The side effects of a drug can often be traced to its interactions with proteins other than its intended target. Michael Fitzgerald, an associate professor at Duke University, and his colleagues developed a method for getting a proteome-wide look at what proteins a drug interacts with. By coupling a ligand-binding assay to mass spectrometry, their proof-of-principle study, published online in PNAS in May, was able to show to which proteins the immunosuppressant drug cyclosporin A binds. By doing such studies early in the drug development process, the researchers say they could keep an eye out for potential side effects before the drug even reaches patients.
"Using this methodology, we think that if you're applying it to newer drugs, you get information up front. And you can start saying, 'Hmm, this drug interacts with a lot of proteins involved with glycolysis, I wonder if this is going to be a problem for diabetics?'" Fitzgerald says. "It's really a discovery tool that will spearhead, hopefully, new investigations that could test drugs more efficiently up front."
The scientists' first step was to develop the ligand-binding assay part of their approach, where they examined the thermodynamic stabilities of the protein, both in the absence and in the presence of the drug. To be able to gauge that stability, they specifically induce methionine oxidation in the sample, as the oxidation rate is related to the thermodynamic stability of the protein — thereby acting as a label. "Fundamentally, that's what we're getting this measure of the stability, the folding free energy of a protein and practically that can give you information about function, and compare the thermodynamic stability of that specific protein that you made this measurement on in presence or absence of the ligand," Fitzgerald says. "Proteins that got more stable presumably are now interacting directly with the drug, or maybe indirectly through another network of proteins, but whatever the case, there's some new interaction that was promoted or induced by the drug." Coupling this to mass spec gives readouts of the proteins whose stability changes.
To test this, Fitzgerald turned to cyclosporin A. This drug, he says, is linked to a propensity for developing diabetes — about 30 percent of organ transplant patients develop the disease — and some of its targets are well-known. With their approach, Fitzgerald and his colleagues were able to detect those known targets of cyclosporin A, as well as a handful of others. Some of those new ones are involved in glycolysis, and Fitzgerald says his team is following up on them.