Picture of the Week
Computer modelling reveals cancer-causing mutations
11 June 2013
Scientists at UCL have used computer simulations to uncover why tiny mutations can trigger certain types of brain and lung cancer. A protein called epidermal growth factor receptor (EGFR), which regulates the cellular functions in response to extracellular messages, can cause cancer when it mutates into a deregulated hyper-active form, as it affects cell survival, migration and division. But studies of the molecule have not, until now, been able to pinpoint exactly why and how these mutations turn the normally inactive EGFR kinase into the hyperactive version.
Proteins do much of the work in living organisms, and they are created by the genetic code in our DNA. Mutant DNA leads to mutant proteins, which can be useless or even harmful to the body. But mutations are not all equal.
“Sometimes you get monster proteins that occur when sections of DNA have major mutations which see whole sections out of order” says Francesco Gervasio (UCL Chemistry and UCL Structural and Molecular Biology). “These proteins typically end up with entire sections of other proteins inserted into them, and it is not difficult to imagine that the resulting chimeras have dramatic effects . Other mutations, like the ones in EGFR kinase, are much smaller, as they are caused by a single letter in the genetic code being changed. Still they are very common in human cancers, implying that they give a substantial advantage to the cells in which they have occurred.”
Mutations like the ones in EGFR kinase are small, as they are caused by a single letter in the genetic code being changed
Understanding why and how small changes in the protein’s makeup can dramatically change their behaviour is a crucial question, as it would explain why these proteins can become potent triggers for cancer.
Research by Gervasio and his colleague Ludovico Sutto (UCL Structural and Molecular Biology), published in the journal PNAS, has shone light on the dramatic changes in EGFR kinase’s shape that arise from very small mutations. Studying the shapes that the protein takes in its normal and mutated forms reveals why they behave differently, and why small mutations can transform them from their passive to active forms.
Complex chemical molecules like proteins do not have a fixed shape. They usually have a particular form which is far more common, but they are not rigid molecules, and they can bend between different structures. As such, static images of their structure (such as those obtained using crystallography), while detailed, do not give the full picture of the proteins as they change over time.
The team therefore constructed computerised models based on the known chemical properties of EGFR kinase in its normal and mutant forms. They ran the models on SuperMUC, the 6th fastest super-computer in the world.to create high-resolution dynamic simulations of the proteins, showing how their shapes evolve over timescales ranging from nanoseconds to milliseconds. These in turn reveal how the small chemical changes in the mutant proteins make them take on profoundly different shapes, dramatically affecting their properties and turning them into active, cancer-causing versions.
The next step for this research is to check the computer simulations are consistent with spectroscopic measurements of the EGFR kinase. Spectroscopic measurements can complement the high resolution but static views of the kinase’s structure obtained by crystallography, allowing insight to their variability in time, which would allow confirmation that the computational models are broadly correct.
In the longer term, better understanding why the EGFR kinase’s shape is changed and activated by these small mutations could lead to new anti-cancer treatments being developed.
The research appears in a paper entitled "The effect of oncogenic mutations on the conformational free energy landscape of EGFR kinase", published online in the journal PNAS on 10 June.
SuperMUC is the 6th-fastest supercomputer in the world, according to Top 500 Supercomputer Sites (top500.org). It is located at the Leibniz Supercomputing Centre in Munich, Germany.
- UCL Department of Chemistry
- Institute of Structural and Molecular Biology
- Leibniz Supercomputing Centre - SuperMUC
- PNAS paper: Effects of oncogenic mutations on the conformational free-energy landscape of EGFR kinase
High resolution images
This image can be reproduced freely providing the source is credited
Francesco Luigi Gervasio
UCL Department of Chemistry
020 7679 3215
Page last modified on 07 jun 13 14:04