Mutations in Sabin 2 strain of poliovirus and stability of attenuation phenotype.

In this study, we attempted to identify the molecular determinants in the genome of the attenuated Sabin 2 vaccine strain of poliovirus that may change during vaccine production and result in an increase in monkey neurovirulence. An extensive search for suitable vaccine lots identified six batches that had failed the monkey neurovirulence test (MNVT). On repeated tests, these batches were found to have acceptable levels of monkey neurovirulence. One of the batches was additionally passaged six times under conditions used in vaccine production, and the resulting high-passage sample was screened for the presence of mutations and tested in monkeys. In addition to the previously described A --> G reversion at nucleotide 481, high-passage stock also contained a mutation in the VP1-coding region (3364 = G --> A) that consistently accumulated in the course of passaging. However, despite the presence of substantial amounts of these mutations, high-passage stock passed the MNVT. Replication of Sabin 2 poliovirus in the central nervous system of transgenic mice susceptible to poliovirus or in cultures of mouse cells, resulted in another mutation (3363 = A --> G). Even though its presence correlated with paralysis in mice, the introduction of 3363-G into the Sabin 2 genome did not increase neurovirulence of the virus. Previous studies identified the 481-G mutation as an important determinant of monkey neurovirulence. We prepared virus samples with varying amounts of genetically defined single mutants at this nucleotide and tested them in monkeys. The results demonstrated that even a 100% substitution at this site introduced into Sabin 2 strain did not increase monkey neurovirulence. The determination of the nucleotide sequence of an alternative strain used for the production of type 2 OPV (Chung 2) showed that it contained 100% of the wild-type 481-G but possessed an extremely low level of neurovirulence. These results demonstrate the remarkable stability of the attenuated phenotype of the Sabin 2 strain and show that (1) no batch of OPV 2 has ever repeatedly failed the MNVT, (2) growing the virus beyond the passage level allowed in vaccine production did not result in increased neurovirulence in monkeys, (3) a test for neurovirulence in transgenic mice may be more sensitive than the MNVT, and (4) determination of the mutational profile of vaccine batches detects inconsistencies in vaccine manufacturing processing that would not be detected by the MNVT. Copyright 1999 Academic Press.