WO1996017594A1

WO1996017594A1 - Sequestration agents

Info

Publication number: WO1996017594A1
Application number: PCT/GB1995/002908
Authority: WO
Inventors: Roger Randal Charles New; Christopher John Kirby
Original assignee: Cortecs Limited
Priority date: 1994-12-09
Filing date: 1995-12-08
Publication date: 1996-06-13
Also published as: IL116311A0; AU4183496A; FI972411A; EP0796086A1; CN1169114A; CA2207274A1; NO972607D0; NO972607L; MX9704272A; FI972411A0; JPH10510207A; GB9424901D0; ZA9510459B

Abstract

The invention provides the use of an agent to reduce direct interaction between a hydrophobic phase and a hydrophilic phase in which the hydrophobic phase is dispersed. Methods for preparing single phase hydrophobic preparations comprising a hydrophilic species wherein such an agent is added are also provided.

Description

SEQUESTRATION AGENTS

The present invention relates to the use of certain compounds aiding retention of hydrophilic molecules, solubilised in a hydrophobic phase in which they would not normally be soluble, in the hydrophobic phase when said hydrophobic phase is dispersed in a hydrophilic phase. In particular the present invention relates to the use of such agents for aiding retention of hydrophilic macromolecules in a hydrophobic phase in which they would not normally be soluble.

For many applications, e.g in the pharmaceutical sciences, in food technology or the cosmetics industry, work with proteins and similar macromolecules presents problems because their hydrophilicity and high degree of polarity limit the extent to which they can interact with or incorporate into lipid phases. Many natural systems employ lipidic barriers (eg skin, cell membranes) to prevent access of hydrophilic molecules to internal compartments; the ability to disperse proteins in lipidic vehicles would open up a new route to introduction of these macromolecules into biological systems, whereby the lipid medium containing the protein can integrate with the hydrophobic constituents of barriers, instead of being excluded by them.

Dispersion of hydrophilic substances in oil phase rather than aqueous media confers other benefits in terms of increasing their stability with respect to temperature- mediated denaturation, hydrolysis, light sensitivity etc. Oils can be chosen which remain fluid over a wider temperature range than aqueous solutions, or that have a higher viscosity, resulting in greater protection against physical damage. In mixed-phase systems, incorporation of hydrophobic substances in oil can limit mutually harmful interactions - eg oxidation - with other agents, both within the oil and in the aqueous phase.

There are examples of formulations containing both macromolecules and oil and one such example is disclosed in EP-A-0366277. The formulation disclosed in this document is an emulsion having both a hydrophobic and a hydrophilic phase, wherein the hydrophobic phase contains chylomicra or chylomicron-forming lipids. However, the macromoiecule is dissolved in the hydrophilic phase not in the hydrophobic phase.

EP-A-0521994 also relates to a composition suitable for the oral delivery of macromolecules which comprises a biologically active material in association with lecithin or a compound capable of acting as a precursor for lecithin in vivo. All of the compositions exemplified are formulations which comprise a hydrophilic and a lipophilic phase. Once again, in this prior art document, the macromoiecule is initially dissolved in the hydrophilic phase rather than in the lipophilic phase.

Although the formulations mentioned above do contain both macromolecules and oils, it is significant that in all cases the macromoiecule is initially dissolved in the hydrophilic rather than in the lipophilic phase. Attempts to form true solutions of macromolecules in oils have met with limited success.

Okahata et al (J. Chem. Soc. Chem. Commun.. 1988, 1392- 1394) disclose a process for solubilising proteins in a hydrophobic solvent. However, in the array of protein surrounded by amphiphile molecules produced by that method the authors stated that the amphiphile molecules reacted with the protein in the liquid medium by hydrogen bonding or via an electrostatic interaction to form a solid precipitate.

UK patent application No. 9323588.5 discloses a process by which a hydrophilic species can be solubilised in a hydrophobic solvent in which it would not normally be soluble. The process relies on the surprising discovery that if a hydrophilic species is mixed with an amphiphile under certain conditions, the resultant composition will be readily soluble in lipophilic solvents such as oils.

However, one potential problem with the product of such a process relates to its use in the production of an emulsion, by dispersion of the single phase hydrophobic preparation in a hydrophilic phase, for instance water. In such a dispersion there will be a tendency, at least in some circumstances, for the solubilised hydrophilic species to "leak" into the hydrophilic phase thus reversing the solubilisation process. There thus exists a need to reduce this effect and ensure that the hydrophilic species remains in the hydrophobic phase even when such a hydrophobic phase is itself subsequently dispersed in a hydrophilic phase.

It has now been surprisingly found that certain compounds can reduce the degree of direct interaction between a hydrophilic species solubilised in a hydrophobic phase and a hydrophilic phase in which the hydrophobic phase is subsequently dispersed, e.g. as an emulsion.

Thus, in a first aspect the present invention provides the use of an agent to reduce the direct interaction between a hydrophilic species solubilised in a hydrophobic phase and a hydrophilic phase in which the hydrophobic phase is dispersed.

In the present invention the term "agent" relates to any species which is capable of reducing direct interaction between a hydrophilic species solubilised in a hydrophobic phase in which it would not normally be soluble, when the hydrophobic phase is itself dispersed in a hydrophilic phase, for example to form an emulsion, and the hydrophilic phase.

In a second aspect, the present invention provides a single phase hydrophobic preparation comprising a hydrophilic species solubilised in a hydrophobic solvent in which it would not normally be soluble and in addition an agent which reduces direct interaction between the hydrophobic species and a hydrophilic phase in which the hydrophobic preparation is dispersed.

It appears that although the hydrophilic species may be able to come into contact with individual molecules of the hydrophilic phase, it cannot come into contact with the bulk of the hydrophilic phase and hence this results in reduced "leakage" of the hydrophilic species into the hydrophilic phase.

In the present invention the term "hydrophilic species" relates to any species which is generally soluble in aqueous solvents but insoluble in hydrophobic solvents. Suitably, the agent can be: (i) an acidic lipid, for example, cholesterol hemisuccinate (Chems) or phosphatidic acid; or

(ii) an emulsion stabiliser which cannot penetrate the hydrophobic phase, e.g. a compound such as casein.

In a third aspect the invention also provides an agent for use in reducing direct interaction between a hydrophilic species, solubilised in a hydrophobic solvent in which it would not normally be soluble, and a hydrophilic phase in which the hydrophobic phase is dispersed, eg as an emulsion.

Suitably the agents described herein are used in a solubilisation process as described in UK patent application No. 9323588.5. Thus, in a further aspect the invention provides a process for the preparation of a single phase hydrophobic preparation comprising a hydrophilic species, in a hydrophobic solvent, the process comprising:

(i) associating the hydrophilic species with an amphiphile in a liquid medium such that, in the liquid medium, there is no chemical interaction between the amphiphile and the hydrophilic species;

(ii) removing the liquid medium to leave an array of amphiphile molecules with their hydrophilic head groups orientated towards the hydrophilic species; and

(iii) providing a hydrophobic solvent around the hydrophilic species/amphiphile array; wherein an agent which reduces the direct interaction between the hydrophilic species and a hydrophilic phase in which the hydrophobic phase is dispersed is added at one or more of the stages.

Preferably, in this method the agent is added with the amphiphile in stage (i) and is preferably an emulsion stabiliser which cannot penetrate the hydrophobic phase, e.g. a compound such as cholesterol hemisuccinate (Chems) or phosphatidic acid (PA) .

In the context of the present invention, the term "chemical interaction" relates to an interaction such as a covalent or ionic bond or a hydrogen bond. It is not intended to include van der aals forces or other interactions of that order of magnitude.

In another aspect, the present invention provides a process for dispersing a single phase hydrophobic preparation, comprising a hydrophilic species in a hydrophobic solvent, in a hydrophilic phase, which comprises the step of adding to the hydrophilic phase an agent which reduces direct interaction between the hydrophilic species and the hydrophilic phase.

In this method, the agent is preferably an emulsion stabiliser which cannot penetrate the hydrophobic phase, e.g. a compound such as casein.

A wide variety of macromolecules can suitably be solubilised using the processes of the present invention. In general, the macromolecular compound will be hydrophilic or will at least have hydrophilic regions since there is usually little difficulty in solubilising a hydrophobic macromoiecule in oily solutions. Examples of suitable macromolecules include proteins and glycoproteins, oligo and polynucleic acids, for example DNA and RNA, polysaccharides and supramolecular assemblies of any of these including, in some cases, whole cells, organelles or viruses (whole or parts thereof) . It may also be convenient to co-solubilise a small molecule such as a vitamin in association with a macromoiecule, particularly a polysaccharide such as a cyclodextrin. Small molecules such as vitamin B12 may also be chemically conjugated with macromolecules and may thus be included in the compositions.

In particular, when the macromoiecule to be stabilised is a protein or polypeptide, the agent is preferably an acidic lipid.

Examples of particular proteins which may be successfully solubilised by the method of the present invention include insulin, calcitonin, haemoglobin, cytochrome C, horseradish peroxidase, aprotinin, mushroom tyrosinase, erythropoietin, somatotropin, growth hormone, growth hormone releasing factor, galanin, urokinase, Factor IX, tissue plasminogen activator, superoxide dismutase, catalase, peroxidase, ferritin, interferon, Factor VIII, melanin and fragments thereof (all of the above proteins can be from any suitable source) . Other macromolecules may be used are FITC-labelled dextran and RNA extract from Torulla yeast.

In addition to macromolecules, the process of the present invention is of use in solubilising smaller organic molecules. Examples of small organic molecules include glucose, ascorbic acid, carboxyfluorescin and many pharmaceutical agents, for example anti-cancer agents, but, of course, the process could equally be applied to other small organic molecules, for example other vitamins or pharmaceutically or biologically active agents. In addition, molecules such as calcium chloride and sodium phosphate can also be solubilised using the process of the invention. Indeed, the present invention would be particularly advantageous for pharmaceutically and biologically active agents since the use of non aqueous solutions may enable the route by which the molecule enters the body to be varied, for example to increase bioavailability.

Another type of species which may be included in the hydrophobic compositions of the invention is an inorganic material such as a small inorganic molecule or a colloidal substance, for example a colloidal metal. The process of the present invention enables some of the properties of a colloidal metal such as colloidal gold, palladium, platinum or rhodium, to be retained even in hydrophobic solvents in which the particles would, under normal circumstances, aggregate. This could be particularly useful for catalysis of reactions carried out in organic solvents.

There are numerous amphiphiles which may be used in the present invention and zwitterionic amphiphiles such as phospholipids are among those which have been found to be especially suitable. Phospholipids having a phosphatidyl choline head group have been used with particular success and examples of such phospholipids include phosphatidyl choline (PC) itself, lyso-phosphatidyl choline (lyso-PC) , sphingo yelin, derivatives of any of these, for example hexadecylphosphocholine or amphiphilic polymers containing phosphoryl choline and halogenated amphiphiles, e.g. fluorinated phospholipids. In the present application, the terms phosphatidyl choline (PC) and lecithin are used interchangeably. Suitable natural lecithins may be derived from any convenient source, for example egg and, in particular, soya. In most cases, it is preferable to select an amphiphile which is chemically similar to the chosen hydrophobic solvent and this is discussed in greater detail below.

The fact that the present inventors have found zwitterionic amphiphiles such as phospholipids to be particularly suitable for use in the process is a further indication of the significant differences between the present invention and the method of Okahata et al. Significantly, the authors of that prior art document concluded that anionic and zwitterionic lipids were completely unsuitable for use in their method and stated that they obtained zero yield of their complex using these lipids.

The hydrophobic solvent of choice will depend on the purpose for which the composition is intended, on the type of species to be solubilised and on the amphiphile. Suitable solvents include non-polar oils such as mineral oil, squalane and squalene, long chain fatty acids with unsaturated fatty acids such as oleic and linoleic acids being preferred, alcohols, particularly medium chain alcohols such as octanol and branched long chain alcohols such as phytol, isoprenoids, e.g. nerol and geraniol, terpineol, monoglycerides such as glycerol monooleate

(GMO) , other esters, e.g. ethyl acetate, amyl acetate and bornyl acetate, diglycerides and triglycerides, particularly medium chain triglycerides and mixtures thereof, halogenated analogues of any of the above including halogenated oils, e.g. long chain fluorocarbons or iodinated triglycerides, e.g. lipidiol.

Optimum results are generally obtained when the hydrophobic solvent and the amphiphile are appropriately matched. For example, with a solvent such as oleic acid, lyso-PC is a more effective amphiphile than PC, whereas the converse is true when the hydrophobic solvent is a triglyceride.

In addition, in some cases it has been found to be advantageous to add a quantity of the amphiphile to the hydrophobic solvent before it is brought into contact with the hydrophilic species/amphiphile array. This ensures that the amphiphile molecules are not stripped away from their positions around the hydrophilic species because of the high affinity of the amphiphile for the hydrophobic solvent.

It is very much preferred that the preparations of the invention are optically clear and this can be monitored by measuring turbidity at visual wave lengths and, in some cases, by checking for sedimentation over a period of time.

The orientation of amphiphile molecules into an array with their hydrophilic head groups facing the moieties of a hydrophilic species can be achieved in several ways and examples of particularly suitable methods are discussed in more detail below.

In a first method, which has a similar starting point to the method described by Kirby et al. (Biotechnology. November 1984, 979-984, and Liposome Technolocrv. Volume I, pages 19-27, Gregoriadis, Ed, CMC Press, Inc., Boca Raton, Florida, USA) a hydrophilic species is mixed with a dispersion of an amphiphile in a hydrophilic solvent, such that the amphiphile molecules form an assembly in which the hydrophilic head groups face outwards towards the hydrophilic phase which contains the hydrophilic species. The hydrophilic solvent is then removed to leave a dry composition in which the hydrophilic head groups of the amphiphile molecules are orientated towards the hydrophilic species.

In the method described by Okahata et al. a solution of a protein was also mixed with a dispersion of an amphiphile in water. However, significantly, the authors of that paper believed that it was necessary to obtain a precipitate which would then be soluble in hydrophobic solvents. Since many of the preferred amphiphiles of the present invention do not form such a precipitate, Okahata et al concluded that they would be of no use. In the process of the present invention, no precipitate is required and, indeed, it is generally thought to be undesirable to allow the formation of a precipitate since this results in a reduced yield of the required product.

In this first method, it is preferred that the hydrophilic solvent is water although other polar solvents may be used.

The form taken by the amphiphile assembly may be micelles, unilamellar vesicles, preferably small unilamellar vesicles which are generally understood to have a diameter of about 25 nm, multilamellar vesicles or tubular structures, for example cochleate cylinders, hexagonal phase, cubic phase or myelin type structures. The form adopted will depend upon the amphiphile which is used and, for example, amphiphiles such as phosphatidyl choline (PC) tend to form small unilamellar vesicles whereas lyso-phosphatidyl choline forms micelles. However, in all of these structures, the hydrophobic tails of the amphiphile molecules face inwards towards the centre of the structure while the hydrophilic head groups face outwards towards the solvent in which the hydrophilic species is dispersed.

The weight ratio of amphiphil :hydrophilic species will generally be in the region of from 1:1 to 100:1, preferably from 2:1 to 20:1 and most preferably about 8:1 for PC and 4:1 for lyso-PC.

These ratios are preferred ratios only and, in particular, it should be pointed out that the upper limit is set by economic considerations which mean that it is preferable to use the minimum possible amount of amphiphile. The lower limit is somewhat more critical and it is likely that ratios of 2:1 or below would only be applicable in cases where the hydrophilic species has a significant hydrophobic portion or is exceptionally large.

Good performance is obtained when the solvent is removed quickly and a convenient method for the removal of the solvent is lyophilisation, although other methods can be used.

In some cases, it may be helpful to include salts in the hydrophilic solution, particularly if the hydrophilic species is a macromolecular compound such as a large protein. However, because the presence of larger amounts of inorganic salts tend to give rise to the formation of crystals and, hence, to a cloudy solution, it may be preferred that organic salts are used rather than inorganic salts such as sodium chloride. Ammonium acetate is especially suitable for this purpose since it has the additional advantage that it is easily removed by freeze drying.

A second method for the preparation of a composition containing an array of amphiphiles with their head groups pointing towards the moieties of the hydrophilic species is to co-solubilise the hydrophilic species and the amphiphile in a common solvent followed by removal of the solvent.

The product of the process of the invention is new and therefore, in a further aspect of the invention there is provided a single phase hydrophobic preparation comprising a hydrophilic species in a hydrophobic solvent obtainable by the process of the invention.

It may also be desirable to include other constituents in the single phase hydrophobic preparation in addition to the hydrophilic species. This is often particularly appropriate when the hydrophilic species is a macromoiecule and, in that case, the preparation may include, for example, bile salts, vitamins or other small molecules which bind to or are otherwise associated with the macromolecules.

Although some macromolecule/amphiphile arrays were disclosed by Kirby et al. supra, the arrays disclosed were all intermediates in the formation of liposomes and, as discussed above, there has been no previous interest in non-liposomal or hydrophobic compositions comprising this type of entity. Therefore, the arrays of the present invention in which the amphiphile is one which does not form small unilamellar vesicles and would therefore not be expected to form liposomes are new.

One advantage of the preparations of the present invention is that they are effectively anhydrous and therefore more stable to hydrolysis. In the case of proteins, they are also stable to freeze-thawing and have greater stability at high temperatures, probably because water must be present in order for the protein to unfold and become denatured. This means that they may be expected to have a much longer shelf life than aqueous preparations of the hydrophilic species.

The solutions of the present invention are extremely versatile and have many applications. They may either be used alone, but preferably they are combined with an aqueous phase to form an emulsion or similar two phase composition which forms yet a further aspect of the invention.

In this aspect of the invention there is provided a two phase composition comprising a hydrophilic phase and a hydrophobic phase, the hydrophobic phase comprising a preparation of a hydrophilic species in a lipophilic solvent obtainable by a process as described herein.

Generally, in this type of composition, the hydrophobic phase will be dispersed in the hydrophilic phase.

The two phase compositions may be emulsions which may either be transient or stable, depending on the purpose for which they are required.

The average size of the emulsion particles will depend on the exact nature of both the hydrophobic and the aqueous phases. However, it may be in the region of 2 μm

Dispersion of the hydrophobic preparation in the aqueous phase can be achieved by mixing, for example either by vigourous vortexing for a short time for example about 10 to 60 seconds, usually about 15 seconds, or by gentle mixing for several hours, for example using an orbital shaker.

In another aspect the present invention provides a process for preparing a dispersion, eg an emulsion, of a hydrophobic phase in which is solubilised a hydrophilic species which comprises dispersing the hydrophobic phase in a hydrophilic phase to which has been added an agent which reduces the direct interaction between the hydrophilic species when so solubilised and the hydrophilic phase in which the hydrophobic phase is dispersed.

The invention will now be described by reference to the following examples, which should not be construed as in any way limiting the inveniton.

EXAMPLE 1

A 0.0015M borate buffer was prepared by dissolving 60mg of sodium tetraborate in 100 ml of distilled water, and adjusting the pH to 8.00. 5 mg of BAPNA was weighed into a B9 glass screw-capped vial and dissolved in 3 ml of methanol. 10 mg of trypsin was weighed into a 15 ml plastic centrifuge and 10 ml borate buffer added with vortexing. The suspension was mixed on a roller mixer, then undissolved material spun down, and the supernatant decanted.

Dilutions of aprotinin (50 μl/well) were dispensed along the rows of the microplate between 0 and 30 μg/ml concentration.

BAPNA solution from above was diluted 20-fold by adding 1 ml to 20 ml of buffer and 100 μl of BAPNA working solution was then introduced into each well and mixed thoroughly. The plate was incubated with shaking at 37°C for forty minutes and then read on a plate reader at 405 nm.

After plotting optical density due to substrate conversion against aprotinin concentration in the well, an inflection is observed at the concentration at which the aprotinin is just sufficient to neutralise the activity of the trypsin. The position of this inflection moves according to the quantity of additional aprotinin introduced into the wells in the test sample, and this concentration can be inferred by comparison with standards. An indication of the proportion of aprotinin released from the oil and accessible to the aqueous phase can thus be obtained.

Aprotinin was solubilised in Miglyol 818 by lyophilising a mixture of 100 μl of soya phosphatidyl choline dispersion (100 mg/ml in distilled water) , sonicated as per the protocol in Example 4, and 25 μl of aprotinin solution (20 mg/ml in distilled water) , followed by addition of 100 μl of Miglyol 818. The concentration of aprotinin was 5 mg/ml in oil . A control solution was prepared as above, in which aprotinin was omitted. Aqueous dispersions of these oils were prepared by 5 vortexing 10 μl of each oil with 1ml of borate buffer for ten seconds. The final concentration of aprotinin in these secondary dispersions was 0 and 50 μg/ml. The dispersions were diluted two-fold and added to the wells of a microplate as described in the method above, where

10 dilutions of 0, 5, 10, 15 and 25 μg/ml were employed. Normalised optical densities are reported in the table below, and in the accompanying graph. Comparison with a control standard of 12.5 μg/ml indicates that at least 50% of the aprotinin is released from the oil into the

15 aqueous phase.

Nature of oil +/- aprot Aprotinin Concentration

0 5 10 15 20 25

- /PC/M818 0.373 0.358 0.343 0.3 0.055 0

Aprot/PC/M818 0.269 -0.004 0.05 0.058 -0.03 0

12.5 μg/ml aprotinin 0.337 0.299 0.135 -0.008 -0.017 0

Aprot/PC/M818 2-fold diln 0 0..333322 0.264 0.201 -0.036 -0.055 0

EXAMPLE 2

Aprotinin was solubilised in Miglyol 818 as described in the example above, except that the phospholipid 5_ dispersion contained 10% by weight of phosphatidic acid in addition to phosphatidyl choline. The dispersions were tested neat, and compared with control standards of 25 μg/ml and 12.5 μg/ml. Comparison with these standards indicates that no more than 25% of the aprotinin is 0 released into the aqueous phase. Aprotinin Concentration 20 15 10 (μg/ml)

Buffer 234 0.26 0 .273 0.283 0 .277

25 μg/ml aprotinin 0 0.003 0 .005 0.068 0, .167 12.5 μg/ml aprotinin 0 -0.009 0, .202 0.258 0, .258 PC:PA/M818 - 0 099 0.182 0. .164 0.194 0. .216 PC:PA/M818 - 50 μg/ml 0 0.059 0. ,182 0.219 0. ,245

EXAMPLE 3

25 Aprotinin was solubilised in Miglyol 818 as described in the examples above, except that the phospholipid dispersion contained 10% by weight of cholesterol hemisuccinate (Chems) in addition to phosphatidyl choline. The dispersions were tested neat, and compared

30 with control standards of 25, 12.5 and 6.25 μg/ml.

Comparison with these standards indicates that no more than 12.5% of the aprotinin is released into the aqueous phase .

60 minute readings Apoprotinin Concentration (μg/ml)

Test samples 0 6 11 15 18 20 22 24 25 30

0 μg/ l 0.272 0.192 0

25 μg/ml 0.265 0.102 0.002 0.008 0.015 0.015 0.008 0.001 0

12.5 μg/ml 0.267 0.266 0.241 0.044 0.007

6.25 μg/ml 0.285 0.269 0.195 0.236 -0.003

PC:Chβms/M81β/- 0.157 0.165 0.156 0.165 0.152 0.167 0.166 0.114 0.138 0

PC: Chems /M818/ 0.192 0.213 0.195 0.22 0.203 0.197 0.14 0.143 0.062 0 λprot (50 μg/ml)

EXAMPLE 4

An aqueous dispersion of soy phosphatidyl choline (soy PC) was prepared, containing lOOmg/g of suspension, flushed thoroughly with nitrogen, and sonicated at an amplitude of 8 microns peak to peak. Each aliquot was subjected to a total sonication time of 4 minutes, in pulses of 30 seconds interspersed by cooling for 30 seconds in an ice slurry bath. The resulting opalescent dispersion of small unilamellar vesicles (SUV) was then centrifuged for 15 minutes to remove particles of titanium.

Colloidal gold sol was prepared as follows. 15μl of 25mM potassium carbonate, 15μl of 1% tannic acid and 50mg of trisodium citrate were made up to a total weight of 22.5g with distilled water and 10ml was transferred to a 25ml stoppered glass conical flask (A) and heated to 60 C in a water bath. 25mg of gold chloride trihydrate was made up to 250mg with distilled water and 50μl of the resulting solution was added to 40ml of distilled water in a 50ml stoppered glass conical flask (B) and heated to 60 C in the same water bath. The contents of flask A were mixed with those of flask B, and heating maintained for

75 minutes during which time a deep red colloidal gold sol was formed. After cooling to room temperature, a 10ml portion was stabilized by mixing with 2mg of bovine serum albumin.

lml of the stabilised gold sol was mixed with 0.6ml of SUV,freeze-dried overnight and the resulting lyophilate dispersed by vortexing with 300mg of Miglyol 818. Within 1 hour, a clear red dispersion of colloidal gold in Miglyol had been formed. Three, 50mg aliquots of this dispersion were added to small glass vials and then 500mg of water, phosphate-buffered glucose solution (300mM glucose containing ImM sodium phosphate, pH 7.4) and SUV were added to the separate vials. The mixtures were emulsified by vortexing for 10 seconds and then observed.

Within 1 hour, creaming of the emulsions had started to occur, and after standing overnight, this had occurred to a substantial extent. In all cases, pink colouration of the lower aqueous phase was observed indicating release of a proportion of the colloidal gold from the oil phase. However, the intensity of colour retention in the upper, oil emulsion phase was significantly higher, and that in the lower aqueous phase correspondingly lower, in the preparation emulsified in the presence of SUV. Thus the presence of the phospholipid SUV dispersion has apparentlyXserved to reduce loss of colloidal gold from the oil phase.

EXAMPLE 5

lml of a 1% solution of insulin (containing 2% acetic acid to aid dissolution) was mixed with lμCi of I¹²⁵- labelled insulin, followed by 3g of cholesterol hemisuccinate-containing SUV prepared as in Example 3. The mixture was freeze-dried and the resulting lyophilate dispersed with 3g of Miglyol 818, mixing for 4 hours on an orbital shaker to produce a clear dispersion of radiolabelled insulin in oil. Two aliquots of 200mg of dispersion, were each mixed with 800mg of phosphate- buffered saline (PBS) , and emulsified by vortexing for 10 seconds and 30 seconds respectively. Each of the resulting o/w emulsions was diluted with a further 9ml of PBS and then centrifuged for 40 minutes at 80000 g to break it down into its component fractions. The surface layer of oil phase was carefully transferred to a vial for counting gamma radioactivity due to residual I¹²⁵- labelled insulin. The supernatent, containing any released I¹²⁵-labelled insulin, was transferred to a separate vessel, leaving behind any pellet that might have been formed. The pellet, representative portions of supernatent and the residual oil fraction, were all counted for radioactivity, making appropriate corrections for any contamination of the oil fraction with supernatant.

Of the 2 emulsions, the one vortexed for 30 seconds showed 45.4% of the radiolabel retained within the oil phase and 3.0% associated with the centrifugal pellet (assumed to be liposomal in nature) , while the corresponding figures for the 10 second vortexed emulsion were 43.3 and 1.3% respectively. In contrast, in 2 separate experiments where the SUV used to prepare the oil dispersion were composed of pure soy PC rather than soy PC/Chems, oil retentions of label were 31% and 28%, with a further 1% in the pellet in each case. Thus it appears that inclusion of Chems in the amphiphile systems used to prepare the protein in oil dispersions, leads to increased retention of protein within the oil when the latter is emulsified to form a w/o secondary dispersions.

EXAMPLE 6

I-¹²⁵-labelled insulin was prepared and incorporated into oil phases as described in Example 3 , but using the compositions listed below. Preparation no. Amphiphile System Oil phase

1 Soy PC SUV Miglyol 818 2 Soy PC/Phosphatidic Miglyol 818 Acid (PA) SUV *

* PC/PA SUV were prepared as in Example 2.

One lOOmg aliquot of Preparation 1 and one of Preparation 2 were weighed into 10ml centrifuge tubes. To each aliquot was added lml of PBS. All were vortexed thoroughly for 10 seconds and then the latter were diluted with 10ml of PBS. The emulsions were then centrifuged and fractionated into their component phases as described in Example 2.

Preparation no. Dispersant % Retention in each fraction

Oil Pellet Supernatant

1 PBS 27.6 1.5 70.9 2 PBS 26.1 20.9 53.0

The pellet obtained is thought to be composed of oil containing a high proportion of phospholipid. In the preparation containing phosphatidic acid, considerably less of the radiolabelled insulin is released into the aqueous supernatant than with the preparation containing just phosphatidyl choline alone.

EXAMPLE 7

125 labelled insulin was prepared and incorporated into oil phases as described in Example 2, but using the composition listed below.

Preparation no. Amphiphile System Oil phase 1 Soy PC SUV Oleic acid

* PC/PA SUV were prepared as in Example 2.

One lOOmg aliquot of Preparation 1 was weighed into a 10ml centrifuge tube, and lml of 0.5% casein solution was added and vortexed thoroughly for 10 seconds. The contents of the tube were then diluted with 10ml of the 0.5% casein. The emulsion was then centrifuged and fractionated into its component phases as described in Example 2.

Preparation no. Dispersant % Retention in each fraction

Oil Pellet Supernatant

1 0.5% casein 63.3 4.7 32

As can be seen, a significant proportion of the radiolabelled insulin is retained within the oil phase.

EXAMPLE 8

0.8ml of 25mM calcium chloride was mixed with 0.8ml of soy PC SUV prepared as in Example l, freeze-dried overnight and the resulting lyophilate mixed with 0.5g of Miglyol 818. After standing overnight, a completely clear dispersion had formed. A portion of the dispersion was coloured by mixing 150 mg together with approximately 0.17 mg of Sudan 4 dye to form a clear, deep-red solution. 2ml of 1% sodium alginate was transferred to a glass test-tube and vortexed briefly while, at the same time, adding the coloured oil dispersion from a pasteur pipette. The resulting opaque emulsion was centrifuged at 500 g for 5 minutes and the upper, pink, oil-rich phase was decanted from the underlying clear aqueous phase and examined under the light microscope. All of the oil was now seen to be present as numerous, small, discrete droplets which showed no signs of coalescence. A proportion of the oil droplets appeared to be surrounded by an outer wall which was presumed to be due to interfacial complexation of the alginate by calcium ions released from the oil droplets. After standing for 12 days, there was still no signs of breakdown of the emulsion, either micro- or macroscopically.

Claims

1. The use of an agent to reduce the direct interaction between a hydrophilic species solubilised in a hydrophobic phase, and a hydrophilic phase in which the hydrophobic phase is dispersed.

2. The use as claimed in claim 1 wherein the agent is an acidic lipid or an emulsion stabiliser which cannot penetrate the hydrophobic phase.

3. The use as claimed in claim 2 wherein the agent is cholesterol hemisuccinate (Chems) , phosphatidic acid (PA) or casein.

4. A single phase hydrophobic preparation comprising a hydrophilic species solubilised in a hydrophobic solvent in which it would not normally be soluble, and an agent which reduces direct interaction between the hydrophilic species and a hydrophilic phase in which the hydrophobic preparation is dispersed.

5. A process for the preparation of a single phase hydrophobic preparation comprising a hydrophilic species, in a hydrophobic solvent, the process comprising:

(iii) providing a hydrophilic solvent around the hydrophilic species/amphiphile array;

wherein an agent which reduces the direct interaction between the hydrophilic species and a hydrophilic phase in which the hydrophobic preparation is dispersed is added at one or all of the stages.

6. A process as claimed in claim 5 wherein the agent is an acidic lipid.

7. A process as claimed in claim 6 wherein the agent is cholesterol hemisuccinate (Chems) or phosphatidic acid

(PA) .

8. A process for dispersing a single phase hydrophobic preparation, comprising a hydrophilic species in a hydrophobic solvent, in a hydrophilic phase, which comprises the step of adding to the hydrophilic phase an agent which reduces the direct interaction between the hydrophilic species and the hydrophilic phase.

9. A process as claimed in claim 8 wherein the agent is an emulsion stabiliser which cannot penetrate the hydrophobic phase.

10. A process as claimed in claim 9 wherein the agent is casein.

11. The use as claimed in any one of claims 1 to 3 , a preparation as claimed in claim 4, or a process as claimed in any one of claims 5 to 10, wherein the hydrophilic species is selected from the group consisting of proteins, glycoproteins, oligo and polynucleic acids, polysaccharides and supramolecular assemblies of any of these.

12. The use, a preparation or the process all as claimed in claim 11, wherein the hydrophilic species insulin, calcitonin, haemoglobin, cytochrome C, horseradish peroxidase, aprotinin, mushroom tyrosinase, erythropoietin, somatotropin, growth hormone, growth hormone releasing factor, galanin, urokinase, Factor IX, tissue plasminogen activator, superoxide dismutase, catalase, peroxidase, ferritin, interferon, Factor VIII, melanin, fragments of any of the above, DNA, RNA, FITC- labelled Dextran or vitamin B12.

13. A process as claimed in any one of claims 5 to 7 wherein the amphiphile is a phospholipid.

14. A process as claimed in claim 13 wherein the phospholipid has a phosphatidyl choline head group.

15. A process as claimed in claim 14 wherein the phospholipid is phosphatidyl choline (PC) , lyso- phosphatidyl choline (lyso-PC) , sphingomyelin, a derivative of one of the above such as hexadecyl phosphocholine or an amphiphile polymer containing phosphoryl choline.

16. A process as claimed in any one of claims 5 to 7 or any one of claims 13 to 15 wherein the hydrophobic solvent comprises a long chain fatty acid, a medium chain alcohol, a branched long chain alcohol, a monoglyceride, diglyceride, medium chain triglyceride or long chain triglyceride.

17. A process as claimed in any one of claims 5 to 7 or any one of claims 13 to 16 wherein the amphiphile comprises PC and the hydrophobic solvent is a triglyceride or wherein the amphiphile comprises lyso-PC and the hydrophobic solvent is oleic acid.

18. A process as claimed in any one of claims 5 to 7 or any one of claims 13 to 16 wherein the hydrophilic species/amphiphile array is formed by mixing the hydrophilic species with a dispersion of an amphiphile in a hydrophilic solvent and removing the hydrophilic solvent.

19. A process as claimed in claim 18 wherein the hydrophilic solvent is water.

20. A process as claimed in claim 18 or claim 19, wherein the amphiphile assembly comprises micelles, unilamellar vesicles, for example unilamellar vesicles, multilamellar vesicles or a tubular structure such as cochleate cylinders, hexagonal phase, cubic phase or myelin type structures.

21. A process as claimed in any one of claims 18 to 20, wherein the hydrophilic solvent is removed by lyophilisation.

22. A process as claimed in any one of claims 5 to 7 or any one of claims 13 to 17 wherein the hydrophilic species/amphiphile array is formed by co-solubilising the hydrophilic species and the amphiphile in a common solvent and subsequently removing the common solvent.

23. A process as claimed in any one of claims 5 to 7 or any one of claims 13 to 17 wherein the hydrophilic species/amphiphile array is formed by emulsifying a solution of the amphiphile in a hydrophobic solvent with a solution of the hydrophilic species in a hydrophilic solvent to give an emulsion and removing the hydrophobic solvent.

24. A process as claimed in claim 22 or claim 23 wherein the weight ratio of amphiphile to hydrophilic species is from about 1:1 to 50:1.

25. A process as claimed in claim 23 wherein the emulsion is water-in-oil emulsion.

26. A process as claimed in claim 23 or claim 24 wherein the hydrophobic solvent is a low boiling point organic solvent such as diethyl ether.

27. A single phase hydrophobic preparation of a hydrophilic species in a hydrophobic solvent, obtainable by a process as claimed in any one of claims 5 to 7 or any one of claims 13 to 25.

28. A two phase composition comprising a hydrophilic phase and a hydrophobic phase, wherein the hydrophobic phase comprises a preparation as claimed in claim 4 or claim 26.

29. A composition as claimed in claim 27 wherein the hydrophobic phase is dispersed in a continuous hydrophilic phase.

30. A composition as claimed in claim 27 or claim 28 which is an emulsion.

31. The use of a preparation as claimed in claim 4 or claim 12, a preparation as claimed in claim 27, or of a composition as claimed in any one of claims 28 to 30 in the oral delivery of a hydrophilic species.