Building on technology that failed the first time around, Protein Forest is ramping up to launch its first commercial product, a chip allowing researchers to separate proteins based on their isoelectric point.
Headquartered in Waltham, Mass., Protein Forest was launched in 2002, and like many startup technology companies, found flaws in its original product design. So it went back to the drawing board, did further development, and after five years, the company is now beta testing the chip in anticipation of its launch onto the commercial market early next year.
Protein Forest received US patent number 7,166,202 titled “Matrixes, arrays, systems and methods” for the chip in January and has two other patents pending.
The chip is based on the company’s digital ProteomeChip technology, which relies on well-established isoelectric principles. The advantage of the dPC technology, the company says, is significantly quicker separation time and ease of use.
Protein Forest began developing the chip in 2003 after receiving its first round of venture capital funding of $12 million. But after a year in development, it became clear that there would not be a market for it. The problem was that proteins could not be identified with the chip, said Russell Garlick, currently the chief technology officer of the company and Protein Forest’s original president and CEO.
“Those were just spots on an image; you really couldn’t identify them,” he said. “We never really did quantitative analysis or characterization [of the proteins], so we never knew how many proteins there were.”
Because of its size, 1 centimeter by 1 centimeter, the chip was too small to produce enough proteins for protein identification, and so couldn’t be used for mass spectrometry sample preparation, he said.
Manufacturing the chip also turned out to be more complex than originally imagined due to its size as well as questions about what materials to use for the chip. “It just never proved to be a product that could meet the manufacturing cost and reproducibility [criteria] that the company set,” Garlick said.
The setback with the original chip delayed the company’s timeline to bring it to market by about two years, Garlick estimated. The key lesson learned, he said: “It’s best to launch the right product into the right market the first time.”
Isolating Proteins by Isoelectric Points
According to its patent abstract, the chip in its current iteration separates proteins in one dimension, based on the protein’s isoelectric point, in an electrical field. The chip has 40 holes, or gel plugs, each filled with a chemical controlling its pH level. Each gel plug has a different pH, ranging from 4 to 8, though ongoing development will bring the upper pH range to 10 in the next few months, Garlick said.
The chip is situated inside a chamber. On one side of the chip is an acidic buffer, the anode, and on the other side, a basic buffer, the cathode. Each buffer is filled with about 1 milliliter of liquid.
By adding a sample of 1 microgram to 2 milligrams and “mixing the samples in one or the other chamber, the proteins are then charged based on the pH, and if they’re in the cathode, or basic, side, the proteins assume a negative charge,” Garlick said. “And when we turn the power on, the proteins will be driven by the electrical field to the anode side of the chip, so they’ll come to the chip and they’ll hit these pH features or gel plugs.”
Because each protein has a unique isoelectric point, a protein will get trapped in the gel plug with a pH level at or near its isoelectric point. With 40 different plugs, each with a different pH level, the separation is done via parallel isoelectric focusing, Garlick said.
After the proteins are separated, they are stabilized in the gel plugs with a solution. The gel plugs can then be placed into a microwell. In-gel trypsin digestion is performed to create peptide fragments, and the peptides are identified through mass spec analysis.
The ProteomeChip is different from other separation chips, such as those offered by Bio-Rad Laboratories and Agilent Technologies, because it utilizes isoelectric focusing technology to separate proteins rather than separating them based on size or SDS-PAGE [sodium dodecyl sulfate polyacrylamide gel electrophoresis], Garlick said.
Small changes in the isoelectric charges of proteins correspond to small changes in molecular weight, but while the dPC chip can detect and separate proteins based on tiny charge differences, other chips are unable to do separation based on such miniscule molecular weight differences, Garlick said.
Getting Back on that Horse
As Protein Forest now sits in anticipation of its first product launch, Garlick cited two reasons for the company’s continuing rebound from the failure of its first chip — a focus on markets that use defined protein detection methods such as mass spec and immuno Western blotting and “improved chip chemistry that enabled resolution of highly complex protein samples under denaturing conditions,” for example urea and detergents.
“It’s best to launch the right product into the right market the first time.”
With its patent in hand, the next step is independent testing of the chip to see how it performs. Earlier this month, beta testing of the chip began at the University of Massachusetts Medical School.At this stage, the goal is to see whether the technology can do what Protein Forest says it can: detect shifts in isoelectric points in proteins.
“pI shift is indicative of modifications of the proteins such as phosphorylation, glycosylation, you name it, so this might be able to give us that very precise measurement,” said Sunny Tam, a research associate professor at the medical school who is leading the beta testing team there. “An important point is that this gel plug is defined enough and is in such a small territory that you can really capture a specific type of protein or even peptide.”
In the coming months, Tam said, he and his team will develop applications to see whether the chip can be used for real-life research. “I have high hopes in terms of different areas looking into signal transduction, process development, and characterization of biomarkers,” he said.
During the summer, testing will be expanded to 10 additional laboratories, Garlick said, focusing on two applications, mass spec fractionation and protein isoform Western blot analysis.
Meanwhile, Protein Forest is hoping to raise $10 million in venture capital funding later this year to commercialize the chip. It would represent the third round of VC funding in the company’s history, following $12 million in Series A funding four years ago, and $4 million raised last year.
Protein Forest has nine full-time employees currently.
As it prepares to launch the chip commercially, Protein Forest is working on developing technology to transfer the proteins from on-the-chip to a second dimension, and an electroeluter for the extraction of protein samples from electrophoresis gels, Garlick said.
Protein Forest was also recently named to the Human Proteome Organization’s Industry Advisory Board to develop standards and improve proteomic technologies [See PM 04/26/07].
“We want to bring reproducibility, speed, and throughput [to proteomics]. HUPO is looking for sophisticated tools for quantitative proteomics and that’s what we have,” Garlick said. “At the end of the day, our goal with our product is to industrialize proteomics.”