Now, the same concept provides clues to unraveling the mystery of how different mutations in different genes can cause diseases with very similar characteristics. In a paper in today's edition of the journal Cell, scientists from Baylor College of Medicine and Harvard Medical School describe a network of proteins that when perturbed by mutations can result in ataxias, characterized by loss of balance and coordination because of degeneration of specific nerve cells. The concept can be used to study many diseases.
"We study degenerative diseases," said Dr. Huda Zoghbi, professor of molecular and human genetics, neurology, neuroscience and pediatrics at BCM, a Howard Hughes Medical Institute investigator and senior author of the paper. "A gene is mutated, and a defective protein is made. We need to know what proteins interact with it, and then we study those interactions to learn how the mutant protein can damage the neuron (or nerve cell)."
"A cell makes tens of thousands of proteins, "said Zoghbi. "One protein interacts with many of them. How do you find all the proteins that are relevant?"
She and the paper's first author, Dr. Janghoo Lim, a postdoctoral trainee at BCM, decided instead to study all the proteins that, when mutated, cause similar clinical problems, in this case ataxias.
"Perhaps the disease happens when these mutant proteins interact with common partners," she said.
To answer that question, Lim began to identify protein partners in the Harvard laboratory of Dr. Marc Vidal, associate professor of genetics and faculty member of the Dana Farber Cancer Institute.
"On the one hand, you have patients with diseases, and those diseases look alike, like the ataxias," said Vidal. "We are in the business of understanding the wiring diagram of cells. In each cell, there are 20,000 to 25,000 proteins, and they interact with one another. We know that many diseases are basically related back to mutations in genes. But what if the whole network was perturbed in a way to cause disease?"
"We picked 23 different disease genes to study," said Lim.
All of them caused the same kind of degeneration in nerve cells, even though they were different genes making different proteins.
"We felt they must have common protein partners," he said.
"We found that the proteins made by those 23 genes all interacted with one another," said Zoghbi. "They either interacted directly or indirectly through common partners. Proteins that we know can change the course of disease in animal models were also identified as common partners."
"In this network are spots where many proteins interact," she said. "These are key proteins."
These proteins, she said, could be extremely important in insuring normal neuronal function.
"When you step back, you realize this is applicable to any human disease," she said. "You start with a handful of genes you know cause disease and then go find their partners, and then their partners and then build a network that will include all the factors that make you susceptible to a disease or cause the symptoms of it."
"Instead of thinking one gene-one disease, you realize that if there are 10 genes that cause the same symptoms, they must do it through some common pathways or interactions with other proteins."
In applying the network theory to understanding disease, Zoghbi and her colleagues have come full circle. Symptoms used to always be the basis for medical treatment.
"Now we are providing a mechanistic basis for understanding why we treat symptoms," she said.
In the future, treatments may be designed to interrupt the cellular missteps that lead to disease.
Others who participated in this research include: Drs. Chad Shaw, Akash J. Patel and Joseph Fi sk, of BCM, Drs. Gábor Szabó, Jean-François Rual, Tong Hao, Ning Li, Alex Smolyar, David E. Hill and Albert-László Barabási of Harvard. (Drs. Barabási and Szabó are now at the University of Notre Dame.)