In the matter of starving a tumor to death by blocking its blood supply, the beat goes on.

It hit a frenzied media crescendo last spring, half a year after anti-angiogenesis pioneer Judah Folkman, of Harvard-affiliated Children's Hospital in Boston, reported that two natural substances, endostatin and angiostatin, stopped the proliferation of tumor-feeding blood vessels. The hospital entered a license agreement with EntreMed Inc., of Rockville, Md. (See BioWorld Today, Oct. 20, 1998, p. 1, and Dec. 2, 1997, p. 1.)

At that time, Folkman's lab had barely enough of the experimental compounds to fuel mouse trials, let alone schedule human clinical trials. And their exact mode of action is still unknown.

Now, two scientific journals report new data on a new anti-angiogenic substance, code-named TNP-470 (standing for "Takeda neoplastic product"), already in extended human trials.

Today's Science, dated Nov. 13, 1998, carries an article titled "Structure of human methionine aminopeptidase-2 complexed with fumagillin." The paper's senior author is chemical biologist Jon Clardy; its lead author is Shenping Liu. Both are in the department of chemistry and chemical biology at Cornell University, in Ithaca, N.Y.

Today, too, the Proceedings of the National Academy of Sciences (PNAS) released an article it had originally slotted for publication in January 1999, but brought forward, an editor's note explained, "to coincide with the release of a paper on similar research in this week's Science."

The PNAS entry's title reads: "Molecular recognition of angiogenesis inhibitors fumagillin and ovalicin by methionine aminopeptidase." Its lead author is Ph.D. candidate Eric Griffith, a biochemist in the Center for Cancer Research at the Massachusetts Institute of Technology (MIT), in Cambridge. Its senior author is biologist and chemist Jun Liu (no relation to Cornell's Liu).

Fumagillin is a small organic molecule emitted by a notoriously pathogenic fungus, Aspergillus fumigata. Griffith described its serendipitous discovery: "It was found way back in the 1940s or 1950s," he told BioWorld Today," and first identified as an antibacteriophage agent.

"But as for the anti-angiogenic activity of it," he went on, "that was discovered in Judah Folkman's lab at Harvard Medical School. I guess what happened was they were growing endothelial cells in culture, and — a bit reminiscent of penicillin — they got a fungal contaminant. And they saw a gradient of cell growth inhibition around this fungus.

"They figured there must be some soluble factor there. So, in collaboration with Takeda Chemical Industries, of Osaka, Japan, they purified it. As they reported in 1990, it turned out to be fumagillin. Then they made some derivatives, and that's where TNP-470 came in."

That drug is in a wide-ranging set of clinical trials at various cancer centers throughout the U.S. and Europe, as a potent inhibitor of angiogenesis — that is, a biochemical foe of endothelial cells, which line blood vessels.

Fumagillin gains that potency by binding to its receptor-like complementary molecule, methionine aminopeptidase (MetAP).

"One could call MetAP a receptor," Griffith observed, "except that it's not on the cell surface. It's an intracellular enzyme, but it seems to bind specifically to the drugs, fumagillin as well as TNP-470. Also ovalicin, which comes from another fungus, Pseudeurotium ovalis, but has the same activity as fumagillin. So these two natural products, produced by different fungi, all bind to this protein.

"All proteins initiate their activity by methylation with a methionine," Griffith explained. "The MetAP enzyme removes that methionine. The enzyme comes in two molecular variants, MetAP 1 and 2. These anti-angiogenic compounds affect only 2, not 1."

Epoxide Groups Come Front And Center

As Griffith's PNAS paper reported, "We wanted to get a better idea of how these drugs were recognizing this enzyme so potently. From earlier experiments, we knew that they were forming actual covalent bonds. To do that we assumed it would be through epoxide residues."

Epoxides are small triangular groups, with an oxygen at one point and two carbons, as a rule, at the other. They give their name to epoxy resin plastics and adhesives.

"All three drugs — fumagillin, ovalicin and TNP-470 — have two epoxides in their structure," Griffith continued. "One of them, the high-energy ring epoxide has been shown in our paper and Clardy's paper to be very important. We made analogs of the drugs, where we replaced that high-energy epoxide with different groups that shouldn't be active. What we found was that if we removed that ring epoxide we got a dramatic loss in potency, and the compound couldn't bind covalently to the protein, whereas, if we remove the other epoxide, we lose 10,000-fold less activity.

"Now we have a good idea how the drug is being recognized by the protein," he went on, "and that should help in the future to potentially give us some idea of how to design a therapeutic that still has this high potency but lacks some of the side effects."

TNP-470 Therapy's Double Edge

Griffith observed that "TNP-470 is in the clinic and has shown some neurotoxicity. Also, its half-life in the blood is fairly low; it's pretty rapidly turned over, and that could be due to a reactive group like an epoxide. But I know there are some cases in the clinical trials where one patient had complete remission from metastatic cervical cancer.

"Angiostatin and endostatin seem to be more potent than TNP-470 in mice," he said, "but they haven't been tried in humans yet. What Takeda and Folkman find is that if you administer TNP-470 with some other things, you will prevent tumor growth. But if you already have a tumor present in a mouse, it won't necessarily shrink it completely away."

Griffith's MIT co-author Jun Liu met Clardy for the first time at a Gordon conference last August in New Hampshire. The PNAS paper thanks him "for generously sharing information on the crystal structure of MetAP2-fumagillin before publication, and for helpful discussions."

"Clardy's paper," Griffith said, "is extremely complementary to ours, in that it basically gives a photographic snapshot of the drug bound to the protein."

Meanwhile, also on the anti-angiogenesis front, a news report that other scientists have been unable to match Folkman's success sent EntreMed's stock (NASDAQ:ENMD) tumbling about 24 percent. The company's shares ended Thursday at $24.88, down $7.75. n