Serological and Immunological Relationships between the 146S and 12S Particles of Foot-and-Mouth Disease Virus

Transcription

1 J. gen. Virol. (198o), 50, Printed #~ Great Britain 369 Serological and Immunological Relationships between the 146S and 12S Particles of Foot-and-Mouth Disease Virus By B. CARTWRIGHT, W. G. CHAPMAN AND F. BROWN Animal [%us Research Institute, Pirbright, Woking, Surrey, U.K. (Accepted 28 May I98O ) SUMMARY Intact 146S particles of the seven serotypes of foot-and-mouth disease virus (FMDV) produce type-specific precipitating, complement-fixing and neutralizing antibodies in cattle and guinea-pigs. However, the 12S structural subunit, produced from the virus particle by mild acid treatment (ph 6) or by heating at 56 C, although stimulating the production of precipitating and complement-fixing antibodies, produces only low levels of neutralizing antibody. Nevertheless, IzS particles were as active as 146S particles in stimulating the production of neutralizing antibody in guinea-pigs primed with a vaccine prepared from L46S particles. Moreover, heterotypic ~zs and I46S particles also boosted the neutralizing antibody response to the first virus. These results point to an antigenic similarity between the I46S particles of each type and to a close antigenic relationship between the r46s and IzS particles. INTRODUCTION Foot-and-mouth disease virus (FMDV) hacvests contain four virus-specific particles: (i) the infective ~46S virus particle, comprising one molecule of ssrna and 6o copies of each of four polypeptides VPI, VPz, VP 3 and VP4; (it) the so-called empty 75S particles, devoid of RNA and comprising 60 copies of each of VPI, VP3 and VPo (precursor of VP2 and VP4); (iii) a IaS particle consisting of VPI, VP2 and VP3 but devoid of VP4 and (iv) the virus infection-associated (VIA) antigen. The I2S particle is also produced by mild acid disruption of the I46S particle or by heating at 56 C. Neutralizing antibody production is associated mainly with the I46S particles (Brown & Crick, I959), whereas the ~2S particles stimulate the production of an antibody which has low neutralizing activity but reacts in precipitin and complement*fixation tests. Controlled absorption of hyperimmune serum with excess IaS particles does not reduce the neutralizing activity of the serum, nor its ability to fix complement with the I46S particles (Cartwright, I962). This suggests that the antigen associated with the ~2S particles is not present on the surface of the I46S particles. The T 2S antigen might be considered a crypto-antigen, revealed only after gentle disruption &the 146S particles. Another possibility is that the I2S antigen is part of the I46S antigenic site, producing neutralizing antibody when intact but not after conversion into I2S particles. The antigenic relationship between the I46S and IaS particles has been examined in more detail and is the subject of the present study. METHODS Viruses. The following strains of FMDV were grown in monolayers of baby hamster kidney cells (BHK2I, clone I3) or pig kidney cells (IB-RS-2) from the Instituto Biologico, Sao Paulo, Brazil: type O, strain I (O-V~) and British field strain t86o (O-BFS ~86o), /80/0o $02.00 Downloaded I980 SGM from by

2 37o B. CARTWRIGHT, W. G. CHAPMAN AND F. BROWN type A, strain 6t (A61) and type SATI, strain RVI 1/37 (SATI). The cells grown in Roux bottles were infected with o-oi to o.i p.f.u./cell and incubated at 37 C in 25 ml Eagle's medium until the cells left the glass surface, usually at I5 to I8 h p.i. The supernatant medium separated from the cell debris provided the crude virus suspension. To prepare radio-labelled virus a monolayer containing ~o s BHK or 5 Io7 IB-RS-2 cells was infected with Io 1 p.f.u, of virus. After I'5 h at 37 C the virus suspension was removed, the cells washed three times with Eagle's medium lacking methionine and then incubated with methionine-free Eagle's medium containing 200/zCi of zss-methionine until the cells left the glass, usually at 4"5 to 6 h p.i. To prepare the I46S virus component the method described by Brown & Cartwright (I963) was used. Heating the purified 146S component at 56 C for a minimum of 30 rain produced the I2S component. Swine vesicular disease virus UK 27/72 was grown in monolayers of IB-RS-2 cells and purified using the method described for FMDV. Vaccines. Purified preparations of the I46S and I2S components were inactivated for 16 h at 26 C with o'o5~ acetyl ethyleneimine (AEI) and the reaction stopped by the addition of sodium thiosulphate to give a final concentration of 2~o (w/v). This procedure inactivates the infectivity of FMDV. All the vaccines, both primary and secondary, were prepared by emulsifying an equal volume of Freund's complete adjuvant with the AEItreated preparations. One ml of the vaccine was inoculated subcutaneously into the left shoulder of Dunkin-Hartley strain guinea-pigs weighing 550 to 600 g. Antisera. Hyperimmune guinea-pig antisera were prepared in the World Reference Centre at this Institute. Guinea-pig-adapted virus was inoculated intradermally into the hind foot pads and a second inoculation was made intramuscularly I z weeks later. Blood was collected to days after the second inoculation. Immune antisera were prepared in guinea-pigs from vaccines produced to the purified I46S and IzS antigens, the animals being held I8 days after inoculation. Serological methods Neutralization test. The antibody was estimated by two methods. In the first, Io-fold dilutions of the virus were mixed with an equal vol. of o'o4 M-sodium phosphate, ph 7"6, or antiserum dilution and incubated for 6o min at I8 C. The mixtures (o'o3 ml) were inoculated i.p. into groups of five 7 day-old mice per Io-fold dilution. The difference between the LDs0 values was taken as the neutralization index of the serum. In the second method, twofold dilutions series of serum were prepared in microtitre plates using 50 pl dilution loops. Virus previously titrated and diluted to contain Ioo TCDs0 in 5o/zl was added to each well. Following incubation for I h at 37 C approx. 6 Io 4 BHK cells in 25 pl were added; the plates were then sealed and incubated at 37 C for 48 to 72 h. Neutralization titres were expressed as the reciprocal of the serum dilution at the 5o~ endpoint. Radioimmune precipitation. Dilutions of the guinea-pig sera were titrated in microtitre plates against portions of 35S-methionine-labelled preparations of I46S and IzS particles. Antigen (5 /d) was allowed to react for I6 h at I5 C with 5opl of antiserum dilution. Then 5o pl of suitably diluted anti-species serum was added and after a futher 2 h the plates were centrifuged at 12oo g for 15 min. Volumes of 75/tl were removed from each well, dried on glass fibre discs and the radioactivity measured. The titres of the sera are expressed as the reciprocal of the dilution required to precipitate 5o70 of the 35S-methionine. Leucocyte sthnulation test. The test was described in detail by Wardley et al. (I979) and their procedure was used without modification.

3 bnmunology of FMD V 37 r Table t. Serotype specificity of the neutralizing activity of FMD V hyperhnmune sera Log10 neutralizing activity of serum A._ Virus serotype O A SATl O 3"9* o'3 -o-i A 0'3 3"5 o-7 SAT1 -o.i -o-i 3'7 * Figures indicate log virus neutralized by t/ioo serum. RESULTS Virus specificity The serotype specificity in FMDV is demonstrated by the results obtained with hyperimmune sera produced with the three serotypes used in this study (Table I). These results are typical of those found with all seven serotypes of FMDV. It has been known for some years that antiserum produced in guinea-pigs by inoculation of I2S particles reacts in complement-fixing and immunodiffusion tests with I2S particles but not with I46S particles. We have now examined the specificity of the reactions of sera produced by inoculation of I46S and ~2S particles using the more sensitive radioimmunoassay procedure described in Methods. Sera produced by inoculating equ~.l amounts of 146S or IzS particles of type A virus into guinea-pigs had titres of about 7oo in tests with the I2S particle. However, the titre of the I23 particle antiserum was ~ 8o with I46S particles. In contrast the titre of the I46S particle antiserum was /> mooo with I46S particles. These results show that there is a large quantitative difference in the antigenicity of the two particles in addition to the qualitative difference. In vitro stimulation of virus-primed cattle leucocytes Recently Wardley et al. (I979) demonstrated that peripheral blood leucocytes from steers that had received three inoculations of a commercial inactivated type O vaccine, prepared from an unfractionated harvest from BHK cells, incorporated 3H-thymidine after addition of purified I46S particles of the same serotype. They further showed that BHK cell proteins, due to their presence in the vaccine, will also stimulate the leucocytes. We have confirmed these observations taking leucocytes from the same animals used by Wardley et al. 0979) and have also shown that a highly purified preparation of the I2S particles produced the same level of 3H-thymidine incorporation as the same mass of protein in the form of I46S particles. An antigenic relationship between particles of different serotypes was demonstrated using leucocytes from the same vaccinated animals and purified I46S and I2S particles of type A virus. Preparations of each heterotypic particle stimulated the uptake of 3H-thymidine to a similar extent but the response was less than that obtained with the corresponding homotypic particles. To check that the responses we had observed were not due to BHK cell proteins associated with the purified virus particles, we prepared particles grown in the pig kidney cell line IB-RS-2, having previously established that extracts of these uninfected cells did not stimulate ZH-thymidine uptake by the cattle leucocytes. Essentially similar results were obtained with purified virus particles prepared from harvests produced in BHK and IB-RS-2 cells, indicating that the stimulus was due to virus antigens. The results of the experiments with virus grown in IB-RS-z cells are given in Fig. I. Swine vesicular disease virus, which belongs to the same family as FMDV but is antigenically unrelated, did not stimulate 3H-thymidine uptake by the cattle leucocytes.

4 372 B. CARTWRIGHT, W. G. CHAPMAN AND F. BROWN I I t I I I J l I I? (a) (b) 100 I = 80 - ~ ~ \ I I I I I I I Log dilution I I o m o Ill Fig. I. Stimulation of type O-BFS-primed cattle lymphocytes with (a) type O and (b) type A virus particles grown in IB-RS-2 ceils. --, I46S particles; I---m, I2S particles. Zero log dilution = 3 x IO 7 IDs0/ml. Production of neutralizing antibody in primed animals [t was assumed that the vaccinated steers had B lymphocytes primed to produce antibodies to either I46S or I zs particles. However, the incorporation of 3H-thymidine observed after the addition of I46S or IzS particles fails to identify which lymphocytes are restimulated. Because of its potential importance in vaccination we were prompted to examine whether homotypic and heterotypic 12S particles would stimulate the in vivo production of neutralizing antibody in animals already primed by previous inoculation of r46s particles. Guinea-pigs which had been inoculated 2o days previously with 2/zg of purified I46S particles of type O virus were divided into four groups, each of four animals. The groups were inoculated with equal masses of(i) I46S, type O; (ii) lzs, type O; (iii) I46S, type A and (iv) IzS, type A. Equal volumes of the individual sera from each group were pooled and the neutralizing activity against virus of type O was determined. The neutralization index at ~ 8 days was 2.t to 2-4 logs; this rose to approx. 3"5 logs 8 days after the second inoculation and was unchanged after a further 6 days. This means that the increases in neutralizing antibody titre after the second inoculation were similar irrespective of whether I46S or ~2S particles were used or whether the antigens were of the same serotype. In another experiment of similar design another group was included in which no second inoculum was given and the neutralizing antibody in the sera of individual animals was determined (Table 2). To eliminate the possibility that the stimulation produced by the I2S particles had been caused by ~46S particles which had survived the heat treatment at 56 C, the tzs particles of type A used in the experiment were separated from any surviving I46S particles by sucrose gradient centrifugation. A preparation of 35S-methionine-labelled I46S particles of serotype A was divided into two equal parts. One was kept as control; the second w~s precipitated at -2o C with 7o~ acetone, resuspended in I ml o'o4 M-sodium phosphate, ph 7"6 and heated at 56 C for I h. It was then centrifuged at 6oooo g for 2.75 h in a I5 to 45 ~ sucrose gradient. All the ass-methionine counts were at the top of the gradient, indicating that all the I46S particles had been converted into IzS particles. This labelled material was non-infectious, providing further evidence that no I46S particles were present in the fraction. Both the control and I2S preparations were incubated with AEI, mixed with Freund's complete adjuvant and inoculated into guinea-pigs which had received an inactivated preparation of type O virus zo days previously. The two groups produced a similar increase in the neutralizing antibody to virus of type O (Table 2, groups 3 and 4). A similar increase in antibody titre was obtained with the I46S and I2S particles of type O.

5 Immunology of FMD V 373 Table 2. Neutralizing activity in the sera* of individual guinea-pigs receiving 146S partieles of type 0 virus followed by homotypic or heterotypic 146S or IzS partieles~f Neutralizing activity against r r Type O virus ix A., Type A virus [8 days 8 days I4 days I4 days Second after first after second after second after second Group inoculum inoculation inoculation inoculation inoculation [ Type O, 146S i.i 2'7 o.i 2-5 2'0 3" '5 o.i o.i 2 Type O, 125 0"5 3"3 3'1 0"3 0"5 0"7 3"7 3"I 3'3 2"9 O'3 0"5 3 Type A, 146S 2"3 o'7 2"3 3"3 o.i 3"1 0'7 I'3 3'I 3"3 o'7 3'5 4"6 4 Type A, [2S 2"5 I"3 3"9 2-I 3"5 2"5 4"9 0"3 I "5 I"7 2'9 3"5 2'9 0"7 0"I 5 None ['7 3"3 3"3 0"5 0"7 I 'I ND~ ND I "3 I "3 NO ND I "7 I '5 ND ND * Sera were from individual guinea-pigs and were used at [/zo dilution. t Titres before AEI treatment: type A 146S particles, IDs0/ml; type A 12S particle~, non-infectious; type O, I46S particles, i05.9 ]DS /ml; type O 12S particles, non-infectious. :~ NO, Not determined. Table 3- Neutralizing activity in the sera* of guinea-pigs receiving I46S particles of type A followed by I46S or I2S particles of type SATI~ Neutralizing activity against A ~. N Type A virus Type SAT1 A 18 days 8 days [4 days I4 days after first after second after second after second Second inoculum inoculation inoculation inoculation inoculation Type SATI, I46S 2"5 3"9 3"3 2.o Type SATI, I2S 3"3 3"9 4"3 ~<o'3 * Sera were pooled from groups of four animals and were used at i/iooo dilution. ~" Titres before AEI treatment: type SATI I46S particles, I07.2 IDso/ml; type SATI I2S particles, non-infectious. When the second inoculum consisted of I2S particles of type A there was no stimulation of neutralizing antibody against virus of this type, whereas 146S particles of type A produced homotypic neutralizing activity at a level expected from a primary inoculation. As expected, the level of neutralizing antibody did not increase in the group that did not receive a second inoculum. To test whether the relationship between the different serotypes O and A demonstrated by these results could also be found with other serotypes of the virus, similar experiments were done with particles from serotypes A and SATI. The latter type was chosen because RNA homology tests and tryptic peptide mapping of the RNA polymerase have indicated that the Southern African Territory viruses are more distantly related to type A virus than are the European and Asia I types (Robson et al. I977; Newman et al. I979). The results in

6 374 B. CARTWRIGHT, W. G. CHAPMAN AND F. BROWN Table 3 show that with this pair of viruses also, inoculation of I46S or I2S particles of the heterotypic SATI serotype stimulated the production of neutralizing antibody against the virus of type A inoculated first. This stimulation was obtained despite the low activity of the 146S particles of type SATI in producing a homotypic neutralizing antibody response. DISCUSSION Although the I46S particles of the seven serotypes of FMDV produce specific neutralizing antibody, the experiments described in this paper demonstrate the ability of the particles to 'trigger' lymphocytes already primed with I46S particles of another serotype. The level of neutralizing antibody against the first virus produced by the heterotypic I46S particles is as great as that produced by homotypic I46S particles, suggesting that there is a similarity in the antigenic structure of the different serotypes. Swine vesicular disease virus, which has a polypeptide structure similar to that of FMDV, does not stimulate the uptake of ahthymidine in the FMDV-primed leucocytes, indicating that the effect is virus-specific. The activity of the t 2S particle in stimulating the production of neutralizing antibody in guinea-pigs which had already been inoculated with I46S particles suggests that there is an antigen on the J2S particles with a structure sufficiently similar to that on the I46S particles to stimulate the membrane-ig receptors of lymphocytes primed to produce type-specific neutralizing antibody. Indeed, the I2S particles appear to be as active as 146S particles in producing neutralizing antibody in primed animals. This is in marked contrast to the response of unprimed guinea-pigs to I2S particles, when the level of neutralizing antibody produced is much less than that produced by 146S particles (Brown & Crick, 1959; Ceglowski, I965). The studies of Meloen et al. (1979) showed that I46S particles elicited a greater neutralizing antibody response than IO times the amount of I2S particles. The virus particle contains four structural polypeptides VPI to VP4 and treatment with trypsin cleaves VPI without apparently affecting the other polypeptides (Wild et al. I969; Burroughs et al. 1971). With several strains of virus this cleavage results in a considerable reduction of the infectivity titre and immunizing activity (Wild & Brown, 1967; Wild et al. 1969; Rowlands et al. 197I). This result provided good evidence that only VPI was involved in the immunizing activity of the I46S particles and the demonstration by Laporte et al. (I973), Bachrach et al. (1975), Kaaden et al. (1977) and Meloen et al. (1979) that isolated VP1 will stimulate the production of neutralizing antibody and protect animals against challenge has provided more direct evidence for its role in immunization. Both the 146S and lzs particles contain the VPI polypeptide, suggesting that the conformation of the polypeptide is different in the two particles and that the configuration required for the production of neutralizing antibody in unprimed animals is largely lost when the 146S particles are converted into 12S particles. This conformational difference might be explained by considering the arrangement of the four polypeptides in the picornavirus particle. In the model that is currently accepted, the picornavirus capsid is composed of 6o subunits, each of which contains four non-identical polypeptides. Cross-linking studies with Mengo virus by Hordern et al. (1979) indicate that these polypeptides occupy discrete domains in the capsid and Dunker (I979) has proposed bonding networks with groupings of five of the c~-polypeptides each equivalent to VPI of FMDV. It seems possible that the entire grouping of the fivevp1 molecules is required for reaction with and stimulation of neutralizing antibody. When the virus is broken down to I2S particles, the quaternary structure of the grouping could be lost although the individual polypeptides would remain intact. The loss of quaternary structure could account for the low level of neutralizing antibody produced in unprimed animals by the I2S particles and the isolated VPI polypeptide. It would be of interest to determine whether a similar dependence on the intact groupings of VPI or ~-polypeptides in other picornaviruses is necessary for immunogenic activity.

7 Immunology of FMD V 375 REFERENCES BACHRACH, H. L., MOORE, D. M., McKERCHER, P. D. & POLATNICK, J. (1975). Immune and antibody responses to an isolated capsid protein of foot-and-mouth disease virus. Journal of Immunology xxs, I636--I64t. BROWN, F. & CARTWRtC;HT, B. 0963). Purification of radioactive foot-and-mouth disease virus. Nature, London r99, I I68-117o. BROWN, F. & CRICK, L (I959). Application of agar gel diffusion analysis to a study of the antigenic structure of inactivated vaccines prepared from the virus of foot-and-mouth disease. Journal of Immunology 82, BURROUGHS, J. N., ROWLANDS, D. J., SANGAR, D. V., TALBOT, P. & BROWN, F. (1971). Further evidence for multiple proteins in the foot-and-mouth disease particle. Journal of General Virology x3, CARTWRIGHT, B, (1962). A study of the antibodies in guinea pigs infected with foot-and-mouth disease virus. Journal of Immunology 88, 128-I33. CEGLOWSKI, W. S. 0965). Antibody response to the noninfectious 7m/~ component of the virus of foot-andmouth disease. Virology 25, o. DUNKER, A. K. 0979). The structure of picornaviruses: classification of the bonding networks. Virology 97, I4I-I50. HORDERN, J. S., LEONARD, J. D. & SCRABA, D. G. 0979)- Structure of the Mengo virion. VI. Spatial relationships of the capsid polypeptides as determined by chemical cross-linking analyses. Virology 97, o. KAADEN, O. R., ADAM, K. H. & STROnMAIER, K. (1977)- Induction of neutralizing antibodies and immunity in vaccinated guinea pigs by cyanogen bromide peptides of VP3 of foot-and-mouth disease virus. Jourpal of General Virology 34, 397-4oo. LAPORTE, J., GROSCLAUDE, J., WANTYGHEM, J., BERNARD, S. & ROUZI~, P. (t973)- Neutralisation en culture cellulaire du pouvoir infectieux du virus de la fi6vre aphteuse par des s6rums provenant de porcs immunis6s 5. l'aide d'une prot6"fne virale purifi6e. Comptes rendus hebdomadaires des SOanees de l'acadomie des Sciences 276D, oi. MELOEN, R. H., ROWLANDS, D. J. & BROWN, F. (1979)- Comparison of the antibodies elicited by the individual structural polypeptides of foot-and-mouth disease and polio viruses. Journal of General Virology 45, NEWMAN, J. F. E., CARTWRIGHT, B., DOLL, T. R. & BROWN, F. 0979). Purification and identification of the RNA-dependent RNA polymerase of foot-and-mouth disease virus. Journal of General Virology 45, 497-5o7. ROBSON, K. J. H., HARRIS, T. J. R. & BROWN, F. (I977). An assessment by competition hybridization of the sequence homology between the RNAs of the seven serotypes of FMDV. Journal of General Virology 37, ROWLANDS, D. J., SANGAR, D. V. & BROWN, F. (I971). Relationship of the antigenic structures of foot-andmouth disease virus to the process of infection. Journal of General Virology x 3, WARDLEY, R. C., CHAPMAN, W. G. & GARLAND, g. J. M. 0979). A blastogenic test for foot-and-mouth disease. Journal of Hygiene 83, 5o WILD, X. V. e~ BROWN, r. 0967). Nature of the inactivating action of trypsin on foot-and-mouth disease virus. Journal of General Virology x, WILD, T. F., BURROUGHS, J. N. & BROWN, F. U969). Surface structure of foot-and-mouth disease virus. Journal of General Virology 4, o. (Received 3 April 198o)