Research Interest:  Early Detection, Cancer Prevention, and Tumor Microenvironment

Use MADM, a mouse genetic mosaic model, to study how tumor cells attack

To treat cancer effectively, great efforts have been devoted to molecularly targeted therapy. While there have been great examples of success, cancer cells often develop drug resistance to evade therapy. To increase the efficacy of cancer therapy, our lab uses a mouse genetic mosaic model termed MADM to study how tumor cells attack in vivo from the tumor-initiating stage and at the single-cell resolution.

The cell of origin for cancer

Since each cell types in our body have their unique “personality”/signaling context, they often respond to the same genetic mutations in entirely different fashion. Using a MADM glioma model, we have successfully identified oligodendrocyte precursor cells (OPCs) as the cell of origin, while other brain cell types fail to transform by the same set of mutations. We are currently investigate the unique signaling properties of OPCs to develop novel prevention/treatment strategies.

Cancer Prevention

Through careful analysis, we found that mutant OPCs massively outcompete WT OPCs long before malignancy. Such competitiveness readily explains the clinical observation of unstoppable progression of low grade glioma. This finding prompts us to look into novel strategies for glioma prevention. Using genetic tricks, we introduced competitive yet non-transforming cells to remove the competitive edge of mutant OPCs, and found that glioma is completely prevented. Excited by this proof-of-principle finding, we plan to dive deeply into the molecular mechanisms of OPC competition and to discover small-molecule compounds that can prevent glioma progression.

Tumor Microenvironment (TME)

Tumor cells are never alone, and their interactions with TME cells play critical roles for tumor initiation and progression. Using a MADM model for medulloblastoma, we found that tumor cells can trans-differentiate into a distinct cell type. After validating the human relevance of our finding, we further demonstrated that tumor-derived TME cells can not only directly support tumor cells, but also activate a second type of TME cells to support tumor progression. Overall, our work revealed an intricately organized TME network in medulloblastoma, shedding light on paradigm-shifting therapeutic strategies to cut off the external support toward tumor cells.

In summary, using high resolution analysis of time, space, and cellular relationships in cancer, we are poised to devise effective therapeutic strategies to detect, prevent, and defeat cancers.