CROSS-REFERENCE TO RELATED APPLICATION
This application claims the benefit of U.S. Provisional Application No. 60/428,435, filed Nov. 21, 2002, the entire contents of which are incorporated by reference herein.
FIELD OF THE INVENTION
The present invention relates to novel fusogenic vesicles as highly efficient and versatile encapsulation systems for delivering a substance of choice, such as nucleic acids, proteins, peptides, antigens, pharmaceutical drugs and cosmetic agents to cells and tissues.
BACKGROUND OF THE INVENTION
Various publications or patents are referred to in parentheses throughout this application to describe the state of the art to which the invention pertains. Each of these publications or patents is incorporated by reference herein.
One of the paramount goals of medical therapy is the efficient delivery of therapeutic substances to the site of disease. While some therapeutic substances can be delivered in free form, others require a carrier or vector in order to reach and enter their final destination, either due to their rapid clearance from the area of introduction or their inability to cross biological barriers, or due to their systemic toxicity. Delivery of substances to cells and tissues requires vectors which are efficient, flexible, easy to prepare and safe. Currently available methods for delivering substances to eukaryotic cells involve the use of either viral or non-viral vectors. Viral vectors are replication-defective viruses with part of their coding sequences replaced by that of a therapeutic gene. Although recombinant viruses are highly efficient gene delivery and expression vectors, they are currently limited to the delivery of nucleic acids and their safety profiles have not yet been established for medical use in humans.
Most non-viral delivery systems operate at the following levels: loading of the delivery vector with a substance of interest (e.g. proteins, peptides, nucleic acids, pharmaceutical or other therapeutic drugs), endocytosis, and in the case of gene delivery, nuclear targeting and entry. The major drawback of non-viral systems, such as liposomes, is their low delivery efficiency to cells (Chu et al., id.; Legendre and Szoka, 1992, Pharmaceut. Res. Vol. 9, P. 1235), presumably due to the absence of fusion-mediating molecules on the surface of the liposomes. A hybrid type of delivery system, the virosome, combines the efficient delivery mechanism of viruses with the versatility and safety of non-viral delivery systems. Virosomes are reconstituted envelopes without the infectious nucleocapsids and the genetic material that can be derived from a variety of viruses. These virosomes are functional, in that their membrane fusion activity closely mimics the well-defined low-pH-dependent membrane fusion activity of the intact virus, which is solely mediated by the viral fusion protein. Like viruses, virosomes are rapidly internalized by receptor-mediated endocytosis. In contrast to viral systems virosomes are safe, since the infectious nucleocapsid of the virus has been removed. Thus, virosomes represent a promising carrier system for the delivery of a wide variety of different substances, either encapsulated in their aqueous interior or co-reconstituted in their membranes. Co-reconstitution of different receptors within the virosomal membrane, furthermore, allows the targeting of virosomes to different cells or tissues. So far, virosomes are mainly used as vaccines by adding antigen onto the surface of the virosomes.
A major limitation of the protocol currently used to prepare virosomes is that it does not result in high encapsulation efficiency. At the lipid concentration at which virosomes are produced (˜1 mM lipid), and given their mean diameter of approximately 200 nm, less than 1% of the aqueous phase will be entrapped within the virosomes (Schoen et al., J. Liposome Res. 3: 767-792, 1993). Such low entrapment rates reduce virosome-mediated efficiency of drug or gene delivery to cells. The development of new/novel, more efficiently loaded vesicles that retain the advantageous fusion properties of virosomes, as well as methods of making, loading, and delivering them would thus be a highly desirable goal in the field of therapeutic drug, protein and gene delivery.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1: This figure shows the simplified scheme representing the overall strategy of the invention. In a first fusion step, chimeric virosomes are fused at 4° C. (pH 5.0) with liposomes containing the substance to be delivered, (in this example DNA) mediated by X-31 HA. Subsequently, the resulting neutralized fusion products are incubated with cells. After receptor-mediated endocytosis, the virosome-like proteoliposomes now containing the substance to be delivered undergo a second round of fusion, triggered by the low pH within endosomes and mediated by PR8/34 HA, releasing the entrapped substance into the cytosol.
FIG. 2: This figure shows the presence of both types of fusion proteins in the membrane of the chimeric virosomes. SDS-PAGE of X-31 and PR8 virus and virosomes. Approx. 10 μg virus and 3 μg virosomes were loaded under non-reducing conditions on a SDS-Tris/Tricine 10% (v/v) polyacrylamide gel (Schäigger et al., Anal. Biochem. 166: 368-379, 1987). The polyacrylamide gel was stained with Coomassie blue. Lane 1: X-31 virus. Lane 2: X-31 virosomes. Lane 3: chimeric virosomes. Lane 4: PR8 virosomes. Lane 5: PR8 virus.
FIG. 3: This figure shows the morphology of the chimeric virosomes by cryo-TEM. They are unilamellar vesicles with diameters ranging from 150-250 nm. Scale bar 100 nm.
FIG. 4: This figure shows the fusion activity of X-31, PR8 viruses, virosomes and X-31/PR8 chimeric virosomes with liposomes. Fluorescence was recorded continuously at excitation and emission wavelengths of 465 nm and 530 nm, respectively. Measurements were carried out with a SPEX-Fluorolog-2 fluorometer (SPEX Industries, Inc., Edison, N.J., USA) equipped with a thermostated cuvette holder and a magnetic stirring device. For calibration of the fluorescence scale the initial residual fluorescence of the liposomes was set to zero and the fluorescence at infinite probe dilution to 100%. The latter value was determined by addition of Triton X-100 (0.5% v/v). 1) X-31 virus and virosomes fused with POPC/POPG-LUVs. 2) PR8/34 virus and virosomes incubated with POPC/POPG-LUVs. 3) Chimeric virosomes incubated with POPC/POPG-LUVs. a) fusion measurements at 4° C.; b) Virus and virosomes were incubated at 4° C. and pH 5.0 for 1 h with POPC/DDAB-LUVs before fusion measurements at 37° C.
FIG. 5: This figure shows the encapsulation efficiency of DNA into POPC/DDAB-liposomes. DNA-containing liposomes were prepared by the freeze/thaw method as described in the examples and digested for 3 h at 37° C. with DNase I/Exonuclease III. After that, encapsulated DNA was phenol/chloroform extracted and subjected to gel electrophoresis on a 1% agarose gel. In lane M 1 Kb plus Ladder marker was loaded. As control 100 ng (lane 1) and 200 ng plasmid-DNA (lane 2) were directly loaded on the gel. As another control untreated (not DNase I digested) DNA-liposomes corresponding to 20 ng (lane 3), 50 ng (lane 4), 100 ng (lane 5) and 200 ng (lane 6) plasmid were phenol/chloroform extracted. The aqueous phase was loaded on the gel. DNase I digested liposomes corresponding to an initial DNA-amount of 50 ng (lane 7) and 100 ng (lane 8) were treated like the samples before. Lane 9 and 10 correspond to the same amount of liposomes, except that liposomes had been solubilized with 1% (v/v) Triton X-100 before DNase I treatment. In lane 11 and 12 DNA was not encapsulated, but added to empty liposomes. Also these liposomes were treated as described above.
FIG. 6: This figure shows the morphology of loaded liposomes by cryo-TEM of. They are unilamellar and measure between 100-150 nm in diameter. Scale bar 100 nm.
FIG. 7: This figure shows the products of the fusion between the chimeric virosomes and the loaded liposomes by cryo-TEM before (upper panels) and after extrusion (lower panels). Before extrusion, the vesicles are of widely varying sizes with some vesicles exceeding 900 nm. This result indicates multiple rounds of fusion. Scale bars upper panels: 500 nm; scale bars lower panels: left: 100 nm; right: 200 nm.
FIG. 8: This figure demonstrates the fusogenicity of the extruded vesicles resulting from the fusion of chimeric virosomes with liposomes. Remarkably, the vesicles retain their fusion activity. The proteoliposomes were extruded through nucleopore membranes. Fluorescence was recorded continuously at excitation and emission wavelengths of 465 nm and 530 nm, respectively. Measurements were carried out with a SPEX-Fluorolog-2 fluorometer equipped with a thermostated cuvette holder and a magnetic stirring device. 100% corresponded to infinite probe dilution, determined after addition of Triton X-100 (0.5% v/v). Fusion measurements were carried out at 37° C. and pH 5.0 with POPC/POPG (4:1)-LUVs containing 0.6% N-NBD-PE and N-Rh-PE.
FIG. 9: This figure shows that the proteoliposomes are rapidly internalized by cells. Rhodamine-labeled proteoliposomes are detected in MDCK and HeLa cells. Fused virosomes were bound at 4° C. for 1 h to HeLa and MDCK cells. After that, unbound material was washed away. Cells were immediately fixed (A), or incubated for another 15 min at 37° C., fixed with 3.7% (v/v) formaldehyde and analyzed by fluorescence microscopy (B).
FIG. 10: This figure shows the rapid uptake of proteoliposomes loaded with Texas Red dextran by HeLa cells. HeLa cells were incubated with Texas Red dextran-containing proteoliposomes for 1 h at 4° C. After that, unbound material was washed away and fixed with 3.7% (v/v) formaldehyde (A). Otherwise, cells were further incubated for 15 min (B), 2 h (C) or 5 h (D) at 37° C. After fixation, cells were analyzed by fluorescence microscopy. Objective magnification: for (A) 63×; for (B), (C), (D) 100×.
DETAILED DESCRIPTION OF THE INVENTION
One of the problems in current virosome technology is the lack of methods for the efficient entrapment of a solute, e.g. protein, nucleic acid, or pharmaceutical drug. The present invention provides a novel approach that circumvents the problems of low virosomal entrapment efficiencies.
Therefore the invention provides a fusogenic vesicle, composed of virosomal and liposomal lipids that is capable of encapsulating at least one therapeutic or immunologically active substance whereby said fusogenic vesicle is comprised of fusion proteins, preferably at least two different fusion proteins or peptides with distinct fusion characteristics. Preferably, the vesicle is unilamellar. The vesicle has a diameter generally in the range of 100 to 600 nm, and preferably a diameter of between 100 nm and 300 nm or a diameter of between 200 nm and 400 nm, depending on the specific vesicle.
The invention takes advantage of the fact that liposomes, which can be prepared at very high lipid concentrations, have high encapsulation efficiencies. Thus, the invention provides a methodology that combines the high loading capacity of liposomes with the efficient cell-fusion and delivery of virosomes, resulting in substantially increased entrapment of solutes, like proteins, nucleic acids, and pharmaceutical drugs, within functional chimeric virosomes of controlled size that are capable of efficiently delivering therapeutic or immunologically active substances to cells.
Accordingly, in a preferred embodiment of the invention, the substance of interest is first loaded into, or encapsulated by, liposomes. The liposomal lipids which can be used in the present invention include cationic lipids, synthetic lipids, glycolipids, phospholipids cholesterol or derivatives thereof, and equivalent molecules known to those of skill in the art. Phospholipids can comprise preferably phosphatidylcholine, sphingomyelin, phosphatidylethanolamine, phosphatidylserine, phosphatidylglycerol, phosphatidic acid, cardiolipin and phosphatidylinositol with varying fatty acyl compositions. Cationic lipids can comprise preferably DOTMA (N-[(1-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride), DOTAP (N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium chloride, DODAC (N,N-dioleyl-N,N,-dimethylammonium chloride), DDAB (didodecyldimethylammonium bromide) and stearylamine or other aliphatic amines and the like. They are generally formulated as small unilamellar liposomes in a mixture with DOPE (dioleoylphosphatidyl ethanolamine) that is widely used as a ‘helper’ lipid to facilitate disruption of the endosomal membrane. Most preferably the liposomal lipids of the liposomes comprise POPC/DDAB.
In a second step, liposomes containing the substance of interest will be fused with chimeric virosomes having different types of fusion proteins in their membranes. The term “fusion protein” refers to peptides or proteins capable of inducing and/or promoting a fusion reaction between the fusogenic vesicle membrane and a biological membrane of the target cell. After resizing, the products of the liposome-virosome fusion will be virosome-like particles, or proteoliposomes, that contain the substance of interest in their internal cavity and are still capable of undergoing a second fusion step, under different conditions, with biological membranes in order to deliver the therapeutic or immunologically active substance.
The fusogenic vesicle of the present invention is capable of encapsulating at least one substance, preferably a therapeutic or an immunologically active substance. Examples of preferred substances suitable for encapsulation into the fusogenic vesicles are DNA, RNA, siRNA, proteins, peptides, amino acids and pharmaceutical active substances or derivatives thereof. Preferably the at least one substance is a pharmaceutical drug, an antigen, a cosmetic agent or a mixture thereof. The substance can also be a precursor to said pharmaceutical drug, antigen, or cosmetic agent that is converted into the final agent by the cell, tissue or interstitium, or by some other mechanism.
Examples of cosmetic substances are well known in the art and may comprise antipsoriatics or antifungals for dermatological use. Examples of therapeutic substances or pharmaceutical drugs are also well known in the art and may comprise anaesthetics, angiogenesis inhibitors, anti-acne preparations, anti-allergica, antibiotics, chemotherapeutics, antihistamines, antiinflammatory or antiinfective agents, antineoplastic agents, antigens, antiprotozoals, antirheumatic agents, antiviral vaccines, antiviral agents, aptoptosis inducing agents, bacterial vaccines, chemotherapeutics, cytostatica, immunosuppressive agents, laxatives, psycholeptics.
Preferred examples of cosmetic agents include tars and nystatin, while preferred examples of pharmaceutical drugs or immuno-active substances include doxorubicin, vinblastine, cisplatin, methotrexate, cyclosporine, ibuprofen, HCV-based T-cell antigens, and tumor-specific and tumor-associated antigens.
Encapsulation of the therapeutic substances, such as proteins, peptides, nucleic acids, pharmaceutical, chemotherapeutic or cosmetic drugs into liposomes can be performed by any method known in the art, including the procedures described in Monnard et al., Biochim. Biophys. Acta 1329: 39-50; 1997, in Wagner et al., J. Liposome Res., 12(3) 271-283, 2002, or in Oberholzer et al., Biochim. Biophys. Acta 1416: 57-68; 1999, among many other well-known methods. In a preferred embodiment of the invention, high liposomal encapsulation efficiencies are achieved by the freeze/thaw technique used to prepare pure lipid vesicles. With this method, approximately 50% of the initial amount of a linearized plasmid over 3 kb can be entrapped within large unilamellar vesicles (LUVs) consisting preferably of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC)/didodecyldimethylammonium bromide (DDAB) (Monnard et al., Biochim. Biophys. Acta 1329: 39-50, 1997).
In one preferred embodiment, the invention provides for the preparation of virosomes containing a binary mixture of viral fusion molecules that display distinct fusion characteristics. In another preferred embodiment of the invention, different viral hemagglutinin (HA) fusion proteins are used to construct the chimeric virosomes. As shown previously (Tsurudome et al., J. Biol. Chem. 267: 20225-20232, 1992), HA fusion proteins from different strains of viruses can display markedly different temperature characteristics of fusion and inactivation. For example, at about pH 5.0, X-31 HA triggers fusion efficiently at low temperature, whereas at the same pH, HA from PR8/34 or A/Singapore virus requires elevated temperature (>25° C.). Hence, in a preferred embodiment of the invention, the chimeric virosomes contain proteins in their membrane that mediate fusion at two distinct temperatures. Different temperature-sensitivity is a particularly advantageous characteristic of the fusion proteins, as it allows convenient and simple control of fusion reactions. As an example, the present invention demonstrates that virosomes containing HA molecules from both X-31 and PR8/34 virions are capable of catalyzing two distinct fusion reactions at pH 5: the first at low temperature (0-4-10° C.), the second at elevated temperature (>25° C.). However, other fusion proteins with distinct fusion characteristics, including sensitivity to temperature, ion concentration, acidity, cell type and tissue type specificity, etc. are well known in the art and may be used for the purposes of the present invention. Fusion proteins with different fusion characteristics can be derived from different influenza strains, such as MRC-11, X-97, NIB24, NIB26, X-47, A/Johannesburg/33 and A/Singapore, to name a few.
In addition, other known viral fusion proteins may be used, such as the vesicular stomatitis virus (VSV) G protein, the Semliki forest virus (SFV) E1 protein, or the Sendai virus F protein, among many others, to construct chimeric virosomes capable of undergoing sequential and separate fusion events. Furthermore, the chimeric virosomes and resulting virosome-like proteoliposomes loaded with the therapeutic substance of interest may contain fusion proteins that target the virosome-like proteoliposomes to specific cells or tissues. Thus, for instance, fusion proteins that are specific for certain cell surface receptors, such as the gp41/gp120 protein of HIV (gp120 binds to CD4 on CD4+ T lymphocytes and cells of the monocyte/macrophage lineage) can be used to target delivery of therapeutic substances to specific cell types or tissues. Other such cell- or tissue-specific fusion proteins are well known in the art and can be conveniently incorporated into the vesicles of the present invention.
In addition to the fusion proteins, the fusogenic vesicle of the present invention can further have, incorporated in the membrane or attached thereto, a cell-surface receptor, a cytokine, a growth-factor, an antibody or an antibody fragment to improve targeting of the fusogenic vesicle to different cells or tissues.
As shown by the present invention, once such chimeric virosomes are constructed, they can be fused at a first specified condition to trigger fusion of the first viral fusion protein or peptide with liposomes containing the encapsulated substance or substances of interest, such as proteins, peptides, nucleic acids, or pharmaceutical drugs. The invention demonstrates for the first time that the fusion of chimeric virosomes with substance-loaded liposomes generates virosome-like proteoliposomes which not only encapsulate the substance of interest, but are still capable of mediating an additional fusion event under different conditions, for example, under different conditions for temperature, ion concentration etc. After proper manipulation to control their size, these substance-containing proteoliposomes can be delivered to cells where they are efficiently taken up. In the environment of the endosome, the second fusion molecules can trigger a second fusion reaction between the virosomal and endosomal membranes. As illustrated in FIG. 1 this multi-step process results in the transfer of liposome-encapsulated substances into the cytosolic compartment of cells.
Procedures for the preparation of virosomes (also referred to in this application as immunopotentiating reconstituted influenza virosomes (IRIV)) are well known to those of skill in the art (Bron et al., Methods Enzymol. 220: 313-331, 1993). Influenza virosomes, for example, can be reconstituted from the original viral membrane lipids and spike glycoproteins after solubilization of intact influenza virus with octaethyleneglycol monododecyl ether, sedimentation of the nucleocapsid (the viral glycoproteins and lipids will remain in the supernatant), and removal of the detergent in the supernatant with a hydrophobic resin (Bio-Beads SM2) (Stegmann et al., Traffic 1: 598-604, 1987). For the preparation of chimeric virosomes containing HAs from X-31 and PR8/34 strains of viruses, as one example, equal amounts of protein of both viruses were solubilized with the non-ionic detergent C12E8. Isolation of other viral fusion proteins is a matter of routine to those of skill in the art. After removal of the detergent with Bio-Beads SM2, new envelopes containing both types of fusion proteins were formed. SDS-PAGE of exemplary chimeric virosomes containing two different types of HA fusion proteins showed (FIG. 2) that very similar amounts of the fusion proteins, here PR8 and X-31 HA, are reconstituted into chimeric virosomes. The HA/phospholipid ratio is approximately 1.4 mg/μmol.
To determine their morphology, the chimeric virosomes constructed can be characterized by cryo-TEM. The reconstituted particles were unilamellar vesicles with a diameter ranging between 150-250 nm; some smaller vesicles were also visible (FIG. 3). The density of hemagglutinin spikes was clearly lower than that typically seen in intact virus particles. This is consistent with the measured HA/phospholipid ratio of 1.4 mg/1 mol for chimeric virosomes, which is somewhat lower than that of intact virions (approximately 2 mg/μmol). Furthermore, HA-spikes are present on both sides of the membrane at about equal density.
The fusion properties of chimeric virosomes with liposomes, such as the dependence of fusion activity on a certain temperature, pH, or other parameter, in vitro can be determined using the lipid mixing assay (Struck et al., Biochemistry 20: 4093-4099, 1981). This assay is based on fluorescence resonance energy transfer (FRET) and makes use of the two fluorophores N-NBD-PE (energy-donor) and N-Rh-PE (energy-acceptor) present within one of the two membranes (e.g. liposomes). A variant of this assay, utilizing the same donor N-NBD-PE but a different acceptor, cholesterol-anthracene-9-carboxylate (CAC), may also be used. Another variant of this assay, utilizing the same acceptor N-Rh-PE but a different donor, Bodipy 530/550-DHPE, may also be used. Upon fusion the two fluorophores move apart and the fluorescence emitted by the donor increases (or that of the acceptor decreases). As an example, the fusion activity of X-31/PR8/34 chimeric virosomes was compared with that of X-31 and PR8/34 virus as well as with the corresponding virosomes (FIG. 4). As target membranes, unilamellar liposomes prepared from POPC/POPG (4:1) and the two fluorophores (0.6% each) were used. As shown in FIG. 4.1a, both X-31 virus and X-31 virosomes fused efficiently with the liposomes at 4° C. and pH 5.0. Because the X-31 and PR8/34 HA fusion proteins both require an acidic pH, no fusion occurred at neutral pH. Preincubation of X-31-HA at low temperature and low pH almost completely abolished its fusion-activity (FIG. 4.1b).
By contrast, at 4° C., pH 5.0, PR8/34 virus and virosomes neither fused nor, importantly, underwent inactivation (FIG. 4.2a). However, they were able to trigger fusion at 37° C. (FIG. 4.2b). According to these data, chimeric virosomes can be designed to perform two distinct fusion reactions, depending on the parameters for which each fusion protein is specific. For example, the PR8/34/X-31 chimeric virosomes can perform one fusion reaction at low temperature (4° C.; FIG. 4.3a) and a second fusion reaction at higher temperature (37° C.; FIG. 4.3b). A surprising and important discovery of the present invention is that the preincubation of chimeric virosomes with target membranes at low temperature and low pH does not eliminate fusion-activity at elevated temperature (FIG. 4.3b). Therefore, the present invention provides for a controlled fusion reaction that results in the formation of loaded virosomes capable of fusing again under different conditions.
The fact that the prepared liposomes are unilamellar is of great importance for the subsequent fusion step with chimeric virosomes, since the loaded substance should be released into the internal cavity of the newly formed vesicle rather than remaining trapped between membrane layers. To examine a possible interaction of the loaded substance with the lipids used and its effect on the liposome morphology, plasmid-containing vesicles were subjected to cryo-TEM. The electron micrographs in FIG. 6 show that the cationic vesicles were unilamellar and 100-150 nm in size. In contrast to other preparations containing cationic lipids and DNA, where multilamellar, tubular structures are formed (Gershon et al., Biochemistry 32: 7143-7151, 1993), advantageously in this procedure homogeneous unilamellar vesicles are generated.
To demonstrate the feasibility of the approach, the encapsulation efficiency of the therapeutic or immunologically active substance of choice, in this case the 6.5 kb model plasmid, was determined. As determined after digestion of free plasmid DNA with DNase I/Exonuclease III digestion and separation by gel filtration the entrapment efficiency achieved was approximately 40% of a circular, 6.5 kb model plasmid. Quantitation of the remaining intact DNA can either be based on radioisotopic labeling of the DNA, or visual comparison with standard amounts on an agarose gel following staining with ethidium bromide, as described in FIG. 5. To ascertain that accessible DNA was degraded quantitatively, control experiments were performed in which all DNA was accessible (i.e. liposomes solubilized with a detergent (FIG. 5, lanes 9 and 10) or DNA on the outside (FIG. 5, lanes 11 and 12)). These results thus demonstrate that the DNase I/exonuclease III-resistant fraction is effectively entrapped in the internal cavity of the liposomes, and not just inaccessible to DNase I and exonuclease III because of (protective) interactions with the cationic lipid. The entrapment achieved (35-40%) is quite significant, thus making this approach suitable for therapeutic applications.
Following the successful preparation of the chimeric virosomes and the unilamellar liposomes loaded with the substance of interest, the present invention provides for the fusion of these two vesicular systems. However, there are several major obstacles that had to be overcome before the desired fusogenic and loaded virosomes were obtained. Because virosomes are capable of fusing both with liposomes and other virosomes, as well as with the primary products of fusion, multiple rounds of fusion were shown to occur in the mixture of loaded liposomes with chimeric virosomes. These uncontrolled fusion events resulted in a significant percentage of large, undefined structures. Such large structures are unsuitable for the delivery of therapeutic substances to cells, as their endocytosis by cells is expected to be severely compromised. Furthermore, the multiple rounds of fusion resulted in populations of virosome-like end products (proteoliposomes) with a wide variety of different sizes. Such heterogeneously sized populations are therapeutically undesirable because the amount and dosage of the encapsulated therapeutic substance cannot be controlled. In addition, encapsulation of the original liposome-entrapped therapeutic or immunologically active substance of interest within an internal compartment of the resulting chimeric virosome-like proteoliposomes has not previously been demonstrated. It was doubtful whether successful encapsulation would be achieved or whether fusion would result in partial or total release of the therapeutic substance of interest. Lastly, it was uncertain whether the fused vesicles would retain fusion capacity under the second set of conditions or parameters, such as elevated temperature.
To investigate the products of the fusion between liposomes and virosomes containing different fusion proteins, samples of fused vesicles were characterized by cryo-transmission electron microscopy (cryo-TEM). The liposome-virosome fusion products showed vesicles with a size distribution between 450 nm and 900 nm (FIG. 7, upper panels). Some smaller vesicles, possibly original virosomes, were detected (FIG. 7, top left). On the other hand, it appeared that all of the liposomes underwent at least one round of fusion. The presence of very large vesicles (>900 nm) is evidence that multiple rounds of fusion take place in the virosome-liposome fusion step. In order to solve the problem of uncontrollable fusion and to generate homogeneously sized vesicles that would allow control over drug or substance delivery dosage, the vesicles would have to be manipulated so as to generate vesicles of uniform size. Ideally, the large diameter vesicles that are unsuitable for endocytosis would have to be reduced in size without causing loss of the encapsulated contents or fusogenicity. Therefore, the present invention further comprises re-sizing the fusogenic vesicles obtained after fusing liposomes encapsulating at least one therapeutic or immunologically active substance with virosomes having fusion proteins with distinct fusion characteristics. In accordance with the invention, said resizing can be carried out by extrusion of the vesicles.
Accordingly, the present invention provides for a size-reduction step of the virosome-like proteosomes obtained through the fusion of chimeric virosomes with loaded liposomes, which remarkably causes neither loss of the encapsulated substance from the vesicles nor inactivation of the remaining functional fusion protein with concomitant loss of fusogenicity. Thus, in a preferred embodiment of the invention, the reaction products of the virosome-liposome fusion reaction are subjected to a nucleopore extrusion step in order to reduce their size. As shown in FIG. 7, lower panels, extrusion through 200 nm pores yields vesicles of approximately half of the original size as shown by cryo-TEM. Most particles are between 100-300 nm, rarely also larger vesicles (500-600 nm) can be detected. The fact that the extrusion step results in re-sized fusogenic virosome-like vesicles without significant loss of encapsulated substance is a rather unexpected and highly advantageous result, as the shearing forces applied during extrusion would be expected to rupture the majority of the membranes, causing loss of entrapped material and inactivation of the fusion proteins. That the extrusion step leads neither to any significant loss of the entrapped plasmid-DNA nor inactivation of the second fusion protein (the PR8-hemagglutinin), was verified by testing the fusogenicity and the DNA-content of extruded proteoliposomes. FRET-measurements showed that the fusogenicity of the loaded proteoliposomes before and after extrusion was the same (FIG. 8).
Further, the plasmid content of the extruded proteoliposomes was determined. Toward this end, liposomes containing radioactive plasmid DNA were prepared (free DNA was digested with DNAse and nucleotides were removed by gel filtration; entrapped DNA=100%). Subsequently, these liposomes were subjected to fusion with virosomes. One aliquot of the fusion products was treated directly with DNase I, the other following extrusion through nucleopore membranes. After removal of the free nucleotides, the radioactivity entrapped within the proteoliposomes was determined. From the results it can be concluded that during the initial fusion step, no more than 15% (±5% s.d.; n=2) and during both the fusion and extrusion step, no more than 25% (±3% s.d.; n=2) of the originally entrapped DNA is lost. Thus, the amount of radioactivity found in proteoliposomes after extrusion is only slightly lower than before extrusion. That the liposome-virosome fusion retains most of the entrapped substance is even more surprising when one considers the reports that HA-mediated membrane fusion is a leaky process (Shangguan et al., Biochemistry 35: 4956-4965, 1996).
Thus, the present invention provides a procedure for the preparation of fusion-competent proteoliposome vesicles that efficiently encapsulate a substance of interest for delivery to cells and tissues. The cellular uptake of these novel proteoliposome vesicles can be traced by the fluorescent lipid rhodamine phosphatidylethanolamine (Rh-PE) stably inserted within the membrane bilayer. The labeled proteoliposomes were added to MDCK or HeLa-cells (approx. 1×105 cells) cultured on cover slips in 24-well plates. Following incubation at 4° C. for 1 h to allow binding to cells, unbound particles were removed and cells were incubated for an additional 15 min at 37° C. After fixation with formaldehyde cells were analyzed by fluorescence microscopy.
After this (short) incubation at 37° C., a clear punctate, perinuclear staining was visible (FIG. 9 B), indicating that rhodamine-labeled proteoliposomes had been rapidly internalized. This process was temperature dependent since at 4° C. all particles remained at the cell surface (FIG. 9 A). Furthermore, it was also dependent on the fusion protein in the vesicle membrane, as protein-free Rh liposomes were not internalized to a detectable extent within 2 hours. These results show that chimeric proteoliposomes are rapidly internalized; in all likelihood by receptor-mediated endocytosis, as is the case for influenza virus. Likewise, dextran can be added to liposomes at a high concentration and because of its smaller size it is assumed that several labeled molecules can be incorporated into one proteoliposome. In fact, theoretically, at the concentration of dextran used (37.5 mg/ml) approximately 170 dextran molecules can be incorporated into one liposome. With the degree of substitution being six fluorophores per dextran molecule, one thousand Texas-red molecules should be available per liposome.
It was expected that if the encapsulated material is released into the cytosol a diffuse, cytosolic staining should be visible by fluorescence microscopy versus a punctate perinuclear staining if the substance remains entrapped. Dextran was entrapped into POPC/DDAB-liposomes as described in the examples. Liposomes were fused with chimeric virosomes and the fusion products added to cells cultured on coverslips. Cells were incubated for 1 h at 4° C., then unbound material was washed away and cells were incubated for different time at 37° C. After fixation with 3.7% (v/v) formaldehyde, cells were analyzed by fluorescence microscopy. Uptake of dextran-proteoliposomes could be detected after a 15 minute incubation at 37° C., as already seen for rhodamine-labeled proteoliposomes. At this stage a substantial amount of dye was at the cell surface (FIG. 10 B). Later (FIGS. 10 C and D) a more perinuclear staining could be observed.
To further demonstrate the feasibility of the approach, the encapsulation efficiencies of other therapeutic substances of choice were determined. Hydrophobic as well as hydrophilic peptides, proteins (e.g. recombinant green fluorescent protein, GFP) and small molecules (e.g. PicoGreen) were used. Table 1 represents a summary of these results. With the conventional preparation procedure (Schoen et al., J. Liposome Res. 3: 767-792, 1993) encapsulation efficiencies below 5% were obtained. However, the new approach resulted in encapsulation efficiencies at least above 15% in every case (Table 1).
Encapsulation efficiencies of different molecules
at subsequent steps during IRIV preparation
(0.2 μm diameter)1