By David N. Leff

By now, p53, years ago proclaimed "molecule of the year," has become the best-known — indeed, ho-hum —tumor suppressor gene, found everywhere in cancerous cells.

Far less familiar is the MDM2 gene, which stifles p53's do-good, antitumor tendencies in the bud.

MDM2, discovered a decade ago by scientists at the University of Pennsylvania, stands for "mouse minute-double." These words describe an oddball, paired extrachromosomal element, without a centromere (a true chromosome's wasp-waisted mid-section), which turns up in many drug-resistance genes.

Its discoverers had no idea what MDM2 was up to in mice, and presumed that its counterpart did not exist in the human genome. When they tried to construct knockout animals lacking the gene, in order to determine its function, the MDM2-minus mice promptly died.

"Now we know why," observed molecular biologist Jiandong Chen, who is at Louisiana State University, in New Orleans. "A mouse without any MDM2 would have so much p53 activity that it could not live.

"I have been working on the MDM2 protein for the past seven years," Chen told BioWorld Today, "and now know something about how it normally works. MDM2 not only binds p53, but its own production is stimulated by that p53, so it acts as a sort of feedback-loop regulator."

MDM2 Minimizes Chemotherapy Effects

That self-stimulating, bite-your-own-tail process is key to MDM2's sinister role in promoting tumor growth. It's also central to a just-published report on an antisense oligonucleotide that knocks that troublemaking protein out of the box.

Today's issue of the Proceedings of the National Academy of Sciences (PNAS), dated Jan. 6, 1998, carries that report, titled: "Synergistic activation of p53 by inhibition of MDM2 expression and DNA damage."

DNA damage is precisely what chemotherapy and radiation therapy seek to inflict on tumor cells, with the help of p53. By complexing with that tumor suppressor, MDM2 minimizes that damage and maximizes tumor growth.

Antisense designer Sudhir Agrawal, a co-author of the PNAS paper, is senior vice president and chief scientific officer of Hybridon Inc., in Cambridge, Mass. He told BioWorld Today, "A year ago, in the February 1997 issue of Chemical & Engineering News, I read a report about how to control p53, and the role of MDM2 and its crystal structure. The article said that if one could down-regulate MDM2 specifically, that would be very useful for cancer therapy."

Agrawal, of course, thought antisense, and contacted Chen. "He agreed to collaborate with us," Agrawal said, "and within six months we had this study almost completed." Chen is the PNAS paper's senior author.

To begin with, Agrawal and his co-authors constructed a 20-mer oligo molecule, targeting the gene. This 25- to 30-kilobase sequence sits on the long arm of human chromosome 12, and expresses a protein 491 amino acids long.

"From the antisense point of view," Agrawal observed, "it shows the specificity. When that 20-mer oligo was regulated, it targeted the MDM2's messenger RNA, which reduced the whole MDM2 protein.

"The concept of antisense," he continued, "is that one should go after the disease-associated gene. But in this case we were not doing that. We were looking at the gene, which is a negative feedback regulator. P53 is untouched, not regulated by, our antisense oligo. It's the MDM2 we were after."

As detailed in PNAS, extensive testing of tumor cells in vitro proved that the antisense oligo performed as expected to free p53 from its disabling MDM2 embrace.

Now, generations of nude mice at the University of Alabama, in Birmingham, are getting the full treatment in vivo, at the hands of research pharmacologist Ruiven Zhang, also a co-author. "We started this animal model in September," Zhang told BioWorld Today, "immediately after submitting the paper to PNAS."

He and his lab staff injected two kinds of tumor cells, osteosarcoma and soft sarcoma, under the skins of the test animals. "They were the same cells we used in vitro," Zhang observed, adding, "They grow pretty well in nude mice."

After a first generation of animals had grown their tumors to a good size, they removed the malignant tissues and diced them into small pieces, averaging 25 mg. Then, using a special needle, they inoculated a second generation of mice with these fragments.

"After a couple of weeks," Zhang recounted, "the tumors will grow up again in pretty good shape. We are trying to get some kind of homogeneity," he explained, "Sometimes, in such trials, some injected cells grow faster, others slower. To choose tumors growing at the same speed, you have to waste a large number of animals. This way," he pointed out, "we passage them through a couple of generations, repeating the tumor subdivision two times."

When the third-generation tumors grew to 200 mg, Zhang's team started to treat them with a chemotherapeutic agent, camptothecin, and the anti-MDM2 oligo, at graduated dosage levels and with varied control animals.

If It Works In Mice, It Should In Folks

Although the trial is still in progress, Zhang recounted, "Initial data show significant dose-dependent tumor inhibition." Moreover, "The antisense and chemotherapy drug had a synergistic effect, inhibiting tumor growth by 80 percent to 90 percent. With antisense itself," he went on, "we cannot entirely inhibit the tumor. But adding relatively low levels of the cytotoxic does so almost completely."

Zhang expects his mouse study to end by March.

As for clinical trials, Chen observed: "If these animal experiments confirm that the antisense oligo is effective in these tumor models in the mice, I know it probably would work in humans." But he noted the many intervening steps:

"The FDA would probably have a panel of outside experts evaluate the in vivo data. Also, run tests to determine the tumor types in which the treatment would be most effective, and do toxicology tests in more complex animal models, like monkeys. After that," he concluded, "they might choose certain cancer patients with whom to do Phase I trials." *