Emergence of a Novel Reassortant Strain of Bluetongue Serotype 6 in Israel, 2017: Clinical Manifestations of the Disease and Molecular Characterization

Reassortment contributes to the evolution of RNA viruses with segmented genomes, including Bluetongue virus (BTV). Recently, co-circulation of natural and vaccine BTV variants in Europe, and their ensuing reassortment, were proposed to promote appearance of novel European BTV strains, with potential implications for pathogenicity, spread and vaccination policies. Similarly, the geographical features of the Mediterranean basin, which spans over portions of three continents, may facilitate the appearance of clinically relevant reassortants via co-circulation of BTV strains of African, Asian and European origins. In August–October 2017, BTV serotype 6 (BTV-6) was identified in young animals exhibiting classical clinical signs of Bluetongue (BT) at Israeli sheep and cattle farms. Sequencing and pairwise analysis of this Israeli BTV-6 isolate revealed the closest sequence homology of its serotype-defining Segment 2 was with that of South African reference BTV-6 strain 5011 (93.88% identity). In contrast, the other viral segments showed highest homology (97.0%–99.47% identity) with BTV-3, -4 and -9 of Mediterranean and African origins. Specifically, four viral segments were nearly identical (99.13%–99.47%), with Tunisian and Italian BTV-3 strains (TUN2016 and SAD2018, correspondingly). Together, our data suggest that Mediterranean co-circulation and reassortment of BTV-3 and BTV-6 drove the emergence of a novel and virulent BTV-6 strain


Introduction
Bluetongue (BT) is an arthropode-born disease of domestic and wild ruminants caused by the Bluetongue virus (BTV) [1,2]. The agent and its associated diseases present wide global distribution [2]. Indeed, BT is one of the major infectious diseases of ruminants and is listed as a notifiable disease by the World Organization for Animal Health (OIE). Transmission between mammalian hosts is mainly carried out by competent Culicoides species [3][4][5][6]. In addition to Culicoides-mediated transmission, a limited repertoire of BTV serotypes has recently been characterized as being able to undergo transplacental

Pan-BTV qRT-PCR
RNA was extracted as previously described for EHDV [30]. A total of 693 field samples were included in the present study. These originated from 402 cattle whole blood EDTA samples, 35 spleens and 7 aborted cattle fetuses (four spleen, one brain, one placenta and one mixed sample, which included brain, spleen, liver and lung), 173 whole blood sheep EDTA samples, 41 spleens from dead sheep and 10 aborted sheep fetuses (seven spleen, two lung, two brain, and two mixed samples, which included brain, spleen, liver and lung). Whole blood samples from wild animals were obtained from camels (14 samples), elephants (3 samples); spleen samples were attained from 4 mountain gazelles, 5 roe deer, 3 Persian fallow deer, 9 Nubian ibexes, 1 alpaca; 2 aborted Persian fallow deer fetuses, 1 adult giraffe and 1 aborted giraffe fetus ( Table 2). Initial assessment of BTV was performed by using VetMAX™ BTV NS3 All Genotypes Kit (Applied Biosystems™, Thermo Fisher Scientific Inc., France), as described by the manufacturer (referred to hereafter as Pan-BTV qRT-PCR). RT-qPCR BTV positive samples (total of 166, Table 2) were inoculated onto 9-11 day old embryonated chicken eggs (ECE) by intravenous delivery, as previously described [30,31], with minor modifications. Briefly, tissue samples were prepared by the same method, which was described above, and supernatant from organ samples was filtrated using 0.45 µm filter (Starstedt, Germany) and stored at 4 • C until use. For isolating BTV from whole blood samples, red blood cells (RBC) were washed 3 times with PBS. Then, 100 µL of washed RBC were resuspended in 900 µL of double distillated water to induce hemolysis. A total of 100 µL from prepared samples were inoculated into each ECE (5 eggs per sample) and eggs were observed for 9 days. Dead ECE (between days 2-9), were homogenized. Supernatant from ECE homogenates was used for RNA extraction and was tested by pan-BTV RT-qPCR; positive in pan-BTV RT-qPCR ECE samples were subsequently passaged 3 or 4 times on baby hamster kidney cells (BHK-21) monolayers (until appearance of a cytopathic effect (CPE)).

BTV Serotype Identification
RNA from ECE homogenates was extracted using Invisorb Spin Virus RNA Mini Kit (STRATEC Molecular GmbH, Berlin, Germany) according to manufacturer's instructions. ECE homogenates which were found positive by Pan-BTV RT-qPCR were further assessed by RT-PCR for the typing of serotypes 2, 3, 4, 5, 8, 12, 15, 16, 24 and 28 (all with in house developed primer pairs), which are known to be present in Israel and neighboring countries, using One-Step RT-PCR kit (Qiagen, Hilden, Germany). Identification of BTV-6 isolates was performed with the following pair of primers (developed in house): 6VP2-124F 5 -TGTAACCCAAATTCCCACGAA-3 and 6VP2-1030R 5 -CAGAGGCGGCTATCATA-3 . Amplified fragments were subsequently sequenced.

Sequencing and Phylogenetic Analysis
Following three passages on BHK-21cells, the BTV strain (ISR-2095/3/17) was sequenced by NGS at Hy Laboratories Ltd., Rehovot, Israel. Sequence gaps were filled following the performance of conventional RT-PCR, employing in house designed primer-pairs. cDNA fragments were purified using MEGAquick-spin™ Total Fragment DNA Purification Kit (iNtRON Biotechnology, Gyeonggi-do, South Korea) and standard Sanger sequencing was performed on ABI 3730xl DNA Analyzer (Hy Laboratories Ltd., Rehovot, Israel).
Nucleotide sequences were assembled and nucleotide (nt) and amino acid (aa) sequences were aligned and pairwise compared using Geneious (version 9.0.5; Biomatters, Auckland, New Zealand). Phylogenetic trees were constructed using Mega 7.1 [32].
The ten segment codding regions of Israeli BTV-6 ISR-2095/17 isolate were mostly sequenced

Clinical Signs in Sheep and Cattle Naturally Infected with BTV-6
From the end of August until mid-October 2017, BTV-6 was identified in clinically affected sheep (n = 6) and cattle milking and fattening farms (n = 5) located in southern, central and northern parts of the county (Figure 1, Tables 3 and 4). In Israeli sheep, BTV-6 was observed mostly in young animals ranging from 4 to 20 months of age, while clinical signs in adult animals were observed only in two farm (farms 5 and 10; Tables 3 and 4). In young sheep, the major clinical sings included pyrexia, skin hyperemia, lameness, rejection of moving, stiffness in leg and back muscles, bloody nasal discharge, recumbency and breath abnormalities; in adult milking ewes clinical signs included milk reduction, Viruses 2019, 11, 633 6 of 15 heavy hyperemia of skin (udder, face), light perinasal and perioral edema. Morbidity in flocks ranged 17%-34% and was observed mostly in lambs. Mortality greatly varied between flocks (from 2% to 83% of identified cases) and appeared to depend on general flock welfare and on symptomatic treatment of ill animals. In cattle, clinical signs were observed in young animals only (ranging from three month to two years old). Inappetence, cachexia, dyspnea and cough were observed in calves for fattening. In addition, field post mortem examination revealed pneumonia in two dead calves. Udder edema, as well as symmetrical and bilateral edema throughout the length of the limbs, were observed in milking heifers, lasting for two weeks (Table 3).
Viruses 2019, 11, x FOR PEER REVIEW 6 of 15 mostly in young animals ranging from 4 to 20 months of age, while clinical signs in adult animals were observed only in two farm (farms 5 and 10; Tables 3 and 4). In young sheep, the major clinical sings included pyrexia, skin hyperemia, lameness, rejection of moving, stiffness in leg and back muscles, bloody nasal discharge, recumbency and breath abnormalities; in adult milking ewes clinical signs included milk reduction, heavy hyperemia of skin (udder, face), light perinasal and perioral edema. Morbidity in flocks ranged 17%-34% and was observed mostly in lambs. Mortality greatly varied between flocks (from 2% to 83% of identified cases) and appeared to depend on general flock welfare and on symptomatic treatment of ill animals. In cattle, clinical signs were observed in young animals only (ranging from three month to two years old). Inappetence, cachexia, dyspnea and cough were observed in calves for fattening. In addition, field post mortem examination revealed pneumonia in two dead calves. Udder edema, as well as symmetrical and bilateral edema throughout the length of the limbs, were observed in milking heifers, lasting for two weeks (Table 3).

BTV Detection and Isolation, 2017
To probe for presence of BTV, we carried out qRT-PCR analysis of field samples. Specifically, 166 out of 693 field samples were positive for BTV by qRT-PCR. Positive samples were found in diseased and dead cattle (99 of 444 samples), sheep (60 of 224 samples) and goat (6 of 33 samples). In addition, three weak positive results were observed from 53 tested samples from zoo and wild animals (one whole blood sample from elephant, one from mountain gazelle and one from Persian roe deer). The rarity of these findings, the low load of BTV RNA in these samples and our inability to isolate the virus or identify serotype, leads us to conclude that the role of wild animals and goats in BTV outbreaks 2017 was negligible. Moreover, we analyzed only a small number of samples from aborted fetuses (n = 20), preventing us from reaching substantive conclusions regarding the role of transplacental transmission in this recent BTV outbreak in Israel.
Overall, in 2017 forty BTV isolates belonging to serotypes 2, 3, 4, 6 and 15 were identified. BTV-2 was isolated from cattle in July 2017, BTV-4 was isolated from sheep and cattle between June and November 2017 in all parts of the country, BTV-3 was found in sheep from three different locations situated in the southern parts of Israel only during autumn months 2017, while BTV-15 was isolated from cattle in late-October-December 2017 from southern and central parts of Israel. In comparison to these relatively sparse cases, BTV-6 showed the highest proportion of virus isolation relative to all other serotypes during late-summer-fall 2017 ( Table 2). To further and quantitatively characterize our findings, we opted to classify the samples according to their Ct value (Table 5) (Table 5). Robust multiple-serotype outbreaks of BTV began in July 2017. Accordingly, August-October 2017 exhibited the highest proportion of positive results (BTV-infected animals compared to the total number of tested ill animals), reaching a peak of more than 50% in September 2017. The notion of a peak in the dynamics of the outbreak is also supported by the high proportion of successful virus isolation from very strong and strong positive samples (>50%). Following the October peak, the proportion of strong and very strong positive samples consequently decreased in November-December 2017; this was correlated with a decrease in the number of acutely diseased animals. Concomitant to the tests aimed at the detection of BTV, we also monitored for presence of non-BTV pathogens by qRT-PCR. In contrast to findings of exams carried out in January-February 2017, where twenty-four whole blood samples from ill cattle were positive for the epizootic hemorrhagic disease virus (EHDV), no such signs of EHDV were observed in August-October 2017. Regarding BEFV, a mini outbreak was registered in one geographic locality situated in the coastal area of Israel, where BEFV was detected in seven milking cows in late July-early August 2017. All samples from cattle, where BTV-6 was identified, bar one which was not tested for BEFV, were negative for BEFV by qRT-PCR. Together, these data allow us to assume that neither EHDV nor BEFV were involved in the identified cases of cattle illness during summer-autumn 2017, at least in the cattle farms where BTV-6 was identified. Table 5. Ct value of positive field samples tested by pan-BTV RT-qPCR for each month during 2017 and consequent BTV isolation.

Discussion
Israel has a long history of presence and effect of BTV on livestock, beginning with the first BT clinical detections in the late 1940s [20]. In particular, BTV-6 was identified in Israel between 1972 and 1989 [20]. Unfortunately, prior to the present study, there was a lack of information about the clinical manifestation of the disease in local Israeli sheep caused by BTV-6. Such a lack of annotated information prevents a systematic comparison of our current findings on the clinical features of BTV-6 in Israeli livestock, with previous documented cases of BTV-6 in Israel. However, a comparison of our data with those obtained from BTV-6 outbreaks in Europe in 2008 reveals considerable differences. There, clinical signs of BT in cows in Netherlands were limited (e.g., to coronitis), a fact which correlated with PCR-based measurements suggestive of very low spread or a very short viremia. This is in sharp contrast to the recent Israeli BTV-6 outbreak, which quickly spread through Israel causing heavy BT clinical manifestations both in sheep and in cattle. We propose that the basis of such a difference stems from the origin and genetic composition of these different viruses, where the European outbreak was most probably derived from a live attenuated South African vaccine strain, which is considered non-virulent, while the Israeli outbreak was caused by the new BTV-6 strain that we describe here [15,16]. A comparison of clinical signs caused BTV-6 with other BTV serotypes circulated in Israeli livestock in 2017 revealed that the most prominent clinical signs both in cattle and in sheep were hyperemia and edema, mostly seen in young animals. BTV-2 and BTV-4 usually cause clinical signs in Israeli cattle (adult milk producing cows) and in the susceptible population of sheep. This is in contrast to BTV-6, where clinical signs in cattle were mostly observed in young animals. According to our analysis, a proposed major genetic source of the Israeli BTV-6 is BTV-3. Within this context, multiple recent studies have identified BTV-3 strains in the Mediterranean Basin [14,[33][34][35]. Notably, the reported clinical BT manifestations of BTV-3 were in sheep only, a markedly different pattern than what was observed for the Israeli BTV-6, as our data support the notion that it affects both sheep and cattle. While serving as a basis for detection and characterization of a new BTV-6 strain in Israel, the present study is most certainly an underestimation of the full extent of Israeli distribution of BT in general, and BTV-6 in particular. This is due to the fact that it is based on passive investigation, using only diagnostic samples collected by veterinary doctors from ill animals (usually a small number of samples from affected animals in each farm, Table 4). Furthermore, the attribution of the clinical phenomena to BTV-6 is also limited by the possibility that the clinical state of the animals may have also reflected other viral and bacterial infections, which could cause pneumonia in fattening calves. Moreover, mixed BTV-3 and -6 infection were registered in at least two out of six BT-affected sheep flocks, raising the possibility that BTV-3 may have also contributed to the observed clinical symptoms.
Genetic analysis of single nucleotide mutation rates and the rate of reassortment (i.e., genetic drift and genetic shift), suggests that the latter is likely to be a major driver of genotypic and phenotypic change in BTV [28]. In this context, the unique features of Israeli BTV-6, such as the magnitude of its spread and/or its ability to induce clinical signs in both sheep and cattle, may reflect its reassortant origin, where a backbone of segments from currently circulating Mediterranean BTV-3 (Seg-1, 4, 5, 6, 7 and -9) was supplemented by contributions from BTV-6 (Seg-2), BTV-4 (Seg-3), BTV-9 (Seg-8) and an untyped BTV isolate (Seg-10). Notably, Seg-2 and Seg-10, which encode for VP2 and NS3 and have been proposed as determinants of virulence [25], were not of BTV-3 origin. Concerning the geographical origin of the proposed contributing BTV strains, we propose that co-circulation in Africa played a predominant role in generating Israeli BTV-6. This proposal also includes the segments presenting highest similitude with Italian BTV-3 isolates, as these have also been proposed to be of North African origin [33]. The notion that BTV distribution in Israel is heavily influenced by strains that circulate in North Africa is further supported by our unpublished results, which show that BTV-3 strains identified in Israel between 2013-2016 (ISR-2019/13, ISR-2153/16 and ISR-2262/2/16) also exhibit high percentages of genetic identity by some of genes with strains originating from Tunisia (Figure 2b, and Figures S2, S3 and S5). In summary, following the observation in the field of farm animals presenting BT-like symptoms, we have isolated and genetically characterized a new reassortant BTV-6 strain. This strain exemplifies how novel disease-causing abilities may stem from co-circulation and reassortment of BTV strains.