Breast Cancer Gene Discovered with New Method

1/9/97

CONTACT: Bill Wells or Rosanne Spector, (415) 723- 6911

BREAST CANCER GENE DISCOVERED WITH NEW METHOD

STANFORD -- By recreating the genetic mayhem that characterizes cancer development, Stanford researchers have isolated a key gene involved in human breast cancer. The gene, called TSG101, was defective in almost half of the breast cancers the researchers studied.

TSG101 is the first gene identified through an innovative strategy called random homozygous knockout (RHKO), devised by Stanford professor of genetics Dr. Stanley N. Cohen and postdoctoral fellow Dr. Limin Li. The researchers report the gene's role in breast cancer in the Jan. 10 issue of the journal Cell.

Using the new method, scientists can simultaneously inactivate, or "knock out," both copies of a gene in a cell without knowing the gene's identity or function. Then they look among millions of these mutant cells for individual cells that are capable of forming cancers. By identifying the gene inactivated in the cancer-forming cells, researchers can isolate genes called tumor suppressors, which normally are necessary to stop cancer development.

"It is an elegant method because you can get directly to the gene," said Dr. Uta Francke, a professor of genetics and Howard Hughes Medical Institute investigator at Stanford, who is collaborating with Li and Cohen.

In this case, the RHKO strategy enabled the researchers to isolate a defective gene in breast cancers that weren't necessarily familial. This is significant because, overall, the vast majority of breast cancers arise through spontaneous mutations rather than through inheritance. In contrast to TSG101, the previously identified BRCA1 and BRCA2 genes are altered in a high percentage of women with relatively rare, familial forms of breast cancer, but are normal in most breast cancers.

Knocking out genes

Related knockout approaches have long been used with bacteria, noted Cohen, who more than 20 years ago invented the basic methods of DNA cloning, or recombinant DNA, together with Herbert W. Boyer of the University of California, San Francisco.

By identifying cells containing mutations that cause the loss of a particular trait, it is possible to find the gene responsible for that trait, Cohen explained. A common approach used with bacteria, which have only a single copy of each gene, is to mutate genes randomly within cells of a population, select the cells that lose a particular trait of interest, and then identify the gene inactivated by the mutation, thereby identifying the gene coding for the missing trait.

This mutation-and-selection procedure is straightforward in bacteria. But animal cells have two copies of each gene -- one from the mother and one from the father. "It is very difficult," said Li, "to mutate two copies simultaneously, and you usually have to mutate both copies to see a change in the cell's traits."

There are some laborious ways to do this if you know the identity of the gene of interest. But the random-hit, bacteria-style experiment was not possible in animal cells until last May, when Li and Cohen first reported their RHKO method in Cell.

In the new method, a specially constructed fragment of DNA jumps randomly into the DNA of the host cell. When the fragment inserts itself into a gene that is actively making proteins, the inserted DNA kicks into action and produces a detectable protein. The researchers then isolate the protein-producing cells, each of which has a different gene blocked by the inserted DNA.

Each cell is still, however, making RNA (the messenger that instructs the construction of proteins) from the other copy of the blocked gene. To knock this RNA out of action, the researchers switch on production of an "antisense" RNA that sticks to the normal RNA and clogs up its ability to be used by the cell.

They do this by instructing their inserted DNA to produce RNA in the opposite direction to the RNA produced by the gene that contains the insert. This antisense RNA sticks to the RNA still being made by the second copy of the blocked gene. Thus, both copies of the gene containing the DNA insert are inactivated.

Isolating TSG101

When Li and Cohen followed this procedure, they found that a few of their mutant cells could grow on an agar surface -- a hallmark of cancer-forming cells. Sure enough, when these cells were injected into mice, they formed tumors. And when the antisense RNA was turned off, the cells lost their cancerous properties.

The researchers then used a biological tag, which they had put into the original DNA insert, to pull out the mouse gene that had been disrupted. They dubbed the gene TSG101 -- tumor susceptibility gene 101-- and found that it was a novel gene whose product may control the expression of other genes.

Li and Cohen next used the mouse gene to isolate the similar human gene. Francke and postdoctoral fellow Dr. Xu Li, also from the Stanford Department of Genetics, joined the team at this point to narrow down the location of the human gene to a single "arm" of a chromosome, and then to a single sliver of that arm.

The scientists were excited to find that the gene was in a chromosomal area that is often missing in various human cancers, particularly breast cancer. Other researchers had attempted, without success, to pinpoint the tumor suppressor in this area. "Using conventional techniques," said Francke, "this can take forever."

With their gene in hand, however, the Stanford researchers could easily test its importance. When they looked at tissue from women with late-stage breast cancer, they found that a piece of the TSG101 gene was missing in 7 out of 15 cancers, whereas the gene was intact in neighboring, non- diseased breast tissue from the same patients.

The scientists note that far more extensive research is needed to get more accurate information about the frequency of TSG101 defects in breast cancer. Such studies should also indicate whether the mutation predisposes cells to become cancerous, or encourages existing cancer cells to become more aggressive. Testing for abnormalities in the gene could then serve as a tool for diagnosing or predicting disease progression.

Meanwhile, scientists using the RHKO method may discover other new, important genes for study. Cohen's lab is examining more candidate tumor suppressor genes. In addition, since publishing their report on the RHKO procedure last May, Li and Cohen have sent the genetic tools they devised to 35 different laboratories worldwide, which are now using the strategy to knock out genes affecting traits they are interested in studying.

###