Simulations Explore Protein-folding Interactions

---

By Salvatore Salamone, Bio-IT World online

Scientists at IBM Corp. (Armonk, NY) and Columbia University (New York, NY) have conducted new simulations to study the basic science that drives the folding of proteins in water.

The results of their investigations appear in the current issue of Science in a paper titled “Hydrophobic Collapse in Multi-domain Protein-folding.” The paper’s authors note that understanding the so-called hydrophobic collapse (of proteins in water) has implications in many areas of research.

“The hydrophobic (interaction) is at the heart of so many processes,” says Bruce Berne, PhD, one of the paper’s authors and a member of Columbia University’s department of chemistry. For example, it’s how soap works, how oil separates from water, and how many biological systems self-assemble.

In fact, in many instances such as protein-folding and the formation of cell membranes, hydrophobic interactions force objects to self-assemble into their final shape or composition.

Through simulations, the IBM and Columbia University researchers were able to better understand the basic science of the hydrophobic interaction.

“The paper is about mechanisms,” says Ruhong Zhou, PhD, another author of the paper and a member of the Computational Biology Center (Yorktown Heights, NY) at the IBM Thomas J. Watson Research Center. “(We are) really trying to understand the physics and chemistry of the interaction.”

The researchers performed the simulations on a fairly large protein: the BphC enzyme. (BphC is what is known as a two-domain protein; the focus of this research was on such multi-domain proteins.) Also, thanks to the computational power available at the labs involved, the researchers were able to perform more sophisticated simulations than were previously possible on such a large protein.

Specifically, the researchers were able to selectively turn on and off various protein-water electrostatic interactions and protein-water van der Waals attractions, all of which drive the protein-folding process.

“You can’t do this in the lab, but you can do this in simulations,” Zhou says. The result is that researchers can better understand the impact of the various forces in the entire process.

Berne notes that this is an example of the power of using a high-performance computer in biology and chemistry. “In the lab, you can't turn processes on and off, you can’t manipulate (experiments) in this way,” Berne says. “But through simulations, we can do thought experiments and unravel why a process takes place.”

The paper’s other authors are Claudio Margulis, PhD and Xuhui Huang, who were both with Columbia University during the study.