Deconstructing DNA

Could Watson and Crick really have been wrong about the structure of DNA? Ayala Ochert meets the artist who has created an astonishing alternative. Published in New Scientist.

SALVADOR Dali, the surrealist painter, had a fascination with DNA and included it in several of his works. He was, like many who followed him, in the thrall of the double helix, calling it “a Jacob’s ladder of genetic angels, the only structure linking man to God”. So it was by no means odd when, three years ago, a British artist named Mark Curtis chose to embark on his own series of drawings and paintings of DNA. But the outcome of his artistic journey is truly radical. If you thought the structure of DNA was a closed book, sealed by James Watson and Francis Crick in 1953, then it may be time to think again.

Seeing his artistic venture as an “investigation into the nature and depiction of visual space”, Curtis never intended to prove anyone wrong. But as sketches and rudimentary physical models began to clutter up his studio, Curtis became at first puzzled, then doubtful, and finally horrified, as he found that the structure proposed by Watson and Crick, and accepted universally by biologists, fails to conform to what he calls “geometrical principles”. Fortunately, wonder has now replaced horror. For Curtis has discovered a striking alternative structure that he claims works far better than Watson and Crick’s.

Is it really possible that scientists have been wrong about DNA for 45 years? Could a lone artist working with pen and pencil, with rope and wooden blocks, discover something that tens of thousands of scientists with computers and microscopes have overlooked? You might think that Curtis must somehow be sadly mistaken. But the story may not be quite that simple.

Traditionally, DNA takes the form of a long twisted ladder. The two legs of the ladder swirl round one another, connected by rungs made of the “base pairs”, the carriers of genetic information. Each rung is made of either the molecule adenine paired up with thymine, or guanine teamed with cytosine. The long twisting legs of the ladder are chemically unrelated to the bases, being built of sugar and phosphate groups linked together into a strong “sugar phosphate backbone” that holds the double helix in shape.

Curtis’s new structure is also a double helix. And in pairing the bases, he has also put adenine with thymine, and guanine with cytosine. The difference comes in the chemistry of fitting the bases together. Curtis has split the Watson-Crick base pairs apart, and reconnected them in a different way, so that the edges that traditionally face outwards now face inwards (see molecular structures). Using these modified base pairs, Curtis has built a totally new double helix with a simpler and more subtle structure.

Whereas the Watson-Crick structure gets its shape from the sugar phosphate backbone, Curtis has found a way to construct the double helix solely from the base pairs themselves. The result is a structure in which the bases don’t sit on opposite sides of the double helix, but gather together more closely on one side. In the Watson and Crick structure, the bonds between base pairs-the rungs of the ladder-cut right through the central axis of the DNA chain. But the twisted ladder picture does not apply to Curtis’s structure. With the base pairs shifted aside, his double helix has a more “open” construction, and a tubular space running down through its centre.

Curtis was driven to create this strange structure by his early attempts to draw some rough sketches of DNA. After many attempts, he found he couldn’t make sense of the two and three-dimensional representations he found in textbooks. They seemed to contradict one another. Frustration finally forced him back to what he refers to as “geometrical first principles”, and in the manner of Renaissance perspectival artists such as Paolo Uccello, he began by making careful scale drawings of DNA.

DNA has ten base pairs to each turn of the helix, so, using these centuries-old methods, Curtis began with a decagon-a 10-sided figure. He then placed a pentagon for each base around it (see Diagram). The pentagon is the only regular shape that can fit round a decagon in this way without leaving any space. And according to Curtis, this figure of 10 pentagons oriented about a decagon is the only geometrical configuration that enables one to create a helix with the known dimensions of DNA. Add some thickness to each of the pentagons, and the 10 pentagons telescope out to form a helix. To build a structurally sound double helix, Curtis reasoned that he would need not just one pentagon, but two joined together . With this insight he was finally able to produce the drawings and paintings that he had originally planned.

This kind of reasoning may seem obscure by modern standards. But for Curtis, it established that a double helix of 10 turns had to be made of twinned pentagons. At this point in his work, he was still thinking in terms of the traditional Watson and Crick structure. But he faced a contradiction: in the standard Watson and Crick chemistry, the base pairs join through hexagonal regions in their structures. Then came Curtis’s moment of truth: “I sat down on the sofa one night and I thought, hang on a second, the molecular structures of the bases also have pentagons in them. And here was I, with two pentagons, building a consistent helix that conformed to the dimensions of the DNA double helix.” By connecting the bases differently, Curtis found that he could naturally form pairs of pentagons in each base pair (see Diagram). “It was placed in my lap. I wasn’t trying to prove anybody wrong. I wasn’t even thinking then that they’ve got it wrong. I was just playing, like artists play.

Five-a-side
Once the bases were arranged in their new configuration, the new structure simply emerged of its own accord. And, for Curtis, this makes it a “better” structure: Watson and Crick established that DNA is a double helix, but Curtis believes he has established why it should take that shape.

So Curtis may have captured beauty, symmetry and simplicity in his creation. But is it really DNA? According to the experts, the answer is a clear and simple no. Richard Dickerson of the Molecular Biology Institute at the University of California, Los Angeles, who has spent the last 17 years working on X-ray crystal structures of DNA, says of Curtis’s structure: “I like it, and it’s within an artist’s licence to interpret DNA in this way, but scientifically it’s wrong. There’s not the slightest question about it.” The reason he is so certain boils down to this-if you break DNA down into its constituent nucleotides (base plus sugar plus phosphate), and study their structures using X-ray crystallography, the results show that the bases pair up Watson and Crick style, not Curtis style. So if Curtis starts with the wrong building blocks, says Dickerson, the final structure can’t be right.

Most other scientists seem to agree. According to Alan Mackay, a crystallographer at Birkbeck College in London, Curtis’s approach is just wrong-headed: “People can get carried away by the geometry-they say, it’s so beautiful it must be right.”

For most scientists, then, X-ray crystallography sounds the death knell for Curtis’s creation. But Curtis insists that he has stumbled upon something of profound importance, and has gained some support from a most unlikely source-Maurice Wilkins, the third man in the DNA story, who shared the Nobel prize with Watson and Crick for their 1953 discovery. “It all depends how you look at science,” says Wilkins, of King’s College, London. “If you take a rather narrow, professional view of the area, then you may conclude that Curtis’s structure is of no value. But from a wider view, it certainly is interesting and requires careful attention.”

But what then is its value? If the structure isn’t DNA, can it ever be anything more that a figment of one artist’s imagination? One possibility is that this hypothetical structure mirrors something that could in principle exist in the real world. Curtis’s approach has a certain resonance with the attitude of the great physicist Paul Dirac. “It is more important,” Dirac felt, “to have beauty in one’s equations than to have them fit experiment. If one is working from the point of view of getting beauty in one’s equations . . . then one is on a sure line of progress.” The beauty in Dirac’s equations led him to predict the existence of antimatter before it was ever discovered in the lab. Is it possible that in the pursuit of aesthetic ideals, Curtis has stumbled onto a strange form of DNA that has yet to be discovered?

Since the 1950s, DNA has proved itself to be a far more complex beast than scientists first imagined. Many forms have been found-circular DNA in bacteria, the left-handed zig-zag helix of Z-DNA. Another peculiar form of DNA may have come up just a few months ago. David Bensimon, Vincent Croquette and their colleagues at the Ecole Normale Supérieure in Paris, working with Richard Lavery of the Institute of Biology and Physical Chemistry, also in Paris, attached one end of a double-stranded DNA molecule to a glass slide, and the other to a magnetic bead. Then they used a magnet to spin the bead and twist the molecule as if it were an elastic band. DNA has a natural twist, but this approach let them study its behaviour when twisted more than usual. They found that whenever the twist on the molecule exceeded 3 per cent of its natural number of turns, it appeared to change suddenly, in some regions, from the ordinary structure in which the base pairs sit inside the external scaffolding of the sugar-phosphate backbone, to one in which the backbone slips inside and runs down the axis of the molecule and the bases, now unpaired, stick out to the sides. Although their evidence is incomplete, it seems as if a twisted strand of DNA can spontaneously change into this other structure so as to reduce some of the stress placed on it. Bensimon and his colleagues believe that the mechanical stress required to generate this “P-DNA” structure might occur in DNA transcription, when the base pairs need to open up for easy reading.

American researchers found a similar structure in viral DNA in 1994. There it was caused not by mechanical twisting, but by the effect of a coat of proteins glued to the outside of the DNA strand. Researchers have also claimed to find triple-helical and parallel-stranded DNA. Could Curtis’s proposed structure be yet another variation? “It has a lot of the properties you would want in a DNA structure,” says Robert Root-Bernstein, a physiologist at Michigan State University with a keen interest in DNA. “It would be quite ridiculous not to consider something reasonable like Curtis’s model. It has novel properties that we need to understand generally, whether we’re biologists or architects.”

Apart from having the right dimensions and base-pairing, Curtis’s DNA has another intriguing feature. Although its overall structure is independent of the sequence of bases, the details of the molecule’s surface are sequence-specific. This is fundamental to DNA as proteins must be able to “recognise” sequences of bases and bind to them, switching genes on or off. Details of this recognition process in standard DNA are only now being established. But Curtis’s structure seems to allow for the same kind of process.

Examples of symmetry abound in both nature and our models of it, so this alone makes Curtis’s double helix intriguing. But this structure isn’t just beautiful or symmetrical, it is also conspicuously functional. Perhaps, as molecular biologists seem certain, this isn’t ordinary DNA. But if it’s not, then what is it?