The findings, published in the March 27th issue of the Proceedings of the National Academy of Sciences, begin to answer an important question: Does an inserted package of genetic material actually incorporate itself into the receiving cell's DNA?

Scientists had worried that if the genetic material did integrate into the cell's own genetic wiring by becoming part of a chromosome, it potentially could disrupt the function of a healthy gene or lead to tumor development. The new research shows that in mice, new genes delivered inside a modified adeno-associated virus dwell independently in cells, reducing the risk of such problems.

"Even though we have never seen any problems from the AAV vector we use for gene therapy, we needed to know where the delivered DNA is going within the cell," said Sihong Song, lead author of the paper and a research assistant professor in the UF College of Medicine's pediatrics department. "We need to understand how it really works, so that we can design safer and more effective gene therapy vehicles."

Composed of DNA, genes are the basic unit of inheritance and provide the instruction manual for how to make the body function. If a person has inherited a faulty version of a critical gene, lifelong disease may result. In gene therapy, researchers seek to reduce or eliminate disease symptoms by providing a patient with working copies of a corrective gene so that cells in the affected tissues may begin following a new set of marching orders.

At UF and elsewhere, a growing number of laboratory experiments and clinical trials are employing AAV to essentially "infect" cells with corrective DNA.

A common virus that has not been linked to any illness, AAV can be modified so that it includes a wide variety of genes in an effort to treat numerous conditions, such as cystic fibrosis, hemophilia, and genetic forms of blindness and heart and lung diseases.

"The main potential safety problem in gene therapy comes from the vector rather than from the gene being delivered," said Dr. Terence R. Flotte, an associate professor of pediatrics, director of UF's Genetics Institute and co-director of the Powell Gene Therapy Center. "The strength of our own center has been the focus on this one particular vector that doesn't cause the side effects, such as inflammation, seen with other vectors.

"But you always have to consider the side effects that could occur," Flotte said. "Theoretically, the greatest potential risk is that the DNA from the virus will insert into the DNA of the cell and in doing so change the characteristics of the cell from being a normal cell to a cancer cell. One way this might occur would be if the new DNA was able to effectively switch on a previously silent tumor-causing gene."

In studies in people and animals conducted during the past decade, UF scientists have never seen tumor development associated with AAV-delivered gene therapy, but they wanted to make sure that wasn't an extremely rare side effect they simply hadn't run across.

In its natural state, AAV does incorporate into a host cell's chromosome, but almost always at the same specific site, where it apparently causes no harm.

However, in its recombinant form - in which much of its natural DNA is spliced away and what remains is combined with a corrective gene - AAV no longer zeroes in on the same target, Flotte said.

Instead, UF's experiment showed that a protein in the receiving cell helps "tie up" the ends of the recombinant AAV so that it acts as if it were a tiny independent chromosome within the cell's nucleus, Flotte said. The protein -- DNA-dependent protein kinase -- is abundant in human cells.

"That puts it into a relatively safe form in which the therapy can work but where it's not likely to trigger the formation of a tumor," he said.

The scientists discovered the protein's role in the process by comparing gene therapy in two strains of mice. A mutation in one of the strains prevents mice from producing the protein.

"In the absence of this protein, over time, much of newly inserted DNA integrates into the host cell's chromosome. Theoretically, that's the kind of event that in some small percentage of cases would be harmful," Flotte said.

"But in the presence of this protein, the new DNA stayed in this independent string form. We saw no detectable integrations. These experiments were conducted in mice, rather than people, but from these mouse models we've been able to get a very good understanding of cellular mechanisms of gene therapy."

Song noted that in other experiments using human cell cultures, the protein also prevented AAV from integrating into a chromosome.

Dr. Kenneth I. Berns, UF's vice president for health affairs and an early pioneer of using AAV as a gene therapy vector, and Philip J. Laipis, a professor and associate chairman of the College of Medicine's biochemistry and molecular biology department, also contributed to the research effort. - By Aaron Hoover

[Contact: Aaron Hoover, Victoria White]

09-Apr-2001