CHAPTER 11: DNA Replication

11.1 Introduction

Replication of duplex DNA is a complicated endeavor involving multiple enzyme complexes. Different activities are involved in the stages of initiation, elongation, and termination. Before initiation can occur, however, the supercoiled chromosome must be relaxed (see the chapter titled Genes Are DNA and Encode RNAs and Polypeptides). This occurs in segments beginning with the replication origin region. This alteration to the structure of the chromosome is accomplished by the enzyme topoisomerase. Replication cannot occur on supercoiled DNA, only the relaxed form. FIGURE 11.1 shows an overview of the first stages of the process.

  • Initiation involves recognition of an origin by a complex of proteins. Before DNA synthesis begins, the parental strands must be separated and (transiently) stabilized in the single-stranded state, creating a replication bubble. After this stage, synthesis of daughter strands can be initiated at the replication fork (see the chapter titled The Replicon: Initiation of Replication).

  • Elongation is undertaken by another complex of proteins. The replisome exists only as a protein complex associated with the particular structure that DNA takes at the replication fork. It does not exist as an independent unit (e.g., analogous to the ribosome), but assembles de novo at the origin for each replication cycle. As the replisome moves along DNA, the parental strands unwind and daughter strands are synthesized.

  • At the end of the replicon, joining and/or termination reactions are necessary. Following termination, the duplicate chromosomes must be separated from one another, which requires manipulation of higher-order DNA structure.

FIGURE 11.1 Replication initiates when a protein complex binds to the origin and melts the DNA there. Then the components of the replisome, including DNA polymerase, assemble. The replisome moves along DNA, synthesizing both new strands.

Inability to replicate DNA is fatal for a growing cell. Mutants for replication must therefore be obtained as conditional lethals. These are able to accomplish replication under permissive conditions (typically provided by the normal temperature of incubation), but they are defective under nonpermissive, or restrictive, conditions (provided by the higher temperature of 42°C). A comprehensive series of such temperature-sensitive mutants in Escherichia coli identifies a set of loci called the dna genes. The dna mutants distinguish two stages of replication by their behavior when the temperature is raised:

  • The members of the major class of quick-stop mutants cease replication immediately upon a temperature increase. They are defective in the components of the replication apparatus, typically in the enzymes needed for elongation (but also include defects in the supply of essential precursors).

  • The members of the smaller class of slow-stop mutants complete the current round of replication, but cannot start another. They are defective in the events involved in initiating a new cycle of replication at the origin.

An important assay that researchers use to identify the components of the replication apparatus is called in vitro complementation. An in vitro system for replication is prepared from a dna mutant and is operated under conditions in which the mutant gene product is inactive. Extracts from wild-type cells are tested for their ability to restore activity. Researchers can purify the protein encoded by the dna locus by identifying the active component in the extract.

Each component of the bacterial replication apparatus is now available for study in vitro as a biochemically pure product, and is implicated in vivo by mutations in its gene. Analogous eukaryotic chromosomal replication systems have largely been developed. Studies of individual replisome components show a high structural and functional similarity with the bacterial replisome.

11.2 DNA Polymerases Are the Enzymes That Make DNA

There are two basic types of DNA synthesis:

  • FIGURE 11.2 shows the result of semiconservative replication. The two strands of the parental duplex are separated, and each serves as a template for synthesis of a new strand. The parental duplex is replaced with two daughter duplexes, each of which has one parental strand and one newly synthesized strand. An enzyme that can synthesize a new DNA strand on a template strand is called a DNA polymerase (or more properly, DNA-dependent DNA polymerase).

  • FIGURE 11.3 shows the consequences of a DNA repair reaction. One strand of DNA has been damaged. It is excised and new material is synthesized to replace it. Both prokaryotic and eukaryotic cells contain multiple DNA polymerase activities. Only a few of these enzymes actually undertake replication; those that do sometimes are called DNA replicases. The remaining enzymes are involved in repair synthesis (discussed in the Repair Systems chapter) or participate in subsidiary roles in replication.

FIGURE 11.2 Semiconservative replication synthesizes two new strands of DNA.

FIGURE 11.3 Repair synthesis replaces a short stretch of one strand of DNA containing a damaged base.

All prokaryotic and eukaryotic DNA polymerases share the same fundamental type of synthetic activity, antiparallel synthesis from 5′ to 3′ from a template that is 3′ to 5′. This means adding nucleotides one at a time to a 3′–OH end, as illustrated in FIGURE 11.4. The choice of the nucleotide to add to the chain is dictated by base pairing with the complementary template strand.

FIGURE 11.4 DNA is synthesized by adding nucleotides to the 3′–OH end of the growing chain, so that the new chain grows in the 5′ to 3′ direction. The precursor for DNA synthesis is a nucleoside triphosphate, which loses the terminal two phosphate groups in the reaction.

Some DNA polymerases, such as the repair polymerases, function as independent enzymes, whereas others (notably the replication polymerases) are incorporated into large protein assemblies called holoenzymes. The DNA-synthesizing subunit is only one of several functions of the holoenzyme, which typically contains other activities concerned with fidelity.

TABLE 11.1 summarizes the DNA polymerases that have been characterized in E. coli. DNA polymerase III, a multisubunit protein, is the replication polymerase responsible for de novo synthesis of new strands of DNA. DNA polymerase I (encoded by polA) is involved in the repair of damaged DNA and, in a subsidiary role, in semiconservative replication. DNA polymerase II is required to restart a replication fork when its progress is blocked by damage in DNA. DNA polymerases IV and V are involved in allowing replication to bypass certain types of damage and are called error-prone polymerases.

TABLE 11.1 Only one DNA polymerase is the replication enzyme. The others participate in repairing damaged DNA, restarting stalled replication forks, or bypassing damage in DNA.

Enzyme Gene Function
I polA Major repair enzyme
II polB Replication restart
III polC Replicase
IV dinB Translesion replication
V umuD’2C Translesion replication

When researchers assay extracts of E. coli for their ability to synthesize DNA, the predominant enzyme activity is DNA polymerase I. Its activity is so great that it makes it impossible to detect the activities of the enzymes actually responsible for DNA replication! To develop in vitro systems in which replication can be followed, researchers therefore prepare extracts from polA mutant cells.

Several classes of eukaryotic DNA polymerases have been identified. DNA polymerases δ and ε are required for nuclear replication; DNA polymerase α is concerned with “priming” (initiating) replication. Other DNA polymerases are involved in repairing damaged nuclear DNA, or in translesion replication of damaged DNA when repair of damage is impossible. Mitochondrial DNA replication is carried out by DNA polymerase γ, whereas chloroplasts have their own replication system (see the section Separate Eukaryotic DNA Polymerases Undertake Initiation and Elongation later in this chapter).

11.3 DNA Polymerases Have Various Nuclease Activities

Replicases often have nuclease activities as well as the ability to synthesize DNA. A 3′–5′ exonuclease activity is typically used to excise bases that have been added to DNA incorrectly. This provides a “proofreading” error-control system (see the section, DNA Polymerases Control the Fidelity of Replication, which follows).

The first DNA-synthesizing enzyme that researchers characterized was DNA polymerase I, which is a single polypeptide of 103 kD (kilodalton). The chain can be cleaved into two parts by proteolytic treatment. The larger cleavage product (68 kD) is called the Klenow fragment. It is used in synthetic reactions in vitro. It contains the polymerase and the proofreading 3′–5′ exonuclease activities. The active sites are approximately 30 Å apart in the protein, which indicates that there is spatial separation between adding a base and removing one.

The small fragment (35 kD) possesses a 5′–3′ exonucleolytic activity, which excises small groups of nucleotides, up to approximately 10 bases at a time. This activity is coordinated with the synthetic/proofreading activity. It provides DNA polymerase I with a unique ability to start replication in vitro at a nick in DNA. (No other DNA polymerase has this ability.) At a point where a phosphodiester bond has been broken in a double-stranded DNA, the enzyme extends the 3′–OH end. As the new segment of DNA is synthesized, it displaces the existing homologous strand in the duplex. The displaced strand is degraded by the 5′–3′ exonucleolytic activity of the enzyme.

FIGURE 11.5 illustrates this process of nick translation. The displaced strand is degraded by the 5′–3′ exonuclease activity of the enzyme. The properties of the DNA are unaltered, except that a segment of one strand has been replaced with newly synthesized material, and the position of the nick has been moved along the duplex. This is of great practical use; nick translation has been a major technique for introducing radioactively labeled nucleotides into DNA in vitro.

FIGURE 11.5 Nick translation replaces part of a preexisting strand of duplex DNA with newly synthesized material.

The coupled 5′–3′ synthetic/3′–5′ exonucleolytic action is used most extensively for filling in short single-stranded regions in double-stranded DNA. These regions arise during lagging strand DNA replication (see the section DNA Polymerases Have a Common Structure later in this chapter), and during DNA repair (see Figure 11.3).

11.4 DNA Polymerases Control the Fidelity of Replication

The fidelity of replication poses the same sort of problem encountered in considering (for example) the accuracy of translation. It relies on the specificity of base pairing. Yet when we consider the energetics involved in base pairing, we would expect errors to occur with a frequency of approximately 10–2 per base pair replicated. The actual rate in bacteria seems to be approximately 10–8 to 10–10. This corresponds to about 1 error per genome per 1,000 bacterial replication cycles, or approximately 10−6 per gene per generation.

Researchers can divide the errors that DNA polymerase makes during replication into two classes:

  • Substitutions occur when the wrong (improperly paired) nucleotide is incorporated. The error level is determined by the efficiency of proofreading, in which the enzyme scrutinizes the newly formed base pair and removes the nucleotide if it is mispaired.

  • Frameshifts occur when an extra nucleotide is inserted or omitted. Fidelity with regard to frameshifts is affected by the processivity of the enzyme: the tendency to remain on a single template rather than to dissociate and reassociate. This is particularly important for the replication of a homopolymeric stretch—for example, a long sequence of dTn:dAn—in which “replication slippage” can change the length of the homopolymeric run. As a general rule, increased processivity reduces the likelihood of such events. In multimeric DNA polymerases, processivity is usually increased by a particular subunit that is not needed for catalytic activity per se.

Bacterial replication enzymes have multiple error reduction systems. The geometry of an A-T base pair is very similar to that of a G-C base pair, as is discussed in the chapter Genes Are DNA and Encode RNAs and Polypeptides. This geometry is used by high-fidelity DNA polymerases as a fidelity mechanism. Only an incoming dNTP that base pairs properly with the template nucleotide fits in the active site, whereas mispairs such as A-C or A-A have the wrong geometry to fit into the active site. On the other hand, low-fidelity DNA polymerases, such as E. coli DNA polymerase IV used for damage bypass replication, have a more open active site that accommodates damaged nucleotides, but also incorrect base pairs. Thus, either the expression or activity of these error-prone DNA polymerases is tightly regulated so that they are only active after DNA damage occurs.

All of the bacterial enzymes possess a 3′–5′ exonucleolytic activity that proceeds in the reverse direction from DNA synthesis. This provides a proofreading function, as illustrated in FIGURE 11.6. In the chain elongation step, a precursor nucleotide enters the position at the end of the growing chain. A bond is formed. The enzyme moves one base pair (bp) farther and then is ready for the next precursor nucleotide to enter. If a mistake has been made, the DNA is structurally warped by the incorporation of the incorrect base that will cause the polymerase to pause or slow down. This will allow the enzyme to back up and remove the incorrect base. In some regions errors occur more frequently than in others; that is, mutation hotspots occur in the DNA. This is caused by the underlying sequence context; some sequences cause the polymerase to move faster or slower, which affects the ability to catch an error.

FIGURE 11.6 DNA polymerases scrutinize the base pair at the end of the growing chain and excise the nucleotide added in the case of a misfit.

As noted in the section DNA Polymerases Are the Enzymes That Make DNA earlier in this chapter, replication enzymes typically are found as multisubunit holoenzyme complexes, whereas repair DNA polymerases are typically found as single subunit enzymes. An advantage to a holoenzyme system is the availability of a specialized subunit responsible for error correction. In E. coli DNA polymerase III, this activity, a 3′ to 5′ exonuclease, resides in a separate subunit, the ε subunit. This subunit gives the replication enzyme a greater fidelity than the repair enzymes.

Different DNA polymerases handle the relationship between the polymerizing and proofreading activities in different ways. In some cases, the activities are part of the same protein subunit, but in others they are contained in different subunits. Each DNA polymerase has a characteristic error rate that is reduced by its proofreading activity. Proofreading typically decreases the error rate in replication from approximately 10−5 to 10−7/bp replicated. Systems that recognize errors and correct them following replication then eliminate some of the errors, bringing the overall rate to less than 10−9/bp replicated (see the chapter titled Repair Systems).

The replicase activity of DNA polymerase III was originally discovered by a conditional lethal mutation in the dnaE locus, which encodes a 130-kD subunit that possesses the DNA synthetic activity. The 3′–5′ exonucleolytic proofreading activity is found in another subunit, ε, encoded by the dnaQ gene. The basic role of the ε subunit in controlling the fidelity of replication in vivo is demonstrated by the effect of mutations in dnaQ: The frequency with which mutations occur in the bacterial strain is increased by greater than 103-fold.

11.5 DNA Polymerases Have a Common Structure

The first DNA polymerase for which the structure was determined was the Klenow fragment of the E. coli DNA polymerase I. From those data, FIGURE 11.7 shows the common structural features that all DNA polymerases share. The enzyme structure can be divided into several independent domains, which are described by analogy with a human right hand. DNA binds in a large cleft composed of three domains. The “palm” domain has important conserved sequence motifs that provide the catalytic active site. The “fingers” are involved in positioning the template correctly at the active site. The “thumb” binds the DNA as it exits the enzyme, and is important in processivity. The most important conserved regions of each of these three domains converge to form a continuous surface at the catalytic site. The exonuclease activity resides in an independent domain with its own catalytic site. The N-terminal domain extends into the nuclease domain. DNA polymerases fall into five families based on sequence homologies; the palm is well conserved among them, but the thumb and fingers provide analogous secondary structure elements from different sequences.

FIGURE 11.7 The structure of the Klenow fragment from E. coli DNA polymerase I. It has a right hand with fingers (purple), a palm (red), and a thumb (green). The Klenow fragment also includes an exonuclease domain.

Data from: Beese, L. S., et al. 1993. “Structure from Protein Data Bank 1KFD.” Biochemistry 32:14095–14101.

The catalytic reaction in a DNA polymerase occurs at an active site in which a nucleotide triphosphate pairs with an (unpaired) single strand of DNA. The DNA lies across the palm in a groove that is created by the thumb and fingers. FIGURE 11.8 shows the crystal structure of the Φ T7 enzyme complexed with DNA (in the form of a primer annealed to a template strand) and an incoming nucleotide that is about to be added to the primer. The DNA is in the classic B-form duplex up to the last two base pairs at the 3′ end of the primer, which are in the more open A-form. A sharp turn in the DNA exposes the template base to the incoming nucleotide. The 3′ end of the primer (to which bases are added) is anchored by the fingers and palm. The DNA is held in position by contacts that are made principally with the phosphodiester backbone (thus enabling the polymerase to function with DNA of any sequence).

FIGURE 11.8 The crystal structure of phage T7 DNA polymerase shows that the template strand takes a sharp turn that exposes it to the incoming nucleotide.

Photo courtesy of Charles Richardson and Thomas Ellenberger, Washington University School of Medicine.

In structures of DNA polymerases of this family complexed only with DNA (i.e., lacking the incoming nucleotide), the orientation of the fingers and thumb relative to the palm is more open, with the O helix (O, O1, O2; see Figure 11.8) rotated away from the palm. This suggests that an inward rotation of the O helix occurs to grasp the incoming nucleotide and create the active catalytic site. When a nucleotide binds, the fingers domain rotates 60° toward the palm, with the tops of the fingers moving by 30 Å. The thumb domain also rotates toward the palm by 8°. These changes are cyclical: They are reversed when the nucleotide is incorporated into the DNA chain, which then translocates through the enzyme to recreate an empty site.

The exonuclease activity is responsible for removing mispaired bases. The catalytic site of the exonuclease domain is distant from the active site of the catalytic domain, though. The enzyme alternates between polymerizing and editing modes, as determined by a competition between the two active sites for the 3′ primer end of the DNA. Amino acids in the active site contact the incoming base in such a way that the enzyme structure is affected by the structure of a mismatched base. When a mismatched base pair occupies the catalytic site, the fingers cannot rotate toward the palm to bind the incoming nucleotide. This leaves the 3′ end free to bind to the active site in the exonuclease domain, which is accomplished by a rotation of the DNA in the enzyme structure.

11.6 The Two New DNA Strands Have Different Modes of Synthesis

The antiparallel structure of the two strands of duplex DNA poses a problem for replication. As the replication fork advances, daughter strands must be synthesized on both of the exposed parental single strands. The fork template strand moves in the direction from 5′–3′ on one strand and in the direction from 3′–5′ on the other strand. Yet DNA is synthesized only from a 5′ end toward a 3′ end (by adding a new nucleotide to the growing 3′ end) on a template that is 3′ to 5′. The problem is solved by synthesizing the new strand on the 5′ to 3′ template in a series of short fragments, each synthesized in the “backward” direction; that is, with the customary 5′–3′ polarity.

Consider the region immediately behind the replication fork, as illustrated in FIGURE 11.9. Researchers describe events in terms of the different properties of each of the newly synthesized strands:

  • On the leading strand (sometimes called the forward strand) DNA synthesis can proceed continuously in the 5′ to 3′ direction as the parental duplex is unwound.

  • On the lagging strand a stretch of single-stranded parental DNA must be exposed, and then a segment is synthesized in the reverse direction (relative to fork movement). A series of these fragments are synthesized, each 5′–3′; they then are joined together to create an intact lagging strand.

FIGURE 11.9 The leading strand is synthesized continuously, whereas the lagging strand is synthesized discontinuously.

Discontinuous replication can be followed by the fate of a very brief label of radioactivity. The label enters newly synthesized DNA in the form of short fragments of approximately 1,000 to 2,000 bases in length. These Okazaki fragments are found in replicating DNA in both prokaryotes and eukaryotes. After longer periods of incubation, the label enters larger segments of DNA. The transition results from covalent linkages between Okazaki fragments.

The lagging strand must be synthesized in the form of Okazaki fragments. For a long time, it was unclear whether the leading strand is synthesized in the same way or is synthesized continuously. All newly synthesized DNA is found as short fragments in E. coli. Superficially, this suggests that both strands are synthesized discontinuously. It turns out, however, that not all of the fragment population represents bona fide Okazaki fragments; some are pseudofragments that have been generated by breakage in a DNA strand that actually was synthesized as a continuous chain. The source of this breakage is the incorporation of some uracil into DNA in place of thymine. When the uracil is removed by a repair system, the leading strand has breaks until a thymine is inserted. Thus, the lagging strand is synthesized discontinuously and the leading strand is synthesized continuously. This is called semidiscontinuous replication.

11.7 Replication Requires a Helicase and a Single-Stranded Binding Protein

As the replication fork advances, it unwinds the duplex DNA. One of the template strands is rapidly converted to duplex DNA as the leading daughter strand is synthesized. The other remains single stranded until a sufficient length has been exposed to initiate synthesis of an Okazaki fragment complementary to the lagging strand in the backward direction. The generation and maintenance of single-stranded DNA is therefore a crucial aspect of replication. Two types of function are needed to convert double-stranded DNA to the single-stranded state:

  • A helicase is an enzyme that separates (or melts) the strands of DNA, usually using the hydrolysis of ATP to provide the necessary energy.

  • A single-stranded binding protein (SSB) binds to the single-stranded DNA, protecting it and preventing it from reforming the duplex state. The SSB binds typically in a cooperative manner in which the binding of additional monomers to the existing complex is enhanced. The E. coli SSB is a tetramer; eukaryotic SSB (also known as RPA) is a trimer.

Helicases separate the strands of a duplex nucleic acid in a variety of situations, ranging from strand separation at the growing point of a replication fork to catalyzing migration of Holliday (recombination) junctions along DNA. There are 12 different helicases in E. coli. A helicase is generally multimeric. A common form of helicase is a hexamer. This typically translocates along DNA by using its multimeric structure to provide multiple DNA-binding sites.

FIGURE 11.10 shows a generalized schematic model for the action of a hexameric helicase. It is likely to have one conformation that binds to duplex DNA and another that binds to single-stranded DNA. Alternation between them drives the motor that melts the duplex and requires ATP hydrolysis—typically 1 ATP is hydrolyzed for each bp that is unwound. A helicase usually initiates unwinding at a single-stranded region adjacent to a duplex. Note that it cannot unwind a segment of duplex DNA; it can only continue to unwind a sequence that has been started (see the chapter titled The Replicon: Initiation of Replication). It might function with a particular polarity, preferring single-stranded DNA with a 3′ end (3′–5′ helicase) or with a 5′ end (5′–3′ helicase). A 5′–3′ helicase is shown in Figure 11.10. Hexameric helicases typically encircle the DNA, which allows them to unwind DNA processively for many kilobases. This property makes them ideally suited as replicative DNA helicases.

FIGURE 11.10 A hexameric helicase moves along one strand of DNA. It probably changes conformation when it binds to the duplex, uses ATP hydrolysis to separate the strands, and then returns to the conformation it has when bound only to a single strand.

Unwinding of double-stranded DNA by a helicase generates two single strands that are then bound by SSB. E. coli SSB is a tetramer of 74 kD that binds single-stranded DNA cooperatively. The significance of the cooperative mode of binding is that the binding of one protein molecule makes it much easier for another to bind. Thus, once the binding reaction has started on a particular DNA molecule, it is rapidly extended until all of the single-stranded DNA is covered with the SSB protein. Note that this protein is not a DNA-unwinding protein; its function is to stabilize DNA that is already in the single-stranded condition.

Under normal circumstances in vivo, the unwinding, coating, and replication reactions proceed in tandem. The SSB protein binds to DNA as the replication fork advances, keeping the two parental strands separate so that they are in the appropriate condition to act as templates. SSB protein is needed in stoichiometric amounts at the replication fork. It is required for more than one stage of replication; ssb mutants have a quick-stop phenotype, and are defective in repair and recombination as well as in replication.

11.8 Priming Is Required to Start DNA Synthesis

A common feature of all DNA polymerases is that they cannot initiate synthesis of a chain of DNA de novo, but can only elongate a chain. FIGURE 11.11 shows the features required for initiation. Synthesis of the new strand can start only from a preexisting 3′–OH end, and the template strand must be converted to a single-stranded condition.

FIGURE 11.11 A DNA polymerase requires a 3′–OH end to initiate replication.

The 3′–OH end is called a primer. The primer can take various forms (see also FIGURE 11.12, which summarizes the types of priming reaction):

  • A sequence of RNA is synthesized on the template, so that the free 3′–OH end of the RNA chain is extended by the DNA polymerase. This is commonly used in replication of cellular DNA and by some viruses.

  • A preformed RNA (often a tRNA) pairs with the template, allowing its 3′–OH end to be used to prime DNA synthesis. This mechanism is used by retroviruses to prime reverse transcription of RNA (see the chapter titled Transposable Elements and Retroviruses).

    FIGURE 11.12 There are several methods for providing the free 3′–OH end that DNA polymerases require to initiate DNA synthesis.

  • A primer terminus is generated within duplex DNA. The most common mechanism is the introduction of a nick, as used to initiate rolling circle replication. In this case, the preexisting strand is displaced by new synthesis.

  • A protein primes the reaction directly by presenting a nucleotide to the DNA polymerase. This reaction is used by certain viruses (see the chapter titled Extrachromosomal Replicons).

Priming activity is required to provide 3′–OH ends to start off the DNA chains on both the leading and lagging strands. The leading strand requires only one such initiation event, which occurs at the origin. There must be a series of initiation events on the lagging strand, though, because each Okazaki fragment requires its own start de novo. Each Okazaki fragment begins with a primer sequence of RNA approximately 10 bases long that provides the 3′–OH end for extension by DNA polymerase.

A primase is required to catalyze the actual priming reaction. In E. coli, this is provided by a special RNA polymerase activity, the product of the dnaG gene. The enzyme is a single polypeptide of 60 kD (much smaller than the RNA polymerase used for transcription). The primase is an RNA polymerase that is used only under specific circumstances; that is, to synthesize short stretches of RNA that are used as primers for DNA synthesis. DnaG primase associates transiently with the replication complex, and typically synthesizes a primer of approximately 10 bases. Primers begin with the sequence pppAG positioned opposite the sequence 3′–GTC-5′ in the template.

There are two types of priming reaction in E. coli:

  • The oriC system, named for the bacterial origin, basically involves the association of the DnaG primase with the protein complex at the replication fork.

  • The Φ X system, named originally for phage Φ X174, requires an initiation complex consisting of additional components, called the primosome. This system is used when damage causes the replication fork to collapse and it must be restarted.

At times, replicons are referred to as being of the Φ X or oriC type. The types of activities involved in the initiation reaction are summarized in FIGURE 11.13. Although other replicons in E. coli might have alternatives for some of these particular proteins, the same general types of activity are required in every case. A helicase is required to generate single strands, a single-strand binding protein is required to maintain the single-stranded state, and the primase synthesizes the RNA primer.

FIGURE 11.13 Initiation requires several enzymatic activities, including helicases, single-strand binding proteins, and synthesis of the primer.

DnaB is the central component in both Φ X and oriC replicas. It provides the 5′–3′ helicase activity that unwinds DNA. Energy for the reaction is provided by cleavage of ATP. Basically, DnaB is the active component required to advance the replication fork. In oriC replicons, DnaB is initially loaded at the origin as part of a large complex (see the chapter titled The Replicon: Initiation of Replication). It forms the growing point at which the DNA strands are separated as the replication fork advances. It is part of the DNA polymerase complex and interacts with the DnaG primase to initiate synthesis of each Okazaki fragment on the lagging strand.

11.9 Coordinating Synthesis of the Lagging and Leading Strands

Each new DNA strand, leading and lagging, is synthesized by an individual catalytic unit. FIGURE 11.14 shows that the behavior of these two units is different because the new DNA strands are growing in opposite directions. One enzyme unit is moving in the same direction as the unwinding point of the replication fork and synthesizing the leading strand continuously. The other unit is moving “backward” relative to the DNA, along the exposed single strand. Only short segments of template are exposed at any one time. When synthesis of one Okazaki fragment is completed, synthesis of the next Okazaki fragment is required to start at a new location approximately in the vicinity of the growing point for the leading strand. This requires that DNA polymerase III on the lagging strand disengage from the template, move to a new location, and be reconnected to the template at a primer to start a new Okazaki fragment.

FIGURE 11.14 A replication complex contains separate catalytic units for synthesizing the leading and lagging strands.

The term enzyme unit avoids the issue of whether the DNA polymerase that synthesizes the leading strand is the same type of enzyme as the DNA polymerase that synthesizes the lagging strand. In the case we know best, E. coli, there is only a single DNA polymerase catalytic subunit used in replication, the DnaE polypeptide. Some bacteria and eukaryotes have multiple replication DNA polymerases (see the section Separate Eukaryotic DNA Polymerases Undertake Initiation and Elongation later in this chapter). The active replicase is an asymmetrical dimer with one unit on the lagging strand and one on the leading strand (see the section DNA Polymerase Holoenzyme Consists of Subcomplexes later in this chapter). Each half of the dimer contains DnaE as the catalytic subunit. DnaE is supported by other proteins (which differ between the leading and lagging strands).

The use of a single type of catalytic subunit, however, might be atypical. In the bacterium Bacillus subtilis, there are two different catalytic subunits. PolC is the homolog to E. coli’s DnaE and is responsible for synthesizing the leading strand. A related protein, DnaEBS is the catalytic subunit that synthesizes the lagging strand. Eukaryotic DNA polymerases have the same general structure, with different enzyme units synthesizing the leading and lagging strands (see the section Separate Eukaryotic DNA Polymerases Undertake Initiation and Elongation later in this chapter).

A major problem of the semidiscontinuous mode of replication follows from the use of different enzyme units to synthesize each new DNA strand: How is synthesis of the lagging strand coordinated with synthesis of the leading strand? As the replisome moves along DNA, unwinding the parental strands, one enzyme unit elongates the leading strand. Periodically, the primosome activity initiates an Okazaki fragment on the lagging strand, and the other enzyme unit must then move in the reverse direction to synthesize DNA. The next sections describe how leading and lagging strand replication is coordinated by interactions between the leading and lagging strand enzyme units.

11.10 DNA Polymerase Holoenzyme Consists of Subcomplexes

We can now relate the subunit structure of E. coli DNA polymerase III holoenzyme (also called a replisome) to the activities required for DNA synthesis and propose a model for its action. The replisome consists of the DNA polymerase III holoenzyme complex and associated proteins, primase and helicase, necessary for replication function. A new model for the structure of the DNA Pol III complex proposes a three-polymerase core structure, with two Pol III catalytic cores responsible for synthesis of the lagging strand and one for the leading strand. Each Okazaki fragment is synthesized by a new alternating core polymerase. The holoenzyme is a complex of 900 kD that contains 10 different proteins organized into four types of subcomplex:

  • There are at least two copies of the catalytic core. Each catalytic core contains the α subunit (the DNA polymerase activity), the ε subunit (the 3′–5′ proofreading exonuclease), and the θ subunit (which stimulates the exonuclease).

  • There are two copies of the dimerizing subunit, τ, which link the two catalytic cores together.

  • There are two copies of the clamp, which is responsible for holding catalytic cores onto their template strands. Each clamp consists of a homodimer of β subunits, the β ring, which binds around the DNA and ensures processivity.

  • The γ complex is a group of seven proteins, encoded by five genes that comprise the clamp loader; the clamp loader places the β clamp on DNA by opening the ring.

FIGURE 11.15 shows one of the models for the assembly of DNA polymerase III. The holoenzyme assembles on DNA in three stages:

  • First, the clamp loader uses hydrolysis of ATP to bind β subunits to a template-primer complex.

  • Binding to DNA changes the conformation of the site on β that binds to the clamp loader, and as a result it now has a high affinity for the core polymerase. This enables core polymerase to bind, and this is the means by which the core polymerase is brought to DNA.

  • A τ dimer binds to the core polymerase and provides a dimerization function that binds a second core polymerase (associated with another β2 clamp). The replisome is an asymmetric dimer because it has only one clamp loader and (at least) two copies of the catalytic core. The clamp loader is responsible for adding a pair of β2 dimers to each parental strand of DNA.

FIGURE 11.15 DNA polymerase III holoenzyme assembles in stages, generating an enzyme complex that synthesizes the DNA of both new strands.

Each of the core complexes of the holoenzyme synthesizes one of the new strands of DNA. The clamp loader is also needed for unloading the β2 clamp from DNA; as a result, the two cores have different abilities to dissociate from DNA. This corresponds to the need to synthesize a continuous leading strand (where polymerase remains associated with the template) and a discontinuous lagging strand (where polymerase repetitively dissociates and reassociates). The clamp loader is associated with the core polymerase that synthesizes the lagging strand and plays a key role in the ability to synthesize individual Okazaki fragments.

11.11 The Clamp Controls Association of Core Enzyme with DNA

The β2-ring dimer makes the holoenzyme highly processive. β is strongly bound to DNA but can slide along a duplex molecule. The crystal structure of β shows that it forms a ring-shaped dimer. The model in FIGURE 11.16 shows the β2 ring in relationship to a DNA double helix. The ring has an external diameter of 80 Å and an internal cavity of 35 Å, almost twice the diameter of the DNA double helix (20 Å). The space between the protein ring and the DNA is filled by water. Each of the β subunits has three globular domains with similar organization (although their sequences are different). As a result, the dimer has sixfold symmetry that is reflected in 12 α-helices that line the inside of the ring.

FIGURE 11.16 The subunit of DNA polymerase III holoenzyme consists of a head-to-tail dimer (the two subunits are shown in red and orange) that forms a ring completely surrounding a DNA duplex (shown in the center).

Reprinted from: Kong, X. P., et al. 1992. “Three-dimensional structure of the β.” Cell 69:425–437, with permission from Elsevier (http://www.sciencedirect.com/science/journal/00928674). Photo courtesy of John Kuriyan, University of California, Berkeley.

The β2-ring dimer surrounds the duplex, providing the “sliding clamp” that allows the holoenzyme to slide along DNA. The structure explains the high processivity—the enzyme can transiently dissociate but cannot fall off and diffuse away. The α-helices on the inside have some positive charges that might interact with the DNA via the intermediate water molecules. The protein clamp does not directly contact the DNA, and, as a result, it might be able to “ice skate” along the DNA, making and breaking contacts via the water molecules.

How does the clamp get onto the DNA? The clamp is a circle of subunits surrounding DNA; thus, its assembly or removal requires the use of an energy-dependent process by the clamp loader. The γ clamp loader is a pentameric circular structure that binds an open form of the β2 ring preparatory to loading it onto DNA. In effect, the ring is opened at one of the interfaces between the two β subunits by the δ subunit of the clamp loader. The binding of δ to the ring destabilizes and opens it, facilitated by ATP. The role of ATP is not clear, whether hydrolysis is used to open the β2 ring or for release of the clamp loader. The SSB proteins that coat the DNA are not passive, but rather are required to stimulate the process.

The relationship between the β2 clamp and the γ clamp loader is a paradigm for similar systems used by DNA polymerases ranging from bacteriophages to animal cells. The clamp is a heteromer (possibly a dimer or trimer) that forms a ring around DNA with a set of 12 α-helices forming sixfold symmetry for the structure as a whole. The clamp loader has some subunits that hydrolyze ATP to provide energy for the reaction.

The basic principle that is established by the dimeric polymerase model is that, while one polymerase subunit synthesizes the leading strand continuously, the other cyclically initiates and terminates the Okazaki fragments of the lagging strand within a large, single-stranded loop formed by its template strand. FIGURE 11.17 draws a generic model for the operation of such a replicase. The replication fork is created by a helicase—which typically forms a hexameric ring—that translocates in the 5′–3′ direction on the template for the lagging strand. The helicase is connected to two DNA polymerase catalytic subunits, each of which is associated with a sliding clamp.

FIGURE 11.17 The helicase creating the replication fork is connected to two DNA polymerase catalytic subunits, each of which is held onto DNA by a sliding clamp. The polymerase that synthesizes the leading strand moves continuously. The polymerase that synthesizes the lagging strand dissociates at the end of an Okazaki fragment and then reassociates with a primer in the single-stranded template loop to synthesize the next fragment.

We can describe this model for DNA polymerase III in terms of the individual components of the enzyme complex, as illustrated in FIGURE 11.18. A catalytic core is associated with each template strand of DNA. The holoenzyme moves continuously along the template for the leading strand; the template for the lagging strand is “pulled through,” thus creating a loop in the DNA. DnaB creates the unwinding point and translocates along the DNA in the “forward” direction.

FIGURE 11.18 Each catalytic core of Pol III synthesizes a daughter strand. DnaB is responsible for forward movement at the replication fork.

DnaB contacts the τ subunit(s) of the clamp loader. This establishes a direct connection between the helicase–primase complex and the catalytic cores. The link has two effects. One is to increase the speed of DNA synthesis by increasing the rate of movement by DNA polymerase core by 10-fold. The second is to prevent the leading strand polymerase from falling off, that is, to increase its processivity.

Synthesis of the leading strand creates a loop of single-stranded DNA that provides the template for lagging strand synthesis, and this loop becomes larger as the unwinding point advances. After initiation of an Okazaki fragment, the lagging strand core complex pulls the single-stranded template through the β2 clamp while synthesizing the new strand. The single-stranded template must extend for the length of at least one Okazaki fragment before the lagging polymerase completes one fragment and is ready to begin the next.

What happens when the Okazaki fragment is completed? All of the components of the replication apparatus function processively (i.e., they remain associated with the DNA), except for the primase and the β2 clamp. FIGURE 11.19 shows that the β2 clamp must be cracked open by the γ clamp loader when the synthesis of each fragment is completed, releasing the loop. We can think of the clamp loader here as a molecular wrench that is modulated by ATP. The clamp loader causes the β2 clamp to alter its conformation to an unstable configuration, which then springs open. A new β2 clamp is then recruited by the clamp loader to initiate the next Okazaki fragment. The lagging strand polymerase transfers from one β2 clamp to the next in each cycle, without dissociating from the replicating complex.

FIGURE 11.19 Core polymerase and the clamp dissociate at completion of Okazaki fragment synthesis and reassociate at the beginning.

What is responsible for recognizing the sites for initiating synthesis of Okazaki fragments? In oriC replicons, the connection between priming and the replication fork is provided by the dual properties of DnaB: It is the helicase that propels the replication fork, and it interacts with the DnaG primase at an appropriate site. Following primer synthesis, the primase is released. The length of the priming RNA is limited to 8 to 14 bases. Apparently, DNA polymerase III is responsible for displacing the primase.

11.12 Okazaki Fragments Are Linked by Ligase

Researchers can now expand their view of the actions involved in joining Okazaki fragments, as illustrated in FIGURE 11.20. The complete order of events is uncertain, but it must involve synthesis of RNA primer, its extension with DNA, removal of the RNA primer, its replacement by a stretch of DNA, and the covalent linking of adjacent Okazaki fragments.

FIGURE 11.20 Synthesis of Okazaki fragments requires priming, extension, removal of RNA primer, gap filling, and nick ligation.

Synthesis of an Okazaki fragment terminates just before the beginning of the RNA primer of the preceding fragment. When the primer is removed, there will be a gap. The gap is filled by DNA polymerase I; polA mutants fail to join their Okazaki fragments properly. The 5′–3′ exonuclease activity removes the RNA primer while simultaneously replacing it with a DNA sequence extended from the 3′–OH end of the next Okazaki fragment. This is equivalent to nick translation, except that the new DNA replaces a stretch of RNA rather than a segment of DNA.

In mammalian systems (where the DNA polymerase does not have a 5′–3′ exonuclease activity), Okazaki fragments are connected by a two-step process. Synthesis of an Okazaki fragment displaces the RNA primer of the preceding fragment in the form of a “flap.” FIGURE 11.21 shows that the base of the flap is cleaved by the enzyme FEN1 (flap endonuclease 1). In this reaction, FEN1 functions as an endonuclease, but it also has a 5′–3′ exonuclease activity. In DNA repair reactions, FEN1 can cleave next to a displaced nucleotide and then use its exonuclease activity to remove adjacent material.

FIGURE 11.21 FEN1 is an exo-/endonuclease that recognizes the structure created when one strand of DNA is displaced from a duplex as a “flap.” In replication it cleaves at the base of the flap to remove the RNA primer.

Failure to remove a flap rapidly can have important consequences in regions of repeated sequences. Direct repeats can be displaced and misaligned with the template; palindromic sequences can form hairpins. These structures can change the number of repeats (see the chapter titled Clusters and Repeats). The general importance of FEN1 is that it prevents flaps of DNA from generating structures that can cause deletions or duplications in the genome.

After the RNA has been removed and replaced, the adjacent Okazaki fragments must be linked together. The 3′–OH end of one fragment is adjacent to the 5′–phosphate end of the previous fragment. The enzyme DNA ligase makes a bond by using a complex with AMP. FIGURE 11.22 shows that the AMP of the enzyme complex becomes attached to the 5′ phosphate of the nick and then a phosphodiester bond is formed with the 3′–OH terminus of the nick, releasing the enzyme and the AMP. Ligases are present in both prokaryotes and eukaryotes.

FIGURE 11.22 DNA ligase seals nicks between adjacent nucleotides by employing an enzyme–AMP intermediate.

The E. coli and Φ T4 ligases share the property of sealing nicks that have 3′–OH and 5′–phosphate termini, as illustrated in Figure 11.22. Both enzymes undertake a two-step reaction that involves an enzyme–AMP complex. (The E. coli and T4 enzymes use different cofactors. The E. coli enzyme uses nicotinamide adenine dinucleotide [NAD] as a cofactor, whereas the T4 enzyme uses ATP.) The AMP of the enzyme complex becomes attached to the 5′ phosphate of the nick, and then a phosphodiester bond is formed with the 3′–OH terminus of the nick, releasing the enzyme and the AMP.

11.13 Separate Eukaryotic DNA Polymerases Undertake Initiation and Elongation

Eukaryotic replication is similar in most aspects to bacterial replication. It is semiconservative, bidirectional, and semidiscontinuous. As a result of the greater amount of DNA in a eukaryote, the genome has multiple replicons. Replication takes place during S phase of the cell cycle. Replicons in euchromatin initiate before replicons in heterochromatin; replicons near active genes initiate before replicons near inactive genes. Origins of replication in eukaryotes are not well defined, except for those in yeast (called autonomously replicating sequences [ARS], in S. cerevisiae). The number of replicons used in any one cycle is tightly controlled. During rapid embryonic development more are activated than in slower-growing adult cells.

Eukaryotes have a much larger number of DNA polymerases. They can be broadly divided into those required for replication, and repair polymerases involved in repairing damaged DNA. Nuclear DNA replication requires DNA polymerases α, β, and ε. All the other nuclear DNA polymerases are concerned with synthesizing stretches of new DNA to replace damaged material or using damaged DNA as a template. TABLE 11.2 shows that most of the nuclear replicases are large heterotetrameric enzymes. In each case, one of the subunits has the responsibility for catalysis, and the others are concerned with ancillary functions, such as priming or processivity. These enzymes all replicate DNA with high fidelity, as does the slightly less complex mitochondrial enzyme. The repair polymerases have much simpler structures, which often consist of a single monomeric subunit (although it might function in the context of a complex of other repair enzymes). Of the enzymes involved in repair, DNA polymerase β has an intermediate fidelity; all of the others have much greater error rates and are called error-prone polymerases. All mitochondrial DNA replication and recombination is undertaken by DNA polymerase γ.

TABLE 11.2 Eukaryotic cells have many DNA polymerases. The replication enzymes operate with high fidelity. Except for the β enzyme, the repair enzymes all have low fidelity. Replication enzymes have large structures, with separate subunits for different activities. Repair enzymes have much simpler structures.

DNA Polymerase Function Structure
High-fidelity replicases
α Nuclear replication 350-kD tetramer
δ Lagging strand 250-kD tetramer
ε Leading strand 350-kD tetramer
γ Mitochondrial replication 200-kD dimer
High-fidelity repair
β Base excision repair 39-kD monomer
Low-fidelity repair
ζ Base damage bypass Heteromer
η Thymine dimer bypass Monomer
ι Required in meiosis Monomer
κ Deletion and base substitution Monomer

Each of the three nuclear DNA replication polymerases has a different function, as summarized in TABLE 11.3.

  • DNA polymerase α/primase initiates the synthesis of new strands.

  • DNA polymerase ε then elongates the leading strand.

  • DNA polymerase δ then elongates the lagging strand.

TABLE 11.3 Similar functions are required at all replication forks.

Function E. coli Eukaryote Phage T4
Helicase
Loading helicase/primase
Single-strand maintenance
Priming		 DnaB
DnaC
SSB
DnaG		 MCM complex
Cdc6
RPA
Polα/primase		 41
59
32
61
Sliding clamp
Clamp loading (ATPase)		 β
γδ complex		 PCNA
RFC		 45
44/62
Catalysis
Holoenzyme dimerization		 Pol III core
T		 Polδ + Pol ε
?		 43
43
RNA removal
Ligation		 Pol I
Ligase		 FEN1
Ligase 1		 43
T4 ligase

DNA polymerase α is unusual because it has the ability to initiate a new strand. It is used to initiate both the leading and lagging strands. The enzyme exists as a complex consisting of a 180-kD catalytic (DNA polymerase) subunit, which is associated with three other subunits: the B subunit that appears necessary for assembly, and two small subunits that provide the primase (RNA polymerase) activity. Reflecting its dual capacity to prime and extend chains, this complex is often called pol α/primase.

FIGURE 11.23 shows that the pol α/primase enzyme binds to the initiation complex at the origin and synthesizes a short strand consisting of approximately10 bases of RNA followed by 20 to 30 bases of DNA (sometimes called iDNA). It is then replaced by an enzyme that will extend the chain. On the leading strand, this is DNA polymerase ε; on the lagging strand this is DNA polymerase δ. This event is called the polymerase switch. It involves interactions among several components of the initiation complex.

FIGURE 11.23 Three different DNA polymerases make up the eukaryotic replication fork. Pol α/primase is responsible for primer synthesis on the lagging strand. The MCM helicase (the eukaryotic homolog of DnaB) unwinds the dsDNA, while PCNA (homolog of α) endows the complex with processivity.

DNA polymerase ε is a highly processive enzyme that continuously synthesizes the leading strand. Its processivity results from its interaction with two other proteins, RFC clamp loader and trimeric PCNA processivity clamp (PCNA was named proliferating cell nuclear antigen for historical reasons).

Table 11.3 illustrates the conserved function of the replication components extends to the clamp loader and processivity clamp as well other functions of the replisome. The roles of RFC and PCNA are analogous to the E. coli γ clamp loader and β2 processivity unit (see the section titled The Clamp Controls Association of Core Enzyme with DNA earlier in this chapter). RFC is a clamp loader that catalyzes the loading of PCNA onto DNA. It binds to the 3′ end of the DNA and uses ATP hydrolysis to open the ring of PCNA so that it can encircle the DNA. The processivity of DNA polymerase δ is maintained by PCNA, which tethers DNA polymerase δ to the template. The crystal structure of PCNA closely resembles the E. coli β subunit: A trimer forms a ring that surrounds the DNA. The sequence and subunit organization are different from the dimeric β2 clamp; however, the function is likely to be similar.

DNA polymerase α elongates the lagging strand. Like DNA polymerase ε on the leading strand, DNA polymerase δ forms a processive complex with the PCNA clamp. The exonuclease FEN1 removes the RNA primers of Okazaki fragments. The complex of DNA polymerase δ and FEN1 carries out the same type of nick translation that E. coli DNA polymerase I carries out during Okazaki fragment maturation (see Figure 11.21). The enzyme DNA ligase I is specifically required to seal the nicks between the completed Okazaki fragments. Currently, it is not known what factor takes on the function of the E. coli τ dimer that dimerizes the polymerase complexes in order to ensure coordinated DNA replication.

11.14 Lesion Bypass Requires Polymerase Replacement

Damage to chromosomes that is not repaired before replication can be catastrophic and lethal. When the replication complex encounters damaged and modified bases such that it cannot place a complementary base opposite it, the polymerase stops and the replication fork may collapse. A cell has two options to avoid death: recombination (see the chapter titled Homologous and Site-Specific Recombination) or lesion bypass. On the leading strand in E. coli, replication can bypass a thymine dimer and can, with the DnaG primase, reinitiate forward DNA synthesis downstream. This leaves a gap behind the fork, which can be repaired by recombination, described as follows.

In addition, bacteria and eukaryotes have multiple error-prone DNA polymerases that have the ability to synthesize past a lesion on the template (see the chapter titled Repair Systems). These enzymes have this ability because they are not constrained to follow standard base pairing rules. Note that this DNA synthesis is not to repair the lesion, but simply to bypass it, to continue replication. That will allow the cell to return to the lesion to repair it.

FIGURE 11.24 compares an advancing replication fork with what happens when there is damage to a base in the DNA or a nick in one strand. In either case, DNA synthesis is halted, and the replication fork either is stalled or is disrupted and collapses. Replication-fork stalling appears to be quite common; estimates for the frequency in E. coli suggest that 18%–50% of bacteria encounter a problem during a replication cycle. E. coli has two error-prone DNA polymerases that can replicate through a lesion, DNA polymerases IV and V (see the chapter titled Repair Systems), plus the repair DNA polymerase II, that are used for translesion synthesis. Eukaryotes have five error-prone DNA polymerases with different specificities.

FIGURE 11.24 The replication fork stalls and may collapse when it reaches a damaged base or a nick in DNA. Arrowheads indicate 3′ ends.

There are two consequences when lesion bypass occurs. First, when the replication complex stalls at a lesion, the polymerase on the strand with the lesion must be removed from the template and replaced by an error-prone polymerase. Second, when the damage has been bypassed, the repair polymerase must be removed and the replication complex reinserted. When used for lesion bypass during replication, these error-prone DNA polymerases replace the replisome and are connected to the PCNA clamp temporarily to allow the lesion bypass polymerase to insert nucleotides opposite the lesion. DNA polymerase III then replaces the error-prone polymerase. The consequences can be different, depending on whether the lesion has occurred on the lagging or leading strand. The replication polymerase on the lagging strand might be more easily replaced.

Alternatively, the situation can be rescued by a recombination event that excises and replaces the damage or provides a new duplex to replace the region containing the double-strand break. The principle of the repair event is to use the built-in redundancy of information between the two DNA strands. FIGURE 11.25 shows the key events in such a repair event. Basically, information from the undamaged DNA daughter duplex is used to repair the damaged sequence. This creates a typical recombination junction that is resolved by the same systems that perform homologous recombination. In fact, one view is that the major importance of these systems for the cell is in repairing damaged DNA at stalled replication forks.

FIGURE 11.25 When replication halts at damaged DNA, the damaged sequence is excised and the complementary (newly synthesized) strand of the other daughter duplex crosses over to repair the gap. Replication can now resume, and the gaps are filled in.

After the damage has been repaired, the replication fork must be restarted. FIGURE 11.26 shows that this can be accomplished by assembly of the primosome, which in effect reloads DnaB so that helicase action can continue. Early work on replication made extensive use of phage ΦX174 and led to the discovery of a complex system for priming. A primosome assembles at a unique phage site on its single-stranded DNA called the assembly site (pas). The pas is the equivalent of an origin for synthesis of the complementary strand of ΦX174. The primosome consists of six proteins: PriA, PriB, PriC, DnaT, DnaB, and DnaC. Two alternative assembly pathways exist, one beginning with PriA and the other with PriC. This might reflect the many types of DNA damage that can occur.

FIGURE 11.26 The primosome is required to restart a stalled replication fork after the DNA has been repaired.

On ΦX174 DNA, the primosome forms initially at the pas; primers are subsequently initiated at a variety of sites. PriA translocates along the DNA, displacing SSB, to reach additional sites at which priming occurs. As in the E. coli oriC replicon, DnaB plays a key role in unwinding and priming in ΦX174 replicons. The role of PriA is to load DnaB, which in turn recruits DnaG primase to prime DNA synthesis for the conversion of single-stranded viral DNA to the double-stranded DNA form.

It has always been puzzling that when replicating in E. coli, ΦX174 origins should use a complex structure that is not required to replicate the bacterial chromosome. Why does the bacterium provide this complex? The answer is provided by the fate of the stalled replication fork. The mechanism used at oriC is specific for origin DNA sequence and cannot be used to restart replication following lesion bypass because each lesion occurs in a different sequence. A separate mechanism employing structural rather than sequence recognition is used.

The proteins encoded by the E. coli pri genes form the core of the primosome. ΦX174 has simply co-opted the primosome for its own replication. The PriA DNA helicase binds first to the single-strand region in cooperation with SSB. The key event in localizing the primosome is the ability of PriA to displace SSB from single-stranded DNA. PriA then recruits PriB and DnaT, which is then able to recruit the DnaB/C complex as described earlier (see the chapter titled The Replicon: Initiation of Replication). The alternate replisome loading system only requires PriC.

Replication fork reactivation is a common (and therefore important) reaction. It can be required in most chromosomal replication cycles. It is impeded by mutations in either the retrieval systems that replace the damaged DNA or in the components of the primosome.

11.15 Termination of Replication

Sequences that are involved with termination are called ter sites. A ter site contains a short, ~23-bp sequence. The termination sequences are unidirectional; that is, they function in only one orientation. The ter site is recognized by a unidirectional contrahelicase (called Tus in E. coli and RTP in B. subtilis) that recognizes the consensus sequence and prevents the replication fork from proceeding. The E. coli enzyme acts by antagonizing the replication helicase in a directional manner by direct contact between the DnaB helicase and Tus. Deletion of the ter sites does not, however, prevent normal replication cycles from occurring, although it does affect segregation of the daughter chromosomes.

Termination in E. coli has the interesting features shown in FIGURE 11.27. The two replication forks meet and halt in a region approximately halfway around the chromosome from the origin. In E. coli, two clusters of five ter sites each, including terK, -I, -E, -D, and -A on one side and terC, -B, -F, -G, and -H on the other, are located ~100 kb on either side of this termination region. Each set of ter sites is specific for one direction of fork movement; that is, each set of ter sites allows a replication fork into the termination region but does not allow it out the other side. For example, replication fork 1 can pass through terC and terB into the region but it cannot continue past terE, -D, and -A. This arrangement creates a “replication fork trap.” If, for some reason, one fork is delayed so that the forks fail to meet in the middle, the faster fork will be trapped at the distal ter sites to wait for the slower fork.

The trapping of the two replication forks in ter leads to transient over-replication. This must be followed by trimming and resection. The two forks must then be joined in a process resembling double-stranded break repair.

The situation is different in eukaryotes because of their linear chromosomes with multiple replicons.

FIGURE 11.27 Replication termini in E. coli are located in a region between two sets of ter sites.

Summary

  • DNA synthesis occurs by semidiscontinuous replication, in which the leading strand of DNA growing 5′–3′ is extended continuously, but the lagging strand that grows overall in the opposite 3′–5′ direction is made as short Okazaki fragments, each synthesized 5′–3′. The leading strand and each Okazaki fragment of the lagging strand initiate with an RNA primer that is extended by DNA polymerase. Bacteria and eukaryotes each possess more than one DNA polymerase activity. DNA polymerase III synthesizes both lagging and leading strands in E. coli. Many proteins are required for DNA polymerase III action and several constitute part of the replisome within which it functions.

  • The replisome contains an asymmetric dimer of DNA polymerase III; each new DNA strand is synthesized by a different core complex containing a catalytic (α) subunit. Processivity of the core complex is maintained by the β2 clamp, which forms a ring around DNA. The clamp is loaded onto DNA by the clamp loader complex. Clamp-clamp loader pairs with similar structural features are widely found in both prokaryotic and eukaryotic replication systems.

  • The looping model for the replication fork proposes that, as one half of the dimer advances to synthesize the leading strand, the other half of the dimer pulls DNA through as a single loop that provides the template for the lagging strand. The transition from completion of one Okazaki fragment to the beginning of the next requires the lagging strand catalytic subunit to dissociate from DNA and then reattach to a β2 clamp at the priming site for the next Okazaki fragment.

  • DnaB provides the helicase activity at a replication fork; this depends on ATP cleavage. DnaB can function by itself in oriC replicons to provide primosome activity by interacting periodically with DnaG, which provides the primase that synthesizes RNA.

  • The Φ X priming event also requires PriA, DnaB, DnaC, and DnaT. The importance of the primosome for the bacterial cell is that it is used to restart replication at forks that stall when they encounter damaged DNA.

References

11.1 Introduction

Research

  1. Hirota, Y., Ryter, A., and Jacob, F. (1968). Thermosensitive mutants of E. coli affected in the processes of DNA synthesis and cellular division. Cold Spring Harbor Symp. Quant. Biol. 33, 677–693.

11.2 DNA Polymerases Are the Enzymes That Make DNA

Reviews

  1. Johnson, A., and O’Donnell, M. (2005). Cellular DNA replicases: components and dynamics at the replication fork. Annu. Rev. Biochem. 74, 283–315.

  2. McHenry, C. S. (2011). DNA replicases from a bacterial perspective. Annu. Rev. Biochem. 80, 403–436.

11.3 DNA Polymerases Have Various Nuclease Activities

Reviews

  1. Hubscher, U., et al. (2002). Eukaryotic DNA polymerases. Annu. Rev. Biochem. 71, 133–163.

  2. Johnson, K. A. (1993). Conformational coupling in DNA polymerase fidelity. Annu. Rev. Biochem. 62, 685–713.

  3. Joyce, C. M., and Steitz, T. A. (1994). Function and structure relationships in DNA polymerases. Annu. Rev. Biochem. 63, 777–822.

Research

  1. Shamoo, Y., and Steitz, T. A. (1999). Building a replisome from interacting pieces: sliding clamp complexed to a peptide from DNA polymerase and a polymerase editing complex. Cell 99, 155–166.

11.7 Replication Requires a Helicase and Single-Stranded Binding Protein

Review

  1. Singleton, M. R., et al. (2007). Structure and mechanism of helicases and nucleic acid translocases. Annu. Rev. Biochem. 76, 23–50.

11.9 Coordinating Synthesis of the Lagging and Leading Strands

Review

  1. Yao, N. Y., and O’Donnell, M. (2010) Snapshot: the replisome. Cell 141, 1088–1088e1.

Research

  1. Dervyn, E., et al. (2001). Two essential DNA polymerases at the bacterial replication fork. Science 294, 1716–1719.

  2. Reyes-Lamothe, R., et al. (2010). Stoichiometry and architecture of active DNA replication machinery in E. coli. Science 328, 498–501.

11.10 DNA Polymerase Holoenzyme Consists of Subcomplexes

Review

  1. Johnson, A., and O’Donnell, M. (2005). Cellular DNA replicases: components and dynamics at the replication fork. Annu. Rev. Biochem. 74, 283–315.

Research

  1. Arias-Palermo, E., et al. (2013). The bacterial DnaC helicase loader is a DnaB ring breaker. Cell 153, 438–448.

  2. Lia, G., et al. (2012). Polymerase exchange during Okazaki fragment synthesis observed in living cells. Science 335, 328–331.

  3. Studwell-Vaughan, P. S., and O’Donnell, M. (1991). Constitution of the twin polymerase of DNA polymerase III holoenzyme. J. Biol. Chem. 266, 19833–19841.

  4. Stukenberg, P. T., et al. (1991). Mechanism of the sliding beta-clamp of DNA polymerase III holoenzyme. J. Biol. Chem. 266, 11328–11334.

11.11 The Clamp Controls Association of Core Enzyme with DNA

Reviews

  1. Benkovic, S. J., et al. (2001). Replisome-mediated DNA replication. Annu. Rev. Biochem. 70, 181–208.

  2. Davey, M. J., et al. (2002). Motors and switches: AAA+ machines within the replisome. Nat. Rev. Mol. Cell Biol. 3, 826–835.

Research

  1. Bowman, G. D., et al. (2004). Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature 429, 724–730.

  2. Jeruzalmi, D., et al. (2001). Crystal structure of the processivity clamp loader gamma (γ) complex of E. coli DNA polymerase III. Cell 106, 429–441.

  3. Stukenberg, P. T., et al. (1994). An explanation for lagging strand replication: polymerase hopping among DNA sliding clamps. Cell 78, 877–887.

11.12 Okazaki Fragments Are Linked by Ligase

Review

  1. Liu, Y., et al. (2004). Flap endonuclease 1: a central component of DNA metabolism. Annu. Rev. Biochem. 73, 589–615.

Research

  1. Garg, P., et al. (2004). Idling by DNA polymerase d maintains a ligatable nick during lagging-strand DNA replication. Genes Dev. 18, 2764–2773.

11.13 Separate Eukaryotic DNA Polymerases Undertake Initiation and Elongation

Reviews

  1. Goodman, M. F. (2002). Error-prone repair DNA polymerases in prokaryotes and eukaryotes. Annu. Rev. Biochem. 71, 17–50.

  2. Hubscher, U., et al. (2002). Eukaryotic DNA polymerases. Annu. Rev. Biochem. 71, 133–163.

  3. Kaguni, L. S. (2004). DNA polymerase gamma, the mitochondrial replicase. Annu. Rev. Biochem. 73, 293–320.

  4. Kunkel, T. A., and Burgers, P. M. (2008). Dividing the workload at a eukaryotic replication fork. Trends Cell Biol. 18, 521–527.

Research

  1. Bowman, G. D., et al. (2004). Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature 429, 724–730.

  2. Karthikeyan, R., et al. (2000). Evidence from mutational specificity studies that yeast DNA polymerases delta and epsilon replicate different DNA strands at an intracellular replication fork. J. Mol. Biol. 299, 405–419.

  3. Kumar, R., et al. (2010). Stepwise loading of yeast clamp revealed by ensemble and single-molecule studies. Proc. Natl. Acad. Sci. USA 107, 19736–19741.

  4. McElhinny, S. A., et al. (2008). Division of labor at the eukaryotic replication fork. Mol. Cell 30, 137–144.

  5. Pursell, Z. F., et al. (2007). Yeast DNA polymerase ε participates in leading-strand DNA replication. Science 317, 127–130.

  6. Shiomi, Y., et al. (2000). ATP-dependent structural change of the eukaryotic clamp-loader protein, replication factor C. Proc. Natl. Acad. Sci. USA 97, 14127–14132.

  7. Waga, S., et al. (2001). DNA polymerase epsilon is required for coordinated and efficient chromosomal DNA replication in Xenopus egg extracts. Proc. Natl. Acad. Sci. USA 98, 4978–4983.

  8. Zuo, S., et al. (2000). Structure and activity associated with multiple forms of S. pombe DNA polymerase delta. J. Biol. Chem. 275, 5153–5162.

11.14 Lesion Bypass Requires Polymerase Replacement

Reviews

  1. Cox, M. M. (2001). Recombinational DNA repair of damaged replication forks in E. coli: questions. Annu. Rev. Genet 35, 53–82.

  2. Heller, R. C, and Marians, K. J. (2006). Replisome assembly and the direct restart of stalled replication forks. Nat. Rev. Mol. Cell Biol. 7, 932–943.

  3. Kuzminov, A. (1995). Collapse and repair of replication forks in E. coli. Mol. Microbiol. 16, 373–384.

  4. McGlynn, P., and Lloyd, R. G. (2002). Recombinational repair and restart of damaged replication forks. Nat. Rev. Mol. Cell Biol. 3, 859–870.

  5. Prakash, S., et al. (2005). Eukaryotic translesion synthesis DNA polymerases: specificity of structure and function. Annu. Rev. Biochem. 74, 317–353.

Research

  1. Furukohri, A., et al. (2008). A dynamic polymerase exchange with E. coli DNA polymerase IV replacing DNA polymerase III on the sliding clamp. J. Biol. Chem. 283, 11260–11269.

  2. Lecointe, F., et al. (2007). Anticipating chromosomal replication fork arrest: SSB targets repair DNA helicases to active forks. EMBO. J. 26, 4239–4251.

  3. Loper, M., et al. (2007). A hand-off mechanism for primosome assembly in replication restart. Mol. Cell 26, 781–793.

  4. Seigneur, M., et al. (1998). RuvAB acts at arrested replication forks. Cell 95, 419–430.

  5. Yeeles, J. T. P., and Marians, K. J. (2011). The Escherichia coli replisome is inherently DNA damage resistant. Science 334, 235–238.

11.15 Termination of Replication

Research

  1. Bastia, D., et al. (2008). Replication termination mechanism as revealed by Tus-mediated polar arrest of a sliding helicase. Proc. Natl. Acad. Sci. USA 105, 12831–12836.

  2. Wendel, B. M., et al. (2014). Completion of replication in E. coli. Proc. Natl. Acad. Sci. USA 111, 16454–16459.