Engineered paramyxovirus soluble fusion (F) proteins and related vaccines

Abstract

The present invention provides engineered soluble F proteins of paramyxoviruses, such as respiratory syncytial virus (RSV), human metapneumovirus (hMPV), and human parainfluenza virus (hPIV). These engineered proteins are stabilized by specific modifications in the wild-type soluble F sequence, e.g., substitutions in the O23 chain and / or introduction of an engineered disulfide bond in the O hairpin in the F1 subunit. The present invention also provides nanoparticle vaccines containing engineered soluble F immunogens displayed on self-assembling nanoparticles. The present invention also provides methods of using such vaccine compositions in various therapeutic applications, e.g., to prevent or treat viral infections, such as RSV, MPV, and PIV infections.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This patent application claims the benefit of priority to U.S. Provisional Patent Application Nos. 63 / 480,261 (filed January 17, 2023; currently pending) and 63 / 488,985 (filed March 8, 2023; currently pending), the entire disclosures of which are incorporated herein by reference in their entirety for all purposes. [Background technology]

[0002] Respiratory syncytial virus (RSV), human metapneumovirus (hMPV), and parainfluenza viruses (PIV) are enveloped, nonsegmented, negative-sense, single-stranded RNA viruses belonging to the Paramyxoviridae family. Among these, RSV has been widely studied. The RSV genome encodes three envelope glycoproteins and eight nonstructural proteins (NS1, NS2, N, P, M, M2-1, M2-2, and L). The three envelope glycoproteins are the attachment (G) protein, the fusion (F) protein, and the small hydrophobic (SH) protein. F and G are important for RSV infectivity and pathogenesis and contain various antigenic determinants that can be recognized by host neutralizing antibodies (NAb). As highlighted in a 2015-2016 report and global surveys by the WHO and the Gates Foundation, RSV is a common cause of acute lower respiratory tract infections (ALRI) in newborns and infants, representing a significant health burden in developing countries. RSV causes acute respiratory infections that result in approximately 66,000-200,000 deaths and 3.5 million hospitalizations worldwide in children under the age of five.

[0003] A series of advances have been made in RSV vaccine development over the past decade. First, a structural understanding of the F protein was achieved in both the prefusion and postfusion states, as well as RSV neutralization by F-specific NAbs. The F protein mediates viral entry and is a primary target for vaccine development. Multiple antigenic sites (AS) on the F protein can be recognized by NAbs. Cocrystallization of NAb D25 with F yielded the first atomic structure of prefusion F, revealing a novel antigenic site (AS-φ) near the trimer apex composed of residues 62–69 of F2 and the solvent-exposed portion of F (α4-helix). This structure has enabled the design of prefusion-stabilizing mutations and structural analysis of other NAbs, such as AM1429 and φ-specific 5C4. Second, both epitope-based and F protein-based strategies are being explored in RSV vaccine development. Early studies demonstrated immunogen design using "epitope grafting" for the RSV motavizumab epitope, leading to proof-of-concept studies of RSV epitope vaccines. Various empirical designs with different mutation sets have been proposed to stabilize the prefusion F structure. Third, recent human vaccine trials have revealed the importance of prefusion F in NAb induction. Compared with failed postfusion F vaccines, a rationally designed prefusion F trimer (DS-Cav1) demonstrated a 10-fold higher serum NAb response. However, despite these advances, current RSV prefusion F designs have various issues. For example, DS-Cav1 exhibited poor expression profiles with many aggregates and other F species after D25 purification. Similarly, another leading vaccine candidate, SC-TM, had only a low trimer yield. Negative-stain EM analysis showed that DS-Cav1 appeared to be entirely monomeric, while SC-TM exhibited closed-structure prefusion trimers mixed with postfusion trimers. A further optimized DS-Cav1 design, termed sc9-10 DS-Cav1, contained an interprotomer disulfide bond to lock F in a fully closed trimeric conformation.However, such interprotomer disulfide bonds can disrupt F nanoparticle (NP) protein folding and assembly of F-displaying NPs, resulting in low yields and protein aggregation. Furthermore, sc9-10 DS-Cav1 contains many mutations generated by random mutagenesis that may or may not be essential for the structure and function of this construct as a vaccine antigen. Summary of the Invention [Problem to be solved by the invention]

[0004] There remains an urgent need in the art for better and more effective vaccines against paramyxoviruses, particularly RSV. The present invention addresses this and other unmet needs in the art. [Means for solving the problem]

[0005] In one aspect, the present invention provides engineered immunogenic proteins derived from or modified from the fusion (F) protein of a paramyxovirus (e.g., RSV). These include modified soluble F sequences with one or more modifications compared to the wild-type soluble F sequence of the paramyxovirus. Typically, the engineered soluble F proteins of the present invention include (1) a substitution of two or more negatively charged residues (D486-A490) around the β23 chain with polar or hydrophobic residues, (2) a deletion of the P27 peptide (E110-R136), and (3) an engineered intraprotomer disulfide bond within the F1 subunit or linking the F2 and F1 subunits. Unless otherwise specified, the amino acid numbering of the various sequence modifications described herein is based on the human RSV A2 strain F protein (UniProt ID P03420). Some of the engineered soluble F proteins are derived from RSV. In some engineered soluble RSV F proteins, the two or more negatively charged residues around the β23 chain are D486 and E487. In some of these embodiments, the substitutions around the β23 chain include D486N / E487Q or D486L / E487L. In some engineered soluble F proteins of the invention, the engineered disulfide bond is S155C / S290C, S62C / K196C, or E60C / K196C.

[0006] In addition to the above modifications (1)-(3), some engineered soluble RSV F proteins of the present invention may further contain (1) one or more furin cleavage sites or (2) a linker portion that replaces the unstructured C-terminus of F2 (Q98-R109) and a portion of the N-terminus of the fusion peptide (FP) (F137-V157). In some of these embodiments, the replaced C-terminus of F2 is residues N104-R109 ( 104 NNRARR 109; SEQ ID NO: 31). In some of these embodiments, the replaced portion of the N-terminus of the fusion peptide (FP) comprises F137-S146. Some engineered soluble F proteins of the invention further comprise a substitution of residue S215. In some of these embodiments, residue S215 is replaced with P. Some engineered soluble F proteins of the invention further comprise a substitution of residue E92. In some of these embodiments, residue E92 is replaced with D, Q, another short polar residue, or a hydrophobic residue. Some engineered soluble F proteins of the invention can further comprise a V185P substitution. Some engineered soluble F proteins of the invention can further comprise an S46G, K462Q, or both substitutions. Some engineered soluble F proteins of the invention can further comprise an engineered intraprotomer disulfide bond S180C / S186C or A177C / T189C in the β3 / β4 hairpin. Some engineered soluble F proteins of the present invention can further comprise an engineered interprotomer disulfide bond A149C / Y458C. In various embodiments, the engineered soluble RSV F proteins of the present invention can have the amino acid sequence set forth in any one of SEQ ID NOs: 17-23, a conservatively modified variant thereof, or a substantially identical sequence thereof. In some embodiments, the engineered soluble F proteins of the present invention can further comprise an N-terminal leader sequence. In some embodiments, the engineered soluble F proteins of the present invention can further comprise a C-terminal foldon motif.

[0007] In a related aspect, the present invention provides nanoparticle vaccines containing the engineered soluble F protein described herein displayed on the surface of self-assembling nanoparticles. In some of these embodiments, the self-assembling nanoparticles comprise a trimeric sequence, and the C-terminus of the engineered soluble F protein is fused to the N-terminus of a subunit sequence of the nanoparticle. In some embodiments, the self-assembling nanoparticles used in the nanoparticle vaccines of the present invention are I3-01 variants. In some of these embodiments, the subunit sequence of the I3-01 variant comprises SEQ ID NO:25 (I3-01v9b) or SEQ ID NO:26 (I3-01v9c).

[0008] In another aspect, the present invention provides polynucleotide sequences encoding the engineered soluble F protein or nanoparticle vaccine described herein. In another aspect, the present invention provides pharmaceutical compositions containing the nanoparticle vaccine or polynucleotide sequence described herein and a pharmaceutically acceptable carrier. In yet another aspect, the present invention provides methods for preventing or treating a paramyxovirus infection in a subject. These methods involve administering a therapeutically effective amount of a pharmaceutical composition described herein to the subject. Some of these treatment methods are directed to treating or preventing RSV infection.

[0009] In yet another aspect, the present invention provides different classes of engineered or redesigned immunogenic polypeptides derived or modified from the fusion protein (F) of a paramyxovirus. These redesigned soluble F immunogens also include modified soluble F sequences with one or more modifications compared to the wild-type soluble F sequence of a paramyxovirus. In some of these embodiments, the modification comprises an intraprotomer-engineered disulfide bond linking a pair of β-sheet-forming amino acids of the β3 / β4 hairpin or equivalent hairpin of the F1 subunit of the soluble F sequence. Unless otherwise specified, the numbering of the hairpins in the immunogens is based on respiratory syncytial virus (RSV).

[0010] Some of these different classes of engineered soluble F immunogens are derived from the wild-type F sequence of human RSV. In these embodiments, an engineered disulfide bond is introduced between the substituted residues S180C / S186C or A177C / T189C in the β3 / β4 hairpin. The amino acid numbering in these embodiments is based on human RSV strain A2, which has UniProt ID P03420. In some of these embodiments, the wild-type soluble F sequence used is set forth in SEQ ID NO: 1, or a conservatively modified variant or substantially identical sequence thereof. In some embodiments, modifications to the wild-type soluble F sequence also include mutations at the unstructured C-terminus of the F2 subunit. For example, a redesigned RSV soluble F immunogen can include a truncation of residues 104-109 (NNRARR) (SEQ ID NO: 31) and / or a P102A substitution at the unstructured F2 C-terminus. In some embodiments, modifications to the wild-type soluble F sequence also include (1) replacing the processed active peptide (P27) (residues E110 to R136 and the N-terminus (residues F137 to S146) of the fusion peptide with a (GS)n linker sequence, where n is any integer from 1 to 5, and / or (2) the amino acid substitutions I379V and M447V. In some exemplary embodiments, the redesigned RSV soluble F immunogen has an amino acid sequence set forth in any one of SEQ ID NOs: 36-43 or a conservatively modified variant thereof.

[0011] Some of these different classes of engineered soluble F immunogens are derived from the wild-type F sequence of human metapneumovirus (hMPV). In some of these embodiments, an engineered disulfide bond is introduced between substituted residues A147C / A159C in the β3 / β4 hairpin. The amino acid numbering in these embodiments is based on hMPV strain CAN97-83, which has UniProt ID Q6WB98. In some of these embodiments, the wild-type soluble F sequence used is set forth in SEQ ID NO:44 or SEQ ID NO:45, or a conservatively modified variant or substantially identical sequence thereof. In some embodiments, the modification to the wild-type soluble F sequence also includes a mutation at the unstructured C-terminus of the F2 subunit. In some of these embodiments, the mutation at the unstructured F2 C-terminus replaces DQLAREEQIENP (SEQ ID NO:60) and cleavage site RQSR (SEQ ID NO:49) at the unstructured F2 C-terminus with a (GS)n linker sequence, where n is any integer between 1 and 5. In some exemplified embodiments, the redesigned hMPV soluble F immunogen has the amino acid sequence set forth in SEQ ID NO: 46 or SEQ ID NO: 47, or a conservatively modified variant thereof.

[0012] Some of these different classes of engineered soluble F immunogens are derived from the soluble F sequence of parainfluenza virus (hPIV). In some of these embodiments, an engineered disulfide bond is introduced between substituted residues Q159C / A171C in the β1 / β2 hairpin. The amino acid numbering in these embodiments is based on the recombinant hPIV3 / hPIV1 virus having UniProt ID O55888. In some of these embodiments, the wild-type soluble F sequence used is set forth in SEQ ID NO:48, or a conservatively modified variant or substantially identical sequence thereof. In some embodiments, the modification to the wild-type soluble F sequence also includes a mutation at the C-terminus of the F2 subunit. In some of these embodiments, the mutation at the C-terminus of F2 replaces the C-terminal sequence NQESNENTDP (SEQ ID NO:50) and cleavage site RTER (SEQ ID NO:51) with a (GS)n linker sequence, where n is any integer between 1 and 6. In some exemplified embodiments, the redesigned hPIV soluble F immunogen has the amino acid sequence set forth in SEQ ID NO: 61 or SEQ ID NO: 62, or a conservatively modified variant thereof.

[0013] In another aspect, the present invention provides engineered or redesigned immunogenic polypeptides derived from the fusion protein (F) of human respiratory syncytial virus (hRSV). In some embodiments, the redesigned hRSV immunogen comprises a modified RSV soluble F sequence that is altered by at least one of the following mutations compared to the wild-type hRSV soluble F sequence: (1) deletion of the P27 peptide (residues E110-R136), (2) modification of the unstructured C-terminus of the F2 subunit (residues Q98-R109), and (3) truncation of the N-terminus of the fusion peptide (residues F137-V157). In these embodiments, the amino acid numbering is based on human RSV strain A2, which has UniProt ID P03420. In some of these embodiments, the wild-type RSV soluble F sequence used is set forth in SEQ ID NO: 1, or a conservatively modified variant or substantially identical sequence thereof. In some of these embodiments, the modification at the C-terminus of unstructured F2 is (1) a truncation of residues 104-109 (NNRARR) (SEQ ID NO: 31) and / or (2) a P102A substitution. In some embodiments, the truncation at the N-terminus of the fusion peptide is a deletion of residues F137-S146.

[0014] In some embodiments, modifications in the redesigned RSV soluble F immunogen compared to the wild-type soluble F sequence can further include (1) a (GS)n linker between the F2 subunit and the F1 subunit in the altered soluble RSV sequence, where n is any integer from 1 to 6, and / or (2) at least one substitution selected from the group consisting of I379V and M447V. In some of these embodiments, the linker comprises the sequence GSGS (SEQ ID NO:27) or GSGSGSGS (SEQ ID NO:28). In some exemplary embodiments, the redesigned hRSV soluble F immunogen has the amino acid sequence set forth in SEQ ID NO:34 or SEQ ID NO:35, or a conservatively modified variant thereof.

[0015] In some embodiments, modifications of the redesigned RSV soluble F immunogen compared to the wild-type soluble F sequence can further include an engineered disulfide bond linking the β-sheet-forming amino acid pair of the β3 / β4 hairpin of the F1 subunit. In some of these embodiments, the engineered disulfide bond is introduced between the substituted residues S180C / S186C or A177C / T189C. Some specific examples of these redesigned hRSV soluble F immunogens have the amino acid sequence set forth in any one of SEQ ID NOs: 36, 38, 40, and 42, or a conservatively modified variant thereof.

[0016] In some embodiments, the modifications in the redesigned RSV soluble F immunogen may further include the substitution of one or two amino acid residues between beta strands β3 and β4 compared to the wild-type soluble F sequence. In some of these embodiments, the amino acid substitutions are S182G and N183P. Some specific examples of these redesigned hRSV soluble F immunogens have the amino acid sequence set forth in any one of SEQ ID NOs: 37, 39, 41, and 43, or conservatively modified variants thereof.

[0017] In addition to the various sequence modifications or mutations described above, some redesigned soluble F immunogens of the present invention can further comprise a trimerization motif at the C-terminus. In some of these embodiments, the trimerization motif used is foldon or the viral capsid protein SHP.

[0018] In another aspect, the present invention provides paramyxovirus vaccine compositions containing a redesigned soluble F immunogen described herein displayed on the surface of a self-assembling nanoparticle. In some embodiments, the self-assembling nanoparticle comprises a trimeric sequence, and the C-terminus of the immunogen polypeptide is fused to the N-terminus of a subunit sequence of the nanoparticle. In another aspect, the present invention provides pharmaceutical compositions containing a redesigned soluble F immunogen or nanoparticle vaccine described herein and a pharmaceutically acceptable carrier. In another aspect, the present invention provides polynucleotide sequences encoding the subunit sequences of a redesigned soluble F immunogen described herein or a vaccine composition displaying a redesigned soluble F immunogen described herein.

[0019] A further understanding of the nature and advantages of the present invention may be realized by reference to the remaining portions of the specification and claims. [Brief explanation of the drawings]

[0020] [Figure 1-1] Analysis of the causes of RSV F metastability. (A) Amino acid sequence and secondary structure alignment of RSV prefusion and postfusion F, SEQ ID NO: 63 (Figure S3, from McLellan et al., Science 2013, 340:1113-1117). Two potential sources of instability, the β3 / β4 hairpin and the β23 strand, are enclosed in dotted boxes. (B) β3 / β4 hairpin (SEQ ID NO: 53). Left: prefusion and postfusion F; Right: Close-up of β3 / β4 in the prefusion and postfusion states (showing potential mutation sites). (C) β23 strand (SEQ ID NO: 64). Left: Side view of β23 in the prefusion F trimer structure; Right: Top view of β23 in the prefusion F trimer structure (top) and close-up of β23 around the three-fold symmetry axis (bottom). [Figure 1-2]Analysis of the causes of RSV F metastability. (A) Amino acid sequence and secondary structure alignment of RSV prefusion and postfusion F, SEQ ID NO: 63 (Figure S3, from McLellan et al., Science 2013, 340:1113-1117). Two potential sources of instability, the β3 / β4 hairpin and the β23 strand, are enclosed in dotted boxes. (B) β3 / β4 hairpin (SEQ ID NO: 53). Left: prefusion and postfusion F; Right: Close-up of β3 / β4 in the prefusion and postfusion states (showing potential mutation sites). (C) β23 strand (SEQ ID NO: 64). Left: Side view of β23 in the prefusion F trimer structure; Right: Top view of β23 in the prefusion F trimer structure (top) and close-up of β23 around the three-fold symmetry axis (bottom). [Figure 1-3] Analysis of the causes of RSV F metastability. (A) Amino acid sequence and secondary structure alignment of RSV prefusion and postfusion F, SEQ ID NO: 63 (Figure S3, from McLellan et al., Science 2013, 340:1113-1117). Two potential sources of instability, the β3 / β4 hairpin and the β23 strand, are enclosed in dotted boxes. (B) β3 / β4 hairpin (SEQ ID NO: 53). Left: prefusion and postfusion F; Right: Close-up of β3 / β4 in the prefusion and postfusion states (showing potential mutation sites). (C) β23 strand (SEQ ID NO: 64). Left: Side view of β23 in the prefusion F trimer structure; Right: Top view of β23 in the prefusion F trimer structure (top) and close-up of β23 around the three-fold symmetry axis (bottom). [Figure 2-1] Rational design of I3-01v9b / c to achieve optimal display of trimeric antigens on nanoparticle surfaces. (A) Structural model of I3-01v9a (SEQ ID NO: 24) with an extended N-terminal helix. (B) Scheme of the process used to design I3-01v9b / c (SEQ ID NO: 25). (C) nsEM analysis of EBOV GP-I3-01v9b trimer. Top: 2D class; bottom: side and top views of the 3D model. [Figure 2-2] Rational design of I3-01v9b / c to achieve optimal display of trimeric antigens on nanoparticle surfaces. (A) Structural model of I3-01v9a (SEQ ID NO: 24) with an extended N-terminal helix. (B) Scheme of the process used to design I3-01v9b / c (SEQ ID NO: 25). (C) nsEM analysis of EBOV GP-I3-01v9b trimer. Top: 2D class; bottom: side and top views of the 3D model. [Figure 2-3] Rational design of I3-01v9b / c to achieve optimal display of trimeric antigens on nanoparticle surfaces. (A) Structural model of I3-01v9a (SEQ ID NO: 24) with an extended N-terminal helix. (B) Scheme of the process used to design I3-01v9b / c (SEQ ID NO: 25). (C) nsEM analysis of EBOV GP-I3-01v9b trimer. Top: 2D class; bottom: side and top views of the 3D model. [Figure 2-4] Rational design of I3-01v9b / c to achieve optimal display of trimeric antigens on nanoparticle surfaces. (A) Structural model of I3-01v9a (SEQ ID NO: 24) with an extended N-terminal helix. (B) Scheme of the process used to design I3-01v9b / c (SEQ ID NO: 25). (C) nsEM analysis of EBOV GP-I3-01v9b trimer. Top: 2D class; bottom: side and top views of the 3D model. [Figure 3-1] Design and negative staining EM analysis of nanoparticles displaying RSV prefusion F trimers. (A) Structural modeling of RSV prefusion F trimers on three 1c-SApNPs containing a ferritin 24-mer and two 60-mers, E2p, and I3-01v9b. (B) EM analysis of DS-Cav1 on three 1c-SApNPs. (C) EM analysis of sc9-10 DS-Cav1 on three 1c-SApNPs. (D) V2-Ext-P2DB6-D-L2 and NQ on FR and NQ on I3-01v9b 1c-SApNP. (E) V2-Ext-P2DB6-GDQ-L2 and NQ on FR and NQ on I3-01v9b 1c-SApNP. (D) and (E) show enlarged views of F-FR nanoparticles with closed-structure pre-F trimers on their surfaces. [Figure 3-2] Design and negative staining EM analysis of nanoparticles displaying RSV prefusion F trimers. (A) Structural modeling of RSV prefusion F trimers on three 1c-SApNPs containing a ferritin 24-mer and two 60-mers, E2p, and I3-01v9b. (B) EM analysis of DS-Cav1 on three 1c-SApNPs. (C) EM analysis of sc9-10 DS-Cav1 on three 1c-SApNPs. (D) V2-Ext-P2DB6-D-L2 and NQ on FR and NQ on I3-01v9b 1c-SApNP. (E) V2-Ext-P2DB6-GDQ-L2 and NQ on FR and NQ on I3-01v9b 1c-SApNP. (D) and (E) show enlarged views of F-FR nanoparticles with closed-structure pre-F trimers on their surfaces. [Figure 4] Negative stain EM of RSV prefusion F trimers on nanoparticle platforms. (A) EM images of DS-Cav1, sc9-10 DS-cav1, and SC-TM RSV F displayed on ferritin nanoparticles. DS-Cav1-FR5 failed to properly form nanoparticles. While sc9-10-DS-Cav-FR5 formed nanoparticles, the F trimer appeared open on the ferritin particle surface (one such particle is highlighted in a red box). SC-TM-FR failed to properly form nanoparticles. (B) EM images of two major F trimer designs displayed on ferritin and E2p particles. Column 1: Purified by V2-Ext-SSGP, D25, and MPE8 on ferritin with a 10GS linker. Column 2: Purified by V2-Ext-AT, D25, and MPE8 on ferritin with a 5GS linker. Column 3: V2-Ext-AT on an E2p 60-mer with a locking domain (LD4) and carrying LD4 and the T cell epitope PADRE. All newly designed F trimers have well-formed closed prefusion conformations on the nanoparticle surface (one such particle, V2-Ext-AT-FR5, is highlighted). DETAILED DESCRIPTION OF THE INVENTION

[0021] I. Overview As a class I viral fusion protein, RSV F has inherent instability and can appear in different forms compared to other class I fusion proteins, such as the HIV-1 envelope glycoprotein (Env). First, prefusion RSV F is highly unstable and tends to change its conformation to a postfusion state, a phenomenon that has been extensively studied in the field. Second, the RSV prefusion F trimer tends to dissociate into monomers or form an open-structure trimer unless it is locked by interproton disulfide bonds, which adversely affects the multivalent display of prefusion F trimers on nanoparticles via gene fusion approaches.

[0022] The present invention is derived in part from the inventors' research to rationally design new stable RSV prefusion F trimers by minimizing F instability. The inventors first investigated the expression, purification, and structure of three known prefusion RSV F designs: DS-Cav1, SC-TM, and sc9-10 DS-Cav1. It was observed that DS-Cav1 and SC-TM produced many aggregates and non-trimeric F species, while sc9-10 DS-Cav1 exhibited a single trimer peak with high yield and purity. Negative staining EM analysis revealed that DS-Cav1 and SC-TM were monomeric and monomer / trimeric mixes, respectively, while sc9-10 DS-Cav1 appeared as a highly purified closed-structure prefusion F trimer. When displayed on protein nanoparticles (NPs), all three prefusion F designs performed poorly with low yield and purity. At the core of the present invention, the inventors investigated novel mutations that can substantially reduce the instability of the prefusion F conformation, such as RSV F mutant proteins in the prefusion conformation achieved by engineered disulfide bonds (e.g., S155C / S290C as exemplified herein). To this end, the inventors discovered that mutations of negatively charged residues in the β23 chain (D486-A490), such as D486 and E487, can significantly stabilize prefusion F in the closed trimeric conformation.

[0023] The inventors further investigated other mutations in the "base" prefusion F structure, which introduce the above mutations in the β23 chain, that could further improve the antigenic profile of the engineered F protein. In addition to the intraprotomer disulfide bond of the β23 chain that locks the RSV F protein in the prefusion conformation, these additional mutations include the unstructured F2 C-terminus, the P27 peptide and a cleavage site linker that replaces the fusion peptide, the S215P mutation, and the E92D mutation. This F construct is referred to as "V2-Ext-PDB6-D." Constructs containing this mutation set, V2-Ext-PDB6-D, were found to produce high-yield and high-purity prefusion F with a small amount of closed-structure trimer. Using this basic design of prefusion F, we further investigated the V185P mutation (K176-S190) in the β3 / β4 hairpin and two mutations to the pair of negatively charged residues D486 and E487 in the β23 strand (D486-A490). The three β23 strands form repulse interactions around the three-fold symmetry axis directly above the α10 coiled-coil. While the V185P mutation may destabilize the postfusion F structure, we hypothesized that the D486-E487 pair might open the prefusion F trimer to facilitate cell entry, and that polar or hydrophobic mutations to these two residues might stabilize prefusion F in a closed trimer conformation. Indeed, negative-stain EM analysis of various constructs confirmed this hypothesis. We then included two mutations, S46G and K465Q, in the first basic design to improve prefusion F folding. This second basic design, designated "V2-Ext-PDB6-GDQ," was used to test the V185P mutation and the D486-E487 pair. S46G and K462Q improved prefusion F folding but also reduced the proportion of closed prefusion F trimers. Both the V2-Ext-PDB6-D and V2-Ext-PDB6-GDQ derivatives can be developed as soluble trimeric vaccines or displayed on single-component self-assembling protein nanoparticles (lc-SApNPs) as virus-like particle (VLP)-type vaccines.Because the F proteins of other paramyxoviruses (e.g., hMPV and PIV) are structurally similar to RSV F, the same design strategy can be applied to other members of the Paramyxoviridae family.

[0024] In further studies, the inventors investigated different sets of minimal mutations that can be introduced into the paramyxovirus fusion (F) glycoprotein to generate stabilized immunogens. As detailed herein, these paramyxovirus F protein trimer immunogens were redesigned by introducing structural modifications into the soluble F sequence that could stabilize the F trimer in the prefusion state. Some of the redesigned soluble F immunogens were stabilized by introducing an engineered disulfide bond into the β-hairpin in the F1 subunit. Some of the redesigned soluble F immunogens were stabilized by introducing other novel mutations into the soluble F sequence. The inventors further displayed the redesigned F trimer immunogens on self-assembling nanoparticles as VLP-type vaccines and developed methods for industrial production and tag-free antigen-specific purification of the immunogens. Because hMPV and PIV F proteins are structurally similar to RSV F, the same stabilization designs validated for RSV F were also utilized to redesign hMPV and PIV3 vaccine immunogens. These studies provide a universal design strategy for paramyxovirus prefusion F-based vaccine design.

[0025] Thus, the present invention provides paramyxovirus immunogens and vaccine compositions according to the design strategies described herein. Related polynucleotide sequences, expression vectors, and pharmaceutical compositions are also provided herein. Unless otherwise specified herein, the vaccine immunogens, coding polynucleotides, expression vectors, and host cells of the present invention, as well as related therapeutic applications, can all be produced or carried out according to procedures exemplified herein or routinely practiced methods known in the art. See, for example, Methods in Enzymology, Volume 289: Solid-Phase Peptide Synthesis, JN Abelson, MI Simon, GB Fields (Editors), Academic Press; 1st edition (1997) (ISBN-13:978-0121821906); U.S. Patent Nos. 4,965,343 and 5,849,954; Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, NY, (3 rdBrent et al., Current Protocols in Molecular Biology, John Wiley&Sons, Inc. (ringbou ed., 2003); Davis et al., Basic Methods in Molecular Biology, Elsevier Science Publishing, Inc., New York, USA (1986); or Methods in Enzymology: Guide to Molecular Cloning Techniques Vol.152, SLBerger and ARKimmerl Eds., Academic Press Inc., San Diego, USA (1987); Current Protocols in Protein Science (CPPS) (John E. Coligan, et.al., ed., John Wiley and Sons, Inc.), Current Protocols in Cell Biology (CPCB) (Juan S. Bonifacino et.al. ed., John Wiley and Sons, Inc.) and Culture of Animal Cells: A Manual of Basic Technique by R.Ian See Freshney, Publisher: Wiley-Liss; 5th edition (2005), Animal Cell Culture Methods (Methods in Cell Biology, Vol. 57, Jennie P. Mather and David Barnes editors, Academic Press, 1st edition, 1998). The following sections provide further guidance for carrying out the compositions and methods of the present invention.

[0026] II. Definition Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The following references provide those skilled in the art with general definitions of many of the terms used in this invention: Academic Press Dictionary of Science and Technology, Morris (Ed.), Academic Press (1999). st ed.,1992);Oxford Dictionary of Biochemistry and Molecular Biology,Smith et al.(Eds.),Oxford University Press(revised ed.,2000);Encyclopaedic Dictionary of Chemistry,Kumar(Ed.),Anmol Publications Pvt.Ltd.(2002);Dictionary of Microbiology and Molecular Biology,Singleton et al. al.(Eds.),John Wiley&Sons(3 rd ed.,2002);Dictionary of Chemistry,Hunt(Ed.),Routledge(1 st ed., 1999); Dictionary of Pharmaceutical Medicine, Nahler (Ed.), Springer-Verlag Telos (1994); Dictionary of Organic Chemistry, Kumar and Anandand (Eds.), Anmol Publications Pvt. Ltd. (2002); and A Dictionary of Biology (Oxford Paperback Reference), Martin and Hine (Eds.), Oxford University Press (4 th ed., 2000). Further clarification of some of these terms as they apply specifically to the present invention is provided herein.

[0027] As used herein, the singular forms "a," "an," and "the" refer to both the singular and the plural unless the context clearly dictates otherwise. For example, "Env-derived trimer" can refer to both a single or multiple Env-derived trimer molecules and can be considered equivalent to the phrase "at least one Env-derived trimer."

[0028] As used herein, the terms "antigen" and "immunogen" are used interchangeably to refer to a substance, typically a protein, that can induce an immune response in a subject. The terms also refer to a protein that is immunologically active in the sense that it can elicit a humoral and / or cellular immune response against the protein when administered to a subject (either directly or by administering to the subject a nucleotide sequence or vector encoding the protein). Unless otherwise specified, the term "vaccine immunogen" is used interchangeably with "protein antigen" or "immunogenic polypeptide."

[0029] The term "conservatively modified variants" applies to both amino acid and nucleic acid sequences. With respect to a particular nucleic acid sequence, conservatively modified variants refers to nucleic acids that encode the same or essentially identical amino acid sequences, or, where the nucleic acid does not encode an amino acid sequence, essentially identical sequences. Due to the degeneracy of the genetic code, a large number of functionally identical nucleic acids encode any given protein. In the case of polypeptide sequences, "conservatively modified variants" refers to variants that have conservative amino acid substitutions, amino acid residues replaced with other amino acid residues having side chains with similar charges. Families of amino acid residues having side chains with similar charges have been defined in the art. These families include amino acids with basic side chains (e.g., lysine, arginine, histidine), acidic side chains (e.g., aspartic acid, glutamic acid), uncharged polar side chains (e.g., glycine, asparagine, glutamine, serine, threonine, tyrosine, cysteine), nonpolar side chains (e.g., alanine, valine, leucine, isoleucine, proline, phenylalanine, methionine, tryptophan), beta-branched side chains (e.g., threonine, valine, isoleucine), and aromatic side chains (e.g., tyrosine, phenylalanine, tryptophan, histidine).

[0030] Epitope refers to an antigenic determinant. These are specific chemical groups or peptide sequences on a molecule that are antigenic and elicit a specific immune response; for example, an epitope is a region of an antigen to which B cells and / or T cells respond. Epitopes can be formed both from contiguous or non-contiguous amino acids juxtaposed by tertiary folding of a protein.

[0031] An effective amount of a vaccine or other agent is sufficient to produce a desired response, e.g., an effective amount of a vaccine or other agent sufficient to reduce or eliminate signs or symptoms of a condition or disease, such as bronchiolitis or pneumonia. For example, this can be the amount necessary to inhibit viral replication or to measurably change the outward symptoms of a viral infection. Generally, this amount will be sufficient to measurably inhibit the replication or infectivity of a virus (e.g., hRSV). When administered to a subject, a dosage that achieves a target tissue concentration shown to achieve in vitro inhibition of viral replication is generally used. In some embodiments, an "effective amount" is one that treats (including prevents) one or more symptoms and / or underlying causes of any disorder or disease, e.g., to treat RSV infection. In some embodiments, an effective amount is a therapeutically effective amount. In some embodiments, an effective amount is an amount that prevents the onset of one or more signs or symptoms of a particular disease or condition, e.g., one or more signs or symptoms associated with bronchiolitis.

[0032] Unless otherwise specified, a fusion protein is a recombinant protein containing amino acid sequences from at least two unrelated proteins joined together via peptide bonds to create a single protein. Therefore, it does not encompass naturally occurring paramyxovirus surface antigens, referred to herein as fusion (F) proteins. Unrelated amino acid sequences can be joined directly to each other or can be joined using a linker sequence. As used herein, proteins are unrelated if their amino acid sequences are not normally found joined together via peptide bonds in their natural environment(s) (e.g., inside a cell). For example, the amino acid sequence of a bacterial enzyme, such as B. stearothermophilus dihydrolipoyl acyltransferase (E2p), and the amino acid sequence of a soluble paramyxovirus F glycoprotein are not normally found joined together via peptide bonds.

[0033] An immunogen is a protein or portion thereof that can induce an immune response in a mammal, such as a mammal infected with or at risk of infection with a pathogen. Administration of the immunogen can result in protective and / or prophylactic immunity against the pathogen of interest.

[0034] An immunogenic composition refers to a composition comprising an immunogenic polypeptide that induces a measurable CTL response against a virus expressing the immunogenic polypeptide, or that induces a measurable B cell response (such as the production of antibodies) against the immunogenic polypeptide.

[0035] Sequence identity or similarity between two or more nucleic acid sequences or two or more amino acid sequences is expressed in terms of the identity or similarity between the sequences. Sequence identity can be measured in terms of percent identity; the higher the percentage, the more identical the sequences. Two sequences are "substantially identical" if they have a specified percentage of the same amino acid residues or nucleotides (i.e., identity over a designated region, or, if not specified, over the entire sequence), when compared and aligned for maximum correspondence over a comparison window or designated region, as measured using one of the following sequence comparison algorithms or by manual alignment and visual inspection. Identity may exist over a region at least about 50 nucleotides (or 10 amino acids) in length, or more preferably over a region 100 to 500 or 1000 or more nucleotides (or 20, 50, 200 or more amino acids) in length.

[0036] Homologs or orthologs of nucleic acid or amino acid sequences have a relatively high degree of sequence identity / similarity when aligned using standard methods. Methods for aligning sequences for comparison are well known in the art. Various programs and alignment algorithms are described in Smith & Waterman, Adv. Appl. Math. 2:482, 1981; Needleman & Wunsch, J. Mol. Biol. 48:443, 1970; Pearson & Lipman, Proc. Natl. Acad. Sci. USA 85:2444, 1988; Higgins & Sharp, Gene, 73:237-44, 1988; Higgins & Sharp, CABIOS 5:151-3, 1989; Corpet et al., Nuc. Acids Res. 16:10881-90, 1988; Huang et al., Computer Appls. in the Biosciences 8, 155-65, 1992; and Pearson et al. al., Meth. Mol. Bio. 24:307-31, 1994. Altschul et al., J. Mol. Biol. 215:403-10, 1990 presents a detailed discussion of sequence alignment methods and homology calculations.

[0037] The term "subject" refers to any animal classified as a mammal, including humans and non-human mammals. Examples of non-human animals include dogs, cats, cows, horses, sheep, pigs, goats, rabbits, etc. Unless otherwise specified, the terms "patient" or "subject" are used interchangeably herein. Preferably, the subject is a human.

[0038] The terms "treating" or "alleviating" include administering a compound or agent to a subject to prevent or delay the onset of symptoms, complications, or biochemical manifestations of a disease (e.g., hRSV infection), alleviate symptoms, or arrest or inhibit further development of a disease, condition, or disorder. Subjects in need of treatment include those already suffering from a disease or disorder as well as those at risk of developing a disorder. Treatment can be prophylactic (to prevent or delay the onset of a disease or to prevent the onset of its clinical or subclinical symptoms) or therapeutic suppression or alleviation of symptoms after the onset of a disease.

[0039] A vaccine refers to a pharmaceutical composition that induces a prophylactic or therapeutic immune response in a subject. In some cases, the immune response is a protective immune response. Typically, a vaccine induces an antigen-specific immune response against an antigen of a pathogen, e.g., a viral pathogen, or a cellular component correlated with a pathological condition. A vaccine can comprise a polynucleotide (such as a nucleic acid encoding a disclosed antigen), a peptide or polypeptide (such as a disclosed antigen), a virus, a cell, or one or more cellular components. In some embodiments of the invention, a vaccine or vaccine immunogen or vaccine composition is expressed from a fusion construct and self-assembles into nanoparticles that display the immunogenic polypeptide or protein on their surface.

[0040] Virus-like particles (VLPs) refer to non-replicative viral shells derived from any of several viruses. VLPs are generally composed of one or more viral proteins, including, but not limited to, proteins referred to as capsid, coat, shell, surface, and / or envelope proteins, or particle-forming polypeptides derived from these proteins. VLPs can spontaneously form upon recombinant expression of proteins in an appropriate expression system. Methods for producing specific VLPs are known in the art. The presence of VLPs following recombinant expression of viral proteins can be detected using conventional techniques known in the art, such as electron microscopy, biophysical characterization, etc. See, e.g., Baker et al. (1991) Biophys. J. 60:1445-1456; and Hagensee et al. (1994) J. Virol. 68:4503-4505. For example, VLPs can be isolated by density gradient centrifugation and / or identified by characteristic density banding. Alternatively, cryo-electron microscopy can be performed on vitrified aqueous samples of the VLP preparation in question and images recorded under appropriate exposure conditions.

[0041] Self-assembling nanoparticles are ball-shaped protein shells with distinct surface geometries formed by identical copies of non-viral proteins that can automatically assemble into nanoparticles with diameters of tens of nanometers and an appearance similar to VLPs. A notable example of a self-assembling nanoparticle is the engineered protein I3-01 (Hsia et al., Nature 535, 136-139, 2016) and its derived variants, including I3-01v9b and I3-01v9c, exemplified herein. Other examples include ferritin (FR), which is conserved across species and forms a 24-mer, as well as B. stearothermophilus dihydrolipoyl acyltransferase (E2p), Aquifex aeolicus lumazine synthase (LS), and Thermotoga maritima encapsulin, all of which form a 60-mer. Self-assembling nanoparticles can form spontaneously upon recombinant expression of proteins in appropriate expression systems. Methods for nanoparticle generation, detection, and characterization can be performed using the same techniques as those developed for VLPs.

[0042] III. Paramyxoviruses and Fusion (F) Glycoproteins The present invention provides novel engineered immunogenic proteins and vaccine compositions containing modified soluble paramyxovirus F glycoprotein sequences. Paramyxovirus F proteins are homotrimers. They contain a hydrophobic fusion peptide (FP), two heptad repeat regions (HRA and HRB), are surface-anchored by a single-pass transmembrane domain (TM), and contain a C-terminal cytoplasmic tail. Using RSV as an illustration, the paramyxovirus F gene encodes a type I integral membrane protein that is synthesized as a 574-amino acid inactive precursor, F0. Three F0 monomers assemble into trimers, and as the trimers transit the Golgi, the monomers are activated by a furin-like host protease. The protease cleaves twice, after amino acids 109 and 136, to generate three polypeptides. The N- and C-terminal cleavage products are the F2 and F1 subunits, respectively (designated in order of size), which are covalently linked to each other by two disulfide bonds. The intervening 27-amino acid peptide, P27, contains two or three N-linked glycans that dissociate after cleavage. The F2 subunit contains two N-linked glycans, while the larger F1 subunit contains a single N-linked glycosylation site. Unlike the others, this F1 glycan is essential for the protein to induce membrane fusion.

[0043] In the general technical field, the wild-type soluble F sequence of a paramyxovirus refers to the entire ectodomain of the fusion glycoprotein (F) of the paramyxovirus. Using the RSV F protein as an example, the soluble F sequence (amino acids 1-529) typically comprises, from its N-terminus to its C-terminus, a leader sequence, followed by the F2 subunit, the processed activation peptide (P27) peptide, and the ectodomain portion of the F1 subunit containing the fusion peptide (FP) at its N-terminus. Deletion of the wild-type soluble F sequence at the C-terminus results in an F construct called Fd (amino acids 1-513), which has been used to determine the structure of the RSV F protein in both prefusion and postfusion conformations. Thus, as used herein, the term wild-type soluble F can refer to Fd, which in some embodiments can be extended by adding additional amino acids to the C-terminus to include the full-length ectodomain portion of the F1 subunit. Unless otherwise specified, the amino acid numbering of the various components of the RSV soluble F sequence is based on the human RSV A2 strain with accession number P03420 (McLellan et al., J. Virol. 85:7788-96, 2011). In some embodiments, the wild-type soluble RSV F sequence from which the engineered soluble F immunogens of the invention are derived is set forth in SEQ ID NO: 1. Similarly, the amino acid numbering in the soluble F sequences of other paramyxoviruses is also based on the specific viral strains and / or secondary structures described herein.

[0044] hRSV A2 strain “wild type” soluble F sequence (SEQ ID NO: 1): [ka]

[0045] As used herein, the unstructured F2 C-terminus of a paramyxovirus F glycoprotein refers to a segment of the amino acid sequence at the C-terminus of its F2 subunit that is flexible and therefore not visible in the three-dimensional structure of the prefusion F protein. Many paramyxoviruses have an unstructured F2 C-terminus. For example, based on the crystal structures determined for various RSV prefusion F constructs, it was found that the F2 C-terminus is always unstructured. The crystal structure solved for the hMPV prefusion F construct shows that the F2 C-terminus of hMPV is unstructured. Similarly, an EM structure has been solved for the PIV3 prefusion F construct, showing that the F2 C-terminus of PIV3 is unstructured.

[0046] IV. Engineered Paramyxovirus Soluble F Immunogens The present invention provides engineered (redesigned or modified) soluble F sequences of paramyxoviruses that can be used to generate vaccine compositions. The redesigned soluble F trimeric immunogens or proteins are stabilized by introducing modifications into the wild-type soluble F sequence of a paramyxovirus. Several specific wild-type soluble F sequences of specific hRSV strains are exemplified herein. Due to functional similarities and sequence homology between different strains of a given paramyxovirus, redesigned soluble F immunogens derived from other known paramyxovirus F protein ortholog sequences can also be generated according to the redesign strategy described herein. There are many known paramyxovirus ortholog or homolog F protein sequences described in the literature. See, e.g., Collins et al., Proc. Natl. Acad. Sci. USA 81:7683-7, 1984; Hause et al., PLoS ONE 12:e0175792, 2017; Chang et al., Viruses 4:613-636, 2012; and Amanda et al., J. Virol. 81:8303-8314, 2007. As detailed herein, the engineered soluble F proteins of the invention contain one or more specific stabilizing mutations in the corresponding wild-type soluble F sequence. These mutations include (a) substitution of two or more negatively charged residues around the β23 chain, as exemplified herein for RSV, (b) deletion of the P27 peptide, (c) engineered intraprotomer disulfide bonds within the F1 subunit or linking the F2 and F1 subunits, and (d) engineered interprotomer disulfide bonds, as exemplified herein for RSV. In some embodiments, the engineered soluble F proteins of the invention can include a combination of any two (e.g., a and d) or three (e.g., a, c, and d) of these mutations. In some embodiments, the engineered soluble F proteins can include all four of these mutations.

[0047] In some embodiments, compared to the wild-type counterpart sequence, the engineered soluble F protein comprises the substitution of two or more negatively charged residues around the β23 chain with polar or hydrophobic residues. As used herein, negatively charged residues around the β23 chain refer to a negatively charged stretch centered around or around β23. It may include the β23 chain and one to two residues upstream and downstream. In addition to these substitutions, the engineered soluble F sequences of the invention typically also lack the processed activation peptide (P27) and / or contain an engineered stabilizing disulfide bond present within the Fs subunit or linking the F2 and F1 subunits, locking the protein in a prefusion conformation. Using the prototype human RSV strain A2 F protein (UniProt ID P03420) as a reference, the P27 peptide corresponds to residues E110-R136, and the negatively charged stretch encompasses the β23 chain D486-A490 and surrounding residues. Notably, residue 485 is Ser (S) in human RSV but Glu (E453) in human MPV. This residue is also included as part of the negatively charged stretch around β23 that forms repulsive interactions around the trimer symmetry axis. In some embodiments, the engineered soluble F proteins of the invention have two residues, D486 and E487, in the β23 chain, replaced with polar or hydrophobic residues. In some embodiments of other paramyxoviruses, such as hMPV, the residue immediately upstream of the β23 chain (E453) can also be replaced with a polar or hydrophobic residue. In some embodiments, the engineered disulfide bond is S155C / S290C present in the F1 subunit. In some other embodiments, the engineered disulfide bond is S62C / K196C or E60C / K196C, linking the F2 and F1 subunits.

[0048] In addition to the above modifications, some engineered soluble F proteins of the present invention may contain one or more additional mutations compared to their wild-type counterparts. In some embodiments, they have an inserted linker moiety that replaces the furin cleavage site. In some other embodiments, a linker moiety is inserted to replace the unstructured C-terminus of F2 and a portion of the N-terminus of the fusion peptide (FP). Again, using human RSV strain A2 as a reference, the replaced C-terminus of F2 corresponds to residues N104 to R109 (NNRARR; SEQ ID NO: 31). The replaced N-terminus of the fusion peptide (FP) (F137-V157) contains residues F137 to S146. In some preferred embodiments, the replaced linker moiety is a GS-rich linker, e.g., (GS)n, where n can be any integer from 1 to about 5. In various embodiments, the linker comprises the sequence GSGS (SEQ ID NO: 27) or GSGSGSGS (SEQ ID NO: 28).

[0049] In some embodiments, the engineered soluble F proteins of the invention can further comprise a substitution of a residue corresponding to S215 in the F glycoprotein of RSV strain A2. In some embodiments, the engineered soluble F proteins of the invention can further comprise a substitution of a residue corresponding to E92 in the F glycoprotein of RSV strain A2. For example, residue E92 in the soluble F protein can be replaced with D, Q, or another short polar residue, or optionally with hydrophobic residues such as L and I, to form hydrophobic interactions with adjacent promoters. In some embodiments, the engineered soluble F proteins of the invention can further comprise substitutions at residues S46 and K462. In some of these embodiments, the substitution is S46G / K462Q. In yet some other embodiments, the engineered soluble F proteins of the invention comprise an engineered disulfide bond linking a β-sheet-forming amino acid pair in the β3 / β4 hairpin or equivalent hairpin in the F1 subunit. As demonstrated herein, this engineered disulfide bond functions to further reduce instability and increase the stability of the prefusion soluble F sequence. Note that this β3 / β4 hairpin in RSV and MPV is equivalent to the β1 / β2 hairpin in PIV. In some of these embodiments, this disulfide bond is formed via the substitutions A177C / T189C, with amino acid numbering based on human RSV strain A2.

[0050] Some specific examples of engineered RSV soluble F sequences or immunogens are shown in SEQ ID NOS: 17 to 23. In addition to these exemplary sequences, the engineered RSV soluble F immunogens of the invention also include conservatively modified variants of these sequences or sequences that are substantially identical.

[0051] In addition to the modifications in the wild-type soluble F sequence described above, some engineered soluble F antigens or immunogenic proteins of the invention may contain an N-terminal leader sequence (or "signal peptide"). In some of these embodiments, the N-terminal leader comprises the sequence MELLILKANAITTILTAVTFCFASG (SEQ ID NO: 2), as exemplified herein. In some embodiments, the engineered soluble F proteins of the invention may also comprise one or more C-terminal structural motifs that promote trimerization. For example, the engineered protein may comprise a C-terminal foldon motif, GYIPEAPRDGQAYVRKDGEWVLLSTFL (SEQ ID NO: 7), as exemplified herein. In some of these embodiments, a restriction site, such as "AS," as exemplified herein, may be added to the N-terminus of the foldon motif.

[0052] V. Engineered paramyxovirus soluble F trimers with different mutation sets In addition to the modifications introduced into the soluble F trimers described above, the present invention also provides engineered paramyxovirus soluble F immunogens containing distinct sets of mutations compared to the wild-type F sequence. Again, several specific wild-type soluble F sequences of specific hRSV, hMPV, and hPIV3 strains are exemplified herein for these additional engineered paramyxovirus soluble F immunogens. Due to functional similarities and sequence homology between different strains of a given paramyxovirus, redesigned soluble F immunogens derived from other known paramyxovirus F protein ortholog sequences can also be generated following the redesign strategy described herein. There are many known paramyxovirus ortholog or homolog F protein sequences described in the literature. See, e.g., Collins et al., Proc. Natl. Acad. Sci. USA 81:7683-7, 1984; Hause et al., PLoS ONE 12:e0175792, 2017; Chang et al., Viruses 4:613-636, 2012; and Amanda et al., J. Virol. 81:8303-8314, 2007.

[0053] As detailed herein (e.g., Examples 8-14), some of these additional redesigned soluble F immunogens of the invention contain engineered disulfides linking pairs of β-sheet-forming amino acid residues in the β3 / β4 hairpin (or equivalent hairpin) in the F1 subunit of the soluble F sequence. Some other redesigned soluble F proteins contain a set of mutations that can stabilize the F trimer in the prefusion state. These include mutations at the C-terminus of F2, deletion of the P27 peptide, and mutations at the N-terminus of the fusion peptide in the F1 subunit. Some other redesigned soluble F proteins may contain engineered disulfide bonds as well as one or more of these mutations.

[0054] In one aspect, the present invention provides engineered or redesigned immunogenic proteins or polypeptides derived from the fusion glycoprotein (F) of any paramyxovirus. These immunogens contain a modified soluble F sequence with modifications compared to the wild-type soluble F sequence of the paramyxovirus. The modifications include an engineered disulfide bond linking a pair of β-sheet-forming amino acids in the β3 / β4 hairpin of the paramyxovirus F protein or the equivalent hairpin in the F1 subunit. This hairpin is the β3 / β4 hairpin in RSV and MPV. In PIV, the equivalent hairpin is the β1 / β2 hairpin. Some of these immunogens are derived from the wild-type soluble F sequence of RSV, such as human RSV (hRSV). In some of these embodiments, the engineered disulfide bond is generated by the amino acid substitutions S180C / S186C in the β3 / β4 hairpin. In some other embodiments, the engineered disulfide bond is generated by the amino acid substitutions A177C / T189C in the hairpin. The amino acid numbering in the redesigned RSV F immunogens of the present invention is based on the F glycoprotein sequence of human RSV strain A2, which has UniProt ID number P03420. In some embodiments, the wild-type soluble RSV F sequence from which the redesigned immunogen is derived is set forth in SEQ ID NO: 1.

[0055] In some redesigned RSV soluble F immunogens of the invention, in addition to engineered disulfide bonds, modifications of the wild-type sequence also include mutations at the unstructured C-terminus of the F2 subunit. For example, the redesigned soluble F sequence can include a truncation at the unstructured F2 C-terminus. As a specific example, residues 104-109 (NNRARR) (SEQ ID NO: 31) at the C-terminus of F2 can be deleted. Additionally or alternatively, amino acid substitutions at the C-terminus of unstructured F2 can be introduced into the redesigned soluble F sequence. For example, some of the redesigned RSV soluble F immunogens of the invention can include a P102A substitution at the F2 C-terminus.

[0056] In some redesigned RSV soluble F immunogens of the present invention, modifications of the wild-type sequence may include one or more other mutations. These include, for example, the substitution of a short GS linker sequence for the processed active peptide (P27) (residues E110-R136) and the N-terminus (e.g., residues F137-S146) of the fusion peptide. In various embodiments, the GS linker has the sequence formula (GS) n where n is any integer from 1 to about 5. Additional modifications can also include further substitutions in the F1 subunit. These include, for example, the substitutions I379V and M447V, as exemplified herein using the wild-type soluble F sequence set forth in SEQ ID NO:1.

[0057] Some specific examples of redesigned RSV soluble F sequences or immunogens are shown in SEQ ID NOS: 36 to 43. In addition to these exemplary sequences, the redesigned RSV soluble F immunogens of the present invention also include conservatively modified variants of these sequences or sequences that are substantially identical.

[0058] Some of the redesigned soluble F immunogens containing engineered disulfide bonds are derived from the wild-type soluble F sequence of a metapneumovirus, such as hMPV. In some of these embodiments, the engineered disulfide bond is generated by the amino acid substitution A147C / A159C in the β3 / β4 hairpin. The amino acid numbering is based on the crystal structure (PDB ID: 5WB0) and the UniProt definition of hMPV strain CAN97-83, which has ID Q6WB98. In some embodiments, the wild-type soluble MPV F sequence from which the redesigned immunogen is derived is set forth in SEQ ID NO: 44 or SEQ ID NO: 45. These two sequences are based on hMPV isolate TN 03.03.19, which has GenBank ID AEZ52364.

[0059] In addition to engineered disulfide bonds, modifications of the wild-type sequence in some redesigned hMPV soluble F immunogens of the invention also include mutations at the unstructured C-terminus of the F2 subunit. In some of these embodiments, the mutation at the unstructured F2 C-terminus is a deletion of DQLAREEQIENP (SEQ ID NO:60) and the cleavage site RQSR (SEQ ID NO:49) at the unstructured F2 C-terminus. Furthermore, the deleted sequence can be replaced with a shorter (GS)n linker sequence as described above. Specific examples of some redesigned hMPV soluble F immunogens of the invention are shown in SEQ ID NOs:46 and 47. In addition to these exemplified sequences, redesigned hMPV soluble F immunogens of the invention also include conservatively modified variants of these sequences or sequences that are substantially identical.

[0060] Several other redesigned soluble F immunogens containing engineered disulfide bonds are derived from the wild-type soluble F sequence of human parainfluenza viruses. These include, for example, human parainfluenza viruses 1-5 (hPIV1-5). Using hPIV3 as an example, some redesigned hPIV3 soluble immunogens have an engineered disulfide bond created by the amino acid substitution Q159C / A171C in the β1 / β2 hairpin. Amino acid numbering is based on the cryo-EM structure (PDB ID: 6MJZ) and the UniProt definition of the recombinant PIV3 / PIV1 virus with ID (O55888). In some embodiments, the wild-type soluble MPV F sequence from which the redesigned immunogens are derived is set forth in SEQ ID NO: 48. This sequence is derived from the F protein of the hPIV3 strain "HPIV3 / USA / 629-D01959 / 2007," which has GenBank ID AGW51052. Due to the substantial structural similarities between different PIVs (e.g., hPIV3 and hPIV5 as exemplified herein), the redesign strategies exemplified herein for hPIV3 can be readily applied to other PIVs.

[0061] In addition to engineered disulfide bonds, modifications of the wild-type sequence in some redesigned PIV soluble F immunogens of the invention can also include mutations at the C-terminus of the F2 subunit. In some of these embodiments, the mutations at the C-terminus of F2 are deletions of NQESNENTDP (SEQ ID NO:50) and the cleavage site RTER (SEQ ID NO:51). Additionally, the deleted sequences can be replaced with shorter (GS)n linker sequences as described above. Specific examples of some redesigned hMPV soluble F immunogens of the invention are shown in SEQ ID NOs:61 and 62. In addition to these exemplified sequences, redesigned PIV soluble F immunogens of the invention also include conservatively modified variants of these sequences or sequences that are substantially identical.

[0062] In another aspect, the present invention provides engineered or redesigned RSV soluble F immunogens or proteins stabilized by a specific set of modifications to the wild-type RSV soluble F sequence. Using amino acid numbering based on human RSV strain A2 (UniProt ID P03420), the modifications to the wild-type soluble RSV F sequence in these redesigned immunogens include (1) deletion of the P27 peptide (residues E110-R136), (2) a modification at the unstructured C-terminus of the F2 subunit (residues Q98-R109), and (3) truncation of the N-terminus of the fusion peptide (e.g., residues F137-V157). In some embodiments, the wild-type soluble RSV F sequence from which the redesigned immunogen is derived is set forth in SEQ ID NO: 1.

[0063] In some of these immunogen proteins, the modification at the C-terminus of unstructured F2 is a truncation of residues 104-109 (NNRARR; SEQ ID NO: 31) and a P102A substitution. In some embodiments, the N-terminal truncation of the fusion peptide is a deletion of residues F137-S146. In some embodiments, the redesigned RSV soluble F immunogen polypeptide comprises a (GS)n linker inserted between F2 and F1. In the linker formula, n can be any integer from 1 to about 5. In various embodiments, the linker comprises the sequence GSGS (SEQ ID NO: 27) or GSGSGSGS (SEQ ID NO: 28). In some embodiments, the redesigned RSV soluble F immunogen polypeptide comprises the amino acid substitutions I379V and / or M447V. Exemplary redesigned RSV soluble F immunogen sequences are set forth in SEQ ID NO: 34 or SEQ ID NO: 35. In addition to these exemplary sequences, the redesigned RSV soluble F immunogens of the present invention also include conservatively modified variants of these sequences or sequences that are substantially identical.

[0064] In addition to the specific mutation sets described above, modifications of the redesigned RSV soluble F immunogens of the present invention compared to the wild-type soluble RSV F sequence can also include engineered disulfide bonds. The engineered disulfides link a pair of β-sheet-forming amino acids in the β3 / β4 hairpin in the F1 subunit, reducing instability and increasing the stability of the prefusion soluble F sequence. In some of these embodiments, the engineered disulfide bond is generated by the amino acid substitution S180C / S186C in the β3 / β4 hairpin. In some other embodiments, the engineered disulfide bond is generated by the amino acid substitution A177C / T189C in the β3 / β4 hairpin. Some examples of these redesigned RSV soluble F immunogen sequences of the present invention are set forth in SEQ ID NOs: 36, 38, 40, and 42. In addition to these exemplary sequences, the redesigned RSV soluble F immunogens of the present invention also include conservatively modified variants of these sequences or sequences that are substantially identical.

[0065] In some other embodiments, modifications in the redesigned RSV soluble F immunogens of the invention may include substitutions of amino acid residues between the two β-strands, β3 and β4. For example, the redesigned immunogen sequence may include the amino acid substitutions S182G and / or N183P. Some specific examples of these redesigned RSV soluble F immunogen sequences are shown in SEQ ID NOs: 37, 39, 41, and 43. In some embodiments, the redesigned RSV soluble F immunogens of the invention may have sequences that are conservatively modified variants of, or substantially identical to, these exemplary sequences.

[0066] VI. Nanoparticle-displayed vaccine compositions The present invention provides vaccine compositions containing heterologous scaffolds that display the stabilized soluble F protein or immunogen of a paramyxovirus described herein. Any heterologous scaffold can be used to present the engineered soluble F protein or immunogen in the construction of the vaccines of the present invention. This includes bacteriophage Q β These include virus-like particles (VLPs) and nanoparticles such as VLPs. Various nanoparticle platforms can be used to prepare the vaccine compositions of the present invention. Generally, the nanoparticles used in the present invention must be formed by multiple copies of a single subunit. The nanoparticles are typically ball-shaped and / or have rotational symmetry (e.g., three- and five-fold axes of symmetry), such as the icosahedral structures exemplified herein. Additionally or alternatively, the amino termini of the particle subunits must be exposed and close to the three-fold axis of symmetry, and the spacing between the three amino termini must closely match the spacing between the carboxyl termini of the displayed trimer-stabilized soluble F protein.

[0067] In various embodiments, the self-assembling nanoparticles used have a diameter of about 25 nm or less (typically assembled from 12, 24, or 60 subunits) and a three-fold axis of symmetry on the particle surface. Such nanoparticles provide a particle platform suitable for producing multivalent vaccines. In some preferred embodiments, paramyxovirus immunogenic proteins or polypeptides can be displayed on self-assembling nanoparticles, such as those derived from I3-01 (I3-01v9b and I3-01v9c) exemplified herein. Other examples of nanoparticles suitable for the present invention include nanoparticles derived from ferritin (FR) or E2p. Ferritin, well known and routinely used in the art, is a globular protein found in all animals, bacteria, and plants. As is well known in the art, it primarily acts to control the rate and location of polynuclear Fe(III)2O3 formation via the transport of hydrated iron ions and protons into and out of the calcified core. The spherical form of ferritin is composed of monomeric subunit proteins (also called monomeric ferritin subunits), which are polypeptides with molecular weights of approximately 17–20 kDa. E2p is a redesigned variant of dihydrolipoyl acyltransferase from Bacillus stearothermophilus that has been shown to self-assemble into thermostable 60-mer nanoparticles. See, for example, He et al., Nat. Commun. 7:12041, 2016. Similarly, I3-01 is an engineered protein capable of self-assembling into ultrastable nanoparticles. See, for example, Hsia et al., Nature 535, 136–139, 2016. Database searches reveal that I3-01 was engineered from a bacterial enzyme with a known crystal structure (PDB ID: 1VLW). The subunit sequences of these proteins are known in the art. See, for example, WO 2017 / 192434. More detailed information regarding the structural and functional properties of various nanoparticle scaffolds and their use in presenting trimeric protein immunogens is provided in the art.See, e.g., WO 2017 / 192434, WO 2019 / 089817, and WO 2019 / 241483. In various embodiments, the paramyxovirus vaccine compositions of the invention can employ any of these known nanoparticles, as well as conservatively modified variants or variants thereof having substantially identical (e.g., at least 90%, 95%, or 99% identical) sequences.

[0068] In addition to the nanoparticle sequences described above, many other nanoparticles or VLPs known in the art can be used in the practice of the present invention, including, for example, Aquifex aeolicus lumazine synthase, Thermotoga maritima encapsulin, Myxococcus xanthus encapsulin, bacteriophage Q beta virus particles, Flock House Virus (FHV) particles, ORSAY virus particles, and infectious bursal disease virus (IBDV) particles.

[0069] In addition to the displayed soluble F immunogen, the nanoparticle vaccine composition of the present invention can contain additional motifs for better biological or pharmaceutical properties. The additional structural components can function to facilitate immunogen display on the surface of the nanoparticle, enhance the stability of the displayed immunogen, and / or improve the yield and purity of the self-assembling protein vaccine. In these embodiments, one or more linkers (linker sequences, motifs, or moieties) can be used to connect the various structural components in the construct. One example of an additional structural component is a trimerization motif, such as foldon, as described above. In some embodiments, coding sequences for polypeptide fragments or motifs that serve as active sites for chemical conjugation can be inserted into appropriate locations in the construct. In some other embodiments, CD4 + T helper epitopes or CD8 +Additional structural components, such as T-cell epitopes, can also be inserted into the nanoparticle construct at appropriate locations, including, for example, the PADRE T-helper epitope as exemplified herein.

[0070] In yet some other embodiments, the nanoparticle vaccine of the present invention can include a locking domain that stabilizes the nanoparticle. The locking domain coding sequence can be fused directly or indirectly to the C-terminus of the nanoparticle subunit coding sequence. The locking domain stabilizes the nanoparticle from the inside, allowing nanoparticles displaying paramyxovirus immunogen polypeptides to remain intact during manufacturing, vaccine formulation, and immunization. Nanoparticle vaccine immunogens constructed in this manner have significantly enhanced stability. Generally, locking domains suitable for the present invention are protein subunits that can naturally form dimers with another protein subunit in solution through non-covalent interactions at an interface. In some preferred embodiments, the two protein subunits are identical in sequence and can form homodimers. In some other embodiments, the two protein subunits can be different proteins, or two distinct domains of a single engineered protein, that can form heterodimers in solution through non-covalent interactions at an interface. Typically, the locking domain is covalently fused to the nanoparticle subunit to which the immunogen polypeptide is linked. Examples of specific locking domains and guidance regarding the use of locking domains (e.g., LD7 or LD4 as exemplified herein) in the construction of nanoparticle-displayed trimeric immunogens can be found in the art, e.g., WO 2019 / 241483. Two specific locking domains, LD4 and LD7, suitable for use in the nanoparticle vaccines of the invention are exemplified herein.

[0071] Locking domain LD4 (SEQ ID NO: 29): FSEEQKKALDLAFYFDRRLTPEWRRYLSQRLGLNEEQIERWFRRKEQQIGWSHPQFEK Locking domain LD7 (SEQ ID NO: 30): SPAVDIGDRLDELEKALEALSAEDGHDDVGQRLESLLRRWNSRRAD Nanoparticles displaying any of the stabilized paramyxovirus soluble F protein immunogens described herein (e.g., stabilized RSV soluble F trimeric immunogens) can be constructed by fusing the immunogenic polypeptide or subunit of a multimeric immunogen protein (e.g., a trimeric immunogen) to a nanoparticle subunit sequence (e.g., the E2p I3-01v9b or I3-01v9c subunit sequence exemplified herein), as well as any other optional or alternative components described herein (e.g., a locking domain or trimerization motif). To construct the nanoparticle-displayed fusion vaccine immunogens of the invention, one or more linker motifs or moieties may be used to facilitate connection and maintain the structural integrity of the different components. Typically, linker motifs comprise short peptide sequences. In various embodiments, the linker or linker motif can be any flexible peptide that connects two protein domains or motifs without interfering with their function. For example, any of these linkers used in the constructs can be a (G a S b ) nwherein a is an integer of about 1 to 5, b is an integer of about 0 to 2, and n is an integer of about 1 to 5. In some embodiments, the linker used comprises the sequence GSGS (SEQ ID NO: 27) or GSGSGSGS (SEQ ID NO: 28). Detailed procedures for recombinant production of the vaccine compositions of the present invention can be based on the protocols described herein and / or other methods described in the art, such as He et al., Nat. Comm. 7, 12041, 2016; Kong et al., Nat. Comm. 7, 12040, 2016; He et al., Sci Adv. 4(11):eaau6769, 2018; WO 2017 / 192434; WO 2019 / 089817, and WO 2019 / 241483.

[0072] VII. Polynucleotides and Expression Constructs The stabilized paramyxovirus soluble F proteins and related vaccine compositions of the present invention are typically produced by first creating an expression construct (i.e., an expression vector) containing operably linked coding sequences for the various structural components described herein. Accordingly, in some related aspects, the present invention provides substantially purified polynucleotides (DNA or RNA) encoding nanoparticles displaying the immunogens described herein (e.g., stabilized RSV soluble F immunogens), as well as expression vectors (e.g., CMV vectors) carrying such polynucleotides and host cells for producing vaccine immunogens (e.g., the HEK293F and ExpiCHO cell lines exemplified herein). Fusion polypeptides encoded by the polynucleotides or expressed from the vectors are also included in the present invention. As described herein, such polypeptides self-assemble into nanoparticle vaccines that display the immunogenic polypeptide or protein on their surface.

[0073] Polynucleotides and associated vectors can be readily prepared using standard molecular biology techniques or the protocols exemplified herein. For example, general protocols for cloning, transfection, transient gene expression, and obtaining stable transfected cell lines are described in the art, see, e.g., Sambrook et al., Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Press, NY, (3 rd ed., 2000); and Brent et al., Current Protocols in Molecular Biology, John Wiley & Sons, Inc. (Ringbou ed., 2003). Introduction of mutations into polynucleotide sequences by PCR is described, for example, in PCR Technology: Principles and Applications for DNA Amplification, H.A. Erlich (Ed.), Freeman Press, NY, NY, 1992; PCR Protocols: A Guide to Methods and Applications, Innis et al. (Ed.), Academic Press, San Diego, CA, 1990; Mattila et al., Nucleic Acids Res. 19:967, 1991; and Eckert et al., PCR Methods and Applications 1:17, 1991.

[0074] The choice of a particular vector depends on the intended use of the fusion polypeptide. For example, the selected vector must be capable of driving expression of the fusion polypeptide in the desired cell type, regardless of whether that cell type is prokaryotic or eukaryotic. Many vectors contain sequences that enable both prokaryotic vector replication and eukaryotic expression of operably linked gene sequences. Vectors useful in the present invention can replicate autonomously, i.e., the vector exists extrachromosomally and its replication is not necessarily directly linked to the replication of the host cell's genome. Alternatively, vector replication can be linked to the replication of the host's chromosomal DNA; for example, the vector can be integrated into the host cell's chromosome, as achieved by retroviral vectors and stably transfected cell lines. Both viral and non-viral expression vectors can be used to produce immunogens in mammalian host cells. Non-viral vectors and systems include plasmids, episomal vectors that typically contain expression cassettes for protein or RNA expression, and human artificial chromosomes (e.g., Harrington et al., Nat. Genet. 15:345, 1997). Useful viral vectors include vectors based on lentiviruses or other retroviruses, adenoviruses, adeno-associated viruses, cytomegaloviruses, herpes viruses, SV40, papillomaviruses, HBP Epstein-Barr virus, vaccinia virus vectors, and Semliki Forest virus (SFV) vectors. See Brent et al., supra; Smith, Annu. Rev. Microbiol. 49:807, 1995; and Rosenfeld et al., Cell 68:143, 1992.

[0075] Depending on the specific vector used to express the fusion polypeptide, a variety of known cells or cell lines can be used in the practice of the present invention. A host cell can be any cell into which a recombinant vector carrying a fusion of the present invention can be introduced, such that the vector is capable of driving expression of the fusion polypeptide and is useful in the present invention. It can be prokaryotic, such as any of several bacterial strains, or eukaryotic, such as yeast or other fungal cells, insect or amphibian cells, or mammalian cells, including rodent, simian, or human cells. Cells expressing the fusion polypeptide of the present invention can be primary cells or established cell lines. Thus, in addition to the cell lines exemplified herein (e.g., CHO cells), numerous other host cell lines known in the art can also be used in the practice of the present invention. These include, for example, various Cos cell lines, HeLa cells, Sf9 cells, HEK293, AtT20, BV2, and N18 cells, myeloma cell lines, transformed B cells, and hybridomas.

[0076] The use of mammalian tissue cell cultures to express polypeptides is generally described, for example, in Winnacker, From Genes to Clones, VCH Publishers, NY, NY, 1987. Fusion polypeptide expression vectors can be introduced into selected host cells by any of several suitable methods known to those skilled in the art. The method used to introduce a vector encoding a fusion polypeptide into mammalian cells depends on the form of the vector. In the case of a plasmid vector, DNA encoding the fusion polypeptide sequence can be introduced by any of several transfection methods, including, for example, lipid-mediated transfection ("lipofection"), DEAE-dextran-mediated transfection, electroporation, or calcium phosphate precipitation. These methods are described in detail, for example, in Brent et al., supra. Lipofection reagents and methods suitable for transient transfection of a wide variety of transformed and non-transformed or primary cells are widely available, making lipofection an attractive method for introducing constructs into eukaryotic cells, particularly mammalian cells in culture. For example, LipofectAMINE™ (Life Technologies) or LipoTaxi™ (Stratagene) kits are available. Other companies that provide reagents and methods for lipofection include Bio-Rad Laboratories, CLONTECH, Glen Research, Life Technologies, JBL Scientific, MBI Fermentas, PanVera, Promega, Quantum Biotechnologies, Sigma-Aldrich, and Wako Chemicals USA.

[0077] Stable expression is preferred for long-term, high-yield production of recombinant fusion polypeptides. Rather than using expression vectors containing viral origins of replication, host cells can be transformed with a sequence encoding the fusion polypeptide under the control of appropriate expression control elements (e.g., promoter, enhancer, sequence, transcription terminator, polyadenylation site, etc.) and a selectable marker. The selectable marker in the recombinant vector confers resistance to selection and enables cells to stably integrate the vector into their chromosomes. Commonly used selectable markers include neo, which confers resistance to the aminoglycoside G-418 (Colberre-Garapin, et al., J. Mol. Biol., 150:1, 1981), and hygro, which confers resistance to hygromycin (Santerre et al., Gene, 30:147, 1984). With appropriate selection, transfected cells can contain integrated copies of the sequence encoding the fusion polypeptide.

[0078] VIII. Pharmaceutical Compositions and Therapeutic Uses In another aspect, the present invention provides pharmaceutical compositions and related methods of treatment using the redesigned paramyxovirus F immunogens and nanoparticle vaccine compositions described herein. In some embodiments, soluble F trimer immunogens against different viruses (e.g., hRSV) can be used to prevent and treat the corresponding viral infections. Some embodiments of the present invention relate to the use of hRSV soluble F-based vaccines to prevent or treat RSV infection in human subjects. Some embodiments of the present invention relate to the use of hMPV soluble F-based vaccines to prevent or treat MPV viral infections. Some embodiments of the present invention relate to the use of hPIV soluble F-based vaccines to prevent or treat PIV viral infections.

[0079] In practicing the various therapeutic methods of the present invention, a subject in need of prevention or treatment of a disease or condition (e.g., hRSV infection) is administered the corresponding nanoparticle vaccine, immunogenic protein or polypeptide, or encoding polynucleotide described herein. Typically, the nanoparticle vaccine, immunogenic protein, or encoding polynucleotide disclosed herein is contained in a pharmaceutical composition. The pharmaceutical composition can be either a therapeutic or prophylactic formulation. Typically, the composition can further comprise one or more pharmaceutically acceptable vehicles and may further comprise other therapeutic ingredients (e.g., antiviral agents). Various pharmaceutically acceptable additives can also be used in the composition.

[0080] Thus, some of the pharmaceutical compositions of the present invention are vaccine compositions. In the case of vaccine compositions, a suitable adjuvant can be further included. Examples of suitable adjuvants include, for example, aluminum hydroxide, lecithin, Freund's adjuvant, MPL™, and IL-12. In some embodiments, the vaccine compositions or nanoparticle immunogens (e.g., hRSV vaccine compositions) disclosed herein can be formulated as controlled-release or sustained-release formulations. This can be achieved in compositions containing slow-release polymers, or via microencapsulated delivery systems or bioadhesive gels. Various pharmaceutical compositions can be prepared according to standard procedures well known in the art. See, for example, Remington's Pharmaceutical Sciences, 1999. th Ed., Mack Publishing Company, Easton, Pa., 1995; Sustained and Controlled Release Drug Delivery Systems, J.R. Robinson, ed., Marcel Dekker, Inc., New York, 1978); U.S. Patent Nos. 4,652,441, 4,917,893; U.S. Patent Nos. 4,677,191 and 4,728,721; and U.S. Patent No. 4,675,189.

[0081] The pharmaceutical compositions of the present invention can be readily used for a variety of therapeutic or prophylactic purposes, for example, to treat hRSV infection or bronchiolitis in a subject, or to induce an immune response against hRSV. In various embodiments, the vaccine compositions can be used to treat or prevent infections caused by the pathogen from which the immunogenic polypeptide displayed in the nanoparticle vaccine is derived. Thus, the vaccine compositions of the present invention can be used in a variety of clinical settings to treat or prevent infections caused by various viruses. Illustratively, a RSV nanoparticle vaccine composition can be administered to a subject to induce an immune response against hRSV, e.g., to induce the production of broadly neutralizing antibodies against the virus. Subjects at risk of developing RSV infection can be administered the vaccine compositions of the present invention to provide prophylactic protection against viral infection. Therapeutic and prophylactic applications of vaccines derived from other immunogens described herein can also be performed. Depending on the specific subject and condition, the pharmaceutical compositions of the present invention can be administered to a subject by various modes of administration known to those skilled in the art, such as intramuscular, subcutaneous, intravenous, intraarterial, intraarticular, intraperitoneal, or parenteral routes. Generally, pharmaceutical compositions are administered to a subject in need of such treatment for a time and under conditions sufficient to prevent, inhibit, and / or ameliorate the selected disease or condition or one or more of its symptoms. For therapeutic use, the composition should contain a therapeutically effective amount of the nanoparticle immunogen described herein. For prophylactic use, the composition should contain a prophylactically effective amount of the nanoparticle immunogen described herein. The appropriate amount of immunogen can be determined based on the particular disease or condition being treated or prevented, the severity of the subject, the age, and other personal attributes of the particular subject (e.g., the general state of the subject's health and the robustness of the subject's immune system). Determination of effective dosages will be further guided by animal model studies followed by human clinical trials and by administration protocols that significantly reduce the occurrence or severity of the targeted disease symptoms or condition in the subject.

[0082] For prophylactic applications, the immunogenic composition is provided prior to any symptoms, e.g., prior to infection. Prophylactic administration of the immunogenic composition serves to prevent or ameliorate subsequent infection. Thus, in some embodiments, the subject being treated is one who has or is at risk of developing an infection (e.g., RSV infection), for example, due to exposure or potential exposure to a virus (e.g., RSV). After administering a therapeutically effective amount of the disclosed therapeutic composition, the subject can be monitored for infection (e.g., RSV infection), symptoms associated with infection (e.g., RSV infection), or both.

[0083] For therapeutic use, the immunogenic compositions are provided at or after the onset of disease or infection symptoms, for example, after the onset of infection symptoms (e.g., RSV infection) or after diagnosis of infection. Thus, the immunogenic compositions can be provided prior to anticipated exposure to the virus, after exposure or suspected exposure to the virus, or after the actual onset of infection to reduce the anticipated severity, duration, or extent of the infection and / or associated disease symptoms. The pharmaceutical compositions of the present invention can be combined with other agents known in the art for treating or preventing infection by the relevant pathogen (e.g., hRSV infection).

[0084] Nanoparticle vaccine compositions containing novel structural moieties (e.g., hRSV vaccines) according to the invention or pharmaceutical compositions of the invention can be provided as components of kits. Such kits may include additional components, including packaging, instructions, and various other reagents, such as buffers, substrates, antibodies or ligands, e.g., control antibodies or ligands, and detection reagents. Optional instructions can additionally be provided with the kit.

[0085] [Example] The following examples are offered to illustrate, but not to limit, the present invention.

[0086] [Example 1] Comparative analysis of existing RSV Prefusion F designs In this study, we compared three known RSV prefusion F designs: DS-Cav1 (McLellan et al., Science 2013, 342:592-598), SC-TM (Krarup et al., Nat Comm 2015, 6:8143), and sc9-10 DS-Cav1 (Joyce et al., Nat Struct Mol Biol 2016, 23:811-820). For these three prefusion F designs, constructs were generated with an enzyme site (amino acid "AS") and a foldon motif attached to the C-terminus. The sequences are shown below.

[0087] MELLILKANAITTILTAVTFCFASG (SEQ ID NO: 2): N-terminal leader.

[0088] QSTPPTNNRARR (SEQ ID NO: 3): Unstructured F2 C-terminus.

[0089] QSTPATNNQAR (SEQ ID NO: 4): F2 C-terminus with mutations.

[0090] ELPRFMNYTLNNAKKTNVTLSKKRKRR (SEQ ID NO: 5): P27 peptide.

[0091] FLGFLLGVGS (SEQ ID NO: 6): fusion peptide (FP).

[0092] GYIPEAPRDGQAYVRKDGEWVLLSTFL (SEQ ID NO: 7): C-terminal foldon.

[0093] A2_DS-Cav1-foldon (PDB ID: 4MMU) (SEQ ID NO: 8) [ka]

[0094] A2_SC-TM-foldon (PDB ID: 5C6B) (SEQ ID NO: 9) [ka]

[0095] A2_sc9-10 DS-Cav1-foldon (PDB ID: 5K6I) (SEQ ID NO: 10) [ka]

[0096] These three constructs were transiently expressed in 25 ml ExpiCHO cells and purified using a D25 antibody column followed by size-exclusion chromatography (SEC) on a Superdex 200 Increase 100 / 300 GL column. The use of D25, a potent neutralizing antibody (NAb) targeting site φ (McLellan et al., Science 2013, 342:592-598), ensures prefusion-specific RSV F purification. In the case of DS-Cav1, the 3.0 Å resolution crystal structure showed a football-shaped, closed prefusion F trimer conformation. Transient expression of DS-Cav1 in ExpiCHO cells resulted in reasonable yields after D25 purification. However, the SEC profile showed a high aggregation peak (at approximately 9 ml) and a second peak corresponding primarily to trimers. Notably, we observed a visible shoulder to the left of the trimer peak in SEC, suggesting the presence of higher-order prefusion F species in the DS-Cav1 sample. To characterize the trimer fraction (approximately 12 ml), we performed negative-stain electron microscopy (nsEM). All 2D classes showed monomers or dimers with no sign of a closed prefusion trimer. In the case of SC-TM, the 2.4 Å resolution crystal structure showed a closed prefusion trimer similar to DS-Cav1. Notably, SC-TM had extremely low yields in ExpiCHO expression and showed both a trimer peak and monomer leakage in SEC. The 2D class images from nsEM showed football-shaped molecules characteristic of a closed prefusion F trimer, as well as monomers and dimers. We used these 2D classes to construct a 3D EM model, which matched the crystal structure almost perfectly and showed unoccupied density at the base of the trimer corresponding to the C-terminal foldon. However, we also found wedge-shaped molecules in the nsEM images corresponding to the post-fusion F trimer, suggesting that the SC-TM cannot prevent the anterior-to-posterior conformational transition.

[0097] The 2.9 Å resolution crystal structure of sc9-10 DS-Cav1 revealed a closed prefusion F trimer nearly identical to that of DS-Cav1 and SC-TM. The sc9-10 DS-Cav1 construct produced high yields and purity in SEC, with trimer peaks over 10- and 250-fold higher than those of DS-Cav1 and SC-TM, respectively. nsEM analysis revealed that almost all 2D classes exhibited a "sucker" shape, suggesting nearly 100% closed prefusion F trimer. 3D structural models constructed from the EM data confirmed this finding. In summary, DS-Cav1 is monomeric in solution, SC-TM expresses prefusion monomers and closed trimers, as well as postfusion trimers, with modest yields, while sc9-10 DS-Cav1 produces closed prefusion trimers with high yield and purity.

[0098] [Example 2] Analysis of the causes of RSV F instability We analyzed the sequence and structure of RSV F to identify potential causes of F instability. The RSV strain A2 (GenBank ID: AAB59858.1, UniProt ID: P03420), previously used to design prefusion F constructs, was used as a template for this study. Briefly, sequence and secondary structure alignment of A2 F in the prefusion and postfusion states revealed several key regions (FIG. 1A). Two such regions are β3 / β4 and β23, both of which undergo secondary structure changes to become α-helices in the postfusion conformation.

[0099] In the case of the β3 / β4 segment (K176-S190), it appears as a β-hairpin in the prefusion state but becomes part of an extended α-helix in the postfusion state (FIG. 1B). Structural analysis suggested that the disulfide bonds S180C / S186C and A177C / T189C, with Cβ-Cβ distances of 4.25 Å and 4.91 Å, respectively, may stabilize RSV F in the prefusion state (this was covered in our previous paramyxovirus patent application). Here, we report another mutation, V185P. In the non-mutated prefusion F (PDB ID: 4JHW), the main chain dihedral angles of V185 are -70.6 (Phi) and 126.4 (Psi), which closely match those of trans-proline: -75 (Phi) and 145 (Psi). We hypothesize that V185P can rigidify the prefusion hairpin structure but introduce a kink into the postfusion helix, thus destabilizing the postfusion conformation (Fig. 1B, right). In the case of the β23 segment (S485-A490), it forms interactions with the β23 segments of the other two promoters around the three-fold symmetry axis, and is located on top of the trimeric coiled coil formed by the three α10 helices, holding the three F promoters in a trimeric conformation (Fig. 1C, left). Cross-sectional analysis revealed that the β23 cluster is located at the bottom of the hollow interior of the football-shaped RSV prefusion F trimer (Fig. 1C, top right). In the wild-type RSV prefusion F trimer, this hollow interior is partially filled with the fusion peptides of the three F promoters, but becomes empty when the fusion peptides are removed, such as in the prefusion F design sc9-10 DS-Cav1. Further analysis of the β23 cluster revealed an unusual pattern of interactions (Fig. 1C, bottom right). More specifically, the short β23 chain contains three negatively charged residues, D486, E487, and D489, which form repulsive charge-charge interactions around the trimer's three-fold symmetry axis.We hypothesize that unfavorable interactions in the β23 cluster promote the rapid opening of wild-type prefusion F trimers on the surface of RSV virus particles, exposing the fusion peptide and accelerating the conformational changes occurring before and after cell entry. In other words, we hypothesize that the β23 cluster is the primary cause of RSV F instability. In this study, we tested our hypothesis by mutating D486 and E487 to (1) polar residues that can form salt bridges, such as D486N and E487Q, and (2) hydrophobic residues that can form hydrophobic clusters, such as D486L and E487L. Other mutations to the β23 segment may further improve trimer stability.

[0100] [Example 3] Characterization of RSV prefusion F constructs based on "V2-Ext-PDB6-D" We used "V2-Ext-PDB6-D" as the base design to generate five soluble F constructs, all of which contained a C-terminal foldon motif (sequences listed below). The first construct is the base design. The second construct incorporates the V185P mutation into the base design to examine the effect of the second proline mutation V185P. The third and fourth constructs incorporate the D486N / E487Q and D486L / E487L mutations, respectively, into the second construct to examine whether removing the repulsive charge-charge interaction at β23 can improve the stability of the RSV prefusion F trimer. The fifth construct incorporates the A149C / Y458C mutation (referred to as SS4) into the base design to test this interprotomer disulfide bond in combination with a minimal set of mutations in the base design. This disulfide bond was used in sc9-10 DS-Cav1 (Joyce et al., Nat Struct Mol Biol 2016, 23: 811-820). Thus, the fifth construct, which uses a covalent bond to lock RSV prefusion F into a closed trimer, provides a "positive control" for the third and fourth constructs, which attempt to hold RSV prefusion F in a closed trimer through engineered non-covalent interactions at β23.

[0101] A2_V2-Ext-PDB6-D-foldon (S215P, DB6=S155-S290, E92D) (SEQ ID NO: 11) [ka]

[0102] GA2_V2-Ext-P2DB6-D-foldon (S215P, DB6=S155-S290, E92D, V185P) (SEQ ID NO: 12) [ka]

[0103] A2_V2-Ext-P2DB6-D-NQ, without N-terminal leader and C-terminal foldon (S215P, DB6 = S155-S290, E92D, V185P, D486N + E487Q) (SEQ ID NO: 17) [ka]

[0104] A2_V2-Ext-P2DB6-D-L2, without N-terminal leader and C-terminal foldon (S215P, DB6 = S155-S290, E92D, V185P, D486L + E487L) (SEQ ID NO: 18) [ka]

[0105] A2_V2-Ext-PDB6-D-SS4-foldon (S215P, DB6=S155-S290, E92D, SS4=A149-Y458) (SEQ ID NO: 13) [ka]

[0106] Five soluble F constructs from the "V2-Ext-PDB6-D" series were transiently expressed in 25 ml ExpiCHO cells and purified using SEC on a D25 column followed by a Superdex 200 Increase 100 / 300 GL column. For the first construct or base design, A2_V2-Ext-PDB6-D-foldon, we observed a single trimer peak in SEC with high yield and purity. However, nsEM analysis of SEC fractions around 12 ml revealed predominantly "open" prefusion F trimers, with two classes exhibiting closed-conformation prefusion trimers. 3D EM models constructed from these two classes closely matched the prefusion F crystal structure. Thus, our results demonstrate that this base design with a minimal set of mutations can produce prefusion F with a low proportion of closed-conformation trimers. For the second construct, A2_V2-Ext-P2DB6-D-foldon, we observed a profile similar to that of the basic design, suggesting that the second proline mutation, V185P, may have little effect on protein properties. For the third construct, A2_V2-Ext-P2DB6-D-NQ-foldon, we observed distinct characteristics compared to the second construct without the D486N / E487Q mutations. More specifically, A2_V2-Ext-P2DB6-D-NQ-foldon yielded a single trimer peak with high yield and purity, but with a slightly increased aggregate peak around 8.5 mL. Notably, nsEM analysis revealed that the majority of the 2D classes, i.e., approximately 73% of the molecules, corresponded to the closed-structure prefusion F trimer. A 3D EM model constructed from the EM data matches the PreFusion F crystal structure with near-perfect fitting, with some deformation around the α10 helix at the bottom of the trimer. For the fourth construct, A2_V2-Ext-P2DB6-D-L2-foldon, we observed overall similar properties to the third construct, with a slightly higher proportion of closed PreFusion F trimers (76% vs. 73%).For the fifth construct, A2_V2-Ext-PDB6-D-SS4-foldon, we observed a trimer peak with high yield and purity, but further increased aggregation compared to constructs 3 and 4. nsEM analysis revealed the highest percentage of closed-conformation prefusion F trimers (88%) among all constructs in this series. The 3D EM model not only matches the crystal structure of the prefusion F trimer, but also shows the density of the C-terminal foldon domain. In summary, our results from constructs 3 and 4 support the hypothesis that the β23 chain is the primary cause of RSV F instability and that elimination of repulsive charge-charge interactions at β23 significantly improves trimer stability at a level similar to that of well-spaced interprotomer disulfide bonds. All five constructs, especially constructs 3, 4, and 5, which have 73-88% closed-conformation trimers, can be developed into RSV F trimer vaccines. The presence of the open trimer is not a major concern, as wild-type prefusion F must be in equilibrium between "closed" and "open" trimers on the surface of the RSV virion, and both states can elicit neutralizing antibodies and block virus entry.

[0107] The third and fourth constructs, both containing engineered non-covalent mutations in β23, were further investigated using known RSV antibodies. First, these two constructs were probed using the postfusion-specific antibody ADI14359. nsEM analysis revealed three types of 2D class images corresponding to unbound ADI14359 Fab, open-conformation prefusion F trimer or F monomer without bound Fab, and closed-conformation prefusion F trimer without bound Fab. In this analysis, we did not find any 2D classes corresponding to postfusion F or the postfusion F / ADI14359 complex. A complete 3D model of the closed-conformation prefusion F trimer was constructed, suggesting that ADI14359 does not perturb these two prefusion F constructs or induce any conformational changes in them. These two constructs were then probed with the prefusion F-specific antibody D25, which was also used to purify the prefusion F protein in this study. nsEM analysis revealed three types of 2D class images corresponding to unbound D25 Fab, open-conformation prefusion F trimer or F monomer bound to D25, and closed-conformation prefusion F trimer bound to D25. A nearly complete 3D structural model of the closed-conformation prefusion F trimer complexed with D25 was constructed from the EM data, providing strong evidence that the A2_V2-Ext-P2DB6-D-NQ-foldon and A2_V2-Ext-P2DB6-D-L2-foldon are ideal candidates for the development of RSV closed-conformation prefusion F trimer-based vaccines.

[0108] [Example 4] Characterization of RSV prefusion F constructs based on the "V2-Ext-PDB6-GDQ" base We generated eight soluble F constructs using "V2-Ext-PDB6-GDQ" as the base design, all of which contained a C-terminal foldon motif (sequences listed below). The first construct is a base design combining V2-Ext-PDB6-D with S46G and K465Q. We hypothesize that the S46G / K465Q mutations can reduce aggregation in some "V2-Ext-PDB6-D" derivatives. The second construct incorporates a second proline mutation, V185P, into the base design. The third and fourth constructs incorporate the D486N / E487Q mutations into the base design, while the fourth construct contains the V185P mutation. Constructs 5 and 6 incorporate the D486L / E487L mutations into the basic design, while construct 6 contains the V185P mutation. Construct 7 incorporates the A149C / Y458C mutation (referred to as SS4) into the basic design. This interprotomer disulfide bond was used in sc9-10 DS-Cav1 (Joyce et al., Nat Struct Mol Biol 2016, 23:811-820). Thus, construct 7, which uses covalent bonds to lock RSV prefusion F into a closed trimer, serves as a "positive control" for constructs 3 through 6, which attempt to maintain RSV prefusion F in a closed trimer through engineered non-covalent interactions at β23. The eighth construct was designed to examine whether the interprotomer disulfide bond, A149C / Y458C, and polar mutations in β23, D486N / E487Q, could be combined into one construct to further stabilize the prefusion F trimer.

[0109] A2_V2-Ext-PDB6-GDQ-foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q) (SEQ ID NO: 14) [ka]

[0110] A2_V2-Ext-PDB6-GDQ-foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q, V185P) (SEQ ID NO: 15) [ka]

[0111] A2_V2-Ext-PDB6-GDQ-NQ, without N-terminal leader and C-terminal foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q, D486N+E487Q) (SEQ ID NO: 19) [ka]

[0112] A2_V2-Ext-P2DB6-GDQ-NQ, without N-terminal leader and C-terminal foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q, V185P, D486N+E487Q) (SEQ ID NO: 20) [ka]

[0113] A2_V2-Ext-PDB6-GDQ-L2, without N-terminal leader and C-terminal foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q, D486L+E487L) (SEQ ID NO: 21) [ka]

[0114] A2_V2-Ext-P2DB6-GDQ-L2, without N-terminal leader and C-terminal foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q, V185P, D486L+E487L) (SEQ ID NO: 22) [ka]

[0115] A2_V2-Ext-PDB6-GDQ-SS4-foldon (S215P, DB6 = S155-S290, S46G + E92D + K465Q, SS4 = A149-Y458) (SEQ ID NO: 16) [ka]

[0116] A2_V2-Ext-P2DB6-GDQ-SS4-NQ, without N-terminal leader and C-terminal foldon (S215P, DB6=S155-S290, S46G+E92D+K465Q, V185P, D486N+E487Q, SS4=A149-Y458) (SEQ ID NO: 23) [ka]

[0117] Seven soluble F constructs from the "V2-Ext-PDB6-GDQ" series were transiently expressed in 25 ml ExpiCHO cells and purified using SEC on a D25 column followed by a Superdex 200 Increase 100 / 300 GL column. For the first construct or basic design A2_V2-Ext-PDB6-GDQ-foldon, we observed a trimer peak with high yield and purity, but also a monomer peak in SEC. nsEM analysis of approximately 11.5 ml of SEC fractions showed a class of "open" prefusion F corresponding to the "closed" trimer. Thus, our results revealed that the mutations S46G and K465Q (presumably S46G) tend to shift the equilibrium toward the "open" conformation while maintaining RSV F in the prefusion state. For the second construct, A2_V2-Ext-P2DB6-GDQ-foldon, we observed a profile similar to that of the basic design. Of note, the slightly anomalous trimer peak may be caused by the high yield of this construct; a similar pattern has been found elsewhere. For the third and fourth constructs, A2_V2-Ext-PDB6-GDQ-NQ-foldon and A2_V2-Ext-P2DB6-GDQ-NQ-foldon, we observed substantial trimer yield and purity without any aggregate peaks in the SEC profiles. nsEM analysis revealed that most of the trimers were in the open conformation, with 12% and 6% closed conformation trimers observed for the third and fourth constructs, respectively. 3D structural models constructed from the EM data further confirmed that these two constructs can form closed conformation prefusion F trimers.

[0118] For the fifth and sixth constructs, A2_V2-Ext-PDB6-GDQ-L2-foldon and A2_V2-Ext-P2DB6-GDQ-L2-foldon, we observed high trimer yield and purity without any aggregate peaks in the SEC profiles. nsEM analysis revealed that the majority of trimers were open, with approximately 29% closed trimers observed in both constructs. 3D structural models confirmed that these two constructs could form closed Prefusion F trimers. For the seventh construct, incorporation of an interprotomer disulfide bond into the "V2-Ext-PDB6-GDQ" base resulted in a slight increase in aggregation in the SEC profiles. nsEM analysis indicated that 76% of the molecules were closed Prefusion F trimers, which was >10% less than the 88% observed when this disulfide bond was incorporated into the V2-Ext-PDB6-D base. Nevertheless, a nearly complete 3D structural model was constructed from the EM data. In the eighth construct, incorporating both the interprotomer disulfide bond SS4 and the noncovalent β23 mutation into the "V2-Ext-PDB6-GDQ" base further increased aggregation, but nsEM analysis did not increase the proportion of closed-conformation prefusion trimers, with 73% of the molecules corresponding to closed-conformation prefusion trimers. A nearly complete 3D structural model was constructed from the EM data. In summary, our results from this systematic comparison supported the hypothesis that the S46G / K465 mutation can minimize aggregation during prefusion F expression. However, we noted a detrimental effect associated with the S46G / K465 mutation, namely a decrease in the proportion of "closed" prefusion F trimers. Furthermore, constructs containing the D486N / E487Q (or "NQ") mutations appear to be more sensitive to the S46G / K465 mutation than constructs containing the D486L / E487L (or "L2") mutations, suggesting that in β23, salt bridges formed by polar residues are less effective than hydrophobic contacts in holding prefusion F in the closed trimer.

[0119] [Example 5] Characterization of various RSV prefusion F constructs by X-ray crystallography Eleven F constructs were structurally characterized by X-ray crystallography. First, the crystal structures of six "V2-Ext-PDB6-D" derivatives were determined. The structure of the "V2-Ext-PDB6-D" base design was determined using 1TD0 with two different C-terminal domains, foldon and a 5GS linker. In both cases, the base design appears as a complete, closed prefusion F trimer, but is mostly open in solution. Crystal structures were then obtained for constructs containing the D486L / E487L ("L2") and D486N / E487Q ("NQ") mutations, confirming that the engineered noncovalent interactions at β23 can indeed stabilize the prefusion F trimer as predicted. Finally, the crystal structures also confirmed that well-placed interprotomer disulfide bonds can effectively lock prefusion F into a closed trimeric conformation. Second, we determined the crystal structures of three "V2-Ext-PDB6-GDQ" derivatives. We focused on constructs containing the D486N / E487Q (NQ) mutation, an interproton disulfide bond, and both. The crystal structures confirmed that either mutations or a combination of both can be used to stabilize the RSV prefusion F trimer. However, no crystal structure was obtained for the "V2-Ext-PDB6-GDQ" base. Third, we determined the crystal structures of three constructs containing the disulfide bond A177C / T189C in the β3 / β4 hairpin. Our crystal structures confirmed that this disulfide bond can be combined with the minimal "V2-Ext-PDB6" base to stabilize the RSV prefusion F trimer.

[0120] [Example 6] Optimization of I3-01v9 nanoparticles for displaying thin-stalk trimeric glycoproteins Previously, we rationally redesigned the I3-01v9 nanoparticle scaffold to optimize the display of monomeric antigens. In I3-01v9a, the N-terminal helix was extended so that its first amino acid was directly above the nanoparticle surface (Figure 2, A). Based on I3-01v9a, we further redesigned the N-terminal helix to achieve optimal display of trimeric antigens, such as the RSV prefusion F trimer (Figure 2, B). First, the 11-aa N-terminal helix of I3-01v9a was truncated to 7 aa. Next, a 13-aa helix-turn fragment (all alanine) was fused to the 7-aa helix of I3-01v9a so that the new N-terminal helix fit within the groove of the two helices that are part of the I3-01 core. Next, several mutations were made to the I3-01 core helix to eliminate steric clashes between the new N-terminal helix and the groove. The helix-turn backbone was relaxed using a computational program called IMO (Zhu et al., Proteins 2006, 65(2):463-79), which was then subjected to the protein structure sampling program CONCOORD to generate 1000 slightly perturbed backbone conformations. An ensemble-based protein design program previously used to optimize HIV gp140 and HCV E2 cores was then used to predict the first 9 aa of the 13-aa fragment using a Cα- and Cβ-based RAPDF scoring function (the 4-aa turn was used as a "GSGS" (SEQ ID NO: 27) linker).

[0121] We selected the final design, I3-01v9b, by combining data from the predictions, and then performed a second design, I3-01v9c, by suddenly changing the flexible turn from "GSGS" (SEQ ID NO: 27) to "GPPS" (SEQ ID NO: 32) to increase its rigidity. After further backbone relaxation, we constructed a structural model of I3-01v9b. The N-terminus of I3-01v9b forms a 12.9 Å triangle, suitable for displaying a trimeric antigen. In a recent study, we used the stabilized Ebola virus (EBOV) GP trimer, GPΔmuc-WL, as a test case.2 P 4 The I3-01v9b / c design was validated using the EBOV GPΔmuc-WL vector (Figure 2C). 2 P 4 The I3-01v9b fusion construct was transiently expressed in HEK293F cells and purified by mAb100 antibody column followed by SEC. nsEM analysis identified a 2D cluster corresponding to the fully formed GP-I3-01v9b fusion protein (Figure 2C, top). A 3D model was constructed from the EM data, showing the complete EBOV GP trimer with a thin stalk displayed on the I3-01v9b trimer (Figure 2C, bottom). The subunit sequences of the I3-01v9a, I3-01v9b, and I3-01v9c nanoparticle scaffolds are shown in SEQ ID NOs: 24–26, respectively. For each of these sequences, an enzyme site, AS, can be added to the N-terminus for fusion with the displayed antigen. A GGGGS (SEQ ID NO: 33) linker can also be inserted after the enzyme site of I3-01v9a.

[0122] I3-01v9a (for monomeric antigen display) (SEQ ID NO: 24) [ka]

[0123] I3-01v9b (for trimeric antigen display; first residue mutated to G) (SEQ ID NO: 25) [ka]

[0124] I3-01v9c (for trimeric antigen display; first residue was mutated to G) (SEQ ID NO: 26) [ka]

[0125] [Example 7] Design and negative stain EM analysis of nanoparticles displaying RSV prefusion F trimers In recent studies, RSV prefusion F (DS-Cav1) was displayed on a two-component nanoparticle platform (Marcandalli et al., Cell 2019, 176(6):1420-1431.e17) and a ferritin 24-mer (Swanson et al., Sci Immunol 2020, 5(47):eaba6466). However, our EM analysis showed that DS-Cav1 tends to be monomeric in solution, which may not be suitable for nanoparticle display. In our study, we investigated the possibility of displaying the known RSV prefusion F on various nanoparticle platforms and then tested a highly optimized prefusion F trimer design for nanoparticle display. We first performed computational modeling to design the RSV prefusion F trimer-displaying 1c-SApNP (Figure 3A). When the RSV prefusion F trimer (PDB ID: 4JHW) was directly fitted to the ferritin (FR) 24-mer, the Cα-RMSD was 11.2 Å, producing 34 nm F-FR nanoparticles. Based on this calculation, a 5GS linker was added to all RSV F-FR constructs between the C-terminus of RSV F and the N-terminal residue D5 of the FR. As a result, the diameter of the F-5GS-FR reached approximately 38 nm. When the RSV prefusion F trimer (PDB ID: 4JHW) was directly fitted to the E2p and I3-01v9b 60-mer, large nanoparticles of 45.4 nm and 47.5 nm were obtained, respectively. As shown in our previous study, such large trimer-presenting 1c-SApNP vaccines can induce stronger and more durable immune responses than small individual antigens.

[0126] We next tested whether DS-Cav1 prefusion F could be displayed on three 1c-SApNP platforms (Figure 3B). The DS-Cav1-5GS-FR, E2p-LD4-PADRE, and I3-01v9b-LD7-PADRE constructs were transiently expressed in ExpiCHO cells and purified using a D25 antibody column. nsEM analysis showed FR nanoparticles with irregular display of F protein mixed with "naked" FR nanoparticles, suggesting protein misfolding (Figure 3B, left). The DS-Cav1-E2p-LD4-PADRE sample appeared to contain only aggregates and debris in EM images (Figure 3B, center), while the DS-Cav1-I3-01v9b-LD7-PADRE construct showed extremely low yields, with no nanoparticles observed in EM images (Figure 3B, right). Due to extremely low yields, the SC-TM 1c-SApNP fusion construct was unable to produce sufficient sample for nsEM analysis.

[0127] Next, we tested whether sc9-10 DS-Cav1 prefusion F could be displayed on three 1c-SApNP platforms (Figure 3C). Due to the presence of interprotomer disulfide bonds, the sc9-10 DS-Cav1 1c-SApNP fusion constructs showed much lower yields than their DS-Cav1 counterparts. Nevertheless, nsEM analysis showed FR nanoparticles displaying closed-structure prefusion F trimers partially mixed with "naked" FR nanoparticles, suggesting that interprotomer disulfide bonds may not form on the nanoparticle surface (Figure 3C, left). Both the E2p and I3-01v9b constructs showed extremely low yields (Figure 3C, center and right), with some partially formed particles mixed with aggregates observed for I3-01v9b (Figure 3C, right).

[0128] After testing previously reported RSV prefusion F designs, we sought to test our "V2-Ext-P2DB6-D-L2 / NQ" design against FR and I3-01v9b 1c-SApNP (Fig. 3D). Notably, EM images showed FR nanoparticles with closed-structure prefusion F trimers for both the "L2" and "NQ" constructs (Fig. 3D, left and center). We observed well-formed, large I3-01v9b nanoparticles with uniformly distributed prefusion F trimers on the surface (Fig. 3D, right). Finally, we tested our "V2-Ext-P2DB6-GDQ-L2 / NQ" design against FR and I3-01v9b 1c-SApNP (Fig. 3E). Similar success was achieved in terms of expression yield and structural integrity. Notably, a well-formed layer of prefusion F trimers can be seen on the surface of I3-01v9 1c-SApNPs ( Figure 3, E , right).

[0129] [Example 8] Additional redesigned RSV prefusion F trimers This example describes additional redesigned RSV prefusion F trimers with different minimal mutation sets to effectively stabilize the prefusion F trimer.

[0130] The sequence of the F protein of human respiratory syncytial virus A (strain A2) was obtained from GenBank under ID (P03420). Numbering is based on the UniProt definition under ID (P03420). Soluble F (also referred to as Fd) here is defined as M1-L513, with M1-G25 being the signal peptide (see SEQ ID NO: 1 or A2N-WT). The non-cleavable version of soluble F was derived from A2-WT by shortening and mutating the C-terminus of unstructured F2 (residues Q98-R109), removing the 27-residue "processed activation peptide" or P27 (residues E110-R136), removing the N-terminus (residues F137-S146) of the fusion peptide (F137-V157), and adding a 4- to 8-residue GS linker. Two mutations, I379V and M447V, were added to improve F protein expression. This F construct design is considered the "base design" (SEQ ID NO: 34 and SEQ ID NO: 35, or A2N-JZ0-V2-Ext and A2N-JZ0-V2-Ext2).

[0131] The uncleavable, prefusion-optimized (UFO) soluble F construct is derived from a "base design" with specific mutations engineered into the "β3 / β4 hairpin" (residues K176-S190) with the hypothesis that this region is the underlying cause of RSV F instability and undergoes the greatest conformational changes during the membrane fusion process. Two types of mutations can be introduced into the β3 / β4 hairpin:

[0132] (i) Disulfide bonds between β-sheet-forming amino acids: Two disulfide bond mutations (S180C / S186C and A177C / T189C) resulted in the shortest C s of 4.25 Å and 4.91 Å, respectively. β -C β These have been experimentally tested to provide the so-called "SS" and "AT" designs.

[0133] (ii) Mutations between the two β-strands: One double mutation, S182G / N183P, has been experimentally tested because this mutation effectively destroys the helix-forming propensity at the turn between β3 and β4, resulting in the so-called “GP” design.

[0134] A total of eight combinations (Ext vs. Ext 2 + SS vs. AT + no GP vs. GP) were tested: SS-based designs: A2N-JZ0-V2-Ext-SS (SEQ ID NO: 36), A2N-JZ0-V2-Ext-SSGP (SEQ ID NO: 37), A2N-JZ0-V2-Ext2-SS (SEQ ID NO: 38), A2N-JZ0-V2-Ext2-SSGP (SEQ ID NO: 39) AT-based designs: A2N-JZ0-V2-Ext-AT (SEQ ID NO: 40), A2N-JZ0-V2-Ext-ATGP (SEQ ID NO: 41), A2N-JZ0-V2-Ext2-AT (SEQ ID NO: 42), A2N-JZ0-V2-Ext2-ATGP (SEQ ID NO: 43) To stabilize the trimer and increase the trimer ratio within the total protein yield, a trimerization motif such as foldon and the viral capsid protein SHP (PDB:1TD0) can be added to the C-terminus of the redesigned F construct with a short GS linker in the middle. A His6 tag can be added to the C-terminus of the trimerization motif to facilitate protein purification by nickel column.

[0135] The C-terminus of the redesigned F construct can be fused to the N-terminus of a nanoparticle-forming subunit, such that the fusion construct, when expressed in an appropriate cell line, can self-assemble into nanoparticles with prefusion F trimers displayed on the nanoparticle surface.

[0136] RSV F construct sequence (based on A2 strain wild-type sequence): MELLILKANAITTILTAVTFCFASG (SEQ ID NO: 2): leader QSTPPTNNRARR (SEQ ID NO: 3): unstructured F2 Ctm ELPRFMNYTLNNAKKTNVTLSKKRKRRFLGFLLGVGSAIASGVAVSKV (SEQ ID NO: 52): P27 peptide + fusion peptide (FP) KAVVSLSNGVSVLTS (SEQ ID NO: 53): β3 / β4 hairpin region.

[0137] SEQ ID NO: 34 (A2N-JZ0-V2-Ext) [ka]

[0138] SEQ ID NO: 35 (A2N-JZ0-V2-Ext2) [ka]

[0139] SEQ ID NO: 36 (A2N-JZ0-V2-Ext-SS) [ka]

[0140] SEQ ID NO: 37 (A2N-JZ0-V2-Ext-SSGP) [ka]

[0141] SEQ ID NO: 38 (A2N-JZ0-V2-Ext2-SS) [ka]

[0142] SEQ ID NO: 39 (A2N-JZ0-V2-Ext2-SSGP) [ka]

[0143] SEQ ID NO: 40 (A2N-JZ0-V2-Ext-AT) [ka]

[0144] SEQ ID NO: 41 (A2N-JZ0-V2-Ext-ATGP) [ka]

[0145] SEQ ID NO: 42 (A2N-JZ0-V2-Ext2-AT) [ka]

[0146] SEQ ID NO: 43 (A2N-JZ0-V2-Ext2-ATGP) [ka]

[0147] [Example 9] Expression and purification of vaccine antigens This example describes the expression and purification of the redesigned prefusion F trimeric vaccine antigen as described in Example 8.

[0148] Cell lines: All F trimers were expressed in HEK293F / Expi293F and ExpiCHO cells, with ExpiCHO showing greater protein yields. All F-displaying nanoparticles were expressed in ExpiCHO cells.

[0149] Purification: After transient expression, RSV F-containing antigens were purified from the supernatant using an antigen-specific antibody column based on (1) the prefusion site-specific neutralizing antibody D25 and (2) the neutralizing antibody MPE8, which recognizes two promoters of the F trimer. MPE8 binds to both prefusion and postfusion F trimers but supports the prefusion structure. Both the D25 and MPE8 antibody columns can effectively purify RSV F trimers and nanoparticles, with D25 showing a higher protein yield. For hMPV and PIV1-5 F trimers, a nickel column is used for His-tag purification. An antibody column is being developed for untagged trimer / nanoparticle purification.

[0150] [Example 10] Study of paramyxovirus F instability using known prefusion F trimers This example describes the characterization of prefusion F trimers reported in the literature.

[0151] We set out to investigate the underlying causes of paramyxovirus F instability and develop a simple, general, and effective prefusion F trimer stabilization strategy that can be applied to all members of the Paramyxoviridae family and enables multivalent display of stabilized F trimers on self-assembling nanoparticles as virus-like particle (VLP) vaccines. RSV was chosen as the focus of this study because it is the most studied in structure-based vaccine design, with three representative prefusion F designs available. We also included hMPV and PIV3 to validate our F stabilization strategy. We first characterized three RSV F designs reported in the literature: DS-Cav1 (McLellan et al., Science 342:592-8, 2013), sc9-10 DS-Cav1 (Joyce et al., Nat Struct Mol Biol 23:811-820, 2016), and SC-TM (Krarup et al., Nat Commun 6:8143, 2015), and compared their expression, trimerization, and purification.

[0152] For expression, the effect of cell line on the yield and purity of RSV F trimers was examined using the transient mammalian cell line HEK293F and ExpiCHO, a transient high-yield version of the industrial CHO-S cell line. For purification, two immunoaffinity columns based on D25, which recognizes the prefusion site Φ, and MPE8, which binds to two adjacent F subunits, were evaluated for their ability as antibody columns for tag-free purification. For trimer stability, the effect of the C-terminal trimerization motif was examined using a soluble F construct without a C-terminal motif and two F constructs with foldon and 1TD0 motifs at their C-termini.

[0153] Three representative RSV prefusion F constructs, DS-Cav1, sc9-10 DS-Cav1, and SC-TM, were transiently expressed in HEK 293F cells and ExpiCHO cells. Our previous study (He et al., Sci Adv 4(11):eaau6769, 2018) showed that HIV-1 gp140 trimers could be expressed in ExpiCHO cells with significantly higher yields and purity than in 293F cells. Based on this finding, a small volume of 25 ml of ExpiCHO cells was used for transfection, followed by purification using D25 and MPE8 antibody columns, while 400 ml and 100 ml of 293F cells were used for transfection, followed by D25 and MPE8 purification, respectively. Because D25 recognizes the Φ site of prefusion F regardless of whether F is in a monomeric or multimeric state (e.g., dimer, trimer, and aggregate), D25 was expected to provide a higher total F protein (and trimer) yield than MPE8 and provide a more complete profile. Therefore, in evaluating the D25 antibody column in combination with 293F cells, a higher volume was used to maximize the usefulness of the data obtained from this combination. After transient expression and antibody purification, the RSV F protein was characterized by size exclusion chromatography (SEC) using a Superdex 200 10 / 300 column.

[0154] The three RSV prefusion F trimers showed distinct design-specific patterns. Overall, the optimized sc9-10 DS-Cav1 design significantly outperformed the original DS-Cav1 and SC-TM designs in almost every aspect examined in this comparison. Of particular note, sc9-10 DS-Cav1 showed the highest trimer yield and purity when fused with a trimerization motif at the C-terminus, but produced primarily monomers without any C-terminal trimerization motif, suggesting that the underlying cause of F instability still resides in sc9-10 DS-Cav1. Regarding the effect of cell line on F expression, DS-Cav1 and SC-TM showed little or no yield in 293F cells, but behaved differently in ExpiCHO cells, with DS-Cav1 showing measurable levels of expression but SC-TM not. Therefore, the three prefusion F-designed constructs can be ranked as follows in terms of protein expression: sc9-10 DS-Cav1>>DS-Cav1>>SC-TM, with ExpiCHO being the more suitable expression system.

[0155] Regarding the antibody column, DS-Cav1 and sc9-10 DS-Cav1 showed quite different patterns: for DS-Cav1, D25 gave somewhat lower yields than MPE8 for constructs without a trimerization motif and with a foldon, whereas for sc9-10 DS-CaV1, D25 gave significantly higher yields than MPE8, as indicated by the UV280 absorbance values ​​in the SEC profile.

[0156] Regarding trimer stability, the SEC profiles clearly showed that both DS-Cav1 and sc9-10 DS-Cav1 required a C-terminal trimerization motif to remain trimerized, suggesting that these F designs either did not improve trimer formation or that the primary source of F instability still resided in these constructs. Finally, regarding the composition of the F protein, DS-Cav1 and sc9-10 DS-Cav1 also showed different patterns: DS-Cav1 appeared to have a high-molecular-weight peak at 11 mL, which merged with a trimer peak at 12 mL when using foldon, but this F species peak was clearly visible for a separate C-terminal trimerization motif, 1TD0. In contrast, sc9-10 DS-Cav1 showed a single trimer peak at 12 mL when using foldon, but produced aggregates when using 1TD0. It is also noteworthy that 1TD0 yielded a narrower trimer peak than that yielded by foldon, suggesting a higher trimer purity for the sc9-10 DS-Cav1 F trimer attached to 1TD0. Blue native polyacrylamide gel electrophoresis (BN-PAGE) showed consistent monomer and trimer bands for the different F constructs.

[0157] [Example 11] Characterization of the redesigned RSV F trimer This example describes the characterization of our redesigned RSV prefusion F trimer vaccine antigen described in Example 8.

[0158] Based on our hypothesis that the β3 / β4 hairpin is the underlying cause of RSV F instability, we investigated two disulfide bond mutations to lock this region in its pre-fusion structure (β-hairpin) and prevent it from transitioning to its post-fusion structure (extended α-helix). These two disulfide bond mutants are referred to as A2N-JZ0-V2-Ext-SS and A2N-JZ0-V2-Ext-AT, or simply V2-Ext-SS and V2-Ext-AT (SEQ ID NO: 36 and SEQ ID NO: 40). These three newly designed F constructs were characterized using the aforementioned protocol. Overall, both disulfide bond mutants outperformed the basic design, with V2-Ext-AT being the most superior, as it produced a significant trimer peak with a high yield without a C-terminal trimerization motif. These results provide compelling evidence that the β3 / β4 hairpin is the underlying cause of RSV F instability and that, when properly introduced, as little as a single disulfide bond mutation in this region can effectively eliminate instability. The design could be expressed in both cell lines, with greater yields obtained from ExpiCHO cells.

[0159] Regarding the antibody column, D25 consistently gave higher yields than MPE8. It is worth noting that MPE8, which recognizes the two F promoters of a trimer, did not favor trimers over other species in the produced F protein. For example, a more pronounced trimer peak was observed for V2-Ext, which lacks any C-terminal trimerization motif, expressed in ExpiCHO cells after D25 purification. Regarding trimer formation, the trimerization motif fused to the C-terminus of V2-Ext can result in more difficult trimer folding, e.g., no yield for the V2-Ext-foldon construct, or aggregation, e.g., a peak corresponding to high molecular weight aggregates (8–10 mL) in ExpiCHO cells for the V2-Ext-1TD0 construct. The first mutant (V2-Ext-SS), with a disulfide bond engineered near the β-turn (one residue apart), showed significantly increased F protein expression in both cell lines, with ExpiCHO slightly outperforming HEK 293F. On the antibody column, D25 and MPE8 gave similar SEC profiles for ExpiCHO-produced proteins, but behaved differently for 293F-produced proteins.

[0160] Furthermore, this disulfide bond mutation appeared to be more effective at stabilizing trimers with foldon than with 1TD0 or without the C-terminal trimerization motif. A second mutant (V2-Ext-AT), in which a disulfide bond was engineered at the distal end of the β3 / β4 hairpin, exhibited the most desirable properties, although its yield was lower than that of the first disulfide bond mutant. In addition to yield, the two disulfide bond mutations differed significantly in their ability to promote trimer folding and their compatibility with different C-terminal trimerization motifs. Specifically, AT appeared to be much more effective than SS in promoting trimer formation when no C-terminal trimerization motif was attached. The AT disulfide bond design also appeared to be more compatible with 1TD0 than with foldon.

[0161] Next, we investigated whether adding a double mutation (GP) to the two disulfide bond mutants in the β-turn would further destabilize the postfusion helix and, consequently, stabilize the prefusion β3 / β4 hairpin. Surprisingly, the GP mutation in the β-turn exerted dramatically different effects on the two disulfide bond mutants. The SSGP design showed a much improved trimer ratio in the produced F protein, which was accompanied by a significant decrease in overall yield. A higher trimer peak was also observed for the V2-Ext-SSGP construct fused with foldon and 1TD0, regardless of the cell line used. In contrast, GP had a primarily negative effect on the AT-containing F construct. Although the ATGP design showed lower monomer and aggregate ratios, the overall yield appeared too low for vaccine production in CHO cells. In summary, V2-Ext-SSGP offers a promising alternative to V2-Ext-AT for the prefusion F trimer design.

[0162] Finally, we investigated whether a longer linker between F2 and F1 could improve the two "best" prefusion F constructs identified to date: V2-Ext-SSGP and V2-Ext-AT. Notably, Ext2 mutations consistently improved trimer ratios without a C-terminal trimerization motif. However, the use of a long cleavage site linker appeared to have a negative effect on any F constructs with a C-terminal trimerization motif, either in terms of trimer ratio or trimer yield. In summary, a long cleavage site linker can be used without a trimerization motif.

[0163] [Example 12] Redesigned hMPV and PIV prefusion F trimers This example describes redesigned hMPV and PIV prefusion F trimers with a minimal set of mutations corresponding to those used in the RSV F trimer described in Example 8 to effectively stabilize the prefusion F trimer. Further details of the study are described in Example 13.

[0164] The sequence of the F protein of human metapneumovirus hMPV (isolate "TN03.03.19") is obtained from GenBank under ID AEZ52364. The numbering is based on the crystal structure (PDB ID: 5WB0) and the UniProt definition of another hMPV strain (strain CAN97-83) with ID Q6WB98. Soluble F is defined as M1-T489, with M1-G18 being the signal peptide (see SEQ ID NO: 44 or TN-WT). The truncated form of soluble F (corresponding to Fd of hRSV) is defined as M1-L481, with M1-G18 being the signal peptide (see SEQ ID NO: 45 or TN-WT-cut). The uncleaved prefusion-optimized (UFO) soluble F construct was based on the TN-WT cut design with specific mutations engineered into the equivalent "β3 / β4 hairpin" (residues E146 to T160) based on the hypothesis that this region is the underlying cause of hMPV F instability and undergoes the greatest conformational changes during the membrane fusion process. Disulfide bond(s) between β-sheet-forming amino acids can be introduced to lock the hMPV F structure in the prefusion state. The disulfide bond-introducing mutations A147C / A159C were tested with two different approaches to manipulating the fusion peptide (FP) (see SEQ ID NO: 46 and SEQ ID NO: 47, or TN-cut-UFO1 and TN-cut-UFO2) to validate the instability hypothesis and the importance of the "β3 / β4 hairpin" to hMPV F.

[0165] To stabilize the trimer and increase the trimer ratio within the total protein yield, a trimerization motif such as foldon and the viral capsid protein SHP (PDB:1TD0) can be added to the C-terminus of the redesigned F construct with a short GS linker in the middle. A His6 tag can be added to the C-terminus of the trimerization motif to facilitate protein purification by nickel column.

[0166] The C-terminus of the redesigned F construct can be fused to the N-terminus of a nanoparticle-forming subunit, such that the fusion construct, when expressed in an appropriate cell line, can self-assemble into nanoparticles with prefusion F trimers displayed on the nanoparticle surface.

[0167] hMPV F construct sequence: MSWKVVIIFSLLITPQHG (SEQ ID NO: 54): leader DQLAREEQIENPRQSRFVLGAIALGV (SEQ ID NO: 55): N-terminus of unstructured F2 + cleavage site + fusion peptide EAVSTLGNGVRVLAT (SEQ ID NO: 56): equivalent β3 / β4 hairpin region SEQ ID NO: 44 (TN-WT) [ka]

[0168] SEQ ID NO: 45 (TN-WT-cut) [ka]

[0169] SEQ ID NO: 46 (TN-cut-UFO1) [ka]

[0170] SEQ ID NO: 47 (TN-cut-UFO2) [ka]

[0171] PIV3 F: The sequence of the F protein of human parainfluenza virus type 3 (PIV3) (strain "HPIV3 / USA / 629-D01959 / 2007") is obtained from GenBank under ID (AGW51052). Numbering is based on the cryo-EM structure (PDB ID: 6MJZ) and the UniProt definition of the recombinant PIV3 / PIV1 virus with ID (O55888). The soluble F sequence (corresponding to Fd in hRSV) is defined as M1-T484, and M1-C18 is the signal peptide (see SEQ ID NO: 48 or PIV3-WT).

[0172] The uncleavable, prefusion-optimized (UFO) soluble F construct is based on the PIV3-WT design after removal of the N-terminus and cleavage site of unstructured F2, with specific mutations incorporated into the "β1 / β2 hairpin" (residues V158-I172) based on the hypothesis that this region is the underlying cause of PIV3 F instability and undergoes the greatest conformational changes during the membrane fusion process. Disulfide bonds between β-sheet-forming amino acids in strands V158-V161 and I169-I172 can be introduced to lock the PIV3 F structure in the prefusion state. It should be noted that the disulfide bond introduced within Q162-L168 may cause structural distortion. The disulfide bond-introducing mutation Q159C / A171C has been tested in two different ways addressing the fusion peptide (FP) region (see SEQ ID NO: 61 and SEQ ID NO: 62, or PIV3-UFO1 and PIV3-UFO2) to verify the instability hypothesis and the importance of the "β1 / β2 hairpin" for PIV3 F.

[0173] To stabilize the trimer and increase the trimer ratio within the total protein yield, a trimerization motif such as foldon and the viral capsid protein SHP (PDB:1TD0) can be added to the C-terminus of the redesigned F construct with a short GS linker in the middle. A His6 tag can be added to the C-terminus of the trimerization motif to facilitate protein purification on a nickel column.

[0174] The C-terminus of the redesigned F construct can be fused to the N-terminus of a nanoparticle-forming subunit, such that the fusion construct, when expressed in an appropriate cell line, can self-assemble into nanoparticles with prefusion F trimers displayed on the nanoparticle surface.

[0175] Considering the high structural similarity between other PIVs: PIV3, PIV5 and other PIVs, UFO trimer constructs can be designed according to the same principles to modify the "β1 / β2 hairpin region."

[0176] PIV3 F construct sequence: MLISILSIITTMIMASHC (SEQ ID NO: 57): leader GLKLQKDVIVTNQESNENTDPRTERFFGGVIGTIALGV (SEQ ID NO: 58): N-terminus of unstructured F2 + cleavage site + fusion peptide VQSVQSSVGNLIVAI (SEQ ID NO: 59): β1 / β2 hairpin, equivalent to the β3 / β4 hairpin region.

[0177] SEQ ID NO: 48 (PIV3-WT) [ka]

[0178] SEQ ID NO: 61 (PIV3-UFO1) [ka]

[0179] SEQ ID NO: 62 (PIV3-UFO2) [ka]

[0180] [Example 13] Characterization of redesigned F trimers from other paramyxoviruses Having established that the β3 / β4 hairpin is the underlying cause of RSV F instability, as described in Examples 8-11, we investigated the possibility of extending this design concept to other members of the Paramyxoviridae family. To investigate this possibility, we generated two sets of UFO constructs for hMPV F (SEQ ID NO: 46 and SEQ ID NO: 47) and PIV3 F (SEQ ID NO: 61 and SEQ ID NO: 62), all of which contain a C-terminal trimerization motif (1TD0) followed by a His6 tag.

[0181] In preliminary studies, all four constructs were transiently expressed in 250 mL of 293F cells and purified using a nickel column prior to SEC. Overall, the UFO2 design significantly outperformed the UFO1 design containing a fusion peptide (FP) for both hMPV and PIV3. For hMPV, a significant trimer peak (11-12 mL) was observed for UFO2 but not UFO1, consistent with BN-PAGE analysis. For PIV3, the UFO2 construct produced a high trimer peak (11-12 mL) with a prominent aggregate peak (8-10 mL), whereas the UFO1 construct produced primarily aggregates in 293F cells, similar to our observations for the hMPV constructs.

[0182] In summary, this comparative analysis confirms the notion that the β3 / β4 and β1 / β2 hairpins are fundamental causes of F instability in hMPV and PIV3, respectively. Given the similarities between PIV1-5, this result also suggests that the UFO2 design may be applicable to other PIVs.

[0183] [Example 14] EM characterization of nanoparticles displaying RSV F trimers Negative staining electron microscopy (nsEM) analysis was performed to characterize nanoparticles displaying RSV F trimers as described in Example 8 (FIG. 4). In this study, two nanoparticle platforms were examined: 24-mer ferritin and 60-mer E2p.

[0184] Three representative RSV F designs were tested using ferritin nanoparticles as a model display system (Fig. 4A). A 5-GS linker was inserted between the C-terminus of F and the N-terminus of the ferritin subunit. Among these three designs, sc9-10 DS-Cav1 was the performer with well-formed nanoparticles (Fig. 4A, center), while the DS-Cav1 and SC-TM ferritin fusion constructs were unable to form nanoparticles or native-like F trimers (Fig. 4A, left and right). Two redesigned RSV F trimers, V2-Ext-SSGP and V2-Ext-AT, were displayed on ferritin with 10-GS and 5-GS linkers, respectively (Fig. 4B, columns 1 and 2). Nanoparticles could be purified using D25 and MPE8 antibody columns (Fig. 4B, rows 1 and 2, columns 1 and 2). EM analysis demonstrated that both redesigned F trimers could be successfully displayed on ferritin nanoparticles, with the shorter (5-GS) linker exhibiting a more visible, intact prefusion F spike on the nanoparticle surface. Most importantly, our newly designed prefusion F spike appeared to have a visually discernible difference in shape compared to the sc9-10 DS-Cav1 trimer spike when both were displayed on ferritin nanoparticles. Specifically, the V2-Ext-AT F trimer spike adopted a "thumb"-like shape with a dense surface, characteristic of the prefusion closed-structure trimer spike, whereas the sc9-10 DS-Cav1 trimer spike appeared "lollipop-like" and hollow, indicating an open-structure conformation (Figure 4, A, column 2, box vs. Figure 4, B, row 1, column 2, box). Our EM data indicate that an inherently more stable F trimer, lacking any characteristic of RSV F instability, is important for the development of an RSV F nanoparticle vaccine. It also reveals that three existing representative RSV prefusion F designs, DS-Cav1, optimized sc9-10 DS-Cav1, and SC-TM, are not suitable for nanoparticle display.Finally, we investigated whether our newly designed RSV prefusion F trimers could be displayed on large 60-mer nanoparticles containing a locking domain (LD) and an integrated T helper epitope. To this end, we expressed and purified two nanoparticle constructs, V1-Ext-AT-E2p-LD4 and V1-Ext-AT-E2p-LD4-PADRE, for EM analysis (Fig. 4B, column 3). Consistently, well-formed nanoparticles with "thumb"-like arrays of prefusion F trimers on the surface were observed for both constructs, confirming that the newly designed prefusion RSV F trimers can be displayed on multilayer nanoparticle platforms as vaccine candidates.

[0185] Thus, the invention has been broadly disclosed and illustrated with reference to the exemplary embodiments set forth above, it being understood that various modifications can be made thereto without departing from the spirit and scope of the invention.

[0186] It is further noted that all publications, sequence accession numbers, patents, and patent applications cited herein are expressly incorporated by reference in their entirety for all purposes, as if each were individually so indicated. Definitions contained in the text incorporated by reference are excluded to the extent that they conflict with definitions in this disclosure.

Claims

1. 1. An engineered soluble fusion (F) protein of a paramyxovirus, comprising an altered soluble F sequence having modifications compared to the wild-type soluble F sequence of said paramyxovirus, wherein the modifications include: (1) replacement of two or more negatively charged residues (D486-A490) around the β23 chain with polar or hydrophobic residues; (2) deletion of the P27 peptide (E110-R136); and (3) an engineered intraprotomer disulfide bond within the F1 subunit or linking the F2 and F1 subunits, wherein the amino acid numbering is based on the human RSV A2 strain F protein (UniProt ID P03420).

2. 2. The engineered soluble F protein of claim 1, wherein the paramyxovirus is RSV.

3. 2. The engineered soluble F protein of claim 1, wherein the two or more negatively charged residues around the β23 strand are D486 and E487.

4. 4. The engineered soluble F protein of claim 3, wherein the substitutions around the β23 strand comprise D486N / E487Q or D486L / E487L.

5. 2. The engineered soluble F glycoprotein of claim 1, wherein the engineered disulfide bond is S155C / S290C, S62C / K196C, or E60C / K196C.

6. 2. The engineered soluble F protein of claim 1, further comprising: (1) one or more furin cleavage sites; or (2) a linker portion that replaces the unstructured C-terminus of F2 (Q98-R109) and a portion of the N-terminus of the fusion peptide (FP) (F137-V157).

7. The replaced C-terminus of F2 is residues N104 to R109 ( １０４ NNRARR １０９ 7. The engineered soluble F protein of claim 6, comprising:

8. 7. The engineered soluble F protein of claim 6, wherein the replaced portion of the N-terminus of the fusion peptide (FP) comprises F137-S146.

9. 2. The engineered soluble F protein of claim 1, further comprising a substitution of residue S215.

10. 10. The engineered soluble F protein of claim 9, wherein residue S215 is replaced with P.

11. 2. The engineered soluble F protein of claim 1, further comprising a substitution of residue E92.

12. 12. The engineered soluble F protein of claim 11, wherein residue E92 is replaced with D, Q, another short polar residue, or a hydrophobic residue.

13. 2. The engineered soluble F protein of claim 1, further comprising a V185P substitution.

14. 2. The engineered soluble F protein of claim 1, further comprising one or both of an S46G and a K462Q substitution.

15. 2. The engineered soluble F protein of claim 1, further comprising an intraprotomer disulfide bond S180C / S186C or A177C / T189C engineered into the β3 / β4 hairpin.

16. 2. The engineered soluble protein of claim 1, further comprising an engineered interprotomer disulfide bond A149C / Y458C.

17. 2. The engineered soluble F protein of claim 1, comprising a sequence set forth in any one of SEQ ID NOs: 17-23, a conservatively modified variant thereof, or a substantially identical sequence thereof.

18. 1. An engineered soluble fusion (F) protein of a paramyxovirus, comprising an altered soluble F sequence having a modification compared to the wild-type soluble F sequence of the paramyxovirus, wherein the modification comprises an engineered disulfide bond linking a pair of beta-sheet-forming amino acids in a β3 / β4 hairpin or equivalent hairpin in the F1 subunit, and wherein the numbering of the hairpins is based on that of respiratory syncytial virus (RSV).

19. 19. The engineered soluble F protein of claim 18, wherein the paramyxovirus is human RSV, the engineered disulfide bond is between substituted residues S180C / S186C or A177C / T189C in the β3 / β4 hairpin, and the amino acid numbering is based on human RSV strain A2, having UniProt ID P03420.

20. 20. The engineered soluble F protein of claim 19, wherein the wild-type soluble F sequence is set forth in SEQ ID NO: 1, or a conservatively modified variant thereof.

21. 20. The engineered soluble F protein of claim 19, wherein the modification further comprises a mutation at the unstructured C-terminus of the F2 subunit.

22. 22. The engineered soluble F protein of claim 21, wherein the mutation at the C-terminus of the unstructured F2 comprises a truncation of residues 104-109 (NNRARR) (SEQ ID NO: 31) and a P102A substitution.

23. 20. The engineered soluble F protein of claim 19, wherein the modifications further comprise: (1) replacing the processed activation peptide (P27) (residues E110-R136) and the N-terminus of the fusion peptide (residues F137-S146) with a (GS)n linker sequence, where n is any integer from 1 to 5; and / or (2) the substitutions I379V and M447V.

24. 20. The engineered soluble F protein of claim 19, comprising the amino acid sequence set forth in any one of SEQ ID NOs: 36-43, or a conservatively modified variant thereof.

25. 19. The engineered soluble F protein of claim 18, wherein the paramyxovirus is human metapneumovirus (hMPV), the engineered disulfide bond is between substituted residues A147C / A159C in the β3 / β4 hairpin, and the amino acid numbering is based on hMPV strain CAN97-83, which has UniProt ID Q6WB98.

26. 26. The engineered soluble F protein of claim 25, wherein the wild-type soluble F sequence is set forth in SEQ ID NO:44 or SEQ ID NO:45, or a conservatively modified variant thereof.

27. 26. The engineered soluble F protein of claim 25, wherein the modification further comprises a mutation at the unstructured C-terminus of the F2 subunit.

28. 28. The engineered soluble F protein of claim 27, wherein the mutation in the unstructured F2 C-terminus comprises replacing the unstructured F2 C-terminus DQLAREQIENP (SEQ ID NO:60) and cleavage site RQSR (SEQ ID NO:49) with a (GS)n linker sequence, where n is any integer from 1 to 5.

29. 26. The engineered soluble F protein of claim 25, comprising the amino acid sequence set forth in any one of SEQ ID NOs: 46 and 47, or a conservatively modified variant thereof.

30. 19. The engineered soluble F protein of claim 18, wherein the paramyxovirus is a human parainfluenza virus (hPIV), the engineered disulfide bond is between substituted residues Q159C / A171C in the β1 / β2 hairpin, and the amino acid numbering is based on a recombinant hPIV3 / hPIV1 virus having UniProt ID 055888.

31. 31. The engineered soluble F protein of claim 30, wherein the wild-type soluble F sequence is set forth in SEQ ID NO: 48, or a conservatively modified variant thereof.

32. 31. The engineered soluble F protein of claim 30, wherein the modification further comprises a mutation at the C-terminus of the F2 subunit.

33. 33. The engineered soluble F protein of claim 32, wherein the mutation at the C-terminus of F2 comprises replacing the C-terminal sequence NQESNENTDP (SEQ ID NO:50) and cleavage site RTER (SEQ ID NO:51) with a (GS)n linker sequence, where n is any integer from 1 to 6.

34. 31. The engineered soluble F protein of claim 30, comprising the amino acid sequence set forth in any one of SEQ ID NOs: 61 and 62, or a conservatively modified variant thereof.

35. 1. An engineered soluble F protein of respiratory syncytial virus (RSV), comprising a soluble RSV F sequence that has been altered by (1) deletion of the P27 peptide (residues E110-R136), (2) modification of the unstructured C-terminus of the F2 subunit (residues Q98-R109), and (3) truncation of the N-terminus of the fusion peptide (residues F137-V157), wherein the amino acid numbering is based on human RSV strain A2, having UniProt ID P03420.

36. 36. The engineered soluble F protein of claim 35, wherein the modification at the unstructured F2 C-terminus comprises at least one of: (1) a truncation of residues 104-109 (NNRARR) (SEQ ID NO: 31); and (2) a P102A substitution.

37. 37. The engineered soluble F protein of any one of claims 1 to 36, further comprising a C-terminal trimerization motif.

38. 38. The engineered soluble F protein of any one of claims 1 to 37, further comprising an N-terminal leader sequence.

39. A paramyxovirus immunogenic composition comprising the engineered soluble F protein of any one of claims 1 to 38 displayed on the surface of a self-assembling nanoparticle.

40. 40. The immunogenic composition of claim 39, wherein the self-assembling nanoparticle comprises a trimeric sequence and the C-terminus of the immunogenic polypeptide is fused to the N-terminus of a subunit sequence of the nanoparticle.

41. A polynucleotide sequence encoding the immunogenic polypeptide of any one of claims 1 to 38 or the immunogenic composition of claim 39.

42. A pharmaceutical composition comprising an immunogenic polypeptide according to any one of claims 1 to 38, an immunogenic composition according to claim 39, or a polynucleotide according to claim 41, and a pharmaceutically acceptable carrier.

43. 43. A method for preventing or treating a paramyxovirus infection in a subject, comprising administering to the subject a therapeutically effective amount of the pharmaceutical composition of claim 42.

44. 44. The method of claim 43, wherein the paramyxovirus is RSV, hMPV, or PIV.

Images