The role of reverse genetics in the development of vaccines against respiratory viruses.

Despite their significance, the only available vaccines against respiratory viruses are those for the prevention of influenza. Attempts have been made to produce vaccines against other respiratory viruses using traditional techniques, but have met with little success. Reverse genetics, although still a relatively new tool for the manipulation of negative-strand RNA viruses, has great potential for the preparation of vaccines against many of the common respiratory viruses. In the preparation of live vaccines, reverse genetics systems allow the direct modification of the specific regions in the genomes of negative-stranded RNA viruses concerned with attenuation; the ultimate goal is the introduction of site-specific mutations through a cDNA intermediate in order to develop strains with the requisite attenuation, antigenic and growth properties needed in a vaccine. These techniques can also be used to disarm potentially highly pathogenic viruses, such as emerging H5N1 avian influenza viruses, in order to facilitate large-scale preparation of viruses for use in inactivated vaccines under conditions of manufacturing safety. Before these vaccines become available, residual issues concerned with intellectual property rights to the technology and its application will need to be resolved.

The growth of attenuated influenza vaccine donor strains in continuous cell lines.

The growth of the Russian live attenuated influenza vaccine donor strains A/Leningrad/134/17/57, A/Leningrad/134/47/57 and B/USSR/60/69 was studied in cells of the VERO and Madin-Darby canine kidney (MDCK) lines as six-well cultures and cell factories infected at different multiplicities of infection. Yields for A/Leningrad/134/17/57 and A/Leningrad/134/47/57 were comparable in either cell line over a range of multiplicities but were about 10-fold lower than in the allantoic fluids of infected chicken embryos. For both A/Leningrad/134/47/57 and B/USSR/60/69, yields from the MDCK line were about 10-fold higher than for the VERO line. For B/USSR/60/69, yields in eggs were approximately 100-fold higher than those obtained in the MDCK line. A feature of the growth of B/USSR/60/69 was its reduced capacity to produce infectious progeny in either cell line at multiplicities of infection of 2.0 or 1.0 pfu/cell. Inhibition was due probably due to the presence of defective-interfering particles and was not detected with A/Leningrad/134/17/57 or A/Leningrad/134/47/57 in cultures of either line infected at the same multiplicities. Yields for both A/Leningrad/134/47/57 and B/USSR/60/69 in cells of the MDCK line were comparable when grown in six-well cultures or cell factories.

The role of cell culture vaccines in the control of the next influenza pandemic.

Pandemic influenza A viruses of avian origin are of particular concern and have crossed the species barrier several times in recent years, giving rise to illness and occasionally death in humans. This situation could become dramatically worse if the infectivity of avian viruses for humans were increased by reassortment between the genes of human and avian viruses. Co-infection of humans or an intermediate host with an avian strain and an existing human strain could produce new viruses of unknown pathogenicity to which the entire population would be susceptible. Inactivated vaccines against influenza have been prepared for many years using viruses grown in embryonated chicken eggs. However, the use of eggs presents difficulties when vaccine supplies need to be expanded at short notice. It seems likely that future vaccines will be prepared in high-yielding cell cultures from continuous lines that are preferably anchorage-independent. At present, only certain preparations of the Vero and Madin-Darby canine kidney cell lines, grown and maintained in serum-free medium, are acceptable to all regulatory authorities. However, this situation is likely to change with increasing need for non-pandemic and pandemic vaccines.

Route of administration is the prime determinant of IgA and IgG2a responses in the respiratory tract of mice to the cold-adapted live attenuated influenza A donor strain A/Leningrad/134/17/57.

Serum antibody and antibody secretory cell (ASC) responses to the cold-adapted (CA) live attenuated influenza A donor strain A/Leningrad/134/17/57 in BALB/c mice were determined in the lungs and mediastinal lymph nodes after administration by the intranasal, subcutaneous and intramuscular routes. Both types of response were greatest when an inoculum consisting of 10(6.5) 50% egg infectious doses (EID(50)) was administered twice intranasally at an interval of 3 weeks. Serum responses by the intramuscular route were much higher than by the subcutaneous route but, at doses of 10(6.5-7.5) EID(50), were still lower than that obtained with two doses of an intranasal inoculum of 10(6.5) EID(50). Virus-specific ASC responses for IgA and IgG2a were obtained in the lungs and mediastinal lymph nodes of mice inoculated with 10(6.5) EID(50) by the intranasal route. However, ASC responses after inoculation by either the subcutaneous or intramuscular routes were barely detectable, even at doses as high as 10(7.5) EID(50). These results confirm that intranasal administration of live vaccines induces far higher virus-specific IgA and IgG2a responses in the respiratory tract of mice than can be achieved by parenteral administration and that serum antibody levels induced by parenteral vaccination are unrelated to the respiratory ASC response.

An evaluation of the genetic stability and pathogenicity of the Russian cold-adapted influenza A donor strains A/Leningrad/134/17/57 and A/Leningrad/134/47/57 in ferrets.

The influenza A components of live attenuated vaccines used in Russia have been prepared as reassortants of the cold-adapted (ca) H2N2 viruses, A/Leningrad/134/17/57-ca (Len/17) and A/Leningrad/134/47/57-ca (Len/47), and virulent epidemic strains. The lesions responsible for attenuation within the six internal genes of each donor strain have been sequenced and described, but relatively little is known as to their stability before and after passage in susceptible hosts. In the work reported in this paper, RT-PCR restriction analysis and limited sequencing of individual genes were used to evaluate the stability of lesions in stocks of the both donor strains after passage in ferrets, which have been used widely as susceptible hosts for assessment of the virulence of influenza strains. Len/47 was shown to possess expected lesions by RT-PCR and restriction analysis. Substitution at position 1066 of the NP gene, which has been previously reported to be unique to Len/47 [Klimov et al., Virology 186 (1992) 795], was also shown to be present in all clones of Len/17. This change was confirmed by limited sequence analysis and was shown to be retained in progeny viruses isolated from the lungs and turbinates of inoculated ferrets. Two other changes in the PB2 and PB1 genes that were present in Len/47 were detected by limited sequence analysis alone. Further previously unreported minor changes were shown to be present for Len/17 and Len/47, but not both, and their significance is unknown. Limited replication of each donor strain occurred in ferrets and minimal clinical signs and histopathology were present. By contrast, the parental strain Len/57 and the recent epidemic strain A/Sydney/6/97 induced clinical signs and histopathology that were typical of influenza disease.

Field evaluations of safety and efficacy of an Australian Marek's disease vaccine.

OBJECTIVE: To demonstrate the safety and efficacy of the Marek's Disease Virus-1 vaccine (strain BH 16) from field studies in comparison with the CVI 988 Rispens vaccine currently available in Australia. STUDY DESIGN: A small field trial was carried out on nine breeder flocks and a larger trial on 21 breeder flocks. All chickens were obtained from a commercial hatchery and each was vaccinated at hatch with cell-associated Herpes Virus of Turkeys vaccine. A group of chickens vaccinated with BH 16 vaccine was placed in one shed per property and the remainder were vaccinated with the Rispens vaccine and placed in the remaining sheds. At 25, 30, 35, and 40 weeks after hatch, the field veterinarian or farm manager examined all birds dying on two consecutive days in the designated placement sheds. RESULTS: In the small trial there was a significantly lower incidence of MD in birds vaccinated with the MDV-1 vaccine compared with the Rispens vaccine (P < 0.001). In a larger trial there was no difference in the incidence of MD between the treatment groups, due possibly to a lower rate of natural challenge. Egg production results and average weekly mortality results for both groups were similar. CONCLUSION: The present study describes an attenuated type 1 MD vaccine which is at least equivalent to a vaccine derived from the CVI 988 Rispens strain in terms of safety and efficacy when used in combination with HVT vaccine.

Derivation, safety and efficacy of a Marek's disease vaccine developed from an Australian isolate of very virulent Marek's disease virus.

OBJECTIVE: To develop a serotype 1 Marek's disease (MD) vaccine from a very virulent MDV (vvMDV) pathotype and demonstrate safety and efficacy against early challenge with very virulent field strains in the presence of maternal antibody. STUDY DESIGN: Strain BH 16 was isolated and attenuated by serial cell culture passage. One of two cloned passages was selected for vaccine development following early laboratory-scale protection trials in commercial birds. Comparative protection trials were carded out on the BH 16 vaccine and on a CVI 988 Rispens vaccine using commercial and SPF chickens. Challenge viruses used were either a low passage strain BH 16 virus, the Woodlands No. 1 strain or MPF 57 strain of MDV. The BH 16 vaccine was back-passaged in SPF chickens six times and virus recovered from the final passage and the original vaccine virus were tested for safety. The immunosuppressive potential of the BH 16 and Rispens vaccines was also assessed in parallel. RESULTS: The BH 16 and Rispens vaccines induced comparable levels of protection when used as monovalent or multivalent vaccines, although protection achieved with the monovalent vaccines was lower. No gross tumour formation was evident in any birds receiving the BH 16 vaccine or bird-passaged virus, although microscopic lesions were present in 2/12 birds that received the bird-passaged virus. In tests for immunosuppression, there was no histological evidence of damage to either the bursa of Fabricius or the thymus. CONCLUSION: The BH 16 vaccine was shown to be safe and at least as protective as the Rispens vaccine against three highly virulent MD challenge viruses.

Preparation and characterisation of attenuated cold-adapted influenza A reassortants derived from the A/Leningrad/134/17/57 donor strain.

The development of a rapid cell culture method for the preparation of cold-adapted (ca) influenza A reassortant viruses is described and compared with a currently used egg method. Mixtures of the ca donor A/Leningrad/134/17/57-ca (A/Len/17) and A/Beijing/32/92 (A/Beij/32), a recent H3N2 epidemic strain, were used to co-infect chicken embryo kidney (CEK) cell cultures; reassortant progeny were selected using an infectious centre assay. The assay was capable of detecting interference where the infectivity ratio for A/Len/17 and A/Beij/32 was 1:7. Progeny viruses were characterised genetically by amplification of defined regions within each of the six internal genes by PCR and identification of the products by restriction enzyme analysis. Reassortants were also tested for ca and temperature-sensitive (ts) phenotype and the identity of the surface antigens. The infectious centre assay was shown to be an effective method for isolating reassortant progeny that possessed the haemagglutinin and neuraminidase surface antigens of A/Beij/32. Reassortants with the six internal genes derived from the donor strain and possessing the ca and ts phenotype were readily obtained when (a) an infectivity ratio of 1:49 was used and (b) two plaque-to-plaque isolations of progeny virus were made after growth at 25 degrees C and one at 34 degrees C, both in the presence of antiserum to the donor strain.

Immunogenic and isotype-specific responses to Russian and US cold-adapted influenza a vaccine donor strains A/Leningrad/134/17/57, A/Leningrad/134/47/57, and A/Ann Arbor/6/60 (H2N2) in mice.

The immunogenicity of the Russian cold-adapted (ca) donor stains, A/Leningrad (Len)/134/17/57 and A/Leningrad/134/47/57, and the US strain A/Ann Arbor (AA)/6/60-ca, were compared in BALB/c mice with their respective wild-type parental viruses. Each ca donor strain was less immunogenic than its wild-type parent. The vaccinating dose, when administered twice, which prevented multiplication of a standard challenge of parental wild-type virus in 50% of mice (the 50% protective dose or PD(50)), was shown for A/Len/134/17/57-ca, A/Len/134/47/57-ca, and A/AA/6/60-ca to be 10(3.77), 10(4.32), and 10(4.70), respectively. These findings were extended by measuring the number of antibody secreting cells induced in the lungs and mediastinal lymph nodes of mice infected with the same ca donors using an ELISPOT assay. When each donor strain was administered twice at a dose of 100 PD(50) over a 3-week interval, the overall immunoglobulin isotype antibody secreting cell profiles were shown to be similar. However, A/Len/134/17/57-ca and A/Len/134/47/57-ca induced significantly higher total immunoglobulin responses in the lungs than A/AA/6/60-ca (P < 0.05). A/Len/134/17/57-ca also induced a significantly greater IgA response in the lungs than A/AA/6/60-ca (P < 0.05). These results suggest that A/Len/134/17/57-ca is a superior immunogen to A/Len/134/47/57-ca which in turn is more immunogenic than A/AA/6/60-ca.

Live attenuated vaccines against influenza; an historical review.

Live attenuated vaccines administered directly to the respiratory tract offer the promise of providing more effective immunity against influenza than subunit or split inactivated vaccines. Evidence has accumulated in recent years that immunological responses relevant to both the prevention of and recovery from influenza are best induced by natural infection. The ease with which the genes of influenza viruses reassort when two or more viruses infect a single cell has been exploited as a means of rapidly producing attenuated vaccines. Donor strains that have been shown by extensive testing to be fully attenuated are used to co-infect cells with contemporary epidemic strains to produce reassortants with the required degree of avirulence and the surface antigens of the epidemic strain. Reassortants prepared from cold-adapted mutants of both influenza A and B viruses have been widely shown from clinical trials in both the United States and Russia over many years to be well tolerated in both adults and children and to be highly efficacious.

Adaptation of Marek's disease virus to the Vero continuous cell line.

Marek's disease virus (MDV) is a highly infectious, cell-associated oncogenic herpesvirus. Production of MD vaccines has been limited to primary chicken and duck embryo fibroblast (CEF and DEF) cultures. These have a limited life span and cannot be readily stored in liquid nitrogen. Moreover, the need to prepare CEF and DEF cells on a regular basis from 10 to 11 day-old embryos derived from a flock that must be tested continuously for the presence of avian pathogens adds to the cost of vaccine production. A continuous cell line that would support MDV replication could have significant advantages for the rapid large-scale preparation of MD vaccines. In this report, we describe the adaptation to growth of CEF-grown preparations of serotype 1 and serotype 3 (herpesvirus of turkeys; HVT) strains of MDV in cells of the Vero continuous cell line. Although both viruses produced typical CPE, higher levels of infectious progeny and more extensive virus-specific immunofluorescence were obtained for HVT than for the serotype 1 virus. PCR and pulsed field electrophoresis (PFE) analysis of the DNA from Vero cells infected with either virus confirmed the presence of virus-specific DNA.

The antigenicity of the chicken anaemia virus protein VP3 (Apoptin).

Chicken anaemia virus protein VP3 (Apoptin) was cloned and expressed as a recombinant protein and evaluated for its suitability as a serodiagnostic reagent. VP3 was expressed as a fusion protein either with glutathione S -transferase or with a six-histidine tag. Both recombinant proteins reacted specifically with anti-VP3 monoclonal antibodies and with serum from vaccinated chickens by Western blot and by enzyme-linked immunosorbent assay (ELISA). However, when testing sera from birds of different ages and genetic backgrounds, high non-specific reactions were evident and false positives were observed, especially in older birds. This suggests that VP3 is poorly immunogenic during infection and low antibody concentrations are masked by non-specific reactions. Thus, VP3 is not suitable for use as antigen in ELISAs.

The development and evaluation of two tissue culture-grown Marek's disease challenge viruses.

Standardized challenge viruses are essential for the evaluation of Marek's disease (MD) vaccines with many MD challenge preparations consisting of lymphocytes or whole blood from infected birds. Virus present in these preparations is difficult to quantify by tissue culture assays and, therefore, the infectious bird dose and long-term storage viability cannot be assured. We report on the properties of two low-passage virulent Australian MD viruses, the Woodlands No. 1 strain and strain MPF 57. Both strains were isolated in chicken embryo kidney cultures and adapted to grow in chicken embryo fibroblast cultures for a maximum of 14 passages. Both strains could be readily assayed in tissue culture and produced titres of 10(3) to 10(4) 50% tissue culture infectious doses per ml (TCID(50)). Birds inoculated at three different doses were observed over 10 weeks, and tissues examined for gross and histological lesions, bursarbody weight ratios and the presence of viraemia. Tissue culture-grown preparations of both strains were only slightly less virulent than the original lymphocyte challenge material and produced similar pathological responses and around 80% death or gross lesions. From bursarbody weight ratios strain MPF 57 appeared to be more virulent than the Woodlands No. 1 strain.

Protective responses in mice to vaccination with multiply administered cold-adapted influenza vaccine reassortants and wild-type viruses.

Protective responses to influenza vaccine reassortants derived from the cold-adapted (ca) donor strains A/Leningrad/134/17/57 and B/USSR/60/69 and wild-type epidemic viruses were studied in two strains of mice. Preliminary experiments revealed that, when mixtures of three viruses were inoculated intranasally to mice with 50 microliters containing 10(6) EID50 per 200 microliters (10(5.4) EID50 per mouse), interference between strains did not occur. However, interference with the growth of the influenza reassortant B/60/32/R took place if its concentration in the mixture was reduced to 10(5) (10(4.4) per mouse) or if it was inoculated at 10(6) EID50 (10(5.4) per mouse) in the presence of the influenza reassortant R/34 and two other influenza A epidemic strains; interference was unrelated to serological responses to infection with B/60/32/R. Despite evidence of interference, mice inoculated with the same mixtures in two identical doses, three weeks apart, were able to clear a challenge from each of seven homotypic and heterotypic influenza A and B strains. Heterotypic clearance of influenza A challenge viruses was greater following mixed infection, indicating that common determinants within the surface antigen glycoproteins contributed to immune responses which were broader than could be expected to be induced by parenteral vaccination.

Isolation of serotype 1 Marek's disease viruses from vaccinated Australian flocks.

Field samples were received for Marek's disease virus (MDV) isolation from clinically affected flocks from several regions of Eastern Australia. Lymphocytes were fractionated in Ficoll-Paque and passaged once or twice in chicken embryo kidney cultures. Serotype-specific virus was detected in infected cultures by indirect immunofluorescence with monoclonal antibodies. Serotype 1 MDV was isolated from 10 flocks. In samples from 5 of these flocks, serotype 2 and 3 vaccine viruses were isolated from the same specimen. In a parallel study, plasmas obtained during lymphocyte isolation were tested for antibodies to MDVs by agar gel precipitin (AGP) tests using serotype 1 and 3 antigen extracts. No correlation was observed between the rate of virus isolation and AGP positivity. The AGP test was incapable of discriminating between the different MDV serotypes.

The immunogenicity of reassortants of the cold-adapted influenza A master strain A/Ann Arbor/6/60 is determined by both the genes for cold-adaptation and the haemagglutinin gene.

Two surface antigen segregants were prepared by co-infection of chicken embryo kidney cell cultures with reassortants of the cold-adapted influenza A master strain A/Ann Arbor/6/60-ca (H2N2) possessing the surface antigens of A/Queensland/6/72 (H3N2) and A/Hong Kong/123/77 (H1N1) and other genes that were common to the master strain. The segregants were shown by serological tests to possess H3N1 and H1N2 surface antigens but it was not possible to determine the presence of H1 or N1 genes by single-stranded RNA polyacrylamide gel electrophoresis. The immunogenic properties of A/Queensland/6/72-ca and H3N1 segregant CR6/35/2/9 were compared by immunising mice intranasally with graded doses of each virus twice at an interval of 3 weeks and then challenging with the wild-type A/Queensland/6/72 (H3N2). Clearance of the challenge virus occurred in mice immunised with the same vaccinating dose, indicating that the immunogenicity of both the ca H3N2 and H3N1 viruses was identical and similar findings were obtained for mice immunised with A/Hong Kong/123/77-ca and the H1N2 segregant CR6/35/1/19 and challenged with the wild-type A/Hong Kong/123/77. Therefore, there appears to be a good correlation between immunogenicity and the inheritance of the haemagglutinin gene.

Avian encephalomyelitis: a review.

Avian encephalomyelitis virus (AEV) is a picornavirus with a predilection for the central nervous system and other parenchymous organs of chickens that is transmitted by the oral-faecal route. The virus may be spread by the vertical and horizontal routes and, because of its great stability, contaminated areas may remain infectious for long periods. The egg-adapted Van Roekel strain is highly neurotropic and does not grow efficiently in the enteric tract of the chicken. Despite this, the virion polypeptides of both naturally-occurring strains and the Van Roekel strain are antigenically identical, and there is only one virus serotype. Natural infection by AEV induces a multi-organ disease with similarities in its pathogenesis for chickens to that of Coxsackie B3 in neonatal and infant mice. Field isolates of AEV can be adapted to grow in chicken embryo brain cultures, but virus growth is highly cell-associated. Simple reliable antibody assays to determine the presence of recent infections or to assess vaccine efficacy were not available until comparatively recently when antibody detection assays, many involving the use of chick embryos, were replaced by enzyme-linked immunosorbent assays. Effective vaccination programmes employing live, highly enterotropic strains of AEV, have greatly reduced the incidence of the disease.

Subgrouping of respiratory syncytial virus strains from Australia and Papua New Guinea by biological and antigenic characteristics.

Strains of respiratory syncytial virus from 3 major areas of Australia and Papua New Guinea (PNG) were analyzed for variations in their antigenic and biological properties and in the molecular weights of their major structural proteins. Seventy-eight strains from infants and young children with LRI were collected from 1981-1984. The RSV season in the Australian cities lasted from April through September, with major peaks in July of each year, while the RSV season in tropical PNG was year-round, with small peaks in March and October of each year coinciding with excessive rainfall. Fifty-six strains were analyzed in detail; 40 were typed by time-resolved fluoroimmunoassay with monoclonal antibodies as group A strains and 16 were group B; both groups were concurrent. Three children of one family had sequential RSV infections 13 months apart, and the etiologic group A strain was identical both years in terms of growth and antigenic properties with strain-specific ferret antisera; the second infection was milder in all three children. On average, the group A strains replicated considerably better than group B strains in HEp2 cells, producing 53% more syncytia and 99% higher infectious virus titers in 31% less time in culture. Ten group A and B reference strains exhibited the same growth patterns as the A and B regional strains, respectively. Differences in antigenicity as measured with hyperimmune antisera to prototype Long strain were even greater. Group A strains exhibited a mean 68% greater IFA staining than B strains, a 71% greater EIA reaction, and were neutralized to 69% higher serum titers than B strains. Again, the reference A and B strains included as controls gave patterns identical to those of the regional strains. Finally, the P phosphoprotein had consistently higher molecular weight in A strains (mean 35,900) than B strains (mean 33,100). Small variations in the sizes of the F and G glycoproteins were not sufficient to suggest grouping on this basis.