A foundation established by actor Michael J. Fox is giving $4.2 million to a University of Pittsburgh Medical Center affiliate called RheoGene Inc. to develop gene therapy for Parkinson's disease.

The therapy would involve not only inserting a beneficial gene into brain cells, but also a technology developed by RheoGene that would make it possible to turn that gene on or off as needed.

The ability to turn off the therapeutic gene would be an important safety feature if the proteins it produced had some unanticipated, harmful effect, said J. Mark Braughler, RheoGene vice president for therapeutics. It also would make it possible to regulate the activity of the gene based on the individual needs of each patient.

"I think [the project] is going to be watched very closely," said Dr. Howard Federoff, a neuro-endocrinologist at the University of Rochester Medical Center.

The grant from the Michael J. Fox Foundation would help prepare the therapy for clinical trials in about four years.

The ability to regulate the activity of genes is something many gene therapy researchers now see as important, particularly when the brain is involved, said Dr. Federoff, who heads the Parkinson's Disease Gene Therapy Study Group, a consortium formed by the National Institute of Neurological Disorders and Stroke.

RheoGene was originally a subsidiary of Rohm & Haas Co., a Philadelphia chemical firm. Its technology emerged from Rohm's research and development efforts, but the company in 2004 donated RheoGene to UPMC because the biotech unit didn't fit in its long-range business plans.

Its technique for regulating gene activity is a "switch" attached to the gene that turns the gene on only when activated by a special molecule. This molecule, called an activator, is taken by the patient each day in the form of a pill.

Other researchers have developed their own switches and activators, some proprietary like RheoGene's. One of RheoGene's collaborators for the Fox Foundation-sponsored project, Martha C. Bohn of Northwestern University, has developed a technique in which the gene is always active, but can be turned off if the patient takes a small dose of the antibiotic doxycycline.

Dr. Braughler said RheoGene will launch a clinical trial to make sure that people can safely ingest the activator for long periods. The company also will adapt its switch to two different genes that have shown promise in animal tests for treating Parkinson's disease.

Parkinson's disease is a chronic, progressive disease of the brain that destroys cells that produce a brain chemical called dopamine. As dopamine levels decline, patients develop such debilitating symptoms as tremors in the arms, legs and face, slowed movement and impaired balance and coordination.

One of the genes RheoGene is adapting produces a chemical called glial cell line-derived growth factor, or GDNF. This growth factor, said Fox Foundation spokeswoman Joyce Oberdorf, "has been likened to Miracle-Gro for your brain." The protein seems to make brain cells more robust and helps prevent the death of ailing cells.

In animal studies, GDNF has been shown to stop the progression of the disease and perhaps even reverse it. The problem has been finding a way to deliver the growth factor, Ms. Oberdorf said. One clinical trial, the results of which have yet to be published, used infusion pumps with less than satisfactory results, she added.

Mr. Fox has become an advocate for Parkinson's disease research since announcing that he suffers from the disease in 1991. His foundation has sponsored more than $60 million of research aimed at finding a cure.

The RheoGene work is one of two projects the foundation has funded that is attempting to use gene therapy to deliver the growth factor. The other project is a small clinical trial, already under way, by a firm called Ceregene.

Dr. Braughler said RheoGene also will study a gene for an enzyme called aromatic L-amino acid decarboxylase, or AADC. In the normal brain, this enzyme converts levodopa, or L-dopa, into dopamine. As Parkinson's disease progresses, however, the brain loses the ability to convert L-dopa. Adding genes that produce AADC thus could help extend the therapeutic effects of L-dopa and permit the use of smaller doses.

In therapeutic uses, the genes would be attached to what's called a viral vector -- a virus that has been altered so that it doesn't cause illness. These vectorized genes would be injected directly into the brain, where the vectors would help the genes get inside the brain cells.

Animal experiments will be conducted at Dr. Bonn's lab at Northwestern and by another gene therapy expert, Dr. Krystof Bankiewicz of the University of California, San Francisco. Dr. Braughler said he isn't sure where the clinical studies of the activator will take place.

The inherent risks associated with this approach have to do both with the biochemical safety of the activator and with the RheoGene switch attached to the genes, said Rochester's Dr. Federoff. It's possible, for instance, that the switching "machinery" will stimulate an immune response or somehow interfere with cell function.