A simpler, more sensitive competitive inhibition ...

Journal of Veterinary Diagnostic Investigation http://vdi.sagepub.com/

A Simpler, More Sensitive Competitive Inhibition Enzyme-Linked Immunosorbent Assay for Detection of Antibody to Malignant Catarrhal Fever Viruses Hong Li, Travis C. McGuire, Uwe U. Müller-Doblies and Timothy B. Crawford J VET Diagn Invest 2001 13: 361 DOI: 10.1177/104063870101300417 The online version of this article can be found at: http://vdi.sagepub.com/content/13/4/361

Published by: http://www.sagepublications.com

On behalf of:

Official Publication of the American Association of Veterinary Laboratory Diagnosticians, Inc.

Additional services and information for Journal of Veterinary Diagnostic Investigation can be found at: Email Alerts: http://vdi.sagepub.com/cgi/alerts Subscriptions: http://vdi.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://vdi.sagepub.com/content/13/4/361.refs.html

>> Version of Record - Jul 1, 2001 What is This?

Downloaded from vdi.sagepub.com by guest on October 11, 2013

J Vet Diagn Invest 13:361–364 (2001)

A simpler, more sensitive competitive inhibition enzyme-linked immunosorbent assay for detection of antibody to malignant catarrhal fever viruses Hong Li, Travis C. McGuire, Uwe U. Mu¨ller-Doblies, Timothy B. Crawford Abstract. An earlier competitive inhibition enzyme-linked immunosorbent assay (CI-ELISA) was developed for detection of specific antibody against malignant catarrhal fever (MCF) viruses (MCFV) in ruminants. In this study, the indirect CI-ELISA was improved by conjugating the monoclonal antibody 15-A directly with horseradish peroxidase and by developing a method of producing precoated, dried antigen plates. This new test is referred to as a direct CI-ELISA. The reformatted test yielded a significantly improved sensitivity, and the time required was reduced to about one-sixth of the previous time. Of 37 MCF cases in cattle that were confirmed by histopathology and polymerase chain reaction (PCR) assay, 37 (100%) were positive by the new test, whereas the indirect CI-ELISA detected only 23 (62%). The direct CI-ELISA detected antibody to MCFV in 100% of 48 sheep that had been defined as infected with ovine herpesvirus 2 (OvHV-2) by PCR, whereas the indirect CI-ELISA detected only 41 (85%). Comparison of antibody titers measured by the 2 assays for sera collected from OvHV-2-infected sheep and from cattle, bison, and deer with clinical sheep-associated MCF revealed that the direct CI-ELISA offered a 4-fold increase in analytical sensitivity over the indirect format. The number of seropositive animals detected by the direct CI-ELISA among apparently normal cattle and bison was 2–3 times greater than the number detected by the indirect CI-ELISA, indicating that a significant percentage of normal cattle and bison are subclincally infected with MCFV. Many new variants of enzyme immunoassays have been developed with the aim of increasing sensitivity and specificity, reducing the test duration, and facilitating assay performance.14 The monoclonal antibody (MAb)-based competitive inhibition enzyme-linked immunosorbent assay (CIELISA) is an accepted, reliable, and sensitive method for measuring antibody responses against a variety of infectious pathogens in both humans and animals.1,5,7 A CI-ELISA was developed using MAb 15-A against a specific epitope broadly conserved among all known strains of malignant catarrhal fever viruses (MCFV), including ovine herpesvirus 2 (OvHV-2), alcelaphine herpesvirus 1 (AlHV-1), and the newly recognized strain from white-tailed deer.8,10 The original assay was an indirect CI-ELISA, because the detection of the MAb was indirect via enzyme-labeled anti-mouse immunoglobulins. Although the indirect CI-ELISA has been useful for epidemiologic studies of MCFV infection,11 significant improvements in sensitivity, efficiency, and usability were needed. In this study, the indirect CI-ELISA was modified by 1) conjugating the MAb 15-A directly with an enzyme label and 2) developing a method whereby the antigen-containing plates are precoated and stored at 4 C for appreciable periods without significant degradation. This modified CI-ELISA is referred to as a direct CI-ELISA. Comparison of the indirect and direct CI-ELISA systems on the same set of defined sera from cattle, bison, deer, sheep, and certain other ruminant species demonstrated that the direct assay was significantly more sensitive for detection of MCFV antibody in animals with clinical sheep-associated malignant catarrhal fever (SAMCF) and in animals with subclinical infection. From the Animal Diseases Research Unit, USDA, ARS (Li), and the Department of Veterinary Microbiology and Pathology (McGuire, Crawford), Washington State University, Pullman, WA 99164, and the Institutes of Virology, University of Zu¨rich, Zu¨rich, Switzerland (Mu¨ller-Doblies). Received for publication May 30, 2000.

MCFV antigen–coated, dried plates were produced based on the following procedures. Wells of flat-bottom platesa were coated at 4 C for 18–20 hours with 50 ml of solution containing 0.2 mg of semipurified MCFV antigens in 50 mM carbonate–bicarbonate buffer (pH 9.6).10 The coated plates were blocked at room temperature (RT, 20–25 C) for 2 hours with 0.05 M phosphate-buffered saline (PBS) containing 2% sucrose, 0.1 M glycine, 0.5% bovine serum albumin, and 0.44% NaCl (pH 7.2). After blocking, wells were emptied by discarding the bulk of the fluid and striking the plate 5 times on a clean paper towel. The plates were then dried in a low humidity environment at 37 C for 18 hours, sealed in plastic bags with desiccant, and stored at 4 C. MAb 15-A in ascitic fluid was precipitated with ammonium sulfate and conjugated with horseradish peroxidase by a commercial company.b The indirect CI-ELISA procedure was as described previously.10 For the direct CI-ELISA, 50 ml of 20% serum in PBS (1:5 dilution) was added in duplicate wells (or quadruplicate wells for the negative control). The plate(s) was covered with parafilm and incubated at RT for 60 minutes. After 3 rapid washes with PBS containing 0.1% Tween-20, 50 ml of a predetermined concentration of conjugated MAb 15-A was added to each well, except for the blank control well. The plates were covered with parafilm and incubated at RT for 60 minutes. After an additional 3 washes, 100 ml of substratec was added to each well and incubated at RT for 60 minutes. The reactions were stopped with 100 ml of 0.18 M sulfuric acid in each well. Optical density at 450 nm (OD450) was determined with an ELISA plate reader.d The results were expressed as the percentage inhibition, derived by the following formula: 100 2 [(mean OD450 of test serum 3 100)/(mean OD450 of negative control)]. To be considered valid, each test had to fall within the following limits: the mean OD of the negative control wells 0.4–2.0, and the inhibition by a defined positive control serum $25.0%. Establishment of the cutoff value for the indirect CI-ELISA was as previously described.10

361

362

Brief Communications

A pool of sera from the 19 OvHV-2–negative sheep ?? used as a negative control. These 19 OvHV-2–negative sheep had been maintained separately from infected sheep for an 18-month period, and their OvHV-2–negative status was reconfirmed monthly by polymerase chain reaction (PCR) assay. The mean OD values from 3 different serum panels from these 19 sheep, collected at different time points over a 6-month period, revealed that there was no significant difference between the panels (60.015SD, n 5 3). The cutoff for the direct CI-ELISA was established by adding 3 SD to the mean OD of the 19 negative sheep sera, which equaled 25% inhibition. Thus, any serum samples having $25% inhibition in the assay were interpreted as seropositive. Among over 300 sequential serum samples taken from these 19 OvHV-2–negative sheep during 18 months in isolation, only 2 samples scored positive by direct CI-ELISA. The inhibitory activity levels declined rapidly in these 2 sheep, returning to a negative status within 2 weeks. The 2 sheep stayed consistently PCR negative for OvHV-2, suggesting that the inhibition yielded by these 2 samples was not specific for MCFV antibody. To determine if the inhibitory activity was associated with immunoglobulin, ammonium sulfate precipitation and protein G–binding experiments were done. The results from both methods indicated that the inhibitory reactivity by these sera was antibody mediated. However, the reason for this transient cross-reactivity in these 2 sheep was not determined. Because well-defined MCFV-negative animals were not available for controls in species other than domestic sheep, the feasibility of using the set of OvHV-2–negative sheep sera as controls for other species was evaluated. A total of 60 MCFV-antibody–negative sera examined by indirect immunofluorescence assay (IFA) from cattle, bison, and deer (20 in each group) were tested by direct CI-ELISA. All sera except 1 from each group yielded inhibition levels less than the cutoff value of 25%. Compared with the OvHV-2–negative sheep sera, the mean inhibition levels of the groups (excluding the single positive serum) were very similar, ranging from 21% to 6.8%. This indicated that a panel of OvHV-2–negative sheep sera legitimately could be used as negative serum controls for cattle, bison, or deer samples when testing by direct CI-ELISA. The ability of the direct and indirect CI-ELISAs to detect antibody was compared using the sera from cattle with clinical SA-MCF and from OvHV-2–infected sheep. Thirty-seven cattle with confirmed clinical SA-MCF were examined. These animals had been used in a previous study on evaluation of indirect CI-ELISA, PCR, and histopathology for diagnosis of acute SA-MCF in cattle.12 All 37 (100%) were positive by the direct format, whereas only 23 (62%) were positive by the indirect format12 (Fig. 1). In addition, of 20 cattle from Switzerland that had clinical SA-MCF confirmed by PCR and histopathology, 19 (95%) were positive by the direct method. These 20 samples were not examined by the indirect method because of inadequate serum volumes. All 48 (100%) OvHV-2–infected adult domestic sheep from Switzerland (n 5 20) and the USA (n 5 28) examined in this study were positive by the direct assay, whereas only 41 (85%) were positive by the indirect assay (Fig. 2A–2C). Among 17 OvHV-2–infected mouflon sheep from the USA,

Figure 1. Comparison of direct and indirect CI-ELISA for detection of MCFV antibody in 37 cattle with clinical SA-MCF. Results from histopathology (pathology) and PCR assay, which were used to confirm the clinical diagnosis, are depicted in the innermost circle. Direct CI-ELISA results are depicted in the adjacent circle, and indirect CI-ELISA results are shown in the outermost circle. Positive test results (solid), uncertain or borderline test results (fine dots), and negative results (open) are shown. Each case is represented by a numbered segment. The larger numbers, within the figure, are the number of animals in that group.

respective figures were 16 (94%) and 8 (47%) (Fig. 2D). Moreover, the percentage of inhibition by individual samples was generally greater in the direct than the indirect test format (Fig. 2). Antibody titers in a cow, bison, and deer with clinical SAMCF and in an OvHV-2–infected sheep were also measured using serial dilutions in both direct and indirect assays. The endpoints by direct CI-ELISA versus indirect CI-ELISA were 1:160 versus 1:40 in bison, 1:320 versus 1:160 in deer, 1:640 versus 1:160 in cattle, and 1:160 versus 1:20 in sheep, respectively. Average titers were thus about 4-fold greater by the direct than by the indirect method. These 4 sera were also titrated in IFA by limiting dilutions. The IFA titers were 1:20 (bison), 1:80 (deer), 1:80 (cow), and 1:20 (sheep), indicating that IFA was at least 4-fold less sensitive than the direct CI-ELISA. Six hundred seventy-two sera from clinically normal cattle (n 5 402) and normal bison (n 5 270) from Canada, Switzerland, and the USA were tested by both direct and indirect CI-ELISAs. As shown in Table 1, 5.2% of all animals tested were positive by indirect CI-ELISA, and 13.8% were positive by direct CI-ELISA. An exception was a group of 31 cattle from Switzerland with no history of contact with sheep; none of these animals were positive by either direct or indirect methods. The overall prevalence was 11% among all normal cattle sera tested and 18% among normal bison. In comparison with the original indirect format assay, the


363

Figure 2. Comparison of direct and indirect CI-ELISA for detection of MCFV antibody in OvHV-2–infected sheep. Infection status was confirmed by PCR assay. Solid bars represent the percentage of inhibition in indirect CI-ELISA (cutoff value 5 35% inhibition). Solid plus hatched bars represent the percentage of inhibition in direct CI-ELISA (cutoff value 5 25% inhibition). A. Sera from 20 domestic sheep obtained from the University of Zurich, Switzerland. B. Sera from 20 domestic sheep collected from the US Sheep Experiment Station, Dubois, Idaho. C. Sera from 8 PCR-positive domestic sheep from the WSU experimental herd. D. Sera from a group of 17 mouflon sheep.

Table 1. Comparison of the sensitivity of indirect and direct CIELISA for detection of MCFV antibody in clinically normal cattle and bison. No. positive (%) Group

Sheep contact

Source

Cattle Cattle Cattle Cattle Bison Bison Bison Total

no yes unknown unknown yes unknown unknown

Switzerland Switzerland USA USA Canada Canada USA

No. tested

31 46 100* 225 70 96 104 672

Indirect CI-ELISA

0 1 5 5 5 16 3 35

(0) (2) (5) (2) (7) (17) (3) (5.2)

Direct CI-ELISA

0 5 17 23 9 23 16 93

(0) (11) (17) (10) (13) (24) (15) (13.8)

* One-hundred cattle sera were selected randomly from the archived serum banks of the Washington Animal Disease Diagnostic Laboratory, having been submitted as normal animals or with clinical diagnoses other than MCF.

direct format provided significant improvements in sensitivity, efficiency, and convenience and a lower unit cost. The improved sensitivity increased the reliability of the test for diagnosis of the acute stage MCF. The sensitivity of the assay for MCFV antibody in cattle with clinical SA-MCF now approximates 95%. The figure for bison is currently somewhat lower, about 80% (unpublished data). The number of sera available for study from acutely ill bison, deer, and other nonbovine ruminants was too limited, however, for a precise sensitivity for the test to be estimated. The specificity of polyclonal antibody–based tests for MCF has always been problematic because of well-documented cross-reactive antigens shared among ruminant herpesviruses.6 The recently described bovine lymphotropic herpesvirus, which is present in most cattle, may well contribute to the nonspecificity inherent in polyclonal assays because it probably is a close relative of the MCFV group, as reflected by the sequence of its polymerase gene.15 The current inability to identify certifiably uninfected animals in species other than sheep hinders the critical definition of

364


specificity in cattle, bison, deer, and other susceptible nonovine species. This specificity was examined during the development of the original assay by testing the binding activity of MAb 15-A on all available ruminant herpesviruses.10 In the current study, sera from cattle, bison, and deer that were clearly IFA negative to MCFV antigens and sera from isolation-raised sheep confirmed to be free of MCFV by PCR were used to estimate the specificity as 94% in these species. This is a minimum figure, representing the worstcase scenario; the sensitivity of the competitive ELISA frequently exceeds that of IFA for herpesviral antibody.13 The improved sensitivity prompted a reexamination of the prevalence of latent or subclinical infection in certain ruminant species that had been previously examined with the original indirect CI-ELISA.9 The percentage of antibodypositive sera among normal cattle and normal bison examined increased about 3-fold using the new format. There may be a relationship between the often inexplicable epidemiologic patterns observed with MCF3,16 and subclinical infection. Numerous reports exist of MCF in herds or on premises where no sheep or other carrier species have been present within several miles for months or even for many years. These reports have prompted speculation about the presence of other etiologic agents or cofactors4 or the possibility of insect vectors2. It is becoming increasingly apparent that very significant levels of subclinical or latent infection are present in normal cattle and bison. Because of inevitable sensitivity limits, the seroprevalence rates found in this study are underestimates of actual infection rates. The unavailability of a ‘‘gold standard’’ to identify truly uninfected clinically susceptible animals is a major impediment at present to a precise definition of the subclinical infection rates in these species. Nevertheless, the prevalence levels that are already recognized would appear to support the hypothesis that recrudescence of latent or subclinical infections represent a significant proportion of cases of MCF seen by practitioners, particularly those that occur in the absence of recent sheep contact. The reformatting of the CI-ELISA significantly reduced the time required, from .24 hours to ,4 hours. With directly conjugated MAb and dried, precoated plates, the direct CIELISA offers greater sensitivity, consistency, economy, and convenience. The 15-A epitope is highly conserved among MCFV, including AlHV-1, OvHV-2,10 the recently reported unique MCFV in white-tailed deer,8 AlHV-2, and other emerging members of lower pathogenicity in domestic goats and certain exotic and domestic wild ruminants (unpublished data). However, preliminary data suggest that this epitope is not present on the bovine lymphotropic herpesvirus (unpublished data), neither is it present on about 13 other common ruminant viruses examined.10 This more sensitive and efficient CI-ELISA thus is a serious candidate for a standard serologic test for MCFV antibody in a variety of species. Acknowledgements. This work was supported by USDAARS grant CWU 5348-32000-013-00D. We thank Dongyue

Zhuang, Lori Fuller, and Janice Keller for excellent technical assistance.

Sources and manufacturers a. Immulon 4, Dynatech Labs, Chantilly, VA. b. VMRD, Pullman, WA. c. TMB Microwell Peroxidase Substrate, Kirkegaard & Perry Laboratories, Gaithersburg, MD. d. Multiskan MCC/340, Fisher Scientific, Pittsburgh, PA.

References 1. Afshar A, Thomas FC, Wright F, et al.: 1987, Comparison of competitive and indirect enzyme-linked immunosorbent assays for detection of bluetongue virus antibodies in serum and whole blood. J Clin Microbiol 25:1705–1710. 2. Barnard BJH, van de Pypekamp HE, Griessel MD, et al.: 1989, Epizootology of wildebeest-derived malignant catarrhal fever in an outbreak in the north-western Transvaal indications of an intermediate host. Onderstepoort J Vet Res 56:135–139. 3. Barnard BJH, and van de Pypekamp HE: 1988, Wildebeestderived malignant catarrhal fever: unusual epidemiology in South Africa. Onderstepoort J Vet Res 55:69–71. 4. Evermann JF, Derse D, Dorn PL: 1991, Interactions between herpesviruses and retroviruses: implications in the initiation of disease. Microb Pathog 10:1–9. 5. Gurtler L: 1996, Difficulties and strategies of HIV diagnosis. Lancet 348:176–179. 6. Heuschele WP, Seal BS: 1992, Malignant catarrhal fever. In: Veterinary diagnostic virology, ed. Castro AE, Heuschele WP, listed., vol. 1, pp. 108–112. Mosby-Year Book, Chicago, IL. 7. Knowles DP, Perryman LE, Kappmeyer LS, et al.: 1991, Detection of equine antibody to Babesia equi merozoite proteins by monoclonal antibody based competitive inhibition enzymelinked immunosorbent assay. J. Clin Microbiol 29:2056-2058. 8. Li H, Dyer N, Keller J, et al.: 2000, Newly recognized herpesvirus causing malignant catarrhal fever in white-tailed deer (Odocoileus virginianus). J Clin Microbiol 38:1313–1318. 9. Li H, Shen DT, Jessup DA, et al.: 1996, Prevalence of antibody to malignant catarrhal fever virus in wild and domestic ruminants by competitive-inhibition ELISA. J Wildl Dis 32:437– 443. 10. Li H, Shen DT, Knowles DP, et al.: 1994, Competitive-inhibition enzyme-linked immunosorbent assay for antibody in sheep and other ruminants to a conserved epitope of malignant catarrhal fever virus. J Clin Microbiol 32:1674-1679. 11. Li H, Snowder D, O’Toole D, et al.: 1998, Transmission of ovine herpesvirus 2 in lambs. J Clin Microbiol 36:223-226. 12. Mu¨ller-Doblies UU, Li H, Hauser B, et al.: 1998, Field validation of laboratory tests for clinical diagnosis of sheep-associated malignant catarrhal fever. J Clin Microbiol 36:2970-2972. 13. Nielsen L, Vestergaard BF: 1996, Competitive ELISA for detection of HHV-6 antibody: seroprevalence in a Danish population. J Virol Methods 56:221-230. 14. Porstmann T, Kiessig ST: 1992, Enzyme immunoassay techniques. J Immunol Methods 150:5-21. 15. Rovnak J, Quackenbush SL, Reyes RA, et al.: 1998, Detection of a novel bovine lymphotropic herpesvirus. J Virol 72:42374242. 16. Schultheiss PC, Collins JK, Austgen LE, et al.: 1998, Malignant catarrhal fever in bison, acute and chronic cases. J Vet Diagn Invest 10:255-262.