Identification of the Genome Segments of Bluetongue Virus Type 26/Type 1 Reassortants Influencing Horizontal Transmission in a Mouse Model

Bluetongue virus serotypes 1 to 24 are transmitted primarily by infected Culicoides midges, in which they also replicate. However, “atypical” BTV serotypes (BTV-25, -26, -27 and -28) have recently been identified that do not infect and replicate in adult Culicoides, or a Culicoides derived cell line (KC cells). These atypical viruses are transmitted horizontally by direct contact between infected and susceptible hosts (primarily small ruminants) causing only mild clinical signs, although the exact transmission mechanisms involved have yet to be determined. We used reverse genetics to generate a strain of BTV-1 (BTV-1 RGC7) which is less virulent, infecting IFNAR(−/−) mice without killing them. Reassortant viruses were also engineered, using the BTV-1 RGC7 genetic backbone, containing individual genome segments derived from BTV-26. These reassortant viruses were used to explore the genetic control of horizontal transmission (HT) in the IFNAR(−/−) mouse model. Previous studies showed that genome segments 1, 2 and 3 restrict infection of Culicoides cells, along with a minor role for segment 7. The current study demonstrates that genome segments 2, 5 and 10 of BTV-26 (coding for proteins VP2, NS1 and NS3/NS3a/NS5, respectively) are individually sufficient to promote HT.


Introduction
Bluetongue is an arboviral disease of ruminants, caused by Bluetongue virus (BTV), the type species of genus Orbivirus. Orbiviruses are transmitted between susceptible hosts by the bites of vector competent, hematophagous arthropods, although vectors have not been identified for all orbiviruses [1]. Most strains of BTV are transmitted between their ruminant hosts via the bites of infected adult female midges (Culicoides spp.). They can infect a number of competent Culicoides species [2], including colony derived adult Culicoides sonorensis (via an oral or intrathoracic route), or a C. sonorensis derived cell line (KC cells).
To date, more than 36 distinct serotypes of BTV have been isolated from ruminants including sheep, cattle and goats [3]. The BTV genome is composed of ten segments of double stranded RNA. BTV serotype is determined by the virus outer capsid and cell attachment protein VP2 (encoded by genome segment 2), which elicits neutralising antibodies in infected vertebrate hosts [4][5][6]. Since 2008, several atypical BTV serotypes have been isolated (primarily from small ruminants, including goats or sheep) that do not with the viruses at a multiplicity of infection (MOI) of 0.1 PFU/cell until full cell lysis was observed (72 h post-infection). The viruses were grown once in BSR cells, then titrated in BSR cells by a plaque assay, as previously described [4,23].

Extraction and Analysis of RNA
Infected BSR cells showing full cytopathic effect (CPE), were pelleted by centrifugation at 2000× g for 10 min at 4 • C. The pellet was dissolved in 1 mL of TRIzol by vigorous shaking, followed by addition of 200 µL of chloroform and further shaking for 1 min. For RNA extraction from mouse blood, 10 µL of blood were dissolved in 1 mL of TRIzol by vigorous shaking [4]. Tubes were left on ice for 30 min prior to centrifugation at 12,000× g for 10 min at 4 • C. The supernatant was collected and mixed vigorously with an equal volume of isopropanol, incubated at −20 • C for 1 h and centrifuged at 18,000× g for 10 min. The supernatant was discarded, and the pellet washed with 75% ethanol. Excess ethanol was removed, and the pellet left to dry. The pellet was dissolved in 50 µL of nuclease-free water.
RNA extracted from BSR cells was subjected to further purification by precipitation in 2 M (final concentration) lithium chloride over night at 8 • C as previously described [24]. Lithium chloride precipitated dsRNA was dissolved in 100 µL of RNase-free water and treated with an equal volume of phenol/chloroform/isoamyl alcohol (25:24:1, Sigma, St. Louis, MO, USA) before being analysed by polyacrylamide gel electrophoresis (PAGE).

PAGE Analysis
PAGE analysis was performed as previously described [25]. Briefly, a 7.5% polyacrylamide gel was prepared in tris/glycine buffer. A dsRNA solution (250 ng/µL) was mixed with an equal volume of Laemmli's 2× buffer [26]. The gel was run at 25 mA for 1 h and 45 min, followed by staining in ethidium bromide.

Preparation of Full-Length cDNAs from dsRNA
Full-length cDNAs were prepared using the FLAC method [27]. Briefly, an anchor primer (GACCTCTGAGGATTCTAAAC/iSp9/TCCAGTTTAGAATCC-OH3 ) was ligated to the 3 ends of the dsRNA. Ligations were carried out in a volume of 20 µL containing 10 µL of total RNA extract, 1 µg (in 1 µL) of anchor primer and 10 units of RNA ligase (New England Bio Labs, Ipswich, MA, USA). The reaction was incubated at 10 • C overnight. Excess non-ligated anchor primer was removed by electrophoresis on 1% agarose gel in TAE buffer. The total dsRNA was recovered from the gel in a volume of 15 µL using RNaid kit, as previously described [24].
Prior to reverse transcription (RT), the dsRNA was heat denatured at 99 • C for 3 min. The RT reaction contained 11 µL of heat denatured dsRNA (1 µg), 2 µL of 10 mM dNTP mix and 4 µL of AMV RT buffer and 1 µL (21 units) of AMV reverse transcriptase (Roche). The reaction was incubated at 37 • C for 1 h, allowing the anchor primer to self-prime, generating a full-length cDNA copies of each dsRNA segment, which were used for PCR.
Full-length cDNAs were PCR amplified using KOD Hot Start DNA Polymerase as described by the manufacturer in presence of primer 5-15-1(5 -GAGGGATCCAGTTTAGAAT CCTCAGAGGTC-3 ).

Reverse Engineering of BTV-1
Forward and reverse primers were designed from the sequences of BTV-1RSArrrr/01 for each of the 10 genome segments. All forward primers included the minimal T7 promoter sequence at their 5 end, (TAATACGACTCACTATA) immediately preceding the viral 5 termini. Forward and reverse primers (Table S1) were designed to amplify each full-length genome segment of BTV-1RSArrrr/01. PCR amplifications were performed using KOD Hot Start DNA Polymerase, as described by the manufacturer, in presence of primers designed for each genome segment. PCR products were analysed by agarose gel electrophoresis and purified using Qiagen PCR purification columns as described by the manufacturer. Capped mRNAs were generated from linear PCR products by in-vitro transcription using the mMessage mMachine T7 Ultra Kit (Life Technologies, Carlsbad, CA, USA) as previously described [18] and purified using the MEGAclear transcription cleanup kit (Ambion, Austin, TX, USA). Figure S1 shows a schematic representation of the PCR amplification and transcription reaction using the PCR products.
BSR cells grown to 80-90% confluence in 6 well plates were lipofected (lipofectamine 2000) with a mixture consisting of 500 ng from each mRNA transcripts corresponding to Seg-1, Seg-2, Seg-3, Seg-4, Seg-5, Seg-8 and Seg-9 of BTV-1. At 16 h post-lipofection, a second set of transcripts (500 ng each) corresponding to all 10 genome segments of BTV-1 were again lipofected into the same wells. The cells were incubated at 37 • C for up to 5 days and buffered if necessary, using 7.5% NaHCO 3 to avoid acidification of the culture medium. On day 5 cells were scraped from the wells, passaged into T25 cell culture flasks, incubated at 37 • C and monitored for CPE. On day 5 cells were scraped and centrifuged. The resulting cell pellets were resuspended in calcium and magnesium-free PBS, and subjected to 10 strokes of Dounce homogenisation, before treating with an equal volume of Vertrel-XF, as previously described [28]. The supernatant was used for plaque purification (twice), as previously described [4,23]. Plaques were collected and the virus amplified once in BSR cells before being assessed in IFNAR (−/−) mice. A clone (BTV-1RG C7 ) which induced milder clinical signs in IFNAR (−/−) mice was selected for reverse genetic studies to generate reassortants of BTV-26 in the BTV-1 backbone.

Reverse Engineering of Reassortants Containing BTV-26 Genome Segments in the BTV-1 Genetic Backbone
Genome segments of the selected BTV-1 clone were converted into cDNA using FLAC as described above. Forward and reverse primers containing enzyme restriction sites were used to facilitate cloning of cDNA copies of each BTV-1 genome segment into plasmids, as shown in Table 1. All forward primers included the minimal T7 promoter sequence at their 5 end, (TAATACGACTCACTATA) immediately preceding the viral 5 termini. Plasmids and PCR products, digested with appropriate restriction enzymes (as described in Table 1), were ligated using T4 DNA ligase and used to transform chemically competent Stbl2™ (Invitrogen, Waltham, MA, USA) bacteria. Plasmid clones were purified from colonies grown in liquid cultures using the Qiaprep miniprep kit and commercially sequenced using the BigDye Terminator v3.1 Cycle Sequencing Kit (Invitrogen).
Plasmids were linearised using the appropriate restriction enzyme (Table 1), then purified using Qiaquick purification kit and used in transcription reactions to generate full length T7 transcripts. The primers used to generate full-length linear PCR products corresponding to each of the 10 individual genome segments of BTV-26 are shown in Table 2. All forward primers included the minimal T7 promoter sequence at their 5 end, (TAATACGACTCACTATA) immediately preceding the viral 5 termini. PCR amplifications were performed using KOD Hot Start DNA Polymerase as described by the manufacturer, in the presence of primers designed for each genome segment. PCR products were analysed by agarose gel electrophoresis and purified using Qiagen PCR purification columns, as described by the manufacturer. Table 1. Primers designed for PCR amplification of full-length cDNAs of genome segments 1-10 of BTV-1, for cloning into the pGEX-4T-2 plasmid. The minimal T7 promoter in the forward primer is shown in bold. Restriction enzyme used to facilitate cloning into the pGEX-4T-2 plasmid are underlined. Enzyme used to linearise the plasmid before T7 transcription is shown in the column to right of the  Table 2. Primers designed for PCR amplification of full-length cDNAs of genome segments 1-10 of BTV-26. The minimal T7 promoter in the forward primer is shown in bold.

Virus/Segment
Primer Sequence 5 →3 BTV26-S1T7FOR TAATACGACTCACTATAGTTAAAATGCAATGGTCGCGATTAC BTV26-S10T7FOR TAATACGACTCACTATAGTTAAAAAGTGTCGCTGCCATG BTV26-S10REV GTAAGTGTGTAGTGCCGCATA Capped mRNAs were generated from PCR products and purified as described above. A schematic representation of cloning the full-length cDNAs of BTV-1 and T7 transcription is shown in Figure S2. Reassortants were generated by lipofecting monolayers of BSR cells grown in 6 well plates as described above, using a mixture of mRNA transcripts of BTV-1 and BTV-26 in the desired combination ( Figure S3). Reassortants were collected, treated with Vertrel XF, plaque purified once on BSR cells and propagated once in BSR cells. This passage was either stored at -80 • C or used to extract dsRNA for PAGE analysis.
Six to fourteen weeks old female IFNAR (−/−) mice were used throughout the experiments. The mice were held in plastic cages with wood-chip bedding, fed ad libitum and kept in standardised conditions (21 • C, 12:12 light/dark cycle) at an SPF animal facility.

Infection of IFNAR (−/−) with Reverse Engineered Reassortant Viruses and Assessment of Horizontal Transmission Potential
Two groups of 4 female mice (6-8 weeks old) were used to assess the potential for horizontal transmission (HT) of BTV-1RG C7 or BTV-26 ( Figure S4). In each group, two mice were inoculated intraperitoneally, either with 10 2 PFU of BTV-1RG C7 or BTV-26/mouse then kept in a cage in direct contact with 2 non-inoculated mice. Blood samples (10-30 µL/mouse) were collected on days 0, 4, 8 and 11. The potential for HT of each reassortant virus was also assessed using the same settings. Experiments assessing HT potential were repeated twice. The HT potential of each of the 10 reassortant viruses was also evaluated in a similar manner. This experiment was repeated once with older animals (10-14 weeks old). Animals were euthanized on day 15 post-inoculation. The WT BTV-1RSArrrr/01 caused clinical signs in mice, ultimately killing them by day 5-6 pi. This would significantly reduce the time that uninoculated mice could be exposed 'in contact' to the infected mice. In contrast, BTV-1RG C7 caused only minor signs of infection, allowing contact with uninfected mice for the duration of the experiment (up to 15 days) providing more opportunity for horizontal transmission. The BTV-1RG C7 backbone was therefore used to reverse engineer reassortants of BTV-26.

Rescue of BTV1-RG and Plaque Purification of Clones
Full-length PCR products of the individual genome segments of BTV-1RSArrrr/01 were successfully generated using the FLAC method ( Figure 1). T7 transcripts generated with the mMessage mMachine T7 Ultra Kit were transfected into BSR cells, resulting in the rescue of BTV-1 by reverse genetics. Vertrel XF was used to release virus particles from cell debris [28,33,34]. Following Vertrel treatment and plaque purification, ten clones were collected, and plaque purified twice in BSR cells. These clones were amplified once in BSR cells and used to infect IFNAR (−/−) mice. Nine clones induced typical signs of infection in inoculated mice, which died by day 6 pi. However, clone 7, which was designated BTV-1RG C7 , induced only minor clinical signs and was used for further reverse genetic studies. The dsRNA from this clone was converted into cDNA, PCR amplified and cloned by restriction digestion/ligation into pGEX-4T-2 plasmid. Plasmid clones corresponding to the 10 genome segments of BTV-1RG C7 were purified by miniprep ( Figure 2).

The Full-Length Sequence of BTV-1RG C7 Genome
The plasmid clones which served as templates for linear PCR products used to generate T7 transcripts, were sequenced using the BigDye Terminator v3.1 Cycle Sequencing Kit. Nucleotide (Nt) and/or amino acid (AA) substitutions were identified in genome segments 1, 2, 3, 4, 5, 6 and 8 of BTV-1RG C7 , compared to the consensus reference sequence of the wild type BTV-1 [BTV-1RSArrrr/01] ( Table 3). No substitutions were detected in segments 7, 9 or 10. As shown in Table 3, most of the depicted nucleotide substitutions are synonymous and thus do not incur changes in the amino acid sequences. However, some nucleotide substitutions incurred amino acid changes including radical or conservative substitutions in the VP2, VP3, VP5 and NS1 (Table 3) as defined by the amino acid exchangeability matrices [35]. There were no insertions or deletions in any of the sequences. The plaque purified clone 3 of BTV-1RG (BTV-1RG C3 , a clone with a virulent phenotype) was sequenced in order to attempt identifying specific residues which might contribute to the attenuated phenotype of BTV-1RG C7 . The sequence of the full-length genome of BTV-1RG C3 , was almost identical to that of the parental BTV-1RSArrrr/01. Only two synonymous nucleotide substitutions were identified in genome segment 1 (G3113A) and 6 (A310T) of this clone, and these two substitutions were also detected in BTV-1RG C7 . This analysis does not allow us to clearly identify which nucleotide and/or amino acid changes contribute to the attenuated phenotype of BTV-1RG C7 . Additional work will be necessary to reverse engineer the various depicted substitutions into the original sequence of BTV-1RSArrr/01.

Generation of Reassortants
Linearised plasmid clones of BTV-1RG C7 were obtained by digesting the plasmids with BsmBI or BsaI enzymes (Table 1). Full-length cDNAs of BTV-26, prepared using FLAC, were PCR amplified with the primers described in Table 2. Linearised plasmids and full-length PCR products (shown in Figure 3) were used for synthesis of T7 RNA transcripts of the BTV-1 or BTV-26 genomes segments. All reassortants were generated by lipofecting T7 transcripts into BSR cells according to the combinations shown in Figure S3.

Horizontal Transmission of BTV-26
In horizontal transmission experiments, the two mice inoculated with BTV-26 showed minor clinical signs of infection with lacrimation on day 4-5 but had recovered by day 6-7. The non-inoculated in-contact mice also showed minor signs of lacrimation on day 8-9 but had recovered by day 11-12. On day 4, both inoculated mice were positive for BTV RNA (Ct values~32-36). On day 8, both non-inoculated, in-contact mice were also positive for BTV RNA (Ct values of~36-38) and they remained positive when retested on day 11 ( Figure 5), demonstrating horizontal transmission of BTV-26. The non-inoculated mice in HT experiments with BTV-1RG C7 did not become infected ( Figure 5). This set of experimental results confirms the usefulness of IFNAR (−/−) mice as a model to study horizontal transmission of BTV-26.

Identification of the BTV-26 Genome Segments That Support Horizontal Transmission
Minor clinical signs (transient lacrimation and scruffy fur) were observed on days 4-5 during HT experiments, in mice inoculated with 10 2 PFU of the BTV-26/ BTV-1RG C7 reassortants. All of the reassortants replicated in the inoculated mice (Table 4), with Ct values on day 4 ranging from 22 to 37, and all of the mice had recovered by day 11-12. Clinical signs consistent with horizontal transmission were also observed in the noninoculated mice in the group receiving reassortants RGBTV1:5 (26), or RGBTV1:10 (26). For RGBTV1:5 (26) , both in-contact mice showed, evidence of viraemia on day 8, while one of these two mice tested negative on day 11 ( Figure 6). For RGBTV1:10 (26) (NS3/NS5 of BTV-26), both in-contact mice were positive on day 8 and were retested positive on day 11 of the experiment (Figure 7). Table 4. Ct values of the RNA extracted from mice inoculated with the reassortants of BTV-26 in the BTV-1RG C7 backbone and from non-inoculated in-contact mice. Ct values ranged from 23 to 37. Positive in-contact mice were identified in the groups inoculated with RGBTV1:2,6,7 (26) , RGBTV1:5 (26) or RGBTV1:10 (26) .   These results indicate that Seg-5/NS1 or Seg10/NS3-NS5 of BTV-26 are both individually sufficient to permit horizontal transmission of these reassortants strains and are therefore likely to influence horizontal transmission of BTV-26, at least in the mouse model. On one occasion, one in-contact mouse was also positive (minor clinical signs and real-time RT-PCR) in the group inoculated with reassortant RGBTV1:2,6,7 (26) (expressing VP2, VP5 and VP7 of BTV-26) (Figure 8), indicating that one or more, or possibly this combination of three BTV-26 genome segments may also play some role in supporting HT of these reassortants. Figure 8. Real-time RT-PCR amplification curves of RNA extracts from blood samples of mice inoculated with BTV-1:2,6,7 (26) . On day 4 both inoculated mice were positive by real-time RT-PCR. On day 8 and 11 one in-contact mouse was positive for BTV RNA.

Discussion
BTV is an arbovirus that is usually transmitted between individual susceptible hosts by competent species of Culicoides biting midges [1,6,36]. Horizontal transmission of vectored BTVs has been linked to contaminated semen from infected animals. Vertical transmission of BTV was first described with attenuated vaccine strains belonging to serotypes 10, 11, 13 and 17 [37]. After BTV-8 emerged in northern-Europe in 2006, it was rapidly observed to cross the placental barrier [38][39][40][41]. However, experimental infection studies with gestating animals that were intended to confirm vertical transmission also revealed potential for the direct transmission of other specific conventional BTV serotypes. For instance, direct transmission of BTV-2 was observed from inoculated to non-inoculated control gestating ewes [11]. The northern-European strain of BTV-8 was shown to infect bovine foetuses in utero and infection of heifers also occurred after ingestion of contaminated tissues, such as placenta [38,42], confirming horizontal transmission by an oral route.
BTV can infect wild carnivores preying on meat/organs of infected animals in Africa [43] and lynx that were fed meat contaminated with the northern-European strain of BTV-8, became infected and died [44]. It has been suggested that this finding may be linked to the use of live attenuated vaccines in the field. This hypothesis is supported by other findings, where vaccination of gestating dogs with a BTV-11vac contaminated multivalent modified live canine distemper virus, adenovirus and parvovirus vaccine lead to abortion [45]. The general hypothesis that transplacental transmission of BTV is linked to live attenuated vaccine strains is now largely accepted [12], although historically these vaccines were widely used and persisted in the field, suggesting that this may now be a more generalized characteristic of BTV field strains.
Horizontal transmission of the northern-European BTV-8, or the South-African BTV-1 vaccine strain from infected to in-contact gestating ewes, could have occurred by ingestion of saliva/blood contaminated feed or water [39,46]. However, in a recent study, UK and French isolates of BTV-8 (UKG2007 and FRA2017 respectively) did not transmit horizontally from infected to non-infected in-contact ewes (non-gestating) [47]. Other studies also showed that strains of BTV-4 do not transmit horizontally in sheep, goat or calves [48].
BTV-25 is the first discovered atypical BTV serotype but was not able to replicate in either mammalian or insect cell cultures and may therefore be different to the other novel BTV serotypes. BTV-25 was not detected in milk, urine, nasal or ocular secretion, or in the faeces of infected goats [49]. These observations suggest that the 'atypical' BTVs can spread efficiently by direct contact, in contrast with the low efficiency contact transmission of conventional vectored BTVs.
Previous studies which reported possible or potential direct transmission of vectored BTVs in ruminants were all conducted with gestating animals, which showed clinical signs of infection. However, the horizontally transmitted BTV-25 caused only limited clinical signs and was not found in biological fluids, including milk, urine, nasal and ocular secretion, or in faeces of infected goats, suggesting that the opportunities for the virus to contaminate animal feed or water would be very limited [49].
Even in the absence of clinical signs, direct contact transmission of BTV-26 or BTV-27 from experimentally infected to in-contact goats, is very efficient [14,15]. It is therefore important to understand if the state of gestation exerts an influence on the way that vectored BTV serotypes (or at least those which can be transmitted horizontally) could be secreted/excreted. The possibility that in-utero infection of the foetus leads to losses/escape of intrauterine fluids contaminated with the virus, promoting oral contamination of incontact animals needs further clarification.
The loss of transmissibility of atypical BTV serotypes by Culicoides vectors [17,18] has been linked to 3 genome segments, each encoding one structural protein: (i) Seg-1 encoding the viral RNA dependent RNA polymerase, VP1, (ii) Seg-2 encoding outer-capsid protein VP2 that binds to cell surface receptors and (iii) Seg-3 encoding the innermost sub-core shell protein VP3(T2) of the virus particle, which is in direct contact with the dsRNA genome. The outer-core VP7 protein (encoded by Seg-7) seemed to play a lesser role in reducing replication of BTV-26 when a mono-reassortant of BTV-26 Seg-7 in a BTV-1 backbone, was used to infect Culicoides cells [18].
Our reverse genetics studies led to isolation of a BTV-1 clone (designated BTV-1RG C7 ), derived from the BTV1-RSArrrr/01, which caused less severe clinical signs of infection in IFNAR (−/−) mice, ultimately with a full recovery. Sequence analysis showed several nucleotide substitutions in genome segments 1, 2, 3, 4, 5, 6 and 8 of BTV-1RG C7 , some of which resulted in amino acid changes. It is therefore likely that the attenuated phenotype of clone BTV-1RG C7 is linked to one or more of these mutations.
In the current study, our assessment of genome segments which influence or support horizontal transmission of an atypical BTV, strongly suggests that genome segments 2, 5 and 10 of BTV-26 can influence vector independent direct horizontal transmission in a mouse model. These genome segments encode outer capsid protein VP2 (Seg-2), nonstructural protein NS1 (Seg-5) and non-structural proteins NS3 and NS5 (Seg-10). VP2 of vectored BTV binds efficiently to insect cell surface while VP2 of non-vectored BTVs fails to do so [17]. VP2 of both groups is functional in mammalian cell cultures. The results of our study suggest that segment 2 (encoding VP2) of a non-vectored BTV may also exert some influence on horizontal, in-contact transmission. However, cellular receptors for BTV VP2 may differ in terms of binding capacity and affinity in the different vertebrate species.
Studies conducted with other viruses including SARS-CoV-2 have shown that lentiviruses pseudo-typed with the variants of the S2 spike protein had different capacity to interact with the well characterised SARS-CoV-2 ACE2 receptor (Angiotensin Converting Enzyme 2: ACE2) of a range of animals. ACE2 of mice came out as the one with poor binding properties [50]. BTV receptors are not fully characterised, and BTV may use several different entry mechanisms for structurally different virus particles, and the two different serotypes that have been studied appear to use different cell entry mechanisms [51,52] although the glycophorin of sheep red blood cells appear to bind the virus with high affinity [53]. There are inherent differences between the VP2 of vectored and non-vectored BTVs in term of their sequences [9]. Identification of specific receptors which bind the outer-capsid protein VP2 of vector independent horizontally transmitted BTVs may be important to fully understand changes to the route of transmission used by atypical BTV serotypes.
As observed with other viruses [54,55], adaptation to specific receptors, could provide an evolutionary advantage for the atypical BTV serotypes (which can no longer replicate in the insect vector), allowing them to directly infect their hosts. Other members of the family Reoviridae including the orthoreoviruses and rotaviruses also infect mammals, via oral-fecal and/or respiratory routes, exhibiting tissue tropism and specificities for different receptors in a strain/serotype dependent manner [56,57]. For instance, Rotaviruses use histo-blood group antigens as receptors, but several other receptors have been proposed for rotavirus cell-entry including heat shock protein 70, integrins, tight junction protein ZO-1, junction adhesion molecule A and occludin [58][59][60][61][62][63][64].
The role of BTV genome segments 5 and 10 in supporting horizontal transmission could relate to the role played by these proteins in cell exit [65], which may determine the availability of virus particles for transmission to other hosts. When NS1 is abundant, the virus leaves the cells by lysis as it is the case usually in mammalian cells during advanced stages of infection. When NS3 is abundant, it favours release of the virus from insect cells [66][67][68]. The virus can also leave the cell in secretory endosomes [69]. Although NS1 is a non-structural protein, earlier studies showed that it is present on the temporary envelope of viral particles leaving the cells [65]. Beyond cell exit, NS1 plays the role of a positive regulator of viral protein synthesis in infected cells. In addition to its role in cell exit [65], NS3 is involved in interferon antagonism [70].
It is noteworthy that BTV-28, which is pathogenic to sheep [16,20], is a reassortant between vectored and atypical bluetongue viruses. BTV-28 does not grow in Culicoides cells and can be transmitted directly from infected to non-infected sheep. It has all three genome segments 2, 5 and 10 related to those of BTV-26. It is therefore possible that these three genome segments of atypical BTV serotypes collectively play a role in entry and exit of tissues which vectored BTVs do not traditionally infect (or infect with lower efficiency), thus facilitating vector independent direct horizontal transmission of an orbivirus.

Conclusions
This study indicates that genome segments 5 and 10 can individually support the horizontal, in-contact transmission of an atypical bluetongue virus (BTV-26) in a mouse model. These are different genome segments to those previously identified as restricting Culicoides infection by BTV-26 (segments 1, 2 and 3, with a lesser effect by segment 7 [18]). There is also some indication that genome segment 2 might exert some influence over horizontal transmission of BTV-26 in mice. Studies in the natural hosts of BTV will be necessary to determine whether these observations are also true in ruminants.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/v13112208/s1, Figure S1: A schematic representation for generation of transcripts from linear PCR products representing the full-length cDNA copies of BTV genome segments (ARCA: anti-reverse cap analogue), Figure S2: A schematic representation for generating transcripts from restriction enzyme linearised plasmids containing the full-length cDNA copies of BTV genome segments (ARCA: anti-reverse cap analogue), Figure S3: Schematic representation showing RNA combinations used to generate reassortants of BTV-26 in the BTV-1 backbone. The reassortants were designated RGBTV1:X (26) (e.g., RGBTV1:1 (26) , RGBTV1:5 (26) ), where RGBTV1 refers to the genetic backbone of BTV1-RG C7 , and X refers to the specific genome segment of BTV-26 introduced into the reassortant, Figure S4: A schematic representation of the horizontal transmission experiment in IFNAR (−/−) mice. Red-coloured mice indicate where replication of a given BTV was detected by real-time RT-PCR, Table S1: Primers designed for PCR amplification of full-length cDNAs of genome segments 1-10 of BTV-1. The minimal T7 promoter in the forward primer is shown in bold.