MIAMI BEACH, Fla. -- On or about Feb. 2, FDA will receive aninvestigational new drug (IND) application, seeking to inject avariety of cancer patients with a protein-engineered bacterialtoxin.

Ira H. Pastan, chief of the National Cancer Institute's MolecularBiology Laboratory, revealed the plan here on Tuesday at the25th Annual Miami Winter Symposium in a session on chimericproteins. Pastan's tumor-killing exotoxin is secreted by acommon pathogen, Pseudomonas aeruginosa. One molecule ofthis single-chain, 66-kilodalton toxin, he told his audience here,"kills one cell -- if it can get inside."

The molecule's lethal payload consists of three stages, ordomains, Pastan determined. Domaine I binds to a receptor onthe target cell; domain II "by a mechanism as yet unknown"penetrates the cell's membrane and is clipped in half in thecytosol. This unleashes domain III, an enzyme that finishes thecell off.

Pastan defused the toxin by deleting its domain I nose cone.Then he chemically conjugated this truncated two-domain toxinto a monoclonal antibody specific for a carbohydrate antigenthat studs the surface of many malignancies -- colorectal, lung,some ovarian and esophageal -- but not normal cells.

This cancer-seeking package wiped out human tumors in nudemice in six or seven days, at doses five times lower than thelethal dose. Primates tolerated the toxin conjugate even better,Pastan reported. "So with this preclinical data, we feel we havea chance of effective therapy in patients, and hope to start ourPhase I trial in a couple of months."

These initial human studies will take place at the NationalInstitutes of Health's clinical center, Pastan told BioWorld."Though designed as a Phase I toxicity trial," he pointed out, "itmay give us some efficacy data because only patients whosetumors react to the conjugate's antibody will be enrolled."

Meanwhile, back at the lab, Pastan has carried his protein-engineered killer conjugate another step forward. Instead ofthe massive 200,000 molecular weight chemically conjugatedconstruct, he is now developing a series of slim-line single-chain models in which the truncated toxin is fused to theantibody of interest via a small peptide linker. All threeelements constitute a single gene sequence, weighing in at ascant 65,000 MW.

By altering the clutch of amino acids at the toxin molecule's tailend, he achieves a ten-to-twentyfold increase in activity.

"With this single-chain model," Pastan said, "we have not beenable to achieve cures" because most cell lines sensitive to it willnot grow in nude mice.

It has proved sensitive to 25 of 26 freshly cultivated humanprostate cancer cells, in work just beginning.

Pastan foresees the single-chain toxin conjugate as being usedprimarily to scavenge residual metastatic cells of many tumortypes, once the main bulk of a solid tumor has been surgicallyremoved.

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.