Serial passage of initially nonlethal Ebola virus (EBOV) in outbred guinea
pigs resulted in the selection of variants with high pathogenicity. Nucleot
ide sequence analysis of the complete genome of the guinea pig-adapted vari
ant 8mc revealed that it differed from wild-type virus by eight mutations.
No mutations were identified in nontranscribed regions, including leader, t
railer, and intragenic sequences. Among noncoding regions the only base cha
nge was found in the VP30 gene. Two silent base changes were found in the o
pen reading frame (ORF) encoding NP protein. Nucleotide changes resulting i
n single-amino-acid exchanges were identified in both NP and L genes. Three
other mutations found in VP24 caused amino acid substitutions, which are r
esponsible for larger structural changes of this protein, as indicated by a
n alteration in electrophoretic mobility. A highly pathogenic EBOV variant
K5 from another passaging series showed an amino acid substitution at nearl
y the same location in the VP24 gene, suggesting the importance of this pro
tein in the adaptation process. In addition, sequence variability of the GP
gene was found when plaque-purified clones of EBOV-8mc were analyzed. Thre
e of five viral clones showed insertion of one uridine residue at the GP ge
ne-editing site, which led to a significant change in the expression of vir
us glycoproteins. This observation suggests that the editing site is a hot
spot for insertion and deletion of nucleotides. not only at the level of tr
anscription but also of genome replication. Irrespective of the number of u
ridine residues at the editing site, all plaque-purified clones of EBOV var
iant 8mc resembled each other in their pathogenicity for guinea pigs, indic
ating either the absence or only supportive role of mutations in the GP gen
e on the adaptation process, (C) 2000 Academic Press.