J. gen. Virol. (I98o), 48, 225-229 22 5 Printed in Great Britain Indirect Enzyme-linked Immtmosorbent Assay (ELISA) for the Detection of Coxsackievirus Group B Antibodies (Accepted 26 November r979) SUMMARY An indirect, solid phase, microplate enzyme-linked immunosorbent assay (ELISA) was found to be highly sensitive and reliable for detecting antibodies to the group B Coxsackieviruses and other picornaviruses. Conditions for obtaining maximum sensitivity and reproducibility of the indirect ELISA are described. Antibody titres were comparable to those obtained by the virus neutralization test and over 50 times higher than those obtained by the complement-fixation test. Purified Wrions used in the indirect ELISA reacted with low levels of cross-reacting heterotypic antibodies elicited by each of the six group B Coxsackieviruses, although homotypic reactions resulted in highest titres. The indirect ELISA, designed for the detection and quantification of antibodies, was first described by Engvall & Perlmann (r972). Since its development, the ELISA has been applied to the study of many different viruses (Gilman & Docherty, 1977; Voller & Bidwell, t976). The present report documents the application of the microplate ELISA for titrating antibodies prepared against the group B Coxsackieviruses (BI to B6) and other picornaviruses. Lack of a sensitive seroiogical assay has made an immunological analysis of the group B Coxsackieviruses difficult to achieve. The only sensitive test available is the neutralization test (Melnick & Ledinko, t95o), which measures type-specifiic antibodies to virus surface antigens and is not suitable for studying the antigenic relationship among these viruses. Serological assays, such as complement-fixation (Kraft & Melnick, i95o), gel immunodiffusion (Schmidt & Lennette, t962) and indirect immunofluorescence (Mac- William & Cooper, r974), are relatively insensitive. The present study shows that the antibody titres obtained by the indirect ELISA for each of the six group B viruses were comparable to those of the virus neutralization test and over fifty times higher than those obtained by the complement-fixation test. In addition the ELISA proved to be a sensitive assay for detection of cross-reacting antibodies among the group B viruses. The following strains of Coxsackieviruses used in this study were obtained from the University of Minnesota: Coxsackievirus BI (Conn.- 5), B2 (Redmond), B3 (Nancy), B4 (Texas ~3), B5 (53 t22 obtained from Dalldorf)(Crowell & Syverton, I96r; Crowell, I963). Coxsackievirus B6 was obtained from Dr. Hammon, University of Pittsburgh. Virus was propagated in suspension cultures of HeLa (Mandel) cells at a multiplicity of zo to 50 p.f.u./cell by methods described previously (Crowell & Phitipson, 197I). Virus was purified by consecutive banding in two caesium chloride gradients as described by Crowell & Philipson (r97i). The protein content of the purified virions was determined by a modification of the micromethod of Lowry et al. (t95 I) using crystalline bovine serum albumin (Miles, Elkhart, Ind., U.S.A.) as a standard. Virus was titrated on HeLa (JJH) cell monolayers grown in 60 t 5 mm plastic Petri dishes as previously described (Crowell & Syverton, I96t). Virus concentrations were described as the number of p.f.u./ml of original virus preparation. For the neutralization test, virus containing IOO p.f.u./o') ml was mixed with an equal volume of a given dilution of antiserum. Serum neutralization titres were defined oo22-i317/80/c, eeo-394~ $02.00 x98o SGM 8-2
226 Short communications as the serum dilution which reduced the plaque-forming titre by one-half as obtained from interpolation of the data plotted in graphic form (Svehag & Mandel, 1964). Each virus was tested for the presence of contaminating viruses by reciprocal cross-neutralization tests. All viruses and their antisera were found to be type-specific. Antisera to the six group B Coxsackieviruses and poliovirus type 2 were prepared in female New Zealand white rabbits. The rabbits were inoculated twice intravenously (i.v.), three weeks apart, with o-2 ml purified virion suspension containing IOO t~g virus protein and io 1 to to n p.f.u. On the same day as the i.v. injections, a mixture of 0.2 ml virus and 0.2 ml Freund's complete adjuvant (FCA; Difco Labs, Detroit) was injected into the middle two toes of the rabbit's hind foot. The rabbits inoculated with Coxsackieviruses B3 and B 5 received a third injection of virions and FCA into the hind footpad. Sera were collected before immunization and shown to be free of antibodies to the viruses tested. Bleedings were repeated postimmunization every 2 to 3 weeks. Hyperimmune sera were pooled, divided into aliquots and frozen at -7o C until used. The microplate adaptation (Voller et al. 1974) of the indirect solid-phase ELISA described by Engvall & Perlmann (1972) was performed with modification. Polystyrene microtitre plates (Cooke Micro ELISA Substrate Plates, Dynatech Labs, Alexandria, Va., U.S.A.) were coated with o.2 ml purified virions (2"5/ g/ml) diluted in phosphate buffered saline, PBS (o.ot M-phosphate buffer; o.85~){, NaCI), ph 7"2. Plates were sealed and incubated in a humidified chamber overnight at 6 C. Prior to use in the assay, the virus coated plates were washed three times, for at least 3 rain per wash with PBS, ph 7"2, containing o'o5,/o Tween- 2o (PBS/Tween). Specific antiviral serum, serially diluted in PBS/Tween, was added in 0.2 ml amounts to the wells, followed by a 90 min incubation at room temperature. After additional washing, o.2 ml of an appropriate dilution of peroxidase-labelled goat antirabbit immunoglobulin (Cappel Laboratories, Cochranville, Pa.) was added for another 9 o rain incubation at room temperature. After an additional three washes, the amount of peroxidase bound was determined by addition of o.2 ml O-phenylenediamine (OPD; Eastman Kodak Co., Rochester, New York) as substrate for 3o min at room temperature. The substrate solution, prepared immediately before use, consisted of o'0o3 ~o H202 and o.ot ~o OPD (w/v). The enzyme-substrate reaction was stopped by 0"05 ml 8 N-H2SO4, and the yellow-brown colour produced was measured at 49o nm in a Beckman DBG spectrophotometer. The ELISA antibody titre was defined as the highest serum dilution giving an absorbance value o.25 absorbance units above control levels. All samples were tested in duplicate with the average value reported. The duplicate values generally agreed within o'o5 absorbance units. Day to day reproducibility of the antibody titres obtained by the ELISA was more than 9o % within one twofold dilution of the antiserum tested. To ensure that the observed ELISA activity was due to specific virus-antibody interactions, the following control studies were performed. A positive serum control which monitored both assay reproducibility and stability of the test reagents was included in each experiment. A normal serum control was also included to detect non-specific reactivity with virus-coated wells. Wells were also coated with PBS alone, or the negatively charged protein, bovine serum albumin (BSA), and reacted against the hyperimmune BI antiserum. This control was done to test whether virus antibodies adsorbed to wells without virus. A conjugate control was included to test for non-specific reactivity of the peroxidase labelled immunoglobulin with virus-coated wells which had not been incubated previously with specific rabbit antiserum. Finally, virus antiserum was reacted against a crude extract and a mock CsCl-purified extract of uninfected HeLa cells. Despite the rather high purity of the virus preparations, these controls were necessary to rule out host cell contamination. All of the control preparations give negative or background levels of reactivity. The results of experiments measuring the antibody titres and antigenic relatedness of the
Short communications 227 Table I. Homotypic and heterotypic reactivity of hyperimmune rabbit antisera prepared against purified group B Coxsackieviruses as detected by the indirect ELISA Antibody* t" A Virus BI B2 B3 B4 B5 B6 Polio T2 BI 40000 (Ioo)t 200 (I) 800 (2) 300 (2) 400 (2) z4oo (0.8) < ioo Bz 400 (I) I5OOO (too) 400 (l) 6oo (4) 80o (3) 16o0 (o'5) < loo B3 B4 3200 (8) 16o0 (4) 800 (5) zoo (I) 35o00 (loo) 600 (z) 8oo (5) 15ooo (loo) [ zoo (5) 6oo (2) 64o0 (2) z4oo (o.8) < Ioo 1oo B5 24oo (6) 800 (5) 16oo (5) 80o (5) 25ooo (Ioo) 64oo (z) < ~oo B6 800 (z) 600 (4) 16oo (5) - 800 (5) 12oo (5) 3oo0oo (too) ioo HeLa uninfected < IOO < IO0 < I00 < 100 < 100 100 < 100 extract * Neutralization titres of the six group B virus antisera were: B1:5oooo; B2:2OOOO; B3:6oooo; B4:4oooo; B5:I3OOO; B6:3ooooo. Complement fixation titres for homotypic group B reactions were: B1:64o; B2:16o; B3:3zo; B4:8o; B5:16o; B6:128o. Complement fixation test was performed according to Lennette et al. (1974). t Numbers in parentheses refer to percentage (%) of cross-reactivity. The percentage was computed from the heterotypic to homotypic antibody titre ratio for each group B antiserum. The ratio was then multiplied by mo to give the percentage. group B viruses are summarized in Table I. Purified virions and their respective hyperimmune antisera were tested in the ELISA assay according to the protocol described above. For each group B virus-antibody reactions the homotypic titres were significantly higher than the heterotypic reactions. A heterotypic to homotypic antibody titre ratio was computed for each of the six group B virus antisera. These ratios revealed that the heterologous activity detected ranged from o'5 % to 8 % of the homologous reaction. Coxsackievirus B6 antiserum, with the highest homotypic titre of 1:3ooooo, showed the lowest relative levels of cross-reactivity of o'5 to 2 /o. In an additional test of assay specificity, high titre hyperimmune antiserum (neutralization titre I:65ooo) to poliovirus type 2, an unrelated enterovirus, was found to be non-reactive with each of the group B viruses. The homotypic antibody titres measured by the ELISA were then compared to the titres obtained by both the virus neutralization and complement-fixation test. The ELISA titres were within a two- to threefold dilution of neutralizing antibody titres. Furthermore, the ELISA was at least 5o times, and more often too times, more sensitive than the complement fixation test for determination of virus antibodies. In another test of the applicability of the ELISA system, antisera to other representative human picornaviruses were examined. Using purified homologous antigens, antibodies to polioviruses types I and 2, Echovirus 6 and human rhinovirus 2 were successfully measured by the ELISA with a sensitivity again comparable to the neutralization test. The results of these studies showed the ELISA to be highly sensitive for the detection of both type-specific and group-reactive antibodies to the group B Coxsackieviruses. As shown in Table r, the levels of cross-reacting antibodies were relatively minor compared to the homotypic response measured by the indirect assay. A commongroup B virus antigen was also detected by Schmidt & Lennette (1962) in gel diffusion studies in which immune human serum was employed. It proved essential to purify the virus antigens for use in the ELISA since experiments performed using commerically prepared crude Coxsackievirus antigens resulted in an insensitive and non-specific assay. Antigen purity has also been found to be critical for the development of a sensitive and specific ELISA for both herpes simplex and rubella viruses (Gilman & Docherty, 1977; Gravell et al. 1977). The results summarized in Table ~ also focus attention on the nature of the antigen in the ELISA. Although native virions were utilized throughout the present study, slight changes in virus structure may have occurred during the overnight adsorption to the microtitre
228 Short communications plates. Since the test is highly sensitive, a small fraction of disrupted virions could have given rise to the heterotypic activity routinely observed in the assay. When virions were disrupted with 3 M-urea or heat (56 C for 30 rain) before use, an increase in virus group-reactivity was detected by the ELISA (M. G. Katze, unpublished results). This observation is in general agreement with.schmidt et al. 0963) who found that the serological activity of virus rich caesium chloride fractions could be converted from virus type-specific to groupreactive by heating at 56 C for 30 rain. The present report describes a rapid and sensitive test for the detection of antibodies to Coxsackievirus and provides a basis upon which human serological studies may be performed. To reduce the potential biohazards using native virions, disrupted virus could be used as antigen. It is likely that a Coxsackievirus group B reactive antigen would be useful in this test to screen for the detection of primary infection by these viruses. Development of an ELISA assay which measures IgM antibody might also prove useful. Because of the high incidence of Coxsackievirus antibodies in the serum of individuals in the general population, the indirect ELISA might best be suited for use in paediatrics for young children who have not been exposed to the many enteroviruses in the environment. This investigation was supported by a U.S. Public Health Service research grant AI-o377I from the National Institute of Allergy and Infectious Diseases. Department of Microbiology and Immunology Hahnemann Medical College Philadelphia, Pennsylvania I9IO2, U.S.A. M. G. KATZE R. L. CROWELL REFERENCES CROWELL, R. L. 0963). Specific viral interference in HeLa cell cultures chronically infected with Coxsackie B5 virus. Journal of Bacteriology 86, 517-526. CROWELL, R. L. & PHILIPSON, L. (I97I). Specific alterations of Coxsackievirus B3 eluted from HeLa cells. Journal of Virology 8, 5o9-515.. CROWELL, R. L. & SYVERTON, J. T. (I961). The mammalian cell virus relationship. VI. Sustained infection of HeLa cells by Coxsackie B3 virus and effect on superinfection. Journal of Experimental Medicine xi3, 419-435. ENGVALL, E. & PERLMANN, V. (I972). Enzyme-linked immunosorbent assay, ELISA. III. Quantitation of specific antibodies by enzyme-labelled anti-immunoglobulin in antigen-coated tubes. Journal of Immunology xo9, I29-135. GILMAN, S. C. & DOCHERTY, J. J. 0977). Detection of antibodies specific for herpes simplex virus in human sera by the enzyme-linked immunsorbent assay. Journal of Infectious Diseases, Supplement x36, $286- $293. GRAVELL, M., DOr~SETT, P., GUTTENSON, O. & LEY, A. C. (I977). Detection of antibody to rubella virus by enzymelinked immunosorbent assay. Journal of Infectious Diseases, Supplement x36, $3oo-$3o3. KRAFT, L. M. & MELNICK, J. L. (I950), Immunological reactions of the Coxsackieviruses. I1. The complement fixation test. Journal of Experimental 3~fedicine 02, 48"~-497. LENNETTE, E. H., MELNICK, J.L. & MAGOFFIN, R.L. 0974). Clinical virology: introduction to method. In Manual of Clinical Microbiology, pp. 667-677. Edited by E. H. Lennette, E.H. Spaulding and J. P Truant. Washington, D.C.: American Society of Microbiology. LOWRY, O. FI., ROSEBROUGH, N. J., FARR, A. L. & RANDALL, R. J. (1951). Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry x93, 265-275. MACWILLrAM, K. g. & COOPER, M. A. (1974). Antibody levels in human sera measured by the fluorescent-antibody technique against the Coxsackie B viruses types i to 5 grown in Hep 2 cells compared with results obtained by neutralization. Journal of Clinical Pathology 27, 825-827. MELNICK, J, L. & LEDINKO, N. (1950). Immunological reactions of the Coxsackievirus. I. The neutralization test: techniques and applications. Journal of Experimental Medicine 92, 463-482. SCFIMIDT, N. J., DENNIS, J., FROMMtlAGEN, L. H. & LENNETTE, E. H. (I963). Serologic reactivity of certain antigens obtained by fractionation of Coxsackie viruses in cesium chloride density gradients. Journal of Immunology 9 o, 654-662.
Short communications 229 SCr~MIDT, N. J. & L~7,rNET-rE, E. ~. 0962). Gel double diffusion studies with group B and group A, type 9 coxsackie viruses. I. The technique and reactions obtained with hyperimmune animal sera and human sera. Journal of Immunology 89, 85--95. SVErrAO, S. E. & MANt)EL, B. (1964). The formation and properties of poliovirus-neutralizing antibody. I. 19S and 7S antibody formation: difference in kinetics and antigen dose requirements for induction. Journal of Experimental Medicine xx9, I-I 9. VOLLER, A. & B~OWELL, O.E. (I976). Enzyme-immunoassays for antibodies in measles, cytomegalovirus infection and after rubella vaccination. British Journal of Experimental Pathology 57, 243-247. roller, A., INDWELL, O., HWLOT, ~. & ENGVALL, E. 0974). A microplate method of enzyme-linked immunosorbent assay and its application to malaria. Bulletin of the Worm Health Organization 5I, 2o9-21 I. (Reeeived I I September I979)