Researchers at the National Cancer Institute's Protein Expression Laboratory have come up with a new, high-throughput method for producing purified proteins from cDNAs containing open reading frames.
With the new method, called pooled ORF expression technology, or POET, researchers insert a mixture of hundreds of different kinds of cDNAs into Escheria coli at one time. After proteins are expressed within the mixture, the mix is deconvoluted by running proteins out on a 2D gel, according to William Gillette, one of the developers of the technique who is the lead author on a paper describing the method that is in currently in press at Molecular & Cellular Proteomics.
"What we're proposing is an alternative, or a complementary technique to the normal technique in the structural biology world, where they subclone cDNAs one at a time, express them one at a time, and purify them one at a time," said Gillette. "You can save a significant amount of time and labor."
The approach is especially useful if researchers are dealing with an ORFeome that is difficult to express, said Jim Hartley, the director of the Protein Expression Laboratory at the NCI.
"POET is an attempt to see only the proteins that succeed, and not the ones that fail," explained Hartley.
"What we're proposing is an alternative, or a complementary technique to the normal technique in the structural biology world, where they subclone cDNAs one at a time, express them one at a time, and purify them one at a time. You can save a significant amount of time and labor."
For example, previous studies have shown that the success rate of converting Caenorhabditis elegans ORFs to purified proteins is only two percent. With POET, the two percent of successful proteins would be found and purified much faster than with traditional methods.
"Instead of subcloning one-by-one 10,000 times and succeeding [in getting a purified protein] only a couple of hundred times, POET is a way to not go through all those failures," said Hartley. "By pooling up front, then subcloning in one big pool, expressing in one big pool, and purifying in one big pool, you can do hundreds of genes, or ORFs, at the same time. After you figure out which proteins are the ones best expressed, you can deconvolute by running the proteins out on a 2D gel."
The POET method is compatible with projects like the Protein Structure Initiative, where scientists are trying to determine the three-dimensional shape of every protein representative of a unique class (see ProteoMonitor 2/18/2005), or the US Department of Energy's GTL project, where scientists are trying to make every protein from every gene of a number of organisms, Hartley noted.
"With both of these big projects, when you start to make proteins from these genes, it's a very wasteful process. The further you get away from E. coli, the lower the hit rate," said Hartley. "The question with POET is, 'Is the work downstream to deconvolute the pool of purified protein worth the effort?'"
The POET method could help to produce proteins that are challenging to express because failures from one experiment can be pooled into a second experiment that is run under different conditions.
"You can go back to the ones that failed, make a new pool, and try a different condition on those, whether it is a different fusion partner, a different incubation condition, or a different detergent for extracting membrane proteins," said Gillette.
Hartley said the POET method has been patented, but it is doubtful that the method will ever be commercialized because the materials for doing POET experiments are so readily available.
While POET kits are not really an option, it would be interesting to commercialize pools of cDNA ORFs, Hartley added.
"If DNAs are available, you could order up your own pools, for example pools of all the G protein-coupled receptors, all the secreted proteins, all the nuclear proteins," he said. "It might be intriguing. … A single pool is certainly biased and non-physiological, but in some sense, it could conceivably represent the metabolome of the human cell. That's a stretch, maybe, but there might be something you could do with such a pool. For example, would it represent the protein-protein interactions you might be interested in?"
As their first proof-of-principle for their new technique, Gillette and his team used the C. elegans ORFeome to produce purified proteins. They pooled together 688 C. elegans ORFs, expressed them in E. coli, deconvoluted the proteins into about 168 spots, and ended up purifying 12 positive clones identified by POET.
As a next step, Gillette said he and his research team hope to perform the same technique on pooled human ORFs.
So far, the team has pooled together six pools of about 540 genes to be expressed, purified, and then deconvoluted by POET, Hartley said.
The researchers are starting out with ORFs that encode proteins not present in the Protein Data Bank, because those are more likely to be proteins that have not been studied, Hartley said.
The new POET method could also conceivably help in making large-scale antibody projects like the HUPO Human Antibody Initiative more high throughput by generating protein antigens from which to make antibodies at a faster rate.
"It depends on how much protein you need," said Hartley. "Can you immunize a mouse with a spot from a 2D gel? I don't know. I guess it's conceivable."
In general, POET is most useful when the protein expression success rate is low, Hartley said.
"The more often you win, the less utility POET has," he said. "If it's not so hard to express, why would you go through all the pooling effort, and all the deconvoluting? You might just as well accept losing half of the time, depending on how much access you have to robots, et cetera."
— Tien-Shun Lee ([email protected])