Science: A new twist on the structure of DNA

DNA is a double helix - isn't it? Yet a young artist reckons that for years science has had the wrong structure for it. He went back to first principles to produce a new view of the most basic molecule of life. Caspar Henderson explains.

The double helix of DNA has become a cultural icon - the book of life, made manifest in a single structure. Its image can be found everywhere from corporate logos to drive-in restaurants. Its very ubiquity implies that everything is known, at least about its shape.

Yet now an artist has suggested a modification to Crick and Watson's pioneering model of 1953. In The Geometry of DNA: a Structural Revision, an exhibition that opens at The Blue Gallery in London tomorrow, Mark Curtis presents drawings, models and paintings of a "more geometrical" DNA that is striking in its simplicity and beauty.

In his modified model the four base pair "building blocks" (adenine, cytosine, guanine and thymine) are joined to their respective complementary bases the opposite way round to that hypothesised by Crick and Watson, in such a way as to create two pentagons, rather than a single octagonal shape, as the core axis of the double helix .

Questioning something so well established is enough for many to brand Curtis as at best ignorant, at worst a crank. After all, for more than 40 years of genetics, science has worked quite happily with the Crick and Watson model.

Not surprisingly, because Curtis is "just" an artist, he has had trouble getting scientists to examine his proposals in detail.

Curtis is undeterred. He says that not only will the model stand up to scrutiny, it may also provide a superior level of understanding to what we already know.

He explains that his work on the nature and depiction of visual space led him to a series of drawings of DNA in the manner of Renaissance perspectival artists such as Paolo Uccello. Using the standard textbook dimensions of the double helix, Curtis says, it struck him that it did not conform to any set of geometrical principles. But a series of relatively simple geometrical relationships underlies every element of the alternative he suggests. And surely nature loves geometrical simplicity? After all, the mathematical Fibonacci ratio is found in the positioning of plant leaves on a growing stem. Why shouldn't something similar be true of DNA?

In the accepted model, the double helix depends on sugar phosphate groups attached on the outside of the base pairs to form an external "backbone", which provides the super-molecule with an overall structural integrity. But Curtis suggests that it is the base pairs themselves that provide that integrity, and he uses geometry to show how it could be done.

Curtis's work shows how 10 joined pairs of regular prismatic pentagons stacked stepwise on top of each other, and turning by one face at each step, produces a single, full turn of a helix in 10 steps. With 10 base pairs per turn of the helix, he was puzzled that the accepted model joins the purine (guanine and adenine) and pyrimidine (thymine and cytosine) base pairs with the pentagonal ring of the purine base oriented away from the central axis.

If, instead, he theorised, the base pairs are reoriented, then two pentamers could be placed at the core of each pair. This modification to the base pairing could exist in both the enol form (where a hydrogen atom is attached to an oxygen) and the keto form (in which the hydrogen is attached to a nitrogen atom).

The pentagonal geometry provides the infrastructure required to make a consistent, stable and uniform helical structure, and also establishes why there should consistently be 10 bases within a single turn of the helix. The hollow centre found in this formation ensures that the bases remain side by side within the structure, allowing access for replication and repair from any point within the helix.

Unfortunately, coming up with an interesting idea is only the first part of any battle to have it accepted. Some scientists who have seen his model say it is plain wrong.

"He has got carried away with the beauty of what he's doing," says Professor Richard Henderson, director of the Medical Research Council's Laboratory of Molecular Biology in Cambridge. He says the idea is "nice, but it doesn't help science at all".

The structure of DNA has been conclusively and objectively determined using X-ray crystallography, he maintains. Images published in 1981 swept away any lingering doubts: every detail of the structure was clearly visible, and Crick and Watson's hypothesis was shown to be correct, he says.

Nevertheless, Curtis is undeterred. He points out that the images published in the early Eighties were of synthetic DNA, built up in the laboratory from just 12 individual base pairs, with the sugar phosphate groups preattached to each base pair so that they were constrained to bond in the Crick and Watson configuration. Such a structure may be sustainable, says Curtis, but that does not necessarily mean it is correct. "Unless they can show me an image from natural DNA confirming Crick and Watson's hypothesis, I will not be convinced".

And he may have an ally. Maurice Wilkins, who shared the 1962 Nobel Prize for the structure of DNA with Crick and Watson, says Curtis's proposal should not be dismissed out of hand.

Although he does not think the geometrical model is correct, Wilkins welcomes the challenge. "Scientists need to step out of their narrow scientific shoes to look at aesthetic issues and the ethical and spiritual questions which are inseparable from them," he says.

Is Curtis a new Copernicus, about to make us view our world from a new perspective? Or a new Columbus, discovering something new that he mistakes for something old? The reassuring thought is that, either way, DNA continues to work. It's just our understanding of it which is incomplete.

Mark Curtis's work is exhibited from tomorrow at The Blue Gallery, which can be contacted on 0171-589 4690.