By David N. Leff

An all-purpose gene therapy system that turns its DNA target on and off at will is two steps closer to human trials.

A couple of years ago, molecular endocrinologist Bert O'Malley reported enabling transgenic mice to express human growth hormone (hGH) at the flip of a chemical switch, taken orally.

That toggle switch was a tiny dose of a notorious drug, mifepristone, much better known as RU-486, a non-surgical abortifacient. (See BioWorld Today, March 26, 1997, p. 1., and Aug. 18, 1994, p. 1.)

That was then - over 22 months ago, a long time in the evolving drama of gene therapy. Now, O'Malley updates his research in the current Proceedings of the National Academy of Sciences (PNAS) dated Jan. 19, 1999. His paper bears the title, "Adenovirus-mediated regulable target gene expression in vivo."

O'Malley chairs the Department of Cell Biology at Baylor College of Medicine in Houston.

"Back in 1997," he told BioWorld Today, "the field of gene therapy was working on how to get DNA into cells, and just making some progress in that area. In fact, it has come along nicely with viral vectors and direct DNA delivery.

"But there was no persistence of gene expression," O'Malley continued. "The viral or direct DNA expression was gone in a few weeks or months. And during this period of time, new adenoviral vectors, a couple of which we're using now, and another adeno-associated virus vector, are beginning to show gene persistence of over a year."

New System Adds Power To RU-486

Armed with a new high-capacity (up to 35 kilobases of DNA) retroviral vector, O'Malley and his co-authors faced their next need: "a good in vivo gene-regulating system. We're using the same 1997 principle," he observed, "but now we've refined the system so that it works on minute amounts of RU-486 - which are a thousand-fold below what would have any pregnancy-aborting effect. This mifepristone is now so powerful that it will turn a gene on at very high levels, and preselected time points.

"Also important, we've engineered it so that the target gene, human growth hormone, is just about completely turned off in the absence of the RU-486 switch, which can be taken by mouth as a simple exogenous pill. There's very little endogenous expression in the absence of swallowing the drug.

"A perfect gene-switching system," O'Malley continued, "would be such that when you have the target gene in there - which in this case is hGH, but in other cases whatever else you would insert - you would want that switched off until you actually take the pill. You want it under your own control.

"And over the intervening period since '97, we've made good strides in that, so now it's really undetectable in the mice when they don't get the medicine."

O'Malley made the point that his original modus operandi "is still a gene-switch molecule that is a mutated progesterone receptor. It can't bind anything in the body, except exogenous antiprogestins, in our case RU-486.

"We chose it for convenience, and engineered the drug so that very small amounts of it will turn the gene on. It does this," he went on, "with a nice dose response. If we give our mice a little bit, the hGH gene is in the on position a little bit; a little bit more, it's on higher."

He compares this homeopathic-strength mifepristone to "regular medicine. You stop the RU-486, the gene goes back off again. Our PNAS paper shows that in mice: turn it on, turn it off, on, off, back and forth. Just by taking an oral pill directly, to target genes on an as-needed basis.

"The reason we can use such small amounts for the gene switch," he explained, "is that it doesn't have to compete with anything. Nothing in the body can bind to it, including endogenous progesterone. So it's mainly a question of gradually perfecting this gene therapy system until it reaches a point where it's ready for prime time - which I think it is now."

O'Malley and his team plan to move on this year from these studies in the mice, using hGH as a marker gene with an easily measurable biological effect. "We'll now put in therapeutic genes, test those in animals, and probably next year try to get into some very early stage human work."

"The possibilities are endless there," O'Malley observed. "It's whatever modern therapeutics would like to make in these conditions. It could be agents against cancer or anemia, or to protect against aging or atherosclerosis."

His first druthers are "to go for agents that might fight cancer, either by shutting off cell replication or inhibiting tumor angiogenesis." Cancer is his first entity of choice "because it will be the most likely thing to get early approval for our system's use, since it's a fatal disease. We hope to begin Phase I studies next year."

O'Malley sees his newly honed gene transfer strategy as equally adapted "to treat any disorder to which you could deliver DNA that would help a metabolic disease. So for instance, delivering to a region of the brain the gene for an enzyme that would make dopamine and alleviate Parkinson's disease."

Two Competitors Verging On Merging

GeneMedicine Inc., of The Woodlands, Texas, which O'Malley co-founded with four other Baylor scientists in 1992, "has licensed this system, including its viral vector, from Baylor," he observed. "They are developing it right now with Merck & Co. Inc. [in Whitehouse Station, N.J.], which will be moving it along commercially. One of the main overseers of that contract is gene therapist Thomas Caskey, a Merck vice president, who had worked on this virus while here at Baylor. Our group is doing the fundamental science," he added, "funded by GeneMedicine."

O'Malley noted that GeneMedicine is this month in the process of finalizing its merger with a rival gene therapy company, Megabios Corp. of Burlingame, Calif. (See BioWorld Today, Oct. 27, 1998, p. 1.) n