraminidase proteins were selected: N2 from 1968 (X-31 strain) and N2 from 2021 (H3N2 seasonal vaccine candidate for 2022 / 23, strain A / Darwin / 6 / 2021). Both proteins expressed at high levels, 142 mg / L and 89 mg / L, respectively. These two proteins differ by 62 residues in total (16%), with 37 of those differences located within the top loops. The loop grafting was performed mainly to demonstrate that the technique can be applied to other neuraminidases, not only for the purpose of increasing yield as in N1.

[0368] Two hybrids were produced: the X-31 hybrid (X-31 loops on an N2 / 21 scaffold) and the N2 / 21 hybrid (N2 / 21 loops on an X-31 scaffold), by transferring 37 residues from loops L01 and L23 of the loop donor onto the scaffold (Figure 18). N2 NA contains five putative N-glycosylation sites: two in the scaffold region and three in the loops, specifically Bl L23, B4L23, and B5L01.

[0369] The X-31 hybrid showed low yield and the protein formed aggregates, appearing as a smear on SDS-PAGE. In contrast, the N2 / 21 hybrid expressed at a two-fold higher level compared to the loop donor NA (Figure 19). This demonstrates that loop grafting can be successful in one direction but may not always work for every loop-scaffold combination.

[0370] EXAMPLE 4

[0371] Comparison of different loop annotations

[0372] In Example 1, loop annotations were made using the msNl scaffold, and using Varghese, 1983, as a basis. These annotations are otherwise referred to as ‘vl’ herein.

[0373] In Example 2, loop annotations were made based on a published N8 structure. The B5L34 loop was also annotated due to its putative immunogenicity. These annotations are otherwise referred to as ‘v2.1’ or ‘v2.1+’ herein.

[0374] In Example 3, loop annotations were made based on a combination of annotations from Varghese 1983, 1991, aligned with N1 and N2 structures to establish a universal annotation system also applicable to N2. This was designed to provide broader coverage for selecting epitopes at loop edges. Furthermore, the CTD annotation shortened, as it forms part of both the interface residues and antigenic epitopes.

[0375] The annotations are displayed in Figure 20.

[0376] Loop annotations should be regarded as reference points and may be adjusted according to the most recently published structure for a given NA subtype. (See discussion in Rijal et al., eLife.)

[0377] In v3, the CTD region was shortened since it overlaps with both interface residues and antigenic epitopes.

[0378] Within a subtype, greater sequence conservation is observed at loop edges, while most differences are concentrated in the loop cores.

[0379] Similarly, for the CTD, sequence variation is primarily located at the distal end.

[0380] Transfer of B5L34 was tolerated in the N8 hybrid, suggesting that, based on B cell epitope predictions and experimental observations, this loop can be considered for grafting. Importantly, B5L34 has also been identified as an epitope targeted by underside mAbs and antigenic drift.

Claims

Claims1. A polypeptide comprising the head region of a first neuraminidase, wherein the L01 and L23 loops are substituted with the L01 and L23 loops of a second neuraminidase.

2. The polypeptide of claim 1, wherein:(a) the head region of the polypeptide that has a total yield that is at least 2 times, at least 3 times, at least 4 times, at least 5 times or at least 10 times greater than the yield of the head region of the second neuraminidase, wherein the yield is calculated based on a construct comprising SEQ ID NO: 8 and the head region and encoded in a pcDNA3.1 vector, expressed in ExpiCHO cells for 8 days at 37°C, 8% CO2;(b) the first neuraminidase has a Tmof at least 60 °C wherein the Tmis calculated based on a construct comprising SEQ ID NO: 8 and the head region of the first neuraminidase in DPBS buffer with calcium, wherein the Tm is measured by nano differential scanning fluorimetry;(c) the head region of the polypeptide that has a total yield that is at least 2 times, at least 3 times, at least 4 times, at least 5 times or at least 10 times greater than the yield of the head region of the second neuraminidase, wherein the yield is calculated based on a construct comprising SEQ ID NO: 48 and the head region and encoded in a pcDNA3.1 vector, expressed in ExpiCHO cells for 8 days at 37°C, 8% CO2; and / or(d) the first neuraminidase has a Tmof at least 60 °C wherein the Tmis calculated based on a construct comprising SEQ ID NO: 48 and the head region of the first neuraminidase in DPBS buffer with calcium, wherein the Tm is measured by nano differential scanning fluorimetry.

3. The polypeptide of claim 1 or 2, wherein the first neuraminidase is derived from an influenza virus.

4. The polypeptide of any one of the preceding claims, wherein the first neuraminidase is derived from a group 1 neuraminidase.

5. The polypeptide of any one of the preceding claims, wherein the first neuraminidase is derived from an N1 neuraminidase.

736. The polypeptide of claim 5, wherein the first neuraminidase is derived from the neuraminidase of H5N1 A / mute swan / England / 053054 / 2021, or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of H5N1 A / mute swan / Engl and / 053054 / 2021.

7. The polypeptide of any one of the preceding claims, wherein the head region comprises the amino acid sequence corresponding to SEQ ID NO: 7.

8. The polypeptide of any one of claims 1 to 4, wherein the first neuraminidase is derived from an N8 neuraminidase.

9. The polypeptide of claim 8, wherein the first neuraminidase is derived from the neuraminidase of chN8 (H5N8|A / chicken / England / 030720 / 2020|EPI_ISL_626652), or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of chN8 (H5N8|A / chicken / England / 030720 / 2020|EPI_ISL_626652).

10. The polypeptide of any one of the preceding claims, wherein the head region comprises the amino acid sequence corresponding to SEQ ID NO: 49 or 50.

11. The polypeptide of any one of claims 1 to 3, wherein the first neuraminidase is derived from a group 2 neuraminidase.

12. The polypeptide of claim 11, wherein the first neuraminidase is derived from an N2 neuraminidase.

13. The polypeptide of claim 12, wherein the first neuraminidase is derived from the neuraminidase of X-31 NA (H3N2 / 1968), or a variant comprising at least 80% identity to the amino acid sequence of the neuraminidase of X-31 NA (H3N2 / 1968).

14. The polypeptide of any one of the preceding claims, wherein the head region comprises the amino acid sequence corresponding to SEQ ID NO: 86.

15. The polypeptide of any one of the preceding claims, wherein the first neuraminidase and the second neuraminidase are of the same neuraminidase subgroup.

16. The polypeptide of any one of the preceding claims, wherein the second neuraminidase is:(a) an N1 neuraminidase, an N8 neuraminidase or an N2 neuraminidase; and / or(b) the neuraminidase of a circulating influenza strain or an influenza strain that is predicted to circulate, or has a reservoir in humans, birds or other animals.

17. The polypeptide of any one of the preceding claims, wherein:(a) the L01 and L23 loops are the amino acid sequences corresponding to residues 26-38, 54-75, 95-97, 115-118, 138-147, 163-170, 188-197, 212-215, 233-266, 283-288,314-318 and 349-356 of SEQ ID NO: 7, optionally wherein the first and / or second neuraminidase is a group 1 neuraminidase, further optionally wherein the first and / or second neuraminidase is a Nl neuraminidase;(b) the L01 and L23 loops are the amino acid sequences corresponding to residues 23-35, 58-75, 95-99, 115-120, 136-147, 163-171, 187-193, 212-219, 236-263, 280-288, 314-320 and 347-356 of SEQ ID NO: 7, optionally wherein the first and / or second neuraminidase is a group 1 neuraminidase, further optionally wherein the first and / or second neuraminidase is a N8 neuraminidase; or(c) the L01 and L23 loops are the amino acid sequences corresponding to residues 23-35, 56-75, 95-99, 115-120, 138-147, 163-171, 188-197, 212-219, 233-267, 281-289, 314-320 and 348-356 of SEQ ID NO: 7, optionally wherein the first and / or second neuraminidase is a group 2 neuraminidase, further optionally wherein the first and / or second neuraminidase is a N2 neuraminidase.

18. The polypeptide of any one of the preceding claims, wherein the B5L34 loop of the first neuraminidase is substituted with the B5L34 loop of the second neuraminidase, optionally wherein the B5L34 loop is the amino acid sequence corresponding to residues 299-307 of SEQ ID NO: 7, further optionally wherein the first and / or second neuraminidase is a group 1 neuraminidase, further optionally wherein the first and / or second neuraminidase is a N8 neuraminidase.

19. The polypeptide of any one of the preceding claims, wherein the C-terminal domain (CTD) of the head region of the first neuraminidase is substituted with the CTD of the second neuraminidase.

20. The polypeptide of claim 20, wherein the C-terminal domain is the sequence corresponding to:(a) residues 372-388 of SEQ ID NO: 7 optionally wherein the first and / or second neuraminidase is a group 1 neuraminidase, further optionally wherein the first and / or second neuraminidase is a Nl or N8 neuraminidase;(b) residues 378-388 of SEQ ID NO: 7, optionally wherein the first and / or second neuraminidase is a group 2 neuraminidase, further optionally wherein the first and / or second neuraminidase is a N2 neuraminidase.

21. The polypeptide of any one of the preceding claims, wherein the L12 and L34 loops are substituted with the L12 and L34 loops of a second neuraminidase.

22. The polypeptide of any one of the preceding claims, further comprising: a vasodilator-stimulated phosphoprotein (VASP) tetramerisation domain; a tetrabrachion tetramerisation domain; a neuraminidase stalk domain, transmembrane domain and cytoplasmic domain; a measles phosphoprotein tetramerisation domain; a Sendai virus phosphoprotein tetramerisation domain; or an Arabidopsis thaliana transcription factor tetramerisation domain.

23. A tetramer comprising four polypeptides that each comprise a neuraminidase head region, wherein at least one of the polypeptides is as defined in any one of the preceding claims.

24. The tetramer of claim 23, wherein all four polypeptides are as defined in any one of claim 1-22.

25. A polynucleotide encoding the polypeptide of any one of claims 1-22, or one or more polynucleotides encoding the tetramer of claim 23 or 24.

26. The polynucleotide of claim 25, which is an RNA molecule.

27. A particle comprising the RNA molecule of claim 26, and a pharmaceutically acceptable excipient or carrier.

28. A virus comprising the polypeptide of any one of claims 1-22, or the tetramer of claim 23 or 24.

29. A virus-like particle (VLP) comprising the polypeptide of any one of claims 1-22, or the tetramer of claim 23 or 24.

30. The virus of claim 28, or the VLP of claim 29, further comprising an influenza hemagglutinin, optionally wherein the influenza hemagglutinin is the influenza hemagglutinin of an influenza strain that is predicted to circulate, circulating or has a reservoir in humans, birds or other animals.

31. A pharmaceutical composition comprising the polypeptide of any one of claims 1- 22, the tetramer of claim 23 or 24, the polynucleotide of any one of claims 25 or 26, the particle of claim 27, the virus of 28 or 30, or the VLP of claim 29 or 30.

32. An immunogenic composition comprising the polypeptide of any one of claims 1- 22, the tetramer of claim 23 or 24, the polynucleotide of any one of claims 25 or 26, the particle of claim 27, the virus of 28 or 30, or the VLP of claim 29 or 30.

33. The composition of claim 31 or 32, further comprising an additional therapeutic agent, wherein the additional therapeutic agent is selected from a vaccine and an immunomodulator.

34. The composition of any one of claims 31-33 for use as a medicament.

35. The composition of any one of claims 31-34 for use in the prevention or treatment of influenza infection in a subject.

36. A method of producing a chimeric protein, the method comprising: aligning the amino acid sequence of the head region of a first neuraminidase and a second neuraminidase to identify the L01 and L23 loops; and substituting the L01 and L23 loops of the first neuraminidase with the L01 and L23 loops of the second neuraminidase.

37. The method of claim 36, wherein the first neuraminidase and / or the second neuraminidase are as defined in any one of claims 1-22.

38. The method of claim 36 or claim 37, wherein the chimeric protein is a polypeptide according to any one of claims 1-22.

Images