An mTOR Inhibitor Discovery System Using Drug-Sensitized Yeast

Abstract

Inhibition of the target of rapamycin (TOR/mTOR) protein kinase by the drug rapamycin extends lifespan and healthspan across diverse species. However, rapamycin has potential off-target and side effects that warrant the discovery of additional TOR inhibitors. TOR was initially discovered in Saccharomyces cerevisiae (yeast) which contains two TOR paralogs, TOR1 and TOR2. Yeast lacking functional Tor1 are viable but are hypersensitive to growth inhibition by TORC1 inhibitors, which is a property of yeast that can be exploited to identify TOR inhibitors. Additionally, yeast lacking FK506-sensitive proline rotamase (FPR1) or containing a tor1-1 allele (a mutation in the Fpr1-rapamycin binding domain of Tor1) are robustly and selectively resistant to rapamycin and analogs that allosterically inhibit TOR activity via an FPR1-dependent mechanism. To facilitate the identification of TOR inhibitors, we generated a panel of yeast strains with mutations in TOR pathway genes combined with the removal of 12 additional genes involved in drug efflux. This creates a drug sensitive strain background that can sensitively and effectively identify TOR inhibitors. In a wildtype yeast strain background, 25 {micro}M of Torin1 and 100 {micro}M of GSK2126458 (omipalisib) are necessary to observe TOR1-dependent growth inhibition by these known TOR inhibitors. In contrast, 100 nM Torin1 and 500 nM GSK2126458 (omipalisib) are sufficient to identify TOR1-dependent growth inhibition in the drug sensitized background. This represents a 200-fold and 250-fold increase in detection sensitivity for Torin1 and GSK2126458, respectively. Additionally, for the TOR inhibitor AZD8055, the drug sensitive system resolves that the compound results in TOR1-dependent growth sensitivity at 100 {micro}M, whereas no growth inhibition is observed in a wildtype yeast strain background. Our platform also identifies the caffeine analog aminophylline as a TOR1-dependent growth inhibitor via selective tor1 growth sensitivity. We also tested nebivolol, isoliquiritigenin, canagliflozin, withaferin A, ganoderic acid A, and taurine, and found no evidence for TOR inhibition using our yeast growth-based model. Our results demonstrate that this system is highly effective at identifying compounds that inhibit the TOR pathway. It offers a rapid, cost-efficient, and sensitive tool for drug discovery, with the potential to expedite the identification of new TOR inhibitors that could serve as geroprotective and/or anti-cancer agents.