| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
YFV, the prototype of this group of related viruses (hereafter called flaviviruses) for the genus Flavivirus,1 occupies an unparalleled position in the history of medical virology. It is the first animal virus scientifically identified to be etiologic agent of human disease during the celebrated study of yellow fever almost a century ago; the first virus confirmed to be transmitted by arthropod vectors; and, retrospectively, as the first hemorrhagic fever-causing virus before the term viral "hemorrhagic fever" came into general use in early 1950s.2,3 Interestingly, in the subsequent classification based on various serologic techniques, YFV was not affiliated with any member of flaviviruses, whereas most other members could be grouped together in antigenic complexes, such as JEV and TBEV complexes.4 For this reason, it was unexpected when the first comprehensive phylogenetic study revealed a close relationship of YFV with Sepik virus (SEPV) of Papau New Guinea, Entebbe bat virus (ENTV) of Africa, Sokuluk virus (SOKV) of Central Asia, and Yokose virus (YOKV) of Far East Asia.5 Later, Wesselsbron virus (WESSV) was also found to be affiliated with YFV. In addition, the aforementioned study also revealed that the major branch that diverges closest from the root of the mosquito-borne group in the phylogenetic tree was composed of three minor branches: a branch containing Entebbe bat virus (ENTV), Sokuluk virus (SOKV), and Yokose virus (YOKV); a branch containing YFV and Sepik virus (SEPV); and a more distantly related branch containing several African viruses (Bouboui virus, Uganda S virus, Banzi virus, Saboya virus, and Potiskum virus) and Austro-Asian viruses (Edge Hill and Jugra viruses). As defined by the current International Committee on Taxonomy of Viruses (ICTV) classification, the viruses in all three minor branches, as well as Wesselsbron virus, are classified as members of YFV group.1
Because the previous phylogenetic study was based on 1-kb segments of the NS5 gene encompassing < 10% of the whole viral genome, it was uncertain if the closer relationship of YFV with those unexpected viruses was biased because of that segment of gene selected for the study. In this study, we focused on the first two minor branches of viruses that had been previously found to be genetically most closely related to YFV.5 Hereafter, these viruses in the first two minor branches are called the members of "YFV cluster" to distinguish them from the members in the aforementioned third minor branch. In this study, we obtained a full genome sequence of SEPV and a near complete sequence of ENTV. Together with the full-genome sequencing of YOKV reported by others,6 we characterized genomic traits of those viruses in comparison with YFV. The data obtained in this study will be useful for discussing not only the evolution of the viruses in the YFV cluster but also for speculating on the history of dispersal of the mosquito-borne flaviviruses in the world in general.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
RT-PCR and sequencing. Viral RNA was extracted directly from infected mouse brain suspension with the QIAmp Viral RNA Mini Kit (Qiagen, Valencia, CA). cDNA was prepared by first incubating 14 µL viral RNA and 1 µL reverse primer and rapidly cooling on ice. For sequencing the genomic region between the 5'-end of genome and conserved sequence (CS2) in the 3'-noncoding region (3'-NCR), primer (VD8)7 was used for cDNA synthesis. A reverse transcription (RT) mixture containing 5 µL RT buffer (5x), 1 µL deoxy-nucleotide mix (dNTPs; 10 mmol/L of each base), 6 units of a reverse transcriptase (RAV-2; Amersham-Pharmacia Bio-tech, Piscataway, NJ), 1 µL reverse primer (100 µmol/L), and water for a total volume of 35 µL was added to 15 µL of heat-treated viral RNA, and the mixture was incubated at 45°C for 45 minutes. Polymerase chain reaction (PCR) was performed using Expand Long Template PCR System kit (Roche Applied Science, Indianapolis, IN). An aliquot (12 µL) cDNA was mixed with 88 µL of reaction mixture containing 10 µL PCR buffer (10x), 2 µL dNTP mix (10 mmol/L of each base), 1 µL each of forward and reverse primers (50 µmol/L), 2 µL Expand Long DNA polymerase, and 72 µL water. The thermocycling program set up in Gene Amp PCR System 9600 thermocycler (Perkin-Elmer, Norwalk, CT) was 1 cycle of 94°C-1 minute/50°C-1 minute/68°C-5 minutes; 3 cycles of 94°C-20 seconds/50°C-1 minute/68°C-4 minute; 10 cycles of 94°C-20 seconds/50°C-30 seconds/68°C-4 minutes, with an increment of 20 seconds per cycle; and 1 cycle of extension at 68°C for 7 minutes. Most of the primers used were designed primarily based on the conserved amino acid motifs among mosquito-borne flaviviruses.8 Amplicons were purified with a Centricep column (Princeton Separations, Adelphia, NJ), and aliquots of ~60–160 ng of the purified DNA templates were used for direct cycle sequencing using PRISM DNA sequencing kit (Big Dye) for dye terminator cycle sequencing with Ampli-Taq FS enzyme (ABI, Foster City, CA), as described previously5 and CEQ 8000 Genetic Analysis System (Beckman Coulter, Fullerton, CA).
5'-prime and 3-prime ends were amplified using 5'-RACE and 3'-RACE kits (Life Technologies, Rockville, MD), respectively. For 3'-RACE, in addition to Adoptor Primer (5'-GGCCACGCGTCGACTAC[T17]-3') supplied in the kit, the following reverse primers were also used: 5'-GCATGCG-GCCGC[T18]AGT-3'; 5'-GCATGCGGCCGC[T18]AGA-3'; 5'- GCATGCGGCCGC [T18]AGC-3'; 5'- GCATGCGGCC-GC[T18]AG-3'. Complete genomes were sequenced in both directions by primer walking, using pairs of primers selected from various regions of flavivirus genome. The sequences of all primers used for amplification and/or sequencing in this study will be provided from the authors on request.
Sequence alignment. The accession numbers of the full-length genome sequences deposited at GenBank are as follows: YFV (X03700); ENTV (DQ837641, including the ORF sequence AY632537); SEPV (DQ837642, including the ORF sequence AY632543); and YOKV (NC005039). The open reading frames (ORFs) of those viruses were aligned first by Clustal X9 followed by manual adjustment with BioEdit (version 5.0.0).10 We applied the "ReGap DNA project" function in the GeneDoc program11 to generate a properly aligned nucleotide sequence file.
Cleavage site determination. Most cleavage sites were identified by following the proteolytic processing scheme for the flavivirus ORFs previously established by Rice and Strauss.12 Junctions of intracellular capsid and premembrane (Ci/prM), premembrane and envelope (prM/E), and envelope and non-structural protein 1 (E/NS1) proteins processed by the host cellular signalase were determined on the basis of the highest cleavage potential score using a computer program SignalP-NN (http://www.cbs.dtu.dk/services/).13
Secondary structure in 3'-NCR and genome cyclization. The secondary structure in the 3'-NCR and cyclization between 3'- and 5'-terminal regions of SEPV were studied using the M-fold program,14 based on the first 200 nucleotides in the 5'-terminal sequence and the entire length of 3'-terminal sequence after the stop codon of the NS5 gene, similar to the study by Khromykh and others.15
Bootscanning of the ORF. As one of the two methods used in this study for analyzing the phylogenetic relationship among the viruses, the Bootscan program in SimPlot program package was used to examine sequence relatedness of the YFV cluster viruses across ORF.16,17 ORFs of selected viruses were scanned against a query virus with a window size of 600 nt to obtain 100 phylogenetic tree replicas using SEQBOOT and DNAPARS of PHYLIP program.18 This procedure was repeated by sliding 10 nt per step for the entire ORF; the bootstrap supports, indicating phylogenetic relatedness among reference viruses and query virus, were tabulated and plotted for each steps. The degree of bootstrap support is expressed on the y-axis as percent permuted trees across the entire length of ORF on the x-axis. The ORF sequences used in this study are the same as those used in our previous study.19
Bayesian inference. Phylogenetic relationship was also studied by Bayesian inference. Using MrBayes program (version 3.1),20 we implemented Markov chain Monte Carlo (MCMC) convergence acceleration of Metropolis coupling technique and adopted the general time reversible model to infer the phylogram that maximizes the posterior probabilities of the branch relationship. As in bootscanning, the 34 ORF sequences analyzed were the same as those used in our previous study.19 The sequences of two insect flaviviruses (CFAV and KRV) were used as outgroups. The majority consensus tree, including branch lengths and posterior clade probability, was calculated from one million MCMC simulations. The phylogenetic tree was produced using Tree view program (version 3.1).21
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) or YFV (3,411 amino acids). Repeated attempts failed to resolve sequence inconsistencies of the 3'-terminal sequence in the 3'-NCR of ENTV experienced in this study. However, we are able to report a near complete sequence (10,510 nucleotides) for ENTV, down to 11 nucleotides downstream from the end of the CS1 in the 3'-NCR. The ORF of ENTV is the same length as the coding region of YFV and encodes 3,411 amino acids. The breakdown of the lengths of genes and genomic regions among the three viruses (Table 1) reveals that the longer ORF length of YOKV is mainly attributed to its longer capsid gene size, because the lengths of other genes are similar. In terms of amino acid identity, SEPV is closest to YFV in all genes, with the only exception of 2K (Table 1).
|
. The amino-termini (N termini) of all sites expected to be cleaved by viral serine protease (virion capsid/intracellular capsid [Cv/Ci], NS2A/NS2B, NS2B/NS3, NS3/NS4A, NS4A/2K, and NS4B/NS5) and host-encoded furin-like protease (pr/M) generally follow C-terminal two basic amino acids (most typically KR, RR, or QR). In the sites cleaved by host signalase, the C-terminal and N-terminal amino acids immediately flanking M/E, E/NS1, and 2K/NS4B are generally similar. The NS1/NS2A site, which is believed to be cleaved by an unknown cellular enzyme, generally follows the sequence V-X-A (in which X is variable) defined by Rice and Strauss.12
|
5'- and 3'-NCRs.
The tandem repeat sequences and CSs in the 3'-NCR of SEPV and ENTV were studied primarily using the nomenclature originally reported by Hahn and others.22 The 15-nucleotide sequence (5'-AACCGGGATA[T/A/C]AAAC-3') in the middle of the tandem repeats of YFV (RYFs), which is also shared by other members of the YFV group, was later designated "core sequence" by Mutebi and others.23 The organization of the secondary structures was studied using the reference system developed by Proutski and others.24 When 3'-NCR is followed in the 5' to 3' direction, SEPV has two tandem repeats (RSEP1, RSEP2) comparable with those of RYF shown earlier,25 and a closer examination revealed the third imperfect or vestigial repeat (RSEP3) preceding CS2 (Figure 1; Table 3). As shown in Figure 1, like the 3'-terminal sequence that assumes a long stable hairpin (3'-LSH), CS2 and repeat sequences, as well as 5'-NCR, are also involved in secondary structure. Although the M-fold program generates energetic-favored multiple folding patterns of secondary structure, cyclization between complementary sequences (3'-CYC and 5'-CYC) apparently restricts the number of possible folding patterns more in the 3'-NCR sequence downstream than upstream of the 3'-CYC (Figure 1). Unlike that of SEPV, the 3'-NCR organization of ENTV has only one imperfect repeat YFV sequence (ImRYF) followed by CS2 and CS1. The 3'-NCR organization of YOKV is the simplest among the four viruses in the YFV cluster, because it has only CS2 and CS1.
|
|
). The third sequence (RSEP3) is most different but can be considered a highly degenerate sequence of RYF. The only sequence in ENTV that has a remote similarity to RYF is a series of noncontiguous segments that collectively resembles that of RYF (Table 3). At the far end of the spectrum of the variation in the YFV cluster is YOKV, which has no similar sequence motif.
Bootscanning of ORFs.
In this analysis, sequences of a group of selected viruses were scanned against the query sequence of any one of the four members of YFV cluster across the entire length of ORF for detecting genomic relatedness. Because the data are voluminous, only the relevant data useful for essential conclusion are presented here. As shown in Figure 2, when YFV is selected as query, the only virus with a significant (> 80% permuted trees) genetic relatedness in multiple genes is SEPV; the moderate (50–65%) range of relatedness shown by ENTV in 5'-terminal segment of the E gene is not considered very strong . When the SEPV sequence is used as query, significant levels of genetic relatedness in multiple genes are shown only by YFV (data not shown). When ENTV is used as query, only moderate levels of genetic relatedness (60–80%) in multiple genes are shown by YFV (data not shown).
|
) shows the tree positions of the four YFV cluster viruses located in the lowest major branch within the mosquito-borne group. As shown in Figure 3, this group of viruses diverged into two minor branches, one containing YFV and SEPV and the other containing ENTV and YOKV.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). However, the repeat sequence could have been lost much earlier for the members in other minor branch. Thus, ENTV retained only a vestigial signal, whereas YOKV lost it entirely. On the other hand, if it was progressive gain, the patterns in ENTV and YOKV would be considered more ancestral than those in YFV or SEPV.
Biologic traits are also useful for studying the evolutionary trend among the members of the YFV cluster. In terms of the extent of epidemic transmission, YFV is an actively proliferating virus in two continents, causing repeated outbreaks of disease. In contrast, although SEPV has been occasionally isolated from mosquitoes in the past four decades, this virus has shown thus far little disease-causing potential and has remained only in Papua New Guinea. Regarding the viruses in other minor branch, neither ENTV nor YOKV has been isolated for nearly half a century. Furthermore, unlike mosquito-borne YFV and SEPV, ENTV, YOKV, and Sokuluk virus (SOKV) have been isolated only from bats; no arthropod vector has ever been found from them. Interestingly, in contrast to the truly vertebrate viruses without a vector that cannot replicate in vector cells (no-vector group), these three bat-associated viruses in the YFV cluster still replicate in mosquito cells.27 This host range specificity is compatible with the grouping of those bat-associated viruses within the mosquito-borne group in the phylogram (Figure 3). Two possibilities have been raised in the past regarding the inability to identify vectors for those bat-associated viruses in the mosquito-borne group: a bias caused by incomplete field study to find the real vector and regressive secondary loss of vector association of formerly mosquito-borne viruses.5
Historically, interest in the geographic origin of YFV in the context of YF spread to the New World has been strong in tropical medicine. African origin of YFV has also been strongly supported by a theory that the domestication of the initially sylvatic Aedes aegypti, leading to the evolution of urban YFV vector originally occurred in Africa.28 More recently, multiple phylogenetic studies of flaviviruses also confirmed the tree position of YFV in a branch closest to the root of the mosquito-borne group. These reports, in turn, raised an interest in speculating the global dispersal of the mosquito-borne flaviviruses originating from Africa. However, this center of origin concept has not been always accepted by all. For example, in a study of the Old World arenaviruses, despite the fact that all members, with the exception of globally distributed lymphocytic choriomeningitis virus, are found only in Africa, the origin of the group was concluded to be elsewhere.29 Similarly, in another analysis of phylogenetic data, it was concluded that the current geographic range of a species was not a reliable indicator of the historical geographic range of the ancestors of the species.30
In one of the pioneering studies to explain the unique pattern of global dispersal of related flaviviruses, Sabin31 speculated on multiple geographic foci involved in dispersal that had developed after the evolution of flaviviruses of different traits from "stem virus" in the unidentified primary focus. His speculation was based on the global dispersal patterns of Aedes-borne viscerotropic and Culex-borne neurotropic flaviviruses. Thus, the current geographic distribution of flaviviruses was explained as the result of further dispersal from the secondary foci. He also raised a possibility of the dispersal of a Culex-borne JEV complex virus into the New World, ultimately evolving to become a genetically distinct virus (SLEV) that still shares neurotropism and Culex-borne mode of transmission.
In considering the geographic origin of the YFV cluster, the complexity of the available information needs to be overcome. The extant members of the yellow fever cluster are found in New Guinea (SEPV), Asia (YOKV, SOKV), and Africa (YFV, ENTV). In the third minor branch of the YFV group, Jugra and Edge Hill viruses are found only in Asia and Australia, respectively, whereas others are found in Africa. To make the analysis more complicated, Wesselsbron virus that is prevalent in Africa has been also isolated from Southeast Asia.32 Thus, it is prudent to take this information into consideration as well as the aforementioned theories on flaviviral dispersal when the geographic origin and subsequent dispersal of the members of the YFV cluster or the mosquito-borne flaviviruses is debated.
Received July 10, 2006. Accepted for publication August 21, 2006.
Acknowledgment We thank S. Vander Vliet for technical assistance.
* Address correspondence to Goro Kuno, PO Box 2087, Fort Collins, CO 80522-2087. E-mail: gok1{at}cdc.gov 
Authors’ address: Goro Kuno and Gwong-Jen J. Chang, PO Box 2087, Fort Collins, CO 80522-2087, Telephone: 970-221-6431, Fax: 970 266-3599, E-mail: gok1{at}cdc.gov.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Huhtamo, N. Putkuri, S. Kurkela, T. Manni, A. Vaheri, O. Vapalahti, and N. Y. Uzcategui Characterization of a Novel Flavivirus from Mosquitoes in Northern Europe That Is Related to Mosquito-Borne Flaviviruses of the Tropics J. Virol., September 15, 2009; 83(18): 9532 - 9540. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Junglen, A. Kopp, A. Kurth, G. Pauli, H. Ellerbrok, and F. H. Leendertz A New Flavivirus and a New Vector: Characterization of a Novel Flavivirus Isolated from Uranotaenia Mosquitoes from a Tropical Rain Forest J. Virol., May 1, 2009; 83(9): 4462 - 4468. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. A. Medeiros, M. R. T. Nunes, P. F. C. Vasconcelos, G.-J. J. Chang, and G. Kuno Complete genome characterization of Rocio virus (Flavivirus: Flaviviridae), a Brazilian flavivirus isolated from a fatal case of encephalitis during an epidemic in Sao Paulo state J. Gen. Virol., August 1, 2007; 88(8): 2237 - 2246. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
