28 May 2013

A new study explains why many drugs come with similar side effects—nausea, vomiting, headaches—bad enough to halt development in its tracks. Researchers from the Georgia Institute of Technology, Atlanta, predict that proteins form fewer than 500 different binding structures, or “pockets.” Published May 20 in the Proceedings of the National Academy of Science online, the results hint that many proteins share the same pocket, and drugs targeting one protein will inevitably hit several, said first author Jeffrey Skolnick. “If there are only a small number of pockets, there is no way to avoid side effects,” he told Alzforum.

Skolnick and coauthor Mu Gao constructed a library of artificial proteins (ART) and compared it to proteins in the Protein Data Bank, a library of real protein structures. They found that each protein pocket in the native collection had a corresponding one in ART, suggesting the artificial library, containing only 500 binding sites, completely represented structures found in the real world of proteins. If a native and an artificial set of pockets overlap significantly, then the artificial set is probably complete, Skolnick explained. If that is the case, and the number of binding sites is truly only 500, then finding drugs that specifically bind only one target out of the 30,000-plus human proteins may be hard to do. “This supports a change in the philosophy from ‘one molecule, one target’ to ‘one molecule, many targets,’” said Skolnick. That concept is gaining traction among scientists studying the varied actions of drugs, he said. However, other researchers working on protein structure believe there are many more than 500 unique sites among the human proteome.

Even if many proteins share the same pocket shape, ligands bind them with differing affinity, depending on the specific amino acid sequence, charge, and hydrophobicity, Skolnick said. That means a drug would bind strongly to some sites and weakly to others, creating background noise or "dirty drugs" while allowing for some degree of specificity.

A finite number of protein pockets makes sense in light of the limited number of proteins in the human system, said Nikolay Dokholyan, University of North Carolina at Chapel Hill. However, the study does not put a damper on drug development, he said. Scientists can exploit the property of differential binding strengths to selectively design compounds. For instance, they can use databases that group similarly shaped pockets, such as the Pocketome currently in development by Ruben Abagyan and colleagues at the UCSD Skaggs School of Pharmacy and Pharmaceutical Sciences, La Jolla, California, to anticipate which important off-target proteins a drug might hit and tweak the drug to diminish those interactions, he said.

One camp of researchers is actually exploring the benefits of “polypharmacology,” the concept of treating a disease with drugs that hit multiple targets. They predict that understanding these varied interactions may help scientists find compounds that act on multiple targets in the same disease pathway. Polypharmacology may also streamline the repurposing of drugs and predict adverse side effects early on in drug development (for a review, see Reddy and Zhang, 2013).—Gwyneth Dickey Zakaib