Like stem cell therapy, gene therapy is a field where there has been much promise but little proof to date. The field’s scientific advances currently are at the level of making the technique work, rather than discovering miracle cures.

In an article in the February 2006 issue of Nature Biotechnology, scientists from the University of California at Berkeley and Ohio State University describe one such scientific advance: the directed evolution of adeno-associated virus to make better gene carriers.

Adeno-associated virus, or AAV, is not affected by the halt the FDA put on a subset of gene therapy trials in 2003. It also is not related to adenovirus, the viral vector type that killed teenager Jesse Gelsinger during a gene therapy clinical trial in 1999. Its name is due to the fact that it originally was isolated as a contaminant in an adenovirus. AAV currently is being used in several gene therapy clinical trials investigating neurodegenerative disorders, hemophilia and other conditions.

The virus itself is not pathogenic, which ironically is both a strength and a weakness for its use in gene therapy. AAV infection is fairly widespread, but "is so harmless that many who have been exposed don’t even know it," senior author David Schaffer, associate professor of chemical engineering, told BioWorld Today. But those who were unknowingly exposed still have antibodies to AAV.

Estimates of how much of the population has antibodies to AAV differ widely, from one in three to 19 in 20. But even at the low end, it is a significant fraction of the population, and their antibodies have a strict nondiscrimination policy: They will attack gene therapy vehicles with the same gusto as AAV from the wild.

Schaffer and his colleagues wanted to make AAV variants that would be able to evade those antibodies. Their approach was to make a library of AAV mutants by employing a method called error-prone PCR (apparently, some processes are even more error-prone than regular PCR) to create a library of AAVs with mutations in the capsid proteins, which make up the viral shell and are antibody targets.

"Re-engineering a virus is a tough business," Schaffer said. "We don’t know all the mechanisms by which they work." But the directed evolution strategy can nevertheless be used to improve desired characteristics because "evolution works without knowing anything about mechanisms."

When the scientists screened the library they had made, which contained about 210 million mutants, they found two viral variants that were able to escape recognition by antibodies at a concentration that was several orders of magnitude greater than that required to render wild-type AAV ineffective. Additionally, in a lucky break, both mutants provided better gene delivery than wild-type AAV.

A patient’s body will start making antibodies to the new and improved AAV as soon as gene therapy is initiated. But Schaffer pointed out that "it does buy us one shot. And sometimes one shot goes a long way. One shot has the potential to cure some disorders."

The scientists also used directed evolution to weaken AAV’s binding to heparin. Heparin exists in two forms: as a soluble anti-clotting protein in the bloodstream, but also on the cell surface.

"A good number of viruses use heparin to gain entry into cells," Schaffer said. The advantage of this is that AAV can infect, and deliver genes to, a number of different cell types - a theoretical advantage that can be a headache in the clinic. Schaffer said that "at this time, one injects the virus in the general area where they want genes delivered, and then the virus enters a wide variety of cell types. It would be better to engineer a virus that can target a specific, desired cell type, which is a problem that we are applying our approach to solve."

Additionally, the soluble heparin also can bind and prevent it from infecting any cell at all. Using their directed evolution technique, Schaffer and his colleagues were able to isolate an AAV mutant that binds heparin less strongly that the wild-type; in practice, this mutant still was able to infect cells while binding to soluble heparin less strongly than its wild-type cousin.

Schaffer’s plan is to bring his research to the small-animal stage and then collaborate with industry for clinical development.

"That’s always been an interest of mine, to make an impact in the clinic," he said. "But it makes sense to let industry do that because that’s what it’s good at."