Simultaneous quantitation of neutralizing antibodies against all four dengue virus serotypes using optimized reporter virus particles

Abstract

Serum-neutralizing antibody titers are a critical measure of vaccine immunogenicity and are used to determine flavivirus seroprevalence in study populations. An effective dengue virus (DENV) vaccine must confer simultaneous protection against viruses grouped within four antigenic serotypes. Existing flavivirus neutralization assays, including the commonly used plaque/focus reduction neutralization titer (PRNT/FRNT) assay, require an individual assay for each virus, serotype, and strain and easily become a labor-intensive and time-consuming effort for large epidemiological studies or vaccine trials. Here, we describe a multiplex reporter virus particle neutralization titer (TetraPlex RVPNT) assay for DENV that allows simultaneous quantitative measures of antibody-mediated neutralization of infection against all four DENV serotypes in a single low-volume clinical sample and analyzed by flow cytometry. Comparative studies confirm that the neutralization titers of antibodies measured by the TetraPlex RVPNT assay are similar to FRNT/PRNT assay approaches performed separately for each viral strain. The use of this high-throughput approach enables the careful serological study in DENV endemic populations and vaccine recipients required to support the development of a safe and effective tetravalent DENV vaccine.

Importance: As a mediator of protection against dengue disease and a serological indicator of prior infection, the detection and quantification of neutralizing antibodies against DENV is an important "gold standard" tool. However, execution of traditional neutralizing antibody assays is often cumbersome and requires repeated application for each virus or serotype. The optimized RVPNT assay described here is high-throughput, easily multiplexed across multiple serotypes, and targets reporter viral particles that can be robustly produced for all four DENV serotypes. The use of this transformative RVPNT assay will support the expansion of neutralizing antibody datasets to answer research and public health questions often limited by the more cumbersome neutralizing antibody assays and the need for greater quantities of test serum.

Keywords: dengue virus; flow cytometry; immunoassays; neutralizing antibodies; viral immunity.

Fig 1

Improvement of DENV RVP production yield due to chimeric WNV-C gene cassettes. (A) The yields of serotype-specific C RVPs for all DENV serotypes and ZIKV. RVPs were produced in HEK 293T cells and harvested 4 dpt and the infectious titer was determined by twofold serial dilution in Raji-DC SIGNR cells, as described in the methods section. * For DENV2 strains New Guinea C (NGC) and Tonga/74, the serotype-specific C contained a T101G mutation to make it compatible with the West Nile virus replicon system (11). Without T101G, infectious RVPs could not be obtained. However, this modification is irrelevant when using the chimeric WNV-C structural gene cassette. (B) Chimeric DENV1-C and WNV-C structural proteins were tested to generate DENV3 RVPs. RVPs were harvested 3, 4, and 5 dpt of HEK 293T cells and the infectious titer was determined by twofold serial dilution on Raji-DC SIGNR cells, as described in the methods section. (C) The yields of WNV-C chimeric RVPs for all DENV serotypes and ZIKV, as described for the serotype-specific C RVPs (see A). (D) Different DENV3/WNV C chimeric CprME cassettes were tested for the ability to produce RVPs in HEK 293T cells. CprME expression plasmids were co-transfected with just the WNV-replicon plasmid alone (all at a 20:1 ratio of replicon to CprME plasmid) or with an additional DENV-C expression plasmid. The RVP yield was determined 4 dpt, as described above. (E) prME expression plasmids with different C-anchor sequences, as indicated, were co-transfected with either DENV3-C or WNV-C expression plasmids, and the RVP yields were evaluated 4 dpt. All RVPs were produced with the GFP-expressing WNV-replicon plasmid. Error bars reflect the standard deviation of at least two independent productions. IU, Infectious Units; LOD, Limit of Detection; WNV, West Nile virus; DENV, dengue virus; ZIKV, Zika virus.

Fig 2

Multicolor RVPs. Fluorescent proteins were selected with minimal spectral overlap to reduce the amount of compensation required. (A) Example scatter plots of all channels of a TetraPlex infected Raji-DCSIGNR cells. Raji-DCSIGNR cells were inoculated with a mix of RVP plasmids for all four serotypes of DENV, fixed after 48 h, and analyzed on BD flow cytometer Fortessa-X20. The signal was compensated for each channel with single color controls of Raji-DCSIGNR cells infected with each of the RVPs individually. (B) RVP production yield is independent of expressed fluorescent protein. RVPs were produced in HEK 293T cells with WNV-replicon plasmids encoding different fluorescent proteins as indicated, harvested after 3 d, and the titers were determined in Raji-DCSIGNR cells by serial-dilution and linear regression analysis. Arrows indicate the fluorescent reporter protein selected for use with each DENV serotype in the TetraPlex Assay. Error bars indicated the standard deviation of three independent replicates. The statistical significance of the difference between fluorescent proteins of each serotype was determined by two-way repeated measures analysis of variance (ANOVA) with Tukey’s multiple comparisons post-test. P-values are shown as *<0.05, **P < 0.01.

Fig 3

RVP infectivity in the presence of all four DENV serotypes. (A) Raji-DCSIGNR cells were infected with different amounts of RVPs for all four DENV serotypes either simultaneously (TetraPlex) or individually (SinglePlex). The infected cells were incubated for 48 h and the percentage of infected cells for each serotype was determined by flow cytometry. Error bars indicate the standard deviation of three independent replicates. The statistical significance of the difference was determined by a two-tailed t-test and p-values are shown as *<0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. IU, infectious units. (B) The fraction of single, double, triple, and quadruple positive cells relative to the total number of positive cells for the TetraPlex infection of Fig. 3A was determined for each RVP dose.

Fig 4

TetraPlex RVPNT complies with the percentage law. (A) The RVPNT₅₀ for a mAb mix (1:1) of C8:EDE1 and C10:EDE1 was determined as described in the methods section using a range of RVP input quantities for each serotype in a TetraPlex assay. The RVP input for each serotype ranged (16-fold range) from 10,000 IU (yielding approx. 15%–20% positive cells per serotype) to 625 IU (yielding approximately 2%–3% positive cells). The mean RVPNT₅₀ values of three independent replicates for each RVP input amount are shown with error bars indicating the standard deviation. (B) The RVPNT₅₀ values obtained from the TetraPlex assay (closed squares, see Fig. 5A.) were compared to the RVPNT₅₀ values obtained from SinglePlex assays (open squares) performed with the same quantity of RVPs for each serotype individually. IU, infectious units; RVPNT₅₀, RVP 50% neutralization titer.

Fig 5

RVPNT₅₀ values of the TetraPlex assay agree with RVPNT₅₀ values obtained in SinglePlex assays for each DENV serotype. The RVPNT₅₀ was determined for cross-reactive mAbs in a TetraPlex assay for all four DENV serotypes as described in the methods section and compared to RVPNT₅₀ values determined individually with SinglePlex assays for each serotype. Statistical differences (paired t-test) between the TetraPlex and SinglePlex assays could not be observed. 625 IU per serotype of RVPs were used. Error bars indicated the standard deviation of two independently performed assays. The lower limit of detection (≤20 RVPNT₅₀) is indicated by the dotted line. RVPNT₅₀, RVP 50% neutralization titer.

Fig 6

Correlation analysis of PRNT and RVPNT. (A) The 50% neutralization titer of 110 human serum samples was determined by the classic PRNT as well as the newly developed TetraPlex RVPNT assay (625 IU of RVPs per serotype of the same strain utilized for the PRNT). Any serum that had a neutralization titer below (≤5 PRNT₅₀ or ≤10 RVPNT₅₀) are above (≥640 PRNT₅₀ or ≥1215 RVPNT₅₀) the detection range was excluded from the correlation analysis. The titers of both assays for each dengue virus serotype (DENV1-4) were compared and a correlation analysis was performed. The trendline (dotted) as well as the twofold higher and lower visual threshold levels (orange lines) are shown in each graph. RVPNT₅₀, RVP 50% neutralization titer. (B) The fold difference of the RVPNT₅₀ and PRNT₅₀ values for each serum was plotted for each serotype. The horizontal bar shows the median fold difference for each serotype.