The development, optimization and validation of an assay for high throughput antiviral drug screening against Dengue virus

Abstract

Dengue virus (DENV) is listed as one of the NIAID Category A priority pathogens. Dengue disease is endemic in most tropical countries, with an estimated 2.5 billion people living in areas at risk of DENV infection. Due to the lack of vaccines and antiviral drugs, it is now a huge public health burden around the world. In order to screen large compound libraries for the identification of novel antivirals targeting DENV, it is essential to develop a high throughput screening (HTS) amenable assay. Here, we present the development, optimization and validation of a cytopathic effect-based assay against Dengue virus serotype-2 (DENV-2). The assay conditions, including cell culturing conditions, DMSO tolerance and the multiplicity of infection, were optimized in both 96- and 384-well plates. Assay robustness and reproducibility were determined under the optimized conditions in 96-well plate, including Z'-value of 0.71, signal-to-background ratio of 6.88, coefficient of variation of 6.3% in mock-infected cells and 12.3% in DENV-2 infected cells. This assay was further miniaturized into a 384-well plate format with similar assay robustness and reproducibility comparing with these in the 96-well plate format. This assay was then validated using the LOPAC(1280) compound library, demonstrating its repeatability with comparable assay robustness and reproducibility. This fully developed and validated HTS amenable assay could be used in future studies to screen large compound libraries for the identification of novel antivirals against dengue disease.

Keywords: CPE; Dengue virus; HTS; antiviral; assay development; assay optimization; assay validation; cytopathic effect; high throughput screening.

Figure 1

Quantitation of CPEs induced in BSR cells post DENV-2 infection. BSR cells were seeded at 5,000 cells/well and then infected with DENV-2 at different MOIs as indicated. Inhibitions of cell viability were determined using the CellTiter-Glo reagent at 120 h.p.i.. Mock infection: BSR cells were mock-infected by adding DMEM media only. MOI: Multiplicity of Infection. Each data point represents the average value of 8 replicates.

Figure 2

Optimization of seeding cell density. BSR cells were seeded at different cell densities in 96-well plates. Cell viabilities were measured at different times post seeding and were expressed as relative luminescent signal units per well using CellTiter-Glo reagent. Each data point represents the average value of 8 replicates.

Figure 3

Effect of FBS in DMEM media on the growth of BSR cells. BSR cells were seeded at 5,000 cells/well in 96-well plates with DMEM containing different concentration of FBS and incubated at 37 °C with 5.0% CO₂. Cell viability were determined at 120 h post seeding using CellTiter-Glo reagent. Line indicated the level of luminescent signals from cells at regular maintenance media containing 5% FBS, which was pre-set as 100%. Each data point represents the average value of 8 replicates.

Figure 4

Evaluation of the DMSO tolerance in BSR cells. BSR cells were seeded at 5,000 cells/well with DMEM containing 1% FBS. The assay media also contain different concentrations of DMSO. Percentages of cell viability were determined using CellTiter-Glo reagent at 120 h post seeding. Line indicated the level of luminescent signals from cells without DMSO treatment, which set as 100%. Each data point represents the average value of 8 replicates.

Figure 5

Optimization of DENV-2 MOI for the induction of CPE in BSR cells. BSR cells were seeded at 5,000 cells/well, and then infected by different DENV-2 at different MOIs. Percentages of cell viabilities were determined using luminescent signals from CellTiter-Glo reagent at 120h post seeding. Each data point represents the average value of 8 replicates.

Figure 6

The percentage of inhibition to DENV-2 induced CPE from each compound in the validation assay using the LOPAC¹²⁸⁰ compound library. Each dot represents one compound from the library, and total of 1280 compounds were plotted here. The lower line at 8.5% inhibition of virus induced CPE was calculated as the average inhibition plus three times standard deviation of % inhibition of all compounds from the entire screen. The higher line indicating the inhibition at 20% of DENV-2 induced CPE.