JOURNAL OF CLINICAL MICROBIOLOGY, June 1989, p. 1320-1323 Vol. 27, No. 6 0095-1137/89/061320-04$02.00/0 Copyright C 1989, American Society for Microbiology

Detection of Bluetongue Antigens in variipennis by Enzyme Immunoassay ABDEL EL HUSSEIN,' CHARLES H. CALISHER,2 FREDERICK R. HOLBROOK,3 RANDAL J. SCHOEPP,' AND BARRY J. BEATY1* Department of Microbiology, Colorado State University,' and Division of Vector-Borne Viral Diseases, Center for Infectious Diseases, Centers for Disease Control,2 Fort Collins, Colorado 80523, and -Borne Diseases Research Laboratory, U.S. Department ofAgriculture, Laramie, Wyoming 820713 Received 6 January 1989/Accepted 1 March 1989

A solid-phase enzyme immunoassay (EIA) was developed to detect bluetongue (BT) virus antigens in infected cell cultures and in suspensions of infected Culicoides variipennis midges. The technique was equally sensitive for detecting the five U.S. BT virus serotypes (2, 10, 11, 13, and 17) in cell cultures. EIA reliably detected about 3.8 loglo median tissue culture infective doses per ml of BT virus in infected cell culture lysates. The EIA readily detected virus antigens in pools of midges infected with BT serotypes 2, 10, 11, 13, and 17 and contained 2.3 to 4.8 loglo median tissue culture infective doses per ml of BT virus. The technique was sensitive enough to detect a single infected midge in a pool with 99 noninfected midges. The EIA may be a sensitive and rapid alternative to virus isolation for surveillance of BT in vector populations.

Bluetongue (BT) is an arthropod-borne virus disease of Virus stocks were titrated in 96-well microculture plates wild and domestic ruminants. BT virus, the etiologic agent of (Becton Dickinson and Co., Lincoln Park, N.J.; 23). the disease, is the prototype virus of the genus Orbivirus, Virus purification. The BT-10 virus used to immunize family (19). Twenty-four serotypes of the virus guinea pigs was purified by polyethylene glycol precipita- are recognized worldwide, and five serotypes (BT-2, -10, tion, tartrate gradient centrifugation, and pelleting through a -11, -13, and -17) have been isolated in the United States (2, 30% sucrose cushion. Purified virus was stored in multiple 6). The disease is widely distributed and assumes additional portions at -70°C. Purified virus and immunoglobulin G economic importance because of restrictions on shipment of protein concentrations were assayed by using a dye-binding and germ plasm products from areas with endemic protein assay (Bio-Rad Laboratories, Richmond, Calif.). BT (21). . (i) Capture . Guinea pigs were hyper- Culicoides variipennis midges (Diptera: ) immunized by a modification of the method of Halonen et al. are the principal vectors of BT virus in the United States (7). In brief, the animals were inoculated intradermally four (17). Procedures currently used for surveillance of BT vi- times at 3-week intervals. The first three injections contained ruses in vectors are unsatisfactory. Virus isolation in 150 p.g of polyethylene glycol-precipitated, sucrose gradient- embryos or in cell lines is the method most fre- purified BT-10 proteins mixed with an equal volume of quently used to detect BT viruses in arthropod vectors (18). Freund incomplete adjuvant. The fourth inoculation con- Although sensitive, virus isolation is laborious, slow, and tained 15 ,ug of virus proteins in the same adjuvant. A expensive, and it may require elaborate laboratory and booster dose, containing 75 ,ug of purified virus proteins containment facilities (18). without adjuvant, was given intramuscularly 3 weeks after Rapid diagnosis of BT virus infections is required to the last inoculation. Animals were bled 1 week later. Serum provide veterinary health authorities adequate time to de- specimens were separated and precipitated with 50% (vol/ velop and deploy control measures, especially during epi- vol) ammonium sulfate, dialyzed against phosphate-buffered zootics. Enzyme immunoassays (EIA) have been used to saline (pH 7.4), and stored at -20°C until used. Purified detect rapidly, efficiently, and cost effectively a variety of immunoglobulin G was obtained from these serum samples in infected (4, 10-13, 22, 24). We by using DEAE-Affi-Gel Blue chromatography (Bio-Rad; report here the development of an antigen capture EIA to 14). detect BT virus antigens in infected cell cultures and in C. (ii) Detector antibody. Antibodies to BT-10 virus were variipennis pools. prepared by hyperimmunization of mice as described by MATERIALS AND METHODS Brandt et al. (3). Ascitic fluid was collected by paracentesis 38 days after immunization began and was stored at -20°C. Virus propagation. The following virus serotypes were (iii) Conjugate. Horseradish peroxidase-conjugated goat propagated in baby hamster kidney (BHK-21) cells: BT-2 anti-mouse (heavy and light chains) antibodies were pur- (strain Ona-B), BT-10 (strain BT-8), BT-11 (station strain, chased from Jackson ImmunoResearch Laboratory (West 1955), BT-13 (strain 6741B), BT-17 (strain 262), and epi- Grove, Pa.). zootic hemorrhagic disease virus (Alberta strain). BHK-21 C. variipennis. Midges used were from the AK colony monolayers in 75-cm2 tissue culture polystyrene flasks (Bruneau strain), which originated from field material col- (Corning Glass Works, Corning, N.Y.) were inoculated with lected in Idaho in 1973 (16). This colony is maintained at the 0.1 PFU and incubated at 34°C. Cells were harvested when U.S. Department of Agriculture Arthropod-Borne Animal approximately half of them exhibited cytopathic effects. Diseases Research Laboratory. The midges were infected with the five serotypes of BT virus by intrathoracic inocu- * Corresponding author. lation with fine-glass needles (15). Inoculated midges were 1320 VOL. 27, 1989 ANTIGEN CAPTURE EIA FOR BLUETONGUE VIRUS 1321 maintained for at least 10 days at 26°C and 40%c relative TABLE 1. Detection of BT virus antigen in infected humidity and then sacrificed and stored at -70°C until used cell cultures by EIA in EIA. For the preliminary tests, the head was severed from Mean absorbance' each midge and examined for virus antigen by immunofluo- Virus dilution BT-13 BT-17 rescence (1). Thoraces and abdomens from midges contain- BT-2 BT-10 BT-11 (6.8)'> (6.9) (6.8) (7.0) (7.0) ing virus antigen were combined with those from uninfected control midges to produce pools of known composition. In 10 00.343 0.369 0.311 0.331 0.343 later tests, whole midges were used. Pools were triturated 10 -2 0.227 0.242 0.175 0.185 0.228 with a mortar and pestle in 1 ml of L-15 (Leibovitz) growth 10 3 0.124 0.130 0.119 0.115 0.131 4 medium with 8% heat-inactivated fetal bovine serum and 10 0.104 0.101 0.105 0.093 0.10( antibiotics. Pools were titrated for infectious virus in BHK- 10 S 0.097 0.094 0.097 0.090 0.097 21 cells in 96-well microculture plates (23). Uninfected cells 0.095 EIA. In preliminary tests, optimal concentrations of re- " Diagnostic criterion: -0.107 (mean absorbance value plus 3 standard agents for EIA were determined by checkerboard titrations. deviations of uninfected control cells). Those selected were the highest dilutions giving maximal " Numbers in parentheses indicate stock virus titer (log,( median tissue absorbance values and lowest background noise. Al] anti- culture infective doses per milliliter). bodies were diluted in phosphate-buffered saline (pH 7.2) containing 4% fetal serum. tests were bovine All performed experiment to maximize possible cross-reactions. No cross- in 96-well polystyrene plates (Immunolon-2; Dynatech Lab- reactivity was observed even the hemor- oratories, Inc., The capture though epizootic Alexandria, Va.). immunoglob- rhagic disease virus stock contained substantially more ulin G was diluted to 500 ng per in of protein 100 infectious virus (7.5 log1, TCID,(5ml) than did the BT virus carbonate-bicarbonate buffer (pH 9.6) and used to coat each preparations (Table 2). well of a 96-well plate (18 to 24 h at room temperature). After Detection of BT virus antigen in parenterally infected coating, the wells were washed five times (using a 500-ml washing bottle) with phosphate-buffered saline (pH 7.2) midges. To determine the minimum number of virus-infected midges detectable by EIA, pools containing one or three containing 0.05% Tween 20 (Sigma Chemical Co., St. Louis, BT-10 virus-infected and different numbers of noninfected Mo.). Portions (50 iul) of virus stock dilutions in L-15 growth midges were constructed (Table 3). These infected midges medium or undiluted Culicoides pools were incubated in the were first tested by immunofluorescence, and thoraces and wells for 1 h at 37°C. Plates were washed as described above, abdomens of those that contained virus antigens were incor- and 50 ,ul of pretitrated mouse anti-BT-10 ascitic fluid diluted porated into the pools of noninfected midges (trial 1). The 1:5,000 in phosphate-buffered saline-4% fetal bovine serum technique was sufficiently sensitive to detect a single in- was then added to each well. After 1 h of incubation at 37°C, fected midge in a pool of 99 noninfected midges (Table 3). the were plates washed five times (as described above), and The titer of one pool was 2.8 log,( TCID5,/ml. The subse- 50 ,u1 of goat anti-mouse antibody conjugated to horseradish quent experiment (trial 2) used the whole body (head, peroxidase (1:10,000) was added to the wells. The plates thorax, and abdomen) of an infected midge and different were incubated for 1 h at 37°C and then washed five times. A numbers of noninfected the 100->tl amount of substrate midges (Table 3). Including r2,2'-azino-bis(3-ethylbenzthia- heads of infected midges, which contain substantial virus zoline 6-sulfonic acid); Sigma] in phosphate-citrate buffer antigen (1), increased the absorbance values for the pool (pH 5.0), containing 2 ,ul of 30%, HO, per 10 ml, was added, containing 1 infected and 49 noninfected midges but seemed and plates were incubated in the dark at room temperature to have little effect on the absorbance values of other pools until green color developed in the wells approximately 30 to (Table 3). 40 min later. was a A410 measured with Dynatech MR600 Midges infected with one of the other BT virus serotypes microplate reader. Each sample was tested in triplicate, and were tested in pools of 50 that contained a single, putatively the three absorbance values were averaged. The mean EIA infected midge (head, thorax, and abdomen) (Table 4). Mean absorbance value was if of the sample considered positive it absorbance values ranged from 0.253 for one BT-17 virus exceeded the mean absorbance value plus 3 standard devi- pool to 0.115 for one BT-13 virus pool. In instances when ations of the corresponding negative control (uninfected cell virus was not isolated, the pools were also negative by EIA culture lysate or uninfected Culicoides pool). The positive (Table 4). Thus, the EIA detected in the midges al of the BT control was a BT-10 virus-infected cell culture lysate. serotypes tested. RESULTS TABLE 2. Specificity of EIA for detecting BT virus antigen Detection of BT virus antigens in infected cell cultures. The EIA was essentially equivalent in sensitivity for detecting Mean absorbance' Virus the five U.S. BT virus serotypes (Table 1). The EIA detected dilution BT-10 BT-11 BT-17 EHD 3.8 logl" of the median tissue culture infective dose (TCID50) (6.6)" (5.9) (6.4) (7.5) per ml of BT-2 and BT-11 viruses, 3.9 log1, of the TCID50 per ml of BTV-10, and 4 log1, of the TCID5, per ml of BT-13 and Undiluted 0.629 0.637 0.631 0.120 10-1 0.390 0.480 0.356 0.090 BT-17 Since the criterion was an (Table 1). diagnostic 10 2 0.153 0.192 0.144 0.090 absorbance value of-0.107, the absorbance values suggest 10 3 0.088 0.100 0.102 that the limit of sensitivity with tissue culture-propagated 10 4 0.083 0.078 0.081 virus was closer to 3 log1o. Specificity of EIA. The specificity of the test for detecting " Diagnostic criterion: -0.128 (mean absorbance value plus 3 standard deviations of the uninfected control cells). EHD. Epizootic hemorrhagic BT viruses in cell cultures was demonstrated by using a disease virus. related orbivirus, epizootic hemorrhagic disease virus. Un- " Numbers in parentheses indicates stock virus titers (log,( median tissue diluted virus antigen preparations were included in this culture infective doses per milliliter). 1322 EL HUSSEIN ET AL. J. CLIN. MICROBIOL.

TABLE 3. Detection by EIA of BT-10 virus antigen TABLE 5. Effect of freezing and thawing on detection by EIA of in pools of C. variipennis BT-10 virus antigen in pools of C. variipennis No. of midges Mean absorbanceb Virus titer range No. of in pool'aiu ie ag midges Virus titer Mean absorbance at given Virus titer (1log, TCID5d/ml) in (1°glo freeze-thaw cycle (1°glo + - Trial 1 Trial 2 pool TCIDo/ml)b TCID5o/ml)C 1 0 0.219 0.230 2.8-3.3 + - 1 2 3 4 5 6 1 9 0.210 0.205 2.5-2.8 1 49 0.154 0.283 3.0-3.3 1 0 0.224 0.233 NTd NT 0.218 NT 1 99 0.201 0.206 2.8-3.8 1 9 0.217 0.219 NT NT 0.188 NT 3 0 0.267 NT 4.5 1 49 3.05 0.281 0.271 0.265 NT NT 0.257 3 7 0.295 NT 3.8 1 99 3.8 0.199 0.185 NT NT 0.177 NT 3 47 0.270 NT 3.8 0 soe 0.072 0.082 0.082 NT 0.084 0.084 3 97 NT 2.8 0.204 a +, Infected; -, uninfected. 0 1 0.154c NT b Before freezing. 0 3 0.130 NT cAfter six freeze-thaw cycles; -, no detectable virus. 0 10 0.150 0.077 d NT, Not tested. 0 50 0.082 0.072 e Diagnostic criterion: greater than or equal to the mean absorbance value 0 100 0.116 0.088 plus 3 standard deviations of respective uninfected control pool. a +, Infected; -, uninfected. b In trial 1, only thoraces and abdomens of infected midges were assayed. detection is comparable to that reported for other viruses (5, In trial 2, whole midges (head, thorax, and abdomen) were assayed. NT, Not 8, 9, 20). However, the test seemed to be more sensitive for tested. detection of BT virus antigens in C. variipennis pools; the C The diagnostic criterion for each pool size in each trial is also presented. sensitivity limit was around 2.3 log1i TCID50/ml. This in- creased sensitivity suggests that more virus antigens or Effect of freeze-thawing on antigen detection in C. variipen- noninfectious virus are produced in the vector than in cell nis pools. The number of freezing-thaw cycles did not seem cultures. to affect the ability of EIA to detect BT-10 virus antigens in The specificity of the EIA for BT viruses was indicated by the pools (Table 5). Pools yielded positive EIA results after the absence of cross-reactivity with epizootic hemorrhagic six such cycles, when virus could no longer be isolated disease virus. This EIA is a BT virus group-reactive test and (Table 5). may be useful in areas where multiple BT virus serotypes are present. A BT virus serotype-specific test could be obtained DISCUSSION by using BT serotype-specific monoclonal antibodies as the 11;3qIAs are simple, sensitive, rapid, and relatively inex- detector antibodies. pensive to perform (11, 13). These advantages are particu- The EIA seems to provide many advantages for surveil- larly desirable for testing of arthropods for arboviruses lance of BT viruses in vectors. The test is sensitive, specific, because many potential vectors usually are assayed, and the and capable of providing diagnostic results within hours rate of vector infection is low (25). rather than weeks. Rapid diagnosis is extremely valuable to Our studies indicate that EIA can be used to detect BT veterinary health practitioners who must respond quickly to virus antigens in both cell cultures and infected C. variipen- impending or ongoing outbreaks of disease. Furthermore, as nis. The technique detected all five U.S. serotypes of BT many as six freeze-thaw cycles did not appear to affect virus in cell cultures, which indicates that the capture antigen detection in the midge pools. Thus, expensive cold antibody is group reactive. EIA also was sensitive enough to chains or other procedures required to isolate virus may not detect a single infected midge in a pool with 99 noninfected be necessary. The EIA may also be beneficial for prelimi- ones. nary screening of pools before other, more complicated and In experiments using cell culture lysates, the sensitivity expensive procedures (e.g., virus isolation) are used. limit was approximately 3.8 log1o TCID5dml. This level of EIA is a promising field technique for detecting BT virus-infected arthropods. However, before the technique is TABLE 4. Detection of BT virus antigen in widely applied in surveillance programs, its diagnostic ef- pools of 50 C. variipennis ficacy needs to be determined in a prospective study com- paring the sensitivity, specificity, and cost of this technique No. of midges BT-2 BT-11 BT-13 BT-17 versus virus isolation in chicken embryos and cell lines. in pool ACKNOWLEDGMENTS - Absor- Virus Absor- Virus Absor- Virus Absor- Virus bance titerb bance titer bance titer bance titer We are grateful to Charles N. McKinnon (Arthropod-borne Ani- mal Diseases Research Laboratory) for his assistance in this project. 1 49 0.156 2.3 0.076 C 0.148 4.8 0.213 4.0 This work was supported by U.S. Department of Agriculture 1 49 0.201 3.0 0.136 3.0 0.142 3.3 0.253 4.5 grant 88-34116-3660, by the Colorado State University Agricultural 1 49 0.202 3.0 0.070 - 0.158 3.0 0.180 3.6 Experiment Station, and by the U. S. Department of Agriculture 1 49 0.241 3.3 0.159 3.3 0.115 2.3 0.216 3.8 Animal Health and Disease Program. A. El Hussein was supported 1 49 0.214 3.0 0.153 3.3 0.129 NDd 0.214 4.0 by the King Faisal Foundation. R. J. Schoepp was supported by 0 50e 0.097 0.083 0.064 0.083 grant 84-07023 from the U.S. Department of Agriculture Coopera- tive State Research Service Agricultural Sciences National Needs a +, Infected; -, Uninfected. Graduate Fellowship Program. b Log1o tissue culture infective doses per milliliter. C-, Uninfected. LITERATURE CITED d ND, Not determined. e Diagnostic criterion: greater than or equal to the mean absorbance value 1. Ballinger, M. E., R. H. Jones, and B. J. Beaty. 1987. The plus 3 standard deviations of respective uninfected control pool. comparative virogenesis of three serotypes of bluetongue virus VOL. 27, 1989 ANTIGEN CAPTURE EIA FOR BLUETONGUE VIRUS 1323

in Cidicoides i'ariipennis (Diptera: Ceratopogonidae). J. Med. 13. Hildreth, S. W., B. J. Beaty, J. M. Meegan, C. L. Frazier, and Entomol. 24:61-65. R. E. Shope. 1982. Detection of LaCrosse antigen in 2. Barber, T. L. 1979. Temporal appearance, geographic distribu- pools: application of chromogenic and fluorogenic tion and species of origin of bluetongue virus serotypes in the enzyme immunoassay system. J. Clin. Microbiol. 15:879-884. United States. Am. J. Vet. Res. 40:1654-1656. 14. Jamnback, T. L., B. J. Beaty, S. W. Hildreth, K. L. Brown, and 3. Brandt, W. E., E. L. Buescher, and F. M. Hetrick. 1967. C. B. Gundersen. 1982. Capture immunoglobulin M system for Production and characterization of arbovirus antibody in mouse rapid diagnosis of La Crosse (California encephalitis) virus ascitic fluid. Am. J. Trop. Med. Hyg. 16:339-347. infections. J. Clin. Microbiol. 16:577-580. 4. Calisher, C. H., P. Auvinen, C. J. Mitchell, C. M. Rice, V. 15. Jochim, M. M., and R. H. Jones. 1966. Multiplication of Hukkanen, and T. Hyypia. 1987. Use of enzyme immunoassay bluetongue virus in Culicoides variipennis following artificial and nucleic acid hybridization for detecting in infection. Am. J. Epidemiol. 84:241-246. infected mosquitoes. J. Virol. Methods 17:229-236. 16. Jones, R. H., and N. M. Foster. 1978. Relevance of laboratory colonies of the vector in arbovirus research. Cidicoides vari- 5. Collins, J. K., A. C. Butcher, Y. A. Teramoto, and S. Winston. ipennis and bluetongue. Am. J. Trop. Med. Hyg. 27:168-177. 1985. Rapid detection of bovine herpesvirus type 1 antigens in 17. Jones, R. H., A. J. Luedke, T. E. Walton, and H. E. Metcalf. nasal swab specimens with an antigen capture enzyme-linked 1981. Bluetongue in the United States. An entomological per- immunosorbent assay. J. Clin. Microbiol. 21:375-380. spective toward control. World Anim. Rev. 38:2-8. 6. Gibbs, E. P. J., E. C. Greiner, W. P. Taylor, T. L. Barber, J. A. 18. Luedke, A. J., and R. H. Jones. 1972. Storage of bluetongue House, and J. E. Pearson. 1983. Isolation of bluetongue virus virus-infected Culicoides iariipennis. Am. J. Vet. Res. 33: serotype 2 from in Florida: a serotype of bluetongue virus 1875-1878. hitherto unrecognized in the Western hemisphere. Am. J. Vet. 19. Mathews, R. E. F. 1982. Classification and nomenclature of Res. 44:2226-2228. viruses. Intervirology 17:1-179. 7. Halonen, P., H. Sarkkinen, P. Arstila, E. Hjertsson, and E. 20. Monath, T. P., and R. R. Nystrom. 1984. Detection of yellow Torfason. 1980. Four-layer radioimmunoassay for detection of fever virus in serum by enzyme immunoassay. Am. J. Trop. adenovirus in stool. J. Clin. Microbiol. 11:614-617. Med. Hyg. 44:151-157. 8. Harmon, M. W., S. Drake, and J. Kasel. 1979. Detection of 21. Mulhern, F. J. 1985. Economic impact of bluetongue and related adenovirus by enzyme-linked immunosorbent assay. J. Clin. orbiviruses, Western hemisphere, p. 21-25. ln T. L. Barber and Microbiol. 9:342-346. M. M. Jochim (ed.), Bluetongue and related orbiviruses. Alan 9. Harmon, M. W., and K. M. Pawlik. 1982. Enzyme immunoas- R. Liss, Inc., New York. say for direct detection of influenza type A and adenovirus 22. Niklasson, B. S., and T. P. Gargan. 1985. Enzyme-linked antigens in clinical specimens. J. Clin. Microbiol. 15:5-11. immunosorbent assay for detection of virus 10. Hildreth, S. W., and B. J. Beaty. 1984. Detection of Eastern antigen in mosquitoes. Am. J. Trop. Med. Hyg. 34:400-405. equine encephalomyelitis virus and Highlands J virus antigens 23. Pantuwatana, S., W. Thompson, D. Watts, and R. Hanson. 1972. within mosquito pools by enzyme immunoassay (EIA). 1. A Experimental infection of chipmunks and squirrels with La- laboratory study. Am. J. Trop. Med. Hyg. 33:965-972. Crosse and trivittatus viruses and biological transmission of 11. Hildreth, S. W., and B. J. Beaty. 1987. Economic comparison of LaCrosse virus by Aedes triseriatits. Am. J. Trop. Med. Hyg. enzyme immunoassay and virus isolation procedure for surveil- 21:476-481. lance of arboviruses in mosquito populations. J. Clin. Micro- 24. Tsai, T. F., R. A. Bolin, M. Montoya, R. E. Bailey, D. B. Francy, biol. 25:976-981. M. Jozan, and J. T. Roehrig. 1987. Detection of St. Louis 12. Hildreth, S. W., B. J. Beaty, H. K. Mayfield, R. F. Gilfillan, and encephalitis virus antigen in mosquitoes by capture enzyme B. J. Rosenau. 1984. Detection of Eastern equine encephalomy- immunoassay. J. Clin. Microbiol. 25:370-376. elitis virus and Highlands J virus antigens within mosquito pools 25. Walter, S. D., S. W. Hildreth, and B. J. Beaty. 1980. Estimation by enzyme immunoassay (EIA). Il. Retrospective field test of of infection rates in populations of organisms using pools of the EIA. Am. J. Trop. Med. Hyg. 33:973-980. variable size. Amr. J. Epidemiol. 112:124-128.