WO2007092457A2 - Ligands for virus channel proteins - Google Patents

Abstract

Ligands that exert biological effects upon viruses by binding to their membrane-spanning protein components, such as membrane channel M2 in influenza virus, are of great pharmacological significance. The invention disclosed herein provides methods of manufacturing of particles that carry such a channel protein on their surface, and oriented in the same way as in viral and cellular membranes. Also disclosed are methods of making antibodies that bind to the channel protein in the particles and uses of such antibodies for treating infections.

Description

LIGANDS FOR VIRUS CHANNEL PROTEINS

Claim of Priority

This application claims priority to United States Provisional Application No: 60/765,712, filed February 6, 2006, titled: "Ligands for Virus Channel Proteins," inventors Tajib Mirzabekov and David Kreimer, herein expressly incorporated fully by reference.

BACKGROUND OF THE INVENTION Pharmacological effects of majority of pharmaceutical drugs on the market and in development are exerted through binding of such a drug compound to a particular protein in patient's body thus affecting the protein's function. Among proteins that are particularly important targets are so called trans-membrane proteins. These proteins typically are involved in vital functions of cells and viruses, and therefore selectively affecting them often brings a desirable outcome in a patient's health. A distinctive feature of trans-membrane proteins is a domain or domains that span the lipid bilayer of a biological membrane, whereas only a portion of such a protein is typically present on the surface of a cell or viral particle. A ligand that binds to such an exposed domain on the surface of a cell or virus might change the protein's function; such a ligand could have therefore a great pharmaceutical potential. Search for ligands that bind such exposed, external domains of membrane proteins remains to be extremely laborious and inefficient process. Two major approaches to the discovery of such ligands are: (1) immunization with preparations of a transmembrane protein, such as cells, viral particles or cellular membranes, or with peptide fragments of a trans-membrane protein, and (2) library screening, in which one tries to isolate from a library containing a large number of potential ligands those components that bind to a preparation of trans-membrane protein, such as cells, viral particles or cellular membranes, or to peptide fragments of a trans-membrane protein.

Both approaches are critically dependent upon the quality of membrane protein preparation. When a trans-membrane protein in a preparation is present at low concentrations, immune responses are generally poor, and isolation of antibodies that bind to the protein may be problematic. Additionally, presence of other protein and/or non-protein contaminants in the preparation, which is typically the case when whole cells, virus particles, or crude membrane preparations are used, can produce a background signal so strong that one may be unable to readily identify the ligands in question. Further, loss of native conformation of such a protein often renders both approaches at best inefficient and often futile. Important requirements for quality of such antibodies are poorly addressed by cell- based, viral particle-based, liposome-based, or membrane-based preparations of membrane proteins. Also, the use of fragments of membrane proteins has proven inefficient because such fragments generally do not maintain native conformation, and even when a ligand that bind such a peptide is discovered, such a ligand often is unable to exert a desirable change upon the function of a membrane protein.

Recently, Magnetic Proteoliposomes (MPLs) disclosed in the US Patent 6,761,902 titled 'Proteoliposomes containing an integral membrane protein having one or more transmembrane domains' by Joseph Sodroski and Tajib Mirzabekov, July 13, 2004 and the US patent application 20040109887, Al, June 10, 2004 titled 'Immunogenic proteoliposomes, and uses thereof by Wyatt, Richard T. et al expressly incorporated herein fully by reference, have been used as membrane protein preparations. MPLs allow one to purify certain membrane proteins in their native, functional conformation, and can stabilize the protein in proper orientation and at high concentration on the surface of easy-to-handle magnetic beads by way of peptide or protein "tags" engineered on the protein. However, in prior art MPLs having tags on each monomer (e.g., see U.S. Patent No: 6,761,902) the tags can stearically hinder proteins in MPLs and can thereby alter the protein into a non-native configuration. Antibodies and other ligands directed against non-natively configured proteins can render the ligand ineffective in modulating the membrane spanning protein in its native state on a virus. Therefore, while MPLs have been proven useful in human antibody development using both transgenic mouse immunization and by selection of antibodies from phage display libraries, there is an ongoing need for improved methods for producing antibodies and other ligands that can affect membrane-spanning proteins.

SUMMARY

In certain aspects of this invention, multimeric membrane spanning proteins can be expressed in a MPL without having tags on all monomers. Thus, a complex comprising a multimeric membrane spanning protein can be successfully expressed in a MPL without the alterations in configuration that can render prior art MPLs ineffective. Manufacture of such complexes can be achieved using expression vectors that provide tags to only a selected number of monomers. In the case of proteins having 4 monomers, one can co-transfect cells with: (1) a vector that encodes a monomer with a tag and (2) a vector that encodes the monomer without a tag. When both vectors are expressed in a cell, the resultant multimeric protein can be assembled from a mixture of tagged and un-tagged monomers, thereby producing a "hetero-multimeric" protein. When the hetero-multimeric protein is expressed in the form of a MPL, the MPL is termed a "hetero-multimeric MPL" or "HMPL." Thus, in a HMPL, due to the lack of bulky tags on each monomer, the protein is less prone to taking on a non-native conformation. However, for multimeric proteins, it is not necessary to provide a tag on each monomer.

Once expressed such hetero-multimeric proteins can be easily captured by a capture reagent (e.g., antibody) designed to recognize a tag- Once a capture antibody recognizes one tag on a hetero-multimeric protein, the entire protein complex can be then formed into an HMPL. Using such HMPLs can dramatically reduce the time for selection of ligands from various libraries, such as chemical library, phage, aptamer, shpigelmer, nanobody, antibody fragment, scFv, minibody, anticalin or other protein scaffold library, cell library, and any other library.

Thus, depending on the relative amounts of expression of the tagged and un- tagged monomers, heteromeric transmembrane protein complexes can be made with fewer than a stoichiometric numbers of tags. In this way, there is less interference of the tag(s) with the proper orientation and structure of the multimeric complex.

Thus, in one example, to prepare a ligand that can affect the function of influenza viruses, the influenza protein M2 can be desirably used. By providing antibodies and/or other types of ligands directed against hetero-multimeric M2, the activity of M2 can be decreased, thereby decreasing infectivity and the mobidity and mortality associated with influenza pandemics.

In other embodiments, antibodies can be directed at a highly conserved "box" structure of the M2 multimer can improve efficacy. ^' Because the box structure is highly conserved and is necessary for formation of functional channels, antibodies directed against this structure are less sensitive to mutations in M2. Antibodies against M2 can bind to the box structure and can block the channel in M2, thereby decreasing the ability of M2 to promote infection. Even mutations that would render viruses drug-resistant need not render the mutant resistant to an anti-M2 antibody.

BRIEF DESCRIPTION OF THE DRAWINGS

This invention is described with reference to specific embodiments thereof. Other features and aspects of this invention can be appreciated by reference to the following figures, in which: FIGS. Ia - Ic depict the structure of the M2 protein.

FIG Ia depicts the M2 protein comprised of tag-carrying and unmodified polypeptide chains of present invention. FIG. Ib depicts a side view the extra-cellular domain (ectodomain) of the M2 protein with the box structure.

FIG. Ic depicts a top view of the extra-cellular domain (ectodomain) of the M2 protein having a box structure. FIG. 2 schematically depicts a HMPL particle of this invention that carries on its surface the M2 protein molecules of present invention having the box structure of this invention in the native conformation in a reconstituted lipid bilayer.

DESCRIPTION OF THE INVENTION Rapid discovery of antibodies and/or other ligands that bind certain membrane proteins of viruses is of crucial importance. It can be the matter of life and death for millions currently infected people already faced with the challenge of clearing rapidly changing viruses such as Influenza, HIV, as well as other viruses such as hepatitis family of viruses. And it is certainly important for avoiding future epidemics and pandemics like the Influenza pandemics of 1918, 1957, and 1964. One of the challenges has been the rapid mutation of many viral proteins, making immunization strategies ineffective over long periods of time.

In fighting viral infections, among the most attractive membrane protein targets for antibodies and other ligands are viral channel proteins. These proteins are critically involved in one or several steps of viral cycle, and due to high evolution pressure imposed by the necessity to carry their intricate function, tend to maintain their sequence highly conserved in order to not change their structure and maintain functionality. For example, the M2 protein of influenza A virus is a proton channel, and is genetically linked to another viral protein, Ml. Therefore, among three major antigens of Influenza A, virus hemagglutinin (HA), neuraminidase (NA), both of which have high level of point mutations and antigenic shifts, a third, the M2 protein, has a low level of sequence variation. The M2 protein is a proton-selective ion channel with pH-inducible activity; it is involved in virus uncoating in the endosome, i.e., in viral fusion and infection, and in virus maturation in trans-Golgi network. The M2 protein is expressed at the plasma membrane of virus-infected cells and is also exposed on the viral surface. Table 1 below illustrates the low level of sequence variation observed in various strains of M2 since 1918. The M2 sequences of viruses that caused pandemics of 1918,

1957, and 1964 are virtually identical. A drug targeting M2 channel would effectively save tens of millions lives. Contemporary viruses have a very similar M2 sequence (e.g., New York 2005 strains) — such a drug would still work. Avian Flu viruses (H5N1) of great current concern have very similar M2 sequences. The strains of Avian Flu viruses

(highlighted) isolated from people in 1997 infect mice directly; this can be a reasonable sequence for discovery of ligands inhibiting influenza infection virtually regardless viral strain. Also highlighted is the 22 aa N-terminus peptide, the sequence of which is additionally conserved due to the genetic link with the Ml protein mentioned above.

Table 1. Sequence Variation in Different Strains of M2

One series of embodiments of present invention includes HMPLs that carry the M2 protein of influenza A virus on their surface. Another embodiment of present invention includes antibody-ligands that can bind the M2 protein exposed on the surface of these preparations. The M2 protein from influenza A virus is an integral membrane protein comprised of four identical, disulfide-1 inked, 97 amino acids polypeptide chains, each containing a TM helix. The M2 protein forms proton-selective channel that is essential to viral function and is the target of the drug Amantadine and similar agents. In contrast to several other viral proteins targeted for anti-viral drag and vaccine development, M2 is highly conserved. Rare mutations in M2 however do occur, and that renders Amantadine inefficient in treatment of several strains of viruses. While Amantadine and similar agents are small molecules whose binding site changes substantially upon such mutations, the binding of antibodies that would have significantly larger volume can be affected by such mutations to a significantly lesser extent. Therefore, the M2 protein appears to be a very attractive target for development of human therapeutic antibodies and universal anti- influenza vaccines.

Recently, there have been reports of promising results in the development of universal anti-influenza vaccine targeting the highly conserved influenza viral membrane protein M2. These results were based on using a small extra-cellular N-terminal peptide of M2, conjugated to unrelated proteins as an antigen for vaccine production. While antibody binding to an extra-cellular fragment might be sufficient for imposing some affect on the virus, functional blockade of the channel (the validated mode of action of Amantadine and similar agents) can only inconsistently be achieved using such an antigen. The very choice of the N-terminal peptide antigen as an immunogen excludes a priori the major portion of the native antigenic surface of native M2 channel residing in the membrane and known to be functionally important. In addition, a large portion of the N-terminal peptide proximal to the M2 trans-membrane region is also excluded from being a useful immunogen.

Thus, some embodiments of the present invention include the use of MPLs that can maintain the native "box structure" of M2. This box structure is a portion of the M2 protein that can be maintained properly protruded into external space from the area of transmembrane domain by maintaining overall native conformation of the protein. Within the native M2 protein, the structure of this portion is presumably constrained by disulfide bonds formed by Cys-17 and Cys-19 and intermolecular interactions between tetramer subunits and interactions with the lipid membrane. Residues 24, 23 and 22 adjacent to the transmembrane (TM)-domain are engaged in a defined 3-D structure that most probably extend farther away due to the inter-chain Cystines at positions 19 and 17. Thus, 28 residues or more can be engaged in a defined structural domain - the box structure adjacent to the TM domain. Thus, several highly conserved, conformation-dependant epitopes presumably exist in the ectodomain in the native M2 protein; the epitopes cannot be targeted using non-structured ectodomain peptides.

The ectodomain fragment structure is thus vastly different from that presented in the bacterially synthesized water-soluble fusion proteins used for vaccine development by the prior art. It is likely that in the 1-24 fragment of the extra-cellular peptide, only first approximately 10 amino acids (unrelated to the channel functions) remain unstructured in the native M2 tetramer. Typically, when N-terminal peptides (20-30 amino acid (aa) long) of trans-membrane proteins such as GPCRs are used for immunization, antibodies that arise bind only to a small portion at the very end of these peptides. Thus, while the current M2- derived N-terminal peptides vaccine trials may show promising results, major and functionally significant parts of the antigenic surface of M2 channel remain untargeted by such vaccines, and the majority of antibodies that can be made in a vaccinated organism are likely to be unrelated to the most effective flu target. Therefore, yet another embodiment of the present invention describes selection of ligands that bind to the external portion of the M2 protein that maintains its native conformation via numerous interactions within transmembrane domains and the box structure.

In contrast to N-terminal peptide approach, the utilization of MPLs with properly maintained box structure for vaccination can result in highly potent antibodies with properties similar to that of Amantadine but with decreased sensitivity to mutations. In other embodiments, antibodies can be fully humanized to decrease the likelihood of adverse immunological reactions to the antibody. Such fully human antibodies of the present invention can bind the native functional M2 channel and some can inactivate it akin to the Amantadine action. Additionally, they can induce an antibody activating body's immune response (ADCC, CDC) against the influenza virus. In addition, our technologies can allow the identification of antibodies able to bind to the M2 channel at both, neutral and low pH. The therapeutic antibodies might be of a broad use because of long blood circulation time (around 1 month). Only one or two injections per person can be sufficient for overcoming influenza epidemic. In case of H5N1 or similar viruses that are threatening broad populations, having stockpiles of such antibodies can be as important as a broad vaccination of the population. Furthermore, by using MPLs and naked particles of this invention more efficient vaccines can be developed.

I. Methods for Manufacturing of MPLs The manufacturing of MPLs in general, is described in US Patent 6,761,902, herein expressly incorporated fully by reference. Briefly, first, paramagnetic particles, for example M-280 Tosylactivated Dynabeads™ produced by Dynal Biotech Inc. are chemically derivatized with a capture agent, using the protocol provided by the Dynal Biotech Inc. A capture agent can be an antibody that is capable of selective binding a respective tag, or streptavidin that can bind a known peptide tag; either of the tags can be attached at the C-terminus of a given membrane protein.

Second, a given membrane protein can be over-expressed in a mammalian cell by transfecting, by using for example a GenePORTER™ transfection reagent and protocol (Gelantis), a line of mammalian cells (can be purchased at ATCC) with a vector (for example, pcDNA3.1, from Invitrogen) carrying the gene of the protein having an appropriate peptide tag at the C-terminus and genes that provide an antibiotic resistance to the cells. For manufacture of hetero-multimeric proteins of this invention, cells can be transfected with vectors that encode tagged monomers and other vectors that encode un-tagged monomers. Alternatively a single vector having two or more expression cassettes (one cassette having a sequence encoding a tagged monomer and another cassette encoding a untagged monomer) can be used. In such systems, a mixture of tagged and untagged monomers can be produced, that when associated with each other, form a hetero-multimeric protein complex.

Antibiotic resistance (for example, resistance to gentamycin (Geneticin™; G418), the feature acquired concomitantly with the capacity to over-express the membrane protein, can be used for selecting over-expressing cells that survive in the presence of added the antibiotic.

Third, cells that over-express the membrane protein can be harvested, and the membranes of the cells can be solubilized in a mixture of detergents and lipids (e.g. phosphatidylcholine, phosphatidylserine, phosphatidethanolamine, or lipid mixtures isolated form tissues or plants).

Fourth, the solublized membrane solution can be clarified by centrifugation solubilization and containing the membrane protein of interest along with numerous other contaminating proteins, can be mixed with the beads carrying a capture agent capable of binding the tag on the protein.

Fifth, washing the beads can remove contaminants. A magnet can be used to hold beads within a vessel (e.g., tube) and washing solutions can be added, to carry away non-bound materials, including contaminants. The beads retaining the protein of our interest in the desirable orientation (i.e., the extracellular portion is exposed on the surface of the bead), can then be dialyzed in the presence of lipids in order to reconstitute a lipid bilayer.

In certain embodiments, MPLs are provided that carry the M2 protein on their surface in the native, tetrameric state. An important feature of the M2 protein is the presence of the box structure that is desirably maintained using special care to obtain MPLs that have the box structure in the proper, native configuration. Such MPLs are desirable because antibodies and other ligands made that bind to M2 proteins having box structures in their native configuration are better suited for binding with M2 in in vivo conditions. Therefore, such antibodies and other ligands can be better reagents for affecting the function of M2. Ligands can be then selected from a library, such as phage antibody library or can be obtained by immunization and selection of an antibody from an antibody pool.

FIG. Ia schematically displays the structure of the M2 protein of present invention 1100. The native protein is comprised of four identical polypeptide chains 1000.

Each polypeptide chain 1000 contains an ectodomain 1001, a transmembrane domain 1002 and a cytoplasmic domain 1003. In some embodiments 1110, it can be desirable to add tag

1004 to C-terminus of cytoplasmic domain 1003. Tag 1004 can be used to attach the protein

1000 to a magnetic bead for constructing MPLs and/or naked particles of the present invention. Alternatively, an antibody against a cytoplasmic portion 1003 of the protein can serve for retention of the protein in the correct orientation on the surface of the particles. Lipid molecules 1102 are depicted as stabilizing the heterotetramer. Additionally, disulfide bridges 1101 are shown

In some embodiments, MPLs can be manufactured utilizing the hetero-tetramer protein. In these embodiments, each of four chains have three identical domains: an ectodomain 1001 of 24 amino acids; a trans-membrane domain 1002 of 19 amino acids; and a cytoplasm domain 1003 of 54 amino acids. In addition, one, two or three chains carry tag 1004. Such a hetero-tetramer protein can be obtained by doing transfection with a mixture of two vectors; one vector encoding an unmodified sequence and one other vector encoding sequence with the tag.

Alternatively, a single vector can contain an open reading frame encoding a native protein and an open reading frame encoding a tagged protein. When expressed, the single vector can produce both native protein and tagged protein. Regardless of how the tagged and native proteins are formed, once made, the monomers can then assemble into the tetrameric form. When at least one tagged protein is assembled with other native protein chains, the structure is termed herein a "hetero-tetramer." By using such a hetero-tetramer, a native structure of the protein with little or no distortion of the M2 protein structure can be achieved. In particular, in some embodiments, the native box structure is maintained. The structure of such hetero-tetramer is shown with lipids 1102 that surround trans-membrane domain 1002. Disulfide bridges 1101 are also shown that link the chains of the protein together to form the typical "box" structure.

FIGS. Ib and Ic depict side and top views, respectively, of M2 protein in a HMPL of this invention. FIG Ib shows the extra-cellular domain (ectodomain) 1103 of the M2 protein with the box structure 1102 that is adjacent to the trans-membrane domain 1002 of the M2 protein. Amino acids 1121 involved in the formation of the box are depicted in white, while amino acids 1111 involved in less structured portion of the ectodomain 1103 are shown in black. Also shown are disulfide bridges 1101 that link the chains of the protein thus forming box structure 1102. Box 1102 leads to the entrance of the channel formed by trans-membrane portions 1002 surrounded by lipids 1104.

FIG Ic is a top view of the extra-cellular domain of M2. Transmembrane domains 1002, "box" amino acids 1121 and "non-box" amino acids 1111 are depicted as for FIG Ib above. Disulfide bridges 1101 are shown in an approximately tetrahedral structure, thereby identifying the "box." Protons flow along arrow into channel 1311.

In order to preserve the native structure of the M2 protein, several variables can be desirably adjusted. First, the density of the capture agent on the surface of the bead and the nature of the tag can be selected. Complete coverage of the bead with the capture agent could result in a distortion arising due to immobilization of more than one tag of the same tetramer that might propagate through the body of the protein's trans-membrane domain into the box area. Therefore, using a hetero-tetramer can help avoid the distortion. Second, in the native protein, Cys-17 and Cys-19 are engaged in inter-chain disulfide bonds. It can be desirable to avoid reducing conditions and reducing agents such as beta-mercaptoethanol and DTT, or use them only at very low concentrations. It can be desirable to express the functional M2 channels in mammalian cells where a high level of the M2 protein expression can be achieved and where posttranslational modification of M2 channel protein and native lipid environment can be easily achieved. Properly selecting these factors can aid in stabilizing the structure of the M2 channel. Two cysteine residues of external N-terminal domain of M2, Cys-17 and Cys-19, lay close to the water filled entry of M2 proton channel. Disulfide bond formation by these residues can be important in the presentation of the M2 channel in the native state. By varying the density of the capture agent and the nature of tag (and respective capture agent), and by maintaining disulfide bridges, one skilful in the art can obtain MPLs carrying the M2 protein with an intact box structure.

FIG. 2 depicts an MPL particle 2001 of this invention carrying the hetero- tetramer M2 protein 1100 on its surface. Paramagnetic particle 2002 is derivatized with a capture agent 2003 capable of binding tag 1004. A lipid bilayer comprised from lipids 2004 can cover the whole particle, as schematically shown with lines 2005.

II. Methods of Immunizing Mice

The use of MPLs for immunization of mice and selection of antibodies is disclosed in the US patent application 20040109887, Al, published June 10, 2004 titled

'Immunogenic proteoliposomes, and uses thereof by Wyatt, Richard T. et al. and is expressly incorporated herein fully by reference. In some embodiments of this invention,

MPLs or HMPLs carrying the M2 protein in the native state can be used to raise mouse antibodies against the ectodomain of the protein. These mouse antibodies can be used for demonstrating the therapeutic activity of the anti-M2 preparations in mouse model.

III. Fully Humanized Antibodies and Other Ligands

By immunizing humanized mice (mice having a complete human immune system) with MPLs carrying the M2 protein in the native state, fully humanized anti-M2 antibodies that bind the ectodomain of the protein can be obtained. These antibodies can be used for therapeutic purposes in humans.

Libraries of fully human antibodies and their fragments, as well as other protein scaffold libraries and libraries of other ligands that do not induce immune reaction in humans can be used for selection of anti-M2 ligands that display anti-viral infection activity in humans. MPLs carrying M2 in its native state as well as naked particles can be used for successful selection of anti-M2 ligands.

IV. Methods for Determining Antibody Presence

The presence of antibodies that bind the ectodomain of the M2 protein can be determined using MPLs or naked particles. Binding of an antibody to such a particle can be determined using a fluorescently labeled secondary antibody directed at the primary M2 antibody. This secondary antibody binds to anti-M2 antibody bound to the particle; this binding can be visualized using FACS or fluorescence microscopic methods. V. Antibody Function

Some antibodies obtained via immunization or from a library may only affect the development of influenza A viral infection indirectly by binding to the M2 protein expressed on the surface of infected cells, and some antibodies can directly affect the viral cycle by inhibiting channel function and other functions via direct binding to the M2 protein on the viral surface. Functional antibodies that bind to or in proximity to the box structure can be more desirable than those that bind a portion of the protein that is distant from the box structure. The box structure leading to the channel can be affected by the antibody binding and such binding can change the conformation of the channel; thereby inhibiting its function. Alternatively, an antibody can hinder entry of ions into the channel through its size. Regardless of the mechanism of action, antibodies produced using MPLs of this invention can be effective at inhibiting the function of M2. Such loss of function can be useful important for halting the infection. Such a functional antibody can act via a mechanism similar to that of Amantadine. It can be desirable to use additional stringency criterion in selection of antibodies with channel inhibitory capacity. These antibodies are desirably stable at acidic pH in order to be able to prevent viral uncoating in the endosome.

Antibodies that bind the M2 ectodomain can directly inhibit channel activity. The function can be demonstrated in patch clamp experiment in which membrane potential in cells expressing the M2 protein is measured as the function of the presence of such an antibody. Alternatively, fluorescent probes sensitive to membrane potential can be used for identifying the functional effect of a given antibody.

VL Inhibitors The inhibition of viral channels by Amantadine is well documented. MPLs of present invention can be used for selection of other small molecules from chemical libraries that can bind to the ectodomain or trans-membrane domain and can inhibit ion flow through channels. Such molecules can be of great pharmaceutical significance because numerous viral strains have evolved that escape inhibitory effect of Amantadine. Upon exposure of MPLs of this invention to a solution containing a small molecule ligand, MPLs are washed using magnetic force for their retention, and then are subjected to denaturing conditions in order to release bound molecules. Because each MPL can carry on its surface up to 100,000 molecules of the M2 protein, there is enough material even in very small amounts of MPLs for identification of bound small molecules by using mass-spectrometry.

VII. Functional Assays Measuring channel inhibition activity can be performed as a functional in vitro assay.

Several assays for in vivo evaluating the effect of a ligand are described in the literature and are routinely performed by those skilful in the art.

VIII. Vaccines Preparations in which the M2 protein is presented in its native state can induce much superior immune responses to those of the prior art. In addition to the presenting of functional M2 channels in particles similar to MPLs (in which paramagnetic core is replaced with a polymer that can be degraded by the body), liposomal formulations of functional M2 channels can be used for intranasal vaccination. Additionally, other virus proteins can be used for antibody and vaccine manufacture.

IX. Combinations

It can be desirable to use antibodies obtained using preparations of M2 in which the box structure is maintained in the native state in combination with antibodies against unstructured portion of the ectodomain of the M2 protein. The use of combined antibodies as therapeutics can dramatically reduce the likelihood of development virus strains resistant to such a treatment. The probability of the M2 protein mutations concomitantly occurring in two or more rather large epitopes for the antibodies is extremely low.

X. MPLs Carrying Native Channels of Other Viruses

Channel proteins are present in several other viruses, such as HIV. The use of preparations disclosed in this invention for the M2 channel can be expanded to immunization and selection of ligands that bind to viral channel proteins of other viruses. Functional ligands that inhibit channel function in these viruses can be used for treatment of these viral infections.

The descriptions above are intended to illustrate aspects of this invention. Other aspects and embodiments can be made by persons of skill in the art without undue experimentation and with a reasonable likelihood of success. All such embodiments are considered to be part of this invention. Each reference cited in this application is expressly incorporated herein fully by reference, as if individually so incorporated.

Claims

We claim:

1. A particle comprising:

(a) a bead having a capture agent attached thereto, said capture agent that binds to a tag on an intracellular portion of an expressed multimeric membrane protein of a pathogen;

(b) at least one expressed multimeric protein attached to said capture agent, said multimeric protein made of monomers, at least one of which, but less than all of which monomers having a tag thereon comprising a heteromultimeric protein; and (c) a lipid bilayer surrounding said bead, at least a portion of said expressed protein being within said lipid bilayer.

2. The particle of claim 1, where in said expressed protein is M2 protein of influenza virus.

3. A method for manufacturing a particle, comprising:

(a) providing a bead of step (a) of claim 1 ;

(b) expressing a heteromeric membrane protein in native conformation, said protein having less than a stoichiometric number of tags attached thereto; (c) attaching said capture agent on said bead to at least one of said tags thereby forming a bead-tag-protein complex; and

(d) encapsulating said bead-tag-protein complex in a lipid bilayer.

4. The method of claim 3, wherein said step of expressing is carried out in a eukaryotic, a procaryotic or a cell-free expression system.

5. A method for manufacturing an antibody, comprising:

(a) providing a particle of claim 1;

(b) administering said particle to an immunologically competent host organism; and

(c) selecting antibodies produced in step (b) based on binding of said antibodies to said protein in native conformation.

6. The method of claim 5, wherein said protein is Influenza M2.

7. The method of claim 5, wherein said protein is M2 protein of Influenza H5N1.

8. The method of claim 5, wherein said multimeric protein comprises 4 monomers.

9. The particle of claim 1, wherein said multimeric protein comprises 4 monomers.

10. Use of an antibody of any of claims 5 to 8, in the manufacture of a medicament for treating a viral infection.

1 1. The use of claim 10, wherein said antibody is humanized.

12. A method of treating an individual having a viral infection, comprising: administering to said individual, an antibody of claim 5 and a pharmaceutically acceptable excipient.

13. The method of claim 12, wherein said antibody is humanized.