Comparison of two high-throughput assays for quantification of adenovirus type 5 neutralizing antibodies in a population of donors in China

Abstract

Background: The presence of various levels of Adenovirus serotype 5 neutralizing antibodies (Ad5NAb) is thought to contribute to the inconsistent clinical results obtained from vaccination and gene therapy studies. Currently, two platforms based on high-throughput technology are available for Ad5NAb quantification, chemiluminescence- and fluorescence-based assays. The aim of this study was to compare the results of two assays in the seroepidemiology of Ad5NAb in a local population of donors.

Methodology/principal findings: The fluorescence-based neutralizing antibody detection test (FRNT) using recombinant Ad5-EGFP virus and the chemiluminescence-based neutralizing antibody test (CLNT) using Ad5-Fluc were developed and standardized for detecting the presence of Ad5NAb in serum samples from the population of donors in Beijing and Anhui provinces, China. First, the overall percentage of people positive for Ad5NAb performed by CLNT was higher than that obtained by FRNT (85.4 vs 69.9%, p<0.001). There was an 84.5% concordance between the two assays for the 206 samples tested (144 positive in both assays and 30 negative in both assays). All 32 discordant sera were CLNT-positive/FRNT-negative and were confirmed positive by western blot. Secondly, for all 144 sera positive by both assays, the two assays showed high correlation (r = 0.94, p<0.001) and close agreement (mean difference: 0.395 log(10), 95% CI: -0.054 log(10) to 0.845 log(10)). Finally, it was found by both assays that there was no significant difference observed for titer or prevalence by gender (p = 0.503 vs 0.818, for two assays); however, age range (p = 0.049 vs 0.010) and geographic origin (p = 0.007 vs 0.011) were correlated with Ad5NAb prevalence in northern regions of China.

Conclusion: The CLNT assay was relatively more simple and had higher sensitivity than the FRNT assay for determining Ad5NAb titers. It is strongly suggested that the CLNT assay be used for future epidemiological studies of Ad5NAb in other localities.

Figure 1. Optimization of conditions for CLNT and FRNT assays.

(A) Dose-response between Ad5-Fluc virus inoculum and luciferase transgene expression. (B) Dose-response between Ad5-EGFP virus inoculum and percent of infected cells. The x-axis indicated virus concentration, ranging from 31–8000 VP/cell, added to 30,000 cells/well in a 96-well plate. Each data point represented the average of three experiments. (C–H) Different amounts of virus and cell density were used in the neutralization assay (▪, 5,000 cells/well; ▴, 30,000 cells/well), and the percentages of neutralization were shown for nine serial 3-fold dilutions of Ad5-vaccinated mouse serum, with CLNT (C–E) and FRNT (F–H) assays. Each data point represents the average of three experiments. (I–J) Ad5NAb titers determination with CLNT (I) and FRNT (J) assays. Each column represents the average of three experiments. (K–P) FACS analysis of Vero cells infected with Ad5-EGFP expressing the EGFP reporter gene. Incubation with optimized viral dose (250VP/cell) and cell density (30,000 cells/well) with a 810-fold (M), 2430-fold (N), 7290-fold (O), 21870-fold (P) dilutions of Ad5-vaccinated mouse serum or without serum (L) resulted in 5.2%, 18.2%, 49.9%, 67.9% or 70.2% EGFP-positive cells, respectively, and uninfected cell control (K) resulted in 0.4% EGFP-positive cells. The mean percentage of infected cells was shown in each plot. Representative data were shown for three independent experiments.

Figure 2. Comparison between CLNT and FRNT measurements.

(A) Correlation between 50% Ad5 neutralizing antibody titers (log₁₀) of 144 human sera positive by CLNT and FRNT assays. The line represents the fitted regression line. (B) Bland-Altman plot of Ad5NAb measurements using the CLNT and FRNT assays. Solid line indicates the mean value, and dashed lines represent 95% confidence limits. The x-axis indicates the average of Ad5NAb titers (log₁₀) of CLNT and FRNT assays. The y-axis indicates the difference between the Ad5NAb titers (log₁₀) of CLNT and FRNT assays.

Figure 3. Confirmation of Ad5NAb by western blotting assay.

Sample S08 was negative and S79 were positive for Ad5NAb titers for both assays. Samples S29, S51, S52, S71 and S85 were CLNT-positive/FRNT-negative sera. The uninfected HEK293 cells as negative control was used while two samples (S71, S29) were performed confirmation by western blotting assay. The Ad5 viral capsid is composed of three major types of proteins: hexon (130 kDa), penton base (82 kDa), and fiber (62 kDa).