← Back to Featured Research Archive

Animating Nature’s Nanomachines

When Barry Grant started his research in molecular biochemistry, he performed his single cell experiments the traditional way: with a microscope. Eventually, however, he became frustrated with techniques that could not provide the rich detail he sought. So he turned to computational simulation.

While he considers himself primarily a biochemist, Grant’s academic career exemplifies a progressive use of cyberinfrastructure:  he got his Master’s degree in Bioinformatics; a PhD in Chemistry and Biophysics; and then a post-doc that was purely about computational methods. “Ultimately, you want to understand how these fascinating protein machines work, but the knowledge of how the protein structure looks only gets you so far. You really want to see the movie of the machine functioning.” As a new faculty member in the Department of Computational Medicine and Bioinformatics, Grant uses simulations to examine how proteins change into unique shapes over time. Other molecules recognize these shapes and only bind when the first molecule takes on a certain form. Visual simulations provide a new way of seeing the structure and assembly of protein machines – as well as a window into how they work, and how they go wrong – malfunctions that contribute to cancerous tumors and neurodegenerative diseases like Alzheimer’s.

Assistant Professor Barry Grant joined U-M in the fall of 2011.

Grant is particularly interested in nature’s nanomachines: molecular motors and molecular switches, which lie at the heart of important biological processes, from the division and growth of cells to the muscular movement of organisms. Grant compares the molecular motors to a biological railway, with an engine on a track. The engine or motor delivers cargo and travels in a forward motion. Grant says that they’ve found that the motor has electrostatic features. “There are very different motors that carry different cargo but they all have a consistent electrostatic feature – this is important for how they bind to the track.”

Grant and his colleagues have learned about how to speed up and slow down the movement of motors along the track, as well as how to inhibit them. This has implications for future nanoengineering applications – scientists may be able to commandeer the motor for faster delivery of therapeutic cargo or restart a motor that has stalled. The molecular “switches” are responsible for signals that tell cells when to divide. Sometimes the switch sends a message to go hyper-on; a malfunction present in about 30 percent of all tumors. Grant’s team is experimenting with a new approach of using multiple drugs at once to hit the pathway and interrupt the signal transmission. “On the computer we’re seeing where these little molecular fragments that are drug-like would like to bind. Think of this as fitting the little shapes all over the surface, but [in the simulation] our protein moves so we can see the different conformations, where they fit, and how the protein responds. We can then begin to target our drugs to the most promising sites – a big advantage to blind screening or only using a single static structure as a guide.”

Creating simulations of the complicated systems Grant and his colleagues study requires high-performance computing methods and thousands of cores and nodes. Having access to the necessary resources is critical. Grant and his colleagues use the Flux high performance computing cluster as well as national supercomputer resources. In leading-edge science, Grant says, there is a lot of collaboration. “A few labs make software tools available for free – we develop tools for analysis and we use and adapt other peoples’ software for simulations. The analysis is often the time consuming part. We make all our code available for free to other researchers, they add to it and hopefully send it back to us.”

With the field moving as quickly as it has, in five years Grant hopes to have fully developed small molecule inhibitors that have some therapeutic benefit, although he admits it will be a challenge. “The idea is that in the field we will have well-developed tools and parameters forming a kind of standard ‘computational microscope’ that will be able to tell you something new and interesting. However, as always, you have to know what you’re looking for and what you’re asking. Being predictive and informing the development of new drugs and new proteins with tailored properties is the real challenge that will push the development of our computational microscope in exciting new directions.”