GNN - Genome News Network  
  Home | About | Topics
How to Recognize a Cancer Cell
By Donna Bernstein

In an effort to use genomics to identify differences between normal cells and cancer cells, researchers have developed a system for identifying genetic changes associated with the disease. The system can reveal whether a specific cancer gene is highly active in a tumor cell—based on the cell’s patterns of gene activity, or genetic signature.

Knowing which cancer genes are active in tumor cells could help researchers dissect the complex biology of the disease. And if the system, which was tested in mice, can be developed for use in people, doctors could use it in selecting treatments for patients.

Diagram showing the different patterns of activity for the three EF2 genes.

Joseph R. Nevins, of the Duke Institute for Genome Sciences and Policy in Durham, North Carolina, and his colleagues defined the genetic signatures associated with three cancer genes in mice. The human versions of these genes, called H-Ras, Myc and E2F, contribute to many cancers, including eye, brain and breast.

After defining the genetic signatures, the researchers used them to distinguish among cancer cells and normal cells. They were even able to tell the difference between cells that had different forms of the E2F gene, which are very similar.

"This is most surprising because previous studies hadn't seen any significant differences between the E2F genes," says Nevins. "It demonstrates the power of the genomic analyses at picking up subtleties."

The method, called “metagene analysis,” is based on a statistical analysis of thousands of genes in cancer cells using “gene chip” technology.

It can be difficult to identify the underlying differences between normal cells and cancer cells, says Chi Van Dang, a cancer researcher at Johns Hopkins University School of Medicine in Baltimore.

“But with genomic profiling, the genetic patterns can be traced backward to see the key switches that are actually causing the cancer,” says Dang, who was not involved in the study.

The Duke team, which collaborated with Georgetown University's Lombardi Cancer Center in Washington D.C., also identified genes not previously associated with cancer that may play a role in the disease. The findings appear in Nature Genetics.

See related GNN article
»Genomic Patterns Reveal Breast Cancer Risks

. . .

Huang, E. et al. Gene expression phenotypic models that predict the activity of oncogenic pathways. Published online in Nature Genetics (May 18, 2003).

Back to GNN Home Page