Mechanism of Ferroptosis
Kenichi was originally trained as a molecular biologist. He received his PhD in the laboratory of Dr. Brent Stockwell at Columbia University. Dr. Stockwell had recently discovered ferroptosis, a non-apoptotic oxidative cell death pathway induced by certain lethal small molecules. Since it’s discovery, ferroptosis has been shown to be involved in multiple diseases. Its potential in cancer therapy has also been suggested by many groups. To discover a novel regulatory mechanism of ferroptosis, he performed a chemical library screen and determined structure-activity relationships to find a new ferroptosis inducer, FIN56. Using target identification and cell biology techniques, Kenichi discovered that FIN56 induces ferroptosis by enhancing lipid peroxidation via two orthogonal mechanisms (Fig. 1A)1. During PhD, Kenichi started to incorporate chemo- and bio-informatics in his research. He also explored to measure the similarities and differences between cell death pathways induced by different chemicals, utilizing the NCI-60 drug testing dataset, containing drug sensitivity data against thousands of lethal molecules as well as basal transcriptomes of a panel of 59 cancer cell lines. Through the analysis, he discovered that lethal drug sensitivity profiles of small molecules were clustered primarily based on their targets (eg, DNA, mitochondria, tyrosine kinases) or cell death phenotypes (e.g., apoptosis, ferroptosis), and that sensitivity to ferroptosis is largely determined by the endogenous level of NADPH (Fig. 1B)2.
1. Shimada K, et al., Global survey of cell death mechanisms reveals metabolic regulation of ferroptosis. Nat Chem Biol. 2016 Jul;12(7):497-503.
2. Shimada K**, et al., Cell-line selectivity improves the predictive power of pharmacogenomic analyses and helps identify NADPH as biomarker for ferroptosis sensitivity. Cell Chem Biol. 2016 Feb 18; 23(2):225-235. (**co-correspondence)
Organ-Level Response to Drugs – Common "Disease States" Induced by 160 Toxic Substances
During his postdoc, Kenichi worked with Dr. Tim Mitchison at Harvard Medical School. He continued to mine large publicly available datasets to generate mechanistic insights relevant to small molecule drug action, disease and therapy. In one of his projects, Kenichi started to extend his interest to explore the chemical biology of organ systems from a large toxicogenomic dataset (TG-GATEs). The general goal of toxicogenomics is to better predict the toxicity of drugs and industrial chemicals, and previous analysis efforts had focused on that goal. Kenichi mined this large and meticulously organized dataset reporting the effects of 160 small molecules on organ transcriptomes, organ histopathology, and whole-body physiology in rats, for disease states and causal pathways (Fig. 2A)3. Some of the discovered disease states were known liver, kidney, or gastrointestinal injuries. But most interestingly, he was able to characterize how the liver adapt to chronic toxin exposure by up-regulation of xenobiotic metabolisms and resistance to ferroptosis. This analysis also highlighted the mechanism of toxin-induced cachexia, a primary indicator of drug toxicity in rats (Fig. 2B). (This article featured this work).
3. Shimada K*, Mitchison TJ. Unsupervised identification of disease states from high dimensional physiological and histopathological profiles. Mol Sys Biol. 2019, 15, e8636 (*correpondence)
shinyDepMap, a Tool to Discover the Next Chemo Target
In another project during postdoc, Kenichi continued to pursue his interest in cell death and resistance in cancer cells, utilizing the Cancer Dependency Map (DepMap). DepMap has performed genome-wide CRISPR and shRNA genetic perturbations in hundreds of cell lines to ask how essential a given gene is in a given cancer cell line. His first effort was to build an experimental biologist-friendly interactive web-tool, shinyDepMap4. The tool’s key contributions are to develop methods to normalize the data between the CRISPR and shRNA screens, generating two simple metrics which quantify the degree to which a given gene is essential in the most sensitive cell lines (efficacy) and selectively required in some lines, but not others (selectivity) (Fig. 3A-C). This allowed to identify 2,492 sufficiently essential genes. Clustering these essential genes by efficacy profiles across cell lines split essential genes into hundreds of functional units, many of which define complexes (e.g., ribosomes) or pathways (e.g., ferroptosis) (Fig. 3D). This analysis has helped academic groups and companies worldwide choose targets for anti-cancer drugs since the best targets will have high selectivity.
4. Shimada K*, Bachman JA, Muhlich JL, Mitchison TJ. shinyDepMap, a tool to identify targetable cancer genes and their functional connections from Cancer Dependency Map data. eLife. 2021 Feb 8;10:e57116. (*correspondence)