Skip to main content
Premium Trial:

Request an Annual Quote

UBC Team Develops High-Throughput Method for Mapping Protein Interactions


Scientists at the University of British Columbia have developed a high-throughput method for measuring protein-protein interactions.

The technique, detailed in a paper in the current edition of Nature Methods, uses a combination of size-exclusion chromatography and SILAC mass spec to detect protein complexes and is significantly faster than commonly used approaches like yeast two-hybrid systems or affinity purification mass spec, Leonard Foster, a UBC researcher and author on the paper, told ProteoMonitor.

The method's speed and use of SILAC labeling, he noted, could prove particularly advantageous for monitoring changes in protein-protein interactions across various biological conditions – a task for which conventional approaches are poorly suited due to their labor-intensiveness.

The technique, Foster said, "relies on the concept that proteins in a complex will co-elute," making it possible to identify and measure protein complexes by quantitating co-eluting proteins.

The researchers used high-resolution size-exclusion chromatography to collect a series of SILAC-labeled sample fractions separated by analyte size. They then analyzed each fraction using LC-MS/MS on a Thermo Fisher Scientific LTQ-Orbitrap XL instrument.

In that way, they developed chromatograms of the levels of "each protein across all the different size exclusion fractions," Foster said. "And where you have chromatograms that look very similar – either across the whole chromatogram or in a region of the chromatogram – then those proteins are likely to be co-eluting."

"Out of that we generate an interaction matrix where we know that proteins that have a very tight co-elution pattern are interacting," he said. "So we generate a matrix of binary interactions, and then out of that [identify] which proteins are in the same complexes together to get a higher order of interaction."

The researchers then used previously validated protein interactions taken from the Corum database of protein complexes to set cut-offs for identifying interactions from their data.

To validate their method, they compared their results to previously reported protein interactions. They also performed an experimental validation in which they spiked into the sample an antibody for a protein identified as belonging to a certain complex, causing a change in that protein's size and, consequently, a shift in its elution pattern. As Foster noted, proteins complexed with that target protein should also see shifts in their elution pattern, allowing the researchers to test the accuracy of their identifications.

Using the technique, the UBC team analyzed HeLa cells in triplicate, identifying 3,400 proteins, which they mapped into 291 co-eluting complexes. They also tested the response of these proteins upon stimulating the cells with epidermal growth factor, identifying 351 proteins whose interactions with a complex were altered by EGF, including a number of proteins known to play roles in EGF signaling.

The most significant advantage of the UBC approach, Foster said, is its speed. He noted that he had previously been involved in an interactome study using affinity purification mass spec on 400 bait proteins that required "many, many person years" to complete. With the method presented in the Nature Methods paper, on the other hand, one person could complete a comparable experiment "in about two weeks' time," he said.

The other major advantage, Foster said, "is that we avoid having to tag the proteins, which is basically a requirement for either yeast two-hybrid or affinity purification [approaches], and so we avoid any artifacts that the tagging itself might introduce."

Foster noted that the technique had been made possible in significant part by recent improvements in size-exclusion chromatography. "Up until about a year ago there was no high enough resolution size-exclusion column that would be useful for the size range most of the protein complexes would fall in," he said. "So there had been a bit of a technical limitation there."

Size exclusion, Foster said, was a desirable approach because it allows for much gentler buffer conditions than chromatographic techniques like cation or anion exchange. "You can resolve protein complexes with pretty high resolution using [those techniques], but as you elute under those conditions, you start having to use higher salt or pH and that often is not convenient for retaining as many interactions as you want."

Foster said that while the approach was not as sensitive as yeast two-hybrid methods, which can pick up very fleeting interactions, it could detect more transient interactions than affinity-purification based techniques.

"The processing of the samples takes far less time, and so it's basically just a kinetic argument – the less time the complex has to dissociate, the more lower-affinity interactions you'll be able to pick up," he said.

Moving forward, Foster said, he plans to use the method to study changes in protein complexes in response to bacterial or viral infections as well as changes in complexes throughout the cell cycle.

In particular, he said, he hopes to use it to study complexes involving membrane proteins – something that he noted, "both yeast two-hybrid and affinity purification mass spec don't do very well."

"As long as detergents are compatible with the [new] system – and some seem to be – then we should be able to use it to profile membrane complexes," he said.