Development of a rubella virus DNA vaccine.

Two rubella virus DNA vaccines were constructed from a cDNA clone of the rubella virus genomic RNA, one which contained the coding sequences for all three virion proteins (C, E2 and E1) and one which contained the two envelope glycoproteins (E2 and E1). When used to immunize mice via gene gun delivery, both constructs induced an antibody response of equivalent titer to that induced by rubella virus that persisted for at least seven months. A booster injection given four weeks after the initial injection increased antibody titers by over thirty-fold. The antibody response in DNA vaccine-injected mice was directed primarily against the E1 glycoprotein, as was the case in rubella virus-injected mice, and neutralizing activity was detected. These DNA vaccines are thus prototypes for a nonreplicating rubella virus vaccine that could be used in specialized circumstances.

Molecular analysis of rubella virus epidemiology across three continents, North America, Europe, and Asia, 1961-1997.

E1 gene nucleotide sequences of 63 rubella virus isolates from North America, Europe, and Asia isolated between 1961 and 1997 were compared phylogenetically. Two genotypes were evident: Genotype I contained 60 viruses from North America, Europe, and Japan, and genotype II contained 3 viruses from China and India. The genotype I isolates prior to 1970 grouped into a single diffuse clade, indicating intercontinental circulation, while most post-1975 viruses segregated into geographic clades from each continent, indicating evolution in response to vaccination programs. The E1 amino acid sequences differed by no more than 3%; thus, no major antigenic variation was apparent. Among 4 viruses from congenital rubella syndrome that occurred following reinfection, only one amino acid substitution occurred in several important epitopes, indicating that antigenic drift is not important in this phenomenon. However, 2 viruses isolated from chronic arthritis exhibited changes in these epitopes. Isolates of the RA 27/3 vaccine strain were readily identifiable by nucleotide sequence.

Genomic sequence of the RA27/3 vaccine strain of rubella virus.

The sequence of the genome of the RA27/3 vaccine strain of rubella virus (RUB) was determined. In the process, several discrepancies between the previously reported genomic sequences of two wild RUB strains (Therien and M33) were resolved. The genomes of all three strains contain 9762 nucleotides (nts), exclusive of the 3' poly A tract. In all three strains, the genome contains (5' to 3'), a 40 nt 5' untranslated region (UTR), an open reading frame (ORF) of 6348 nts that encodes nonstructural proteins, a 123 nt UTR between the two genomic ORFs, a 3189 nt ORF that encodes the structural proteins, and a 62 nt 3' UTR. The 5' end of the subgenomic RNA was found to correspond to a uridine residue at nt 6436 of the genomic RNA. At the nucleotide level, the sequence of the three strains varied by 1.0 to 2.8%, while at the amino acid level, the sequence varied by 1.1 to 2.4% over both ORFs. The RA27/3 sequence will be of use in identification of the determinants of its attenuation, in vaccine production control and in development of second generation RUB vaccines based on recombinant DNA technology.

Improvement of the specific infectivity of the rubella virus (RUB) infectious clone: determinants of cytopathogenicity induced by RUB map to the nonstructural proteins.

A plasmid, Robo102, which contains a cDNA copy of the rubella virus (RUB) genomic RNA from which infectious transcripts can be synthesized in vitro, was recently developed (C. Y. Wang, G. Dominguez, and T. K. Frey, J. Virol. 68:3550-3557, 1994). To increase the specific infectivity of Robo102 transcripts (approximately 5 plaques/10 microg of transcripts), a modified reverse transcription-PCR method was used to amplify nearly 90% of the RUB genome in three fragments, which were then used to replace the corresponding fragments in Robo102. Replacement of a fragment covering nucleotides (nt) 5352 to 9759 of the RUB genome yielded a construct, Robo202, which produced highly infectious transcripts (10(4) plaques/microg), indicating the presence of an unrecognized deleterious mutation (or mutations) in this region of the Robo102 cDNA. Robo102 was based on the w-Therien strain of RUB, which forms opaque plaques in Vero cells, while the PCR replacement fragments were generated from a variant, f-Therien, which produces clear plaques in Vero cells. Although Robo202 contains over 4,000 nt from f-Therien, Robo202 virus produces opaque plaques. However, when the other two PCR fragments amplified from f-Therien (nt 1 to 1723 and nt 2800 to 5352) were introduced into Robo202, the resulting construct, Robo302, yielded transcripts that produced a virus that formed clear plaques. This indicates that the determinants of plaque morphology map to the regions of the genome covered by these two fragments, both of which are in the nonstructural open reading frame. Generation of Robo202/302 chimeras indicated that the most 5' terminal fragment (nt. 1 to 1723) had the greatest effect on plaque morphology. The plaque morphology was correlated with the ability of the viruses to kill infected cells. The only difference at the molecular level detected among the viruses was that the more cytopathic viruses produced more nonstructural proteins than did the less cytopathic viruses. This finding, as well as the mapping of the genetic determinants to the region of the genome encoding these proteins, indicates that the nonstructural proteins can mediate cell killing.

Identification of strain-specific nucleotide sequences in the RA 27/3 rubella virus vaccine.

A polymerase chain reaction-based protocol was developed for determination of the sequence of 1300 nucleotides in the E1 coding region of the genomes of multiple strains of rubella virus. From a collection of sequences of 9 independent strains and isolates, characteristic nucleotides were identified that distinguished the RA 27/3 strain, which is the currently used attenuated vaccine strain. These characteristic nucleotides were maintained in virus recovered from recent vaccinees. Both the assay and the knowledge of these characteristic nucleotides should be of use in determining the origin of virus isolated from individuals who have suspected vaccine-associated complications. The nucleotide sequences of the independent rubella virus strains differed by 0.7%-3.6%. Phylogenetic tree analysis of the sequences indicated the existence of at least three distinct genetic lineages.

Effect of antiviral antibody on maintenance of long-term rubella virus persistent infection in Vero cells.

A Vero cell line with a long-term rubella virus persistent infection was maintained for 45 weeks in the presence of anti-rubella virus antibody of sufficient titer to completely neutralize the virus in the culture fluid to determine the effect of the presence of antibody on the maintenance of the persistent infection. Prior to antibody treatment, virus was continuously detected as plaque-forming units in the persistently infected culture fluid. Virus clones that were plaque purified from the persistently infected culture fluid were temperature sensitive and exhibited a reduced efficiency of replication and ability to induce cytopathic effects in Vero cells at the persistently infected culture temperature compared with the standard virus used to initiate the persistently infected culture. Defective interfering RNAs were the major intracellular virus-specific RNA species present in the persistently infected cells. Treatment with antibody failed to cure the persistently infected culture of virus, and the cells retained the ability to release virus after antibody treatment was discontinued. Interestingly, the presence of antibody led to the selection of a population of virus which was markedly less cytopathic for Vero cells than the virus population which was selected during persistent infection in the absence of antibody.

Time course of virus-specific macromolecular synthesis during rubella virus infection in Vero cells.

Virus specific macromolecular synthesis was studied in Vero cells infected with plaque-purified rubella virus under one-step multiplication conditions. Under these conditions, the rate of virus production was found to increase rapidly until 24 hr postinfection after which time the rate of virus production rose more slowly, reaching a peak level at 48 hr postinfection. This peak rate of virus production was maintained through 72 hr postinfection. A majority of the cells remained alive through 96 hr postinfection, although a 20 to 30% decrease in the number of living cells occurred between 24 and 48 hr postinfection, the time period at which cytopathic effect was first observed. The virus structural proteins were first detected intracellularly at 16 hr postinfection. The rate of synthesis of these proteins was already maximal at 16 hr postinfection and remained constant through 48 hr postinfection. By immunofluorescence, cells expressing virus proteins were first observed at 12 hr postinfection. At 24 hr postinfection, 35 to 50% of the cells in the infected culture were exhibiting immunofluorescence, at 36 hr postinfection, 65 to 90% of the cells were exhibiting immunofluorescence, and at 48 hr postinfection, all of the cells were exhibiting immunofluorescence. The virus genomic and subgenomic RNA species were first detectable by 12 hr postinfection. The rate of synthesis of both of these species peaked at 26 hr postinfection. Rubella virus infection was found to have no effect on total cell RNA synthesis. However, a modest inhibition of total cell protein synthesis which reached 40% by 48 hr postinfection was observed. When Northern analysis of RNA extracted from infected cells was performed, a negative-polarity, virus-specific RNA probe hybridized only to the virus genomic and subgenomic RNA species. A positive-polarity, virus-specific RNA probe hybridized predominantly to a negative-polarity RNA of genome length indicating that both the genomic and subgenomic RNAs are synthesized from a genome-length negative-polarity template. Defective interfering (DI) RNAs were not detected in infected cells through 96 hr postinfection or in cells onto which virus released through 96 hr postinfection was passaged. Thus, the generation of DI particles by rubella virus appears to play no role in the slow, noncytopathic replication of this virus or in the ability of rubella virus-infected cells to survive for extended periods of time.