For the first time, scientists are closing in on the molecular mechanism that causes Parkinson’s disease, giving sufferers of that debilitating disease new hope.

The tremors, rigid posture and shuffling gait of Parkinson’s disease have been associated for decades with the die-off of dopamine-producing neurons in the brain. These neurons tend to be riddled with suspicious protein clumps. What scientists haven’t known, until recently, was how these protein plaques were connected to the disease. Now a team of scientists led by Igor Tsigelny, a project scientist in chemistry and biochemistry at the San Diego Supercomputer Center at the University of California- San Diego (UCSD), has demonstrated the molecular mechanism behind Parkinson’s disease for the first time. Their modeling work, performed on supercomputers at Argonne National Laboratory in DuPage County, Illinois, has already yielded a promising treatment for Parkinson’s disease. At the same time, it has provided other researchers with tools to aid the study of other disorders associated with abnormally aggregated proteins, including Alzheimer’s and prion diseases.

The protein clumps in Parkinson’s disease consist primarily of a protein called alpha synuclein (aS). Long a prime target for Parkinson’s research, aS has resisted conventional protein analysis because it won’t stand still. “It doesn’t have a stable conformation,” Tsigelny said. “Like Alice in Wonderland, it’s changing shape all the time.” Undaunted, Tsigelny decided to study aS in motion. A computer modeling approach known as molecular dynamics enabled him to track the twists, spins and gyrations of every atom in a solution of aS proteins and solvent molecules. His goal: to catch a close-up look at aS proteins in the act of aggregating.

However, tracking the motions of so many particles is too taxing for today’s fastest desktop computers. So Tsigelny requested an allocation of supercomputing time through the U.S. Department of Energy’s Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program. His proposal was granted 1.2 million hours of processor time on Argonne’s Leadership Computing Facility’s IBM Blue Gene/P system, which debuted in June as the world’s fastest computer dedicated to unclassified research.

Tsigelny developed a simulation involving some 800,000 atoms. After using 16,000 processor hours in 2006 and 75,000 hours in 2007, Tsigelny observed aS forming ringlike pentamers and identified the protein conformation involved in oligomer formation. His finding bolstered a previous hypothesis suggesting that aS proteins aggregate to form pores that pierce the membranes of otherwise healthy nerve cells. The pores allow a flood of calcium ions to enter the neuron, triggering a cascade that leads to further clumping of aS and other proteins and ultimately results in cell death. “We couldn’t have done this without the supercomputers,” said Tsigelny. “They gave us the power to track enough molecules over time to see the interactions we were looking for.”

Tsigelny’s virtual view of pore formation enabled him to identify the protein-binding sites on aS proteins in exquisite detail. Further studies revealed that beta synuclein, a brain protein very similar to aS, appeared to inhibit aS molecules from linking together. Working with Eliezer Masliah, a professor of Neurosciences and Pathology at the UCSD School of Medicine, Tsigelny has used these findings to develop a compound capable of clogging aS binding sites, halting pore formation. Tsigelny used his computer models to design and test the compound first. Masliah’s later laboratory tests on cultured mouse neurons show the compound has a protective effect, and the drug appears to be nontoxic so far. If approved, “it will be the first drug to treat the cause of Parkinson’s disease instead of the symptoms,” Tsigelny said. The treatment would offer hope to the more than 1 million people estimated to be living with Parkinson’s disease today.

But Tsigelny couldn’t demonstrate with existing molecular dynamics programs how the aS rings insinuated themselves into nerve cell membranes. Instead, he created his own modeling application called MAPAS (Membrane-Associated Protein Assessments). With MAPAS, he was able to identify which regions of the aS proteins were most likely to interact with cell membrane lipids. The program is now in high demand among other scientists studying membrane and protein interactions. Tsigelny himself is now applying MAPAS in his studies of the mechanisms that underlie kidney disease, cancer and other disorders.

Parkinson’s is one of about 16 diseases associated with abnormal protein clumping. These include Alzheimer’s, prion diseases (such as spongiform encephalopathy), type 2 diabetes and rheumatoid arthritis. Tsigelny’s molecular modeling findings and analysis techniques should help scientists better understand how proteins form plaques and how to block their formation.