Antibodies of patterns 2 and 3 recognized all subgroup ... the same reactivity as pattern 4 except for not recognizing the non-I non-II equine strain. Pattern 6.
Journal of General Virology(1990), 71, 1395-1398. Printedin Great Britain
1395
Monoclonal antibodies to the VP6 of porcine subgroup I rotaviruses reactive with subgroup I and non-subgroup I non-subgroup II strains F. Liprandi, 1. G. Lopez, 1 I. Rodriguez, 1 M . Hidalgo, ~ J. E. Ludert ~ and N. Mattion 2 1Centro de Microbiologia y Biologia Celular, Instituto Venezolano de Investigaciones Cientificas (IVIC), Apartado 21827, Caracas 1020 A, Venezuela and 2Centro de Virologia Animal CEVAN-CONICET, Serrano 661, 1414 Capital Federal, Argentina
A panel of 10 monoclonal antibodies produced after immunization with two porcine subgroup I rotavirus strains (OSU and A46), and directed against the major inner capsid protein (VP6), fell into six patterns of reactivity when tested against a collection of human and animal group A rotavirus strains. Monoclonal antibodies of pattern 1 recognized all rotavirus strains. Antibodies of patterns 2 and 3 recognized all subgroup II strains and some, but not all, subgroup I strains. Pattern 4 antibodies identified all subgroup I strains and two strains (H2, equine; C C l l 7 , porcine) not reactive with reference subgroup monoclonal anti-
bodies (strains non-I non-II). Pattern 5 antibody exhibited the same reactivity as pattern 4 except for not recognizing the non-I non-II equine strain. Pattern 6 antibodies reacted exclusively with subgroup I and non-I non-II rotaviruses of porcine origin. By competitive binding assays, monoclonal antibodies of patterns 4, 5 and 6 appeared to recognize a single antigenic site, which included at least three overlapping epitopes. In immunoblots all monoclonal antibodies, except one, recognized only the trimeric, but not the monomeric form of VP6.
The most abundant protein of the rotavirus virion is the product of gene 6, VP6, a 45K polypeptide which forms approximately 50 to 6 0 ~ of the virion mass (Novo & Esparza, 1981; Liu et al., 1988). This protein is highly conserved, bearing a group antigen common to all group A rotaviruses (Greenberg et al., 1983). Because of its abundance and its antigenic cross-reactivity, this protein is probably the main target of the various immunological tests used for rotavirus detection. In addition to the common group reactivity, VP6 of most mammalian rotaviruses has been shown to carry subgroup specificities that permitted establishment of the first antigenic classification of rotaviruses (Kapikian et al., 1981) into either subgroup I (comprising most animal and some human rotaviruses) or subgroup II (including mainly human and some porcine rotaviruses). Availability of subgroup I- and subgroup II-specific monoclonal antibodies (MAbs) (Greenberg et al., 1983) has allowed study of the distribution of these antigenic markers among human and animal rotaviruses and detection of two additional types of rotavirus strains, either exhibiting both specificities or neither of them (non-I non-II) (Greenberg et al., 1983; Hoshino et al., 1987; Mattion et al., 1989). Little information is available on other specificities on rotavirus VP6 (Hoshino et al., 1987). The present communication reports the presence on the VP6
of porcine subgroup I rotaviruses of five additional types of antigenic determinants, detected by a panel of MAbs, differing from the common group and the classical subgroup I. One subset of MAbs recognized all subgroup I strains and also some strains classified as non-I non-II, suggesting a subgroup I specificity different from the classical one (Greenberg et al., 1983). Another subset of MAbs appears to be specific for subgroup I rotaviruses of porcine origin. Rotavirus strains DSI, Wa, Gottfried, EDIM and H2 were kindly supplied by Y. Hoshino and M. Gorziglia (NIAID, Bethesda, Md., U.S.A.). Strain YM was provided by C. Arias, Universidad Autonoma de Mexico, Cuernavaca. Isolation and characterization of porcine rotavirus isolates from Argentina (C95, C134, C135, C117 and CC117) have been described (Mattion et al., 1989). Characterization of porcine rotavirus isolates from Venezuela (strains A46, A253, A130) will be described in a separate report (F. Liprandi, M. Hidalgo, L. Palencia, I. Rodriguez, J. E. Ludert, C. Pina, A. Sanchez, J. Esparza, unpublished results). All strains were grown in MA 104 cells essentially as previously described (Gorziglia et al., 1985). The stocks of all strains used were pretested by subgroup-specific ELISA and polyacrylamide electrophoresis of genomic RNA. Rotavirus antigens used in the ELISAs consisted of clarified cell
0000-9318 © 1990SGM
1396
Short communication
culture lysates, unless otherwise indicated. Two types of sandwich ELISAs were used. In the first (PCE) a rabbit polyclonal antibody was used for capturing the antigen and MAb preparations (hybridoma supernatants) as detecting antibody. In the MAb capture assay (MCE) the solid phase was coated with a dilution of ascitic fluid (1/100 to 1/1000) or of afffinity-purified MAbs (0-2 to 2 ~tg/well), previously determined as optimal for each MAb. Bound virus was detected by a rabbit anti-OSU rotavirus hyperimmune serum and a peroxidase-conjugated goat anti-rabbit IgG serum (Sigma). The cut-off values for PCE and MCE were absorbance values of 0.1 and 0.2, respectively. Virus samples were classified as previously described (Greenberg et al., 1983; Liprandi et al., 1987), using subgroup I- and subgroup II-specific MAbs (255/60 and 631/9 respectively), kindly provided by Dr H. Greenberg, Stanford University, Stanford, Ca., U:S~A. Purified preparations of the porcine rotavirus strains OSU and A46 were used for immunization of BALB/c female mice. A standard immunization schedule and fusion protocol were used (Greenberg et al., 1983; Liprandi et al., 1986). Nine different MAbs from four separate fusion experiments were selected for further characterization on the basis of ELISA reactivities more restricted than the one exhibited by MAbs directed to common group A specificities. In all assays a MAb against a common group determinant (O-4B2; Liprandi et al., 1986) was included as a reference antibody. The collection of MAbs was tested by immunoblotting, performed as described (Gorziglia et al., 1988), using single-capsid particles with or without heating the protein dissociation mixture (Fig. 1a). In the electrophoretic profile of the sample heated at 100 °C five polypeptides are evident, of approximate sizes of 120K (VP1), 96K (VP2), 92K (a possible cleavage product of VP2), 88K (VP3; Liu et al., 1988) and 45K (VP6) (Fig. 1 a). Omitting the heating step, an additional polypeptide of approximately 135K appears which replaces VP6 and corresponds to its trimeric form (Gorziglia et al., 1985; Sabara et al., 1987). In immunoblots nine of the 10 MAbs described and the subgroup I reference MAb recognized the trimeric but not the monomeric form of VP6 (Fig. 1b). The remaining antibody, O-2C5, failed to recognize either form of VP6. Monoclonal antibodies were tested as ascitic fluid and as affinity-purified immunoglobulins by MCE against a panel of human and animal rotavirus strains. Table 1 summarizes the established reactivity patterns. The MAbs studied could be subdivided into six different patterns of reactivity. MAb O-4B2 (pattern 1) identified all rotaviruses tested. MAbs A-1C5 and O-2C5 (patterns 2 and 3) recognized all subgroup II strains, and some (porcine and, in the case of MAb O-2C5, rhesus monkey)
(a)
(b) 1
2
1
2
3 4 5
6
7
8
9 10
Fig. 1. Western blot recognition of the trimeric form of OSU VP6 by MAbs. (a) PAGE analysis of rotavirus single-capsid proteins, stained with Coomassie blue. Proteins in the dissociation mixture were either incubatedat t00 °C (lane 1) or kept at room temperature (lane 2). (b) A mixture of single-capsid proteins incubated in the two conditions were separated by PAGE, electroblotted and reacted with antibodies. Lane 1, rabbit hyperimmune serum; lane 2, O-4B2; lane 3, A-4B5; lane 4, A-3E3; lane 5, A-2B9; lane 6, O-1G3; lane 7, A-4E10; lane 8, 255/60; lane 9, O-2C5; lane 10, control ascitic fluid.
but not other (bovine, human) subgroup I rotavirus strains. Given the high degree of conservation of VP6, the patterns of reactivity of these MAbs can be taken to suggest a division of subgroup I rotaviruses into bovinehuman, simian and porcine evolutionary lineages as has been proposed by Hoshino et al., (1987). MAb O-2C5 recognized subgroup I porcine strains from U.S.A. (OSU), Argentina (C134, C95, C135, Cl17), Mexico (YM) and Venezuela (A253, A130), but did not react with eight porcine subgroup I cell-adapted strains from Venezuela, mostly isolated from the same farm. This indicates that variation of the corresponding epitope occurs in rotaviruses circulating within one host species. Pattern 4 was represented by MAbs A-4B5, A-3E3 and O-1G3 which recognized all subgroup I strains but differed from the subgroup I reference antibody 255/60 in that they also reacted with the porcine and equine nonI non-II strains (CC117 and H2). This reactivity pattern is suggestive of the presence of a subgroup I epitope different from the one recognized by the reference antibody. Although different MAbs have been described with subgroup I specificity, no comparisons have been reported to assess whether they recognize the same or different epitopes. Sequence data indicate that more than one region of VP6 may contribute to epitopes of each subgroup (Gorziglia et al., 1988). The reactivity of
Short communication
1397
Table 1. ELISA reactivity patterns of MAbs directed to the VP6 of porcine rotaviruses with strains of human and animal rotaviruses M A b representative of each reactivity pattern*
Virus strain
Species
Subgroupt
O-4B2 l
A-1 C5 2
O-2C5 3
A-4B5 4
OSU A46 CC 117 NCDV DS1 MMU18006 H2 Wa Gottfried EDIM
Porcine Porcine Porcine Bovine Human Simian Equine Human Porcine Murine
I 1 nlnlI I I I nlnlI II II nlnlI
+ + + + + + + + + +
+ + + + + + -
+ + +/+ + + +
+ + + + + + + -
A-2B9 5 + + + + + +/-
A-4EI0 6 + + + -
* ELISA results, obtained using M A b s as capture antibodies (MCE), are expressed as - , + / - , + indicating A values of respectively < 0.2, >/0.2 and < 0.5,/> 0.5. Prefixes O and A assigned to MA bs indicate the rotavirus strain used for i mmuni z a t i on of mice (strains OSU and A46, respectively). t Subgroup is defined by the reaction with MAbs 631/9 (subgroup II) and 255/60 (subgroup I) (Greenberg et al., 1983). Strains not reactive with either of the two antibodies are indicated as n l n l I .
Table 2. Competitive ELISAs measuring binding of biotinylated compared to unlabelled VP6-specific MAbs to OSU rotavirus*
Pattern1" 1 4 5 6
Unlabelled competing antibody
Biotinylated M A b O-4B2
A-4B5
A-3E3
A-2B9
A-4E 10
O-4B2 A-4B5 A-3E3 A-2B9 A-4E 10
10 30 300 > 3000 > 3000
30 < 10 30 10 1000
30 < 10 < 10 10 1000
1000 < 10 < 10 < 10 10
< 10 < 10 < 10 1_.00
* Results are expressed as the concentration (ng/weU) of the purified unlabelled competitor antibody required to produce > 6 6 ~ inhibition of binding of a fixed amount of biotinylated antibody. Homologous reactions are underlined. t Based on the pattern of reactivity shown in Table 1. :~ E n hancemen t of binding ( > 80 ~ ) in a range of concentrations of the unlabelled competitor between 10 and 3000 ng/well.
the two non-I non-II strains with pattern 4 MAbs suggests that these strains are evolutionarily related to subgroup I strains, as has been shown for strain H2 on the basis of sequence homology data (Gorziglia et al., 1988). The reactivity of our group 4 MAbs with human non-I non-II strains of the long electropherotype (Svensson et al., 1988) remains to be established. Pattern 5 MAbs (A-2B9, A-1G7) recognized the subgroup I strains and the porcine but not the equine non-I non-II strains. The sixth group of MAbs (A-4E 10, A-5C7) detected only porcine rotavirus subgroup I strains and the non-I non-II porcine strain CC117. The great majority of circulating porcine strains are either of subgroup I or of subgroup II (Bohl et al., 1984; Liprandi et al., 1987; Mattion et al., 1989). The porcine origin of strain CC117 is suggested by its reactivity with the porcine-specific MAbs of pattern 6. The topographic relationship between the subgroup-
and porcine-specific determinants was studied by competitive ELISAs. MAbs, purified from ascitic fluids by affinity chromatography on Protein A-Sepharose, were biotinylated with the N-hydroxysuccinimide ester of biotin as described by Goding (1986). Labelled MAbs were used at the lowest concentration yielding maximal binding. For competition, 25 ~tl samples of increasing concentrations of unlabelled MAbs were added to microtitre plate wells coated with 0-1 ~tg of purified single-capsid rotavirus (OSU strain). After 30 min incubation at 37 °C, 25 ~1 of biotinylated MAbs were added and the plates incubated for a further 30 min. Binding of biotinylated MAbs was detected with streptavidin-horseradish peroxidase conjugate (Sigma), using ABTS (Sigma) as substrate. The determinants recognized by MAbs of patterns 4, 5 and 6 appeared to be related as shown by the degree of competition among them (Table 2). MAbs of patterns 4 and 5 competed in a
1398
Short communication
reciprocal fashion and all competed with the MAb of pattern 6. In the reverse reaction MAb A-4EI0 (pattern 6) competed with MAb A-2B9 (pattern 5) but not with MAbs A-4B5 and A-3E3 (pattern 4). The common group MAb O-4B2 competed efficiently with MAbs A-4B5 and A-3E3, failed to compete with MAb A-2B9 and produced an up to 1609/ooenhancement of binding of the biotinylated MAb A-4E10. Within the limitations of the technique which tends to overestimate physical proximity (Yewdell & Gerhardt, 1981) these results support the idea that pattern 4, 5 and 6 antibodies define a number of topographically overlapping epitopes within the same antigenic site. This site seems to include a porcinespecific region contiguous to a subgroup 1-specific region. The conformational nature of these determinants is suggested by the immunoblot results in which MAbs of patterns 4, 5 and 6 recognized only the trimeric but not the monomeric form of VP6. A similar observation has been made by Gorziglia et al. (1988) with subgroupspecific MAbs. In conclusion the patterns of reactivity of the MAbs described here suggest the existence of at least seven types of operationally defined epitopes on the VP6 of group A rotavirus strains. MAbs directed to a site, specific to one or a few animal species, on the most abundant protein of the rotavirus virion, might be of practical use to identify the species of origin of the virus and possibly to corroborate animal-to-human transmission in suspected cases, like subgroup I long electropherotype human isolates. We wish to thank our colleagues for providing rotavirus strains and Dr M. Gorziglia for many helpful discussions. The expert technical assistance of C. Pina and A. Sanchez is gratefully acknowledged. This work was supported by the International Foundation for Science, grant B/903-2.
References BOHL, E. H., THEIL, K. W. & SAIF, L. (1984). Isolation and serotyping of porcine rotaviruses and antigenic comparison with other rotaviruses. Journal of Clinical Microbiology 19, 312-319. GODING, J. W. (1986). Immunofluorescence. In MonoclonalAntibodies: Principles and Practice, pp. 241 280. London: Academic Press.
GORZIGLIA, M., LARREA, C., L1PRANDI, F. & ESPARZA, J. (1985). Biochemical evidence for the oligomeric (possibly trimeric) structure of the major inner capsid polypeptide (45K) of rotaviruses. Journalof General Virology 66, 1889-1900. GORZIGLIA,M., HOSHINO,Y., NISHIKAWA,K., MALOY,W. L., JONES, R. W., KAPIKIAN,A. Z. & CHANOCK,R. M. (1988). Comparative sequence analysis of the genomic segment 6 of four rotaviruses each with a different subgroup specificity. Journalof General Virology69, 1659 1669_ GREENBERG,H., McAULIFFE, V., VALDESUSO,J., WYATT,R., FLORES, J., KAL1CA,A., HOSHINO,Y. & S]NGH, N. (1983). Serological analysis of the subgroup protein of rotavirus using monoclonal antibodies. Infection and Immunity 39, 91-99. HOSHINO,Y., GORZIGLIA,M., VALDESUSO,J., ASKAA,J., GLASS,R. & KAPIKIAN, A. Z. (1987). An equine rotavirus (FI-14 strain) which bears both subgroup I and II specificities on its VP6. Virology 157, 488-496. KAPIKIAN, A. Z., CLINE, W. L., GREENBERG, H. B., WYATT, R. G., KALICA,A. R., BANKS,C. E., JAMES,H. D., FLORES,J. ~; CHANOCK, R. M. (1981). Antigenic characterization of human and animal rotaviruses by immune adherence hemagglutination assay (IAHA): evidence for distinctness of IAHA and neutralization antigens. Infection and Immunity 33, 415-425. LIPRANDI, F., BRITO, B., PALENCIA, L. & ESPARZA, J. (1986). Derivation of a monoclonal antibody against the group specific antigen of rotaviruses and its use in a diagnostic enzymatic immunoassay. Acta cientifica venezolana 37, 432 436. LIPRANDI, F., GARCIA, n., BOTERO, t., GORZIGLIA, i . , CAVAZZA, M. E., PEREZ-SCHAEL,I. & ESPARZA,J. (1987). Characterization of rotaviruses isolated from pigs with diarrhea in Venezuela. Veterinary Microbiology 13, 35-45. LIU, M., OFEIT, P. A. & ESTES, M. K. (1988). Identification of simian rotavirus SA11 genome 3 segment product. Virology 163, 26-32. MATTION, N. i . , BELLINZONI,R. C., BLACKHALL,J. O., LA TORRE, J. L. & SCODELLER,E. A. (1989). Antigenic characterization of swine rotaviruses in Argentina. Journal of Clinical Microbiology 26, 795-798. Novo, E. & ESPARZA, J. (1981). Composition and topography of structural polypeptides of bovine rotavirus. Journal of General Virology 56, 325-335. SABARA,M., READY, K. F. M., FRENCHICK,P. J. & BABIUK, L. A. (1987). Biochemical evidence for the oligomeric arrangement of bovine rotavirus nucleocapsid protein and its possible significance in the immunogenicity of this protein. Journal of General Virology 68, 123-133. SVENSSON, L., GRAHNQUIST, L., PETTERSON, C., GRANDIEN, M., STINTZING, G. & GREENBERG, H. B. (1988). Detection of human rotaviruses which do not react with subgroup I- and If-specific monoclonal antibodies. Journal of Clinical Microbiology 26, 1238-1240. YEWDELL,J. W. & GERHARD,W. (1981). Antigenic characterization of viruses by monoclonal antibodies. Annual Review of Microbiology 35, 185-206.
(Received 10 October 1989; Accepted 13 February 1990)
