This article is the third in a series of bi-weekly columns focusing on the area of "high-throughput biology," or new techniques for high-volume cell based screening and imaging that biopharma is using to validate targets generated through genomics or proteomics, screen for toxicity, and replace other traditional assays.
This article has been corrected from a previous version, which contained incorrect information about anisotropy.
NEW YORK, Jan. 12 (GenomeWeb News) - When Chris Shumate, a veteran of the high-content cellular screening industry, recently started talking to an old high school classmate, Evan Cromwell, an expert in laser systems, the two quickly realized that combining their expertise might just yield a new approach to cellular analysis.
Since the two had graduated from high school in the suburbs of St. Louis over 20 years ago, their careers had followed parallel tracks. Shumate had become an analytical chemist and moved out to California to get into the bioanalytical equipment industry, most recently co-founding Cytoprint, a company that focused on algorithms to extract data from cellular assays. Cromwell, a physical chemist, had also ended up in the fertile crescent of California's technology industry, but in the area of materials characterization and optical inspection.
Together with a third founder, Paul Comita, a chemist who came from semiconductor industry giant Applied Materials, Shumate and Cromwell began to talk about combining Cromwell's knowledge of lasers with Shumate's experience in the cell-based assay arena to develop a laser system that would be able to read high-throughput fluorescent cell-based assays in a novel manner. They also wanted it to be affordable and not to produce so much data that complex algorithms would be needed to make sense of it.
Their ideas revolved around anisotropy - a principle in physics and chemistry in which the measurement of an object differs with directionality. The founders wanted to borrow this concept, as applied to the polarization of light, to use it as a way to derive information about the state of a fluorescently-tagged molecule in the cell. Given that lasers can measure the different properties of light in different directions (polarization), they figured out that they had a powerful tool for measuring this quality in fluorescence assays.
"We started to realize that lasers are inherently polarized, and so you could do laser scanning cytometry in [a] polarization domain," Shumate said. "The idea is that you can take a fluorescent topography of cells or beads and be able to tell whether the signal is rotationally hindered-if its freely rotating in solution, or its rotating in the cytoplasm or it's bound to a membrane or bead."
The group also figured out how to "take a picture" of the object's fluorescence along its lifetime, and combined this with the anisotropy measure, as well as the intensity measurement to arrive at a concept they call "dynamic fluorimetry."
Dynamic fluorimetry, said Cromwell, is "really an object-based measurement where you're characterizing the properties of the specific objects of that cell and gleaning information from that [cell], that has really not been available to the drug discovery community in a high throughput system."
While the team has not yet arrived at the final specs of the laser scanning instrument they intend to sell, they have figured out a way to scale it so they can scan over a well-plate from the standard 96- and 384-well level up to 3,000 to 6,000 well-plates, in as little as 10 seconds at the lowest resolution. The data derived from such a screen is not stored as an image, as with other high-throughput cellular screening systems, but in the form of entries in a database that record the various measurements of anisotropy, lifetime, total intensity, peak intensity, average intensity, brightness, and perimeter. "It's something like a super-plate reader," said Shumate. "You get 20 numbers per well per color."
In being measurement-based, the system does not offer the rich image-based data provided high-throughput imaging technologies such as Reify's Visible Discovery System. (See GenomeWeb News 12-30-03). But Shumate sees this as an advantage. "Other high-content screening systems store images "and those images fill up terabyte servers very quickly," he said. "We don't have that problem since we don't save any image." As such, he said, Blueshift's instrument will not require any complex image processing algorithm.
The system is about the same size as current microplate fluorimeters, and the company is planning to sell each one for less than $100,000. The goal, said the founders, is to get their new concept of dynamic fluorimetry rapidly adopted in the lab.
Specific applications of the system include primary cellular screening, while other high-content systems generally target secondary screening due to the high volume of data generated per cell, Shumate said.
The company's scientific advisors, David Piston from Vanderbilt University, and Tobias Meyer from Stanford, have provided the underlying inventions and associated IP for this new kind of fluorescence screening, and are also testing out novel applications of the technology.
Piston has discovered that system can be used to circumvent some problems with FRET (Fluorescence Resonance Energy Transfer) assays, in which the transfer of energy between a fluorescently tagged donor and acceptor molecule, Cromwell said. In the past, problems such as spectral overlap between emissions from donor and acceptor molecules have made the assays difficult, but with the company's instrument, an alternative measurement of the donor's fluorescence lifetime shift can circumvent this problem, he said.
Meyer, meanwhile, "is interested in the application of this technology to more RNAi knockout functional genomics studies in which you randomly add knockdown genes and you actually ask over a series of cell lines, perhaps, 'did any phenotype change?,'" Cromwell said. "We have a tool that says you don't have to know a priori what the phenotype change will be."
The company also thinks its system will be good for doing basic translocation assays, as the anisotropy and other measurements can track the fluorescence movement from cytoplasm to nucleus.
So far, Blueshift has managed to bring down a celestial concept to the domain of the drug discovery assay, but the company still has yet to make the final leg of the journey from demo model to commercial product.
Currently, the team is talking to "the usual suspects" in the VC community and elsewhere to raise their initial round of financing, Shumate said. They are also talking to established companies to try to lure a partner in to funding development of specific applications.
Cromwell and Shumate are slated to speak at the Cambridge Healthtech's upcoming High-Content Analysis meeting in San Francisco, which is taking place Jan. 29 and 30 (their talk is slated for 6:15 Friday), and at the ALA's Laboratory Automation 2004, at 8AM on Feb. 3. At these events, Blueshift will be up against the heavyweights of high-content screening such as Cellomics (see GenomeWeb News, 12-08-03), Amersham, and others.
But the old friends believe their product has a good chance of finding a niche as the newest screening technology out there. "The world is looking at fluorescence and live-cell assays as the future," said Shumate, "and that's right where we're positioned."