J Vet Diagn Invest 2:44-50 (1990)
Laboratory diagnosis of African horse sickness: comparison of serological techniques and evaluation of storage methods of samples for virus isolation Carol House, Peter E. Mikiciuk, Mary Lou Beminger Abstract. Five serological methods of diagnosing African horse sickness were evaluated, using a battery of serum samples from experimental horses vaccinated and challenged with each serotype of African horse sickness virus (AHSVl through AHSV9): agar gel immunodiffusion (AGID), indirect fluorescent antibody (IFA), complement fixation (CF), virus neutralization (VN), and enzyme-linked immunosorbent assay (ELISA). The 5 tests were also compared using a panel of field samples, convalescent equine sera with antibodies to domestic equine viral diseases, and sera from horses awaiting export. The ELISA described in this paper was group specific. It did not require calibration with a standard positive serum but did yield elevated values with negative sera that were repeatedly frozen and thawed or heat inactivated. The IFA test was sensitive but could not be used on some field sera as the control cells exhibited fluorescence, possibly due to the animal being recently vaccinated with cell culture material. Sixty-two experimental sera were compared by VN, CF, AGID, and ELISA. Forty sera, 10 positive and 30 negative, were correctly classified by the 5 serologic assays. The 22 remaining sera gave mixed reactions. The AGID had no false positive results but had false negative results for up to 20% of the samples, depending upon the comparison. The VN, CF, and ELISA were similar in their variability. The length of time that virus could be recovered from a viremic blood sample was compared in an evaluation of storage methods for virus isolation samples. Washed erythrocytes were held at 4 C, washed erythrocytes plus stabilizer were held at -70 C, and blood that was drawn into a preservative (oxalate/phenol/glycerol) was held at 4 C. Virus was isolated for 12 months from a sample stored as washed erythrocytes at 4 C but for only 6 months from the same blood stored by the other 2 methods.
Detection of antibody to African horse sickness virus (AHSV) has traditionally utilized agar gel immunodiffusion (AGID), complement fixation (CF),2,12 and virus neutralization (VN)11 techniques. The indirect fluorescent antibody (IFA)5 assay and enzyme-linked immunosorbent assay (ELISA) calculated using a reference serum17 have been reported, but both have lacked standardization and acceptance. The hemagglutination test15 has rarely been used. In this report we describe an improved ELISA test and delineate its parameters, compare the AGID, CF, VN, IFA, and ELISA diagnostic techniques, and compare the common methods of blood storage to preserve AHSV. Materials and methods Experimental horses. Fifteen clinically healthy adult horses and ponies obtained from local farms were housed in high containment (P3 or greater) animal rooms at the Foreign Animal Disease Diagnostic Laboratory (FADDL). Animals were observed, and temperatures were taken daily. Animals
From the Foreign Animal Disease Diagnostic Laboratory, USDA, APHIS, PO Box 848, Greenport, NY 11944. Received for publication August 7, 1989.
were bled and then each was vaccinated with 1 of 9 monovalent vaccines (1 ml subcutaneously). Because of unexpected results, 2 additional horses were vaccinated, 1 with AHSV5 and 1 with AHSV6. Blood samples were collected on days 7, 14, and 21 postvaccination (dpv); some horses were also sampled on 10 and 17 dpv. The immunity of the horses was challenged on 21 dpv by subcutaneous inoculation with 2 ml of homologous challenge virus. Blood samples were subsequently collected at 3-4 day intervals, up to day 32 postchallenge (dpc). Four additional horses were each inoculated subcutaneously with 1 ml of a challenge virus, AHSV1, -2, -3, or -4. Blood samples were drawn on days 0, 7, 10, and 14 after inoculation. Horses found dead or euthanized when severely ill were necropsied, and tissues were examined histopathologically. Clinically asymptomatic horses were necropsied for comparison. Vaccines. Monovalent vaccines were prepared in 1969 (Y. Ozawa and A. H. Dardiri, unpublished data) by passaging each virusa 10l-103 times in suckling mice by intracranial (IC) inoculation.l,9 The brain tissue suspension vaccine was adjusted to contain l04 mouse lethal doses-50% (MICLD50) per ml and was lyophilized in 2-ml quantities. The vaccine strains were designated AHSVl (A50l), AHSV2 (OD), AHSV3 (L), AHSV4 (Vryheid), AHSV5 (VH), AHSV6 (114), AHSV7 (Karen), AHSV8 (18/60), and AHSV9 (60). Vaccines were determined to be potent in 1988 by intracranial inoculation (IC) titration in suckling mice (SM) (C. House and
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness
45
Table 1. Serological, pathological, and clinical observations of experimental horses, AHSV1.
Table 2. Serological, pathological, and clinical observations of experimental horses, AHSV2.
J. A. House, unpublished data), with log,, titers of 2.8-4.0 SMICLD50. Challenge viruses. Viruses were prepared in 1968 and 1969 (Y. Ozawa and A. H. Dardiri, unpublished data) by passaging each virusb l-6 times in SM by IC inoculations. The brain tissue suspension (10%) was lyophilized in 2-ml quantities and held at 4 C. The strains were designated AHSVl 29/62 (4 mouse passages, 103.5 SMICLD50), AHSV2 MAT (2 mouse passages, 103.5 SMICLD50), AHSV3 13/63 (1 mouse passage, 104 SMICLD50), AHSV4 12/60 (5 mouse passages, 103 SMICLD50 ), AHSV5 30/62 (4 mouse passages, 104.5 SMICLD50), AHSV6 3887 (6 mouse passages, l03 SMICLD50), AHSV7 3893 (4 mouse passages, l03 SMICLD50), AHSV8 18/60 (4 mouse passages, l03 SMICLD50), AHSV9 7/60 (4 mouse passages, 103.5 SMICLD50). Domestic horse sera. A panel of convalescent sera from horses recovering from diseases found in the USA and a panel of normal horse sera randomly assembled from samples submitted to the National Veterinary Services Laboratories (NVSL) were supplied.c A third panel of 45 horse sera was assembled from samples submitted to the FADDL for testing prior to exportation. Recovery of AHSV from whole blood or washed erythrocytes. One horse, inoculated with AHSVl challenge virus, was bled on day 6 postinoculation, and its heparinized blood was stored as washed erythrocytes at 4 C, as washed erythrocytes plus an equivalent volume of virus stabilizer (20% NZ Amine NS,d 0.07% L-glutamic acid, 0.125% K2HPO4; 0.05% KH2PO4; 20% lactose at pH 7.2) at -70 C, and as heparinized blood plus an equal volume of OPG (0.5% ox-
alate, 0.5% phenol, 50% glycerol)6 at 4 C.14 On days 0, 90, 180, 270, and 360 of storage, blood stored by each method was used for virus isolation. The washed erythrocytes at 4 C and washed erythrocytes plus virus stabilizer at -70 C were inoculated onto confluent monolayers of Vero cell cultures, incubated at 37 C, and observed daily for development of CPE. Cells were harvested by scraping cells into the media when CPE appeared or at 1 wk postinoculation. One milliliter of cell/media suspension was inoculated onto a second confluent monolayer of Vero cell culture, incubated at 37 C, and observed daily for CPE for 7 days. The OPG samples were treated in the same manner after removal of the OPG fluid by centrifugation and 3 washes with sterile 0.01 M phosphate buffered saline (PBS) at pH 7.4. Agar gel immunodiffusion (AGID) test. Antigen was prepared by growing AHSV5 in Vero cell cultures, harvesting cells and fluid by freezing and thawing when the cytopathic effect (CPE) was observed in at least 80% or greater of the monolayer. The fluid was clarified by centrifugation at 1,300 x g for 30 min at 4 C, and supernatant fluid was filtered through a prefilter and a 5.0-µm filter. The filtrate was concentrated approximately 100-fold by dialysis using a 14,000 nominal molecular weight limiting membrane. The fluid was further concentrated 2-4-fold by dialysis against solid polyethylene glycol. The resulting antigen was titrated by preparing 2-fold dilutions in PBS and testing against known positive and negative sera in the AGID test. An optimal dilution was determined by observing the intensity and central location of the precipitin line after after 24 hr incubation at room temperature.
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
46
House, Mikiciuk, Beminger
Table 3. Serological, pathological, and clinical observations of experimental horses, AHSV3.
Table 4. Serological, pathological, and clinical observations of experimental horses, AHSV4.
Test plates were prepared by dissolving 0.8% agarose in 0.85% sodium chloride (NaCl) solution and pouring 18 ml/ 100 mm petri dish. Wells 3 mm in diameter and spaced 2 mm apart were made in the solidified agar; approximately 40 µ1 of test serum, positive control serum, or antigen were placed in wells in a standard AGID pattern (antigen centrally and alternating positive control serum and test sera peripherally). The plates were placed in a humidified atmosphere at 25 C for 24 hr and read in a darkened room, using a light source angled beneath the plate. A test serum was considered positive if it formed a line of identity or caused a specific bend with the positive control precipitin line. Complement fixation (CF) test. Sera were diluted 1:1011 and inactivated at 56 C for 30 min. Titrations of positive sera were performed using additional 2-fold dilutions. The CF antigen and control antigen (normal mouse brain) were diluted the same, with the CF containing 4-8 antigen units in veronal buffered saline containing 1% gelatin (VBSG); guinea pig complement was diluted in VBSG to contain 4 complement hemolytic units-50% (CH50). Sera, complement, and antigen were reacted in 96-well round-bottom microtiter plates at 4 C for 18 hr. Hemolysin was diluted to contain 2 hemolytic units and used to sensitize washed sheep erythrocytes (SRBC’s). The SRBC’s were standardized to 3% concentration and added to all wells on the microtiter plate. The test was incubated for 30 min at 37 C. The plates were then centrifuged (200 x g), and the wells were scored for the presence of hemolysis. The following controls were used: 1) serum and complement; 2) serum and SRBC; 3) CF antigen
and control antigen each with 4 CH50, 2 CH50, and 1 CH50, of complement; 4) CF antigen and SRBC’s; 5) control antigen and SRBC’s; 6) complement dilutions of 4 CH50, 2 CH50, and 1 CH50; and 7) SRBC’s. The inverse of the highest dilution of serum specifically fixing complement with the CF antigen was called the titer. Antigens for complement fixation (CF) test. The vaccine virus AHSV9 (S2) was propagated in suckling mice by IC inoculation. Brain tissue was collected from moribund mice for antigen preparation. Fifty grams of tissue were homogenized with 200 ml of 8.5% sucrose at 4 C, and the suspension was added dropwise to 4 liters of acetone at 4 C. The precipitate was collected by centrifugation at 900 x g for 5 min. The supernatant fluid was discarded, and the pellet was reextracted in 4 liters of acetone at 4 C for 1 hr. Sediment was collected by centrifugation at 900 x g for 5 min. Supernatant fluid was discarded and the pellet allowed to dry at 25 C. The resulting powder was rehydrated with 100 ml of 0.85% NaCl and held for 1 hr in an ice bath. The preparation was clarified by centrifugation at 12,100 x g for 30 min and the pellet discarded. Supernatant fluid, called CF antigen, was aliquoted, lyophilized, and stored at 4 C. A similar preparation was made from normal mouse brain and called control antigen. Indirect fluorescent antibody (IFA) slide test. Vero cells were grown on chamber slidese incubated at 37 C and 95% humidity until the cells were 80% confluent. Each chamber was inoculated with virus sufficient to produce scattered foci (3-8 foci/field at 160 x magnification) after 18-22 hr of incubation. Negative-control chamber slides were prepared from uninoculated cell cultures. Slides were decanted and fixed in
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness
47
Table 5. Serological, pathological, and clinical observations of experimental horses, AHSV5.
Table 6. Serological, pathological, and clinical observations of experimental horses, AHSV6.
acetone at 4 C for 5 min. Fixed, dried slides were stored for up to 18 mo at -70 C. Experimental sera were diluted 1:10 for screening. Slides of both positive or infected and negative or control chambers were warmed to 25 C, and diluted serum was applied to each chamber. The slides were then incubated at 25 C in a humid chamber for 30 min and subsequently washed in PBS for 5 min. Rabbit anti-horse IgG FITC conjugatef was diluted 1:32 in PBS and applied to the slide for 30 min at 25 C. Slides were rinsed in PBS and distilled water and dried. Specificity of immunofluorescence was evaluated by first examining the negative control slide. If no fluorescence was seen in the control chamber, the test chamber was examined for the presence of fluorescent inclusion bodies, seen as bright perinuclear bodies. Experimental samples were tested on slides infected with the same serotype as the vaccination and/or challenge (IFA homologous) and on slides infected with different serotypes from the vaccination and/or challenge (IFA heterologous). Virus neutralization (VN) test. Stocks of viruses were prepared by passaging the following mouse brain suspension stock virusg once in Vero cell culture followed by freezing and thawing at the time of maximal CPE. After clarification by centrifugation at 200 x g for 10 min, 30% virus stabilizer was added and aliquots were stored at -70 C. Mouse brain passage 102, monkey kidney cell line13 (MS) passage 5 stocks were used for AHSV1 (A50l), AHSV2 (OD), AHSV3 (L), AHSV4 (Vryheid), AHSV5 (VH), AHSV6 (114), AHSV7 (Karen MS), and AHSV9 (S2). The stock of AHSV8 (18/60) (mouse brain passage 102, MS passage 5) was prepared by 3 sequential passages in Vero cell culture and was stored at -7 C. Each cell culture stock virus was titrated in Vero cell culture, and the dilution required to yield 30-100 TCID50/0.025 ml was calculated for use in the microtiter VN test. Stock
viruses were diluted, and 0.025 ml was added to each of 4 microtiter wells containing 0.025-ml serum dilutions. For screening, a final serum dilution of 1:1011 was used. Doubling dilutions were used for titrations. Sera were screened for cell culture toxicity by adding 0.025 ml of the 1:10 serum dilution to each of 4 microtiter wells. Serum and virus mixtures were incubated for 60 min at 37 C prior to the addition of 0.1 ml of Vero cell suspension (200,000 cells/ml) to each test well. A back titration of the virus stock was prepared for each test using 4 wells per lo-fold dilution, 0.025 ml per well. Test plates were incubated at 37 C, 5% CO2, 95% humidity for 4-5 days until the back titration indicated the stock virus contained 30-100 TCID50. The plates were then fixed and stained in a solution of 0.15% crystal violet in 2% glutaraldehyde and rinsed. If the serum alone exhibited toxicity, the test was repeated, inoculating the virus/serum mixture onto confluent cell sheets rather than adding a cell suspension to the serum virus mixture. The 50% endpoint titer of the serum was calculated by the Spearman-Karber method4 and expressed as the negative log,,.
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
House, Mikiciuk, Beminger
48
Table 7. Serological, pathological, and clinical observations of an experimental horse, AHSV7.
* Serotype of AHSV indicated in parentheses. † dpv = days postvaccinate. ‡ dpc = days postchallenge.
ELISA antigen preparation. Confluent Vero cell cultures in 750-cm2 roller bottles were rinsed with sterile PBS and then inoculated with either AHSVl, AHSV2, AHSV3, AHSV5, or AHSV6, using a multiplicity of infection of 1:200 in serum-free medium (50 ml/culture). After determining that the ELISA reaction was group specific (see section below), AHSV5 antigen was used routinely. Inoculated cultures were incubated at 37 C until 80-100% CPE was observed, usually between days 2 and 4 postinoculation. The cultures were then frozen at -70 C, thawed, and the fluid clarified by centrifugation at 200 x g for 5 min. Polyethylene glycol 6000 (8% w/v) was added to the fluid and stirred overnight at 4 C. Precipitate was concentrated by centrifugation at 10,400 x g for 30 min and extracted twice with 2 mM Tris, pH 8.8, using 2.5% of the original volume. Supernatant fluid was pooled from each extraction and Triton X-100 (10%) was added to a final concentration of 1% v/v. The preparation was held for 10 min at 4 C and then pelleted through a 3-ml cushion of 40% (w/v) sucrose in 2 mM Tris, pH 8.8, in a swinging bucket rotor (SW28) at 27,000 rpm for 90 min at 4 C. The pellet was resuspended in 2 mM Tris, pH 8.8, totaling 8% of the pooled supernatant volume or 0.4% of the volume of the original harvested fluid. Aliquots were held at -70 C. ELISA normal host cell (NHC) antigen (control antigen) preparation. Confluent Vero cell cultures (750-cm2 bottles) were rinsed, and 50 ml of serum-free media was added. The bottles were frozen at -70 C and processed as described for ELISA antigen preparation. ELISA test. The microtiter ELISA technique7,8 was performed using platesd coated with the appropriate dilutions of ELISA antigen and the control antigen preparations and held at -70 C for up to 6 mo. This technique was a modification of a previously reported method,16 using a different antigen coating procedure.3 The serum and conjugate diluent was 0.05 M Tris buffer (0.15 M NaCl, 0.001 M ethylenedi-
Table 8. Serological, pathological, and clinical observations of an experimental horse, AHSV8.
* Serotype of AHSV indicated in parentheses. = days postvaccinate. ‡ dpc = days postchallenge. § ND = not done.
aminetetraacetic acid with 1% bovine serum albumin and 0.05% Tween 20, and the wash solution was 0.01 M PBS with 0.05% Tween 20 (pH 7.2). The conjugate was horseradish peroxidase-conjugated anti-horse IgG (H&L).c Sera were screened at at dilution of 1:100. Plates were read on a ELISA readerg (test filter wavelength 410 nm, reference filter wavelength 490 nm) or another readerh (test filter 405 nm, reference filter 492 nm) 15 min after the addition of substrate. The average optical density (OD) value for the control antigen was subtracted from the average OD value for the ELISA antigen; corrected values exceeding 0.100 were considered positive. Serial serum titrations from 1:10 were performed using 2-fold dilutions. Determination of the group-specific nature of the ELISA reaction. Sera from 5 ponies vaccinated and challenged with AHSVl, AHSV2, AHSV3, AHSV6, and AHSV9 were titrated in the ELISA test using ELISA antigens prepared with AHSVl, AHSV2, AHSV3, AHSV6, and AHSV5, and the results compared.
Results
Recovery of AHSV from whole blood or washed erythrocytes. Samples stored in OPG at 4 C or as washed erythrocytes mixed with virus stabilizer at -70 C yielded AHSVl when tested at 0, 90, and 180 days poststorage. Tests on samples 270 and 360 days poststorage were negative. Samples stored as washed erythrocytes at 4 C yielded AHSVl when tested at 0, 90, 180, 270, and 360 days poststorage.
Determination of group-specific nature of the ELISA reaction. Titers obtained with homologous antigen varied by no more than a 2-fold dilution from titers obtained with heterologous antigens on all samples tested.
Serological, pathological, and clinical observations
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness Table 9. Serological, pathological, and clinical observations of an experimental horse, AHSV9.
49
Table 10. Comparison of ELISA, CF, VN, and AGID results on controversial sera tested for AHSV antibodies. Number of negative sera
* Serotype of AHSV indicated in parentheses. v = days postvaccinate. days postchallenge. § ND = not done.
of experimental horses. Results are shown in Tables 1-9
Sero1ogy of domestic horse sera. All domestic sera (nonAHSV) were negative by AGID, CF, and VN test procedures. Some sera from horses awaiting exportation showed varying degrees of nonspecific lines in the AGID test and marginal reactions to control antigen in the CF test. A few domestic sera exhibited toxicity on the cell control in the VN test; this was avoided by inoculating confluent monolayer cell cultures rather than cell suspensions. Sera that had been frozen and thawed more than 3 times or heat inactivated were marginally positive (corrected OD ≥ 0.100) when tested by the ELISA procedure; all other control sera were negative by this technique. Seven samples from the exportation panel (15%) reacted with Vero cell control in the IFA test and could not be tested by this technique; all other sera were negative by this technique. Comparison of serological test systems. Sixty-two experimental sera were compared by ELISA (1:100 dilution), CF (1:10 and 1:20 dilutions), VN (1:10 and 1:20 dilutions), and AGID. Ten sera were positive by all test systems. Thirty sera were negative by all test systems. The results for the 22 sera that exhibited mixed test reactions are shown in Table 10. Discussion Laboratory diagnosis of AHSV depends upon both virus isolation and serology. The classical method of drawing blood into OPG solution is useful in field conditions where aseptic technique is difficult and cer-
tainly preserves AHSV for usual shipping times. Washed erythrocytes held at 4 C appear to retain AHSV for the longest storage interval, perhaps because this treatment is the gentlest to the erythrocyte. The domestic serum panels and experimental horse sera provided a unique battery of samples to compare the standard serological tests and evaluate the ELISA procedure described in this paper. The ELISA appears to detect group specific antibodies as early as 14 days after primary exposure to virus (horse #l, inoculated with AHSVl, Table 1) when a screening dilution of 1:100 is used. The ELISA as previously described17 requires calculation against a standard to achieve reproducible results whereas our test does not. We observed that when repeatedly frozen and thawed samples were used, rising values were encountered, perhaps due to aggregates. We recommend against using abused serum in the ELISA screening test, and that any ELISA-positive sample be confirmed by a second method. The merit of using a variety of serological methods was underscored in this study. The IFA test gave an unacceptable degree of nonspecific staining when field cases for exportation were tested. These animals were probably recently vaccinated with a variety of products, possibly hyperimmunizing them against cell culture components. The IFA test is sensitive, detects antibodies as early as 10 days after primary exposure to live virus (horse #l, inoculated with AHSVl), and has been used to show that close to 50% of zebras sampled in Africa are seropositive to AHSV5. However, this test’s usefulness should be limited to screening by experienced personnel, with all positive results confirmed by other techniques. Each test system has particular drawbacks. The VN test in cell culture is sensitive to serum toxicity, inhibiting cell growth, and would require 9 separate tests to ascertain the seronegativity of 1 sample. The CF test utilizes a laboriously prepared antigen; samples can be reactive with many components of the test, requiring scrupulous care in the inclusion of controls. The AGID test uses a crude antigen, and nonspecific
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
House, Mikiciuk, Beminger
50
lines can mask critical areas on the plate during reading. Using the criteria that a positive titer for CF or VN is greater than 10, the ELISA screening test gave false negative results on 15% of the experimental samples. Some2,10 argue that a titer of 20 is needed to be positive on CF or VN; this would decrease the false negatives of the ELISA to 3% and 10%, respectively. The false positive errors of the ELISA were 3% compared to CF ≥ 10, 2% compared to VN ≥ 10, 8% compared to CF to AGID. The AGID test gave 20% false negative results when compared to CF and VN at the 1:10 dilution, and 3% (CF) and 13% (VN) false negative-results at the 1:20 dilution. The selection of the AGID test as the method of choice for screening animals has been partly dictated by the simplicity and speed of the technique. However, the AGID antibody was the last serological reaction to become positive in 11 of the 12 experimental horses who seroconverted; 2 animals (#2, #22) completed 26 and 28 days postchallenge without becoming AGID positive. A more sensitive screening test is needed, with confirmation by a second test (CF or VN). It is notable that only 1 horse (#2) began to exhibit CF and VN antibodies on day 21 postchallenge, justifying and emphasizing the need for the prolonged (60 day) quarantine for equines returning to the USA from regions endemic for AHSV. Acknowledgments We greatly appreciate the technical assistance of M. G. Llewellyn and T. Zaveski, the animal care by Dr. W. White
and the Scientific Services staff, the pathology studies by Dr. T. Wilson, and the guidance by Drs. R. J. Yedloutschnig and J. A. House.
Sources and manufacturers a. The virus strains originated from the Onderstepoort Laboratory, South Africa, and were received at the Plum Island Animal Disease Center in 1960. b. The virus strains originated from the Onderstepoort Laboratory, South Africa, and were received at the Plum Island Animal Disease Center from the Razi Serum and Vaccine Institute, Teheran, Iran, in 1960. c. Dr. J. Pearson, Diagnostic Virology Section, NVSL, Ames, IA. d. Immulon #l, Sheffield Products, Memphis, TN. e. Nunc, Inc., Naperville, IL. f. ICN Immuno Biologicals, Lisle, IL. g. Dynatech Laboratories, Inc., Alexandria, VA. h. Flow Laboratories, Inc., McLean, VA. The use of a particular manufacturer’s product does not constitute an endorsement on behalf of the USDA.
References 1. Alexander RA, Neitz NO, DuToit PJ: 1936, Horsesickness immunizations of horses and mules in the field during the season 1934-1935 with a description of the technique of preparation of polyvalent mouse neurotropic vaccine. Onderstepoort J Vet Sci Anim Ind 7: 17-30. 2. Blackburn NK, Swanepoel R: 1988, Observations on antibody levels associated with active and passive immunity to African horse sickness. Trop Anim Health Prod 20:203-210. 3. Bommeli WR, Kihm U, Lazarowicz M, Steck F: 1980, Rapid detection of antibodies to infectious bovine rhinotracheitis (IBR) virus by micro enzyme-linked immunosorbent assay (micro ELISA). Proc 2nd Int Symp Vet Lab Diagn 2:235-239. 4. Cottral GE: 1978, Manual of standardized methods for veterinary microbiology. Cornell University Press, Ithaca, NY. Bull Off Int Epizoot 63:1607-1625. 5. Davies FG, Lund LJ: 1974, The application of fluorescent antibody techniques to the virus of African horsesickness. Res Vet Sci 17:128-130. 6. Edington A: 1903, Note on the co-relation of several diseases occurring among animals in South Africa. J Hyg 3:137-l54. 7. Eernisse KA, Erickson GA, Snyder ML: 1987, Recommended minimum standards for an enzyme-linked immunosorbent assay (ELISA) in pseudorabies (PR) serodiagnosis. National Veterinary Services Laboratories protocol (August 15,1987), Ames, IA. 8. Engval E, Perlmann P: 1972, Enzyme-linked immunosorbent assay, ELISA. III. Quantification of specific antibodies by enzyme labeled anti-immunoglobulin antigen-coated tubes. J Immunol 109:129-135. 9. Erasmus BJ: 1969, The attenuation of horsesickness virus: problems and advantages associated with the use of different host systems. Proc 2nd Int Conf Equine Infect Dis, Paris, 1969. S. Karger, Basel, Switzerland. Pp. 208-213. 10. Hazrati A, Mirchamsy H, Bahrami S: 1973, Comparative studies on the serological responses of horses to African horsesickness virus. Proc 3rd Int Conf Equine Infect Dis, Paris, 1972. S. Karger, Basel, Switzerland. Pp. 69-80. 11. Hazrati A, Ozawa Y: 1965, Serologic studies of African horse sickness virus with emphasis on neutralization in tissue culture. Can J Comp Med 29:173-178. 12. McIntosh BM: 1956, Complement fixation with horsesickness viruses. Onderstepoort J Vet Res 27: 165-169. 13. Ozawa Y, Hazrati A: 1964, Growth of African horsesickness virus in monkey kidney cell cultures. Am J Vet Res 25:505511. 14. Ozawa Y, Salama SA, Dardiri AH: 1973, Methods for recovering African horsesickness from horse blood. Proc 3rd Int Conf Equine Inf Dis, Paris, 1973. S. Karger, Basel, Switzerland. Pp. 58-62. 15. Pavri KM: 1961, Haemagglutination and haemagglutination inhibition with African horse sickness virus. Nature (Lond): 189-249. 16. Snyder ML, Stewart WC: 1977, Application of an enzyme labeled antibody test in pseudorabies. Proc Annu Meet Am Assoc Vet Lab Diagn 20: 17-32. 17. Williams R: 1987, A single dilution enzyme-linked immunosorbent assay for the quantitative detection of antibodies to African horsesickness virus. Onderstepoort J Vet Res 54:6770.
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness: comparison of serological techniques and evaluation of storage methods of samples for virus isolation Carol House, Peter E. Mikiciuk, Mary Lou Beminger Abstract. Five serological methods of diagnosing African horse sickness were evaluated, using a battery of serum samples from experimental horses vaccinated and challenged with each serotype of African horse sickness virus (AHSVl through AHSV9): agar gel immunodiffusion (AGID), indirect fluorescent antibody (IFA), complement fixation (CF), virus neutralization (VN), and enzyme-linked immunosorbent assay (ELISA). The 5 tests were also compared using a panel of field samples, convalescent equine sera with antibodies to domestic equine viral diseases, and sera from horses awaiting export. The ELISA described in this paper was group specific. It did not require calibration with a standard positive serum but did yield elevated values with negative sera that were repeatedly frozen and thawed or heat inactivated. The IFA test was sensitive but could not be used on some field sera as the control cells exhibited fluorescence, possibly due to the animal being recently vaccinated with cell culture material. Sixty-two experimental sera were compared by VN, CF, AGID, and ELISA. Forty sera, 10 positive and 30 negative, were correctly classified by the 5 serologic assays. The 22 remaining sera gave mixed reactions. The AGID had no false positive results but had false negative results for up to 20% of the samples, depending upon the comparison. The VN, CF, and ELISA were similar in their variability. The length of time that virus could be recovered from a viremic blood sample was compared in an evaluation of storage methods for virus isolation samples. Washed erythrocytes were held at 4 C, washed erythrocytes plus stabilizer were held at -70 C, and blood that was drawn into a preservative (oxalate/phenol/glycerol) was held at 4 C. Virus was isolated for 12 months from a sample stored as washed erythrocytes at 4 C but for only 6 months from the same blood stored by the other 2 methods.
Detection of antibody to African horse sickness virus (AHSV) has traditionally utilized agar gel immunodiffusion (AGID), complement fixation (CF),2,12 and virus neutralization (VN)11 techniques. The indirect fluorescent antibody (IFA)5 assay and enzyme-linked immunosorbent assay (ELISA) calculated using a reference serum17 have been reported, but both have lacked standardization and acceptance. The hemagglutination test15 has rarely been used. In this report we describe an improved ELISA test and delineate its parameters, compare the AGID, CF, VN, IFA, and ELISA diagnostic techniques, and compare the common methods of blood storage to preserve AHSV. Materials and methods Experimental horses. Fifteen clinically healthy adult horses and ponies obtained from local farms were housed in high containment (P3 or greater) animal rooms at the Foreign Animal Disease Diagnostic Laboratory (FADDL). Animals were observed, and temperatures were taken daily. Animals
From the Foreign Animal Disease Diagnostic Laboratory, USDA, APHIS, PO Box 848, Greenport, NY 11944. Received for publication August 7, 1989.
were bled and then each was vaccinated with 1 of 9 monovalent vaccines (1 ml subcutaneously). Because of unexpected results, 2 additional horses were vaccinated, 1 with AHSV5 and 1 with AHSV6. Blood samples were collected on days 7, 14, and 21 postvaccination (dpv); some horses were also sampled on 10 and 17 dpv. The immunity of the horses was challenged on 21 dpv by subcutaneous inoculation with 2 ml of homologous challenge virus. Blood samples were subsequently collected at 3-4 day intervals, up to day 32 postchallenge (dpc). Four additional horses were each inoculated subcutaneously with 1 ml of a challenge virus, AHSV1, -2, -3, or -4. Blood samples were drawn on days 0, 7, 10, and 14 after inoculation. Horses found dead or euthanized when severely ill were necropsied, and tissues were examined histopathologically. Clinically asymptomatic horses were necropsied for comparison. Vaccines. Monovalent vaccines were prepared in 1969 (Y. Ozawa and A. H. Dardiri, unpublished data) by passaging each virusa 10l-103 times in suckling mice by intracranial (IC) inoculation.l,9 The brain tissue suspension vaccine was adjusted to contain l04 mouse lethal doses-50% (MICLD50) per ml and was lyophilized in 2-ml quantities. The vaccine strains were designated AHSVl (A50l), AHSV2 (OD), AHSV3 (L), AHSV4 (Vryheid), AHSV5 (VH), AHSV6 (114), AHSV7 (Karen), AHSV8 (18/60), and AHSV9 (60). Vaccines were determined to be potent in 1988 by intracranial inoculation (IC) titration in suckling mice (SM) (C. House and
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness
45
Table 1. Serological, pathological, and clinical observations of experimental horses, AHSV1.
Table 2. Serological, pathological, and clinical observations of experimental horses, AHSV2.
J. A. House, unpublished data), with log,, titers of 2.8-4.0 SMICLD50. Challenge viruses. Viruses were prepared in 1968 and 1969 (Y. Ozawa and A. H. Dardiri, unpublished data) by passaging each virusb l-6 times in SM by IC inoculations. The brain tissue suspension (10%) was lyophilized in 2-ml quantities and held at 4 C. The strains were designated AHSVl 29/62 (4 mouse passages, 103.5 SMICLD50), AHSV2 MAT (2 mouse passages, 103.5 SMICLD50), AHSV3 13/63 (1 mouse passage, 104 SMICLD50), AHSV4 12/60 (5 mouse passages, 103 SMICLD50 ), AHSV5 30/62 (4 mouse passages, 104.5 SMICLD50), AHSV6 3887 (6 mouse passages, l03 SMICLD50), AHSV7 3893 (4 mouse passages, l03 SMICLD50), AHSV8 18/60 (4 mouse passages, l03 SMICLD50), AHSV9 7/60 (4 mouse passages, 103.5 SMICLD50). Domestic horse sera. A panel of convalescent sera from horses recovering from diseases found in the USA and a panel of normal horse sera randomly assembled from samples submitted to the National Veterinary Services Laboratories (NVSL) were supplied.c A third panel of 45 horse sera was assembled from samples submitted to the FADDL for testing prior to exportation. Recovery of AHSV from whole blood or washed erythrocytes. One horse, inoculated with AHSVl challenge virus, was bled on day 6 postinoculation, and its heparinized blood was stored as washed erythrocytes at 4 C, as washed erythrocytes plus an equivalent volume of virus stabilizer (20% NZ Amine NS,d 0.07% L-glutamic acid, 0.125% K2HPO4; 0.05% KH2PO4; 20% lactose at pH 7.2) at -70 C, and as heparinized blood plus an equal volume of OPG (0.5% ox-
alate, 0.5% phenol, 50% glycerol)6 at 4 C.14 On days 0, 90, 180, 270, and 360 of storage, blood stored by each method was used for virus isolation. The washed erythrocytes at 4 C and washed erythrocytes plus virus stabilizer at -70 C were inoculated onto confluent monolayers of Vero cell cultures, incubated at 37 C, and observed daily for development of CPE. Cells were harvested by scraping cells into the media when CPE appeared or at 1 wk postinoculation. One milliliter of cell/media suspension was inoculated onto a second confluent monolayer of Vero cell culture, incubated at 37 C, and observed daily for CPE for 7 days. The OPG samples were treated in the same manner after removal of the OPG fluid by centrifugation and 3 washes with sterile 0.01 M phosphate buffered saline (PBS) at pH 7.4. Agar gel immunodiffusion (AGID) test. Antigen was prepared by growing AHSV5 in Vero cell cultures, harvesting cells and fluid by freezing and thawing when the cytopathic effect (CPE) was observed in at least 80% or greater of the monolayer. The fluid was clarified by centrifugation at 1,300 x g for 30 min at 4 C, and supernatant fluid was filtered through a prefilter and a 5.0-µm filter. The filtrate was concentrated approximately 100-fold by dialysis using a 14,000 nominal molecular weight limiting membrane. The fluid was further concentrated 2-4-fold by dialysis against solid polyethylene glycol. The resulting antigen was titrated by preparing 2-fold dilutions in PBS and testing against known positive and negative sera in the AGID test. An optimal dilution was determined by observing the intensity and central location of the precipitin line after after 24 hr incubation at room temperature.
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
46
House, Mikiciuk, Beminger
Table 3. Serological, pathological, and clinical observations of experimental horses, AHSV3.
Table 4. Serological, pathological, and clinical observations of experimental horses, AHSV4.
Test plates were prepared by dissolving 0.8% agarose in 0.85% sodium chloride (NaCl) solution and pouring 18 ml/ 100 mm petri dish. Wells 3 mm in diameter and spaced 2 mm apart were made in the solidified agar; approximately 40 µ1 of test serum, positive control serum, or antigen were placed in wells in a standard AGID pattern (antigen centrally and alternating positive control serum and test sera peripherally). The plates were placed in a humidified atmosphere at 25 C for 24 hr and read in a darkened room, using a light source angled beneath the plate. A test serum was considered positive if it formed a line of identity or caused a specific bend with the positive control precipitin line. Complement fixation (CF) test. Sera were diluted 1:1011 and inactivated at 56 C for 30 min. Titrations of positive sera were performed using additional 2-fold dilutions. The CF antigen and control antigen (normal mouse brain) were diluted the same, with the CF containing 4-8 antigen units in veronal buffered saline containing 1% gelatin (VBSG); guinea pig complement was diluted in VBSG to contain 4 complement hemolytic units-50% (CH50). Sera, complement, and antigen were reacted in 96-well round-bottom microtiter plates at 4 C for 18 hr. Hemolysin was diluted to contain 2 hemolytic units and used to sensitize washed sheep erythrocytes (SRBC’s). The SRBC’s were standardized to 3% concentration and added to all wells on the microtiter plate. The test was incubated for 30 min at 37 C. The plates were then centrifuged (200 x g), and the wells were scored for the presence of hemolysis. The following controls were used: 1) serum and complement; 2) serum and SRBC; 3) CF antigen
and control antigen each with 4 CH50, 2 CH50, and 1 CH50, of complement; 4) CF antigen and SRBC’s; 5) control antigen and SRBC’s; 6) complement dilutions of 4 CH50, 2 CH50, and 1 CH50; and 7) SRBC’s. The inverse of the highest dilution of serum specifically fixing complement with the CF antigen was called the titer. Antigens for complement fixation (CF) test. The vaccine virus AHSV9 (S2) was propagated in suckling mice by IC inoculation. Brain tissue was collected from moribund mice for antigen preparation. Fifty grams of tissue were homogenized with 200 ml of 8.5% sucrose at 4 C, and the suspension was added dropwise to 4 liters of acetone at 4 C. The precipitate was collected by centrifugation at 900 x g for 5 min. The supernatant fluid was discarded, and the pellet was reextracted in 4 liters of acetone at 4 C for 1 hr. Sediment was collected by centrifugation at 900 x g for 5 min. Supernatant fluid was discarded and the pellet allowed to dry at 25 C. The resulting powder was rehydrated with 100 ml of 0.85% NaCl and held for 1 hr in an ice bath. The preparation was clarified by centrifugation at 12,100 x g for 30 min and the pellet discarded. Supernatant fluid, called CF antigen, was aliquoted, lyophilized, and stored at 4 C. A similar preparation was made from normal mouse brain and called control antigen. Indirect fluorescent antibody (IFA) slide test. Vero cells were grown on chamber slidese incubated at 37 C and 95% humidity until the cells were 80% confluent. Each chamber was inoculated with virus sufficient to produce scattered foci (3-8 foci/field at 160 x magnification) after 18-22 hr of incubation. Negative-control chamber slides were prepared from uninoculated cell cultures. Slides were decanted and fixed in
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness
47
Table 5. Serological, pathological, and clinical observations of experimental horses, AHSV5.
Table 6. Serological, pathological, and clinical observations of experimental horses, AHSV6.
acetone at 4 C for 5 min. Fixed, dried slides were stored for up to 18 mo at -70 C. Experimental sera were diluted 1:10 for screening. Slides of both positive or infected and negative or control chambers were warmed to 25 C, and diluted serum was applied to each chamber. The slides were then incubated at 25 C in a humid chamber for 30 min and subsequently washed in PBS for 5 min. Rabbit anti-horse IgG FITC conjugatef was diluted 1:32 in PBS and applied to the slide for 30 min at 25 C. Slides were rinsed in PBS and distilled water and dried. Specificity of immunofluorescence was evaluated by first examining the negative control slide. If no fluorescence was seen in the control chamber, the test chamber was examined for the presence of fluorescent inclusion bodies, seen as bright perinuclear bodies. Experimental samples were tested on slides infected with the same serotype as the vaccination and/or challenge (IFA homologous) and on slides infected with different serotypes from the vaccination and/or challenge (IFA heterologous). Virus neutralization (VN) test. Stocks of viruses were prepared by passaging the following mouse brain suspension stock virusg once in Vero cell culture followed by freezing and thawing at the time of maximal CPE. After clarification by centrifugation at 200 x g for 10 min, 30% virus stabilizer was added and aliquots were stored at -70 C. Mouse brain passage 102, monkey kidney cell line13 (MS) passage 5 stocks were used for AHSV1 (A50l), AHSV2 (OD), AHSV3 (L), AHSV4 (Vryheid), AHSV5 (VH), AHSV6 (114), AHSV7 (Karen MS), and AHSV9 (S2). The stock of AHSV8 (18/60) (mouse brain passage 102, MS passage 5) was prepared by 3 sequential passages in Vero cell culture and was stored at -7 C. Each cell culture stock virus was titrated in Vero cell culture, and the dilution required to yield 30-100 TCID50/0.025 ml was calculated for use in the microtiter VN test. Stock
viruses were diluted, and 0.025 ml was added to each of 4 microtiter wells containing 0.025-ml serum dilutions. For screening, a final serum dilution of 1:1011 was used. Doubling dilutions were used for titrations. Sera were screened for cell culture toxicity by adding 0.025 ml of the 1:10 serum dilution to each of 4 microtiter wells. Serum and virus mixtures were incubated for 60 min at 37 C prior to the addition of 0.1 ml of Vero cell suspension (200,000 cells/ml) to each test well. A back titration of the virus stock was prepared for each test using 4 wells per lo-fold dilution, 0.025 ml per well. Test plates were incubated at 37 C, 5% CO2, 95% humidity for 4-5 days until the back titration indicated the stock virus contained 30-100 TCID50. The plates were then fixed and stained in a solution of 0.15% crystal violet in 2% glutaraldehyde and rinsed. If the serum alone exhibited toxicity, the test was repeated, inoculating the virus/serum mixture onto confluent cell sheets rather than adding a cell suspension to the serum virus mixture. The 50% endpoint titer of the serum was calculated by the Spearman-Karber method4 and expressed as the negative log,,.
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
House, Mikiciuk, Beminger
48
Table 7. Serological, pathological, and clinical observations of an experimental horse, AHSV7.
* Serotype of AHSV indicated in parentheses. † dpv = days postvaccinate. ‡ dpc = days postchallenge.
ELISA antigen preparation. Confluent Vero cell cultures in 750-cm2 roller bottles were rinsed with sterile PBS and then inoculated with either AHSVl, AHSV2, AHSV3, AHSV5, or AHSV6, using a multiplicity of infection of 1:200 in serum-free medium (50 ml/culture). After determining that the ELISA reaction was group specific (see section below), AHSV5 antigen was used routinely. Inoculated cultures were incubated at 37 C until 80-100% CPE was observed, usually between days 2 and 4 postinoculation. The cultures were then frozen at -70 C, thawed, and the fluid clarified by centrifugation at 200 x g for 5 min. Polyethylene glycol 6000 (8% w/v) was added to the fluid and stirred overnight at 4 C. Precipitate was concentrated by centrifugation at 10,400 x g for 30 min and extracted twice with 2 mM Tris, pH 8.8, using 2.5% of the original volume. Supernatant fluid was pooled from each extraction and Triton X-100 (10%) was added to a final concentration of 1% v/v. The preparation was held for 10 min at 4 C and then pelleted through a 3-ml cushion of 40% (w/v) sucrose in 2 mM Tris, pH 8.8, in a swinging bucket rotor (SW28) at 27,000 rpm for 90 min at 4 C. The pellet was resuspended in 2 mM Tris, pH 8.8, totaling 8% of the pooled supernatant volume or 0.4% of the volume of the original harvested fluid. Aliquots were held at -70 C. ELISA normal host cell (NHC) antigen (control antigen) preparation. Confluent Vero cell cultures (750-cm2 bottles) were rinsed, and 50 ml of serum-free media was added. The bottles were frozen at -70 C and processed as described for ELISA antigen preparation. ELISA test. The microtiter ELISA technique7,8 was performed using platesd coated with the appropriate dilutions of ELISA antigen and the control antigen preparations and held at -70 C for up to 6 mo. This technique was a modification of a previously reported method,16 using a different antigen coating procedure.3 The serum and conjugate diluent was 0.05 M Tris buffer (0.15 M NaCl, 0.001 M ethylenedi-
Table 8. Serological, pathological, and clinical observations of an experimental horse, AHSV8.
* Serotype of AHSV indicated in parentheses. = days postvaccinate. ‡ dpc = days postchallenge. § ND = not done.
aminetetraacetic acid with 1% bovine serum albumin and 0.05% Tween 20, and the wash solution was 0.01 M PBS with 0.05% Tween 20 (pH 7.2). The conjugate was horseradish peroxidase-conjugated anti-horse IgG (H&L).c Sera were screened at at dilution of 1:100. Plates were read on a ELISA readerg (test filter wavelength 410 nm, reference filter wavelength 490 nm) or another readerh (test filter 405 nm, reference filter 492 nm) 15 min after the addition of substrate. The average optical density (OD) value for the control antigen was subtracted from the average OD value for the ELISA antigen; corrected values exceeding 0.100 were considered positive. Serial serum titrations from 1:10 were performed using 2-fold dilutions. Determination of the group-specific nature of the ELISA reaction. Sera from 5 ponies vaccinated and challenged with AHSVl, AHSV2, AHSV3, AHSV6, and AHSV9 were titrated in the ELISA test using ELISA antigens prepared with AHSVl, AHSV2, AHSV3, AHSV6, and AHSV5, and the results compared.
Results
Recovery of AHSV from whole blood or washed erythrocytes. Samples stored in OPG at 4 C or as washed erythrocytes mixed with virus stabilizer at -70 C yielded AHSVl when tested at 0, 90, and 180 days poststorage. Tests on samples 270 and 360 days poststorage were negative. Samples stored as washed erythrocytes at 4 C yielded AHSVl when tested at 0, 90, 180, 270, and 360 days poststorage.
Determination of group-specific nature of the ELISA reaction. Titers obtained with homologous antigen varied by no more than a 2-fold dilution from titers obtained with heterologous antigens on all samples tested.
Serological, pathological, and clinical observations
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
Laboratory diagnosis of African horse sickness Table 9. Serological, pathological, and clinical observations of an experimental horse, AHSV9.
49
Table 10. Comparison of ELISA, CF, VN, and AGID results on controversial sera tested for AHSV antibodies. Number of negative sera
* Serotype of AHSV indicated in parentheses. v = days postvaccinate. days postchallenge. § ND = not done.
of experimental horses. Results are shown in Tables 1-9
Sero1ogy of domestic horse sera. All domestic sera (nonAHSV) were negative by AGID, CF, and VN test procedures. Some sera from horses awaiting exportation showed varying degrees of nonspecific lines in the AGID test and marginal reactions to control antigen in the CF test. A few domestic sera exhibited toxicity on the cell control in the VN test; this was avoided by inoculating confluent monolayer cell cultures rather than cell suspensions. Sera that had been frozen and thawed more than 3 times or heat inactivated were marginally positive (corrected OD ≥ 0.100) when tested by the ELISA procedure; all other control sera were negative by this technique. Seven samples from the exportation panel (15%) reacted with Vero cell control in the IFA test and could not be tested by this technique; all other sera were negative by this technique. Comparison of serological test systems. Sixty-two experimental sera were compared by ELISA (1:100 dilution), CF (1:10 and 1:20 dilutions), VN (1:10 and 1:20 dilutions), and AGID. Ten sera were positive by all test systems. Thirty sera were negative by all test systems. The results for the 22 sera that exhibited mixed test reactions are shown in Table 10. Discussion Laboratory diagnosis of AHSV depends upon both virus isolation and serology. The classical method of drawing blood into OPG solution is useful in field conditions where aseptic technique is difficult and cer-
tainly preserves AHSV for usual shipping times. Washed erythrocytes held at 4 C appear to retain AHSV for the longest storage interval, perhaps because this treatment is the gentlest to the erythrocyte. The domestic serum panels and experimental horse sera provided a unique battery of samples to compare the standard serological tests and evaluate the ELISA procedure described in this paper. The ELISA appears to detect group specific antibodies as early as 14 days after primary exposure to virus (horse #l, inoculated with AHSVl, Table 1) when a screening dilution of 1:100 is used. The ELISA as previously described17 requires calculation against a standard to achieve reproducible results whereas our test does not. We observed that when repeatedly frozen and thawed samples were used, rising values were encountered, perhaps due to aggregates. We recommend against using abused serum in the ELISA screening test, and that any ELISA-positive sample be confirmed by a second method. The merit of using a variety of serological methods was underscored in this study. The IFA test gave an unacceptable degree of nonspecific staining when field cases for exportation were tested. These animals were probably recently vaccinated with a variety of products, possibly hyperimmunizing them against cell culture components. The IFA test is sensitive, detects antibodies as early as 10 days after primary exposure to live virus (horse #l, inoculated with AHSVl), and has been used to show that close to 50% of zebras sampled in Africa are seropositive to AHSV5. However, this test’s usefulness should be limited to screening by experienced personnel, with all positive results confirmed by other techniques. Each test system has particular drawbacks. The VN test in cell culture is sensitive to serum toxicity, inhibiting cell growth, and would require 9 separate tests to ascertain the seronegativity of 1 sample. The CF test utilizes a laboriously prepared antigen; samples can be reactive with many components of the test, requiring scrupulous care in the inclusion of controls. The AGID test uses a crude antigen, and nonspecific
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
House, Mikiciuk, Beminger
50
lines can mask critical areas on the plate during reading. Using the criteria that a positive titer for CF or VN is greater than 10, the ELISA screening test gave false negative results on 15% of the experimental samples. Some2,10 argue that a titer of 20 is needed to be positive on CF or VN; this would decrease the false negatives of the ELISA to 3% and 10%, respectively. The false positive errors of the ELISA were 3% compared to CF ≥ 10, 2% compared to VN ≥ 10, 8% compared to CF to AGID. The AGID test gave 20% false negative results when compared to CF and VN at the 1:10 dilution, and 3% (CF) and 13% (VN) false negative-results at the 1:20 dilution. The selection of the AGID test as the method of choice for screening animals has been partly dictated by the simplicity and speed of the technique. However, the AGID antibody was the last serological reaction to become positive in 11 of the 12 experimental horses who seroconverted; 2 animals (#2, #22) completed 26 and 28 days postchallenge without becoming AGID positive. A more sensitive screening test is needed, with confirmation by a second test (CF or VN). It is notable that only 1 horse (#2) began to exhibit CF and VN antibodies on day 21 postchallenge, justifying and emphasizing the need for the prolonged (60 day) quarantine for equines returning to the USA from regions endemic for AHSV. Acknowledgments We greatly appreciate the technical assistance of M. G. Llewellyn and T. Zaveski, the animal care by Dr. W. White
and the Scientific Services staff, the pathology studies by Dr. T. Wilson, and the guidance by Drs. R. J. Yedloutschnig and J. A. House.
Sources and manufacturers a. The virus strains originated from the Onderstepoort Laboratory, South Africa, and were received at the Plum Island Animal Disease Center in 1960. b. The virus strains originated from the Onderstepoort Laboratory, South Africa, and were received at the Plum Island Animal Disease Center from the Razi Serum and Vaccine Institute, Teheran, Iran, in 1960. c. Dr. J. Pearson, Diagnostic Virology Section, NVSL, Ames, IA. d. Immulon #l, Sheffield Products, Memphis, TN. e. Nunc, Inc., Naperville, IL. f. ICN Immuno Biologicals, Lisle, IL. g. Dynatech Laboratories, Inc., Alexandria, VA. h. Flow Laboratories, Inc., McLean, VA. The use of a particular manufacturer’s product does not constitute an endorsement on behalf of the USDA.
References 1. Alexander RA, Neitz NO, DuToit PJ: 1936, Horsesickness immunizations of horses and mules in the field during the season 1934-1935 with a description of the technique of preparation of polyvalent mouse neurotropic vaccine. Onderstepoort J Vet Sci Anim Ind 7: 17-30. 2. Blackburn NK, Swanepoel R: 1988, Observations on antibody levels associated with active and passive immunity to African horse sickness. Trop Anim Health Prod 20:203-210. 3. Bommeli WR, Kihm U, Lazarowicz M, Steck F: 1980, Rapid detection of antibodies to infectious bovine rhinotracheitis (IBR) virus by micro enzyme-linked immunosorbent assay (micro ELISA). Proc 2nd Int Symp Vet Lab Diagn 2:235-239. 4. Cottral GE: 1978, Manual of standardized methods for veterinary microbiology. Cornell University Press, Ithaca, NY. Bull Off Int Epizoot 63:1607-1625. 5. Davies FG, Lund LJ: 1974, The application of fluorescent antibody techniques to the virus of African horsesickness. Res Vet Sci 17:128-130. 6. Edington A: 1903, Note on the co-relation of several diseases occurring among animals in South Africa. J Hyg 3:137-l54. 7. Eernisse KA, Erickson GA, Snyder ML: 1987, Recommended minimum standards for an enzyme-linked immunosorbent assay (ELISA) in pseudorabies (PR) serodiagnosis. National Veterinary Services Laboratories protocol (August 15,1987), Ames, IA. 8. Engval E, Perlmann P: 1972, Enzyme-linked immunosorbent assay, ELISA. III. Quantification of specific antibodies by enzyme labeled anti-immunoglobulin antigen-coated tubes. J Immunol 109:129-135. 9. Erasmus BJ: 1969, The attenuation of horsesickness virus: problems and advantages associated with the use of different host systems. Proc 2nd Int Conf Equine Infect Dis, Paris, 1969. S. Karger, Basel, Switzerland. Pp. 208-213. 10. Hazrati A, Mirchamsy H, Bahrami S: 1973, Comparative studies on the serological responses of horses to African horsesickness virus. Proc 3rd Int Conf Equine Infect Dis, Paris, 1972. S. Karger, Basel, Switzerland. Pp. 69-80. 11. Hazrati A, Ozawa Y: 1965, Serologic studies of African horse sickness virus with emphasis on neutralization in tissue culture. Can J Comp Med 29:173-178. 12. McIntosh BM: 1956, Complement fixation with horsesickness viruses. Onderstepoort J Vet Res 27: 165-169. 13. Ozawa Y, Hazrati A: 1964, Growth of African horsesickness virus in monkey kidney cell cultures. Am J Vet Res 25:505511. 14. Ozawa Y, Salama SA, Dardiri AH: 1973, Methods for recovering African horsesickness from horse blood. Proc 3rd Int Conf Equine Inf Dis, Paris, 1973. S. Karger, Basel, Switzerland. Pp. 58-62. 15. Pavri KM: 1961, Haemagglutination and haemagglutination inhibition with African horse sickness virus. Nature (Lond): 189-249. 16. Snyder ML, Stewart WC: 1977, Application of an enzyme labeled antibody test in pseudorabies. Proc Annu Meet Am Assoc Vet Lab Diagn 20: 17-32. 17. Williams R: 1987, A single dilution enzyme-linked immunosorbent assay for the quantitative detection of antibodies to African horsesickness virus. Onderstepoort J Vet Res 54:6770.
Downloaded from vdi.sagepub.com at FLORIDA INTL UNIV on May 28, 2015
