NEW YORK (GenomeWeb) – University of California, San Diego researchers have used an enzymatic tagging approach to map in vivo tissue and subcellular-specific proteomes in Caenorhabditis elegans.
The study, published this week in Science Advances, used spatially restricted enzymatic tagging to identify and localize more than 3,000 proteins in C. elegans, and is the first example of the technique being applied to live multicellular organisms, said Aaron Reinke, a UCSD researcher and first author on the paper.
Spatially restricted enzymatic tagging uses proteins expressing a peroxidase to facilitate the biotin tagging of nearby proteins. In the presence of hydrogen peroxide and biotin-phenol, peroxidase will generate a phenolic radical that will modify surrounding proteins with biotin. These biotin-tagged proteins can then be pulled down using streptavidin beads and identified and quantified using mass spectrometry.
Because only proteins in close proximity to the peroxidase will be labeled, it is possible, assuming researchers know where the peroxidase is expressed, to localize the biotin-tagged proteins measured by mass spec.
In the Science Advances study, Reinke and his colleagues used the protein soybean ascorbate peroxidase (APX), which, they noted, has previously been used for spatially restricted enzymatic tagging to localize proteins in various compartments of human cells as well as to identify proteins present in the mitochondria in dissected fruit flies.
It had not previously been done in a living multicellular organism where researchers were then able to gather tissue and subcellular-localization information without dissection, Reinke said. "That is one of the main breakthroughs of [the study]," he said.
Reinke said he and his colleagues pursued the approach in C. elegans as this is the model organism they do much of their work in.
"We want to know where every protein is, both in individual cells and also in a subset of locations in the organism," he said, noting that, traditionally, researchers have pursued this sort of information by tagging individual proteins by fusing them with fluorescent agents that can be tracked via microscopy.
"And that works well with a single cellular organism," Reinke said. "You can do it on a pretty wide scale. But on an actual multicellular organism it can't be done yet. You can't make transgenic organisms that express every protein tagged so that you know where it is. It just hasn't been possible."
Using spatially restricted enzymatic tagging, the researchers were able to make different C. elegans strains expressing APX in different tissues and cellular compartments and then analyze the proteins tagged by APX due to their proximity.
The researchers wanted to look at proteins both in specific tissues and specific subcellular compartments within those tissues, Reinke said. To do this, they generated forms of APX with tissue-specific promoters as well as subcellular localization tags. To make sure the protein was actually being expressed where they thought, they tagged it using green fluorescent protein, "so we could visualize within the animal that the [APX] protein was actually localizing to the correct place."
Reinke said the approach could be used for a variety of purposes, such as investigating changes in protein localization due to post-translational modifications or outside stimuli like changes in growth conditions or stresses.
"For example, you could have worms growing in rich food and then under starvation conditions, and you could see if proteins are relocating between the cytoplasm and nucleus in response to these different conditions," he said.
He cited a study he and his colleagues published earlier this year in Nature Communications that made use of the technique to identify proteins secreted by the pathogen microsporidia when it infects C. elegans.
"We wanted to know what proteins that pathogen was using to interact with the C. elegans host," he said. "And that's difficult because this pathogen has its own proteins inside its own membrane, and then it's also going to have proteins that it is going to secrete into C. elegans. So, we were able to use this technology to identify the pathogen proteins that were secreted into C. elegans. And there was really no other way that you would have been able to do that."
Another recent application of the approach was published last month in Cell in a study by UC, San Francisco researchers who used it to identify interaction partners of G-protein coupled receptors, which have been largely intractable to conventional methods of protein-interaction analysis.
In that case, the researchers genetically inserted an ascorbic acid peroxidase (APEX) into their GPCRs of interest and then examined the proteins tagged by APEX to determine likely interactors. One of the authors on that paper was Stanford University professor Alice Ting, who led development of the spatially restricted enzymatic tagging method while at the Massachusetts Institute of Technology.
Reinke said that despite his team's success in applying the technique to C. elegans, there were certain limitations. One issue they ran into was getting good uptake by the organism of the biotin-phenol reagents required for the approach.
"That substrate is very easy to get into human cells, so [in those cells] you get a lot of signal," he said. "But in C. elegans it has historically been very difficult to get a small compound [into the organism.] So, we had to work for a long time to figure that out."
Overall, they were able to identify proteins to between 7 percent and 21 percent of the number of mRNAs present in a given tissue. While, as the authors noted, there is not thought to be strong correlation between protein and mRNA levels, this nonetheless suggests that many proteins are not being observed by the technique.
This is in large part due to the limits of mass spec sensitivity, Reinke suggested.
"I think from a practical standpoint, the limiting factor ends up being mass spectrometry," he said. "Even though mass spectrometry has gotten way more sensitive than in the past, I think we're labeling way more proteins than we're able to actually detect."