A High Throughput Screen For Inhibitors Of Rtk Post-Translational Processing

Receptor tyrosine kinase (RTK) signaling promotes tumor cell proliferation and resistance to ionizing radiation. Inhibition of RTK post-translational modifications such as N-linked glycosylation (NLG) reduce levels of over-expressed RTKs and sensitize tumor cells to radiation therapy. We therefore developed a high throughput screening (HTS) strategy to identify compounds that inhibit endoplasmic reticulum (ER) protein processing and thus block RTK function. A cell-based, gain of function bioluminescent reporter (ER-LucT) that detects inhibition of ER post-translational processes such as NLG was developed for HTS formats. A total of 18,149 compounds from the NCI Diversity set and the ChemBridge DIVERSet were screened at 10uM. Secondary screening to identify false positives was performed with a cytoplasmic luciferase (LucT). Tertiary screening for NLG inhibition was performed using positive and negative growth selection strategies with ConA and PHA lectins. Potential lead compounds were further analyzed by western blot analysis in D54, H1975, and H1650 tumor lines to determine effects on EGFR and MET protein levels, ER stress indicators, and downstream signaling through MAPK and PI3K/Akt. The primary HTS was optimized for the D54 ER-LucT cell line. Tunicamycin was used as a positive control for the Chembridge library and yielded a mean luminescence induction of 6.1-fold (SD ± 0.93) over DMSO alone. The Z-factor for the HTS was 0.71 (SD ± 0.04) indicating robust statistical differentiation to identify potential hits. Luciferase activity was scored as percent effect relative to positive controls and the mean percent effect was -1.1% (range, -25.3% to 87.7%). This dataset formed a normal distribution at ∼0% effect (SD ± 3.9%). Compounds with activity 2-3 SDs above the mean were triaged, with a hit rate of 0.6 and 1.2% for each library. Further compounds were eliminated via the secondary screen, with a final hit rate of 0.2%. Despite the tiered luciferase screening method, several compounds had benzimidazole structures, consistent with non-specific luciferase activators. One compound was demonstrated to reduce EGFR and MET protein levels at nM concentrations without induction of Grp78 and without changes in HSP70 expression in all cell lines tested. Inhibition of NLG enzymes catalyzing ER mannosylation or ER glucosylation were excluded as the potential mechanism of action through lectin selection experiments. We have developed a cell-based HTS to select for pharmacologic inhibitors of RTK post-translational processing and maturation, as well as secondary and tertiary analyses to complement this screening strategy. The screen has identified a compound that acts in an ER-specific manner and reduces RTK protein levels in tumor cells.