Genetic relationships and evolution of genotypes of yellow fever virus and other members of the yellow fever virus group within the Flavivirus genus based on the 3' noncoding region

Abstract

Genetic relationships among flaviviruses within the yellow fever (YF) virus genetic group were investigated by comparing nucleotide sequences of the 3' noncoding region (3'NCR). Size heterogeneity was observed between members and even among strains of the same viral species. Size variation between YF strains was due to duplications and/or deletions of repeated nucleotide sequence elements (RYF). West African genotypes had three copies of the RYF (RYF1, RYF2, and RYF3); the Angola and the East and Central African genotypes had two copies (RYF1 and RYF3); and South American genotypes had only a single copy (RYF3). Nucleotide sequence analyses suggest a deletion within the 3'NCR of South American genotypes, including RYF1 and RYF2. Based on studies with the French neurotropic vaccine strain, passage of a YF virus strain in cell culture can result in deletion of RYF1 and RYF2. Taken together, these observations suggest that South American genotypes of YF virus evolved from West African genotypes and that the South American genotypes lost RYF1 and RYF2, possibly in a single event. Repeated sequence elements were found within the 3'NCR of other members of the YF virus genetic group, suggesting that it is probably characteristic for members of the YF virus genetic group. A core sequence of 15 nucleotides, containing two stem-loops, was found within the 3'NCR of all members of the YF genetic group and may represent the progenitor repeat sequence. Secondary structure predictions of the 3'NCR showed very similar structures for viruses that were closely related phylogenetically.

FIG. 1.

Alignment of nucleotide sequence for the 3′NCR of YF virus strains showing structural regions for the complete 3′NCR. (A) Predicted secondary structure for the entire 3′NCR of YF virus. (B) Nucleotide sequence and identity of sequence elements in the predicted secondary structure shown in panel A. The bottom line indicates the consensus sequence for all strains. Regions involved in base pairing are boxed, and the letters above the boxes mark specific stem regions. Regions involved in formation of pseudoknot structures are marked by hatched boxes. The cyclization signal is marked by a grey box. The RYF repeat regions are marked as RYF1, RYF2, and RYF3. Boxes A and B show pentanucleotide sequences in Brazil35, which are very similar to the sequences at the 5′ end of RYF1 and the 3′ end of RYF2, respectively. Box C highlights nucleotide sequence similarity in the 3′-flanking region of RYF3 between West African and South American genotypes. Box D highlights nucleotide sequence similarity in the 3′ flanking region of RYF3 among East and Central African genotypes. A dash indicates spacing introduced to optimize alignment, while a dot indicates nucleotides identical with the Ghana27 sequence.

FIG. 1.

Alignment of nucleotide sequence for the 3′NCR of YF virus strains showing structural regions for the complete 3′NCR. (A) Predicted secondary structure for the entire 3′NCR of YF virus. (B) Nucleotide sequence and identity of sequence elements in the predicted secondary structure shown in panel A. The bottom line indicates the consensus sequence for all strains. Regions involved in base pairing are boxed, and the letters above the boxes mark specific stem regions. Regions involved in formation of pseudoknot structures are marked by hatched boxes. The cyclization signal is marked by a grey box. The RYF repeat regions are marked as RYF1, RYF2, and RYF3. Boxes A and B show pentanucleotide sequences in Brazil35, which are very similar to the sequences at the 5′ end of RYF1 and the 3′ end of RYF2, respectively. Box C highlights nucleotide sequence similarity in the 3′-flanking region of RYF3 between West African and South American genotypes. Box D highlights nucleotide sequence similarity in the 3′ flanking region of RYF3 among East and Central African genotypes. A dash indicates spacing introduced to optimize alignment, while a dot indicates nucleotides identical with the Ghana27 sequence.

FIG. 1.

Alignment of nucleotide sequence for the 3′NCR of YF virus strains showing structural regions for the complete 3′NCR. (A) Predicted secondary structure for the entire 3′NCR of YF virus. (B) Nucleotide sequence and identity of sequence elements in the predicted secondary structure shown in panel A. The bottom line indicates the consensus sequence for all strains. Regions involved in base pairing are boxed, and the letters above the boxes mark specific stem regions. Regions involved in formation of pseudoknot structures are marked by hatched boxes. The cyclization signal is marked by a grey box. The RYF repeat regions are marked as RYF1, RYF2, and RYF3. Boxes A and B show pentanucleotide sequences in Brazil35, which are very similar to the sequences at the 5′ end of RYF1 and the 3′ end of RYF2, respectively. Box C highlights nucleotide sequence similarity in the 3′-flanking region of RYF3 between West African and South American genotypes. Box D highlights nucleotide sequence similarity in the 3′ flanking region of RYF3 among East and Central African genotypes. A dash indicates spacing introduced to optimize alignment, while a dot indicates nucleotides identical with the Ghana27 sequence.

FIG. 1.

Alignment of nucleotide sequence for the 3′NCR of YF virus strains showing structural regions for the complete 3′NCR. (A) Predicted secondary structure for the entire 3′NCR of YF virus. (B) Nucleotide sequence and identity of sequence elements in the predicted secondary structure shown in panel A. The bottom line indicates the consensus sequence for all strains. Regions involved in base pairing are boxed, and the letters above the boxes mark specific stem regions. Regions involved in formation of pseudoknot structures are marked by hatched boxes. The cyclization signal is marked by a grey box. The RYF repeat regions are marked as RYF1, RYF2, and RYF3. Boxes A and B show pentanucleotide sequences in Brazil35, which are very similar to the sequences at the 5′ end of RYF1 and the 3′ end of RYF2, respectively. Box C highlights nucleotide sequence similarity in the 3′-flanking region of RYF3 between West African and South American genotypes. Box D highlights nucleotide sequence similarity in the 3′ flanking region of RYF3 among East and Central African genotypes. A dash indicates spacing introduced to optimize alignment, while a dot indicates nucleotides identical with the Ghana27 sequence.

FIG. 2.

Predicted secondary structure for the 3′NCR of Ghana27 virus. The folding for stem-loops LSH, E, and F1 is based on that of Proutski et al. (28). The terminal base pairing has been modified based on our sequence comparison. Stem-loops B, C, and D are based on the model proposed by Olsthoorn and Bol (23). The location of RYF1, RYF2, and RYF3 repeat elements is indicated, and the core elements are highlighted in bold print. The repeated stem-loop structures present in the RYF repeats are indicated as F1, F2, F3, G1, G2, and G3. Hatched boxes mark nucleotides involved in pseudoknot formation (pk1 to pk3). Arrows indicate the locations of positions of nucleotide variation. Nucleotides in solid boxes mark changes that leave stem formation unchanged. Dotted boxes indicate nucleotide changes that are expected to disrupt base pairing. Nucleotide insertions are marked by open arrows. The cyclization element is indicated as a shaded box. The locations of the conserved CS2 sequences are marked by gray lines. AR1, A-rich region 1 (28).

FIG. 3.

Nucleotide sequence alignment for 3′NCR of four strains of FNV and the parent FVV (or Senegal27). For further details, see the legend for Fig. 2.

FIG. 4.

Alignment of nucleotide sequences for the 3′NCR of YF (Ghana27), BAN, SEP, and UGS viruses, showing the nucleotide sequence and identity of sequence elements in the predicted secondary structure. Regions involved in base paring are boxed, and the letters above the boxes mark specific stem regions. For further details, see the legend for Fig. 2.

FIG. 5.

A neighbor-joining tree generated from the 3′NCR nucleotide sequences of 13 strains of flaviviruses. The numbers at the nodes are bootstrap values for 1,000 replicates.

FIG. 6.

Predicted stem-loop structure for the repeats and flanking nucleotide sequences in the 3′NCR of flaviviruses in the YF virus genetic group. A dash indicates a missing nucleotide.

FIG. 7.

Predicted RNA structures for the 3′NCR of members of the YF virus subgroup. The predicted structures based on computer-based folding and phylogenetic analysis for the 3′NCR upstream of the CS2 sequence for SEP(A), BAN (B), and UGS (C) viruses are shown. The repeat regions for each virus are indicated. The nomenclature of the RNA structures follows that used for YF virus.