Computational identification of novel therapeutic candidates for Streptococcus pyogenes and influenza A coinfections through transcriptomic-based drug repositioning.

BACKGROUND: Influenza A virus (IAV) causes severe illness with a high mortality rate, and secondary bacterial infections can lead to severe pneumonia. Despite the availability of antibiotics and antivirals, treatment of concurrent IAV and invasive group A streptococcal infections remains challenging. As bioinformatic drug repurposing represents a cost- and time-effective manner for discovering novel treatments, this study aimed at the identification of novel therapeutic options for the S. pyogenes-IAV coinfection through this computational approach. RESULTS: Following in vitro infections of pharyngeal epithelial cell lines with either IAV or S. pyogenes serotypes M1 or M49, transcriptomic changes in host cells were analyzed by RNA-seq, obtaining patterns of differentially expressed genes for each infection. These genes were then queried against the LINCS L1000 small molecule database to find compounds capable of reversing infection-induced molecular phenotypes. In this manner, we identified through computational analyses the antitumoral and antibacterial compound mitoxantrone as well as three kinase inhibitors: AT7519 and flavopiridol, which act against cyclin-dependent kinases, and BI2536, an inhibitor of the Polo-like kinase 1 (PLK1), as the main candidates to treat these coinfections. CONCLUSIONS: Using computational drug repurposing we identified four compounds that have the potential to act as suitable drugs in IAV-S. pyogenes coinfections. BI2536 and flavopiridol have been previously confirmed as active against IAV infections in vitro, while mitoxantrone is effective against S. pneumoniae. These results validate our approach, which offers a cost-effective alternative to large-scale drug screenings to find suitable candidate compounds.