Yale researchers have devised a technique for expressing activated phosphoproteins in Escherichia coli, using it to generate phosphorylated versions of the protein MEK1, a key molecule in cell proliferation, development, differentiation, and oncogenesis.
According to Yale assistant professor Jesse Rinehart, one of the leaders of the study – which was published in the August edition of Science – the technique could prove an important source of reagents for phosphoproteomic research and provides an early glimpse of synthetic biology's potential to aid proteomics more generally.
Phosphoproteins are a significant area of proteomics research, but generating activated phosphoproteins for study is difficult due to the fact that phosphorylation is a post-translational modification not encoded in the genetic code.
To get around this problem, the Yale team used an O-phosphoserine-tRNA synthetase found in methanogenic archaea to incorporate phosphoserine, Sep, into proteins directed by the UAG codon, allowing them to insert Sep residues in a precise and controlled manner.
"Phosphorylation is obviously a big [area of research]," Rinehart told ProteoMonitor. "But here's the problem: Labs have published hundreds and thousands of phosphorylation sites, but we are decades away from understanding how cells accomplish [protein phosphorylation]. We don't really understand, I would argue, what the kinases are that create most of the phosphorylated proteins in a human cell or organ."
This is problematic, Rinehart said, because researchers interested in studying phosphoproteins typically need some understanding of how they are created in vivo. The Sep insertion method, he said, is a tool to help scientists "bypass all these gaps" in knowledge.
"My group is interested in a number of kinases and some of their substrates, and we have [proteomic] projects where we may find interesting phosphoproteins that are changing dynamically and we have no idea how that is achieved," he said. "Now, we can go ahead and synthesize these proteins and do next-step-type research on how they function. Maybe ten years down the road we'll understand how [their activation] happens in the cell, but it's no longer a roadblock" to further research.
In the past researchers have devised other methods for generating phosphoproteins, but, Rinehart said, "they are largely inadequate." For instance, New England Biolabs, which was a collaborator on the paper, has a technology for attaching pieces of a phosphopeptide to proteins, but, Rinehart said, it doesn't offer the precision of the genetic method. Likewise, efforts to generate phosphoproteins by mixing proteins with cocktails of kinases in vitro suffered from a lack of control and specificity.
"You would take your protein of interest and you would mix it with a cocktail of kinases and hope that one of them would replicate or partially replicate the natural modification," Rinehard said. "But if you turn one of the kinases loose on a substrate, it may phosphorylate the site you're interested in, but it may also phosphorylate ten other sites that are completely irrelevant."
"So that's an important aspect of our technology, that it's site specific," he added. "We can say, 'Put one here. Put two in these places, and nowhere else.'"
To ensure the Sep insertions were occurring where the researchers intended, Rinehart's team developed multiple-reaction monitoring mass spec assays on an AB Sciex QTRAP 5500 to verify the phosphoproteins they produced. As they generate additional phosphoproteins with the technique, they are building a database of MRM assays to be used for this purpose.
"We're using MRM almost exclusively with this technology," Rinehart said. "We've been doing MRM to study phosphoproteins for quite some time now, so for us it's really a go-to technique and integrating it into this workflow has really been transformative."
"We just plug [the proteins] into our standard workflow, pull out the peptides, find the [phosphorylation] site, and confirm it, so it's very easy," he said. "The only other alternative [to mass spec] is an antibody approach, and antibodies can take months if not years to develop. We can take phosphoproteins and characterize them in one week" using mass spec.
The technique described in the Science paper was specific to Sep, which is the most common phosphoamino acid, showing 7.3 and 48 times higher relative abundance than phosphothreonine and phosphotyrosine, respectively, in an analysis of HeLa cells. The researchers are also pursuing similar techniques for encoding for phosphothreonine and phosphotyrosine, Rinehart said.
Thus far, he noted, his lab has generated between 20 and 30 phosphoproteins using the technique, and, he said, it should be a broadly applicable to any proteins incorporating Sep.
"As long as you can clone your protein of interest and put it in E. coli you're off and running," he said. "We're synthesizing phosphoproteins easily and of great diversity."
Rinehart declined to say specifically what additional phosphoproteins his lab had synthesized, but said they were molecules linked to hypertension, cancer, and diabetes. His lab has patented the technique, but, he said, it was too early to discuss any potential strategies for commercialization of the method.
Beyond the specific instance of Sep, the research demonstrates the broader potential of synthetic biology in proteomics, Rinehart said.
"It's important to understand that this is a general approach to incorporate other types of amino acids," he said. "So I can imagine that when our lab and other labs get better at doing this we might see novel amino acids incorporated into proteins that might be very beneficial to proteomics. You can imagine labeling or bringing in modified amino acids that are relevant to protein stability and detection and enzymatic activity. So there's a greater field and strategy here."
Rinehart said he thought that currently proteomics researchers were by and large unaware of the potential that synthetic biology holds for their field, noting that he expects there to be "a real positive growth trend in how synthetic biology can really feed back and help proteomics efforts."
Have topics you'd like to see covered in ProteoMonitor? Contact the editor at abonislawski [at] genomeweb [.] com.