Skip Navigation
News and Publications
 
 
 
In This Section      
Print This Page

Promise and Progress - How to Build a Better Mouse Model

A Cancer Revolution

How to Build a Better Mouse Model

Date: December 1, 2004


It’s uncommon for a scientist to announce, “I think we’re going about this the wrong way,” particularly in regard to their own research. But that’s exactly what Rhoda Alani, M.D., did when she reported in the October 20, 2003, issue of Cancer Cell that a common, cancer-linked gene believed by her and other investigators around the country to control blood-vessel growth (angiogenesis) in tumor cells may not be that useful after all. The problem, she reports, is in a commonly used, but somewhat flawed, research model.

The focus of this controversy is the Id1 gene, determined by Alani and others through several earlier studies in mouse models, to be a key step in tumor angiogenesis, or the development of connecting blood vessels required for tumors to thrive and grow. The investigators believed that Id1 had to be activated for tumor angiogenesis, and they had sound laboratory data to back them up. 

In previous studies, investigators had used a popular animal model in which tumor cells are injected into mice and specific genes are evaluated to see how turning them off or on impacts the growth of tumor cells. In the Id1 studies, the tumors grew in the animals with active Id1 genes but were dormant in mice with inactive Id1 genes, leading the research team to believe that Id1 activation contributes to the angiogenesis and the growth of tumors. Yet, when Alani set out to confirm these findings, part of the checks and balances inherent to the research process, she used a different model and got completely different results.

“People don’t get cancer by having tumors injected into their bodies,” says the assistant professor of oncology, dermatology, and molecular biology. “We get cancer through a series of genetic events that occur over time,” says Alani. These events are triggered by both internal factors, such as inherited genetic mutations, and external factors like exposures to carcinogens in diet and environment.When Alani allowed her laboratory mice to develop cancer gradually through exposure to carcinogens, a much lengthier research process but one more closely aligned to how people get cancer, the results were surprising. In fact they were the exact opposite of her earlier findings. All of the mice with inactive Id1 genes developed more tumors than those with the Id1 gene turned on.

That’s when Alani announced to the world that it was time for researchers to take a step back and look at their approach. “Clues to promising cancer drug development are only as good as the model in which you study a process,” she warns. “If knocking out the Id1 gene in two different models produces two different results, then we need to re-evaluate the role that Id1 plays in angiogenesis,” she says. And, if this is the case with the Id1 gene, it is likely the case with other cancer-related genes and may help explain why some therapies that look very promising in mice never pan out in humans, say experts.

Alani says investigators like the tumor transplantation approach, where tumor cells are injected into animals, because it is quicker and easier to perform in the laboratory. And, she is not suggesting investigators abandon this approach. In fact, she believes this model is more analogous to and will help shed light on how tumors metastasize or spread to other parts of the body. However, to get a clear picture of how cancer genes interact to cause both the initiation and progression of cancer she recommends investigators incorporate both models into their research.

Clearly, her approach will take more time, and she is keenly aware of the importance of time in battling cancer. But, she points out, moving ahead too rapidly and promoting new cancer therapies based on the wrong research model will ultimately cost everyone time.

© The Johns Hopkins University, The Johns Hopkins Hospital, and Johns Hopkins Health System. All rights reserved.

Privacy Policy and Disclaimer