August 2001

From University of North Carolina School of Medicine

UNC scientists develop new way to calculate stability of proteins

Pioneering study may influence protein design for drug development

CHAPEL HILL, NC - Scientists from the University of North Carolina schools of medicine and pharmacy have teamed up to develop a new way to calculate the stability of cellular proteins.

The collaboration has resulted in a pioneering report in the Journal of Molecular Biology, published Aug. 24. The report should eventually have an impact on the way proteins for new drug development are designed and engineered.

With the human and mouse genomes essentially sequenced, determining from that complex blueprint how proteins function and interact in cells of the body has taken center stage in the ongoing effort of basic biomedical science to understand the regulation of important physiological processes in health and disease.

Charles W. Carter Jr, PhD, professor of biochemistry and biophysics at UNC-CH School of Medicine and the report's lead author, approaches these issues by looking at how proteins are built.

"I study the architecture of proteins," he said. "And because I do that, I'm also interested in how they function. How they function depends to a large extent on what makes them stable."

Why is protein stability important?

"The function of a protein depends on the fact that it can sustain the same structure over an extended period of time," Carter said. "Thus, for purposes of drug design, the whole area of protein engineering is highly dependent on knowing the contributing factors that make a protein stable."

Getting to know a protein structure begins with understanding its unique chemistry, amino acid building blocks strung together end-to-end like beads via peptide bonds. What makes each protein unique is the specific order of these different building block. This sequence, along with the length of the string, forms the unknown code that determines how fast it folds, what its final shape looks like and how it functions. Carter and his colleagues, Professors Marshall Edgell and Alex Tropsha want to crack that code.

Knowing a protein's unique amino acid arrangement isn't enough information to reliably predict its 3-dimensional architecture. So the next step involves techniques such as nuclear magnetic resonance (NMR) imaging and X-ray crystallography to determine protein structure in atomic resolution. Experiments based on the resultant 3-D images have provided a database of how all the atoms in the building blocks interact in many thousands of different proteins. These imaging experiments sometimes facilitate the analysis of what happens to proteins if specific amino acids are altered to create protein mutations or variants.

Using this database of known protein structures, Carter and his co-authors were able to relate protein structure to stability by making computational manipulations. This allowed them to predict the effects on stability of 'virtual' amino acid mutations introduced into the hydrophobic (water avoiding) core of known protein structures.

When proteins fold, certain types of amino acid side-chains, have a propensity to aggregate or clump together and exclude water. This "hydrophobic effect" is similar to what's observed when pouring oil into water or seeing oil separate from vinegar.

The aggregation takes place in the protein's hydrophobic core. And proteins derive a considerable portion of their stability from the interactions between these amino acids, Carter explained.

"For many years it's been difficult to provide a quantitative description of how this hydrophobic force is mediated," he said. "I think we have made an important step in that direction. We are able to go from the structure of a protein and its variants - by which I mean discrete mutations introduced into the hydrophobic core - to an algorithm [formula] for calculating the effect these mutations have on stability ."

The report's authors emphasize that the three-dimensionality of proteins is key to understanding their stability. "We argue and support the argument with the data in our paper that it's important - perhaps essential - to treat the interactions within the internal volume of a protein in subsets of four at a time to produce an effective accounting for the energetics, including their equilibrium stability."

According to the UNC researchers, the best way to accomplish this is to consider the internal volume of a known protein structure as tiled with tetrahedra, each having an amino acid side-chain at each of their four vertices. And it's the amino acid compositions of these tetrahedral tiling motifs that is apparently important for the stability of proteins, Carter said.

"Extracting and computationally manipulating useful structural properties from the database to help us understand protein stability, represents what is now called 'structural genomics.'" Carter said.

Co-authors of the report, along with Carter, are Drs. Marshall Hall Edgell, Kenan Professor of Microbiology and Immunology and Alexander Tropsha of the Laboratory for Molecular Modeling at UNC School of Pharmacy, Drs. Stephen A. Cammer a former research student, and Brendan C. LeFebvre, an undergraduate NSF fellow.

The research was supported by grants from the National Institute of General Medical Sciences at NIH and the National Science Foundation.

media note:
Contact Dr. Edgell at 919-962-0147,
After Sept. 10 , contact Dr. Carter at 919-966-3263,
School of Medicine contact: Leslie Lang, 919-843-9687,

This article comes from Science Blog. Copyright 2004

Archives 2001 E