Rescue of Recent Virulent and Avirulent Field Strains of Bluetongue Virus by Reverse Genetics

Since 1998, Bluetongue virus (BTV)-serotypes 1, 2, 4, 9, and 16 have invaded European countries around the Mediterranean Basin. In 2006, a huge BT-outbreak started after incursion of BTV-serotype 8 (BTV8) in North-Western Europe. More recently, BTV6 and BTV11 were reported in North-Western Europe in 2008. These latter strains are closely related to live-attenuated vaccine, whereas BTV8 is virulent and can induce severe disease in ruminants, including cattle. In addition, Toggenburg orbivirus (TOV) was detected in 2008 in Swiss goats, which was recognized as a new serotype of BTV (BTV25). The (re-)emergency of known and unknown BTV-serotypes needs a rapid response to supply effective vaccines, and research to study this phenomenon. Recently, orbivirus research achieved an important breakthrough by the establishment of reverse genetics for BTV1. Here, reverse genetics for two recent BTV strains representing virulent BTV8 and avirulent BTV6 was developed. For this purpose, extensive sequencing of full-genomes was performed, resulting in the consensus sequences of BTV8/net07 and BTV6/net08. The recovery of ‘synthetic BTV’, respectively rgBTV8 and rgBTV6, completely from T7-derived RNA transcripts was confirmed by silent mutations by which these ‘synthetic BTVs’ could be genetically distinguished from wild type BTV, respectively wtBTV6 and wtBTV8. The in vitro and in vivo properties of rgBTV6 or rgBTV8 were comparable to the properties of their parent strains. The asymptomatic or avirulent properties of rgBTV6 and the virulence of rgBTV8 were confirmed by experimental infection of sheep. Reverse genetics of the vaccine-related BTV6 provides a perfect start to develop new generations of BT-vaccines. Reverse genetics of the virulent BTV8 will accelerate research on the special features of BTV8, like transmission by species of Culicoides in a moderate climate, transplacental transmission, and pathogenesis in cattle.


Introduction
Bluetongue virus (BTV) belongs to the family Reoviridae, genus Orbivirus [1]. BTV-transmission between ruminants, including cattle, sheep, and goats, occurs in majority by bites of species of Culicoides. Bluetongue (BT) is listed as a 'notifiable disease' by the Office International des Epizooties (OIE) [2] causing severe hemorrhagic disease with fever, lameness, coronitis, swelling of the head (particularly the lips and tongue) and death. There are 24 BTV-serotypes recognized as defined by cross neutralization assays.
The genome of BTV consists of ten linear double-stranded RNA genome segments Seg-1 to Seg-10 encoding structural proteins VP1 to VP7, and nonstructural proteins, NS1, NS2 and NS3/NS3a, for reviews see Roy et al. [3,4]. The virus particle composes three shells of proteins, the inner shell consists of VP3 encoded by Seg-3, the middle shell consists of VP7 encoded by Seg-7, and the outer shell is formed by VP2 (Seg-2) and VP5 (Seg-6). The BTV-particle further contains three enzymatic proteins, VP1 (Seg-1), VP4 (Seg-4) and VP6 (Seg-9), and one copy of each of the ten genome segments Seg1 to Seg-10 in the inner shell. The nonstructural proteins NS1, NS2, and NS3/3a encoded by Seg-5, Seg-8, and Seg-10, respectively, are not part of the BTV particle.
These incursions and the re-emergency of BTV-serotypes in Europe are addressing the need for safe and effective vaccines, but also for research on the recent development of Bluetongue in areas with a moderate climate. Spread of Bluetongue is restricted to the expansion of the competent vector, specific species of Culicoides. BTV8/net06 must be transmitted by endemic European Culicoides spp. [12,13,14], as the common vector C. imicola was not found in the affected area. Further, Bluetongue is generally believed to be a 'sheep disease', but BTV8/net06 is causing significant clinical signs in cattle [15]. Finally, BTV8/net06 is passing the placenta of pregnant heifers and ewes, although this is restricted to BTvaccines or cell-adapted BTV-strains [16]. It was therefore suggested that BTV8/net06 should be related to a BT-vaccine or a laboratory strain of BTV8. To study the specific properties of BTV8/net06, a method to genetically manipulate BTV is required. For BTV8/net06 and BTV6/net08, a genetic modification system based on the uptake of a plasmid-derived T7transcript was published [17]. However, this system is not efficient, laborious, and needs positive selection to remove wild type virus. Recently reverse genetics was developed based on T7 RNApolymerase driven run-off transcripts for BTV1 [18]. This reverse genetics system was successfully used to genetically modify BTV [19,20,21]. The reverse genetics system offers many advantages, like the efficient rescue of virus with fully defined genomes, recovery of mutant viruses with delayed growth characteristics, and tailor-made BTV-reassortants. Here the rescue of completely 'synthetic' virulent BTV8 (rgBTV8), and completely 'synthetic' avirulent BTV6 (rgBTV6) is described. Both 'synthetic' BTVs are studied in sheep in order to confirm their properties compared to their wild type strains, wtBTV8 and wtBTV6.
BTV8/net07 was isolated from the Holstein Frisian cow NL441689187 from Bavel, the Netherlands, which was sampled for export purposes on July 24th, 2007 [24]. Isolation from EDTA-blood was performed on eggs (e1) and subsequent three passages on BHK-21 cells (e1/bhk3) or subsequent three passages on KC cells (e1/kc3). BTV6/net08 was isolated from the Holstein Frisian cow NL415834681 (24th October 2008) from Heeten, the Netherlands [7]. Isolation from EDTA-blood was performed on eggs and subsequent three passages on BHK-21 cells and once on BSR cells.
All virus stocks were obtained by infection of BSR cells at low multiplicity of infection (MOI) and harvested after 100% cytopathic effect (CPE) was observed. Virus titers were determined by endpoint dilution and expressed as plaque forming units per ml. Viral stocks were stored at 280uC.
Contig sequences of BTV8 strains were checked manually and compared to sequences of BTV-8nt(2006/04) [5] with genbank accession numbers AM498051-AM498060) and to sequences of BTV8/Neth2006 submitted by A.C. Potgieter ([25] with genbank accession numbers FJ183374-FJ183383 for Seg-1 to Seg-10. Consensus sequences of BTV8 directly isolated from EDTAblood were determined after RNA-isolation using the TRIZOL method, and reverse transcription and amplification by a one-tube system (Qiagen one-step RT-PCR kit). Overlapping DNA amplicons were sequenced using BigDyeH Terminator v1.1 Cycle Sequencing Kit in a ABI PRISMH 3130 Genetic Analyzer Applied Biosystems.
Construction of T7 plasmids with cDNA-inserts derived from BTV genome segments cDNA of genome segments were synthesized by Genscript Corporation (Piscataway, NJ). Genome segments Seg-1 to Seg-10 of BTV1 were based on the sequences as submitted to Genbank with accession numbers FJ969719-FJ969728 [18]. Genome segments Seg-1 to Seg-10 of BTV8/net07 and of BTV6/net08 were based on the consensus sequences described in this paper. cDNAs were cloned in pUC-derivatives, pUC18, 19 or 57, or pJET1.2 under control of the T7 RNA-polymerase promoter and a recognition site for a restriction enzyme at the 39-terminus for defined run-off transcription ( Figure 1, depicted from Boyce et al., [18]). Plasmids were maintained in E.coli DH5a, and were purified using the QIAfilter Plasmid Midi Kit (Qiagen).

In vitro synthesis of BTV-transcripts
Plasmid DNA was digested at the 39-terminus with the respective restriction enzyme, and was purified by standard procedures. One mg digested plasmid DNA was used for in vitro RNA run-off transcription with 59-cap analogue using the MESSAGE mMACHINE T7 Ultra Kit (Ambion). In this reaction, a ratio of 4:1 of anti-reverse cap analogue to rGTP Figure 1. Schematic overview of plasmids containing a full-length BTV genome segment. A full-length BTV genome segment flanked by a T7 promoter and an unique site for a restriction enzyme (here BsmBI is indicated as an example), which defines the BTV 39end sequence during transcription (depicted from Boyce et al., [18]. The nucleotides of the ultimate 59-and 39-ends of the BTV genome segment are presented in bold symbols. The sequence of the T7 promoter is italicized, and that of the unique site is underlined. The positions of the start of transcription and digestion by restriction enzymes for run-off transcription are indicated by arrows. doi:10.1371/journal.pone.0030540.g001 was used. Synthesized RNA was purified by use of MEGAclear columns (Ambion) according to the manufacturer's instructions, and eluted RNA was stored at 280uC.
Transfection of T7-derived RNA transcripts to rescue 'synthetic' BTV1, BTV6 and BTV8 Monolayers of 10 5 BSR cells per 2 cm 2 were transfected with equimolar amounts of RNA of BTV segments encoding VP1, VP3, VP4, NS1, VP6, NS2. In total, 600 ng RNA was transfected using 1 mg lipofectamine TM 2000 (1:2.5; 1 mg/ml Invitrogen) in Opti-MEMH I Reduced Serum Medium according to manufacturer's conditions. Eighteen to twenty hours post transfection, monolayers were transfected again with 600 ng equimolar amounts of ten RNA segments. At 4 hrs post transfection, the transfection mix was replaced with 1 ml DMEM supplemented with 5% FBS and 1% of Penicillin/Streptomycin/Fungizone. Supernatants were harvested from monolayers developing cytopathogenic effect (CPE) at 48 hrs after the second transfection. BTV-specific CPE was confirmed by immunostaining of fixed monolayers with monoclonal antibody (Mab) produced by ATCC-CRL-1875 directed against VP7 according to standard procedures. In case of monolayers without visible CPE, duplicate wells were passed 1:5 to rescue virus from less efficiently transfected monolayers. After incubation of three days, monolayers were screened for CPE and VP7-expression as described above.

Characterization of rescued BTVs
Viral RNA of virus stocks was isolated from 200 ml of supernatant with the High Pure Viral RNA kit (Roche). Genome segments Seg-2 and Seg-6 were identified by amplification of (parts of) the genome segments with serotype-specific primers, which are listed in Table 1. Template RNA (6 ml) was denaturated at 94uC for 3 min and immediately cooled on ice. A one-step RT-PCR kit (Qiagen) was used to reversely transcribe RNA and amplification of cDNA in a RT-PCR containing both primers. The reaction mix contained 10 ml of 5xQIAGEN one-step RT-PCR buffer, 2 ml of dNTP mix, 0.6 mM of each primer and 2 ml of the enzyme mix (containing RT and PCR reaction enzymes). RNase-free water was added to a total volume of 44 ml. Six microliters of denatured RNA was added to the mix. The RNA was reversely transcribed at 45uC for 30 min. This was followed by an activation step at 94uC for 15 min. Forty amplification cycles were then carried out (94uC for 1 min, 45uC for 1 min and 72uC for 2 min), followed by a terminal extension step at 72uC for 10 min. The cDNA products were analyzed by 0.9% agarose gel electrophoresis and visualized under UV light after staining with ethidium bromide.
For Seg-10, the in-house developed serogroup-specific diagnostic PCR-test was used [15]. Sequencing of amplicons was carried out to discriminate between different Seg-10 sequences by using the BigDyeH Terminator v3.1 Cycle Sequencing Kit in a ABI PRISMH 3130 Genetic Analyzer (both supplied by Applied Biosystems, Foster City, IA, USA).

Growth curves and plaque morphology of 'synthetic' BTV6 and 8 (rgBTV6 and rgBTV8)
Confluent BSR-monolayers in M24-well plates were infected in duplicate at a multiplicity of infection (moi) of 0.1. After attachment to cells for 1.5 h at 37uC, the medium was removed and refreshed with 1 ml of DMEM with 5% FBS, 1% Penicillin/ Streptomycin/Fungizone and incubation was continued. At 24, 48 and 72 h post infection (hpi), samples of the supernatants were harvested and stored at 280uC. Virus titers were determined by plaque-titration on BSR cells. Therefore, BSR cells were infected with tenfold dilutions of samples, and grown for 48 h under EMEM overlay medium (containing 1% methylcellulose). Infectious centers (hereafter referred to as plaques) were detected by immunostaining with a VP7-specific Mab ATCC-CRL-1875. Virus titers were expressed as pfu/ml.
To study plaque size, BSR monolayers were infected by tenfold dilutions of indicated viruses, and grown for two days under overlay medium (EMEM complete with 1% methylcellulose). Monolayers were fixed with methanol/aceton (1:1) and immunostained with Mab ATCC-CRL-1875. Plaques in appropriate dilutions were compared.

Experimental infection with rgBTV6 and rgBTV8
All experiments with live animals were performed under the guidelines of the European Community (86/609) and were approved by the Committee on the Ethics of Animal Experiments of the Central Veterinary Institute (Permit Number: 2011-003). Sixteen female Blessumer sheep of 6-24 months old and free of BTV and BTV-antibodies were commercially sourced from the same flock of a Dutch farm. The sheep were randomly allocated to four groups of four animals. On day 0 (0 dpi), a total of 1 ml of 10 5 TCID 50 /ml of wtBTV8 or rgBTV8 was administered subcutaneously to four sheep per virus between the shoulder blades left and right from the spinal cord. Four sheep were intramuscularly injected in the neck with 1 ml of 10 5 TCID 50 /ml rgBTV6. The fourth group served as control group. EDTA-blood samples were collected daily during the first week after inoculation and every other day until the end of the trial at 21 dpi. Serum samples were collected more frequently in the second week post inoculation. Samples were tested by the in-house PCR-test for BTV [15] for the detection of BTV RNA in EDTA-blood and by blocking ELISA (ID-VET) for the detection of BTV-specific antibodies in serum. Body temperature was recorded daily, and fever was defined as the average temperature plus two times the standard deviation. Clinical signs were recorded daily according to the clinical score table for BTV8 animal trials (Table S1).
Clinical signs were quantified following challenge by an adapted clinical reaction index (CRI) as described by Huismans [26]. A maximum score of 12 was given to the cumulative total of fever readings (a) as described above from days 3 to 14 post inoculation (dpi), a clinical score according the clinical score table and normalized to a maximum score of 12 (b). An additional 4 points were added to the sum of a and b if death occurred within 14 dpi.

Recovery of synthetic BTV1
Rescue of infectious BTV1 from all ten in vitro synthesized RNAs has been previously demonstrated [18]. With the aim of the

Recovery of rgBTV8
cDNAs from genome segments Seg-1 to Seg-10 of BTV8/net06 were synthesized based on genbank accession numbers AM498051-AM498060 [5]. However, new sequences became available for this strain [25] showing differences with the mentioned accession numbers. In order to start with a welldocumented strain of BTV8 we decided to extensively sequence the BTV8-isolate of the first reported case in 2007 [24]. Consensus sequences were determined of BTV8/net07/e1, BTV8/net07/ e1/bhk3, and BTV8/net07e1/kc3 by deep sequencing. Consensus sequences were compared, and only in BTV8/net07/e1/kc3 two nucleotide differences were found. In Seg-1 at nucleotide position 3277 CRT leads to amino acid change ARV and in Seg-2 position 1215ARG leads to amino acid change RRG. The consensus sequences determined by deep sequencing of BTV8/ net07/e1 and BTV8/net07/e1/bhk3 were the same as that of blood-derived BTV8/net07 which was determined by conventional RTPCR and sequencing and compared to sequences of BTV8/Neth2006 (Table 2).
The consensus sequence of BTV8/net07 was used to generate cDNAs encoding the correct amino acid sequence and contain the correct 59-and 39-UTR. Therefore, differences in the 59-and 39-UTR and changes resulting in amino acid mutations in the originally derived cDNAs (based on BTV8/Net2006) were corrected by standard procedures with respect to the consensus sequence from BTV8/net07. Consequently, 16 silent mutations remained in cDNAs used for rescue of BTV8/net07. T7 run-off transcripts from these corrected cDNAs were transfected according to the modified double transfection protocol. CPE appeared between 24-48 h after the second transfection. In general, CPE seemed to be slightly delayed compared to BTV1.
To confirm the rescue of 'synthetic' BTV8 (rgBTV8), a straightforward method was used to distinguish rgBTV8 from wild type BTV8 (wtBTV8). A part of Seg-2 encompassing a SnaBI recognition site that is specific for rgBTV8 (position 242) was amplified by RT-PCR using primers btv8-S2f and btv8-S2r6. Digestion by SnaBI of amplicons derived from wtBTV8 or rgBTV8 resulted in fragments of 1000 bp or 760 and 240 bp respectively (Fig. 2A, lane 1 wtBTV8 vs lane 2 rgBTV8). This demonstrated that a SnaBI site was uniquely present in rgBTV8. Successful digestion by XhoI confirmed the specificity, and comparability of both amplicons (lane 3-4). The amplicons were also sequenced using the btv8-S2f primer to indeed confirm the introduction of this distinguishable sequence (Fig. 2B).
The genetic stability of rgBTV8 was studied by sequencing the entire genome, except for the ultimate sequences at the 59and 39ends. RNA was isolated from the supernatant of the third passage, sequenced by the described method, and compared to sequences of the cDNAs used for rescue. Except for one silent mutation in Seg-3 at nucleotide position 260, the determined sequences of Seg-1 to Seg-10 were confirmed, including all silent mutations. In conclusion, 'synthetic' BTV8 (rgBTV8) was recovered, is genetically stable, and confirmed by the presence of silent mutations in the RNA segments of rgBTV8. These silent mutations, 17 in total, and divided over 6 genome segments, are unique and distinguish rgBTV8 from wild type BTV8 (wtBTV8).

Recovery of rgBTV6
Sequences of Seg1 to Seg-10 derived from BTV6/net08 e1/ bhk3/bsr1 were 100% identical with those of BTV6/Net2008/05 (accession numbers QG506472-QG506481) isolated by one passage on KC cells and three passages on BHK-21 cells. These consensus sequences were used for cDNA-synthesis. Two mutated LguI recognition sites of which one was changed to an NlaIV-site were introduced in the cDNA of Seg-2 without affecting the encoded amino acid sequence.
T7 RNA-transcripts from the digested BTV6/net08 plasmids were transfected according to the optimized protocol as described. CPE appeared between 1-2 days post the second transfection. The presence of rgBTV6 was confirmed by digestion with NlaIV instead of the LguI at position 910 in Seg-2. Therefore, a part of Seg-2 segments of rgBTV6 and wtBTV6, encompassing the respective region were partially amplified by RT-PCR using primers btv6-S2f8 and btv6-S2-2R. Digestion of these amplicons demonstrated that an extra NlaIV site was introduced in rgBTV6 compared to wtBTV6 (Fig. 3A, lanes 1  and 2). Digestion with LguI further confirmed that this recognition site was present in wtBTV8 and mutated in rgBTV8 (Fig. 3A, lanes 3-4). Sequences of the respective amplicons indeed confirmed the introduction of the distinghuishable, but silent, mutation (Fig. 3B). In conclusion, 'synthetic' BTV6 (rgBTV6) was recovered, and could be differentially identified by the presence of silent mutations.
Plaque morphology of the 'synthetic' viruses, rgBTV6 and rgBTV8, grown under overlay medium for two days was similar to that of the wtBTVs (Fig. 4B). In conclusion, rgBTV6 and rgBTV8 derived from ten in vitro synthesized T7 transcripts are indistinguishable from wtBTV6 respectively wtBTV8 with respect to growth characteristics in BSR cells. Furthermore, the introduced silent mutations rgBTV6 (two mutations in Seg-2) and in rgBTV8 (16 mutations in genome segments Seg-2, Seg-3, Seg-4, Seg-5, Seg-6 and Seg-8) did not influence these growth characteristics. Virulence of rescued 'synthetic' BTVs rgBTV6 and rgBTV8 All sheep of the control group remained negative by PCR, and ELISA. None of the sheep in the control group developed fever and only very mild clinical signs were observed, like nasal/ocular discharge and upper airway distress. Sheep infected (i.m.) with rgBTV6 developed very mild signs of BT (Fig. 5A), like loss of appetite for one day, mild ocular lacrimation and upper airway distress and/or nasal discharge for more than one day. All infected sheep developed fever between days 8 and 11 post infection (p.i.) and were PCR-positive for BTV from five dpi onwards the end (21 dpi) (Fig. 5B). All infected sheep seroconverted in the blocking ELISA (Fig. 5C) from 8 dpi onwards.
Sheep infected (s.c.) with wtBTV8 developed mild to moderate clinical signs of BT (Fig. 5A). Except for one sheep, which had no clinical signs at all, all sheep scored mild clinical signs as mentioned for rgBTV6. More severe clinical signs were observed from 9 dpi onwards. Two of the animals were listless, showed local oedema and red mucous in the mouth for several days. All sheep developed fever between 6 and 11 dpi. All sheep became PCRpositive between day 3-6 dpi and remained PCR-positive until the end of the trial (21 dpi) (Fig. 5B). Sheep seroconverted by ELISA from 7 dpi onwards (Fig. 5C).
Sheep infected (s.c.) with rgBTV8 developed similar clinical signs as the wtBTV8-infected sheep (Fig. 5A), except for one sheep, which showed no clinical signs at all. One sheep died at 8 dpi. This sheep also showed moderate clinical signs like apathy, salivation, red mucous in the mouth and conjunctivitis. All sheep in this group developed fever between 8-10 dpi. All sheep were tested PCR-positive from 3-9 dpi onwards (Fig. 5B), and seroconverted for VP7 antibodies from 7 dpi onwards (Fig. 5C). The sheep in both groups, infected with wtBTV8 and rgBTV8 which did not show clinical signs, developed fever, became PCRpositive and seroconverted, which confirmed successful infection.
The adapted clinical reaction index (CRI) was determined for each sheep and averaged for the different groups to further quantify clinical signs in groups (in particular wt-and rgBTV8). For the control group the CRI (mean 6 SE) was 1.160.4, for the rgBTV6 group 3.960.8, for the wtBTV8 group 8.762.2 and for rgBTV8 group the CRI was 7.963.3. This showed a significant difference (p,0.01) between the control group and the inoculated groups. There was also a significant difference in the CRI for rgBTV6 and both BTV8-groups, although the variation in the BTV8-groups was high. There was no significant difference in CRI between the wtBTV8-and the rgBTV8 -group.
Blood samples were used to determine the sequence of parts of Seg-2, as described above, and a part of Seg-3 for only rgBTV8 and wtBTV8. For all viruses, the determined sequences matched the sequence of the BTV administered to the sheep.
For this purpose, the complete sequences were determined by deep sequencing of three different strains originating from the same BTV8-isolate from the same cow. In addition, to exclude possible artefacts by the isolation procedure or the method of sequencing, RNA of BTV8 was directly isolated from EDTAblood and almost completely sequenced by conventional sequencing. Except for three passages in KC cells (BTV8/ net07/e1/kc3), the sequences of BTV8/net07/e1 and BTV8/ net07/e1/bhk3 completely match the sequence of the original blood-isolate. Indeed, BTV8/net07/e1/bhk3 induces similar d BTV8/net07/e1/bhk3 (described in this paper) was isolated on eggs and was further passaged 3 times on BHK21 cells before analysis. This sequence matches completely to sequences determined by Maan et al (2010). See also footnote a. e Amino acid position numbering from strain BTV8/Neth2007. doi:10.1371/journal.pone.0030540.t002  clinical signs as the blood-isolate (manuscript in preparation). In contrast, two differences were found after three passages in KC cells, which could be involved in adaptation to this cell line. Previously derived cDNAs were corrected with respect to this consensus sequences, but the remaining 16 nucleotide differences did not result in change of encoded amino acids or mutations in UTR-sequences. Consequently, rescued virus harbours genetic markers in six genome segments to identify the presence of synthetic BTV8 (rgBTV8). It has been shown for Sindbis-and poliovirus that silent mutations can affect the attenuation [27,28]. Similarly, remaining genetic differences could change the characteristics of rgBTV8 compared to wtBTV8, but no differences were observed in growth kinetics, plaque morphology and virulence. It can therefore be concluded that these remaining silent mutations had no effect on the characteristics of BTV8. Furthermore, all silent mutations remained present in rgBTV indicating that there is no strong positive selection pressure for the original genetic background of BTV8/net07. Finally, one extra silent mutation was found in Seg-3 at nucleotide position 260.
Except for Seg-7 and Seg-10 (no mutations), and Seg-2 and Seg-3 (silent mutations), obvious (nonsilent) variations were found in all other genome segments of BTV8/Neth2006 compared to the consensus sequences of BTV8/Neth2007 and strains derived from BTV8/net07, as listed in Table 2. Remarkably, all these genome segments (Seg-1, -4, -5, -6, -8, and -9) showed variation in the sequences of UTR-sequences, and some of these are located in the highly conserved ''ACTTAC'' sequence [29]. Of course, these differences could be involved in translation and/or replication of the respective genome segment. In addition, genome segments (Seg-1, -4, -8, and -9) showed variation in the encoded amino acid sequences. For Seg-1, the amino acid (aa) difference at position 493 (Q493E) is located between the N-terminal domain (NTD: residues 1-373) and the Polymerase domain (PD; residues 582-882) of VP1. The other aa difference T875A is located within the PD, but is located outside the conserved sequence motifs A-D [3]   at the end of the PD in the thumb region. For Seg-4, aa difference M340L is located in the 2'OMT domain (residues 155-377) of VP4. The aa difference F439G is located in the second region of the N7MTase domain (residues 110-154 and 378-509). For Seg-8, aa difference D59N is located in the N-terminal region of NS2 (1-115) which is involved in formation of viral inclusion bodies, binding of ssRNAs, and replication. Finally, most aa differences were found in Seg-9. These aa differences, M3L, A43T, D81R and A133V are not located in any of the conserved motifs that are characteristic for RNA helicases of the DExH type [30]. Summarizing, the cause and effect of these observed variations in several determined sequences of closely related strains of BTV8/Neth06 are unknown.
Most of the clinical signs associated to Bluetongue were observed after experimental infection of sheep with BTV8, and were consistent with clinical signs reported most frequently in the field in 2006 [31]. The most prominent of these were fever, apathy, erosions of the oral mucosa, salivation, local oedema and some problems with feet. To minimize interbreed susceptibility to BTV, sheep from the same flock were used. However, the severity of the clinical signs varied within a group, with moderate clinical signs in two of the sheep, but relatively mild (one sheep) to no clinical signs in the other sheep.
For the BTV6/net08 the consensus sequence was determined of BTV6/net08/e1/bhk3/bsr2 and was 100% identical to that of BTV6/Neth2008/05 (E1/KC1/BHK1) [10]. Virus rgBTV6 was synthesized based on this consensus sequence, except for two silent mutations in Seg-2, which were introduced for differentiation  between rgBTV6 and wtBTV6. In fact, BTV6/net08 is likely a reassortant of two live-attenuated vaccines, since 9 out of 10 genome segments are almost identical (99.7 to 100%) to these of live-attenuated vaccine for serotype 6 (MLV6), whereas the variable genome segment Seg-10 showed the greatest identity (98.4%) to that of MLV2 (RSAvvv2/02) [10]. Indeed, BTV6/ net08 hasn't shown obvious clinical signs in the field and after experimental infection [8,32]. Furthermore, transmission of BTV6/net08 was very low in N-W European conditions resulting in a natural decline, and finally BTV6/net08 disappeared in the winter of 2008/2009 without control measures were undertaken. Similar growth characteristics of wtBTV6 and rgBTV6 were observed (Fig. 4), and the absence of virulence of rgBTV6 was similar to previous observations after experimental infection of sheep, the most susceptible and sensitive ruminant for Bluetongue. This vaccine-related BTV6/net08 caused fever in all sheep, but showed very mild clinical signs in sheep compared to rgBTV8/ net07. Reverse genetics was developed for two recent field isolates of BTV, the virulent strain of BTV8 and the vaccine-related strain of BTV6. These BTV-isolates differ in many aspects, like serotype, virulence, pathogenesis in sheep and cattle, and transmission by Culicoides species which are endemic for N-W European countries with a moderate climate. Reverse genetics can be used to generate desired reassortants of these different viruses. Reassortment is a natural process of viruses with a segmented genome. Reassortants have also been described for BTV for a long time (reviewed in [33]). However, these 'natural' reassortants will arise by positive selection, even in cell culture. Previously, we have used this process by transfection of synthetic RNA in a BTV-infected cell, followed by screening for the desired 'synthetic' resassortant meaning with the synthetic genome segment [17]. However, this method was not very efficient and limits the possibilities. Reverse genetics can increase this method by formation of sets of desired genome segments for rescue of viable BTV. Even less fit virus will be rescued due to the absence of negative selection. In order to investigate this application, we have rescued multiple reassortants of BTV1 and BTV8 during the process of developing reverse genetics (Table 3). These reassortants were all viable indicating the flexibility of combination of sets of ten genome segments of BTV1 and BTV8.
However the 'synthetic' BTVs, rgBTV6 and BTV8, are genotypically distinguishable from their 'ancestor' viruses, BTV6/net08 and BTV8/net06, these 'synthetic BTVs' are phenotypically indistinguishable from their 'ancestors' as studied for several in vitro parameters and virulence in sheep. Other aspects, including transplacental transmission and replication in the vector, are not studied in detail yet. Reverse genetics as genetic modification system, here demonstrated by the introduction of many silent mutations, will be a powerful tool for in-depth investigations by site-directed mutagenesis of viral genes involved in these intriguing phenomena of bluetongue virus and other orbiviruses.

Supporting Information
Table S1 Clinical score table BTV animal trial. The clinical signs described in this table were based on findings in the field [31] as well as experimental data [15] and scored daily depending on severity from 0-3 points during 3-15 days post inoculation. (DOCX)