Reducing Animal Testing Using Stem Cells

Medicinal chemists know their latest compound won't be taken seriously as a drug lead until it's proven its punch in animal models. Stem cells, however, may turn out to be a promising alternative to animal testing.

Although animal models are an established substitute for humans, cancer has been "cured" so many times in mice that even the average layperson holds a healthy skepticism about animal proxies. The problem is that other mammals aren't always a great model for us, especially when it comes to reflecting a drug's potential liver and heart toxicity. To make things worse, animal models are expensive, and they take time to produce results.

Enter embryonic stem cells. They can grow and differentiate in a petri dish into the variety of cells that build a human organ. These in vitro versions of human tissue are superior to dishes of a single cell type to assess the toxicological impact of a drug, and they provide a human impact profile, not a mouse's.

Embryonic stem cells also hold promise on the front end of drug discovery. Disease genes are inserted into embryonic stem cells, which are then induced to differentiate into human disease tissues that can be used to screen for drugs.

The kinds of in vitro models of human tissue based on embryonic stem cells "are as diverse as the cell types they generate," says Gabriela G. Cezar, an assistant professor of animal sciences at the University of Wisconsin, Madison.

Although many stem cell researchers in both academia and industry see the primary promise of stem cells as direct therapeutics, Cezar disagrees. "Embryonic stem cells are in the spotlight for the great promise they have in regenerative medicine. This is not hype," she says. "But the long-term benefits of using stem cells to improve drug testing will have wider impact." The argument goes as follows: Almost everyone will take a drug at some point, but not everybody will need stem cells directly as therapeutics.

Last June, Douglas A. Melton, a cell biologist, and George Q. Daley, a biological chemist, both at Harvard University, initiated a multi-million-dollar endeavor, sponsored entirely by private funders, to create embryonic stem cell models of human disease. "The idea is to take an embryonic stem cell and, in a petri dish, tell it to become the neuron," Melton says. "You are watching normal development, not in a person but in a petri dish."

And then in the next dish, the researchers will create an embryonic stem cell line using the genes from a Parkinson's patient and watch that cell become a dopaminergic neuron, the kind that shows that disease's degenerative symptoms. "You compare the first dish to the second, and you ask where the second cell line screws up," Melton tells C&EN. "Then you screen for drugs that can slow the process. In the ideal case, you would prevent it, but tha's a higher hurdle."

The eventual plan is to study a whole host of diseases that have a mix of genetic and environmental roots, including diabetes and Alzheimer's disease. Right now, the group is ironing out the details on embryonic-stem-cell-derived mouse models of two spinal cord diseases, spinal muscular atrophy and Lou Gehrig's disease, says Lee L. Rubin, a neurobiologist leading the drug screening at the Harvard Stem Cell Institute.

For biotech and pharmaceutical firms, using stem cells to test potential drug leads for toxicity could help them avoid wasting time on harmful compounds. "The goal is to have a more clinically reliable screening system that one can apply in the drug discovery process," says H. Ralph Snodgrass, chief executive officer of Vistagen Therapeutics, which has a major focus in developing stem cells for toxicological testing. "Stem cell screens would be part of a prioritization of compounds early on in the discovery and development process." AstraZeneca is working with Göteborg, Sweden-based Cellartis on human embryonic toxicology screens. Many stem cell companies—including the Madison, Wis., start-up Cellular Dynamics International and Geron of Menlo Park, Calif.—also have programs in this area.

Snodgrass points out that some drugs have harmful effects in certain ethnic subpopulations, which could never be predicted by conventional techniques because there are no animal models of race. "Various genetic differences make a big difference in the way a cell responds to a drug. The classic example of that is people of Asian extraction, who respond differently to a wide range of drugs compared to northern Europeans." Embryonic stem cells with a specific ethnic genome could be used to develop ethnic cell populations that could be screened to check the toxicity of potential drug leads.

Currently, the most successful development of stem cells as in vitro models for toxicology testing is in human cardiac tissue. Many drugs have been removed from the market as a consequence of cardiac toxicity—an infamous example is terfenadine, an antihistamine that caused close to 100 deaths as a consequence of adverse cardiac events when it was released on the market. Embryonic stem cells have since been differentiated into cardiac tissue that shows potential as a toxicological model of disease, and many companies involved in using embryonic stem cells have research programs in this area.

Developing stem cell models of the liver, the other major organ that animal models fail to emulate, has been more difficult but is nevertheless a main focus of many researchers and companies working in this area.

Where the promise of stem cells falls short is in their modeling of systemic toxicity. Sometimes one organ alters a drug in some way, but this subsequent metabolite is toxic only to a different organ. Since stem cells in a dish reflect only a single organ's response to a drug, there would be no way to check the impact of subsequent metabolites on the whole organism. "Stem cells won't replace animals," Cezar says. "But they will reduce and refine our use of them."

Whether they are being used as an alternative to animals in toxicology testing or disease modeling, embryonic stem cells will, many researchers hope, offer an in vitro representation of human response second only to the real thing.