- Volume 76, Issue 1, 1995

Surveillance for influenza A virus infection of seals has continued following the association of influenza A virus with epizootics of pneumonia in seals off the New England coast in 1979–1980 and 1982–1983. In January 1991 and January to February 1992, influenza A viruses were isolated from seals that died of pneumonia along the Cape Cod peninsula of Massachusetts. Antigenic characterization identified two H4N6 and three H3N3 viruses. This was the first isolation of H3 influenza viruses from seals, although this subtype is frequently detected in birds, pigs, horses and humans. Haemagglutination inhibition assays of the H3 isolates showed two distinct antigenic reactivity patterns: one more similar to an avian reference virus (A/Duck/Ukraine/1/63) and one more similar to a human virus (A/Aichi/2/68). The haemagglutinin (HA) genes from two of the H3 seal viruses showing different antigenic reactivity (A/Seal/MA/3911/92 and A/Seal/MA/3984/92) were 99.7% identical, with four nucleotide differences accounting for four amino acid differences. Phylogenetic analysis demonstrated that both of these sequences were closely related to the sequence from the avian H3 virus, A/Mallard/New York/6874/78. This indicates that influenza A viruses of apparent avian origin, including the H3 subtype viruses, continue to infect seals.

Thirty-one strains of human influenza A (H1N1) viruses isolated in Europe, mostly in Finland, from 1978–1992 were compared with respect to their nucleotide sequences coding for the HA1 portion of haemagglutinin. In 1984, at least two sublineages of H1N1 subtype viruses co-circulated in Finland. The viruses isolated after 1986 formed three sequential phylogenetic clusters. Loss of glycosylation sites, on the globular head of the HA1 portion suggests that oligosaccharides at these sites are not necessarily advantageous for the human virus. Isolation of a herald strain in Finland in June 1988 raised the question as to whether the virus was able to survive in Europe throughout the non-epidemic summer period. Demonstration of highly conserved strains, found over two continents in 1988, is further evidence of the existence of infection chains whose viruses have not been subjected to random sampling or selection events.

The nature and distribution of hepatitis C virus (HCV) genotypic variants present in south-east Asia have not been extensively investigated. We analysed HCV RNA obtained from 67 clinical serum samples from Singapore, Thailand, Indonesia, the Philippines and South Korea. All samples were amplified by semi-nested RT-PCR and the nucleotide sequence determined for four regions within the E1, E2/NS1, NS4 and NS5 genes. Each isolate had a unique nucleotide and deduced amino acid sequence, consistent with the genetic heterogeneity of this virus. There was remarkably little amino acid sequence variation between isolates of the same genotype, apart from variable domains within putative envelope glycoproteins that are likely to be under immune pressure. All isolates could be classified according to the currently recognized genotypes of HCV, with the exception of one Singapore isolate that defined a new group 3 subtype. The 1b genotype, which predominates in Japan, was the most widely distributed genotype and accounted for 58% of all isolates sequenced. Regional variations in HCV genotype distribution were observed, with type 3a being found almost exclusively in Thailand. By contrast, the 1a genotype, which predominates in the USA was the most prevalent genotype in the Philippines. Genotype 1a was found less commonly among the Thai isolates, presumably having been introduced from the West in stored blood products or by sporadic transmission. The significant prevalence of HCV types 2 and 3 restates the need for variant genotypes to be included in immunodiagnostic and vaccine development strategies. This study reveals that the 1b genotype of HCV, previously found to be the major variant present in east Asia, also predominates in the south-east Asian region, and may be the major HCV type found worldwide.

Monoclonal antibodies (MAbs) prepared against the yellow fever virus (YF) vaccine strain 17D (17D YF) envelope E protein were used to investigate Fc piece involvement in antibody-mediated protection against YF encephalitis in mice. 17D YF passaged either in Vero cells or in mouse brain (P-YF) to increase neurovirulence was used. To avoid uncertainty concerning antibody clearance and blood–brain barrier penetration, and to directly compare protective activity with neutralization in vitro, pre-formed antibody–virus complexes were injected intracerebrally or assayed for plaque formation in parallel. F(ab′)2 fragments of an IgG2a MAb that strongly neutralized both YF strains retained molar equivalent neutralizing activity in vitro, but did not protect. However, further incubation of such F(ab′)2-virus antibody complexes with rabbit IgG, but not F(ab′)2 anti-mouse IgG resulted in protection. To unambiguously test for Fc piece involvement in this model we derived an IgG2a isotype switch variant from a protective IgG1 MAb-secreting hybridoma and prepared F(ab′)2 fragments of the derivative. Intact and fragmented antibodies exhibited weak neutralizing activity. The variant antibody failed to protect against P-YF, but against considerably less neurovirulent 17D YF its protective capacity was 10-fold higher than that of its IgG1 parent. F(ab′)2 fragments of the variant did not protect. Together, these results provide strong evidence of an in vivo protective function for the anti-virion antibody Fc piece and indicate that in vitro neutralizing activity as a predictor of antibody protective capacity is dependent on Fc piece integrity and isotype.

A porcine rotavirus strain, CN86, originally isolated from rotavirus-infected piglets in Argentina, has been shown to possess unique characteristics. It was the first animal strain described to be antigenically related to human serotype G1 and the standard counterpart of another porcine strain showing rearrangement of genome segment 11. Owing to these features, molecular characterization of this virus seemed relevant. The gene encoding the major inner capsid protein, VP6, was cloned and its nucleotide sequence was determined. Comparative analysis of the deduced amino acid sequence of CN86 VP6 with those representing the four different subgroups showed that it is more closely related to subgroup II human Wa and porcine Gottfried strains, albeit to a lesser extent than they are to each other. Despite exhibiting sequence divergence, CN86 VP6 has 12 out of the 14 residues expected to be conserved in strains bearing subgroup II specificity. Interestingly, CN86 VP6 shows a high degree of homology with VP6 of porcine strain YM rotavirus which, although being closely related to subgroup II strains, has been serologically characterized as subgroup I. Subgroup II reactivity of CN86 strain, predicted by sequence analysis, was confirmed by ELISA with subgroup-specific monoclonal antibodies. Taken together, our results provide evidence for the existence of a human–pig lineage for rotavirus gene 6.

The nucleotide sequence of the 5′-terminal P1 coding region of an aphid-transmissible isolate of zucchini yellow mosaic virus (ZYMV; strain FL/AT), a mild isolate (strain MD) and a severe isolate (strain SV), all from Florida, were compared with two other ZYMV isolates. The ZYMV MD and SV isolates and an isolate from California (ZYMV CA) had 95–98% sequence similarities to FL/AT, whereas an isolate from Reunion Island (ZYMV RU) had a 60% sequence similarity to FL/AT. ZYMV MD had an 18 nucleotide insert following the start codon of the P1 coding region. The P1 proteins of all ZYMV isolates shared conserved amino acids in areas of the C terminus similar to those reported for other potyviruses. Polyclonal antisera were prepared to the P1 proteins of ZYMV FL/AT and RU expressed in Escherichia coli. The FL/AT and RU P1 antisera showed varying degrees of reactivity in Western blots with extracts of pumpkin (Cucurbita pepo L.) singly infected with a number of distinct ZYMV isolates. The reaction of the FL/AT P1 antiserum with isolate RU-infected tissue extracts was very weak compared to the homologous reaction. Neither antiserum reacted with extracts from plants singly infected with three other potyviruses, a potexvirus, or a cucumovirus. The P1 proteins of ZYMV isolates ranged in molecular mass from 33 kDa to 35 kDa. The P1 protein of strain MD was larger (35 kDa) than that of FL/AT (34 kDa). Indirect immunofluorescence tests with FL/AT P1 antiserum indicated that the P1 protein aggregates in ZYMV-infected tissues. The antisera to the ZYMV P1 proteins have potential as serological probes for identifying ZYMV and for distinguishing ZYMV isolates by immunoblotting.

Particles of several symptom-modulating TCV coat protein (CP) mutants were pretreated at pH 5·5, 7·5 or 8·5 and their conformations compared by agarose gel electrophoresis to those of wild-type particles. Particles of two mutants were swollen under conditions in which wild-type particles remained contracted; particles of one mutant were contracted under conditions in which wild-type particles were swollen; a portion of the particles of one mutant was contracted and another portion swollen under conditions in which wild-type particles remained contracted; and particles of one mutant, that elicited wild-type symptoms, comigrated with wild-type particles under all conditions tested. The results of in vitro translation experiments with mutant particles were essentially similar to those with wild-type particles, despite conformational differences at pH 5·5 and 8·5. These results suggest that more than the swollen conformation is required for in vitro translation, and that particle conformation may play a role in symptom elicitation.