PREPARATION OF ' IMMUNOSTIMULATING COMPLEXES . ( ISCOM ) CONTAINING BOVINE HERPESVIRUS 1 PROTEINS

Franz J., J. Hampl, J. Steplinek, B. Smid: Preparation of ImmunostimuJ.ating Complexes (ISCOM) Containing BOfJine HerpesfJirus 1 Proteins. Acta vet. Bmo, 61,1992: 37-41. A method for obtaining lSCOMs with incorporated bovine herpesvirus 1 (BHV-l) proteins from various amounts of cholesterol, phosphatidylcholine and Quil A is described. The highest virus protein incorporation rate obtained was 25 %. Morphology of ISCOMs depended on amounts of Quil A, cholesterol and phosphatidylcholine used. A high immunogenicity ofBHV-I-ISCOM, as compared with free virusproteins, was confirmed experimentally in mice. BOfJine herpesfJirusBHV-1, fJirus protein, ISCOM, antibody response One of the essential means for the control of virus diseases offarm animals are vaccines, capable of in· ducing specific resistance against the causal agent in individual animals and animal populations. Current technologies use suspensions of complete virus particles propagated in cell cultures or laboratory animals. In addition to the specific immunogenic antigen, such suspensions contain a number of other proteins,originsting from culture media, disintegrated cells or even microbial contaminants. Such a broad antigenic spectrum, reaching far beyond 1ihe purpose of vaccination, is unwanted and therefore alternative methods for producing quality biologicals are developed currently. Progressive trends in the control of virus diseases are focussed on the development of subunit vaccines. Protein subunits are prepared by highly effective purification procedures or by biotechnological methods, such as chemical synthesis of peptides or DNA recombination. However, the development of subunit vaccines is complicated by the rather low antigenicity of separated immunostimulants or adjuvants (Liu and Cepica 1990; Rowland 1986; Zanetti et aI. 1987; Dougan 1985). On the other hand, procedures involving incorporation of virus subunits with specific antigenic determinants into submicroscopical structures, generally called immunostimulating complexes (ISCOMs), appear to be prospective (Morein et al. 1984). The spatial arrangement of ISCOMs imitates that of small virus particles, the high immunogenicity of which has been confirmed repeatedly (Trudel et aI. 1987; Morein et a1. 1987; Berezin et aI. 1988; Merza et al. 1988; Cook eta!. 1990; Lovgren and Morein, 1991; Thapar et a1. 1991). The aim of our experiments was to develop an efficient procedure for the incorporation of bovine hC11'esvirus 1 (BHV-l) protein subunits into ISCOM and to examine the morphology and immunogenicity of the complexes formed. Materials and Methods Virus propagation and purification Strain Los Angeles of BHV-l, the causal agent of infectious bovine rhinotracheitis, was propagated in primary calf kidney cell cultures. The virus was harvested after repeated freezing-thawing as soon as CPE became apparent, i. e. after 2 3 days of incubation. Cell debris was removed by low-speed centrifugation, the infective culture fluid was concentrated by ultrafiltration and the concentrated suspension was purified by ultracentrifugation (Beckman L8-80 M) in potassium tartrate density gradient (Trudel et a1. 1987). Virus-containing fractions were pooled and centri-


Virus propagation and purification
Strain Los Angeles of BHV-l, the causal agent of infectious bovine rhinotracheitis, was propagated in primary calf kidney cell cultures.The virus was harvested after repeated freezing-thawing as soon as CPE became apparent, i. e. after 2 -3 days of incubation.Cell debris was removed by low-speed centrifugation, the infective culture fluid was concentrated by ultrafiltration and the concentrated suspension was purified by ultracentrifugation (Beckman L8-80 M) in potassium tartrate density gradient (Trudel et a1. 1987).Virus-containing fractions were pooled and centri-fugated and the sediment was resuspended in TEN buffer.The virus content was checked by • titration in cell cultures and by electron microscopy.

Virus protein subunits
The suspension of the purified virus was incubated with the non-ionogenic detergent 1-0-n-octyl--glucopyranoside.Nucleocapsids were separated by ultracentrifugation in saccharose density gradient and the protein content in the supernatant was determined by Lowry's (1951) method.

ISCOM
Quil A, cholesterol (C) and phosphatidylcholine (PC) were added in various weight proportions to the superuatant containing saccharose and virus protein subunits and the mixture was dialyzed against a large volume of 0.1 M ammonium acetate (pH 7.0) for 48 hours.The ISCOMs formed were separated into the sediment by 4 hours ultracentrifugation at 40 000 r. p. m. using the rotor SW 55Ti.The rate of protein incorporation into ISCOM was determined by the addition of a standard amount of 126I-Iabelled homologous protein per 500 pg virus protein subunits.0.0 pg, 50 pg, 250 pg, 500 pg, 1 000 pg or 1 500 pg of C and PC, and 2 000, pg or -in one case -4 000 pg Quil A were used in individual incorporation experiments.

• Electron microscopy
The morphology of ISCOMs was examined by electron microscopy.Negative staining with 2 % ammonium molybdate (pH 6.8-7.0) was used.

Immunization
Immunogenicity of BHV-I-ISCOMs, prepared under various conditions, was examined in Balb/c mice.Eight groups oflO-16 mice were inoculated subcutaneously with 1.0 pg BHV -1 protein incorporated into ISCOM and the same dose was repeated on Day 21 p. i. Blood sera for the determination of antibodies to BHV-I were obtained by successive exsanguinations of mice 21, 35 and 49 days after the first treatment.Mice in the control group were inoculated and reinoculated with 10 pg of free virus protein.

Antibody determination by ELISA
The indirect method with complete, solid phase-bound BHV-l as antigen was used.The conjugate was prepared by labelling porcine anti-mouse antibodies, purified by affinite chromatography with horse radish peroxidase.Hydrogen peroxide and tetramethylbenzidine were used as substrate.Blood sera were examined in serial twofold dilutions starting with 1 : 50.The results • were expressed as mean titres of BHV-l antibodies.

Results
Weight proportions of ISCOM components, i. e. virus subunit proteins, Qui! A, C and PC and respective rates of virus protein incorporation are given in Table 1.The highest incorporation rate, i. e. 25 %, at the .constantamounts of 500 ftg virus protein and 2 000 {lg Quil A was reached when 250 ftg C and 250 ftg PC were used.Only 9.8 and 7.8 % of the protein was incorporated into ISCOMs prepared without C and PC and with 50 !Jog C and 50 ftg PC, respectively.
Excess of C and PC (1000 or 1 500 ftg) reduced the incorporation rate to 15 %, but the use of 4 000 I'g Quil A resulted in an increase to 23.3 %.
Electron microscopic examinations confirmed the formation of typical spatial • cage-like structuresin all experiments (Plate XXIII, Fig. 1).The size of ISCOM particles rose from approx.20 to 60 nm when the amounts of C and PC increased (Fig. 2,3,5).Aggregation of particles or formation of extensive ISCOM complexes were observed when 1 000 or 1 500 ftg C and PC were combined with 2 000 ftg Quil A (Plates XXIV., XXV., Fig. 4, 6).The formation of the aggregates and ~omplexes was prevented by increasing the amount of Qui!A to 4 000 ftg (Fig. 5).
A very• weak antibody response was recorded in the control group inoculated with free BHV-J proteins.Antibodies were detectable on Day 49 p. i. only.The most intensive antibody response was recorded in Group 3 (Table 1), treated with ISCOMs with the highest virus "protein incorporation rate.This applied

Discussion
The starting point of our experiments was the information of Morein et al. (1984)3 who described a new type of a subunit vaccine based on the immunostimulating complex -ISCOM -in which the antigen is arranged into a spatial structure formed by an adjuvant and a lipid component.The ISCOM matrix is formed by the glycoside Quil A3 cholesterol and3 as the case may be3 phospha-tidy1choline in optimal proportions.Specific microbial antigens can be incorporated into such matrices by hydrophobic interactions.As confirmed for pseudorabies (Morein et a1. 1989)3 influenza (Lovgren et a1. 1990;Weiss et a1. 1990)3 bovine leucosis (Merza et al. 1989) and BHV-l (Trudel et al. 1988~ Merza et al. 1988) viruses3 Toxoplasma gondii (Overnes et al. 1991) and other causal agents, antigens arranged in such a manner show a high immunogenicity.Envelope proteins of BHV-13 the causal agent of a widespread and serious disease of cattle, were used as a model in our experiments.Preliminary experiments showed a rather low rate of BHV-l protein incorporation into ISCOM (less than 10 %), however.Therefore we attempted to enhance the incorporation by defined amounts of cholesterol and phosphatidy1choline.The incorporation rate reached its maximum of 25 % when 250 I-lg cholesterol3 250 I-lg phosphatidy1choline and 2 000 I-lg Quil A per 500 ttg virus protein were used.Such BHV-I-ISCOMs showed also the highest immunogenicity in mice.Similar results were reported by Lovgren and Morein (1989)3 who stressed that the matrix-forming lipids are essential when highly purified proteins are used as the antigen.
All ISCOMs, prepared by any of the modification of the procedure3 showed the typical cage-like structures3 as described by Ozel et al. (1989)3 e. g.However, elevated amounts of cholesterol and phosphatidy1choline at the constant amount of proteins led to an increase of the size if ISCOMs up to the formation of large aggregates.This fact has not been described in available literature.In similar experiments~ Lovgren and Morein (1988) found a partial reduction of the sedimentation constant~ but no morphologic!!l deviations of ISCOMs.It is noteworthy in this conection that immune responses of mice inoculated with BHV-I-ISCOM aggregates were very weak or absent.We suppose that the specific antigenic determinants in ISCOM aggregates are not readily accessible to immunity-mediating cells.The formation of the aggregates could be preven'ted by a twofold increase of the proportion of Quil A~ but such combination appeared to be toxic in our experiments.
The results of our experiments and other available information let us conclude that ISCOMs are a prospectful component for further development of subunit vacCines.This applies especially to the use of specific proteins obtained either by DNA recombination~ or by chemical synthesis.

Table 1
Rate incot-poration of DHV-l protein into ISCOM and antibody responses in mice after subcutaneous administration All mice were reinoculated on Oay 21 p. i.The dose was 1 pg protein ISCOM in all anitnals.The dose of free virus protein was 10 PII.•C = cholesterol, PC = phosphatidylcholine ••NO = not examined -all mice died within 24 hours for both Day 35 and Day 49 p. i. Very low antibody titres or absence of antibodies were observed in mice treated with ISCOMs arranged into large aggregates (Groups 5 and 6).A twofold elevation of the content of Quil A (Group 7) resulted in an increase of the incorporation rate to a level comparable with that used in Group 53 but all treated mice died within 24 hours.
I Number I Jig I % I ELISA titre daY'! p. i. Group of mice C• I PC• I Qui!A I Protein BHV-1 incorporation I 35 I Franz J. et al.: Preparation of ... pp.37-41.