The Double Helix Doubles Up

So familiar is the double-stranded structure of DNA that biologists often refer to the molecule simply as the double helix. But a paper in this week's Proceedings of the National Academy of Sciences shows that in the test tube, the double helix can sometimes get doubled up, forming four-stranded DNA. Together with an earlier glimpse of a four-stranded structure, it suggests that quadruple DNA may play a role in living cells, say the researchers.

Scientists already knew that DNA strands with unnaturally repetitive sequences, such as all C bases or many G's and C's in a row, can clump into four-stranded structures in the lab. Two years ago, a British research group found that a more diverse sequence of seven nucleotides could also form a foursome. Even so, many biochemists argued that the unnatural laboratory conditions--the researchers had to crystallize a concentrated solution of DNA to detect the unusual structure--could explain the four-stranded configuration.

Now Stephen Salisbury and his colleagues at the Cambridge Crystallographic Data Center in the United Kingdom, the University of Göttingen in Germany, and the University of Barcelona in Spain have coaxed an entirely different DNA sequence into a four-stranded structure. The four "strands" are really the ascending and descending portions of two different loops, each one eight bases long. Two AT sequences on opposite sides of one loop bind to complementary TA sequences on the matching sides of a second loop, and a positively charged sodium ion unites the foursome by pulling together four negatively charged oxygens, one on each of the four T's.

Salisbury says that because the structure has the same basic characteristics as the earlier one, in spite of the sequence difference, four-stranded motifs might well turn up in the DNA of living cells. "If you get one of these structures, you might think it's some bizarre effect of the high concentration of nucleotides in the crystal lattice," says Salisbury. "But we're arguing that now the balance is in favor of this as a natural structure."

The most likely setting for four-stranded DNA to form, he and his colleagues say, might be the genetic sorting that takes place during the formation of sperm and eggs. During this process, matching chromosomes--each one made of double-stranded DNA--pair up and swap parts, and the four-stranded motif may help the complementary sequences align. But Walter Chazin, a molecular biologist at the Scripps Research Institute in La Jolla, California, says the odds are against four-stranded DNA in nature. In order for such four-stranded complexes to form, two DNA duplexes must first partially unzip to allow all four strands to knit together. "In real genes, their mechanisms won't happen because [DNA] duplexes are too stable," he says. Salisbury says his group hopes to settle the question by finding the biloop four-stranded structure in living cells.