About

Abstract

Telomerase activity, while absent in most somatic cells, is a hallmark of numerous cancers, as upregulation in its expression confers replicative immortality in the cell. Broadly, this project endeavours to find therapeutics to target proteins that mediate telomerase activity and function . By studying the protein-protein interaction (PPI) network of yeast (closely related to the human interactome), we strive to identify novel targets with clinical applications in cancer therapy.

Network Theory: Community and Target Finding

From the outset, it seems unnatural to use a binary network to model the complicated dynamic interactions of proteins; a binary network is a static object with no care for biology. In the network, for example, we lose all sense of space and time, things of course which protein interactions rely on. Simple network measures, such as shortest paths, are therefore far too naive to capture the complexities of a living cell. Our hope, here, is to use methods which can somehow retain the probabilistic nature of protein interactions.

We begin by exploring the PPI network with WalkTrap, a community finding algorithm based on random walks. This allows us to identify the telomerase community and its closest functional groups. We then use betweenness to visualise the local topology of the network about the telomerase complex, allowing us to identify naive protein targets which seem most important for the interaction of telomerase with the rest of the cell. A more complicated distance measure, diffusion distance, based on random walks, gives us targets which may be more biologically informed.

Validation

A Pearson correlation coefficient allows us to mathematically validate these targets to ensure removing them is not lethal to the cell. We then validate these proteins biologically. This will be done via CRISPR/Cas9 mediated knockouts of lead proteins selected from the mathematically validated list. These proteins are chosen due to their cell localization with telomerase and associated functions. These knockouts will then undergo a yeast viability assay, where those that do not show immediate cell death will undergo a telomere assay to determine if the size of the telomere changed. Knockouts successful in reducing telomere size can then undergo the PROTAC workflow.

Drug Development: From Protein To PROTAC

We have also developed a workflow to develop therapeutics for these validated targets. PROTACs are an exciting therapeutic that have demonstrated enormous potential in cancer treatment. They are chimeric molecules that hijack a cell’s own protein degradation machinery in order to degrade a specific protein.

In order to produce a PROTAC to degrade a target, we first must identify and test small molecule ligands for their binding affinity to the target. When a reasonable candidate is found, they can be conjugated to a linker and E3 ligase, producing a PROTAC whose efficacy can be evaluated both in vivo and in vitro.

As a proof of concept, this report details the application of this PROTAC development workflow to two proteins: ATM and POP1. However the goal is to utilise the overall strategy in order to develop PROTACs targeting any strong candidates obtained through analysis of the PPI network around telomerase.

A flow chart of the process of developing the therapeutics, starting from target identification with network theory, and ending with PROTAC testing is shown below:

Team

This project was created by Lucy Arnold, Alex Loustau, Simone Titterton (Biochemistry Team); Kate Fiumara, David He, Russell He, and Amelie Skelton (Mathematics Team) to satisfy MATH3888/BCMB3888 capstone requirements at the University of Sydney.

Lucy Arnold: Wrote the following sections: ATM (within Target Proteins), Linking Telomerase to Cancer, Experimental Evaluation of PROTACs (Introduction, Cell Line Preparation, Western Blot, Flow Cytometry, Mouse Model Experiments), Methods for the Protein Validation.
Alex Loustau: Wrote the following sections: Telomerase and its Biological Function, Advantages of PROTACs, Evaluation of Ligands and Integration into PROTACs (overexpression and purification of POP1 and ATM, ITC analysis of ligands, and PROTAC Design). Also made the final presentation.
Simone Titterton: Wrote the following sections: POP1 (within Target Proteins), Limitations of Current Treatment, Finding Ligands, Terminal Restriction Fragment Analysis (within Experimental Evaluation of PROTACs) and the justification for the Protein Validation.
Kate Fiumara: Implemented WalkTrap algorithm; researched and implemented Pearson’s correlation coefficient, date hub/party hub differentiation; identified the final proteins; wrote the corresponding parts in the report; helped make the final presentation.
David He: Researched and implemented diffusion distance in R; wrote the corresponding section; researched and implemented the Jaccard and Restricted community distance measures to determine the most relevant communities; helped make the final presentation.
Russell He: Designed and implemented the website; created the telomerase interactive visual; wrote the motivation for community finding and random walks; helped make the final presentation.
Amelie Skelton: Helped implement the website; made the website network model and flow chart; wrote the telomerase in the community section and made the community network visualisation; wrote the section on betweenness and made the betweenness visualisations; researched similarity measures; drafted the abstract; edited the sections on “Why Network Theory” and “Why Community Finding”; helped make the final presentation.