With $19 million of Series A venture capital in its pocket as of last week, and a former Perkin-Elmer executive at its head, separations start-up Protein Forest is hoping to replace the use of 2D gels in pharma’s and biotech’s proteomics core laboratories with its miniaturized ProteomeChip, set for launch in Q4 2004. The Watertown, Mass.-based company closed the $19 million series last week, with four firms — S.R. One, Boston Millennia Partners, IDG Ventures, and Novo A/S — leading the charge.
Large-scale pharma and biotech proteomics labs “fit perfectly into our marketing plans because it gives them similar technology to what they already know — how to separate by size and charge — but miniaturized, and faster, much faster,” Russell Garlick, CEO of Protein Forest, told ProteoMonitor. Garlick, who left PerkinElmer in March 2002, has been CEO since Protein Forest co-founder PureTech Ventures — for which Garlick consulted on Protein Forest’s technology — got involved with the project in June 2002. The three Israeli physicists who developed the ProteomeChip technology — Shmuel Bukshpan, Uri Halavee, and Gleb Zilberstein — first started up the company in February 2002. Protein Forest is the seventh company that the trio has founded; previous start-ups include Rehovot, Israel-based drug discovery company Proteologics, and San Antonio, Texas-based IT company X Technologies. Garlick said that because Protein Forest has “very strong investors right now,” the company can comfortably move to the next stage without having to raise more venture capital. “The next step for us is to invest in product development — meaning the engineering of the product, commercializing the product, getting beta testing going — all the good things you do to launch a product in the life sciences,” he said. “This [money] allows us to do this.”
Protein Forest plans to manufacture and market two different chips: a “research chip,” and a “mass spectrometry chip.” Both have essentially the same design, but the research chip works with sample masses as little as 10 ng, while the mass spec chip works with 100 to 500 µg — scales at which researchers can pick spots from the gel and the mass spec can identify proteins. The chip is similar to a 2D gel, but uses an array of pHs for first dimension solution-phase separation instead of the usual linear IPG strip used in traditional gels. Each pH unit is separated into 50 discrete spots on a chip “half the size of a microscope slide,” with each spot leading to an acrylamide gel lane so that for each pH unit, 50 lanes run in parallel. The chip sits in about 1mL of buffer. One run takes about 20 minutes for the research chip and 1 hour for the mass spec chip. This format “allowed us to not only do solution phase kinetics, which is much faster, but it’s allowed us to miniaturize it. By doing both, we’re able to get down to input mass of proteins in the sample in the low nanograms,” Garlick said.
The selling point of the technology is clearly the sensitivity — Garlick cited an experiment, for example, in which scientists were able to use the system to identify several proteins from a single red blood cell using the research chip. But there are limitations. Because the research chip deals with quantities too small to feed into a mass spec, the only way of detecting the proteins is “using the same methods everyone else does,” — that is, fluorescent and silver labeling and traditional imaging technology. In this way, the dynamic range problems common to other 2D gels are not solved, and mass spec-based protein identification of these tiny samples is still not possible. “Our restrictions are the same as anyone else using these detection systems,” Garlick acknowledged. Still, he said that the precision of the chip made it a preferable method compared with liquid chromatography. “We have precision down to .02 units. You don’t get that precision with liquid chromatography,” he said.
Garlick is also jumping on the low abundance protein bandwagon with ProteomeChip. “If you look at the old method [2D gels], it’s slow, tedious, and it misses the important proteins,” Garlick said. “You’re unable to detect low abundance proteins, and low abundance proteins are the important drug targets, the important regulatory proteins.” In the market for concentrating low abundance proteins, Protein Forest will compete with Invitrogen’s Zoom system (see story, p. 4). Although Garlick would not comment on how ProteomeChip measures up to Zoom, he did say that the company would compare the two systems during beta testing.
Customers will have to wait until this time next year to see whether Protein Forest’s ProteomeChip takes root in the market.