What is the difference between the leading strand and the lagging strand in DNA replication What are Okazaki fragments and which strand are they formed on?

Discontinuous Replication Generates Okazaki Fragments

Okazaki fragments in bacteria and in bacteriophage T4 are 1000–2000 nucleotides long, but are only about 100–300 nucleotides in eukaryotes. Because DNA polymerases cannot initiate DNA synthesis, each Okazaki fragment is primed with a short RNA. The coordination of leading and lagging strand replication and the synthesis of RNA primers for lagging strand replication are explained elsewhere in this encyclopedia. For some organisms including Escherichia coli and bacteriophage T4, the same DNA polymerase is responsible for both leading and lagging strand DNA replication. For yeast and presumably all eukaryotes, there are different DNA polymerases for leading and lagging strand DNA replication. DNA polymerase epsilon (ε) is primarily responsible for leading strand replication and DNA polymerase delta (δ) is responsible for synthesis of Okazaki fragments and lagging strand replication.

Okazaki fragment joining requires removal of the RNA primer, DNA replication to complete synthesis, and processing of the ends by nucleases to create a ‘nick’ that can be closed by the action of DNA ligase (Figure 2(b)).

Deoxyribonucleotide Polymerization

The precursors for synthesis of DNA are 5′-deoxyribonucleotide triphosphates. DNA polymerase creates a phosphodiester bond by cleaving off a pyrophosphate from the precursor and attaching it to the free 3′-hydroxy group on the growing polypeptide (Fig. 15-6). This leaves a free 3′-hydroxy group on the newly added nucleotide ready to receive the next 5′-deoxyribonucleotide.

Because of the nature of the polymerase, the direction of DNA synthesis is always in the 5′ to 3′ direction; thus nucleotide sequences are typically represented with the 5′ end on the left and the 3′ end on the right. The DNA strand that is synthesized with the 3′ end advancing in the same direction as the replication fork is called the leading strand. As might be expected, leading strand synthesis is continuous, or highly processive. In prokaryotes, this is catalyzed by DNA polymerase III (Table 15-2), which encircles the DNA helix and moves along the template strand to add the new nucleotides to the growing daughter strand.

The direction of the opposite strand creates a replication dilemma. It requires that the direction of synthesis is initiated and extended away from the direction in which the replication fork is advancing (Fig. 15-7). This strand is called the lagging strand and, because it must be continuously restarted, this process is referred to as discontinuous synthesis (compared with processive synthesis with the leading strand). Lagging strand synthesis requires sequential action of series of enzymes that initiate, elongate, and join pieces of DNA of about 1000 nucleotides called Okazaki fragments.

The synthesis of a new DNA molecule may lead to some errors, which, if not corrected, would lead to mutations in the genome. DNA polymerase needs to be attached to a helix with one complete turn, requiring from 9 to 10 nucleotides. This initial helix is provided by synthesizing an RNA primer that is subsequently removed, instead of a DNA primer that would be permanent. The removal of the primer and replacement with the corresponding DNA sequence allows for high-fidelity base pairing and a reduction in potentially damaging mutations. The primer for each new Okazaki fragment is synthesized in the 5′ to 3′ direction by primase (a DNA-dependent RNA polymerase), which is also component of the primosome along with helicase and other DNA binding proteins (Fig. 15-8). Thus each primer originates at or near the replication fork and is extended in the opposite direction. The primosome functions to keep lagging strand synthesis in synchrony with leading strand synthesis at the replication fork.

Extension of the new Okazaki fragment is accomplished by DNA polymerase III (a DNA-dependent DNA polymerase). The polymerization of deoxynucleotides continues until it reaches the 3′ hydroxyl of the primer on the prior Okazaki fragment. The primer on the prior Okazaki fragment is removed one base at a time by DNA polymerase I, which has 5′ to 3′ exonuclease activity. Each ribonucleotide is replaced with the corresponding deoxyribonucleotide, and any errors associated with the RNA primer are corrected. The last deoxyribonucleotide is joined by a different enzyme, DNA ligase, which uses one ATP to join the Okazaki fragment into the growing lagging strand.

Microbiology

12.2 Synthesis of Eukaryotic DNA

The same general principles of bacterial replication apply to eukaryotic replication, although there are differences in detail from the bacterial scheme. In eukaryotes, semi-conservative replication occurs, and as was seen in bacteria, one new strand is made continuously and the other in fragments. Both strands are made simultaneously by a replisome consisting of a helicase plus DNA polymerase holoenzyme. Within the holoenzyme, the sliding clamp loader holds each core enzyme and two sliding clamps. In addition, an RNA primer is required to initiate each new DNA strand.

Three DNA polymerases (α, δ, and ɛ) are involved in eukaryotic chromosome replication. DNA polymerase α (Polα-primase) and two associated small proteins are responsible for initiation of new strands. First, the complex makes an RNA primer. Then, polymerase α elongates the RNA primer by making a short piece of DNA about 20 nucleotides long (the initiator DNA, or “iDNA”). The single-stranded regions of the replication fork are covered by replication protein A (RPA), the eukaryotic equivalent of single-strand binding protein. Together, this whole assembly is known as the replisome (Fig. 10.29).

The functional equivalent to the bacterial sliding clamp loader is called replication factor C (RFC) and binds to the iDNA via other proteins in the pre-LC. RFC then loads the sliding clamp (PCNA protein) plus one of two DNA polymerases onto each strand of DNA. DNA polymerase ɛ is used to make the leading strand, whereas DNA polymerase δ is used for the lagging strand. These two DNA polymerases elongate the two new strands. The sliding clamp of animal cells is a trimer (not a dimer as in bacteria) that forms a ring surrounding the DNA. It was named PCNA, for proliferating cell nuclear antigen, before its role was known. Although two different polymerases are usually found in the replisome, at least in yeast, replication can take place without polymerase ɛ—apparently polymerase δ can substitute if necessary. The various families of DNA polymerase are listed in Box 10.03.

Box 10.03

DNA Polymerase Families

DNA polymerases all perform the same basic function, catalyzing the connection of a 5′-phosphate group from the incoming nucleotide to the 3′–OH group on the ribose or deoxyribose of the upstream nucleotide. Yet, there are so many different variations of structure and other functions that DNA polymerases actually fall into seven different families: (Table 10.02).

Linking of the Okazaki fragments differs significantly between animal and bacterial cells. In animals, there is no equivalent of the dual function DNA Pol I of bacteria. The RNA primers are removed by the exonucleases Fen1 and/or Dna2, and the gaps are filled by the DNA polymerase δ that is already working on the lagging strand. As in bacteria, the nicks are sealed by DNA ligase.

Eukaryotic replisomes are more complex than bacterial replisomes. They contain extra regulatory proteins and multiple different forms of DNA polymerase (α, ɛ, and δ).

12.2 Synthesis of Eukaryotic DNA

The same general principles of bacterial replication apply to eukaryotic replication, although there are differences in detail from the bacterial scheme. In eukaryotes, semi-conservative replication occurs, and as was seen in bacteria, one new strand is made continuously and the other in fragments. Both strands are made simultaneously by a replisome consisting of a helicase plus DNA polymerase holoenzyme. Within the holoenzyme, the sliding clamp loader holds each core enzyme and two sliding clamps. In addition, an RNA primer is required to initiate each new DNA strand.

In animal cells, the double helix first needs to be separated into two single strands at the origin of replication. The pre-loading complex (see Fig.10.27) recruits the helicase, minichromosome maintenance (MCM), which then moves along the helix in a 3′ to 5′ direction, separating the two strands of DNA (Fig. 10.28). MCM is similar to the bacterial helicase DnaB, since both molecules consist of multiple subunits. Three DNA polymerases (α, δ, and ɛ) are involved in eukaryotic chromosome replication. DNA polymerase α (Polα-primase) and two associated small proteins are responsible for initiation of new strands. First, the complex makes an RNA primer. Then, polymerase α elongates the RNA primer by making a short piece of DNA about 20 nucleotides long (the initiator DNA, or “iDNA”). The single-stranded regions of the replication fork are covered by replication protein A (RPA), the eukaryotic equivalent of single-strand binding protein.

The bacterial functional equivalent to the sliding clamp loader is called replication factor C (RFC) and binds to the iDNA via other proteins in the complex and loads a sliding clamp (PCNA protein) plus one of two DNA polymerases onto each strand of DNA. DNA polymerase ɛ is loaded onto the DNA for the leading strand, whereas DNA polymerase δ is used for the lagging strand. These two DNA polymerase assemblies elongate the two new strands. The sliding clamp of animal cells is a trimer (not a dimer as in bacteria) that forms a ring surrounding the DNA. It was named PCNA, for proliferating cell nuclear antigen, before its role was fully known. Although two different polymerases are usually found in the replisome, at least in yeast, replication can take place without polymerase ɛ—apparently polymerase δ can substitute if necessary. Eukaryotic replisomes also contain other regulatory proteins such as Cdc45 and a complex of four proteins called GINS for the four protein names, Go, Ichi, Nii, and San. The function of these within the replisome is still under investigation.

Linking of the Okazaki fragments differs significantly between animal and bacterial cells. In animals, there is no equivalent of the dual function DNA polymerase I of bacteria. The RNA primers are removed by the exonucleases Fen1 and/or Dna2, and the gaps are filled by the DNA polymerase δ that is already working on the lagging strand. As in bacteria, the nicks are sealed by DNA ligase.

Eukaryotic replisomes are more complex than bacterial replisomes. For example, eukaryotic replisomes have more regulatory proteins (cdc45 and GINS), and the replisome has different forms of DNA polymerase (α, ɛ, and δ) to synthesize the new DNA.

The Sequence of Primer RNA

In Escherichia coli, large amounts of Okazaki fragments can be isolated from double mutants carrying temperature-sensitive lesions in RNase H and the 5′-exonuclease domain of polymerase I. When these cells are arrested shortly after the initiation of replication, the primer RNA attached to the Okazaki fragments is found to be 11±1 nucleotides long. The primer was found to be initiated with ATP fives times more often than with other nucleoside triphosphates. A phosphodiester bond is formed between this initiating nucleotide and either ATP or GTP to make the initial diribonucleotide. In E. coli, primer synthesis begins complementary to the template trinucleotide sequence 5′-d(CTG)-3′. The guanine in the trinucleotide does not serve as a template for synthesis but is required for directing primase to the cytosine and thymine so that it will synthesize pppApG.

Primase and transcriptional RNA polymerase exhibit a high level of initiation specificity (Table 1). In all cases, once the specific purine-rich diribonucleotide has been synthesized, the rest of the primer sequence is determined by whatever the template sequence happens to be. There are several exceptions to this general rule; P4 primase recognizes only a dinucleotide template sequence; T7 primase synthesizes primarily trinucleotides and tetranucleotides ribopolymers with distinct template sequence specificity (not shown); and second of the two Thermoanaerobic tengcongensis primases completely lacks initiation specificity. This latter case is interesting because sequencing efforts have revealed a number of bacterial genomes that code for two primases, where the second one is less sequence conserved. The role of these additional primases is unclear.

Table 1. Initiation specificity of selected primases and RNA polymerases as established in biochemical assaysa

The biochemical features of primer RNA synthesis have provided a number of insights into the control of DNA replication. For bacterial and eukaryotic primases, the rate-determining step is either the rate of formation of the first phosphodiester bond or some step preceding it. Rate-limiting steps are usually subject to control. In bacteria, DnaB helicase is able to stimulate primase activity greatly and, because it unwinds duplex DNA, results in the synthesis of primers at the DNA replication fork.

In eukaryotes, replication protein A is a single-stranded DNA binding protein that is able to stimulate eukaryotic primase. After catalyzing the formation of the first bond, primases synthesize the next 10 or so bonds rapidly but then slow down. During this brief elongation phase, bacterial and eukaryotic primases readily incorporate deoxyribonucleotides into the primer to create mixed ribo- and deoxyribo-oligomers. In the absence of other replication enzymes, bacterial primer RNA is 12 or more nucleotides and eukaryotic primer RNA is 8 or more nucleotides. When either DNA polymerases or replicative helicases and their substrates are added to the primase assay mixture, the primers are limited to a length of 7–12 nucleotides. These primer lengths are the same as those observed at the ends of Okazaki fragments isolated from living organisms.

Leading and Lagging Strands

•

The lagging strand synthesis is done discontinuously. Okazaki fragments are initiated by creation of a new RNA primer by the primosome. To restart DNA synthesis, the DNA clamp loader releases the lagging strand from the sliding clamp, and then reattaches the clamp at the new RNA primer. Then DNA polymerase III can synthesize the segment of DNA.

One strand of the DNA double helix is easy to replicate because DNA polymerase III can continuously slide along this strand moving towards the fork, all the while inserting complementary bases 5´ to 3´. This strand is called the leading strand. However, on the opposite side, DNA polymerase is still required to synthesize DNA in a 5´ to 3´ direction, but the movement of the replisome relative to the template strand is 5´ to 3´. This strand is called the lagging strand. Since the newly synthesized strand must still be antiparallel to the template, the lagging strand template DNA is looped out away from the replisome and replicated in sections. The sliding clamp loader attaches the sliding clamp to the RNA primer to begin synthesis of the fragments. At the end of each fragment, the sliding clamp releases the DNA and is loaded onto the next RNA primer. Each section begins with an RNA primer. This discontinuous synthesis results in the generation of fragments on the lagging strand called Okazaki fragments.

DNA polymerase I recognizes a “nick” or break in the phosphate backbone, and then removes each RNA primer and fills the gaps with DNA. DNA ligase then covalently links the phosphate backbone.

After the replisome has passed, the lagging strand contains intervening RNA primer sequences along with gaps (missing nucleotides) in the strand and nicks (missing the covalent bond between adjacent nucleotides). More enzymes are needed to clean up the DNA. DNA polymerase I removes the RNA primer and fills in the gaps with DNA. However, DNA polymerase I cannot catalyze the reaction to remove the nicks. Another enzyme, DNA ligase, seals the nicks by forming the phosphodiester bond, thus generating a continuous sugar-phosphate backbone for the lagging strand.

2.3.2.3 DNA Ligases

DNA ligases are best known for their role in joining adjacent Okazaki fragments at the lagging strand of the replication fork; however, they are essentially involved in any process that requires sealing of phosphodiester bonds from the DNA backbone. Ligases catalyze the formation of phosphodiester bond between 3′OH and 5′phosphate on various substrates such as DNA nicks, DNA fragments with various lengths cohesive ends, DNA fragments with blunt ends and some DNA/RNA hybrids. Fig. 2.14 depicts the basics of ligase action at a DNA nicked substrate. The most widely used viral ligase is the T4 DNA ligase, which is routinely employed in cloning applications. T4 DNA ligase exhibits highest activity ligating DNA nicks and fragments with overhangs longer than 2 nucleotides, whereas fragments with blunt ends, shorter overhangs or nicks containing mismatches are processed less efficiently requiring higher enzyme concentrations and extended ligation times. Interestingly, T4 ligase can also join the 3′ end of RNA strand to a DNA strand with 5′ phosphate when the DNA fragment is annealed to a complementary DNA. Many commercial formulations of T4 ligase are available, which differ in enzyme concentration, presence of proprietary ligase activity enhancers, or buffer conditions compatible with various downstream applications such as transformation of chemical or electrocompetent cells. DNA ligases from bacteriophages T3 and T7 are also commercially available, as well as several nonviral DNA ligases, some of which are thermostable. T7 DNA ligase is the enzyme of choice when sticky ends only need to be sealed since its affinity to sticky ends grossly outweighs the one for blunt ends. T3 ligase is more salt tolerant than T4 and T7 ligases and it is the preferred ligase in applications requiring high ionic strength conditions. The use of DNA ligases in cloning is on decline due to the development of ligation-independent cloning approaches (Fig. 2.9) and recombination-based cloning technologies (Chapter 3), however, their overall value as molecular biology reagents is increasing as more and more techniques employing adapter ligation (for example, SAGE, Fig. 2.7) are being developed. Some of the new-generation sequencing methods also employ T4 DNA ligase to attach adapters to genomic DNA fragments to be sequenced.

Ligase itself drives sequencing by ligation, the newest concept in the fast changing landscape of DNA sequencing. Currently, two major platforms for whole genome sequencing employ the principle: SOLiD (Thermo Fisher, Inc.) and Complete Genomics (Beijing Genomics Institute). The SOLiD technology immobilizes DNA fragments subject to sequencing on beads attached to a slide. Bead/DNA attachment is accomplished via universal adapter. An anchor sequence, complementary to the adapter provides the 5’ phosphate group to be ligated to the 3’ hydroxyl group of the upcoming probe. Fluorescent sequencing probes consist of two known nucleotides in position 1 and 2 (depicted in color in Fig. 2.15), followed by a stretch of degenerate or universal bases depicted with N and Z. At each cycle only the probe with a perfect match of the two known nucleotides will ligate. The slide is imaged and the unextended molecules are capped by unlabeled probes to maintain cycle synchronization. The fluorophore and the terminal degenerated bases are cleaved off the probe, resulting in net 5 nt extension. The process is repeated 10 times resulting in a string of nucleotides in which two out of every five bases are identified. All ligated probes are then removed, and the entire process of probe binding, ligation, imaging, and cleavage is repeated four times, each with different anchors. At each round the new anchor is offset with one nucleotide allowing the entire sequence to be deciphered. The sequence is assembled computationally based on the imaging snapshots taken after each cycle.

The Complete Genomics technology employs fluorescent probes containing only one known nucleotide and several degenerate nucleotides. The DNA genome of interest is fragmented, cloned with flanking synthetic adapters, and then amplified to form DNA nanoballs. Each DNA nanoball is a long concatemer, product of rolling circle DNA amplification with bacteriophage phi29 DNA pol. Arrays of nanoballs are assembled in a flow cell, which is imaged after every sequencing cycle. Each cycle starts with the hybridization of an anchor sequence to one of the synthetic adapters. Then a pool of fluorescently labeled probes is flown allowing complementary probe hybridization and ligation to take place. The pool of unligated probes is washed away, and the flow cell is imaged. The same steps are performed repeatedly using a set of anchors with offset of one nucleotide, and the DNA sequence is assembled computationally. A separate sequence-specific set of anchors is used for each adapter. Sequences derived from different library clones are assembled into the sequence of the entire genome.

Cellular Functions of Eukaryotic DNA Ligases

As DNA joining is required to complete DNA replication, DNA repair, and genetic recombination, it appears likely that the multiple species of DNA ligase have evolved to participate in specific DNA transactions. Insights into the cellular functions of the different DNA ligases have been obtained by various approaches including phenotypic analysis of DNA ligase-deficient cells and the identification of interacting proteins (Table 1). The results of these studies are summarized below.

5.2.4 Okazaki Fragment Maturation

To assemble the continuous daughter DNA strand complementary to the lagging-strand template from discontinuous Okazaki fragments, an RNA primer from each of them has to be replaced by DNA to leave nicks, and the nicks need to be ligated, in the process known as Okazaki fragment maturation (OFM). To achieve this, E. coli may potentially use any of three proteins to remove the RNA primers: Pol I, RNase HI, and/or RNase HII. Pol I is equipped to replace the primers with DNA and DNA ligase A's role is to seal the remaining nick by making a phosphodiester bond between free 3′-OH and 5′-phosphate groups.

Pol I was the first DNA polymerase discovered [307]. It bears the classical 5′–3′ polymerase and 3′–5′ proofreading exonuclease activities in a large C-terminal fragment that can be generated by proteolysis, and has been termed the Klenow fragment [68,308,309]. The smaller N-terminal fragment houses its unusual 5′–3′ exonuclease activity [310,311], which is implicated in the removal of RNA primers. The two nuclease activities present in Pol I are fundamentally different; the 3′–5′ nuclease is a classic exonucleolytic proofreader (like ɛ) that cleaves the last phosphodiester bond in a nonbase-paired (mismatched or melted) 3′-end. The 5′–3′ nuclease cleaves a bond only in a base-paired region.

Traditionally it has been thought that Pol I removes primers by nick translation, where the sequential (one-by-one) removal of 5′-rNMPs (or dNMPs) at a nick occurs concomitantly with replacement DNA synthesis in the 5′–3′ direction. However, concurrent DNA synthesis was shown to stimulate the 5′–3′ exonuclease domain to occasionally remove oligo- rather than mononucleotides [312], and subsequently, the small N-terminal fragment of Pol I was shown to be capable of specific hydrolysis of the phosphodiester bond at the junction of a 5′-single-stranded overhang (or flap), leaving a nick. This process is known as flap endonuclease (FEN) activity. FEN activity of Pol I is consistent with identified homology between E. coli Pol I and both human FEN1 and yeast RAD27 FENs [313]. This is further reinforced by the homology between the E. coli Pol I FEN domain and the structurally characterized FEN domain from T. aquaticus Pol I that retains the structural fold of other known FENs [314]. The removal of oligonucleotides could operate by coupling Pol I strand-displacement activity [6] that generates ssDNA flaps at the 5′-end and subsequent FEN activity that removes them to leave a nick.

Finally, the nick is covalently sealed by the activity of DNA ligase A to form a contiguous nascent strand [71,72]. Interestingly, both Pol I and ligase are suggested to exert their activities during OFM by binding to the β2 sliding clamp that is likely left in the wake of the replication fork following Pol III HE dissociation from the template-bound Okazaki fragment [69].

It appears that at least one functional FEN, that could be associated with OFM, is required for cell viability in bacteria [315]. Many eubacteria encode a second FEN paralog in addition to that present in Pol I [316]. FEN in bacteria is encoded by the xni (Exo nine) gene and the enzyme from Firmicutes contains clustered conserved acidic residues that comprise binding sites termed Cat1 and Cat2 that coordinate one divalent metal ion each. The Cat1 metal-binding site is involved in catalytic activity whereas the Cat2-bound metal ion stabilizes the enzyme–substrate complex and may not be involved directly in catalysis of the phosphoryl transfer reaction [317,318]. However, a subset of genera including E. coli encodes a different FEN (also termed Exo IX) that lacks the three aspartate residues that make up the Cat2 site [315,316]. Indeed, the structure of E. coli Exo IX bound to a flapped DNA template shows that the Cat2 site is absent, whereas the conserved Cat1 site contains a pair of oxo-bridged Mg2 + ions [319], lending further support to the notion that the previously proposed two-metal ion mechanism based on the structure of FEN from T. aquaticus Pol I [314] is conserved in FENs across phyla.

Whereas consensus was reached for Pol I being the essential enzyme required for processing RNA primers during OFM, in vivo studies later showed that the polA gene (encoding Pol I) is dispensable in E. coli, although deletion of polA leads to high mutation frequencies and a temperature-sensitive phenotype [315,320]. Further studies showed that it is the small (FEN) domain of Pol I that is more important for the viability of E. coli than the Klenow fragment [315,321]; the presence of only the FEN-coding region of the polA gene is sufficient to provide full viability [321]. Moreover, the FEN domain of Pol I is absolutely required for cell viability in Streptococcus pneumoniae [322] and Synechococcus elongates [315], species that appear not to produce other FEN paralogs.

It has been further demonstrated that while the FEN domain from either DNA Pol I or Exo IX is sufficient for cell viability in E. coli, double null-mutants in both FEN-encoding domains are not [315]. Considering that it has been shown that in T5FEN, a FEN homolog of Exo IX in bacteriophage T5, Cat1 is essential and sufficient for endonucleolytic flap cleavage whereas both sites are required for 5′–3′ exonuclease activity [323] on flapped templates, it can be concluded that the most indispensable activity of Pol I is not its polymerization or 3′–5′ exonuclease activities, but is in fact its FEN activity. It could thus be anticipated that the FEN endonuclease activity of Pol I is utilized during the normal OFM process and this way of processing Okazaki fragments must be critical for cell survival.

The other potential Okazaki fragment processing enzymes in E. coli, RNase HI and RNase HII, have been shown to be dispensable in the presence of Pol I in live E. coli cells [315]. In addition, the fact that unlike eukaryotic RNase H, E. coli RNase HI can only digest RNA from the 3′–5′ direction and leaves at least one ribonucleotide behind [324,325], combined with the lethal phenotype of the polA, xni double null mutant further strengthens the status of Pol I in the OFM process and raises doubt that RNase HI and HII have any role in it.