A Predictive Nuclear Translocation Assay for Spliced X-Box–Binding Protein 1 Identifies Compounds with Known Organ Toxicities

Abstract

Compound toxicity is still the main cause of attrition, emphasizing the need for novel predictive assays to identify toxic compounds early during drug development. Endoplasmic reticulum (ER) stress has recently been discovered as a molecular event that links cellular dysfunction to drug-induced organ toxicity in humans. Among higher organisms the inositol-requiring transmembrane kinase/endoribonuclease pathway plays a major role in mediating the ER stress response. Inositol-requiring transmembrane kinase/endoribonuclease achieves this through its endoribonuclease activity causing a frameshift in the translation of the X-box–binding protein 1 (XBP1) to produce spliced XBP1 (XBP1s), which translocates into the nucleus, where it initiates transcription of ER stress response genes. Based on this biology, we have designed a novel β-galactosidase–based XBP1s-enzyme fragment complementation assay, which enables identification of compound-induced ER stress in human U2OS cells. The XBP1s-enzyme fragment complementation assay was established in a 384-well format and validated using a library of 1280 pharmacologically active compounds. Importantly, the library of pharmacologically active compounds screen identified both well-established ER stress inducers and several compounds that are known organ toxicants but not previously reported to induce ER stress. Implementation of this assay to assess compound-induced ER stress will facilitate decision making for compound selection and we believe that it will significantly increase the ability to reduce toxicity of preclinical drug candidates.