By David N. Leff

Under the current criminal correction system, there are two competing schools of thought: incarceration versus rehabilitation.

One would lock up repeat felons for life, with no possibility of parole. The other urges prison programs to rehabilitate offenders and return them to society reformed.

When it comes to the second most-wanted serial killer - cancer - the throw-away-the-key approach practices surgery, radiation, chemotherapy and immunotherapy - all aimed at terminating tumors with extreme prejudice. While these anticancer modalities are indispensable, some research oncologists are looking for ways to cure tumors of their malignant ways and turn them back into healthy cells.

One such investigator is molecular biologist Paul Fisher, who directs the neurology/oncology program at Columbia University's College of Physicians & Surgeons, in New York.

Fisher is senior author of a paper in today's Proceedings of the National Academy of Sciences (PNAS), dated Nov. 23, 1998. Its title: "The cancer growth suppressor gene mda-7 selectively induces apoptosis in human breast cancer cells and inhibits tumor growth in nude mice."

The name of Fisher's game is differentiation. To him this means "taking a cancer cell and literally reverting it back to an early state, where it develops the properties of a benign terminally differentiated cell. In essence," he explained, "one of the hypotheses as to how cancers develop is by aberrations in cell programs of normal differentiation.

"For example, in melanoma, you start off with a melanocyte that then goes through different changes and ultimately becomes a metastatic melanoma. What we were able to do a number of years ago," Fisher recalled, "was determine that recombinant interferon and mezarin, an antileukemic agent, when used together in malignant melanoma cells, reverted their growth and phenotype to a more normal state, something like a healthy melanocyte." (See BioWorld Today, Sept. 2, 1997, p. 1.)

"Those changes," Fisher went on, "were actually quite astonishing, because these cells had lost their malignant properties. They stopped growing irreversibly, and all of a sudden developed traits that suggested that they were cells that no longer had the cancerous properties.

"With that," he added, "we assumed - correctly, as it turned out - that a plethora of genes would be changed as a consequence of this reversion from cancer back to normal. And we also postulated that in many scenarios cancer cells might have lost the expression of genes that in a normal context would prevent a cancer cell from being a cancer cell."

So, he and his co-authors used an approach Fisher had patented about two years earlier for subtracting complementary DNAs (cDNAs). They used this combination of subtraction agents to induce the whole population of melanoma cells to lose growth control, terminally differentiate, and revert to normal.

Fisher explained: "We tried to define and subtract away - what an actively growing melanoma would do - those genes that would now be up-regulated as a function of that cancer growth arrest.

Melanoma First, Breast Cancer Now, Other Tumors Later

"Since we had this library of subtracted cDNAs," he said, "we asked the simple question: 'Are there cDNAs now that are up-regulated when we induce this change, but that are not normally expressed when melanoma cells are malignant and actively growing?' That is, are there tumor suppressor genes present that have the potential to reverse cancer, and only get switched on when we induce this process?"

Whereupon, the team randomly plated out their whole cDNA library.

"When we started off," Fisher recalled, "we had maybe a million cDNAs. After subtraction, this came down to about 8,000. From these we pulled 70 clones at random that we thought might have inserts in them and found 23 that were differentially expressed - which is a pretty high proportion.

"Of the 23, we were able to identify nine genes that were novel, not in the database, that no one else had discovered. The bottom line is, of the nine genes, we identified a novel genetic element, which we called melanoma-differentiation-associated gene-7 - mda-7.

"That mda-7 gene came up when we induced a melanoma cell to terminally differentiate.

"When we put it back into cancer cells, we got a dramatic increase in their ability to form colonies and grow. When we put it into normal cells, we saw no biological effect.

"This information, and the fact that mda-7 didn't work only in melanoma cells, but also in breast cancer, brain cancer, colon cancer and other diverse tumor cells, suggested to us that it might be a ubiquitous cancer suppressor.

"So, we decided to put this mda-7 protein-coding gene into an efficient delivery system that one could use ultimately in patients. We chose a replication-incompetent adenovirus vector, into which we inserted the mda-7 protein-coding cDNA, and put this construct first into cancer cells, then into nude mice carrying human breast cancer tumors. The essence of our paper in today's PNAS is the fact that this cancer-suppressing gene works not only in melanoma, but will also work in other defined cancers - in this case, breast cancer."

Fisher observed that "when we inserted the mda-7 expression vector into normal mammary epithelial cells, even at high levels, we got forced expression of its protein, which had no biological effect in the normal cells."

"It suggests," he pointed out, "that one could use it as a direct therapeutic for patients with either primary tumors or metastatic diseases. And it will have no effect on normal cells."

DISHing Out Cancer Gene Therapy

Fisher calls his patented technology DISH - for "differentiation induction, subtraction hybridization." The university has licensed it to the company he formed in 1995, Seattle-based GenQuest Inc. It has now sublicensed the invention to Introgen Therapeutics Inc., of Austin, Texas, which aims to take it into clinical trials.

"We are about a year and a half away from Phase I studies," Fisher estimated. "The license agreement with IntroGen is such that they are using the FDA-approved adenovirus vector to put the mda-7 gene back in. They've repeated our studies, and are getting the same induction of apoptosis in diverse cancer cells. The Phase I dose-escalation trial they are planning will probably be for head-and-neck cancer." n