Flaviviruses (family Flaviviridae, genus Flavivirus) are enveloped viruses with a 10–11-kb positive-strand RNA genome. The genome contains a single open reading frame encoding three structural (C, preM/M, and E) and seven nonstructural (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) proteins flanked by 5′ and 3′ untranslated regions [16, 27]. Most flaviviruses are arthropod-borne viruses that are transmitted by mosquitoes or ticks, except for a group of vertebrate-infecting flaviviruses with unknown vectors or no vectors. A group of flaviviruses that lacks the capacity to replicate in vertebrates was found in a cell line and/or natural populations of mosquitoes [46, 8, 1013, 23, 28]. These viruses are now recognized as “insect-specific flaviviruses” that are specifically adapted to their host mosquitoes and are probably maintained by vertical transmission in nature [2, 17, 22].

Culex flavivirus (CxFV) is an insect-specific flavivirus isolated for the first time from Culex pipiens, Cx. tritaeniorhynchus, and Cx. quinquefasciatus mosquitoes in Japan and Indonesia during 2003–2004 [10, 12]. Many strains of CxFV were subsequently detected in these and other Culex spp. mosquitoes (e.g., Cx. restuans, Cx. tarsalis, and Cx. interrogator) worldwide [1, 2, 7, 9, 15, 18, 21]. However, to date, CxFV has not been found in any other mosquito species, such as Aedes or Anopheles spp., suggesting that CxFV is strictly maintained in a host-genus-specific manner. Insect-specific flaviviruses, including CxFV, have often been detected in important vector species of mosquito-borne flaviviruses worldwide, although little is known of their own biological characteristics and their potential ecological effects on other viruses and host mosquitoes [14, 19].

Reverse genetics is a powerful molecular biology technique for analyzing the genetic determinants of viral growth and virulence in positive-strand RNA viruses, such as flaviviruses [3]. The construction of infectious cDNA clones and production of recombinant progeny viruses have been reported for multiple arthropod-borne flaviviruses, and these constructs have been utilized in molecular genetic studies or vaccine development. However, the production of infectious flavivirus cDNA clones is often difficult because of the genetic instability of the cloned viral cDNA in Escherichia coli [20, 24, 29]. To date, there have been no reports of the successful establishment of a system for the production of recombinant insect-specific flaviviruses, including CxFV. In this study, we report the first successful construction of a full-length infectious cDNA clone of CxFV.

A CxFV strain NIID21 (designated CxFV-NIID21; GenBank accession no. AB377213) isolated from Cx. pipiens mosquitoes in Tokyo in 2003 was used for the construction of a full-length CxFV cDNA clone. This viral strain was then propagated in C6/36 mosquito cells as described previously [10]. Viral RNA was extracted from the culture supernatant using a QIAamp Viral RNA kit (QIAGEN, Hiden, Germany) and used for viral cDNA synthesis with SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA). Four regions of the virus genome were amplified by PCR using thermostable high-fidelity DNA polymerase KOD plus ver. 2.0 (Toyobo Co. Ltd., Tokyo, Japan) and serially ligated into the low-copy-number plasmid pMW119 (Nippon Gene Co., Tokyo, Japan) [25, 26], which was then introduced into E. coli DH5α or Stbl2 (Invitrogen) as follows (Fig. 1). The 5′ terminal-NS1 region of the CxFV genome was amplified using the primers T7-5T and CV3038r during the initial PCR and primers SI-AI-T7 and CV3038r during the second PCR (primers used for the construction of the clones are listed in Supplementary Table 1). Note that the primers T7-5T and SI-AI-T7 contained a complete T7 polymerase promoter sequence for in vitro transcription. The PCR product was subcloned into pMW119 at SphI-SalI sites (5′T-NS1/pMW119; Fig. 1). The NS1–NS4B region of the CxFV genome was amplified with the primers CV2426f and CV6956rBI, and the PCR product was subcloned into the NheI-BamHI sites of 5′T-NS1/pMW119 (5′T-NS4B/pMW119; Fig. 1). The NS5-3′ terminal region of the CxFV genome was amplified with the primers NS5-GSP1 and KI-3T. The primer KI-3T contained an additional KpnI site to produce run-off transcripts. The PCR product was subcloned into the BamHI (blunt-ended)-KpnI sites of 5′T-NS4B/pMW119 (5′T-NS4B-3′T/pMW119, Fig. 1). The NS3–NS5 region of the CxFV genome was amplified with the primers CV5925f and CV9925r, and the PCR product was subcloned into the Eco47III-EagI sites of 5′T-NS4B-3′T/pMW119 to construct a complete recombinant CxFV clone (rCxFV/pMW119, Fig. 1). All transformants were incubated at low temperature (25–30°C) in the presence of low levels of antibiotic (25 μg mL−1 ampicillin) to minimize unexpected mutations in the viral sequence of the recombinant plasmid [24]. The nucleotide sequence of the viral genome region in the recombinant plasmid was confirmed after several passages in E. coli, and no sequence mutations were detected. This suggested that the full-length CxFV cDNA clone was stable in E. coli under the culture conditions.

Fig. 1
figure 1

Schematic representation of the construction of a full-length CxFV cDNA clone using the low-copy-number plasmid pMW119. The upper panel of the figure shows the genome organization of CxFV and the RT-PCR primers and cDNA fragments used for cloning. The T7 RNA polymerase promoter and restriction enzyme sites used for plasmid construction are indicated

To prepare a template for in vitro transcription, the rCxFV/pMW119 plasmid was digested at the 3′ end of the viral genome with KpnI and purified using a GenElute PCR Clean-Up Kit (Sigma-Aldrich, St. Louis, MO), and 1 μg of linearized DNA was transcribed using an mMESSAGE mMACHINE RNA transcription kit (Ambion, Austin, TX) containing m7G(5′)ppp(5′)G cap analog according to the manufacturer’s instructions. After DNase I treatment, synthesized RNA was purified using an RNeasy MinElute Cleanup Kit (QIAGEN). An aliquot was electrophoresed on a 1% denaturing formaldehyde agarose gel to verify the efficient transcription of the full-length product (Supplementary Fig. 1). C6/36 cells (1 × 106 cells) were subsequently transfected with 3 μg of in vitro-transcribed RNA using Lipofectamine 2000 reagent (Invitrogen) as described previously [25, 26], and the cells were incubated at 28°C in an atmosphere of 5% CO2 for 6 days. After two additional blind passages to increase the virus titer, culture supernatants were harvested and subjected to RNA extraction. Propagation of the recombinant progeny virus was confirmed by RT-PCR using a primer set specific for part of the E region of CxFV and sequencing the resultant RT-PCR product (data not shown). The recombinant progeny virus produced a mild cytopathic effect that was indistinguishable from the one caused by the parental virus at 5–6 days after infection (Fig. 2a-c). Negative-staining electron microscopy of infected culture supernatant confirmed the production of the recombinant virus particles, which were morphologically indistinguishable from the wild-type parent virus (Fig. 2d and e) [10].

Fig. 2
figure 2

a-c. Phase-contrast micrographs of control (mock-infected) C6/36 cells (a), parental CxFV-infected cells (b), and recombinant CxFV-infected cells (c) at 5 days post-infection. Scale bar, 100 μm. d and e. Negative-contrast electron micrographs of parental (d) and recombinant (e) CxFV particles from the culture supernatant at 5 days post-infection. Scale bar, 20 nm

Because plaque formation assays are not applicable to CxFV [10], the viral titer was measured by quantitative PCR (qPCR) to determine the copy number of viral genomic RNA. We initially prepared viral RNA standards. In brief, a 353-bp region within the E gene (sequence position 91–443) was initially amplified by RT-PCR. A second PCR was conducted to add a T7 promoter sequence at the 5′ terminus of the first PCR product, which was purified and subjected to in vitro transcription as described above (primer sequences are available upon request). The resultant transcripts were quantified and serially diluted 10-fold before being used as templates for standard curve analysis. To examine the growth properties of the recombinant progeny virus, C6/36 cells were seeded at a density of 0.5 × 106 cells per well in a 6-well plate and inoculated with 1.7 × 109 RNA copies of the parent or recombinant CxFV. The inoculum was removed after 1 h of virus adsorption, and cells were rinsed with phosphate-buffered saline before incubation at 28°C in 2 mL of minimum essential medium supplemented with 2% fetal bovine serum. A 140-μL aliquot of the culture supernatant was collected daily on days 0–8 after infection, and equal volumes of fresh culture medium were added to maintain the sample volume. RNA extracted from each culture supernatant was reverse-transcribed to cDNA using the PrimeScript RT Master Mix (Takara Bio, Shiga, Japan). qPCR was performed using 12 μL of reaction mixture containing 1 μL of cDNA solution, 5 pmol of each forward and reverse primer for the partial E region (E129f and E280r, Supplementary Table 1), 0.25 μL of ROX Reference Dye II (Takara Bio), and 6 μL of SYBR premix Ex Taq II (Takara Bio) on the 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA). qPCR conditions were as follows: one cycle at 95°C for 30 s followed by 40 cycles at 95°C for 5 s and 60°C for 34 s. Viral RNA levels were estimated from the threshold cycle and standard curve. As shown in Fig. 3, growth properties of the recombinant progeny virus were similar to those of the parental virus. The yields of the parent and progeny virus reached peak titers within 5–6 days after infection.

Fig. 3
figure 3

Comparison of growth characteristics in C6/36 cells of the parental virus and recombinant progeny virus. Each point represents the mean titer of triplicate experiments

In conclusion, we describe the construction of a stable full-length CxFV cDNA clone using a low-copy-number plasmid. RNA transcribed from the clone was infectious when introduced into susceptible mosquito cells. The recovered progeny virus was indistinguishable from the parental virus in terms of its growth kinetics, cytopathogenicity, and virion morphology. This is the first report of the production of an infectious insect-specific flavivirus clone. This infectious clone provides a useful basic tool for producing deletion, insertion, and amino acid substitution mutants or chimeric viruses to elucidate the molecular determinants of replication, virulence, and host range specificity of CxFV. The reverse genetics system for CxFV established in this study will offer new opportunities for molecular genetic studies directed at understanding the biological nature and evolution of insect-specific flaviviruses.