In the study of protein-protein interactions, there exists no greater goal than to develop a method for accurately mapping all the interactions in an organism. Industry promises aside, the field still has a ways to go. But in the June issue of Nature Biotechnology, researchers describe a technique for studying interactions that may at least address one of the challenges. Using two specific protein fragments, Stephen Michnick and his colleagues at the University of Montreal have found a way to study protein-protein interactions in vivo, while at the same time measuring the extent to which these interactions occur.
Michnick’s technique is based on a strategy called the protein fragment complementation assay (PCA), and involves a molecule, ß-lactamase, that can be split into two fragments and linked to proteins likely to interact with each other in an assay. If an interaction does occur, the two fragments of ß-lactamase are brought together, allowing them to re-fold. A colorimetric or fluorescent reporter molecule, such as cephalosporin nitrocefin, can then interact with the refolded ß-lactamase, causing its fluorescence to change from green to blue. Based on the intensity of the fluorescence, compared with the background blue, Michnick’s team can also gauge the number of interacting proteins.
Michnick and other scientists studying protein-protein interactions say the technique is straightforward experimentally compared to methods employing fluorescence resonance energy transfer (FRET), which often require matching the expression levels of interacting proteins. In addition, using ß-lactamase in a PCA experiment allows researchers to identify protein-protein interactions with a higher degree of confidence than with similar experiments using techniques such as the yeast two-hybrid system, Michnick said.
“The really key thing here, if you compare [the ß-lactamase PCA experiment] to FRET or yeast two-hybrid, is the process by which a signal has to be produced,” Michnick said. “You’ve got these fragments of the enzyme that have to refold exactly into the three-dimensional structure that is active, so you really have to bring them together. You get a stereo- and regiospecificity to the reaction that you don’t get from other techniques.”
Of course, a technique such as the ß-lactamase PCA experiment is not as amenable to fishing expeditions for unexpected protein-protein interactions. In those cases, a researcher might resort to yeast two-hybrid, or affinity pull-down techniques similar to those used by scientists at MDS Proteomics and Cellzome in their back-to-back studies in Nature published earlier this year. But Michnick maintained that his method might offer additional insights because it can be used to study interactions in vivo. In addition, studying all the members of a pathway can result in some surprises. “The funny thing [was] that we found even in our preliminary studies [of biochemical pathways that] we turned up all sorts of connections that no one had ever seen,” he said.
Other researchers agreed that the technique’s ability to study protein-protein interactions in vivo is one of its strong points. “In this case you can actually look in the microscope, and see that the interaction is actually happening in this cell, but not in that cell,” said Bruce Mayer of the University of Connecticut Health Center in Farmington, Conn., who uses in vitro co-immunoprecipitation techniques to study protein-protein interactions. “Even better than that, you can treat [a cell] with a growth factor or stimulus, and [if] you see the interaction, now you can actually learn something about signaling pathways. Where they’re going is a way to validate interactions involved in signaling pathways,” he said.
That’s not to say there isn’t room for improvement, however. Mayer said he suspected the resolution of the detection technique is “not very good,” and that as it now stands it wouldn’t allow a researcher to pinpoint with accuracy where in time and space an interaction occurs in the cell. Currently the method has signal-to-background ratios ranging between 10:1 and 250:1. Randall Mrsny, who studies membrane proteins at the University of Cardiff in Wales, added that setting up the reporter system requires a substantial amount of genetic and protein engineering. “It’s still a lot of work,” he said.
But Michnick already has a potential commercial user in the form of Odyssey Pharmaceuticals, a San Ramon, Calif.-based drug discovery company. Michnick, a co-founder who serves as the chair of the company’s scientific advisory board, said Odyssey has already licensed the technique for commercial applications, and plans to use it to study changes in signaling pathways associated with cells’ response to drugs. — JSM