During October–November 2009, samples of whole blood from 8 febrile horses (H1–8; temperatures 39.5°C–42°C) in Israel were collected into EDTA tubes and analyzed at the Koret School of Veterinary Medicine (Hebrew University, Rehovot, Israel). Vero cell (American Type Culture Collection, Manassas, VA, USA) culture results of blood from H3, H5, and H8 were positive for EEV (Table 1; Figure 1).

Total RNA was extracted from the fifth and sixth passages of all 3 samples by using the QIAamp Viral RNA Mini Kit (QIAGEN, Valencia, CA, USA) to obtain sufficient viral load for the subsequent analyses and replications. RNA was reverse transcribed into cDNA by using the Verso cDNA Kit (Thermo Fisher Scientific, Epsom, UK). PCR amplification of the gene encoding NS3 (Seg-10) was performed on the 3 isolates by using GoTaq Green Master Mix (Promega, Madison, WI, USA) with the following primers: 5′-¹GTT AAG TTT CTG CGC CAT GT²³-3′, 5′-⁷⁴¹GTA ACA CGT TTC CGC CAC G⁷⁶⁰-3′. Thermal cycling conditions for the PCR were as previously described (9); the primer annealing temperature was modified to 53.5°C. PCR products were purified by using a cDNA purification kit (ExoSAP-IT; USB, Cleveland, OH, USA), and sequencing was conducted by BigDye terminator cycle sequencing chemistry (Applied Biosystems, Foster City, CA, USA) in an ABI 3700 DNA Analyzer (Applied Biosystems) by using ABI data collection and sequence analysis software. Further analysis of the NS3 sequence was performed with Sequencer software, version 4.8 (Gene Codes Corp., Ann Arbor, MI, USA). Sequences were deposited in GenBank under accession nos. HQ441245 for H5, HQ441246 for H3, and HQ441247 for H8. The NS3 genes (Seg-10) were compared with those of different EEVs (9) and other related orbiviruses (Table 2). Phylogenic trees were generated by using the neighbor-joining and maximum likelihood methods (Phylip Inference Package version 3.68, Seqboot Program; J. Felsenstein, University of Washington, Seattle, WA, USA) to create 100 datasets (bootstrapping) and the DNA Maximum Likelihood Program version 3.5 (http://cmgm.stanford.edu/phylip/dnaml.html) to construct the trees. Finally, the Consense program version 3.5c (http://cmgm.stanford.edu/phylip/consense.html) was used to create a final consensus tree for our dataset. Broadhaven virus, a tick-borne orbivirus, was used as the outgroup in the phylogram for the gene encoding NS3.

The phylogenetic analyses of EEV Seg-10 grouped the Israeli isolates with other EEV isolates but as a distinct group with no close relation to African horse sickness virus, BTV, or epizootic hemorrhagic disease virus. Within the EEV group, 3 discrete clusters (A, B, C) were recognized; the Israeli isolates formed one of these clusters (C; Figure 2). The Israeli isolates have 85%–86% nt identity to cluster A and 75%–76% nt identity to cluster B.

In addition, full-length cDNA copies of individual EEV (from H3 and H8) genome segments were synthesized and amplified by reverse transcription PCR by using the anchor spacer–ligation method as described (10,11). Partial sequences (for the upstream 450 bp) of Seg-2 from the different Israeli isolates were identical, showing 92.3% nt and 95.7% aa sequence identity with Seg-2 and VP2 of the Kaalplaas isolate, the reference isolate of EEV-3 (GenBank accession numbers are listed in Table 2). Previous phylogenetic comparisons of Seg-2/VP2 from different BTV types showed a maximum of 71% nt and 78% aa acid identity between serotypes (6), indicating that the isolates from Israel also belong to EEV type 3.

Equine encephalosis virus has long been enzootic to southern Africa, but it has not been isolated in other parts of the world. We report the characterization of an EEV strain isolated outside of Africa. Phylogenetic analysis of Seg-2 showed 92% sequence identity to EEV-3 (Kaalplaas).

Analysis of Seg-10 (the gene encoding NS3) of different orbiviruses showed 2 clusters of South African EEV strains (A and B), in agreement with previously published studies (9). These 2 clusters appear to correlate with the geographic origins of the viruses in South Africa, independent of their isolation date. It has been suggested that the 2 EEV Seg-10 clusters in South Africa are related to the distribution of their Culicoides spp. midge vectors, C. imicola (senso stricto) and C. bolitinos. The former is the most abundant Culicoides spp. midge in Israel (12). However, the EEV isolates from Israel group as a distinct cluster (C) with similar distances to the 2 South African clusters, raising questions concerning the geographic origin of this virus. A similar finding has been observed in African horse sickness virus Seg-10, which also forms into 3 distinct groups (13).

The question of how and when the virus was initially introduced to Israel remains unanswered. Because the clinical manifestations of equine encephalosis are usually mild, it is often overlooked and underdiagnosed. EEV could have been introduced to Israel before the virus was first isolated in 2009. Alternatively, the virus might have been introduced into neighboring countries and transmitted into Israel by infected vectors carried by winds, as described for other orbiviruses (14,15). The fact that the Israeli strain of EEV-3 grouped in a different cluster than the 2 South African strains, supports the idea that it has evolved in this region for a sufficient time to accumulate these changes and most likely was not recently introduced into Israel from South Africa.