Abstract
The GenBank/EMBL/DDBJ accession numbers for the sequences reported in this paper are EU564827, EU564828, EU564829, EU564830 and EU564831 for Bwamba virus, Pongola virus, Fort Sherman virus, Xingu virus and Shokwe virus S segment RNAs, respectively.
The virus genus Orthobunyavirus, of family Bunyaviridae, includes over 100 viruses, more than 25 of which are associated with a variety of human illnesses ranging from mild to severe. Orthobunyaviruses are arthropod-borne and are distributed throughout both hemispheres; various small mammals and mosquito species generally serve as amplifying hosts and principal vectors of transmission, respectively (Barrett & Shope, 2005; Calisher & Karabatsos, 1988; Elliott et al., 2000). Viruses of the genus Orthobunyavirus are enveloped, containing a tripartite genome of negative polarity. The three genomic segments, L, M and S, can occur in end-hydrogen-bonded circularized forms and encode an RNA-dependent RNA polymerase, envelope glycoproteins (Gn and Gc) and nucleocapsid (N) and non-structural proteins (NSs), respectively. Orthobunyaviruses are diverse, with members classified into individual serogroups, subtypes and complexes according to serological reactivity (Barrett & Shope, 2005; Calisher & Karabatsos, 1988; Elliott et al., 2000; Shope & Causey, 1962). The majority of human pathogens within the genus Orthobunyavirus are classified within three serogroups: (i) the predominantly North American and European California serogroup, (ii) the exclusively New World Group C viruses and (iii) the predominantly African, Central and South American Bunyamwera serogroup (Barrett & Shope, 2005; Calisher & Karabatsos, 1988; Elliott et al., 2000; Shope & Causey, 1962). To date, genetic classification of orthobunyaviruses is limited by a lack of available nucleotide sequence data.
To help facilitate genetic classification, we have determined the full-length S segment RNA sequences of five human pathogens of the genus Orthobunyavirus. We describe the S segment sequences of Fort Sherman, Shokwe and Xingu viruses of the Bunyamwera serogroup, as well as those of Bwamba and Pongola viruses, the only known members of the relatively small, exclusively African, Bwamba serogroup. Human infections with these viruses have generally been associated with febrile illnesses that can present in a nondescript fashion (associated with Fort Sherman, Shokwe and Xingu viruses) or with arthritis (associated with Pongola virus) or a rash (associated with Bwamba virus). Subsequent characterizations and phylogenetic analyses of newly determined sequences in comparison with previously established S segment sequences of selected orthobunyaviruses indicate that the African Bwamba serogroup viruses are more closely related to North American and European viruses of the California serogroup than to other African viruses of the genus Orthobunyavirus.
Viruses were provided by the Arbovirus Diseases Branch of the Division of Vector-Borne Infectious Diseases, National Center for Zoonotic, Vector-Borne and Enteric Diseases (NCZVED), Centers for Disease Control and Prevention (CDC), World Health Organization (WHO) arthropod-borne virus reference collection. Low-passage preparations of the following virus strains were used in this study: Fort Sherman 86MSP18 strain, isolated from a human in Panama in 1985; Shokwe SAAr 4042 strain isolated from Aedes cuminsii in Ndumu, South Africa, in 1962; Xingu BeH 388464 strain, provided to the WHO reference collection in 1987 and isolated from a human in Brazil of unknown date; Bwamba M459 strain, isolated from a human in Uganda in 1937; and Pongola SAAr1 strain, isolated from Aedes (Neomelaniconion) circumluteolus in the Pongola river area of South Africa in 1955.
All procedures described in this manuscript were performed according to manufacturer's instructions unless stated otherwise. Viral RNAs were extracted using the QIAamp Viral RNA mini kit (Qiagen). Purified viral RNAs were subjected to amplification using a single primer designed according to the conserved, complementary terminal ends of the orthobunyavirus S segment (5'-AGTAGTGTRCTCCAC-3'). Viral RNA (5 µl) and 100 pmol primer were heated to 65 °C for 5 min and immediately placed on ice. OneStep RT-PCR mastermix (45 µl; Qiagen) was added to the RNA/primer mixture. Reaction mixtures were amplified by RT-PCR using the following conditions: 45 °C for 60 min, 95 °C for 15 min, followed by 45 cycles of 94 °C for 30 s, 45 °C for 1 min and 72 °C for 2 min. Reactions were terminated with a final extension step at 72 °C for 10 min.
Products of amplification were evaluated by electrophoresis in a 1.5 % agarose gel in 0.04 M Tris/acetate/0.001 M EDTA (TAE) buffer. Visualization of DNA bands was achieved through ethidium bromide staining and UV transillumination. DNA bands were excised from the agarose gel and purified using the QIAquick gel extraction kit (Qiagen).
Purified DNAs were ligated into the pCR2.1-TOPO vector (Invitrogen) and ligated plasmids were introduced into TOP 10F' competent Escherichia coli (Invitrogen) through chemical transformation. Transformed cells were spread onto imMedia Amp Blue agar (Invitrogen) and incubated overnight at 37 °C. White colonies were picked and each white colony evaluated was used to inoculate 3 ml imMedia (Invitrogen). Inoculated media were incubated, shaking, at 37 °C overnight. Plasmid DNAs were then extracted from 1.7 ml of these E. coli preparations using the minprep kit (Qiagen).
To verify the presence, and determine the sequence, of DNA inserts, extracted plasmid DNAs were sequenced with M13 forward and reverse primers provided with the pCR2.1-TOPO vector (Invitrogen) using the BigDye Terminator v3.1 ready reaction cycle sequencing mix (Applied Biosystems). Sequencing reactions were purified using the DyEx 2.0 spin kit (Qiagen). Nucleotide sequences were determined by running purified sequencing reactions on the ABI 3130 Genetic Analyzer (Applied Biosystems). At least three plasmid preparations, originating from three separate colonies, were sequenced for each virus. All sequences were confirmed by direct sequencing of purified RT-PCR products using virus-specific primers designed from plasmid-derived sequence. 5'- and 3'-terminal end sequences were determined through poly(A) tailing of purified viral RNAs (Ambion) and 5'/3' RACE kit (Roche), followed by nucleotide sequencing. By using the described methods, all presented sequences were confirmed from multiple preparations of viral RNAs and RT-PCR amplifications, in full-length, in both the 5' and 3' directions.
General characteristics of newly derived and previously determined S segment sequences of human pathogens representing all medically important serogroups of the genus Orthobunyavirus for which S segment sequence data has been generated, are presented in Table 1. S segment sequences of medically important viruses of the California, Bwamba, Bunyamwera, Group C and Simbu serogroups of the genus Orthobunyavirus are described in Table 1. The total and non-coding region (NCR) nucleotide sequence lengths vary among the compared viruses, with the African Bwamba virus having the longest S segment total and 3' NCR sequences (1096 and 310 nucleotides, respectively) (Table 1). In addition, the Bwamba serogroup virus S segments have comparatively high A+U content; the Bwamba virus S segment has the highest overall A+U content of nearly 65 mol%, which is contributed to by the long, A+U-rich, 3' NCR (Table 1). The significance of these findings is not known. The 5' and 3' termini of the newly derived Bwamba and Bunyamwera serogroup S segment sequences are highly conserved; only the Shokwe and Xingu virus S segments differ from the other presented sequences by the last two nucleotides of the 3' terminal region (AC for CT) and the 3' terminal nucleotide of the S segment (C for T), respectively (data not shown). The coding regions of viruses in the Bunyamwera serogroup (the Central and South American Fort Sherman and Xingu viruses and the African Shokwe virus) have identical N and NSs gene ORF nucleotide and amino acid sequence lengths, 702 nucleotides (nt)/233 amino acids (aa) and 306 nt/101 aa, respectively, compared to those of the prototype African Bunyamwera virus strain. Overall, these N and NSs sequence lengths are generally conserved among viruses of the Bunyamwera serogroup (Table 1). It is of interest that the newly determined N and NSs nucleotide and amino acid sequences of the African Bwamba viruses, Bwamba and Pongola, have nearly identical lengths compared to those of the majority of represented North American California serogroup viruses (Table 1). When the coding regions of Bwamba and Pongola viruses are directly compared to the North American California serogroup La Crosse virus (strain Human, 1978), only the N gene of Bwamba virus differs by 3 nt (705/708) and 1 corresponding amino acid (234/235; Table 1).
Table 1. Characteristics of newly derived Bwamba and Bunyamwera serogroup S segments shown and previously derived full-length S segments of representative human pathogens of the California, Bunyamwera, Group C and Simbu serogroups GenBank accession numbers appear next to virus names. Apeu and Oropouche viruses are included as Group C and Simbu serogroup references, respectively.
The percentage identities shared among newly derived and previously determined N gene ORF nucleotide and amino acid sequences are presented in Table 2. Percentage identities were calculated by using the MEGA version 4 software pairwise distance calculation function using a number of differences only model for both nucleotide and amino acid analyses (Tamura et al., 2007; Table 2). Fort Sherman, Xingu and Shokwe viruses share between 74.1 and 90 % nucleotide and 74.2 to 97.4 % amino acid sequence identities with viruses of the Bunyamwera serogroup (Table 2). Among all compared viruses, the African Shokwe and Bunyamwera viruses share the highest nucleotide and amino acid sequence identities of 90 and 97.4 %, respectively (Table 2). The Bwamba serogroup viruses, Bwamba and Pongola, share nearly 80 % nucleotide and 83 % amino acid sequence identies within the serogroup (Table 2). Interestingly, the Bwamba serogroup viruses also share, on average, over 70 % nucleotide and nearly 67 % amino acid sequence identities with the California serogroup reference La Crosse virus; these identities are higher than those shared between any other compared viruses of heterologous serotypes (Table 2).
Table 2. N gene nucleotide and amino acid sequence identities among human pathogenic California, Bwamba, Bunyamwera, Group C and Simbu serogroup viruses GenBank accession numbers appear next to virus names. Apeu and Oropouche viruses are included as Group C and Simbu serogroup references, respectively. Nucleotide and amino acid sequence identities are shown above and below the diagonal, respectively. Emboldened values represent comparisons highlighted within the text of the Bwamba and California serogroups. Underlined values represent highlighted comparisons between viruses of like serogroup.
Phylogenetic analyses of newly determined N and NSs gene ORF sequences in comparison to previously determined S segment sequences of selected members of the genus Orthobunyavirus were conducted using MEGA version 4 (Tamura et al., 2007). Sequences of viruses of the Bunyamwera, Bwamba, California, Group C and Simbu serogroups of the genus Orthobunyavirus were aligned using CLUSTAL W (MEGA version 4). Both neighbour-joining (NJ) and maximum-parsimony (MP) trees were generated and analysed with 2000 replicates for bootstrap testing of each grouping (Saitou & Nei, 1987; Eck & Dayhoff, 1966; Felsenstein, 1985). Trees generated by NJ and MP methods displayed nearly identical topologies with highly comparable bootstrap values for groupings generated from both the N (NJ method shown in Fig. 1) and NSs (data not shown) ORF nucleotide sequences. Within the NJ-inferred evolutionary history, distances were computed using the maximum composite likelihood method (Tamura et al., 2004; Fig. 1). From this analysis, three major lineages of orthobunyaviruses are depicted with extreme bootstrap support (99–100 %). The three major lineages are mostly composed of (i) the predominantly North American California serogroup, (ii) the exclusively New World Group C viruses and (iii) the predominantly African Bunyamwera serogroup. As expected, Fort Sherman, Shokwe and Xingu viruses group within the Bunyamwera serogroup of lineage (iii) with strong bootstrap support (Fig. 1). However, while the American and African viruses generally cluster with viruses from similar geographical regions, there is relatively weak support for some of these groupings, suggesting a limited association between geographical and genetic distinctions among some viruses within the Bunyamwera serogroup (Fig. 1). Surprisingly, the African Bwamba serogroup viruses, Bwamba and Pongola, cluster together in one of two groups represented in lineage (i); the California serogroup viruses (from North America and Europe) comprise the second, major group from lineage (i), grouping with extreme bootstrap support (Fig. 1).
|
On the African continent, viruses of both the Bwamba and Bunyamwera serogroups of the genus Orthobunyavirus have been associated with predominantly febrile human illnesses (Barrett & Shope, 2005; Kalunda et al., 1985; Lanciotti & Tsai, 2007; Lutwama et al., 2002; Smithburn et al., 1941). While there is some evidence of Tahyna virus circulation in Africa (Chastel et al., 1983; Kuniholm et al., 2006), all other human pathogens of the California serogroup, and the primarily encephalitic human illnesses associated with these viruses, are exclusively known to occur in North America and Europe (Barrett & Shope, 2005; Gonzales-Scarano & Nathanson, 1996; Lanciotti & Tsai, 2007). These molecular and phylogenetic characterizations of the N and NSs ORFs suggest that a common S segment ancestor is shared between the serologically, clinically and geographically disparate Bwamba and California groups of the genus Orthobunyavirus. We speculate that ancestral subpopulations of unidentified origin became isolated through host migration and geographical separation of unknown mechanisms. As a function of this isolation, varying selective pressures among ecologically distinct vertebrate and arthropod hosts are proposed to have driven the divergent evolution of these viruses, resulting in the unique and defining clinical and serological characteristics of the California and Bwamba serogroups. Although no information regarding the M segments is available, we hypothesize that, as a consequence of this divergence, the M segment sequences of Bwamba serogroup viruses will be more variable than the presented S segment sequences compared to those of the California serogroup viruses, because of the association between M segment-encoded glycoproteins, vertebrate and arthropod host cell attachment and entry (Barrett & Shope, 2005; Ludwig et al., 1991).