By Lisa Seachrist

Washington Editor

The challenge in science is often finding order in the midst of chaos. That's precisely what Cyrus Safinya and his colleagues achieved when they began to look at how liposomes and DNA interact. As a result, they have generated new insight into the biggest problem facing gene therapy: delivering genes into cells.

"The Human Genome Project is leaps and bounds ahead of attempts at gene therapy," said University of California at Santa Barbara biophysicist Safinya, who noted that neither of the current methods of gene delivery, using viruses or synthetic molecules, has been optimized.

Since 1987, scientists have known that they could introduce DNA into cells using cationic liposomes, lipid strings with positive heads that form a bilayer very similar to cell membranes. Cell membranes, however, carry a negative charge so the positive liposomes will stick to the membranes and presumably allow the DNA to enter the cell. Because DNA is negatively charged, the assumption was that the positive head of the cationic lipid simply lined up on the "string" of the DNA helix.

"We thought we had a very uncomplicated, unstructured interaction," said Safinya. "But, as it turns out, like most biology, the underlying structure is very important and complex."

Safinya and his colleagues mixed cationic lipids, neutrally-charged lipids and DNA in various concentrations and examined the resulting liposomes with fluorescent in situ microscopy and X-ray diffraction. Instead of beads on a string, they found molecular sandwiches of lipid bilayer and DNA.

"These structures were highly-ordered, self-assembled sandwiches," Safinya said.

As the researchers report in today's issue of Science, when they varied the concentrations of the liposome components, they found that the structure of the molecule sandwich changed. At certain concentrations, the distance between the DNA layers was as close as 2 angstroms or greater than 30 angstroms.

"By simply varying the ratio between lipid and DNA, we can determine the structure of these liposomes," Safinya said.

One of the biggest problems with delivering DNA to cells via liposomes is that the procedure is tremendously unpredictable. Safinya and colleagues have begun to look at whether tightly packed or more loosely packed cells enter with enhanced efficiency. "We have very preliminary evidence that neither extreme is the most efficient," Safinya said.

Safinya did note in the paper that, for reasons still unknown, the distance between the DNA layers did not progressively increase but took a jump from 35 angstroms to 45 angstroms.

"After the best transfection parameters have been determined, we could look to including signals that would direct the DNA to the nucleus where it can produce protein," Safinya said.

"What this work does is allow us to begin to rationally investigate how these components interact and to optimize it," said Joel Schnur, director of the Center for Bio-Molecular Science and Engineering at the Naval Research Laboratory, in Washington.

Schnur pointed out that the work is exciting not only in reference to the obvious medical applications for these lipid/DNA complexes, but also because DNA is beginning to be seen as biomaterial. "People are looking into using DNA as a biosensor and this type of information may facilitate the development of this technology," he said.

However, Schnur added that much more work on lipid/DNA complexes need to be done with more precise probing methods in order to understand why the complexes form the way that they do.

Nevertheless, Safinya noted that several pharmaceutical and biotech companies already have liposome programs and that he has consulted with several of them. *