A Parallel Phenotypic Versus Target-Based Screening Strategy for RNA-Dependent RNA Polymerase Inhibitors of the Influenza A Virus

Abstract

Influenza A virus infections cause significant morbidity and mortality, and novel antivirals are urgently needed. Influenza RNA-dependent RNA polymerase (RdRp) activity has been acknowledged as a promising target for novel antivirals. In this study, a phenotypic versus target-based screening strategy was established to identify the influenza A virus inhibitors targeting the virus RNA transcription/replication steps by sequentially using an RdRp-targeted screen and a replication-competent reporter virus-based approach using the same compounds. To demonstrate the utility of this approach, a pilot screen of a library of 891 compounds derived from natural products was carried out. Quality control analysis indicates that the primary screen was robust for identification of influenza A virus inhibitors targeting RdRp activity. Finally, two hit candidates were identified, and one was validated as a putative RdRp inhibitor. This strategy can greatly reduce the number of false positives and improve the accuracy and efficacy of primary screening, thereby providing a powerful tool for antiviral discovery.

Keywords: Influenza A virus; RNA dependent RNA polymerase inhibitor; phenotypic screen; target-based screen.

Figure 1

Optimization of cell-based influenza RNA-dependent RNA polymerase (RdRp) assay in a high-throughput screen system. 293T cells transiently expressing viral RdRp proteins (PB2, PB1, PA) and NP, together with a minigenome RNA carrying the firefly luciferase gene, were seeded into 96-well plates (16 replicates) and subjected to a luciferase assay at the indicated times post transfection. The signal-to-background (S/B) ratios, coefficient of variations (CVs), as well as the Z’ factors for each batch of data, were calculated.

Figure 2

Establishment of a high-throughput screen system using the replication-competent recombinant influenza virus carrying Gaussia luciferase, PR8-Gluc. (a) Signal range of the PR8-Gluc virus-based infection assay. Madin-Darby canine kidney (MDCK) cells were infected with the reporter influenza virus PR8-Gluc virus at multiplicities of infection (MOIs) of 0.001, 0.01, 0.1, and 1, respectively. Gluc activities were measured at 24 and 36 hr post-infection (p.i.), and those of the mock-infected cells were regarded as background. Data are shown as the average ± standard deviation (n = 3). (b) Quality control of the PR8-Gluc virus-based screening system. MDCK cells growing in 96-well plates (24 replicates) were infected with a PR8-Gluc virus at an MOI of 0.01, at 36 hr p.i. Gluc activities were measured and the S/B ratio, CV, and Z’ factor were calculated. Mock infected cells were measured for Gluc activities as background.

Figure 3

Experimental design of the parallel high-throughput screening strategy. (a) RdRp-targeted and PR8-Gluc virus-based screen were sequentially or simultaneously carried out to evaluate the same library of compounds, and the data of each screen were collected for comparison. (b) Upper panel: For each plate in both screens, columns 1 and 12 were treated with baloxavir acid and DMSO as positive and negative controls, respectively. Lower panel: A compound that shows an inhibition of >80% in an assay plate is selected as hits (black). Hits shared by two screens are selected as putative RdRp inhibitors; hits specific to PR8-Gluc virus-based screen are secondary hits and not discussed in this study, while hits specific to RdRp-targeted screen are regarded as false positives.

Figure 4

Pilot screen using the parallel high-throughput screen approach. A pilot screen was carried out with a library containing 891 natural products. The quality of primary screen was controlled by assessing the (a) S/B ratios, (b) CVs, and (c) Z’ factors of each plate. The range of the three parameters for each screen were presented by box-and-whisker plots. (d) The signal distribution of all 891 samples. (e) Primary hit identification strategy. The percent inhibition of each test compound against PR8-Gluc infection was plotted with that against RdRp activity. Compounds showing >80% inhibition in both screens (shared “hits”) are shortlisted for a second screening. (f) Second screening. Shared hits identified by the primary screen were subjected to a second round of the parallel screen as well as a cytotoxicity screen. Toxic hits indicate the compounds that reduce cell viability by >30%.

Figure 5

Antiviral determination of JL-5001 and JL-5002. (a–c) Dose-response curves of (a) JL-5001, (b) JL-5002, and (c) baloxavir acid upon PR8-Gluc virus replication, RdRp activity, as well as viabilities of MDCK and 293T cell lines. The inhibitory effects were analyzed using GraphPad Prism 5. (d) MDCK cells were infected with wild-type PR8 virus at an MOI of 0.01 and incubated with indicated concentrations of JL-5001, JL-5002 or baloxavir acid. At 36 hr p.i., the virus titers were determined. ns, no significance; *, p < 0.05; **, p < 0.01; student’s t test.