No abstract

Mutations that consistently accumulated in the attenuated Sabin 2 strain of poliovirus during propagation in cell cultures were identified by sequence heterogeneity assay and quantified by mutant analysis by PCR and restriction enzyme cleavage (MAPREC). Eight additional sites previously identified in stool isolates were also examined by MAPREC in the virus passages. The pattern of selectable mutations and the rate of their accumulation depended on the type and confluence of the cell culture and the temperature of virus growth. Five unstable genomic sites were identified in Sabin 2 virus passaged 10 times at 34 degrees in African green monkey kidney (AGMK) cells, with the mutations accumulating in the range 1 to 24%. Accumulation of these mutations did not appear to result in a loss of attenuated phenotype since the virus passaged under these conditions passed the monkey neurovirulence test (MNVT). The content of the 481-G revertant known to be related to neurovirulence in monkeys did not increase. Thus, our results suggest that upon growth of Sabin 2 virus in AGMK cells at 34 degrees, the key determinant(s) of attenuation remained stable, and the mutations that occurred did not affect monkey neurovirulence. In virus passaged 10 times at 37 degrees in AGMK cells, 4 unstable genomic sites were identified, in some of them accumulating up to 12% of the mutants. This virus sample severely failed the MNVT. Virus passaged in Vero cells at 34 and 37 degrees accumulated mutants at 7 and 14 genomic sites, respectively, including 481-G in both cases, with almost complete substitution of the original nucleotides at some of the sites. We tested 44 commercial monopools of Type 2 OPV and found out that all of them contained 481-G revertants in the range 0.4-1.1%. An increase in the 481-G revertants in passaged viruses to the level of 4% and above correlated with failure of these samples by the MNVT. Since the pattern of selectable mutations differed in viruses grown in the two cell cultures used in this study, specific mutation profiles should be determined for each cell substrate used for vaccine production to assess manufacturing consistency.

Quantitation of virus revertants by PCR and restriction enzyme cleavage may give nonlinear results and, in some cases, produce artifacts caused by nucleotide misincorporation and heteroduplex formation, occurring during PCR. Modifications of the procedure allowed us to overcome these problems and develop a highly sensitive and reliable method of mutant quantitation. This procedure can be used to assess the quality of live vaccines and to study heterogeneity of viral and bacterial populations.

Replication of attenuated poliovirus strains results in their partial deattenuation. Recently we identified mutations accumulating in the Sabin 1 poliovirus in cell cultures. Here we report genetic changes occurring in this virus during replication in the central nervous system (CNS) of monkeys. Viruses isolated from different parts of the CNS of rhesus monkeys (inoculated into the spinal cord) were screened for sequence heterogeneities and newly identified mutations were independently confirmed and quantified using mutant analysis by PCR and restriction enzyme cleavage (MAPREC). All consistently accumulating mutations identified in this study ,vere located in untranslated regions: GU -~ AU or GU -~ GC substitution at a complementary pair formed by nucleotides 480 and 525, U -~ C substitution at nucleotide 612, and GU -~ AU or GU~GC substitution of a base pair formed by the nucleotides 7427/7441 immediately preceding the poly(A) tract. All these mutations except one (7427) were previously identified in cell culture passages or stool isolates from vaccinees. Sequencing of 11 CNS isolates also identified a few random silent mutations that accumulated as neutral 'passengers', passively coselected with genuinely selectable mutations present on the same RNA molecule. One isolate also contained the wild-type base at nucleotide 2741 (Ala 88 ~ Thr in VP1).Our results demonstrate a remarkable genetic stability of the Sabin 1 poliovirus in the CNS of monkeys, suggesting that deattenuation is determined by a very limited number of mutations. These mutations can be assayed by MAPREC to monitor the consistency of oral poliovirus vaccine (OPV) production. IntroductionThe high mutation rates inherent in viral RNA replication provide a practically unlimited source of variants for the selection of virus particles with the highest replicative capacity under given growth conditions. It has been demonstrated that replication of attenuated Sabin strains ofpoliovirus both in vivo and in vitro results in gradual quantitative changes in attenuation markers such as reduced replicative capacity at an increased temperature (rct40) or at a reduced concentration of sodium bicarbonate (d-marker) and more importantly, increased neurovirulence in monkeys (Benyesh-Melnick & Melnick, 1959; Sabin, 1957Sabin, , 1961. Because of these changes, every batch of oral poliovirus vaccine (OPV) must be safety-tested in monkeys before it can be released for use in humans (WHO, 1990). Monkey tests are very expensive and require sacrificing primates, so * Author for correspondence. Fax + 1 301 594 6530. e-mail CHUMAKOV@HELIX.NIH.GOV the development of alternative safety tests to directly measure the content of neurovirulent mutations in vaccine lots directly is an important objective. In the past decade significant progress has been made in identifying the molecular determinants of poliovirus attenuation and neurovirulence (see Minor et al., 1993 ; Macadam et al., 1994 for reviews). Studies with recombinants between revertant strains of poliovirus isolated...

Sabin strains of oral poliovirus vaccine (OPV) undergo limited genetic changes during replication in cell cultures, the gastrointestinal tract of vaccinees, and the central nervous system of monkeys. Some of these changes are associated with loss of attenuation markers. Here we report the dynamics of mutant accumulation in the Sabin strain of poliovirus type 3 inoculated intraspinally into monkeys. Thr --> lle reversion in amino acid 6 of VP1 (2493 C --> U) occurred within the first few days postinoculation (p.i.), but decreased on later days and completely disappeared by Day 17 p.i. 472 U --> C reversion in the 5'-untranslated region appeared to accumulate slower and by Day 17 completely substituted for the vaccine-type nucleotide at this site. These results indicate that experimental infection of the central nervous system of monkeys consists of early and late phases in which a different genetic constitution of the virus is favored. In several isolates one additional neurovirulent revertant was found: a Phe --> Ser at amino acid 91 of VP3 (2034 U --> C). Since this mutation was never detected in vaccine lots and is strongly selected against in cell cultures at temperatures below 38.5 degrees, it does not threaten the safety of OPV.