Cross-neutralizing Antibodies Induced by Single Serotype Vaccination of Cows with Rotavirus

Transcription

1 J. gen. Virol. (1988), 69, Printed in Great Britain 1647 Key words: rotavirus/vaccination/cows, passive immunity Cross-neutralizing Antibodies Induced by Single Serotype Vaccination of Cows with Rotavirus By HARALD BRUSSOW,* ISABELLE WALTHER, VINCENT FRYDER, JOSETTE SIDOTI AND ANNE BRUTTIN Nestec Ltd., Nestl~ Research Centre, Vers-chez-les-Blanc, 1000 Lausanne 26, Switzerland (Accepted 29 March 1988) SUMMARY Single serotype vaccination of mature cows with nine different strains of bovine, simian and human rotaviruses induced heterotypic milk and serum neutralizing antibodies against two bovine and four human rotavirus serotypes. Immunization with single-shelled simian rotavirus SA11 increased milk and serum neutralization titres fivefold over those of control cows, without inducing antibodies to outer shell polypeptides of rotavirus. Vaccination with double-shelled SA 11 virions also elicited cross-reacting antibodies to the outer shell proteins VP3 and/or VP7 which neutralized rotavirus seven times more efficiently than antisera to single-shelled SAIl virus. A related rotavirus similar to simian rotavirus SA 11, but from a different host, might thus be an attractive vaccine for immunization of pregnant cows to confer passive immunity to calves. INTRODUCTION Rotaviruses are a major cause of diarrhoea in young animals and children (Flewett & Woode, 1978). Serologically they are quite complex: to cite only two mammalian species, five serotypes have been identified among rotaviruses from children (Hoshino et al., 1984; Matsuno et al., 1985), and at least two serotypes have been isolated from calves (Bridget & Brown, 1984; Briissow et al., 1987 a; Ihara et al., 1983; Murakami et al., 1983 ; Ojeh et al., 1984; Snodgrass et al., 1984; Woode et al., 1983). Two bovine rotavirus genes coding for two different outer shell proteins cosegregated with virus neutralization specificities (Offit & Blavat, 1986). The serotype specificity of neutralizing antibodies elicited by the minor neutralizing antigen of bovine rotavirus can differ from that of neutralizing antibody elicited by the major neutralizing antigen (Hoshino et al., 1985). Thus a binary system similar to that used for influenza A virus has been proposed for rotavirus serotype description (Hoshino et al., 1984, 1985). This serological complexity of rotaviruses evidently poses problems for immunization (Woode et al., 1987). However, the relevance of rotavirus serotype distinction for practical immunization procedures in cattle has been questioned (Briissow et al., 1987b; Snodgrass et al., 1984). We therefore studied the serotype specificity of milk and serum antibodies of cows immunized with two bovine and four human rotavirus serotypes. METHODS Viruses. Bovine rotavirus strains UK, NCDV (both serotype 6; Hoshino et al., 1984, 1985) and V1005 (Bachmann & Hess, 1981) were obtained from P. Bachmann (Institute of Medical Microbiology, Munich, F.R.G.). Rotavirus V1005 differs from rotavirus UK in having 20-fold differences in neutralization titres with guinea-pig hyperimmune sera (Briissow et al., 1987 a), but serotype assignation has not been established. Bovine rotavirus 678 was obtained from D. Snodgrass (Moredun Research Institute, Edinburgh, U.K.). It is distinct from serotype 6 UK virus (Snodgrass et at., 1984). Human rotavirus strains I(U and S-2 (Urasawa et al., 1982), O 1988 SGM

2 1648 H. BRUSSOW AND OTHERS representative of serotype 1 and 2 rotaviruses respectively (Wyatt et al., 1983; Beards & Flewett, 1984), were provided by S. Urasawa (Sapporo Medical College, Sapporo, Japan). Y. Inaba (National Institute of Animal Health, Tsukuba, Ibaraki, Japan) donated human rotavirus strains Toi, Ito and Hochi (Sato et al., 1982). Rotaviruses Ito and Hochi are of serotypes 3 and 4 respectively (Wyatt et al., 1983; Beards & Flewett, 1984). Rotavirus Toi may be of serotype 3 (Sato et al., 1982). Human rotavirus Wa and simian rotavirus SA 11, prototypes of serotypes 1 and 3 respectively (Hoshino et al., 1985), were gifts from P. Offit (Children's Hospital, Philadelphia, Pa., U.S.A.). Immunization of cows. Pregnant German Brown and Holstein-Fresian cows from our experimental dairy farm at Albffihren, F.R.G. (Nestl6 Germany) were hyperimmunized during their second to fourth pregnancy with the cellfree culture supernatant of MA-104 cells (macaque rhesus monkey kidney cells) that had been infected with the indicated rotavirus strains. Alternatively, cows were immunized with single-shelled or double-shelled SA11 virus recovered from CsCI gradients. Immunization started 6 weeks before the calculated date of delivery with a subcutaneous injection near the lymph nodes (10 ml of virus and 10 ml of Freund's incomplete adjuvant; Difco). For the next 2 weeks cows were immunized weekly by intracisternal infusion through the four teat channels (25 ml of virus, four times). Three weeks before delivery, cows were immunized subcutaneously via the retromammary lymph nodes [20 ml of virus and 20 ml of 2~ AI(OH)3]. Two weeks before delivery, cows were given an intramuscular injection, deep into the hip (10 ml of virus and 10 ml of incomplete Freund's adjuvant), followed by a slow intravenous injection [20 ml of virus and 320 ml of AI(OH)3] 1 week before delivery. Antigen was administered to each cow by each of the routes indicated and on the same time schedule. Each cow received only one rotavirus strain. Control cows were immunized with five Escherichia coli serotypes responsible for infantile gastroenteritis, O18 :K76[B20], O20:K 1711], 026 :K60[B6], O44 : K74[L] and O55 :K59[B5]. The bacteria were inactivated by treatment with formalin (0.5 ~). The vaccine was diluted to contain 108 bacteria/ml. The immunization protocol was as for immunization with rotavirus. The lengthy immunization protocol was necessary because we were trying to obtain high-titre bovine milk (Briissow et al., 1987b) for use in paediatrics (Hilpert et al., 1987). The primary aim was not vaccination of cows for prevention of calf diarrhoea. ELISA. Single-shelled simian rotavirus SA11 grown in MA-104 cells and purified by CsCI density gradient centrifugation was coated onto microtitre plates. Serum and milk samples were diluted in a twofold dilution series in PBS-Tween 20 (0.05 ~). Bound antibodies were revealed with affinity-purified antibody to bovine IgG (heavy and light chains) coupled to alkaline phosphatase (Kirkegaard & Perry, Gaithersburg, Md., U.S.A.). Sigma 104 substrate was used according to the manufacturer's instructions. The positive-negative cut-off value was set at three times the absorbance value for the sample on the plate with control antigen from uninfected MA-104 cells, and the buffer blank on the plate with SA11 antigen. Neutralization test. The indicated rotavirus strain (100 TCIDs0) was incubated with a twofold dilution series of serum or milk. The neutralization test (NT) was done as described by Gerna et al. (1984). NT titres are expressed as the reciprocal of serum dilution reducing the number of infected cells by 90~. Extracellular rotavirus purification. Extracellular rotavirus particles were recovered from the cell-free culture supernatant of infected cells by high-speed centrifugation (90000g for 2 h at 4 C) through a 20~ sucrose cushion. Single-shelled and doubled-shelled SAIl rotavirus particles were purified by CsC1 density gradient centrifugation (SB405 rotor, International; g for 18 h; initial density p = g/ml). The viral particles banding at 1.38 g/ml and 1.36 g/ml, respectively, were recovered with a Pasteur pipette, diluted with 0.9~ NaCI containing 10 mm-cac12 and pelleted by a further high-speed centrifugation ( g for 1 h at 4 C). Western blot analysis. Polypeptides of extracellular SA 11 rotavirus were separated on SDS-polyacrylamide gels according to Laemmli (1970) under non-reducing conditions and then transferred onto nitrocellulose membranes (Towbin et al., 1979). Membrane strips were incubated with a 1:200 dilution in PBS-Tween 20 (0-5~) of pooled milk or serum of cows identified by their code number in Fig. 3. ELISA and NT titres of those samples are given in Tables 2, 4, 5 and 6. The immune reaction was revealed (Blake et al., 1984) by alkaline phosphatase-conjugated goat antibodies to bovine immunoglobulins (see ELISA). Immune electron microscopy. Double-shelled and single-shelled SA11 rotaviruses were incubated with a 1:500 dilution of first milking colostrum of cow 25 immunized with E. coli control antigen (NT titre to SAIl: 3200; ELISA: 5000), cow 24 (NT: 8000; ELISA: ) immunized with single-shelled SA11 virus, cow 21 (NT: ; ELISA: ) immunized with double-shelled SAIl virus and cow 07 (NT: ; ELISA: 80000) immunized with the supernatant of a cell culture infected with rotavirus Hochi. Skimmed colostrum was cleared by centrifugation (10000 g for 15 min) before use. Alternatively, sera diluted 1:50 and cleared by centrifugation ( g for 1 h) were used. Incubation was for 1 h at 37 C and 1 h at 4 C. Viral particles were deposited on copper grids coated with Formvar film and carbon. They were negatively stained with a 2~ solution of phosphotungstic acid, ph 7.0. Particles were viewed with a Philips EM 300 electron microscope at 80 kv accelerating voltage.

3 Immunization of cows with rotavirus 1649 RESULTS status of cows and immunization with bovine rotaviruses All adult cows used in our study showed serological evidence of natural rotavirus exposures (Tables 1 to 6). In the pooled milk from control cows of the first experimental series, we measured a mean ELISA titre of 800 and mean NT titres between 140 and 640, against two different bovine rotavirus serotypes and four different human rotavirus serotypes (Table 1). Comparable ELISA titres (640) and N T titres (160 to 340) were observed in sera of control cows (Table 3). On immunization with bovine rotaviruses NCDV or V1005 or both strains, we observed in pooled milk a 35-fold increase in ELISA and an 11- to 61-fold increase in NT titre to the four different bovine rotavirus strains tested in comparison with control cows immunized with E. coli (Table 1). Immunization with bovine strains also induced a 10- to 14-fold increase in NT titre to four different human rotavirus serotypes (Table 1). Immunization of cows with human rotavirus In a second experimental series, cows were immunized with rotavirus strains representing the four human rotavirus serotypes. In pooled milk we observed a 27-fold increase in ELISA titres and a 13- to 37-fold increase in NT titres to the four human rotavirus serotypes (Table 2). Human rotavirus also induced a six- to 25-fold increase in NT titres to the four strains of bovine rotavirus representing two bovine rotavirus serotypes (Table 2). As in the preceding experimental series, extensive cross-neutralizations were observed: only a single cow (04) showed a 20-fold higher NT titre to the homologous virus (4200) than to a heterologous rotavirus serotype (200 to rotavirus 678). Table 1. NT titres to bovine, simian and human rotaviruses in pooled milk from cows immunized with bovine rotaviruses or E. coli * Reciprocal of highest dilution of pooled milk from the first 8 days after calving, that neutralized the indicated rotavirus strain. 1" According to Hoshino et al. (1984). Pooled milk of indicated number of cows, mean of duplicates. Arithmetic mean value for all cows immunized with bovine rotaviruses (BRV). II Arithmetic mean value for all cows immunized with E. coli. Strain serotypet,k Number " NCDV UK V Wa S-2 SA11 Hochi ~ Antigen of cows 6 6?? used for used for Mean neutralization titre* ( 10-3) to indicated rotavirus immunization each group r ~ ~ ELISA NCDV :~ NCDV NCDV V VI NCDV + V , NCDV + V NCDV + V E. coli :~ E. coti E. coli E. coli , E. coli E. coli Mean, BRV Mean, E. cohql Titre ratio (BRV/E. coli; milk) 6l

4 1650 H. BRUSSOW AND OTHERS Table 2. NT titres to bovine, simian and human rotaviruses in pooled milk from cows immunized with human rotaviruses or E. coli Strain serotype ), Antigen Wa S-2 SAll Hochi NCDV UK V used for Code ?? immunization number Neutralization titre ( x 10-3) to indicated rotavirus (serotype) of cow c ~" ~ ELISA KU (1) * KU (1) * t 4.0 Wa (I) * S-2 (2) * t 8-0 SA11 (3) * Ito (3) * t Hochi (4) * Hochi (4) * Toi (?) Toi (?) E. coli E. coli E. coli Mean, HRV~ Mean, E. coli Titre ratio (HRV/E. coli; milk) * Homologous titre. t Homologous/heterologous titre _> 10. :~HRV, Human rotavirus. Cross-neutralizing antibodies to rotavirus are not a peculiarity of bovine milk. Immunization of cows with human rotavirus also gave a 12- to 49-fold increase in serum antibodies neutralizing four human and two bovine rotavirus serotypes (Table 3). Immunization of cows with purified single-shelled and double-shelled simian rotavirus The cell-free culture supernatant of MA-104 cells infected with simian rotavirus SAIl (serotype 3) induced high-titre neutralizing antibodies to all four human and two rotavirus serotypes in bovine milk and bovine serum (cow 05 in Tables 2 and 3). Next, we were interested in investigating whether these cross-reacting antibodies were stimulated by single-shelled or double-shelled virions. Cows were therefore immunized with single-shelled and double-shelled rotavirus SA11 separated by CsCI density gradient centrifugation. Double-shelled and single-shelled SA 11 rotavirus induced a comparable increase in milk (20- and 14-fold increase, respectively; Table 4) and serum (15- and 34-fold increase, respectively; Tables 5 and 6) ELISA antibody titres. However, double-shelled virions elicited a greater increase in neutralizing antibody titres than single-shelled virions both in milk (62-fold and sixfold, respectively; Table 4) and in serum (14-fold and fivefold, respectively; Tables 5 and 6). Note that the increase in neutralizing antibody induced by double-shelled virions is about sevenfold higher than that induced by single-shelled virions when the ratio of ELISA antibody increases is taken into account (for serum 14:5/15 : 34 = 6.4; for milk 62 : 6/20 : 14 = 7-2). lmmune electron microscopy serum of cow 32 moderately aggregated double-shelled simian rotavirus SA11 (Fig. 1 d). Immune serum of cow 05 immunized with simian rotavirus SAIl induced large aggregates of double-shelled SA1 l virions which were heavily coated with antibodies (Fig. 1 c). Comparable aggregation of double-shelled SA11 virions was observed with immune sera of cows 02, 04 and 07 immunized with rotavirus KU (Fig. 1 b), S-2 (Fig. 1 e) and Hochi (Fig. l f).

5 Immunization of cows with rotavirus 1651 Table 3. NT titres to bovine, simian and human rotaviruses in sera from cows immunized with human rotaviruses or E. coli Strain serotype A F Antigen Wa S-2 SA 11 Hochi NCDV UK V1005 used for Code ? immunization number Neutralization titre (x 10-3) to indicated rotavirus (serotype) of cow r KU (1) 01 12'00" 2"40 6"40 KU (1) 02 12"00" 2"00 3"20 Wa (1) * Wa (1) " S-2 (2) * 3.20 S-2 (2) ' 1.60 S-2 (2) " 6-40 SAIl (3) * Ito (3) * Hochi (4) E. coli E. coli E. coli E. coli E. coli Foetal calf serum (Flow) <0.01 < Foetal calf serum (Seromed) 0-02 <0.01 <0.01 Mean, HRV Mean, E. coli Titre ratio (HRV/E. coli; sera) * Homologous titre. t Homologous/heterologous titre > ? ~", ELISA "60 3"20 9" "80 2" " ' "60 16" ' "00 3' "20 25" " ' "00 7"50* 0-50 t " ' "10 0" ' " ' ;32 0' ' < <0.01 < <0.01 <0.01 < Table 4. NT titres to bovine, simian and human rotaviruses in pooled milk from cows immunized with double-shelled (DS) or single-shelled (SS) SAIl rotavirus or E. coli ( Strain serotype A "x SAIl Wa S-2 Hochi NCDV UK V Antigen Code ?? used for number Neutralization titre ( x 10-3) to indicated rotavirus immunization of cow c "~ ~ ELISA SA11 DS SA11 DS SA11 SS SA11 SS E. coti t E. coli E. coli E. coli E. coli Mean, DS Mean, SS Mean, E. coli Titre ratio (DS : E. coil/ SS :E. coli) 105/3 59/3 6/1 40/3 49/15 32/11 172/4 29/5 20/14 Double-shelled and single-shelled SAll virions were coated and aggregated by the first milking colostrum of cow 21 immunized with double-shelled SA11 rotavirus (Fig. 2e and f, respectively). Colostrum of cow 24 immunized with single-shelled SA11 rotavirus aggregated single-shelled (Fig. 2d) but not double-shelled (Fig. 2 c) SA11 virus. At the 1 : 500 dilution used in

6 1652 H. BRLISSOW AND OTHERS Table 5. NT titres to bovine, simian and human rotaviruses in sera from cows immunized with double-shelled rotavirus SAl l Code number of cow Serum 31 Immune 32 Immune 33 Immune 34 Immune 21 Immune 22 Immune Mean, immune sera Mean, preimmune sera Titre ratio (immune/ preimmune sera) Strain serotype f A SAll Wa S-2 Hochi NCDV UK V ?? Neutralization titre ( 10-3) to indicated rotavirus r A ~ ELISA , ' "1 0"2 0"2 0"40 0"~ 0"2 0"4 0' " " "0 0" " "40 0"2 0"4 1" '0 3"2 100" "0 100"0 16"0 3" '8 6'4 3'20 1"60 3'2 6'4 3' '6 1"6 1" "2 6"4 8"0 0"4 0"4 0'2 0'4 0'05 0"05 0'1 0'8 0' "0 0"4 8"0 5' "4 0'8 32'0 0"8 0"4 0'4 0'6 0'80 0"10 0"8 0'1 1'6 18" "8 6"4 3" "2 6"4 32"0 1"8 0"8 0"4 0" "4 0'4 1'6 47"0 10"60 1"8 24"00 7" "50 20'30 22'70 1"2 0"95 0'4 1'45 1'0 0"45 0'81 1'42 1" Table 6. NT titres to bovine, simian and human rotaviruses in sera from cows immunized with sing&-shelled rotavirus SAl l Code number of cow Serum 41 Immune 42 Immune 43 Immune 44 Immune 24 Immune 23 Immune Mean, immune Mean, preimmune Titre ratio (immune/ preimmune sera) Strain serotype f SAll Wa S-2 Hochi NCDV UK V ?? Neutralization titre ( x 10-3) to indicated rotavirus r ~ ~ ELISA 1"6 0"8 0"4 0"8 0"8 0"40 0"8 1'6 8" this experiment, colostrum of cow 25 immunized with E. coli did not react with double-shelled (Fig. 2a) or single-shelled (Fig. 2b) SA11 rotavirus. Double-shelled SA11 rotavirus was also aggregated by colostrum of cow 07 immunized with heterologous rotavirus Hochi (Fig. 2g) but not by a hyperimmune serum of a guinea-pig immunized with rotavirus Hochi (Fig. 2h; NT titre of to Hochi and 200 to SAll rotavirus, ELISA titre of 12800).

7 Immunization of cows with rotavirus 1653 (d) Fig. 1. Immune electron microscopy analysis of cows' serum immune response to rotavirus. Doubleshelled rotavirus SAI 1 was incubated with buffer alone (a), preimmune sera of cow 32 (d) or immune serum of cows immunized with rotavirus KU (cow 02, b), S-2 (cow 04, e), SA11 (cow 05, c) or Hochi (cow 07, f). Antibody titres of cows' sera are given in Table 3 (cows 02 to 07) and in Table 5 (cow 32). Bar marker represents 0.5 rtm. Western blot analysis sera of cows reacted with VP2 (Fig. 3 lane 1) or VP2 and VP6 (not shown). Immunization with single-shelled SA11 rotavirus stimulated a strong antibody response to VP6 (Fig. 3 lane 2), whereas double-shelled SA 11 rotavirus also induced antibodies to the outer shell proteins VP3 and VP7 (Fig. 3 lane 3). Milk of control cows at a 1:500 dilution showed almost no reactivity in Western blots (Fig. 3 lane 4). Milk of cows immunized with single-shelled SA11 virions reacted strongly with VP6 and weakly with VP2 and VP7 (Fig. 3 lanes 5 and 7). Milk of cows immunized with double-shelled SA11 virions, however, reacted strongly with both outer shell proteins VP3 and VP7 in addition to VP6 (Fig. 3 lanes 6 and 8). Milk of cows immunized with rotaviruses representing the four human serotypes also cross-reacted with either VP3 or VP7 of rotavirus S A l l (Fig. 3 lanes 9 to 12).

8 1654 H. BRfJSSOW AND OTHERS Fig. 2. Immune electron microscopy analysis of cows' immune response to single-shelled and doubleshelled SA11 rotavirus. Colostrum of cow 25 immunized with E. coli (a and b), colostrum of cow 24 immunized with single-shelled SA11 rotavirus (c and d) and colostrum of cow 21 immunized with double-shelled SA11 (e and f ) were incubated with double-shelled (a, c and e) and single-shelled SA11 rotavirus (b, d and f). Colostrum of a cow 07 immunized with human rotavirus Hochi (serotype 4) (g) and hyperimmune guinea-pig serum to Hochi (h) were incubated with double-shelled SA 11 rotavirus. Antibody titres of cows ~ milk are given in Table 2 (cow 07) and Table 4 (cows 21, 24 and 25).

9 Immunization o f cows with rotavirus i?"? =: i i i?):iii,:i?:, il= =?"~= ~!? 14 VP3 ~ ii i:i:i ~ ~iii ~ VP ! ~ 9 Fig. 3. Western blot analysis of cows' immune response to rotavirus polypeptides. After electrophoresis extracellular rotavirus SA11 was transferred onto a nitrocellulose membrane and incubated with sera (lanes 1 to 3) and milk (lanes 4 to 12) of cows identified by their code number and immunizing antigen (O, preimmune; E. coli, control antigen; SS, single-shelled SA 11; DS, double-shelled SA11 ; Wa, S-2, SA 11 and Hochi, cell culture supernatant of MA- 104 cells infected with the respective rotavirus strain). Lane 1, 24 O; lane 2, 24 SS; lane 3, 21 DS; lane 4, 25 E. coli; lane 5, 24 SS; lane 6, 21 DS; lane 7, 23 SS; lane 8, 22 DS; lane 9, 03 Wa; lane 10, 04 S-2; lane 11,05 SA11 ; lane 12, 07 Hochi. Antigen recognition pattern was revealed using phosphatase-conjugated antibody to bovine IgG. Lanes 13 and 14 show, for comparison, the polypeptide composition of [3sS]methionine-labelled double-shelled (lane 13) and single-shelled (lane 14) SA11 rotavirus as revealed by autoradiography. DISCUSSION Single serotype vaccination of mature cows with nine different bovine, simian or human rotavirus strains induced heterotypic milk and serum antibodies against at least six different serotypes of bovine and human viruses. Our results are an extension of the work of Snodgrass et al. (1984) who observed both homotypic and heterotypic increases in serum antibody titres after monovalent vaccination of cows with one serotype of bovine rotavirus. According to the theory of 'original antigenic sin' the induction of cross-reacting antibodies should be expected in individuals with pre-existing heterologous immunity (Davenport et al., 1953). Snodgrass et al. (1984) found preimmune titres to two bovine rotavirus serotypes and human rotavirus serotypes 1, 2 and 3, but not to human serotype 4, in cows. They also observed a six- to 10-fold increase of serum neutralizing titres to both bovine rotavirus serotypes and heterologous human rotavirus serotypes 1, 2 and 3, but not to human serotype 4, when cows were immunized with one bovine rotavirus serotype. Our immunization also induced antibodies cross-reacting with serotype 4 rotavirus, which agrees with the observation that our cows also had preimmune titres to serotype 4 rotavirus. Snodgrass et al. (1984) proposed that cows have acquired this heterologous immunity by natural exposure to bovine rotavirus serotypes resembling human rotaviruses.

10 1656 H. BRUSSOW AND OTHERS From studies of influenza A infections, it is well known that infection with one influenza virus subtype may influence the subsequent response to another subtype which does not cross-react serologically but which shares common internal antigens (Jennings et al., 1974). Subsequently, it was shown that T cells recognizing some internal component of the virion are capable of cooperation with B cells in the production of haemagglutination-inhibiting antibodies (Russell & Liew, 1979). It is interesting to note that internal components of rotaviruses have some intrinsic activity to induce neutralizing antibodies both in serum and in milk (Tables 6 and 4, respectively) without inducing antibodies to outer shell components (Fig. 2 c and d, Fig. 3, lanes 5 and 7). Two other groups also reported minor neutralizing epitopes on VP6 (Bastardo et al., 1981 ; Killen & Dimmock, 1982). Double-shelled SA11 virions induced antibodies which were seven times more efficient in neutralizing rotaviruses than those induced by immunization with single-shelled SA 11 virions (Tables 4 to 6). Thus the majority of the cross-neutralizing antibodies are directed against VP3 and/or VP7 of the outer shell of rotaviruses, as shown by immune electron microscopy (Fig. 2) and Western blot analysis (Fig. 3). The induction of heterotypic antibodies by single serotype vaccination of mature cows might also be of practical importance. Active rotavirus immunization of calves failed to induce protection in most (Acres & Radostits, 1976; De Leeuw et al., 1980) but not all (Thurber et al., 1977) field studies. These failures have been variously attributed to difficulties in proving vaccination efficiency in double-blind placebo-controlled trials due to problems with herd immunity (Thurber et al., 1977). Alternatively, it was argued that vaccination-induced immunity may not be fully developed when calves become infected with virulent rotavirus shortly after birth (De Leeuw et al., 1980). Finally, specific antibodies in the colostrum may hinder the 'take' of the oral rotavirus vaccine (De Leeuw & Tiessink, 1985). The last two problems could be circumvented by immunizing dams with rotavirus and feeding immune colostrum to calves. However, a field trial failed to prove a significant difference between calves from placebo-treated and vaccine-treated dams. This failure is hardly a valid argument against the feasibility of passive protection of calves: the commercial vaccine used in this study did not induce even a twofold increase in milk antibody titre in comparison with control cows (Walther- Toews et al., 1985). Encouraging results have been obtained in field trials using bovine rotavirus V1005 as a vaccine (Eichhorn et al., 1982). In a "Jennerian' approach to vaccination (Kapikian et al., 1986), i.e. the use of a related virus from a different host, simian rotavirus SAIl for example might be an attractive immunizing antigen for vaccination of pregnant cows. We wish to thank the late Professor P. Bachmann, Professor Y. Inaba, Professor S. Urasawa, Dr D. Snodgrass and Dr P. Offit for their generous gifts of different rotavirus strains, Dr E. and M. Hermann from the Nestl6 Experimental Dairy Farm, Albfiihren, F.R.G. and Miss E. Cano for their excellent assistance, Dr W. Wunderli for help in the initial phase of the work, Dr H. Hilpert for constant support of the project, and Miss M. Kalberlah for typing the manuscript. REFERENCES ACRES, S. D. & RADOSTITS, O. M. (1976). The efficacy of a modified live reolike virus vaccine and an E. coli bacterin for prevention of acute undifferentiated neonatal diarrhoea of beef calves. Canadian Veterinary Journal 17, BACHMANN, P. A. & HESS, R. G. (1981). Routine isolation and cultivation of bovine rotaviruses in cell culture. American Journal of Veterinary Research 42, BASTARDO, J. W., McKIMM-BRESCHKIN, J.-L., SONZA, S., MERCER, L. D. & HOLMES, I. H. (1981). Preparation and characterization of antisera to electrophoretically purified SA11 virus polypeptides. Infection and Immunity 34, 641~547. BEARDS, G. M. & FLEWETT, T. H. (1984). Serological characterization of human rotaviruses propagated in cell culture. Archives of Virology 80, BLAKE, M. S., JOHNSTON, K. H., RUSSELL-JONES, G. J. & GOTSCHLICH, E. C. (1984). A rapid, sensitive method for detection of alkaline phosphatase-conjugated anti-antibody on Western blots. Analytical Biochemistry 136, BRIDGER, J. C. & BROWN, J. F. (1984). Antigenic and pathogenic relationships of three bovine rotaviruses and a porcine rotavirus. Journal of General Virology 65,

11 Immunization of cows with rotavirus 1657 BR~)SSOW, H., MARC-MARTIN, S., EICHHORN, W., SIDOTI, J. & FRYDER, V. (1987 a). Characterization of a second bovine rotavirus serotype. Archives of Virology 94, BRUSSOW, H., HILPERT, H., WALTHER, J., SIDOTI, J., MIETENS, C. & BACHMANN, P. (1987b). Bovine milk immunoglobulins for passive immunity to infantile rotavirus gastroenteritis. Journal of Clinical Microbiology 25, DAVENPORT, F. M., HENNESSY, A. V. & FRANCIS, T., JR (1953). Epidemiological and immunological significance of age distribution of antibody to antigenic variants of influenza virus. Journal of Experimental Medicine 98, 641~50. DR LEEUW, P. W. & TIESSINK, J. W. A. (1985). Laboratory experiments on oral vaccination of calves against rotavirus or coronavirus induced diarrhoea. Zentralblatt J~r Veteriniirmedizin B 32, DE LEEUW, P. W., ELLENS, D. J., TALMON, F. B. & ZIMMER, G. N. (1980). Rotavirus infections in calves: efficacy of oral vaccination in endemically infected herds. Research in Veterinary Science 29, E1CHHORN, W., BACHMANN, P. A., BALJER, G., PLANK, P. & SCHNEIDER, P. (1982). Vakzinierung hochtr~ichtiger Rinder mit einem kombinierten Rotavirus/E. cell K99--Impfstoff zur Prophylaxe von Durchfallerkrankungen bei neugeborenen K~ilbern. Tieri~rztliche Umschau 37, ELEWETT, T. H. & WOODE, G. N. (1978). The rotaviruses. Archives of Virology 57, GERNA, G., BATTAGLIA, M., MILANESI, G., PASSARANI, N., PERCIVALLE, E. & CATTANEO, E. (1984). Serotyping of cell culture-adapted subgroup 2 human rotavirus strains by neutralization. Infection and Immunity 43, H1LPERT, H., BRLISSOW, H., MIETENS, C., SIDOTI, J., LERNER, L. & WERCHAU, H. (1987). Use of bovine milk concentrate containing antibody to rotavirus to treat rotavirus gastroenteritis in infants. Journal oflnfectious Diseases 156, HOSHINO, Y., WYATT, R. G., GREENBERG, H. B., FLORES, J. & KAPIKIAN, A. Z. (1984). Serotypic similarity and diversity of rotaviruses of mammalian and avian origin as studied by plaque-reduction neutralization. Journal of Infectious Diseases 149, HOSI-IINO, Y., SERENe, M. M., MIDTHUN, K., FLORES, J., KAPIKIAN, A. Z. & CHANOCK, R. M. (1985). Independent segregation of two antigenic specificities (VP3 and VP7) involved in neutralization of rotavirus infectivity. Proceedings of the National Academy of Sciences, U.S.A. 82, IHARA, T., SAMEJIMA, T., KUWAHARA, H. & TAYIMA, M. (1983). Isolation of new serotypes of bovine rotavirus. Archives of Virology 78, JENNINGS, R., POTTER, C. w. & McLAREN, C. (1974). Effect of preinfection and preimmunization on the serum antibody response to subsequent immunization with heterotypic influenza vaccines. Journal of Immunology 113, KAPIKIAN, A. Z., FLORES, J., HOSHINO, Y., GLASS, R. I., MIDTHUN, K., GORZIGLIA, M. & CHANOCK, R. M. (1986). Rotavirus: the major etiologic agent of severe infantile diarrhoea may be controllable by a "Jennerian" approach to vaccination. Journal of Infectious Diseases 153, KILLEN, H. M. & DIMMOCK, N. J. (] 982). Identification of a neutralization-specific antigen of a calf rotavirus. Journal of General Virology 62, LAEMMLI, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, London 227, MATSUNO, S., HASEGAWA, A., MUKOYAMA, A. & INOUYE, S. (1985). A candidate for a new serotype of human rotavirus. Journal of Virology 54, MURAKAMI, Y., NISHIOKA, N., HASHIGUCHI, Y. & KUNIYASHU, C. (1983). Serotypes of bovine rotaviruses distinguished by serum neutralization. Infection and lmmunity 40, OFFIT, P. A. & BLAVAT, G. (1986). Identification of the two rotavirus genes determining neutralizing specificities. Journal of Virology 57, OJEH, C. K., SNODGRASS, D. R. & HERRING, A. J. (1984). Evidence for serotypic variation among bovine rotaviruses. Archives of Virology 79, RUSSELL, S. M. & LmW, r. Y. (1979). T cells primed by influenza virion internal components can cooperate in the antibody response to haemagglutinin. Nature, London 2,80, SATe, K., INABA, Y., MIURA, Y., TOKUHISA, S. & MATUMOTO, M. (1982). Antigenic relationship between rotaviruses from different species as studied by neutralization and immunofluorescence. Archives of Virology 73, SNODGRASS, D. R., OJEH, C. K., CAMPBELL, I. & HERRING, A. J. (1984). Bovine rotavirus serotypes and their significance for immunization. Journal of Clinical Microbiology 20, THURBER, E. T., BASS, E. P. & BECKENHAUER, W. H. (1977). Field trial evaluation of a reo-coronavirus calf diarrhoea vaccine. Canadian Journal of Comparative Medicine 41, TOWBIN, H., STAEHELIN, T. & GORDON, J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proceedings of the National Academy of Sciences, U.S.A. 76, URASAWA, S., URASAWA, T. & TANIGUCHI, K. (1982). Three human rotavirus serotypes demonstrated by plaque neutralization of isolated strains. Infection and Immunity 38, WALTNER-TOEWS, D., MARTIN, S. W., MEEK, A. H., McMILLAN, L & CROUCH, C. F. (1985). A field trial to evaluate the efficacy of a combined rotavirus~-coronavirus/escherichia cell vaccine in dairy cattle. Canadian Journal of Comparative Medicine 49, 1-9. WOODE, G. N., KELSO, N. E~, SIMPSON, T. g., GAUL, S. K., EVANS, L. E. & BABIUK, L. (1983). Antigenic relationship among some bovine rotaviruses: serum neutralization and cross-protection in gnotobiotic calves. Journal of Clinical Microbiology 118,

12 1658 H. BRf3SSOW AND OTHERS WOODE, G. N., ZHENG, s., ROSEN, B. I., KNIGHT, N., GOURLEY, N. E. K. & RAMIG, R. F. (1987). Protection between different serotypes of bovine rotavirus in gnotobiotic calves: specificity of serum antibody and coproantibody responses. Journal of Clinical Microbiology 25, WYATT, R. G., JAMES, H. D., PITTMAN, A. L., HOSHINO, Y., GREENBERG, H. B., KALICA, A. R., FLORES, J. & KAPIKIAN, A. Z. (1983). Direct isolation in cell culture of human rotaviruses and their characterization into four serotypes. Journal of Clinical Microbiology 18, (Received 8 September 1987)