Direct isolation in eggs of influenza A (H1N1) and B viruses with haemagglutinins of different antigenic and amino acid composition

Transcription

1 Journal of General Virology (1991), 72, Printed in Great Britain 185 Direct isolation in eggs of influenza A (H1N1) and B viruses with haemagglutinins of different antigenic and amino acid composition J. S. Oxford, 1. R. Newman, 2 T. Corcoran, 2 J. Bootman, 2 D. Major, 2 P. Yates, 2 J. Robertson 2 and G. C. Schild 2 1Department of Medical Microbiology, The London Hospital Medical College, Whitechapel, London E1 2AD and 2The National Institute for Biological Standards and Control, Blanche Lane, South Mimms, Hertfordshire EN6 3QG, U.K. Influenza A (H1N1) and influenza B viruses from clinical samples were isolated in the amniotic cavity of embryonated hens' eggs by classical techniques and propagated in the allantoic cavity. Virus progeny from different eggs which had been inoculated with virus material from the same clinical sample possessed antigenically distinguishable haemagglutinins (HAs). Virus progeny of some eggs possessed HAs which were serologically identical to those of virus isolated in parallel in mammalian (MDCK) cells. These egggrown viruses possessing HAs with the antigenic phenotype of mammalian cell-grown viruses appeared to be antigenically related to epidemic influenza virus because post-infection human sera reacted to high titre with the virus HA. Specific nucleotide changes were detected in the HAs of the viruses isolated directly in eggs at positions 163 and 189 for influenza A (H1N1) viruses or positions 141 and 196 to 198 for influenza B viruses. Egg-isolated viruses which possessed the antigenic phenotype of mammalian cell-grown viruses retained glycosylation sites at positions 163 and 196. The viruses isolated directly in embryonated hens' eggs which possessed the HA antigenic phenotype and glycosylation sites of MDCK cell-grown virus can, unlike the latter viruses themselves, be used as candidate influenza vaccine viruses. Virus isolation in the amniotic cavity of embryonated hens' eggs is a classical method for the direct recovery of influenza virus from a clinical specimen although, more recently, mammalian cells such as MDCK have also been recognized as a sensitive substrate for primary virus isolation and subsequent virus cultivation (reviewed by Stuart-Harris et al., 1985). However, strong selective pressures are exerted on presumed heterogeneous populations of influenza virions (Holland et al., 1982) in clinical samples by cultivation of the viruses in MDCK cells compared to their cultivation in embryonated hens' eggs (Schild et al., 1983). Mammalian cell-grown viruses possess haemagglutinins (HAs) which are antigenically (Schild et al., 1983; Oxford et al., 1987) and biochemically (Robertson et al., 1985, 1987; Wang et al., 1989) distinguishable from those of viruses isolated from the same clinical specimen in embryonated hens' eggs. Nucleotide sequence changes were noted in and around the receptor-binding pocket (Wiley & Skehel, 1987) on the HAs of mammalian cell-grown viruses compared to those of egg-grown viruses (Robertson et al., 1985, 1987; Oxford et al., 1987) and, concomitantly, antigenic changes were noted in the HA of the virus because of the close physical positioning of antigenic site B on the periphery of the receptor-binding pocket. The mamma- lian cell-grown viruses are more related antigenically to natural epidemic influenza viruses because post-infection human sera react preferentially with the HA of virus grown in MDCK cells (Oxford et al., 1987). Experimental influenza virus vaccines produced using MDCK cell-grown virus induce antibody in animals which appears to be more serologically reactive with epidemic influenza virus HA than does a conventional vaccine containing HA of egg-grown virus (Wood et al., 1989). However, MDCK cells are not a licensed substrate for the cultivation of influenza vaccine viruses for human administration. We now report that influenza A and B viruses, adapted to grow directly in embryonated hens' eggs without previous passage in MDCK cells, may possess an HA with an antigenic phenotype identical to that of the HA of an MDCK cell-cultivated virus. These viruses can be used as candidate influenza vaccine viruses for human administration. Furthermore, the isolation in eggs of at least two viruses with differing antigenic and biochemical properties from a clinical sample has important implications in the continued use of eggs for epidemiological studies of virus antigenic drift. Influenza A and B viruses were isolated from clinical samples obtained at Christ's Hospital School (Horsham, SGM

2 186 Short communication Table 1. Antigenic analysis of HAs from independent influenza A (H1N1) virus isolates from a single clinical sample HI titre with human HI titre with anti-ha monoclonal antibodies polyclonal sera Passage Virus history* G aml all < < > M MDCK1 < < > E al 1 > > < < 100 < < 10 * am, Amniotic cavity; al, allantoic cavity (numbers denote passages). U.K.) (Oxford et al., 1983; Hoskins et al., 1979) using classical methods of cultivating influenza virus in the amniotic and allantoic cavities of 10-day-old embryonated hens' eggs. Unless otherwise stated the same egg was inoculated in both the amniotic and allantoic cavities. Amniotic and allantoic fluids were pooled and tested for virus after 3 days incubation at 35 C by agglutination of 0.5~ human erythrocytes; viruses were thereafter passaged once or twice in the allantoic cavity of 10-day-old hens' eggs. Parallel virus isolation was also carried out directly from the clinical specimen in MDCK cells in the presence of trypsin in the cell culture medium (Schild et al., 1983). Monoclonal antibodies to virus HA were obtained by standard techniques and mice were immunized with purified and concentrated (10 mg virus protein/ml) influenza viruses cultivated in MDCK cells or in embryonated hens' eggs. For nucleotide sequence analysis, purified virus was disrupted in 10 mm-tris- HCI buffer ph 7.5 containing 5 mm-edta, 0"5~o SDS and treated with 50 ~tg/ml proteinase K (BDH) for 30 min at 37 C. Total virion RNA was extracted twice with phenol-chloroform, once with chloroform and was then ethanol-precipitated. Nucleotide sequences were obtained by the dideoxynucleotide chain termination technique (Sanger et al., 1977) using reverse transcriptase (Anglian Biotechnology), synthetic oligonucleotide primers and virion RNA as described previously (Robertson et al., 1985; Caton et al., 1982; Krystal et al., 1982). Reaction products were analysed on 8 ~ polyacrylamide, Tris-borate-urea sequencing gels. A serological analysis, using monoclonal and polyclonal antibodies, of the HA of three independently obtained influenza A (H 1N 1) isolates from a single nasal wash of a boy at Christ's Hospital School is shown in Table 1. Monoclonal antibodies which reacted with the HA of the MDCK cell-grown virus (l 57M), such as 7, 23, 41, 61 and 725, also reacted with HA of the virus 157G which had not been passaged in MDCK cells but been isolated directly from the same clinical sample in the amniotic and allantoic cavity of embryonated hens' eggs. Conversely, monoclonal antibodies 3 and 11 failed to react with the HA of virus 157G. In addition, the HA of virus 157G was serologically indistinguishable from the HA of the MDCK cell isolate (157M) by polyclonal antibody analysis (Table 1). In contrast, another virus, 157E, isolated directly in the allantoic cavity of embryonated hens' eggs from the same clinical sample, could be distinguished serologically both from the 157M and 157G virus HAs, particularly with monoclonal antibodies 3 and 11 and by a lower, or absence of, serological reactivity with monoclonal antibodies 7, 23, 41, 61 and 725. Human post-infection antibodies also distinguished virus 157E from both 157M and 157G isolates. In subsequent experiments both 157G and 157E were reisolated from the original clinical sample by direct amniotic inoculation followed by allantoic cavity passage with identical results. Influenza B viruses from a single clinical nasal wash sample were also able to adapt and grow in embryonated hens' eggs and had serologicauy distinguishable HAs (Table 2). Thus, monoclonal antibodies 238 and 21, which normally recognize the HA of egg-grown viruses (Schild et al., 1983), failed to react with the HA of virus B/EFR/83G, although this virus was recovered directly in eggs from the same clinical specimen as B/EFR/83E (Oxford et al., 1990). Polyclonal human sera reacted to high haemagglutination inhibition (HI) titres with the MDCK cell isolate and with B/EFR/83G but reacted with a significantly lower titre with B/EFR/83E. Thus, influenza B viruses, in addition to influenza A (H1N1) viruses, may adapt and replicate directly in hens' eggs and have HAs possessing an antigenic phenotype similar to that of virus isolated on MDCK cells. Nucleotide sequence analyses were carried out on the HAs of the various egg- and MDCK cell-grown influenza A (H1N1) viruses (Table 3). The HA of the MDCKgrown virus 157M possessed typical nucleotide sequence differences compared to egg-grown virus 157E at

3 Short communication 187 Table 2. Antigenic analysis of HA from influenza B viruses isolated in eggs or in MDCK cells from a single clinical sample HI titre with anti-ha monoclonal antibodies HI titre with human polyclonal sera Passage Virus* history t B/EFR/83M MDCK < < B/EFR/83E HET am < B/EFR/83G all < < * Note that B/EFR/83E and B/EFR/83G viruses were designated pool D and B respectively in previous studies (Oxford et al., 1990) and are redesignated here to allow a comparison with 157M, 157E and 157G (Table 1). t am, Arnniotic cavity; al, allantoic cavity; HET, human embryo trachea organ culture (Oxford et al., 1990). Table 3. Amino acid sequence of HA1 of influenza A (H1N1) viruses isolated from a single clinical sample in MDCK cells (157M) or in eggs (157E and 157G) Virus HA 157M 157E 157G 157M 157E 157G Amino acid sequence* NGSYPNLSKSYVNNKEKEVLVLWGVHHPSNIED S K N E QKTIYRKENAYVSVV SSHYNRRFTPEIAKRPKVRDQ D G * For isolates 157E and 157G only the amino acids at positions known to vary are shown. The remaining sequence is identical. positions 163, and also at position 189 of HA1 (H3 numbering) (Robertson et al., 1987). In the former position a glycosylation site is present on the HA of MDCK cell-grown virus (157M) but not in the conventional egg-grown virus, 157E. A single nucleotide difference was detected between the HA of 157G and 157M viruses, at amino acid position 225 in the receptorbinding pocket of the HA molecule, which resulted in an altered amino acid residue (Table 3). Such a substitution presumably counteracts the changes at positions 163 and 189 in the egg-grown virus 157E and permits replication of the 157G virus in the eggs, whereas 157M replicates very poorly or not at all in eggs (data not presented). We have reported previously that influenza A (H1N1) viruses, after direct isolation in MDCK cells and subsequent passage into embryonated hens' eggs, could retain an HA with typical antigenic and biochemical properties of the MDCK-grown virus. The new finding of the present study is that such viruses may be isolated directly in eggs without prior passage in MDCK ceils. Nucleotide sequence analyses of the two direct virus isolates of influenza B virus in hens' eggs showed that the isolate which possessed an MDCK cell-like antigenic phenotype for the HA, namely B/EFR/83G, possessed an amino acid substitution (arginine substituted for glycine) at position 141 on the loop near the receptorbinding site of the HA molecule (Oxford et al., 1990). In addition, this egg-grown virus B/EFR/83G possessed the glycosylation sequence Asn-X-Thr at positions 196 to 198, which is normally detected only in influenza B viruses cultivated in MDCK cells and which was not present, for example, in B/EFR/83E. The substitution of glycine by arginine would result in an increased positive charge at this site, presumably affecting binding and allowing this virus (and not the antigenically identical, but MDCK cell-grown virus) to replicate in eggs. Previous studies by Robertson et al. (1987) showed that influenza A (H1N1) viruses isolated firstly on MDCK cells and then passaged into hens' eggs could maintain an HA with an antigenic phenotype closely similar to that of virus cultivated directly on MDCK ceils. However, the previous passage on MDCK cells, as well as acting as an amplification step, prevented the use of such viruses as potential human vaccine viruses on

4 188 Short communication safety grounds. The observations reported here have enabled us to develop a new method of selecting influenza viruses for use as a vaccine (British Patent No ); HAs of viruses isolated directly in embryonated eggs and which possess the presumed optimal antigenic phenotype of HA of an MDCK cell-grown virus are identified. Two reassortant viruses, designated NIB-14 and NIB-17, were constructed from the viruses 157G and 157E and used to prepare experimental inactivated influenza vaccines. Although the HAs of the two viruses differed by only three amino acids in the sequence of HA1, a different immune response was detected in students immunized with the two vaccines (C. W. Potter, R. Jennings and J. S. Oxford, unpublished results). Further studies are required to define precisely the optimal conditions under which such MDCK cell-like viruses may be selected and grown in hens' eggs. A possible factor may be genetic variability in the embryonated hens' eggs which exerts different selective pressures on the virus. In a separate study of multiple influenza A (H1N1) viruses isolated during an epidemic at Christ's Hospital School (Oxford et al., 1987), we estimated, retrospectively, that approximately half of the viruses isolated directly in the amniotic and allantoic cavities of eggs were of the MDCK cell antigenic phenotype. Burnet & Clarke (1942) were the first to describe the biological heterogeneity of influenza A viruses in clinical samples when they derived at least two such virus populations by limited passage in the allantoic versus the amniotic cavity of a single embryonated hens egg. The 'O' (original) and 'D' (derived) viruses differed in their ability to agglutinate chicken erythrocytes but no antigenic differences were detected between the HAs of the two subpopulations of viruses. Recently, we have analysed 'O' and 'D' variants of the above influenza A (H1N1) viruses and have observed that selective pressures do occur and result in antigenic variants but the viruses are different from those resulting in HAs with an 'egg-like' and 'MDCK cell-like' antigenic phenotype described here (A. Azzi, O. Bartolomei and J. S. Oxford, unpublished results). Our current hypothesis to explain the findings detailed above is that such clinical samples contain a heterogeneous virus population (Patterson & Oxford, 1986) and that cells of developing chick embryos, themselves of mixed genetic background, exert differential selection, so allowing the replication and emergence of variants with HAs of differing amino acid sequence and antigenicity. The viruses possessing an 'MDCK cell-like' antigenic phenotype, isolated directly in eggs, are not stably maintained on passage in the allantoic cavity (J. S. Oxford, unpublished data). It would appear likely that this naturally occurring phenomenon can be exploited in the future to select the most appropriate virus strain for incorporation into influenza vaccines. Current influenza vaccines have an estimated 70 to 80 9/00 protective efficacy (reviewed by Stuart-Harris et al., 1985; Zuckerman et al., 1990) and a contributory factor towards this rather low efficacy could be antigenic mismatching of vaccine viruses and epidemic viruses. Our findings may also have direct relevance to epidemiological interpretations of the yearly antigenic changes in the HAs of influenza viruses cultivated in eggs noted over the last 4 decades. If differential selective pressures are exerted at the first stage of virus replication in avian cells, these may result in the emergence and subsequent laboratory analysis of viruses which are antigenically unrepresentative of natural field isolates. We are now extending our study to include antigenic and genetic analysis of other influenza viruses isolated directly in eggs, including influenza A (H3N2) viruses. Wang et al. (1989) have isolated a group of influenza A (H3N2) viruses directly in eggs and separately in MDCK cells. A single antigenic phenotype was identified in the HA of the MDCK-grown viruses, whereas five distinct antigenic phenotypes were identified in the egg-grown counterparts. Furthermore, most of the egg-grown viruses were mixtures of one or more antigenic phenotypes. We would like to thank our colleagues Drs E. Grilli and J. Smith, PHLS Laboratories, Guildford, U.K., Dr D. A. J. Tyrrell, MRC Common Cold Unit, Salisbury, U.K. and Dr A. Azzi and the late Dr O. Bartolomei, University of Florence, Italy, and Dr R. G. Webster, Children's Hospital, Memphis, Tenn., U.S.A. for influenza viruses and monoclonal antibodies. Financial assistance is provided to J.S.O. by the Oretta Bartolomei-Cruzi Research Foundation, Florence, Italy, and by the Medical Research Council. References BURNET, F. M. & CLARKE, E. (1942). Walter and Eliza Hall Institute, Monograph No. 4. Melbourne: Macmillan. CATON, A. J., BROWNLEE, G. G., YEWDELL, J. W. & GERHARD, W. (1982). The antigenic structure of the influenza virus A/PR/8/34 haemagglutinin (HI subtype). Cell 31, HOLLAND, J., SPINDLER, K., HORODYSKI, F., GRAEAU, E., NICHOL, S. & VAN DE POL, S. (1982). Rapid evolution of RNA genomes. Science 215, HOSKINS, T. W., DAVIES J. R., SMITH, A. J., MILLER, C. & ALLCHIN, A. (1979). Assessment of inactivated influenza A vaccine after three outbreaks of influenza A at Christ's Hospital. Lancet i, KRYSTAL, M., ELLIOTT, R. M., BENZ, E. W., YOUNG, J. F. & PALESE, P. (1982). Evolution of influenza A and B viruses: conservation of structural features in the hemagglutinin genes. Proceedings of the National Academy of Sciences, U.S.A. 79, OXFORD, J. S., ABBO, H., CORCORAN, T., WEBSTER, R. G., SMITH, A. J., GRILLI, E. A. & SCHILD, G. C. (1983). Antigenic and biochemical analysis of field isolates of influenza B virus: evidence for intra- and inter-epidemic variation. Journal of General Virology 64, OXFORD, J. S., CORCORAN, T., KNOTT, R., BATES, J., BARTOLOMEI, O., MAJOR, D., NEWMAN, R. W., ROBERTSON, J. S., WEBSTER, R. G. & SCHILD, G. C. (1987). Serological studies with influenza A (H1N1) viruses cultivated in eggs or in a canine kidney cell line (MDCK). Bulletin of the World Health Organization 65,

5 Short communication 189 OXFORD, J. S., SCFIILD, G. C., CORCORAN, T., NEWMAN, R., MAJOR, D., ROBERTSON, J. S., BOOTMAN, J. S., HIGGINS, P. G., NAKIB, W. A., BARROW, I. & TYRRELL I). A. J. (1990). A host cell selected variant of influenza B virus with a single nucleotide substitution in HA affecting a potential glycosylation site was attenuated in virulence for volunteers. Archives of Virology 110, PATTERSON, S. & OXFORD, J. S. (1986). Analysis of antigenic determinants of internal and external proteins of influenza virus and identification of antigenic subpopulations of virions in recent field isolates using monoclonal antibodies and immunogold labelling. Archives of Virology 88, ROBERTSON, J. S., NAEVE, C. W., WEBSTER, R. G., BOOTMAN, J. S., NEWMAN, R. & SCmLD, G. C. (1985). Alterations in the hemagglutinin associated with adaptation of influenza B virus to growth in eggs. Virology 143, ROBERTSON, J. S., BOOTMAN, J. S., OXFORD, J. S., DANIELS, R. S., WEBSTER, R. G. & SCmLD, G. C. (1987). Structural changes in the hemagglutinin which accompany egg adaptation of an influenza A (H1N1) virus. Virology 160, SANGER, F., NICKLEN, S. & COULSON, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences. U.S.A. 74, SCHILD, G. C., OXFORD, J. S., DE JONG, J. C. & WEBSTER, R. G. (1983). Evidence of host cell selection of influenza virus antigenic variants. Nature, London 303, STUART-HARRIS, C. H., SCHILD, G. C. & OXFORD, J. S. (1985). Influenza. The Viruses and the Disease. London: Edward Arnold. WANG, M., KATZ, J. M. & WEBSTER, R. G. (1989). Extensive heterogeneity in the hemagglutinin of egg-grown influenza viruses from different patients. Virology 171, WILEY, D. C. & SKEHEL, J. J. (1987), The structure and function of the hemagglutinin membrane glycoprotein of influenza virus. Annual Review of Biochemistry 56, WOOD, J. M., OXFORD, J. S., DUNLEAVY, V., NEWMAN, R. W., MAJOR, D. & ROBERTSON, J. S. (1989). Influenza A (H1N1) vaccine efficacy in animal models is influenced by two amino acid substitutions in the hemagglutinin molecule. Virology 171, ZUCKERMAN, M., WOOD, J., TAYLOR, J. OXFORD, J. S. (1990). The influenza A (H3N2) component of vaccine induces antibody to the current virus. Lancet i, 179. (Received 6 July 1990; Accepted 4 October 1990)