MODEL COMPOUND Watson (left) and Crick pose with their model of DNA in 1953. A. BARRINGTON BROWN/PHOTO RESEARCHERS INC.

Although the B-form DNA structure that James D. Watson and Francis H. C. Crick proposed in their April 25, 1953, letter to Nature made only a limited splash at the time, the subsequent ripples revolutionized biology--particularly genetics--turning it into a molecular science. People pretty much agree about the impact that the double helix has had on biology, but what about chemistry? How have chemists been inspired by this structure?

For one, Watson and Crick's model for DNA structure demonstrated that chemists could play a role in solving biological problems.

Before the a-helix model suggested by Linus Pauling for proteins and the double-helix model for DNA, the view of the structure of biological macromolecules was "blurry," says Albert Eschenmoser, a chemistry professor at Scripps Research Institute and emeritus professor at the Swiss Federal Institute of Technology, Zurich. "Pauling's -helix and Watson-Crick's double helix were the first time that scientists could look at the structure of biopolymers as closely, as sharply, as organic chemists had done at the time with the small molecules."

	RECOGNITION A polyamide dimer containing four pairs of rings (Im/Py, Hp/Py, Py/Hp, and Py/Im, where Im = imidazole, Py = pyrrole, and Hp = hydroxypyrrole) can read the minor groove of the DNA double helix and distinguish each of the Watson-Crick pairs, G-C, T-A, A-T, and C-G. COURTESY OF PETER DERVAN

FOR CHEMISTS, the adage about a picture being worth a thousand words is true. "Many scientists, myself included, design and interpret their experiments based on pictures they carry around in their heads of the molecules that are involved," says Thomas R. Cech, who received the Chemistry Nobel Prize in 1989 for his role in discovering the catalytic properties of RNA. Cech is president of the Howard Hughes Medical Institute, and professor of chemistry and biochemistry at the University of Colorado, Boulder.

"The better picture you have, the better experiment you can design to figure out how nature works and the better you can interpret your results," Cech says.

"Watson and Crick provided the picture at just the right level of detail that allowed this whole field to explode," he says. "Until a biological phenomenon is defined at the chemical level, it's very hard for the chemist to get interested or involved. But the moment you have a specific structure of sugar, phosphate, and heterocyclic base, all of a sudden you're in the world of chemistry."

Eric T. Kool, a professor of chemistry at Stanford University, agrees. "Chemists often delve into biological problems once they know enough chemical information to get their foot in the door. That [DNA structure] was the foot in the door for chemists. I can speak for most organic chemists. They love to look at pictures. This was the picture that really said this is a chemical problem."

Interest in DNA chemistry predates Watson and Crick's structure. H. Gobind Khorana, emeritus professor of chemistry and biology at Massachusetts Institute of Technology, was trained as an organic chemist and was already working on ways to synthesize dinucleotides and oligonucleotides in the early 1950s.

"I began to think about nucleotide synthesis and hooking two nucleotides together to make an internucleotide bond as a starting point in dinucleotide and oligonucleotide synthesis," Khorana recalls. "It was very clear at this time that one had to start this work to put nucleotides in a specific sequence." He developed the first practical, but laborious, technique for synthesizing nucleic acids, known as the phosphodiester method.

"All my time in the 1950s was spent making oligos, methods that we developed in Vancouver, British Columbia, and later in Madison" at the University of Wisconsin, Khorana says. He painstakingly combined his chemical techniques with enzymatic methods developed by Arthur Kornberg to synthesize a 126-nucleotide transfer-RNA gene. Khorana shared the 1968 Nobel Prize in Physiology or Medicine with Robert W. Holley and Marshall W. Nirenberg for their contributions to deciphering the genetic code. Later, in the 1960s and early 1970s, Robert L. Letsinger at Northwestern University, Marvin H. Caruthers at the University of Colorado, and others developed methods of solid-phase synthesis of DNA that are still used in modern automated DNA synthesizers.

"Chemical synthesis of DNA is the fundamental technology that has led the molecular biology revolution," Cech says. DNA synthesis of probes and primers is a necessary first step for DNA sequencing, genetic engineering, and the polymerase chain reaction. "All of a sudden, it was up to the chemist to provide the essential tools to allow the field to fly forward," Cech says. "Chemists provided not just little trinkets for the biologists, but actually the core technologies that enabled molecular biology and biotechnology to go forward."

Peter B. Dervan, a chemistry professor at California Institute of Technology, believes that DNA research can reveal new chemical principles. "We chemists are different from biologists and physicists in that in addition to studying and trying to unravel principles of how the natural world works, we are also inventors. In the process of trying to create new matter, new materials, we come up against the limitations of our field. It helps define the next chemical question where our understanding may need to be enriched," Dervan says. "In the exercise of trying to invent or discover new materials that would mimic a biological system, we get to phrase new questions and sometimes arrive at unanticipated discoveries that are inherently chemically interesting."

DNA HAS INSPIRED Dervan to ask fundamental questions about molecular recognition. His challenge was to invent a new polymer code that could read the edges of the Watson-Crick base pairs in either the major or minor groove of the helix.

"Our first effort to create a language different from nature's proteins was to use a triple helix in the major groove. We could not push beyond a two-letter code," Dervan says. His team was able to recognize the purines--adenine (A) and guanine (G)--but couldn't recognize the pyrimidines cytosine (C) and thymine (T). "It just showed the limitations of our understanding of molecular recognition," Dervan says.

In parallel to the triple-helix work, Dervan also investigated the minor groove as a recognition site. "The breakthrough was an unanticipated discovery. We discovered that an unsymmetrical pair of imidazole and pyrrole stacked side by side would distinguish GC from CG." Dervan's students invented a new ring pair of hydroxypyrrole and pyrrole that distinguishes TA from AT.

	UNNATURAL PAIR Solution structure (measured by high-resolution NMR) of the nonpolar thymine analog paired opposite adenine in a DNA duplex. Such analogs that don't form Watson-Crick hydrogen bonds have been useful in the study of DNA replication and repair. COURTESY OF ERIC KOOL

Kool was drawn to DNA research in the 1980s by a talk given by Dervan that "totally captivated" him. "I was fascinated by the possibility that suddenly this is organic chemistry. It's not just biology, but it is organic chemistry," Kool says.

Kool is using DNA as a launching point to design new base pairs that don't hydrogen bond but retain some of the properties of DNA. In natural DNA bases, the purine adenine forms strong hydrogen bonds exclusively with the pyrimidine thymine. In a similar fashion, the bases guanine and cytosine hydrogen bond with one another. Kool mimics the shape but not the bonding ability of DNA bases by keeping the shape and size as close as possible to those of the original base but removing the strongly polarized atoms and functional groups. Kool's mimics are much less polar than natural DNA bases.

The DNA mimics have revealed when the Watson-Crick hydrogen bonds are important and when they are not. "In the basic double helix--the two strands binding to one another in a specific way--we still believe that Watson-Crick hydrogen bonds are pretty important," Kool says. "Our molecules make it fairly clear that Watson-Crick hydrogen bonds are important for assembling the helix and keeping it together."

Surprisingly, however, the hydrogen bonds turn out to be less important when enzymes synthesize new DNA strands. "It's quite clear that a number of polymerase enzymes--the basic replicators of nature--can copy DNA base pairs highly accurately and quite efficiently without any Watson-Crick hydrogen bonding in a given base pair," Kool remarks. "You can't remove them all forever, but you can remove some of them in an isolated way and it still works. Although the hydrogen bonds are necessary to help hold the helix together, the enzymes that copy DNA don't really care about them very much."

The mimics do not easily pair with opposite natural bases. "Despite the fact it looks like the natural base, a mimic of adenine does not especially like to pair opposite thymine. When you pair them in this way, it's strongly destabilizing to the DNA," Kool says. "If you replace the partner of one of those modified bases with another modified base, it's not as destabilizing. We have a couple of cases where those base pairs are as stable as a natural base pair."

Kool uses these new bases as structural probes. Because the bases don't change the overall shape of the DNA, they can be used to understand the energetics and mechanisms of a variety of interactions, such as enzyme recognition of DNA or nucleotides, protein bending or binding of DNA, or water interaction with DNA. "We're studying various DNA polymerase enzymes and various DNA repair enzymes using these molecules as probes of their mechanisms."

On the more applied side, Kool is using modified DNA to design diagnostic tools. For example, he is designing molecules that fluoresce in the presence of specific genetic sequences. "We have DNA [molecules] that can join themselves together to create a longer DNA. When this bond is formed, it releases a fluorescence quencher and causes the molecules to light up with a specific color. They can only join together when the correct genetic sequence is in the cell." He anticipates that such assays could be used for genetically typing bacteria or cancer mutations.

INORGANIC CHEMISTS have also been drawn to DNA. For example, Stephen J. Lippard, chemistry professor at MIT, first became interested in DNA in the 1960s, when he worked on methods to sequence DNA using metal atoms and electron microscopy. "Let's say we wished to identify the position of guanine nucleotide in DNA," Lippard explains. "The idea was to attach a heavy atom label to every guanine and then spread the DNA out on an electron microscope grid and use the electron microscope to read where the heavy atoms were. The process would be repeated for the other nucleotides. The metal-DNA labeling chemistry was successful, but the electron-beam energy applied to the microscope grids made the heavy atoms move so that the picture was too fuzzy to ascertain the sequence.

However, the experience with metal-nucleic acid chemistry was valuable when Lippard turned his attention to cisplatin, an anticancer platinum compound that targets DNA. His group has been able to determine the structure of the intrastrand cross-linked adducts that form between DNA and platinum.

"The current model that we are using to explain the anticancer activity is that the formation of specific adducts blocks transcription on DNA," Lippard says. "When the RNA polymerase gets stalled at the platinum cross-link, it triggers a variety of cellular responses that lead to the death of the cancer cell."

Lippard's group is now working on "groove-crossing" compounds. "The platinum forms its major adducts by binding to two guanines in the DNA major groove," he says. "The proteins that bind to platinated DNA tend to recognize the widened shallow minor groove that occurs as a consequence of the platinum adduct formation." Such protein binding inhibits repair of platinum adducts and sensitizes cells to cisplatin. Lippard would like to design a single molecule that could simultaneously form guanine cross-links in the major groove and fill the minor groove, thus interfering with repair processes and still blocking transcription.

Lippard is now studying the repair of platinum adducts in larger, more complicated DNA structures, such as nucleosomes, which consist of DNA wrapped around a histone core. "These are very complicated molecules to synthesize," he says. "We can't use genetics to put platinum in DNA. We have to make the DNA, platinating it in a specific place as part of the synthesis, and then reconstitute it into higher order structures. By doing that, we believe we're getting closer to understanding how the cell really processes DNA."

Jacqueline K. Barton, a chemistry professor at Caltech, was originally drawn to DNA because of its structure. "When I was starting graduate school, I was very interested in beautiful structures. DNA has an absolutely beautiful structure, and it has this interesting characteristic of being a polymer as well, a polymer of p-stacked base pairs."

BARTON IS AN EXAMPLE of a chemist who was inspired by DNA's structure to ask chemical questions. "We started out asking a fundamental chemistry question: How are electrons or electron holes transported through a DNA helix? Is this molecule particularly facile for electron transfer and transport chemistry? To the first order, that question has nothing to do with biology," Barton points out. "What we found is, in fact, because of the stacked base pairs, that structure does indeed facilitate charge transport, and it's really sensitive to the structure. From that point, we've come up with applications, such as developing new DNA-based diagnostics to look at mutations and lesions in DNA. Now we're trying to understand the biological consequences of that chemistry."

A large part of Barton's work involves developing molecules that can recognize various sequence-dependent conformations of DNA. An early example of such work involved chiral metal complexes in which right-handed metal complexes preferentially bound to the right-handed DNA helix, taking advantage of DNA's simplest recognition element, the handedness of its helix.

"Now we're interested in designing molecules that recognize specific sites. We're interested in seeing if we can target mismatches, mistakes in DNA, in different kinds of ways," Barton says. "We're taking advantage of the structural and thermodynamic distortions that are associated with mismatches to see if we can design transition-metal complexes that allow us to target some sites on DNA with specificity. It turns out that analogs of these complexes also carry out interesting electron-transfer chemistry. We've exploited these metal complexes as probes of that charge-transport chemistry on DNA."

With the metal complexes, Barton has found that DNA is capable of long-range charge transport. "I could bind a metal complex to site A, use that metal complex to inject a hole into the DNA, and oxidatively damage DNA at a distant site," she says. "We've been able to demonstrate this oxidative damage at a distance as long as 200 Å."

A number of younger chemists and biochemists who cut their teeth on DNA are now studying RNA.

"Once a chemist gets hold of a nucleic acid, whether it starts as RNA or DNA, it can end up looking like something halfway in between," Cech says. "After all, the main difference between the two is the 29-hydroxyl group, and as soon as chemists get their hands on DNA or RNA, they tend to start mucking around with those 29-hydroxyl groups," he quips. "Those of us who live in the RNA world consider DNA to be just a derivative of RNA anyway."

Scott A. Strobel, a professor in the departments of chemistry and molecular biophysics and biochemistry at Yale, got his start as a graduate student in Dervan's lab working on triple-helix DNA structures. "Part of the reason I moved from DNA to RNA is that RNA doesn't follow the Watson-Crick rules," Strobel says. "In RNA, because one strand isn't used to template another, you can actually have a molecule that folds back on itself. You end up with all kinds of things that are exceptions to the Watson-Crick rules of base pairing and structure. We're interested in what those pairings are and what kinds of structural and functional implications there are as a result of having things that aren't simple duplexes."

Strobel uses a technique called nucleotide analog interference mapping (NAIM) to determine what parts of RNA catalytic structures are important for their function. "We're considering a nucleic acid to be a chemical. If it's a chemical, then it's one that we can modify," he says.

Of course, the simplest way to modify a nucleic acid is just to switch one base for another. "The problem is that that's too sharp a sword on one side and too dull a sword on the other," he laments. "If you mutate a G to a C, you've changed almost everything about the base, but you've changed almost nothing about the backbone. It's still a ribose sugar; it's still a phosphate. No matter what mutagenesis you do with a nucleic acid, the changes are too big on the base and too little on the sugar."

Instead, Strobel makes more subtle changes to the nucleotides. So far, he has made about 70 or 75 different nucleotides that are variations of the original adenine, guanine, cytosine, or uracil. The changes have included deleting 29- hydroxyl groups on the ribose, substituting fluorine for the hydroxyl, and changing functional groups on the base. "We can modify or manipulate almost every functional group on the base and most of them on the sugar. The caveat is that not only do you have to be able to make them as a triphosphate, but they also have to be incorporated by an RNA polymerase to be used in this assay."

STROBEL HAS EXTENDED NAIM so that not only can he identify which functional groups are important in a particular RNA molecule, but he can tell which nucleotides interact with one another. "We knock out a functional group that we know is important, but we knock it out in every molecule," he says. "If you repeat the original interference experiment in this background where one of the functional groups that you knew was important was deleted, the outcome is that its partner tends to no longer show interference. You can actually define tertiary interaction partners. Within the group I intron, we've systematically found about a dozen of these that effectively bring the active site very tightly woven together with multiple layers--not unlike an onion--to create an active center."

Strobel considers nucleic acid interference suppression, as the second technique is called, to be complementary to such structural techniques as X-ray crystallography and nuclear magnetic resonance. "The beauty of it is that it's functional information that then has structural implications."

Anna Marie Pyle, another professor in Yale's department of molecular biophysics and biochemistry, is studying RNA structure. She started out working on small-molecule probes of DNA structure in Barton's lab while Barton was at Columbia University; Pyle moved to Yale in 2002.

"I became very interested in the information content on the surface of the DNA major groove, which got me interested in nucleic acid structure in general," Pyle says. "At the time, about 1990, it became very clear through Tom Cech's work that RNA adopted a much more complex structure than DNA. I decided to start studying RNA. It has a very interesting tertiary architecture that is superimposed on its duplex structure."

At Yale, her research is focusing on the group of ribozymes--RNA molecules that catalyze reactions--known as group II introns. "The discovery of ribozymes underscored the complexity of RNA structure in a way that nothing else ever had," Pyle says. "To make an active site for chemistry, something has to have a pretty interesting organization." She realized that ribozymes were an optimal way to study nucleic acid structure because they report on their structure through their activity.

"We work on the folding pathway of tertiary-structured RNA. We work on how it gets to the folded state, and then we also work on the architectural arrangement of the folded molecule," Pyle says. "In the RNA field, we are just beginning to understand the alphabet of tertiary interactions in RNA. We are trying to identify new types of interaction, to characterize their thermodynamic signatures, and to understand how strong they are and what kind of context they require for formation."

OF COURSE, biological applications of nucleic acids are important to chemists, but DNA has also become more than just a biological molecule.

"The DNA molecule today can be viewed as a chemical reagent, quite apart from its role in biology, although much of the work is inspired by what might be the positive consequences for medicine or drug development or diagnosis of disease," Lippard says. "For even neglecting its biological importance, there would still be a lot of interest in DNA as a fundamental molecule for chemists to study and use as an object for chemical building blocks, sensing, polymer design, etcetera, without any medical application whatsoever."

	DNA AS MATERIAL Seeman uses DNA to make structures such as this cube. COURTESY OF NADRIAN SEEMAN

For example, Nadrian C. Seeman, a chemistry professor at New York University, is using DNA as a building block to make nanometer-scale structures. "Although I have interests in biology, mostly I think about DNA as a chemical," he says. "It has properties that can be exploited, particularly on the nanometer scale." Seeman's interest originally sprang from a desire to characterize Holliday junctions, four-armed DNA structures that are a key step in genetic recombination.

"I didn't really feel that thinking about the DNA structure, except in a biological context, was a really useful thing until I started up my own program somewhere around 1980," Seeman recalls. "At that time, DNA was biology--period. Many of my early papers were rejected; occasionally, it still happens when some referee says, 'Where's the biology?' The first thing we did that was of any prominence [was to make] a DNA molecule that was connected like a cube. We sent it to a prominent journal that will remain nameless. One referee said, 'This is spectacular,' and the other referee said, 'Where's the biology?' People just could not get away from the idea that DNA had to be associated with biology."

Seeman utilizes "sticky ends," single-stranded overhang at the ends of double-stranded DNA, to direct the assembly of DNA objects. He needs to keep several things in mind when devising the sequences that he uses.

"I come from a crystallographic background, so I think of things in terms of symmetry," he says. "To a first approximation, we trash symmetry. We divide each strand into a series of overlapping elements. A typical example is a 16-mer. We break it up into a series of 13 overlapping tetramers. We insist that each of those be unique. We also insist that for any of the tetramers going around the bend of the branch point, there be no complement to them, so they couldn't form a double helix if they wanted to." Seeman has used these methods to make a variety of structures, including cubes, truncated octahedrons, and two- and three-dimensional periodic arrays.

Despite 50 years of research since Watson and Crick described the double helix, there are still questions to be answered about higher order nucleic acid structures. "We are still just at the beginning in terms of understanding the driving forces for RNA folding and tertiary-structure formation," Pyle says. "We're still just at the beginning of defining the tertiary building blocks by which an RNA folds, such that we can look at the sequence and know what it's going to do."

Kool has had three-dimensional models of DNA sitting on his desk since 1988. "At first I wondered, how much can you look at this and continue to understand something new," he muses. "I'm always amazed at needing to come back to the structure and look at some small nuance I had forgotten. Yes, we're making modifications to it and making new synthetic molecules, but we come back to this structure a lot."

Kool sees one of the great challenges as being whether people can design new genetic systems, information-encoding systems that work in organisms or perhaps only in test tubes. "If we understand enough about the structure and function of DNA, we ought to be able to design our own," he notes.

Maybe we'll have the answers to those questions in time for the centennial of the double helix.

MORE ONLINE

A list of various meetings and events celebrating the 50th anniversary of the discovery of DNA's double helical structure

Related resources: DNA weblinks and books of interest

Chemical & Engineering News
Copyright © 2003 American Chemical Society