Molecular consistency monitoring of live viral vaccines ♦ 823

Abstract

The replication of viral RNA is error-prone, resulting in a high degree of genetic diversity within natural populations of viruses. Live viral vaccines may undergo significant genetic changes during passage in vitro and after replication in the bodies of vaccinees, often resulting in some loss of attenuation. A classic example of the plasticity of viral RNA genomes can be seen with Sabin's attenuated strains of Oral Poliovirus Vaccine (OPV), which partially revert to neurovirulence for monkeys when serially passaged in cell cultures-especially at elevated temperatures and in the intestinal tracts of vaccine recipients. This genetic instability dictates that production conditions for OPV must be strictly monitored to assure that the vaccine maintains the attenuating phenotype. Thus every batch of OPV is required to undergo rigorous consistency testing in monkeys. Recently we demonstrated that quantification of the amount of a neurovirulent mutation in batches of type-3 OPV accurately predicted the results of the monkey neurovirulence test (MNVT), and that this can be used to monitor consistency of type-3 OPV production. In this communication we summarize recent efforts to predict the neurovirulence of type-1 and type-2 OPV by quantifying selected mutations. Monitoring of two mutations in complementary nucleotides 480 and 525 of the 5′-non coding region of type-1 OPV detected breaks in consistency of vaccine production with much greater sensitivity than did the MNVT. We also demonstrated that the attenuating phenotype of type-1 OPV is much more stable than those of the other two types of OPV. This higher stability results from the relatively weak contribution of 480-G->A and 525-U >C mutations to neurovirulence, and the virtual absence of any other genetically unstable sites. For type-2 OPV, an A->G mutation at nucleotide 481 plays a major role in reversion to virulence, but its virulence is modulated by the presence of two additional mutations (3363-A->G and 3364-G->A) in the gene coding the VP1 capsid protein. However, genetically engineered viral constructs containing these newly identified mutations without 481-A >G reversion had only low levels of neurovirulence, suggesting that they play only an accessory role in the neurovirulence of type-2 poliovirus vaccine strains. We propose that the monitoring of molecular consistency of other live attenuated viral vaccines quantifying both mutant nucleotides of known virulence as well as other unstable nucleotides that tend to accumulate during production may provide a useful test to control vaccine quality.