Single amino acid codon changes detected in ... - Semantic Scholar
Journal of General Virology (1993), 74, 931-935. Printed in Great Britain
931
Single amino acid codon changes detected in louping ill virus antibodyresistant mutants with reduced neurovirulence W . R. J i a n g , 1 A. L o w e , ~ S. H i g g s , ~ H . Reid 2 and E. A. G o u l d ~* ~NERC Institute of Virology and Environmental Microbiology, Mansfield Road, Oxford OX1 3SR and 2Moredun Research Institute, 408 Gilmerton Road, Edinburgh EH17 7JH, U.K.
Seven mutant viruses were derived from a Scottish strain of louping ill virus using a virus envelope-specific neutralizing monoclonal antibody. None of the mutants was neutralized and immunofluorescence microscopy confirmed that they did not bind to this antibody. Four mutants showed reduced mouse neurovirulence compared with parent virus and two mutants failed to induce protective immune responses in mice challenged with virulent tick-borne encephalitis virus. The mutants with the lowest virulence showed poor or undetectable haemagglutinating activity. The nucleotide sequence of the envelope glycoprotein gene of each of the seven
mutants was determined and the deduced amino acid sequence was compared with parent virus. For each mutant, only a single amino acid codon change was detected and all the amino acid substitutions occurred within amino acid positions 308 to 311. A change from the amino acid aspartate to asparagine at amino acid position 308, which represented a potential glycosylation site, was the most effective substitution in reducing mouse neurovirulence. The results demonstrate the importance of critical sites within the envelope glycoprotein as determinants of virus virulence.
Louping ill (LI) virus, a tick-transmitted member of the virus family Flaviviridae (Westaway et al., 1985), causes an encephalitic or encephalomyelitic disease of sheep and grouse. The virus is prevalent in the upland sheep grazing areas of Scotland, northern England, Wales and southwest England. Other wildlife species can also become infected with LI virus if they serve as hosts for the feeding tick vector Ixodes ricinus, but it is believed that these vertebrates are dead-end hosts because they do not develop high viraemic titres (for review, see Reid, 1987). LI virus is genetically (Shiu et al., 1991) and antigenically (De Madrid & Porterfield, 1974; Porterfield, 1980; Heinz, 1986; Calisher et al., 1989) closely related to other European tick-borne encephalitis (TBE) viruses, many strains of which produce severe and often fatal encephalitis in humans. The factors that determine why some strains of TBE virus are virulent have not been determined yet but studies with antibody-resistant mutants have shown that the viral envelope (E) protein contributes significantly to flavivirus virulence (Holzmann et al., 1990; Cecilia & Gould, 1991). It is believed that the global climate is changing and that in the future the United Kingdom is likely to experience an increase in temperature. This will inevitably lead to altered land usage and the likelihood of redistributed arthropod populations, including ticks. In addition, there is an increasing possibility that virulent strains of TBE could be introduced into the U.K. from
mainland Europe. Thus, there is a need to understand the molecular basis of flavivirus virulence so that appropriate action can be taken to control these viruses in the environment. In this paper we report the preparation and characterization of seven mutants of LI virus which were derived using a potent neutralizing monoclonal antibody (MAb). Nucleotide sequencing of the viral E gene of each mutant has revealed an important amino acid that appears to be closely associated with LI virus virulence. Continuous lines of porcine kidney (PS) cells were grown in Eagle's MEM with 5 % fetal calf serum. LI virus (strain 369/T2) was described previously (Shiu et al., 1991). It was plaque-purified in PS cells and then passaged once through suckling mouse brains before selection of antibody-resistant mutants, as described below. TBE virus was a central European strain kindly supplied by Dr J. S. Porterfield. Its antigenic characteristics were reported previously (Gould et al., 1985). MAbs were selected for their capacity to bind the E protein of LI virus. They have all been described previously (Gould et al., 1985; Venugopal et al., 1992). MAbs 4.2 and 7.1 neutralize the infectivity of LI virus in plaque reduction neutralization tests and MAb 7.1, but not MAb 4.2, also shows haemagglutination (HA)inhibiting activity with LI virus. Escape mutants were prepared using plaque-purified LI virus, which was diluted to 100 p.f.u./ml and mixed with the selecting
0001-1384 © 1993 SGM
932
Short communication
Table 1. Biological and molecular characteristics o f L I virus escape mutants Virus
HA (units)
Infectivity (log10 p.f.u./ml)
522 525 523 529 520 521 526 Parent
< 2 < 2 < 2 8 64 16 64 32
8'98 8.40 8'60 8.30 7.54 8.48 9'00 8.18
LDs0 (loga0 p.f.u. to kill 50% mice) > 6.5 5.8 > 6.1 > 5.5 2.3 3'0 3.6 2.2
Plaque size (mm) 1-1.5 1-1.5 3-5 3 5 3-5 3-5 3-5 3 5
Nucleotide change* G to A G to A U to C U to C A to C G to U G to C None
(922) (922) (928) (928) (931) (933) (933)
Amino acid codon change* D to N (308) D to N (308) S to P (310) S to P (310) K to Q (311) K to N (311) K to N (311) None
* N u m b e r s in parentheses indicate position of mutation in the E gene.
MAb 4.2 (diluted 1 : 100; neutralization titre 1 : 10000) in Eagle's MEM. The mixture was incubated at 37 °C for 1 h and was then adsorbed onto PS cell monolayers, again for 1 h at 37 °C. The cell monolayers were washed and medium containing the same concentration of MAb 4.2 and 0.8 % agarose was added. Non-neutralized virus was recovered and plaque-purified twice more under the same experimental conditions. The mutant viruses were tested again for their binding capacity with MAb 4.2. They were all negative. For H A tests, the parent L1 virus and each escape mutant was grown in PS cells. Supernatant medium from the infected cells was harvested at day 5 post-infection and was concentrated using 7 % polyethylene glycol and 0.4 M-NaC1. The concentrated virus was precipitated by centrifugation and was resuspended in borate-buffered saline at pH 9.0 (Clarke & Casals, 1958) at a concentration of 50-fold. The HA test was carried out at pH 6.2 (Clarke & Casals, 1958). For neurovirulence tests, groups of 10 3-week-old female mice (strain TO, Tuck and Son) were inoculated either intracerebrally (i.c.) or subcutaneously (s.c.) with serial 10-fold dilutions of parent or mutant virus. Sickness and mortality were recorded for 18 days. The mice that survived s.c. inoculation of parent or mutant virus were challenged s.c. with TBE virus at 100 LDs0/mouse. For viral RNA preparation and nucleotide (nt) sequence analysis, suckling mouse brain preparations of plaque-purified MAb-resistant mutants were produced using newborn mice. The viral R N A was extracted directly from the infected suckling mouse brain suspensions using guanidinium thiocyanate (Chomczynski & Sacchi, 1987) and first strand c D N A was synthesized using either the downstream primer EXP2 (TGCag a t c t C T A C T A C G C T C C C A C C C C G A G A G T , nt positions 2341 to 2358) or LIE4 (CCCggtaccAAGCTGC A T C T C T A T G A , nt positions 1982 to 1998) and then amplified by PCR using EXP1 (CTTagatctaatATGG T C G C C G T T G T G T G G C T A , nt positions 781 to 801) and EXP2, or LIE3 ( T T T g g t a c c C C T C A T G C T G T C A A G A T G , nt positions 1609 to 1629; Shiu et al., 1991)
and LIE4. The PCR-amplified D N A was cloned using p U C l l 8 . At least two subclones of each virus were sequenced in both directions by double-stranded dideoxynucleotide sequencing (Sanger et al., 1977). Seven plaque-purified neutralization-resistant mutants of LI virus strain 369/T2 (Shiu et al., 1991) were derived using MAb 4.2 as described above. Each mutant was then grown in PS cells both in the presence and absence of MAb 4.2, the selecting MAb. There were no major differences in the infectivity titres of the mutant viruses when grown under these conditions. PS cell monolayers, infected with the mutant viruses, were analysed by indirect immunofluorescence (IIF) microscopy for reactivity with MAbs that were known to react with standard LI virus (Venugopal et al., 1992). Each mutant reacted with all the MAbs except MAb 4.2 and MAb 7.1 suggesting that the epitopes represented by these two MAbs are very closely associated. It is important to note that MAb 7.1, but not MAb 4.2, inhibits HA by LI virus and these two MAbs are therefore considered to represent different but closely associated epitopes on the envelope glycoprotein of LI virus (Venugopal et al., 1992). Cultures of the parent 369/T2 strain of LI virus and each of the seven escape mutants were concentrated from the supernatant culture medium of infected PS cell monolayers as described above. Each preparation of virus was then tested for H A activity. The parent virus and four escape mutants (520, 526, 521 and 529) were positive in H A tests but the other three escape mutants (522, 525 and 523) were negative (Table 1). The size of plaque produced by each mutant when grown in PS cells under carboxymethylcellulose overlay was compared with the parent virus, which produced plaques of 3 to 5 mm in 4 days at 37 °C. Two mutants (522 and 525) produced small plaques (1 to 1.5 ram), whereas the other mutants were similar to the parent virus. The neurovirulence of parent 369/T2 LI virus and each of the seven escape mutants was analysed in newborn and 3-week-old mice by i.c. challenge and also in 3-week-old mice by s.c. inoculation of serial 10-fold
Short communication
Table 2. Detection o f antibody response in mice following inoculation with escape mutants
Challenge virus
Proportion of mice positive by IIF
522 525 523 529 520 521 526 Parent
1/5 2/5 3/4 7/7 5/5 4/5 5/5 5/5
Neutralization titre of serum < 10 40 < 10 < 10 40 160 10 40
Proportion of mice surviving TBE virus challenge* 2/49 3/25 17/20 18/19 13/14 13/13 13/15 13/13
*Mice givena sub-lethaldose of LI virus were subsequentlychallenged with a lethal s.c. dose of TBE virus. dilutions of each virus. Four of the escape mutants (522, 523, 525 and 529) had reduced neurovirulence compared with parent virus, whereas three mutants (520, 521 and 526) had retained the virulence of the parent virus. All viruses were virulent for newborn mice and 3-week-old mice if the virus was given i.c., demonstrating that even the low virulence mutants had retained the capacity to infect and multiply in susceptible mouse brain cells. The neurovirulence titrations for subcutaneously infected 3-week-old mice are shown in Table 1. Groups of 3-week-old mice were inoculated s.c. with an estimated 100 p.f.u, of each virulent escape mutant (520, 521 and 526) or each avirulent mutant (522, 523, 525 and 529). Three weeks later, serum was collected from five surviving mice of each group. These sera were tested for the presence of antibody against LI virus using an IIF test on cells infected with LI virus. The results (Table 2) show that mice inoculated with sub-lethal doses of virulent LI virus escape mutants contained detectable antibody. On the other hand, two of the escape mutants with low neurovirulence for mice (522 and 525) induced detectable serum antibody responses in only a low proportion of the mice. The surviving mice were then challenged s.c. with 100 LDs0 of TBE virus. All mice initially given sub-lethal doses of the virulent escape mutants and which possessed detectable antibody survived subsequent challenge with TBE virus, whereas all mice initially given the avirulent mutants, and which were antibody-negative, were fatally infected by the TBE virus (Table 2). These results imply that two of the avirulent mutants (522 and 525) did not multiply significantly in the mice. The complete nucleotide sequence of the gene encoding the viral E glycoprotein of two escape mutants (520 and 522) was determined. The results obtained were aligned and compared with the published nucleotide sequence for LI virus strain 369/T2. Single nucleotide changes were detected corresponding to amino acid positions
933
308 and 311 for mutants 522 and 520 respectively. The remaining escape mutants were therefore sequenced over a region of the viral E protein spanning the area in which the nucleotide substitutions had been identified (nt 756 to 1110). The nucleotide changes and the corresponding amino acid codon changes, only one in each mutant, are shown in Table 1. The two least virulent mutants (522 and 525) had amino acid substitutions at position 308. The two mutants (523 and 529) that showed partially reduced virulence had amino acid substitutions in a nearby amino acid (amino acid 310). Mutants 520, 521 and 526, which showed either small or no significant reductions in virulence, had amino acid substitutions at position 311. The change of aspartate to asparagine at amino acid position 308 produced a potential glycosylation site. We have produced and analysed the nucleotide sequences of the E protein of seven neutralization escape mutants derived using a single neutralizing MAb. Although they were isolated in separate experiments, two pairs of mutants showed identical base substitutions and could therefore represent the progeny of single clones. For each mutant analysed, only a single nucleotide change was detected in the E gene and in each case this correlated with an amino acid codon change. All codon changes occurred within a tetrapeptide (amino acid sequence 308 to 311) suggesting that the epitope identified by MAb 4.2 could be linear, and data obtained by immunoblotting MAb 4.2 with LI virus under reducing conditions support this possibility (data not presented). MAb 7.1 also lost the capacity to bind to the escape mutants implying that it identifies the same epitope as MAb 4.2. This seems unlikely since MAb 7.1, in contrast with MAb 4.2, inhibits H A activity. A similar situation was recently reported by Arbiza et al. (1992) with human respiratory syncytial virus. Their data were consistent with the concept that overlapping MAbs might recognize sites with specific local conformational differences. Therefore, it is possible that at least a part of the MAb 4.2 epitope is present in the MAb 7.1 epitope and this region of the epitope could have local conformational rearrangements. The flavivirus E glycoprotein is a major structural protein which exhibits H A activity and induces HAinhibiting, neutralizing and protective antibodies (for review, see Gould et al., 1990). It is believed to participate in many important viral functions, such as receptor binding and the fusion of virus membrane with cellular membranes. Antibody-resistant escape mutants in the E protein of both TBE and Japanese encephalitis (JE) virus have previously been described and shown to exhibit reduced neurovirulence for mice (Holzmann et al., 1990; Cecilia & Gould, 1991). It is worth noting that in both previous reports, reduced virulence was not seen with all
934
Short communication
escape mutants. Thus, most mutations leading to attenuation (Hahn et al., 1987, 1988; Lobigs et al., 1990) were found in the B domain (as defined by Heinz et al., 1990) of the E protein. All of the LI virus mutants induced by neutralizing MAb 4.2 had substitutions within a tetrapeptide in the B domain, but they were at different positions from those seen and reported for TBE virus (Holzmann et al., 1990) and JE virus (Cecilia & Gould, 1991). Moreover, neurovirulence was reduced when the LI virus amino acid substitution occurred either at position 308 or 310 but not when the substitution occurred at position 311. Thus, the expression of virulence in different flaviviruses probably reflects changes in the E protein at different and possibly multiple sites. It was noticeable that none of the mutants isolated by plaque purification had a substitution at amino acid position 309. This may reflect its importance in maintaining the structural integrity of the E protein. This is supported by data demonstrating that amino acid 309 is strictly conserved in all of the tick-borne ftavivirus E proteins for which sequence data are available (Venugopal et al., 1992; Gritsun et al., 1993 and unpublished). Further evidence will become available when infectious clones are applied to study the neurovirulence of the virus. Three mutants with amino acid substitutions at position 311 were similar to the parent virus in their HA activity, their ability to induce protective antibody responses, their plaque size and their neurovirulence for mice. Four other mutants demonstrated reduced neurovirulence in mice compared with the parent virus. Two of these mutants, with amino acid substitutions at position 310, showed intermediate HA activities and induced protective antibody responses, whereas the two mutants that had substitutions at position 308 showed low virulence, small plaque size, poor HA activity and poor capacity to induce protective antibody responses. These results seem to confirm that in the parent virus at least part of the epitope identified by MAb 4.2 is situated in an exposed position on the E protein, and in the most attenuated escape mutants the changes in protein folding could affect the ability of the mutants to attach to receptors. The mouse protection experiments raise an important issue. The low virulence mutants did not induce protective immune responses in mice. Thus, not all attenuated viruses are suitable candidates for use as vaccines. It has not been possible to test directly if the change to a potential glycosylation site (NKS; aspartic acid to asparagine) by the amino acid substitution at position 308 is responsible for the attenuation of this mutant. However, in SDS-polyacrylamide gel analysis and also in tunicamycin experiments we failed to demonstrate differences between parent and mutant viruses in the E
protein (data not presented). Thus, it seems unlikely that the mutant protein has additional glycosylation. MAb 4.2 demonstrated potent neutralizing and protective ability when tested against LI virus, and we have also tested this antibody against TBE virus. MAb 4.2 protected the mice against virus challenge and was positive in neutralization tests. Moreover, a MAb 4.2resistant escape mutant of TBE virus was obtained and a single nucleotide substitution, which changed the amino acid codon at position 311 (lysine to threonine), was observed (unpublished results). These data imply that TBE virus has the same epitope in the same position as that identified above for LI virus. Therefore, the epitope represented by this antibody is probably also important in these viruses and it remains to be seen whether or not it has a significant role in determining their pathogenic or ecological characteristics. It is recognized that mutations in viral envelope proteins do exert profound changes in virus virulence (Dietzschold et al., 1983; Russell et al., 1989). Moreover, it seems unlikely that mutations other than those in the E gene of LI virus would have had any effect on the virulence of mutants 522 and 525, since only a very low proportion of the TBE and JE virus escape mutants reported previously (Holzmann et al., 1990; Cecilia & Gould, 1991) showed altered virulence characteristics. Furthermore, mutants 520, 521 and 526, which had different base changes but the same amino acid codon change, showed similar changes in biological characteristics implying that significant functional mutations outside the E gene had not occurred in these mutants. On the basis of the mouse virulence, mouse protection, virus plaque formation characteristics and HA studies, the epitope represented by the amino acids at position 308 to 311 appears to play an important role in determining the pathogenicity of LI virus. Whether or not it represents a functional component of the viral receptor remains to be determined. Moreover, formal proof that a single nucleotide substitution in the E protein is sufficient to modify the neurovirulence of a flavivirus awaits the use of infectious clones. The authors thank Dr S. Shiu, Dr K. Venugopal and Dr T. Gritsun for helpful advice and discussions in carrying out this work.
References ARBIZA,J., TAYLOR,G., LdPEz, J. A., FURZE,J., WYLD, S., WHYTE,P., STOTT, E. J., WERTZ, G., SULLENDER,W., TRUDEL, M. & MELERO, J. A. (1992). Characterization of two antigenic sites recognized by neutralizing monoclonal antibodies directed against the fusion glycoprotein of human respiratory syncytial virus. Journal of General Virology 73, 2225~234. CALISHER, C. H., KARABATSOS,N., DALRYMPLE,J. M., SHOPE, R. E., PORTERFIELD, J.S., WESTAWAY, E.G. & BRANDT, W.E. (1989). Antigenic relationships between flaviviruses as determined by crossneutralization tests with polyclonal antisera. Journal of General Virology 70, 37~43.
Short communication
CECILIA, D. & GOULD, E. A. (1991). Nucleotide changes responsible for loss of neuroinvasiveness in Japanese encephalitis virus neutralization-resistant mutants. Virology 181, 70-77. CHOMCZYNSKI,P. & SACCHI, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Analytical Biochemistry 162, 156-159. CLARKE,D. H. & CASALS,J. (1958). Techniques for hemagglutination and hemagglutination-inhibition with arthropod-borne viruses. American Journal of Tropical Medicine and Hygiene 7, 561-573. DE MADRID,A. Z. & PORTEREIELD,J. S. (1974). The flaviviruses (group B arboviruses): a cross-neutralization study. Journal of General Virology 23, 91-96. DIETZSCHOLD, B., WUNNER, W.H., WIKTOR, T.J., LOPES, A.D., LAFON, M., SMITH, C.L. & KOPROWSKI, H.L. (1983). Characterization of an antigenic determinant of the glycoprotein that correlates with pathogenicity of rabies virus. Proceedings of the National Academy of Sciences, U.S.A. 80, 70-74. GOULD, E. m., BUCKLEY,m., CAMMACK,N., BARRETT,A. D. T., CLEGG, J. C. S., ISHAK, R. & VARMA,M. G. R. (1985). Examination of the immunological relationships between flaviviruses using yellow fever virus monoclonal antibodies. Journal of General Virology 66, 1369-1382. GOULD, E.A., BUCKLEY,A., HIGGS, S. & GAIDAMOVICH,S. (1990). Antigenicity of flaviviruses. Archives of Virology Supplementum 1, 137-152. GRITSUN,T. S., LASHKEVICH,V. A. & GOULD,E. A. (1993). Nucleotide and deduced amino acid sequence of the envelope glycoprotein of Omsk haemorrhagic fever virus; comparison with other flaviviruses. Journal of General Virology 74, 287-291. HAHN, C. S., DALRYMPLE,J. M., STRAUSS,J. H. & RICE, C. M. (1987). Comparison of the virulent Asibi strain of yellow fever virus with the 17D vaccine strain derived from it. Proceedings of the National Academy of Sciences, U.S.A. 84, 20195023. HAHN, C.S., GALLER, R., HUNKAPILLER, T., DALRYMPLE, J.M., STRAUSS, J.H. & STRAUSS,E.G. (1988). Nucleotide sequence of dengue 2 RNA and comparison of the encoded proteins with those of the other flaviviruses. Virology 162, 167 180. HEINZ, F.X. (1986). Epitope mapping of flavivirus glycoproteins. Advances in Virus Research 31, 103 167.
935
HEINZ, F. X., MANDL, C. W., GUIRAKHOO,F., HOLZMANN,H., TUMA, W. & KUNZ, C. (1990). The envelope protein E of tick-borne encephalitis virus and other flaviviruses: structure, function and evolutionary relationships. Archives of Virology Supplementum 1, 125-135. HOLZMANN, H., HEINZ, F.X., MANDL, C.W., GUIRAKHOO, F. & KUNZ, C. (1990). A single amino acid substitution in envelope protein E of tick-borne encephalitis virus leads to attenuation in the mouse model. Journal of Virology 64, 5156-5159. LOBIGS, M., USHA, R., NESTOROWICZ, A., MARSHALL, I. D., WEIR, R.C. & DALGARNO,L. (1990). Host cell selection of the Murray Valley encephalitis virus variants altered at an RGD sequence in the envelope protein and in mouse virulence. Virology 176, 58%595. PORTERFIELD,J. S. (1980). Antigenic characteristics and classification of Togaviridae. In The Togaviruses, pp. 1346. Edited by R.W. Schlesinger. New York & London: Academic Press. REID, H. W. (1987). Louping ill. In The Arboviruses: Epidemiology and Ecology, vol. 3, pp. 117 135. Edited by T. P. Monath. Boca Raton: CRC Press. RUSSELL,D. L., DALRYMPLE,J. M. & JOHNSTON,R. E. (1989). Sindbis virus mutations which coordinately affect glycoprotein processing, penetration and virulence in mice. Journal of Virology 63, 1619-1629. 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, 5463-5467. SHIU, S.Y.W., AYRES, M.D. & GOULD, E.A. (1991). Genomic sequence of the structural proteins of louping ill virus: comparative analysis with tick-borne encephalitis virus. Virology 180, 411415. VENUGOPAL,K., BUCKLEY,A., REID, H. W. & GOULD, E. A. (1992). Nucleotide sequence of the envelope glycoprotein of Negishi virus shows very close homology to louping ill virus. Virology 190, 515 521. WESTAWAY,E. G., BRINTON,M. A., GAIDAMOVICH,S. YA., HORZINEK, M.C., IGARASHI,A., K~_)~RI~,NEN,L., LVOV, D.K., PORTERFIELD, J.S., RUSSELL,P.K. & TRENT, D.W. (1985). Flaviviridae. Intervirology 24, 183 192.
(Received 1 October 1992; Accepted 24 December 1992)
931
Single amino acid codon changes detected in louping ill virus antibodyresistant mutants with reduced neurovirulence W . R. J i a n g , 1 A. L o w e , ~ S. H i g g s , ~ H . Reid 2 and E. A. G o u l d ~* ~NERC Institute of Virology and Environmental Microbiology, Mansfield Road, Oxford OX1 3SR and 2Moredun Research Institute, 408 Gilmerton Road, Edinburgh EH17 7JH, U.K.
Seven mutant viruses were derived from a Scottish strain of louping ill virus using a virus envelope-specific neutralizing monoclonal antibody. None of the mutants was neutralized and immunofluorescence microscopy confirmed that they did not bind to this antibody. Four mutants showed reduced mouse neurovirulence compared with parent virus and two mutants failed to induce protective immune responses in mice challenged with virulent tick-borne encephalitis virus. The mutants with the lowest virulence showed poor or undetectable haemagglutinating activity. The nucleotide sequence of the envelope glycoprotein gene of each of the seven
mutants was determined and the deduced amino acid sequence was compared with parent virus. For each mutant, only a single amino acid codon change was detected and all the amino acid substitutions occurred within amino acid positions 308 to 311. A change from the amino acid aspartate to asparagine at amino acid position 308, which represented a potential glycosylation site, was the most effective substitution in reducing mouse neurovirulence. The results demonstrate the importance of critical sites within the envelope glycoprotein as determinants of virus virulence.
Louping ill (LI) virus, a tick-transmitted member of the virus family Flaviviridae (Westaway et al., 1985), causes an encephalitic or encephalomyelitic disease of sheep and grouse. The virus is prevalent in the upland sheep grazing areas of Scotland, northern England, Wales and southwest England. Other wildlife species can also become infected with LI virus if they serve as hosts for the feeding tick vector Ixodes ricinus, but it is believed that these vertebrates are dead-end hosts because they do not develop high viraemic titres (for review, see Reid, 1987). LI virus is genetically (Shiu et al., 1991) and antigenically (De Madrid & Porterfield, 1974; Porterfield, 1980; Heinz, 1986; Calisher et al., 1989) closely related to other European tick-borne encephalitis (TBE) viruses, many strains of which produce severe and often fatal encephalitis in humans. The factors that determine why some strains of TBE virus are virulent have not been determined yet but studies with antibody-resistant mutants have shown that the viral envelope (E) protein contributes significantly to flavivirus virulence (Holzmann et al., 1990; Cecilia & Gould, 1991). It is believed that the global climate is changing and that in the future the United Kingdom is likely to experience an increase in temperature. This will inevitably lead to altered land usage and the likelihood of redistributed arthropod populations, including ticks. In addition, there is an increasing possibility that virulent strains of TBE could be introduced into the U.K. from
mainland Europe. Thus, there is a need to understand the molecular basis of flavivirus virulence so that appropriate action can be taken to control these viruses in the environment. In this paper we report the preparation and characterization of seven mutants of LI virus which were derived using a potent neutralizing monoclonal antibody (MAb). Nucleotide sequencing of the viral E gene of each mutant has revealed an important amino acid that appears to be closely associated with LI virus virulence. Continuous lines of porcine kidney (PS) cells were grown in Eagle's MEM with 5 % fetal calf serum. LI virus (strain 369/T2) was described previously (Shiu et al., 1991). It was plaque-purified in PS cells and then passaged once through suckling mouse brains before selection of antibody-resistant mutants, as described below. TBE virus was a central European strain kindly supplied by Dr J. S. Porterfield. Its antigenic characteristics were reported previously (Gould et al., 1985). MAbs were selected for their capacity to bind the E protein of LI virus. They have all been described previously (Gould et al., 1985; Venugopal et al., 1992). MAbs 4.2 and 7.1 neutralize the infectivity of LI virus in plaque reduction neutralization tests and MAb 7.1, but not MAb 4.2, also shows haemagglutination (HA)inhibiting activity with LI virus. Escape mutants were prepared using plaque-purified LI virus, which was diluted to 100 p.f.u./ml and mixed with the selecting
0001-1384 © 1993 SGM
932
Short communication
Table 1. Biological and molecular characteristics o f L I virus escape mutants Virus
HA (units)
Infectivity (log10 p.f.u./ml)
522 525 523 529 520 521 526 Parent
< 2 < 2 < 2 8 64 16 64 32
8'98 8.40 8'60 8.30 7.54 8.48 9'00 8.18
LDs0 (loga0 p.f.u. to kill 50% mice) > 6.5 5.8 > 6.1 > 5.5 2.3 3'0 3.6 2.2
Plaque size (mm) 1-1.5 1-1.5 3-5 3 5 3-5 3-5 3-5 3 5
Nucleotide change* G to A G to A U to C U to C A to C G to U G to C None
(922) (922) (928) (928) (931) (933) (933)
Amino acid codon change* D to N (308) D to N (308) S to P (310) S to P (310) K to Q (311) K to N (311) K to N (311) None
* N u m b e r s in parentheses indicate position of mutation in the E gene.
MAb 4.2 (diluted 1 : 100; neutralization titre 1 : 10000) in Eagle's MEM. The mixture was incubated at 37 °C for 1 h and was then adsorbed onto PS cell monolayers, again for 1 h at 37 °C. The cell monolayers were washed and medium containing the same concentration of MAb 4.2 and 0.8 % agarose was added. Non-neutralized virus was recovered and plaque-purified twice more under the same experimental conditions. The mutant viruses were tested again for their binding capacity with MAb 4.2. They were all negative. For H A tests, the parent L1 virus and each escape mutant was grown in PS cells. Supernatant medium from the infected cells was harvested at day 5 post-infection and was concentrated using 7 % polyethylene glycol and 0.4 M-NaC1. The concentrated virus was precipitated by centrifugation and was resuspended in borate-buffered saline at pH 9.0 (Clarke & Casals, 1958) at a concentration of 50-fold. The HA test was carried out at pH 6.2 (Clarke & Casals, 1958). For neurovirulence tests, groups of 10 3-week-old female mice (strain TO, Tuck and Son) were inoculated either intracerebrally (i.c.) or subcutaneously (s.c.) with serial 10-fold dilutions of parent or mutant virus. Sickness and mortality were recorded for 18 days. The mice that survived s.c. inoculation of parent or mutant virus were challenged s.c. with TBE virus at 100 LDs0/mouse. For viral RNA preparation and nucleotide (nt) sequence analysis, suckling mouse brain preparations of plaque-purified MAb-resistant mutants were produced using newborn mice. The viral R N A was extracted directly from the infected suckling mouse brain suspensions using guanidinium thiocyanate (Chomczynski & Sacchi, 1987) and first strand c D N A was synthesized using either the downstream primer EXP2 (TGCag a t c t C T A C T A C G C T C C C A C C C C G A G A G T , nt positions 2341 to 2358) or LIE4 (CCCggtaccAAGCTGC A T C T C T A T G A , nt positions 1982 to 1998) and then amplified by PCR using EXP1 (CTTagatctaatATGG T C G C C G T T G T G T G G C T A , nt positions 781 to 801) and EXP2, or LIE3 ( T T T g g t a c c C C T C A T G C T G T C A A G A T G , nt positions 1609 to 1629; Shiu et al., 1991)
and LIE4. The PCR-amplified D N A was cloned using p U C l l 8 . At least two subclones of each virus were sequenced in both directions by double-stranded dideoxynucleotide sequencing (Sanger et al., 1977). Seven plaque-purified neutralization-resistant mutants of LI virus strain 369/T2 (Shiu et al., 1991) were derived using MAb 4.2 as described above. Each mutant was then grown in PS cells both in the presence and absence of MAb 4.2, the selecting MAb. There were no major differences in the infectivity titres of the mutant viruses when grown under these conditions. PS cell monolayers, infected with the mutant viruses, were analysed by indirect immunofluorescence (IIF) microscopy for reactivity with MAbs that were known to react with standard LI virus (Venugopal et al., 1992). Each mutant reacted with all the MAbs except MAb 4.2 and MAb 7.1 suggesting that the epitopes represented by these two MAbs are very closely associated. It is important to note that MAb 7.1, but not MAb 4.2, inhibits HA by LI virus and these two MAbs are therefore considered to represent different but closely associated epitopes on the envelope glycoprotein of LI virus (Venugopal et al., 1992). Cultures of the parent 369/T2 strain of LI virus and each of the seven escape mutants were concentrated from the supernatant culture medium of infected PS cell monolayers as described above. Each preparation of virus was then tested for H A activity. The parent virus and four escape mutants (520, 526, 521 and 529) were positive in H A tests but the other three escape mutants (522, 525 and 523) were negative (Table 1). The size of plaque produced by each mutant when grown in PS cells under carboxymethylcellulose overlay was compared with the parent virus, which produced plaques of 3 to 5 mm in 4 days at 37 °C. Two mutants (522 and 525) produced small plaques (1 to 1.5 ram), whereas the other mutants were similar to the parent virus. The neurovirulence of parent 369/T2 LI virus and each of the seven escape mutants was analysed in newborn and 3-week-old mice by i.c. challenge and also in 3-week-old mice by s.c. inoculation of serial 10-fold
Short communication
Table 2. Detection o f antibody response in mice following inoculation with escape mutants
Challenge virus
Proportion of mice positive by IIF
522 525 523 529 520 521 526 Parent
1/5 2/5 3/4 7/7 5/5 4/5 5/5 5/5
Neutralization titre of serum < 10 40 < 10 < 10 40 160 10 40
Proportion of mice surviving TBE virus challenge* 2/49 3/25 17/20 18/19 13/14 13/13 13/15 13/13
*Mice givena sub-lethaldose of LI virus were subsequentlychallenged with a lethal s.c. dose of TBE virus. dilutions of each virus. Four of the escape mutants (522, 523, 525 and 529) had reduced neurovirulence compared with parent virus, whereas three mutants (520, 521 and 526) had retained the virulence of the parent virus. All viruses were virulent for newborn mice and 3-week-old mice if the virus was given i.c., demonstrating that even the low virulence mutants had retained the capacity to infect and multiply in susceptible mouse brain cells. The neurovirulence titrations for subcutaneously infected 3-week-old mice are shown in Table 1. Groups of 3-week-old mice were inoculated s.c. with an estimated 100 p.f.u, of each virulent escape mutant (520, 521 and 526) or each avirulent mutant (522, 523, 525 and 529). Three weeks later, serum was collected from five surviving mice of each group. These sera were tested for the presence of antibody against LI virus using an IIF test on cells infected with LI virus. The results (Table 2) show that mice inoculated with sub-lethal doses of virulent LI virus escape mutants contained detectable antibody. On the other hand, two of the escape mutants with low neurovirulence for mice (522 and 525) induced detectable serum antibody responses in only a low proportion of the mice. The surviving mice were then challenged s.c. with 100 LDs0 of TBE virus. All mice initially given sub-lethal doses of the virulent escape mutants and which possessed detectable antibody survived subsequent challenge with TBE virus, whereas all mice initially given the avirulent mutants, and which were antibody-negative, were fatally infected by the TBE virus (Table 2). These results imply that two of the avirulent mutants (522 and 525) did not multiply significantly in the mice. The complete nucleotide sequence of the gene encoding the viral E glycoprotein of two escape mutants (520 and 522) was determined. The results obtained were aligned and compared with the published nucleotide sequence for LI virus strain 369/T2. Single nucleotide changes were detected corresponding to amino acid positions
933
308 and 311 for mutants 522 and 520 respectively. The remaining escape mutants were therefore sequenced over a region of the viral E protein spanning the area in which the nucleotide substitutions had been identified (nt 756 to 1110). The nucleotide changes and the corresponding amino acid codon changes, only one in each mutant, are shown in Table 1. The two least virulent mutants (522 and 525) had amino acid substitutions at position 308. The two mutants (523 and 529) that showed partially reduced virulence had amino acid substitutions in a nearby amino acid (amino acid 310). Mutants 520, 521 and 526, which showed either small or no significant reductions in virulence, had amino acid substitutions at position 311. The change of aspartate to asparagine at amino acid position 308 produced a potential glycosylation site. We have produced and analysed the nucleotide sequences of the E protein of seven neutralization escape mutants derived using a single neutralizing MAb. Although they were isolated in separate experiments, two pairs of mutants showed identical base substitutions and could therefore represent the progeny of single clones. For each mutant analysed, only a single nucleotide change was detected in the E gene and in each case this correlated with an amino acid codon change. All codon changes occurred within a tetrapeptide (amino acid sequence 308 to 311) suggesting that the epitope identified by MAb 4.2 could be linear, and data obtained by immunoblotting MAb 4.2 with LI virus under reducing conditions support this possibility (data not presented). MAb 7.1 also lost the capacity to bind to the escape mutants implying that it identifies the same epitope as MAb 4.2. This seems unlikely since MAb 7.1, in contrast with MAb 4.2, inhibits H A activity. A similar situation was recently reported by Arbiza et al. (1992) with human respiratory syncytial virus. Their data were consistent with the concept that overlapping MAbs might recognize sites with specific local conformational differences. Therefore, it is possible that at least a part of the MAb 4.2 epitope is present in the MAb 7.1 epitope and this region of the epitope could have local conformational rearrangements. The flavivirus E glycoprotein is a major structural protein which exhibits H A activity and induces HAinhibiting, neutralizing and protective antibodies (for review, see Gould et al., 1990). It is believed to participate in many important viral functions, such as receptor binding and the fusion of virus membrane with cellular membranes. Antibody-resistant escape mutants in the E protein of both TBE and Japanese encephalitis (JE) virus have previously been described and shown to exhibit reduced neurovirulence for mice (Holzmann et al., 1990; Cecilia & Gould, 1991). It is worth noting that in both previous reports, reduced virulence was not seen with all
934
Short communication
escape mutants. Thus, most mutations leading to attenuation (Hahn et al., 1987, 1988; Lobigs et al., 1990) were found in the B domain (as defined by Heinz et al., 1990) of the E protein. All of the LI virus mutants induced by neutralizing MAb 4.2 had substitutions within a tetrapeptide in the B domain, but they were at different positions from those seen and reported for TBE virus (Holzmann et al., 1990) and JE virus (Cecilia & Gould, 1991). Moreover, neurovirulence was reduced when the LI virus amino acid substitution occurred either at position 308 or 310 but not when the substitution occurred at position 311. Thus, the expression of virulence in different flaviviruses probably reflects changes in the E protein at different and possibly multiple sites. It was noticeable that none of the mutants isolated by plaque purification had a substitution at amino acid position 309. This may reflect its importance in maintaining the structural integrity of the E protein. This is supported by data demonstrating that amino acid 309 is strictly conserved in all of the tick-borne ftavivirus E proteins for which sequence data are available (Venugopal et al., 1992; Gritsun et al., 1993 and unpublished). Further evidence will become available when infectious clones are applied to study the neurovirulence of the virus. Three mutants with amino acid substitutions at position 311 were similar to the parent virus in their HA activity, their ability to induce protective antibody responses, their plaque size and their neurovirulence for mice. Four other mutants demonstrated reduced neurovirulence in mice compared with the parent virus. Two of these mutants, with amino acid substitutions at position 310, showed intermediate HA activities and induced protective antibody responses, whereas the two mutants that had substitutions at position 308 showed low virulence, small plaque size, poor HA activity and poor capacity to induce protective antibody responses. These results seem to confirm that in the parent virus at least part of the epitope identified by MAb 4.2 is situated in an exposed position on the E protein, and in the most attenuated escape mutants the changes in protein folding could affect the ability of the mutants to attach to receptors. The mouse protection experiments raise an important issue. The low virulence mutants did not induce protective immune responses in mice. Thus, not all attenuated viruses are suitable candidates for use as vaccines. It has not been possible to test directly if the change to a potential glycosylation site (NKS; aspartic acid to asparagine) by the amino acid substitution at position 308 is responsible for the attenuation of this mutant. However, in SDS-polyacrylamide gel analysis and also in tunicamycin experiments we failed to demonstrate differences between parent and mutant viruses in the E
protein (data not presented). Thus, it seems unlikely that the mutant protein has additional glycosylation. MAb 4.2 demonstrated potent neutralizing and protective ability when tested against LI virus, and we have also tested this antibody against TBE virus. MAb 4.2 protected the mice against virus challenge and was positive in neutralization tests. Moreover, a MAb 4.2resistant escape mutant of TBE virus was obtained and a single nucleotide substitution, which changed the amino acid codon at position 311 (lysine to threonine), was observed (unpublished results). These data imply that TBE virus has the same epitope in the same position as that identified above for LI virus. Therefore, the epitope represented by this antibody is probably also important in these viruses and it remains to be seen whether or not it has a significant role in determining their pathogenic or ecological characteristics. It is recognized that mutations in viral envelope proteins do exert profound changes in virus virulence (Dietzschold et al., 1983; Russell et al., 1989). Moreover, it seems unlikely that mutations other than those in the E gene of LI virus would have had any effect on the virulence of mutants 522 and 525, since only a very low proportion of the TBE and JE virus escape mutants reported previously (Holzmann et al., 1990; Cecilia & Gould, 1991) showed altered virulence characteristics. Furthermore, mutants 520, 521 and 526, which had different base changes but the same amino acid codon change, showed similar changes in biological characteristics implying that significant functional mutations outside the E gene had not occurred in these mutants. On the basis of the mouse virulence, mouse protection, virus plaque formation characteristics and HA studies, the epitope represented by the amino acids at position 308 to 311 appears to play an important role in determining the pathogenicity of LI virus. Whether or not it represents a functional component of the viral receptor remains to be determined. Moreover, formal proof that a single nucleotide substitution in the E protein is sufficient to modify the neurovirulence of a flavivirus awaits the use of infectious clones. The authors thank Dr S. Shiu, Dr K. Venugopal and Dr T. Gritsun for helpful advice and discussions in carrying out this work.
References ARBIZA,J., TAYLOR,G., LdPEz, J. A., FURZE,J., WYLD, S., WHYTE,P., STOTT, E. J., WERTZ, G., SULLENDER,W., TRUDEL, M. & MELERO, J. A. (1992). Characterization of two antigenic sites recognized by neutralizing monoclonal antibodies directed against the fusion glycoprotein of human respiratory syncytial virus. Journal of General Virology 73, 2225~234. CALISHER, C. H., KARABATSOS,N., DALRYMPLE,J. M., SHOPE, R. E., PORTERFIELD, J.S., WESTAWAY, E.G. & BRANDT, W.E. (1989). Antigenic relationships between flaviviruses as determined by crossneutralization tests with polyclonal antisera. Journal of General Virology 70, 37~43.
Short communication
CECILIA, D. & GOULD, E. A. (1991). Nucleotide changes responsible for loss of neuroinvasiveness in Japanese encephalitis virus neutralization-resistant mutants. Virology 181, 70-77. CHOMCZYNSKI,P. & SACCHI, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Analytical Biochemistry 162, 156-159. CLARKE,D. H. & CASALS,J. (1958). Techniques for hemagglutination and hemagglutination-inhibition with arthropod-borne viruses. American Journal of Tropical Medicine and Hygiene 7, 561-573. DE MADRID,A. Z. & PORTEREIELD,J. S. (1974). The flaviviruses (group B arboviruses): a cross-neutralization study. Journal of General Virology 23, 91-96. DIETZSCHOLD, B., WUNNER, W.H., WIKTOR, T.J., LOPES, A.D., LAFON, M., SMITH, C.L. & KOPROWSKI, H.L. (1983). Characterization of an antigenic determinant of the glycoprotein that correlates with pathogenicity of rabies virus. Proceedings of the National Academy of Sciences, U.S.A. 80, 70-74. GOULD, E. m., BUCKLEY,m., CAMMACK,N., BARRETT,A. D. T., CLEGG, J. C. S., ISHAK, R. & VARMA,M. G. R. (1985). Examination of the immunological relationships between flaviviruses using yellow fever virus monoclonal antibodies. Journal of General Virology 66, 1369-1382. GOULD, E.A., BUCKLEY,A., HIGGS, S. & GAIDAMOVICH,S. (1990). Antigenicity of flaviviruses. Archives of Virology Supplementum 1, 137-152. GRITSUN,T. S., LASHKEVICH,V. A. & GOULD,E. A. (1993). Nucleotide and deduced amino acid sequence of the envelope glycoprotein of Omsk haemorrhagic fever virus; comparison with other flaviviruses. Journal of General Virology 74, 287-291. HAHN, C. S., DALRYMPLE,J. M., STRAUSS,J. H. & RICE, C. M. (1987). Comparison of the virulent Asibi strain of yellow fever virus with the 17D vaccine strain derived from it. Proceedings of the National Academy of Sciences, U.S.A. 84, 20195023. HAHN, C.S., GALLER, R., HUNKAPILLER, T., DALRYMPLE, J.M., STRAUSS, J.H. & STRAUSS,E.G. (1988). Nucleotide sequence of dengue 2 RNA and comparison of the encoded proteins with those of the other flaviviruses. Virology 162, 167 180. HEINZ, F.X. (1986). Epitope mapping of flavivirus glycoproteins. Advances in Virus Research 31, 103 167.
935
HEINZ, F. X., MANDL, C. W., GUIRAKHOO,F., HOLZMANN,H., TUMA, W. & KUNZ, C. (1990). The envelope protein E of tick-borne encephalitis virus and other flaviviruses: structure, function and evolutionary relationships. Archives of Virology Supplementum 1, 125-135. HOLZMANN, H., HEINZ, F.X., MANDL, C.W., GUIRAKHOO, F. & KUNZ, C. (1990). A single amino acid substitution in envelope protein E of tick-borne encephalitis virus leads to attenuation in the mouse model. Journal of Virology 64, 5156-5159. LOBIGS, M., USHA, R., NESTOROWICZ, A., MARSHALL, I. D., WEIR, R.C. & DALGARNO,L. (1990). Host cell selection of the Murray Valley encephalitis virus variants altered at an RGD sequence in the envelope protein and in mouse virulence. Virology 176, 58%595. PORTERFIELD,J. S. (1980). Antigenic characteristics and classification of Togaviridae. In The Togaviruses, pp. 1346. Edited by R.W. Schlesinger. New York & London: Academic Press. REID, H. W. (1987). Louping ill. In The Arboviruses: Epidemiology and Ecology, vol. 3, pp. 117 135. Edited by T. P. Monath. Boca Raton: CRC Press. RUSSELL,D. L., DALRYMPLE,J. M. & JOHNSTON,R. E. (1989). Sindbis virus mutations which coordinately affect glycoprotein processing, penetration and virulence in mice. Journal of Virology 63, 1619-1629. 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, 5463-5467. SHIU, S.Y.W., AYRES, M.D. & GOULD, E.A. (1991). Genomic sequence of the structural proteins of louping ill virus: comparative analysis with tick-borne encephalitis virus. Virology 180, 411415. VENUGOPAL,K., BUCKLEY,A., REID, H. W. & GOULD, E. A. (1992). Nucleotide sequence of the envelope glycoprotein of Negishi virus shows very close homology to louping ill virus. Virology 190, 515 521. WESTAWAY,E. G., BRINTON,M. A., GAIDAMOVICH,S. YA., HORZINEK, M.C., IGARASHI,A., K~_)~RI~,NEN,L., LVOV, D.K., PORTERFIELD, J.S., RUSSELL,P.K. & TRENT, D.W. (1985). Flaviviridae. Intervirology 24, 183 192.
(Received 1 October 1992; Accepted 24 December 1992)
