THE JOUR~AL OF INFECTIOUS DISEASE. VOL. 136, SUPPLEMENT. OCTOBER 1977 © 1977 by the University of Chicago. AU rights reserved.
Enzyme-Linked Immunosorbent Assay for Hepatitis B Surface Antigen From the Scientific Development Group of Organon International B. V., Oss, The Netherlands
Gerrit Wolters, Leo P. C. Kuijpers, Jovan Kacaki, and Anton H. W. M. Schuurs
Our laboratory has worked for several years on the development of a sensitive immunochemical technique, the enzyme immunoassay (enzymelinked immunosorbent assay, ELISA) [1-3]. This method has also been studied by many other groups of investigators . This paper describes the results obtained with a solid-phase ELISA based on the "sandwich" principle for the detection of hepatitis B surface antigen (HB s Ag) in blood of donors and in blood of patients with hepatitis B. Materials and Methods
Test reagents. Details of the preparation of the reagents have been described elsewhere . In brief, the solid-phase antibody consisted of Cooke Microtiter" plates (Greiner, Niirtingen, 'Vest Germany), the wells of which were coated with antibody to HB s Ag (anti-Hls.). The coated plates were stored under dry conditions at 4 C and were stable for at least 12 months. The enzyme-labeled antibody (conjugate) consisted of anti-Hls, coupled to horseradish peroxidase. We thank Mr. J. Croese and Mr. F. Rodes for their technical assistance, and Mr. S. Stulemeyer for his statistical help. Please address requests for reprints to Dr. Gerrit Wolters, Biochemical Research and Development Laboratories, Organon Scientific Development Group, Kloosterstraat 6, Oss, The Netherlands.
The conjugate could be stored frozen or lyophilized for at least 12 months. Test performance. The sample (0.1 ml), either plasma or serum, was placed in a well of an antibody-coated Microtiter plate. Control sera, either containing HB s Ag (positive control) or free of HB s Ag (negative control), were included in each test run. The plates were covered with parafilm and incubated for 2 hrat 37 C or for 16-24 hr at room tern perature. The wells were washed with 0.2 M Tris buffer (pH 7.4) containing 0.05% Tween. To each well 0.1 ml of the conjugate solution was added carefully. The plates were covered again and incubated for 2 hr at 37 C. The wells were then washed thoroughly with the same Tris buffer. The enzymatic activity was determined by addition of 0.1 ml of a freshly prepared solution of o-phenylenediamine and urea peroxide to each well. After 50 min at room temperature the reaction was stopped by addition of 0.05 ml of 4 N H 2S0 4 , The yellow color of the reaction product was compared with negative controls by the naked eye and colorimetrically by measurement of absorbance at 492 nm. Colorimetric results were considered positive if the signal-to-noise ratio was at least 2.1, that is, the absorbance test was at least 2.1 times the average absorbance of five negative controls. In general, reading by eye approached the sensitivity of the colorimetric reading, but was less reliable in borderline cases.
Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on July 1, 2015
A solid-phase enzyme-linked immunosorbent assay (ELISA), based on the "sandwich" principle with use of microtiter plates, was developed for the detection of hepatitis B surface antigen (HB s Ag). Results could be read within one day by the naked eye or by colorimeter. The detection level was ===5-10 ng of HB s Ag/ml. The sensitivities of ELISA and radioimmunoassay were about the same in dilution series and in a followup study of 19 patients with acute hepatitis B infection. In II European medical centers where >50,000 samples were tested, ELISA detected significantly more HBs Ag-positive samples than a reversed hemagglutination test. No significant difference in sensitivity between ELISA and radioimmunoassay could be demonstrated. On the average, 2.2% of readings were false-positive reactions. Falsely positive samples were identified by a confirmatory test.
ples giving discrepant results between test systems were sent to us for further investigation. Results
Reference panels. The German reference panel consisted of 50 samples (two positive sera in 20 dilutions and 10 negative samples). The concentration of HB s Ag in each sample and the results of 34 West German laboratories' solidphase RIA were unknown to us before testing. When the code was broken, it was seen that the concentration of HB s Ag in the positive samples ranged from 16,000 to 0.03 ng/ml for subtype ad and from 10,000 to 0.02 ng/ml for subtype ay. The results obtained with the positive samples are summarized in table 1. The negati ve sam ples did not react in ELISA. For BOB panel no. 3, all 17 positive samples labeled A, B, and C were found positive in ELISA, both by eye reading and colorimetrically. Both positive samples labeled C, which were described by the BOB as "occasionally reactive by third-generation test methods" (such as RIA), were positive in one ELISA and equivocal in another. The samples labeled D and described
Table 1. Results of enzyme-linked immunosorbent assay (ELISA) and radioimmunoassay (RIA) of samples from the German National Reference Panel that were positive for hepatitis B surface antigen (HB s Ag). HBs Ag subtype, amount" (ng/ml)
Percentage positive in RIAt
Signal-to-noise ratio in ELISA
Subtype ad ~23.5
10.5 3.7 1.2