By David N. Leff

A virus that never harmed a human being is under contract to deliver therapeutic genes into muscle cells.

Not only is adeno-associated virus (AAV) harmless; it's been known only as the side-kick of the nastier adenovirus (AV), which causes respiratory and other infections. Out in the real world, AAV needs AV just to get activated, but in the University of Pennsylvania's Institute for Human Gene Therapy (IHGT), it's about to take on an independent career as the vector for a smart gene that a simple daily pill turns on and off.

This dual function -- a muscle-bound vector inserting a self-regulating gene -- combines the technologies of Philadelphia-based Genovo Inc. and Ariad Pharmaceuticals Inc., of Cambridge, Mass. Gene therapist James Wilson, co-founded Genovo, which became active in 1995.

"Genovo's joint venture with Ariad," Wilson told BioWorld Today, "brings together our technology for gene transfer into muscle with their regulated gene expression." (See BioWorld Today, Feb. 19, 1997, p. 1.)

Wilson is principal author of an article in the March issue of Nature Medicine, titled: "Recombinant adeno-
associated virus for muscle-directed gene therapy." Ariad's president, chairman and CEO, Harvey Berger, told BioWorld Today: "Jim's paper in Nature Medicine was really the scientific basis for our new partnership."

Wilson described Ariad's share in the project as "a system you can incorporate into your vectors that allows the gene to be regulated in response to a cue -- a pill. This orally available small molecule is a form of rapamycin, a common immunosuppressant drug.

"Our concept," he continued, "is to inject the vectors into muscles, for stable, efficient and apparently nonimmunogenic gene transfer."

The gene products Wilson has under current study include "those that control hematopoiesis, such as erythropoietin, some other cytokines, and -- an obvious target -- proteins for treating hemophilia."

Blood-clotting Factor IX, which many hemophiliacs lack, tops the list of proteins being prepped for the new gene-therapy paradigm. Research pediatrician Catherine High, at Children's Hospital in Philadelphia, a co-author of the Nature Medicine article, is working with Wilson and another co-author, Krishna Fisher, to bring Factor IX from preclinical to clinical testing. "We want to get this into humans as fast as we possibly can," Wilson said. "Our goal is to get it in within 12 months, or sooner.

One-Shot Needle; Daily Gulp

"We hope to change the paradigm of therapeutic proteins," Wilson continued. "In the past, it's meant repeated injections of EPO for anemia, insulin for diabetes, clotting factors for hemophilia. We aim for a one-time injection of the vector that carries the regulated gene expression, then a daily pill of the drug that confers the regulation."

Fisher told BioWorld Today that he and High "had been working on expressing Factor IX in the liver, with not very good results." So they loaded their AAV vector with Factor IX genes, and injected them into the muscle cells of mice. These in vitro tests showed the AAV vector could deliver, so now they have in vivo preclinical trials in progress.

Fisher raised one paradoxical hurdle: "If you were to treat a human with hemophilia who had never seen Factor IX, his immune system might reject it. This is a problem we can deal with," he added, "but it really hits at all of gene therapy."

Virologists discovered AAV in the 1960s, only because it lurked in the shadow of AV. Fisher explained that association: "What wild-type AV does is kick an infected cell in the butt and say, 'Come on, you've got work to do. You've got to activate this AAV.' " But the recombinant vectors of the muscle experiments needed no kick-start from AV.

"At 4,600 base pairs, AAV's genome is very small and simple," Fisher went on. "By comparison, AV's runs 36,000 bp." That AAV mini-genome ends with two inverted terminal repeat sequences. They don't encode any genes, but carry information necessary for the virus to replicate and integrate.

"And between those two sequences," he said, "are the AAV's viral genes for replication and virion formation. With one simple step, we could lift those right out and substitute the genes we want to express."

French Prove Pill Works, But Not Ex Vivo Therapy

A companion paper in the March Nature Medicine reports on a different, but similar, attempt to transfer genes into muscle fibers. Its title: "Long-term control of erythropoietin secretion by doxycycline [tetracycline] in mice transplanted with engineered primary myoblasts." Its principal author is virologist Jean-Michel Heard, who heads the Pasteur Institute's Retroviral and Gene-Transfer Laboratory, in Paris.

Using two retroviral vectors, the French team transferred cDNA expressing erythropoietin into explanted mouse myoblasts. After ex vivo culture, they returned these transduced cells to mice, which then got tetracycline in their drinking water. Gene expression increased 200-fold, in response to both myogenic cell differentiation and the tetracycline stimulation.

For five months, EPO secretion could be repeatedly turned on and off, controlled by the oral dose of the antibiotic.

This regulatory system, Heard explained to BioWorld Today, "is the way that bacteria develop resistance to tetracycline naturally. We used the same molecules, but modified them, completely changing inhibitor to activator But this activator does its activating only when tetracycline is present, not when it is absent."

As for its clinical potential, Heard said, "Very frankly, the situation described in our paper is not relevant for human use. This technique for gene transferring to muscle cannot really be proposed for humans, because it supposes that muscle cells are grown in culture for a while, then returned to the patient. All the groups that have tried to grow human myoblasts or muscle cells ex vivo, then put them back, have failed."

Heard now plans to "use the tetracycline regulatory system in the context of the AAV vector, instead of the retroviral vector, and in vivo instead of ex vivo." *