High Throughput Screens Identify Activators of UPR Signaling and Apoptosis

Despite technological advances in surgery, radiotherapy and chemotherapy for patients who suffer from head and neck squamous cell carcinoma (HNSCC), the survival rate has remained un-improved in the last two decades indicating our ability to treat patients has reached a plateau. Previously we have reported that the proteasome inhibitor PS-341 (Velcade) induced endoplasmic reticulum (ER) stress-dependent apoptosis in a variety of HNSCC cell lines. ER stress leads to a physiological response known as the unfolded protein response (UPR) consisting of distinct parallel genetic programs to restore homeostatic protein folding or lead to apoptosis when stress is prolonged or robust. OBJECTIVE: We hypothesized high throughput screening (HTS) of large diverse chemical libraries might identify novel small molecule activators of the apoptotic arm of the UPR and induce death in HNSCC. METHOD: To identify small molecules that modulate the apoptotic subpathway of the UPR we developed two complementary cell-based assays using stably transfected CHO-K1 cells that individually report on the PERK/eIF2-alpha/CHOP pathway and the IRE1/XBP1 pathway. Identifying compounds that specifically activate only CHOP and not XBP1 has allowed us to rule out many compounds that are generally toxic or non-specific. RESULTS: Interestingly, we have identified several families of compounds (generally considered thiriums or tharidins), each with common structural backbones, that potently induce UPR signaling and apoptosis in a variety of transformed cells as evidenced by Western blot and real time PCR analysis for the UPR markers CHOP, ATF4 and XBP1, and biochemical assays to measure viable, dead and apoptotic (caspase 3/7) cells. Conclusion: The data support the use of HTS to identify novel modulators of UPR signaling and apoptosis. Currently we are creating analog libraries of these compounds to identify more potent therapeutic candidates and to develop structure activity relationship models. This work is supported by the Howard Hughes Medical Institute (RJK)