High throughput detection of bluetongue virus by anew
Shared by: cpd16778
J Vet Diagn Invest 18:7–17 (2006) High throughput detection of bluetongue virus by a new real-time ﬂuorogenic reverse transcription–polymerase chain reaction: Application on clinical samples from current Mediterranean outbreaks Miguel Angel Jimenez-Clavero,1 Montserrat Aguero, Elena San Miguel, Tomas Mayoral, ´ ¨ ´ ´ ´ ´ Maria Cruz Lopez, Marıa Jose Ruano, Esther Romero, Federica Monaco, Andrea Polci, ´ ´ Giovanni Savini, Concepcion Gomez-Tejedor Abstract. A real-time reverse transcription–polymerase chain reaction (RT-PCR) assay was developed for the detection of bluetongue virus (BTV) in blood samples. A combination of primers speciﬁc for a highly conserved region in RNA segment 5 (based on Mediterranean BTV sequences) and a DNA probe bound to 5 - Taq nuclease-3 minor groove binder (TaqMan MGB) was used to detect a range of isolates. This real-time RT-PCR assay could detect 5.4 10 3 tissue culture infectious doses (TCID50) of virus per milliliter of sample, which was comparable to our current BTV diagnostic nested RT-PCR assay. The assay detected all recent Mediterranean isolates (including serotypes 2, 4, and 16), BTV vaccine strains for serotypes 2 and 4, and 15 out of the 24 BTV reference strains available (all serotypes), but did not detect the related orbiviruses epizootic hemorrhagic disease and African horse sickness viruses. Following assay evaluation, the ability of this assay to identify BTV in recent isolates (2003, 2004) from ovine and bovine samples from an epizootic outbreak in Spain was also tested. Minor nucleotide changes (detected by sequencing viral genomes) within the probe- binding region were found to have a profound effect on virus detection. This assay has the beneﬁts of being fast and simple, and the 96-well format enables large-scale epidemiological screening for BTV, especially when combined with a high-throughput nucleic acid extraction method. Key words: Bluetongue virus; diagnosis; high throughput; real-time reverse transcription–polymerase chain reaction. Introduction graphically distinct evolutionary lineages, or topo- types,11,12 allowing the differentiation of BTV variants Bluetongue virus (BTV) is the prototype member of from different origins, even within the same serotype. the genus Orbivirus within the family Reoviridae, and The emergence of new BTV variants is considered to its primary route of transmission among sheep and be driven by reassortment of RNA segments (i.e., ge- other ruminants is via the arthropod vector Culicoides netic shift),24,30,31 whereas individual gene segments (midge). Cattle are suspected to be the predominant evolve independently by genetic drift due to immu- reservoir for this virus, as BTV infection is asymptom- nological pressure from the infected host.6 atic in cattle, but results in a prolonged viremia.8,9 The The distribution of Bluetongue disease is deter- BTV genome consists of 10 double-stranded RNA mined by the geographic distribution of the arthropod segments that encode 7 structural proteins (VP1 to vector and includes both tropical and temperate areas, VP7) and 4 nonstructural (NS1, NS2, NS3, and NS3a) including Africa, southern Asia, Australia, the Middle proteins. Not only is there marked genetic variation East, and the Americas. Due to the serious socioeco- among the 24 known BTV serotypes, but also among nomic consequences of BTV outbreaks on the inter- the virus strains within each serotype.7,12 In addition, national trade of animals and animal products, it has this group of viruses exhibits genetically and geo- been included in the Ofﬁce International des Epizo- oties (OIE) list of notiﬁable diseases (formerly List From the Laboratorio Central de Veterinaria, Algete, Spain A).25 Epizootic episodes of BTV, which occur infre- (Jimenez-Clavero, Aguero, San Miguel, Mayoral, Cruz Lopez, Ru- ´ ¨ ´ quently, result in large losses through required control ´ ano, Romero, Gomez-Tejedor), and Istituto Zooproﬁlattico Speri- measures of infected and cohoused livestock. In au- mentale dell’Abruzzo e del Molise ‘‘G. Caporale,’’ Teramo, Italy tumn 2003, several outbreaks of BTV serotype 4 virus (Monaco, Polci, Savini). 1Corresponding author: Miguel Angel Jimenez-Clavero, Departa- ´ were reported in the western Mediterranean region mento de Enfermedades Emergentes, Laboratorio Central de Veter- (Corsica, Sardinia, and Menorca) (OIE-Handistatus II, inaria, Ctra. Algete km 8, 28110, Algete, Madrid, Spain. http://www.oie.int/hs2 [accessed June 2005]) and BTV 7 8 ´ Jimenez-Clavero et al. serotype 2 outbreaks had been previously reported in Union (EU) reference laboratory for BTV (Pirbright, UK) this area in 1999–2000.13,22,38 Although BTV had never (http://www.iah.bbsrc.ac.uk/dsRNA virus proteins/ReoID/ been detected in this area before 1999, the presence of BTV-isolates.htm [accessed June 2005]); ﬁeld isolates from the Culicoides vectors23,32,37 and theoretical modeling the Mediterranean region: SPA2000 (Majorca, Spain) and based on climate data37 and satellite remote sensing35 SAD2000 (Sardinia, Italy), both belonging to serotype 2; SPA2003 (Menorca, Spain), SAD2003 (Sardinia, Italy), predicted that this region was at high risk for BTV ´ SPA2004 (Cadiz, Spain), and MOR2004 (Morocco), belong- epizootic infections. In autumn 2004, these predictions ing to serotype 4; ITL2002 (Puglia, Italy), belonging to se- were borne out by new epizootic outbreaks of BTV rotype 16; and live attenuated monovalent vaccines for se- serotype 4, ﬁrst reported in northern Morocco, and rotypes 2 and 4 (OBPa, Onderstepoort, South Africa). Virus- soon after in the Iberian Peninsula, which had been es were propagated and virus concentrations were deter- free of BTV since the 1950s. Other areas of the Med- mined by microtitration assay28 using Vero cells, which were iterranean region where BTV outbreaks have been re- cultured in Eagle medium (EMEM) supplemented with L- ported in the last 7 years include Albania, Algeria, glutamine, antibiotics (100 U/ml penicillin, 100 g/ml strep- Bosnia and Herzegovina, Bulgaria, Croatia, Cyprus, tomycin) and 5% fetal calf serum. Assay speciﬁcity was as- Greece, Israel, Italy (mainland), Kosovo, Macedonia, sessed using phylogenetically or symptomatically related vi- Serbia and Montenegro, Tunisia, and Turkey (OIE- ruses (or both): African horse sickness virus (AHSV sero- Handistatus II, http://www.oie.int/hs2 [accessed June type 4, isolate SPA’87, from the National Reference Laboratory for African Horse sickness virus, Central Veter- 2005]). inary Laboratory, Algete, Spain); epizootic hemorrhagic dis- The speed of the assays to detect BTV is directly ease virus (EHDV North American prototype serotypes 2 related to the ability to detect viremic animals and and 4)c; foot-and-mouth disease virus, FMDV serotypes A, therefore to limiting the spread of disease. Viremia can O, C, Asia-1, SAT-1, SAT-2, and SAT-3d; and contagious be demonstrated in blood of infected animals by either ecthyma virus (ﬁeld isolate, Central Veterinary Laboratory, virus isolation or nucleic acid detection methods. Al- Algete, Spain, unpublished). though virus isolation in embryonated chicken eggs is reliable and sensitive, it is technically cumbersome and Samples requires signiﬁcantly longer to yield a result than nu- Ethylenediamine tetraacetic acid was used to stop the co- cleic acid–based detection assays. For these reasons, agulation of blood collected from sheep or cattle suspected the latter have become the method of choice for BTV of having BTV and from sentinel cattle during recent out- detection. Several reverse transcription–polymerase breaks in Spain during either October–December 2003 (Mediterranean island of Menorca, Balearic Islands) or Oc- chain reaction (RT-PCR) methods have previously tober–December 2004 (mainland [Andalusia and Extremad- been described,1–4,10,15,17,33 and the BTV segment 5 (en- ura]) (OIE-Handistatus II, http://www.oie.int/hs2 [accessed coding the NS1 protein) is the most commonly tar- June 2005]). Tissue samples (spleen, heart blood, and gut) geted gene due its high degree of conservation across from a sheep fatally infected with BTV during one of these BTV serogroups (segment 10 has also been used). outbreaks were also tested. Virus isolated from this sheep Current RT-PCR methods for BTV detection are rapid was identiﬁed as serotype 4 by the European Reference Lab- and show satisfactory sensitivity and speciﬁcity; how- oratory for BTV at Pirbright, UK, and is referred to as ever, this methodology is not well suited to high- SPA2003 isolate in the Results section. Control ovine blood throughput analysis, limiting its use in large-scale ep- samples were obtained from 50 healthy BTV-seronegative idemiological surveillance. For these reasons, we have (as determined by enzyme-linked immunosorbent assay examined the use of real-time ﬂuorogenic RT-PCR [ELISA]b testing) sheep from areas where BTV has never methodologies in a 96-well format for high-throughput been reported. Control bovine blood samples were obtained sample monitoring. from 300 healthy BTV ELISAb-negative cattle, used in the Spanish surveillance program for BTV, in areas where the We developed a real-time ﬂuorogenic RT-PCR disease was absent at the time of testing. method using a 5 -Taq nuclease-3 minor groove bind- er18 (TaqMan MGB) DNA probe for the detection of Sample processing and total nucleic acid extraction BTV in a 96-well format. We then evaluated the use- Single tube-spin protocol. Five hundred microliters of fulness of the assay using laboratory and previously each blood sample was lysed with 1 ml of sterile water, characterized BTV serotypes (isolated from western incubated for 10 minutes on ice, and then centrifuged at Mediterranean outbreaks) and tested the assay against 16,000 g for 5 minutes at 4 C. The supernatant was then samples from recent Spanish BTV outbreaks. removed and the pellet was resuspended in 0.2 ml of binding buffer that had been provided by the High Pure Viral (HPV) Materials and methods nucleic acid extraction kit.e The total RNA extraction was Viruses, vaccines, cells, and virus propagation in completed as indicated in the manufacturer’s instructions us- cell cultures ing 50 l as the elution volume. Tissue samples (approxi- The following BTV strains were used in this study: pro- mately 0.5 g) were homogenized in 1 ml of phosphate-buff- totype strains for all 24 BTV serotypes from the European ered saline (PBS) and following a 10-minute centrifugation High throughput detection of bluetongue virus by real-time RT-PCR 9 at 16,000 g, 1 volume (0.2 ml) of supernatant was mixed nucleic acid extraction was performed as described in the with 1 volume of binding buffer (from the same kit as single tube-spin protocol (see above). above), and nucleic acid extraction was achieved in the same way. Total RNA extraction of infected cell culture superna- Sequence analysis tants containing BTV serotypes or control viruses at titers To analyze the detection range of the real-time RT-PCR of at least 1 104 TCID50 /ml (starting volume, 20–200 l), method in more detail, the RNA regions corresponding to was achieved in the same way. nucleotide positions 316–790 (numbered according to Nucleic acid extraction in 96-well format. Two separate GenBank AY137387) within segment 5 were sequenced for formats were used: 1) The Nucleospin Multi-96 virus nucleic the following isolates: BTV-2 SPA2000, BTV-4 SPA2003, acid extraction kitf was applied as follows: 0.1 ml of blood BTV-4 SPA2004, BTV-4 MOR2004, BTV-4 ITL2002, BTV- and 0.2 ml of distilled water was added to each well in the 4 (vaccine straina), BTV-4 reference strain, BTV-16 refer- 96-well lysis microplate. Following centrifugation as above, ence strain, BTV-24 reference strain, BTV-19 reference the cell pellet that was resuspended in 0.4 ml of RAV1 buffer strain, BTV-10 reference strain, BTV-13 reference strain, plus 20 l (1 g/ l) of proteinase K and RNA extraction and BTV-7 reference strain. RNA sequence was determined completed according to the manufacturer’s instructions. 2) following ampliﬁcation by nested RT-PCR,2 which was bi- High Pure viral nucleic acid extraction kite in 96-well format directionally sequenced using the inner primers of the nested (HPV-96): samples were prepared essentially according to reaction, the Big Dye Terminator (version 3.1) Cycle Se- the manufacturer’s instructions except that Whatman 96-well quencing Kit, j and an ABI 3730 XL DNA Analyzer.k Elec- Uniﬁlter microplatesg were used instead of the spin columns tropherograms were analyzed using the Sequencing Analysis provided in the kit. The ﬁnal protocol was as follows: 0.1- (version 5.1) softwarel and sequences have been submitted ml blood samples were lysed in the same way as described to GenBank under accession numbers AY741396 (BTV-2 above for Nucleospin Multi-96, but using a Whatman 800- SPA2000), AY741397 (BTV-4 SPA2003), DQ098094 (BTV- l round-bottom 96-well Uniplate.h The cell pellet was re- 4 ITL2002), DQ098092 (BTV-4 MOR2004), DQ098093 suspended in 0.1 ml of binding buffer plus 20 l (1 g/ l) (BTV-4 SPA2004), DQ098095 (BTV-4 vaccine straina), of proteinase K (provided in the kit), and subjected to a 10- DQ098096 (BTV-4 reference strain), DQ098097 (BTV-7 ref- minute incubation at 72 C, followed by the addition of 50 erence strain), DQ098098 (BTV-10 reference strain), l of isopropanol to each well. After gentle mixing, the en- DQ098099 (BTV-13 reference strain), DQ098100 (BTV-16 tire contents of each well was removed from the 96-well reference strain), DQ098101 (BTV-19 reference strain), and Uniplate, transferred to a Uniﬁlter microplate,g and placed DQ098102 (BTV-24 reference strain). The sequences ob- over another 800- l round-bottom 96-well Uniplateh for col- tained were aligned with the following BTV NS1 sequences lection of the ﬁltrates, following centrifugation for 15 min- available in GenBank: AY137387 (BTV-2 Corsica 2002), utes at 2,250 g. The ﬁltrates were discarded and 250 l Y00422 (BTV-10, US prototype strain), X17041 (BTV-17, of inhibitor buffer (provided in the kit) was added to each US prototype strain), M97681 (BTV-11, US prototype ﬁlter well and the plate was centrifuged for 4 minutes at strain), M97762 (BTV-13, US prototype strain), M97680 2,250 g. The ﬁlter wells were then washed twice with 250 (BTV-2, US prototype strain), AY462225 (BTV-2 KN, Tai- l of wash buffer (provided in the kit) and subjected to 2 wan), and X56735 (BTV-20, Australia) using the CLUSTAL centrifugation steps for 4 minutes and 15 minutes at 2,250 W 1.6 program.36 g. The ﬁltrate was discarded and the ﬁlters were then dried by centrifugation at 2,250 g for a further 5 minutes. The Real-time ﬂuorogenic RT-PCR recovery plate was changed to a Whatman 250- l V-bottom Oligonucleotide primers and a Taqman-MGB ﬂuorogenic 96-well Uniplate,i and 50 l of nuclease-free water was add- probe were designed using Primer Express (version 2.0.0) ed to each ﬁlter well and the plate centrifuged for 10 minutes softwarem based on the BTV-2 (SPA2000) and BTV-4 at 2,250 g to elute the nucleic acids from the ﬁlters. The (SPA2003) isolate sequences; however, homology to the eluted nucleic acids were either immediately analyzed by same region of the BTV genome (GenBank accession num- RT-PCR or stored at 20 C for future use. bers AY137387, AY138895, M97762, X17041, X15891, Comparison between single-tube and 96-well nucleic acid Y00422, M97681, M97680, and X56735) were also con- extraction methods. BTV-seropositive blood samples (1 ﬁrmed. The CLUSTAL W 1.6 program36 was used to design from sheep, 2 from cattle) were serially diluted in BTV- conserved sequence primers: forward (5 -GTTGAGAGA- seronegative blood from the same species, whereas the vac- CAAATTAACACATGTCC-3 ) and reverse (5 -AATGCT cine was diluted in PBS. All samples were subjected to nu- TCGCAAAATCATCCAT-3 ). These primers ampliﬁed a re- cleic acid extraction in parallel. gion comprising nucleotides 606–723 of BTV segment 5, and a ﬂuorogenic TaqMan-MGB probe (6-FAM-CGATT- Nested RT-PCR for BTV and RT-PCR for CAGCTGATCAAT-MGB) was designed to hybridize nucle- other viruses otides 676–692 (nucleotide position numbering according to Nested RT-PCR for BTV detection was performed ac- the AY137387 sequence). The ﬁnal protocol consisted of an cording to the method described by Aradaib et al.2 We also initial denaturation step during which primers (ﬁnal concen- used PCR or RT-PCR methods already described for the de- tration: 0.5 M each) and viral RNA (2 l) were heated at tection of African horse sickness virus,34 epizootic hemor- 95 C for 5 minutes in a volume of 7 l of RNAase-free rhagic disease virus,2 contagious ecthyma virus,14 and foot- water, and then chilled on ice. This denaturation step was and-mouth disease virus.29 In the case of control viruses, followed by the addition of 18 l containing (per reaction) 10 ´ Jimenez-Clavero et al. Figure 1. Comparative sensitivity analysis of BTV-2 RNA (SPA2000 isolate) using ﬂuorogenic real-time RT-PCR and nested RT-PCR methods. Tenfold dilutions of the viral extract were subjected to both real-time ﬂuorogenic RT-PCR and nested RT-PCR analysis in parallel. The results of the ﬂuorogenic RT-PCR analysis are plotted as curves of ﬂuorescence (matching the course of the ampliﬁcation) as a function of PCR cycle number. The insert within the plot shows the result of the nested RT-PCR as a photograph of gel electrophoresis stained with ethidium bromide. The equivalent viral units (TCID50/ml) of each dilution are indicated both on the right side of the plot and above the corresponding lanes in the gel. The same experiment was repeated twice with identical results. 12.5 l of TaqMan 2 universal PCR master mix, 0.6 l of dilution to yield a Ct value below 40. All determinations 40 Multiscribe and RNase inhibitor mix, both from the were conﬁrmed by repeating the experiment at least once. commercial RT-PCR ampliﬁcation kit TaqMan One-step RT- The sensitivity of the assay including the RNA extraction PCR Master Mix Reagents,n 0.12 l of the TaqMan-MGB was also assessed by repeating the above experiment on 10- probe (ﬁnal concentration: 0.25 M), and RNase-free water. fold dilutions of the viral stock, which were then processed The ﬁnal reaction volume was 25 l. for nucleic acid extraction and analyzed by both the real- The reactions were analyzed in 96-well optical PCR plates time and nested RT-PCR methods. using an ABI PRISM 7000 Sequence Detection System (SDSo) apparatus and the SDS 7000 (version 1.1) softwarep. Results Reverse transcription was carried out at 48 C for 25 minutes followed by 10 minutes at 95 C (‘‘hot start’’) and 40 PCR Detection of BTV by a new real-time RT-PCR cycles of 15 seconds at 95 C and 1 minute at 61 C (total method was compared with an existing nested RT-PCR reaction time 2 hours 11 minutes). Samples were considered method (Fig. 1). Both methods were able to detect a positive for BTV if they yielded Ct values lower than 40 10 5 dilution of the BTV RNA (equivalent to 0.0054 (threshold set automatically following the instructions of the TCID50 /ml of sample) extracted from 1,080 TCID50 of SDS 7000 analysis software, version 1.1). The sensitivity of cell-cultured BTV serotype 2 (SPA2000). As shown in the ﬂuorogenic real-time RT-PCR assay was assessed using Fig. 2, the detection achieved by the real-time RT-PCR 10-fold dilutions from the control BTV-2 isolate RNA assay showed a linear relationship between signal and (SPA2000 isolate), comparing detection of viral extract for the concentration of template present in the sample both real-time ﬂuorogenic RT-PCR and nested RT-PCR anal- ysis in parallel. (correlation coefﬁcient of 0.9948, and a slope of Sensitivity was determined as the highest dilution giving 3.334) over a range of 102 to 10 3 TCID50 /ml. To a positive result: for nested RT-PCR this was the highest assess whether the nucleic acid extraction step affected dilution at which a band was visible on an ethidium bro- the performance of the method, 10-fold dilutions of mide–stained gel; for real-time RT-PCR this was the highest the viral stock (instead the puriﬁed RNA) were also High throughput detection of bluetongue virus by real-time RT-PCR 11 Figure 2. Linearity and efﬁciency of the real-time RT-PCR assay for BTV. The experiment shown in Fig. 1 is plotted here as Ct (threshold value) as a function of quantity of viral RNA (equivalent TCID50 infectious units per milliliter) on a log scale. analyzed. A comparable level of sensitivity (0.001 all recent BTV ﬁeld variants from the Mediterranean TCID50 /ml) and linearity (dynamic range, 102 to 10 3; region (serotypes 2, 4, and 16); vaccine strains from r2, 0.9941; slope, 3.592) was obtained, indicating BTV serotypes 2 and 4, and 15 out of 24 serotype that the nucleic acid extraction does not signiﬁcantly BTV reference strains. No cross-reactivity was ob- affect the overall performance of the method, at least served for non-BTV virus samples by our real-time with this type of sample. RT-PCR assay, regardless of whether they belong to The speciﬁcity of the new method was determined the Orbivirus genus or have similar pathogenic se- by analyzing reference BTV isolates from all 24 se- quelae. However, the integrity of the nucleic acids iso- rotypes (obtained from the EU reference laboratory for lated from the negative control viruses was conﬁrmed BTV disease, Pirbright, UK), as well as vaccine strains by positive results for their own speciﬁc PCR or RT- (OBP, Onderstepoort, South Africa) and ﬁeld variants PCR detection assays (not shown). isolated in recent years from the Mediterranean region. Although the BTV serotype 4 reference strain was We also determined whether our real-time RT-PCR not detected by our real-time RT-PCR assay, all other method cross-reacted with other members of the Or- BTV serotype 4 ﬁeld isolates and the vaccine strain bivirus genus (AHSV serotype 4, EHDV serotypes 2 were detected (Table 1). Similarly, a ﬁeld isolate of and 4); viruses that cause symptoms similar to blue- BTV serotype 16 was detected even though the ref- tongue (contagious ecthyma virus, a member of Pox- erence strain was not (Table 1). These results indicate viridae); or foot-and-mouth disease virus (serotypes A, that the inability to detect the reference strains does O, C, Asia, SAT1, SAT2, and SAT3; Picornavirus fam- not rule out its effectiveness at detecting these sero- ily) (Table 1). The real-time RT-PCR assay detected types in the ﬁeld. The failure of our assay to detect 12 ´ Jimenez-Clavero et al. Table 1. Speciﬁcity of the real-time ﬂuorogenic RT-PCR assay in the BTV-4 reference strain, for which a single mis- for BTV. Analysis of BTV cell culture supernatants, vaccines, and match in the probe region and 2 more in the forward other viruses to assess speciﬁcity. (5 ) primer resulted in a loss of detection. In other Virus (serotype) Isolate/origin Result nondetected reference strains (serotypes 7, 10, 13, 16, BTV-1 494 2E (reference strain) positive 19, and 24), the differences in nucleotide sequence BTV-2 reference strain positive ranged from 1 to 3 mismatches in the forward (5 ) BTV-2 ﬁeld isolate SPA/2000 positive primer, 1 to 5 mismatches in the probe, and at least 1 BTV-2 ﬁeld isolate SAD/2000 positive mismatch in the reverse (3 ) primer. It is remarkable BTV-2 vaccine (South Africa) positive that there was a low homology in the target region BTV-3 reference strain positive BTV-4 reference strain negative between the BTV-16 reference strain and the BTV-16 BTV-4 ﬁeld isolate SPA2003 positive ﬁeld isolate used in this study. Table 2 shows the nu- BTV-4 ﬁeld isolate SAD2003 positive cleotide sequence alignments of BTV serotypes used BTV-4 ﬁeld isolate SPA2004 positive in this study, but also includes the available sequences BTV-4 ﬁeld isolate MOR2004 positive for prototype strains from US BTV serotypes, BTV-2 BTV-4 vaccine (South Africa) positive BTV-5 reference strain positive from Taiwan, and BTV-20 from Australia. Given the BTV-6 reference strain positive homology, we would expect viruses showing 100% BTV-7 reference strain negative identity within the probe region (US BTV-11, -13, and BTV-8 reference strain positive -17) and almost complete nucleotide sequence identity BTV-9 reference strain positive within the primer regions would be detected, whereas BTV-9 vaccine (South Africa) positive BTV-10 reference strain negative those BTV strains showing at least 1 mismatch within BTV-11 reference strain positive the probe region (US BTV-2, and BTV-10, Australian BTV-12 reference strain positive BTV-20, and Taiwan BTV-2) would likely remain un- BTV-13 reference strain negative detected. These expectations are supported by a much BTV-14 reference strain positive larger number of mismatches in the primer and probe BTV-15 reference strain positive BTV-16 reference strain negative regions in published sequences for EHDV and AHSV BTV-16 puglia 2002 positive (not shown), which were not detected by our real-time BTV-17 reference strain positive RT-PCR assay (Table 1). BTV-18 reference strain positive To assess the performance of the new BTV detec- BTV-19 reference strain negative tion method under normal diagnostic laboratory con- BTV-20 reference strain negative BTV-21 reference strain negative ditions, 76 sheep and 21 cattle blood samples from the BTV-22 reference strain positive 2003 Menorca outbreak (previously identiﬁed as BTV- BTV-23 reference strain positive positive by the nested RT-PCR method) were analyzed BTV-24 reference strain negative by real-time ﬂuorogenic RT-PCR. In addition, spleen, AHSV-4 SPA/1987 negative heart blood, and gut homogenates from a symptomatic EHDV-2 reference strain negative EHDV-4 reference strain negative sheep that died as a consequence of BTV disease in Contagious ecthyma virus ﬁeld positive sample negative this outbreak were also tested. As a control, blood FMDV-A reference strain negative samples from 50 BTV-seronegative sheep in BTV-free FMDV-O reference strain negative areas were also tested by nested RT-PCR and the new FMDV-C reference strain negative real-time ﬂuorogenic RT-PCR. Similarly, 300 BTV-se- FMDV-Asia reference strain negative FMDV-SAT1 reference strain negative ronegative blood samples collected from sentinel cattle FMDV-SAT2 reference strain negative in BT-free areas during the 2004 Spanish BTV sur- FMDV-SAT3 reference strain negative veillance program were also tested. As shown in Table 3, there was complete agreement between the results obtained with both nested and real-time ﬂuorogenic some BTV isolates was investigated by comparing the RT-PCR methods. All 97 BTV blood samples found sequence of the primer and probe target regions from positive by the nested RT-PCR assay were also posi- undetected BTV isolates and a selection of the detect- tive by the real-time ﬂuorogenic RT-PCR method. ed viruses (Table 2). Although the primer binding re- BTV was detected by both methods from spleen and gions were found to allow 1 (e.g., BTV-2 SPA2000) heart blood, but not gut tissue samples. All BTV-neg- or 2 (e.g., BTV-4 and BTV-2 vaccine strains, BTV-16 ative samples yielded negative results in both assays ITL2002) mismatches without loss of detection, higher (i.e., a zero false-positive rate). stringency was found for the probe-binding sequence, The real-time ﬂuorogenic RT-PCR method can be as a single mismatch resulting in loss of detection for performed in a 96-well format, and this feature opens 1 isolate. The lowest number of nucleotide sequence the possibility of using RT-PCR methodology in the changes that rendered a strain nondetectable was found high-throughput screening of ﬁeld samples. As nucleic Table 2. Primer and probe-binding sequences from genome segment 5 from different BTV isolates, and correlation with real-time RT-PCR (RRT-PCR) detection. GenBank Primer 3 (reverse accession RRT-PCR Primer 5 Taqman MGB-probe complementary sequence) BTV isolate† number detection 606-631* 676–692* 702–723* BT2 COR2002 AY137387 not tested GTTGAGAGACAAATTAACACATGTCC CGATTCAGCTGATCAAT ATGGATGATTTTGCGAAGCATT BT4 SPA2003 AY741397 —————————————————————————— ————————————————— —————————————————————— BT4 SPA2004 DQ098093 —————————————————————————— ————————————————— —————————————————————— BT4 MOR2004 DQ098092 —————————————————————————— ————————————————— —————————————————————— BT2 SPA2000 AY741396 ———————————————————————C—— ————————————————— —————————————————————— BT2 VAC AY138895 ————————G————————————————— ————————————————— —————————————————A———— BT4 VAC DQ098095 ———————————————————————C—— ————————————————— ————C————————————————— BT4 REF DQ098096 —————————————————T—————C—— ——————————A—————— —————————————————————— BT16 IT2002 DQ098094 ————————G————————————————— ————————————————— —————————————————A———— BT16 REF DQ098100 ——A—————G————————T———————— —A——C——AA————A——— ———————————C————————C— BT24 REF DQ098102 —————————————————T—————C—— ——————————A—————C —————————————————————— BT19 REF DQ098101 ———————————G—————T—————C—— ——————————A—————— —————————N—NNN———————— BT10 REF DQ098098 ————————G——————————————C—— —————————————T——— —————————————————————— BT13 REF DQ098099 ————————G————————————————— ————————T———————— —————————————————A———— BT7 REF DQ097097 ————————G——————————————C—— —N————NNT———————— —————N——N————————A———— BT10 US Y00422 not tested ———————————————————————C—— ——————————A—————— —————————————————————— BT17 US X17041 not tested ———————————————————————C—— ————————————————— —————————————————————— BT11 US M97681 not tested ———————————————————————C—— ————————————————— ———————————————————G—— BT13 US M97762 not tested ———————————————————————C—— ————————————————— —————————————————————— BT2 US M97680 not tested ———————————————————————C—— —A——————————————— —————————————————————— BT2 Taiw AY462225 not tested ————————G————————T———————— —A——C——AA————A——— —————————————————————— BT20 AUS X56735 not tested ————————G——G——C—————T——C—— —A—————A——A——A——C ——————————————A——A———— * Nucleotide position (numbering according to GenBank AY137387 sequence). High throughput detection of bluetongue virus by real-time RT-PCR † Isolates marked as ‘‘VAC’’ are vaccine strains; ‘‘REF’’ are the prototype strains from the EU reference collection; ‘‘US’’ are US prototype strains, ‘‘Taiw’’ is a BTV-2 isolate from Taiwan, and ‘‘AUS’’ is the Australian BTV-20 isolate. 13 14 ´ Jimenez-Clavero et al. Table 3. Assessment of the performance of the real-time ﬂuorogenic RT-PCR assay for BTV detection using clinical samples from the Menorca 2003 outbreak, as well as BTV-free control samples. Real time ﬂuorogenic Nested RT-PCR RT-PCR Type of sample Origin n Positive Negative Positive Negative ovine blood outbreak 76 76 0 76 0 ovine blood BTV-free 50 — — 0 50 bovine blood outbreak 21 21 0 21 0 bovine blood BTV-free (sentinel) 300 — — 0 300 ovine heart BTV-free (sentinel) 1 1 0 1 0 ovine gut BTV-free (sentinel) 1 0 1 0 1 acid extraction represents a bottleneck when large timization of the nucleic acid extraction step may be numbers of samples are to be analyzed by RT-PCR, required for these types of samples. These ﬁndings the efﬁcacies of 2 high-throughput (96-well format) were conﬁrmed by reanalysis on a separate day. nucleic acid extraction methods were also evaluated. Based on the satisfactory correlation between results Tenfold dilutions of each of 4 different samples, 3 of from the new ﬂuorogenic real-time RT-PCR method them ﬁeld BTV-positive blood samples from the recent and the RT-PCR method currently used in our labo- BTV-4 Menorca outbreak (1 from sheep, 2 from cat- ratory, and the high-throughput capacity of the new tle), and a control BTV-2 attenuated vaccine isolate assay, a decision was made to apply the real-time ﬂuo- were subjected to nucleic acid extraction by the 3 rogenic RT-PCR in the large-scale monitoring of BTV methods in parallel, and analyzed by real-time RT- circulation in Spain during 2004 and 2005. A total of PCR for BTV RNA. As shown in Table 4, the results 47,532 blood samples (34,170 cattle and 12,142 sheep) obtained with both high-throughput RNA isolation were tested between January 2004 and June 2005. methods were comparable to those obtained with the Some of these samples came from the surveillance conventional single-tube method; if differences in start program established in Spain in 2004 using strategi- volume and elution volumes are taken into account, cally located sentinel cattle. This surveillance was suc- the high-throughput methods appear to have a slightly cessful in detecting a BTV outbreak in cattle on the higher yield than the single spin-column protocol (Ta- Spanish mainland in October 2004, 9 days before it ble 4). However, Ct values for 1 of the bovine blood was detected in sheep. In the period between October samples (cattle blood-1) and the vaccine control were 2004 and March 2005, after the BTV outbreak in pen- higher than expected at the higher dilutions for all insular Spain was ﬁrst detected, 26,787 blood samples RNA extraction methods, suggesting that further op- have been tested in our laboratory by this technique, Table 4. Comparison of the efﬁciency of 3 nucleic acid extraction methods, 2 high-throughput (96-well format) systems (Nucleospin- 96 and HPV-96), and 1 conventional single tube-spin column (HPV-col) kit. HPV-col HPV-96 (Ct*) Nucleospin-96 (Ct*) (Ct*) Sample Dilution Observed Observed Normalized† Observed Normalized† BTV-2 vaccine 1:10 28.83 27.92 25.62 26.93 23.63 1:100 33.40 31.82 29.52 33.39 30.09 1:1000 37.87 37.01 34.71 38.80 35.50 1:10000 not detected not detected not detected Sheep blood-1 1:1 33.16 35.08 32.78 35.82 32.52 1:10 35.64 38.21 35.91 38.82 35.52 1:100 not detected not detected not detected Cattle blood-1 1:1 29.50 29.76 27.40 31.15 27.85 1:10 37.21 35.01 32.71 35.56 32.26 1:100 not detected not detected not detected Cattle blood-2 1:1 32.83 33.70 31.40 34.00 30.70 1:10 35.72 38.90 36.60 37.92 34.62 1:100 not detected not detected not detected sample volume ( l) 500 100 100 elution volume ( l) 50 50 100 * Ct: Threshold cycle. † The normalization accounts for the different starting and elution volumes of each method, indicated in the bottom of the table. High throughput detection of bluetongue virus by real-time RT-PCR 15 which has reached as high as 2,635 samples in a single essarily indicate the presence of contagious virus in week. the animal. The data presented herein demonstrates that this Discussion new method is capable of detecting a wide range of Real-time RT-PCR methods (especially those based BTV isolates, including those currently circulating in on ﬂuorogenic probe hybridization, such as TaqMan the western Mediterranean. However, 9 out of 24 ref- probes,19 molecular beacons,20 and ﬂuorescence reso- erence BTV serotypes strains were not detected. Al- nance energy transfer (FRET) probes26) are gaining though EU reference strains of serotypes 4 and 16 popularity in the diagnosis of animal viral diseases. As were among those not detected, our assay did detect well as providing higher sensitivity and comparable, all 4 BTV-4 ﬁeld isolates previously identiﬁed, 1 vac- or even higher, speciﬁcity than traditional nucleic acid cine BTV-4 strain, and the only BTV-16 ﬁeld isolate detection methods, they are much faster and less cum- available. We determined the primer and probe se- bersome because they do not require electrophoresis quence requirements for detection by comparing the in agarose gels. In addition, some real-time instru- gene segment 5 sequences between detected and un- detected BTV strains. Although this region is fairly ments allow the use of 96-well plate formats, which highly conserved, it does exhibit enough sequence var- further increases the capacity and speed of analysis. iation among isolates to render some of them unde- The BTV diagnostic assay developed in this study is tectable. Our investigation found that minor nucleotide suitable for screening large numbers of samples con- changes within the probe region were more detrimen- sistent with those analyzed in recent BTV outbreaks. tal to detection by the assay than changes in the prim- The only real-time RT-PCR method for BTV previ- er-binding region. It is important to note, however, that ously described in the literature is a recent protocol the fewest changes resulting in loss of signal included that restricted its analysis to the differentiation of vac- 1 mismatch within the probe region as well as 2 mis- cine and ﬁeld variants of a single BTV serotype (BTV- matches in the primer-binding region, whereas a single 2).26 When attenuated BTV vaccines are used to pro- nucleotide mismatch within each of the primer-hybrid- tect livestock from BTV infection, the ability to dis- izing regions did not cause loss of detection. The BTV- tinguish between wild-type and vaccine strains of BTV 4 undetectable EU reference BTV-4 strain and the de- makes it possible to know whether an RT-PCR positive tectable vaccine strain differed only by 2 nucleotides result is derived from an acquired wild-type virus or within the target region, 1 of which is located in the is the result of vaccination. While the real-time RT- probe-hybridizing region. The ability of this method PCR that distinguishes wild-type infections from vac- to detect a majority of BTV serotypes is likely due to cinations is a useful diagnostic tool, it has a different its detection of RNA segment 5, rather than segment purpose than our real-time RT-PCR method, which is 2, which contains the serotype determinant sequence.27 the ﬁrst method designed to screen ﬁeld samples for a It is noteworthy that all nondetected BTV variants had wide range of BT viruses. Furthermore, the method also been previously isolated from regions that were developed in this study represents a great improvement geographically or temporally distant (or both) from the over previous conventional (i.e., nonﬂuorogenic) RT- variants currently circulating in the Mediterranean re- PCR methods, because it can be performed on a large gion. scale to monitor virus circulation in animal popula- Taken together, these results suggest that the devel- tions. The sensitivity of the new assay was comparable opment of a real-time RT-PCR with ﬂuorogenic probes to that of the established nested RT-PCR protocol (Fig. that is capable of detecting all BTV variants might not 1) that can detect 0.1 fg of BTV RNA.2 This level of be an easy task. However, our assay is currently used sensitivity is also comparable to virus isolation in em- as the standard to identify new BTV outbreaks and bryonated chicken eggs,2 which takes considerably monitor the spread of epizootic infections in Spain. longer to yield results. However, results of PCR-based This approach can be adapted to other geographic lo- BTV diagnosis, including the ﬂuorogenic method de- cations by 1) identifying the local BTV variants cir- scribed here, need to be interpreted with caution as culating in the affected area, 2) obtaining sequence BTV has been detected by RT-PCR from the blood of data from conserved regions (here we propose a short, infected calves and sheep for at least 30 days, and conserved stretch within RNA segment 5) of the BTV sometimes longer (up to 90 days) after the virus iso- variants identiﬁed, and 3) designing a set of primers lation ceases to be positive.5,16,21 Transmission of BTV and ﬂuorogenic Taqman probe for real-time RT-PCR, by biting midges fed with this noninfectious blood, taking into account that a single nucleotide mismatch even though it is RT-PCR positive, is poor.5,16,21 Thus in the probe-binding region could result in a dramatic detection of virus-speciﬁc nucleic acid by these meth- loss of detection. In this regard, the use of an MGB ods indicates recent viral infection, but does not nec- group linked to the Taqman probe can be of great help, 16 ´ Jimenez-Clavero et al. because it increases the binding strength of the probe ing of the introduction and spread of BTV in a large to the nucleic acid,18 allowing the selection of shorter population of cattle and sheep at risk. This technique target sequences. In this study, this approach was suc- is currently being used to monitor BTV outbreaks in cessful at detecting all BTV variants currently circu- Spain. lating in the western Mediterranean. Whether or not our assay could be applied to other BTV epizootics Acknowledgements occurring in different parts of the world would depend We are indebted to O. Brad Spiller for critically reviewing on the nucleotide sequence similarity within the primer the manuscript, C. Hambling for providing BTV prototype and probe-binding regions in RNA segment 5 of the strains, M. El Harrak for providing the ﬁeld samples from BTV variant(s) involved. For instance, based on nu- which the MOR2004 BTV isolate was obtained, and W. C. Wilson for providing EHDV isolates. We thank Roche Di- cleotide sequence similarity, it is predicted that only 3 agnostic Systems Spain, for advice and technical assistance out of 5 US BTV prototype strains (BTV-11, -13, and in the adaptation of the HPV nucleic acid extraction kit from -17) are likely to be detectable by our assay. single column to 96-well format. The speciﬁcity of the new real-time ﬂuorogenic RT- PCR assay was further conﬁrmed by the lack of de- Sources and manufacturers tection of non-BTV orbiviruses (AHSV and EHDV) a. BTV live attenuated vaccines (serotypes 2 and 4), OBP Onder- and phylogenetically unrelated viruses that cause sim- stepoort Biological Products, Onderstepoort, South Africa. ilar symptoms to BTV (foot-and-mouth disease virus b. Bluetongue Virus Antibody test kit, VMRD, Pullman, WA. and contagious ecthyma virus). Consequently, this as- c. EHDV North American prototype serotypes 2 and 4, kindly pro- vided by W. C. Wilson, Arthropod-Borne Animal Disease Re- say could be of great help in the differential diagnosis search Laboratory, Laramie WY. of bluetongue disease, once it is ascertained that BTV d. Foot-and-mouth disease virus, FMDV serotypes A, O, C, Asia- variants circulating in the affected region are detect- 1, SAT-1, SAT-2, and SAT-3, provided by the World Reference able by this method. Laboratory for foot-and-mouth disease, Institute for Animal The ﬂuorogenic real-time RT-PCR assay described Health, Pirbright, UK. e. High Pure Viral (HPV) Nucleic Acid extraction kit, Roche Di- here yields levels of sensitivity and speciﬁcity com- agnostics, Indianapolis, IN. parable to virus isolation and conventional nested RT- f. Nucleospin Multi-96 Virus nucleic acid extraction kit, Machin- PCR methods used for BTV nucleic acid detection. In ¨ ery-Nagel, Duren, Germany. addition, the new assay presents 2 major advantages g. 96-Well Uniﬁlter GF/F microplate, Whatman, Clifton, NJ. over conventional nested RT-PCR methods. First, it h. 800- l round-bottom 96-well Uniplate, Whatman, Clifton, NJ. i. 250- l V-bottom 96-well Uniplate, Whatman, Clifton, NJ. can be used in large-scale screening because of its abil- j. Big Dye Terminator (version 3.0) Cycle sequencing kit, Applied ity to simultaneously analyze up to 96 samples per run. Biosystems, Branchburg, NJ. In fact, the nucleic acid extraction protocol for 2 k. 3730 XL DNA Analyzer, Applied Biosystems, Branchburg, NJ. blocks of 96 samples can be carried out in less than 2 l. Sequence Analysis (version 5.1) software, Applied Biosystems, hours, which together with sample preparation and the Branchburg, NJ. m. Primer Express (version 2.0.0) software, Applied Biosystems, time required for each real-time RT-PCR (2 hours 11 Branchburg, NJ. minutes), gives a processing capacity of 192 samples n. TaqMan One-step RT-PCR Master Mix Reagents, Applied Bio- in approximately 6 hours. Currently, testing capacity systems, Branchburg, NJ. in the authors’ laboratories has been increased to up o. ABI PRISM 7000 Sequence Detection System SDS, Applied to 500 blood samples per day by using 2 real-time Biosystems, Branchburg, NJ. p. SDS 7000 (version 1.1) software, Applied Biosystems, Branch- thermal cyclers working in parallel with 96-well burg, NJ. plates. With a working scheme like this, testing ca- pacity reached 2,600 samples per week in peak periods References during the recent BTV outbreaks in Spain. Further op- 1. Akita GY, Chinsangaram J, Osburn BI, et al.: 1992, Detection timization, including a fully automated system for nu- of bluetongue virus serogroup by polymerase chain reaction. J cleic acid extraction, and adaptation to a recently avail- Vet Diagn Invest 4:400–405. able 384-well format, is ongoing in the authors’ lab- 2. Aradaib IE, Smith WL, Osburn BI, Cullor JS: 2003, A multiplex oratories and testing capacity is expected to reach PCR for simultaneous detection and differentiation of North American serotypes of bluetongue and epizootic hemorrhagic 1,000 samples per day. Second, the real-time ﬂuoro- disease viruses. Comp Immunol Microbiol Infect Dis 26:77–87. genic RT-PCR is less susceptible than nested RT-PCR 3. Aradaib IE, Schore CE, Cullor JS, Osburn BI: 1998, A nested to potential contamination problems, which often pre- PCR for detection of North American isolates of bluetongue vent many laboratories from using nested RT-PCR virus based on NS1 genome sequence analysis of BTV-17. Vet methods. Microbiol 59:99–108. 4. Billinis C, Koumbati M, Spyrou V, et al.: 2001, Bluetongue In summary, the real-time ﬂuorogenic RT-PCR virus diagnosis of clinical cases by a duplex reverse transcrip- method developed and applied in this study represents tion PCR: a comparison with conventional methods. J Virol a new analytical tool that allows large-scale monitor- Methods 98:77–89. High throughput detection of bluetongue virus by real-time RT-PCR 17 5. Bonneau KR, DeMaula CD, Mullens BA, MacLachlan NJ: parison of virus isolation, PCR assay, and in vitro feeding of 2002, Duration of viraemia infectious to Culicoides sonorensis Culicoides variipennis. Arch Virol 136:1–8. in bluetongue virus-infected cattle and sheep. Vet Microbiol 88: 22. Mellor PS, Whittmann EJ: 2002, Bluetongue virus in the Med- 115–125. iterranean basin 1998–2001. Vet J 164:20–37. 6. Bonneau KR, Mullens BA, MacLachlan NJ: 2001, Occurrence 23. ´ ´ Miranda MA, Borras D, Rincon C, Alemany A: 2003, Presence of genetic drift and founder effect during quasispecies evolution ´ of Culicoides imıcola and Culicoides obsoletus in the Balearic of the VP2 and NS3/NS3A genes of bluetongue virus upon pas- Islands. Med Vet Entomol 17:1–4. sage between sheep, cattle, and Culicoides sonorensis. J Virol 24. Oberst RD, Stott JL, Blanchard-Channell M, Osburn BI: 1987, 75:8298–8305. Genetic reassortment of bluetongue virus serotype 11 strains in 7. Bonneau KR, Zhang N, Zhu J, et al.: 1999, Sequence compar- the bovine. Vet Microbiol 15:11–18. ison of the L2 and S10 genes of bluetongue viruses from the 25. OIE Terrestrial animal health code, 13th ed., 2004: OIE Orga- United States and the People’s Republic of China. Virus Res nization International des epizooties. Ofﬁce International des 61:153–160. Epizooties: 2004, Terrestrial Animal Health Code. Thirteenth 8. Brewer AW, MacLachlan NJ: 1992, Ultrastructural characteriza- Edition, Ofﬁce International des Epizooties, Paris. tion of the interaction of bluetongue virus with bovine eryth- 26. Orru G, De Santis P, Solinas F, et al.: 2005, Differentiation of rocytes in vitro. Vet Pathol 29:356–359. Italian ﬁeld and South African vaccine strains of bluetongue 9. Brewer AW, MacLachlan NJ: 1994, The pathogenesis of blue- virus serotype 2 using real-time PCR. J Virol Methods 122:37– tongue virus infection of bovine blood cells in vitro: ultrastruc- 43. tural characterization. Arch Virol 136:287–298. 27. Pritchard LI, Gould AR: 1995, Phylogenetic comparison of the 10. Dangler CA, De Mattos CA, De Mattos CC, Osburn BI: 1990, serotype-speciﬁc VP2 protein of bluetongue and related orbi- Identifying bluetongue virus ribonucleic acid sequences by the viruses. Virus Res 39:207–20. polymerase chain reaction. J Virol Methods 28:281–292. 28. Reed LJ, Muench H: 1938, A simple method of estimating ﬁfty 11. Gould AR: 1987, The complete nucleotide sequence of blue- percent endpoints. Am J Hyg 27:493–497. tongue virus serotype 1 RNA3 and a comparison with other 29. Reid SM, Ferris NP, Hutchings GH, et al.: 2000, Primary di- geographic serotypes from Australia, South Africa and United agnosis of foot-and-mouth disease virus by reverse transcription States of America, and with other orbivirus isolates. Virus Res PCR. J Virol Methods 89:167–176. 7:169–183. 30. Samal SK, El-Hussein A, Holbrook FR, et al.: 1987. Mixed 12. Gould AR, Pritchard LI: 1990, Relationships amongst blue- infection of Culicoides variipennis with bluetongue virus sero- tongue viruses revealed by comparisons of capsid and outer coat types 10 and 17: evidence for high frequency reassortment in protein nucleotide sequences. Virus Res 17:31–52. the vector. J Gen Virol 68:2319–2329. 31. Samal SK, Livingston Jr CW, McConnell S, Ramig RF: 1987, ` 13. Gregory M, Zientara S, Hendrikx P: 2002, La ﬁevre catarrhale Analysis of mixed infection of sheep with bluetongue virus se- ´ du mouton en Corse en 2000 et 2001. Bulletin Epidemiologique rotypes 10 and 17: evidence for genetic reassortment in the ver- ´ ˆ du Ministere de l’Agriculture et de la Peche 4:1–3. tebrate host. J Virol 61:1086–1091. 14. Inoshima Y, Morooka A, Sentsui H: 2000, Detection and diag- 32. ´ Sarto I, Monteys V, Saiz-Ardanaz M: 2003, Culicoides midges nosis of parapoxvirus by the polymerase chain reaction. J Virol in Catalonia (Spain), with special reference to likely bluetongue Methods 84:201–208. virus vectors. Med Vet Entomol 17:288–293. 15. Johnson DJ, Wilson WC, Paul PS: 2000, Validation of a reverse 33. Shad G, Wilson WC, Mecham JO, Evermann JF: 1997, Blue- transcriptase multiplex PCR test for the serotype determination tongue virus detection: a safer reverse-transcription polymerase of U.S. isolates of bluetongue virus. Vet Microbiol 76:105–115. chain reaction for prediction of viremia in sheep. J Vet Diagn 16. Katz J, Alstad D, Gustafson G, Evermann J: 1994, Diagnostic Invest 9:118–124. analysis of the prolonged bluetongue virus RNA presence found 34. Stone-Marschat M, Carville A, Skowronek A, Laegreid WW: in the blood of naturally infected cattle and experimentally in- 1994, Detection of African horse sickness virus by reverse tran- fected sheep. J Vet Diagn Invest 6:139–142. scription-PCR. J Clin Microbiol 32:697–700. 17. Katz JB, Alstad AD, Gustafson GA, Moser KM: 1993, Sensitive 35. Tatem AJ, Baylis M, Mellor PS, et al.: 2003, Prediction of blue- identiﬁcation of bluetongue virus serogroup by a colorimetric tongue vector distribution in Europe and North Africa using dual oligonucleotide sorbent assay of ampliﬁed viral nucleic satellite imagery. Vet Microbiol 97:13–29. acid. J Clin Microbiol 31:3028–3030. 36. Thompson JD, Higgins DG, Gibson TJ: 1994, CLUSTAL W: 18. Kutyavin IV, Afonina IA, Mills A, et al.: 2000, 3 -minor groove improving the sensitivity of progressive multiple sequence binder-DNA probes increase sequence speciﬁcity at PCR exten- alignment through sequence weighting, position-speciﬁc gap sion temperatures. Nucl Acids Res 28:655–661. penalties and weight matrix choice. Nucleic Acids Res 22:4673– 19. Lanciotti RS, Kerst AJ, Nasci RS, et al.: 2000, Rapid detection 4680. of West Nile virus from human clinical specimens, ﬁeld collec- 37. Whittmann EJ, Mellor PS, Baylis M: 2001, Using climate data tion mosquitoes, and avian samples by a TaqMan reverse tran- to map the potential distribution of Culicoides imicola (Diptera: scriptase-PCR assay. J Clin Microbiol 38:4066–4071. Ceratopogonidae) in Europe. Rev Sci Tech Off Int Epiz 20:731– 20. Lanciotti RS, Kerst AJ: 2001, Nucleic acid sequence-based am- 740. pliﬁcation assays for rapid detection of West Nile and St. Louis 38. Zientara S, Sailleau C, Dauphin G, et al.: 2002, Identiﬁcation encephalitis viruses. J Clin Microbiol 39:4506–4513. of bluetongue virus serotype 2 (Corsican strain) by reverse-tran- 21. MacLachlan NJ, Nunamaker RA, Katz JB, et al.: 1994, Detec- scriptase PCR reaction analysis of segment 2 of the genome. tion of bluetongue virus in the blood of inoculated calves: com- Vet Rec 150:598–601.