VACCINES This invention relates to liposomes, a process for their preparation, and pharmaceutical compositions containing them. It is known that while a variety of inactivated viruses are good immunogens they are also pyrogenic which presents a.serious disadvantage to their use as vaccines. One example of this are current influenza vaccines. These are composed of whole virus and suffer from problems of pyrogenicity as well as sensitization to egg proteins. An alternative is to use influenza vaccines composed of virus sub-units but these are poorly immunogenic and stimulate poor protection compared to live infection.

Generally, the poorest responses to influenza vaccines are observed in elderly patients who are most at risk from complications and death following infection with influenza. In addition to these problems, influenza vaccines are unpopular aε they are conceived to be ineffective and because of fear of injections. GB-A-1564500 discloses antigenic preparations containing a plurality of unila ellar microvesicles, otherwise known as virosomes, each microvesicle comprising a single lipid bilayer upon the exterior surface of which is bound an antigenic protein derived from a virus. GB-A- 1564500 is related to two U.S. Continuation-in-Part Patents, US-A-4196191 and US-A-4148876. US-A-4148876 discloses antigenic virosome preparations of the type disclosed in GB- A-1564500 in which the antigenic protein is bound by hydrophobic bonding and iε a haemagglutinin and neuraminidase sub-unit of a protective surface antigen derived from a myxovirus and having a hydrophobic region. We have now found that influenza virosomes which comprise reconstituted virus envelopes and which have been treated to inactivate neuraminidase are highly immunogenic when administered intranasally. Significant IgA responses were observed in the lung lavage fluid of mice immunised

intranasally but not parenterally. These findings have general applicability. Accordingly, the present invention provides liposomes which have present on their surfaces a polypeptide capable of binding to a mucosal cell surface of a human or animal and which are substantially free of active neuraminidase. The liposomes are typically virosomes.

Liposomes are lipid vesicles enclosing a three- dimensional space. Envelope viruses comprise a lipid envelope. Liposomeε according to the present invention may therefore be made of the lipid of an envelope virus. The virus envelope may be reconstituted after an envelope virus has been disrupted, for example by a detergent, thereby to form liposomes.

Useful liposomes may also be made of natural or synthetic phosphocholine-containing lipids having one fatty acid chain of from 12 to 20 carbon atoms and one fatty acid claim of at leaεt 8 carbon atoms, for example 12 to 20 carbon atoms. Such lipids include dimyriεtoylphosphatidyl- choline, dioleoylphosphatidylcholine, dipalmitoylphos- phatidylcholine, dipalmitoylphosphatidylglycerol, diste ^' aroylphosphatidylcholine, phosphatidylcholine, phosphatidylserine and sphingomyelin. Another lipid may also be included in the liposomes, for example cholesterol, which is preferably present as less than 30% w/w of the whole lipid composition. The lipids may further comprise a material to provide a positive or negative charge, such aε phosphatidic acid, dicetyl phosphate, phosphatidyl serine or phosphatidyl inositol to provide a negative charge or stearyl amine or other primary amines to provide a positive charge.

The liposomes used in the present invention may be either unilamellar or multilamellar, preferably unilamellar. They are typically biodegradable. The lipid of which they are composed is generally non-antigenic. The liposomes may encapεulate a substance, for example an antibody, antigen or drug. They may therefore be used as a delivery system for

the enca ila-eed component. The liposomes can be used as a general delivery system.

Typically the environment within the liposomes is an aqueous environment. A variety of substances can be encapsulated within the liposomes, such as peptides, proteins or adjuvants. The substance may be a substance against which it is wished to induce an immune response. Substances which may be encapsulated include antigenic subunits prepared from many types of virus such as herpes simplex virus, hepatitis A virus and hepatitis B virus.

Proteins or peptides containing class 1 T-cell epitopes may be used. Encapsulation of this material within virosomes may help to generate a cytotoxic T-cell response against them. The liposo eε are preferably in a form which iε εuitable for intranasal administration. Preferably, therefore, the mucosal cell surface-binding polypeptide imparts on the liposomes the ability to bind to the nasal mucosa or to the mucoεa of the lungs. Preferably the diameter of the liposomes is from 5 to lOOOnm, for example 10 to 400 n and most preferably from 20 to 100 nm.

The polypeptide capable of binding to a mucosal cell surface may be glycosylated or unglycosylated. The polypeptide may therefore be in the form of a glycoprotein. Preferably the polypeptide renders the liposomes fuεogenic so that they are able to fuse with, rather simply bind to, host cell membranes. These membranes may be either the outer membrane of the membrane of endosomes following endocytosis. The polypeptide is typically a virus envelope polypeptide or is derived from a virus envelope polypeptide. All envelope viruses have a surface-binding function. The polypeptide may therefore be a polypeptide which is naturally present on the surface of an envelope virus and which provides the liposomes with the capability of binding to a cell surface. The virus may be a myxovirus such as influenza, mumps or measles virus. In particular the

polypeptide may be or be derived from an influenza virus envelope protein, for example of influenza virus type A, B or C.

The polypeptide capable of binding to a cell surface may for example be a haemagglutinin. Haemagglutinin is an integral membrane glycoprotein present in myxoviruseε which is commonly composed of three monomers or sub-units. During infection of host cells, it serves two functions. Firstly, it attaches the virus to the cell by the binding of sialic acid residueε present on cellular glycoproteins and glycolipidε. Second, after internalization of virus into cellular endosomeε the εubεequent acidification triggers conformational changes in the haemagglutinin which lead to the fusion of viral and cellular membranes. Haemagglutinins are antigenic and stimulate the production of antibodies in hoεtε.

Another type of polypeptide capable of binding to a cell surface may be a bacterial adhesive protein such as the 3-εubunit of cholera toxin (CTB) or the heat-labile enterotoxin 3-subunit of E. coli (LTB) . Thiε may alεo be used aε an adjuvant in combination with haemagglutinin. Neuraminidase iε another glycoprotein which iε found aε an integral membrane protein in myxoviruses. Thiε functions to cleave sialic acid residueε and prevent the irreverεible binding of viruε to a host cell membrane by haemagglutinin. If active neuraminidase is present in the liposomes, then a significantly lower immunological response iε obεerved. If active neuraminidaεe would otherwise be present in the liposomes, it must be inactivated. Neuraminidase may be inactivated by heat or by incubation with a neuraminidase inhibitor such as 2,3-dehydro-2-deoxy- N-acetylneuraminic acid (DDAN) .

The present liposomes are prepared by a process which comprises forming liposomes which have present on their surfaces a polypeptide capable of binding to a mucosal cell surface of a human or animal and which are

substantially free of active neuraminidase.

The polypeptide capable of binding to a cell surface may be added to the lipid materials before, during or after formation of the liposomes. Alternatively, virosomes can be prepared using the natural lipid of the envelope of an envelope viruε to provide the necessary lipid component. If the polypeptide does not. naturally associate with lipids it may be coupled to a fatty acid such as phosphatidylethanolamine (PE) by the use of a cross-linking agent such as succinimidyl 4-(p-maleimidophenyl)butyrate (SMPB) .

Liposomes may for example be prepared by dissolving the lipid starting material in a solvent and evaporating the solvent. The lipid layer is then dispersed with aqueous saline or a buffer (if it iε intended to incorporate the polypeptide into the liposomes after vesicle formation) or with an aqueous suspension of the polypeptide (if it is intended to form vesicles in the presence of the polypeptide) . The dispersion is then agitated, for example by εonication. Polypeptide may then be added where it is not already incorporated in the surface of the liposomes and the vesicles again agitated.

An alternative method is to add the lipid starting material to an aqueous phase and slowly heat the mixture. It is then agitated to form liposomes. The aqueous phase may contain the polypeptide or it may be added subsequently.

A further method of preparing liposomes comprises the rapid injection of an ethanolic solution of lipid into aqueous saline or a buffer which has previously been purged with nitrogen. The resulting lipoεome preparation is then concentrated by ultrafiltration with rapid stirring under nitrogen at low presεure to avoid the formation of larger non-heterogeneous liposome. The ethanol may be removed from the vesicle fraction by analyεis or washing with an ultra- filter. The polypeptide may be present in aqueous solution or alternatively the liposome fraction obtained after

ultrafiltration may be lightly sonicated with the polypeptide.

The liposome preparations obtained in the manner described above comprise aqueous dispersions of the lipid 5 vesicles.

If the liposomes comprise neuraminidase then this must be inactivated. This may be achieved by heating the aqueous dispersion of liposomes comprising active neuraminidase to a temperature of for example from 30 to

10 60*C, for example from 50 to 60°C, more preferably from 53 to 58°C and most preferably about 55'C. The length of time required for neuraminidase inactivation will depend on the strain of virus and the temperature but is typically from 5 minutes to 5 hours, for example from 15 minutes to 3 hours.

15 At low temperatures eg. 30'C a longer period of heating iε required, whilεt at higher temperatureε a shorter period is required. We have found for influenza virus that heating at 55°C must be for 120 minutes or more, for example up to 180 minutes, in order to achieve an optimum effect. At

20 56°C, however, the optimal period for heating is from 6 to 10 minutes, for example about 8 minutes.

Alternatively, active neuraminidase may be deactivated by incubation of the liposomeε with a neuraminidase inhibitor such as DDAN. As a further

25 alternative active neuraminidase may be inactivated by heat or incubation with a neuraminidase inhibitor prior to incorporation into the liposomes.

A suitable way of preparing liposomes comprises:

(a) diεrupting a myxoviruε and removing the viral genome 30 and internal viral protein or proteinε; and

(b) forming liposomeε in the presence of the material remaining, especially the envelope protein or proteins; and

(c) inactivating the neuraminidase present in the thus- formed liposomeε.

35 Step (a) may be achieved be detergent solubilisation of viral particles and removal of internal

viral proteins and RNA. In an alternative way of preparing liposomes, th" cell surface-binding polypeptide may be prepared by a recombinant DNA methodology. It will then necessarily be provided free of neuraminidase, so liposomes substantially free of active neuraminidase are necessarily obtained.

The liposomes of the present invention may be administered in the form of a pharmaceutical or veterinary composition which additionally comprises a suitable pharmaceutically or veterinarily acceptable carrier or diluent. The compositions are suitable for administration intranasally.

The compositions are preferably provided in a sterilised form. They may take the form of an aerosol. The compositions may further comprise preservativeε, stabilisers and other conventional vaccine excipients if required.

The dosage of liposomes will vary depending upon a variety of factors. These include the nature of the cell surface-binding protein, the recipient (human or animal) , the vaccination schedule and the extent of adjuvanticity conferred by the preparation. In general a doεe of liposomes may be administered intranasally ^' as a single unit or as a multiplicity of a sub-dosage over a period of time. Typically the unit dose for intranasal delivery to a human iε from 2 to 500 μg.

We have also found that inactivated influenza virus which is substantially free of inactive neuraminidase is highly immunogenic when administered intranasally. This finding also has general applicability. The invention therefore further provides:

- an influenza virus which is not infectious and which is substantially free of active neuraminidase, for use as an influenza virus; and

- use of an influenza virus which is not infectious and which is substantially free of active neuraminidase in the preparation of a medicament for use as an influenza vaccine.

The influenza virus may be any influenza virus, for example type A, B or C. The virus is the virus against which it is wished to vaccinate. The neuraminidase may be inactivated by heating or specific inhibitors. .An aqueous dispersion of the virus may be heated. Heating may be carried out at a temperature of for example from 30 to 60°C, more preferably from 53 to 58°C and most preferably about 55°C.

The length of time for which heating must be conducted to ensure neuraminidase inactivation will depend upon the strain of virus and the temperature but is typically from 5 minutes to 5 hours, for example from 15 minutes to 3 hours. At low temperatures, e.g. 30°C, a longer period of heating is required than at higher temperatures. We have found that heating at 55°C must be for 120 minutes of more, for example up to 180 minutes, in order to achieve an optimum effect. At 56°C, however, the optimal period for heating iε from 6 to 10 minutes, for example 8 minutes. The influenza virus is inactivated. In particular, viral infectivity is inactivated. Thiε may be achieved by the heating to inactivate the neuraminidaεe. Typically, however, it iε achieved by irradiation with ultraviolet light to provide a fail-safe inactivation procedure. Thiε may be carried out before, simultaneously with or after treatment to inactivate the neuraminidase. Irradiation is carried out for at leaεt 5 minuteε, for example for from 5 to 60 minutes, at 400μW/cm ² at a short wavelength, for example from 240 to 250 nm. The ultraviolet-inactivated, heated virus is grown in the allantoic fluid of embryonated hens eggε, for 2-3 days, recovered and purified on sucrose gradients.

The inactivated influenza virus substantially free of active neuraminidase is administered in the form of a pharmaceutical composition which additionally comprises a suitable pharmaceutically acceptable carrier or diluent.

The compositic are suitable for administration intranasclly.

The c ^positions are preferably provided in a sterilised form. They may take the form of an aerosol. The compositions may further comprise preservatives, stabilisers and other conventional vaccine excipients if required.

. The dosage of inactivated virus will vary depending upon a variety of factors. An effective amount of the inactivated influenza virus substantially free from active neuraminadase is administered to a person in need of vaccination, in particular in need of vaccination against the said virus. Factors which need to be taken into account in assessing dosage include the age of the recipient, the vaccination schedule and the extent of adjuvanticity conferred by the preparation. In general a dose may be administered intranasally as a single unit or as a multiplicity of a sub-dosage over a period of time. Typically the unit dose for intranaεal delivery is from 2 to 500 μg. The vaccines of the invention exhibit advantages over current influenza vaccines. These include immunogenicity, the convenience of intranasal adminiεtration and the production of local mucosal immunity.

The invention will now be further illustrated by means of the following Example. In the accompanying drawingε:

Figure 1A εhows the ELISA titreε against X31 influenza virus in sera from Balb/c mice immunised intranasally (i.n.) with heated and acid-treated virosomes or virus;

Figure IB shows the ELISA titres against denatured virus in sera from mice immunised i.n. with heated and acid- treated virosomes or virus;

Figure 1C shows the neutralisation titres of sera from mice immunised i.n. with heated and acid-treated virosomes or virus;

Figure ID shows the HAI titres against virus in sera from mice immunised i.n. with heated and acid-treated virosomes or virus;

Figure 2A shows the ELISA titres against virus in sera from mice immunised i.n. with heated and acid-treated virus;

Figure 2B shows the neutralisation titres against virus in sera from mice immunised i.n. with heated and acid- treated virus; Figure 3A shows the ELISA titres against virus in sera from mice immunised i.n. with heated and acid-treated viroso eε;

Figure 3B shows the neutraliεation titreε against virus in sera from mice immunised i.n. with heated and acid- treated virosomeε;

Figure 4A shows the ELISA titres against virus in sera from mice immunised i.n. with heated and acid-treated viroεomeε encapεulating the internal proteins and RNA ("virosome-coreε") ; Figure 4B shows the neutralisation titres against virus in sera from mice immunised i.n. with heated and acid- treated viroso e-cores;

Figure 5A shows the ELISA titres against viruε in εera from mice im uniεed i.n. with heated and acid-treated viroεomeε encapεulating ovalbumin ("ova-virosomes") ;

Figure 5B εhowε the neutraliεation titres against virus in sera from mice immunised i.n. with heated and acid- treated ova-virosomes;

Figure 6A showε the individual ELISA titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated and acid-treated virus;

Figure 6B shows the individual neutralisation titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated and acid-treated virus;

Figure 7A showε the individual ELISA titres against

viruε in sera (2 de ~ post-challenge bleed) from mice immunised i.n. with neated and acid-treated virosomes; Figure 7B shows the individual neutralisation titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated and acid-treated virosomes;

Figure 8A shows the .individual ELISA titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated and acid-treated virosome cores; Figure 8B shows the individual neutralisation titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated/acid-treated virosome- cores;

Figure 9A shows the individual ELISA titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated and acid-treated ova-virosomeε;

Figure 9B shows the individual neutralisation titres against virus in sera (2 days post-challenge bleed) from mice immunised i.n. with heated/acid-treated ova- virosomes;

Figure 10A shows ELISA titres showing the effect of heating on the immunogenicity of virosomeε administered i.n. ;

Figure 10B shows the neutralisation titres showing the effect of heating on the immunogenicity of virosomes administered i.n.;

Figure 11 compares the anti-virus and neutralising antibody response following immunisation with heated virosomes. Figure 12A shows the effect of heating at 55°C on the immunogenicity of virosomes administered i.n.;

Figure 12B shows the effect of heating at 55°C on the immunogenicity of virosomes administered i.n.; Figure 13 shows the effect of pre-treating virosomes or mice with gangliosides on the immunogenicity of virosomes given i.n. ;

Figure 14A shows the ELISA results of sera from mice immunised i.n. on days 0 and 43 with different doseε of influenza viroεomeε; and

Figure 14B shows the neutralisation reεultε of εera from mice immunised i.n. on days 0 and 43 with different doses of influenza virosomeε.

EXAMPLE

1. METHODS

Preparation of viroεomes The procedure for making the reconstituted virus envelopes was similar to that described by Metsikko et al. (EMBO J. 5., 3429-3435, 1986) and Stegmann et al. (EMBO J. 6., 2651-2659, 1987) . A pellet of X31 influenza viruε (5mg) waε solubilised in 0.7ml of lOOmM octaethyleneglycol monododecylether (C ₁₂ Es) ---■ dialyεiε buffer (145mM NaCl, 5mM Hepes, pH 7.4) for 20 min at room temperature. The mixture was centrifuged at 170,000g from 30 min to remove the internal proteins and RNA. 0.56ml of the supernatant was added to 160mg of wet Bio-Beads SM-2 and shaken on a rotating table (approx. 400 rpm) for 1 hour at room temperature. The supernatant was removed from the beads with a 23g needle attached to a 1ml syringe and added to 80mg of wet Bio-Beads SM-2 and shaken on a rotating table (approximately 500-600rpm) for 8 min yielding a turbid suεpenεion. The εupernatant waε removed with a 23g needle and syringe. The virosomeε were εeparated from unincorporated protein by diεcontinuouε sucrose gradients (40%/5% or 40%/20%/5%) spun at 170,000g for 90min. The morphology of the virosomeε waε analysed by electron microscopy using negative staining with phosphotungstate.

Viroεomes containing encapsulated proteins, e.g. ovalbumin (Virosomeε + ova) , were made as described above except that lOOμl of 200 mg/ml ovalbumin was added prior to adding the SM-2 beads. Virosomeε-coreε were made as described above except that the interal proteins and RNA

were not removed by centrifugation.

ELISA assays

Anti-virus antibodies in the serum from vaccinated mice were measured by ELISA (enzyme-linked immunoadsorbent assay) . The virus antigen was diluted in carbonate coating buffer pH 9.5: 1/50 dilution of allantoic fluid from hens eggs inoculated with virus or lμg/ml of purified egg-grown virus. Microtitre plates were coated with antigen and left at 37°C for 1 hour and then overnight at 4°C. After washing the plates 3 times in 0.05% Tween 20 in PBS lOOμl of 1% BSA was added and left at 37°C for 1 hour to block the plates. The antisera to be tested was diluted down or across the plate in doubling or half log dilutions in 1% BSA in PBS and left at 4°C overnight. The plates were washed with Tween/PBS before adding the enzyme-conjugated second antibody at 1/500-1/1000 in 1% BSA in PBS. The plates were left at 37°C for 2 hours and washed in Tween/PBS. The subεtrate, o-phenylenediamine dihydrochloride (OPD) (lOmg/lOO l) in citrate buffer with 0.01% H ₂ 0 ₂ waε added to the plates and the reaction stopped in H2SO4. The plates were read on a microplate reader at 492nm. The titreε were end point titreε determined by taking the titre at which the OD value waε equal to the mean OD value obtained with 1/10 dilution of control normal sera plus 2 standard deviations.

In Vitro Neutralisation Assay

We have established a microtitre plate-based neutralisation assay on MDCK cells. Serial dilutions of antibody were incubated with 2 logs of virus for 1 hour at 37°C. These were transferred to microtitre plates with 70- 90% confluent MDCK cells in MEM media without serum. After incubation at 37°C for 1 hour the supernatant was removed and fresh MEM added with lOμg/ml trypsin. The plates were stained after 48-72 hours and the neutralisation titres read by eye.

HAI - Haemagglutination inhibition assay

Haemagglutination and haemagglutination inhibition asεayε were performed aε described by Fazekas de St. Groth and Webster, (J. Exp. Med. 124.; 331-345, 1966).

Experiments

Experiments were carried out as follows, referring to the Figures:

Fiσure 1

Dose - 5μg of X31 viruε/viroεomeε per mouεe in 30μl volume given i.n.

All virosomes were uv. inactivated for 5 min (400μW/cm ² )

Heating - heating carried out at 55°C for 20 min.

Acid treatment - 1/lOOth volume of 3M acetate buffer pH 4.8 was added to the virosomeε. Theεe were left at 37°C for 15 min before neutraliεing the acid with IM Tris pH 7.5.

Second immunization - 6 weeks

Bleed tested - 12 weeks

Figures 2-5

Doεe - 5μg of X31 virus/virosomeε per mouεe in 30μl volume given i.n.

Viroεomeε were uv inactivated for 5 min (400μW/cm ² ) Heating - heating carried out at 55°C for 20 min. Acid treatment - l/100th volume of 3M acetate buffer pH 4.8 waε added to the viroεomeε. Theεe were left at 37°C for 15 min before neutraliεing the acid with IM Triε pH 7.5. Second immunization - 6 weekε

Figures 6-9

Dose - 5 μg of X31 virus/virosomes per mouse in 30μl volume given i.n. Virosomes were uv inactivated for 5 min (400μW/cm ² )

Heating - heating carried out at 55°C for 20 min.

Acid treatment - l/lOOth volume of 3M acetate buffer pH 4.8 was added to the virosomes. These were left at 37°C for 15 min before neutralising the acid with IM Tris pH 7.5. Second immunization - 6 weeks Bleed tested - 12 weeks

Figure 10

Dose - 3 μg of X31 virosomes per mouse in 30μl volume given i.n.

Virosomes were uv inactivated for 5 min (400μW/cm ² ) Heating - heating carried out at 55°C for specified times.

Second immunization - 6 weeks

Figure 11

Dose - 3 μg of X31 virosomes per mouse in 30μl volume given i.n. Viroεomes were uv inactivated for 5 min (400μW/cm ² )

Heating - heating carried out at 55°C for specified times.

Second immunization - 6 weeks

Bleed tested - 8 weeks

Figure 12 Dose - 3μg of X31 viroεomes per mouse in 30μl volume given i.n.

Virosomeε were uv inactivated for 5 min (400μW/cm ² )

Heating - heating carried out at 55°C for εpecified times.

Second immunization - 6 weeks Bleed tested - 12 weeks

Figure 13 preincubation with gangliosides and antibody

The virosomes were dialysed against Hepes buffer (145mM NaCl, 5mM Hepes pH 7.4 plus 3mM EDTA) . These virosomes were heated at 55°C for 1 hour. Dose - 3μg/mouse in 30μl volume.

Incubation with gangliosides - Virosomes were incubated at

37°C for 1 hour and then overnight at 4°C with a 12 Molar excesε of ganglioεides to viral haemagglutinin.

Pretreatment of mice with gangliosides - 100 Molar excess of gangliosideε to viral haemagglutinin.

Incubation with antibody - 20μg of virosomes were incubated in 40μg of purified HC2 antibody or 20μg HC2 Fab fragments for 2 hours at 37°C

Administration of viroεomes with CTB

2μg of CTB (B-subunit of cholera toxin) was given together with 3μg of virosomes to each mouse.

Second immunization - 6 weeks Bleed tested - 12 weeks

Treatment with DDAN

20μg of virosomeε were incubated with ImM DDAN for 1 hour at 37°C and then at 4°C overnight, doεe per mouεe = 3μg in 30μl volume i.n. Second immunization - 6 weekε Bleed teεted - 12 weekε

Figure 14 CDose response) Doεe - variable doεe of X31 virosomes in 30μl volume given i.n.

Virosomeε were uv inactivated for 5 min Heating - heating carried out at 55°C for specified times. Second immunization - 6 weeks

2. RESULTS

Effect of acid-treatment on the immunogenicity of virus and viroεomeε

Influenza viruε, influenza virosomes, or influenza virosomes containing cores (HBcAg) or ovalbumin were treated

with acid (pH 4.8) for 30 min. at 37 ' C . Acid-treatment of virus or virosomes led to a dramatic reduction in immunogenicity of virus or virosomes given by the intranasal route aε assayed by serum ELISA titres against native virus (Figures 1A and 2A-5A) . This was not due to the fact that acid destroys some of the neutralisation epitopes on haemagglutinin because lower responses were also observed when the sera were tested against SDS-denatured virus (Figure IB) . In addition, the levels of neutralising antibodies induced were considerably reduced if the inoculum was acid-treated (Figure 1C and 2B-5B) .

There appeared to be some protection against acid- inactivation of virosomes containing cores but this may be due to insufficient acidification of the boost inoculum (see Figure 4A & B) . When the responεe of individual animalε was analysed there was a consistent reduction in response if the virus or virosomes were acid treated (Figures 6 - 9) . We also looked at the haemagglutination inhibition (HI) activity of the sera (Figure ID) , which also show a reduction in the titre of antibody stimulated when virosomeε were acid treated before inoculation.

Acid-treatment (pH 4.8) of virus abrogates the ability of virus to fuse with cells while virus attachment is unaffected. This iε due to the irreverεible conformational shift in the conformation of haemagglutinin that normally occurs inside the endosome after uptake of the viruε within coated pitε. Theεe results suggest that the virus or virosomeε muεt not only bind to the mucosal surfaceε but also fuse with the epithelial cells to stimulate optimal responseε.

Effect of heating at 55°C on the immunogenicity of viruε Mice inoculated intranasally with X31 influenza virus heated for 20 min. showed significantly greater serum ELISA, HAI or neutralising antibody titres than mice receiving unheated virus (Figures 1A-D) . Both the ELISA

titres against native virus and the neutralising titres were approaching those observed following immunization with the same dose of infectious virus (Figure 2) . In a further experiment virus was heated for only 8 ins. and again this led to an increase in response following intranasal inoculation (Table 1) .

Responses to both the heated or infectious virus were observed at least 21 days before responses to inactivated virus. When the response of individual animals was examined there was an increase in response when the virus was heated and the ELISA titres paralleled the neutralising titres (Figure 6). There was, however, considerable variation probably due to the efficiency of inoculation.

Effect of heating at 55°C on the immunogenicity of viroεomeε In a preliminary experiment, mice were inoculated intranaεally with viroεomeε, or viroεomeε containing cores or ovalbumin, that had been heated for 20 min. These heated virosomes stimulated comparable or greater serum ELISA or neutralising titreε than mice receiving unheated virosomeε (Figureε 1-5) . Figure 4 shows that uv-inactivated, heated viroεomes containing viral cores stimulate a much earlier responεe than uv-inactivated or acid-treated viroεomeε. In addition, when the reεponse of individual animals was examined there was little variation within the animal groups (Figure 7) . The neutralisation titres showed greater variation but parallelled the ELISA titres. It should be noted that these virosomes were stored at 4°C, so it is possible that much of the neuraminidase activity was lost. Mice were immunized with fresh virosomeε that were stored in 50% glycerol at -20"C. A dramatic increase in immunogenicity was observed if the virosomes were heated for up to 128 mins. (Figures 10A and 11, Table 1). The levels of neutralizing antibodies showed a more dramatic increase with increaεing periodε of heating (Figures 10B and 11) .

Animals immunized with unheated virosomes had undetectable levels of neutralising antibodies suggesting that the high responses observed in the previous experiments with unheated virosomes were due to partial inactivation of neuraminidase activity during storage at 4°C. In addition, when serum from individual mice were analyzed a significant increase in ELISA and neutralising antibody titre was observed in sera from mice receiving virosomes heated for increasing periods of time (Table 2, Figure 12).

Effect of Specific Inactivation of Neuraminidase on the Immunogenicity of Virosomes Given Intranasally

We have carried out an experiment to determine whether the increase in immunogenicity observed with heating of virosomes was due to inhibition of neuraminidase (NA) . Thus, the NA in virosomes was specifically inactivated with the neuraminic acid analogue, DDAN. DDAN-treated virosomes stimulated a greater response than untreated virosomes (log titre of 2.2 cf 1.6) showing that inhibition of neuraminidase leads to an increase in immunogenicity of intranasally administered virosomes.

Effect of Specific Blocking of Virosome Attachment by Pre- incubation with Gangliosides on the Immunogenicity of Virosomes given Intranasally In order to study the effect of blocking virosome attachment on the immunogenicity of intranasally administered virosomes we have pre-incubated virosomes in various sialic acid-containing gangliosideε. The immunogenicity of influenza virosomes administered intranasally (i.n.) could be partially abrogated by pre- treating the virosomes with GMl or GDla gangliosides but not by pre-treating with GTlb or a ganglioside mixture (Figure 13). Similarly, we have found that the virosome-mediated haemagglutination waε partially inhibited only by the GMl and GBla gangliosides. These experiments show that binding

of virosomeε to sialic acid receptorε is critical for their immunogenicity.

Effect of pre-treating mice with gangliosideε on the reεponεe of mice to virosomes given intranasally We have studied the effect on the response to virosomeε of increaεing the viral receptors on the respiratory mucosal surfaces through intranasal pre- treatment of mice with various gangliosides (Figure 13) . Pre-treat ent with GMl ganglioside but not GDla GTlb or a ganglioεide mixture led to an increase in response presumably because of an increase in density of receptors or replacement with higher affinity receptors on the mucosal εurfaces facilitating greater binding and uptake of the virosomeε.

Effect of Specific Blocking of Viroεo e Attachment by Pre- incubation with Neutralising Monoclonal Antibodieε on the Immunogenicity of Virosomes Given Intranasally

Pre-incubation of virosomes with a neutralising monoclonal antibody (either whole or Fab fragmentε) completely inhibited haemagglutination. However, the immunogenicity of the virosomes was unaffected by prior incubation of virosomeε in whole antibody or Fab fragmentε or i.n. inoculation of Fab fragmentε 2 hourε after inoculation with untreated viroεomes. Thiε reεult is surprising and may indicate that some neutralising antibodies do not prevent virus binding or entry into mucosal epithelia but some later event (e.g. secondary uncoating) . Studying the fate of antibody-treated virus or virosomes should provide some inεight into the mechanisms of humoral immunity in the respiratory tract. Moreover, with regards intranasal vaccination of humans, it is encouraging if the presence of local neutralising antibody fails to reduce the immunogenicity of intranasally-administered virosomes.

Effect of Administering the B-subunit of Cholera Toxin (CT- B. Intranasally Together with Virosomes on the Immunogenicity of the Virosomes

We investigated whether CTB could enhance the responses to virosomes administered intranasally. Figure 13 shows that there was a dramatic increase (10-fold) in response to the virosomes when given with CTB) .

Dose response study of virosomes administered intranasally

We have studied the response to a range of doses of virosomes administered intranasally. These virosomes were unheated and fresh so the antibody responses are relatively low. A clear dose response effect was observed with the minimal immunogenic dose being lμg (Figure 14) . We have repeated this experiment using heated virosomes which we would expect to be much more immunogenic.

Protection against Challenge

All the animals receiving intranasal immunisations of virus or virosomes have been challenged with live virus and the lungs were removed 2 days later. Preliminary challenge lung titres indicate that animals immunised with virus or virosomes that were not acid-treated were completely protected against infection.

Table 1 - INTRANASAL IMMUNISATIO W H

Table 2 - INTRANASAL IMMUNISATION OF BALB/C MICE WITH INFLUENZA VIROSOMES - RESPONSE OF INDIVIDUAL MICE Effect of heating at 55'C on Immunogenicity

ELISA TITRES MEAN TITRE

Group Antigen Heat Animal NV . E DV 63A VIROSOMES Not heated G 1.21 1.23 Y 1.98 1.89 M 1.26 1.14 1.84 1.44

W 1.48 1.23 R 1.31 1.10

63B VIROSOMES heated 2min M <1.00 <1.00 W 1.38 1.02 G 2.02 1.38 1.64 1.24 R 1.24 1.22 Y 1.78 1.42

ELISA TITRES MEAN TITRE

Group Antigen Heat Animal NV DV NV DV 63C VIROSOMES heated 8min R Y G 1.95 1.69

M

63D VIROSOMES heated 32min M Y R 2.12 2.04 G W

63E VIROSOMES heated 128min R Y

M 2.31 2.46 W G

Bleed tested: 30/3/90 - 4 weeks after second immunisation.