Published online 9 February 2010 | Nature 463, 719 (2010) | doi:10.1038/463719a

News

Testing time for stem cells

Industry turns to cell lines for drug screens.

A deal between Roche and two academic institutions will focus on stem-cell-based drug screening.Roche

The drug industry is keener on stem-cell technologies than ever before — and not just as a source of new treatments. A wave of new partnerships aims to use stem cells as a way to screen other potential drug candidates.

In the latest such example, Roche last week announced a deal worth some US$20 million with Harvard University in Cambridge, Massachusetts, and Massachusetts General Hospital in Boston. Roche, based in Basel, Switzerland, will use cell lines and protocols developed by academic researchers to screen for drugs to treat cardiovascular disease and other conditions.

Because relevant human cell types are often unavailable, current screens tend to use cells from rodents or human tissues other than the ones researchers want to target. The hope is that stem cells could provide exactly the type of cells relevant for an assay.

The deal is the latest in a string of similar partnerships. Within the past 15 months, Pfizer of New York and GE Healthcare of Chalfont St Giles, UK, signed deals geared towards using stem cells in drug discovery with the California companies Novocell of San Diego and Geron of Menlo Park, respectively. In 2008, GlaxoSmithKline teamed up with the Harvard Stem Cell Institute for research in neuroscience, heart disease, cancer, diabetes, musculoskeletal diseases and obesity. And in 2006, AstraZeneca of London began collaborating with Cellartis of Gothenburg, Sweden, to use stem cells to make human liver and heart cells for safety tests.

Although using stem cells for drug screening and early research should be easier than developing them into replacement tissues, even the most ardent advocates admit that it won't be straightforward. "At the moment it's all really early days," says Stephen Minger, who left King's College London last year to head GE Healthcare's efforts to develop drug-screening tests with cells derived from human embryonic stem cells.

"What needs to be demonstrated is the actual application of the technology," adds John Walker, the chief executive of iPierian, a stem-cell company in San Francisco, California. The firm has created motor neurons using induced pluripotent stem (iPS) cells derived from people with and without spinal muscular atrophy, a neurodegenerative disease. Company scientists are investigating whether drug candidates can disrupt telltale protein clumps seen in neurons derived from affected individuals. Eventually, this work could lead to what Walker calls an "in vitro clinical trial" in which iPS cells derived from a wide variety of individuals could be used to predict patients' response to a drug.

Most drug companies remain to be convinced that new screens can predict drugs' properties as well as or better than existing methods. They also don't know whether cells can be produced reliably in sufficient quantities for screening, and whether regulatory agencies will accept data derived from them.

ADVERTISEMENT

Roche says that it is already running screens based on stem cells to test drugs for cardiotoxicity and effects on neurogenesis. Matthias Steger, the company's global business development director for stem-cell alliances, says that under the latest deal, Harvard and Massachusetts General Hospital will work on embryonic stem cells and iPS cells derived from individuals with various forms of cardiovascular disease. The institutions will receive milestone payments as screens are validated, and the collaboration will last 3–5 years.

Ruth McKernan, chief scientific officer of Pfizer's regenerative medicine unit, notes that although screens based on stem cells will be useful, they are not the main reason that Pfizer is interested in these cells. "This progress, though important, is incremental when compared to the real promise in using these cells as therapeutics," she says. 

Commenting is now closed.