Molecular Neurogenetics

To systematically model the genetic spectrum affecting human glioblastoma (GBM), we employ human induced pluripotent stem cell (iPSC)-derived organoids to elucidate the genetic heterogeneity within GBM. Utilizing CRISPR/Cas9, we generated a comprehensive spectrum of mutation combinations (Pten, Trp53, Cdkn2a, Cdkn2b, Nf1, Rb1, Egfr, Tert, etc.), which are the most frequently mutated genes in human GBM, in human iPSCs. Subsequently, we induced GBM organoids from these cells, which we refer to as LEGO: Laboratory Engineered Glioma Organoid (npj Preci Onco, 2024). These organoids effectively recapitulate key molecular features of human GBM.

In the LEGO^2.0 model, we further enhance its fidelity by incorporating barcodes, immune cells, and blood vessels. This comprehensive model enables us to systematically dissect the interaction between genetic heterogeneity and functional heterogeneity within GBM. Additionally, drug screening is underway to establish a genotype-based drug reference for personalized treatment of GBM

Tumor organoids are pivotal tools in cancer research, yet current models face limitations that hinder their application in predicting therapeutic responses. To address this challenge, we have developed a sophisticated culture system (IPTO, individualized patient tumor organoid, Cell Stem Cell, 2025) that accurately simulates the cellular and molecular pathology of human brain tumors. Patient-derived tumor explants were cultured within induced pluripotent stem cell (iPSC)-derived cerebral organoids, enabling the cultivation of a diverse range of human tumors within the central nervous system (CNS), including adult, pediatric, and metastatic brain cancers. Histopathological, genomic, epigenomic, and single-cell RNA sequencing (scRNA-seq) analyses revealed that the IPTO model accurately recapitulates cellular heterogeneity and molecular features of the original tumors. 

Notably, we demonstrated that the IPTO model accurately predicts patient-specific drug responses, including resistance mechanisms, in a prospective patient cohort. Collectively, the IPTO model represents a significant advancement in preclinical modeling of human cancers, paving the way for personalized cancer therapy. The IPTO model is also utilized for target discovery, validation, and T cell-based cancer therapies. The high-quality drug response data generated from the IPTO models are being employed to train predictive generative AI models for patient response prediction.

We generated a series of mouse models and human organoid models that allow us to precisely manipulate gene expression in neural stem cells and brain tumor stem cells. In this project, we are dedicated to unravel how the dynamic transcriptional regulations are involved in regulating the identity, activity and fate of NSCs. We are systematically analyzing the role of essential transcription factors and chromatin regulators in neurogenesis in order to generalize the molecular principles of cell identity determination. In particular we are interested in studying transcription regulators like Nr2e1 and Chd7 which are mutated in human neurological disorders and brain cancers.