Journal Article

174 High-throughput high-plex proteomic profiling of hepatocyte toxicity to oncologic drug compounds: a platform for toxicity assessment, immune risk prediction, and drug development acceleration

Edwardson Kiran, Dagher Milad, Rosenbloom Alyssa, Robichaud, Nathaniel, Rashidi Narges, Eberlein Jens

2025

Overview

BackgroundThe rapid and necessary expansion of cancer drug development requires accessible and scalable strategies for early toxicity screening and immune risk profiling. Liver plays a central role in drug metabolism, detoxification, and immune modulation. Robust detection of liver-specific responses to oncologic therapeutics is essential to identify and mitigate hepatotoxicity and immune-related adverse events (irAEs) which could limit clinical development. Here, we demonstrate a high-throughput, high-plex proteomic platform as a powerful approach to evaluate functional perturbations of hepatocytes across a spectrum of cancer drugs.MethodsUtilizing the novel Omni 1000 proteomic solution, based on nELISA technology, we profiled 1,000 proteins in >6500 samples of primary human hepatocytes following exposure to a diverse panel of over 550 distinct drug compou nds at 3 doses per compound. Well-chosen Omni 1000 protein target content enables simultaneous quantification of stress response markers, inflammatory cytokines, metabolic enzymes, and signaling pathway effectors. This allows for evaluation of both cytotoxic and immunomodulatory signatures, revealing potential off-target effects and mechanistic insights.ResultsPerturbation with tripolide, doxorubicin, and cobimetinib resulted in dose dependent cell toxicity, detected via increased levels of intracellular proteins such as GAPDH and Casp3 released into the supernatant, consistent with clinically observed hepatotoxicity. We also detect increased Galectin-3, Galectin-1, and VEGFA at non-toxic perturbation doses, indicative of cell stress and potential toxicity at higher doses. In addition, we concurrently observe loss of proteins associated with protection against apoptosis, such as ST6GAL1, a potent inhibitor of FAS-mediated apoptosis. Across the 510 compounds screened, 68 showed similar signs of toxicity, including compounds with dose-limiting liver toxicity in the clinic. Of particular note is PHA-793887, a potential therapeutic candidate, which failed clinically due to unexpected hepatotoxicity, which here was captured in a pre-clinical in vitro model.We further demonstrated the platform’s ability to identify likely irAEs through detection of dose dependent increases in key cytokines. Hepatocytes perturbed with Imatinib and Pomalidomide demonstrated increases in IL-6, IL-10, and IL-2, indicative of potential irAE and consistent with published literature for these drugs.ConclusionsHigh-throughput high-plex proteomic analysis of hepatocyte responses offers a significant advance in preclinical toxicology and immune safety, and enables an early stage indication of drug-induced liver injury (DILI) risk and immunotoxic potential, two major barriers in oncology drug development. As the pharmaceutical industry increasingly utilizes systems-level screening, such proteomic platforms will be instrumental in enhancing predictive toxicology, reducing clinical attrition, and accelerating the path from discovery to approval.

• Publisher: BMJ Publishing Group LTD
• Subject: Apoptosis / Cytokines / Drug development / Drug dosages / Liver / Proteins / Toxicity / Toxicology