During February 2010, sheep and goats in the Abdali region of Kuwait, close to the Iraq border, showed the following clinical signs consistent with BT: lameness, coughing, mouth lesions, stillbirth, congenital abnormalities, pneumonia, enteritis, and hepatitis. Forty-six of 48 serum samples were positive for BTV-specific antibodies by competitive ELISA (Investcare-Vet, London, UK) at the Veterinary Diagnostic Laboratory and Animal Research Center in Kuwait, or by double antigen-recognition ELISA (ID Vet, Montpellier, France) at the Institute for Animal Health in the United Kingdom.

Twenty-six EDTA-treated blood samples, 4 spleens, and 1 liver sent for analysis for BTV to the World Organisation for Animal Health reference laboratory at the Institute for Animal Health (Woking, UK) all gave negative results by real-time reverse transcription–PCR (rRT-PCR) targeting either BTV Seg-1 (6) or Seg-1 and Seg-5 (7). However, 2 blood samples (from animals 364 and 374) were positive in assays selective for Seg-10 (8), with cycle threshold (C_t) values of 35 and 28, respectively. Previous attempts to isolate BTV from blood samples with C_t values >32 were usually unsuccessful, and no further work was done with animal 364. Washed blood from animal 374 (reference collection sample KUW2010/01) (9) was injected into embryonated chicken eggs. Although no hemorrhages were detected, the virus was passaged twice in BHK-21 cells (isolate KUW2010/02), producing atypical cytopathic effects.

KUW2010/02 RNA analyzed by agarose gel electrophoresis generated a genome segment migration pattern (electropherotype) typical of BTV (Figure 1). Although KUW2010/02 was negative by BTV-specific rRT-PCR selective for Seg-1 (6), it had C_t values of 16.8 in Seg-10–specific assays (8). Identification of KUW2010/02 as BTV was confirmed by indirect antigen-sandwich ELISA selective for core protein VP7(T13) (10) with optical density at 490 nm values >0.15 (equivalent BTV titer of 6.75 log₁₀ 50% tissue culture infective dose/mL).

RNA from KUW2010/01 and KUW2010/02, extracted by using TRIzol (Invitrogen, Carlsbad, CA, USA) (11), gave uniformly negative results by type-specific rRT-PCRs targeting Seg-2 of BTV serotypes 1–25 (kits supplied by Laboratoire Service International, Lissieu, France). However, full-length cDNA copies of Seg-2 (2,929 bp) and Seg-3 (2,773 bp) from KUW2010/02 were synthesized and then sequenced as described (11), showing conserved terminal regions typical of BTV (5′-GUUAAA...........ACUUAC-3′) (2). BLAST (www.ncbi.nlm.nih.gov/BLAST) analysis of KUW2010/02 Seg-2 and Seg-3 sequences (GenBank accession nos. HM590642 and HM590643) showed highest identity with equivalent genome segments of other BTVs, although for Seg-2 the search algorithm was changed from megablast (highly similar sequences) to blastn (somewhat similar sequences).

Phylogenetic analysis of KUW2010/02 Seg-3/VP3(T2), conducted by using neighbor-joining methods and p-distance matrices (12), showed nucleotide/amino acid identity levels of 73.7%/87.6% to 76.6% /88.9% with other BTVs. Seg-3 of none of the previously characterized BTVs showed close relationships to KUW2010/02; BTV-1/GRE2001/05 and BTV-25/TOV were most closely related (GenBank accession nos. DQ186822 and GQ982523, respectively), which indicates that KUW2010/02 represents a distinct geographic cluster or topotype (4).

Seg-2/VP2 of KUW2010/02 showed nucleotide/amino acid identity levels of 42.8%/28.3% to 63.9%/61.5% with previously recognized BTV serotypes and was most closely related to BTV-25 (nucleotype K, Figure 2). Reference strains of BTV-10 and BTV-17 from the United States (nucleotype A) were the next most closely related, with 61.8%/58.1% and 62%/57.7% nt/aa identity, respectively. This places KUW2010/02 as a distinct virus type within a novel 12th Seg-2 nucleotype L (4; Figure 2). ClustalX (www.clustal.org) alignments and neighbor-joining trees also confirmed the identity of KUW2010/02 as a BTV. Virus neutralization tests of KUW2010/02 (13) that used reference guinea pig immune serum against BTV types 1–24 and antiserum from goats previously infected with BTV-25 showed no reduction in KUW2010/02 infectivity (data not shown), demonstrating that it does not belong to BTV serotypes 1–25.

Most serum samples tested from sheep and goats in Kuwait showing clinical signs of disease were seropositive for BTV-specific antibodies. However, BTV RNA was detected in only 2 sheep serum samples (animals 364 and 374), suggesting that the clinical signs were not caused by ongoing BTV infection. All samples were also tested for peste des petits ruminants virus by rRT-PCR (14), but results were uniformly negative.

BTV RNA was detected by using a BTV Seg-10–specific rRT-PCR (8) previously used to detect BTV-25 in Switzerland (15). However, BTV Seg-1–specific or Seg-1– and Seg-5–specific assays (6,7) failed to detect KUW2010/02, identifying it is an unusual or atypical BTV strain. Alignment of the Seg-1–specific and Seg-5–specific primers and probes with KUW2010/02 sequences identified numerous mismatches that would prevent detection of the viral RNA (data not shown). However, the probe and primers designed by Orrù et al. (8) showed a perfect match with Seg-10 of KUW2010/02, demonstrating the need for appropriate rRT-PCR protocols to detect this virus. Agarose gel electrophoresis analysis of KUW2010/02 genome segments generated a migration pattern typical of BTV (Figure 1). Levels of nucleotide/amino acid identity of Seg-3 (up to 76.6%/89%) of KUW2010/02 with other BTV isolates also identified it as BTV.

At peak viremia, a previously unexposed sheep infected with a virulent BTV strain could be expected to show C_t values ≈20. The C_t of 28 obtained with blood of animal 374 (KUW2010/01) indicates a low viremia, suggesting that the severe clinical signs observed were not caused by BTV. Experimental infection of previously unexposed sheep with KUW2010/02 also caused only mild clinical signs (data not shown).

Different isolates of a single BTV serotype show >68.4%/72.6% nt/aa identity in Seg-2/VP2, with 40.5%/22.1% to 71.5%/77.8% identity between different serotypes (4). KUW2010/02 showed only 42.8%/28.3% to 63.9%/61.5% identity in Seg-2/VP2 with recognized BTV serotypes, consistent with membership of a novel 26th BTV type and Seg-2 nucleotype (L) (4). These conclusions were supported by virus neutralization test results. The sequence data presented here will help support development of new diagnostic tools (RT-PCR–based typing assays) to determine the incidence and distribution of this novel serotype, as well as natural reservoir(s) and insect vectors.