With two rounds of VC funds expected to last into 2005, and a recent licensing deal with F. Hoffman-La Roche under its belt, Hanover, Germany-based BioVision is looking to take low molecular weight biomarkers to the next level. The company plans to develop methods to steeply scale up sample numbers for primary biomarker discovery.
BioVision was founded by current chief scientific officer Peter Schulz-Knappe and current chief technology officer Michael Schrader in 1997 as a spin-off of the Lower Saxony Institute of Peptide Research, with the mandate of commercializing technologies that examine the “peptidome.” Schulz-Knappe defined this term as “everything below 20 kDa in size,” including molecules such as cytokines, growth factors, and insulin, that might fall below the 20 kDa range but that by many would still be considered small proteins. What makes BioVision’s peptidomics different than low molecular weight proteomics, Schulz-Knappe said, is that BioVision only uses top-down approaches. “We fear if we do digestion of the native peptides, we will destroy biological information,” Schulz-Knappe said.
But BioVision has narrowed its goals still further. “We’re not interested in all peptides or proteins in the sample. We are only interested in those that are different,” Schulz-Knappe said, emphasizing that the company did not wish to make catalogues of peptides. BioVision’s flagship technology, Differential Peptide Display, is a collection of wet lab and software methods that finds peptides differentially expressed between sample sets based on analysis of mass spectra. The company uses proprietary pattern recognition and statistical analysis software called Spectromania to analyze the differences. The company then uses traditional tandem mass spec techniques to identify the selected peptides. BioVision provides access to its technology through direct fee-for-service, licensing deals and research collaborations, but does not sell any products. The company also provides sample preparation services for reducing complexity and removing high molecular weight proteins, using a variety of traditional chromatography methods as well as “some proprietary methods that can really achieve reduction of total protein content by nearly 98 percent,” Schulz-Knappe said.
Schulz-Knappe calls BioVision the “pioneer of peptidomics” — a technique that he said has only recently caught on with the advent of the biomarker craze and current efforts to examine the low molecular weight proteome. The company has had some success in becoming entrenched in the field: In addition to securing venture capital through funding rounds in 1998 and 2001, BioVision has ongoing collaborations with AstraZeneca, Novo Nordisk, Solvay Pharmaceuticals, Applied Biosystems, Matrix Science, and Roche. Schulz-Knappe said that the company had additional collaborations, primarily with pharma, which he declined to disclose. In addition, BioVision is participating in the HUPO brain and plasma proteome projects. “From what we know, we are the only ones that are specifically addressing peptides either in plasma or in the brain.” Schulz-Knappe said the company is aiming to reach cash break-even by 2005 or 2006, solely through continued service and collaboration deals.
The collaboration with ABI, which began in 1999 with ABI’s German subsidiary and grew in 2001 into a longer term collaboration with the larger company, focuses on combining DPD with ABI’s mass specs and software to look for biomarkers. “We can show that with their instruments, biomarker search[es] can be done in a very good way. Therefore, it’s a marketing effect,” Schulz-Knappe said. On the other side of things, “we value their expertise and their instrumentation power,” he said.
Schrader said additionally that by testing the Q-STAR and other ABI instruments for their ability to do top-down proteomics on peptides, “we really push [the instruments] to the limits … so this is really valuable for them to go forward with their development.”
As part of the deal, BioVision also serves as a European Reference Laboratory for Proteomics and Peptidomics Technologies for ABI.
The F. Hoffman-La Roche deal, announced in March, provides Roche with a license to the patent for DPD. The deal followed on an ongoing biomarker discovery R&D collaboration with Roche Diagnostics for Alzheimer’s disease. Although the Roche research collaboration is currently only geared toward biomarker discovery, “depending on the set-up of the experiment one is doing, the technology is useful for either pure biomarker work, or for biomarker plus lead discovery work,” Schulz-Knappe said, ”because we believe there are more bioactive peptides out there to be discovered and [that] can serve as protodrugs or maybe even drug compounds themselves.”
Now BioVision is looking to scale up the number of samples it can handle in a single biomarker discovery experiment, with the goal of reducing the time required for follow-up validation. “Our goal is to analyze, within one study, 1,000 samples or more. So far most people fall short of 100 samples per study,” Schulz-Knappe said. Small numbers at the discovery stage lead to validation bottlenecks as the biomarker candidates — many of them false positives — are each tested. “We believe that you have to go to high sample numbers with a good quality and sensitive method in order to find just the few really very good biomarker candidates,” he said.
The company is attacking this problem from a number of angles, including switching to MALDI-TOF/TOF for biomarker discovery, adding additional chromatography dimensions and automation, and developing new software capable of handling higher numbers of samples, according to Schulz-Knappe.
As it strives to break even next year, BioVision intends to do what it can to capitalize on the biomarker craze. “We expect due to the increased awareness of the biomarker field, increased activity at the FDA to demand evidence-based medicine — and therefore [to] favor the combination of biomarker plus drug — this is a good playground to be in,” Schulz-Knappe said.
— KAM