Science Stories

Discover More about XSEDE-enabled science, programs, trainings, and more.

Key Points
See how XSEDE supports science

A Basic Switch

Bridges Simulations Explain Lab Results in pH-Dependent Life Process

At high pH, the simulations show, a tight, hairpin-shaped bend in the protein (left) gives way to a wobbly, random coil (right).

By: Ken Chiacchia, Pittsburgh Supercomputing Center

Large changes in protein-chain folding accompany an acidity-dependent switch between the two modes of an important life process called electron transport, Georgia Tech scientists have suggested. Using simulations on the XSEDE-allocated Bridges system at the Pittsburgh Supercomputing Center (PSC), the team has explained otherwise mysterious lab results. The finding sheds light on a process central to plants building sugars from sunlight, higher organisms using those sugars for energy and the development of some human diseases.

Why It's Important:

Though it's far from a household phrase, electron transport is a big deal in higher organisms. Without it, the Earth would not have an oxygen-rich atmosphere. Without it, we could not gain extra energy from our food by inhaling that oxygen. Without it, our DNA could not reproduce itself properly—and so neither could we. It's safe to say that, without electron transport, the most complex organisms on Earth would be bacteria. Understanding electron transport may be central in solving problems as different as some degenerative neurological diseases in humans, survival of crop plants in arid or hot climates and the growth of the tuberculosis bacterium.

"Charge separation is one of the first steps in photosynthesis. The plant uses a photon [of light] to charge a certain molecule. That charged particle, an electron, can then go to another protein, which then uses it to carry out another step and make ATP [the cell's energy-storage molecule] eventually. Here we're looking at a very simplified system that recreates this step."—JC Gumbart, Georgia Institute of Technology.

JC Gumbart and his colleagues at the Georgia Institute of Technology wanted to understand how the two types of electron transport work. One type shuffles energy between molecules in a living cell by transporting a negatively charged electron. The other transports an electron plus a positively charged hydrogen atom, which is basically a proton surrounded by water molecules. They also wanted to learn how the structure of the proteins that carry these processes out affect them. To do this, they turned to a mix of experiments in the laboratory of Georgia Tech professor Bridgette Barry and simulation of two small artificial proteins on the XSEDE-allocated resource Bridges at PSC.

How XSEDE Helped:

Electron transport helps plants to capture energy from the sun. It also helps higher organisms to recover that energy from their food. It accomplishes both by transferring energy into high-energy electrons. These electrons, in turn, can store the energy as sugars, which can then be used for most of the functions of life. In their experiments and simulations, Gumbart, Barry and their teams turned to two artificial proteins, Peptide A and Peptide C. These proteins were designed as simplified versions of the ones that carry out electron transport in living organisms.

"These peptides were designed to be representative of actual parts of proteins involved in these [electron transport] reactions—photosynthesis, respiration and DNA synthesis. What's special about them is that a step in the reaction involves tyrosine, which, depending on the pH, participates in different ways."—JC Gumbart, Georgia Institute of Technology.

Peptides A and C are pared-down proteins that only contain the structures needed to reproduce an important step in the two types of electron transport. The electron-only type works in a relatively basic (less acidic) environment. The electron-plus-proton type works in a relatively neutral (more acidic) environment. Previous work by Barry and research scientist Cynthia Pagba suggested how changes in the pH—the level of acidity—alter the behavior of an amino acid in the protein called tyrosine. The changes in tyrosine, in turn, change how the protein folds. But this had never been directly observed.

"What I like the most about this study is the strong coupling between experimentation and simulation. You have these UVRR [ultraviolet resonance Raman spectroscopy] experiments that give you the spectra … but it's impossible to work out the dynamics. Why do the spectral lines have this dip at pH 8.5 and a bigger dip at pH 11? With the simulations, we could actually see what was going on underneath."—JC Gumbart, Georgia Institute of Technology.

With the help of XSEDE ECSS expert Marcela Madrid at PSC, the scientists employed Bridges' 800 "regular memory" nodes in simulating the proteins at different pHs. These nodes were important for the simulations partly by providing a lot of computation at one time. Bridges also provided an environment in which it was possible for the scientists to run multiple copies of the simulation at once, and to run them for longer simulated times. Bridges' power also helped in a unique approach that the team took. Instead of only modeling the molecules in their standard state, they carried out quantum mechanical calculations that helped them determine how electrical charges would be distributed in the critical tyrosine under different conditions and different steps in the process.

The simulations on Bridges reproduced changes predicted by the earlier work exactly. This reassured the scientists that their simulations were on target. At the high pH value, the simulations show that a tight, hairpin-shaped bend in the protein gives way to a wobbly, random coil. The change in overall structure hints that such processes may be important to natural electron-transport proteins. We might want to encourage electron transport, in cases such as crop photosynthesis. Or we might want to discourage it, such as in stopping tuberculosis-bacterium growth. In either case, the insight provides an entry point for further study and eventual targeting by drugs. The Georgia Tech team reported their results in the Journal of Physical Chemistry B in 2017 and have another forthcoming paper that delves into the system in more detail.

This work was supported by NSF CLP 12-13350 and NSF MCB-1452464, with partial support by the Georgia Tech GAANN program in Molecular Biophysics and Biotechnology. Computational resources were provided via the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by NSF grant number OCI-1053575.