NEW YORK (GenomeWeb) – A team led by researchers at Cellzome, a part of GlaxoSmithKline's R&D division, has developed a thermal profiling-based method for measuring protein-drug binding in cell membrane proteins.
Detailed in a paper published this week in Nature Methods, the approach is an expanded version of a thermal-profiling method the group published last year in Science, and expands the classes of proteins that can be studied using the technique.
The Nature Methods study also provided evidence that the method could prove useful for investigating not only protein-drug binding events, but also the downstream effects of these events, Gerard Drewes, Cellzome's head of science and senior author on the study, told GenomeWeb.
The basis for the method is the fact that, when bound to a ligand like a drug, proteins tend to have higher thermal stability. Taking advantage of this phenomenon, Drewes and his colleagues devised a process wherein they established thermal profiles for thousands of human proteins using mass spectrometry.
Upon denaturation, proteins clump together, which means they are not introduced into the mass spec and therefore are not detected. Using a Thermo Fisher Q Exactive instrument, the researchers built thermal profiles by tracking the decrease in the amount of a given protein detected, which corresponds to the increase in its denaturation, across the series of 10 temperature points.
They were then able to treat cells with drugs of interest and, by looking for proteins whose thermal profiles shifted in response to treatment, identify the proteins binding to the drug.
In the initial study, the researchers were not able to look at membrane proteins due to the challenges involved in extracting these molecules for measurement. Drewes noted at the time, though, that this was a goal of future research. Speaking this week, he said that the problem turned out to be easier than he and his colleagues had initially expected.
The concern was that the detergents required for extract membrane proteins would affect their thermal profiles. "But, actually, when we started working with detergent, we saw that if you use a mild detergent there is nothing to worry about, and the [profiles] are not altered very much," he said.
In the case of studies in intact cells, the detergent has essentially no effect, he said, because the ligand-binding and melting portions of the process occur before detergent is used to extract the membrane proteins.
This is not the case in cell lysate work. But, Drewes said, the effect of the detergent on protein thermal stability can be accounted for.
In general, he and his colleagues prefer to use intact cells, he said, but there are cases where cell lysates are useful. In particular, cell lysates can be helpful in that they allow researchers to separate the effects of the initial protein-ligand binding from subsequent downstream effects.
"One potential problem with doing the experiment in the intact cell is that you are also going to pick up changes in thermal stability that are downstream effects," Drewes said. "There is no downstream signaling going on in the lysate, so that is where the lysate experiment is still useful."
One way of dealing with this issue without using lysate experiments is to expose the cells to the ligands being studied for only a very short period of time before adding heat, allowing no time for downstream effects to take place, he said.
Drewes noted that the researchers are exploring whether the method might, in fact, prove useful for studying such downstream effects. In the Nature Methods paper, they looked at this question in Jurkat cells stimulated with pervanadate, finding that they could see "shifts in a lot of the known downstream effectors all the way down to the nuclear sectors that are known to be playing a role," he said.
"So we think this is a potential method to look at downstream signaling," he noted, adding, however, that sensitivity was still an issue, as was the question of identifying the specific nature of the downstream effects leading to changes in thermal stability.
"We believe, for instance, that phosphorylation may come quite often with a change in thermal stability, and maybe this is true for other modifications," he said. "But also there is obviously other stuff — proteins moving in and out of complexes, proteins maybe associating with DNA downstream, and all of this could play a role and change thermal stability. So, in the end it may be difficult to deconvolute."
The fact that the method uses mass spec could help deconvolute these processes, Drewes said, noting that he and his colleagues are currently working on correlating phosphoproteomics data with thermal stability data.
"With mass spec, you can look specifically at phosphopeptides and check if they have different melting properties, and you can measure those sites and their changes," he said. "You can also search for other modifications. For instance, we have seen that acetylation leads to changes in thermal stability. I think this is the power of mass spec, that in the same [thermal stability] data set you can also score post-translational modifications."
The Cellzome thermal stability profiling platform uses Thermo Fisher Scientific's Q Exactive mass spec and 10-plex TMT isobaric tags. This high level of multiplexing is key, Drewes said, in that it allows the researchers to collect enough data points in a single experiment to generate high-quality thermal profiles.
He said the researchers are also interested in ongoing efforts to expand the limits of multiplexing with isobaric tags. For instance, a 20-plex assay would allow the researchers to combine both their treated and untreated samples in a single experiment.
"At the moment we multiplex 10 temperatures in one experiment without the drug, and then we do the same experiment with the drug," Drewes said. "So, we have two multiplex experiments that we then need to cross correlate with each other. So higher multiplexing is something that would be an ideal fit with this method."
In September,Harvard University researchers led by Steve Gygi published on a new form of isobaric tags that could ultimately allow for multiplexing of up to 28 samples at once.
Drewes said he and his colleagues also aim to increase the method's sensitivity. Currently, he said, they can reproducibly quantify in the range of 5,000 to 7,500 proteins with the method.
"Often with this method we are looking to find new targets for a compound, so you want to be covering as many proteins as possible," he said.