NEW YORK (GenomeWeb) – Researchers at the University of California, San Francisco have developed a mass spec-based approach to studying protein-protein interactions that can observe such interactions in vivo with high spatial and temporal resolution.
In a study published last week in Cell, they used the method to study the interactome of G-protein coupled receptors, which, said Nevan Krogan, professor of cellular and molecular pharmacology at UCSF and senior author on the paper, have been largely intractable to conventional methods of protein-interaction analysis.
The approach combines proximity labeling mediated by engineered ascorbic acid peroxidase (APEX) with quantitative mass spectrometry to identify protein interactors.
Researchers genetically insert the APEX tag into their protein of interest. Upon stimulation with hydrogen peroxide, this tag releases biotin-phenoxyl radicals that tag nearby proteins in the cell, and these tagged proteins can then be pulled out of the sample using streptavidin-based enrichment and analyzed using mass spec. Because the APEX tagging process takes place very rapidly, it can be used to monitor protein changes occurring on the order of seconds.
The approach has proved particularly useful for applications like studying the protein components of organelles, Krogan said. However, because the APEX system tags whatever proteins are nearby, it has generally been considered poorly suited to protein-protein interaction work.
"The thought in the field has really been that this can't be used for protein-protein interactions because of the high degree of background," Krogan said. However, by using mass spec measurements of proteins of known location, he and his colleagues were able to account for this background, making it possible to distinguish between true interactors and proteins merely in close proximity to the analyte being investigated.
"What we did was identify spatial reference [proteins], and then we used those to normalize away the background that we saw," he said. The idea, essentially, is that if APEX tagging of surrounding proteins is proximity-dependent, then it should be possible to assess the location of unknown proteins by comparing the extent to which they are tagged to that of proteins of known location.
This, Krogan noted, requires a priori knowledge of the location of these reference proteins, though he said that he and his colleagues were also exploring using the technique in reverse to localize target proteins.
"If you had a reference protein for each sub-compartment and then you carried out a mass spec experiment, you could line that up and say, '[The target protein] is here 50 percent of the time and in this other sub-compartment 50 percent of the time,'" he said. "So, if you think of it that way, you could actually use this approach to localize your [target] protein in a very high resolution manner."
The method is also based on the assumption that APEX labeling is strongly proximity dependent, but Krogan noted there could be other factors involved in why a particular protein is more or less strongly labeled than another.
"On paper it is supposed to be just proximity, but in practice there are more things going on," he said. "So, we need to investigate many different types of proteins using this approach to get a handle on that exact question."
Based on the Cell work, though, the technique appears effective at identifying interaction partners of GPCRs, Krogan said, which, in addition to being difficult to analyze via conventional methods, are major drug targets.
Using their approach, the researchers analyzed the interactome of the beta-2 adrenergic receptor, a well-studied GPCR whose interaction partners have been fairly thoroughly characterized. Their analysis found that they could identify both direct and indirect interactors of the receptor and with sufficient resolution to observe agonist-dependent interactions.
They also used the approach to study interactors of the GPCR DOR, identifying known interactors as well as novel interactors that appear to play a role in sorting DOR to the lysosome, which, they noted, offers "insight into the molecular mechanism of opioid receptor downregulation."
While development of the APEX-mass spec approach is still in its early stages, Krogan said it offers a look at GPCR function, and potentially that of other proteins, that was previously not possible.
Currently, Krogan and his colleagues are using the method for research into psychiatric drugs and disorders as part of the Psychiatric Cell Map initiative he is heading.
Using the approach to look at, for instance, dopamine or serotonin receptors, "we can now study antidepressants in a way that nobody could before at the molecular level," he said. " We can study, for instance, how the effects of different mutations in these GPCRs result in different psychiatric disorders. So, in terms of studying these kind of neuropsychiatric disorders, this is a fantastic new tool."
The UCSF researchers are also using the approach to study the progression of infectious diseases like tuberculosis and HIV," he said. "There are key proteins in the host that we know are being highjacked in the course of an infection, and we want to get this high temporal and spatial resolution information to see how these [protein] interactions change in the context of infectious diseases."
There are limitations to the method — notably its tendency to detect more transient interactions, due, perhaps, to the fact that in the case of more stable interactions the protein sites APEX would label are already bound.
Krogan cited the example of work he and his colleagues did on a large chromatin complex. "We got a lot of transient interactions, which turned out to be very exciting, but some of the stable stoichiometric ones that we know are in close proximity, they were not there," he said.
"This is not going to usurp traditional affinity purification-MS methods," Krogan said. "But it is another tool in the toolbox that can be used especially for cases where proteins are moving around the cell, and especially where you are getting changes in interactions at a sub-minute level."