GNN - Genome News Network  
  Home | About | Topics
Targeted drug discovery
A novel screen for potential cancer drugs
Edward R. Winstead

Featured article.

Scientists have developed an automated system for screening thousands of drug compounds against human cancer cells with common genetic mutations. They used the system to identify four compounds (from a set of 30,000) that distinguished mutant cells from normal cells grown in the same culture. One of the compounds inhibited the growth of tumors in mice, making it a candidate for further study.

Fluorescent micrographs of normal and mutant cells exposed to test compounds. View full

Bert Vogelstein, a Howard Hughes Medical Institute investigator at Johns Hopkins University School of Medicine in Baltimore, and colleagues developed the system to discover drug leads. Using colon cancer as a test case, the researchers identified compounds that inhibited the growth of cells with a mutation in the K-Ras gene. This mutation disrupts signaling within the cell and leads to uncontrolled growth.

"Our screen is designed to find drugs that selectively kill mutant cells but leave other cells alone," says Christopher J. Torrance, who led the research at Johns Hopkins. "We hope that finding compounds with greater selectivity will lead to drugs with fewer side effects for the patient."

‘In one week, we can screen more than 8,000 drugs.’

A long-term goal of the research is to move cancer medicine toward personalized therapies. The idea is that physicians will one day use a patient's DNA profile to select the appropriate treatment.

The new strategy has two components. The first is the use of paired cell colonies, or lines, that differ by a single mutant gene—a gene that is altered in the type of tumor being studied. The second is the use of fluorescent proteins as markers that allow researchers to distinguish between mutant and control cells. Genes for yellow or blue fluorescent proteins are inserted into each cell line at the outset.

When both cell lines are cultured together, fluorescent spectroscopy is used to track in real time the relative growth of each line. A drug that primarily inhibits the growth of mutant cells produces images with more blue cells than yellow cells, for example.

The key to the study was having two cell lines that were identical in every respect except the mutant K-Ras gene. Both lines originated with a colon cancer patient, who, in addition to having a normal K-Ras gene, had a second, mutant version. In the early 1990s, researchers at Kyushu University in Fukuoka, Japan, used the parental line to create a second line by 'knocking out' the mutant gene.

The result was two cell lines that grew well under similar conditions and had one genetic difference between them. Vogelstein's group used versions of these lines in their study, which is reported in Nature Biotechnology.

Detail from schematic of cancer drug screen. View full

"The new strategy is an elegant combination of fluorescent protein technology and gene-knockout technology," says George C. Prendergast, of DuPont Pharmaceuticals in Glenolden, Pennsylvania, who wrote an editorial accompanying the study. "This was something I hadn't seen before, and it represents a logical progression in the field."

The new method is simple, rapid, and accessible, says Prendergast, senior director of the Cancer Research Group at DuPont. He notes that the biggest impediment to success may be selecting the right target gene. Many cancer genes are involved in cell growth, and knocking them out prevents any growth at all—making it impossible to create a line of control cells.

The new method is a perfect marriage of information and technology.

Still, says Prendergast, the new method has no particular bias and can produce biologically interesting drug leads. "Once you find the compound of interest, you still have to figure out why it works," he says. "But at least you don't have to consider all possibilities. This method gives you a smaller haystack in which to search, if you will."

The new strategy is designed to save time and money. "By growing the cell lines together rather than separately, you essentially reduce the work by half," says Torrance. Cells were grown in plates containing 96 individual wells, and the researchers tested 80 drugs per plate. "In one week, we can screen more than 8,000 drugs," he says.

Growing the cells together also reduces the chance that a result is due to laboratory errors rather than natural biology. If a robotic arm, for example, fails to place a compound on the cell cultures, both cell lines are affected and the error becomes evident.

“The hardest thing about this strategy is creating the cell lines in which genes have been knocked out,” says Torrance. To knock out a human gene, researchers insert foreign DNA into the cell and hope that it will go to the right place in the genome. But the DNA rarely hits the target gene, according to Ben Park, a researcher developing cancer cell knockouts in the laboratory of Bert Vogelstein.

“The DNA goes to the wrong place in the genome 99 percent of the time,” says Park. “But we are working on improving the efficiency of the technique and recently have discovered all sorts of tricks.” He predicts that if knocking out cancer genes becomes routine, then so will targeted drug screens.

The new method, says Park, is a perfect marriage of information and technology: “We know what the genetic changes in cancer are. Now we can use those changes to screen for drugs.”

. . .

Prendergast, G.C. Knockout drug screens. Nat Biotechnol 19, 919-920 (October 2001).
Torrance, C. et al. Use of isogenic human cancer cells for high-throughput screening and drug discovery. Nat Biotechnol 19, 940-945 (October 2001).
Shirasawa, S. et al. Altered growth of human colon cancer cell lines disrupted at activated Ki-ras. Science 260, 85-88 (April 2, 1993).

Back to GNN Home Page