NUCLEOTIDE-SEQUENCE VARIATION OF THE ENVELOPE PROTEIN GENE IDENTIFIES2 DISTINCT GENOTYPES OF YELLOW-FEVER VIRUS

The evolution of yellow fever virus over 67 years was investigated by comparing the nucleotide sequences of the envelope (E) protein genes o f 20 viruses isolated in Africa, the Caribbean, and South America. Uni formly weighted parsimony algorithm analysis defined two major evoluti onary yellow fever virus lineages designated E genotypes I and II. E g enotype I contained viruses isolated from East and Central Africa. E g enotype II viruses were divided into two sublineages: IIA viruses from West Africa and IIB viruses from America, except for a 1979 virus iso lated from Trinidad (TRINID79A). Unique signature patterns were identi fied at 111 nucleotide and 12 amino acid positions within the yellow f ever virus E gene by signature pattern analysis. Yellow fever viruses from East and Central Africa contained unique signatures at 60 nucleot ide and five amino acid positions, those from West Africa contained un ique signatures at 25 nucleotide and two amino acid positions, and vir uses from America contained such signatures at 30 nucleotide and five amino acid positions in the E gene. The dissemination of yellow fever viruses from Africa to the Americas is supported by the close genetic relatedness of genotype IIA and LLB viruses and genetic evidence of a possible second introduction of yellow fever virus from West Africa, a s illustrated by the TRINID79A virus isolate. The E protein genes of A merican IIB yellow fever viruses had higher frequencies of amino acid substitutions than did genes of yellow fever viruses of genotypes I an d IIA on the basis of comparisons with a consensus amino acid sequence for the yellow fever E gene. The great variation in the E proteins of American yellow fever virus probably results from positive selection imposed by virus interaction with different species of mosquitoes or n onhuman primates in the Americas.