Blockade-of-Binding Activities toward Envelope-Associated, Type-Specific Epitopes as a Correlative Marker for Dengue Virus-Neutralizing Antibody

Abstract

Humans infected with dengue virus (DENV) acquire long-term protection against the infecting serotype, whereas cross-protection against other serotypes is short-lived. Long-term protection induced by low levels of type-specific neutralizing antibodies can be assessed using the virus-neutralizing antibody test. However, this test is laborious and time-consuming. In this study, a blockade-of-binding enzyme-linked immunoassay was developed to assess antibody activity by using a set of neutralizing anti-E monoclonal antibodies and blood samples from dengue virus-infected or -immunized macaques. Diluted blood samples were incubated with plate-bound dengue virus particles before the addition of an enzyme-conjugated antibody specific to the epitope of interest. Based on blocking reference curves constructed using autologous purified antibodies, sample blocking activity was determined as the relative concentration of unconjugated antibody that resulted in the same percent signal reduction. In separate DENV-1-, -2-, -3-, and -4-related sets of samples, moderate to strong correlations of the blocking activity with neutralizing antibody titers were found with the four type-specific antibodies 1F4, 3H5, 8A1, and 5H2, respectively. Significant correlations were observed for single samples taken 1 month after infection as well as samples drawn before and at various time points after infection/immunization. Similar testing using a cross-reactive EDE-1 antibody revealed a moderate correlation between the blocking activity and the neutralizing antibody titer only for the DENV-2-related set. The potential usefulness of the blockade-of-binding activity as a correlative marker of neutralizing antibodies against dengue viruses needs to be validated in humans. IMPORTANCE This study describes a blockade-of-binding assay for the determination of antibodies that recognize a selected set of serotype-specific or group-reactive epitopes in the envelope of dengue virus. By employing blood samples collected from dengue virus-infected or -immunized macaques, moderate to strong correlations of the epitope-blocking activities with the virus-neutralizing antibody titers were observed with serotype-specific blocking activities for each of the four dengue serotypes. This simple, rapid, and less laborious method should be useful for the evaluation of antibody responses to dengue virus infection and may serve as, or be a component of, an in vitro correlate of protection against dengue in the future.

Keywords: ELISA; blockade of binding; dengue virus; neutralizing antibody.

FIG 1

Relationship between the neutralizing antibody titer and the type-specific epitope-blocking activity in macaque blood samples taken 30 days after infection with recent clinical isolates or attenuated strains. Scatterplots show the neutralizing antibody titers (PRNT₅₀) and type-specific epitope-blocking activities (RC [relative concentration]) on a log₁₀ scale to facilitate the visual inspection of all values. The Spearman correlation coefficient (rho) derived from nontransformed data is shown on each plot together with the 95% confidence interval in parentheses, the P value, and the number of macaques (n).

FIG 2

Relationship between the neutralizing antibody titer and the type-specific epitope-blocking activity in pooled blood samples. Scatterplots show the relationship between the neutralizing antibody titer (PRNT₅₀) and the type-specific epitope-blocking activity (RC [relative concentration]). Colored lines represent regression lines for each macaque. The repeated-measures correlation coefficient (r_rm) is shown on each plot together with the 95% confidence interval in parentheses, the P value, and the numbers of macaques and blood samples (n), respectively.

FIG 3

Relationship between the neutralizing antibody titer, the 513 epitope-blocking activity, and DENV-2-specific epitope-blocking activities in pooled blood samples. Scatterplots and regression lines show the relationship between the 513 epitope-blocking activity and the neutralizing antibody titer (A to D, G, and H) and between the 513 epitope-blocking activity and DENV-2-specific epitope-blocking activities (E and F) for different sample sets. DENV-2 strain 16681 or prE203A was employed as a binding target in panels A to F; DENV-1 strain Hawaii was employed in panels G and H. The repeated-measures correlation coefficient (r_rm) is shown on each plot together with the 95% confidence interval in parentheses, the P value, and the numbers of macaques and blood samples (n), respectively.

FIG 4

Relationship between the neutralizing antibody titer, the C10 epitope-blocking activity, and the DENV-2-specific epitope-blocking activity in macaque blood samples. Scatterplots and regression lines show the relationship between the C10 epitope-blocking activity and the neutralizing antibody titer (A to D and H) and between the C10 epitope-blocking activity and the DENV-2-specific epitope-blocking activity (E and F) or the 513 epitope-blocking activity (G) for different sample sets. DENV-2 strain 16681 or prE203A was employed as a binding target in panels A to G; a DENV-4 strain was used in panel H. The repeated-measures correlation coefficient (r) is shown on each plot together with the 95% confidence interval in parentheses, the P value, and the numbers of macaques and blood samples (n), respectively.

FIG 5

Relationship between the neutralizing antibody titer and binding activity in pooled blood samples. Scatterplots and regression lines show the relationship between the neutralizing antibody titer and the DENV-2 particle-binding activity (A) or the DENV-2 EDIII-binding activity (B) for the DENV-2 infection/immunization sample set. The repeated-measures correlation coefficient (r_rm) is shown on each plot together with the 95% confidence interval in parentheses, the P value, and the numbers of macaques and blood samples (n), respectively. EPT, endpoint titer.