- Email: [email protected]
Analysis of the haemagglutinin gene of current wild-type canine distemper virus isolates from Germany
Virus Research 48 (1997) 165 – 171
Analysis of the haemagglutinin gene of current wild-type canine distemper virus isolates from Germany L. Haas a,*, W. Martens a, I. Greiser-Wilke a, L. Mamaev b, T. Butina b, D. Maack a, T. Barrett c a
Veterinary School Hanno6er, Institute of Virology, Bu¨nteweg 17, D-30559 Hanno6er, Germany b Limnological Institute, Ulan-Bayorskaya-3, Irkutsk, Russia c Institute for Animal Health, Pirbright Laboratory, Woking, Surrey GU 24 ONF, UK Received 10 September 1996; accepted 10 January 1997
Abstract The haemagglutinin (H) gene sequences from three wild-type canine distemper viruses (CDV) isolated during 1994–1995 were sequenced to determine whether contemporary strains had undergone significant genetic changes relative to the currently used vaccine strains. The new isolates were closely related to each other (\99%) and displayed about 90–91% sequence homology to the Onderstepoort and Convac vaccine strains. There were one to four additional potential glycosylation sites compared to the vaccine strains which were also present in a German dog CDV isolate dating from 1990. However, only a very slight reduction in neutralizing titre against the new isolates was found when compared with the Onderstepoort and Rockborn vaccine strains. Cysteine and proline residues were well conserved indicating a conserved three dimensional structure for the protein. By phylogenetic analysis the recent isolates showed a narrow clustering close to the previous canine isolates indicating a linear pattern of evolutionary changes. A comparison with published CDV H gene sequences suggested the presence of different lineages of CDV on a global scale and possible cocirculation of more than one genotype of CDV. © 1997 Elsevier Science B.V. Keywords: Canine distemper virus; Haemagglutinin gene; Neutralizing antibody titers
Canine distemper virus (CDV) is the the causative agent of a serious infection of dogs and many other carnivores (Appel, 1987). It is classified in the Morbilli6irus genus of the family * Corresponding author.
Paramyxoviridae. In developed countries with a high rate of vaccination against CDV clinical cases occur in an epidemic manner. Different suggestions are put forward to explain these outbreaks. It may be that after several years of absence of disease people become negligent about
0168-1702/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 8 - 1 7 0 2 ( 9 7 ) 0 1 4 4 9 - 4
166
L. Haas et al. / Virus Research 48 (1997) 165–171
vaccination. There may also be problems with the vaccination schedule, e.g. vaccinating puppies still possessing maternal antibodies which interfere with the generation of immunity. On the other hand it is occasionally assumed that the vaccines in use would not protect properly against current field virus infections even if applied correctly. Since the most widely used vaccines are modified live CDV vaccines which became available in the early 1960s (Chappuis, 1995), it is sometimes suggested that field viruses may have structurally and antigenically changed since that time and that these vaccines would thus no longer confer efficacious immunity. Immunological and sequence data have clearly shown that the major envelope glycoprotein (haemagglutinin or H protein) is the most variable of the morbillivirus proteins and therefore is the most suitable candidate to look for genetic changes in the virus (O8 rvell et al., 1990). During winter 1994/95 there were reports of CDV infections in some northern and eastern German cities. We isolated CDV from several cases in Hamburg and Berlin, amongst them isolates from young animals which according to the anamnesis showed clinical signs in spite of a previous vaccination record. The viruses were obtained from the buffy coat fraction of heparinized blood samples. White blood cells were co-cultivated on Vero, A72 and MDCK cell lines for up to three passages. RNA was prepared from infected cells when a cpe was visible using a RNA isolation kit (RNeasy®, Quiagen) following the instructions of the manufacturer. The H gene RNA of the three field isolates (animal no. 404, Berlin; nos. 2544 and 4513, Hamburg) was amplified by reverse transcriptase-polymerase chain reaction (RT-PCR) as decribed by Curran et al. (1991). We amplified the entire H gene (:1.9 kb) sequence using this method. As a comparison we also amplified the H gene from an older virulent CDV isolate (Snyder Hill). Amplicons were cloned into a specialised PCR cloning vector (pCRII®, Invitrogen). Dideoxy sequencing of double stranded plasmid DNA was carried out either using conventional radioactive sequencing techniques or an automatic ALF sequencer (Pharmacia) with the AutoRead T7 sequencing kit
(Pharmacia). Labelled M13 universal forward and reverse primers as well as sequence-specific internal H gene primers were used in the reactions. Sequences were analysed and aligned using the Seqaid II (version 3.81) and the Align Plus (Scientific and Educational Software, State Line, PA) software packages. For phylogenetic analysis the PHYLIP programs were used (Felsenstein, 1993). A single large open reading frame (ORF) coding for 607 amino acids was identified in each case and the deduced amino acid sequences aligned with other CDV H protein sequences obtained from the Genbank database (Fig. 1). Homologies to the published CDV vaccine strains Onderstepoort and Convac were between 90 and 91%. A high level of homology (\ 99%) was found between the new isolates. As expected, the three field viruses closely resembled an earlier German dog isolate dating from 1990 (5804/Han90). The extracellular glycosylation pattern of their H proteins showed some differences when compared with the vaccine strains. Four glycosylation sites (positions 19–21; 149–151; 422–424; 587–589) were shared by all CDVs but there were four additional potential glycosylation sites when compared with the Onderstepoort strain and one when compared with the Convac strain. The latter glycosylation site (position 309–311) was found in all recent virus field isolates but was absent in the older Snyder Hill strain. The potential glycosylation site at the extreme C-terminus (position 603–605) is unlikely to be used. It is not present in all field viruses and is missing from the Onderstepoort vaccine due to truncation of this protein at position 604. Amino acids such as proline and cysteine, which are known to be important in determining protein secondary structure, were highly conserved. Over 80% of the prolines and almost all 12 cysteine residues were completely conserved, the only exception being the Javelina CDV isolate where the cysteine residue at position 390 was replaced by a tryptophane. The H sequence results described here clearly show that the current isolates represent typical CDVs. The variation in sequence compared to the vaccine strains is most probably due to changes that have occurred in the field viruses over the past 60 or so years since the progenitors of the
L. Haas et al. / Virus Research 48 (1997) 165–171
vaccine strains were in circulation. Cell culture adaptation is unlikely to have radically changed the virus sequences since, in the case of the related rinderpest virus, it has been shown that the vaccine strain derived by multiple cell culture passages was more than 99% similar to the virulent
167
parental virus (Baron and Barrett, 1995). The H gene of current field isolates of rinderpest virus differ by about 10–12% in nucleotide sequence from the vaccine strain (Barrett, unpublished work) and similar levels of variation between recent field isolates and the vaccine strains have
Fig. 1. Alignment of the deduced amino acid sequences of recent German CDV dog isolates (H404, H2544, H4513) and the virulent Snyder Hill strain (Snyhill) with various CDV isolates available on the genbank database. Non-conservative substitutions are shown in white on a black background while conservative substitutions are shown in black on a grey background. Other virus sources: Ondvac, vaccine strain Onderstepoort (D00758); Convac, vaccine strain Convac (Z35493); H5804, German dog isolate (X85000); Seal, Siberian seal (X84998); Dkdog, Danish dog (GBVRL:CDVH3GEN); Gldog, Greenland dog (GBVRL:CDVH4GEN); Java, javelina (collared peccary) (GBVRL:CDVH11GEN); Usdog, American dog (GBVRL:CDVH9GEN); Bleo, American black leopard (GBVRL:CDVH10GEN); Racc, American raccoon (GBVRL:H13GEN), Dkmink, Danish mink (GBVRL:CDVH3GEN); Ferret, German ferret (X84999).
168
L. Haas et al. / Virus Research 48 (1997) 165–171
Fig. 1. (Contd.)
been reported for the H gene of measles virus (Tamin et al., 1994). However, structurally important amino acids were well conserved. Cysteine residues, which probably play a major role in the determining the antigenic structure and processing of H molecules (Hu and Norrby, 1994) were, with one exception, completely conserved. The role, if any, that the additional potential glycosylation site at position 309 – 311 found in the recent isolates plays in the immune response remains to be determined. An additional potential
glycosylation site was also observed in wild-type measles virus H genes compared with the vaccine strain, Moraten (Rota et al., 1992). A conspicuous narrow clustering of predicted amino acid changes around potential glycosylation sites was found in measles virus field isolates (Rota et al., 1992). While amino acid changes were scattered across the H protein in the case of CDV, slightly more amino acid changes were noted in the region around and between the two glycosylation sites located at positions 422–424 and 456–458, which
L. Haas et al. / Virus Research 48 (1997) 165–171
might indicate sites in the protein subject to a greater antigenic pressure and selective force. The receptor binding site for measles virus has been located around position 470 – 490 (Lecouturier et al., 1996) and this region of the protein may form part of the globular head of the mature protein. Some differences in cross-neutralization were detectable in the case of the earlier German canine field isolate (5804/Han90) using sera directed against the Rockborn vaccine strain (Harder et al., 1993) and similar neutralisation differences have been reported in the case of rinderpest virus (Rossiter, 1988). To test whether or not CDV positive sera from infected or vaccinated animals differ in their reactivity to current vaccine strains and an actual CDV isolate, respectively, neutralizing peroxidase-linked assays (NPLA) were performed as described by Zaghawa et al. (1990). The neutralizing antibody titres against isolate 2544 were simultaneously tested and compared to those of the Onderstepoort and Rockborn vaccine strains. Sera from naturally infected or vaccinated carnivores from our diagnostic unit were used and, as can be seen in Fig. 2, there was a slight but not significant reduction of neutralizing antibody titres measured against isolate 2544. It is therefore clear that the sequence changes identified do not have a significant effect on the ability of the vaccine to protect against the new strains. Phylogenetic analysis indicated a slowly progressive evolution in wild-type CDV based on changes in the H gene. The data were analysed by three different PHYLIP programes (NEIGHBOR, FITCH and KITCH) and a distance matrix derived using DNADIST. The same branching patterns were obtained with all three methods. As expected, the topology of the phylogenetic trees of the H gene sequences revealed a close clustering of the recent isolates and the previous German canine isolate and they clearly represent one lineage of the virus (Fig. 3). The German isolates were also closely related to CDV isolated from a Danish dog. Interestingly, a German ferret CDV and a Danish mink CDV formed a separate branch which might indicate a mustelid lineage of CDV. This is supported by similar phylogenetic relationships which were based on part of the F gene sequences of different mustelids (Harder,
169
Liermann and Haas, unpublished results). There was also a clustering of the American isolates, irrespective of the species from which the viruses originated. Interestingly, the more distantly related seal CDV isolate obtained from a Siberian seal (Phoca sibirica) in Lake Baikal in 1987 (Mamaev et al., 1995) was most related to CDV isolated from a Greenland dog in 1988 and might indicate a separate lineage circulating across the arctic ecosystem in susceptible species such as polar bear and arctic foxes. The Baikal seal virus is sometimes referred to as phocid distemper virus-2 (PDV-2), although it is now known to be CDV, to distinguish it from the first morbillivirus isolated from a seal (PDV-1) during the European seal morbillivirus epizootic and which is distinct from CDV (Visser et al., 1990; Haas et al., 1991).
Fig. 2. Neutralizing antibody titres (NPLA) measured using field sera with either CDV vaccine strain Rockborn (top) or Onderstepoort (bottom) compared with neutralising titres obtained with CDV field isolate 2554.
170
L. Haas et al. / Virus Research 48 (1997) 165–171
To summarize, although it is not possible to rule out that some of the immunization failures in properly vaccinated dogs may have been due to changes in the antigenic nature of current field strains, our data do not substantiate this. This is in agreement with previous antigenic studies performed on CDV field isolates in an urban dog population (Blixenkrone-Møller et al., 1993). Since CDV infections can realistically only be controlled by vaccination (Chappuis, 1995) care must be taken to apply a rigid vaccination schedule in combination with other prophylactic measures. The nucleotide sequences reported in this paper have been submitted to GenBank/EMBL. Accession numbers: virus 404 (Z77671); virus 2544 (Z77672); virus 4513 (Z77673) Fig. 3. Phylogenetic analysis of the nucleotide sequences of the coding regions of CDV H proteins. The tree was constructed using the PHYLIP DNADIST and KITSCH programes (Felsenstein, 1993). For comparison other CDV H gene sequences were extracted from the Genbank database (see legend Fig. 1).
Geographically distinct lineages have been described for other morbilliviruses such as rinderpest (Chamberlain et al., 1993; Wamwayi et al., 1995) and measles viruses (Taylor et al., 1991). In the case of measles virus, it was proposed that several lineages or genotypes co-circulate at a given time (Rima et al., 1995). Our phylogentic analysis shows an analogous situation for CDV since in Europe different lineages have been identified in canines and in mustelids. It is interesting to note that the virulent Snyder Hill virus sequence was the closest to the attenuated vaccine viruses. Most probably this relationship is due to the time of isolation of the Snyder Hill CDV which is closer to that of the vaccine viruses than to all the other isolates analysed in this tree. The data presented shows that RT-PCR of H genes in combination with fast sequencing methods is a sensitive tool to precisely characterise current CDV field viruses. In contrast, epitopic mapping studies on field isolates when compared with vaccine strains revealed only limited variability (Blixenkrone-Møller et al., 1992; Appel et al., 1994).
Acknowledgements The authors wish to thank Dr B. Rima, Belfast, for supplying the Snyder Hill CDV strain. This work was supported by the Deutscher AkademischerAustauschdienst DAAD, Bonn (British–German Research Cooperation programme). Dr L. Mamaev was the recipient of a Royal Society Fellowship.
References Appel, M.A. (1987) Canine distemper virus. In: M. Appel (Ed.), Virus Infection of Carnivores, Elsevier, New York, pp. 133 – 159. Appel, M.A., Yates, R.A., Foley, G.L., Bernstein, J.J., Santinelli, S., Spelman, L.H., Miller, L.D., Arp, L.H., Anderson, M., Barr, M., Pearce-Kelling, S. and Summers, B.A. (1994) Canine distemper epizootic in lions, tigers, and leopards in North America. J. Vet. Diagn. Invest. 6, 277 – 288. Baron, M. and Barrett, T. (1995) The sequence of the N and L genes of rinderpest virus, and the 5% and 3% extra-genic sequences: the completion of the genome sequence of the virus. Vet. Microbiol. 44, 175 – 186. Blixenkrone-Møller, M., Svansson, V., Appel, M., Krogsrud, J., Have, P. and O8 rvell, C. (1992) Antigenic relationships between field isolates from different carnivores. Arch. Virol. 123, 279 – 294. Blixenkrone-Møller, M., Appel, M., Pedersen, I.R., Dietz, H.H. and Henriksen, P. (1993) Studies on manifestations
L. Haas et al. / Virus Research 48 (1997) 165–171 of canine distemper virus infection in an urban dog population. Vet. Microbiol. 37, 163–173. Chamberlain, R.W., Wamwayi, H., Hockley, E., Subbarao, S.M., Goatley, L., Knowles, N.J. and Barrett, T. (1993) Evidence for different lineages of rinderpest virus reflecting their geographic isolation. J. Gen. Virol. 72, 443–447. Chappuis, G. (1995). Control of canine distemper. Vet. Microbiol. 44, 351 – 358. Curran, M.D., Clarke, D.K. and Rima, B.K. (1991) The nucleotide sequence of the gene encoding the attachment protein H of canine distemper virus. J. Gen. Virol. 73, 1189 – 1194. Felsenstein, J. (1993) Phylip 3.5p Manual. University of Washington, Seattle. Haas, L., Subbarao, S.M., Harder, T., Liess, B. and Barrett, T. (1991) Detection of phocid distemper virus RNA in seal tissues using slot hybridisation and the polymerase chain reaction amplification assay: genetic evidence that the virus is distinct from canine distemper virus. J. Gen. Virol. 74, 825 – 832. Harder, T.C., Klusmeyer, K., Frey, H.-R., O8 rvell, C. and Liess, B. (1993) Intertypic differentiation and detection of intratypic variants among canine and phocid morbillivirus isolates by kinetic neutralization using a novel immunoplaque assay. J. Virol. Methods 41, 77–92. Hu, A. and Norrby, E. (1994) Role of individual cysteine residues in the processing and antigenicity of the measles virus haemagglutinin protein. J. Gen. Virol. 75, 2173 – 2181. Lecouturier, V., Fayolle, J., Caballero, M., Carabana, J., Celma, M.L., Fernandez-Munoz, R., Wild, T. and Buckland, F. (1996). Identification of two amino acids in the haemagglutinin glycoprotein of measles virus (MV) that govern haemadsorption, HeLa cell fusion and CD46 downregulation: phenotypic markers that differentiate vaccine and wild-type MV strains. J. Virol. 70, 4200–4204. Mamaev, L.V., Denikina, N.N., Belikov, V.E., Volchkov, V.E., Visser, I.K.G., Fleming, M., Kai, C., Harder, T.C., Liess, B., Osterhaus, A.D.M.E. and Barrett, T. (1995)
.
171
Characterisation of morbilliviruses isolated from Lake Baikal seals (Phoca sibirica). Vet. Microbiol. 44, 251 – 259. O8 rvell, C., Blixenkrone-Møller, M., Svansson, V and Have, P. (1990). Immunological relationships between phocid and canine distemper virus studied with monoclonal antibodies. J. Gen. Virol. 71, 2085 – 2092. Rima, B.K., Earle, J.A.P., Yeo, R,P., Herlihy, L., Baczko, K., ter Meulen, V., Carabana, J., Caballero, M., Celma, M.L. and Fernandes-Munoz, R. (1995) Temporal and geographical distribution of measles virus genotypes. J. Gen. Virol. 76, 1173 – 1180. Rossiter, P.B. (1988). Antigenic variation between three strains of rinderpest virus detected by kinetic neutralisation and competitive ELISA using early rabbit sera. Vet. Microbiol. 16, 195 – 200. Rota, J.S., Hummel, N.B., Rota, P.A. and Bellini, W.J. (1992) Genetic variability of current wild-type measles virus isolates. Virology 188, 145 – 142. Tamin, A., Rota, P.A., Wang, Z., Heath, J.L., Anderson, L.J. and Bellini, W.J. (1994). Antigenic analysis of current wild type and vaccine strains of measles virus. J. Infect. Dis. 170, 795 – 801. Taylor, M.J., Godfrey, E., Baczko, K., ter Meulen, V., Wild, T.F. and Rima, B.K. (1991) Identification of different lineages of measles virus. J. Gen. Virol. 72, 83 – 88. Visser, I.K.G., Kumarev, V.P., O8 rvell, C., Vries, P. de, Broders, H.W.J., Bildt, M.W.G.van de, Groen, J. and Teppema, J.S. (1990) Comparison of two morbilliviruses isolated from seals during outbreaks of canine distemper in North West Europe and Siberia. Arch. Virol. 111, 149 – 164. Wamwayi, H.M., Fleming, M. and Barrett, T. (1995). Characterisation of African isolates of rinderpest virus. Vet. Microbiol. 44, 151 – 163. Zaghawa, A., Liess, B. and Frey, H.R. (1990). Antiserum raised in pigs against canine distemper virus and its utility in diagnostic procedures for morbillivirus infections (canine distemper, phocid distemper and rinderpest). J. Vet. Med. B, 37, 353 – 362.
Analysis of the haemagglutinin gene of current wild-type canine distemper virus isolates from Germany L. Haas a,*, W. Martens a, I. Greiser-Wilke a, L. Mamaev b, T. Butina b, D. Maack a, T. Barrett c a
Veterinary School Hanno6er, Institute of Virology, Bu¨nteweg 17, D-30559 Hanno6er, Germany b Limnological Institute, Ulan-Bayorskaya-3, Irkutsk, Russia c Institute for Animal Health, Pirbright Laboratory, Woking, Surrey GU 24 ONF, UK Received 10 September 1996; accepted 10 January 1997
Abstract The haemagglutinin (H) gene sequences from three wild-type canine distemper viruses (CDV) isolated during 1994–1995 were sequenced to determine whether contemporary strains had undergone significant genetic changes relative to the currently used vaccine strains. The new isolates were closely related to each other (\99%) and displayed about 90–91% sequence homology to the Onderstepoort and Convac vaccine strains. There were one to four additional potential glycosylation sites compared to the vaccine strains which were also present in a German dog CDV isolate dating from 1990. However, only a very slight reduction in neutralizing titre against the new isolates was found when compared with the Onderstepoort and Rockborn vaccine strains. Cysteine and proline residues were well conserved indicating a conserved three dimensional structure for the protein. By phylogenetic analysis the recent isolates showed a narrow clustering close to the previous canine isolates indicating a linear pattern of evolutionary changes. A comparison with published CDV H gene sequences suggested the presence of different lineages of CDV on a global scale and possible cocirculation of more than one genotype of CDV. © 1997 Elsevier Science B.V. Keywords: Canine distemper virus; Haemagglutinin gene; Neutralizing antibody titers
Canine distemper virus (CDV) is the the causative agent of a serious infection of dogs and many other carnivores (Appel, 1987). It is classified in the Morbilli6irus genus of the family * Corresponding author.
Paramyxoviridae. In developed countries with a high rate of vaccination against CDV clinical cases occur in an epidemic manner. Different suggestions are put forward to explain these outbreaks. It may be that after several years of absence of disease people become negligent about
0168-1702/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 8 - 1 7 0 2 ( 9 7 ) 0 1 4 4 9 - 4
166
L. Haas et al. / Virus Research 48 (1997) 165–171
vaccination. There may also be problems with the vaccination schedule, e.g. vaccinating puppies still possessing maternal antibodies which interfere with the generation of immunity. On the other hand it is occasionally assumed that the vaccines in use would not protect properly against current field virus infections even if applied correctly. Since the most widely used vaccines are modified live CDV vaccines which became available in the early 1960s (Chappuis, 1995), it is sometimes suggested that field viruses may have structurally and antigenically changed since that time and that these vaccines would thus no longer confer efficacious immunity. Immunological and sequence data have clearly shown that the major envelope glycoprotein (haemagglutinin or H protein) is the most variable of the morbillivirus proteins and therefore is the most suitable candidate to look for genetic changes in the virus (O8 rvell et al., 1990). During winter 1994/95 there were reports of CDV infections in some northern and eastern German cities. We isolated CDV from several cases in Hamburg and Berlin, amongst them isolates from young animals which according to the anamnesis showed clinical signs in spite of a previous vaccination record. The viruses were obtained from the buffy coat fraction of heparinized blood samples. White blood cells were co-cultivated on Vero, A72 and MDCK cell lines for up to three passages. RNA was prepared from infected cells when a cpe was visible using a RNA isolation kit (RNeasy®, Quiagen) following the instructions of the manufacturer. The H gene RNA of the three field isolates (animal no. 404, Berlin; nos. 2544 and 4513, Hamburg) was amplified by reverse transcriptase-polymerase chain reaction (RT-PCR) as decribed by Curran et al. (1991). We amplified the entire H gene (:1.9 kb) sequence using this method. As a comparison we also amplified the H gene from an older virulent CDV isolate (Snyder Hill). Amplicons were cloned into a specialised PCR cloning vector (pCRII®, Invitrogen). Dideoxy sequencing of double stranded plasmid DNA was carried out either using conventional radioactive sequencing techniques or an automatic ALF sequencer (Pharmacia) with the AutoRead T7 sequencing kit
(Pharmacia). Labelled M13 universal forward and reverse primers as well as sequence-specific internal H gene primers were used in the reactions. Sequences were analysed and aligned using the Seqaid II (version 3.81) and the Align Plus (Scientific and Educational Software, State Line, PA) software packages. For phylogenetic analysis the PHYLIP programs were used (Felsenstein, 1993). A single large open reading frame (ORF) coding for 607 amino acids was identified in each case and the deduced amino acid sequences aligned with other CDV H protein sequences obtained from the Genbank database (Fig. 1). Homologies to the published CDV vaccine strains Onderstepoort and Convac were between 90 and 91%. A high level of homology (\ 99%) was found between the new isolates. As expected, the three field viruses closely resembled an earlier German dog isolate dating from 1990 (5804/Han90). The extracellular glycosylation pattern of their H proteins showed some differences when compared with the vaccine strains. Four glycosylation sites (positions 19–21; 149–151; 422–424; 587–589) were shared by all CDVs but there were four additional potential glycosylation sites when compared with the Onderstepoort strain and one when compared with the Convac strain. The latter glycosylation site (position 309–311) was found in all recent virus field isolates but was absent in the older Snyder Hill strain. The potential glycosylation site at the extreme C-terminus (position 603–605) is unlikely to be used. It is not present in all field viruses and is missing from the Onderstepoort vaccine due to truncation of this protein at position 604. Amino acids such as proline and cysteine, which are known to be important in determining protein secondary structure, were highly conserved. Over 80% of the prolines and almost all 12 cysteine residues were completely conserved, the only exception being the Javelina CDV isolate where the cysteine residue at position 390 was replaced by a tryptophane. The H sequence results described here clearly show that the current isolates represent typical CDVs. The variation in sequence compared to the vaccine strains is most probably due to changes that have occurred in the field viruses over the past 60 or so years since the progenitors of the
L. Haas et al. / Virus Research 48 (1997) 165–171
vaccine strains were in circulation. Cell culture adaptation is unlikely to have radically changed the virus sequences since, in the case of the related rinderpest virus, it has been shown that the vaccine strain derived by multiple cell culture passages was more than 99% similar to the virulent
167
parental virus (Baron and Barrett, 1995). The H gene of current field isolates of rinderpest virus differ by about 10–12% in nucleotide sequence from the vaccine strain (Barrett, unpublished work) and similar levels of variation between recent field isolates and the vaccine strains have
Fig. 1. Alignment of the deduced amino acid sequences of recent German CDV dog isolates (H404, H2544, H4513) and the virulent Snyder Hill strain (Snyhill) with various CDV isolates available on the genbank database. Non-conservative substitutions are shown in white on a black background while conservative substitutions are shown in black on a grey background. Other virus sources: Ondvac, vaccine strain Onderstepoort (D00758); Convac, vaccine strain Convac (Z35493); H5804, German dog isolate (X85000); Seal, Siberian seal (X84998); Dkdog, Danish dog (GBVRL:CDVH3GEN); Gldog, Greenland dog (GBVRL:CDVH4GEN); Java, javelina (collared peccary) (GBVRL:CDVH11GEN); Usdog, American dog (GBVRL:CDVH9GEN); Bleo, American black leopard (GBVRL:CDVH10GEN); Racc, American raccoon (GBVRL:H13GEN), Dkmink, Danish mink (GBVRL:CDVH3GEN); Ferret, German ferret (X84999).
168
L. Haas et al. / Virus Research 48 (1997) 165–171
Fig. 1. (Contd.)
been reported for the H gene of measles virus (Tamin et al., 1994). However, structurally important amino acids were well conserved. Cysteine residues, which probably play a major role in the determining the antigenic structure and processing of H molecules (Hu and Norrby, 1994) were, with one exception, completely conserved. The role, if any, that the additional potential glycosylation site at position 309 – 311 found in the recent isolates plays in the immune response remains to be determined. An additional potential
glycosylation site was also observed in wild-type measles virus H genes compared with the vaccine strain, Moraten (Rota et al., 1992). A conspicuous narrow clustering of predicted amino acid changes around potential glycosylation sites was found in measles virus field isolates (Rota et al., 1992). While amino acid changes were scattered across the H protein in the case of CDV, slightly more amino acid changes were noted in the region around and between the two glycosylation sites located at positions 422–424 and 456–458, which
L. Haas et al. / Virus Research 48 (1997) 165–171
might indicate sites in the protein subject to a greater antigenic pressure and selective force. The receptor binding site for measles virus has been located around position 470 – 490 (Lecouturier et al., 1996) and this region of the protein may form part of the globular head of the mature protein. Some differences in cross-neutralization were detectable in the case of the earlier German canine field isolate (5804/Han90) using sera directed against the Rockborn vaccine strain (Harder et al., 1993) and similar neutralisation differences have been reported in the case of rinderpest virus (Rossiter, 1988). To test whether or not CDV positive sera from infected or vaccinated animals differ in their reactivity to current vaccine strains and an actual CDV isolate, respectively, neutralizing peroxidase-linked assays (NPLA) were performed as described by Zaghawa et al. (1990). The neutralizing antibody titres against isolate 2544 were simultaneously tested and compared to those of the Onderstepoort and Rockborn vaccine strains. Sera from naturally infected or vaccinated carnivores from our diagnostic unit were used and, as can be seen in Fig. 2, there was a slight but not significant reduction of neutralizing antibody titres measured against isolate 2544. It is therefore clear that the sequence changes identified do not have a significant effect on the ability of the vaccine to protect against the new strains. Phylogenetic analysis indicated a slowly progressive evolution in wild-type CDV based on changes in the H gene. The data were analysed by three different PHYLIP programes (NEIGHBOR, FITCH and KITCH) and a distance matrix derived using DNADIST. The same branching patterns were obtained with all three methods. As expected, the topology of the phylogenetic trees of the H gene sequences revealed a close clustering of the recent isolates and the previous German canine isolate and they clearly represent one lineage of the virus (Fig. 3). The German isolates were also closely related to CDV isolated from a Danish dog. Interestingly, a German ferret CDV and a Danish mink CDV formed a separate branch which might indicate a mustelid lineage of CDV. This is supported by similar phylogenetic relationships which were based on part of the F gene sequences of different mustelids (Harder,
169
Liermann and Haas, unpublished results). There was also a clustering of the American isolates, irrespective of the species from which the viruses originated. Interestingly, the more distantly related seal CDV isolate obtained from a Siberian seal (Phoca sibirica) in Lake Baikal in 1987 (Mamaev et al., 1995) was most related to CDV isolated from a Greenland dog in 1988 and might indicate a separate lineage circulating across the arctic ecosystem in susceptible species such as polar bear and arctic foxes. The Baikal seal virus is sometimes referred to as phocid distemper virus-2 (PDV-2), although it is now known to be CDV, to distinguish it from the first morbillivirus isolated from a seal (PDV-1) during the European seal morbillivirus epizootic and which is distinct from CDV (Visser et al., 1990; Haas et al., 1991).
Fig. 2. Neutralizing antibody titres (NPLA) measured using field sera with either CDV vaccine strain Rockborn (top) or Onderstepoort (bottom) compared with neutralising titres obtained with CDV field isolate 2554.
170
L. Haas et al. / Virus Research 48 (1997) 165–171
To summarize, although it is not possible to rule out that some of the immunization failures in properly vaccinated dogs may have been due to changes in the antigenic nature of current field strains, our data do not substantiate this. This is in agreement with previous antigenic studies performed on CDV field isolates in an urban dog population (Blixenkrone-Møller et al., 1993). Since CDV infections can realistically only be controlled by vaccination (Chappuis, 1995) care must be taken to apply a rigid vaccination schedule in combination with other prophylactic measures. The nucleotide sequences reported in this paper have been submitted to GenBank/EMBL. Accession numbers: virus 404 (Z77671); virus 2544 (Z77672); virus 4513 (Z77673) Fig. 3. Phylogenetic analysis of the nucleotide sequences of the coding regions of CDV H proteins. The tree was constructed using the PHYLIP DNADIST and KITSCH programes (Felsenstein, 1993). For comparison other CDV H gene sequences were extracted from the Genbank database (see legend Fig. 1).
Geographically distinct lineages have been described for other morbilliviruses such as rinderpest (Chamberlain et al., 1993; Wamwayi et al., 1995) and measles viruses (Taylor et al., 1991). In the case of measles virus, it was proposed that several lineages or genotypes co-circulate at a given time (Rima et al., 1995). Our phylogentic analysis shows an analogous situation for CDV since in Europe different lineages have been identified in canines and in mustelids. It is interesting to note that the virulent Snyder Hill virus sequence was the closest to the attenuated vaccine viruses. Most probably this relationship is due to the time of isolation of the Snyder Hill CDV which is closer to that of the vaccine viruses than to all the other isolates analysed in this tree. The data presented shows that RT-PCR of H genes in combination with fast sequencing methods is a sensitive tool to precisely characterise current CDV field viruses. In contrast, epitopic mapping studies on field isolates when compared with vaccine strains revealed only limited variability (Blixenkrone-Møller et al., 1992; Appel et al., 1994).
Acknowledgements The authors wish to thank Dr B. Rima, Belfast, for supplying the Snyder Hill CDV strain. This work was supported by the Deutscher AkademischerAustauschdienst DAAD, Bonn (British–German Research Cooperation programme). Dr L. Mamaev was the recipient of a Royal Society Fellowship.
References Appel, M.A. (1987) Canine distemper virus. In: M. Appel (Ed.), Virus Infection of Carnivores, Elsevier, New York, pp. 133 – 159. Appel, M.A., Yates, R.A., Foley, G.L., Bernstein, J.J., Santinelli, S., Spelman, L.H., Miller, L.D., Arp, L.H., Anderson, M., Barr, M., Pearce-Kelling, S. and Summers, B.A. (1994) Canine distemper epizootic in lions, tigers, and leopards in North America. J. Vet. Diagn. Invest. 6, 277 – 288. Baron, M. and Barrett, T. (1995) The sequence of the N and L genes of rinderpest virus, and the 5% and 3% extra-genic sequences: the completion of the genome sequence of the virus. Vet. Microbiol. 44, 175 – 186. Blixenkrone-Møller, M., Svansson, V., Appel, M., Krogsrud, J., Have, P. and O8 rvell, C. (1992) Antigenic relationships between field isolates from different carnivores. Arch. Virol. 123, 279 – 294. Blixenkrone-Møller, M., Appel, M., Pedersen, I.R., Dietz, H.H. and Henriksen, P. (1993) Studies on manifestations
L. Haas et al. / Virus Research 48 (1997) 165–171 of canine distemper virus infection in an urban dog population. Vet. Microbiol. 37, 163–173. Chamberlain, R.W., Wamwayi, H., Hockley, E., Subbarao, S.M., Goatley, L., Knowles, N.J. and Barrett, T. (1993) Evidence for different lineages of rinderpest virus reflecting their geographic isolation. J. Gen. Virol. 72, 443–447. Chappuis, G. (1995). Control of canine distemper. Vet. Microbiol. 44, 351 – 358. Curran, M.D., Clarke, D.K. and Rima, B.K. (1991) The nucleotide sequence of the gene encoding the attachment protein H of canine distemper virus. J. Gen. Virol. 73, 1189 – 1194. Felsenstein, J. (1993) Phylip 3.5p Manual. University of Washington, Seattle. Haas, L., Subbarao, S.M., Harder, T., Liess, B. and Barrett, T. (1991) Detection of phocid distemper virus RNA in seal tissues using slot hybridisation and the polymerase chain reaction amplification assay: genetic evidence that the virus is distinct from canine distemper virus. J. Gen. Virol. 74, 825 – 832. Harder, T.C., Klusmeyer, K., Frey, H.-R., O8 rvell, C. and Liess, B. (1993) Intertypic differentiation and detection of intratypic variants among canine and phocid morbillivirus isolates by kinetic neutralization using a novel immunoplaque assay. J. Virol. Methods 41, 77–92. Hu, A. and Norrby, E. (1994) Role of individual cysteine residues in the processing and antigenicity of the measles virus haemagglutinin protein. J. Gen. Virol. 75, 2173 – 2181. Lecouturier, V., Fayolle, J., Caballero, M., Carabana, J., Celma, M.L., Fernandez-Munoz, R., Wild, T. and Buckland, F. (1996). Identification of two amino acids in the haemagglutinin glycoprotein of measles virus (MV) that govern haemadsorption, HeLa cell fusion and CD46 downregulation: phenotypic markers that differentiate vaccine and wild-type MV strains. J. Virol. 70, 4200–4204. Mamaev, L.V., Denikina, N.N., Belikov, V.E., Volchkov, V.E., Visser, I.K.G., Fleming, M., Kai, C., Harder, T.C., Liess, B., Osterhaus, A.D.M.E. and Barrett, T. (1995)
.
171
Characterisation of morbilliviruses isolated from Lake Baikal seals (Phoca sibirica). Vet. Microbiol. 44, 251 – 259. O8 rvell, C., Blixenkrone-Møller, M., Svansson, V and Have, P. (1990). Immunological relationships between phocid and canine distemper virus studied with monoclonal antibodies. J. Gen. Virol. 71, 2085 – 2092. Rima, B.K., Earle, J.A.P., Yeo, R,P., Herlihy, L., Baczko, K., ter Meulen, V., Carabana, J., Caballero, M., Celma, M.L. and Fernandes-Munoz, R. (1995) Temporal and geographical distribution of measles virus genotypes. J. Gen. Virol. 76, 1173 – 1180. Rossiter, P.B. (1988). Antigenic variation between three strains of rinderpest virus detected by kinetic neutralisation and competitive ELISA using early rabbit sera. Vet. Microbiol. 16, 195 – 200. Rota, J.S., Hummel, N.B., Rota, P.A. and Bellini, W.J. (1992) Genetic variability of current wild-type measles virus isolates. Virology 188, 145 – 142. Tamin, A., Rota, P.A., Wang, Z., Heath, J.L., Anderson, L.J. and Bellini, W.J. (1994). Antigenic analysis of current wild type and vaccine strains of measles virus. J. Infect. Dis. 170, 795 – 801. Taylor, M.J., Godfrey, E., Baczko, K., ter Meulen, V., Wild, T.F. and Rima, B.K. (1991) Identification of different lineages of measles virus. J. Gen. Virol. 72, 83 – 88. Visser, I.K.G., Kumarev, V.P., O8 rvell, C., Vries, P. de, Broders, H.W.J., Bildt, M.W.G.van de, Groen, J. and Teppema, J.S. (1990) Comparison of two morbilliviruses isolated from seals during outbreaks of canine distemper in North West Europe and Siberia. Arch. Virol. 111, 149 – 164. Wamwayi, H.M., Fleming, M. and Barrett, T. (1995). Characterisation of African isolates of rinderpest virus. Vet. Microbiol. 44, 151 – 163. Zaghawa, A., Liess, B. and Frey, H.R. (1990). Antiserum raised in pigs against canine distemper virus and its utility in diagnostic procedures for morbillivirus infections (canine distemper, phocid distemper and rinderpest). J. Vet. Med. B, 37, 353 – 362.