A high-throughput human tissue model for respiratory viruses: Automating the use of human airway epithelial tissues for faster drug discovery

Abstract

Animal testing for drug discovery is expensive and the decision to test a compound in an in vivo model should be carefully considered. In addition, the FDA Modernization Act has resulted in the allowance of alternatives to animal models for testing the safety and efficacy of drug candidates. Among these alternatives are human tissue models that provide a human-relevant context. Specialized cell types can be produced from primary human cells and used for basic research and drug discovery purposes. One of these is a 3D model for respiratory disease research, consisting of human-derived tracheal/bronchial epithelial cells. Though this translational Human Airway Epithelial (HAE) model is currently being employed by many researchers, this work is primarily done using individual tissue inserts placed in 6-well plates. This low-throughput approach is labor-intensive, time-consuming and expensive (high cost per each compound screened). We have established a high-throughput HAE assay that can be used for compound screening to advance respiratory virus programs. This is significant as it allows a higher number of compounds to be evaluated before being tested in vivo, allowing a more comprehensive comparison of candidates at this later discovery stage. We developed 96-well assays to evaluate compounds for Influenza, Respiratory Syncytial Virus (RSV) and Coronaviruses (including SARS-CoV-2), and these assays are easily adaptable to other respiratory viruses like Human metapneumovirus (HMPV). The development process involved performing a titration of each virus for 50 % tissue culture infectious dose (TCID₅₀) calculations and determining the optimal HAE infection time in a time course experiment, with every step of the process automated to increase speed and precision. Following infection of the HAE tissues, the amount of infectious virus in apically collected sample was assessed in a Cytopathic Effect (CPE) or Virus Titer Reduction (VTR) assay in an appropriate cell line for that particular virus. The optimized assays consistently showed Z’ values > 0.75 and were used to test reference compounds relevant to each antiviral assay. Potency values for oseltamivir and ribavirin against H3N2 A/Udorn/72 were 100 nM and 5.3 µM, respectively. The streamlined assay development process highlights the benefits of miniaturizing HAE assays from individual tissue inserts placed inside a 6-well plate to a 96-well format, providing a high-throughput solution for human 3D in vitro respiratory tissue models.