Considering Signaling Pathway Kinetics and Protein Flexibility in Designing Protein Kinase Inhibitors

Document Type



Master of Science


Biochemistry & Biotechnology

Date of Defense


Graduate Advisor

Chung F. Wong


Dr. Michael Nichols

Dr. Wendy Olivas


Protein kinases are involved in various control activities in the cell, such as directing localization of proteins and controlling the function of proteins in most cellular pathways. Hence, protein kinases are very important molecules in the cell and their mutation or mis-regulation can lead to many diseases. Many forms of cancer result from improper functioning of protein kinases in the MAPK (Mitogen-activated Protein Kinase) pathways. In this work, we focused on studying the EGF-EGFR pathway that contains a MAPK pathway. We studied the impact of applying drugs to this pathway computationally using single-cell models based on three different cell lines. In addition, we used an agent-based model to study multi-cellular interactions in lung-cancer cells, in order to study the impact of applying drugs to the EGF-EGFR pathway on cell proliferation and migration. Unique in this study is the explicit account of drug-binding kinetics on suppressing cell signaling; previous computational studies focused only on studying drug-binding thermodynamics. The goals of this study were to examine: 1) whether drug-binding kinetics, in addition to thermodynamics, could also play a role in determining the effectiveness of the drug in attenuating the pathway, 2) Which protein kinases might be most useful targets, and 3) whether applying two drugs to two protein kinases in the pathways could provide synergistic effects. Another aspect of this thesis was to develop structural models of protein kinases to help design drug candidates. Many previous studies used only one structure of a protein kinase obtained experimentally or by homology modeling in computer-aided drug design. Such studies thus do not account for the conformational flexibility of the protein. In this work, we developed an automated pipeline to build multiple structural models for a human protein kinase by using a larger number of available structures of human protein kinases in the PDB (Protein Data Bank). For illustration of the automated pipeline, we built a multiple conformational model of the protein kinase SYK containing 1785 structures. The automated pipeline can be used to build multiple conformational models for any of the ~500 protein kinases in humans.

This document is currently not available here.