PRNT.

The serum dilution PRNT was performed with Vero cells, as previously described (2). The following viruses were used to represent the three viral genera in all tests: EEE strain NJ/60, SLE strain TBH-28, and LAC strain Original. Endpoints were determined at a 90% plaque-reduction level.

MAC-ELISA.

This test was a modification of the assay previously reported by Beaty et al. (2). Goat anti-human IgM (PerImmune, Inc., Rockville, Md.) was used as capture antibody, and aliquots were stored at −70°C long-term, thawed once, and held at 4°C thereafter. Carbonate-bicarbonate buffer (0.015 M sodium carbonate, 0.035 M sodium bicarbonate, pH 9.6) was used as coating buffer. Coated plates (Immulon II 96-well microtiter plates; Dynatech Industries, Inc.) were incubated overnight at 4°C in a humidified container and were stable for 2 weeks. A blocking step using phosphate-buffered saline with 0.5% Tween 20 and 5% nonfat dry milk was added to reduce overall background and increase test sensitivity. Blocking and coating buffers and undiluted conjugates after reconstitution were stored at 4°C. A standard five washes with a Skatron microplate washer (Skatron Instruments, Sterling, Va.) were used except before the addition of substrate where 10 washes reduced overall background. Sera and homologous positive and negative antibody controls were titrated using 10 twofold dilutions starting at 1:100 to evaluate the optimum screening dilution. The diluent used for test sera and antigens was phosphate-buffered saline with 0.5% Tween 20 without fetal bovine serum as recommended by Beaty et al. (2). Diluted antibody was stable for only 7 to 10 days and was discarded thereafter (personal observation). Viral and normal antigens, prepared as sucrose-acetone extracts of infected suckling mouse brains (2), were obtained from the DVBID reference collection. Reconstituted undiluted antigen was stored at −20°C, and antigen dilutions were used once and discarded. An overnight incubation at 4°C was used for the antigen step to increase test sensitivity, after observing a decrease in measured absorbance (A₄₅₀) values when using a 2-h antigen incubation. Group-reactive MAbs were purified and conjugated to horseradish peroxidase by Jackson Immunological Laboratories, Inc. (West Grove, Pa.), and used as detector antibodies. The following MAbs were used: 2A2C-3 for alphaviruses (15), 10G5.4 for California group viruses (a gift from George Ludwig via Barbara Israel, University of Wisconsin) (18), and 6B6C-1 for flaviviruses (22). Use of MAb detectors eliminated the necessity for a secondary antibody addition. Commercially prepared 3,3′,5,5′-tetramethylbenzidine (TMB-ELISA; GIBCO Bethesda Research Laboratories, Inc., Gaithersburg, Md.) was used, and the substrate reaction was stopped with 1 N sulfuric acid. Reactions were measured using a Bio-Rad microplate reader (Bio-Rad Laboratories, Hercules, Calif.) at an absorbance of 450 nm.

All reagents were titrated individually by using a twofold dilution series that provided reagent excess in the first well of the series. Reagent dilutions which resulted in optical densities between 0.8 and 1.0 were chosen.

Calculation of P/N values.

Calculations that were performed followed guidelines set in the work of Beaty et al. (2) with the following modifications. The negative serum control P/N must be <2.0 and the positive serum control P/N must be ≥2.0 for the test to be valid. A positive test result was obtained when the P/N of the test serum was >2.0 and the mean of the A₄₅₀ values of the test serum reacted on viral antigen was at least twice the mean of the A₄₅₀ values of serum reacted on normal mouse brain antigen. When the latter criterion was not met due to nonspecific reaction with the normal mouse brain antigen, the result was reported as uninterpretable.

Determination of the MAC-ELISA screening dilution.

Titration curves for the representative antiviral serum samples are shown in Fig. ​Fig.1.1. The resultant curves were typical of ELISA endpoint titrations, which correlated well with concurrently run PRNT for the same sera. Figure ​Figure11 demonstrates the flat line produced by negative serum specimens even though in some cases the A₄₅₀ values were sufficient for negative P/N values to be in the range of 2.0 to 2.5. To determine the optimum dilution for screening sera, P/N values for these sera at 1:100, 1:400, 1:1,600, and 1:25,600 dilutions were compared to endpoint titers. Using chi-square analysis for all three viruses tested, a 1:400 screening dilution generated P/N ratios in the positive range that correlated well with measured endpoint titers. Only 2 of 20 positive EEE antibody specimens were positive by endpoint but negative by 1:400 dilution screening. One of 12 positive LAC antibody specimens was positive by 1:400 dilution screening and negative by endpoint titration. The 22 specimens testing positive to SLE in the screening MAC-ELISA were also positive by endpoint titration. All antibody-negative control specimens were negative by both testing procedures for all three viruses.

Open in a separate window

FIG. 1

MAC-ELISA titration curves. Log₁₀ serum dilution versus A₄₅₀ is presented. Representative specimens of high-titered sera (■), medium-titered sera (●), low-titered sera (), and negative sera () are plotted.

We thank Alan Dupuis and Rebecca Deavours for their help with the preparation of the manuscript.