To understand how nucleic acids function, we must understand their chemical properties as well as their structures. The role of DNA as a repository of genetic information depends in part on its inherent stability. The chemical transformations that do happen are generally very slow in the absence of an enzyme catalyst. The long-term storage of information without alteration is so important to a cell, however, that even very slow changes in DNA structure can be physiologically significant. Other, nondestructive alterations of DNA do occur and are essential to function, such as the strand separation that must precede replication or transcription. For RNAs, chemical modifications can play significant roles in ensuring correct structure and function.

196

In addition to providing insights about physiological processes, our understanding of nucleic acid chemistry gives us a powerful array of technologies that have applications in molecular biology, medicine, and forensic science. We examine here the chemical properties of DNA and RNA, as well as some of these technologies. Many more techniques and applications are discussed in Chapter 7.

Solutions of carefully isolated, double-stranded DNA are highly viscous at pH 7.0 and room temperature (25°C). When such a solution is subjected to extremes of pH or to temperatures above 80°C, its viscosity decreases sharply, indicating that the DNA has undergone a physical change. This change is due to denaturation, or melting, of the double-helical DNA and can also occur with RNA. Disruption of both the hydrogen bonding between paired bases and the base stacking causes the double helix to unwind, forming two single strands that are completely separate from each other along the entire (or partial) length of the molecule. No covalent bonds in the nucleic acid are broken during denaturation (Figure 6-28).

Renaturation of a DNA or RNA molecule is a rapid, one-step process, as long as a double-helical segment of at least a dozen residues still unites the two strands. When the temperature or pH is returned to the range in which most organisms live, the unwound segments of the two strands spontaneously rewind to yield the intact duplex. This process, called annealing, involves re-formation of all the base pairs in the double helix. If the strands were completely separated, renaturation occurs in two steps. In the first step, which is relatively slow, complementary sequences in the two strands “find” each other by random collisions and form a short segment of double helix. The second step is much faster: the remaining unpaired bases successively come into register as base pairs, and the two strands “zipper” themselves together to form the double helix.

197

The close interaction of stacked bases in a nucleic acid has the effect of decreasing its absorption of UV light relative to that of a solution with the same concentration of free nucleotides, and the absorption is further decreased by the pairing of two complementary strands. Hydrogen bonding and base stacking in the double helix limit the resonance of the aromatic rings of the bases, thereby decreasing UV light absorption. This is known as the hypochromic effect (Figure 6-29). When DNA is denatured, the base pairs are disrupted and the two strands separate into randomly coiled chains. The resonance of the bases in each strand is no longer constrained, as it is when the bases are part of a double helix. As a result, the UV light absorption of single-stranded DNA is approximately 40% higher than that of double-stranded DNA at the same concentration. This increase in absorption is called the hyperchromic effect. The transition from double-stranded DNA to the single-stranded, denatured form can thus be detected by monitoring the absorption of UV light.

DNA molecules in solution denature when they are heated slowly. Each species of DNA has a characteristic denaturation temperature, or melting point (T_m), defined as the temperature at which half the DNA is denatured. In general, the higher the content of G≡C base pairs in the DNA, the higher its melting point. This is because G≡C base pairs, with three hydrogen bonds, require more heat energy to dissociate than do A=T base pairs. Careful determination of the melting point of a DNA specimen, under fixed conditions of pH and ionic strength, can yield an estimate of its base composition. If denaturation conditions are carefully controlled, regions that are rich in A=T base pairs will specifically denature while most of the DNA remains double-stranded. Such denatured regions, or bubbles, can be visualized with electron microscopy (Figure 6-30). Strand separation of DNA must occur in vivo during processes such as DNA replication and transcription. As we will see in Chapter 11, the DNA sites where these processes are initiated are often rich in A=T base pairs.

RNA duplexes or RNA-DNA hybrid duplexes can also be denatured. Notably, RNA duplexes are more stable than DNA duplexes. At neutral pH, the denaturation of a double-helical RNA often requires higher temperatures, by 20°C or more, than those required to denature a DNA molecule with a comparable sequence. The stability of an RNA-DNA hybrid is generally intermediate between that of double-stranded RNA and DNA.

198

The ability of two complementary DNA strands to pair with each other can be used to detect similar DNA sequences in different species or within the genome of a single species. If double-stranded DNAs isolated from human cells and from mouse cells are completely denatured by heating, then mixed and kept at 25°C for many hours, much of the DNA will anneal. Most of the mouse DNA strands anneal with complementary mouse DNA strands to form mouse duplex DNA; similarly, most human DNA strands anneal with complementary human DNA strands. However, some strands of the mouse DNA will associate with human DNA to yield hybrid duplexes, in which segments of a mouse DNA strand form base-paired regions with segments of a human DNA strand (Figure 6-31). The ability to hybridize DNA from different species is a valuable laboratory tool for exploring evolutionary relationships. Different species have proteins and RNAs with similar functions—and often similar structures. In many cases, the DNAs encoding these proteins and RNAs have similar sequences. The closer the evolutionary relationship between two species, the more extensively their DNAs will hybridize. For example, human DNA hybridizes much more extensively with mouse DNA than with DNA from yeast.

The hybridization of DNA strands from different sources forms the basis for powerful techniques used in classical molecular genetics. Although these hybridization techniques are used less often as new, high-throughput approaches based on DNA sequencing become available (as described in Chapters 7 and 8), they are still widely used in research laboratories worldwide. Such techniques have laid the foundation for understanding the study of nucleic acids.

A specific gene’s DNA or RNA sequence can be detected in the presence of many other sequences by hybridization with a probe, a carefully chosen nucleic acid sequence complementary to the gene of interest. To be visualized in the laboratory, the probe must be labeled in some way, usually radioactively or with a fluorophore (a compound carrying a fluorescent group). The probe that is selected depends on what is known about the gene under investigation. Sometimes, a gene from another species that has sequence similarity to the gene of interest makes a suitable probe. If the protein product of a gene has been purified, probes can be designed and synthesized by working backward from the amino acid sequence, deducing the DNA sequence that would code for it. Or, researchers can often obtain the DNA sequence information necessary for creating a probe from sequence databases that detail the structure of millions of genes from a wide range of organisms. Because base-pairing stability is sensitive to pH and temperature, these parameters can be adjusted experimentally to detect nucleic acid sequences with varying degrees of complementarity to the probe. The technique is sensitive enough to reveal sequences that differ by a single base pair. This can be critically important in medical and forensic applications, given that two people can share genetic information that differs at only one or a few base pairs, called single-nucleotide polymorphisms (SNPs).

199

In other common hybridization methods, gel electrophoresis is used to separate DNA or RNA molecules by size (Figure 6-32a). A variation of gel electrophoresis, used to detect proteins under denaturing conditions, was discussed in Chapter 4 (see Highlight 4-1). Here, the gel matrix is not denaturing but instead is made of agarose, a kelp-derived material that does not disrupt nucleic acid base pairing. The starting nucleic acid sample, in solution in a test tube, is applied to a slot at one end of the gel, and a voltage is applied. Because DNA and RNA molecules are negatively charged, they migrate toward the positive end of the gel matrix in the electric field. Larger molecules tend to move more slowly than smaller ones, so this provides a means of separating nucleic acids by size. Following electrophoresis, the DNA fragments are transferred to a nitrocellulose membrane so that their positions in the gel relative to each other are preserved on the membrane. Once on the membrane, the nucleic acid can be hybridized with a DNA or RNA probe, labeled so that it can be detected by measuring radioactivity or fluorescence.

When used to detect DNA, this method is known as Southern blotting, named for Edwin Southern, who invented the technique at the University of Edinburgh. When used for detecting RNA, the technique is called Northern blotting, because of its similarity to the Southern method. Applications of these techniques include identifying a person on the basis of a single hair left at the scene of a crime or predicting the onset of a disease in an individual decades before symptoms appear. Northern blotting can also be used to detect the levels of a particular type of RNA in different body tissues (Figure 6-32b), providing fascinating insight into how cells regulate the expression of genes. Notably, these classical methods of Southern and Northern blotting are still used to answer specific experimental questions, despite the development of high-throughput strategies based on DNA sequencing technology (see Chapters 7 and 8).

Purines and pyrimidines, and the nucleotides of which they are a part, can undergo spontaneous alterations in their covalent structure. The rate of these reactions is generally very slow, but as noted earlier, they are physiologically significant because of the cell’s low tolerance for changes in its genetic information. Alterations in DNA structure that produce permanent genetic changes are known as mutations. Extensive evidence suggests an intimate link between the accumulation of mutations in an individual organism and the processes of aging and carcinogenesis.

200

Several nucleotide bases undergo deamination, a spontaneous loss of their exocyclic amino groups. For example, under typical cellular conditions, deamination of cytosine (in DNA) to uracil occurs in about 1 of every 10⁷ C residues in 24 hours (Figure 6-33). This corresponds to about 100 spontaneous deamination events per day, on average, in a mammalian cell. Deamination of adenine and guanine occurs at about 1/100th this rate.

The slow cytosine deamination reaction seems innocuous enough, but it is almost certainly the reason that DNA contains thymine rather than uracil. In DNA, uracil is the product of cytosine deamination, and it is readily recognized as foreign and is removed by a DNA repair system (see Chapter 12). If DNA normally contained uracil, the recognition of U residues resulting from cytosine deamination would be more difficult, and unrepaired uracils would lead to permanent sequence changes as they were paired with adenines during replication (cytosine normally pairs with guanine, so introduction of uracil into DNA effectively changes a C≡G base pair to a U–A base pair). Establishing thymine as one of the four bases in DNA may well have been a crucial turning point in evolution, making the long-term storage of genetic information possible.

Cytosine deamination also provides innate cellular defense against viral infection. A family of human proteins called APOBECs catalyze cytosine deamination in the viral genome during the initial round of replication by HIV. This hypermutation results in many nonviable viral particles, eventually destroying the coding capacity of the virus. In HIV and related viruses, the viral protein Vif binds to APOBECs and triggers their degradation. Vif has therefore become an important antiviral target, because viruses lacking this protein are much less capable of establishing chronic infection in human cells.

Certain nucleotide bases in DNA molecules are enzymatically methylated, usually after DNA synthesis is complete. Adenine and cytosine are methylated more frequently than guanine (Figure 6-34a). Methylation is generally confined to certain sequences or regions of a DNA molecule. For example, more than half of all CpG sequences in mammalian genomes are methylated on the C residue. Methylation tends to inhibit gene expression, because the methylated DNA is not efficiently copied into RNA. In many cancers, gene regulatory regions in DNA become abnormally hypermethylated. This can result in the silencing of genes that would otherwise control cell growth. DNA methylation may affect gene transcription by physically blocking the binding of proteins that facilitate transcription. Other proteins, however, can specifically bind to methylated DNA and recruit additional proteins that help form compact, inactive regions of chromosomal DNA.

201

All known DNA methylases (methyltransferases) use S-adenosylmethionine as a methyl group donor (see Figure 6-13). E. coli has two prominent methylation systems. One serves as part of a defense mechanism that helps the cell distinguish its DNA from foreign DNA by marking its own DNA with methyl groups and destroys foreign, nonmethylated DNA, a process known as restriction modification (discussed in Chapter 7). The other system methylates A residues in the sequence 5′-GATC-3′ to form N⁶-methyldeoxyadenosine. Methylation in this case is mediated by the Dam (DNA adenine methylation) methylase, a component of a system that repairs the mismatched base pairs that occasionally form during DNA replication (see Chapter 12).

In eukaryotic cells, about 5% of all C residues in DNA are methylated to form 5-methyldeoxycytidine (see Figure 6-34a). As noted above, methylation is most common at CpG sequences, producing methyl-CpG symmetrically on both strands of the DNA. The observed 5% methylation frequency of C residues is lower than would be expected based on the random presence of CpG sequences in a genome, and, in fact, CpG dinucleotides occur much less often in vertebrate genomes than predicted by chance. For example, in the human genome, which has a G≡C content of ∼40%, the frequency of CpG would be predicted as 0.2 × 0.2 = 0.04, or 4%, but the frequency is closer to just 1%. The extent of methylation of CpG sequences varies with molecular region in large eukaryotic DNA molecules. In addition to the tendency of methylation of C residues to inhibit gene expression, methylation suppresses the migration of segments of DNA called transposons (see Chapter 14).

Like DNA, many functional RNAs are posttranscriptionally modified at specific nucleotides (Figure 6-34b). Some of the first examples were discovered in ribosomal and transfer RNAs. In some cases, modifications involve the addition of a functional group to an existing nucleotide in the sequence. For example, a methyl group can be added to the 2′ hydroxyl of ribose, thereby blocking its ability to form a hydrogen bond. In bacteria, some C residues of tRNAs are modified to N⁴-acetylcytidine in a process thought to contribute to the accuracy of protein synthesis. In other cases, the base itself is changed, or its linkage to the ribose—the glycosidic bond—is altered. For instance, inosine, 4-thiouridine, and pseudouridine are relatively common in tRNAs and rRNAs.

Many of the enzymes that catalyze these chemical modifications of RNA have been identified. They are often evolutionarily conserved, indicating that RNA modification has been occurring in biological systems for a long time. More difficult to figure out is the function of these chemical changes in RNA. Molecular biologists can produce unmodified versions of RNAs in the laboratory and compare their functions with those of the chemically altered counterparts isolated from cells. This approach has only rarely discerned much of an effect of a modified base. However, genetic experiments in which an RNA-modifying enzyme is mutated or deleted from an organism suggest that these enzymes give cells a subtle but important selective advantage over organisms that do not modify their RNA. Some evidence supports the hypothesis that RNA modifications stabilize RNA structures and help RNAs interact with proteins in the cell.

Knowledge of DNA and RNA chemistry provided the basis for devising methods to synthesize nucleic acids in the laboratory. This technology has paved the way for many biochemical advances that depend on the ability to synthesize oligonucleotides with any chosen sequence. The chemical methods for synthesizing nucleic acids were developed primarily by H. Gobind Khorana and his colleagues in the 1970s. Refinement and automation of these methods have made possible the rapid and accurate synthesis of DNA strands.

DNA (or RNA) synthesis is carried out with the growing strand attached to a solid support (Figure 6-35). First, a nucleotide is attached to the support, a glass or polystyrene bead, through its 3′-hydroxyl group, and polynucleotide synthesis proceeds in the 3′→5′ direction. This is the opposite of the direction of biological polynucleotide synthesis by polymerase enzymes, which is 5′→3′. Functional groups on the bases and phosphates, including hydroxyl and amine groups, are transiently protected with chemical groups that are readily removed after synthesis is complete. The 5′-hydroxyl group is temporarily protected by a dimethoxytrityl (DMT) group; the DMT group is removed from the end of the growing polymer chain at the beginning of each cycle (step 1 in Figure 6-35) to permit extension of the chain by another nucleotide (step 2). Oxidation of the phosphite linkage between the nucleotides completes the cycle (step 3). When chain synthesis is complete, protecting groups are removed from the bases and phosphates, and the oligonucleotide chain is cleaved from its solid support (steps 4, 5, 6). The efficiency of each addition step is very high, allowing the routine laboratory synthesis of polymers containing 70 to 80 nucleotides and, in some laboratories, much longer strands.

202

Oligonucleotide synthesis is very useful for techniques such as Southern and Northern blotting and for the polymerase chain reaction (PCR) and DNA sequencing, which are discussed in Chapter 7. In addition, chemical synthesis makes it possible to incorporate chemical modifications in the polymer product, such as biotin groups, extra phosphates, sulfhydryl groups, and methyl groups. These functional groups are useful for such applications as specific labeling of a DNA strand or stabilization of an RNA oligonucleotide against enzymatic degradation in cells.

203

204

205

206

207