Standardization of the filovirus plaque assay for use in preclinical studies

Abstract

The filovirus plaque assay serves as the assay of choice to measure infectious virus in a cell culture, blood, or homogenized tissue sample. It has been in use for more than 30 years and is the generally accepted assay used to titrate virus in samples from animals treated with a potential antiviral therapeutic or vaccine. As these animal studies are required for the development of vaccines and therapeutics under the FDA Animal Rule, it is essential to have a standardized assay to compare their efficacies against the various filoviruses. Here, we present an evaluation of the conditions under which the filovirus plaque assay performs best for the Ebola virus Kikwit variant and the Angola variant of Marburg virus. The indicator cell type and source, inoculum volumes, length of incubation and general features of filovirus biology as visualized in the assay are addressed in terms of the impact on the sample viral titer calculations. These optimization studies have resulted in a plaque assay protocol which can be used for preclinical studies, and as a standardized protocol for use across institutions, to aid in data comparison. This protocol will be validated for use in GLP studies supporting advanced development of filovirus therapeutics and vaccines.

Figure 1

Vero E6 cells are suitable for quantitation of EBOV plaques. (A) Vero and Vero E6 cells produce EBOV plaques. (B) EBOV titers are similar in ATCC Vero E6 cells plated 24, 48 or 72 hours prior to plaque assay. This experiment was performed twice, and one representative graph is shown. Each bar represents an average of 5 replicates. The * indicates p = 0.006 between 24 and 72 hour samples for passage 29.

Figure 2

Filovirus plaques produced on Vero E6 cells from two sources are similar in appearance and titer. (A) EBOV plaques on ATCC Vero E6 cells plated at (1) 24 hours before assay, and (2) 72 hours before assay. EBOV plaques on BEI Vero E6 cells plated at (3) 24 hours before assay, and (4) 72 hours before assay. (5) MARV plaques on Vero cells from (5) ATCC and (6) BEI plated 24 hours before assay. (B) EBOV titers are similar in Vero E6 cells from ATCC and BEI when measured independently by two operators. This experiment was performed twice with up to 3 operators (data not shown), and one representative graph is shown. Each bar represents an average of 7 replicates.

Figure 3

(A) EBOV titers in ATCC or BEI Vero E6 cells of various passage ages. This experiment was performed by three independent investigators, four replicates per cell type. The combined data are presented, where each bar represents 12 replicates. * indicates a significant difference in value between BEI passage 40 and passage 27 (p = 0.0003), ** a difference between ATCC passage 54 and 28 (p = 0.007), and *** a difference between BEI passage 40 and ATCC passage 28 (p = 0.000002). For these experiments, p value cutoff was ≤0.008). (B) Analysis of EBOV titer changes in cells of various passages. The arrows point out data from passages plotted in (A).

Figure 4

A wide range of monolayer confluencies returns a similar EBOV titer. This experiment was performed by three independent operators with similar results, and one representative graph is shown. Bars represent an average of 5 replicates.

Figure 5

Evaluation of optimal inoculum volume. (A) EBOV titers may be reduced if the inoculum volume is too large. (B) 300 µL inoculum volume appears optimal for EBOV plaque formation. * indicates a significant difference in value between the 300 µL and the 500 µL conditions, where plates were rocked during incubation. Each bar represents six replicates (two experiments). (C) 300 µL inoculum volume appears optimal for MARV plaque formation. * indicates comparisons between 300 and 400 µL for Experiments 1 and 2, and across Experiments 1 and 2 resulted in p < 0.008. Each bar represents three replicates.

Figure 6

Gibco neutral red produced consistent results when tested at varying concentrations. (A) Three neutral red staining products provided different levels of staining quality. (B) Gibco neutral red produced consistent staining results when tested at varying concentrations. These experiments were performed twice with three replicates per experiment.