Studies on the origin of pandemic influenza: III. Evidence implicating duck and equine influenza viruses as possible progenitors of the Hong Kong strain of human influenza

Two strains of influenza virus isolated from horses and ducks in 1963, A/equine/Miami/1/63 (Heq2 Neq2) and A/duck/Ukraine/1/63 (Hav7 Neq2) were found to possess hemagglutinin subunits which cross-reacted in hemagglutination-inhibition and immunodiffusion tests with those of the Hong Kong strain of human influenza A/Hong Kong/1/68 (H3 N2).Peptide maps of the heavy polypeptide chains from the hemagglutinin subunits of these three strains showed a number of differences, but maps of the light chains were almost identical, indicating that the light polypeptide chains from the hemagglutinin subunits of these animal, avian and human viruses had practically the same amino acid sequence.One explanation of these results is that the three viruses arose, by genetic recombination, from a common ancestor.     

Hemagglutinin molecules of Hong Kong, equine-2, and duck/Ukraine influenza viruses lack N-terminal aspartic acid.

We have found that hemagglutinin molecules from equine-2 (Heq2Neq2) and duck/Ukraine (Hav7Neq2) influenza viruses do not possess N-terminal aspartic acid and that the N-terminus of HA1 from these viruses seems to be blocked. In this respect, these hemagglutinin molecules are similar to those of Hong Kong influenza virus (H3N2) and its variants and unlike those of every other influenza virus examined (a total of 12 strains, including those of the Asian H2N2 series) which have been found to possess N-terminal aspartic acid (or asparagine) on the hemagglutinin polypeptides.     

Ortho- and paramyxoviruses from migrating feral ducks: characterization of a new group of influenza A viruses.

Ortho- and parainfluenza viruses isolated from the cloacas of migrating feral ducks shot on the Mississippi flyway included three strains of influenza. A virus (Hav6 Nav1, Hav6 Nl, Hav7 Neq2) as well as Newcastle disease virus. One influenza virus, A/duck/Memphis/546/74, possessed Hav3 haemagglutinin, but the neuraminidase was not inhibited by any of the known influenza reference antisera. The neuraminidase on this virus was related to the neuraminidases on A/duck/GDR/72 (H2 N?), A/turkey/Ontario/7732/66 (Hav 5 N?), A/duck/Ukraine/1/60 (Hav3 N?) and A/turkey/Wisconsin/68. We therefore propose that the neuraminidase on this group of influenza viruses be designated Nav6. The A/duck/Memphis/546/74 influenza virus caused an ocular discharge in 1 of 5 ducks and was shed in faeces for 10 days; it was stable in faecal samples for up to 3 days at 20 degrees C. These results suggest that ecological studies on influenza in avian species should include attempts to isolate virus from faeces. Faecal-oral transmission is an attractive explanation for the spread of influenza virus from feral birds to other animals.     

Isolation and characterization of influenza A viruses from wild ducks in northern Japan: appearance of HSW1 antigens in the Japanese duck population.

Twenty-six influenza A viruses were isolated from cloacal and tracheal samples of 235 resident and 396 migratory ducks in Miyagi prefecture, Japan, in 1977--78. Of these, twelve were antigenically related to the avian-origin HSW1 virus, A/duck/Alberta/35/76 (HSW1N1), but their neuraminidase antigens were characterized as Nav2-3, Nav4 or N2. These antigenic configuration have not previously been reported. In addition, one strain in which the neuraminidase antigen was identified as Nav4, was demonstrated to be a mixture of two haemagglutinins, HSW1 and Hav7. Two distinct strains were separated from the mixture and characterized as HSW1Nav4 and Hav7Nav4. The antigenic identification of an additional 13 influenza A viruses revealed the presence of six haemagglutinin subtypes (Hav1, Hav3, Hav4, Hav6, Hav7, and Hav8) and five neuaraminidase subtypes (Nav1, Nav2-3, Nav4, Neq2, and N2) in various combinations. The results suggest that the avian influenza A viruses among feral ducks may be isolated in various combinations of haemagglutinins and neuraminidase subtypes in Japan, and that feral ducks may be the site of genetic recombination occurring as a result of dual infection with different subtypes of influenza A virus.     

Isolation of influenza a viruses from commercial ducks on a farm in Norfolk between August 1979 and March 1980

Investigation of respiratory disease and high mortality which occurred on a commercial duck fattening farm between August 1979 and March 1980 resulted in the isolation of 10 influenza A viruses. The viruses were characterised as Hav6 N2 (three isolates), Hav4 Navl (four isolates), Hav4 Nl (two isolates) and Hav7 Neq2 (one isolate) subtypes by haemagglutination-inhibition and neuraminidase-inhibition tests. A Newcastle disease virus isolate was also obtained from the ducks. All isolates had low intravenous pathogenicity indices in 6-week-old chickens.     

Novel influenza A viruses isolated from Canadian feral ducks: including strains antigenically related to swine influenza (Hsw1N1) viruses.

Twelve influenza A viruses, antigenically related to the Ho, H1 and Hsw1 subtypes, were isolated from cloacal samples of feral ducks in Canada. Antigenic comparisons showed that these viruses were most closely related to the recent HSW1N1 isolates from man and pigs, whereas in vivo pathogenicity tests revealed differences between the Hsw1N1 viruses from the ducks and those from humans and pigs. Antigenic characterization of 94 additional influenza A viruses from the ducks showed four haemagglutinin subtypes (Hav1, Hav4, Hav5 and Hav7), an unclassified haemagglutinin, and six neuraminidase subtypes (N1, N2, Neq2, Nav1, Nav2 and Nav5) in various combinations, some of which are novel and have not previously been reported. Three of these duck influenza viruses possessed a haemagglutinin antigenically related to that of classical fowl plaque virus. A much higher percentage of virus isolations were from juvenile ducks (18.5%) than from adults (5%). All of the ducks, from which viruses were isolated, appeared healthy at the time of sampling. Serological studies on a limited number of humans and domestic birds living in close proximity to the Canadian ducks revealed no evidence of interspecies transmission. Our findings suggest that these birds serve as a substantial reservoir of antigenically diverse influenza viruses, including isolates antigenically related to the current human and animal influenza viruses. This reservoir in nature may be perpetuated by a cycle involving annual infection of juvenile birds followed by transmission to the remaining susceptible birds until the next congregation during the breeding season.     

Characterisation of influenza A viruses isolated from turkeys in England during March-May 1979

During the early spring of 1979 turkeys on at least twelve sites in England became infected with influenza A viruses. On five of these sites no virus was isolated but birds were shown to have antibodies to Havl (four sites) and Hav2 antigenic subtypes of influenza A viruses. The eight viruses isolated were typed: A/turkey/England/192-328/79 (Havl Nav2/3), A/turkey/England/192-329/79 (Hav1 N2), A/turkey/England/199/79 (Hav1 Neq1), A/turkey/ England/214/79 (Hav1 Neq1), A/turkey/England/250/79 (Hsw1 N1), A/turkey/England/262/79 (Hav1 Nav2/3), A/turkey/England/272/79 (Havl Neq1), A/turkey/England/384/79 (Hav2 Nav4). Pathogenicity index tests in 6-week-old chickens agreed with the clinical signs seen in turkeys in the field. Three of the isolates: 199, 214 and 272 were of extremely high virulence, 384 showed intermediate virulence, while the other isolates were of low virulence.     

Properties of Hav6Neq2 and Hswl(H0)Hav2 influenza viruses isolated from waterfowl in southern Turkmenia

Three influenza A virus strains were isolated from shorebirds in October, 1977, in southern Turkmenia, in the vicinities of Tedzhen water reservoir. From a common tern, A/Sterna hirundo/Turkmenia/45/77 strain was isolated with the antigenic formula Hav6Neq2, from a teal and a black-headed gull influenza A/Anas crecca/Turkmenia/4/77 and A/Larus ridibundus/Turkmenia/13/77 strains with previously unknown combination of surface antigens Hswl(H0)Nav2 were recovered. By the molecular weight of the heavy (HA1 59,000 d) and light (HA2 24,000 d) chains of hemagglutinin, the Turkmenian viruses A/Larus ridibundus/Turkmenia/13/77 and A/Anas crecca/Turkmenia/4/77 are similar to each other and to the strains having H0 hemagglutinin: A/PR8/34 (H0N1) and A/Whale/PO/19/76 (H09Nav2). The Turkmenian viruses are characterized by a low content of the light hemagglutinin chain (HA2) which is typical of the viruses with Hsw1 hemagglutinin: A/New Jersey/8/76 (Hsw1N1) and A/SW/Wisk/68 (Hsw1N1).     

Host range recombinants of fowl plague (influenza A) virus

Recombinants between the influenza virus strains fowl plague virus (FPV, Hav1N1) and Hong Kong (H3N2) have been isolated which form plaques on MDCK cells but not on chick embryo cells, although they carry the hemagglutinin of FPV. These host range recombinants have been characterized and one of them has been used for a second recombination with virus N (Hav2Neq1) or equi 2 (Heq2Neq2) on chick embryo cells. In this way, recombinants were obtained with a mixed genome which have regained the natural host range of FPV and some pathogenic properties for chicken. The results are discussed as a possible mechanism for a pandemic influenza strain to survive in an animal reservoir by changing its host range by recombination, and to regain the original host range by a second recombination but always keeping the same hemagglutinin.     

Isolation of the influenza virus from ducks in the Aral Sea region

Influenza A/duck/Aralsk/1/78 virus was isolated from ducks in southwestern Kazakhstan. Typing of this virus neuraminidase indicated its appurtenance to the 2nd avian type. Hemagglutinin of A/duck/Aralsk/1/78 virus typed by the HI test showed no relationship with any one of the strains used in the test including H0, H1, H2, H3, Hsw1, Heq1, Heq2, viruses and strains with all 10 avian hemagglutinins. This suggests that hemagglutinin of A/duck/Aralsk/1/78 virus (Hav?Nav2) belongs to a new, hitherto unknown type.     

Evidence for antigenic variation in influenza A nucleoprotein.

Nucleoprotein (NP) antigens isolated from sodium sarcosyl detergent-disrupted influenza A virus particles by cellulose acetate electrophoresis were used to prepare specific immune sera. Antigenic analysis of the nucleoproteins by immuno-double-diffusion and antibody absorption tests revealed antigenic differences in the nucleoprotein antigens of human A/PR8/34 (HON1) virus and viruses of the Hong Kong (H3N2) subtype. The nucleoprotein antigens of avian A/duck/Ukraine/1/63 (Hav7Neg2) and A/swine/Iowa/15/30 (Hsw1Nl) viruses resembled more closely the NP of A/PR8/34 virus than the NP of human H3N2 strains. Antigenic analysis of recombinant strains prepared from A/PR8/34 and H3N2 parent viruses indicated that the parental origin of the NP antigens could be clearly identified. Identification of the nucleoproteins of parental and recombinant influenza A viruses by migration rate analysis of NP polypeptides and RNA genes by electrophoresis on polyacrylamide gels gave results which correlated with the antigenic characterization of their NP antigens. The findings suggested that antigenic “drift” occurs in the nucleoproteins of human influenza A viruses.     

Detection of the determinants of influenza virus H3 hemagglutinin by the technic of radioimmunologic analysis

The interrelations between H3/73 hemagglutinin of human influenza virus and the other 16 mammalian and avian hemagglutinin subtypes (a total of 50 strains) were studied by the method of radioimmunologic analysis (RIA). The antigenic relations of H3, Hav7 and Heq2 were confirmed, certain common determinants were also found in H3/73 hemagglutinin and avian viral Hav6 and Hav9 hemagglutinins. No interrelations were revealed with previously circulating human influenza viruses H0, H1, H2 as well as with swine influenza virus and avian viruses Hav1-Hav5, Hav8. It has been shown that the H3/73 determinant in some avian viruses evolves similarly to drift-variants of human influenza virus. The method can be recommended for fine analysis of influenza virus antigenic structure as it allows detecting small antigenic quantities.     

Type A influenza virus strains isolated from free living ducks in Czechoslovakia during 1978-1981

Eight influenza virus A strains were isolated from 269 cloacal swabs taken from wild ducks (Anas platyrhynchos), shot during their autumn migrations in the years 1978-1981. One strain was identified as subtype A-H3N8N6 (Hav7Neq2Nav1), the remaining seven as subtype A-H4N6 (Hav4Nav1).     

Characterization of an influenza A virus from seals

An influenza A virus antigenically similar to A/FPV/Dutch/27 (Hav1Neq1) [H7N7] was isolated from harbor seals (Phoca vitulina) that had died of acute hemorrhagic pneumonia on Cape Cod Peninsula, beginning in the winter of 1979–1980. High titers of virus were obtained from the lungs and lower titers from the brains of the seals. Although antigenic analyses and characterization of the RNAs show that all of the genes and gene products are closely related to different avian influenza viruses, biologically the virus behaves more like a mammalian strain. The seal virus replicated and produced pneumonia in experimentally infected harbor seals, but the clinical course and pathology were less severe than in the natural infection; the virus also replicated in ferrets, cats, and pigs but produced no disease. In avian species, the seal influenza virus replicated poorly, produced no disease signs, and was not shed in the feces. Although the seal influenza virus can cause conjunctivitis in humans who have known contamination of the eyes from infected animals, serological studies detected no evidence of seroconversion among persons working with infected seals or with the virus. Preliminary studies detected antibodies to this virus in harbor seals on the New England coast but not in harbor seals, gray seals, or fur seals from other areas, suggesting that this virus may be a new introduction to this species. An Hav1Neq1 [H7N7] virus was also isolated from feral ducks in Iceland in 1980, but the two viruses could be distinguished by analysis of their RNAs and host range. The A/Seal/Mass/1/80 influenza virus provides the first evidence suggesting that a strain deriving all of its genes from one or more avian influenza viruses can be associated with severe disease in a mammalian population in nature. Whether this breach of species specificity represents a unique event in influenza evolution remains to be determined, but raises the possibility that human or animal influenza viruses may be derived directly from avian strains.     

Antigenic determinants of human influenza viruses among the influenza viruses isolated from animals

Comparative studies of the antigenic properties of hemagglutinin (HA) of animal and human viruses revealed both similarities between them and complete differences in the composition of antigenic determinants. Avian influenza viruses A/chicken/Kamchatka/12/71, A/pintail/Primorie/730/76, and A/bat/Alma-Ata/73/77 were completely identical with human strains of influenza virus. Influenza A/horse/Miami/63 contains one antigenic determinant H3.1.HA of A/tern/Turkmenia/18/73 (Hav7) viruses has a peculiar set of antigens. Apart from two antigenic determinants H3.1 and H3.3 inherent in human virus strains, HA of A/tern/Turkmenia/18/73 virus contains an antigenic determinant the population of antibodies to which shows no relation to HA of subtypes Hav2-Hav9.     

Growth of avian and human influenza viruses in organ cultures of duck and chicken colons

Summary Colons from ducks and chicken 1, 7, 14 and 28 days old maintained near-normal morphology up to 48 and 96 hours respectively in a system using NCTC 135 medium (1 part)+Dulbecco's modified Eagle's medium (9 parts), at 37° C and 95 per cent 0₂/5 per cent CO₂. In the colon of 1 and 28 day-old ducks, duck influenza virus (Hav7N2) and budgerigar influenza virus (Hav4Nav1) grew to peak titer by hour 72, whereas human influenza virus (H3N2) did not grow. In the colon of 1 day-old chicken, the three viruses grew in the order of first duck virus, then budgerigar virus and then human virus, but in the colon of 28 day-old chicken, the growth of human virus was much less. Specific fluorescence was demonstrated in the mucosal epithelium of the colon of ducks and chicken, and intensity of fluorescence correlated with virus yield. The fact that the avian and not the human influenza viruses showed good growth in the duck clons coincided with the fact that influenza viruses possessing avian hemagglutinin subtypes have frequently been isolated in nature from duck intestines.With 3 Figures     

Isolation of an influenza A virus from seals

Summary Influenza A virus of serotype Hav1 Neq1 (H7N7 by the 1980 revised influenza typing system proposed by WHO experts) was repeatedly isolated from lung and brain tissues taken from harbor seals (Phoca vitulina) found suffering from pneumonia on Cape Cod Peninsula (U.S.A.) in the winter of 1979–1980. The seal isolates, although of a serotype identical to some fowl plague virus strains, were harmless to chickens and turkeys in transmission experiments. An earlier human infection by a Hav1 Neq1 influenza virus and the serologic relatedness of this avian serotype with the equine 1 serotype are cited in support of the view that influenza viruses with these antigenic characteristics seem to have a facility to pass from birds to mammals.     

On the origin of the human influenza virus subtypes H2N2 and H3N2

The RNA of the human influenza virus Singapore (H2N2) strain has been labeled in vivo by phosphorus-32 and separated by polyacrylamide gel electrophoresis into eight segments, which were correlated to the corresponding gene functions and/or proteins. The base sequence homology between the individual genes (segments) of the H2N2 virus and those of different influenza A strains has been determined by molecular hybridization. Segments 1, 5, 7, and 8 of the Singapore strain exhibit a base sequence homology of almost 100% as compared to the FM1 strain (HlNl), while the homology between the other segments is significantly lower (24–76%). For the Singapore and Hong Kong (H3N2) strains all segments except that coding for the hemagglutinin (HA, 24%) exhibit a homology close to 100%. The ³²P-labeled segment 4 (HA-gene) of the avian influenza A strain duck Ukraine (Hav7Neg2) shows a homology of 92% to Hong Kong, while the homology of at least two other segments is significantly lower. These results are taken as an indication that the H2N2 subtype is derived from the HlN1 subtype by a recombination event retaining four H1N1 segments, while the other four segments are gained from another yet unknown strain. The H3N2 subtype is presumably derived from a H2N2 subtype, retaining seven segments of the H2N2 subtype, while the gene coding for the HA is obtained from the duck Ukraine or another highly related strain.     

Genes involved in the virulence of an avian influenza virus

A/chicken/Japan/24 (Japan) (HavlNegl) isolated from an outbreak of fowl plague-like illness among chickens is comparable to fowl plague viruses in its virulence, causing paralysis and death of chickens within 3 days. It also kills chick embryos after allantoic infection of embryonated eggs. Another avian influenza virus, A/duck/Ukraine/1/63 (Ukraine) (Hav7Neg2), in contrast, causes no illness or death in either chickens or chick embryos. In an attempt to determine the genes involved in the observed difference in virulence, we prepared a series of recombinants between Japan and Ukraine viruses. Parental derivation of genes in the recombinants was determined by urea-polyacrylamide gel electrophoresis of viral RNA. The virulence was assessed by three criteria: (1) lethality for chickens after intracerebral inoculation, (2) lethality for embryos after allantoic inoculation into 15-day-old embryonated eggs, and (3) lethality for embryos after allantoic inoculation into 10-day-old embryonated eggs. The recombinants in which genes coding for hemagglutinin (HA), neuraminidase (NA), and matrix (M) proteins had been derived from Japan virus and five other genes from Ukraine virus were fully virulent by all three criteria. The recombinants which had received HA and NA genes from Japan virus showed a slightly but definitely reduced virulence in chickens and 15-day-old embryos but retained the virulence in 10-day-old embryos unchanged. The recombinants which had received only HA gene from Japan virus had a markedly diminished virulence in chickens and 15-day-old embryos but were as lethal for 10-day-old embryos as Japan virus. It was concluded that HA gene was the key determinant of virulence of Japan virus, but, in addition, NA and M genes were required for the full expression of virulence. Requirement of HA and NA genes was confirmed by reconstitution of virulence after reuniting the two genes.