By AMY ADAMS
Researchers at the medical center have devised a gene therapy technique that eliminates one of the major health risks of this treatment approach. The technique, published in the Oct. 15 advanced online edition of the journal Nature Biotechnology, overcomes the need for viral vectors that have plagued gene therapy trials, while retaining the ability to insert therapeutic DNA into specific sites on the chromosome.
"Our approach provides an alternative that didn’t exist before," said Michele Calos, PhD, associate professor of genetics and the study’s lead author.
The goal of gene therapy is to insert a healthy copy of a gene into a cell where it can take over for a faulty version. If the therapeutic DNA does not integrate into the human chromosome, it produces its protein for a short time before being turned off or broken down within the cell. For a long-term cure, the gene has to wedge itself into a chromosome where it gets passed on when the cell divides.
Current gene therapy approaches that cause genes to integrate use a viral vector to sneak the therapeutic DNA into the host cell, Calos said. However, the DNA inserts itself at random positions. In one recent French gene therapy trial, the randomly inserted DNA activated a neighboring oncogene, apparently causing a patient to develop leukemia. "That sort of puts another cloud over the existing gene therapy trials," Calos said.
Calos’ technique avoids the pitfalls of other gene therapy approaches by integrating DNA without using viral vectors, inserting the DNA at known locations. This new technique can also handle genes that are too large to fit into a viral package, such as the gene for Duchenne’s muscular dystrophy, Calos said.
In developing this approach, Calos hijacked a mechanism used by a bacteria-infecting virus to integrate its genes into bacteria. The virus makes a protein called integrase that inserts the viral genes into a specific DNA sequence on the bacteria chromosome. It turns out that humans also have a version of that DNA sequence. When the researchers insert a copy of the therapeutic gene and a gene coding for integrase into a human cell, the integrase inserts the gene within the human sequence.
Calos and her team, in collaboration with Mark Kay, MD, PhD, professor of pediatrics and of genetics, tested the technique using a gene that makes Factor IX — a protein that is missing in the blood of people with one form of hemophilia. They injected mice with a piece of DNA containing the Factor IX gene plus a stretch of DNA that acts as an "insert me" signal to integrase. At the same time, they injected a gene for integrase.
Within a week, the injected mice made 12 times more Factor IX than their littermates that did not receive integrase. Further experiments confirmed that the Factor IX gene had successfully integrated into the mouse DNA.
Although the mouse genome contains at least 53 potential integration sites, Calos and her team found the Factor IX gene in only two spots, one being by far more common. She said that for each tissue there may be a particular site that is the most likely insertion point. Her group is testing the technique in different tissue types to ensure that no human integration site is near a potential oncogene. "We need to look in different tissues to see where the hot spot is," Calos said.
Calos is also modifying the integrase so it targets specific integration sites that her team knows are safe. "We mutated the enzyme and evolved it so it will prefer one place over another," she said.
Calos said this approach should be effective for treating diseases in several different human organs including skin, retina, blood, muscle and lung. She hopes to start human trials for the technique in a fatal childhood skin disease called recessive dystrophic epidermolysis bullosa, which she has already treated in mice. "If that trial shows that it is safe then that will open the door for trials in other diseases," Calos said. She has collaborations under way testing the technique for use in Duchenne’s muscular dystrophy and cystic fibrosis, among others.
Contributing researchers to the study include graduate students Eric Olivares and Thomas Chalberg and postdoctoral scholar Roger Hollis, PhD.
Stanford Report, October 16, 2002