Research Topics

The overarching goal of this research is to improve the efficacy of precision cancer therapies, including chemotherapy and immunotherapy. This work is guided by two primary aims: (1) identifying biomarkers that predict responsiveness to specific therapies, and (2) elucidating mechanisms by which tumors develop therapeutic resistance. These challenges are addressed through integrative analyses combining single-cell transcriptomic data with high-dimensional spatial imaging. For imaging, Kenichi employs cyclic immunofluorescence (CyCIF), a technology developed in the LSP at Harvard Medical School.

During his postdoctoral training, Kenichi pursued a project focused on cell death and therapeutic resistance in cancer cells using the Cancer Dependency Map (DepMap). DepMap systematically performs genome-wide CRISPR and shRNA genetic perturbation screens across hundreds of cancer cell lines to assess gene essentiality in a context-dependent manner.

As part of this work, Kenichi developed shinyDepMap, an interactive, experimental biologist–friendly web tool designed to enable intuitive exploration of large-scale functional genomics datasets. A key contribution was the development of methods to normalize data across CRISPR and shRNA screens, yielding two interpretable metrics that quantify (i) the degree to which a gene is essential in the most sensitive cell lines (efficacy) and (ii) the extent to which a gene is selectively required in a subset of cell lines (selectivity) (Fig. 3A–C). Using this framework, 2,492 genes were identified as sufficiently essential.

Clustering these essential genes based on their efficacy profiles across cell lines revealed hundreds of functional gene modules, many of which correspond to known molecular complexes (e.g., ribosomes) or biological pathways (e.g., ferroptosis) (Fig. 3D). This work has helped both academic groups and industry partners prioritize therapeutic targets, as genes with high selectivity represent particularly attractive candidates for anti-cancer drug development.

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)

During his postdoctoral training, Kenichi worked with Dr. Tim Mitchison at Harvard Medical School, where he leveraged large publicly available datasets to derive mechanistic insights into small-molecule drug action, disease processes, and therapeutic response. In one project, he expanded his interests into the chemical biology of organ systems by analyzing a large toxicogenomic resource, TG-GATEs.

Toxicogenomics aims to improve the prediction of drug and chemical toxicity, and prior analyses of TG-GATEs primarily focused on this objective. Kenichi reanalyzed this extensively curated dataset—which profiles the effects of 160 small molecules on organ-specific transcriptomes, histopathology, and whole-body physiology in rats—to identify disease states and underlying causal pathways (Fig. 2A). While some inferred states corresponded to known liver, kidney, or gastrointestinal injuries, a particularly notable finding was the characterization of an adaptive liver response to chronic toxin exposure, marked by upregulation of xenobiotic metabolism and resistance to ferroptosis.

This analysis also elucidated a mechanism underlying toxin-induced cachexia, a major indicator of drug toxicity in rats (Fig. 2B). This work was featured in Harvard News.(

Kenichi was originally trained as a molecular biologist and received his PhD in the laboratory of Dr. Brent Stockwell at Columbia University. During his doctoral training, Kenichi contributed to the early discovery and mechanistic characterization of ferroptosis, a non-apoptotic, oxidative form of cell death induced by specific lethal small molecules. Since its identification, ferroptosis has been implicated in a wide range of diseases, and its potential relevance to cancer therapy has been widely explored.

To identify novel regulatory mechanisms of ferroptosis, Kenichi performed a chemical library screen coupled with structure–activity relationship analyses, leading to the discovery of a new ferroptosis inducer, FIN56. Using a combination of target identification and cell biological approaches, he demonstrated that FIN56 induces ferroptosis by promoting lipid peroxidation through two orthogonal mechanisms (Fig. 1A)¹.

During his PhD training, Kenichi also began integrating chemo- and bioinformatics approaches into his research. He investigated similarities and differences among cell death pathways induced by diverse small molecules using the NCI-60 drug testing dataset, which includes drug sensitivity profiles for thousands of compounds alongside baseline transcriptomes from 59 cancer cell lines. This analysis revealed that lethal drug sensitivity profiles cluster primarily by molecular targets (e.g., DNA, mitochondria, tyrosine kinases) or cell death phenotypes (e.g., apoptosis, ferroptosis), and further showed that sensitivity to ferroptosis is largely determined by endogenous NADPH levels (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)