Computational analysis of cell-cell communication in the tumor microenvironment

• Main Author: Kumar, Manu Prajapati.
• Other Authors: Douglas A. Lauffenburger.
• Format: Others
• Language: English
• Published: Massachusetts Institute of Technology 2019
• Subjects: Biological Engineering.
• Online Access: https://hdl.handle.net/1721.1/123063

Thesis: Ph. D., Massachusetts Institute of Technology, Department of Biological Engineering, 2019 === Cataloged from PDF version of thesis. === Includes bibliographical references (pages 147-168). === Cell-cell communication between malignant, immune, and stromal cells influences many aspects of in vivo tumor biology, including tumorigenesis, tumor progression, and therapeutic resistance. As a result, targeting receptor-ligand interactions, for instance with immune check-point inhibitors, can provide significant benefit for patients. However, our knowledge of this complex network of cell-cell interactions in a tumor microenvironment is still incomplete, and there is a need for systematic approaches to study cell-cell communication. This thesis presents computational approaches for characterizing cell-cell communication networks in three different experimental studies. In the first study, we modeled metastatic triple negative breast cancer in the liver using a microphysiological system and identified inflammatory cytokines secreted by the microenvironment that result in the proliferation of dormant metastases. In the second study, we used single-cell RNA sequencing (scRNA-seq) to quantify receptor-ligand interactions in six syngeneic mouse tumor models. To identify specific receptor-ligand interactions that predict tumor growth rate and immune infiltration, we used receptor-ligand interactions as features in regression models. For the third study, we extended our scRNA-seq approach to include inferences of single-cell signaling pathway and transcription factor activity. We then identified protein-protein interaction networks that connect extra-cellular receptor-ligand interactions to intra-cellular signal transduction pathways. Using this approach, we compared inflammatory versus genetic models of colorectal cancer and identified cancer-associated-fibroblasts as drivers of a partial epithelial-to-mesenchymal transition in tumor cells via MAPK1 and MAPK14 signaling. Overall, the methods developed in this thesis provide a foundational computational framework for constructing "multi-scale" models of communication networks in multi-cellular tissues. === by Manu Prajapati Kumar. === Ph. D. === Ph.D. Massachusetts Institute of Technology, Department of Biological Engineering