WO1995015746A1

WO1995015746A1 - Liposome delivery systems

Info

Publication number: WO1995015746A1
Application number: PCT/GB1994/002702
Authority: WO
Inventors: Gregory Gregoriadis; Brenda Mccormack
Original assignee: The School Of Pharmacy
Priority date: 1993-12-10
Filing date: 1994-12-09
Publication date: 1995-06-15

Abstract

The invention provides liposomes containing in the aqueous phase inside the liposome at least one complex, wherein the complex comprises at least one molecule non-covalently bound to at least one receptor. Preferably this binding is effective to modify the interactions of said molecule with other components of the liposome, for instance solubility properties. This system may be used to include in liposomes hydrophobic non-water-soluble molecules and/or molecules which tend to leak out of liposomes when not bound to a receptor. The bound molecule may be a pharmaceutical active, in which case the liposomes may be used as a drug delivery system.

Description

Liposome Delivery Systems Liposomes may be used in a variety of pharmaceutical applications, in particular as delivery systems for drugs and other substances. Drugs and other substances may be entrapped in the liposomes, for instance by dissolution in the aqueous phase of the liposomes or incorporation in the lipid phase (the bilayer) . Delivery of drugs and other substances, for instance intravenously, orally or transdermally, using liposomes can result in various benefits, for instance increased half-life of drugs in blood circulation and targeting of drugs or other substances to specific sites in the body, for instance diseased cells or lymph nodes (particularly useful in the administration of vaccines) . Liposomes are also known to be useful for topical administration of actives (useful for instance in the cosmetics industry) .

In general the solvent contained within the liposomes is water-based. Therefore entrapment of water soluble or hydrophilic substances is not problematic. Such substances may be dissolved in the aqueous phase. Difficulties are however encountered when it is attempted to entrap substances having low water solubility.

Attempts have been made to develop a liposome-based delivery system for hydrophobic substances.

It has been attempted to entrap such substances by incorporation into the lipid bilayer. This approach has certain disadvantages. Firstly, the presence of the drug or other substance in the lipid bilayer may destabilise the liposome structure. Secondly, the amount of drug or other substance which may be incorporated into any one liposome by inclusion in the lipid bilayer is more limited than the amount which may be incorporated by solvation in the aqueous phase. Attempts have been made to incorporate hydrophobic substances into liposomes by more permanent methods. Such substances have been covalently linked to phospholipids; the hydrophobic substance - phospholipid conjugate can then form an element of the lipid bilayer. Again this has the disadvantage that only a limited amount of the substance can be incorporated. This method is also time-consuming and complicated.

It has also been attempted to provide liposomes in which DMSO is the encapsulated solvent rather than one which is water-based. DMSO is capable of solvating hydrophobic substances. This method is however not useful in practice. Large amounts of solvent are required to solvate the hydrophobic substances (larger amounts than the amount of water required to solvate an equivalent molar amount . of a hydrophilic substance) making the method expensive. In addition the liposomes will be large, which is not preferred. Furthermore, DMSO can be toxic in large amounts; for instance because it interacts with cell membranes.

There exists a problem with the use of liposomes as delivery systems for some water-soluble substances, in particular substances of low molecular weight. Such substances tend to escape, or leak out of, the liposomes. Attempts have been made to provide liposomes in which the lipid bilayer has reduced permeability but these attempts have not proved very successful in solving this problem. There are systems known which are considered useful for the delivery of hydrophobic substances. These systems are based on cyclodextrins. Cyclodextrins, first described by Villiers in 1891, are non-reducing cyclic oligosaccharides composed of α-(1,4)-linked D glucopyranosyl units. There are three naturally-occurring cyclodextrins, known as o-, β- and γ- cyclodextrin, made up from 6, 7 and 8 glucopyranose units respectively.

Cyclodextrin molecules have a geometry which may be described as annular or "bucket" shaped. The geometry and electronic structure of a cyclodextrin molecule is such that the interior surface of the "bucket" is hydrophobic and the external surface of the "bucket" is hydrophilic. A cyclodextrin molecule is illustrated schematically in Figure 1, showing the hydrophobic cavity l and the hydrophilic outer surface 3.

A particularly useful feature of cyclodextrins is their ability to form inclusion complexes ("host-guest" complexes) with a wide range of substances. Molecular encapsulation of guest molecules within host cyclodextrins allows modification of the apparent physical and chemical properties of these guest molecules. In particular solubility properties may be modified. The cyclodextrins are water-soluble. This characteristic derives from the location of all three hydroxyl groups of each successive glucose unit on the rims of the bucket-shaped molecules - the C₆ primary hydroxyls 5 on the narrower side 7 and the C₂ and C₃ secondary hydroxyls 9 occupying the wider side 11. These two hydrophilic planes confer hydrophilicity upon the entire molecule. As mentioned above, however, the interior surface of the molecule is hydrophobic. Thus hydrophobic molecules may be encapsulated by the cyclodextrin molecule. The driving force for formation of such a host-guest complex is displacement of water molecules from the hydrophobic cavity by the more hydrophobic guest molecule to attain a polar-apolar association and decrease of cyclodextrin ring strain, resulting in a more stable lower energy state. cr-, β-, and 7-cyclodextrins have a "depth" along the curved wall of about 7.8 angstroms. An α-cyclodextrin has a cavity of diameter about 5.7 angstroms and a total diameter on the wider side of about 13.7 angstroms. These dimensions in a β-cyclodextrin are around 7.8 angstroms and 15.3 angstroms respectively. In a γ-cyclodextrin they are around 9.5 angstroms and 16.9 angstroms respectively.

Cyclodextrins and their uses and properties are discussed by Parrish in "Cyclodextrins - a review", Speciality Chemicals 1(6), 366 to 380 (1987) and more recently by Uekama and Otagiri in CRC Critical Reviews in

Therapeutic Drug Carrier Systems, vol 3 1987 ppl-40. Cyclodextrins have also been used to improve a variety of other properties of drugs. For instance, the encapsulation of active ingredient may protect it from its environment and thus from reactions which may adversely affect its storage stability, such as hydrolysis, oxidation or volatilisation. Encapsulation may also mask unpleasant tastes and reduce local irritancy and haemolysis.

Cyclodextrins have been incorporated into vesicles by Bellanger and Perly described in "Amphiphilic Cyclodextrin Derivatives As Potential Vectors For Hydrophobic Drugs", Sixth International Cyclodextrin Symposium, Chicago 1992. Cyclodextrin molecules were covalently attached to phospholipids to give compounds named "lollipops", the lipid parts of which can be incorporated into the lipid bilayer of organised structures such as liposomes, with the cyclodextrin moiety thus attached to the inner or outer surface of the lipid bilayer of such liposomes and extending therefrom.

At the same Symposium, Loftsson et al described formation of microcapsules containing cyclodextrins, in "Microcapsules of drug-cyclodextrin complexes". The preparation was described of microcapsules having a polymeric coating which encapsulates a solution of a 2- hydroxypropy1-β-cyclodextrin complex ofhydrocortisone. The polymeric coating is dissolved by gastric fluid, allowing release of the cyclodextrin complex.

International Patent Publications 093/25195 and W093/25194, both published after the priority date of the present invention, describe the production of nanocapsules of which the membranes are formed of acyl-modified cyclodextrin. The nanocapsules contain oil, in which may be dissolved an active molecule. The nanocapsules also optionally comprise surfactant. It is stated that the system may be used as a carrier for pharmaceutical and other molecules. A disadvantage of this system as a drug delivery formulation is that it requires the presence of components such as the oil, which may not be compatible with somatic systems.

Attempts have been made to deliver drugs and other substances whose molecules are hydrophobic in the form of cyclodextrin drug complexes. Cyclodextrin delivery systems experimented with so far have been found to possess certain disadvantages.

Cyclodextrins and cyclodextrin complexes tend to be eliminated unchanged by the kidneys very rapidly, thus not allowing release of the encapsulated substance into the body. Further it has been found that CD's can be toxic to the kidneys.

A further disadvantage results from competition with other molecules in vivo; the molecule of the drug or other substance may be replaced in the cyclodextrin by, for instance, components in the plasma (e.g plasma proteins) gastric contents and other biological fluids. Thus the drug or other substance may be released prematurely.

Cyclodextrins may react with a complex lipophilic component in the body. Cyclodextrins tend to extract components of bio-surfaces they encounter, for example the surfaces of erythrocytes and also lipoproteins. This behaviour also results in the membrane solubilising properties of cyclodextrins, which can be detrimental if large amounts of cyclodextrin are administered.

Methods used in the prior art for forming CD-guest complexes and recovering the complex and unco plexed guest molecule separately from the reaction mixture have tended to rely on the differential solubility of the complex from the active ingredient (guest molecule) itself. For instance where the active ingredient is relatively insoluble in water the non-complexed active ingredient can be recovered for instance by filtration or by centrifugation. Where the active ingredient is soluble in a solvent in which the reaction is carried out then the complex may have to be recovered by causing it to precipitate, for instance by cooling the product mixture and/or by carrying out the reaction with the CD at or near saturation. There are limitations to this approach in that it may not be suitable for producing complexes which are very soluble from actives which are also very soluble. According to a first aspect of the invention there are provided liposomes containing in the aqueous phase inside the liposome at least one complex, wherein the complex comprises a molecule non-covalently bound to a receptor. By "in the aqueous phase" it is meant that the complex is free in the aqueous phase and is not covalently attached to any part of the lipid bilayer.

Preferably the non-covalent binding of said molecule to said at least one receptor is effective to modify the interactions of said molecule with other components of the liposome, more preferably effective to modify the solubility properties of said molecule and/or the interaction of said molecule with the lipid bilayer. Two embodiments of the invention in particular are important.

According to the first important embodiment there are provided liposomes containing in the aqueous phase inside the liposome at least one complex, wherein the complex comprises a hydrophobic molecule non-covalently bound to a receptor, and wherein the complex is hydrophilic.

According to the second important embodiment there are provided liposomes containing in the aqueous phase inside the liposome at least one complex, wherein the complex comprises a molecule having a tendency to escape from liposomes, non-covalently bound to a receptor and wherein the complex has a tendency to escape from liposomes which is lower than that of the molecule.

There is also provided the use of such liposomes as a delivery system for drugs or other substances into the body.

Use of the liposome system of the invention allows the solubilisation of hydrophobic substances in the aqueous phase of liposomes and hence facilitates their encapsulation in said liposomes. Thus the benefits obtainable with delivery using liposomes are accessible for hydrophobic substances as well as for hydrophilic substances. These benefits include increase in the half- life of substances in the circulation, compared with the half-life of the cyclodextrin - active substance complexes mentioned above or of the active substances alone, and direction to particular regions of the body. The system of the invention allows the incorporation into liposomes of larger amounts of hydrophobic substances than has previously been possible.

The use of the system of the invention allows the inclusion in the aqueous phase of a single liposome drugs which would normally be incompatible in that environment. The drugs may be rendered inactive or reduced in activity towards each other by their complexation with a receptor or receptors.

With the second important embodiment of the invention the problem of leakage of small molecules from liposomes is greatly reduced. A small molecule is non-covalently bound to a receptor which is not able to escape easily from the liposomes through the lipid bilayer. Thus the molecule is trapped inside the liposome.

Preferred aspects of the invention will now be described in more detail. The liposomes may be unilamellar or multilamellar. They may have mean diameter up to 50μm but preferably have a mean diameter of 200nm or less. The liposomes may be produced from liposome forming material, preferably lipid. Lipids used may be one or more of phosphatidylcholine, cholesterol , phosphatidylglycerol , phosphatidylethanolamine, phosphatidylserine, phosphatidic acid, sphingomyelin or derivatives, for instance distearyl, dirmyristoyl and dipal itoyl derivatives, of these lipids, preferably phosphatidylcholine or cholesterol. We find that the percent entrapment of complex into and stability of the resulting liposomes is influenced by the choice of lipid material used to make the liposomes. In particular, we find that it is advantageous to include cholesterol in the liquid bilayer of liposomes made from phospholipids having low gel-liquid crystallisation temperature (Tc) , such as phosphatidylcholine and dimyristoyl phosphatidylcholine. We also find that the absence of cholesterol is advantageous in liposomes whose lipid bilayer is made from high Tc phospholipids, such as dipalmitoyl phosphatidylcholine. Both systems can result in greater entrapment values. Furthermore, the systems tend to exhibit reduced loss of receptor without drug from the vesicle. It is believed that the respective presence and absence of cholesterol in these environments gives these results due to promotion of bilayer rigidity.

Examples of substances which may be complexed with a receptor and subsequently entrapped in the liposomes are pharmaceuticals, vaccines, genetic materials, enzymes, hormones, vitamins, carbohydrates, proteins/peptides, lipids, organic molecules and inorganic molecules or atoms. More specific examples are anti-tumour and anti-microbial agents, enzymes, hormones, vitamins, metal chelators and genetic material, preferably carbohydrates or proteins/peptides. Specific examples include morphine, indomethacin, Naproxen, Ketoprofen, tin etiopurpurin, pilocarpine, hydrocortisone, oestrogen, progesterone, prostaglandins, cholesterol. The complexed and entrapped molecule of the first important embodiment of the invention is hydrophobic and hence has low water solubility. Examples of such substances are dehydroepiandrosterone (DHEA) , retinoic acid, retinol, chlorambucil, dexamethasone, indomethacin, /3-Tocopherol, Vitamin E, Vitamin D. The second important embodiment of the invention covers the complexation and entrapment of small molecules with a tendency to leak from liposomes. Examples of such substances are melaphalon, vincristine. The liposome systems of the invention are useful for both systemic and topical administration. The receptor used for the first important embodiment of the invention is usefully any substance capable of forming a complex with a hydrophobic molecule in order to modify the apparent solubility properties of the said hydrophobic molecule. Complexation is preferably by engulfment or encapsulation of the hydrophobic molecule. More than one hydrophobic molecule may be bound by one receptor. Conversely more than one receptor may bind one hydrophobic molecule. More than one type of molecule may be bound by a single receptor. Examples of such substances are binding proteins having a hydrophobic pocket into which a small hydrophobic molecule may be received and an external surface which is hydrophilic, hence rendering the binding protein and the complex of the hydrophobic molecule with the binding protein water soluble.

A particularly suitable group of substances for use as the receptor are those which have an annular molecular geometry. The molecules are "bucket-like", having a cavity which can receive molecules of suitable size and character. An example of a substance of this type is cyclopolygalacturonic acid (CPGA) . The CPGA molecule has a ring structure which allows engulfment of lipids or lipid-like materials. Further substances of this type are calixarene based substances. Calixarenes have a "bucket- like" geometry. They are based on phenol molecules joined in a ring by alkyl groups. The benzene rings from the walls of the "bucket". A further suitable class of receptors of this type is the class of cyclodextrins and cyclodextrin derivatives. This class of material is the most preferred for use in the first important embodiment of the invention.

The receptor for use in the second important embodiment of the invention is usually any substance capable of forming a complex with a small molecule with a natural tendency to escape from liposomes said complexation being effective to modify the apparent leakage properties of the molecule. The receptor itself should have a low tendency to escape from liposomes, hence the receptor-drug complex will have a low tendency to escape.

Similar types of receptor as are suitable for the first important embodiment of the invention are also suitable for this second important embodiment. Binding proteins having a hydrophilic pocket can form complexes with small hydrophilic molecules. Binding proteins have a low tendency to leak from liposomes due to their large molecular weight. Thus the complex itself remains within the liposome.

In this second important embodiment of the invention also it is preferred to use as the receptor a substance having annular molecular geometry, in particular a cyclodextrin or cyclodextrin derivative. The system of the invention pre erably includes as the receptor a cyclodextrin or cyclodextrin derivative. This system has advantages over previously tested delivery systems based on administration of cyclodextrin-drug complexes alone. Elimination of cyclodextrins and cyclodextrin complexes by the kidneys and toxicity of cyclodextrins to the kidneys are no longer problems, firstly because the cyclodextrins which might lead to problems are encapsulated inside liposomes and secondly because administered liposome preparations tend not to travel to the kidneys.

Problems of competition with other biomolecules and premature release are greatly reduced since the cyclodextrin-drug complex is encapsulated inside the liposome. Similarly problems of cell membrane solubilisation by cyclodextrin are greatly reduced.

Cyclodextrins, like sugars in general, do not tend to leak out through the lipid bilayer, even when attempts have not been made to decrease the permeability of the bilayer. Thus small molecules with a tendency to leak from the liposomes can be rendered "non-leaky" by inclusion in a cyclodextrin host-guest complex. Natural cyclodextrins may be used, that is α-, β- or 7- cyclodextrin. β-cyclodextrin is the most readily available and lowest-priced of these. Natural cyclodextrins are produced from starch by the action of cyclodextrin glycosyltransferase (CTG) , an amylase-type enzyme usually obtained from cultures of the micro-organism Bacillus Macerans. Natural cyclodextrins are commercially available, for instance from American Maize-Products Company. Other methods of encapsulation are applicable, for instance the microfluidation method, which would not normally be appropriate for the encapsulation of non- ccmplexed water-insoluble substances; detergent dialysis may also be used, as may methods involving solid lipids. Cyclodextrins may be derivatised in various ways. For instance they may be chemically substituted to give methylated cyclodextrins or hydroxypropyl cyclodextrins, which have a much greater water solubility than their corresponding parent cyclodextrins. Methods of substituting cyclodextrins include monosubstitution on primary-hydroxyl side, monosubstitution on secondary- hydroxyl side, disubstitution on adjacent glucopyranose units, disubstitution on alternate glucopyranose units, disubstitution on diametrically opposite glucopyranose units, appending, capping, double capping, duplexing with single bridge and duplexing with double bridge. Derivatisation may be carried out by conventional chemical methods known in the art.

It appears that hydroxypropyl derivatives are less toxic than their natural counterparts and that toxicity of cyclodextrins increases in the order 7 < < β. In particular haemolytic activities of cyclodextrin sulphates and 2-hydroxypropyl-β-cyclodextrin are lower than those of their parent cyclodextrins.

Derivatives which can be used include methyl, ethyl, pentyl, hydroxypropyl, 2-hydroxypropyl, hydroxyethyl, amino, deoxy, glucosyl, maltosyl, heptakis-2,6-dimethyl, O- carboxymethyl-O-ethyl and sulphate derivatives. Derivatisation can be very useful in modifying the properties of cyclodextrins, for instance their solubility, toxicity and inclusion properties.

Soluble polymers and dimers of cyclodextrins or their derivatives may also be used as the receptor. Any of the above mentioned cyclodextrins may be used in dimerised or polymerised form provided such dimerisation or polymerisation is possible. Polymers of j8-cyclodextrin are commercially available, for instance from Cyclolab (Budapest, Hungary) .

Molecular weight may vary from for instance 1,900 to 13,000, generally from around 4,000 to around 9,100. The number of cyclodextrin units present in the polymer may be any number from 2 upwards, and is usually not more than 10 or 15 units. Most usually the cyclodextrin polymers contain from 3 to 8 units on average per polymer molecule. Commercially available 3-cyclodextrin polymers have average molecular weights of for instance 4,000 to 4,500 and around 8,700. Cyclodextrin polymers may be homopolymers of one type of cyclodextrin monomer. Alternatively they may be copolymers or terpolymers of more than one type of cyclodextrin monomer. Alternatively or additionally further comonomers may be present which are not cyclodextrins.

The polymers may be graft copolymers of cyclodextrin onto a polymeric backbone containing suitable groups. This backbone may be for instance polyvinylimidazole. Alternatively random copolymers may be produced, for instance with epichlorhydrin monomer. The presence of epichlorhydrin or other bifunctional agents such as diepoxide, diester and diisocyanate as comonomer can give cross-linking of the cyclodextrin polymer. Random copolymerisation and graft copolymerisation can be promoted by the introduction of polymerisable groups into the cyclodextrin or cyclodextrin derivative. Examples of such groups include the (meth)acryloyl group. Conventional methods of polymerisation may be used for production of homopolymers or copolymers. These include reverse phase emulsion polymerisation and inverse suspension bead polymerisation. Examples of polymers which are useful include those prepared by reacting a mono-substituted 6-o-(3-chloro-2- hydroxypropyl) 3-cyclodextrin derivative with the amine groups of polyvinylimidazole to give a graft copolymer. A further exemplary polymer may be produced by introducing polymerisable methacryloyl groups into α-, β- and 7- cyclodextrins or their hydroxyalkylated derivatives and carrying out radicalic inverse suspension polymerisation of the water-soluble monomers with or without comonomers to give a water-swelling bead polymer of particle size 5-100 μm.

There are various methods known for the production of inclusion complexes. In one method of inclusion by slurry mixing the appropriate amount of guest compound (alone or dissolved in an appropriate solvent) is added to a slurry of cyclodextrin prepared with 0.3 to 3.0 (generally l.o to 2.0) times its own weight of water and mixed thoroughly for 0.5 hour to several hours using homogeniser, mortar etc. In this method the viscosity may increase giving a paste. The complex is appropriately washed and dried. Complexation may also be effected by additions of water-insoluble active, dry or dissolved in a non-aqueous solvent, to an aqueous solution of cyclodextrin (active usually in excess although not always) and mixing, for instance by sonication or stirring. Complexation may be evidenced by the dissolution of the water-insoluble active substance.

Before use the active-cd complex may be separated from uncomplexed components. This may be achieved by filtration or centrifugation, for instance, if the complex is insoluble. Filtration may be by any suitable method, for instance through packed glass wool. For the water-soluble complexes of the present invention chromatographic separation may be used, for instance using a Sephadex G¹⁰ column.

Complexes of cyclodextrins with hydrophilic substances may be formed by a method of saturation in aqueous solution, in which a water-soluble compound is added directly to a saturated aqueous solution of cyclodextrin and the mixture is slowly agitated for from 0.5 hour to several hours to form the complex. The complex precipitates at room temperature or on cooling and is isolated by filtration and drying. This method may be used for water-insoluble compounds, which are first dissolved in the minimum quantity of a suitable solvent, for instance acetone, before addition to the cyclodextrin solution.

The receptor:drug molar ratio is usually 1:1 but complexes may be formed with a higher or lower proportion of receptor.

The stability of the receptor-drug complex is very important. A balance is required between lability, replacement of the active substance by another molecule in vivo and hence premature release of the administered substance and very high stability leading to retarded or incomplete release of the administered substance in vivo. Stability constants within the desired range may be found by a skilled person. If the receptor is a cyclodextrin, stability constants may be modified by selective derivatisation of natural cyclodextrins.

As explained above it is possible to use chromatographic separation for separating complexed receptor and molecules from non-complexed molecule. According to a further aspect of the invention there is provided a process comprising

(a) contacting an annular receptor molecule with a guest molecule in solution to form a complex, and

(b) subjecting a solution of the complex formed in step (a) to gel permeation chromatography to separate the complex from non-complexed guest molecules. This process is especially useful in the preparation of complexes which are suitable for incorporation into the liposomes of the invention.

In the process it is preferred for the product solution from step a) to be passed straight to step b) , although in some circumstances it is preferred for the product solution to be subjected to intermediate purification or recovery steps. Furthermore it is preferred for the product solution to be transferred to the step b) without further additions, although in some cases it may be desirable for additional solvent or other diluent to be added.

In process of this further aspect of the invention the encapsulation of the active ingredient in the first step may be achieved by dissolving the receptor molecule into a solvent and then adding the guest molecule, either as a solid or slurry or as a solution in the same or another solvent. The mixture is then mixed thoroughly for 0.5 hour to several hours until the complex formation has taken place.

The separation step, step b) , is carried out using the usual apparatus and arrangement of gel. Usually the gel is supported in a column. We have found that a particularly suitable column is a Sephadex G column. The receptor:guest molar ratio is usually 1:1 but complexes may be formed with a higher or lower proportion of receptor molecule.

The receptor molecule may be any of the receptors useful in the liposomes of the invention, such as cyclodextrin-based compounds. These include natural cyclodextrin monomer, derivatised cyclodextrin monomer and cyclodextrin-containing polymer.

The guest molecule may be any of the active molecules useful in the liposomes of the invention. The guest molecule may be a hydrophobic molecule. It may be a molecule which has a tendency to escape from liposomes when non-complexed. The invention will be illustrated with reference to specific examples. These illustrate entrapment of the hydrophobic substances when complexed with cyclodextrin and the stability of the resulting liposomes in blood plasma. Entrapment of drv HPgcd inclusion complexes into liposomes Example 1 Entrapment of dehydroepiandrosterone (DHEA) hydroxypropyl-j8-cyclodextrin inclusion complex into the aqueous phase of liposomes. 22.4mg DHEA were dissolved in 2ml CHC1₃ into which 7 x lo⁵ dpm of ³H-labelled DHEA (3H-DHEA) were added. Following evaporation of the organic solvent, the dry DHEA was dissolved with bath sonication in a 1ml aqueous solution of hydroxypropyl β cyclodextrin (HP-3cd) (450mg) also containing 2.3 x 10 dpm of C-labelled HP-/8cd 0.05ml of the solution, containing 1.12mg DHEA and 22.5mg of UP-βcd were used for entrapment in liposomes by the dehydration- rehydration method (dehydration- rehydration vesicles; DRV liposomes) with modifications (unless otherwise stated; see Table 1) as follows: 32μmoles of phospholipid alone or together with 32μmoles of cholesterol were dissolved in 2ml CHC1₃. Following rotary evaporation of the solvent the thin lipid film was suspended in 2ml water at a temperature (Ta) above the gel-liquid crystalline transition temperature (Tc) of the phospholipid. The suspension was then probe sonicated at Ta to produce small unilamellar vesicles (SUV) . These were allowed to stand for 60min at Ta and then mixed with 0.05ml of the HP-jffcd-DHEA complex. The mixture was diluted to 3 or 10ml with H₂0 and freeze-dried overnight. To the freeze-dried material 0.1ml of H₂0 was added at Ta and the sample swirled vigorously and allowed to stand for 30min at Ta. To the suspension, 0.1ml 15mM sodium phosphate buffer containing 0.9% BaCl, pH7.4 (PBS) was added at Ta and again allowed to stand for 30min at Ta. The solution was finally mixed with 0.8ml PBS at Ta and, after 30min at Ta, 8ml PBS was added and it was centrifuged at 35,000g for 30min at 4°C. The pellet consisting of DRV and entrapped HP-?cd- DHEA complex was suspended in 5ml PBS and centrifuged again. The washed pellet was then suspended in 1.0ml PBS. ³H and ^UC radioactivities were assayed in portions of the suspended DRV and supernatants and the HP-j8cd-DHEA complex entrapment estimated (% of complex used) (Table 1) .

Table 1. Entrapment of DHEA/HP-jS cd inclusion complex into Liposomes

PC = egg phosphatidylcholine DSPC = distearoylphosphatidylcholine DMPC = dimyristoylphosphatidylcholine SM = sphingomyelin HPC = hydrogenated phosphatidylcholine PA = phosphatidic acid CHOL = cholesterol.

DHEA-HP-j8cd complex was entrapped in DRV liposomes composed of the lipids shown (phospholipid to cholesterol and phospholipid to phosphatidic acid molar ratios were 1:1 and 1:0.1 respectively). ³H: C ratios approximating 1.0 signify similar entrapment values for the two isotopes, i.e. entrapment of the intact inclusion complex. Note poor entrapment values and/or 3H:1-4C ratios when volume of freeze-dried liposomes is low.

Example 2 Entrapment of carboxyfluorescein (CF)-HP-øcd complex into the aqueous phase of liposomes. 0.2ml of 0.2M CF was mixed with 0.05ml of HP-3cd (22.5mg) into which 1.15 x 10⁵ dpm of ⁴C-labelled HP-/3cd had been previously added. Formation of the CF-HP-βcd inclusion complex was verified by molecular sieve chromatography on a Sephadex G10 column: both CF and C-HP-j8cd eluted in the same fractions (labels both peak at fraction 8) (where free HP-/3cd also eluted when chromatographed as such) , whereas free CF eluted at subsequent fractions peaking at fraction 15. 1.18ml H₂0 containing the above quantities (0.2ml 0.2MCF and 22.5mg HP-3cd) of CF-^UC KP-βcd inclusion complex was used for entrapment in DRV by the procedure described in Example 1. Entrapment values for the inclusion complex were derived from CF and C HP-3cd measurements (Table 2) . Table 2. Entrapment of CF/HP-β-cd inclusion complex into liposomes

Liposomes Temperature of Freeze-dried Complex <entrapped CF/14C preparation volume (ml) (°C) CF% 14C%

2.1 DSPC/CHOL 60°C 10 37.9 36.0 1.05

2.2 PC/CHOL 20°C 10 51.4 36.1 1.42

CF/HP /3-cd complex was entrapped in liposomes of the compositions shown.

Example 3 Entrapment of retinoic acid (RA)-HP-_/8cd inclusion complex into the aqueous phase of liposomes. lOmg retinoic acid were dissolved in 2ml CHC1₃ into which 6.3 x 10⁶ dpm of 3H-labelled RA (3H-RA) were added.

Following evaporation of the solvent, the dry RA was dissolved in 2ml solution of HP-/3cd (400mg) containing 6.5 x 10⁵ dpm of ^uC-HP-/3cd by stirring at 37°C for 3 days. The solution formed (milky suspension) was centrifuged at 100,000g for 1 h. Most of the inclusion complex of RA-HP- jScd was recovered in the supernatant. Formation of the inclusion complex was verified by molecular sieve chromatography on a Sephadex G10 column: the fraction containing ^UC-HP-j8cd (peak at fraction 10) also contained most of the 3H-RA. 0.75mg RA and 30mg HP-/3cd (0.15ml of the RA-HP-Scd complex in the supernatant; see above) was used for entrapment in DRV as in Example 1 (for conditions see Table 30) . Entrapment of RA-HP-3cd complex was estimated on the basis of ³H-RA and ^HC-HP-3cd radioactivity (Table 3) .

Table 3. Entrapment of RA/HP-3-cd inclusion complex into liposomes

Liposomes Temperature of Freeze-dried Complex entrapped 3H/14C preparation volume (ml) (°C) 3H% 14C%

3.1 DSPC 60 10 20.7 26.8 0.78

3.2 DSPC/CHOL 60 10 13.9 15.1 0.92

3H-RA/14C-HP/3cd inclusion complex was entrapped in liposomes of the compositions shown.

Example 4 Entrapment of retinol (R)-HP-3cd inclusion complex into the aqueous phase of liposomes. lO g of retinol were dissolved in 2ml CHC1₃ into which 7.2 x 10° dpm of ³H-labelled R(³H-R) were added. Following evaporation of the solvent, the dry R was dissolved in 2ml of HP-3cd (lOOmg) containing 6.5 x 10⁵ dpm of ^uC-HP-/3cd by stirring at 37°C for 2 days. Verification of inclusion complex formation in the clear solution was carried out by molecular sieve chromatography using Sephadex G10: most of the inclusion complex of H-R- C-HP-/_?cd was recovered in the fraction also containing the C-HP-3cd. 0.5ml of the R-HP-3cd inclusion complex (2.5mg of R and 25mg of HP-/3cd) was used for entrapment into DRV as in Example 1 (for conditions see Table 4) . Entrapment of R-HP-/3cd inclusion complex was estimated on the basis of H-R and ¹⁴C-HP-/Scd radioactivity (Table 4) .

Example 5 Entrapment of DHEA 3-cyclodextrin polymer inclusion complex into the aqueous phase of liposomes. 2.24 mg DHEA mixed with 3.77 x 10⁶ dpm [³H]-DHEA and 19 mg of each of /5-cd polymers 2009 (molecular weight 4,000 to 4,500) and 2010 molecular weight 8,700, each containing 52% of their weight 0-cd, available from Cyclolab (Budapest, Hungary) was mixed and stirred for 2 days at 30°C. Initial molar ratios, based on the 8-cd content (52% w/w) of the polymers, were 1:2 (polymer 2009) and 1:1 (polymer 2010). The solutions formed were filtered through packed glass wool to remove non-solubilised matter. Final molar ratios were not estimated because radio-labelled polymers were unavailable.

Entrapment was carried out into DRV as in Example 1 using 1.0 ml each of the complexes with polymer 2009 and 2010. Again, entrapment ratios were not established because the polymers were not radio-labelled. Table 4. Entrapment of R/HP-/3-cd inclusion complex into liposomes

Liposomes Temperature of Freeze-dried Complex entrapped 3H/14C preparation volume (ml) Ratio (°C) 3H% 14C%

4.1 PC/CHOL 20 10 3.2 5.6 0.57

4.2 DSPC 60 10 17.9 18.3 0.98

4.3 DSPC/CHOL 60 10 18.2 22.9 0.80

H.R/ HP-3-cd inclusion complex was entrapped in liposomes of the composition shown. Note t low entrapment values obtained with PC/CHOL liposomes.

Example 6 Release of RA-HP-3cd inclusion complex from liposomes in the presence of blood plasma. 0.3ml of liposome-entrapped RA-HP-/Scd inclusion complex, radiolabelled with ³H(RA) and ^UC-(HP-Scd) , was mixed with 0.6ml rat blood plasma or 0.6ml PBS and incubated at 37°C for time intervals. 0.4ml and 0.5ml samples were taken at 2 min and 60 min respectively from the incubated samples and tested for inclusion complex release by centrifugation at 100,000g for 25 min of the sample diluted to 5ml with water. Results in Table 5 (release values) , show that most of the inclusion complex ( 3H and 14C radi.oacti.vi.ti.es) was recovered in the liposomal pellet for both PBS and rat blood plasma incubation media, indicating little to modest release of the complex of liposomes. Table 5. Release of RA/HPS-cd complex from liposomes in the presence of PBS or plasma at 37°C

Liposomes Incubation % release medium

2 min 60 min

3H% 14C% 3H% 14C%

5.1 DSPC PBS 12.6 4.7 17.5 5.6

5.2 DSPC Plasma 14.7 4.3 26.6 3.2

5.3 DSPC/CHOL PBS 28.4 8.8 29.2 8.5

5.4 DSPC/CHOL Plasma 32.2 5.4 35.1 11.9

DRV liposomes containing RA/HP-3-cd complex (see Example 3) were incubated at 37°C in the presence of PBS (control) or rat blood plasma. Values denote released ³H and ¹⁴C as % of amounts incubated. Note the small additional effect of plasma when compared to that of PBS. Results indicate that, in terms of RA release, DSPC/CHOL liposomes are less stable than those made of DSPC.

As release of HPS-cd from liposomes is less pronounced, it appears that some RA escapes from the HP/3-cd and behaves independently. RA escape appears greater with DSPC/CHOL liposomes, possibly because some of the RA within HP8-cd is replaced by cholesterol.

Example 7 Release of R-HP-/3cd inclusion complex from liposomes in the presence of PBS or blood plasma 0.3ml of liposome-entrapped R-HP-8cd inclusion complex, radiolabelled with ³H(R) and ^UC-(HP-3cd) was mixed with 0.6ml rat blood plasma or 0.6ml PBS and incubated at 37°C for time intervals. 0.4ml and 0.5ml samples were taken at 2 min and 60 min respectively from the incubated samples and tested for inclusion complex release by centrifugation at 100,000g for 25 min. Results in Table 6 (release values) , show that most of the inclusion complex ( 3H and 14C radio activities) was recovered in the liposomal pellet for both PBS and rat blood plasma incubation media indicating little to modest release of the complex from liposomes. Table 6. Release of R/HP/S-cd complex from liposomes in the presence of PBS or plasma at 37°C.

Liposomes Incubation % release medium

2 min 60 min

3H% 14C% 3H% 14C%

6.1 DSPC PBS 12.6 1.1 15.3 1.8

6.2 DSPC Plasma 12.6 0.7 26.8 2.5

6.3 DSPC/CHOL PBS 17.1 5.2 25.9 3.8

6.4 DSPC/CHOL Plasma 25.0 2.5 35.6 2.0

DRV liposomes containing R/HP/S-cd complex (see Example 4) were incubated at 37°C in the presence of PBS or rat blood plasma. Values denote released 3H and 14C as % of amounts incubated. Note the small additional effect of plasma when compared to that of PBS. Results indicate that, in terms of R release, DSPC/CHOL liposomes are less stable than those made of DSPC. For other comments see Example 6. These examples indicate that hydrophobic or water- insoluble substances may be entrapped in the aqueous phase of liposomes by means of the formation of a cyclodextrin inclusion complex. The resulting liposomes are stable in blood plasma. (Usually the majority of liposomes have left the blood plasma 5 minutes after entry) . It is expected that it would be possible to achieve greater than 90% complex retention at 12 minutes using these systems. Example 8 Release of DXM/HP-/3cd inclusion complex from liposomes in the presence of PBS or blood plasma.

³H-DXM/¹⁴C-HP-/3cd complexes were prepared using methods as described above. The complexes were entrapped in DRV liposomes comprised of DSPC only, using methods as described above. The liposome preparations were incubated in the presence of PBS or rat plasma at 37°C for 2 and 60 minutes. At the end of the incubation period samples were centrifuged to sediment liposome. Released radioactivity (³H and ^HC) was measured in the supernatant.

Table 7 Stability of dexamethasone/HP/fcd complex entrapped in DRV liposomes

Preparation % Released ± SD Incubated

2 mins 60 mins

³H ^uc ³H ^MC

Liposomal Complex in PBS 7.7 ± 3.0 7.3 ± 3.9 10.1 ± 1.6 2.4 ± 0.1

Liposomal Complex in Rat Plasma 3.4 ± 0.8 0.9 ± 0.1 11.0 ± 3.2 1.4 ± 0.2

There is less drug released compared with data from tests on R and RA complexes with HP-/3cd. Drug release appears to depend at least in part on the characteristics of the drug, as well as on the environment.

Example 9 Behaviour of drug-cd complexes in vivo Inclusion complexes were prepared of DHEA, retinol (R) and dexamethasone (DXM) , each labelled with H, were prepared with ^uC-(HP-Scd) as described above. The molar ratios of DHEA to HP/Scd and DXM to HP-Scd were 1:1 and the molar ratio of R to KP-βcd was 1:3. Some of each complex was entrapped in DRV liposomes as described above. The liposomes were made of DSPC only.

Rats in groups of four were injected intravenously with 1.0 ml of PBS containing free or liposomal complex. In each case the dose of free DHEA/HP-/3cd comprised 5 mg DHEA and 2 mg HP-/Scd. The dose of liposomal DHEA/HP-/3cd comprised 2 mg DHEA and 0.45 mg HP-Scd. The dose of free R/HP-/3cd comprised 0.2 mg R and 4.5 mg P-βcd. Each dose of liposomal R/HP-|8cd complex comprised 0.05 mg R and 0.6 mg HP-jScd. Each dose of free DXM/HP-Scd complex comprised 0.2 mg DXM and 0.6 mg HP-jScd. Each dose of liposomal DXM/HP-/3cd complex comprised 0.1 mg DXM and 0.2 mg HP-/3cd. Some animals were killed at 30 minutes after injection. Others were killed at 24 hours after injection. Content of drug ( H content) and KP-βcd ( C content) were measured in the blood plasma, liver, spleen and kidneys of the animals.

Results are shown in Tables 8 to 11 below.

Table 8 Recovery of Η-druq and ^MC-HP Jed in total blood plasma of rats after intravenous injection

% of Injected Dose Recovered ± SD

Preparation

30 min 24 hrs Injected

3„ 14_c 3_H/14_C 3„ 14c 3„/14_c (no S .D. given) (no S . D . )

Free DHEA/HPβcd complex 6.2 ± 0.5 8.3 ± 0.7 0.7 0.7 ± 0.2 o. i i o.o 7.0

Liposomal DHEA/HPβcd 4.7 ± 1.3 0.4 ± 0.2 11.75 0.9 ± 0.4 0.6 ± 0.4 1. 5 complex

Free R/HPβcd complex 16.3 ± 1.7 14.9 ± 1.3 1.1 1.7 ± 0.2 0.05 ± 0.01 34.0

Liposomal R/HPβcd complex 1.3 ± 0.2 1.0 ± 1.0 1.3 0.9 ± 0.1 0.1 ± 0.0 9.0

10 Free DXM/HPβcd complex 9.9 ± 0.4 7.2 ± 0.8 1.4 0.2 ± 0.0 0. 1 ± 0.0 2 .0

Liposomal DXM/HPβcd 1.0 ± 0. 1 0.25 ± 0.0 4.0 0.2 ± 0.1 0. 1 ± 0.0 2.0 complex

Table 9 Recovery of ³H-druα and ^uC-HPβcd from the liver of rats after intravenous injection

% of Injected Dose Recovered ± SD

Preparation

30 min 24 hrs Injected

3„ 14c 3_H/14_C 3„ 14_c 3„/14_c (no S.D. (no S.D.) given)

Free DHEA/HPβcd complex 19.4 ± 2.2 1.0 ± 0.1 19.4 0.9 ± 0.2 0.3 ± 0.1 3.0

Liposomal DHEA/HPβcd 32.6 ± 2.3 82.5 ± 3.6 0.4 0.8 ± 0.9 69.6 ± 7.9 0.01 complex

Free R/HPβcd complex 19.2 ± 1.5 2.0 t 0.4 9.6 1.6 ± 0.2 0.25 1 0.06 6.4

Liposomal R/HPβcd 47.2 ± 3.2 67.3 ± 8.4 0.7 26.8 ± 3.5 77.7 ± 7.9 0.3

10 complex

Free DXM/HPβcd complex 25.6 ± 2.1 1.0 ± 0.2 25.6 0.6 ± 0.5 0.4 ± 0.2 1.5

Liposomal DXM/HPβcd 53.9 ± 1.5 60.5 ± 2.1 0.9 51.0 ± 23.4 75.6 ± 18.8 0.7 complex

Table 10 Recovery of ³H-drug and "c-HPβcd from the spleen of rate after intravenous injection

% of Injected Dose Recovered ± SD

Preparation

30 min 24 hrs Injected

3„ 1 c 3„/14_c 3„ 14_c 3_H/14_C (no S .D. (no S .D . ) given)

Free DHEA/HPβcd complex 0.1 ± 0.0 0.1 ± 0.0 1.0 0.0 0.0 1.0

Liposomal DHEA/HPβcd 2.0 ± 0.2 9.5 ± 1.9 0.2 0.2 ± 0.2 7.0 t 1.8 0.03 complex

Free R/HPβcd complex 1.2 ± 0.6 0.3 ± 0.1 4.0 0.2 ± 0.1 0.1 ± 0.0 2.0

Liposomal R/HPβcd 8.2 ± 1.7 12.9 1 2.0 0.6 4.5 ± 1.2 10.8 ± 0.6 0.4

10 complex

Free DXM/HPβcd complex 0.5 ± 0.2 0.2 ± 0.0 2.5 0.1 ± 0.0 0.1 ± 0.0 1.0

Liposomal DXM/HPβcd 5.5 ± 0.8 6.9 ± 0.8 0.8 5.2 ± 2.1 9.3 ± 2.7 0.6 complex

Table 11 Recovery of ³H-drug and '"c-HPβcd from the kidney of rats after intravenous injection

10

Example 10 Recovery of drug/HP/Scd inclusion complexes in urine of rats Rats were injected as in Example 8. The animals which were killed at 24 hours were kept individually in metabolic cages with facility for 24 hour urine collection. Levels of ³H (drug) and ^UC (HP-?cd) were measured. In the urine collected over the 24 hour period. Results are shown in Table 11 below.

3 14

Table 12 Recovery of H-drug and C-HP3cd in the urine of rats after intravenous injectio

% of Injected Dose Recovered ± SD

Preparation Injected ³H c ³H/^UC (no S.D. given)

Free DHEA/HP/3cd complex 26.4 ± 8.2 73.9 ± 2.9 0.35

Liposomal DHEA/HPjScd complex 25.1 ± 2.1 6.4 ± 1.2 3.9

Free R/HPj8cd complex 40.4 ± 5.0 81.7 ± 16.0 0.5

Liposomal R/HPj8cd complex 26.1 ± 5.9 20.8 ± 0.0 1.25

Free DXM/HPβcd complex 48.2 ± 15.9 95.2 ± 7.2 0.5

10 Liposomal DXM/HPScd complex 22.7 ± 7.7 9.9 ± 6.1 2.3

Examples 9 and 10 show that liposome-entrapped inclusion complexes have a reduced tendency to be excreted rapidly in the urine compared with free inclusion complexes and have a greater tendency to travel to the tissues. Table 8 shows that the non-entrapped complexes have a greater tendency to remain in the circulation than the liposome-entrapped complexes. The results shown in Tables 9 and 10 illustrate that the liposomal complexes when removed from the circulation have a tendency to travel to the tissues, where the drugs will of course be required. This is illustrated by their presence in the liver and spleen in greater amounts at 30 minutes than the free inclusion complexes. It will also be seen that levels of drug remain high in the tissues even after 24 hours when the drug has been injected in liposome-entrapped complex form. The results shown in Tables 11 and 12 illustrate that free inclusion complexes when removed from the circulation have a far greater tendency to be excreted rapidly by the kidneys. Levels of free complex in the kidneys at 30 minutes were greater than levels of liposomal complex in the kidneys at that stage. Over a 24 hour period consistently larger amounts of drug and HP-j8cd are excreted in the urine when the drug is injected in the free rather than liposomal form.

Claims

1. Liposomes containing in the aqueous phase inside the liposome at least one complex, wherein the complex comprises at least one molecule non-covalently bound to at least one receptor.

2. Liposomes according to claim 1, wherein the non- covalent binding of said molecule to said at least one receptor is effective to modify the interactions of said molecule with other components of the liposomes.

3. Liposomes according to claim l or claim 2, wherein the non-covalent binding of said molecule to said at least one receptor is effective to modify the solubility properties of said molecule and/or the interaction of said molecule with the lipid bilayer.

4. Liposomes according to any preceding claim wherein the complex comprises a molecule, having a tendency to escape from liposomes, non-covalently bound to a receptor and wherein the complex has a tendency to escape from liposomes which is lower than that of the molecule alone.

5. Liposomes according to any preceding claim, wherein the complex comprises a hydrophobic molecule non-covalently bound to a receptor, and wherein the complex is hydrophilic.

6. Liposomes according to any preceding claim, wherein the receptor has annular molecular geometry such that it provides a cavity which is capable of receiving said at least one molecule.

7. Liposomes according to claim 6, wherein the receptor comprises a cyclodextrin moiety or cyclodextrin derived moiety.

8. Liposomes according to claim 7, wherein the receptor is a dimer or polymer comprising one or more cyclodextrins or cyclodextrin derivatives.

9. Liposomes according to any preceding claim, wherein the molecule is a pharmaceutical, a vaccine, a genetic material, an enzyme, a hormone, a vitamin, a metal chelator, an anti-tumour agent or an anti-microbial agent.

10. A preparation suitable for topical administration comprising liposomes according to any of claims 1 to 8.

11. A pharmaceutical preparation suitable for intravenous injection comprising liposomes according to any preceding claim.

12. Use of liposomes of any of claims 1 to 9 as a drug delivery system.

13. A process comprising

(a) contacting an annular receptor molecule with a guest molecule in solution to form a complex and

(b) subjecting a solution of the complex formed in step (a) to gel permeation chromatography to separate the complex from non-complexed guest molecule.

14. A process according to claim 13 in which the product solution from step (a) is passed directly to step (b) .

15. A process according to claim 13 or claim 14 wherein the receptor molecule comprises a cyclodextrin moiety or cyclodextrin-derived moiety.

16. A process according to claim 15 wherein the receptor is a polymer of one or more cyclodextrins or cyclodextrin derivatives.

17. A process according to any of claims 13 to 16 wherein the guest molecule is a pharmaceutical, a vaccine, a genetic material, an enzyme, a hormone, a vitamin, a metal chelator, an anti-tumour agent or an anti-microbial agent.

18. A process according to any of claims 13 to 17 in which the gel is supported in a Sephadex G column.