The present invention relates to a method of treating or preventing inflammatory and other conditions associated with undesirably elevated levels of cytokines, particularly proinflammatory cytokines and other cytokines which when elevated induce an injurious effect.

Cytokines are inducible, soluble proteins produced by a variety of cells involved in immune inflammatory responses including T cells, B cells and macrophages. There are many different cytokines, including the families of the interleukins, colony stimulating factors, chemokines, interferons and tumour necrosis factors. These are produced in response to a wide range of stimuli including injury, infection, inflammation and tumour states and serve a variety of functions, including as immunoregulators, growth factors and differentiation factors.

The immune system comprises an extremely complex network with numerous interactions between cytokines and host cells, other cytokines and regulatory molecules. Each cell type of the immune system produces a distinct repertoire of cytokines typical of that cell type, with some overlap in production between the various cell types since most cytokines are produced by more than one cell type. Cytokines act pleiotropically, i.e. they act on a variety of target cells within the host, where their effects are exerted by means of high affinity membrane receptors, to produce an effect dependent upon the particular target cell . This network therefore ensures that a single cytokine interacts with more than one cell type, that individual cytokines have multiple biological activities, that several cytokines can act as factors

mediating a common effect (often in a seemingly synergistic manner) and that the effects of cytokines can be various and widespread (as cytokine receptors are present on multiple cells) . Cytokines may be induced extremely rapidly and, when induced at normal levels, provide benefit to the host by mediating the metabolic and biochemical changes in response to challenges, including infection, injury and inflammation which are essential to the body's defence against such challenges and to the healing process. However, the production of cytokines to undesirably elevated levels can mediate some of the most lethal and widespread chronically debilitating diseases known to man, including sepsis, cachexia, rheumatoid arthritis and asthma.

Of particular relevance in this regard are the group of cytokines intrinsically linked to the generation and maintenance of inflammation, the so- called proinflammatory cytokines, which includes tumour necrosis factor alpha (TNF-α) , interleukin one alpha and beta (IL-lα and IL-lβ) , interleukin six (IL-6) , interleukin eight (IL-8) , and interleukin-12 (IL-12) as well as other cytokines which act as promoters of allergic inflammation, including interleukin four (IL-4) and interleukin five (IL-5) . Overproduction of any of these cytokines is associated with both acute and chronic inflammatory pathologies.

The pathophysiological effects of these proinflammatory cytokines, and the cytokine promoters of allergic inflammation, are often characterised by the exaggeration of the response they normally elicit, as would be expected by overproduction. This may lead to hyper-production of endothelial factors causing excess vasodilation, leucocyte migration, endothelial hyporesponsiveness, smooth muscle constriction, diapedesis and further release of soluble factors or stimulation of cells, e.g. macrophages and mast cells,

capable of further inflammatory responses.

A variety of disease states are associated with undesirably elevated cytokine levels, and often these are exacerbated by synergy between the various cytokines, which lead to a diversity of pathophysiological effects. Examples of such diseases include:

- Septic shock, a particular problem encountered with hospitalised patients, often those with life- threatening diseases, believed to be caused by infection with bacteria, where the cell wall component, the endotoxin LPS, is responsible for initiating the disease.

The destructive inflammatory response associated with infection is the result of stimulation by LPS of TNF-α and the other proinflammatory cytokines. TNF-α has been shown to be a primary causative agent of sepsis, (Glauser et al. , Clinical Infectious Diseases 18, S205-216 (1994)). The acute or chronic overproduction of TNF-α has widespread effects on various systems; it can result in disseminated intravascular coagulation, an increase in vascular adhesion molecules, and hence efflux of cells, and an increase in prostaglandins and leukotrienes, which can effect the vascular tone and platelet aggregation.

TNF-α mediates the progression of bacterial infection to systemic inflammatory response syndrome (SIRS: also known as sepsis) , which can result from other insults such as trauma, through severe sepsis (SIRS plus raised temperature, tachycardia, lactic acidosis, perfusion abnormalities and oliguria) to septic shock (severe sepsis plus hypotension) . This plethora of effects takes the form of a cascade system that has TNF-α as a fulcrum with other proinflammatory cytokines such as IL- 1, IL-6 and IL-8 acting in association in a cumulative process. In particular, IL-1 and IL-6 play critical roles in mediating SIRS (Glauser et al. , (1994);

Dinarelli, Eur. Cytokine Netw. , 5.(6), 517-531 (1994) ; Borden & Chin, J. Lab. Clin. Med., I2£(6) , 824-829 (1994) ) .

- The acute phase response (APR) . This comprises predetermined, well-orchestrated local and systemic reactions resulting from infections, trauma, neoplasms or other disorders which put a stress on homeostasis (Borden & Chin, J. Lab. Clin. Med 123., 824-9 (1994)) . Local reactions at the site of injury include aggregation of platelets and clot formation, dilation and leakage of blood vessels and accumulation and activation of granulocytes and monocytes/macrophages that release a number of cytokines including IL-1, TNF-α and IL-6 in particular. Systemic reactions include fever, leukocytosis, activation of complement and clotting cascades, and changes in concentrations of "acute phase" plasma proteins generated by the liver. Undesirably elevated levels of IL-6 have been noted in patients with SIRS, burns injuries, trauma and after organ transplantation, giving an early indication of organ rejection in the latter case.

- Arthritic diseases, osteoarthritis (OA) and rheumatoid arthritis (RA) , which are controlled by a complex cytokine network, in which the proinflammatory cytokines IL-1, TNF-α and IL-6 are of major importance (Sipe et al . , Mediators of Inflammation, 3., 243-256 (1994) ) . All three cytokines have been detected at undesirably elevated levels in the synovial fluid, synovium and cartilage from RA patients, whilst IL-1 and IL-6 have been found in the latter tissues from OA patients. IL-8 is also important in RA. IL-15 has also been associated with RA.

Indeed certain therapies which reduce TNF-α or IL-1 levels have been shown to have a positive benefit to the swollen joints and overall pain in rheumatoid arthritis.

- Allergic inflammation.

Many allergies are associated with overproduction

of IL-4 and IL-5 by Th2-type T helper cells, which are involved in the pathogenesis of allergic inflammation in humans (Kumar & Busse, 1995) ; Romagnami, Current Opin. Immunology, 6. ( 6 ) , 838-846 (1994)) and which are associated with humoral immunity.

Both allergic and non-allergic asthma are associated with undesirably elevated levels of IL-4 and IL-5 (Kumar Sc Busse, Scientific American: Science & Medicine, March/April, 38-47 (1995)) . IL-6 and TNF-α are also involved in the pathogenesis of asthma.

Other common allergies fundamentally linked to the overproduction of the "allergic" inflammatory promoters IL-4 and IL-5 and rapid mediators such as histamine include, but are not limited to, pollen allergies (hay fever) , non-specific allergic rhinitis, house dust mite allergies and animal dander allergies.

- Cachexia, a condition of severe weight loss and tissue wasting which is associated with chronic invasive diseases such as cancer and parasitic diseases as well as HIV infection, characterised by continued lipid and protein catabolism out of balance with nutritional requirement and food intake in which TNF-α, IL-1 and IL- 6 are all implicated as humoral mediators.

- Certain types of cancer, eg. B cell neoplasias such as multiple myeloma, which have been shown to be associated with increased levels of IL-6 (Akira et al. , Adv. Immunology, ΞA. 1-63, (1993)) .

Other inflammatory diseases include ulcerative colitis, inflammatory bowel disease, with which IL-12 has been associated, and atherosclerosis.

Despite what might be perceived as an extremely diverse range of symptoms exhibited by patients suffering from these diseases, they are all characterised by the underlying involvement of these cytokines.

Current therapeutic strategies are based on targeting the effect of individual cytokines (or second

messengers) by seeking to block or interfere with the interaction of the cytokine with its receptor using, for example, a binding partner to one or other component such as an antibody to the receptor or the cytokine. Such strategies are extremely limiting in concentrating only on a single mediator and are often short-lived. Furthermore, removing or inactivating just one agent in the complex network of interacting pathways between cytokines could simply act as a switch between pathways resulting in a different cytokine mediating the inflammatory response.

In addition, treatment regimes which rely on administration of antibodies to humans are not problem- free and often the undesirable side effects of such treatments outweigh any benefit.

There is accordingly a need for a new therapeutic method capable of pleiotropically affecting multiple cytokines. The present invention provides such a method. Thus viewed from one aspect, the present invention provides a method for combating conditions associated with undesirably elevated levels of one or more cytokines which when elevated induce an injurious effect comprising administering to a subject an effective amount of non-ionic surfactant vesicles (NISV) .

In a related aspect, the present invention provides the use of NISV in the manufacture of an agent for use in combating conditions associated with undesirably elevated levels of one or more cytokines which when elevated induce an injurious effect.

Such conditions include septic shock and severe sepsis, cachexia as well as inflammatory conditions including rheumatoid arthritis, asthma as well as topical allergic conditions such as eczema and psoriasis, bacterial endotoxaemia, SIRS, ulcerative colitis, inflammatory bowel disease, Crohn's disease, atherosclerosis, osteoporosis, diabetes, leukaemia,

multiple myeloma, cystic fibrosis, pulmonary fibrosis, acute meningococcal infections, alcoholic hepatitis, various allergies, systemic lupus erythrymatosus, multiple sclerosis and treatment of these disease constitute particular aspects of the invention.

As used herein, the term 'combating' includes both prophylaxis and therapy.

We have found that NISV are able to reduce independently the levels of the proinflammatory cytokines TNF-α, IL-1, IL-6 and IL-8, as well as the cytokine mediators of allergic inflammation, IL-4 and IL-5. This therapeutic activity at the cellular level translates to an observable benefit at the physiological level, in terms of management of those diseases which are associated with raised levels of these cytokines. In this context, the undesirably elevated cytokine levels refers to levels greater than observed in 'normal' subjects which do not exhibit any signs of inflammation or responses characteristic thereof. We have demonstrated the reduction in levels of these cytokines in vitro and ex vivo in both human and animal cells. We have demonstrated also a positive therapeutic effect in animal models of cachexia.

In one aspect, the invention relates to the use of NISV for reducing the level of TNF-α produced by cells in an inflammatory response. This is of particular importance in the therapy and prophylaxis of septic shock, and of cachexia. The invention also relates to the use of NISV for reducing the levels of one or more of IL-1, IL-6, IL-8, IL-12 and/or IL-4 and/or IL-5 produced by cells involved in inflammatory and immune processes.

NISV are known, for example as carriers e.g. for drugs and also as components of cosmetics. With antigen entrapped, such vesicles are also known as potent immunological adjuvants, as described in International patent applications numbers W093/19781 and W095/09651.

We are not aware however of any prior recognition of the therapeutic potential of the vesicles themselves, and particularly as an immunomodulating agent in the absence of an antigen. Thus, the present invention may be distinguished from prior art therapies involving NISV in that the NISV may be used as the sole therapeutically active agent, rather than a carrier. The NISV are active without any other biologically active agent being entrapped or associated with them. In other words they may be used "empty" .

The vesicles used according to the invention may comprise non-ionic surfactants alone and may optionally include other components such as molecules which have the ability to transport or facilitate the transport of fats, fatty acids and lipids across mucosal membranes for example bile salts as described in WO95/09651.

Indeed compositions based on vesicles comprising such molecules with these transporting capabilities are ne .

Examples of such vesicles are described in the above-mentioned WO95/09651 which also describes preparative processes.

The invention is applicable to all types of NISV vesicular structures, including unilamellar vesicles

(comprised of a single bilayer) , multilamellar vesicles (comprised of more than one bilayer) and multivesicular vesicles, which may comprise unilamellar and/or multilamellar vesicles. Methods for preparing NISV are well known in the art and described in the literature, including for example in W093/19781 and WO95/09651.

The non-ionic surfactant used to form the NISV may be any pharmacologically acceptable material with the appropriate surface active properties. Preferred examples of such materials are glycerol esters. Such glycerol esters may comprise one or two higher aliphatic

acyl groups e.g. containing at least ten carbon atoms in each acyl moiety. Glycerol monoesters are preferred, particularly those containing a C ₁₂ -C ₂₀ alkanoyl or alkenoyl moiety, for example caproyl, lauroyl, myristoyl, palmitoyl, oleyl or stearoyl. A particularly preferred surfactant is 1-monopalmitoyl glycerol.

Ether-linked surfactants may also be used as the non-ionic surfactant of which the NISV according to the invention are comprised. Preferred examples of such materials are ether-linked surfactants based on glycerol or a glycol preferably a lower aliphatic glycol of up to 4 carbon atoms, most preferably ethylene glycol. Surfactants based on such glycols may comprise more than one glycol unit, preferably up to 5 glycol units and more preferably 2 or 3 glycol units, for example diglycol cetyl ether or polyoxyethylene-3-lauryl ether. Glycol or glycerol monoethers are preferred, particularly those containing a C ₁₂ -C ₂ o alkyl or alkenyl moiety, for example capryl, lauryl, myristyl, cetyl, oleyl or stearyl.

The ethylene oxide condensation products usable in thiε invention include those disclosed in WO88/06882, i.e. polyoxyethylene higher aliphatic ether and amine surfactants. Particularly preferred ether-linked surfactants are 1-monocetyl glycerol ether and diglycol cetyl ether. However, for use in the present invention it is necessary to select pharmacologically acceptable materials, preferably those which are readily biodegradable in the mammalian system. For this reason, we prefer the aforementioned glycerol esters for preparing vesicles to be administered by injection, either subcutaneous, intramuscular, intradermal intraperitoneal or intra-articular, or via the mucosal route such as by oral, nasal, bronchial, urogenital, rectal, intrapulmonary or ocular administration, oral administration being particularly preferred, especially in the case of vesicles containing bile salts or

equivalent compounds.

In one aspect, the invention provides a pharmaceutical composition for combating conditions associated with undesirably elevated levels of cytokines which when elevated induce an injurious effect comprising NISV together with a pharmaceutically acceptable carrier or excipient in a form suitable for intrapulmonary or intra-articular administration.

In the case of the aforementioned vesicles which additionally comprise molecules which have the ability to transport or facilitate the transport of fats, fatty acids and lipids across membranes, (hereinafter "transport enhancers") a variety of such molecules may be used such as those described in WO95/09561. Cholesterol derivatives in which the C ²³ carbon atom of the side chain carries a carboxylic acid, and derivatives thereof are particularly preferred.

Amongst such derivatives are the "bile acids" cholic acid and chenodeoxycholic acid, their conjugation products with glycine or taurine such as glycocholic and taurocholic acid, and derivatives including deoxycholic and ursodeoxycholic acid, and salts of each of these acids.

Also preferred as "transport enchancers" are acyloxylated amino acids, preferably acyl carnitines and salts thereof particularly those containing C ₆ . ₂ o alkanoyl or alkenoyl moieties, such as palmitoyl carnitine. As used herein, the term acyloxylated amino acid is intended to cover primary, secondary and tertiary amino acids as well as α, β &γ amino acids.

Acylcarnitines are examples of acyloxylated γ amino acids.

These vesicles may, naturally, comprise more than one type of "transport enhancer" in addition to the non- ionic surfactants for example one (or more) different bile salts and one (or more) acylcarnitines.

For effective vesicle formation, the non-ionic

surfactant may need to be admixed with an appropriate hydrophobic material of higher molecular mass capable of forming a bi-layer, particularly a steroid, e.g. a sterol such as cholesterol . The presence of material such as cholesterol assists in forming the bi-layer on which the physical properties of the vesicle depend.

The NISV may also incorporate a charge-producing amphiphile, to cause the NISV to take on a charge. Acidic materials such as higher alkanoic and alkenoic acids (e.g. palmitic acid, oleic acid) ; or other compounds containing acidic groups, e.g. phosphates such as dialkyl, preferably di (higher alkyl) , phosphates, e.g. dicetyl phosphate or phosphatidic acid or sulphate monoesters such as higher alkyl sulphates, e.g. cetyl sulphate, may all be used for this purpose.

The steroid may e.g. comprise 20-120 percent by weight of the non-ionic surfactant, preferably 60-100 percent. The amphiphilic material producing a charge may e.g. comprise 1-30 percent by weight of the non- ionic surfactant.

The charge-producing amphiphilic material stabilises the structure of the vesicles and provides effective dispersion.

The non-ionic surfactant and membrane-forming hydrophobic material may be converted to NISV by hydration in the presence of shearing forces. Apparatus to apply such shearing forces is well known, suitable equipment being mentioned e.g. in WO88/06882. Sonication and ultra-sonication are also effective means to form NISV or to alter their particle size.

Essentially, the reagents to form the vesicles are brought into contact conveniently at temperatures in the range of 80 to 150°C to 'melt' the surfactants, and conveniently in the presence of a suitable medium such as a buffer or aqueous solution and mixed to ensure a homogenous suspension. Such mixing may be by standard well known techniques including vortexing or the use of

homogenisation apparatus, such as is common in the pharmaceutical industry.

An effective method for the production of NISV is that disclosed by Collins et al . , J. Pharm. Pharmocol . 12, 53 (1990) . This involves melting a mixture of the NIS, steroid, and amphiphile (if used) and hydrating with vigorous mixing in the presence of aqueous buffer. The suspension may then be extruded several times through microporous polycarbonate membranes at an elevated temperature sufficient to maintain the NISV- forming mixture in a molten condition.

It is also possible to form NISV by rotary film evaporation from an organic solvent, e.g. a hydrocarbon or chlorinated hydrocarbon solvent such as chloroform. The resulting thin film may then be hydrated optionally in phosphate-buffered saline in the presence of any material to be entrapped and optionally another surfactant (Russell and Alexander, J. Immunol. 140 , 1274 (1988) ) . Where vesicles of specific size are required, these may be prepared by sequential extrusion through polycarbonate filters as described in Nayar et al. , (Biochem. Biophys. Acta 986 200-206 (1989)) or by other methods known in the art such as mixing and homogenising the reagents under particular conditions, for example, for different times and at different speeds, which may be appropriately determined for each system and size desired.

Without wishing to be bound by theory, it is believed that the therapeutic agents of the invention act at the level of the cells responsible for production of the relevant cytokines, typically macrophages and monocytes, as well as other immune cells. Down¬ regulation of cytokine production at this level represents a considerable advance over current therapies which seek to block the activity of individual cytokines, particularly since according to the invention

a single agent may be used to down regulate more than one cytokine.

Viewed from another aspect, the present invention provides a method of combating cachexia comprising administering to a subject suffering from or liable to cachexia an effective amount of NISV.

In a related aspect the present invention provides the use of NISV in the manufacture of an agent for use in the treatment or prophylaxis of cachexia. In a further aspect the present invention provides a method of treating septic shock comprising administering to a subject suffering from or liable to septic shock an effective amount of NISV.

In a related aspect, the present invention provides the use of NISV in the manufacture of an agent for use in the treatment or prophylaxis of septic shock.

Viewed from another aspect, the present invention provides a method of combating arthritis comprising administering to a subject suffering from or liable to arthritis an effective amount of NISV.

In a related aspect the present invention provides the use of NISV in the manufacture of an agent for use in the treatment or prophylaxis of arthritis.

Viewed from another aspect, the present invention provides a method of combating asthma comprising administering to a subject suffering from or liable to asthma an effective amount of NISV.

In a related aspect the present invention provides the use of NISV in the manufacture of an agent for use in the treatment or prophylaxis of asthma.

"Treatment" and "treating" as used herein refer both to the alleviation of existing morbid conditions and to the prophylactic prevention thereof by timely administration of NISV before proinflammatory cytokine levels have become dangerously elevated. The onset of such conditions can often be foreseen, but there has hitherto been no effective method of prophylaxis

available.

The therapeutic agents according to the invention may be administered by all conventional methods including parenterally (e.g. intraperitoneally, subcutaneously, intramuscularly, intradermally or intravenously) , topically (e.g. as a cream to the skin, intra-articularly, mucosally (e.g. orally, nasally, vaginally, rectally and via the intra-ocular route) or by intrapulmonary delivery for example by means of devices designed to deliver the agents directly into the lungs and bronchial system such as inhaling devices and nebulisers, and formulated according to conventional methods of pharmacy optionally with one or more pharmaceutically acceptable carriers or excipients, such as for example those described in Remingtons

Pharmaceutical Sciences, ed. Gennaro, Mack Publishing Company, Pennsylvania, USA (1990) .

Such compositions are conveniently formulated in unit dosage form eg. for mucosal, parenteral or oral administration.

Although NISV are known per se, it has not previously been proposed to divide such preparations into unit dosages.

Therefore, in a further aspect, the present invention provides a pharmaceutical composition comprising empty NISV together with a pharmaceutically acceptable carrier or excipient, conveniently in unit dosage form.

In this context, "empty NISV" are NISV which have no active agent entrapped or associated with them.

Actual treatment regimes or prophylactic regimes, formulations and dosages will depend to a large extent upon the individual patient and may be devised by the medical practitioner based on the individual circumstances.

The type of formulation will be appropriate to the route of administration. For example, parenteral

administration of NISV by subcutaneous or intramuscular injection may be with a sterile aqueous suspension of NISV in PBS or water for injection, provided in ampoules, vials or as measured doses in pre-filled syringes or in the form of a lyophilisate for reconstitution with PBS or water for injection prior to administration.

The dosage of NISV for subcutaneous injection may include from 2.5 to 50 mg eg. 2.5 to 25 mg of vesicles formulated as described above for example in PBS or water. Administration regimes for the subcutaneous route may be determined by the duration of action in specific clinical situations. Frequency of administration may range from daily injections to injections weekly or fortnightly. Typical administration regimes may for example comprise two doses at fourteen day intervals, three doses at fourteen day intervals, three doses at intervals of 0, 28 and 84 days, or three doses at seven day intervals. Mucosal adminstration may for example follow these regimes. The dosages for oral administration may be between 2.5 to 50 mg of vesicles or considerably higher. Suitable oral formulations include flavoured liquid suspension of syrups, liquid/powder filled capsules. For administration to the respiratory tract (nasally or orally) metered spray inhalers or nebulisation of an aqueous suspension of NISV may be used.

Although primarily of applicability to humans, the invention may also be used in veterinary medicine for example to treat companion animals such as cats and dogs, and livestock, eg. poultry.

In general, the size of the vesicles is not critical and the method is applicable to a wide size range of vesicles, appropriate for administration by the above-mentioned routes. A wide range of NISV sizes has been described in the literature ranging for example in the order of about 100 nm to several micrometers and can

be used. We have however found that whilst the levels of the proinflammatory cytokines and the cytokine mediators of inflammation are affected by the vesicles of the invention, the degree of modulation of cytokine levels may be influenced by the size of the vesicles and in particular that certain vesicle sizes may enhance the reduction in cytokine levels which may be observed. Thus we have found that effective down regulation of pro-inflammatory cytokine production is particularly pronounced above a threshold which our experiments have shown to be within the range 150-215 nm. This effect may vary depending on the system used ie. the nature of the vesicle, the condition being treated and even the animal species concerned. Appropriate vesicle sizes for achieving this enhanced beneficial effect, may be determined by appropriate tests. In vitro, murine and human cells appear to show similar size thresholds for enhancement of the beneficial effects of NISV. Thus, for example, in the case of mice and humans our experiments have shown that the beneficial enhanced effect may be obtained with vesicles of greater than about 200 nm. In particular larger vesicles, typically of mean diameter greater than approximately 200 or 215 nm up to several micrometers (eg. 750-3000 nm) have been shown to cause greater down-regulation of IL-5 and IL-1 as compared with smaller vesicles, for example those of mean diameter approximately 160 nm. In vivo experiments have demonstrated that vesicles up to several micrometers are effective. Thus, a preferred aspect of this invention comprises all the methods, uses and compositions of the invention wherein the vesicles are of mean diameter greater than 200 nm, preferably greater than 215 nm and more preferably greater than 250 nm. This finding of the role of NISV size may be important in all the cases of allergic and inflammatory diseases associated with elevated levels of

proinflammatory cytokines and cytokine mediators of inflammation.

If desired, the size distribution of vesicles within a preparation may be modified for example to reduce variability, exclude vesicles of certain size ranges or obtain a homogenous preparation. This may be achieved by extrusion through polycarbonate membranes with pores of known diameter as described in Nayar et al, Biochem. Biophys. Acta, 986, 200-209 (1989) . It will of course be appreciated that actual size of the vesicles produced by this extrusion method may differ from the stated pore diameter of the membrane used, and it is therefore desirable to further characterise the size of the vesicles once formed by other methods known to those skilled in the art, such as photon correlation spectroscopy (PCS) and electron microscopy (EM) . Other methods of preparing vesicles of relatively homogeneous size include the use of a French press, microfluidisation/ homogenisation and sonication as described in Lasic, Liposomes: from physics to applications, Elsevier, Amsterdam (1993) . The size of such preparations may be confirmed by the techniques described above. If other preparative methods are used, vesicles of desired size may be fractionated from a more heterogeneous size population by a variety of techniques known to those skilled in the art, including for example size exclusion chromatography, centrifugation etc.

The above discussion of cytokine pathways and interactions is only a summary of the existing knowledge about this highly complex and rapidly-advancing field. It should be clearly appreciated that our invention is not tied to any specific theory of cytokine activity but is firmly based on experimental observations. It will be appreciated that the activities of certain cytokines vary according to cell type, and that cytokines may have different properties according to the

condition and its stage of development .

The method of the invention is applicable to the production of therapeutic agents comprising NISV as the sole therapeutic agent, and also to combinations together with one or more other agents useful in the treatment of the diseases or conditions concerned, for example anti-inflammatory agents such as corticosteroids, antihistamines and anti-cytokine antibodies, therapeutic cytokines such as IL-2, IL-12 and IFN-γ, and in the case of septic shock, antibacterial agents. Such additional agents may be administered simultaneously or sequentially with the surfactant vesicles, preferably NISV and in the case of simultaneous administration, the agents may be provided in admixture with the vesicles, and/or entrapped therein.

Thus viewed from a yet further aspect, the invention provides a product containing NISV and at least one other pharmaceutically active agent as a combined preparation for simultaneous separate or sequential use in therapy. The pharmaceutically active agent will be selected according to the therapy, and examples are given herein.

The NISV may also be administered in combination with an agent (s) in order to counteract the unwanted side effects of that agent (s) without removing the therapeutic effect of the agent (s) . For example, in the treatment of cancer, the administration of NISV in a combination therapy may be used to reduce the undesirable side effects, such as cachexia, of chemotherapeutic drugs.

Furthermore, in such combination therapies, entrapping an agent (s) within NISV may, in addition to providing the beneficial therapeutic effects of the NISV, be used to improve the pharmacokinetic profile of that agent (s), for example by providing a sustained release vehicle for the agent (s) or by protecting the

agent (s) from degradation or rapid clearance from the system.

Methods by which other agents may be entrapped within preformed NISV include the dehydration- rehydration method (Kirby & Gregoriadis, Biotechnology, 2., 979-984 (1984)) in which the agent present in the aqueous phase is entrapped in pre-formed vesicles by flash freezing followed by lyophilisation, and the freeze-thaw technique (Pick, Arch. Biochem. Biophys. 212 , 195-203 (1981)) . In the latter technique, vesicles are mixed with the agent concerned and repeatedly flash frozen in liquid nitrogen and e.g. warmed to temperatures of the order of 60°C (ie. above the transition temperature of the relevant surfactant) . In the case where NISV are prepared by homogenisation, the agent may be entrapped during the homogenisation process itself. In such a method the agent is dissolved in the aqueous phase prior to homogenisation.

The invention will now be described by way of the following non-limiting Examples, with reference to the Figures which show:

Figure 1 : The ability of NISV to reduce weight-loss associated with T.Gondii-induced cachexia in mice. Figure la: A bar chart showing the number of T.Gondii cysts in brains of mice infected with T.Gondii after immunisations with PBS, NISV, soluble tachyzoite antigen (STAg) , STAg entrapped in NISV (NISV/STAg) or STAg mixed with NISV (NISV+STAg) .

Figure lb: Graphs showing the level of weight loss after infection with T.Gondii in experimental mice actively immunised with PBS, NISV, STAg or NISV/STAg.

Figure 2 : The ability of NISV to reduce weight loss associated with FK-565 induced cachexia in mice.

Figure 2a: Graph showing weight loss in mice after intraperitoneal injection with PBS, NISV, the acyltripeptide FK-565 in PBS or FK-565 entrapped in NISV.

Figures 2b and 2c: Bar charts showing the levels of IL-2 (Figure 2b) and IL-12 (Figure 2c) produced in ConA stimulated splenocytes isolated from mice treated as per Fig. 2a. Figure 3 : Reduction by NISV in levels of cytokines produced by stimulated murine cells.

Figure 3a: Bar chart showing TNF-α production by the stimulation of the murine macrophage cell line J774 with LPS+IFN-γ, IFNγ, LPS, NISV+IFNγ+LPS, NISV+IFN-γ, NISV+LPS or NISV.

Figure 3b: Bar chart showing TNF-α production by the stimulation of murine peritoneal macrophages with PBS, LPS, LPS+NISV, LPS+IFN-γ or LPS+NISV+IFN-γ.

Figure 3c: Bar chart showing IL-6 production by the stimulation of murine peritoneal macrophages with

LPS+NISV or LPS+PBS as compared to IL-6 produced by unstimulated cells.

Figure 4 : Reduction by NISV in levels of cytokines produced by stimulated human cells. Figures 4a and 4b: Graphs showing levels of IL-6

(Figure 4a) and TNF-α (Figure 4b) in human peripheral blood leucocytes treated with PBS, LPS or LPS+NISV.

Figure 5 : Reduction by NISV in levels of cytokines produced by stimulated human cells. Figures 5a and 5b: Graphs showing levels of IL-6 (Figure 5a) and TNF-α (Figure 5b) in human peripheral blood leucocytes treated with PBS, LPS or LPS+NISV.

Figures 5c and 5d: Graphs showing the level of IL- lα (Figure 5c) and IL-lβ (Figure 5d) in human peripheral blood leucocytes treated with PBS, LPS, NISV or

LPS+NISV.

Figure 6: The effect of size of NISV on their immunomodul tory effect.

Figures 6a and 6b: IL-2 and IL-5 (Fig. 6a) and IL-5 and IFNγ (Fig. 6b) produced by con-A stimulated lymph node cells collected from mice treated with ovalbumin entrapped in NISV prepared by extrusion though membranes

of pore size 800 nm, 400 nm, 200 nm and 100 nm, and in PBS.

Figures 7a and 7b: Bar charts showing levels of IL- lα (Figure 7a) and IL-lβ (Figure 7b) in cells from human leucocyte pro-macrophage cell line U937 and PBLS from healthy volunteers treated with PBS, LPS, NISV extruded through 200 nm pore size membrane or non-extruded NISV;

Figures 8, 9 and 10: Suppresion of LPS induced cytokine production by NISV in mice. Figure 8 is a bar chart showing serum IL-6 production in mice in response to LPS administered 1, 4 or 14 days after subcutaneous injection of NISV. Figure 8 shows results from an NISV dose of 17 mg/kg, 3 hours after LPS challenge (Figure 8a) or 6 hours after LPS challenge (Figure 8b) .

Figure 9 shows results from an NISV dose of 80 mg/kg 3 hours after LPS challenge (Figure 9a) or 6 hours after LPS challenge (Figure 9b) . (*: p≤0.05 v control; **: p≤0.025 v control; *** : p≤0.0005 v control) Figure 10 shows results from an NISV dose of 80 mg/kg after a second challenge with LPS at days 15, 18 or 28 days after the NISV dose. (*: p≤0.05 v control)

Figures 11 and 12: Reduction by NISV in levels of TNFα and IL-6 in LPS-stimulated human PBLs. Figures Ila and lib are graphs showing TNFα levels in LPS-stimulated (lib) or non-stimulated (Ila) PBLs extracted from a human volunteer before (dotted line) and after (solid line) administration of NISV.

Figures 12a and 12b are graphs showing IL-6 levels in LPS-stimulated (12b) or non-stimulated (12a) PBLs extracted from a human volunteer before (dotted line) and after (solid line) administration of NISV.

EXAMPLES

Example 1

The modulation of cachexia

Cachexia is a serious condition characterised by pronounced weight-loss caused by several underlying conditions (e.g. chronic infections with microorganisms such as viruses and parasites, tumours, congestive cardiac failure) but mediated by proinflammatory cytokines, in particular TNF-α and IL-6. The induction of this weight-loss is independent of the nutritional status of the animal, which could be well fed, and is characterised by loss of body mass. NISV were investigated as an immunomodulator in two models for prevention of cachexic weight-loss, one in which the cachexia was induced by parasitism and the other in which it was induced by a peptide drug.

A. Reduction in the level of weight-loss associated with Toxoplasma σondii-induced cachexia in mice treated with NISV

Materials and Methods

Vesicle preparation: Vesicles were prepared by the methods previously described by Brewer and Alexander (Immunology, 25_, 570-575 (1992)) . 1-Mono palmitoyl glycerol (MPG) , cholesterol (CHOL) and dicetyl phosphate (DCP) (all Sigma, Poole, Dorset, U.K.) were mixed in a 15 ml pyrex test tube in the molar ratio 5:4:1 (MPG: CHOL: DCP) to a total of 150 μmoles and then heated to 130°C in a dry-block (Grant) until melted. Empty vesicles were formed when 5 ml aqueous buffer (PBS, pH 7.4) was added and the resulting suspension vortexed vigorously for 1 minute and the suspension shaken at

60°C for 2 hours. The vesicles were then ready for storage or in vi tro assay. Previous studies have shown that vesicles prepared by this method yielded vesicles of approximately 2 micron diameter.

Antigen entrapment: Antigen entrapment into preformed vesicles was achieved by the dehydration-rehydration technique as described by Kirby and Gregoriadis, (Biotechnology, 2 , 979-984 (1984)) . Briefly, 5 ml (150 μmoles) of vesicle solution were mixed with 2 ml antigen in PBS (5 mg/ml) in polypropylene centrifuge tubes (Elkay Products Inc., Shrewsbury, MA, U.S.A.) and flash frozen as a thin shell by swirling in liquid nitrogen. Preparations were then lyophilised in a freeze drier at 0.1 torr overnight before rehydration in 0.5 ml distilled water.

Animals and inoculations: BALB/K mice were in-house bred and inoculated when 8-10 weeks old. Groups of 5 mice were immunised subcutaneously with either 50 μg soluble tachyzoite antigen (STAg) (as prepared by Roberts & Alexander, Parasitology, 104. 19-23 (1992)) in PBS emulsified with 100 μl FCA, 50 μg STAg entrapped within NISV, or 50 μg STAg mixed with empty NISV. Control groups were immunised with the same volume of

PBS or empty NISV. The inoculations were repeated after 2 weeks.

Mice were infected with 20 viable T. gondii cysts 2 weeks after the second immunisations. Four weeks later cyst burdens were enumerated manually from brain suspensions. The mean body weights of the mice were measured for 32 days post infection.

RESULTS

Active immunisation with STAg alone, STAg entrapped in NISV (i.e. NISV/STAg) or STAg mixed with NISV (i.e. NISV & STAg) resulted in a significant reduction in the number of parasites encysted within the mice brains (Figure la) . There was no reduction in and no significant difference between the number of cysts per brain in the control groups treated with PBS or NISV alone. NISV alone had little effect on the numbers of cysts able to infect each mouse, i.e. they did not provide any direct protection against the parasites or directly affect the parasites themselves.

Figure lb shows the level of weight-loss post-infection with the T. gondii cysts. Weight-loss (cachexia) was severe in the control group receiving PBS. Mice in this group lost approximately 12% of their body weight within 20 days post-infection. In contrast, the administration of NISV in the other control group prevented any significant metabolic weight loss in infected mice, without altering their degree of parasitism. The weights of mice treated with NISV alone were similar to those exhibiting active, protective immunity after inoculation with STAg, NISV/STAg or NISV-STAg.

The administration of NISV alone did not prevent infection by Toxoplasma gondii parasites or effect parasite viability as shown by the number of cysts in the brain. However, NISV administration did prevent the considerable weight-loss (cachexia) associated with this condition. In contrast, mice in the PBS control group exhibited radical losses of weight over 20 days as well as high levels of parasitism.

B. Reduction in the level of weight-loss associated with FK-565-induced cachexia in mice treated with NISV

FK-565 is an experimental acidic acyltripeptide known to have potent antitumour and antibacterial effects. It enhances anti-tumour host defence activity by inhibiting tumour growth. Repeated intraperitoneal injections of FK-565 significantly activates the cytotoxicity of murine peritoneal macrophages and natural killer (NK) cells towards tumours and also augments their killing potential. FK-565 exhibits further antitumour activity by increasing the release of TNF-α, a potent anti-tumour cytokine. However, the increased production of TNF-α and other proinflammatory cytokines by FK-565 is associated with a significant degree of drug-induced cachexia.

Materials and Methods

NISV were prepared as described by Brewer and Alexander (Supra) .

Drug entrapment: FK-565 was entrapped in preformed NISV using the freeze-thaw method (Pick, Arch. Biochemistry, Biophysics 212, 195-203 (1981)) . The antigen vesicle mixture was frozen in liquid nitrogen and then thawed to 60 ^β C. This was repeated five times. The suspension was shaken for a further 2 hours at 60°C. The level of entrapment of FK-565 was assessed using a standard ninhydrin assay.

Animals and inoculations: Female BALB/c mice were in- house bred and inoculated when 8-10 weeks old. Groups of 5 mice were immunised intraperitoneally with either 100 μl PBS, empty NISV (100 μl; 5 mg) , FK-565 in PBS (200 μl; 20 μg in toto) or FK-565 entrapped in NISV (200

μl; 6 μg in toto in 10 mg vesicles) . Mouse weights were recorded prior to injection and at 1, 2 and 3 days after injection of the above agents.

Cytokine assays: Spleens were collected from the mice 3 days after injection and pooled in RPMI 1640 culture medium supplemented with 2 mM L-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, 0.05 mM β- mercaptoethanol and 10% (v/v) foetal calf serum (FCS) (all Gibco, Paisley, U.K.) . Splenocyte suspensions were prepared by gently teasing the spleens apart with forceps after which the suspensions were centrifuged at 200 x g for 10 minutes and resuspended in 0.5ml of medium. Viable cells were enumerated by Trypan Blue exclusion test and cell suspensions adjusted to 5xl0 ⁶ cells/ml. 100 μl/well aliquots of cell suspension, containing 5xl0 ⁵ cells, were added to 96-well flat- bottomed tissue culture plates (Costar, Cambridge, M.A. , USA) , followed by 100 μl/well aliquots of Con A (5 μg/ml) in triplicate. Cultures were then incubated for

60 hours at 37°C in a 5% C0 ₂ atmosphere after which 150 μl aliquots of cell culture supernatants were removed and stored at -70°C for cytokine assay.

Cytokines (IL-2 and IL-12) were detected by monospecific ELISA. Flat-bottomed polystyrene plates (Dynatech, Alexandria, VA, USA) were coated overnight at 4°C with 50 μl/well anti-mouse cytokine monoclonal antibodies at optimum concentrations (determined at 2 μg/ml in each case) . (IL-2 reagents were obtained from Pharmingen,

San Diego, CA, USA, and IL-12 reagents from Wistar Institute, USA) . Plates were washed three times with PBS/Tween (PBST) (pH 7.4, 0.05% Tween 20) and blocked with 200 μl/well 10% (v/v) FCS in PBS for 60 minutes at 37°C. Plates were washed three times in PBST, 100 μl samples of supernatants and standards (IL-2, 0-62.5 units/ml; IL-12, 0-500 pg/ml) added in duplicate to the

wells and incubated for 2 hours at 37°C. After four washes with PBST, 100 μl/well of biotinylated anti-mouse cytokine monoclonal antibody (all 1 μg/ml) was added and the plates incubated for 45 minutes at 37 °C. After washing six times with PBST, 100 μl/well alkaline phosphatase-streptavidin conjugate (Pharmingen) , diluted 1/2000 in 10% (v/v) FCS in PBS, was added and the plates incubated for 30 minutes at 37°C. Plates were washed eight times in PBST and 100 μl/well of para-nitrophenyl- phosphate (pNPP) substrate (Sigma) , prepared in glycine buffer (0.1 M; pH 10.4), added. Plates were incubated for 30 minutes at 37°C in darkness before the resulting absorbances were read at 405 nm on a Titertek Multiskan plate reader (Flow Laboratories, Irvine, Ayrshire, U.K.) . Cytokine concentrations in the cell cultures were determined from the standard curve (regression coefficient, r = 0.990 or better) . Comparisons between groups were made using a Student ' s T test .

RESULTS

Figure 2a shows the level of weight-loss after intraperitoneal administration of FK-565. The control groups which received either PBS or NISV exhibited normal, unaltered body weights throughout. Mice treated with FK-565 in PBS lost 16% of their body weight 3 days post-administration of the drug. Mice which received FK-565 entrapped in NISV showed a significant reduction in weight-loss, with the mean body weight stabilising after 1 day and rising 3 days after injection. The mice lost approximately 5% mean body weight over 3 days.

IL-2 and IL-12 production in Con A stimulated splenocytes isolated from mice after the various treatments was compared (Figures 2b and 2c, respectively) . The administration of FK-565 entrapped in NISV elicited a predominantly Thl-type immune

response as evidenced by the significant production of IL-2 and IL-12. These cytokines were not produced in levels significantly greater than controls after the delivery of FK-565 in PBS which tends to produce cytokines involved with a Th2 response and inflammation.

The administration of FK-565 entrapped in NISV did not completely prevent the cachexic weight-loss attributable to FK-565 but did considerably reduce it (from 16% to 5% over 3 days) . NISV, as is apparent from the data presented below, prevented cachexia by direct down¬ regulation in the production of the major proinflammatory cytokines including TNF-α, IL-6 and IL-1 (α and β) .

Example 2

Reduction of LPS-induced TNF-α and IL-6 levels in in vi tro murine macrophage models treated with NISV

The results in Example 1 show that NISV has a significant therapeutic ability to reduce whole-body metabolic weight-loss caused by proinflammatory cytokines. TNF-α is a primary proinflammatory cytokine associated with cachexic weight-loss. An experiment was set up to confirm that this reduction was mediated by NISV and the ability of NISV to reduce the levels of TNF-α.

Materials and Methods

NISV were prepared by the method described in Example 1.

A. In vitro study using the murine macrophage cell line J774.

The murine macrophage cell line, J774, harvested from

BALB/c/NIH mice, were obtained as a gift from Professor H. Harris/Dr R. Sutherland (Sir William Dunn School of Pathology, Oxford, U.K.) . The cells were maintained in RPMI 1640 culture medium supplemented with 10% (v/v) FCS, 2 mM L-glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin (all Gibco) at 37°C in a 5% C0 ₂ atmosphere. Cells were washed extensively by centrifugation at 1000 x g before activation with various agents in Dulbecco's modified Eagles medium (DMEM) supplemented with 3 M D-glucose and 10% FCS.

100 μl/well aliquots of cells, at a concentration of approximately lxlO ⁶ cells/ml, were established in 96-well flat-bottomed tissue culture plates (Costar) . Seven groups (5-6 replicates/group) of cells were treated with 10 μl of the following preparations and incubated at 37°C in a 5% C0 ₂ atmosphere:

Group 1 LPS (40 ng/ml) + IFNγ (10 units/ml) Group 2 IFNγ (10 units/ml) Group 3 LPS (40 ng/ml) Group 4 NISV (0.2 mg) + IFNγ (10 units/ml) + LPS (40 ng/ml)

Group 5 NISV (0.2 mg) + IFN (10 units/ml) Group 6 NISV (0.2 mg) + LPS (40 ng/ml) Group 7 NISV (0.2 mg)

Cytokine assay: After 48 hours of stimulation, supernatants were removed from the test culture wells and tested for TNF-α production using a monospecific ELISA (Pharmingen) . Flat-bottomed polystyrene plates (Dynatech) were coated overnight at 37°C in a 5% C0 ₂ atmosphere with 50 μl/well rat anti-mouse TNF-α monoclonal antibody. Plates were washed three times with PBST and blocked with 200 μl/well 10% (v/v) FCS in PBS for 60 minutes at 37°C. Plates were washed three times with PBST, 100 μl samples of supernatants and

standards (20 ng/ml TNF-α) , diluted in 10% (v/v) FCS in PBS, added and incubated for 2 hours at 37°C. After four washes with PBST, 100 μl/well of biotinylated rat anti-mouse TNF-α monoclonal antibody (0.5 mg/ml in 10% (v/v) FCS in PBS) was added and the plates incubated for 45 minutes at 37°C. Detection of TNF-α levels using an alkaline phosphatase-streptavidin conjugate and pNPP substrate was exactly as described in the cytokine assays in Example 1.

B. In vi tro study using murine peritoneal macrophages.

The above study was carried out in an established murine cell culture (J774) . In order to assess that the same immunomodulatory effect was seen in cells freshly removed from mice, which should arguably provide a more accurate model of a true in vivo system, a parallel study was performed using exudated murine peritoneal macrophages.

Peritoneal macrophages were removed from BALB/c mice by introduction of a nominal volume of RPMI culture medium and subsequent removal via a fine-gauge needle. Cells were washed as described in (A) in RPMI culture medium supplemented with 3 M D-glucose and 10% FCS.

100 μl/well aliquots of cells, at a concentration of approximately 2.5xl0 ⁶ cells/ml, were incubated in 96- well flat-bottomed tissue culture plates (Costar) for 4 hours at 37°C in a 5% C0 ₂ atmosphere. Non-adherent cells were removed by aspirating the wells with 2 volumes of RPMI culture medium supplemented with FCS. Eight groups (3 replicates/group) of cells were treated with the following preparations and incubated at 37°C in a 5% C0 ₂ atmosphere:

Group 1 Control (PBS, pH 7.4) Group 2 LPS (200 ng/well) Group 3 LPS (200 ng/well) + NISV (0.2 mg) Group 4 LPS (200 ng/well) + IFN-γ (10 units/well) Group 5 LPS (200 ng/well) + NISV (0.2 mg) - IFN-γ (10 units/well)

Group 6 LPS (200 ng/well) + NISV (0.2 mg) Group 7 LPS (200 ng/well) + control (PBS, pH 7.4) Group 8 No stimulation (background)

Cytokine assay: After 24 hours stimulation, 100 μl of supernatants were removed from each test culture-well and tested for:

i) TNF-α production using a monospecific ELISA

(Pharmingen) exactly as described in (A) above (Groups 1-5) .

ii) IL-6 production using a monospecific ELISA (Pharmingen) exactly as described in (A) above (Groups 6-8) .

RESULTS

Figure 3a shows the levels of TNF-α release after stimulation of J774 cells. Treatment of J774 murine macrophages with NISV down-regulated the LPS-induced production of the pivotal proinflammatory cytokine, TNF- α. LPS treatment of the murine macrophages resulted in levels of TNF-α that are significantly higher than those elicited after the treatment of J774 cells with the other agents. Co-administration of NISV and LPS to J774 macrophages significantly reduced (p < 0.025) the production of TNF-α, compared to the administration of LPS alone. A similar reduction in TNF-α release in J774 macrophages was observed after the co-administration of NISV and IFN-γ.

Figure 3b shows the levels of TNF-α release after LPS stimulation of murine peritoneal macrophages (treatment Groups 1-5) . The administration of LPS alone to the murine macrophages induced considerable release of TNF-α from the cells. The co-administration of LPS and IFN-γ caused a greater increase, although not significant, in TNF-α production than LPS alone. Co-administration of NISV and LPS substantially reduced (by approximately 50%) the level of TNF-α released. Similarly, the reduction of TNF-α release from the peritoneal macrophages was even greater after the co-administration of NISV and IFN-γ (approximately 75%) .

Figure 3c shows the levels of IL-6 release after LPS stimulation of murine peritoneal macrophages (treatment Groups 6-8) . By co-administering NISV with LPS, the level of IL-6 produced by murine peritoneal macrophages was significantly reduced as compared to the administration LPS in PBS. Unstimulated peritoneal macrophages did not produce IL-6.

This study shows that NISV can significantly reduce the levels of TNF-α in cultured murine macrophages after stimulation with LPS, which mimics an inflammatory event. The results from the peritoneal macrophages also indicate that this effect is not specific to established murine cell cultures and demonstrate the immunomodulatory capacity of NISV to reduce the levels of two pivotal proinflammatory cytokines, TNF-α and IL-6, after an inflammatory-type stimulus in cells freshly removed from mice and maintained for 24 hours.

This data provides an important link between the cellular events that result in prevention of inflammatory weight-loss (cachexia) , the whole-body effect observed in Example 1. It indicates that the immunomodulatory effects of NISV act to significantly

reduce both TNF-α and IL-6. This reduction in proinflammatory cytokines at a cellular level can result in a benefit in inflammatory-mediated conditions, at the physiological level. In vivo both the effects on TNF-α and IL-6 would occur in tandem and act to down-regulate the inflammatory response to a greater extent than either acting alone.

Example 3

Reduction in the levels of LPS-induced TNF-α and IL-6 and in an in vi tro human peripheral blood leucocyte model treated with NISV

This study was carried out to demonstrate that the ability of NISV to down-regulate proinflammatory cytokine production was not exclusive to murine systems and that human cells responded equally as readily to the anti-inflammatory effect of NISV.

Materials and Methods

NISV were prepared by the method described in Example 1.

Peripheral blood mononuclear leucocytes (PBLs) were prepared from heparinised/citrated venous blood from healthy adult volunteers. Approximately 10 ml of whole human venous blood was withdrawn into a heparinised syringe and carefully layered onto 10 ml Ficoll/Paque, room temperature-equilibrated in a sterile centrifuge tube, with minimal perturbation of the interface between the fluids. The blood/Ficoll gradient was centrifuged at 1000 x g for 30 minutes at 20°C. The PBLs appeared as a tight band of cells layered above the pelleted erythrocytes and polymorphonuclear leucocytes and below the straw-coloured pool of plasma and platelets. The plasma and platelets were carefully removed and

discarded leaving a very thin film of plasma above the leucocytes. These were aspirated into a sterile 20 ml Nunc tube containing 10 ml RPMI 1640 culture medium supplemented with 10% (v/v) FCS in PBS and stored at 4°C. The cells plus medium were re-centrifuged for 20 minutes at 25°C prior to use and the cell pellet resuspended in 10 ml RPMI 1640 supplemented with FCS as before. The cells were re-washed in 10 ml RPMI 1640 supplemented with FCS and re-suspended in 10 ml DMEM supplemented with FCS and suitable antibiotics (1% w/v tetracycline; 1% w/v streptomycin). The cells were then ready for in vi tro assay.

2 ml/well aliquots of human PBLs, at a concentration of approximately 2xl0 ⁶ cells/ml, were established in 24-well flat-bottomed tissue culture plates (Costar) . Three groups (3 replicates/group) of cells were treated with the following preparations and incubated at 37°C in a 5% C0 ₂ atmosphere:

Group 1 Control (PBS, pH 7.4) Group 2 LPS (40 ng/ml) Group 3 LPS (40 ng/ml) + NISV (1 mg)

The cells were maintained for 72 hours with aliquots of the cell culture supernatants removed at 4, 24, 48 and 72 hours post-stimulation and stored at -70°C for subsequent cytokine assay.

The supernatants from the cell cultures were assayed by monospecific human ELISA for IL-6 (Genzyme) and TNF-α (Pharmingen) according to the manufacturers' instructions. The ELISA plates were developed using a suitable colorimetric solution and read at 405 nm on a Titertek Multiskan plate reader (Flow Laboratories) .

Cytokine concentrations in the cell culture supernatants were determined from appropriate standard curves

(regression coefficient r = 0.990 or better) . Preliminary statistical analyses were carried out. Any significance mentioned in the text refers to standard deviation measurements (not shown) .

RESULTS

This study illustrates the response of human PBLs to in vi tro stimulus with LPS, which mimics an inflammatory response. Figures 4a and 4b indicate that the introduction of LPS to the cells resulted in the production of the cytokines IL-6 and TNF-α, respectively.

Figure 4a shows the release of IL-6. After stimulation with LPS, the levels of IL-6 increased rapidly and peaked at 2.5 ng/ml 24 hours after administration. Stimulation of the PBLs with PBS resulted in negligible levels of IL-6 release. The co-administration of NISV with LPS completely prevented IL-6 release. Measurable levels of IL-6 after NISV treatment were not significantly different from those achieved after the control administration of PBS.

Figure 4b shows the release of TNF-α. TNF-α was released very rapidly from human PBL after stimulation with LPS and was present in high levels only 4 hours after stimulation. Co-administration of NISV with LPS reduced the level TNF-α release at 4 hours post- introduction, as compared with the group receiving LPS alone, from 5.5 ng/ml to 3 ng/ml .

These results illustrate the considerable ability of NISV to reduce the levels of the pivotal proinflammatory cytokines IL-6 and TNF-α released by human PBLs in vi tro after stimulation with LPS. Both IL-6 and TNF-α are potential therapeutic targets for modulation, e.g. in

the treatment of chronic diseases such as rheumatoid arthritis as well as acute conditions such as SIRS. In this study, NISV were observed to almost completely prevent IL-6 release from stimulated PBL and greatly reduce levels of TNF-α release, as rapidly as 4 hours after co-administration with LPS.

Example 4

Reduction in the levels of LPS-induced TNF-α. IL-6 and IL-1 in human PBLs with NISV.

Materials and Methods

NISV were prepared by the method described in Example 1.

PBLs were obtained from heparinised/citrated venous blood from healthy adult volunteers and also from the Scottish Blood Transfusion Service. PBLs were prepared by density-dependent centrifugation using a

Ficoll/Plaque gradient medium as described in Example 3.

2 ml/well aliquots of human PBLs, at a concentration of approximately 2xl0 ⁶ cells/ml, were established in 24-well flat-bottomed tissue culture plates (Costar) . Four groups (3 replicates/group) of cells were treated with the following preparations and incubated at 37°C in a 5% C0 ₂ atmosphere:

Group 1 Control (PBS, pH 7.4) Group 2 LPS (40 ng/ml) Group 3 NISV (1 mg) Group 4 LPS (40 ng/ml) + NISV (1 mg)

The cells were maintained for 48 hours with aliquots of the cell culture supernatants removed at 1.5, 4, 24 and 48 hours post-stimulation and stored at -70°C for

subsequent cytokine assay.

The supernatants from the cell cultures were assayed by monospecific human ELISA for IL-1 (α and β) (Dynatech) (all treatment groups) , IL-6 (Genzyme) and TNF-α

(Pharmingen) (treatment groups 1, 2 & 4) , according to the manufacturers' instructions. The ELISA plates were developed using a suitable colorimetric solution and read at 490nm on a Titertek Multiskan plate reader (Flow Laboratories) . Cytokine concentrations in the cell culture supernatants were determined from appropriate standard curves (regression coefficient r = 0.990 or better) . Preliminary statistical analyses were carried out. Any significance mentioned in the text refers to standard deviation measurements (not shown) .

RESULTS

This study shows the ability of NISV to reduce the levels of proinflammatory cytokines from stimulated human PBL, in vi tro . Figures 5a and 5b indicate that the introduction of LPS to the cells resulted in the production of the cytokines IL-6 and TNF-α, respectively. Notably, the control PBLs in this particular study happened to be derived from an individual with an ongoing inflammatory response. As such, the control PBLs consistently produced a detectable level of the proinflammatory cytokines IL-6 and TNFα.

Figure 5a shows the release of IL-6. LPS-stimulated PBLs produced peak levels of IL-6 after 24 hours of culture. These levels were similar to those observed in Figure 4a. Control PBLs also produced measurable levels of IL-6. The co-administration of NISV with LPS reduced the level of IL-6 below control levels.

Figure 5b shows the release of TNF-α. After stimulation with LPS, the levels of TNF-α increased rapidly and peaked 4 hours post-administration. The co- administration of NISV with LPS reduced the level of TNF-α by approximately 50%. Control PBLs also produced measurable levels of TNF-α in this system.

Figure 5c shows the release of IL-lα. NISV failed to induce significant levels of IL-lα in the absence of a stimulatory signal (LPS) . The administration of LPS resulted in a considerable production of IL-lα. The co- administration of NISV with LPS substantially reduced the release of IL-lα over 48 hours.

Figure 5d shows the release of IL-lβ. NISV failed to induce any significant levels of IL-lβ in the absence of a stimulatory signal (LPS) . The administration of LPS resulted in a considerable production of IL-lβ. The co- administration of NISV with LPS reduced the release of IL-lβ over 48 hours.

This study represents a repeat investigation in which the results from Example 3, showing the considerable ability of NISV to reduce the pivotal proinflammatory cytokines, were confirmed.

Interestingly, the vesicle preparation, at the concentration used in this study, had no substantial affect on the kinetics of release of these pro- inflammatory cytokines (except in instances of complete prevention of release) .

The control PBLs from this study represent a cell population that are similar to those that would be encountered in individuals afflicted by an ongoing inflammatory response. As such, the ability of NISV to reduce the "baseline" levels of one of the major

proinflammatory cytokines, IL-6, can only be regarded as beneficial and significant.

The results also demonstrate that NISV have the ability to reduce the overall levels of IL-1 elicited after LPS stimulation of human PBLs. NISV have the ability to significantly reduce the level of IL-lα produced in this system and, in addition, reduce the amounts of IL-lβ elicited, both of which are important in the inflammatory response.

Example 5

Effect of size on the immunomodulatory effect of NISV

This study was designed to assess if the size of vesicles influenced the level of therapeutic immunomodulation achieved with NISV.

A. The effect of size of NISV on IL-5 levels.

Whether the size of the vesicles could influence the therapeutic potential of NISV was determined from immunogenicity studies using entrapped ovalbumin (OVA) as an antigen. IL-5, a cytokine implicated in the onset of asthma, was measured.

Materials and Methods

NISV were prepared by the method described in Example 1.

Antigen entrapment: OVA (grade V, Sigma) was entrapped in preformed NISV using the freeze-thaw method (Pick, 1981) . The antigen/vesicle mixture was frozen in liquid nitrogen and then thawed to 60°C. This was repeated five times. The suspension was shaken for a further 2 hours at 60°C and vesicle preparations of different

sizes prepared by sequential extrusion through decreasing pore size polycarbonate filters (Costar) at 60°C in a thermobarrel extruder (Lipex Biomembranes Inc., Vancouver, Canada). The free antigen was removed by washing at 1000000 x g for 40 minutes at 4°C. The protein concentration was measured by nitrogen assay (Brewer et al. , vaccine 13.(5), 1441-1444 (1995)).

Electron microscopy: The vesicles were examined by electron microscopy as follows. A small sample of vesicle suspension (approximately 10 μl) was sandwiched between clean copper plates (Balzers High Vacuum, Milton Keynes, U.K.) and fast-frozen by plunging into liquid propane at -190 °C. Samples were then transferred to a cold stage at -100°C in a diffusion pumped vacuum system operating around 4x10" ⁶ torr. The support plates were fractured apart and the exposed surfaces shadowed immediately with evaporated platinum/carbon at 45° followed by a second strengthening coat of carbon applied at 90° to the exposed fracture faces. The vesicle preparations were removed from the replica by sequential washing in acetone/distilled water solution of several decreasing acetone concentrations from pure acetone. Finally, after several washes in distilled water, the replicas were collected onto copper grids, dried and examined under a transmission electron microscope.

Animals and inoculations: Female BALB/c mice were in- house bred and inoculated when 8-10 weeks old. Groups of 5 mice were inoculated in the footpad with 10 μl of the following:

Group 1: 10 μg OVA entrapped in NISV extruded through a 800 nm pore size membrane Group 2: 10 μg OVA entrapped in NISV extruded through a

400 nm pore size membrane Group 3: 10 μg OVA entrapped in NISV extruded through a

200 nm pore size membrane Group 4: 10 μg OVA entrapped in NISV extruded through a

100 nm pore size membrane Group 5: 10 μg OVA in PBS (control)

Draining inguinal and popliteal lymph nodes were collected from the mice 10-14 days after the treatments. Inguinal and popliteal lymph nodes were aseptically removed and cell suspensions prepared and enumerated as described in Example 1. 100 μl/well aliquots of cell suspension, containing 5xl0 ⁵ cells, were added to 96-well flat-bottomed tissue culture plates (Costar) , followed by 100 μl/well aliquots of Con A (5 μg/ml) or OVA (2000 μg/ml) in triplicate. Cultures were then incubated for 60 hours at 37°C in a 5% C0 ₂ atmosphere after which 150 μl aliquots of cell culture supernatants were removed and stored at -70°C for cytokine assay.

Cytokines (IL-2, IL-5 and IFN-γ) were detected by monospecific ELISA. Flat-bottomed polystyrene plates (Dynatech) were coated overnight at 4°C with 50 μl/well anti-mouse cytokine monoclonal antibodies at optimum concentrations (determined at 2 μg/ml in each case; Pharmingen) . Plates were washed three times with PBST and blocked with 200 μl/well 10% (v/v) FCS in PBS for 60 minutes at 37°C. Plates were washed three times in PBST, 100 μl samples of supernatants and standards (IL- 2, 0-62.5 units/ml; IL-5, 0-1.6 ng/ml; IFN-γ, 0-70 units/ml) added in duplicate to the wells and incubated for 2 hours at 37°C. Detection of cytokine levels using an alkaline phosphatase-streptavidin conjugate and pNPP substrate was exactly as described in the cytokine assays in Example 1.

RESULTS

Other experiments using photon correlation spectroscopy (data not shown) confirmed that the actual size of vesicles extruded through a polycarbonate filter of pore diameter 200 nm is in the range 152-157 nm, and that of vesicles extruded through a 400 nm pore diameter membrane within the range 200-242 nm.

Figure 6a shows the levels of IL-2 and IL-5 production in Con A stimulated lymph node cells . Figure 6b shows the levels of IL-5 and IFN-γ production in Con A and OVA stimulated, respectively, lymph node cells. NISV extruded through 400 nm membranes elicited the largest amounts of IL-2 and IFN-γ, whilst reducing the levels of IL-5 to below (but not significantly) those achieved in the control OVA group. NISV extruded through 800 nm membranes produced levels of IL-5 that were similar to those achieved with OVA alone. NISV extruded through 100 & 200 nm membranes produced significantly greater levels of IL-5 than those elicited after the administration of NISV extruded through 400 & 800 nm membranes or the OVA control .

The size of the NISV used in a therapeutic application may be important as the ability to act as an immunomodulator appears to be influenced by this parameter. In the system described in this Example NISV of of mean diameter greater than 200 nm are particularly well suited to generate a good therapeutic effect on the down regulation of IL-5 and, also by implication IL-4 whose production is linked to that of IL-5.

B. The effect of size of NISV on IL-1 levels.

The size of NISV was studied in human cell lines to confirm that larger (i.e. > 200 nm) vesicles may also be

more suitable for therapeutic indications in human cells. Parallel studies were carried out in a cultured human macrophage cell line and in human PBLS, freshly derived from volunteers.

Materials and Methods

NISV were prepared by the method described in Example 1 and extruded through polycarbonate filters (Costar) as described previously.

Human U937 cell line: Cells from the human leucocyte pro-macrophage cell line U937 were maintained in RPMI 1640 culture medium supplemented with 10% (v/v) FCS, 2 mM L-glutamine, 100 units/ml penicillin and 100 μg/ml streptomycin (all Gibco) at 37°C in a 5% C0 ₂ atmosphere. Cells were washed extensively by centrifugation at 1000 x g before activation with various agents in DMEM supplemented with 10% FCS.

Human PBLs: Human PBLs were prepared from healthy volunteers as described in Example 3.

2 ml/well aliquots of either U937 or PBL cells, at a concentration of approximately 2 xlO ⁶ cells/ml, were established in 24-well flat-bottomed tissue culture plates (Costar) . Four groups (3 replicates/group) of each cell type were treated with the following preparations and incubated at 37 ^C C in a 5% C0 ₂ atmosphere:

Group 1: Control (PBS, pH 7.4) Group 2: LPS (40 ng/ml)

Group 3: NISV extruded through a 200 nm pore size membrane (1 mg) ("small")

Group 4: NISV non-extruded (approximately 1.6 μm in diameter) (1 mg) ("normal")

The cells were maintained for 72 hours with aliquots of the cell culture supernatants removed at 24, and 48 hours (PBLs) and 72 hours (U937 cells) post-stimulation and stored at -70°C for subsequent cytokine assay.

The supernatants from both the U937 and PBL cell cultures were assayed by monospecific human ELISA for IL-lα and IL-lβ, (Pharmigen) according to the manufacturers instructions. The ELISA plates were developed using a suitable colorimetric solution and read at 490 nm on a Titertek Multiskan plate reader (Flow Laboratories) . Cytokine concentrations in the cell culture supernatants were determined from appropriate standard curves (regression coefficient r = 0.990 or better) .

RESULTS

Figures 7a and 7b show the levels of IL-lα and IL-lβ, respectively. As expected, LPS stimulated the release of IL-1 from both types of cells between 24 and 72 hours, except that U937-release of IL-lα appeared to be refractory to LPS stimulation. Small, approximately 160 nm, NISV elicited the release of low levels of IL-lα from U937 cells and human PBLs. However, larger, approximately 1.6 μm, vesicles did not stimulate the release of IL-lα to any detectable extent.

Small, approximately 160 μm, NISV elicited significant levels of IL-lβ in both U937 and human PBL cells. The levels of this cytokine elicited by small NISV are similar to those generated after the stimulation of these cells with LPS. Larger, approximately 1.6 μm, NISV elicited levels of IL-lβ similar to those observed in the control PBS group.

The results exhibited after the stimulation of human U937 pro-macrophages and human PBLs with small and large

NISV correlate with the results described for murine in vivo systems. NISV greater than approx. 200 nm in mean diameter have a particularly good therapeutic profile for immunomodulation in human cells.

Example 6

Suppression of LPS induced cytokine production by NISV, administered by subcutaneous injection in mice

This study was carried out to demonstrate that the ability of NISV to down regulate LPS-induced cytokine production in vi tro was retained in a comparable in vivo test system.

Materials and Methods

NISV were prepared from cholesterol, 1-monopalmitoyl glycerol and dicetyl phosphate in a ratio 4:5:1 by weight. These components were melted together at 135°C, diluted with PBS to a final volume of 750 ml and a concentration of 25 mg/ml at 65 ^β C as in Example 1 then homogenised at 8000 rpm for 30 minutes and sterilised by autoclaving. NISV were administered to groups of BALB/c mice (5 in each group) by subcutaneous injection at one of two dose levels 2.5 mg (equivalent to approximately 83 mg/kg bodyweight) or 0.5 mg (equivalent to approximately 17 mg/kg bodyweight) . Three groups of animals received each dose and a further two groups were left untreated.

LPS was administered by interperitoneal injection (4 μg in 200 μl) at different time points after NISV injection.

Treatment Group Day of LPS challenge (s) (n=5) in relation to NISV dose 1st 2nd

1 - untreated control

2 - 0.5 mg NISV +1

3 - 0.5 mg NISV +4

4 - 0.5 mg NISV +14

5 - untreated control

6 - 2.5 mg NISV +1 +15

7 - 2.5 mg NISV +4 +18

8 - 2.5 mg NISV +14 +28

Animals allocated to 2.5 mg NISV received two LPS challenges at 14 days intervals, whereas those allocated to 0.5 mg NISV received one LPS challenge. The respective control groups were similarly challenged.

Blood samples were taken for measurement of IL-6 levels at 3 and 6 hours after each LPS challenge. The samples were analysed using 'Quantikine' ELISA kits (R+D

Systems, Abingdon, Oxon) according to the manufacturer's instructions .

RESULTS

The results are presented in Figures 8, 9 and 10. As expected, the control group showed elevation of serum IL-6 levels post LPS challenge, the highest levels occurring at the 3 hour post-challenge time point. Both dose levels of NISV partially inhibited the increase in IL-6 levels.

Fig. 8 shows the effect of NISV at 17 mg/kg bodyweight. The suppressive effect was greatest at 1 day post NISV and lasted for at least 4 days (IL-6 levels at 14 days post NISV were comparable to control levels) .

Fig. 9 shows the effect of NISV at 83 mg/kg bodyweight. At this higher dose the onset of effect appeared slower and was greatest in the group that received NISV 14 days prior to LPS challenge.

Fig. 10 shows the response to the second LPS challenge in the high dose NISV groups. The inhibition of IL-6 response to LPS appears maximal in the group challenged 15 days after NISV administration, but persists in part for 28 days.

These results demonstrate that subcutaneous injection of NISV inhibits the cytokine response to a noxious stimulus (LPS) in vivo and that the duration of action is dose dependent, suggesting a 'depot' effect when administered by this route.

Example 7

Reduction in the levels of TNFα and IL-6 in LPS stimulated human PBLs

In a volunteer study a preliminary evaluation of cytokine production ex vivo was undertaken using PBLs stimulated with LPS.

Materials and Methods

Pre and post dose (+ 24 hr) venous blood samples were taken from a volunteer who received a subcutaneous injection of 25 mg NISV (prepared as in Example 6) suspended in 1 ml of phosphate buffered saline. Peripheral blood mononuclear leucocytes were separated from the heparinised sample by density dependent centrifugation using a Ficoll/Paque gradient and the leucocyte layer further processed as described in Example 3 to produce a cell suspension in Dulbecco's

Minimal Essential Medium.

Aliquots from this suspension (containing approximately 2xl0 ⁶ cells/ml) were established in 24 well flat bottomed tissue culture plates. Two sets of cells (3 replicates/ set) from each sample were treated as follows and incubated at 37°C for 48 hours.

Set 1 Unstimulated cell culture Set 2 LPS (40 ng/ml) stimulated cell culture

Aliquots of the cell culture supernatant were removed at 0, 1.5, 4, 24 and 48 hours of culture and stored at -70°C until assayed for IL-6 and TNFα using 'Quantikine' kits (R&D Systems, Abingdon, Oxon) . Cytokine concentrations were determined from standard curves based on results from assay standards.

RESULTS

The results are shown in Figures 11 and 12. Increases in both TNFα and IL-6 levels occurred irrespective of LPS stimulation in the pre-NISV dose samples. However, in the volunteer, a demonstrable difference was observed in comparing PBLs obtained prior to and subsequent to NISV administration. A suppression of TNFα levels was seen in PBLs extracted at both 4 and 24 hours post NISV administration (see Figure 11) and suppression of IL-6 levels was seen in PBLs extracted at 24 hours post NISV administration (see Figure 12) .

These preliminary data suggest that subcutaneous injection of NISV suppresses the responsiveness of human PBLs to LPS challenge.