Patients suffering from diseases as varied as Type II diabetes, Alzheimer's, Parkinson's and dozens of lesser known maladies have one thing in common: they suffer from a large build up of amyloids, tissue that's created when millions upon millions of misfolded proteins stick together and form a mass that the body can't get rid of on its own.
Doctors don't yet understand whether amyloids cause disease or result from it, but the fact that they are present in very different diseases affecting millions of people points to the need for improved understanding of the basic processes of protein folding, one of the most complicated and least understood of all biological phenomena.
Research appearing in the Oct. 8 issue of the Journal of Molecular Biology, describes a new technique that may help scientists predict which proteins are prone to misfold and at what point the folding process is likely to break down. The research could support efforts to find the causes for diseases involving amyloids, and it could prove useful for researchers studying proteins involved in even more prevalent diseases like cancer and heart disease.
“We know now that most diseases involve proteins going wrong in one of two ways,” said lead researcher Cecilia Clementi, assistant professor of chemistry at Rice University. “In the first, proteins don't function correctly because they fold into the wrong shape. This is something we see in sickle-cell anemia, for instance, because of genetic flaws that cause the amino acid sequence to be incorrectly synthesized.
“The second way proteins go wrong is by not folding at all, which is what we find in diseases involving amyloids. In these situations, the misfolded proteins assemble together into macroscopic aggregates.”
All the basic functions of life are carried out by proteins, and the DNA in each of our cells contains the blueprints for all the proteins we need. Every protein has a characteristic shape, and it folds itself into that shape very soon — generally in less than a second — after it is made. To carry out their tasks, proteins interact with one another, bind with some molecules, cleave others into pieces and fuse other molecules together.
Since the function of a protein is often based upon specific chemical interactions — enzymes, for instance, are proteins that make or disrupt chemical bonds in other molecules — individual atoms of a protein must be aligned just so if they are to function properly. Consequently, there is a direct relationship between a protein's shape and its function.
The study of amyloids is complicated by the fact that the shape of very few proteins is known, the mechanics of protein folding are a mystery, and protein folding is incredibly complex; even the fastest supercomputer in the world would take decades to simulate all of the chemical interactions that take place when a single protein folds itself into its characteristic shape.
Despite this mystery and complexity, Clementi and colleagues believe they are creating a statistical method that will help scientists predict which proteins are prone to misfold.
“In designing a computer model for protein folding, you can't take everything into account because there are too many variables,” said Vijay Pande, assistant professor chemistry and of structural biology at Stanford University and founder of the [email protected] distributed computing project. “By designing simplified models that retain the essential physical and chemical features of protein dynamics, Clementi's team is making excellent progress in quantitative prediction — work that's highly complimentary to the detailed simulations we're creating through [email protected]”
Basic thermodynamics dictates that the entire process of protein folding can be seen as the systematic progression of the unfolded protein into its lowest possible free energy state. To identify whether a protein is a candidate for misfolding, Clementi has devised a dynamic interplay between theory and experiment — with each informing the other as an experiment is carried out — to construct a profile of the energy states a protein progresses through while it folds.
When this energy profile is plotted on a multidimensional graph, Clementi and her colleagues study the image like a mountain range, looking for low-lying valleys — reduced energy states where the protein is most likely to become sidetracked before it can finish folding into its proper shape.
Ultimately, Clementi hopes the technique will be refined and used to catalog the folding energies of proteins that have already been implicated in specific diseases. She believes the profiles could offer new clues to doctors and drug companies about which proteins are good candidates for drug therapy.
Clementi co-authored the latest study with graduate student Silvina Matysiak. Two other graduate students in Clementi's group, Payel Das and Alexei Tcherniak, and undergraduate student Yoav Kallus, are working with Clementi and Matysiak to test and improve the technique for practical applications.
The research is sponsored by the National Science Foundation, the Welch Foundation and the Texas Advanced Technology Program.