Evidence of reduced viremia, pathogenicity and vector competence in a re‐emerging European strain of bluetongue virus serotype 8 in sheep

Summary The outbreak of bluetongue virus (BTV) serotype 8 (BTV‐8) during 2006–2009 in Europe was the most costly epidemic of the virus in recorded history. In 2015, a BTV‐8 strain re‐emerged in France which has continued to circulate since then. To examine anecdotal reports of reduced pathogenicity and transmission efficiency, we investigated the infection kinetics of a 2007 UK BTV‐8 strain alongside the re‐emerging BTV‐8 strain isolated from France in 2017. Two groups of eight BTV‐naïve British mule sheep were inoculated with 5.75 log10TCID50/ml of either BTV‐8 strain. BTV RNA was detected by 2 dpi in both groups with peak viraemia occurring between 5–9 dpi. A significantly greater amount of BTV RNA was detected in sheep infected with the 2007 strain (6.0–8.8 log10 genome copies/ml) than the re‐emerging BTV‐8 strain (2.9–7.9 log10 genome copies/ml). All infected sheep developed BTV‐specific antibodies by 9 dpi. BTV was isolated from 2 dpi to 12 dpi for 2007 BTV‐8‐inoculated sheep and from 5 to 10 dpi for sheep inoculated with the remerging BTV‐8. In Culicoides sonorensis feeding on the sheep over the period 7–12 dpi, vector competence was significantly higher for the 2007 strain than the re‐emerging strain. Both the proportion of animals showing moderate (as opposed to mild or no) clinical disease (6/8 vs. 1/8) and the overall clinical scores (median 5.25 vs. 3) were significantly higher in sheep infected with the 2007 strain, compared to those infected with the re‐emerging strain. However, one sheep infected with the re‐emerging strain was euthanized at 16 dpi having developed severe lameness. This highlights the potential of the re‐emerging BTV‐8 to still cause illness in naïve ruminants with concurrent costs to the livestock industry.


| INTRODUCTION
Bluetongue (BT) is an infectious haemorrhagic disease of ruminants caused by the bluetongue virus (BTV) which is transmitted via Culicoides biting midges (Carpenter, Groschup, Garros, Felippe-Bauer, & Purse, 2013). BTV is the type species of the genus Orbivirus within the family Reoviridae and is a serologically and genetically diverse virus. The BTV genome is comprised of 10 double-stranded RNA segments encoding several structural and non-structural proteins.
Since 1998, a number of BTV serotypes have caused both sporadic and widespread incursions into the EU (Belbis et al., 2017).
In August 2006, a BTV serotype 8 (BTV-8) strain of sub-Saharan origin (Maan et al., 2008)  In August 2015, BTV-8 re-emerged in France and subsequently spread throughout the entire country . How BTV-8 persisted in areas that were thought to have been free of virus transmission remains unknown, although it has been recently proposed that low-level circulation of BTV-8 occurred in France prior to the detection in 2015 (Courtejoie et al., 2018). Importantly however, anecdotal observation suggested that changes had occurred in the epidemiology of the re-emerging strain. Whereas the [2007][2008][2009] BTV-8 strain caused widespread clinical signs in cattle and sheep (Elbers, Spek, & Rijn, 2009;Zanella et al., 2013), the re-emerging BTV-8 strain has thus far only caused mild clinical illness . In addition, the rate of spread of the virus in France appeared slower than the original BTV-8 strain. At the time of the BTV-8 re-emergence in 2015, it had been estimated that the ruminant herd immunity in France was 18% . (1), 4 (1), 8 (1), 9 (3) and 10 (1). The high degree of amino acid similarity between the 2007 and 2015 strains underpins the hypothesis that this is not a new introduction of BTV-8 .
In this study we directly compare the infection of sheep with the original and re-emerging European BTV-8 strains by assessing viraemia, clinical signs, antibody production and transmission to Culicoides biting midges. By using British mule sheep, representative of the UK flock from a region not originally exposed to BTV-8 infection, the aim is to understand if the changes in epidemiology and pathogenicity observed in France will also occur in regions beyond those affected by outbreaks from 2006 to 2009. The findings from this study will inform response in both the UK and other countries previously unaffected by BTV-8 and allow for evidence-based surveillance and control measures to be implemented.   2,3,5,6,7,8,9,10,12,14,16,19, and 20/21 dpi. Sheep body temperatures were recorded daily and clinical scoring was performed throughout as described previously (Darpel et al., 2007).

| Vector competence
Culicoides sonorensis from a colony maintained at The Pirbright Institute (PIRB-s-3 strain) were fed on two sheep from each group at 6 dpi (UKG2007) or 7 dpi (FRA2017) to coincide with determined peak viraemia as described previously (Baylis, O'Connell, & Mellor, 2008). Engorged individuals were incubated at 25°C for 8 days and surviving individuals were homogenized in GMEM using a tissue lyser (Qiagen) and made up to a final volume of 1 ml per sample (Veronesi et al., 2013). Culicoides samples were analyzed in pools of eight individuals for BTV genome detection by using RT-qPCR as described below. Pools that reported a cycle threshold (C T value) less than that of the original blood meal in midges harvested on the day of feeding were considered to contain one or more individuals competent for BTV transmission.   using the Express One-Step qRT-PCR kit (ThermoFisher) on an Applied Biosystems 7500 Fast instrument (ThermoFisher). A log-dilution series of the plasmid (1 × 10 0 -1 × 10 6 copies per μl) was included in triplicate on each RT-qPCR run.

|
BTV RNA copies were determined by comparing sample C T values to the standard curves and expressed as log 10 genome copies/ml.

| Serological analyses
Whole blood samples were centrifuged at 3,000 g for 5 min and the serum was decanted and stored at +4°C until analysis. BTV antibodies were detected using the ID Screen ® Bluetongue Competition kit (ID Vet, Grabels, France) in accordance with the manufacturer's instructions. The serum neutralization test (SNT) against BTV-8 was performed as described previously (Batten et al., 2012).

| Virus isolation
EDTA blood cells were washed 3× with PBS and sonicated as described in the World Organisation for Animal Health (OIE) manual (Savini, 2014). KC cells were inoculated with 100 μl of washed blood and incubated at 26°C in Schneiders media (1% Amphotericin B, 1% Penicillin/Streptomycin and 10% FBS). Media was replenished 24 hr post infection and cells were incubated for a further 6 days. Following the 7-day incubation, cells were harvested, centrifuged at 3,000 g for 5 min and the supernatant tested for BTV using RT-qPCR as described above. Virus was considered to have been isolated if the BTV C T value of the harvested material was 3-C T values less than that of the original EDTA-blood inoculum.

| Statistical analysis
Body temperatures and clinical scores were analyzed by comparing the maximum values and the dpi at which they occurred for each strain using the Wilcoxon rank-sum (also known as Mann-Whitney Two sheep were euthanized on welfare grounds during the study, one in each group. The sheep from the UKG2007-inoculated group was euthanized at 9 dpi for reaching the humane endpoint (at the end of the moderate clinical spectrum) of the moderate study protocol. In this sheep, typical haemorrhagic lesions on mucosal membranes, coronary band, lymph nodes, pulmonary artery as well as the epithelium of the reticulum and omasum were observed during necropsy. The sheep from the FRA2017-inoculated group developed acute lameness on all four feet late in infection and was euthanized on humane grounds at 17 dpi.
The two transmission control sheep remained clinically unaffected other than exhibiting a mild nasal discharge over a 2-day period during the experiment. Figure 2 shows the BTV concentrations in EDTA blood over the study period. In five UKG2007-inoculated sheep and in three FRA2017inoculated sheep BTV RNA was detected by using RT-qPCR at 2 dpi.

| Molecular analyses
BTV RNA was detected in all other inoculated sheep at 5 dpi with the exception of Sheep 17 (that had been inoculated with FRA2017).
Peak viraemia occurred earlier for UKG2007-inoculated sheep (between 5 and 6 dpi) than for FRA2017-inoculated sheep (5-9 dpi), but this difference was not statistically significant (p = 0.11). At peak viraemia, the BTV concentrations in blood (log 10 genome copies/ml) were significantly (p = 0.021) higher for UKG2007-inoculated sheep  higher on the day of Culicoides feeding than for those infected with FRA2017 (8.3 and 8.0 compared with 7.3 and 6.6, respectively). An increase in competence was associated with an increase in titre, but this was not statistically significant (estimate for coefficient, a 1 : 2.6; 95% CI: −0.1, 3.6).

| Virus isolation
Virus isolation from EDTA blood was performed for all inoculated sheep at 2, 5, 7, 10, 12, 14, 16 and 21 dpi. For UKG2007-inoculated sheep, virus was isolated at 2 dpi (3/8 sheep), 5 dpi (7/8 sheep), 7-10 dpi (8/8 sheep) until 12 dpi (1 sheep). For FRA2017-inoculated sheep, virus was isolated at 5-7 dpi (6/8 sheep) until 10 dpi (4/8 sheep). Interestingly, one of the sheep inoculated with the re-emerging BTV-8 strain did not develop detectable viraemia throughout the entire study, yet seroconverted and developed a neutralizing antibody response. This suggests that the sheep did receive the inoculum. Historically studies have previously reported that ruminants experimentally-inoculated with BTV did not develop detectable viraemia (in the presence or absence of seroconversion), although viraemia detection was based on virus isolation rather than the more sensitive use of real-time PCR (Baylis et al., 2008;Flanagan, Wilson, Trueman, & Shepherd, 1982;Parsonson et al., 1987). Other studies have further noted seroconversion of individual BTV-inoculated sheep or cattle either in the presence of very low detectable RNA levels by RT-qPCR (van der Sluijs et al., 2013) or in the absence of detectable viraemia, however, the latter was associated with either a low-titre (Baylis et al., 2008;Di Gialleonardo, Migliaccio, Teodori, & Savini, 2011) or a high-passage inoculum (Janowicz et al., 2015).

| DISCUSSION
Interestingly, a recent study also highlighted that a highly-passaged BTV-8 strain infected and was detectable in the skin and draining lymph nodes of inoculated sheep however did not result in detectable RNA levels in the systemic blood (Melzi et al., 2016). Overall all these studies, including the data presented here, demonstrate that to 2.5 years following vaccination (Batten, Edwards, & Oura, 2013) and it has been estimated that they can persist for 5-6 years following infection or vaccination . The sheep used in our study were at an advanced age (>7 years old), as reported by the farmer and by assessing the general condition of the sheep.   (Darpel et al., 2007;Di Gialleonardo et al., 2011;Martinelle et al., 2011). An experimental comparison of BTV-8 strain-specific virulence (similar to our study here) is likely to be more challenging in cattle.
Considering that morbidity rates in field conditions are likely to be higher than what we have reported here and coupled with animal movement restrictions, it is probable that future incursions of the reemerged BTV-8 will have an economic impact on the livestock industry. Therefore, appropriate surveillance and vaccination strategies should be considered in advance of the likely incursion of this re-emerged BTV-8 strain.