Engineering Peptides that Target Disordered Regions in PAX3-FOXO1–Positive Alveolar Rhabdomyosarcoma
Alveolar rhabdomyosarcoma is an aggressive childhood cancer, and many children with this disease carry a mutation that combines two different proteins, PAX3 and FOXO1, into one protein, PAX3-FOXO1. This creates a protein with new function, causing changes in gene expression that leads to cancer. Children with this mutation face poor outcomes. This fusion protein is highly dynamic and samples many different structures, too many to reasonably count in fact. This makes the protein hard to drug because we traditionally target stably folded regions of proteins. The flexible regions in PAX3-FOXO1 recruit the cell’s gene-activation machinery. This presents a possible vulnerability that we could target if we had the ability to drug these dynamic regions and inhibit these cancer-causing gene programs. If we could find a way to stop this recruitment, we could possibly have a new way to treat these patients. My project focuses on a new strategy: designing tiny peptides that can stick to these disordered regions despite them being dynamic. This “dynamic binding” approach is fundamentally different from classic drug design and may finally allow us to interfere with the fusion protein’s cancer-driving activity. By combining advanced computer modeling with in vitro experiments, we aim to learn the rules that govern how these dynamic proteins interact and use that information to create the first generation of molecules capable of targeting fusion oncoproteins like PAX3-FOXO1.
Project Goals
The goal of this project is to create a completely new way to block the cancer-causing protein PAX3-FOXO1, which drives a particularly aggressive childhood cancer called alveolar rhabdomyosarcoma. This protein is highly dynamic and lacks a fixed shape. Our traditional approaches to developing drugs don’t work well for these types of proteins. My project aims to solve this problem by designing small, dynamic peptides that can bind to these dynamic regions of PAX3-FOXO1 and disrupt its transcriptional activity. We will identify peptides that can weaken or block PAX3-FOXO1’s ability to drive tumor growth by combining in vitro measurements with deep-learning-based predictions of binding using only peptide amino acid sequences and measured binding affinities. This will provide the first direct strategy for targeting this fusion oncoprotein, something current treatments have failed to do, and hopefully one day lead to new therapies to treat children with alveolar rhabdomyosarcoma.
