Abstract
Triosephosphate Isomerase deficiency (TPI-Df) is a devastating untreatable childhood metabolic disease resulting in anemia, severe locomotor impairment, and premature death. Numerous single amino acid substitutions in TPI are pathogenic and result in rapidly progressing multisystem disease. Importantly, all known pathogenic TPI-Df mutations result in a protein that retains function, and pathogenesis is known to result from decreased steady state levels of the functioning protein. There are no small molecule therapies for TPI-Df; current treatments are limited to symptomatic support and dietary interventions. We reasoned that a phenotypic screen was most appropriate to capture agents that stabilize mutant TPI and developed a human cellular TPI-Df assay based on a cellular model of the "common" TPI(E105D) mutant protein fused with a GFP and a fluorescent ROS biosensor. The assay was implemented for high-content, high-throughput imaging, optimized to full HTS standards, and used to screen a 2,560 compound pilot library and the 220,700 compound NIH MLSMR compound collection to identify candidate compounds for development into small molecule TPI-Df therapies. Hits were validated in dose-response, TPI-Df patient cells, and various orthogonal assays. Limited SAR revealed three promising compound series, which were evaluated for potential mechanisms of action. The lead series had previously been identified as inducers of HIF1 alpha, spawning a novel hypothesis that HIF1 alpha activation might be a potential avenue to treat TPI-Df patients. A lead molecule was moved into preliminary mouse studies to evaluate pharmacokinetics and tissue distribution and was shown to be moderately brain-penetrant. The lead compound is now positioned for target identification studies and efficacy testing in vivo TPI Df models, including a newly validated mouse model.