WO1997004748A9 - Enhanced artificial viral envelopes for cellular delivery of therapeutic substances - Google Patents

Abstract

This invention provides artificial viral envelopes and other lipid vesicles that encapsulate therapeutic substances, such as expression vectors, targeted to mammalian cells. Polynucleotides may be packed into the envelopes by compressing them beforehand with a short peptide with a predominant positive charge. The compression step not only facilitates encapsulation, it also increases the number of vesicles containing nucleic acid, minimizes the amount of free nucleic acid, and may also increase the size and complexity of plasmids that can be encapsulated. The vesicles may be provided with a tissue targeting component that helps direct it towards certain tissue sites in an animal. The vesicles may also be provided with a fusogenic component that facilitates delivery of the therapeutic substance into the cell. The materials and reagents of this invention are effective, for example, in increasing expression of model proteins in both isolated cells and intact animals, and are expected to be useful for gene therapy.

Description

ENHANCED ARTIFICIAL VIRAL ENVELOPES FOR CELLULAR DELIVERY OF THERAPEUTIC SUBSTANCES

STATEMENT OF RIGHTS TO INVENTIONS MADE UNDER FEDERALLY SPONSORED RESEARCH

This invention was made in part during work supported by grants P50 HL19153, ROl HL 45151 and HL 43167 from the National Institutes of Health. The government has certain rights in the invention.

BACKGROUND OF THE INVENTION

A number of important human diseases involve a perturbation in the normal level of gene expression. Certain genes are either underexpressed or overexpressed in the affected cells. There is a long-standing interest in developing a treatment modality capable of altering genetic expression levels. Many important genes have been cloned and sequenced, and expression systems that have been developed should allow them to play a therapeutic role once delivered inside the affected cells.

One ofthe key challenges in gene therapy is finding a means to deliver the polynucleotide expression vector inside the intended target cells. Transfection may be a routine matter in certain in vitro culture conditions, but it is considerably more difficult in a whole individual. Compositions may be diluted by endogenous substances, and degraded or eliminated before they have a chance to exert their pharmacological effect. They may also modify cells other than the intended target.

An effective delivery vehicle will therefore have several properties. At the very least, it should protect the polynucleotide from degradation long enough to arrive at the intended target intact, and have no substantial adverse effect on the subject. If the polynucleotide exerts its therapeutic effect inside the cell, then the delivery vehicle should also be capable of promoting entry ofthe polynucleotide across the cell membrane. Ideally, the delivery vehicle will also have an ability to exert its effect on the target tissue in preference to unaffected tissues in the same individual.

A major research emphasis for gene delivery in recent years has been viral-based vectors based on such viruses as retroviruses, adenovirus, and adeno-associated viruses (Drumm et al., Rosenfeld et al., Muzyczka et al.). Viral vectors have several complications. First, a large amount of native virus sequence is provided, with the virus modified to be replication negative. There is always the risk of recombination with other infective agents with undesirable results. Second, each gene to be administered must be genetically engineered into the virus. Third, both retrovirus and adeno-associated virus vectors have low packaging efficiency, and there is a packaging limit on the size of heterologous polynucleotide sequence that can be included. Fourth, adenovirus vectors have been associated with a significant inflammatory response.

Because of these concerns, non- viral approaches to gene therapy are an emerging area of interest (reviewed, e.g., by Ledley, by Schofield et al., and by Chonn et al.). Gene therapy may be conducted by administering naked DNA, particularly for immunization purposes. DNA may be complexed with proteins that package them into certain structures. For example, the B. subtilis nucleoid-associated protein HPB12 compacts DNA into slightly curved rods (Schultz-Gahmen et al.), and a valine tripeptide condenses circular DNA into torus-shaped particles (Vengerov et al.). DNA may also be complexed with proteins that promote localization near the nucleus (reviewed by Dingwall et al. and Behr et al.).

In addition, the DNA may be associated with an additional component to assist it in binding or transcending the outer membrane ofthe intended target cell. Examples of such complexes are: a) a ligand-polylysine complex (Wilson et al.), in which the ligand is specifically recognized by a receptor such as the asialoglycoprotein receptor on hepatocytes, and facilitates binding to the target cell; b) an antibody-polylysine complex (Trubetskoy et al.), in which the antibody facilitates binding to a cell-surface antigen recognized by the antibody; and c) a complex in which the DNA is conjugated to a peptide such as hemagglutinin (Bongartz et al.) designed to destabilize the target cell membrane. In any of these DNA-protein formulations, the DNA is not protected from the effects of enzymes and other solutes in the environment that may degrade the DNA before it reaches its destination.

Another non-viral approach is cationic liposomes, particularly liposomes comprising DC-cholesterol (Ledley). Positively charged liposomes are pre-formed, and then combined with the polynucleotide, which associates with the liposomes by charge- association. When the complex contacts the target cell wall, the positively charged liposome membrane facilitates entry ofthe polynucleotide into the cell without initiating endocytosis. The complex may optionally also comprise a polypeptide that may target the polynucleotide to the nucleus once inside the cell (see, e.g., Conary et al.; WO 95/34647). However, since the polynucleotide is associated with the outside ofthe liposomes, rather than being encapsulated within them, it is still at the mercy ofthe environment while en route to the target cell.

To both protect the polynucleotide and enhance its delivery, the polynucleotide can be encapsulated inside a liposome. Particularly suitable for encapsulation are liposomes with lipid membranes that are anionic or neutral in charge. The lipid envelope protects the polynucleotide during storage and after administration before contact with the target cell.

A number of publications describe a variety of methods for preparing liposomes of different structure and lipid composition. See, for example, Gregoriadis (1988 & 1993), Watwe et al., Vemuri et al., Elorza et al., and U.S. Patents 4,737,323 and 5,008,050. Preformed liposomes may then be filled with a therapeutic substance by dehydrating and rehydrating them in the presence ofthe substance (WO 95/12387); forming a liposome film and then presenting an aqueous mixture ofthe substance (U.S. Patent 4,515,736); or treating them with a bile-acid detergent to render them permeable to the substance (EP Patent 0274581). Placing a therapeutic substance in a liposome may in itself enhance targeting to specific locations in the body, such as tumors (U.S. Patent 5,019,369).

However, providing a polynucleotide in a liposome encapsulated form suitable for therapeutic use has a number of special challenges. These include: a) sufficient incoφoration of polynucleotide into liposomes of suitable composition; b) evasion ofthe reticulo-endothelial system when used in vivo long enough to reach the intended location; c) delivery ofthe therapeutic contents into the target cell; d) therapeutic availability of the polynucleotide once delivered.

Many ofthe methods for filling a pre-formed liposome with a therapeutic substance are generally unsuitable for use with polynucleotides, because the polynucleotides are typically large and highly charged. Usually, polynucleotides are placed into liposomes by forming or reconstituting the liposome in the presence ofthe polynucleotide. This considerably affects the liposome preparation methods that can be used, and the characteristics ofthe liposomes that emerge. Methods of encapsulating polynucleotides or oligonucleotides in liposomes is described, e.g., in U.S. Patent 4,078,052, WO 92/06192, and by Baru et al. and Thierry et al. Prior art methods have been limited by the size ofthe liposome relative to that ofthe polynucleotide. Liposomes encapsulating polynucleotide expression systems (such as bacterial plasmids) are typically quite large. For example, those described in U.S. 4,078,052 are about 1000 nm; those described in Thierry et al. are up to 3000 nm in diameter. In order for liposomes administered in vivo to reach the intended therapeutic location, they must first avoid uptake by the reticulo-endothelial system; particularly cells such as Kupffer cells that have direct contact with circulating particles. Larger liposomes are especially susceptible to uptake. Liposomes of particular composition apparently evade the RES and have prolonged circulation time: for a general review, see Oko et al. Particularly effective lipid compositions are those with a large proportion of sphingomyelin and cholesterol (WO 95/35094), or a combination of sphingomyelin and phosphatidylinositol (U.S. Patent 4,920,016). Liposomes with prolonged circulation time also include those that comprise the monosialoganglioside GM1 (Mumtaz et al.), glucuronide, or polyethylene glycol (Allen). Such compositions are sometimes referred to as "stealth liposomes". Alternatively, the RES system may be pre-blocked by a first administration of empty liposomes, which is then followed by a second administration of liposomes encapsulating the therapeutic compound (U.S. Patent 5,435,989).

Once the liposome has reached the intended location, the liposome must then deliver its payload across the membrane into the interior ofthe cell. Unlike their positively-charged counterparts, anionic liposomes do not necessarily have the ability to disrupt the cell membrane. Anionic liposomes (when they are taken up by the cell at all) typically enter via an endocytic pathway. This in itself does not succeed in delivering the payload to the cytoplasm, since endocytic compartments are topologically equivalent to the outside ofthe cell. Liposomes remaining in endocytic compartments or delivered to lysosomes may therefore be therapeutically ineffective. Instead, the liposome should fuse with the membrane, either on the outside ofthe cell or in an endocytic compartment, in order to reach full effectiveness.

Lipid assymetry can make vesicles fusion-competent by inhibition of thermal undulations (Devaux et al.) Considerable progress towards therapeutically effective anionic liposomes occurred upon the invention of artificial viral envelopes (AVE; U.S. Patent 5,252,348). These are liposomes made to resemble the bilayer composition of viral envelopes, such as the HIV envelope. AVE with a unilamellar bilayer comprising a phospholipidxholesterol molar ratio of about 0.8:1.2 are inherently fusogenic.

Liposomes may also be provided with molecules at the surface that help them find or treat the cell of interest. Small molecules may be attached by incorporating into the lipid bilayer a functionalized phosphoiipid (U.S. Patent 5,052,421) or a functionalized cholesterol (U.S. Patent 4,544,545). Polypeptides may be attached covalently to the lipid bilayer (EP Patent 0036277), to a glycophospholipid (U.S. Patent 5,374,548), to a carboxylated phosphoiipid (U.S. Patent 4,762,915), to a derivatized sterol (U.S. Patent 5,000,960), or to a peptide anchor (U.S. Patent 5,109,113). Alternatively, if the polypeptide comprises a hydrophobic domain, it may be incoφorated directly into the lipid bilayer, either by forming the liposome in its presence, or by preforming the liposome and inserting the polypeptide subsequently using a suitable detergent (Tranum- Jensen et al., EP Patent 0047480, U.S. Patent 5,252,348).

Liposomes have been prepared with mammalian-derived peptides such as cytokines (U.S. Patent 5,258,49), transferrin (Stavridis et al.), antibody (Laukkanen et al.), asialofetuin and other galactose-terminated side chains (Ishihara; Ghosh et al.), a fusogenic protein from rat brain microsomal membranes (Rakowska et al.), and surfactant protein A (Walther et al.). Liposomes have been prepared with artificial peptides, such as a 14-residue amphipathic sequence which is a fusogenic GALA-type peptide (Puyal et al.). Liposomes have also been prepared with viral components: for example, the F and G glycoprotein of respiratory syncytial virus (RSV) (U.S. Patent 5,252,348), reovirus M. cell attachment protein (Rubas et al.), influenza virus surface protein (WO 92/19267, EP 0047480, Nussbaum et al.), viral membrane fusion proteins, particularly hemagglutinin (WO 95/32706), and the influenza hemagglutinin D loop and K loop peptides (Friede et al.). Although some useful progress has been made, previous disclosures still do not provide all the desirable attributes of a therapeutic composition together in a single liposome.

For example, one long-recognized problem is that liposomes are difficult to pack efficiently with polynucleotides (Fraley et al. 1980 & 1981). Even when packaging is accomplished, the number of liposomes containing a polynucleotide is low, which decreases the effectiveness ofthe preparation. As a consequence ofthe low packing rate, a large proportion ofthe polynucleotide supplied for the process is wasted. The size and plurality of polynucleotide inside each liposome is necessarily constrained, which limits the range of therapeutic compositions that can be prepared. The lipid envelopes that do encapsulate a polynucleotide are typically so large that they are easily captured by the reticuloendothelial system, and are consequently unsuitable for systemic administration. Another problem is that encapsulated polynucleotides have not been targeted with adequate efficiency to particular cell types of interest, such as those in the lung.

The packing problem is solved in this invention by providing compounds and methods for compressing the polynucleotide. Once compressed, the polynucleotide is easily encapsulated into a small-diameter liposome. Strategies presented in this disclosure are particularly suited for targeted delivery of many classes of therapeutic substances to particular cell types.

SUMMARY OF THR INVENTION

This invention provides reagents and methods that can be used to compress polynucleotides to a suitable radius for encapsulation in neutral and anionic liposomes. It has been discovered that small peptides with a predominant positive charge, particularly due to lysine side chains, considerably reduce the average radius of DNA when analyzed by electron microscopy. As a result, polynucleotides are much more readily encapsulated in to liposomes. It has also been discovered that liposomes pre-formed with compressed polynucleotides inside can subsequently be inserted with tissue targeting and fusogenic components, and that these components work in concert to promote delivery ofthe polynucleotide into the cell. After delivery, the polynucleotide is unpackaged and decompressed, and can thereafter modulate the expression of an encoded protein.

An important benefit of this invention is that the proportion of liposomes in a given preparation that encapsulate a polynucleotide is higher than what was previously possible. This technique has been used to encapsulate nucleotides of up to about 8,000 base pairs into liposomes of 250-1000 nm. Compared with the encapsulation of uncompressed polynucleotides in like-sized liposomes, compression increases the uptake into the liposomes from about 24% to about 64%. In addition, the number of liposomes containing genetic material increases from about 31% to about 73%. Thus: a) precious polynucleotides may be more efficiently encapsulated; b) more polynucleotide is present per liposome, meaning that less liposome need be given per dose, increasing the therapeutic potential; c) potentially, larger polynucleotides containing more genes and controlling elements, and more complex polynucleotide mixtures can be encapsulated than before. It is predicted that the maximum size of a nucleic acid in this type of composition for use in gene therapy will be limited not by the effectiveness ofthe compression or the size ofthe liposome, but by the practical limit size of a functioning plasmid, which is ofthe order of 12,000 base pairs.

Another important benefit of this invention is the efficiency by which the material encapsulated by a liposome can be delivered to a target cell. By providing the liposome with a tissue targeting component, it accumulates in the vicinity ofthe target. By providing the liposome with a fusogenic polypeptide, the rate of fusion with the target cell membrane is enhanced. The effect is synergistic. Compared with unmodified liposomes, those containing a fusogenic component show about 2.5 x higher elevation in expression ofthe encapsulated polynucleotide. Those containing both fusogenic and tissue targeting components show about 8 x higher elevation in expression.

Particular embodiments of this invention include a liposome comprising a synthetic lipid vesicle and a compressed polynucleotide encapsulated by the vesicle.

Preferred vesicles are essentially unilamellar, between about 100 nm and 1000 nm in size, and have a cholesterol :phospholipid molar ratio between about 0.5 and 1.2. The polynucleotide is preferably compressed with a polypeptide comprising a linear sequence of at least 7 amino acids of which at least 50% ofthe amino acids have a side chain bearing a positive charge at pH 7. The liposome also preferably comprises a tissue targeting component, specific for lung endothelial cells or another cell type.

Other embodiments of this invention are liposomes with serotonin on the outer surface. The serotonin acts as a tissue targeting component for lung endothelial cells. The liposomes may encapsulate any material, including polynucleotides (particularly a compressed polynucleotides), peptides, drugs, and toxins. Further embodiments of this invention are bifunctional liposomes, comprising a synthetic lipid vesicle, a tissue targeting component, and a fusogenic component, wherein the tissue targeting component and the fusogenic component do not naturally occur together on a single molecule or a single viral particle. The two components may be independently anchored in the lipid bilayer, present together as a fusion protein, or the tissue targeting component may be on the surface while the fusogenic component is encapsulated. Preferred fusogenic components are modeled on fragments of the influenza hemagglutinin peptide, particularly those having the sequence shown in SEQ. ID NO:7.

Preferred bifunctional liposomes are obtainable by one ofthe following processes: 1. Encapsulating the polynucleotide in the lipid vesicle; contacting the lipid vesicle with the tissue targeting component and with the fusogenic component in any order in the presence of a detergent; and removing the detergent. 2. Encapsulating the polynucleotide in a lipid vesicle comprising a phosphoiipid or sterol to which the tissue targeting component is covalently attached; contacting the lipid vesicle with the tissue targeting component in the presence of a detergent; and removing the detergent. 3. Encapsulating the fusogenic component and the polynucleotide in a lipid vesicle comprising a phosphoiipid or sterol to which the tissue targeting component is covalently attached; contacting the lipid vesicle with the fusogenic component in the presence of a detergent; and removing the detergent. Bifunctional liposomes prepared by other methods are contemplated and included in the invention. Yet another embodiment is a method of compressing a polynucleotide, comprising the step of contacting the polynucleotide with a positively charged polypeptide, preferably comprising a linear sequence of at least 7 amino acids of which at least 50% ofthe amino acids have a side chain bearing a positive charge at pH 7. Preferably at least about half of the amino acids in the comprised sequence are lysine.

Another embodiment of this invention is a method of delivering a polynucleotide to a cell, comprising contacting the cell with a liposome of this invention. Another embodiment is a method for up- or down-regulating expression of a protein by a cell, comprising contacting the cell with a liposome of this invention comprising a compressed polynucleotide comprising an encoding or antisense sequence. Further embodiments are methods for delivering a substance to a cell in an individual, comprising administering the substance encapsulated in a liposome of this invention. Pharmaceutical compositions and methods of gene therapy are also embodied in the invention.

These and additional embodiments ofthe invention are outlined in the description that follows.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 is a half-tone reproduction of transmission electron micrographs, showing identical plasmids, first before and then after addition of a compressing peptide. The plasmid comprised an expressible polynucleotide sequence and was approximately 8 kilobases in size. The compressing peptide was about 12 amino acids in length, having the sequence of SEQ. ID NO:2. The magnification in the Right Panel is about 2.5 times that in the Left Panel.

Figure 2 is a half-tone reproduction of transmission electron micrographs, showing plasmids identical to those in Figure 1 , compressed with polylysine or histone protein.

Figure 3 is a half-tone reproduction of a gel stained with ethidium bromide, showing that the electrophoretic mobility of a plasmid is altered after addition ofthe compressing peptide.

Figure 4 is a series of graphs obtained from FACS analysis, showing the percentage of liposomes containing fluorescent DNA. Fig. 4a is the analysis of empty liposomes; Fig. 4b is the analysis of liposomes containing uncompressed DNA; the Fig. 4c is the analysis of liposomes containing compressed DNA.

Figure 5 is a half-tone reproduction of fluorescence micrographs, showing the location of fluorescent DNA in transfected cells. The DNA was prepared in the absence or presence of a nuclear targeting signal.

Figure 6 is a graph showing the level of expression ofthe CAT reporter gene transfected cells. The plasmid used for transfection was prepared with increasing amounts of a nuclear targeting signal.

Figure 7 is a graph showing the degree of tissue staining due to aerosol administration of live animals with liposome-encapsulated compressed CAT reporter gene (hatched bars) or empty liposomes (solid bars).

Figure 8 is a half-tone reproduction of lung staining due to intravenous administration of live animals with liposome-encapsulated PAP gene or empty liposomes. Neither the plasmid nor control-treated animals showed significant staining in the lung. Figure 9 is a half-tone reproduction of kidney staining in the same animals. PAP staining is present in sections from the gene transfected animal but not the control.

Figure 10 is a half-tone reproduction of lung and kidney staining in an animal intravenously administered with liposome-encapsulated PAP gene. In this case, the liposomes also comprised serotonin, a tissue targeting component that was specific for lung endothelial cells. Significant staining was observed both in the lung and the kidney.

Figure 11 is a bar graph showing that artificial viral envelopes (AVE) are stable in pH (open bars) and diameter (solid bars) during extended storage refrigerated or at room temperature.

Figure 12 is a scheme for the preparation of serotonin conjugated to cholesterol, which can be incoφorated into liposomes for targeting to the lung.

Figure 13 is bar graph showing that the level of expression ofthe marker gene PAP is enhanced in the lungs of treated animals when PAP is delivered with AVE comprising a fusogenic component (stippled bars). The level of expression is enhanced over 8-fold when delivered with bifunctional AVE, comprising both a fusogenic component and the lung-specific tissue targeting component serotonin. Figure 14 is four photomicrographs taken under phase-contrast and fluorescence illumination, showing that an αj-antitrypsin expression vector delivered using AVE accumulates near the nucleus of treated cells.

Figure 15 is a bar graph showing that α^-antitrypsin is secreted by treated cells when the expression vector is delivered using an AVE with a fusogenic component.

Figure 16 is a series photomicrographs showing cells containing fluorescently labeled antisense oligonucleotides compressed with a positively charged polypeptide and delivered in an AVE.

DETAILED DESCRIPTION

It is an object of this invention to provide improved compositions and methods for delivering a therapeutic agent, particularly a polynucleotide. The therapeutic agent is enveloped in a lipid vesicle which has an enhanced ability to deliver its encapsulated material to a cell or region of therapeutic interest.

Where the therapeutic agent is a polynucleotide, small synthetic peptides are used to compress it into a three-dimensional conformation with a smaller average radius, and also partly neutralize the charge. This permits the polynucleotide to be more readily encapsulated into a liposome, which in turn is capable of promoting translocation ofthe polynucleotide across the outer membrane of a target cell.

The liposome compositions of this invention may optionally comprise a molecule such as a member of a receptor-ligand pair that is accessible from the outside ofthe lipid envelope, and acts as a tissue targeting component. This enhances the ability ofthe liposome to localize near certain target cells when used in vitro, or at certain tissue sites when admimstered to a mammal, due to an affinity for the other member ofthe receptor- ligand pair. Various types of tissue targeting components are described in this enclosure, including the novel lung-specific targeting molecule serotonin.

Alternatively or in addition, the liposome compositions may optionally comprise a fusogenic component. This is a peptide attached to or embedded in the lipid envelope and promotes the ability ofthe liposome to deliver its encapsulated material into the target cell. Liposome compositions of this invention may also optionally comprise a polypeptide attached to the polynucleotide being delivered, whereby the polynucleotide, once inside the cell, is directed towards certain regions ofthe cell, such as the nucleus. The intracellular targeting component may be part ofthe same polypeptide which is used to compress the polynucleotide before its encapsulation.

Liposomes prepared according to this method may be used whenever it is desirable to provide therapeutic compounds to a cell, such as for antisense or gene therapy. They may be employed for treatment of either isolated cells or intact animals, particularly humans, that are in need of such therapy. Accordingly, this invention also embodies liposomes in pharmaceutical compositions, and their use.

This invention fulfills a long-felt need by solving many ofthe problems described earlier for the previously known compositions. The invention represents a significant advancement over prior art liposome compositions and gene delivery mechanisms for the following reasons: 1. The lipid vesicles are particularly suitable for use in biological systems. They protect their therapeutic contents from the environment until delivery to the target cell. Any therapeutic compound capable of envelopment in the vesicles may be delivered, from a large gene vector to a chemical hapten. While the compositions may comprise compression, targeting and fusogenic components, these need only be present in the amount required for activity.

No potentially immunogenic structural proteins are present, as they are in viral packages. 2. By compressing a polynucleotide by combining with positively charged polypeptides, they can be encapsulated within liposomes at a previously unpredictable efficiency, and a suφrisingly high density. Vesicles may be packed with large therapeutically useful expression vectors, then sized by extrusion down to a diameter as small as 100-300 nm without losing the polynucleotide. Small vesicles are particularly useful in avoiding uptake by the reticuloendothelial system in vivo, increasing their circulation time and enhancing uptake by the target tissue. 3. Vesicles containing compressed polynucleotides are remarkably resistant to subsequent manipulation. In particular, they may be inserted with outward- oriented proteins, such as fusogenic or tissue targeting components, by mild detergent treatment. The detergent treatment does not result in release ofthe compressed polynucleotide. As a result, the proteins are functionally accessible at the vesicle surface, and do not occupy space inside the vesicle where they might interfere with the polynucleotide.

4. Before this invention was made, it was not predictable that vesicles containing a compressed polynucleotide, including those of small diameter or with a rigid structure, would be able to effectively display a tissue targeting molecule, thereby enhancing their accumulation at the target cell. In addition, it was not previously predictable that vesicles, having reached the surface ofthe target cell, would be capable of delivering their encapsulated material inside the cell. This disclosure shows that delivery to the target is considerably enhanced by a functionally accessible tissue targeting component. This even includes haptens that target to cell surface receptors like the serotonin receptor.

5. Before this invention was made, it was not predictable that a compressed polynucleotide, once delivered to a target cell, could subsequently be unpackaged and decompressed. The association between the polynucleotide and the compression peptides are strong enough to maintain the polynucleotide in the compressed form throughout encapsulation. However, it was found that once they are delivered to the cell, compressed expression vectors are efficiently used to express the proteins they encode. This indicates that the peptides dissociate from the polynucleotide inside the cell, allowing the polynucleotide to unfold and participate in the normal process of transcription.

6. Bifunctional liposomes, comprising both fusogenic and tissue targeting components, are remarkably efficient in delivering their encapsulated material to a target cell. Previously described viral particles like influenza have proteins that enhance cell surface binding and membrane fusion. These proteins have deliberately evolved and are oriented so as to work together to enhance infectivity ofthe particle. The present disclosure shows that fusogenic and tissue targeting components that are: a) from completely different sources, and b) inserted into the viral envelope artificially, may nonetheless work together on the surface of an artificial lipid envelope. The ability of these components to work in concert results in an 8-fold improvement in delivery and expression ofthe encapsulated polynucleotide.

Definitions

As used herein, a "liposome" "lipid envelope" or "lipid vesicle" is a small vesicle bounded by at least one and possibly more than one bilayer lipid membrane. It is made artificially from phospholipids, glycolipids, steroids such as cholesterol, related molecules, or a combination thereof by any technique known in the art, included but not limited to sonication, extrusion, or removal of detergent from lipid-detergent complexes. A liposome may optionally contain within the lipid membrane an additional element, such as a nucleic acid, a polypeptide, or a drug, that it may be desirable to deliver to a target cell. A liposome may also optionally comprise additional components associated with the outer surface, such as a tissue targeting component. A "synthetic" lipid vesicle refers to a vesicle assembled from lipids and other components rather than being extracted directly from viruses or cells, although their composition may be modeled on virus or cell membranes.

It is understood that a "lipid membrane" or "lipid bilayer" need not consist exclusively of lipids, but may additionally contain any percentage of other components, included but not limited to cholesterol and other steroids, proteins of any length, and other amphipathic molecules, providing the general structure ofthe membrane is a sheet of two hydrophilic surfaces sandwiching a hydrophobic core. For a general discussion of membrane structure, see "The Encyclopedia of Molecular Biology" by J. Kendrew (1994).

As used herein, the terms "artificial viral envelope" and "AVE" indicate a liposome that bears a lipid or lipid/cholesterol composition not derived from but resembling that of a naturally occurring virus or viral particle. AVE generally comprise at least two and preferably at least three phospholipids and a sterol, usually cholesterol or a cholesterol derivative. Preferred AVE are rigid, stable structures with a unilamellar lipid envelope. The cholesterol :phospholipid molar ratio is usually at least about 0.2:1 to 2.0:1, preferably about 0.5:1 to 1.2:1, more preferably about 0.8:1 to 1.2: 1, and even more preferably about 1 :1. AVE of this invention also comprise a tissue targeting component functionally accessible from the outside.

A liposome or AVE is said to "encapsulate" an additional component, such as a polynucleotide, a peptide, or a drug, if the lipid bilayer separates the additional component from the external environment. This may be tested, for example, by determining whether the liposome is capable of protecting the component from an externally provided reagent which is effective against the component but not against the bilayer. For example, an encapsulated DNA will be protected from digestion by exogenously added DNAse; an encapsulated RNA will be protected from exogenously added RNAse; an encapsulated peptide will be protected from binding by an exogenously added antibody directed against it. It is understood that the liposome may still be permeable to certain lipid-soluble peptides, chemicals and drugs, and that the encapsulated component will still be susceptible to such compounds, or to other compounds should the integrity ofthe lipid membrane be compromised.

The "payload" or "encapsulated material" of a liposome refers to material encapsulated by the liposome according to the preceding definition. Encapsulated material used in the liposomes of this invention may be an agent, substance, compound, or mixture, useful for any other purpose including but not limited to treatment, diagnosis, experimentation, or to act as a control. No limitation is implied when an embodiment is illustrated with one of these terms, unless explicitly required. Preferably the material is "deliverable" to the interior ofthe target cell by the liposome. The ability of a liposome to deliver material may be measured, for example, by labeling the material with a fluorescent marker, or (if the material is an expression vector) by measuring expression of the encoded protein by the treated cell.

"Targeting" is the process by which a compound or complex is permitted to accumulate in a particular locale in greater preference over other locales than would otherwise be the case. This may be accomplished by providing the compound or complex with a component, called a "targeting component" which promotes accumulation in an area near the target. Accumulation near the target may be promoted, for instance, via specific transport towards the target, or specific retention in the neighborhood ofthe target. A cell to which it is desired that a liposome localize or deliver its contents is described as a "target cell".

A "tissue targeting component" is a component of a complex, particularly an AVE or other liposome, that enhances its accumulation at certain tissue sites in preference to others when administered to an intact animal, artificial organ, or cell culture. A tissue targeting component is generally accessible from outside the liposome, and is therefore generally either bound to the outer surface or inserted into the outer lipid bilayer. A tissue targeting component may be inter alia a peptide, a region of a larger peptide, an antibody, a nucleic acid, a carbohydrate, a region of a complex carbohydrate, a special lipid, or a small molecule such as a drug, hormone, or hapten, attached to any ofthe aforementioned molecules. A "fusogenic" peptide or component is a peptide or component that enhances the ability of an agent to fuse with a cell membrane, including but not limited to the outer cell membrane and membranes in the endocytic pathway, including those of endosomes and lysosomes. A fusogenic peptide attached to or contained in a liposome enhances the ability ofthe liposome to deliver its encapsulated material once it has reached a target cell. The term includes any delivery method, including but not limited to fusion ofthe lipid envelope with the outer membrane ofthe cell, or fusion ofthe lipid envelope with the membrane of an endosomal compartment subsequent to endocytosis ofthe liposome into the cell.

A tissue targeting component or fusogenic component is said to be "inserted" into a liposome if it is attached to the liposome in a way that permits it to exercise its targeting or fusogenic function. Generally, this implies that the component is anchored into or grafted onto the membrane, but is still at least partly accessible from outside the external surface.

A tissue targeting component is said to be "functionally accessible" from outside a liposome if it confers a targeting ability upon the liposome when tested either in vivo or with isolated cells, or if it confers upon the liposome the ability to bind a soluble receptor or antibody against the component.

An "intracellular targeting component" is a component of a complex, particularly comprising a polynucleotide, that enhances its accumulation at certain subcellular sites in preference to others when administered cells in culture or in an intact animal. Exemplary intracellular targeting components embodied in this invention are small peptides, small regions of a larger peptide, or protein complexes. Two examples of intracellular targeting components are nuclear localization signals (NLS) and mitochondria localization signals (MLS). The term "polynucleotide" refers to a polymeric form of nucleotides of any length, either deoxyribonucleotides or ribonucleotides, or analogs thereof. A polynucleotide is an example of a nucleic acid. Polynucleotides may have any three-dimensional structure, and may perform any function, known or unknown. The following are non-limiting examples of polynucleotides: a gene or gene fragment, exons, introns, messenger RNA (mRNA), transfer RNA, ribosomal RNA, ribozymes, cDNA, recombinant polynucleotides, branched polynucleotides, plasmids, vectors, isolated DNA of any sequence, isolated RNA of any sequence, nucleic acid probes, and primers. A polynucleotide may comprise modified nucleotides, such as methylated nucleotides and nucleotide analogs. If present, modifications to the nucleotide structure may be imparted before or after assembly ofthe polymer. The sequence of nucleotides may be interrupted by non-nucleotide components. A polynucleotide may be further modified after polymerization, such as by conjugation with a labeling component.

The term polynucleotide, as used herein, refers interchangeably to double- and single-stranded molecules. Unless otherwise specified or required, any embodiment of the invention described herein that is a polynucleotide encompasses both the double-stranded form and each of two complementary single-stranded forms known or predicted to make up the double-stranded form.

In the context of polynucleotides, a "linear sequence" or a "sequence" is an order of nucleotides in a polynucleotide in a 5' to 3' direction in which residues that neighbor each other in the sequence are contiguous in the primary structure ofthe polynucleotide. A "partial sequence" is a linear sequence of part of a polynucleotide which is known to comprise additional residues in one or both directions.

An "expressible" gene is a polynucleotide with an encoding sequence, which is capable of producing the functional form ofthe encoded molecule in a particular cell. For a sequence encoding a polypeptide, the gene is capable of being transcribed and translated. For an anti-sense molecule, the gene is capable of producing replicate transcripts comprising anti-sense sequence. For a sequence encoding a ribozyme, the gene is capable of producing catalytic RNA.

"Recombinant", as applied to a polynucleotide, means that the polynucleotide is the product of various combinations of cloning, restriction or ligation steps, and other procedures that result in a construct that is distinct from a polynucleotide found in nature. A recombinant virus is a viral particle comprising a recombinant polynucleotide. The terms respectively include replicates ofthe original polynucleotide construct and progeny ofthe original virus construct. A "compressed", "collapsed", or "compacted" polynucleotide is a polynucleotide that has been modified so as to reduce its average three-dimensional radius in a solution of interest, such as physiologically buffered solutions. A preferred means of compressing a polynucleotide in this invention is through association with a "compressing" polypeptide, which is a polypeptide or region of a larger polypeptide capable of effecting the compression by association with the polynucleotide. The compressed polynucleotide may continue to be associated with the compressing polypeptide, but this is not necessary if the polynucleotide remains in the compressed form. For example, if a polynucleotide is compressed with a polypeptide, then encapsulated in a liposome that prevents it from expanding to its original radius, the polynucleotide is still said to be compressed regardless of whether the polypeptide is still part of the complex.

A "control element" or "control sequence" is a nucleotide sequence involved in an interaction of molecules that contributes to the functional regulation of a polynucleotide, including replication, duplication, transcription, splicing, translation, or degradation of the polynucleotide. The regulation may affect the frequency, speed, or specificity ofthe process, and may be enhancing or inhibitory in nature. Control elements are known in the art. For example, a promoter and an enhancer are two examples of control elements. A promoter is a DNA region capable under certain conditions of binding RNA polymerase and initiating transcription of a coding region usually located downstream (in the 3' direction) from the promoter.

"Operatively linked" refers to a juxtaposition of genetic elements, wherein the elements are in a relationship permitting them to operate in the expected manner. For instance, a promoter is operatively linked to a coding region if the promoter helps initiate transcription ofthe coding sequence. There may be intervening residues between the promoter and coding region so long as this functional relationship is maintained. An "expression vector" is a vector comprising a region which encodes a polypeptide of interest, and is used for effecting the expression ofthe protein in an intended target cell. An expression vector also comprises control elements operatively linked to the encoding region to facilitate expression ofthe protein in the target.

"Heterologous" means derived from a genotypically distinct entity from that ofthe rest ofthe entity to which it is being compared. For example, a polynucleotide introduced by genetic engineering techniques into a plasmid or vector derived from a different species is a heterologous polynucleotide. A promoter removed from its native coding sequence and operatively linked to a coding sequence with which it is not naturally found linked is a heterologous promoter.

"Genetic alteration" refers to a process wherein a genetic element is introduced into a cell other than by mitosis or meiosis. The element may be heterologous to the cell, or it may be an additional copy or improved version of an element already present in the cell. Genetic alteration may be effected, for example, by transfecting a cell with a recombinant plasmid or other polynucleotide through any process known in the art, such as electroporation, calcium phosphate precipitation, or contacting with a polynucleotide- liposome complex. Genetic alteration may also be effected, for example, by transduction or infection with a DNA or RNA virus or viral vector.

A cell is said to be "inheritably altered" if a genetic alteration is introduced which is inheritable by progeny ofthe altered cell. Preferably, the genetic element is introduced into a chromosome or mini-chromosome in the cell; but any alteration that changes the phenotype and/or genotype ofthe cell and its progeny is included in this term. The terms "polypeptide", "peptide" and "protein" are used interchangeably herein to refer to polymers of amino acids of any length. The polymer may be linear or branched, it may comprise modified amino acids, and it may be interrupted by non-amino acids. The terms also encompass an amino acid polymer that has been modified; for example, disulfide bond formation, glycosylation, lipidation, acetylation, phosphorylation, or any other manipulation, such as conjugation with a labeling component.

In the context of polypeptides, a "linear sequence" or a "sequence" is an order of amino acids in a polypeptide in an N-terminal to C-terminal direction in which residues that neighbor each other in the sequence are contiguous in the primary structure of the polypeptide. A "partial sequence" is a linear sequence of part of a polypeptide which is known to comprise additional residues in one or both directions.

A linear sequence of amino acids is "essentially identical" to another sequence if the two sequences have a substantial degree of sequence identity. It is understood that the folding and the biochemical function of proteins can accommodate insertions, deletions, and substitutions in the amino acid sequence. Thus, linear sequences of amino acids can be essentially identical even if some ofthe residues do not precisely correspond or align. Sequences that correspond or align more closely to the invention disclosed herein are more preferred. It is also understood that some amino acid substitutions are more easily tolerated. For example, substitution of an amino acid with hydrophobic side chains, aromatic side chains, polar side chains, side chains with a positive or negative charge, or side chains comprising two or fewer carbon atoms, by another amino acid with a side chain of like properties can occur without disturbing the essential identity ofthe two sequences. Methods for determining homologous regions and scoring the degree of homology are well known in the art; see for example Altschul et al. (1986) Bull. Math.

Bio. 48:603-616; and Henikoff et al. (1992), Proc. Natl. Acad. Sci. USA 89:10915-10919. Well-tolerated sequence differences are referred to as "conservative substitutions". Thus, sequences with conservative substitutions are preferred over those with other substitutions in the same positions; sequences with identical residues at the same positions are still more preferred. Protein "expression" refers to the amount of protein present in a cell or secreted by a cell. It may be increased, for example, by increasing the rate of translation of mRNA encoding the protein, which in turn may be accomplished by increasing the rate of transcription ofthe corresponding gene or increasing the persistence ofthe RNA. Protein expression may be decreased, for example, by decreasing the rate of translation of the mRNA encoding the protein, which in turn may be accomplished by decreasing the rate of transcription ofthe corresponding gene or decreasing the availability ofthe RNA to the ribosomal apparatus capable of translating it.

A "fusion polypeptide" is a polypeptide comprising regions in a different position in the sequence than occurs in nature. The regions may normally exist in separate proteins and are brought together in the fusion polypeptide; or they may normally exist in the same protein but are placed in a new arrangement in the fusion polypeptide. A fusion polypeptide may be created, for example, by chemical synthesis, or by creating and translating a polynucleotide in which the peptide regions are encoded in the desired relationship.

An "antibody" (interchangeably used in plural form) is an immunoglobulin molecule capable of specific binding to a target, such as a polypeptide, through at least one antigen recognition site, located in the variable region ofthe immunoglobulin molecule. As used herein, the term encompasses not only intact antibodies, but also fragments thereof, mutants thereof, fusion proteins, humanized antibodies, and any other modified configuration ofthe immunoglobulin molecule that comprises an antigen recognition site ofthe required specificity.

An "immunogenic" compound is a compound capable of stimulating production of an antibody when injected into a suitable host, usually a mammal. Immunogenic compounds include certain proteins, complex carbohydrates, complex lipids, polynucleotides, drugs, haptens, and other chemicals. Compounds which are identical to or closely mimic compounds that are part ofthe animal they are being administered to are generally not immunogenic, except in a disease condition or a state of immune hyper- responsiveness.

A substance is said to be "selective" or "specific" if it reacts or associates more frequently, more rapidly, or with greater duration with a particular cell or substance than it does with alternative cells or substances. For example, a liposome equipped with a targeting molecule may be "specific" for its intended target if the targeting molecule causes the liposome to react with or accumulate in the neighborhood ofthe target to a greater extent than would otherwise occur.

"α_rantitrypsin", "LDL receptor", "apolipoprotein", "Factor VIII", and other proteins, when discussed in the context of gene therapy and the compositions therefor, refer to the respective intact polypeptide, or any fragment or genetically engineered derivative thereof, that retains the desired biochemical function ofthe intact protein.

An "isolated" polynucleotide, polypeptide, or other substance refers to a preparation ofthe substance devoid of at least some ofthe other components that may also be present where the substance or a similar substance naturally occurs or is initially prepared from. Thus, for example, an isolated substance may be prepared by using a purification technique to enrich it from a source mixture. Enrichment can be measured on an absolute basis, such as weight per volume of solution, or it can be measured in relation to a second, potentially interfering substance present in the source mixture. Increasing enrichments ofthe embodiments of this invention are increasingly more preferred. Thus, for example, a 2-fold enrichment is preferred, 10-fold enrichment is more preferred, 100-fold enrichment is more preferred, 1000-fold enrichment is even more preferred.

An "individual" refers to vertebrates, particularly members of a mammalian species, and includes but is not limited to domestic animals, sports animals, and primates, including humans. An "effective amount" of a composition of this invention is an amount sufficient to obtain a beneficial or desired result, measured by an assay appropriate to monitor the effect of an active ingredient ofthe composition, or by clinical improvement. An effective amount may be given in single or divided doses.

"Treatment" of an individual or a cell is any type of intervention in an attempt to alter the natural course ofthe individual or cell at the time the treatment is initiated. For example, treatment of an individual may be undertaken to decrease or limit the pathology caused by any pathological condition, in.iuding (but not limited to) an inherited or induced genetic deficiency, infection by a viral, bacterial, or parasitic organism, a neoplastic or aplastic condition, or an immune system disfunction such as autoimmunity or immunosuppression. Treatment includes (but is not limited to) admimstration of a composition, such as a pharmaceutical composition, and administration of compatible cells that have been treated with a composition. Treatment may be performed either prophylactically or therapeutically; that is, either prior or subsequent to the initiation of a pathologic event or contact with an etiologic agent.

General techniques

The practice ofthe present invention will employ, unless otherwise indicated, conventional techniques of molecular biology, microbiology, recombinant DNA, and immunology, which are within the skill ofthe art. Such techniques are explained fully in the literature. See, for example, "Molecular Cloning: A Laboratory Manual", Second Edition (Sambrook, Fritsch & Maniatis, 1989), "Oligonucleotide Synthesis" (MJ. Gait, ed., 1984), "Animal Cell Culture" (R.I. Freshney, ed., 1987); the series "Methods in Enzymology" (Academic Press, Inc.); "Handbook of Experimental Immunology" (D.M. Weir & CC. Blackwell, eds.), "Gene Transfer Vectors for Mammalian Cells" (J.M. Miller & M.P. Calos, eds., 1987), "Current Protocols in Molecular Biology" (F.M.

Ausubel et al., eds., 1987); and "Current Protocols in Immunology" (J.E. Coligan et al., eds., 1991).

All patents, patent applications, articles and publications mentioned herein, both supra and infra, are hereby incoφorated herein by reference.

Preparing polynucleotides

Polynucleotides and oligonucleotides for use with this invention may be prepared by any technique in the art. This includes isolation of polynucleotides from natural sources, gene libraries, or cDNA libraries, amplification of polynucleotide sequences by such techniques as the polymerase chain reaction, the cloning of polynucleotides or polynucleotide fragments from any of these methods, and any combination thereof.

Polynucleotides may also be prepared by chemical synthesis. Several methods of synthesis are known in the art, including the triester method and the phosphite method. In a preferred method, polynucleotides are prepared by solid-phase synthesis using mononucleoside phosphoramidite coupling units. A typical solid-phase synthesis involves reiterating four steps: deprotection, coupling, capping, and oxidation. This results in the stepwise synthesis of an oligonucleotide in the 3' to 5' direction. See, for example, Hirose et al. (1978), Tetra. Lett. 19:2449-2452; Beaucage et al. (1981), Tetra. Lett 22:1859-1862; and U.S. Patent No. 4,415,732.

Where it is desired that the polynucleotide confer upon the target cell an ability to express an encoded polypeptide, the polynucleotide is provided in the form of an expression vector with transcription and translation sequences in a suitable orientation and appropriate for the target cell. Expression vectors generally are replicable polynucleotide constructs that encode a polypeptide operatively linked to suitable transcriptional and translational controlling elements. Examples of transcriptional controlling elements are promoters, enhancers, transcription initiation sites, and transcription termination sites. Examples of translational controlling elements are ribosome binding sites, translation initiation sites, and stop codons. Protein processing elements may also be included: for example, regions that encode leader or signal peptides and protease cleavage sites required for translocation ofthe polypeptide across the membrane or secretion from the cell. The elements employed would be functional in the host cell used for expression. The controlling elements may be derived from the same DNA polymerase gene used in the vector, or they may be heterologous (i.e., derived from other genes and/or other organisms), as long as they are effective to a desirable degree in the target cell.

Preparing polypeptides and antibodies

Polypeptides of this invention, including compressing peptides, tissue targeting components, and intracellular targeting components, may be prepared by any suitable method known in the art. These include isolation of peptides from natural sources, enzymatic cleavage of larger proteins, and expression of a polynucleotide encoding the polypeptide in a suitable expression system, such as by operatively linking the coding strand to a suitable promoter, and transfecting into a suitable host cell. The host cell is then cultured under conditions that allow transcription and translation to occur, and the polypeptide is subsequently recovered. Polypeptides may also be prepared directly from sequence data by chemical synthesis. Several methods of synthesis are known in the art. A preferred method is the solid-phase Merrifield technique. This may be accomplished in automatic peptide synthesizers, or by supplying the amino acid sequence data to a commercial organization equipped to perform this technique.

Antibodies to be used as targeting components may be prepared by isolating a molecule such as a surface protein from the intended target cell, and using it as an immunogen to raise antibodies. It is often preferable to enhance the immunogenicity of a polypeptide by such techniques as polymerization with glutaraldehyde, or combining with an adjuvant, such as Freund's adjuvant. The immunogen is injected into a suitable experimental animal. Sera harvested from the immunized animals provide a source of polyclonal antibodies. Alternatively, immune cells such as splenocytes can be recovered from the immunized animals and used to prepare a monoclonal antibody-producing cell line. See, for example, Harrow & Lane (1988), U.S. Patent Nos. 4,472,500 and. 4,444,887. Antibodies can be purified from sera, tissue culture supematants, and ascites fluids by such techniques known in the art, such as protein A chromatography, ammonium sulfate precipitation, ion exchange chromatography, high-performance liquid chromatography and immunoaffinity chromatography on a column ofthe immunizing polypeptide coupled to a solid support.

Choice of material for encapsulation

A variety of substances can be delivered using this invention. Suitable substances include but are not limited to: polynucleotide expression vectors, small oligonucleotides, proteins, and small drugs and toxins. Substances may be delivered for therapeutic puφoses, to alter a cell for experimental puφoses, or as a diagnostic aid, for example, in radioimaging.

Polynucleotide expression vectors can be used in the compositions of this invention, typically for the puφoses of gene therapy. Gene therapy may be conducted either in vivo or ex vivo. Expression vectors can be used to either increase or decrease the level of expression of a polynucleotide or polypeptide by a cell for a variety of therapeutic puφoses. The design of expression vectors and their use in gene therapy is described in more detail in a later section.

Smaller oligonucleotides may be provided which have a more immediate and transient effect. One example is messenger RNA, comprising translation initiation and termination elements and an appropriate encoding region, for immediate translation within the target cell. Another example is antisense oligonucleotides that are capable of direct interference in transcription or translation of a particular gene. Antisense gene therapy, described in a later section, is preferable for prolonged down-regulation, whereas oligonucleotide antisense therapy is preferable for a more instantaneous effect. Antisense oligonucleotides are designed so as to form hybrid duplexes with the target polynucleotide inside the cell that are stable in the cell environment. The sequence can be selected so as to interact with a critical portion ofthe target, such as a region encoding a catalytic site. Generally, oligonucleotides are only long enough to form a duplex of sufficient specificity and stability, to increase the number of active molecules per gram of nucleic acid. Preferably, antisense oligonucleotides are about 15-100 bases in length, more preferably they are about 20-50 bases in length, and even more preferably they are about 20-30 bases in length. To improve the resistance ofthe oligonucleotides to enzymatic degradation and enhance their longevity, they may be synthesized with a non- naturally occurring backbone structure. Examples are methylphosphonate and phosphorotioate oligonucleotides.

Target genes for antisense oligonucleotide therapy of particular interest are oncogenes, and genes that encode cytokines, growth factors, enzymes producing lipid mediators, kinases, and viral enzymes. Particular examples are genes for lipopolysaccharide receptor, 5-lipooxygenase, prostaglandin G/H synthetase, proto- oncogenes, viral DNA and RNA polymerases, viral proteases, growth factors, tumor necrosis factor (TNF-α), interleukin- 1, interleukin-6, and platelet activating factor.

Lipid vesicles of this invention may also be used to deliver proteins for therapeutic puφoses. Particularly relevant are proteins and protein fragments that possess a binding or enzymatic activity capable of modulating a metabolic pathway within the cell. Examples include synthetic enzymes that participate in the pathways of gene transcription or translation, or intracellular pathways that are regulated by activation or cell surface receptors.

Lipid vesicles of this invention may also be used to deliver small drugs or toxins. While many small drugs do not require a delivery vehicle, some do, particularly where it is desirable either to target delivery to a particular target organ or cell type, or alternatively to sequester the drug during transit to the target (either because of its fragility, or because it is too toxic for other tissues). In order for a small drug to be suitable for delivery in a lipid vesicle, it should have the property of being unable to partition through the lipid membrane after encapsulation until it reaches the target cell. Small molecules of particular interest for delivery with this invention are azidothymidine (AZT), Taxonol, nucleic acid analogs, ricin A, methotrexate, vincristine, and other chemotherapeutics used in cancer therapy.

Compressing the polynucleotide using a positively charged peptide sequence

Where the material encapsulated by a liposome of this invention is a polynucleotide or oligonucleotide, it is desirable to reduce its radius, thereby facilitating encapsulation. This is accomplished by associating the polynucleotide with a suitable peptide or peptide complex. The peptide has the property of being able to bind to a polynucleotide, thereafter either directly pulling together regions ofthe polynucleotide to reduce the overall average radius, or inducing a folding change in the polynucleotide with the same effect. If the peptide is small in relation to the polynucleotide, then a number of peptides may be involved in the association reaction in order to compress the polynucleotide. Different peptides are suitable for the compression reaction. For example, proteins involved in the packaging of nucleic acids may be suitable, including histone proteins and genome-associated proteins present in the core of different viruses.

Especially preferred are short artificial peptides or peptide fragments with a predominant positive charge. These are readily obtained and more effective in reducing the overall size of a polynucleotide than any naturally occurring intact protein tested. The positive charge permits the peptide to associate readily with the polynucleotide. The positive charge may also play a role in permitting the compression to occur, since it partly neutralizes the predominant negative charge ofthe polynucleotide, thus allowing different regions ofthe strand to approach each other in three-dimensional space without the usual degree of electrostatic repulsion. The spacing of positive charges at intervals along the length ofthe compressing polypeptide may provide an overall avidity for the polynucleotide that is much higher than is possible for small ions, which in turn allows the formation of a more stable complex and a longer lived compression. In addition, the partial neutralization ofthe intrinsic charge on the polynucleotide is an advantage where the polynucleotide is to be encapsulated in a liposome with a partly anionic surface charge.

Compression has been achieved using synthetic polypeptides in which five lysine residues were present at the C-terminus, and believe that compression may be possible with as few as 3 consecutive lysines. Preferably, the compressing sequence is five consecutive lysine residues, or seven or more amino acids in length of which at least about 50% have a side chain that is positively charged at physiological pH (pH 7.0-8.0, especially pH 7.4). The natural amino acids lysine and arginine are examples of such positively charged amino acids. Other preferred examples are analogues of lysine that have a shorter or longer side chain with a titratable amino or imino group near the terminus. It is believer' that the compression is due largely to a charge effect, and that both D- and L- amino ds are suitable. Preferably, the sequence is at least 3 and less than about 100 amino acids in length, more preferably it is about 5 but no more than about 50 amino acids in length; more preferably, it is at least 7 but no more than about 30 amino acids in length; even more preferably, it is between 7 and about 20 amino acids in length. Preferably, at least about 50% ofthe amino acids in the sequence have a positively charged sequence at physiological pH; more preferably, at least about 65% of the amino acids in the sequence have a positive charge; more preferably, at least about 80% have a positive charge; even more preferably, essentially all of the amino acids in the sequence have a positive charge.

If the composition is to be used in intact animals, it is desirable to select a compressing polypeptide that is not immunogenic to the intended host. Immunogenicity may be predicted by comparison with other known amino acid sequences, particularly those of host origin. It may be tested experimentally either in the intended host species or in an animal model by administering small quantities of the isolated peptide in the same manner as is intended for the liposome preparation. To avoid immunogenicity, small peptides are preferred, as are peptides that are identical to or mimic other peptides or peptide regions which are natural host components.

The sequence ofthe compressing region may be attached to another sequence of a larger peptide. The additional sequence may play a spacer role, or it may have an additional function, such as providing an intracellular targeting signal.

Preferred examples of compressing polypeptides are shown in Table 1 :

TABLE 1: Exemplary Polypeptide Sequences for Compressing Polynucleotides

Sequence SEQ. ID NO:

KKKKK K 1

PKKKR KVKKK KK 2

PKKKR KVLKK KKK 3

In some applications, it may be desirable to deliver the polynucleotide encapsulated in the liposome to a particular location in a cell. This may be achieved by complexing the polynucleotide with a peptide comprising an intracellular localization sequence. In preferred embodiments of this invention, a peptide is used in the preparation ofthe polynucleotide that comprises both the properties required for compression and a localization sequence. Nuclear localization sequences are described, for example, in Dingwall et al. and Goldfarb et al. However, any peptide can be tested to determine whether it promotes localization to the desired intracellular organelle. In addition, as is known in the art, a nuclear localization peptide can be determined by observing its effect on the intracellular sorting of other proteins when they are attached to them by recombinant DNA methods (Lanford et al.).

Several preferred examples of nuclear localization peptides are provided in Table 2: TABLE 2: Intracellular Targeting Peptide sequences

Sequence SEQ. ID NO: Description:

PKKKR KV 4 SV40 virus nuclear localization sequence

KRPRE DDDGE PSERK RER 5 plant nuclear localization sequence

MLFNL RILLD DAAFR DGKKK 6 mitochondrial localization sequence

PKKKR KVKKK KK 2 SV40 NLS with additional polynucleotide compressing sequence

PKKKR KVLKK KKK 3

SEQ ID NO:4 provides a typical nuclear localization peptide similar to the SV40 T-antigen nuclear localization sequence. Any nuclear localization peptide can include a polylysine tail, which may also serve in the compressing ofthe polynucleotide as described above. Therefore, SEQ ID NO:2 is the sequence of SEQ ID NO:4 with an extension of 5 lysine residues. Fewer or more lysines can be utilized. Typically, about 3 to about 10 lysines can be used. A preferable length is about 4-5 lysines. SEQ ID NO:3 additionally contains a leucine residue inserted between the nuclear localization region and the polylysine tail, causing the peptide to be less linear. Another example of a nuclear localization peptide is shown in SEQ ID NO:5, which is a sequence utilized by plants to localize proteins to the nucleus.

Mitochondrial localization signals are known in the art to be characterized by being about 12-80 amino acids in length and to form amphipathic α-helical structures in the cytoplasm in which positively charged residues line up on one side ofthe helix while uncharged hydrophobic residues line up toward the opposite side. An example of a mitochondrial localization signal is listed herein as SEQ ID NO:6. A polylysine tail of a desired length, for example, from about 3 to about 10 lysine residues, can be included to enhance the ability ofthe peptide to compress the polynucleotide.

The polynucleotide is compressed by incubating it briefly with the polypeptide, for example, at room temperature or physiological temperature. An incubation at room temperature for 10-15 minutes is sufficient for the compression to occur. Suitable ratios are about 1 :5 to about 1 :200 peptide :polynucleotide on a wt wt basis; about 1 : 10 to about 1:100 is more usual, and 1 :30 is typically an effective ratio. Since the molecular weight ofthe polynucleotide is about several hundred fold higher than that ofthe polypeptide, the reaction is performed in molar excess of the peptide. It is predicted that the compressed polynucleotide— olypeptide complex comprises several short peptide per polynucleotide molecule when the polynucleotide is several kilobases in size or larger. There may be at least 5, or at least about 20, or even at least about 50 polypeptides associated with each polynucleotide. Optionally, excess peptide not consumed in the compression reaction may be removed by a technique such as gel filtration chromatography or ultracentrifugation, before the polynucleotide is used in a subsequent procedure, such as liposome encapsulation.

With little adjustment, the ratio of peptide:polynucleotide on a wt/wt basis described in the preceding paragraph has been found to provide effective compression of a variety of different polynucleotides. Preferred embodiments of this invention comprise compressed plasmids or expression vectors of double-stranded DNA of at least about 2 kilobases in size, preferably about 5 kilobases to about 8 kilobases in size, or even those of about 10 kilobases in size and larger, in either linear form or closed-circular form. The same ratios are also effective in compressing small single-stranded oligonucleotides of about 20-50 bases, such as might be used in antisense therapy. It is believed that similar peptide :nucleic acid ratios will also be effective in compressing RNA, and any other polynucleotide or polynucleotide analog, including branched or modified nucleic acid structures that might be suitable for therapeutic puφoses.

The exact ratio optimal for a particular application may be determined by routine experimentation. Efficacy ofthe compressing reaction may be evaluated using several techniques available to practitioners of ordinary skill, such as transmission electron microscopy (Example 2). Thus, the polynucleotide and polypeptide are combined at different ratios, and compared for the average overall radius compared to the naked polynucleotide in solution. Where the compressing reaction is being performed to enhance encapsulation in liposomes, a more relevant test may be the effectiveness ofthe preparation in liposome-mediated gene transfection experiments. The polynucleotide and polypeptide are combined at different ratios, encapsulated into liposomes, used to transfect target cells, and then the expression ofthe polynucleotide is measured to identify the ratio that provides the highest transfection levels. Formation of the liposome

The compressed polynucleotide may be encapsulated in a liposome of any composition. The lipid bilayer making up the liposome may comprise phospholipids, glycolipids, steroids, and their equivalents; amphipathic proteins, and lipid-soluble chemicals. Preferably, a composition is chosen that allows the envelope to be formed with reproducible qualities, such as diameter, and is stable in the presence of elements expected to occur where the liposome is to be used, such as physiological buffers and circulating molecules. Preferably, the liposome is resilient to effects of manipulation by storage, freezing, and mixing with pharmaceutical excipients.

In one preferred embodiment, the lipid bilayer ofthe liposome is formed primarily from phospholipids. More preferably, the phosphoiipid composition is a complex mixture, comprising a combination of phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylethanolamine (PE), and sphingomyelin (SM). The envelope may further comprise additional lipids such as phosphatidylinositol (PI), phosphatidic acid (PA), or cardiolipin (diphosphatidylglycerol). If desired, SM may be replaced with a greater proportion of PC, PE, or a combination thereof. PS may optionally be replaced with phosphatidylglycerol (PG). Preferably, at least PC and PE are included; more preferably, at least three ofthe group PC, PS, PE, and SM are included. The composition is chosen so as to confer upon the lipid envelope both stability during storage and administration, and fusogenic properties, especially for an outer or endosomal membrane ofthe target cell.

Practitioners of ordinary skill will readily appreciate that each phosphoiipid in the foregoing list may vary in its structure depending on the fatty acid moieties that are esterified to the glycerol moiety ofthe phosphoiipid. Generally, most commercially available forms of a particular phosphoiipid can be used. However, phospholipids containing particular fatty acid moieties may be preferred for certain applications. A particularly preferred phosphoiipid composition is shown in Table 3 : TABLE 3: Preferred Phosphoiipid Mixture

Phosphoiipid Approximate relative molar quantity

Dioleylphosphatidylcholine (DOPC) 0.259

Dioleylphosphatidylethanolamine (DOPE) 0.253

Sphingomyelin (SM) 0.287

Phosphatidylserine (PS) 0.175

Phosphatidylinositol (PI) 0.126

Cholesterol (CH) 1.000

Preferably, the envelope also includes cholesterol or a related steroid to improve the rigidity ofthe membrane, and enhance fusogenicity. Any amount of cholesterol may be used. A preferred ratio of total cholesterol to lipid is between about 0.5 and about 1.2 moles of cholesterol per mole of lipid. More preferred is a molar ratio of about 0.8 to about 1.2:1 ; even more preferred is a molar ratio of about 0.9 to about 1.1:1; still more preferred is a molar ratio of about 1.0:1.0. Other molecules that can be used to increase the rigidity ofthe membrane include cross-linked phospholipids.

It is particularly preferable to use a phosphoiipid mixture and the phospholipidxholesterol ratio of a viral particle which is fusogenic with eukaryotic cells. It is possible to extract lipid from a preparation of virus or defective viral particles, which is then reconstituted into liposomes, with or without viral proteins. More preferably, the lipid is a synthetic mixture of isolated components, modeled on the composition of a naturally occurring viral particle. An especially preferred viral model for designing the lipid composition is the human immunodeficiency virus (HIV). Optionally, this mixture can be further adjusted to enhance its fusogenic properties; for example, by introducing a degree of lipid assymetry in the liposome ultimately formed.

Other preferred liposomes for use in vivo are those with an enhanced ability to evade the reticuloendothelial system, thereby giving them a longer period in which to reach the target cell. Effective lipid compositions in this regard are those with a large proportion of SM and cholesterol, or SM and PI. Liposomes with prolonged circulation time also include those that comprise the monosialoganglioside GM1, glucuronide, or polyethylene glycol. Compositions that combine the attributes of stealth liposomes and viral particles are also encompassed by this invention. For example, cholesterol may be added at the ratios indicated above to a lipid mixture consisting of any combination of SM, PI, glucuronide, polyethylene glycol, and other suitable components.

The lipid vesicles may be prepared by any suitable technique known in the art. Methods include but are not limited to: microencapsulation, microfluidization, LLC method, ethanol injection, freon injection, the "bubble" method, detergent dialysis, hydration, sonication, and reverse-phase evaporation (reviewed in Watwe et al.). The selection of a particular method is made taking into consideration a number of criteria: a) size and lamellarity ofthe vesicles formed; b) stability ofthe vesicles; c) suitability for the intended payload; d) reproducibility; e) suitability for ultimate use, particularly scale- up and human administration, if contemplated; and f) convenience. For example, ultrasonication and dialysis methods generally produce small unilamellar vesicles; extrusion and reverse-phase evaporation generally produce larger sized vesicles. Techniques may be combined in order to provide vesicles with the most desirable attributes. One particularly preferred method is dialysis. The dialysis method generally produces vesicles of a more constrained size variation, and is especially preferred where smaller sized envelopes or envelopes with more reproducible characteristics are desired. A more detailed description ofthe dialysis method may be found, for example, in U.S. Patent 5,252,348. Briefly, the phospholipid/cholesterol envelope is prepared by solubilization ofthe lipids and cholesterol with sodium cholate or other appropriate detergent as the solubilizing agent, followed by removal ofthe detergent by exhaustive dialysis against phosphate-buffered saline (PBS). Other useful detergents are well known to those skilled in the art, and include any of those with a critical micelle concentration suitable for a dialysis technique. Non-limiting examples of suitable non-ionic detergents like CHAPSO™ and octylglucoside, and mild anionic detergents, particularly bile salts like sodium cholate, deoxycholate, taurocholate, and so on. The optimum detergent: lipid ratio is from about 10:1 to about 100:1, more usually about 30:1 to about 60:1 (w/w) and is preferably from about 40:1 to about 50:1 and is most preferably approximately 45:1. It should be noted that relatively small changes in these ratios can have significant effects. The skilled artisan can manipulate the procedures described in order to determine the optimum ratio. Removal ofthe detergent can be carried out using any of a number of techniques which are known to those skilled in the art. For example, bag, disc, flow-through, and counter-flow dialysis techniques and apparatus may be utilized. The time of dialysis and the volume of dialyzing buffer will depend on the relative micelle concentration that is an inherent property ofthe buffer being used. Detergents that partition easily between the micelle and the fully solubilized form dialyze more easily, and are consequently more preferred. Products ofthe dialysis method using the phospholipid-cholesterol mixtures suggested herein will be approximately 250 nm in diameter. The method is flexible so that batch sizes in a range of less than 5 ml to liter quantities can be prepared reproducibly and under sterile conditions using, for example, either teflon dialysis cells or flow-through hollow fiber dialysis apparatus.

Another particularly preferred method for preparing liposomes is the extrusion method. This can be performed rapidly and in small quantities, and is useful for preparing and evaluating various compositions, for example, in a laboratory setting. It may also be easily scaled up to commercial production.

For a description ofthe extrusion method, see for example, Schreier et al. (1992) and U.S. Patent Nos. 4,737,323 and 5,008,050. Briefly, small samples of phospholipid- cholesterol solution are prepared in a suitable solvent such as water/tert-butanol, and then dried by lyophilization. The solutions are frozen at -40°C, and dried for 24 h at -25^βC followed by 10 h at +10°C, by attachment to a vacuum source sufficient to keep the solution frozen. Alternatively, the solution may be prepared in an organic solvent such as chloroform, and then dried by rotoevaporation. The lipid coprecipitate is reconstituted by injecting sterile water under laminar flow and agitating gently for a few seconds. The concentration of lipid is about 100 mg/ml. This creates a preparation of lipid vesicles which are large and multilamellar. The dispersion is then extruded through a polycarbonate membrane of about 0.4 μm using an extruder, such as from Lipex Biomembranes, Vancouver BC. If desired, consecutive lyophilization, reconstitution and extrusion cycles are performed through membranes of about 0.2 μm. Preferably, the preparation is also subjected to several freeze-thaw cycles in baths of about -80°C and +40°C, to decrease the proportion of any remaining multilamellar vesicles. If desired, the liposome preparation can also be dialyzed to remove unwanted contaminants. The ultrastructure ofthe liposomes formed may be determined, for example, by transmission or freeze-fracture electron microscopy. The ultrastructure formed is preferably unilamellar. However, oligolamellar vesicles may also be acceptable for some puφoses. The diameter of liposomes may also be determined by electron microscopy. Liposome size may also be estimated by other techniques known in the art, such as quasi- elastic light scattering, gradient centrifugation, or gel filtration, using appropriate standards.

The size range of lipid envelopes is usually 50 to about 2000 nm, more preferably about 100 to about 1000 nm in diameter, and depends on the puφose for which they are being prepared. Vesicles encapsulating therapeutic substances of small molecular weight (< 3000 Da) are preferably about 50 to about 750 nm, more preferably about 75 to about 500 nm, more preferably about 100 to about 300 nm, and even more preferably approximately 200 nm in diameter. Vesicles encapsulating large polynucleotides or vectors are generally formed with larger diameters, typically 500 nm to 2000 nm, but may be sized to smaller diameters by extrusion. Smaller vesicles may be preferred where it is desirable that the liposome be able to penetrate complex tissue architecture or transgress barriers with a restricted pore size. For example, liposomes targeted to the hepatocyte must slip through sinusoidal fenestrations, and are preferably less than about 500 nm, more preferably less than about 300 nm, more preferably about 100-200 nm in diameter. Smaller vesicles may also have an advantage in evading the reticuloendothelial system, particularly Kupffer cells in the liver when administered to the general circulation. Liposomes containing large polynucleotides and targeted to other sites or designed for local administration are typically larger in size, generally about 200 to about 1000 nm, more preferably about 500 to about 1000 nm, even more preferably about 800 nm. In general, lipid envelopes in the preferred size range comfortably encapsulate the payload, and are of superior physical stability with an average size, and size distribution, that remain essentially unchanged over several months when stored under refrigeration.

Should the vesicles be undesirably large after the initial formation step, they may be reduced in size by shearing through a suitable filter, preferably a polycarbonate filter of about 0.2 μm. If lyophilized and reconstituted, vesicles that are sized in this fashion will reconstitute to a smaller size than they would otherwise. Cycles of extrusion, lyophilization, and reconstitution may be repeated until the vesicles reconstitute into a more desirable size. If it is necessary to adjust the size ofthe envelopes, it is preferable to do this early, since sizing may disturb encapsulated or surface-bound components. To obtain vesicles with an average diameter of about 250 nm, it is generally not necessary to adjust the size of vesicles obtained via the dialysis method, whereas vesicles obtained by the extrusion method may require several cycles of sizing, drying, and reconstitution.

The lipid envelopes can be freeze-dried and thereby preserved for extended periods of time. Freeze-drying, or other means of preservation, can be done either before or after further modification ofthe vesicles, such as the adding of tissue targeting components described in a later section. Freeze-drying may reduce the need for keeping preparations refrigerated. The stability ofthe lipid vesicles can also be improved by polymerization of one or more ofthe phosphoiipid components. Thus, a large reserve of concentrated lipid envelopes can be prepared and stored, and used to prepare individual batches with different surface proteins or other constituents when required.

Method of encapsulation

The encapsulated material to be delivered to the target cell may be placed inside a lipid envelope by including it in the solution used during initial formation ofthe liposomes. Altematively, it may be more efficient or more convenient to insert the payload into the liposomes after they are formed and optionally purified or characterized. Certain small molecules may be readily introduced into preformed liposomes by softening the liposomes with a detergent. The detergent is then removed from the solution, sealing the small molecules within the envelope. Larger molecules, particularly polynucleotides and proteins, may also be encapsulated either during initial vesicle formation, or subsequently.

Encapsulation during envelope formation is conveniently performed by using the dialysis method is used to form the vesicles. If the encapsulated material is a nucleic acid, it is first compressed with a suitable peptide, and then added to a lipid-detergent suspension as described in the previous section. The detergent is then dialyzed away as described, with the result that the vesicles form around the nucleic acid molecules that are present.

Alternatively, in the reconstitution method, lipid envelopes formed, for example, by the dialysis or extrusion method, are dried, for example, by evaporation on the surface of a suitable vessel, such as a glass-bottomed flask. A preferred method of drying is lyophilization. The compressed polynucleotide is then added to the vessel, which causes the lipid envelopes to reconstitute. During the reconstitution process, the polynucleotides are encapsulated into the envelopes. In a third method, vesicles are formed de novo by adding an aqueous solvent to a lipid film dried in a glass vessel. By including the compressed polynucleotide in the solvent, it becomes encapsulated as part ofthe process. One ofthe particular benefits of encapsulating polynucleotides compressed according to the methods described earlier is that large or heterogeneous liposomes may subsequently be manipulated, for example, by extrusion or freeze-thawing, with a suφrisingly low loss ofthe polynucleotide cargo. Liposomes encapsulating polynucleotides made by either ofthe techniques outlined in the previous paragraph may be from 100 nm to 5000 nm in size. Preferably, they are between 100 nm and 2000 nm in size; more preferably, they are between about 100 nm and 1000 nm in size; more preferably, they are between about 200 nm and 800 nm in size. Large liposomes may be reduced in size by extrusion through a suitable filter, such as a 0.2 μm polycarbonate filter. However, size reduction after encapsulation is avoided if not necessary, to further minimize polynucleotide release. Liposomes made up ofthe lipid-cholesterol mixtures suggested herein and sized to 250 nm during initial formation will generally reconstitute to about a 750 nm size during the encapsulation step. Conditions may be established without extensive experimentation, and will be reproducible once established.

The efficacy ofthe encapsulation reaction may be determined by any suitable technique available to the ordinary practitioner. Suitable techniques include transmission electron microscopy, labeling experiments, and liposome-mediated gene transduction experiments. For example, the polynucleotide may be provided with a radiolabel (such as ³²P) or a fluorescent label (such as TOTO™-l). The polynucleotide is then encapsulated into liposomes. Encapsulated polynucleotide is separated from non-encapsulated polynucleotide, such as by centrifugation, gel filtration chromatography, or fluorescence-activated sorting, and the proportion of label associated with the liposome fraction is determined. The encapsulated polynucleotides may also be evaluated for their ability to transfect a suitable target, such as test cells grown in culture. The degree of polynucleotide expression in the liposome-treated cells is compared with that for cells treated with unencapsulated polynucleotide alone.

During the encapsulation of a compressed polynucleotide, it is probable that the compressing peptide will also be enveloped into the liposome. In principle, the peptide may no longer be needed for compression puφoses once the polynucleotide is encapsulated. However, so long as the peptide does not interfere with the function ofthe polynucleotide once inside the cell, its presence in the lumen ofthe liposome is not an encumbrance. On the contrary: the compressing peptide may also comprise a sequence that performs a function once the liposomal contents enter the cell, such as targeting the polynucleotide intracellularly.

Tissue targeting components

The liposomes of this invention are designed for delivering the encapsulated material efficiently into cells. They may do this effectively without any additional components. For example, unmodified liposomes may be used directly to transfect cells in tissue culture. They may also be used for disseminated polynucleotide delivery in an individual, or may accumulate in particular target sites due to properties of size, charge, or natural adherence ofthe lipid envelope. They may also have a local effect if administered locally: for example, encapsulated polynucleotides given as an aerosol will generally exert their primary effect in the cells ofthe lung.

In preferred embodiments of this invention, liposomes are designed for more specific target delivery when given parenterally. This means that they accumulate in greater prevalence at certain tissue types in the animal they are administered to than related compositions. This specific accumulation may be a result ofthe size, charge, or solubility properties ofthe liposomes. More usually, the accumulation is due to a tissue targeting component present on the liposome that is capable of interacting with a host component that is more prevalent at the tissue than elsewhere. This interaction increases the duration during which the liposome remains in the vicinity, preventing its recirculation or promoting a more rapid functional interaction ofthe liposome with the target cell. In some applications, it may be preferable that the material encapsulated by the liposome be preferentially delivered to certain tissue sites. The liposome may also optionally be provided with a tissue targeting component that assists the liposome in preferentially localizing near the site ofthe target cell.

Suitable targeting components include but are not limited to: a) surface components that are present on tissue specific viruses or other pathogens; b) ligands and ligand analogues for which the target cell has receptors, adherence proteins, transmembrane transporters, or other specific recognition units; c) isolated naturally occurring recognition units from exogenous sources that are capable of distinguishing between cell types, such as lectins; and d) antibodies and antibody equivalents that have been raised against a similar tissue type as the target.

Relevant components in category a) include components of viruses, bacteria, and parasites, such as those listed in U.S. Patent 5,252,348. Useful components of such organisms include any complex molecule that facilitates binding ofthe virus to the cell, since incoφoration of these onto a liposome will assist it in localizing near the target cell. The component may be either an integral part ofthe model organism, or released by an organism, such as a toxin or toxoid subunit, which may be modified to permit it to be inserted into the liposome. Preferred examples are surface proteins of the Respiratory Syncytial Virus (RSV). This is a human pathogen ofthe paramyxovirus family, comprising a single-stranded RNA that replicates in the cytoplasm, and a pleomoφhic lipid-containing envelope. Liposomes targeted using RSV Glycoprotein G are expected to preferentially locate in the epithelial cells ofthe lung, especially if given by aerosol. Relevant components in category b) include cytokines of various sorts, such as growth factors like GM-CSF, interleukins, interferons, and TGF; hormones such as insulin and adrenaline; neurotransmitters such as serotonin; cell adhesion proteins such as ICAM and ELAM; cell recognition units, such as the CD antigens and their respective ligands; and carbohydrate ligands for mammalian lectins, such as complex carbohydrates with terminal gaiactose residues that is recognized by the hepatocyte asialoglycoprotein receptor.

Of particular interest are molecules that target to the lung. One example is triamcinolone acetonide phosphate (TAP), a ligand for the glucocorticoid receptor suitable for targeting liposomes (Gonzalez-Rothi et al.). A second example is surfactant protein A, suitable for targeting liposomes to alveolar type II cells (Walther et al.).

Especially preferred is the molecule serotonin, for targeting liposomes from the general circulation to lung endothelial cells. These cells are known to accumulate and inactivate serotonin (Block et al.). Specific binding with a high affinity of about 8 nM has been observed in mitochondrial fractions of lung tissue (Das et al.). Vesicles have been described with serotonin in the vesicle interior and directed towards hepatocytes (U.S. Patent 4,761,287). The present disclosure shows for the first time that serotonin attached to the liposome membrane is capable of directing the liposome for delivery to the lung. Delivery is suφrisingly effective, despite the fact that: a) serotonin receptors are apparently predominantly inside the lung endothelial cell, not on the cell surface; and b) since serotonin is a small molecule, it would not a priori be sufficiently accessible from the surface of a liposome. It is readily appreciated that serotonin analogs and other molecules capable of binding the same receptor may have the same ability to target liposomes to the lung. Suitable tissue targeting components for the lung may be identified by conducting inhibition studies, using the analog to compete with the binding of labeled serotonin, or preferably with serotonin-targeted liposomes, to lung endothelial cells. Liposomes targeted with molecules capable of competing with serotonin in this fashion are also encompassed in this invention.

Relevant components in category c) include plant lectins of various kinds that may be useful in targeting specifically to cells bearing certain carbohydrate structures.

Relevant components in category d) include antibodies and antibody analogues raised against a cell-surface antigen ofthe intended target cell. They may be directed against a particular autoantigen that is specifically associated within the host with a particular tissue type; such as the CD4 antigen on helper T cells. They may also be directed against an antigen that is associated with particular malignancies, such as the carcinoembrionic antigen. A third example is a monoclonal antibody against the lung endothelial anticoagulant protein thrombomodulin, which can be used to direct immunoliposomes to the lung (Mori et al.).

As illustrated, the tissue targeting component may be, for example, a lipid, a protein, a protein fragment, a glycoprotein, or a small molecule such as a hapten, or a combination of any of these. The component may be homogeneous, or it may be a cocktail of related components or components with different functions.

The targeting component is added to the liposome during the initial formation of the liposome, during a reconstitution step, or once it is already formed. Tissue targeting molecules that are not lipophilic are prepared by conjugation to a second molecule which can be assembled into the lipid bilayer. The example section below provides illustrations wherein serotonin is covalently coupled to phosphatidyl ethanolamine or to cholesterol. The conjugate may include a spacer between the lipophilic moiety and the targeting component that enhances accessibility.

Targeting components that are chemically related to lipids or sterols, or attached thereto, are generally more conveniently added to the liposome during initial vesicle formation. The incoφoration ofthe targeting component is performed by mixing the component or conjugate with the other phospholipids before the initial formation. As illustrated in the example section, serotonin conjugated to cholesterol can be used as a component in the dried lipid film, which is then reconstituted by an aqueous solution into a preparation of liposomes displaying them on the surface. In an alternative approach, vesicles are formed comprising a functionalized lipid (U.S. Patent 5,059,421). The tissue targeting component is subsequently coupled onto the liposome by activating the functionalized lipid appropriately.

Proteins and small peptides generally must comprise a lipophilic region in order to be incoφorated into the vesicle in a stable manner. If they do not already comprise such a region, they may be conjugated or cross-linked to a second peptide with this property, or synthesized as a fusion protein having a membrane spanning region. Alternatively, peptides may be conjugated to a phosphoiipid or steroid. Conjugation of peptides onto myristic or palmitic acid is preferred, and is illustrated in the example section. Appropriately prepared peptides may be added to the lipid mixture during formation ofthe liposomes or during reconstitution. More preferably, the protein is inserted into the membrane by partial micellation after the liposome has already been formed. This results in an asymmetric distribution ofthe protein, with a majority being oriented outwards and thus available to exert its targeting role. This has the combined advantage of improving the number of accessible molecules per gram used in the preparation, and not taking up space in the interior which would otherwise contain material for delivery to the target cell.

To conduct the insertion method, the membrane is treated with sodium deoxycholate or other appropriate detergent at an approximate ratio of 8:1 and removal of the detergent by exhaustive dialysis. Appropriate detergents are those suitable for dialysis, particularly anionic detergents, and preferably bile acids. As will be appreciated by a person skilled in the art, the term "partial micellation" refers to a vesicle membrane which is "softened" to the point that the vesicle flattens out and acquires a disc- or dumb¬ bell-like shape which reverses into a vesicular structure upon removal ofthe detergent; however, the vesicles are not solubilized (micellized) to the point that they lose their intrinsic bilayer structure and become true mixed micelles again.

This process can conveniently be controlled by monitoring the scattering of light ofthe vesicles using a laser light scattering instrument. The selection of the detergent is determined taking into account the compatibility of a particular detergent with the surface protein to be inserted. Enough detergent is introduced into the vesicle dispersion to maintain the light scattering signal. Loss ofthe light scatter signal indicates true solubilization, thus excess of detergent and loss ofthe vesicular structure in favor of a micellar structure. In the example section that follows, the preferred molar ratio of detergent:lipid that maintains the partially micellated disc-like vesicular structure was found to be between 20:1 and 3:1, preferably 12:1 to 5:1, and more preferably about 8:1. The protein is then added to the softened liposome particles and allowed to insert into the lipid surface. Conditions that optimize the amount of tissue targeting component can be determined by routine experimentation using several techniques available to practitioners of ordinary skill. These include conducting experiments in which the molecule to be inserted is provided with an appropriate label, such as a radioisotope. The insertion reaction is conducted at various targeting component: liposome ratios. After the insertion reaction, liposomes are separated from unincoφorated targeting component, for example, by centrifugation or chromatography, and the proportion of label associated with the liposomes is determined. Altematively, antibodies specific for the targeting component or cell surface protein from the intended target cell may be used with each preparation to detect the presence of incoφorated targeting component that is expressed at the surface. In another approach, liposomes with an encapsulated expression vector and various amounts of targeting component are used to transfect a suitable test cell line. The expression levels are then compared between samples and those treated with liposome encapsulated vectors into which a targeting component has not been added.

In principle, it is possible to perform the encapsulation step and the insertion of a tissue targeting protein in either order with acceptable results. The optimal order for any particular application of this technology will depend on a number of different factors, including the method used for initial formation ofthe liposomes, the method used for encapsulation, the nature ofthe material to be encapsulated, the stability ofthe targeting component to detergents and subsequent manipulations, and the method (such as refrigeration or freeze-drying) whereby the preparation is to be preserved. Where the material to be encapsulated is a large molecule such as a compressed polynucleotide, two general approaches are preferred. For the purposes of forming small quantities for evaluation puφoses, the liposomes are preferably formed by extrusion, sized to a smaller diameter, inserted with a tissue targeting protein, freeze-dryed, reconstituted in the presence of a material to be encapsulated, and then re-sized if necessary. For commercial production, the liposomes are preferably formed by dialysis in the presence of material to be encapsulated, and then inserted with a tissue targeting protein by micellation as a final step.

Fusogenic components

Once a liposome localizes near a target cell, delivery ofthe encapsulated material inside the cell may be facilitated if a fusogenic component is present. While not wishing to be bound by theory, it is envisioned that a fusogenic component disrupts the membrane ofthe target cell, permitting entry ofthe liposome or fusion ofthe liposome membrane with that ofthe cell. Fusion may occur at the cell surface, or in an intracellular compartment such as an endosome or a lysosome. Particular fusogenic components are only active at the lower pH of an endosome.

A number of fusogenic components are suitable for use with this invention. Mammalian and a wide variety of viral proteins with fusion properties are known in the art and suitable for use in this invention. Examples are listed in the Background section of this application. Another class of fusogenic molecules are artificial peptides with sequences modeled on those previously known to be fusogenic. To maximize fusogenic potential per gram while minimizing immunogenicity, it is preferable to use only the fragment ofthe relevant protein which contains the fusogenic activity. Particular fusogenic proteins of interest are: gp41 of Human Immunodeficiency

Virus- 1, members of the RSV surface Glycoprotein superfamily, papilloma virus Ll protein, Coronavirus S Protein, Rabies virus Glycoprotein G, Glycoproteins D, B, H, and L ofthe Heφesvirus family (especially VSV), the two Rhino virus pH-dependent fusogens, the Paramoxa Virus F protein petide region 59-140, the Murine Leukemia Virus R Protein, and the Flavovirus pH 6.3 E & M proteins. A preferred fusogenic protein is hemagglutinin of influenza, or a fusogenic fragment thereof, particularly a peptide consisting essentially ofthe sequence shown in SEQ. ID NO:7.

In certain embodiments of this invention, the fusogenic component is encapsulated within the liposome and not accessible from the cell surface. Encapsulated fusogens may promote delivery of liposomal contents by any of a number of mechanisms. For example, once the liposome reaches the target cell, the component may tunnel through the liposome and through the cell membrane, and be followed by the encapsulated therapeutic substance. Altematively, the component may be released by disintigration ofthe liposome in an endosomal compartment, and go on to penetrate the endosomal membrane along with the payload. Example 13 illustrates that a fusogenic fragment of hemagglutinin encapsulated in a liposome enhances expression ofthe vector payload in the target cell.

More usually, the fusogenic component is anchored into the membrane. Accordingly, a fragment is selected with a lipophilic region, or else prepared with a lipophilic region attached, as described earlier for tissue targeting peptides. Fusogenic peptides may be prepared so as to be oriented either C-terminal or N-terminal outwards, and can optionally comprise additional spacer regions which preferably are non- immunogenic.

Either fusogenic or tissue targeting components may be used individually to enhance the ability of a liposome to exert its therapeutic effect. Example 12 illustrates that fusogenic and tissue targeting components, when both present, may have an additive or even a synergistic effect. Accordingly, preferred embodiments of this invention are bifunctional liposomes having both fusogenic and tissue targeting components. Particularly preferred are bifunctional liposomes wherein the tissue targeting component and the fusogenic component do not naturally occur together on a single molecule or a single viral particle.

Where both are present, fusogenic and tissue targeting components may be inserted into the membrane simultaneously, or through separate process, as appropriate for their chemical composition. Fusogenic and tissue targeting components may be comprised in a single molecule, particularly a fusion proteins, optionally through a polypeptide linker region. More usually, the fusogenic component and the tissue targeting component are present as separate molecules and anchored separately into the liposome membrane.

Use of liposomes containing compressed polynucleotides for gene therapy

Embodied in this invention are liposome compositions comprising polynucleotides with a therapeutically relevant genetic sequence. Encapsulated polynucleotides of this invention can be used for administration to an individual for puφoses of gene therapy. Suitable diseases for gene therapy include but are not limited to those induced by viral, bacterial, or parasitic infections, various malignancies and hypeφroliferative conditions, autoimmune conditions, and congenital deficiencies.

Gene therapy can be conducted to enhance the level - expression of a particular protein either within or secreted by the cell. The liposome-encapsulated polynucleotides provided herein can be employed to transfect cells either for gene marking, replacement of a missing gene, or insertion of a therapeutic gene. For example, marker genes can be used to monitor the state of disease, the longevity of undesirable cells in the diseased tissue, or the longevity ofthe modified or transplanted cells in the diseased tissue. The invention can also be used to supply a cytokine or mediator important in enhancement or attenuation of an immunological or inflammatory reaction, such as may occur during viral infection, autoimmune disease, or septic shock. An example of a cytokine of particular interest in this regard is α-interferon.

Altematively, the invention can be used in the replacement of a defective or missing gene in a human cell to correct its malfunction. One example is replacement of the adenosine deaminase gene in ADA deficiency. Another example is the correction of the genetic defect of cystic fibrosis, by supplying a properly functioning cystic fibrosis transmembrane conductance regulator (CFTR) or biologically active fragment thereof to the airway epithelium. An a antitrypsin encoding sequence may be used for treatment of null-type hereditary emphysema. A sequence encoding arantitrypsin or superoxide dismutase may be used for treatment of certain inflammatory conditions. α_rantitrypsin may be used as an adjunct therapy for cystic fibrosis. Adult respiratory distress syndrome (ARDS) may be amenable to treatment by a combination of polynucleotides encoding human αi-antitrypsin, PGH synthase, and a cocktail of inflammatory inhibitors. In a further example, a gene related to prostaglandin synthesis may be used for the treatment of asthma.

To enhance expression of a particular protein, a polynucleotide is provided with a sequence encoding the protein, operatively linked to transcription and translation elements that are likely to be active in the target cell. Such transcription and translation elements may be those that occur naturally in the corresponding gene. Altematively, an expression cassette may be constructed by recombinant nucleic acid chemistry to place encoding sequences in operative linkage with a heterologous promoter that is intrinsically active in the target cells, or that can be induced by a suitable agent. Minimally, the encoding region will encode the region ofthe protein that provides the desired biological activity. Mutations from the natural sequence that enhance catalysis, improve specificity, or avoid unwanted regulatory activity may optionally present. Preferably, the encoded protein will include elements that participate in its transport to a desired location within or outside the cell, such as a signal sequence for protein secretion. Alternatively, a polynucleotide may be provided to the cell that decreases the level of expression. This may be used for the suppression of an undesirable phenotype, such as the product of a gene amplified or overexpressed during the course of a malignancy, or a gene introduced or overexpressed during the course of a microbial infection. One approach to decreasing expression is antisense gene therapy. The therapeutic polynucleotide comprises a sequence or complementary sequence that is capable of forming a stable hybrid with either the target gene itself, or more typically RNA transcribed therefrom, operatively linked to a suitable promoter. In this way, multiple copies may be transcribed in the target cell, which in rum form stable hybrids with polynucleotides encoding the target protein. The antisense polynucleotide need not be the exact complement ofthe target polynucleotide to be effective, so long as stable hybrids form under physiological conditions.

A second approach to decrease expression is to provide a polynucleotide that contains or encodes a ribozyme capable of cleaving the relevant mRNA. Typically, the therapeutic polynucleotide contains an encoding region operatively linked to a suitable promoter that allows it to be transcribed into a ribozyme in the target cell. The ribozyme comprises a catalytically active segment flanked by nucleotide recognition sequences that serve to anneal the ribozyme to the RNA in a site-specific manner. Only an amount of complementarity sufficient to form a duplex with the target RNA and to allow the catalytically active segment ofthe ribozyme to cleave at the target sites is necessary. The enzymatic RNA molecule is formed in a hammerhead motif (Rossi et al., AIDS Res. Hum. Retrovir. 8:183, 1992), a haiφin motif (Hampel et al.., Biochem. 28:4929, 1989; Nucl. Acids Res. 18:299, 1990), or any other motif known in the art and capable of providing the desired activity (e.g., Cech et al., US 4,987,071). Therapeutically useful ribozymes may be selected using the sequence of the target polynucleotide from a library of ribozymes flanked by random sequences cloned into the loop region of a ribozyme expression cassette.

Compositions of this invention may be used to conduct gene therapy either ex vivo or in vivo. Ex vivo gene therapy is outlined in US Patent 5,399,346. In the present context, cells are removed from a donor (or obtained from a cultured cell line), genetically altered with a liposome-encapsulated polynucleotide of this invention, and then administered to a recipient. The cells are obtained from the donor in the form of a blood sample, bone marrow aspirate, biopsy, surgical excision, or other clinically suitable procedure. The cells are optionally purified or otherwise subfractionated, and then treated with a liposome. After transduction, the cells are optionally cultured or otherwise manipulated, and then administered to the recipient. Preferably, the cell donor is the same as the recipient ofthe transduced cells (an autologous transplant). However, the transfer of cells from one individual to another is permissible, or even preferred where the recipient does not have sufficient donor cells for autologous treatment. The donor is preferably histocompatible and blood group identical or compatible with the recipient, although this may be less important for administration of cells that are normally immunologically privileged, such as those in the liver. Non-limiting examples of tissue types suitable for ex vivo gene therapy are stem cells, obtained either from bone marrow or peripheral blood, such as CD34+ cells, and hepatocytes.

Liposome-encapsulated polynucleotides of this invention may also be used for administration directly to an individual for purposes of gene therapy in vivo. The method comprises administering an effective amount via one ofthe modes of administration described in the following section. What constitutes an effective amount depends on the condition ofthe recipient and the objective of treatment. Where a low percentage of transduction can cure a genetic deficiency, then the objective of treatment is generally to meet or exceed this level of transduction. In some instances, this level of transduction can be achieved by transduction of only about 1 to 5% ofthe target cells, especially if the gene is normally expressed by a small proportion of cells or at a modest level, or if the therapeutic gene is provided under control of a more active promoter. In other instances, the treatment will provide a better degree of protection or a longer lasting effect if a large percentage of cells in the target tissue are modified. In these instances, a sufficient number of liposomes should be admimstered preferably to genetically alter at least about 20% ofthe cells ofthe desired tissue type, usually at least about 50%, preferably at least about 80%, more preferably at least about 95%, and even more preferably at least about 99% ofthe cells ofthe desired tissue type. Generally, dosage will approximate that which is typical for the administration of a polynucleotide, particularly one that remains extrachromosomal, and is typically in the range of about 50-500 μg DNA per kg. The treatment can be repeated every two or three weeks or as required by the attending physician. The effectiveness ofthe genetic alteration can be monitored by clinical features, and by determining whether the cells express the function intended to be conveyed by the therapeutic polynucleotide. Samples removed by biopsy or surgical excision may be analyzed by in situ hybridization, immunohistology, or immunofluorescent cell counting.

Preparation of pharmaceutical compositions and their administration

Liposomes of this invention comprising any suitable encapsulated material may be prepared for administration to an individual in need thereof, particularly humans, in accordance with generally accepted procedures for the preparation of pharmaceutical compositions. Preferred methods for preparing the liposomes described herein are sufficiently flexible that batch sizes from 5 ml to several liters or more may be prepared reproducibly and under sterile conditions, using (for example) either teflon dialysis cells or flow-through hollow fiber dialysis apparatus. General procedures for preparing pharmaceutical compositions are described in

Remington's Pharmaceutical Sciences, E.W. Martin ed., Mack Publishing Co., PA. Liquid pharmaceutically administrable compositions can, for example, be prepared by dispersing a liposome in a liquid excipient, such as water, saline, aqueous dextrose, glycerol, or ethanol. The composition may optionally also contain other medicinal agents, pharmaceutical agents, adjuvants, carriers, and auxiliary substances. Pharmaceutical compositions for oral, intranasal, or topical administration can be supplied in solid, semi-solid or liquid forms, including tablets, capsules, powders, liquids, and suspensions. Compositions for injection can be supplied as liquid solutions or suspensions, as emulsions, or as solid forms suitable for dissolution or suspension in liquid prior to injection. For administration via the respiratory tract, a preferred composition is one that provides either a solid or liquid aerosol when used with an appropriate aerosolizer device. Although not required, pharmaceutical compositions are preferably supplied in unit dosage form suitable for administration of a precise amount.

As will be readily appreciated, components for preparing a pharmacological composition of this invention may be provided in kit form. One example is a kit comprising in separate containers liposomes and a pharmaceutical excipient. Another example is a kit comprising a compressed polynucleotide in one container, and a lipid film in another container that reconstitutes into liposomes when the polynucleotide solution is added. Other combinations are possible. Packaged compositions and kits of this invention optionally include instructions for storage, preparation and administration ofthe composition.

The route of administration of a pharmaceutical composition depends, inter alia, on the intended target site, the presence or absence of a tissue targeting component, and the nature ofthe condition being treated. Possible routes of systemic administration include parenteral, intramuscular, subcutaneous, intradermal, intravenous, oral, intraperitoneal, and intranasal routes. Local administration is preferred when the liposome lacks a tissue targeting component and the target area is small. Liposomes may be administered near to or directly into a pathologically affected tissue site, either by direct injection, or through an indwelling catheter.

Where the liposome is intended for administration to the epithelium, a preferred method of administration is by aerosol. The aerosol may be either solid or liquid. Liquid aerosols may be prepared by suspending the liposomes in a suitable excipient that may also contain a preservative, and supplied in a suitable aerosolizer. Solid aerosols may provide a more stable form ofthe product, and may therefore be preferred. They can be prepared by drying down a liposome preparation in the presence of a carrying agent, such as lactose or gaiactose. A preferred method of drying is lyophilization. The lyophilized powder may be ground and compressed into a capsule. It is then placed in a suitable device to provide a particulate spray. The liposomes will then reconstitute when they contact the fluid bathing the alveolar surface.

The decision of whether to use in vivo or ex vivo therapy, and the selection of a particular composition, dose, and route of administration will depend on a number of different factors, including but not limited to features ofthe condition and the subject being treated. The assessment of such features and the design of an appropriate therapeutic regimen is ultimately the responsibility ofthe prescribing physician.

The foregoing description provides, inter alia, a detailed explanation of how liposomes can be prepared that contain compressed nucleic acids or other deliverable material. The liposomes may optionally also comprise a tissue targeting component and an intracellular targeting component. It is understood that variations may be made to the composition ofthe liposome and the methods used for preparing and using them without departing from the spirit of this invention.

The examples presented below are provided as a further guide to a practitioner of ordinary skill in the art, and are not meant to be limiting in any way.

EXAMPLES

Example 1: Preparation of polynucleotide plasmids with reporter sequences

Plasmids containing cDNA encoding either chloramphenicol acetyltransferase (pCMV4-CAT) or prostaglandin G/H synthase (pCMV4-PGH) were constructed in the eukaryotic expression vector pCMV4 driven by the promoter sequence ofthe cytomegalovirus major immediate early gene (Conary et al., Canonico et al.). pCMV4 is a plasmid made up of double-stranded DNA. The plasmid RV-PAP, which contains the gene for placental alkaline phosphatase driven by the Rous sarcoma virus promoter (pRV- PAP), was obtained from L. Culp, Case Western Reserve. The plasmids were replicated in E. Coli strain NM522 and purified using Qiagen Gigapreps (Chatsworth, CA) according to the manufacturer's protocol. Purity ofthe plasmid and integrity ofthe cDNA insert was determined by electrophoresis using 1% agarose gels (D. Voytas (1992) pp. 2.13-2.14 in: Short Protocols in Molecular Biology, F.M. Ausubel et al., John Wiley & Sons, NY).

For localization experiments, plasmids were labeled by dye intercalation with acridine orange or with the dimeric cyanine nucleic acid stain, TOTO-1™ (Molecular Probes, Eugene, OR). The integrity ofthe labeled plasmid and the absence of any unincoφorated dye was determined by agarose gel electrophoresis. The labeling procedure was as follows: Plasmid at a concentration of 0.5 μg/ml in sterile water was combined with 10 μl of a 10 mg/ml solution of acridine-orange and incubated for 45 minutes in the dark. The plasmid DNA was precipitated with sodium acetate and ethanol, collected by centrifugation, washed once with 70% ethanol and resuspended in sterile water at a concentration of approximately 1 mg/ml. TOTO-1™ labeling was performed in 40 mM Tris acetate, pH 8.0, containing 2 mM EDTA. DNA at a concentration of 1 mg/ml in the Tris acetate buffer was added to a 1 : 10 dilution ofthe dye at a ratio of 1 :3 v/v. The DNA was precipitated with sodium acetate and ethanol, collected by centrifugation, washed with 70% ethanol and resuspended in Tris acetate buffer at a concentration of 0.5 mg/ml. The absence of unincoφorated dye was confirmed by electrophoresis in 1 % agarose gels.

Fluorescein- 11-dUTP (FluoroGreen, Amersham, Arlington Heights, IL) also was used to label plasmid DNA. The plasmid DNA was linearized by digestion with the restriction endonuclease Xba 1 and purified by chromatography over a Select 6-L spin column (5 Prime 3 Prime, Boulder, CO). Fluorescein- 11-dUTP was added to the 3' ends ofthe plasmid through the action of terminal deoxytransferase, the plasmid was purified by chromatography through a Select 6-L column, precipitated with ethanol, collected by centrifugation, washed with 70% ethanol and resuspended in sterile water. An aliquot 10 μg of plasmid DNA was digested with a restriction endonuclease to linearize the plasmid. Some experiments were conducted using radiolabeled plasmid DNA. To perform the labeling, the 5' ends of the DNA were labeled using [³²P]-ATP via enzymatic action of T4 polynucleotide kinase. The new labeled plasmid DNA was separated from unreacted labeled ATP and T4 kinase by column chromatography.

Example 2: Compressing of the polynucleotide plasmids

Synthetic peptides with sequences corresponding to SEQ. ID NOS:2-4 were ordered and obtained from Research Genetics (Huntsville, AL). The peptides were combined with plasmid DNA at varying DNA to peptide ratios and incubated at 23 °C for 10-15 minutes. These proteins were combined with plasmid at a ratio of 1 :30

(Protein: Plasmid, wt/wt) and incubated for 10 minutes at room temperature. Poly-L- lysine (Sigma, St. Louis, MO) and whole histones (Boehringer Mannheim, Indianapolis, IN) also were used in some experiments under similar conditions.

The ability ofthe peptides to compress the plasmids was evaluated by transmission electron microscopy (TEM). Carbon coated grids were prepared by resistance evaporation of carbon thread (Bal-Tec, Hudson, NH) onto freshly cleaved mica strips. Films were floated off on water and picked up on 400 mesh nickel grids that had been treated with grid glue (transparent tape extracted with chloroform) and dried. CLPCs were allowed to adsorb to the surface ofthe carbon film for 20 min; grids were floated face down on a drop of 1% OsO₄ for 1 h, washed, dried and rotary shadowed with carbon-platinum at 10° angle by beam evaporation in a Balzers MED010 evaporator. Specimens were observed with a Hitachi H-7000 electron microscope.

An experiment was performed with the pCMV-PGH plasmid, which is about 7,800-7,900 base pairs in size. The images showed that before compression, the plasmids were predominantly in an open supercoiled form (Figure 1 , Left Panel). When the plasmid was combined with the peptide with the sequence of SEQ. ID NO:2, it was found to be substantially compressed (Figure 1, Right Panel). Note that the magnification ofthe Right Panel of Figure 1 is about 2.5 times that ofthe Left. The addition of either poly-L- lysine or intact histones to the plasmid altered the structure in a similar fashion (Figure 2, Left and Right Panels, respectively), but the compression was found generally to be less than with the small peptide.

The net charge on plasmid DNA and on plasmid DNA/peptide complexes was determined by a gel electrophoresis gel shift experiment. Plasmid DNA and synthetic peptides were combined as described above. After the incubation period, the plasmid subjected to electrophoresis (7 V/cm). After 90 minutes, DNA in the gel was detected by staining with ethidium bromide and irradiation with ultraviolet light.

The results are shown in Figure 3. Lane 1 was loaded with 3 μg ofthe pCMV- PGH plasmid that had not been compressed. Lanes 2-5 were loaded with the same amount of plasmid compressed with 15 ng, 100 ng, 200 ng, and 400 ng ofthe compressing peptide. The horizontal lines in each lane at about 8.2 on the scale indicate the top ofthe gel. The uncompressed intact plasmid appears as a doublet at about 8.8 and 9.2 in Lane 1. With increasing amounts of peptide, the plasmid resolves to a single band at lower relative mobility. This is consistent with the plasmid having adopted an altered configuration.

Example 3: Preparation of artificial viral envelopes Phospholipids were purchased from the following sources: egg phosphatidylcholine (PC) (lot #37F-8420), phosphatidylserine (PS) (lot #99F-83561) from bovine brain, egg phosphatidylethanolamine (PE) (lot #58F-8371), cholesterol from porcine liver (lot #36F-7040), deoxycholic acid (lot #108F-0331) and sodium cholate (lot #78F-0533), were from Sigma Chemical Co., St. Louis, MO. Egg sphingomyeline (SM) (lot #ESM-22) was from Avanti Polar Lipids, Pelham, AL. The composition of phosphate buffered saline (PBS) was NaCl 137 mM, KCl 2.7 mM, Na₂HPO₄ 8.1 mM, KH₂PO₄ 1.5 mM, with 0.5 mM sodium azide (lot #13F-0600) (Sigma). Spectra Por 2 (mol. cut-off 12-14,000) membrane discs were used for dialysis in teflon dialysis cells.

Stock lipid solutions were prepare as shown in Table 4. Briefly, enough cholesterol or phospholipids were dissolved individually in 10 ml chloroform to give the concentrations indicated in Table 4. Sodium cholate stock solution was prepared in methanol.

TABLE 4: Lipid C Composition a nd Stock Soi utiσns mg/10 ml MW μ moles/ mole % of HCCI₃ 10 ml total PL cholesterol (CH) 38 386 98.4 phosphatidylcholine (PC) 20 786 25.4 23.7 phosphatidylethanolamine (PE) 18 743 24.2 22.6 phosphatidylserine (PS) 23 832 27.6 25.7 sphingomyelin (SM) 22 731 30.1 28.1

TOTAL phosphoiipid (PL) 83 107.3 100.0

TOTAL lipid (including CH) 121 205.7 sodium cholate 2000¹ 430.6 4644.7 in methanol

As shown in Table 5, the ratio of total phospholipid holesterol was approximately 1 :1. The detergen total lipid ratio was 9300 μmol/10 ml to 200 μmol/10 ml, or approximately 45: 1. TABLE 5: Lipid:Choleβterol and Lipid: Detergent Ratios

Concentration (μmol 10 ml)

Total Cholesterol Molar Ratio Phosphoiipid

| Initial 107.3 98.4 0.92

Recovered after Dialysis 79.4 (74.0%) 69.2 (70.3%) 0.87

¹Total amount of detergent used: 4 mg (20 μl) = 9300 μ oles

The phosphoiipid composition was based on the reported composition of the natural HIV-l envelope (L.M. Gordon et al. (1988), pp. 255-294 in: Lipid Domains and the Relationship to Membrane Function, R.C Aloia et al., Alan R. Liss, Inc., NY), as shown in Table 6. The minor fractions of 2.1 mole% phosphatidylinositol and 0.9 mole% phosphatidic acid, and the 5 mole% of "other" lipids were substituted by a larger fraction (25.7 mole% vs. 15.1 mole%) of PS.

TABLE 6: Lipid Composition of Artificial and Natural HIV-1 Envelope

Mole % of Total Phospholipids

LIPID C PE SM PS PI² PA^J Other

Natural¹ 23.8 24.6 28.3 15.1 2.1 0.9 5.0

Artificial 23.7 22.6 28.1 25.7 n.a. n.a n.a.

¹ Gordon et al., supra. PI = phosphatidylinositol ³PA = phosphatidic acid n.a. = not added

Of every lipid stock solution, 500 μl were combined in a round-bottom flask and 1000 μl ofthe sodium cholate stock solution were added. The organic solvent was removed under a stream of nitrogen.

The lipid/detergent film was dispersed in 5.0 ml 10 mM PBS and sonicated for 10 minutes in a bath sonicator (Lab Supplies, Hicksville, NY) until solubilization ofthe lipids was completed. The clear liquid was transferred to a teflon dialysis cell equipped with a Spectra Por 2 membrane (MW cut-off 12-14,000) and dialyzed against 2 liters of PBS with 5 buffer changes after 5, 8, 16, 24, and 48 hours. The buffer was purged with N₂ over the entire time of dialysis. The samples were removed form the dialysis cell after a total dialysis time of 54-56 hours and stored at 4°C.

The size and size distribution ofthe artificial envelopes was analyzed using a NICOMP™ Model 370 laser particle sizer (Particle Sizing Systems, St. Barbara, CA). A typical example of a homogenous population of vesicles had an average size of 216 nm ± 82 nm (S.D.) and achi² value of 1.39. The reproducibility of preparation was remarkable. A total of 15 samples prepared was found to have an average diameter of 250 mm with an extremely narrow standard deviation ofthe mean of 26 nm. The ultrastructure ofthe vesicles was determined by freeze-fracture electron microscopy. The results ofthe electron microscopy showed perfectly unilamellar artificial envelopes.

Cholesterol was determined according to the method of A. Zlatkis et al. (1953), J. Lab. Clin. Med. 41:486-492. A total of 267.1 μg CH/ml, corresponding to 76.3% ofthe original total amount of CH, were recovered.

For phosphoiipid analysis a sample was extracted according to the method of E.G. Bligh et al. (1959), Can. J. Biochem. Phys. 39:911-917. A quantitative phosphoiipid assay was performed according to the method of J.C.M. Stewart (1980), Anal. Biochem. 104:10-14. In a typical experiment, a total of 613.8 μg PL/ml, corresponding to 74.0% of the original total amount of PL, was recovered.

The final phosphoiipid: cholesterol ratio was 0.87, only slightly different than the original ratio of 0.92.

Example 4: Demonstration that compressed polynucleotides are more readily encapsulated

Compressed polynucleotides are encapsulated into AVE particles by a reconstitution method. AVE particles are formed beforehand; the empty AVE are then dried down into a lipid film in a round bottom flask. A solution containing the compressed polynucleotide is added, and as the AVE become suspended in the solution, the polynucleotide is encapsulated. The AVE formed for this Example, and for Examples 8, and 9, were prepared by the extrusion method. The phospholipid-cholesterol mixture used was as shown in Table 7:

TABLE 7: Preferred Phosphoiipid Mixture

Phosphoiipid Relative molar quantity

Palmitoleylphosphatidylcholine (POPC) 0.259

Egg Phosphatidylethanolamine (PE) 0.253

Sphingomyelin (SM) 0.287

Phosphatidylserine (PS) 0.175

Phosphatidylinositol (PI) 0.126

Cholesterol (CH) 1.000

Phospholipids were purchased from Avanti or Genzyme. The phospholipid- cholesterol mixture was prepared in ~5 ml chloroform, dried down in the bottom of a glass flask by rotoevaporation at 40°C. PBS was added at the level of 1 ml/10 mg residue, and the AVE were extruded through a 0.2 μm polycarbonate filter under N₂ pressure. They were then transferred to a small vial, and lyophilized for 18-24 h until dry. The plasmid solution was added, and the flask was shaken to reconstitute the vesicles and encapsulate the plasmid.

To determine the proportion of AVE particles that take up compressed plasmids in the encapsulation step, studies were performed using plasmid which had been labeled by intercalation of the fluorescent DNA dye TOTO- 1 ™ (Example 1 ). The labeled DNA was combined with peptide and incubated as described in Example 2. The compressed polynucleotide was then encapsulated into AVE particles by the reconstitution method. The particles were extruded through a sizing filter to reduce their size down to about 250 nm. Quantitation of fluorescence was performed in a FACScan fluorescent cell sorter (Becton Dickinson), and analyzed using Lysis 2 software.

Figure 4 shows the results of this analysis. The left side of each panel shows the size ofthe particle (Y-axis) versus relative fluorescence (X-axis). The right side of each panel shows the frequency (Y-axis) versus relative fluorescence (X-axis). Fluorescent AVE particles were defined as any particle having a relative fluorescence of 10 or more (region "Ml "). The sample analyzed in Fig. 4a was AVE formed in the absence of DNA. This shows the background fluorescence due to the AVE envelope alone. The sample analyzed in Fig. 4b was AVE containing fluorescent DNA which had not been compressed before the encapsulation step. The percentage of particles containing DNA molecules was 31.80%. In other words, less than 1/3 ofthe liposomes contained DNA. Fig. 4c was AVE containing fluorescent DNA which had been compressed with polypeptides with an amino acid sequence of SEQ. ID NO:2. The percentage of particles containing fluorescent molecules was 73.40%. In other words, almost 3/4 ofthe liposomes contained DNA. To determine the proportion of plasmid that is taken up by AVE, experiments were conducted in which the plasmid was labeled with ³ P. The labeled plasmid was combined with the compressing peptide and incubated as before. The compressed plasmid was then encapsulated in AVE according to the reconstitution method. Encapsulation efficiency was determined by treating the AVE suspension with DNAse I for 30 minutes at 37°C. (encapsulated plasmid DNA is protected from DNAse digestion), removing digested plasmid DNA by column chromatography, and measuring the remaining radioactivity. Since encapsulated plasmid DNA is protected from DNAse digestion, the amount of label remaining in the high molecular weight fraction is proportional to the amount of encapsulated plasmid. The results of this analysis showed that when the peptide was combined with the plasmid DNA the efficiency of encapsulation by the AVE increased from 24% to 64%.

Example 5: Preparation of AVE with a tissue targeting component from RSV

Isolated RSV glycoproteins F and G were obtained from E. Walsh, Rochester,

NY, and were purified by affinity chromatography according to standard methods. Purity was assessed by SDS polyacrylamide gel electrophoresis and Coomassie Blue stain. Western blots showed no cross-reactivity of F glycoprotein with G glycoprotein and vice versa. The glycoprotein stock solutions contained 175 μg/ml G glycoprotein (-90,000 Da) and 350 μg/ml F glycoprotein (-48,000 Da), respectively. The glycoproteins were inserted into AVE prepared by the dialysis method as in Example 3. The insertion was performed by a second dialysis step before encapsulation ofthe polynucleotide. The pre-formed envelopes were filtered through 0.22 μm filters (Acrodisc) and 2.5 ml of these were mixed under aseptic conditions with 0.5 of a filtered aqueous solution of deoxycholate (lipid:detergent molar ratio about 8) and incubated at room temperature for 1 hour. Partial solubilization was observed with electron microscopy of a vesicle sample treated similarly. The glycoprotein solution was added aseptically, gently mixed, and kept for 45 minutes at room temperature. The mixture was then dialyzed in the cold (4°C) against 2 liters of Tris (10 mM, pH 7.8, containing 0.5 mM NaN₃) with 5 buffer changes at 4, 8, 16, and 48 hours. The buffer was purged with N₂ for the entire time of dialysis. The sample was removed after 56 hours and analyzed for size and inclusion ofthe glycoprotein on the outer vesicle surface.

Samples were prepared with 46.3 μl ofthe G glycoprotein stock solution only, one sample with 8.2 μl ofthe F glycoprotein stock solution only, or one sample with both 46.3 μl ofthe glycoprotein and 8.2 μl ofthe F glycoprotein stock solution. The corresponding lipid:protein ratios are shown in Table 8. Prior to use in subsequent experiments, the artificial RSV envelopes were centrifuged for 10 minutes in an Eppendorff microfuge and the CF solution replaced with PBS.

TABLE 8: Lipid :Proteln Ratio of Artificial RSV Envelopes

Glycoprotein Lipid Cone, Glycoprotein Cone, Molar Ratio (μmole/ml) (μmole/ml)

G 20.6 9 x 10 ^s 2.3 x 10^s

F 20.6 6 x 10^s 3.4 x 10^s

G + F 20.6 1.5 X 100-⁴ 1.4 x 10^s

Example 6: Efficacy ofthe tissue targeting component

HEp-2 cells were grown on sterile coverslip flasks at 37°C and 5% CO₂. When the cells were approximately 50% confluent, they were washed with PBS and then used to perform fusion experiments. The AVE were prepared according to the description in Example 5, except that they were labeled by preparing them in the presence of 6- carboxyfluorescein (6-CF) as an aqueous space marker, rather than a polynucleotide.

Fusion experiments were conducted as follows: To HEp-3 cells in Petri dishes were added 0.5 ml of one ofthe solutions containing the artificial envelopes without protein, with G glycoprotein only, with F glycoprotein only, or with both G and F glycoprotein as described in the previous example. A sample of AVE without glycoprotein was used as a negative control. An additional control was a solution containing 6-CF diluted 1 :20,000 only (no lipid control). Cells were replenished with 3 ml of a 1% DMEM cell culture medium and incubated at 37°C in 5% CO₂. Cells were viewed after 1 , 2, 4, and 24 hours under a fluorescent microscope at 40X magnification and photographed under phase and fluorescent light.

Cells incubated with diluted CF solution did not fluoresce. Cells incubated with artificial envelopes without protein showed only faint occasional fluorescence. There was no detectable fluorescence from a field of cells incubated with lipid envelopes without protein (lipid control). Also, cells incubated with artificial envelopes containing RSV G glycoprotein or F glycoprotein only showed some, but relatively faint, fluorescence. However, practically all cells of a batch that had been incubated with the complete artificial RSV envelopes were fluorescent after 1 hour incubation. Fluorescence is in all cases diffuse within the cytoplasm ofthe cells, which confirms that the transfer process was a fusogenic process rather than a phagocytic process which would result in punctate fluorescence confined to intracellular vacuoles. These results demonstrate that the complete artificial RSV envelope is fusogenic and may therefore be used as a drug carrier to deliver polynucleotides directly into the cytoplasm of infected cells.

Example 7; Efficacy of the intracellular targeting component

Experiments were conducted to determine the effectiveness of nuclear localization signals by preparing complexes of fluorescently labeled plasmids (Example 1), a polypeptide with the amino acid sequence shown in SEQ. ID NO:2. In this experiment, the polynucleotide was delivered into the cells not via AVE, but using small cationic liposomes. The polypeptide used was the nuclear localization signal (NLS) from the large T antigen of SV40. Once the three components were combined, the plasmids bound to the outside ofthe cationic liposomes, which provide a means of egress across the outer membrane ofthe target cell that is different from AVE.

Cells used for this series of experiments included the transformed human epithelial cell line 2-CFSME₀-, containing the 6508 mutation in the cystic fibrosis transmembrane conductance regulator (obtained from D. Gruenert, U. California San Francisco). These cells were cultured in D-MEM/F-12 at 37°C in 5% CO₂. Cells used in microscopic studies were grown for 12 to 14 h on glass coverslips. Cultures used for expression studies were grown in 60 mm tissue culture dishes and transfected at 80-90% confluence. Also used were bovine pulmonary artery endothelial cells (BPAE) (Conary et al.).

Cationic liposomes consisted of 3β-[N-(N^,,N'-dimethylaminoethane)-carbamoyl]- cholesterol (DC -Choi) complexed with Dioleylphosphatidylethanolamine (DOPE) in a 1 :1 (w/w) ratio: see X. Gao et al. (1991), Biochem. Biophys Res. Comm. 179:280-285. DC-Choi was obtained from L. Huang, University of Pittsburgh, PA. DOPE was purchased from Avanti Polar Lipids (Alabaster, AL). Cationic liposomes consisting of N- [l-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride (DOTMA) and dioleylphosphatidylethanolamine (DOPE) (1 :1 wt/wt) (LIPOFECTIN™) were purchased from Life Technologies (Gaithersburg, MA). DOTMA also was received from Syntex Research (Palo Alto, CA).

DOTMA/DOPE liposomes were prepared by dissolving 5 mg of each component (6:5 DOTMA DOPE molar ratio) in 3 ml of CHC1₃, removing the organic solvent by rotoevaporation at 40°C for 45 min, and resuspending the lipid film in sterile double- distilled water under gentle swirling. The crude dispersion was agitated on a wrist-action shaker for 1 h at room temperature. Liposomes were reduced in diameter to <200 nm by high pressure homogenization with an Emulsiflex-B3 (Avestin, Ottawa, Canada), using approximately 15 cycles at 16,000 psi nitrogen pressure. The final lipid concentration was determined calorimetrically according to a modification (13). Liposomes were prepared at stock concentrations of 10 mg/ml (15 μmoles/ml) total lipid; working concentrations were 1 mg/ml or less. Liposomes were stored in glass vials at 4°C under nitrogen. DC-Chol/DOPE liposomes were prepared by mixing 4.5 mg of each component (6:4 DC-Chol/DOPE molar ratio) in 2 ml CHC1₃. Rotoevaporation and dispersion in water were done as described above. The flask was briefly sonicated to ensure removal of all lipid from the vessel wall and hydration continued in the refrigerator overnight before the preparation was homogenized and analyzed as described above.

CAT or PAP plasmids were pre-incubated with the intracellular targeting peptide. Liposome and plasmid solutions were then combined under gentle swirling following dilution of both components with water. Typically, the final lipid concentration was 180 μg/ml and the coπesponding plasmid concentration was 60 μg/ml. The final mixture was incubated for 15 minutes at room temperature before use.

Transfection experiments were performed on cells grown in 60 mm cell culture dishes to between 80 and 90% confluence. Each plate received 20 μg of plasmid DNA combined with 60 μg of cationic liposome. Location ofthe fluorescently labeled plasmid within the cells was performed by low-light fluorescent microscopy. A cooled CCD camera (Star 1 camera, Photometries, Tucson, AZ) and an Axo Video Program (Axon Instruments, Foster City, CA) was used. The tissue culture dish was mounted in a heated (37°C) microscope stage with the coverslip directly over the microscope objective. The shutter system on both the camera and the fluorescent light source was programmed to record images at 2 minute intervals. Cells were transfected with labeled DNA (6 μg) combined with 18 μg of DOTMA/DOPE, 60 sec before the first image was recorded.

These visualization studies showed that 24 h after transfection of BPAE cells with complexes without the targeting component, only low levels of fluorescence could be detected within the cells (Figure 5, Top Panel). When NLS was included as part ofthe complexes, the nuclei were fluorescent and distinct from the other cell components (Figure 5, Bottom Panel). When 2-CFME₀- cells were transfected with RV-PAP, the expression of enzymatically active PAP could be determined with a histochemical stain, and was augmented when the plasmid had been prepared using the NLS peptide.

To quantitate the degree to which the improved targeting ofthe plasmid to the cell nucleus resulted in improved expression ofthe plasmid, experiments were conducted in which the CAT plasmid was used as a reporter sequence. Two days following transfection the 2-CFME₀- cells, CAT activity was determined as previously described (K.L. Brigham et al. (1989), Am. J. Respir. Cell Mol. Biol. 1 :95-100). All assay data was normalized for protein using the BioRad Dye Reagent (BioRad Laboratories, Inc., Melville, NY) with bovine serum albumin as a standard.

Figure 6 shows the results of this experiment (Con = untransfected cells; Lipo = nucleic acid/liposome complex only; Pep = complexes also containing nuclear targeting peptide in the amount shown (relative to 30 units of DNA, wt/wt)). Expression ofthe CAT plasmid in the cells was augmented when the cells were transfected with complexes comprising the NLS peptide. The effect was concentration dependent. The peak expression observed (0.5:30, peptide:DNA, wt/wt) was nearly 8 times greater than that observed for the complexes without the peptide.

Example 8: AVE with compressed nucleic acid enhances protein expression in vitro

In vitro transfer using AVE containing a compressed plasmid was performed as follows: AVE were prepared by the extrusion method as described in Example 4.

Glycoprotein F ofthe RSV was added by the method described in Example 5. The AVE were then reconstituted in the presence of plasmid pRV-PAP. The encapsulated plasmid was placed into 35 mm Petri dishes containing confluent cultures of cells from lung tissue. The cells were grown from a human patient with cystic fibrosis. After an additional 48 h in culture, the cells were washed twice with saline, fixed with glutaraldehyde/formaldehyde, heat treated to inactivate the endogenous alkaline phosphatase (placental alkaline phosphatase is heat stable). The enzymatically active alkaline phosphatase was then stained with SIGMAFAST RED™.

When stained sections were viewed under the microscope at ~40x magnification, fields of about 100 cells were seen. From the cultures treated with pRF-PAP containing AVE, typically about 8 cells showed substantial amounts of red staining; the rest ofthe cells in the field remained unstained. From the cultures treated with empty AVE, none of the cells showed detectable staining. Example 9: AVE with compressed nucleic acid increases protein expression in vivo

In vivo transfer using AVE containing a compressed plasmid was performed as using both the CAT and PAP plasmids. The AVE were prepared and the plasmid was encapsulated by the reconstitution method, as described in Example 8.

In one experiment, pCMV4-CAT was compressed using a peptide with the sequence shown in SEQ. ID NO:2. The F surface glycoprotein of RSV was prepared by extrusion, then inserted into the AVE as a tissue targeting component, and used to encapsulate the compressed plasmid. Rats were anesthetized and a dose of 100 μg ofthe plasmid (in 180 μl total volume) instilled into their lungs via a syringe connected to a thin endotracheal tube. Control animals were treated identically with AVE inserted with glycoprotein F and containing saline, rather than a plasmid.

Animals were sacrificed after 48 hours and lungs, liver, and kidneys were analyzed for CAT activity. The organs were flash-frozen in liquid nitrogen, minced and extracted. CAT expression is quantitated by the conversion of [' C]-choramphenicol added as a substrate to chloramphenicol acetate, when incubated with the tissue extract. Data were normalized for the number of cells (< oπesponding to the amount of protein) present and are expressed as cpm/mg protein/hr. Only values > 400 cpm/mg/hr are considered to be significantly above background. Results are shown in Figure 7 for 3 animals in both the CAT plasmid treated

(EXP) and control (CONTR) groups. The lungs showed significant expression of CAT above background levels. As one would expect, there was practically no expression in the liver. Expression of CAT in the kidney is most likely due to spill-over ofthe gene product via the lymph into the systemic circulation from where it would reach the kidneys.

In a second experiment, AVE containing an expression plasmid was injected intravenously into rats to examine the requirement for tissue targeting from the general circulation. The PAP plasmid was compressed as described above for CAT, and then encapsulated into an AVE. The AVE had been previously inserted with the F glycoprotein of RSV . Anesthetized rats were injected in a tail vein with a dose of 600 μl AVE containing 100 μg ofthe PAP gene. Control animals were treated with AVE containing saline. The animals were allowed to recover, and were given food and water ad libitum for 60 h. They were then sacrificed and organs were preserved in paraforaldehyde. Thin frozen sections were heated to 60°C for 10 min to inactivate endogenous alkaline phosphate activity. They were then incubated with SIGMAFAST RED™ histochemical stain for 1 h.

Figure 8 Top and Bottom panels show lung sections from animals treated respectively with the plasmid-containing AVE or control AVE (40x magnification). The field shows alveolar epithelial cells, endothelial cells, and the vascular bed, along with erythrocytes located in the vasculature. There is essentially no staining of any ofthe cells. This indicates that there is no expression ofthe PAP gene, probably because the AVE were blocked from reaching the epithelial cells by the endothelial layer.

Figure 9 Top & Bottom panels show kidney sections from the same animals as in Figure 8 (40 x magnification). In the animal treated with plasmid-containing AVE (Top panel), PAP expression is evident by the dark-staining patches that occur around some of the open tubule sections. There was only a relatively low level of staining in the animal treated with control AVE (Bottom panel).

In a third experiment, AVE containing the PAP plasmid were again used for intravenous administration. However, in this case, the AVE were specifically targeted to the endothelium ofthe lung using serotonin.

In order to provide AVE expressing serotonin on the surface, serotonin was first linked covalently to a phosphoiipid. N-glutaraldehyde substituted phosphatidyl ethanolamine was obtained from Sigma Chemical Co., St. Louis. It was mixed with serotonin in the presence of ethylenediamine carbodiimide (EDC) to form a linkage from the C-terminal of serotonin to the N-group ofthe substituted PE. Reaction conditions were essentially as described in V. Weissig et al. (1993), pp. 231-234 in Liposome Technology (2nd ed.) G. Gregoridas. The serotonin-linked PE was then included as a minor component in the phospholipid-detergent matrix during the initial formation ofthe lipid envelopes. The compressed PAP plasmid was then encapsulated in the AVE by the reconstitution method, as before. Figure 10 (Top Panel) shows a lung section from an animal treated 60 h previously with 600 μl of AVE containing 100 μg PAP plasmid. The section is stained with SIGMAFAST RED™ to display PAP enzymatic activity. There is staining in individual cells scattered throughout the section away from the alveolar lumen, which is the expected pattern for lung endothelial cells. Figure 10 (Bottom Panel) shows a kidney section from a similarly treated animal. The staining pattern is similar to that in the Top Panel of Figure 9, indicating that serotonin-targeted AVE have a tendency to localize in both the lung and the kidney.

Example 10: Stability of artificial viral envelopes

In this study, the stability of AVE was assessed at room temperature and when refrigerated. AVE were prepared by dissolving phospholipids and cholesterol in chloroform, and removing the solvent by rotoevaporation. The resulting film was suspended by gentle swirls in iso-osmotic sterile-filtered glycerol (292 mmol/kg, adjusted to pH 7.01), followed by 1 h of agitation at room temperature on a wrist-action shaker. The AVE were then subjected to several cycles of freeze-thawing. The vesicle size was reduced to -800 nm by nitrogen pressure extrusion (Lipex Biomembranes, Vancouver, Canada), using a polycarbonate membrane (Poretics Corp., Livermore CA). The AVE encapsulated no material other than solvent, and comprised no targeting molecule.

Stability studies were conducted on AVE suspensions at 50 mg/mL, 10 mg/mL, and 1 mg/mL, both at room temperature (20-36°C) and refrigerated (4-8°C). AVE were stored in Wheaton's clear borosilicate glass serum bottles capped with Wheaton's gray butyl rubber stoppers. A vacuum was created using an Edward's freeze-dryer. A small amount of nitrogen gas was introduced into the lyophilizer chamber before capping the bottles.

The mean diameter of stored AVE was determined weekly by laser light scattering using the Nicomp Model 370 laser particle sizer (Particle Sizing Systems, Santa Barbara CA). Phosphoiipid composition was assessed by chomatographing AVE on AL SIL G plates, comprising a s250 μm layer of silica gel on an aluminum backing (Whatman), using a solvent of CHCl₃:MeOH: water of 65:25:4, and developing with iodine. Figure 11 shows a representative result in which AVE were stored at 10 mg/mL in a refrigerator (upper panel) or at room temperature (lower panel). The pH of suspensions kept at room temperature or refrigerated was stable throughout the study (open bars). The size ofthe AVE was also stable (filled bars). Some alteration ofthe phosphoiipid composition was noticeable after 8 weeks, but did not impact the vesicle size or the pH. Sample sterility was verified weekly by striking a rich standard solid media plate with a 10 μL loop, and incubating at 37°C. All samples were negative for growth at 24 h.

Example 11: Preparation of bifunctional AVE targeted with serotonin

A synthetic chemist of ordinary skill will readily appreciate that a number of options are available to link molecules like serotonin onto lipids or sterols for constitution into liposomes. In the previous example, serotonin was linked onto PE. This example illustrates a method for linking serotonin onto cholesterol. When using glutarate as the linking group, cholesterol glutarate may be prepared first, and then serotonin attached. Alternatively, serotonin glutaramide may be prepared and then attached to cholesterol. The first method is generally more convenient, and is described here. The method is shown schematically in Figure 12.

Cholesterol glutarate (compound 1 in the Figure) is prepared as follows: In a dry 50 mL round-bottomed flask equipped with a CaCl₂ guard tube and a stirring system, the following are added: 2.5 g (6.5 mmol) cholesterol; 0.741 g (6.5 mmol) glutaric anhydride; in 20 mL anhydrous chloroform. After stirring for 2 min, 8.1 mg 4- dimethylaminopyridine is added. After stirring another 4 h at 25°C, a further 30 mL chloroform is added. Product is washed with 20 mL 2 N HCl, twice with 20 mL water, then dried over sodium sulfate. Evaporate. The crude product is purified on silica gel, eluting with 2%-4% methanol in chloroform.

N-[2-(3'(5'-hydroxyindole))ethyl]-6-(3-cholesteryloxy)-2,5-dioxohexylamine (compound 2 in the Figure) is prepared as follows: 350 mg (0.7 mmol) cholesterol glutarate is dissolved in 5 mL DMF, and a 1 M solution of dicyclohexylcarbodiimide in 0.8 mL chloroform is added. After stiπing for 5 min, 170 mg (0.8 mmol) serotonur.HCl is added. After stirring for 4 h, 20 mL chloroform is added, and the product is washed twice with 50 mL of a 3% solution of sodium bicarbonate and then with 50 mL water, then dried over sodium sulfate. Evaporate. The crude product is purified on a column of silica gel, eluting with 2% methanol in chloroform. Altematively, the product may be purified by crystalization. Identity ofthe product was confirmed by elemental analysis, mass spectrometry, and proton and ^I3C NMR. The purified product was nearly 100% pure.

Artificial viral envelopes comprising serotonin as a tissue targeting component have been prepared by mixing N-[2-(3^,(5'-hydroxyindole))ethyl]-6-(3-cholesteryloxy)- 2,5-dioxohexylamine into the phospholipidxholesterol mixture before formation ofthe liposomes. Between about 0.1 and 5.0 mole percent may be used, with 0.25-0.5 mole percent being typical. For encapsulating nucleic acid, AVE are formed from the lipid mixture in the presence of a compressed polynucleotide.

Artificial viral envelopes have been prepared comprising the 23-amino acid peptide modeled in influenza hemagglutinin (SEQ. ID NO:7) as a fusogenic component. A synthetic chemist of ordinary skill will readily appreciate that a number of options are available to link peptides onto liposomes. Peptides already comprising a lipophilic region may be inserted into preformed liposomes, for example, as outlined in Example 5. Peptides not naturally comprising a lipophilic region may be synthesized with an additional transmembrane spanning region. Altematively, peptides may be synthesized attached to a non-amino-acid lipid-soluble component, such as palmytic or myristic acid. The structure of SEQ. ID NO:7 linked to palmytic acid is shown below:

H(CH₂)₁₅(CO)-GLFEAIEGFIENGWEGMIDGWYG-OH

Peptides are linked to palmytic or myristic acid as part ofthe peptide synthesis process. The peptide is synthesized on solid phase, starting from the C-terminus, and the lipophilic acid is added as the last unit at the amino acid end ofthe peptide. During this process, palmitic or myristic acid is activated the same was as an amino acid and coupled to the peptidic residue. The peptide is then cleaved form the solid phase and deprotected, as usual. Details of preparation ofthe fusogenic component are as follows: Solid phase synthesis according to the method of Merrifield (J. Am. Chem. Soc. (1963) 85:2149) was conducted on a Model 432A SYNERGY™ peptide synthesizer from Applied Biosystems, Foster City, CA. The peptidic chain being synthesized was linked via the C-terminus to a lightly cross-linked polystyrene resin, and each amino acid was added according to the following procedure: first, the 9-fluorenylmethoxycarbonyl (Fmoc) protective group at the amino terminus ofthe growing peptide was removed with piperidine. The next amino acid in line (its amino function Fmoc protected) was then activated with a benzotriazolyl moiety and coupled to the growing peptide. This sequence is repeated until all amino acids ofthe sequence are added, followed by the addition of palmitic acid itself. The peptide is removed from the resin and secondary protecting groups are removed by reacting with trifluoroacetic acid (FT A). The peptide is then precipitated from the TFA using either, and filtered off. The peptide is dissolved in water containing 10% acetonitrile, and lyophilized. Between about 1 and 5 mole percent of palmytic acid-peptide conjugate may be constituted into liposomes, with about 1.6-2.5 mole percent being typical. The conjugate is constituted into liposomes either by mixing the conjugate into the solution used to form the liposomes, or more usually by inserting it into preformed liposomes by detergent softening. Bifunctional AVE targeted with serotonin are typically prepared as follows: A plasmid comprising a PAP reporter gene encoding sequence are compressed using a compression peptide consisting essentially of SEQ. ID NO:2, as described in Example 2. A lipid film containing phosphoiipid, cholesterol, and serotonin-linked cholesterol are dried as a lipid film on a round-bottom flask. An aqueous solution ofthe compressed polynucleotide is added, and liposomes form encapsulating the polynucleotide. To increase the proportion of unilamellar vesicles, the preparation is subjected to 5-6 cycles of freeze-thawing by incubating alternately in a dry ice/acetone bath and a 40-42°C water bath. Vesicles are sized by extrusion to 600-1000 nm. A conjugate ofthe 23 amino-acid fusogenic component shown in SEQ. ID NO: 7 linked to palmytic acid is inserted into the outer liposome membrane by insertion similar to the method outlined in Example 5. The AVE are incubated with the conjugate in the presence of deoxycholate, and then the deoxycholate is removed by dialysis. Liposomes are characterized for lipid, peptide, and nucleic acid content. Routine size determination is conducted by laser light scattering.

Example 12: Components of bifunctional AVE work in concert to promote delivery of the encapsulated material

Liposomes were prepared that comprised: a) a phospholipidxholesterol envelope characteristic of AVE; b) serotonin as a tissue targeting component; c) a hemagglutinin 23-mer fusogenic component; d) a plasmid with the PAP reporter gene. A second preparation of liposomes comprised the fusogenic component but not the targeting component.

Three groups of five female Sprague Dawley rats received an intravenous injection of either: 1) saline; 2) AVE with the fusogenic component; or 3) AVE with both the fusogenic and targeting components. Two days later, animals were sacrificed, and expression ofthe reporter gene was analyzed in tissue homogenates. PAP encoded in the expression vector is thermally stable, whereas endogenous alkaline phosphatase is heat labile. The level of expression was measured by heating the tissue aliquot at 65 °C for 1 h, and then conducting a standard enzyme assay using the fluorometric substrate 4- methylumbelliferyl phosphate. Results were coπected for the protein concentration ofthe tissue aliquot and further analyzed by statistical analysis.

Results of targeted gene delivery in the lung of treated animals are shown in Figure 13. Basal expression determined in the saline treated animals is shown by the solid bar. PAP expression determined in animals treated with AVE having only fusogenic components is shown by the stippled bar. PAP expression determined in animals treated with AVE having both fusogenic and targeting components is shown by the open bar.

The approximate levels of expression observed were as follows: For the AVE comprising the fusogenic component but not serotonin, activity in nmol/mg protein/h was: lung, 0.46; liver, 2.28; kidney, 0.73; brain, 0.41; heart, 2.39; spleen, 1.62; ovary, 0.64. For the bifunctional AVE, activity was: lung, 1.51; liver, 9.18; kidney, 1.71 ; brain, 1.00; heart, 1.15; spleen, 1.56; ovary, 0.96. The results show that AVE comprising the PAP expression vector and a fusogenic component increase expression in the lung (the tissue of interest) over basal levels by about 2.3 fold. AVE comprising both the fusogenic component and serotonin as a tissue targeting component increase expression over basal levels by about 8.1 fold.

Example 12: An expression vector for αi-antitrypsin

This example provides a pCMV4 expression vector including a coding sequence for human α,.antitrypsin (AAT), SEQ. ID NO:8. The plasmid can be incorporated into the liposome for targeting, for example, to the lung or liver. The plasmid is then capable of expression of AAT by the cells without incorporation into the chromosome.

The construct includes a short transcription augmenter sequence 5' to AAT encoding region, which increases the rate of translation. The construct also includes a human growth hormone 3' untranslated region, which stabilizes the transcript. Adult respiratory distress syndrome (ARDS) is thought to involve a relative deficiency of AAT activity. Therefore, the delivery of this plasmid to the lungs may be therapeutic in many human conditions characterized by injury of the lungs. The plasmid made in accordance with this example does not replicate in eukaryotic cells. Therefore, the increased expression ofthe gene is transient. The plasmid is not readily incoφorated into the host DNA. Both of these characteristics enhance safety for human administration.

The plasmid construct was prepared as follows: Two oligonucleotide primers of twenty to thirty nucleotides were synthesized. One nucleotide was homologous to the 5' untranslated region immediately upstream (5') ofthe initiation codon. The second oligonucleotide was complementary to the 3' untranslated region immediately downstream (3') ofthe stop transcription codon. Both oligonucleotides had a one or two base substitution, creating a different restriction enzyme site in the untranslated regions of the amplified gene. The new restriction enzyme sites were approximately eight nucleotides downstream from the 5' end ofthe oligonucleotide. An EcoRl-EcoRl fragment of AAT cDNA from a phAT85 vector (gift of Dr. S.

Wog Baylor University) was used as a template. The AAT gene was amplified using Vent DNA polymerase, 100 ng of target DNA, a programmable temperature cycler, and standard reaction conditions. A PCR reaction was conducted in a buffer of lOmM KCl, lOmM (NH4)₂SO4, 20mM Tris pH 8.8, 2mM MgSO₄, 0.1% Triton X-100, 2 units of Vent DNA polymerase, 200 mM each dNTP, and 1.0 mM each primer. Denaturing was done at 93.5° C, annealing at 56° C and extension at 75° C.

The PCR amplification product was cleaved with restriction enzymes Clal and Smal. The amplified gene was separated from small fragments nucleotides by gel filtration through an S-400 spin column. pCMV4 (Andersson et al. (1989) J. Biol. Chem. 264:8222-8224) was used as the expression vector. The vector backbone is pTZ18R (Pharmacia, Piscataway, NJ). The promoter consists of nucleotides -760 to +3 ofthe promoter-regulatory region ofthe human cytomegalovirus (Towne strain) major immediate early gene. Transcription termination and polyadenylation signals are supplied by sequence 1533 to 2157 of the human growth hormone. A synthetic fragment of DNA coπesponding to the 5'- untranslated region ofthe alfalfa mosaic virus 4 RNA is located in front ofthe polylinker region site and acts as a translational enhancer. pCMV4 was cleaved with Clal and Smal, then ligated with the AAT encoding sequence. After ligation, the pCMV4-AAT construct was transfected into fresh competent bacteria (E. coli NM522), prepared by standard methods (Hanahan, D. (1980) J. Mol. Biol. 166:557-580). After transfection, the bacteria were selected on plates containing carbenicillin. The colonies were grown up as individual liquid cultures for restriction nuclease analysis.

Properly constructed plasmid was purified by lysis ofthe bacteria and precipitation with polyethylene glycol. Plasmid was further purified by ultracentrifugation in isopycnic CsCl for 40 hours at 45,000 rpm. The isolated plasmid was precipitated with ethanol and resuspended in sterile water. It is important to deplete endotoxin (lipopolysaccharide A from gram negative bacterial). The method outlined reduces the level to about 45 pg of endotoxin/5μg of plasmid.

Plasmid DNA was purified from a 2 L liquid culture of E. coli strain XL-1 Blue by standard methods (Cuπent Protocols in Molecular Biology, pp. 1.7.1-1.7.1 1). Plasmid purity was determined by electrophoresis on submarine agarose gels. Example 13: Delivery of an αj.antitrypsin expression vector into cells using AVE

Plasmid DNA comprising an AAT expression vector was prepared according to the previous example. Purified plasmid DNA was labeled with the intercalating dye TOTO™-l using a method similar to that outlined in Example 1. DNA and dye were coincubated at a ratio of 1 : 1000 (w/w) at room temperature for 15 min, and free dye was removed by precipitation with sodium acetate and ethanol. The labeled DNA was then compressed by combining with the compression peptide (SEQ. ID NO:2) at a ratio of 1 :30 (w/w) and incubating for 10 min at 37°C. This plasmid-peptide complex was added to dried lipid film and vesicles were formed by gentle shaking. Unilamellar vesicles were enriched by repeated freeze-thawing. The vesicles were extruded through a polycarbonate membrane and characterized by laser light particle sizing. No tissue targeting or fusogenic component was used in this experiment. AVE containing 20 μg DNA were applied to plates of HepG2 cells at about 60% confluence. Presence ofthe labeled DNA was observed by fluorescent light at 495 nm every hour.

Results of this experiment is shown in Figure 14. Figures A & B show the cells under normal (phase) illumination. Figures C & D show the cells photographed under fluorescent illumination. The presence of fluorescence in and around the nucleus indicates that the DNA had been delivered into the cell, and at least partially localized to the nucleus. This demonstrates that the AVE can be used to deliver an AAT expression vector.

Another experiment was conducted to determine whether delivering the AAT vector to the cells resulted in secretion of AAT. AVE encapsulating the pCMV-AAT expression vector were prepared as before, except that the DNA was not labeled. In one preparation, AVE contained the vector, but no tissue targeting or fusogenic component. In another preparation, AVE encapsulated not only the vector, but also a tetramer ofthe 23 amino-acid fusogenic peptide according to SEQ. ID NO:7. In this case, the fusogenic peptide was not attached to a lipophilic anchor; instead, it was encapsulated inside the vesicles by including it in the aqueous phase added to the lipid film during vesicle formation.

Each ofthe two preparations was altematively used to transfect cultured BHK cells (20 μg DNA per plate). AAT expression was determined at various times after transfection by culturing the cells in serum-free medium. Twenty-one h after transfection, the culture medium was removed and replaced with medium free of phenol red and fetal bovine serum. The serum-free medium was collected 6 h later, aliquoted, and stored at - 80°C. The cells were returned to normal tissue culture medium, and then a second serum- free supernatant was collected in the same fashion at 45 h after transfection. An enzyme-linked immunosorbant assay (ELISA) was designed that identified α, .anti trypsin using the double criteria of: a) ability to bind trypsin; and b) ability to bind antibodies to human AAT. An Immunon IV 96-well plate was coated with 50 μL/well bovine trypsin at 50 μg/mL in phosphate-buffered saline (PBS) overnight at room temperature, or 2 h at 37°C. After washing with PBS, the wells were blocked with 50 μL PBS containing 0.25 g BSA, 50 μL Tween-20, and 1 mL 20% sodium azide per 100 mL, and rewashed. 50 μL culture supernatant or AAT control (6.25-500 ng/mL) was incubated in the wells for 1 h at 37°C. Plates were overlaid with 50 μL/well goat anti- human AAT diluted to 1 :500 in blocking buffer, followed by 50 μL/well rabbit anti-goat IgG diluted 1 :30,000. Plates were washed 5 times with water, then developed with substrate solution (97 mL diethanolamine, 101 mg MgCl₂ and 1 mL 20% sodium azide per liter, adjusted to pH 9.8) for 1-2 h at 37°C, and the optical densities were measured at 405 nm.

Results of this experiment are shown in Figure 15. Levels of AAT detected were -2.5 fold higher after 24 h when the AVE comprised a fusogenic peptide (open bars) compared with AVE without a fusogenic peptide (filled bars). At 48 h, the levels obtained using the fusogenic peptide were -6.4 times higher. AAT may also be detected on a cell-by-cell basis by histochemical staining using anti- AAT.

The results show that the AVE can be used to deliver DNA, which in turn results in expression and secretion of bioactive protein by the treated cells. The presence of a fusogenic component enhances expression. Example 14: Delivery of an antisense oligonucleotide using AVE

In this example, oligonucleotides complementary to human αi.antitrypsin were delivered to cultured cells in order to down-regulate AAT expression. This is of therapeutic interest for clinical conditions in which AAT is overexpressed. Cells used for this experiment were the human hepatocyte cell line HepG2, which has a constitutive level of AAT expression.

A series of overlapping DNA oligonucleotides of 20-50 residues in length are prepared by standard oligonucleotide synthesis techniques. The oligonucleotides are complementary to the naturally occurring AAT encoding sequence, focusing on the 3' half. Model oligonucleotides are listed in Table 9:

TABLE 9: Exemplary antisense oligonucleotides

Target gene Approximate SEQ. Sequence corresponding ID NO; nucleotide nos. α^antitrypsin 1240-1220 9 MAI I I I I GGϋ lGϋϋAI ICA α_rantitrypsin 1120-1090 10 TTTCTCGTCGATGGTCAGCACAGCCTTATG

Separate preparations of AVE were prepared, using compressed or uncompressed oligonucleotides. Antisense oligonucleotides were compressed using "Peptide A", having the amino acid sequence shown in SEQ. ID NO:2, at a ratio of 1 :30 (w/w) peptide:DNA. Oligonucleotides were then encapsulated into AVE by adding the solution to a dried lipid film, as described in previous examples.

In some experiments, oligonucleotides are synthesized incoφorating a CY3 dye molecule (Pharmacia) at the 5' end. AVE containing oligonucleotides labeled in this way are used to quantitate the delivery to HepG2 cells on an hourly basis by fluorescent microscopy.

Results of a representative experiment is shown in Figure 16. Figures A & B, labeled oligonucleotides were encapsulated directly into AVE; for Figures C & D, labeled oligonucleotides were compressed with Peptide A. In neither case did the AVE have fusogenic or targeting components. An estimated dose of 20 μg of encapsulated labeled oligonucleotide was added to each flask of HepG2 cells, and the pattern of fluorescence inside the cells was photographed after 2 h.

In other experiments, unlabeled oligonucleotides is delivered to the cells in the AVE, and the cells are used to prepare serum-free culture supernatant as in Example 13. The supernatant is assayed by ELISA for the level of AAT expression. Downwards modulation of AAT expression is expected, compared with untreated cells or cells treated with empty AAV. Effectiveness of various fusogenic components and liver cell specific targeting components are compared to optimize the composition of bifunctional AVE.

Example 15: Delivery of small molecules using bifunctional AVE

Virtually any substance of therapeutic interest may be delivered using the liposomes of this invention, providing: a) they can be encapsulated within the liposomes; and b) they are sufficiently lipid insoluble so as not to partition into the liposome membrane and thereby out ofthe liposome, but remain encapsulated.

AVE comprising small molecules are modeled using propidium iodide, a 632 mol wt lipid-insoluble fluorescent marker. Encapsulation is performed by adding a solution of propidium iodide to a dried film of phosphoiipid and cholesterol, allowing the AVE to form. Unencapsulated propidium iodide is removed by gel filtration chromatography. The AVE formed in the presence of small molecules are generally smaller in diameter (100-200 nm) compared with those formed in the presence of polynucleotides several kb in size. Consequently, it is not necessary to down-size the AVE by extrusion, which may result in loss of some ofthe encapsulated material. Bifunctional AVE are prepared according to the methods described elsewhere in this disclosure. In one example, bifunctional AVE are prepared with encapsulated propidium iodide, N-[2-(3'(5'- hydroxyindole))ethyl]-6-(3-cholesteryloxy)-2,5-dioxohexylamine included in the lipid film as a lung-specific tissue targeting component, and a peptide comprising SEQ. ID NO: 7 inserted into the membrane by detergent softening as a fusogenic component. These AVE are tested for delivery in a culture model of lung tissue. Uptake of propidium iodide into the cells is measured in terms of both fluorescence intensity and percent of positively staining cells, and compared between AVE of different composition. REFERENCES:

Articles

1. Allen TM. Long-circulating (sterically stabilised) liposomes for enhanced drug delivery. Trends Pharmacol. Sci. 15:215-220, 1994.

2. Altschul et al. Bull. Math. Bio. 48:603-616, 1986

3. Baru M. et al. Liposome-encapsulated DNA-mediated gene transfer and synthesis of human factor IX in mice. Gene 161 :143-150, 1995. 4. Behr J.P. et al. Efficient gene transfer into mammalian primary endocrine cells with lipopolyamine-coated DNA. Proc. Natl. Acad. Sci. USA 86:6982-6986, 1989.

5. Behr J.P. et al. Synthetic gene-transfer vectors. Ace. Chem. Res. 26:274-278, 1993.

6. Block E. et al. Effect of plasma membrane fluidity on serotonin transport by endothelial cells. Am. J. Physiol. 22:C672-C678, 1987. 7. Bongartz J. et al. Improved biological activity of antisense oligonucleotides conjugated to a fusogenic peptide. Nuc. Ac. Res. 22:4681-4688, 1994.

8. Canonico A.E. et al. Aerosol and intravenous transfection of human αl-antitrypsin gene to lungs of rabbits. Am. J. Respir. Cell Mol. Biol. 10:24-29, 1994.

9. Chonn A. et al. Recent advances in liposomal drug-delivery systems. Curr. Op. Biotech. 6:698-708, 1995.

10. Conary J.T. et al. J. Clin. Invest. 93:1834-1840, 1994.

11. Das D. et al. Specific binding of serotonin in rat lung. Am. J. Physiol. 248:E58-E63, 1985.

12. Devaux P. et al. How lipid asymmetry can make vesicles fusion-competent by inhibition ofthe thermal undulations. Biochem. Soc. Transactions 21 :276-280, 1993.

13. Dingwall et al. Annu. Rev. Cell Biol. 2:367-390, 1986.

14. Drumm M.L.et al. Cell 562:1227-1233, 1990.

15. Elorza B. et al. Comparison of particle size and encapsulation parameters of three liposomal preparations. J. Microencapsulation 10:237-248, 1993. 16. Fraley R. et al. Biochemistry 20:6978-6987, 1981

17. Fraley R. et al. J. Biol. Chem. 256:10431-10435, 1980 18. Friede M. et al. Selective induction of protection against influenza virus infection in mice by a lipid-peptide conjugate delivered in liposomes. Vaccine 12:791-797, 1994.

19. Ghosh P. et al. Targeting of liposomes to hepatocytes. In: Liver diseases, targeted diagnosis and therapy using specific receptors and ligands. G. Wu et al., eds. Marcel Dekker, NY pp 87-103, 1991.

20. Goldfarb D.S. et al. Nature 322:641-644, 1986.

21. Gonzalez-Rothi R. et al. Liposomes as a targeted drug delivery system to the lungs (abstract). Pharmaceut. Res. 12:S-314, 1995.

22. Gregoriadis G. (editor). Liposome technology 2nd edition. Volume I: Liposome preparation and related techniques. CRC Press, Boca Raton, 1993.

23. Gregoriadis G. (editor). Liposomes as drug caπiers. Wiley, New York, 1988.

24. Henikoff et al. Proc. Natl. Acad. Sci. USA 89:10915-10919, 1992

25. Inex Pharmaceuticals Coφoration. Initial Public Offering Preliminary Prospectus, January 31, 1996. 26. Ishihara H. et al. Preparation of asialofetuin-labeled liposomes with encapsulated human interferon-gamma and their uptake by isolated rat hepatocytes. Pharm. Res. 7:542-546, 1990.

27. Lanford R.E. et al. Cell 37:801-813, 1984.

28. Laukkanen M. et al. Functional immunoliposomes harboring a biosynthetically lipid- tagged single-chain antibody. Biochem. 33:11664-11670, 1994.

29. Ledley F. Nonviral gene therapy: the promise of genes as pharmaceutical products. Human Gene Ther. 6:1129-1144, 1995.

30. Mori A. et al. Characterization of organ-specific immunoliposomes for delivery of 3^,,5'-O-dipalmitoyl-5-fluoro-2'-deoxyuridine in a mouse lung-metastasis model. Cancer Chemother. Pharmacol. 35:447-456, 1995.

31. Mumtaz S. et al. Design of liposomes for circumventing the reticuloendothelial cells. Glyconiol. 1:505-510., 1991.

32. Muzyczka N. Cuπ. Top. Micro. Immunol. 158:97-129, 1992

33. Nussbaum O. et al. Reconstitution of Functional influenza virus envelopes and fusion with membranes and liposomes lacking virus receptors. J. Vir. 61 :2245-2252, 1987. 34. Oku N. et al. Long-circulating liposomes. Crit. Rev. Ther. Drug Caπier Systems 11:231-270, 1994.

35. Puyal C. et al. Design of a short membrane-destabilizing peptide covalently bound to liposomes. Biochimica et Biophysica Acta 1195:259-266, 1994. 36. Rakowska M. et al. A fusogenic protein from rat brain microsomal membranes: partial purification and reconstitution into liposomes. J. Membrane Biol. 142:35-42, 1994.

37. Rosenfeld M.A. et al. Cell 68:143-155, 1993

38. Rubas W. et al. Incoφoration ofthe reovirus M cell attachment protein into small unilamellar vesicles: incoφoration efficiency and binding capability to L929 cells in vitro. J. Microencapsulation 7:385-395, 1990.

39. Schofield J. et al. Non- viral approaches to gene therapy. Brit. Med. Bulletin 51 :56-71, 1995.

40. Schreier H. et al. Int. J. Pharmaceut. 87:183-193, 1992. 41. Schultz-Gahmen B. et al. The bacillus subtilis nucleoid-associated protein HPB 12 strongly compacts DNA. J. Bacteriology. 176:50-60, 1994.

42. Thierry A. et al. Systemic gene therapy: biodistribution and long-term expression of a transgene in mice. Proc. Natl. Acad. Sci. USA 92: 9742-9746, 1995.

43. Tranum-Jensen J. et al. Membrane topology of insulin receptors reconstituted into lipid vesicles. J. Membrane Biol. 140:215-223, 994.

44. Trubetskoy V.S. et al. Biochim. Biophys. Acta 1131:311-313, 1993.

45. Vemuri S. et al. Preparation and characterization of liposomes as therapeutic delivery systems: a review. Pharm. Acta Helvetiae 70:95-111 , 1995.

46. Vengerov Y.Y. et al. Torus-shaped particles formed due to intermolecular condensation of circular DNA upon interaction with synthetic tripeptide. FEBS

180:81-84, 1985.

47. Walther FJ et al. Uptake of antioxidants in surfactant liposomes is enhanced by SP-A. Am. J. Physiol. 265:L330-339, 1993.

48. Watwe R. et al. Manufacture of liposomes: A review. Cuπ. Sci. 68:715-724., 1995. 49. Wilson J.M. et al. J. Biol. Chem. 267:11483-11489, 1992. Patent Publications

EP 0036277 Papahadjopoulos et al. (liposome-protein complexes) EP 0047480 Thibodeau et al. (immunosome, detergent insertion) EP 0273085 Myers et al. (internalizing nucleic acids) EP 0274581 MPI Medizinische (encapuslation method) EP 0366770 Konigsberg et al. (liposomes coupled to hormones) U.S. 4,078,052 Papahadjopolos et al. (liposomes encapsulating DNA) U.S. 4,515,736 Deamer (encapsulation method) U.S. 4,544,545 Ryan et al. (modified cholesterol) U.S. 4,737,323 Martin et al. (Liposome Tech)(liposome extrusion method) U.S. 4.761,287 Geho et al. (serotonin-containing vesicles) U.S. 4,762,915 Kung et al. (liposome anchor) U.S. 4,895,719 Radhakrishnan et al (dehydrated liposomes, inhalation) U.S. 4,920,016 Allen et al. (enhanced circulation liposomes) U.S. 5,000,960 Wallach et al. (liposome anchor) U.S. 5,008,050 Cullis et al. (Liposome Co) (liposome extrusion method) U.S. 5,019,369 Presant et al. (Vestar) (liposomes for tumor targeting) U.S. 5,059,421 Loughrey et al. (Liposome Co)(targeted liposomes) U.S. 5,109,113 Caras et al. (membrane anchor fusion peptides) U.S. 5,252,348 Schreier et al. (aritifical viral envelopes) U.S. 5,258,499 Konigsberg et al. (liposomes targeted with cytokine) U.S. 5,374,548 Caras (Genentech) (glycophospholipid anchor) U.S. 5,435,989 Presant et al. (liposome targeting, RES blocking) WO 92/06192 Scolaro et al. (liposomes with anti-sense oligos) WO 92/19267 Gluck et al. (influenza virosomes) WO 92/19730 Brigham et al. (construct for in vivo expression) WO 93/12756 Debs et al. (aerosolized transgene delivery) WO 95/12387 Collins (encapsulation method) WO 95/25234 Kan et al. (targeting specific tissue) WO 95/35094 Webb et al. (Inex) (enhanced circulation liposomes)

WO 95/32706 Wilschut et al. (Inex) (virosome with hemagglutinin)

WO 95/34647 Conary et al. (nucleic acid delivery)

SEQUENCE LISTINGS:

Sequence Description Typ pee

1 KKKKK K Compressing peptide Peptide

2 PKKKR KVKKK KK Compressing peptide & nuclear targeting Peptide

3 PKKKR KVLKK KKK Compressing peptide & nuclear targeting Peptide

4 PKKKR KV Nuclear targeting (SV40 sequence) Peptide

5 KRPRE DDDGE PSERK RER Nuclear targeting (plant sequence) Peptide

6 MLFNL RILLD DAAFR DGKKK Mitochondrial targeting Peptide

7 GLFEA IEGFI ENGWE GMIDG WYG Fusion peptide modeled from influenza Peptide hemagglutinin

8 (as shown) human c^-antitrypsin encoding region Nucleic acid

9 M A I I ! I I GGGTGGGATTCA α_rantitrypsin antisense oligonucleotide Nucleic acid

10 TTTCTCGTCGATGGTCAGCACAGC α^antitrypsin antisense oligonucleotide Nucleic acid CTTATG

SEQUENCE DESCRIPTION: SEQ ID N0: 1:

Lys Lys Lys Lys Lys Lys

1 5

SEQUENCE DESCRIPTION: SEQ ID NO:2:

Pro Lys Lys Lys Arg Lys Val Lys Lys Lys Lys Lys 1 5 10

SEQUENCE DESCRIPTION: SEQ ID NO:3:

Pro Lys Lys Lys Arg Lys Val Leu Lys Lys Lys Lys Lys 1 5 10

SEQUENCE DESCRIPTION: SEQ ID NO:4:

Pro Lys Lys Lys Arg Lys Val 1 5

SEQUENCE DESCRIPTION: SEQ ID NO:5:

Lys Arg Pro Arg Glu Asp Asp Asp Gly Glu Pro Ser Glu Arg Lys Arg 1 5 10 15

Glu Arg

SEQUENCE DESCRIPTION: SEQ ID N0:6: Met Leu Phe Asn Leu Arg He Leu Leu Asp Asp Ala Al a Phe Arg Asp 1 5 10 15

Gly Lys Lys Lys 20

SEQUENCE DESCRIPTION: SEQ ID NO:7:

Gly Leu Phe Glu Ala He Glu Gly Phe He Glu Asn Gly Trp Glu Gly 1 5 10 15

Met He Asp Gly Trp Tyr Gly 20

SEQUENCE DESCRIPTION: SEQ ID NO:8:

CTGGGACAGT GAATCGACAA TGCCGTCTTC TGTCTCGTGG GGCATCCTCC TGCTGGCAGG 60

CCTGTGCTGC CTGGTCCCTG TCTCCCTGGC TGAGGATCCC CAGGGAGATG CTGCCCAGAA 120

GACAGATACA TCCCACCATG ATCAGGATCA CCCAACCTTC AACAAGATCA CCCCCAACCT 180

GGCTGAGTTC GCCTTCAGCC TATACCGCCA GCTGGCACAC CAGTCCAACA GCACCAATAT 240 CTTCTTCTCC CCAGTGAGCA TCGCTACAGC CTTTGCAATG CTCTCCCTGG GGACCAAGGC 300

TGACACTCAC GATGAAATCC TGGAGGGCCT GAATTTCAAC CTCACGGAGA TTCCGGAGGC 360

TCAGATCCAT GAAGGCTTCC AGGAACTCCT CCGTACCCTC AACCAGCCAG ACAGCCAGCT 420

CCAGCTGACC ACCGGCAATG GCCTGTTCCT CAGCGAGGGC CTGAAGCTAG TGGATAAGTT 480

TTTGGAGGAT GTTAAAAAGT TGTACCACTC AGAAGCCTTC ACTGTCAACT TCGGGGACAC 540 CGAAGAGGCC AAGAAACAGA TCAACGATTA CGTGGAGAAG GGTACTCAAG GGAAAATTGT 600

GGATTTGGTC AAGGAGCTTG ACAGAGACAC AGTTTTTGCT CTGGTGAATT ACATCTTCTT 660

TAAAGGCAAA TGGGAGAGAC CCTTTGAAGT CAAGGACACC GAGGAAGAGG ACTTCCACGT 720

GGACCAGGTG ACCACCGTGA AGGTGCCTAT GATGAAGCGT TTAGGCATGT TTAACATCCA 780

GCACTGTAAG AAGCTGTCCA GCTGGGTGCT GCTGATGAAA TACCTGGGCA ATGCCACCGC 840 CATcπcπc CTGCCTGATG AGGGGAAACT ACAGCACCTG GAAAATGAAC TCACCCACGA 900

TATCATCACC AAGTTCCTGG AAAATGAAGA CAGAAGGTCT GCCAGCTTAC ATTTACCCAA 960

ACTGTCCATT ACTGGAACCT ATGATCTGAA GAGCGTCCTG GGTCAACTGG GCATCACTAA 1020

GGTCTTCAGC AATGGGGCTG ACCTCTCCGG GGTCACAGAG GAGGCACCCC TGAAGCTCTC 1080

CAAGGCCGTG CATAAGGCTG TGCTGACCAT CGACGAGAAA GGGACTGAAG CTGCTGGGGC 1140 CATGTTTTTA GAGGCCATAC CCATGTCTAT CCCCCCCGAG GTCAAGTTCA ACAAACCCTT 1200

TGTCTTCTTA ATGAπGAAC AAAATACCAA GTCTCCCCTC πCATGGGAA AAGTGGTGAA 1260

TCCCACCCAA AAATAACTGC CTCTCGCTCC TCAACCCCTC CCCTCCATCC CTGGCCCCCT 1320

CCCTGGATGA CATTAAAGAA GGGTTGAGCT GG 1352 SEQUENCE DESCRIPTION: SEQ ID NO:9: πATTTTTGG GTGGGATTCA 20

SEQUENCE DESCRIPTION: SEQ ID NO:10:

TTTCTCGTCG ATGGTCAGCA CAGCCTTATG 30

Claims

What is claimed as the invention is:

A liposome comprising a synthetic lipid vesicle and a compressed polynucleotide encapsulated by said vesicle.

2. The liposome of claim 1, wherein the synthetic lipid vesicle is essentially unilamellar, between 100 nm and 1000 nm in size, and has a cholesterol :phospholipid molar ratio between 0.5 and 1.2.

3. The liposome of claim 1 , which is an artificial viral envelope (AVE).

4. The liposome of claim 1, also comprising a polypeptide bound to the polynucleotide.

5. The liposome of claim 4, wherein the polypeptide comprises a linear sequence of at least 7 amino acids of which at least 50% of the amino acids have a side chain bearing a positive charge at pH 7.

6. The liposome of claim 1, also comprising a tissue targeting component.

7. The liposome of claim 6, wherein the tissue targeting component is specific for lung endothelial cells.

8. The liposome of claim 6, wherein the tissue targeting component is serotonin.

9. The liposome of claim 1, wherein the polynucleotide encodes αl-antitrypsin or α-interferon.

10. A liposome comprising a lipid vesicle with an inner and outer surface, and a tissue targeting component exposed on the outer surface which is serotonin.

11. The liposome of claim 10, encapsulating a substance selected from the group consisting of polynucleotides, peptides, drugs, and toxins.

12. The liposome of claim 11 , wherein me substance is a compressed polynucleotide.

13. A bifunctional liposome, comprising a synthetic lipid vesicle, a tissue targeting component, and a fusogenic component, wherein the tissue targeting component and the fusogenic component do not naturally occur together on a single molecule or a single viral particle.

14. The bifunctional liposome of claim 13, wherein the synthetic lipid vesicle is essentially unilamellar, between 100 nm and 1000 nm in size, and has a cholesterol :phospholipid molar ratio between 0.5 and 1.2.

15. The bifunctional liposome of claim 13, which is an artificial viral envelope (AVE).

16. The bifunctional liposome of claim 13, wherein the tissue targeting component and the fusogenic component are covalently linked to each other or to a single polypeptide.

17. The bifunctional liposome of claim 13, wherein the tissue targeting component and the fusogenic component are independently anchored in the synthetic lipid vesicle.

18. The bifunctional liposome of claim 13, wherein the fusogenic component comprises an influenza hemagglutinin peptide.

19. The bifunctional liposome of claim 13 comprising an encapsulated polynucleotide, obtainable by a process comprising the steps of: a) encapsulating the polynucleotide in the lipid vesicle; b) contacting the lipid vesicle formed in step a) with the tissue targeting component in the presence of a detergent; c) contacting the lipid vesicle formed in step a) with the fusogenic component in the presence of a detergent; and d) removing the detergent(s) used in steps b) and c).

20. The bifunctional liposome of claim 19, wherein the encapsulated polynucleotide is compressed by contacting the polynucleotide with a polypeptide comprising a linear sequence of at least 7 amino acids of which at least 50% ofthe amino acids have a side chain bearing a positive charge at pH 7.

21. The bifunctional liposome of claim 13 comprising an encapsulated polynucleotide, obtainable by a process comprising the steps of: a) encapsulating the polynucleotide in a lipid vesicle comprising a phosphoiipid or sterol to which the tissue targeting component is covalently attached; b) contacting the lipid vesicle formed in step a) with the tissue targeting component in the presence of a detergent; and c) removing the detergent.

22. A bifunctional liposome of claim 13 comprising an encapsulated polynucleotide, obtainable by a process comprising the steps of: a) encapsulating the fusogenic component and the polynucleotide in a lipid vesicle comprising a phosphoiipid or sterol to which the tissue targeting component is covalently attached; b) contacting the lipid vesicle formed in step a) with the fusogenic component in the presence of a detergent; and c) removing the detergent.

23. A method of compressing a polynucleotide, comprising the step of contacting the polynucleotide with a polypeptide comprising a linear sequence of at least 7 amino acids of which at least 50% of the amino acids have a side chain bearing a positive charge at pH 7.

24. A method of preparing an encapsulated compressed polynucleotide, comprising the steps of: a) providing a polynucleotide compressed according to the method of claim 23; and b) forming a lipid vesicle of between 100 and 1000 nm that encapsulates the compressed polynucleotide.

25. The method of claim 24, further comprising inserting a tissue targeting component into the lipid vesicle following step b).

26. A method of delivering a polynucleotide to a cell, comprising contacting the cell with the liposome of claim 1.

27. A method of modulating expression of a protein by a cell, comprising contacting the cell with the liposome of claim 1, wherein said compressed polynucleotide comprises a nucleic acid sequence encoding the protein or a fragment thereof, or a nucleic acid sequence complementary to a sequence encoding the protein or a fragment thereof.

28. A method of delivering a polynucleotide to a cell in an individual, comprising administering to the individual the liposome of claim 1.

29. A method of delivering a substance to a lung endothelial cell in an individual, comprising administering to an individual the liposome of claim 10 encapsulating the substance.

30. A pharmaceutical composition comprising the liposome of claim 1. 31. A pharmaceutical composition comprising the bifunctional liposome of claim 13.

32. A method for preparing a pharmaceutical composition, comprising mixing the liposome of claim 1 with a pharmaceutically compatible excipient.

33. A method of gene therapy, comprising the steps of: a) obtaining cells from an individual in need ofthe gene therapy; b) genetically altering the cells by delivering to the cells a polynucleotide by the method of claim 26; and c) readministering the genetically altered cells to the individual.

34. A method of gene therapy, comprising administering to an individual an effective amount ofthe pharmaceutical composition of claim 30.

AMENDED CLAIMS

[received by the International Bureau on 10 June 1 97 (10.06.97), original claims 1, 13, 23 and 24 amended; remaining claims unchanged (5 pages)]

1 A liposome compπsing a svnthetic neutral or anionic lipid vesicle and a compressed polynucleotide encapsulated by said vesicle

2 The liposome of claim 1 wherein the svnthetic lipid vesicle is essentialh unilamellar. between 100 nm and 1000 nm m size and has a cholesterol phosphoiipid molar ratio between 0 5 and I 2

3 The liposome ol claim 1. having a lipid composition resembling that ol a naturally occurring virus or viral panicle

4 The liposome of claim 1 also compπsing a polypeptide bound to the polvnucleotide

5 1 he liposome oi claim 4 wherein the polypeptide comprises a linear sequence ot at least 7 amino acids ot which at least 50% ot the ammo acids have a side chain bearing a positive charge at pH 7

6 The liposome of claim 1. also compπsing a tissue targeting component

7 The liposome of claim 6. wherein the tissue targeting component is specific for lung endothelial cells

8 The liposome of claim 6 wherein the tissue targeting component is serotonin

9 The liposome ol claim 1 w herein the po nucleotide encodes t-1 -anturvpsιn or (x- terferon 10. A liposome comprising a lipid vesicle with an inner and outer surface, and a tissue targeting component exposed on the outer surface which is serotonin.

1 1 . The liposome of claim 10. encapsulating a substance selected from the group consisting of polynucleotides. peptides. drugs, and toxins.

12. The liposome of claim 1 1. wherein the substance is a compressed polynucleotide.

13. A bifunctional liposome. comprising a synthetic lipid vesicle, a tissue targeting component, and a fusogenic component, wherein the tissue targeting component is distinct from an antibody and does not naturally occur together with the fusogenic component on a single molecule or a single viral particle.

14. The bifunctional liposome of claim 13, wherein the synthetic lipid vesicle is essentially unilamellar. between 100 nm and 1000 nm in size, and has a cholesterol:phospholipid molar ratio between 0.5 and 1.2.

15. The bifunctional liposome of claim 13. having a lipid composition resembling that of a naturally occurring virus or viral particle.

16. The bifunctional liposome of claim 1 . wherein the tissue targeting component and the fusogenic component are covalently linked to each other or to a single polypeptide.

17. The bifunctional liposome of claim 13. wherein the tissue targeting component and the fusogenic component are independently anchored in the synthetic lipid vesicle.

18. The bifunctional liposome of claim 13, wherein the fusogenic component comprises an influenza hemagglutinin peptide.

19. The bifunctional liposome of claim 13 comprising an encapsulated polynucleotide, obtainable by a process comprising the steps of: a) encapsulating the polynucleotide in the lipid vesicle: b) contacting the lipid vesicle formed in step a) with the tissue targeting component in the presence of a detergent: c ) contacting the lipid vesicle formed in step a) with the fusogenic component in the presence of a detergent; and d) removing the detergent(s) used in steps b) and c).

20. The bifunctional liposome of claim 19. wherein the encapsulated polynucleotide is compressed by contacting the polynucleotide with a polypeptide comprising a linear sequence of at least 7 amino acids of which at least 50% of the amino acids have a side chain bearing a positive charge at pH 7.

21. The bifunctional liposome of claim 13 comprising an encapsulated polynucleotide. obtainable by a process comprising the steps of: a) encapsulating the polynucleotide in a lipid vesicle comprising a lipid to which the tissue targeting component is covalently attached; b) contacting the lipid vesicle formed in step a) with the fusogenic component in the presence of a detergent: and c) removing the detergent.

22. A bifunctional liposome of claim 13 comprising an encapsulated polynucleotide. obtainable by a process comprising the step of encapsulating the fusogenic component and the polynucleotide in a lipid vesicle comprising a lipid and the tissue targeting component.

23. A method of compressing a polynucleotide, comprising the steps of contacting the polynucleotide with a polypeptide comprising a linear sequence of at least 7 amino acids of which at least 50% of the amino acids have a side chain bearing a positive charge at pH 7 to produce a polypeptide:polynucleotide complex, and then forming a lipid envelope around the complex.

A method of preparing an encapsulated compressed polynucleotide. compπsing compressing a polynucleotide according to claim 23. wherein the lipid envelope is a lipid vesicle of between 100 and 1000 nm that encapsulates the compressed polynucleotide

The method of claim 24. further comprising inserting a tissue targeting component into the lipid vesicle following step b)

A method of deliveπng a polynucleotide to a cell, compπsing contacting the cell with the liposome of claim 1

A method of modulating expression ot a protein

a cell, compπsing contacting the cell with the liposome of claim 1 wherein said compressed polynucleotide comprises a nucleic acid sequence encoding the protein or a fragment thereof, or a nucleic acid sequence complementary to a sequence encoding the protein or a fragment thereof

A method of dehveπng a polynucleotide to a cell in an individual, compπsing administeπng to the individual the liposome of claim 1

A method of delivering a substance to a lung endothelial cell in an individual, compπsing administeπng to an individual the liposome of claim 10 encapsulating the substance

A pharmaceutical composition compπsing the liposome of claim 1

A pharmaceutical composition compπsing the bifunctional liposome of claim 13

A method for prepaπng a pharmaceutical composition, compπsing mixing the liposome of claim 1 with a pharmaceutically compatible excipient

33. A method of gene therapy, comprising the steps of: a) obtaining cells from an individual in need of the gene therapy; b) genetically altering the cells by delivering to the cells a polynucleotide by the method of claim 26; and c) readministering the genetically altered cells to the individual.

34. A method of gene therapy, comprising administering to an individual an effective amount ofthe pharmaceutical composition of claim 30.