GB2535753A

GB2535753A - Particles comprising fusion proteins

Info

Publication number: GB2535753A
Application number: GB1503218.8A
Authority: GB
Inventors: Alexander Roesen Nicholas; Hannemann Holger; Hale Geoffrey
Original assignee: Native Antigen Co
Current assignee: Native Antigen Co
Priority date: 2015-02-26
Filing date: 2015-02-26
Publication date: 2016-08-31
Also published as: GB201503218D0

Abstract

Provided is a virus-like particle comprising a membrane and a fusion polypeptide, wherein the membrane is a lipid bilayer 5 and the fusion polypeptide comprises: (a) a core polypeptide enveloped by the membrane 4; and (b) a membrane-spanning polypeptide which crosses said membrane at least once 2. The core polypeptide may be a protein such as ferritin or lipoamide acetyltransferase, or a viral capsid such as that of hepatitis E, Norwalk virus, Parvovirus B19, Rabbit haemorrhagic disease virus, murine polyoma virus or human papilloma virus. The fusion polypeptide may also comprise a linker polypeptide 3 and/or a polypeptide tag 1. Also provided are nucleic acids and vectors coding for the said particle and methods for producing the said particle. Also provided are methods for using the said particle or said nucleic acids as an immunogen or as a vaccine or for the production of a pharmaceutical composition for use in therapy.

Description

PARTICLES COMPRISING FUSION PROTEINS

BACKGROUND

Transmembrane proteins are of great importance in biology and medicine. They make up 27% of the total human proteome (Almen 2009) but are targets for more than 50% of modern drugs (Overington 2006). However, the study of transmembrane proteins presents formidable technical obstacles because they are frequently expressed at low levels and they are usually very difficult to purify (Smith 2011). As a result, transmembrane proteins represent fewer than 2% of the three-dimensional protein structures available in the protein data bank.

Integral membrane proteins are permanently attached to the membrane and can only be separated and purified using detergents, nonpolar solvents, or denaturing agents. Transmembrane proteins are integral membrane proteins that span across the membrane at least once. The membrane-spanning peptides are most commonly alpha-helix bundles (found in all types of membrane). Beta-barrel proteins are found only in bacteria and outer membranes of mitochondria and chloroplasts.

It would be highly desirable to have a method for the efficient production of purified transmembrane proteins which did not require extraction with detergents, nonpolar solvents or denaturing agents. Avoidance of such reagents would greatly reduce the risk of damage or loss of functional integrity. The availability of purified transmembrane proteins in a native state would be greatly beneficial to biomedical research and development of pharmaceuticals. For example, they could facilitate structural and functional studies and provide candidate vaccines or antigens for raising antibodies.

Particles containing multiple copies of protein subunits are often highly immunogenic, probably because the immune system is attuned to recognition of pathogens with repetitive structures and because particles of about 40 to 50 nm are optimally taken up by dendritic cells (Fifis 2004). Virus-like particles (VLPs) are particles which resemble viruses but do not contain nucleic acid and are therefore non-infectious. They commonly contain one or more virus capsid or envelope proteins which are capable of self-assembly to form the VLP. VLPs composed of the small hepatitis B derived surface antigen (HBsAg) occur naturally in patient sera (Bayer 1968). More recently, VLPs have been produced from components of a wide variety of virus families (Noad 2003, Grgacic 2006). Some VLPs have been approved as therapeutic vaccines, for example Engerix-B (for hepatitis B), Cervarix and Gardasil (for human papilloma viruses). VLPs can be engineered to display heterologous antigens on their surface by creation of fusion proteins between a self-assembling VLP protein and a foreign protein. For example, the extracellular domain of the influenza A M2 ion channel protein has been fused with the hepatitis virus core protein (HBc) to create a VLP which gave mice protective immunity against a lethal challenge with influenza A virus (Neirynck 1999). Or again, a fragment of gonadotropin releasing hormone has been attached to bacteriophage AP205 coat protein to create a VLP which produced neutralising antibodies and concomitant reduction of testosterone levels in male mice (Tissot 2010). Many other examples are described in the literature (Grgacic 2006, Ludwig 2007).

In a few instances, particles have been created based on soluble self-assembling protein subunits not derived from viruses. For example the lipoyl acetyltransferase (E2) core of the pyruvate dehydrogenase multienzyme complex assembles to form an icosahedral structure which can be used as a scaffold to display a variety of soluble proteins and polypeptides (Domingo 2001). Artificial self-assembling proteins have also been created for similar purposes (Raman 2009).

However, VLPs based on non-enveloped viruses or on soluble self-assembling proteins are capable of displaying only foreign antigens which are soluble proteins. In contrast, display of a transmembrane protein in its native configuration requires the presence of a lipid membrane.

Enveloped viruses contain a lipid membrane derived from the host cell and might in principle be capable of displaying heterologous transmembrane proteins. Virus-like particles have been created based on capsid or envelope proteins of such enveloped viruses and used to display epitopes of heterologous proteins (for example Deml 2005, Grgacic 2006, Ludwig 2007, W02013/122262). However, as explained below, substantial technical obstacles prevent the use of such VLPs for display of heterologous transmembrane proteins to meet the needs of biomedical scientists.

W02013/122262 describes the use of VLPs based on an envelope protein of an alphavirus or a flavivirus. Certain sufficiently small heterologous peptide epitopes could be displayed by insertion of the desired sequence into an exposed loop of the envelope protein. However, as described in W0213/122262, the attempt to incorporate a full length heterologous transmembrane protein was unsuccessful. It can be readily understood that such a construct would not be able to self-assemble due to steric interference between the heterologous transmembrane protein and the virus envelope protein.

Retroviruses such as human immunodeficiency virus (HIV) are a different class of enveloped virus which have been used as the basis for construction of VLPs. In this case the gag polyprotein complex can provide a self-assembling capsid around which the lipid membrane is organised (Deml 2005). Two strategies have been used for the incorporation of heterologous polypeptides. In so-called Type 1 VLPs sufficiently small peptide epitopes are fused within the gag polyprotein. This is analogous to the case of the alphavirus or flavivirus described above except that the epitope is now inside the membrane rather than outside. Such VLPs are suited for invoking cellular rather than humoral immunity. So-called Type 2 VLPs approach the problem of expression of transmembrane proteins by expression of the transmembrane protein separately from the gag protein. For example, US7763258 describes a VLP with an enveloped virus core and a heterologous multiple membrane spanning protein. However, the heterologous transmembrane protein is not specifically associated with the VLP; it is simply expressed on the cell membrane and is instead non-specifically gathered into the membrane of the VLP along with other cell membrane proteins. In order for the VLP to have a useful quantity of the desired heterologous protein, the latter must be expressed at a high level on the cell surface. Such VLPs do not contain a single type of transmembrane protein, as they also contain a sample of other cell-derived transmembrane proteins.

Surprisingly, we have found that it is possible to produce purified transmembrane proteins in their native state. In particular, we have shown that self-assembling fusion polypeptides, which are processed in the endoplasmic reticulum of a host cell, can be used to produce particles that contain a transmembrane protein embedded in a lipid membrane.

STATEMENT OF INVENTION

In the first aspect, the invention provides a virus-like particle comprising a membrane and a fusion polypeptide, wherein the membrane is a lipid bilayer and the fusion polypeptide comprises: (a) a core polypeptide enveloped by the membrane; and (b) a membrane-spanning polypeptide which crosses said membrane at least once.

In some embodiments the core polypeptide is capable of self-assembly to form an oligomeric structure greater than 400 kDa in size. The oligomer may be a hetero-oligomer, of two or more different subunits, or a homo-oligomer of identical subunits.

In some embodiments, the core polypeptide may be selected from the group consisting of: ferritin, encapsulin, dihydrolipoylsuccinyltransferase (EC 2.3.1.61), dihydrolipoamide branched chain transacylase (EC 2.3.1.168), lipoyl acetyltransferase (EC 2.3.1.12) or may be a polypeptide fragment or a polypeptide homologue of a member of the said group.

In some embodiments, the core polypeptide is ferritin. In some embodiments, the core polypeptide is lipoyl acetyltransferase.

Alternatively, in some embodiments, the core polypeptide may be a capsid protein or a polypeptide fragment or a polypeptide homologue of the capsid protein wherein the N-terminus or the C-terminus of the core polypeptide is accessible on the outside of the assembled capsid.

In some embodiments the core polypeptide may be selected from a group consisting of: Newton's T=3 capsid protein, Bacteriophage GA capsid protein, Bacteriophage PP7 capsid protein, Bacteriophage AP205 capsid protein, Ryegrass mottle virus capsid protein, Turnip yellow mosaic virus capsid protein, Melon necrotic spot virus capsid protein, Pyrococcus furiosus virus-like protein, Norwalk virus capsid protein, Hepatitis E virus capsid protein, Rabbit haemorrhagic disease virus capsid protein or may be a polypeptide fragment or a polypeptide homologue of a member of the said group, In some embodiments, the membrane-spanning polypeptide may be any transmembrane polypeptide. In some embodiments, it may be one or more selected from a group consisting of: Type I membrane proteins, Type II membrane proteins, Type III membrane proteins, Type IV-A membrane proteins, Type IV-B membrane proteins or is a polypeptide fragment or a polypeptide homologue of a member of the said group.

In some embodiments, the fusion polypeptide may also comprise one or more linker polypeptides. For example, a linker polypeptide may be fused at its N-terminus to the C-terminus of the membrane-spanning polypeptide and at its C-terminus to the N-terminus of the core polypeptide. Alternatively a linker polypeptide may be fused at its N-terminus to the C-terminus of the core polypeptide and at its C-terminus to the N-terminus of the membrane-spanning polypeptide.

In some embodiments, the fusion polypeptide also comprises a polypeptide tag, including an affinity, epitope or fluorescent tag. The tag may incorporate a protease recognition site.

In a further aspect, the invention provides a polynucleotide coding for the fusion polypeptide described above. The polynucleotide may be DNA or RNA.

In a further aspect, the invention provides a vector comprising the said polynucleotide.

In a further aspect, the invention provides a host cell comprising the said polynucleotide, wherein said host cell is capable of expressing the particle described above. The host cell may have been transfected with the polynucleotide or the polynucleotide may have been inherited upon cell division.

In a further aspect, the invention provides a method of preparing the particle described above comprising: culturing the said host cell with a culture medium; and purifying the particle from the culture medium or from the host cell. The method optionally includes biosynthesis of the particle in the endoplasmic reticulum of the host cell. This reduces the possibility of non-fused (that is non-specific) membrane proteins transecting the membrane of the particle.

In further aspects, the invention provides pharmaceutical and vaccine compositions comprising the particle described above and DNA vaccine compositions comprising the polynucleotide described above.

In further aspects, the invention provides a method of producing an antibody, comprising contacting the particle and/or the polynucleotide described above to a bird or mammal and a method of selecting an antibody, comprising screening an antibody library for binding to the particle described above.

In a further aspect, the invention provides an antibody so produced or selected, as well as a pharmaceutical composition comprising the antibody produced or selected as described above.

In a further aspect, the invention provides a medicament for use in therapy comprising any of the pharmaceutical, vaccine or DNA vaccine compositions described above.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described with reference to the following, non-limiting figures, wherein: Figure 1 shows an example of one embodiment of the particle of the invention wherein an oligomerised core polypeptide (4) has been genetically fused to a single linker polypeptide (3), genetically fused to a single transmembrane protein (1+2). The core is surrounded by a lipid bilayer membrane (5), the membrane is transected by a single membrane-spanning domain (2). The membrane-spanning polypeptide displays an exodomain (1) externally. For simplicity, six copies of the fusion protein have been represented, though it is envisaged that there will typically be more copies in total, for example 24, 60 or a multiple of 60 such as 120, 180, 240 or 420.

Figure 2 shows an example of one embodiment of the particle of the invention wherein an oligomerised core polypeptide (4) has been genetically fused to a single linker polypeptide (3), genetically fused to a single transmembrane protein (1+2). The core is surrounded by a lipid bilayer membrane (5), the membrane is transected by two membrane-spanning domains (2). The membrane-spanning polypeptide displays a transmembrane protein exodomain (1) externally and a cytoplasmic domain internally. For simplicity, six copies of the fusion protein have been represented, though it is envisaged that there will typically be more copies in total, for example 24, 60 or a multiple of 60 such as 120, 180, 240 or 420.

Figure 3 shows an example of one embodiment of the particle of the invention wherein an oligomerised core polypeptide (4) has been genetically fused to a single linker polypeptide (3), genetically fused to a single transmembrane protein (1+2). The core is surrounded by a lipid bilayer membrane (5), the membrane is transected by three membrane-spanning domains (2). The membrane-spanning polypeptide displays two exodomains (1) externally and a cytoplasmic domain internally. Some of the fusion proteins are truncated (6) and have only a single membrane-spanning domain. For simplicity, six copies of the fusion proteins have been represented, though it is envisaged that there will typically be more copies in total, for example 24, 60 or a multiple of 60 such as 120, 180, 240 or 420.

Figure 4 is a diagrammatic representation of the topology of some examples of fusion proteins which contain a core polypeptide (Core) linked by a linker polypeptide (Linker) to a transmembrane protein. Five different types of transmembrane proteins are illustrated. They contain domains which are exposed to the lumen of the endoplasmic reticulum (and eventually the outside of the particle) or the cytosol (and eventually the inside of the particle). These are linked by a membrane-spanning domain, either an internal stop-transfer anchor sequence (STA) or an internal signal-anchor sequence (SA). Type I transmembrane proteins have no SA domain and instead have a signal sequence at the N-terminus which is often cleaved off during processing in the endoplasmic reticulum. Note that the core polypeptide must be linked to a cytosolic domain of the transmembrane protein. Type IV-A transmembrane proteins have such a domain at each end of the polypeptide chain, so the core polypeptide may be attached to the N-terminus or C-terminus.

Figure 5 is a diagram of plasmid pAC1, an acceptor vector for linking core polypeptides to the N-terminus of a membrane polypeptide.

Figure 6 is a diagram of plasmid pAC3, an acceptor vector for linking core polypeptides to the C-terminus of a membrane polypeptide.

DETAILED DESCRIPTION OF THE INVENTION

Transmembrane proteins are of great importance in biology and medicine but are very difficult to purify in a native form. This invention provides biosynthetically-produced, self-assembling membrane-bound virus-like particles which are capable of displaying specific transmembrane proteins. These virus-like particles can be isolated and used in a range of applications, including for example as antigens, vaccines or as a source of transmembrane protein for biochemical or structural studies.

Virus-like particles based on non-enveloped viruses or on soluble self-assembling proteins are capable of displaying only foreign antigens which are soluble proteins, or soluble domains of membrane proteins. In contrast, display of a transmembrane protein in its native configuration requires the presence of a lipid membrane.

As the art shows, it is not trivial to provide a lipid membrane without incorporating large number of unwanted cellular membrane proteins. These can interfere with use of the VLP as a means to generate immune responses (for example by displaying unwanted epitopes or antigens). The presence of additional or un-specified cellular membrane proteins can also take up much needed space on the membrane itself and can even lead to steric hindrance. This is significant in nano-sized particles. Furthermore, the art does not provide a core polypeptide fused to the transmembrane polypeptide. Accordingly, until this invention, the problem of producing purified transmembrane proteins for structural or functional studies or for use as antigens or vaccines had not been solved.

The present invention allows for the efficient production of purified transmembrane proteins without extraction with detergents, nonpolar solvents or denaturing agents, thus reducing the risk of damage or loss of functional integrity. We provide purified transmembrane proteins which are in a native state. This is of great use in biomedical research and the development of pharmaceuticals. These native trans-membrane antigens can be used, for example, in structural and functional studies and can provide candidate vaccines or antigens for raising antibodies. The native state of the transmembrane protein is typically that found in a wild-type or un-transfected cell for the particular host or transformant species (e.g. bird or mammal, in particular chickens, mice, hamsters, rats and other rodents, goats, rabbits, llamas, sheep or primates including humans, and so forth). This native state is achieved by using self-assembling fusion polypeptides which are processed in the endoplasmic reticulum (ER).

The transmembrane protein embedded in a lipid membrane may be a single transmembrane protein in the sense that just one copy of a particular transmembrane protein is provided. However, multiple copies of the same transmembrane protein may also be provided, provided that they are fused with the core. Indeed, two or more different transmembrane proteins may also be provided, with the same proviso. What distinguishes the present selected or specified transmembrane proteins from the art, where numerous cellular proteins are found in the lipid membrane of the particle, is that the present transmembrane proteins are fused with the core during formation of the particle as it buds off the endoplasmic reticulum (ER). The fusion typically occurs via a linker as discussed elsewhere. It is envisaged that the linker may be cleaved after particle formation (budding off from the ER), but it is preferred that this does not occur and that the core remains fused with the transmembrane protein (i.e. that the linker remains intact).

Five different types of transmembrane proteins are exemplified as illustrated in Figure 4. In some embodiments, the membrane-spanning polypeptide comprises one or more domains which are exposed to the lumen of the endoplasmic reticulum. These will eventually present on the outside of the particle. Alternatively, in some embodiments, the membrane-spanning polypeptide comprises one or more domains which are exposed to the cytosol (and will eventually present on the inside of the particle). In either case, these may be linked, in some embodiments, by a membrane-spanning domain, either an internal stop-transfer anchor sequence (STA) or an internal signal-anchor sequence (SA).

Certain embodiments of the invention can be visualised in Figures 1-3. Any or all of the following features are preferred and in any combination. A core polypeptide which is capable of forming oligomers has been genetically fused to a single linker polypeptide. The linker has been genetically fused to a single membrane-spanning polypeptide. The core is surrounded, encompassed or enveloped by or within a lipid membrane. The core may, therefore, be inside, not through, the membrane.

The membrane is transected at least once by the membrane-spanning polypeptide so that the membrane-spanning polypeptide has at least one single membrane-spanning domain. In some embodiments, the membrane-spanning polypeptide is encoded with more than one membrane-spanning domain and the exposed domains on the ER lumen or cytosolic sides are cleaved post translationally or after insertion in the membrane.

The membrane-spanning polypeptide may, in some embodiments, displays transmembrane protein exodomains externally.

The core is, in some embodiments, an oligomerised polypeptide. It may be fused to a linker polypeptide. The linker may also be fused to the membrane-spanning polypeptide. In either or both cases, this fusion is, in some embodiments, genetic, i.e. encoded by a single polynucleotide.

The membrane-spanning polypeptide may comprise or be further linked or fused to a polypeptide tag.

The particle is preferably in the nanometre range, i.e. typically in the range of 1 nm to 1000 nm. This distinguishes it greatly from a cell (aside from the absence of non-specified cellular membrane proteins). Preferably, the particle may be 8 nm to 500 nm. In some embodiments, the particle may be 10 nm to 400 nm. In some embodiments, the particle may be 15 nm to 200 nm. In some embodiments, the particle may be 10 nm to 150 nm. In some embodiments, the particle may be 10 nm to 100 nm. In some embodiments, the particle may be 10 nm to 40 nm. In some embodiments, the particle may be 15 nm to 30 nm. In some embodiments, the particle may be 20 nm to 40 nm. The particles measured in the Examples, for instance, were about 18 nm in diameter which is consistent with the size expected for a ferritin core (11 nm) surrounded by a lipid membrane.

This invention is not limited to specific compositions or process steps, as such may vary. Whenever appropriate in this specification and the appended claims, terms used in the singular form will also include the plural and vice versa.

Unless defined otherwise, all technical and scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which this invention is related. For example, the Concise Dictionary of Biomedicine and Molecular Biology, Juo, Pei-Show, 2nd ed., 2002, CRC Press; The Dictionary of Cell and Molecular Biology, 3rd ed., 1999, Academic Press; and the Oxford Dictionary Of Biochemistry And Molecular Biology, Revised, 2000, Oxford University Press, provide general dictionaries of many of the terms used in this invention.

Amino acids may be referred to by either their commonly known three letter symbols or by the one-letter symbols recommended by the I UPAC-IUB Biochemical Nomenclature Commission. Nucleotides, likewise, may be referred to by their commonly accepted single-letter codes.

The meaning of certain terms used in this invention may be different from the meanings of similar terms used in the literature. Definitions of such terms, in the context of this invention, are provided below.

Accessible. The N-terminus or C-terminus of a core polypeptide is considered to be "accessible" if it can form a peptide bond with another polypeptide, for example a linker polypeptide or a membrane-spanning polypeptide, and the resulting fusion protein is still capable of capable of self-assembly to form an oligomeric structure.

Antibody. An "antibody" is a molecule with one or more protein chains containing at least one domain which can potentially bind to an antigen (the "antigen-binding domain"). It may be an intact immunoglobulin of any isotype (for example IgG, IgE, IgM, IgD, IgA or IgY), subtype (for example IgG1, IgG2, IgG2a, IgG2b, IgG2c, IgG3, IgG4, IgAl and IgA2) or allotype. It may be a fragment of an immunoglobulin (for example Fab, F(ab1)2,scFv, dAb). It may be of any species (for example human, monkey, camel, llama, goat, sheep, rabbit, rat, mouse, hamster or chicken) or it may be a hybrid derived from more than one species. It may be a polyclonal antibody or a monoclonal antibody. It may be naturally occurring or may be created by genetic engineering or phage display (for example a chimeric antibody, humanised antibody, camelised antibody, intrabody, domain antibody, single-chain variable region).

Antibody Library. An "antibody library" is a collection of different antibodies. The antibodies may be produced by any means, for example, from animal sera, from hybrid myeloma cell lines, from isolated lymphocytes, or by bacteriophage display.

Capsid Protein. A "capsid protein" is a protein which comprises the shell (capsid) of a virus or a virus-like particle or a peptidic nanoparticle which resembles a virus-like particle. It comprises multiple identical subunits which are capable of self-assembly (possibly together with other capsid proteins) to form a regular structure, for example an icosahedron or a helical cylinder.

Core Polypeptide. A "core polypeptide" is a polypeptide which lacks a membrane-spanning domain or a signal peptide and is capable of self-assembly to form an oligomeric structure. It may therefore form the core or central part of a particle of this invention.

Culture medium. "Culture medium" is an aqueous mixture which is capable of supporting the growth of a host cell and permitting the expression of a product encoded by a gene contained within its transfected nucleic acid.

Fusion polypeptide. A "fusion polypeptide" (sometimes known as a fusion protein or a chimeric protein) is a polypeptide created through the joining of two or more genes that originally or in the wild type coded for separate polypeptides. Transcription and translation of this fusion gene results in a single or multiple polypeptides with properties derived from each of the original proteins. Naturally occurring fusion polypeptides can be found in cancer cells. In the context of this invention, the term "fusion polypeptide" is taken to mean a fusion polypeptide created by genetic engineering, and which may include genes or gene segments which are altered (compared with an original gene which coded for a separate polypeptide). A fusion polypeptide may consist of single or multiple polypeptides, any of which may be post-translationally modified.

Host Cell. A "host cell" is a cell which is capable of being transfected with nucleic acid to produce a cell which expresses the product of a gene contained within the nucleic acid.

Icosahedral Virus. An "icosahedral virus" possesses one or more capsid proteins which assemble with 5:3:2 rotational symmetry having 20 equilateral triangular faces and 12 regular pentagon faces. Examples of icosahedral viruses are available at the VIPER database: http://viperdb.scripps.edu/ (Natarajan 2005) Immunomodulation. "Immunomodulation" is the adjustment of an immune response to a desired level. Various forms of immunomodulation may be used as therapeutic interventions for treatment of disease. They may include: (a) immunopotentiation, induction or enhancement of an immune response, for example for the prevention or treatment of infectious disease or cancer, (b) immunosuppression, reduction or abrogation or an immune response, for example for the treatment of autoimmune disease, (c) induction of immunological tolerance, for example for the prevention or treatment of transplant rejection or autoimmune disease.

Linker Polypeptide. A "linker polypeptide" (sometimes known as a spacer) is a sequence of amino acids inserted into a fusion polypeptide between two other polypeptide segments for the purpose of ensuring or improving on the folding, activity, or spacing between the said segments. Various methods are well known in the art for design of flexible, rigid or cleavable linker polypeptides (for example Arai 2001, Wriggers 2005, Chen 2013). Non-limiting examples that we used include Gly-Ser linkers (SEQ ID NO: 33 to SEQ ID NO: 37), lipoyl domain from lipoamide acetyltransferase (SEQ ID NO: 45), EAAAK linker (SEQ ID NO: 46), and modified Moloney murine leukaemia virus gag protein (SEQ ID NO: 47).

Lipid membrane. A "lipid membrane" (sometimes known as lipid bilayer) is a thin polar membrane made of two layers of lipid molecules. Natural lipid membranes are found in all living organisms and usually comprise phospholipids with a hydrophilic head and two hydrophobic fatty acid tails and may include other lipids, for example cholesterol. Such lipid membranes comprise part of a particle of this invention.

Membrane Protein. A "membrane protein" is a protein which interacts with a lipid membrane. There are many types and classifications of membrane proteins. Integral membrane proteins are permanently attached to the membrane. Transmembrane proteins (otherwise known as polytopic integral membrane proteins) span the membrane. Monotopic integral membrane proteins are attached to one side. Peripheral membrane proteins are attached to the membrane or to integral membrane proteins by electrostatic or hydrophobic forces. Transmembrane proteins may be further classified in relation to the number of membrane-spanning domains and the position of the N-and C-terminal domains as follows (Lodish et al. 2007). Types I, II, and III are single-pass molecules. Type I transmembrane proteins are anchored to the lipid membrane with a stop-transfer anchor (STA) sequence and have their N-terminal domains targeted to the endoplasmic reticulum (ER) lumen during synthesis (and the extracellular space, if mature forms are located on plasmalemma). Type II and III are anchored with a signal-anchor (SA) sequence, with type II being targeted to the ER lumen with its C-terminal domain, while type III have their N-terminal domains targeted to the ER lumen. Type IV-A are multi-pass molecules with an N-terminal domain targeted to the cytosol and Type IV-B are multi-pass molecules with an N-terminal domain targeted to the lumen.

Membrane-spanning domain. A "membrane-spanning domain" (sometimes known as a transmembrane domain) is a peptide which can potentially cross a lipid membrane once and make up part of a transmembrane protein. It typically consists of an alpha helix with between 5 and 25 mostly hydrophobic amino acids (the core) flanked by 1 to 5 mostly charged amino acids (the caps) on either side (Krogh 2001). Non-limiting examples that we have used include membrane-spanning domains from human transferrin receptor 1 (SEQ ID NO: 38) and influenza haemagglutinin (SEQ ID NO: 42).

Membrane-spanning polypeptide. A "membrane-spanning polypeptide" is a transmembrane protein or a fragment thereof which includes at least one membrane-spanning domain.

Passive Immunization. "Passive immunization" is the administration of an antibody to a human or animal subject with the intention of altering the immunity of the subject towards one or more antigens. Passive immunization may be used for the treatment or prophylaxis of disease, for example infectious disease, autoimmune disease, cancer, transplant rejection, cardiovascular disease.

Polypeptide Fragment. A "polypeptide fragment" is a polypeptide which has fewer amino acids than a similar naturally-occurring polypeptide and still retains a desirable characteristic of the naturally-occurring polypeptide (for example the ability to self-assemble into oligomers, exhibit an antigenic epitope or express a physiological activity). Compared with the similar naturally-occurring polypeptide, the polypeptide fragment may have one or more amino acid residues deleted from the N-terminus, the C-terminus, elsewhere in the sequence, or any combination of such deletions.

Polypeptide Homologue. A "polypeptide homologue" is a polypeptide which has statistically-significant sequence similarity to a naturally-occurring polypeptide and shares a desirable characteristic with the naturally-occurring polypeptide (for example the ability to self-assemble into oligomers, exhibit an antigenic epitope or express a physiological activity) Polypeptide tag. A "polypeptide tag" is a peptide sequence which is genetically grafted onto a recombinant polypeptide in order to provide a novel property. An affinity tag (for example, polyhistidine, chitin-binding protein, maltose binding protein, glutathione-S-transferase) may allow purification using an affinity binding technique. An epitope tag (for example, FLAG-tag, V5-tag, Myc-tag, and HA-tag) may allow binding of an antibody. A fluorescent tag (for example, green-fluorescent protein or its variants) may allow visualisation of the recombinant polypeptide. A protein tag may allow post-translational modification (for example biotin ligase target sequence) or chemical modification (for example, fluorescein-arsenical helix binder FlAsH). Optionally, a tag may incorporate a protease recognition site for cleavage from the recombinant protein. Non-limiting examples that we have used include polyhistidine tag (SEQ ID NO: 39), FLAG tag (SEQ ID NO: 12) and V5 tag (SEQ ID NO: 20).

Post-translational modification. "Post-translational modification" is a step in protein biosynthesis whereby a polypeptide chain is modified in any of a number of ways including for example folding, cutting, cross-linking, modification of amino acid residues or addition of other molecules. Many different types of post-translational modification are known to exist in nature, some of the most common being N-linked glycosylation, phosphorylation, acetylation, methylation, palmitoylation, citrullination, sulfation, 0-linked glycosylation and amidation (Khoury 2011) as well as formation of disulphide bonds, folding, specific proteolytic cleavages and assembly of multimeric proteins.

Protease cleavage site. A "protease cleavage site" is a peptide sequence which is amenable to specific cleavage by a protease.

Vector. In the context of this invention a "vector" is a polynucleotide used as a vehicle to artificially carry foreign genetic material into another cell, where it can be replicated and/or expressed. The host cell may be a prokaryote but is preferably a eukaryote. The vector may be a plasmid, a viral vector, a cosmid or an artificial chromosome. A vector for expression of a polypeptide in a eukaryotic cell typically comprises: (a) sequences necessary for replication in bacteria including an origin of replication and an antibiotic resistance gene (for example for resistance to kanamycin or ampicillin), (b) sequences necessary for expression of the foreign genetic material in eukaryotic cells including a multicloning site for insertion of the foreign genetic material, at least one promoter sequence to drive expression of the foreign genetic material, a polyadenylation signal, a ribosome recognition site eg a Kozak sequence and optionally (c) a sequence coding for a marker selectable in eukaryotic cells to confer, for resistance to an antibiotic such as G418, hygromycin or puromycin. A non-limiting example of an expression vector we have used is pSV-CMV (SEQ ID NO: 5).

Virus-like particle. A "virus-like particle" is a particle which resembles a virus but is not infectious, normally because it does not contain viral nucleic acid.

Overview Particles of the invention ideally have any, and preferably all, of the following features: 1. They contain a core polypeptide which is capable of self-assembly to form a stable oligomeric structure 2. The core polypeptide is surrounded by a lipid membrane 3. A membrane-spanning polypeptide is embedded in the lipid membrane 4. The core polypeptide and the membrane-spanning polypeptide have been part of the same fusion polypeptide at some point.

In certain embodiments, the core polypeptide and the membrane-spanning polypeptide are derived from different sources. They are not parts of a single polypeptide which can be found in nature. Rather, they have been joined together in a fusion polypeptide.

In certain embodiments, the intact fusion polypeptide may be found in the particle of the invention or in other embodiments it may have been subjected to proteolytic cleavage (either during biosynthesis or subsequently) so that the membrane-spanning polypeptide and the core polypeptide need not be covalently connected in the particle.

In certain embodiments the fusion polypeptide may contain any number of other parts which facilitate the assembly and/or use of the particle. These other parts may, for example, be selected from the group including: linker polypeptide, membrane-spanning domain, polypeptide tag, protease cleavage site.

In certain embodiments, the particle of this invention is produced by a eukaryotic cell wherein the fusion polypeptide is subjected to processing within the endoplasmic reticulum. As a result, it may be subject to post-translational modification in a manner which is characteristic of the eukaryotic cell.

The core polypeptide: cellular proteins In certain embodiments the core polypeptide is derived from a cellular protein. The protein is ideally capable of self-assembly. Examples of such proteins include ferritin, dihydrolipoylsuccinyltransferase (EC 2.3.1.61), dihydrolipoamide branched chain transacylase (EC 2.3.1.168), lipoyl acetyltransferase (EC 2.3.1.12). These proteins may be derived from any prokaryote or eukaryote, including for example bacteria, plants, invertebrates, vertebrates, mammals, humans. Other examples of potential core polypeptides include encapsulin and lumazine synthase (EC 2.5.1.78) which may be derived for example from bacteria. The core polypeptide may be the same as the natural form of the soluble protein or it may be altered in any number of the following ways: mutation of one or more amino acid residues, deletion of one or more amino acid residues, insertion of one or more amino acid residues, post-translational modification. Thus the core polypeptide may be a polypeptide fragment or a polypeptide homologue provided that it retains the capability of self-assembly into an oligomeric structure.

In preferred embodiments the core polypeptide is derived from lipoyl acetyltransferase. This enzyme (sometimes called dihydrolipoyl acetyltransferase or E2) forms the self-assembling core of the pyruvate dehydrogenase multienzyme complex (Perham 1991). In some species (for example E. coli and Azotobacter vinelandit) the core consists of 24 subunits with octahedral symmetry whereas in other species (for example Bacillus stearothermophilus, Bacillus subtilis, Streptococcus faecalis, Saccharomyces cerevisiae and mammals including humans) the core consists of 60 subunits with icosahedral symmetry. Lipoyl acetyltransferase which form cores of 60 subunits is preferred because such a particle will, in principle, be larger than a core consisting of 24 subunits and so may be better able to display a membrane-spanning polypeptide.

Lipoyl acetyltransferase monomers consist of one, two or three highly homologous lipoyl domains (each about BkDa), linked together by long flexible peptides, followed by a binding domain of about 5 kDa which binds peripheral enzyme subunits and which is linked in turn to a C-terminal core domain of about 29 kDa. It is this core domain which forms the self-assembling core of the complex and contains the enzyme active site (Perham 1991). The E2 core enzymes (dihydrolipoylsuccinyltransferase and dihydrolipoamide branched chain transacylase) found in other 2-oxo acid dehydrogenase multienzyme complexes have similar structures, though there are differences in the number of lipoyl domains between enzymes of different species.

The domain structure of the E2 enzymes provides an opportunity to construct core polypeptides of different sizes which may be suitable for displaying different membrane-spanning polypeptides. In certain embodiments the core polypeptide consists of only the core domain and in other embodiments the core polypeptide consists of the binding domain linked to the core domain or one or more lipoyl domains linked to the binding domain linked to the core domain. In principle a larger core polypeptide will provide a greater surface area of lipid membrane to support larger membrane-spanning polypeptides, especially those which contain multiple membrane-spanning domains.

The core polypeptide: viral proteins In certain embodiments the core polypeptide is derived from a viral capsid protein which is capable of self-assembly. There are many such proteins known in the literature. The most common self-assembled structures are icosahedra and helical cylinders. Not all viral capsid proteins may be suitable candidates for a core polypeptide. For example, some may have an N-terminus and C-terminus which is buried within the structure or exposed only on an inner surface and incapable of providing an attachment point for a fusion protein without disrupting the assembly. Ultimately it is necessary to test the suitability of a proposed capsid protein experimentally. This can be done, for example, using methods described in the Examples. But when the three-dimensional structure of the capsid protein (or a structurally related family member) is already known, it is possible to select potential candidates by examination of the structure to identify the position of the ends of the protein chain, for example as described in Example 2 and Example 3.

By 2005 more than 70 three dimensional structures of icosahedral viral capsids were known and deposited in the VIPER database (Nataragan 2005). By 2008 the database contained 256 entries from 28 families of virus (Carillo-Tripp 2008). In January 2015 the database contained 544 entries including 271 unique structures from 38 families of viruses. As described in Example 3, it is possible to interrogate the database to identify short-lists of viral capsids with properties that may make them preferable candidates for use as a core polypeptide. For example, one can identify a list of candidates which have amino acid residues at the N-terminus or the C-terminus or both the N-terminus and the C-terminus of the protein chain which are likely to be exposed on the outside of the oligomeric particle. This approach has some limitations. For example: (1) The ends of the protein chain may not be defined in the crystal structure. However, even a truncated end can be informative, since it is unlikely that unresolved residues are buried or cross the capsid shell. (2) Not all structures have been solved to high resolution and therefore the structural data for chain ends may not be definitive. (3) Attachment of protein chains to the N-terminus or C-terminus to create fusion proteins might disrupt folding and/or assembly of the viral capsid. For these reasons, short-listed candidates need to be experimentally tested to determine whether they are suitable for use as a core polypeptide. Nevertheless, the process of creating a short-list by examination of structures reduces the amount of experimental work and may also facilitate the design of fusion proteins.

In preferable embodiments the oligomeric capsid protein has a diameter greater than 20 nm. The surface area of a core polypeptide made from a smaller capsid may be insufficient for the presentation of substantial membrane polypeptides without the addition of polypeptide spacers. In some embodiments the oligomeric capsid protein has a diameter in the range 20 nm to 100 nm. In some embodiments it has a diameter in the range 20 nm to 80 nm. In some embodiments it has a diameter in the range 20 nm to 60 nm. In some embodiments it has a diameter in the range 20 nm to 40 nm. Using data available in the VIPER database It is straightforward to refine a short-list to include only capsid proteins within a desirable range of diameters.

Following is a non-exclusive short list of viruses or virus-like particles which may provide capsid proteins that are candidates for consideration as core polypeptides by virtue of their size and potential exposure of N-terminus and/or C-terminus: Bacteriophage GA, Bacteriophage PP7, Bacteriophage AP205, Ryegrass mottle virus, Turnip yellow mosaic virus, Melon necrotic spot virus, Pyrococcus furiosus virus-like protein, Norwalk virus, Hepatitis E virus, Rabbit haemorrhagic disease virus, The core polypeptide: purpose-designed proteins In certain embodiments the core polypeptide is a protein which has been designed to self-assemble into a suitable oligomeric structure. For example, EP1594469 describes peptidic nanoparticles which consist of aggregates of a continuous chain comprising two peptidic oligomerization domains connected by a linker segment. If a trimer-forming domain is linked to a pentamer-forming domain, the resulting aggregates may contain a significant proportion of icosahedral particles (Raman 2009) and thus may be suitable candidates for core polypeptides of the present invention, particularly because both the N-and C-termini of the polypeptide are both exposed near the outside of the particle. For example, the polypeptide identified as Newtons T=3 capsid (ID upana1503) in the VIPER database or P6c peptide (Yang 2012) is a candidate for consideration as a core polypeptide.

The polypeptide may be further engineered to introduce other desirable properties. For example it may be stabilised by introduction of Cys residues in suitable positions to form inter-chain disulphide bonds, the net charge may be manipulated by altering the content of acidic or basic amino acid residues, the overall size may be manipulated by altering the length of the oligomerization domains or other modifications may be made as further described below.

Modified core polypeptides In some embodiments the sequence of a native protein providing a core polypeptide is altered, for example by deletion or mutation of amino acid residues, to create a polypeptide fragment or polypeptide homologue of the original core polypeptide. Such a modified core polypeptide may have superior properties compared with the original core polypeptide, for example: a higher level of expression in mammalian cells, improved stability with respect to thermal, oxidative and/or proteolytic degradation, the ability to form larger particles and/or expressible fusion proteins with a greater variety of transmembrane polypeptides.

In some embodiments, two native core polypeptides may be fused together with a suitable linker polypeptide to improve the stability of the particle, for example as described by Caldeira (2010).

Core polypeptides: unpredictability of formation of particles As discussed above, candidate core polypeptides can be selected from published literature based on rational principles such as the accessibility of N-and C-termini in the three-dimensional structure, the size and stability of the particle and ease of expression in mammalian cells. As shown in Example 13 and Example 15 there is a reasonable correlation between the theoretical predictions and the measured expression of particles. Nevertheless, there are exceptions; surprisingly some theoretically plausible particles are not detected and others are detected even though they were considered to be theoretically implausible. Therefore it remains necessary to experimentally test a candidate core polypeptide to determine whether it is capable of forming fusion polypeptides and particles suitable for a particular application. We provide several examples of core polypeptides suitable for different applications and straightforward methods for identification of others.

The membrane-spanning polypeptide An especially useful feature of this invention is the ability for a range of different membrane-spanning polypeptides to be incorporated into the particles of this invention. Naturally occurring transmembrane proteins have a number of different structures and many of them can be incorporated into the particles of this invention with a suitable choice of core polypeptide, and, if necessary, the inclusion of suitable linker polypeptide(s) and/or suitable protease cleavage site(s).

Throughout the following description, it is to be understood that a membrane-spanning polypeptide may be a naturally-occurring transmembrane protein or a polypeptide fragment of polypeptide homologue of such a naturally-occurring transmembrane protein. Examples of different topologies which are discussed below are illustrated in Figure 4.

In certain embodiments the membrane-spanning polypeptide is a Type I transmembrane protein (for example low-density lipoprotein receptor, influenza haemagglutinin protein, insulin receptor, growth hormone receptor and many others well known in the art).

In certain embodiments the membrane-spanning polypeptide is a Type II transmembrane protein (for example asialoglycoprotein receptor, transferrin receptor, DC-SIGN and many others well known in the art) In certain embodiments the membrane-spanning polypeptide is a Type III transmembrane protein (for example cytochrome P450 and many others well known in the art) In certain embodiments the membrane-spanning polypeptide is a Type IV-A or Type IV-B transmembrane protein (for example a G-protein coupled receptor, glucose transporter, voltage-gated Cat' ion channels, ABC small molecule pump, and many others well known in the art) In certain embodiments the membrane-spanning polypeptide has an odd number of membrane-spanning domains with an N-terminal domain which is targeted towards the lumen of the endoplasmic reticulum and a C-terminal domain which remains in the cytosol (for example when the membrane-spanning polypeptide is a Type I, Type III, or Type IV-B transmembrane protein). In such cases the fusion polypeptide of the invention is preferably constructed with the membrane-spanning polypeptide linked, either directly or with a linker polypeptide, to the N-terminus of the core polypeptide.

In other embodiments the membrane-spanning polypeptide has an odd number of membrane-spanning domains with an N-terminal domain which remains in the cytosol while the C-terminal domain is targeted to the lumen of the endoplasmic reticulum (for example when the membrane-spanning polypeptide is a Type II transmembrane protein). In such cases the fusion polypeptide of the invention is preferably constructed with the membrane-spanning polypeptide linked, either directly or with a linker polypeptide, to the C-terminus of the core polypeptide.

In other embodiments the membrane-spanning polypeptide has an even number of membrane-spanning domains with both N-terminal and C-terminal domains which remain in the cytosol (for example when the membrane-spanning polypeptide is a Type IV-A transmembrane protein). In such cases the fusion polypeptide of the invention may be constructed with the membrane-spanning polypeptide linked, either directly or with a linker polypeptide, to either the N-terminus or the C-terminus of the core polypeptide.

In certain embodiments the topology of a transmembrane protein may be altered by addition or deletion of membrane-spanning domains so that the membrane-spanning polypeptide is a polypeptide fragment or polypeptide homologue with a different, and potentially more convenient topology than the transmembrane protein on which it is based. So for example, a Type II transmembrane protein might be modified by addition of a stop-transfer anchor membrane-spanning domain to its C-terminus in order to retain the C-terminal domain in the cytosol which could be linked to the N-terminus of the core polypeptide. Or for example, a Type I, Type III or Type IV-B transmembrane protein might be modified by additional of an internal signal anchor sequence to its N-terminus in order to retain the N-terminal domain in the cytosol which could be linked to the C-terminus of the core polypeptide.

Mixtures of fusion polypeptides containing different membrane-spanning polypeptides It can be appreciated that there is likely to be an upper limit to the number of membrane-spanning domains which can be included within a given fusion polypeptide. This limit is likely to depend on the size of the core polypeptide and the diameter of the assembled particle which it can form. Too many membrane-spanning domains would create steric hindrance to the successful assembly of the particle. Therefore, in certain embodiments the particle consists of a mixture of at least two types of fusion protein where all types have a similar or identical core polypeptide but at least one type has the desired membrane-spanning polypeptide (with 2, 3, 4, 5, or more membrane-spanning domains) and at least one type has a membrane-spanning polypeptide with a smaller number of membrane-spanning domains. An example of such a particle is illustrated in Figure 3. In preferred embodiments, the proportion of the different types of fusion polypeptide in the particle is adjusted to give optimal expression of the larger membrane-spanning polypeptide, for example as described in Example 18.

Linker polypeptides In certain embodiments the particle of this invention may contain one or more linker polypeptides, each of which may have different properties and purposes.

In certain embodiments the core polypeptide and the membrane-spanning polypeptide are connected by means of a linker polypeptide. Such a linker polypeptide is located inside the lipid membrane and may facilitate the assembly of the particle of this invention. For example, it may provide flexibility so that the core polypeptide and the membrane-spanning polypeptide may each be aligned in a configuration which is favourable for self-assembly of the particle. Or for example, it may provide a relatively rigid spacer so that the core polypeptide and the membrane-spanning polypeptide may be separated by a sufficient distance to be favourable for self-assembly of the particle.

In certain embodiments a linker polypeptide may link two domains of the membrane-spanning polypeptide, or it may link the membrane-spanning polypeptide to a polypeptide tag. Such a linker polypeptide is located on the outside of the lipid membrane of the particle of this invention. It may facilitate the folding or display of the membrane-spanning polypeptide or the folding or display of a polypeptide tag.

It is well known in the art that the construction of fusion proteins involves the linking of two proteins or two domains of a protein by means of a peptide linker. Suitable selection of the linker is important for the correct folding and function of the fusion protein. There are several studies of linker design and selection (for example Argos 1990, Alfthan 1995, Robinson 1998, Crasto 2000, Arai 2001, George 2002, Wriggers 2005, Chen 2013). Two independent studies (Argos 1990, George 2002) examined databases of naturally-occurring peptide linkers to identify common properties and preferred structures. Other investigators have designed empirical linkers (for example flexible linkers, rigid linkers, cleavable linkers) for different applications (reviewed by Chen 2013). Flexible linkers are generally composed of small non-polar (eg Gly) or polar (eg Ser or Thr) amino acids. Incorporation of Ser or Thr maintains stability in aqueous solution. Commonly used flexible linkers consist of repeating units (GGGGS)n. However, the genetic stability and expression of a fusion protein containing many repetitive units might be compromised and in such a case a flexible linker more closely based on a natural linker peptide would be preferred. Rigid linkers may be advantageous when it is desired to maintain a certain separation between the core polypeptide and the membrane-spanning polypeptide. Peptides which form an alpha-helical structure are preferred and are often stabilised by salt bridges. Commonly used rigid linkers include repeating units of (EAAAK)n (Arai 2001). Another type of rigid linker includes repeating units (XP)n where X may be any amino acid but preferably Ala, Lys or Glu (George 2003).

In certain embodiments the linker polypeptide may comprise one or more naturally occurring protein domains, for example the lipoyl domain of lipoamide acetyltransferase (SEQ ID NO: 45) or a polypeptide homologue of a naturally occurring protein domain, for example modified Moloney murine leukaemia virus gag protein (SEQ ID NO: 47).

Because of their relative accessibility compared with other parts of the fusion polypeptide, linker polypeptides may be susceptible to cleavage by proteases. This may or may not be desirable. For example, a stable linker polypeptide may be required for favourable assembly of the particle of this invention and display of the membrane-spanning polypeptide. In such a case, the sequence of the linker polypeptide is preferably selected to avoid potential protease cleavage sites. Or for example, a cleavable linker polypeptide may be required in order that the membrane-spanning polypeptide be released and displayed in a particular (for example native) conformation. In such a case, the sequence of the linker polypeptide is preferably selected to introduce a desirable protease cleavage site.

In certain embodiments the linker polypeptide includes a furin cleavage site based on the recognition sequence Arg-Xaa-Arg*Lys-Arg, where Xaa represents any amino acid and * indicates the cleavage site. Furin is mainly localised in the trans-Golgi network and thus may be able to cleave such a linker polypeptide as it is made within the cell.

In certain embodiments the linker polypeptide includes a recognition site for a protease which is not normally found within the host cell. This may be a broad-specificity protease for example chymotrypsin, papain, plasmin, or trypsin or more preferably a protease with more restricted sequence specificity for example: factor Xa protease (11e-(Glu/Asp)-Gly-Arg*), enterokinase (enteropeptidase) light chain (Asp-Asp-Asp-Asp-Lys*), tobacco etch virus protease (Glu-AsnLeu-Tyr-Phe-GlnYSer), thrombin (Leu-Val-Pro-Arg*Gly-Ser) or HRV 3C protease (PreScission) (Leu-Glu-Val-Leu-Phe-Gln*Gly-Pro).

Polypeptide tags In certain embodiments the particle of this invention may contain one or more polypeptide tags, each of which may have different properties and purposes. Many such tags are well known in the art (Terpe 2003).

In certain embodiments the polypeptide tag is connected to the core polypeptide, either directly or by means of a linker polypeptide. Such a polypeptide tag is located within the lipid membrane and in the intact particle of this invention it will generally be inaccessible to polar substances. Such a polypeptide tag might be useful for example for targeting degradation products of the particle in vivo to a particular cellular compartment.

In certain embodiments the polypeptide tag is connected to or inserted within the membrane-spanning peptide, either directly or by means of a linker polypeptide in such a way that it is located on the outside of the lipid membrane of the particle of this invention. Such a polypeptide tag will be accessible to a range of polar substances, including antibodies, ligands and chemical reagents for protein modification and accordingly can be useful for different purposes as exemplified below.

In certain embodiments the polypeptide tag comprises an epitope that can be recognised by an antibody. This may be useful for affinity purification, detection or visualisation of the particles of the invention. Examples of antibody epitope tags include: CD52 mimotope (TSSPSAD), E-tag (GAPVPYPDPLEPR), Flag-tag (DYKDDDDK), HA-tag (YPYDVPDYA), His-tag (HHHHHH), Myctag (EQKLISEEDL), S-tag (KETAAAKFERQHMDS), Softag 1 (SLAELLNAGLGGS), Softag 3 (TQDPSRVG), Ty1-tag (EVHTNQDPLD), V5-tag (GKPIPNPLLGLDST), VSV-tag (YTDIEMNRLGK), Xpress tag (DLYDDDDK). These are SEQ ID NO: 10 to SEQ ID NO: 22.

In certain embodiments the polypeptide tag comprises a peptide which binds to a non-antibody ligand. These may also be useful for purification, detection or visualisation of the particles of the invention. Examples of such peptide tags include: Arg-tag (RRRRR), calmodulin-binding peptide (KRRWKKNFIAVSAANRFKKISSSGAL), Glu-tag (EEEEE), His-tag (HHHHHH), SBP-tag (M DEKTTGWRGGHVVEGLAGELEQLRARLEHHPQGQREP) Strep-tag II (WSHPQFEK). These are SEQ ID NO: 23 to SEQ ID NO: 28.

In certain embodiments the polypeptide tag comprises a protein which confers a useful property on the particle of the invention. Examples of protein tags and some of their possible uses include: biotin carboxyl carrier protein (BCCP), a protein domain biotinylated by biotin ligase enabling recognition by avidin or streptavidin; cellulose-binding domain for affinity purification; chitin-binding domain for affinity purification, glutathione-S-transferase, a protein which binds to immobilized glutathione allowing convenient purification by affinity chromatography; green fluorescent protein, a protein which is spontaneously fluorescent; maltose binding protein, a protein which binds to amylose agarose allowing convenient purification; Fc derived from immunoglobulin Fc domain allows dimerization and can be used for affinity purification.

In certain embodiments the polypeptide tag comprises a peptide which is amenable to post-translational modification, either in vivo or in vitro. Such tags may permit the addition of a useful chemical moiety to a specific site on the particle of the invention. Examples of tags which are amenable to post-translational modification and some of the possible modifications include: AviTag, a peptide allowing biotinylation by the enzyme biotin ligase enabling recognition by avidin or streptavidin (GLNDIFEAQKIEWHE); fluorescein-arsenical helix binder FlAsH for site-specific modification with a fluorochrome (CCXXCC); Isopeptag, a peptide which binds covalently to pilinC protein (TDKDMTITFTNKKDAE), SpyTag, a peptide which binds covalently to SpyCatcher protein (AHIVMVDAYKPTK). These are SEQ ID NO: 29 to SEQ ID NO: 32.

In certain embodiments any of the above polypeptide tags may be linked to a linker polypeptide which contains a protease cleavage site (for example factor Xa protease, enterokinase, tobacco etch virus protease, thrombin or HRV 3C protease) in order to facilitate removal of the tag when it is no longer required.

Nucleic acids In another aspect, this invention also provides nucleotide sequences encoding the fusion polypeptide of this invention. The polynucleotides of the invention may be obtained, and the nucleotide sequence of the polynucleotides determined, by any method known in the art. Preferably the codon usage in the nucleotide sequence is optimised to facilitate expression of the fusion polypeptide in the chosen host cell. Various methods for codon optimisation are known in the art (for example see US 8,825,411).

Vectors In another aspect, this invention also provides vectors that contain a polynucleotide encoding a fusion polypeptide of the invention. In an exemplary embodiment, a polynucleotide that encodes a fusion polypeptide may be incorporated into a vector in order to express particles of the invention in a suitable host cell. A variety of expression vectors may be used. They may comprise self-replicating extra-chromosomal vectors or vectors which integrate into a host genome. Expression vectors are constructed to be compatible with the host cell type. As is known in the art, a variety of expression vectors are available, commercially or otherwise, that may find use for expressing fusion polypeptides. Preferred expression vectors include those which enable expression of fusion polypeptides in eukaryotic cells, for example yeast, plant, insect or mammalian cells. Especially preferred expression vectors include those which enable expression of fusion polypeptides in mammalian cells, including human cells, for example the vector pSV-CMV (SEQ ID NO: 5). An example of expression vector pSV-CMV adapted for linking core polypeptides to the N-terminus of a membrane polypeptide is illustrated in Figure 7 (plasmid pAC1, SEQ ID NO: 1). An example of expression vector pSV-CMV adapted for linking core polypeptides to the C-terminus of a membrane polypeptide is illustrated in Figure 8 (plasmid pAC3, SEQ ID NO: 3).

Host cells In another aspect, this invention also provides host cells comprising a vector as described above which can be used for expression of a fusion polypeptide of the invention. If the fusion polypeptide is not glycosylated and does not contain disulphide bonds which are critical for its correctly folded structure, then suitable host cells may include prokaryotes such as Escherichia coli or Bacillus subti/is. For expression of a glycosylated or extensively disulphide-bonded fusion polypeptide, a eukaryotic host cell is preferable. Suitable eukaryotes may include yeast (for example Saccharomyces cerevisiae, Pichia pastoris), plants (for example Nicotiana species), insects (for example Spodoptera frugiperda) or mammalian cell lines. Preferred host cells include mammalian cell lines. Suitable mammalian cell lines available as hosts for expression of recombinant polypeptides are well known in the art and include immortalized cell lines available from the American Type Culture Collection (ATCC), including but not limited to Chinese hamster ovary (CHO) cells, HeLa cells, baby hamster kidney (BHK) cells, monkey kidney cells (COS), human hepatocellular carcinoma cells (Hep G2), and human embryonic kidney (HEK) 293 cells.

The vector is introduced into the host cell by transfection or transduction using any suitable technique known in the art. An example of transfection of HEK293 cells is provided in Example 6.

In certain embodiments, fusion polypeptides and particles of the invention are expressed in a cell line with stable expression of the fusion polypeptide. In other embodiments, fusion polypeptide and particles of the invention are expressed in a cell line with transient expression of the fusion polypeptide. The cell line, either stable or transiently transfected, is maintained in cell culture medium and conditions well known in the art resulting in the expression and production of particles of the invention.

Preparation of particles of the invention In another aspect, this invention also provides a method for preparation of particles of the invention. Host cells as described above are cultured to allow the expression of the fusion polypeptide under suitable conditions as are well known in the art (for examples see Vicente 2011, Zeltins 2013). General principles of particle purification are well known and include the following steps: (i) lysis of the cell to transfer the synthesized particles to the solution, (ii) clarification to remove cell debris and other large aggregates; (iii) concentration and (iv) purification to remove residual impurities. Preferably the particles of the invention are secreted into the culture medium in which case lysis of the cells is not necessary. In the event that the particles are retained within the cells, then a suitable method is required to obtain the cell-free extract. For example, eukaryotic (including mammalian and human) cells can be lysed with a suitable detergent. Bacterial, yeast or plant cells may require mechanical treatment such as mills, French press, ultrasonication or repeated freeze/thawing. A suitable method for a particular case may be determined by analysis of the yield and integrity of the particles in the lysate. To protect the particles from oxidation or proteolytic degradation during the purification process, the extraction buffer may be supplemented with reducing and chelating agents and protease inhibitors. Nucleases may be added at this stage to reduce the amount of nucleic acids in the extracts.

The cell extract or culture supernatant containing the particles may be clarified by low-speed centrifugation or filtration or by low-speed centrifugation followed by filtration. Preferably centrifugation conditions (time and g-force) are optimised to obtain a high yield of particles in the supernatant whilst pelleting a high proportion of cells and/or cellular debris. Filtration may be accomplished in one or more steps. A coarse (depth) filter may be used first to remove cells and large debris, followed by a filter with smaller pore size to remove cell fragments and small debris. A preferred pore size is selected with reference to the anticipated size of the particles to avoid their retention. A suitable pore size for a particular case may be determined by analysis of the yield and integrity of the particles in the filtrate.

The clarified cell extract or culture supernatant may be concentrated by any of a number of methods known in the art, for example ultrafiltration or precipitation with an agent such as ammonium sulphate or polyethylene glycol. Suitable conditions for a particular case may be determined by analysis of the yield and integrity of the particles in the concentrate.

Culture medium containing the product may be separated from host cells by centrifugation or by filtration or by centrifugation followed by filtration. Alternatively, if particles are retained within the cells they may be lysed and the product solubilised using methods known in the art.

The particles may be purified by methods known in the art for purification of viruses and virus-like particles, for example, by chromatography (for example ion exchange, affinity, and size exclusion chromatography), centrifugation, differential solubility, or by any other standard technique for the purification of proteins.

The preferred methods used for the final purification of the particles depend on their intended application and the scale of the process. For example, laboratory-scale purifications for experimental or research use may be accomplished using ultracentrifugation in density gradients of caesium chloride or sucrose. However, this may be relatively labour intensive. For industrial-scale processes other methods such as chromatography may be preferable. Depending on the particular properties of the particles, ion-exchange, size-exclusion or affinity chromatography may be useful. Size-exclusion chromatography is a generic method which may be used for the majority of different particles, but is less easy to scale than other methods. Incorporation of a suitable polypeptide tag into the particle may greatly facilitate the use of ion-exchange or affinity chromatography.

If the particles are to be used for the preparation of a pharmaceutical composition or a vaccine composition, then additional purification steps well known in the art may be required to remove actual or potential impurities or contaminants and to comply with regulatory requirements.

Compositions and medicaments for use in therapy In another aspect, this invention provides for pharmaceutical or vaccine compositions. In certain embodiments pharmaceutical or vaccine compositions comprise the particle of this invention. In other embodiments DNA vaccine compositions comprise the polynucleotide of the invention. Any of these compositions may also comprise a pharmaceutically acceptable diluent, excipient or carrier. Vaccine or DNA vaccine compositions may also comprise a pharmaceutically acceptable adjuvant.

In certain embodiments this invention provides for medicaments comprising the particle of this invention for use in therapy. For example, such medicaments may be used for immunomodulation, for prophylaxis or treatment of infectious disease, autoimmune disease, cancer or transplant rejection. Or for example, such medicaments may be used for vaccination, for prevention or treatment of infectious disease or cancer.

Particles of this invention may be particularly advantageous for the manufacture of pharmaceutical or vaccine compositions or medicaments for immunomodulation or vaccination compared with previously available substances because they are capable of displaying a single type of membrane-spanning polypeptide in an oligomeric particle of a suitable size for effective recognition by the immune system. Furthermore, depending on the particular host cell used for manufacture of the particles of this invention, it is possible to ensure that any post-translational modifications (for example glycosylation) are of a desirable (for example native) type or structure.

In certain other embodiments this invention provides for medicaments comprising the polynucleotide of this invention for use in therapy. Preferably the polynucleotide is comprised within a vector which is capable of causing expression of the fusion polypeptide and particles of the invention in vivo. Such medicaments may be used for vaccination, for example for the treatment of infectious disease or cancer.

In certain embodiments any of the core polypeptide, linker polypeptide or polypeptide tag comprised within the medicaments described above may in turn comprise one or more peptide sequences which can be processed by antigen-presenting cells and presented to CD4+ T cells to provide help for an antibody response against the membrane-spanning polypeptide comprised within the medicament.

Antibodies In another aspect, this invention provides for a method for producing antibodies directed against a transmembrane protein. Such antibodies may be very desirable as reagents for research, diagnosis or therapy. However, it was previously difficult to produce them due to the difficulty of obtaining sufficiently pure or sufficient amounts of purified transmembrane protein. Particles of this invention are particularly advantageous for immunisation compared with previously available substances because they are capable of displaying a single type of membrane-spanning polypeptide in an oligomeric particle of a suitable size for effective uptake by antigen-presenting cells, for example macrophages or dendritic cells. In addition, a suitable core polypeptide could provide T-cell help for the generation of an antibody response against the membrane-spanning polypeptide. In addition, particles of this invention which display a single membrane-spanning polypeptide in a physiological format are likely to be especially useful reagents for screening, detection and analysis of antibodies which recognise the membrane-spanning polypeptide.

In certain embodiments animals (for example llama, goat, sheep, rabbit, rat, mouse, hamster or chicken) are immunised with particles of this invention to raise polyclonal antibodies. Methods for immunisation, collection of sera (or eggs) and purification of polyclonal antibodies are well known in the art (for example Hanly 1995; Howard 2001).

In other embodiments animals (for example llama, rabbit, rat, mouse, hamster) are immunised with particles of this invention and antibody-producing cells (for example B-lymphocytes from the spleen) are immortalised to produce monoclonal antibodies. Methods for immunisation, collection of antibody-producing cells, immortalisation and cloning of cells and production of monoclonal antibodies are well known in the art (for example Shepherd 2000; Howard 2001; Ossipow 2014).

In certain embodiments any of the core polypeptide, linker polypeptide or polypeptide tag may comprise one or more peptide sequences which can be processed by antigen-presenting cells and presented to CD4+ T cells to provide help for any of the antibody responses described above.

In certain embodiments antibodies are selected from phage display libraries (produced for example from human lymphocytes) using particles of this invention for selection of the desired antibodies. Methods for production of antibodies from phage display libraries are well known in the art (for example Winter 1994; Barbas 2001).

In certain embodiments polyclonal antibodies, or more preferably monoclonal antibodies raised and/or selected using particles of this invention may be used in the manufacture of pharmaceutical compositions and medicaments for the treatment of disease including for example, cancer, autoimmune disease, transplant rejection, infectious disease, cardiovascular disease, neurological disease.

EXAMPLES

The examples below are given so as to illustrate the practice of this invention. They are not intended to limit or define the entire scope of this invention.

Example 1 Selection of candidate core polypeptides by searching primary literature The PubMed database (http://www.ncbi.nlm.nih.gov/pubmed) was used to search for primary scientific literature reporting proteins that self-assemble into multimeric structures greater than 20 monomers in size with evidence for self-assembly of the protein following expression in prokaryotic and/or eukaryotic cells. Protein sequences were retrieved using the UniProt database (http://www.uniprot.org/). Non-exclusive lists of candidates are shown in Table 1 and Table 2.

Table 1 Candidates for core polypeptides identified by searching primary literature.

Genbank Entry UniProt Entry Name Reference CAZ61323 F8J354 Aspergillus fumigatus partitivirus-1 coat protein Ochoa 2008 CAA41364 Q65664 brome mosaic virus KU-1 coat protein Kao 2011, Moon 2014 AFU48534.1 K0IB68 barley yellow dwarf virus PAV coat protein nal BAA02061.1 P18027 cucumber mosaic virus Y coat protein Moon 2014 NP 056749.1 Q76QV9 potato leaf roll virus coat protein Lamb 1996 223380 P03606 satellite tobacco necrosis virus coat protein Lane 2011 NP_062899.1 P11689 tomato bushy stunt virus coat protein Kumar 2009 WP_010880027.1 066529 Aquifex aeolicus lumazine synthase Ladenstein 2013 WP_000269761.1 P0A998 Escherichia coli ferritin Kim 2011 NP_001922.2 P10515 human pyruvate dehydrogenase E2 (lipoamide acetyltransferase) Zhou 2001 selected because of its similarity to potato leaf roll virus Table 2 Candidates for core polypeptides identified by searching primary literature.

Genbank Entry UniProt Entry Name Reference AAB47793 P07299 Human parvovirus B19 Rueda 1999 Q9IVZ8.1 Q9IVZ8 Hepatitis E Li 2005 NP 056821.2 Q83884 Norwalk virus Debbink 2013 AAM21587.1 Q8JZ14 Rabbit haemorrhagic disease virus Luque 2012 P03090.3 P03090 Murine polyomavirus Simon 2014 P04012.1 P040121 Human papillomavirus Mossadegh 2014 Example 2 Selection of candidate core polypeptides by searching the Protein Data Bank The Protein Data Bank (http://www.rcsb.org/pdb) was searched for homomeric multi-subunit proteins with octahedral or icosahedral symmetry, other than virus capsids. The three-dimensional models were retrieved as.pdb files and viewed using Discovery Studio Client 3.5 Visualizer software (Accelrys Software Inc). The position and accessibility of the N-and C-termini of the proteins was assessed. Where at least one terminus appeared to be accessible on the outside of the particle, the protein was considered as a potential candidate for a core polypeptide. A non-exclusive list of candidates is shown in Table 3.

Table 3 Candidates for core polypeptides identified by searching the Protein Data Bank PDB Name Subunits Diameter (nm) Exposed Exposed entry N-terminus C-terminus lfha Human ferritin 24 15 Thr-5 inner surface 1hqk Aquifex aeolicus lumazine synthase 60 16 Met-1 Arg-154 1b5s Bacillus stearothermophilus dihydrolipoyl transacetylase (E2), catalytic domain 60 24 Ala-184 Met-426 3dkt Thermotoga maritima encapsulin 180 24 inner Lys-267 surface 4pt2 Myxococcus xanthus encapsulin 180 32 inner Leu-238 surface Based on particle size and apparent accessibility of both the N-and C-termini, the E2 catalytic domain appeared to be a good potential candidate for a core polypeptide.

Example 3 Selection of candidate core polypeptides by searching the VIPER database The VIPER database (http://viperdb.scripps.edu) contains structural data determined by X-ray crystallography or cryo-electron microscopy for a large number of icosahedral viruses. Using the database browser provided on the VIPER website, a set of data was extracted which included the following parameters: PDB entry, virus name, genus, family, number of protein chains, minimum and maximum radius of the capsid, radius and solvent-accessible surface area in the intact capsid (sasa_bound) of the endmost amino acid residues. Data for a total of 507 structures were extracted. (The endmost amino acid residues in the database may not always be the N-and C-termini of the protein chains since these are sometimes not visualised in the three-dimensional structure.) The dataset was loaded into a Microsoft Excel spreadsheet and the relative solvent-accessible surface area (sasa_bound%) was calculated for each amino acid residue as a percentage of the average solvent-accessible surface area for the corresponding reference amino acid (sasa_reference) in the VIPER database. The dataset was then filtered to identify viral capsid proteins which might provide candidate core polypeptides. The filter criteria were as follows: * No more than one type of protein chain in the assembly * sasa_bound% greater than 50% * radius of residue within 30 Angstroms of the maximum radius of the capsid It is possible to use more or less stringent filter criteria to select a smaller or larger list of candidates. For example, the terminal NH2 or COOH of a protein chain might be sufficiently exposed even if sasa_bound% for the residue were less than or equal to 50%. Or an exposed terminal residue might be adequately accessible on the edge of a hole in the virus capsid even though it might be more than 30 Angstroms from the boundary sphere enclosing the capsid.

The three-dimensional models of each candidate were retrieved as.pdb files and viewed using Discovery Studio Client 3.5 Visualizer software (Accelrys Software Inc). The position and accessibility of the N-and C-termini of the proteins was assessed. Where there were several similar structures for a particular virus, or virus family, then a representative was selected which appeared to have the most accessible NH2 and/or COOH group at the ends of the protein chain. From the reduced short-list, capsids were selected with a maximum radius in the range 125 to 250 Angstroms (equivalent to a diameter of 25 to 50 nm).

A non-exclusive list of the final set of candidates is shown in Table 4.

Table 4 Candidates for core polypeptides identified by searching the VIPER database VDB entry Name Subunits Diameter (nm) Exposed Exposed N-terminus C-terminus upana1503 Newton's T=3 capsid 180 28 His-1 Arg-104 1 gav Bacteriophage GA 180 28 Ala-1 Ala-130 ldwn Bacteriophage PP7 180 30 Ala-1 Arg-127 2izw Ryegrass mottle virus 180 31 buried Gln-235 lauy Turnip yellow mosaic virus 180 31 buried Thr-189 2zah Melon necrotic spot virus 180 34 buried Ala-390 2e0z Pyrococcus furiosus viruslike protein 180 38 buried Glu-345 1ihm Norwalk virus 180 40 buried Ser-520 Example 4 Selection of additional candidate core polypeptide Among the candidates identified in the VIPER database (Example 3) were a number of bacteriophage from the levivirus family, including bacteriophage MS2, FR, GA, Qbeta, PRR1 and PP7. These all have homologous structures (Tars 1997) with N-and C-termini close to the outer surface of the capsid. Previous workers have been able to display short peptides inserted into an exposed loop of such capsids, including MS2 and PP7 (Caldeira 2010) although successful insertion of desired epitopes remained unpredictable (Tissot 2010) and creation of fusion proteins linked at the N-or C-terminus has not been reported. Bacteriophage AP205 is an RNA bacteriophage which infects Acinetobacter species and has little sequence similarity with other RNA phages. Its three-dimensional structure has not been reported but it has been shown that fusion proteins having up to 55 amino acid residues attached to either the N-or C-terminus of the coat protein can be expressed and assembled into VLPs (Tissot 2010; EP2351770). Therefore, despite the lack of structural information, bacteriophage AP205 may be considered as another candidate core polypeptide.

Example 5 Synthesis of vectors containing genes coding for model transmembrane proteins as a tool for screening candidate core polypeptides Acceptor vectors containing inserts coding for model transmembrane polvpeptides with a HIS tag A set of acceptor vectors pAC1 to pAC4 was engineered by insertion of specific coding sequences into the mammalian expression vector pSF-CMV (Oxford Genetics) so that fusion proteins could be created between a desired core polypeptide and a membrane polypeptide (Table 5). Vector pSF-CMV (SEQ ID NO: 5) contains a multiple cloning site which includes unique sites for the restriction enzymes Not1 (857/859), HindlIl (869/870), EcoR1 (885/886), Xba1 (937/938) and Nhe1 (1002/1003).

Table 5 Acceptor Vectors Plasmid For fusion to Transmembrane domain MMLV gag late domains DNA sequence of insert pAC1 C-terminus SEQ ID NO: 38 n/a SEQ ID NO: 1 pAC2 C-terminus SEQ ID NO: 38 SEQ ID NO: 40 SEQ ID NO: 2 pAC3 N-terminus SEQ ID NO: 42 No SEQ ID NO: 3 pAC4 N-terminus SEQ ID NO: 42 SEQ ID NO: 40 SEQ ID NO: 4 Acceptor vectors for joining the N-terminus of the membrane polypeptide to the C-terminus of a core polypeptide Vectors pAC1 and pAC2 were designed for fusion of the N-terminus of a membrane polypeptide to the C-terminus of a core polypeptide. They contained DNA coding for: (a) a 12 amino-acid flexible Gly-Ser linker (SEQ ID NO: 33), (b) the human transferrin receptor 1 (Genbank entry NP_003225.2) transmembrane domain, amino acid residues 65-89 (SEQ ID NO: 38) flanked on both sides by the amino acid pair Lys-Glu (c) a second 12 amino-acid flexible Gly-Ser linker (SEQ ID NO: 34), (d) a 10 amino-acid His tag (SEQ ID NO: 39), (e) a stop codon. Vector pAC2 also contained an additional Gly-Ser linker (SEQ ID NO: 35) and a modified sequence derived from the late domains of the Moloney murine leukemia virus (MMLV) gag protein (Genbank entry NP_057934) (amino acids 98-170) upstream of the N-terminal flexible linker and the transmembrane domain (SEQ ID NO: 47).

DNA inserts for vectors pAC1 and pAC2 were codon optimized for expression in human cells (SEQ ID NO: 1 and SEQ ID NO: 2). The DNA was synthesized and provided in pBSK (pBSKSequencel and pBSK-Sequence2). It was digested with HindlIl and Xbal, and the resulting 0.2kb or 0.4kb fragment was ligated to the 4.2kb fragment of pSF-CMV digested with HindIll and Xbal, resulting in pAC1 (Figure 5) and pAC2 respectively.

Acceptor vectors for joining the C-terminus of the membrane polypeptide to the N-terminus of a core polypeptide Vectors pAC3 and pAC4 were designed for fusion of the C-terminus of a membrane polypeptide to the N-terminus of a core polypeptide. They contained DNA coding for: (a) human serum albumin (Genbank entry AAQ89947.1) signal peptide, residues 1-18 (SEQ ID NO: 41), (b a 10 amino-acid His tag (SEQ ID NO: 39), (c) a 7 amino-acid flexible Gly-Ser linker (SEQ ID NO: 36), (c) the negatively charged amino acid cluster Asp-Gly-Glu-Glu-Gly-Gly-Gly, (d) the human influenza virus H3N2 (Genbank PAA43200,1) haemagglutinin (HA) transmembrane domain, residues 163-183 (SEQ ID NO: 42), (e) a cluster of positively charged amino acids Gly-Gly-GlyArg-Gly-Arg-Lys and (f) a flexible 8 amino acid glycine-serine linker (SEQ ID NO: 37). Vector pAC4 also contained an additional Gly-Ser linker (SEQ ID NO: 33) and a sequence derived from the late domains of the Moloney murine leukemia virus (MMLV) gag protein (Genbank entry NP 057934) (amino acids 98-170) including the PTAP/PSAP and PPxY motifs downstream of the transmembrane domain.

DNA inserts for vectors pAC3 and pAC4 were codon optimized for expression in human cells (SEQ ID NO: 3 and SEQ ID NO: 4). The DNA was synthesized and provided in pBSK (pBSKSequence3 and pBSK-Sequence4). It was digested with Notl and EcoRl, and the resulting 0.25kb or 0.45kb fragment was ligated to the 4.3kb fragment of Notl and EcoRl digested pSF-CMV, resulting in pAC3 (Figure 6) and pAC4 respectively.

DNA coding for core polvpeptides The coding sequence of selected core polypeptides (for example as listed in Table 1, Table 2, Table 3Table 4) were codon optimized for expression in human cells. To the 5' end of each coding sequence (starting with ATG) the nucleotides GCGGCCGCGGAATTCTTCCCACC (SEQ ID NO: 8) were added to provide both a Notl and an EcoRl restriction enzyme site and a consensus Kozak sequence. At the 3' end of each coding sequence the naturally occurring stop codon was replaced by the nucleotides GGAAGCTTGTAGTGAGCTAGC (SEQ ID NO: 9) to provide a HindlIl site, two stop codons and an Nhe1 site. The complete sequence was synthesized and provided in the pBSK cloning vector (Agilent Technologies, catalogue number 212205).

Expression vectors Expression vectors pAC1-E2 and pAC2-E2 were constructed by digesting acceptor vectors pAC1 or pAC2 and the DNA coding for the E2 core polypeptide (SEQ ID NO: 6) with the restriction enzymes Notl and Hindil and ligating the resulting coding fragments with the result that the sequence coding for the core polypeptide was inserted upstream of the sequence coding for the membrane polypeptide.

Expression vectors pAC3-E2 and pAC4-E2 were constructed by digesting acceptor vectors pAC3 or pAC4 and the DNA coding for the E2 core polypeptide (SEQ ID NO: 6) with the restriction enzymes EcoRl and Nhel and ligating the resulting coding fragments with the result that the sequence coding for the core polypeptide was inserted downstream of the sequence coding for the membrane polypeptide.

Expression vectors for other core polypeptides were constructed in the same way.

Expression vectors for fusion proteins tagged with green fluorescence protein (GFP) were prepared in a similar way except that a gene coding for GFP (Genbank entry AAF62891.1) (SEQ ID NO: 44) was substituted for the HIS tag.

Transfection grade plasmid vectors were produced using Qiagen Plasmid Midi Kit. Purity was assessed by measurement of the ratio of absorbance at 260 and 280 nm (A260/A280).

Example 6 Transfection and expression of particles in HEK-293FT cells Cell culture HEK-293FT cells (Life Technologies) were grown in Freestyle293 Expression Medium (Life Technologies) in suspension on a platform shaker in a humidified 37°C CO2 incubator with rotation at 140 rpm. Cells were maintained between 4 x 105 and3 x 106 cells/mL in a volume not exceeding 20% of the total volume of the culture flask. Cultures were maintained for at least three passages to ensure stable growth prior to performing transfections.

Stock solutions of reagents A stock solution of linear polyethyleneimine (PEI) molecular weight 25,000 (Polysciences) was prepared at 1 mg/mL in 25mM HEPES pH 7.5 containing 150 mM NaCI. The PEI was added to the buffer and vortex mixed until completely dissolved. Valproic acid (VPA) (Sigma-Aldrich) was diluted in water to a concentration of 0.5M. The reagents were sterile filtered, aliquotted and stored frozen at -20°C.

Transfection HEK-293FT cells at a viability of >95% were pelleted by centrifugation at 300 x g for 10 min and resuspended in fresh Freestyle293 Expression Medium to give a density of 3 x 106 cells/mL. Purified DNA of the desired expression vector (with A260/A280 ratio greater than 1.8) was added to the cells to give a final concentration of 3 pg/mL. The cells were incubated for 5 min under shaking, and then PEI was added to give a final concentration of 9 pg/ml. The cells were incubated for 24 h then diluted with an equal volume of fresh medium containing 8 mM VPA, 100 U/mL penicillin and 100 pg/mL streptomycin.

Expression and harvest: Four days after the transfection, the cells were removed by centrifugation at 4000 x g. The culture supernatant was passed through a 0.22 pm filter and analysed by ELISA or Western blotting, or stored at -80°C. Optionally, the culture supernatant was concentrated up to 100 fold by tangential flow filtration using a 300 kDa cut-off membrane. The concentrated culture supernatant was stored frozen at -80°C.

Example 7 Purification of particles containing a HIS tag HEK-293FT cells were transfected with an expression vector (Example 5) and culture supernatant was collected as described in Example 6. The particles were purified by affinity chromatography using a cobalt affinity column. (HiTrap Talon crude from GE Healthcare cat no 28-9537-66). A 1 mL Talon column was equilibrated with phosphate buffered saline (PBS) pH 7.4 (Life Technologies cat no 10010-015), and 50 mL of concentrated culture supernatant was applied at a flowrate of 1 mL per min. The column was washed with 20 mL of PBS, and particles were eluted with 5 mL of 100 mM imidazole in PBS. The column was washed with 20 mL of PBS and the process was repeated 3 times. Eluates were pooled and analysed by SDSpolyacrylamide gel electrophoresis alongside a set of molecular weight markers. The gel was stained with Coomassie brilliant blue.

When this was done using an expression vector coding for the fusion protein [ferritin-linkermembrane-spanning domain-linker-his tag] (SEQ ID NO: 43) there was a single stained band in the eluate samples with an apparent molecular weight of about 25 kDa closely corresponding to the molecular weight of 26 kDa predicted for the fusion polypeptide.

Example 8 Purification of particles having a net positive surface charge Particles having a net positive surface charge are purified by cation exchange chromatography on carboxymethyl Sepharose (HiTrap CM FF from GE Healthcare cat no 17-5056-01) optionally followed by size exclusion chromatography on hydroxylated methacrylic polymer (Toyopearl HW65S from Tosoh Bioscience cat no 07467). A 1 mL CM FF column is equilibrated in binding buffer (eg 20 mM sodium phosphate, pH 7.4) and culture supernatant or concentrated culture supernatant is applied at a flow rate of 1 mL per min. The column is washed with binding buffer until the absorbance of the eluate at 280 nm is stable. The particles are eluted with a linear gradient of binding buffer containing from 0 to 1 M sodium chloride over approximately 20 mL. Fractions of approximately 1 mL are collected. Samples of each fraction are analysed by SDS polyacrylamide gel electrophoresis and the fractions containing protein bands of the predicted molecular weight are pooled. Optionally the particles may be further purified by size exclusion chromatography. An XK 16/100 column (GE Healthcare cat no 28-9889-47) is packed with approximately 189 mL of Toyopearl HW-65S and equilibrated with phosphate-buffered saline. The pooled eluate from the cation exchange column is applied and eluted with phosphate-buffered saline at 1.5 mUmin. Fractions of approximately 1 mL are collected and analysed by SDS polyacrylamide gel electrophoresis as before. Fractions containing protein bands of the predicted molecular weight are pooled.

Example 9 Purification of particles having a net negative surface charge Particles having a net negative surface charge are purified by anion exchange chromatography on diethylaminoethyl (DEAE) Sepharose (HiTrap DEAE FF from GE Healthcare cat no 17-505501) optionally followed by size exclusion chromatography on hydroxylated methacrylic polymer (Toyopearl HW-65S from Tosoh Bioscience cat no 07467). A 1 mL DEAE FF column is equilibrated in binding buffer (eg 20 mM triethanolamine hydrochloride, pH 7.4) and culture supernatant or concentrated culture supernatant is applied at a flow rate of 1 mL per min. The column is washed with binding buffer until the absorbance of the eluate at 280 nm is stable. The particles are eluted with a linear gradient of binding buffer containing from 0 to 1 M sodium chloride over approximately 20 mL. Fractions of approximately 1 mL are collected. Samples of each fraction are analysed by SDS polyacrylamide gel electrophoresis and the fractions containing protein bands of the predicted molecular weight are pooled. Optionally the particles may be further purified by size exclusion chromatography as described in Example 8.

Example 10 SDS-PAGE and Western blotting HEK-293FT cells were transfected with vectors encoding various fusion polypeptides (Example 5) and culture supernatant was collected as described in Example 6.

Samples of the cells and culture supernatant were analysed by SDS polyacrylamide gel electrophoresis. 20 yL of cell suspension was centrifuged for 1 min at 10,000 x g and the pellet was resuspended in 50 pL of NuPAGE LDS sample buffer (Life Technologies cat no NP0007) containing 2.5 pL XT reducing agent (Biorad cat no 161-0792). 14 yL of supernatant was mixed with 5 pL of NuPAGE LDS sample buffer and 1 pL of XT reducing agent. The samples were incubated at 95°C for 5 min then loaded onto a Novex NuPAGE Bis-Tris 4-12% Midi polyacrylamide gel (Life Technologies cat no NP0321) and electrophoresed at 200V for approximately 45 min. Proteins were blotted onto nitrocellulose using Transblot Turbo membranes (Bio-Rad cat no 1704159) and a Transblot Turbo blotting system (Bio-Rad). Membranes were washed in PBS containing 0.1% Tween 20 (PBS-T) for 5 min. All steps were conducted under gentle rocking at room temperature. The membranes were blocked for 1 h in PBS-T containing 5% dried non-fat milk (blocking buffer) and then incubated for 1 h in blocking buffer containing 1 pg/mL mouse antihistidine-tag monoclonal antibody (Serotec cat no MCA 1396). The membrane was washed 3 times for 5 min with PBS-T then incubated for 1 h in blocking buffer containing 1 pg/mL peroxidase-labelled goat anti-mouse-19G (Serotec cat no 103005). The membrane was washed 3 times for 5 min with PBS-T and developed using TMB membrane peroxidase substrate (Insight Biotechnology cat no 50-77-18 KPL). Results from a representative experiment are shown in Table 8.

The apparent molecular weight of the principal bands was compared with the theoretical molecular weight of the fusion polypeptide construct calculated from its amino acid sequence. A sample was considered positive if a band was clearly visible in the culture supernatant with an apparent molecular weight within ± 5 kDa of the theoretical molecular weight of the fusion polypeptide, except for polypeptides containing a lipoyl acetytransferase (E2) core where, due to its unusual structure, the apparent molecular weight by SDS-PAGE exceeds the true molecular weight by approx. 15 kDa (Stephens 1983) and so an appropriate correction was made.

Example 11 ELISA

HEK-293FT cells were transfected with vectors encoding various fusion polypeptides (Example 5) and culture supernatant was collected as described in Example 6.

pL of filtered supernatant was mixed with 100 pL of a 200 mM sodium bicarbonate pH9.4 (coating buffer) on Maxisorp plates (Nunc) and incubated for 1 h. All incubations were carried out at room temperature. The plate was washed three times with PBS containing 0.1% Tween 20 (PBS-T) and blocked with 50 pL/well of coating buffer containing 2% bovine serum albumin (BSA) for 1 h. The plate was washed three times with PBS-T and incubated for 1 h with PBS-T containing 0.5 pg/ml mouse anti-histidine-tag monoclonal antibody. The plate was washed three times with PBS-T and incubated for 1 h with PBS-T containing a 1:2000 dilution of peroxidaselabelled goat anti-mouse-IgG. The plate was washed three times with PBS-T and developed using TMB microwell peroxidase substrate (Insight Biotechnology cat no 53-00-01). The absorbance at 450 nm (A450) was measured using an Infinite F50 microplate reader (Tecan).

Example 12 Analysis of particles by electron microscopy Particles containing the fusion protein [ferritin-linker-membrane-spanning domain-linker-his tag] (SEQ ID NO: 43) purified as described in Example 7 were immobilised on coated copper grids and stained with 2% uranyl acetate for 1 min and washed three times with ultrapure water. After drying the grids for 10 min they were analysed using a A-7650 tomographic transmission electron microscope (Hitachi). Approximately spherical particles with diameters in the range 17.3 nm to 19.3 nm could be clearly seen, consistent with the size of a ferritin molecule (11 nm) surrounded by a lipid membrane.

Example 13 Screening candidate core polypeptides by ELISA Candidate core polypeptides were identified by literature review as described in Example 1. An experiment was carried out to determine which of the candidates might be capable of supporting the expression of a VLP when fused either at the N-terminus or the C-terminus to a membrane-spanning polypeptide. For this experiment a model membrane-spanning polypeptide was used which contained a single membrane-spanning domain and a HIS polypeptide tag to facilitate detection, similar to that illustrated in Figure 1.

HEK-293FT cells were transfected with vectors encoding various fusion polypeptides (Example 5) containing candidate core polypeptides as listed in Table 1. Culture supernatant was collected as described in Example 6 and tested by ELISA as described in Example 11. The results are shown in Table 6.

Table 6 Expression of HIS-tagged fusion proteins in culture supernatant of transfected cells Plasmid Core polypeptide Orientation N to C MMLV gag A450 none N/A N/A N/A 0.043 pFCK5 Aquifex aeolicus lumazine synthase TM-CORE 0.012 pFCK6 Aquifex aeolicus lumazine synthase TM-CORE + 0.007 pFCK7 Aquifex aeolicus lumazine synthase CORE-TM - 0.011 pFCK8 Aquifex aeolicus lumazine synthase CORE-TM + 0.005 pFCK9 Escherichia coli ferritin TM-CORE - 0.005 pFCK10 Escherichia coil ferritin TM-CORE + 0.005 pFCK11 Escherichia coli ferritin CORE-TM 0.463 pFCK12 Escherichia coli ferritin CORE-TM + 0.026 pFCK13 Human lipoamide acetyltransferase (E2) TM-CORE 0.014 pFCK14 Human lipoamide acetyltransferase (E2) TM-CORE + 0.007 pFCK15 Human lipoamide acetyltransferase (E2) CORE-TM 0.430 pFCK16 Human lipoamide acetyltransferase (E2) CORE-TM + 0.514 pFCK17 Aspergillus fumigatus partitivirus coat protein TM-CORE - 0.003 pFCK18 Aspergillus fumigatus partitivirus coat protein TM-CORE + 0.004 pFCK19 Aspergillus fumigatus partitivirus coat protein CORE-TM - 0.012 pFCK20 Aspergillus fumigatus partitivirus coat protein CORE-TM + 0.006 pFCK21 Brome mosaic virus coat protein TM-CORE 0.007 pFCK22 Brome mosaic virus coat protein TM-CORE + 0.007 pFCK23 Satellite tobacco necrosis virus coat protein TM-CORE 0.008 pFCK24 Satellite tobacco necrosis virus coat protein TM-CORE + 0.010 pFCK25 Tomato bushy stunt virus coat protein TM-CORE - 0.004 pFCK26 Tomato bushy stunt virus coat protein TM-CORE + 0.027 pFCK27 Barley yellow dwarf virus coat protein CORE-TM - 0.009 pFCK28 Barley yellow dwarf virus coat protein CORE-TM + 0.010 pFCK29 Potato leafroll virus coat protein TM-CORE - 0.010 pFCK30 Potato leafroll virus coat protein TM-CORE + 0.013 pFCK31 Potato leafroll virus coat protein CORE-TM 0.008 pFCK32 Potato leafroll virus coat protein CORE-TM + 0.007 pSF6 GFP (negative control) 0.012 pSF28 Dengue virus NS1 (positive control) 0.776 An absorbance (A450) of greater than 0.05 was taken to indicate the presence of HIS-tagged protein in the culture supernatant, indicating that the cell was secreting VLPs containing the fusion protein. Plasmid pSF6 coding for intracellular GFP served as a negative control and plasmid pSF28 coding for HIS-tagged Dengue virus NS1 served as a positive control. They gave the expected results. Negative results were obtained with constructs containing core polypeptides derived from virus capsids. Positive results were obtained with constructs containing core polypeptides derived from E coli ferritin and human lipoamide aceytltransferase (E2) when the fusion polypeptide had the orientation NH2-Core-Transmembrane-COON, but not with the opposite orientation NH2-Transmembrane-Core-COOH. The presence of a spacer polypeptide between the core and the membrane-spanning domain obliterated the signal with the ferritin core but not the E2 core.

Three-dimensional models of the assembled proteins were retrieved as.pdb files from the Protein Data Bank or the VIPER database as described in Example 2 and Example 3 and viewed using Discovery Studio Client 3.5 Visualizer software to assess the position and accessibility of the Nand C-termini. If a structure was not available then a homologue was examined if possible. (No structures were found for luteoviridae such as barley yellow dwarf virus or potato leafroll virus.) The results are shown in Table 7.

Table 7 Exposure of terminal residues in three-dimensional structures PDB Name Subunits Diameter Exposed Exposed entry (nm) N-terminus C-terminus 1hqk Aquifex aeolicus lumazine synthase 60 16 Met-1 Arg-154 leum Escherichia coli ferritin 24 12 Leu-2 inner surface 1b5s Bacillus stearothermophilus dihydrolipoyl transacetylase (E2), catalytic domain (homologue of human E2) 60 24 Ala-184 Met-426 3iym Penicillium stoloniferum partitivirus coat protein (homologue of Aspergillus fumigatus partitivirus coat protein) 120 38 inner surface inner surface 1js9 Brome mosaic virus coat protein 180 28 inner surface buried 2buk Satellite tobacco necrosis virus coat protein 60 20 inner surface buried 1f15 Cucumber mosaic virus coat protein 180 30 inner surface buried 2tbv Tomato bushy stunt virus coat protein 180 35 inner Leu-387 surface In the light of this information the negative results obtained with core polypeptides from A fumigatus partitivirus, brome mosaic virus, satellite tobacco ncecrosis virus and cucumber mosaic virus could have been predictable. However, the negative results with fusions to the N-terminus of E coli ferritin, human E2, the C-terminus of tomato bushy stunt virus and both termini of A aeolicus lumazine synthase could not have been predicted and the positive result obtained with a fusion to the C-terminus of E coli ferritin was highly surprising considering that it is found on the inside of the assembled protein shell.

Example 14 Comparison of different polypeptide tags HEK-293FT cells were transfected with vectors encoding fusion polypeptides (Example 5) containing either ferritin or lipoamide acetyltransferase (E2) core polypeptide, membrane-spanning domains and either polyhistidine (HIS) or green fluorescent protein (GFP) polypeptide tag fused to the C-terminus of the core proteins. Culture supernatant was collected as described in Example 6 and tested by SDS-gel electrophoresis and Western blotting as described in Example 10. The results are shown in Table 8.

Table 8 Visualisation of fusion proteins by SDS-PAGE and Western blotting Plasmid Sample type Core Tag Predicted molecular weight Band polypeptide seen? pFCK11 lysate ferritin HIS 26 kDa 25 kDa pFCK16 lysate E2 HIS 73 kDa 80 kDa pFCK11 supernatant ferritin HIS 26 kDa 25 kDa pFCK16 supernatant E2 HIS 73 kDa 80 kDa pFCK35 lysate ferritin GFP 51 kDa 48 kDa pFCK36 lysate E2 GFP 99 kDa 115 kDa pFCK35 supernatant ferritin GFP 51 kDa 48 kDa pFCK36 supernatant E2 GFP 99 kDa 115 kDa Specific protein bands of the predicted molecular weights were detected in the supernatants of all transfectants, indicating that either HIS or GFP may be used as a polypeptide tag with either E. coli ferritin or human lipoamide acetyltransferase (E2) may be used as a core polypeptide.

Example 15 Screening candidate core polypeptides by Western blotting A second set of candidate core polypeptides were identified by literature review as described in Example 1. They were screened as described in Example 13 except that the supernatant samples were tested by SDS-PAGE and Western blotting instead of by ELISA.

HEK-293FT cells were transfected with vectors encoding various fusion polypeptides (Example 5) containing candidate core polypeptides as listed in Table 2. All constructs included a HIS tag. Culture supernatant was collected as described in Example 6 and tested by SDS-gel electrophoresis and Western blotting as described in Example 10. The results for the supernatant samples are shown in Table 9.

Table 9 Visualisation of fusion proteins by SDS-PAGE and Western blotting Plasmid Core polypeptide Orientation N to C MMLV gag Predicted molecular weight Band seen? pFCK139 human parvovirus B19 VP2 core-TM 67 kDa (65 kDa)1 pFCK140 human parvovirus B19 VP2 core-TM + 74 kDa (74 kDa) pFCK141 human parvovirus B19 VP2 TM-core 67 kDa None pFCK142 human parvovirus B19 VP2 TM-core + 74 kDa None pFCK144 heptatitis E virus ORF2 core-TM + 67 kDa 70 kDa pFCK145 heptatitis E virus ORF2 TM-core - 60 kDa None pFCK146 heptatitis E virus ORF2 TM-core + 67 kDa None pFCK147 Norwalk virus VP1 core-TM 70 kDa 65 kDa pFCK148 Norwalk virus VP1 core-TM + 77 kDa None pFCK149 Norwalk virus VP1 TM-core 70 kDa 65 kDa pFCK150 Norwalk virus VP1 TM-core + 77 kDa None pFCK151 Rabbit haemorrhagic disease virus VP60 core-TM - 67 kDa 70 kDa pFCK152 Rabbit haemorrhagic disease virus VP60 core-TM + 74 kDa None pFCK153 Rabbit haemorrhagic disease virus VP60 TM-core - 67 kDa 70 kDa pFCK155 murine polyoma virus VP1 core-TM 49 kDa None pFCK156 murine polyoma virus VP1 core-TM + 56 kDa (60 kDa) pECK157 murine polyoma virus VP1 TM-core - 49 kDa None pFCK158 murine polyoma virus VP1 TM-core + 56 kDa None pFCK160 human papilloma virus 11 L1 core-TM + 69 kDa None pFCK161 human papilloma virus 11 L1 TM-core - 62 kDa None pECK162 human papilloma virus 11 L1 TM-core + 69 kDa None 1 Results in brackets indicate that only a weak band was seen.

Very weak or negative results were obtained for core polypeptides derived from human parvovirus, murine polyoma virus and human papilloma virus. Positive results were obtained for hepatitis E virus in the orientation NH2-Core-Spacer-Transmembrane-COOH and for Norwalk virus and rabbit haemorrhagic disease virus in either orientation provided that the MMLV gag spacer polypeptide was omitted. This confirmed the observation in Example 13 that some core polypeptides (eg E2, hepatitis E) could tolerate the presence of a MMLV gag spacer but others (eg ferritin, Norwalk virus, rabbit haemorrhagic disease virus) could not.

Three-dimensional models of the assembled proteins were retrieved as.pdb files from the Protein Data Bank or the VIPER database as described in Example 2 and Example 3 and viewed using Discovery Studio Client 3.5 Visualizer software to assess the position and accessibility of the Nand C-termini. The results are shown in Table 10.

Table 10 Exposure of terminal residues in three-dimensional structures PDB Name Subunits Diameter (nm) Exposed Exposed entry N-terminus C-terminus 1s58 human parvovirus B19 VP2 60 27 inner surface buried 2ztn hepatitis E virus ORF2 60 28 inner surface Ala-606 3iyo hepatitis E virus ORF2 180 41 inner surface Ala-606 lihm Norwalk virus VP1 180 40 Asp-29 buried Ser-520 3jip Rabbit haemorrhagic disease virus VP60 180 55 Asn-45 Leu-569 inner surface lsid murine polyoma virus VP1 360 31 inner surface Gly-3821 1dzI human papilloma virus 11 L1 60 32 Lys-201 inner surface 110t human papilloma virus 11 L1 360 60 Met-1 inner surface Residues appear to be partly exposed on the edge of a hole in the capsid In the light of this information the weak or negative results obtained with core polypeptides from human parvovirus, murine polyoma virus and human papilloma virus could have been predictable. Likewise, the positive results with fusions to the C-terminus of core polypeptides from hepatitis E, Norwalk virus and rabbit haemorrhagic disease virus were not unexpected.

However, the positive results obtained with a fusion to the N-terminus of core polypeptides from Norwalk virus and rabbit haemorrhagic disease virus could not have been predicted considering that the termini visible in the crystal structure were buried or on the inner surface of the capsid. However, it was noted that in both cases the true N-terminus of the protein was not visualized in the three-dimensional structure.

Example 16 Comparison of different linker polypeptides HEK-293FT cells were transfected with vectors encoding various fusion polypeptides (Example 5) containing ferritin or lipoamide acetyltransferase (E2) core polypeptide, membrane-spanning domains and a HIS tag. A variety of alternative linkers were inserted between the core polypeptide and the membrane-spanning domain including: (a) 4, 5 or 6 copies of alternating fragments of the MMLV gag late domain (SEQ ID NO: 4549 and 50), (b) 10, 20, 30, 40 or 50 repeats of the EAAAK peptide (SEQ ID NO: 46), (c) 4 repeats of the E2 lipoyl domain (SEQ ID NO: 45). Culture supernatant was collected as described in Example 6 and tested by SDS-gel electrophoresis and Western blotting as described in Example 10. The results for the supernatant samples are shown in Table 11.

Table 11 Visualisation of fusion proteins by SDS-PAGE and Western blotting Plasmid Core polypeptide Linker Copies Predicted molecular weight Band seen? pFCK34 human lipoamide acetyltransferase (E2) Gly-Ser 1 73 kDa 78 kDa pFCK82 MMLV gag fragments 4 85 kDa 85 kDa pFCK84 MMLV gag fragments 5 90 kDa 90 kDa pFCK86 MMLV gag fragments 6 95 kDa 95 kDa pFCK210 EAAAK 20 76 kDa 80 kDa pFCK202 EAAAK 50 90 kDa 90 kDa pFCK203 lipoyl domain 4 92 kDa 95 kDa pFCK207 E coil ferritin EAAAK 10 31 kDa 30 kDa pFCK206 EAAAK 20 35 kDa 35 kDa pFCK205 EAAAK 30 40 kDa 40 kDa pFCK204 EAAAK 40 44 kDa 45 kDa pFCK200 EAAAK 50 49 kDa 55 kDa pFCK201 lipoyl domain 4 51 kDa 70 kDa Specific protein bands of the predicted molecular weights were detected in the supernatants of all transfectants, indicating that any of the tested combinations of core polypeptide and linker polypeptide may be used.

Example 17 Expression of Type IV-A membrane polypeptide with polypeptide tags HEK-293FT cells were transfected with vectors encoding various fusion polypeptides (Example 5) containing lipoamide acetyltransferase (E2) core polypeptide linked at the C-terminus by linker SEQ ID NO: 35 to a modified MMLV gag sequence (SEQ ID NO: 47) linked by linker SEQ ID NO: 35 to a transmembrane polypeptide which consisted of (a) a signal anchor (SA) membrane-spanning domain (SEQ ID NO: 38), (b) a polypeptide tag, either FLAG (SEQ ID NO: 12) or V5 (SEQ ID NO: 20) linked by linker SEQ ID NO: 34 to (c) a stop transfer anchor (STA) membrane-spanning domain (SEQ ID NO: 48). A negative control vector pFCK16 contained the E2 core polypeptide with a single pass HIS-tagged transmembrane polypeptide as described in Example 14, Table 8. This provided a system to test the expression of a transmembrane polypeptide which spanned the membrane twice, displaying a peptide epitope on the outside and terminating inside the membrane as illustrated in Figure 2. Culture supernatant was collected as described in Example 6 and tested by ELISA as described in Example 11 and by SDS-gel electrophoresis and Western blotting as described in Example 10. In each case the anti-HIS tag antibody was substituted by an appropriate mouse anti-FLAG or mouse anti-V5 antibody. The results for the supernatant samples are shown in Table 12.

Table 12 Visualisation of fusion proteins by ELISA, SDS-PAGE and Western blotting Plasmid Polypeptide tag Anti-FLAG Anti-V5 Predicted Band A450 A450 molecular weight seen? pFCK34 negative control 0.010 0.008 73 kDa 70 kDa pFCK136 FLAG 0.548 0.009 77 kDa 75 kDa pFCK137 V5 0.004 0.706 78 kDa 75 kDa Bands of the expected size were seen for all three constructs and positive results were obtained by ELISA for the FLAG peptide or the V5 peptide when present in the construct. This indicated that VLPs containing a Type IV-A membrane polypeptide could be expressed and that a variety of polypeptide tags could be incorporated in the fusion protein.

Example 18 Expression of a larger Type IV-A membrane polypeptide including three membrane-spanning domains and expression of VLPs containing a mixture of different of different fusion proteins It can be appreciated that there is likely to be an upper limit to the number of membrane-spanning domains which can be included within the membrane-spanning polypeptide. This limit is likely to depend on the size of the core polypeptide and the diameter of the assembled particle which it can form. Too many membrane-spanning domains would create steric hindrance to the successful assembly of the VLP. An experiment was carried out to test whether a polypeptide containing two membrane spanning domains could be supported by the comparatively small E. coil ferritin core polypeptide and if not, to test the possibility that a mixture of fusion proteins containing either full-size or truncated membrane-spanning polypeptides might allow expression of a VLP similar to that illustrated in Figure 3.

HEK-293FT cells were co-transfected with different ratios of vectors pFCK73 and pFCK11. Vector pFCK73 encoded a fusion polypeptide containing the E. coli ferritin core polypeptide, followed by a fragment of human SCN9A protein containing two membrane spanning domains (SEQ ID NO: 51) followed by a six amino acid HIS tag (SEQ ID NO: 26). Vector pFCK11 encoded a fusion polypeptide containing the same E. coli ferritin core polypeptide, a single membrane spanning domain (SEQ ID NO: 38) or followed by a six amino acid HIS tag (SEQ ID NO: 26). Culture supernatant was collected as described in Example 6 and tested by SDS-gel electrophoresis and Western blotting as described in Example 10. Relative intensities of the stained bands corresponding to the expected molecular weight of the fusion protein were estimated visually. The results are shown in Table 13.

Table 13 Visualisation of fusion proteins by Western blotting following transfection with mixtures of plasmids Transfection 1 Transfection 2 Transfection 3 Transfection 4 Plasmid vector pFCK73 pFCK11 pFCK73 pFCK11 pFCK73 pFCK11 pFCK73 pFCK11 Proportion in the vector mix 100% 0% 50% 50% 10% 90% 5% 95% Visible fusion protein in the cell pellet Very none high high medium high weak high high Visible fusion protein in the supernatant none none medium medium weak high none high Transfection with a single plasmid pFCK73 encoding a fusion protein with two transmembrane domains resulted in transcription of the protein (as indicated by the strong band in the cell pellet sample) but there was no evidence for formation and secretion of particles (as indicated by lack of a band in the supernatant sample). However, co-transfection with mixtures of the plasmids pFCK73 and pFCK11 in the proportions 50:50 or 10:90 resulted in visible bands corresponding to both fusion proteins in the supernatant samples, indicated that VLPs were formed and secreted. Thus although a small particle based on a ferritin core polypeptide may not be able to accommodate 24 copies of a membrane polypeptide with two membrane-spanning domains, it can support approximately 12 copies of such a membrane polypeptide.

REFERENCES

Alfthan K et al. (1995) Properties of a single-chain antibody containing different linker peptides. Protein Engineering vol.8 pp.725-731.

Almen MS et al. (2009) Mapping the human membrane proteome: a majority of the human membrane proteins can be classified according to function and evolutionary origin. BMC Biology 2009, vol.7 pp.50 doi:10.1186/1741-7007-7-50.

Arai R et al (2001) Design of the linkers which effectively separate domains of a bifunctional fusion protein. Protein Engineering vol.14 pp.529-532.

Argos P. (1990) An Investigation of Oligopeptides Linking Domains in Protein Tertiary Structures and Possible Candidates for General Gene Fusion. J. Mol. Biol. vol.211 pp.943-958.

Barbas CF et al. (2001) Phage Display: A Laboratory Manual. pub. Cold Spring Harbor Laboratory Press.

Bayer ME et al. (1968) Particles associated with Australia antigen in the sera of patients with leukaemia, Down's Syndrome and hepatitis. Nature vol.218 pp.1057-1059.

Caldeira et al. (2010) Immunogenic display of diverse peptides, including a broadly cross-type neutralizing human papillomavirus L2 epitope, on virus-like particles of the RNA bacteriophage PP7. Vaccine vol. 28 pp. 4384-4393.

Carillo-Tripp M et al. (2008) VIPERdb2: an enhanced and web API enabled relational database for structural virology. Nucleic Acids Research vol.37 pp.D436-D442 Chen X et al. (2013) Fusion protein linkers: Property, design and functionality. Advanced Drug Delivery Reviews vol.65 pp.1357-1369 Crasto CJ and Feng J (2000) LINKER: a program to generate linker sequences for fusion proteins. Protein Engineering vol.13 pp.309-312.

Debbink K et al. (2013) Human norovirus detection and production, quantification, and storage of virus-like particles. Curr. Protoc. Microbiol. vol. 31 pp. 15K.1.1-15K.1.45.

Deml L et al. (2005) Recombinant HIV-1 Pr55gag virus-like particles: potent stimulators of innate and acquired immune responses. Molecular Immunology vol.42 pp.259-277 Fifis T et al (2004) Size-Dependent Immunogenicity: Therapeutic and Protective Properties of Nano-Vaccines against Tumours. J. Immunol. vol.173 pp.3148-3154.

George RA and Heringa J. (2003) An analysis of protein domain linkers: their classification and role in protein folding. Protein Engineering voll5 pp.871-879.

Grgacic EVL et al. (2006) Virus-like particles: Passport to immune recognition. Methods vol.40 pp.60-65.

Hanly WC et al. (1995) Review of Polyclonal Antibody Production. Procedures in Mammals and Poultry. ILAR Journal vol.37 pp.93-118.

Howard GC and Bethel DR (eds) (2001) Basic methods in antibody production and characterization. pub. CRC Press LLC.

Kao CC et al. (2011) The coat protein leads the way: an update on basic and applied studies with the Brome mosaic virus coat protein. Mol. Plant. Pathol. vol 12 pp 403-12.

Khoury GA et al. (2011) Proteome-wide post-translational modification statistics: frequency analysis and curation of the swiss-prot database. Scientific Reports Vol.1, Article number: 90 pp.1-5. doi:10.1038/srep00090 Kim SE et al (2011) Fluorescent ferritin nanoparticles and application to the aptamer sensor. Anal. Chem. vol. 83 pp. 5834 43.

Krogh A et al. (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. vol.305 pp.567-580.

Kumar S, et al (2009) Tomato bushy stunt virus (TBSV), a versatile platform for polyvalent display of antigenic epitopes and vaccine design. Virology vol. 388 pp 185-90.

Ladenstein R et al. (2013) The lumazine synthase/riboflavin synthase complex: shapes and functions of a highly variable enzyme system. FEBS J. vol. 280 pp. 2537-63.

Lamb JW et al. (1996). Assembly of virus-like particles in insect cells infected with a baculovirus containing a modified coat protein gene of potato leafroll luteovirus. J. Gen. Virol. vol. 77 pp. 1349-58.

Lane SW et al. (2011) Construction and crystal structure of recombinant STNV capsids. J. Mol. Biol. vol. 413 pp. 41-50.

Li TC et al. (2005) Essential elements of the capsid protein for self-assembly into empty virus-like particles of hepatitis E virus. J Virol. vol. 79 pp. 12999-3006.

Lodish et al. (2007) Molecular Cell Biology, pub MacMillan 6th edition, p 546.

Ludwig C and Wagner R (2007) Virus-like particles -universal molecular toolboxes. Current Opinion in Biotechnology vol.18 pp.537-545.

Luque D et al. (2012). Epitope insertion at the N-terminal molecular switch of the rabbit haemorrhagic disease virus T = 3 capsid protein leads to larger T = 4 capsids. J. Virol. vol. 86 pp. 6470-80.

Moon KB et al. (2014) Overexpression and self-assembly of virus-like particles in Nicotiana benthamiana by a single-vector DNA replicon system. Appl. Microbiol. Biotechnol. vol. 98 pp. 8281-90.

Mossadegh N et al. (2004) Codon optimization of the human papillomavirus 11 (HPV 11) L1 gene leads to increased gene expression and formation of virus-like particles in mammalian epithelial cells. Virology. vol. 326 pp. 57-66.

Natarajan P et al. (2005) Exploring icosahedral virus structures with VIPER. Nature Reviews Microbiology vol.3 pp.809-817.

Neirynck S et al. (1999) Universal influenza A vaccine based on the extracellular domain of the M2 protein. Nature Medicine vol.5 pp.1157-1163.

Noad R and Roy P (2003) Virus-like particles as immunogens. Trends in Microbiology vol.11 pp.438-444.

Ochoa WF et al. (2008) Partitivirus structure reveals a 120-subunit, helix-rich capsid with distinctive surface arches formed by quasisymmetric coat-protein dimers. Structure. vol. 16 pp 776-86.

Ossipow V and Fischer N (eds.) (2014) Monoclonal Antibodies Methods and Protocols. Series: Methods in Molecular Biology vol.1131. 2nd ed. pub. Humana Press Overington JP et al. (2006). How many drug targets are there? Nat Rev Drug Discov vol.5 pp.993-6.

Perham RN (1991) Domains, motifs and linkers in 2-oxo acid dehydrogenase multienzyme complexes: a paradigm in the design of a multifunctional protein. Biochemistry vol.30 pp.8501-8512.

Raman S et al. (2009) Design of Peptide Nanoparticles Using Simple Protein Oligomerization Domains. The Open Nanomedicine Journal vol. 2 pp15-26.

Robinson CR and Sauer RT (1998) Optimizing the stability of single-chain proteins by linker length and composition mutagenesis. Proc. Natl. Acad. Sci. USA vol.95 pp. 529-5934.

Rueda P et al. (1999) Minor displacements in the insertion site provoke major differences in the induction of antibody responses by chimeric parvovirus-like particles. Virology. vol. 263 pp 89-99.

Simon C et al. (2014) Disulfide linkage and structure of highly stable yeast-derived virus-like particles of murine polyomavirus. J. Biol. Chem. vol. 289 pp. 10411-8.

Shepherd P and Dean C. (2000) Monoclonal Antibodies: A practical approach. pub Oxford University Press.

Smith SM. (2011) Strategies for the purification of membrane proteins. Methods Mol Biol. vol.681 pp.485-96. doi: 10.1007/978-1-60761-913-0_29.

Stephens PE et al. (1983) The Pyruvate Dehydrogenase Complex of Escherichia coli K12 Nucleotide Sequence Encoding the Dihydrolipoamide Acetyltransferase Component. Eur. J. Biochem. vol 133 pp 481-489.

Tars K et al. (2007) The crystal structure of bacteriophage GA and a comparison of bacteriophages belonging to the major groups of Escherichia coli leviviruses. J. Mol. Biol. vol 271 pp 759-773.

Terpe K (2003) Overview of tag protein fusions: from molecular and biochemical fundamentals to commercial systems. Appl Microbiol Biotechnol vol.60 pp.523-533.

Tissot AC et al. (2010) Versatile virus-like particle carrier for epitope based vaccines. Plos One vol 5 article e9809. doi: 10.1371/journal.pone.0009809.

Vicente T et al. (2011) Large-scale production and purification of VLP-based vaccines. Journal of Invertebrate Pathology vol.107 pp.S42-S48.

Winter G (1994) Making Antibodies by Phage Display Technology. Annual Review of Immunology vol.12 pp.433-455.

Wriggers Wet al. (2005) Control of Protein Functional Dynamics by Linkers. Peptide Science vol.80 pp.736-746.

Yang Y & Burkhard P (2012) Encapsulation of gold nanoparticles into self-assembling protein nanoparticles. Journal of Nanobiotechnology vol. 10 pp. 42-53.

Zeltins A (2013) Construction and Characterization of Virus-Like Particles: A Review. Mol Biotechnol vol.53 pp.92-107.

Zhou ZH et al (2001) The remarkable structural and functional organization of the eukaryotic pyruvate dehydrogenase complexes. Proc. Natl Acad. Sci. U S A. vol. 98 pp. 14802-7.

Claims

CLAIMSWe claim: 1. A virus-like particle comprising a membrane and a fusion polypeptide, wherein the membrane is a lipid bilayer and the fusion polypeptide comprises: (a) a core polypeptide enveloped by the membrane; and (b) a membrane-spanning polypeptide which crosses said membrane at least once.
2. The particle according to claim 1 wherein the core polypeptide is selected from a group consisting of: ferritin, encapsulin, dihydrolipoylsuccinyltransferase (EC 2.3.1.61), dihydrolipoamide branched chain acetyltransferase (EC 2.3.1.168), lipoyl acetyltransferase (EC 2.3.1.12) or is a polypeptide fragment or a polypeptide homologue of a member of the said group.
3. The particle according to claim 1 wherein: (a) the core polypeptide is a capsid protein or is a polypeptide fragment or a polypeptide homologue of a capsid protein and (b) the N-terminus or the C-terminus of the core polypeptide is accessible on the outside of the assembled capsid.
4. The particle according to claim 1 wherein the core polypeptide is selected from a group consisting of: Newtons T=3 capsid protein, Bacteriophage GA capsid protein, Bacteriophage PP7 capsid protein, Bacteriophage AP205 capsid protein, Ryegrass mottle virus capsid protein, Turnip yellow mosaic virus capsid protein, Melon necrotic spot virus capsid protein, Pyrococcus furiosus virus-like protein, Norwalk virus capsid protein, Hepatitis E virus capsid protein, Rabbit haemorrhagic disease virus capsid protein or is a polypeptide fragment or a polypeptide homologue of a member of the said group,
5. The particle according to any of claims 1 to 4 wherein the membrane-spanning polypeptide is selected from a group consisting of: Type I membrane proteins, Type II membrane proteins, Type III membrane proteins, Type IV-A membrane proteins, Type IV-B membrane proteins or is a polypeptide fragment or a polypeptide homologue of a member of the said group.
6. The particle according to any of claims 1 to 5 wherein the fusion polypeptide also comprises one or more linker polypeptides.
7. The particle according to any of claims 1 to 6 wherein the fusion polypeptide also comprises a polypeptide tag.
8. A polynucleotide coding for the fusion polypeptide according to any of the claims 1 to 7.
9. A vector comprising the polynucleotide according to claim 8.
10. A host cell comprising the polynucleotide according to claim 8, wherein said host cell is capable of expressing the particle according to any of claims 1 to 7.
11. A method of preparing the particle according to any of claims 1 to 7 comprising: culturing the host cell of claim 10 with a culture medium; and purifying the particle from the culture medium or from the host cell.
12. A pharmaceutical composition comprising the particle according to any of claims 1 to 7.
13. A vaccine composition comprising the particle according to any of claims 1 to 7.
14. A DNA vaccine composition comprising the polynucleotide according to claim 8.
15. A method of producing an antibody, comprising contacting the particle according to any of claims 1 to 7 and/or the polynucleotide according to claim 8 to a bird or mammal.
16. A method of selecting an antibody, comprising screening an antibody library for binding to the particle according to any of claims 1 to 7.
17. A pharmaceutical composition comprising the antibody produced according to claim 15 and/or selected according to claim 16.
18. A medicament for use in therapy comprising any composition according to claim 12, 13, 14 or 17.