After helping start several companies while at the Stanford Research Institute, Luke Schneider finally decided five years ago that it was time to go out and do his own thing. He figured proteomics would be his best bet, not only because it was fairly new, but also because he had first-hand experience with the frustrations of running 2D gels and realized there was room for a company offering an alternative.
As a chemical engineer with a PhD from Princeton, Schneider started working on a protein separation and identification system that employs capillary electrophoresis (CE) and mass spectrometry. When his ex-wife introduced him to Jeffrey Peterson, who had worked in various capacities for General Electric and Abbott Laboratories and was looking to “do something entrepreneurial from the ground up,” the two founded Target Discovery in April of 1999. As the company’s name implies, Target Discovery’s long-term goal is to discover drug targets from differences in protein expression patterns.
Seeking More Money
Two months after founding Target, Peterson and Schneider raised $2.5 million in startup funding from angel investors, enabling the company to hire staff and start developing its platform technology. Last December, the company moved into 4,000 square feet of space in Palo Alto, Calif., and now has a staff of ten. Over the past year, it has raised $2 million in a second private financing round, and hopes to attract at least an additional $7 million from institutional investors this year, Peterson said.
Using CE instead of 2D gels to separate proteins has several advantages, he claimed. In theory, the resolution capacity for three dimensions of CE is over 30,000 proteins, compared to about 7,000 for gels, he said. Target uses isoelectric focusing, size separation, and capillary zone electrophoresis — which separates proteins according to their mass/charge ratio — as its three dimensions. Unlike 2D gels, highly charged proteins are not lost in the isoelectric focusing step, and since CE can be run in an organic phase, it is also amenable to hydrophobic proteins.
To detect and identify the denatured proteins that pass through the capillaries, Target tags each of them with a label containing a fluorophore and an undisclosed component that takes advantage of “an insight from subnuclear physics” that is different from an isotopic mass label, Peterson said. After the separation, the researchers record protein peaks with a laser-induced fluorescence detector that others have shown to be sensitive enough to detect a single protein copy, he added. With regard to run-to-run variability, the company hopes to achieve under one percent deviation within the next few months.
To identify the proteins, Target uses “inverted mass ladder sequencing,” a mass spectrometry technique that sequences proteins by comparing fragments of increasing mass, starting with the label. The patent-pending method uses electrospray time-of-flight mass spectrometry, fragmenting the intact protein as it enters the instrument in a process similar to collision-induced dissociation.
Data collection in the mass spec currently takes only 90 seconds, Peterson pointed out, and he expects to make this faster. Using the unique mass and “subnuclear characteristic” of the label, Target researchers have developed an algorithm for simplifying the mass spectrum, allowing them to identify the terminal five to seven amino acids. The company is expecting its first patent on the technology within the next few weeks.
Onward to Humans
So far Target Discovery has used its platform to separate “hundreds of proteins” from both E. coli and wheat germ and is about to test it on human plasma. The company is also currently preparing a platform technology paper for a peer-reviewed journal that will be submitted as soon as the first patent is issued. Nine other patents covering various aspects of the protein sequencing and CE methods — including an application for studying low molecular weight metabolites — are still pending.
In the long-term, Target — like many others — hopes to become a drug discovery company, by leveraging the “vast proteomic databases” Peterson is hoping to build. But will the platform replace 2D gels? “There is certainly development work ahead before there is an instrument platform ready to displace them on a broad scale, but that’s certainly the hope,” said Peterson.