DNA replication must be very rapid, given the sizes of the genomes and the rates of cell division. The E. coli genome contains 4.6 million base pairs and is copied in less than 40 minutes. Thus, 2000 bases are incorporated per second. Enzyme activities need to be highly coordinated to replicate entire genomes precisely and rapidly.

We begin our consideration of the coordination of DNA replication by looking at E. coli, which has been extensively studied. For this organism, with a small genome, replication begins at a single site and continues around the circular chromosome. The coordination of eukaryotic DNA replication is more complex than bacterial replication because there are many initiation sites throughout the genome and because an additional enzyme is needed to replicate the ends of linear chromosomes.

In E. coli, DNA replication starts at a unique site within the entire 4 × 10⁶ bp genome. This origin of replication, called the oriC locus, is a 245-bp region that has several unusual features.

The oriC locus contains four copies of a sequence that are preferred binding sites for the origin-recognition protein DnaA. In addition, the locus contains a tandem array of 13-bp sequences that are rich in AT base pairs (Figure 34.11A). The binding of DnaA molecules to the oriC locus begins the building of the replication complex. The DnaA proteins then oligomerize, wrapping the origin around themselves (Figure 34.11B). Additional proteins, such as DnaB, then join DnaA. The DnaB protein is a helicase that utilizes ATP hydrolysis to unwind the duplex, including the AT-rich regions. Single-strand-binding proteins (SSB) then bind to the newly generated single-stranded regions, preventing the two complementary strands from re-forming the double helix (Figure 34.11C). The result of this process is the generation of a structure called the prepriming complex, which makes single-stranded DNA accessible for other enzymes to begin the synthesis of the complementary strands.

Even with the DNA template exposed and the prepriming complex assembled, there are still obstacles to DNA synthesis. DNA polymerases can add nucleotides only to a free hydroxyl group; they cannot start a strand de novo. Therefore, a primer is required. How is this primer formed? An important clue came from the observation that RNA synthesis is required for the initiation of DNA synthesis. In fact, RNA primes the synthesis of DNA. A specialized RNA polymerase called primase joins the prepriming complex in a multisubunit assembly called the primosome. Primase synthesizes a short stretch of RNA (about 10 nucleotides) that is complementary to one of the template DNA strands (Figure 34.12). The RNA primer is removed by a 5′ → 3′ exonuclease; in E. coli, the exonuclease is present as the third active site in DNA polymerase I.

Both strands of parental DNA serve as templates for the synthesis of new DNA. The site of DNA synthesis is called the replication fork because the complex formed by the newly synthesized daughter strands arising from the parental duplex resembles a two-pronged fork (Figure 34.13). Recall that the two strands are antiparallel; that is, they run in opposite directions. On cursory examination, both daughter strands appear to grow in the same direction, as shown in Figure 34.13. However, this appearance of same-direction growth presents a conundrum, because all known DNA polymerases synthesize DNA in the 5′ → 3′ direction but not in the 3′ → 5′ direction. How then does one of the daughter DNA strands appear to grow in the 3′ → 5′ direction?

This dilemma was resolved when careful experimentation found that a significant proportion of newly synthesized DNA exists as small fragments. These units of about a thousand nucleotides, called Okazaki fragments (after Reiji Okazaki, who first identified them), are present briefly in the vicinity of the replication fork (Figure 34.14). As replication proceeds, these fragments become covalently joined through the action of DNA ligase, an enzyme that uses ATP hydrolysis to power the joining of DNA fragments to form one of the daughter strands. The other new strand is synthesized continuously. The strand formed from Okazaki fragments is termed the lagging strand, whereas the one synthesized without interruption is the leading strand. Both the Okazaki fragments and the leading strand are synthesized in the 5′ → 3′ direction. The discontinuous assembly of the lagging strand enables 5′ → 3′ polymerization at the nucleotide level to give rise to overall growth in the 3′ → 5′ direction.

We have been introduced to the enzymes that participate in DNA replication. They include helicases and topoisomerases, and we have examined the general properties of polymerases by focusing on DNA polymerase I. We now examine the enzyme responsible for the rapid and accurate synthesis of DNA in E. coli—the holoenzyme DNA polymerase III. The hallmarks of this multi-subunit assembly are not only its fidelity, but also its very high catalytic potency and processivity. The term processivity refers to the ability of an enzyme to catalyze many consecutive reactions without releasing its substrate. The holoenzyme catalyzes the formation of many thousands of phosphodiester linkages before releasing its template, compared with only 20 for DNA polymerase I. The DNA polymerase III holoenzyme grasps its template and does not let go until the template has been completely replicated. Another distinctive feature of the holoenzyme is its catalytic prowess: 1000 nucleotides are added per second compared with only 10 per second for DNA polymerase I. This acceleration is accomplished with no loss of accuracy. The greater catalytic prowess of polymerase III is largely due to its processivity; no time is lost in repeatedly stepping on and off the template.

The source of the processivity—the ability to stay “on task”—is the β₂ subunit, which has the form of a star-shaped ring (Figure 34.15). A 35-Å-diameter hole in its center can readily accommodate a duplex DNA molecule yet leaves enough space between the DNA and the protein to allow rapid sliding and turning in the course of replication. A catalytic rate of 1000 nucleotides polymerized per second requires the sliding of 100 turns of duplex DNA (a length of 3400 Å, or 0.34 mm) through the central hole of β₂ per second. Thus, β₂ plays a key role in replication by serving as a sliding DNA clamp.

How does DNA become trapped inside the sliding clamp? Polymerases also include assemblies of subunits that function as clamp loaders. These enzymes grasp the sliding clamp and, utilizing the energy of ATP, pull apart the two subunits of the sliding clamp on one side. DNA can move through the gap, inserting itself through the central hole. ATP hydrolysis then releases the clamp, which closes around the DNA.

Replicative polymerases such as DNA polymerase III synthesize the leading and lagging strands simultaneously at the replication fork (Figure 34.16). DNA polymerase III begins the synthesis of the leading strand starting from the RNA primer formed by primase. The duplex DNA ahead of the polymerase is unwound by a helicase. Single-strand-binding proteins bind to the unwound strands, keeping the strands separated so that both strands can serve as templates. The leading strand is synthesized continuously by polymerase III. Topoisomerase II concurrently introduces negative supercoils to avert a topological crisis.

The mode of synthesis of the lagging strand is necessarily more complex. As mentioned earlier, the lagging strand is synthesized in fragments so that 5′ → 3′ polymerization leads to overall growth in the 3′ → 5′ direction. Yet the synthesis of the lagging strand is coordinated with the synthesis of the leading strand. How is this coordination accomplished? Examination of the subunit composition of the DNA polymerase III holoenzyme reveals an elegant solution (Figure 34.17). The holoenzyme includes two copies of the polymerase core enzyme, which consists of the DNA polymerase itself, a 3′-to-5′ proofreading exonuclease, two copies of the dimeric β₂-subunit sliding clamp, and several other proteins. The core enzymes are linked to a central structure, which consists of the clamp loader, and subunits that interact with the single-strand-binding protein. The entire central structure interacts with helicase DnaB.

The lagging-strand template is looped out so that it passes through the polymerase site in one subunit of polymerase III in the 3′ → 5′ direction. After adding about 1000 nucleotides, DNA polymerase III lets go of the lagging-strand template by releasing the sliding clamp. A new loop is then formed, a sliding clamp is added, and primase again synthesizes a short stretch of RNA primer to initiate the formation of another Okazaki fragment. This mode of replication has been termed the trombone model because the size of the loop lengthens and shortens like the slide on a trombone (Figure 34.18). The gaps between fragments of the nascent lagging strand are filled by DNA polymerase I. This essential enzyme also uses its 5′ → 3′ exonuclease activity to remove the RNA primer lying ahead of the polymerase site. The primer cannot be erased by DNA polymerase III, because the enzyme lacks 5′ → 3′ editing capability. Finally, DNA ligase connects the fragments. DNA ligase catalyzes the formation of a phosphodiester linkage between the 3′-hydroxyl group at the end of one DNA chain and the 5′-phosphate group at the end of the other (Figure 34.19). ATP is commonly used as the energy source to drive this thermodynamically uphill reaction.

Replication in eukaryotes is mechanistically similar to replication in bacteria but is more challenging in three ways. One challenge is sheer size: E. coli must replicate 4.6 million base pairs, whereas a human diploid cell must replicate 6 billion base pairs. The second challenge is the fact that, while the genetic information for E. coli is contained on 1 chromosome, human beings have 23 pairs of chromosomes that must be replicated. Finally, whereas the E. coli chromosome is circular, human chromosomes are linear. The third challenge arises because of the nature of DNA synthesis on the lagging strand. Linear chromosomes are subject to shortening with each round of replication unless countermeasures are taken.

The first two challenges are met by the use of multiple origins of replication, which are located between 30 and 300 kilobase pairs (kbp) apart. In human beings, replication of the entire genome requires about 30,000 origins of replication, with each chromosome containing several hundred. Each origin of replication represents a replication unit, or replicon. The use of multiple origins of replication requires mechanisms for ensuring that each sequence is replicated once and only once. How are replicons controlled so that each replicon is replicated only once in each cell division? Proteins, called licensing factors because they permit (license) the formation of the DNA synthesis initiation complex, bind to the origin of replication. These proteins ensure that each replicon is replicated only once in each round of DNA synthesis. After the licensing factors have established the initiation complex, these factors are subsequently destroyed. The license expires after one use.

Two distinct polymerases are needed to copy a eukaryotic replicon. An initiator polymerase called polymerase α begins replication. This enzyme includes a primase subunit, used to synthesize the RNA primer, as well as an active DNA polymerase. After this polymerase has added a stretch of about 20 deoxynucleotides to the primer, it is replaced by DNA polymerase δ, a more processive enzyme and the principal replicative polymerase in eukaryotes. This process is called polymerase switching because one polymerase has replaced another.

Whereas the genomes of essentially all bacteria are circular, the chromosomes of human beings and other eukaryotes are linear. The free ends of linear DNA molecules introduce several complications. First, the unprotected termini of the DNA at the end of a chromosome are likely to be more susceptible to digestion by exonucleases if they are left to freely dangle at the end of the chromosome during replication. Second, the complete replication of DNA ends is difficult because polymerases act only in the 5′ → 3′ direction. The lagging strand would have an incomplete 5′ end after the removal of the RNA primer. Each round of replication would further shorten the chromosome, which would, like the Cheshire cat, slowly disappear (Figure 34.20).

What characteristics of the chromosome ends, called telomeres (from the Greek telos, meaning “an end”), mitigate these two problems? The most notable feature of telomeric DNA is that it contains hundreds of tandem repeats of a hexanucleotide sequence. One of the strands is G-rich at the 3′ end, and it is slightly longer than the other strand. In human beings, the repeating G-rich sequence is AGGGTT. This simple repeat precludes the likelihood that the sequence encodes any information, but the simple repeat does facilitate the formation of large duplex loops (Figure 34.21). The single-stranded region at the very end of the structure has been proposed to loop back to form a DNA duplex with another part of the repeated sequence, displacing a part of the original telomeric duplex. This loop-like structure is formed and stabilized by specific telomere-binding proteins. Such structures would nicely protect the end of the chromosome from degradation.