The covalent structure of nucleic acids accounts for their ability to carry information in the form of a sequence of bases. Other features of nucleic acid structure facilitate the process of replication—that is, the generation of two copies of a nucleic acid from one. DNA replication is the basis of cell duplication, growth, and, ultimately, reproduction. Replication depends on the ability of the bases found in nucleic acids to form specific base pairs in such a way that a helical structure consisting of two strands is formed.

The discovery of the DNA double helix has been chronicled in many articles, books, and even a movie. James Watson and Francis Crick, using data obtained by Maurice Wilkins and Rosalind Franklin and simple molecular models, inferred a structural model for DNA that was the source of some remarkable insights into the functional properties of nucleic acids (Figure 33.11).

612

The features of the Watson–Crick model of DNA are

How is such a regular structure able to accommodate an arbitrary sequence of bases, given the different sizes and shapes of the purines and pyrimidines? Restricting the combinations in which bases on opposite strands can be paired—A with T and C with G—alleviates this problem. When guanine is paired with cytosine and adenine with thymine to form base pairs, these pairs have essentially the same shape (Figure 33.12). These base pairs, often called Watson-Crick base pairs, are held together by hydrogen bonds. These base-pairing rules mean that, if the sequence of one strand is known, the sequence of the other strand is defined. We will see the implications of this below and in Chapter 34. The base-pairing also helps to stabilize the double helix.

The stacking of bases one on top of another also contributes to the stability of the double helix in two ways (Figure 33.13). First, the double helix is stabilized by the hydrophobic effect: hydrophobic interactions between the bases drive the bases to the interior of the helix, resulting in the exposure of the more polar surfaces to the surrounding water. This arrangement is reminiscent of protein folding in which hydrophobic amino acids are interior in the protein and hydrophilic amino acids are exterior. Second, stacked bases attract one another through van der Waals forces, a phenomenon called base stacking. Energies associated with van der Waals interactions are quite small, such that typical interactions contribute from 2 to 4 kJ mol⁻¹ (0.5 to 1.0 kcal mol⁻¹) per atom pair. In the double helix, however, a large number of atoms are in van der Waals contact, and the net effect, summed over these atom pairs, is substantial. As stated earlier, the base sequence is the information content of nucleic acids. Note that, in DNA, this information is tucked into the relative safety of the interior of the double helix.

The double-helical model of DNA and the presence of specific base pairs immediately suggested how the genetic material might replicate. The sequence of bases of one strand of the double helix precisely determines the sequence of the other strand. Thus, the separation of a double helix into its two component strands would yield two single-stranded templates onto which new double helices could be constructed, each of which would have the same sequence of bases as the parent double helix. Consequently, as DNA is replicated, one of the strands of each daughter DNA molecule would be newly synthesized, whereas the other would be passed unchanged from the parent DNA molecule. This distribution of parental atoms is achieved by semiconservative replication (Figure 33.14). We will investigate DNA replication in detail in Chapter 34.

614

How do we know that DNA replication is semiconservative? Matthew Meselson and Franklin Stahl carried out an ingenious experiment to establish this fundamental point. They labeled the parent DNA with ¹⁵N, a heavy isotope of nitrogen, to make it denser than ordinary DNA. The labeled DNA was generated by growing E. coli for many generations in a medium that contained ¹⁵NH₄Cl as the sole nitrogen source. After the incorporation of heavy nitrogen was complete, the bacteria were abruptly transferred to a medium that contained ¹⁴N, the ordinary, lighter isotope of nitrogen. The question asked was: What is the distribution of ¹⁴N and ¹⁵N in the DNA molecules after successive rounds of replication?

The distribution of ¹⁴N and ¹⁵N was revealed by the technique of density-gradient equilibrium sedimentation. A small amount of DNA was dissolved in a concentrated solution of cesium chloride having a density close to that of the DNA (1.7 g cm⁻³). This solution was centrifuged until it was nearly at equilibrium, where the opposing forces of sedimentation and diffusion created a gradient in the concentration of cesium chloride across the centrifuge tube. The result was a stable density gradient, ranging from 1.66 to 1.76 g cm⁻³. The DNA molecules in this density gradient were driven by centrifugal force into the region where the solution’s density was equal to their own, yielding a narrow band of DNA that was detected by its absorption of ultraviolet light. ¹⁴N DNA and ¹⁵N DNA molecules in a control mixture differ in density by about 1%, enough to clearly distinguish the two types of DNA (Figure 33.15).

DNA was then extracted from the bacteria at various times after they had been transferred from a ¹⁵N to a ¹⁴N medium and centrifuged. Analysis of these samples showed that there was a single band of DNA after one generation. The density of this band was precisely halfway between the densities of the ¹⁴N DNA and ¹⁵N DNA bands (Figure 33.16). The absence of ¹⁵N DNA indicated that parental DNA was not preserved as an intact unit after replication. The absence of ¹⁴N DNA indicated that all of the daughter DNA derived some of their atoms from the parent DNA. This proportion had to be half, because the density of the hybrid DNA band was halfway between the densities of the ¹⁴N DNA and ¹⁵N DNA bands.

615

After two generations, there were equal amounts of two bands of DNA. One was hybrid DNA, and the other was ¹⁴N DNA. Meselson and Stahl concluded from these incisive experiments “that the nitrogen in a DNA molecule is divided equally between two physically continuous subunits; that following duplication, each daughter molecule receives one of these; and that the subunits are conserved through many duplications.” Their results agreed perfectly with the Watson–Crick model for DNA replication (Figure 33.14).

During DNA replication and transcription, the two strands of the double helix must be separated from one another, at least in a local region. In other words, sometimes the archival information must be accessible to be useful. In the laboratory, the two strands of the double helix can be separated by heating a solution of DNA. The thermal energy causes the DNA molecules to move to such a degree that the weak forces holding the helix together, such as the hydrogen bonds between the bases, break apart. The dissociation of the double helix is called melting or denaturation, and it occurs relatively abruptly at a certain temperature. The melting temperature T_m is defined as the temperature at which half the helical structure is lost. Strands can also be separated by adding acid or alkali to ionize the nucleotide bases and disrupt base-pairing.

Separated complementary strands of nucleic acids spontaneously reassociate to form a double helix when the temperature is lowered below T_m. This renaturation process is sometimes called annealing. The facility with which double helices can be melted and then reassociated is crucial for the biological functions of nucleic acids.