DNA polymerases can only add deoxynucleotides to 3' ends (i.e., strands are synthesized only in a 5' to 3' direction), but both strands at replication forks are synthesized at the same time. Thus, at the replication fork (Figure H2), the leading strand is synthesized continuously, but the lagging strand is made discontinuously as Okazaki fragments.
There are five DNA polymerases in E. coli. We have already discussed DNA polymerase I. DNA polymerase III is the replicative enzyme, also called the replicase (we will not discuss the other three DNA polymerases; they are primarily involved in various pathways to repair damaged DNA). The Table compares DNA polymerases I and III.
|DNA polymerase III||DNA polymerase I|
|Structure||DNA Pol III holoenzyme is an asymmetric dimer; i. e., two cores with other accessory subunits. It can thus move with the fork and make both leading and lagging strands.||DNA Pol I is a monomeric protein with three active sites. It is distributive, so having 5'-to-3' exonuclease and polymerase on the same molecule for removing RNA primers is effective and efficient.|
|Activities||Polymerization and 3'-to-5' exonuclease, but on different subunits. This is the replicative polymerase in the cell. Can only isolate conditional-lethal dnaE mutants. Synthesizes both leading and lagging strands. No 5' to 3' exonuclease activity.||Polymerization, 3'-to-5' exonuclease, and 5'-to-3' exonuclease (mutants lacking this essential activity are not viable). Primary function is to remove RNA primers on the lagging strand, and fill-in the resulting gaps.|
|Vmax (nuc./sec)||250-1,000 nucleotides/second. This is as fast as the rate of replication measured in Cairns' experiments. Only this polymerase is fast enough to be the main replicative enzyme.||20 nucleotides/second. This is NOT fast enough to be the main replicative enzyme, but is capable of "filling in" DNA to replace the short (about 10 nucleotides) RNA primers on Okazaki fragments.|
|Processivity||Highly processive. The beta subunit is a sliding clamp. The holoenzyme remains associated with the fork until replication terminates.||Distributive. Pol I does NOT remain associated with the lagging strand, but disassociates after each RNA primer is removed.|
|Molecules/cell||10-20 molecules/cell. In rapidly growing cells, there are 6 forks. If one processive holoenzyme (two cores) is at each fork, then only 12 core polymerases are needed for replication.||About 400 molecules/cell. It is distributive, so the higher concentration means that it can reassociate with the lagging strand easily.|
Note: Processivity is generally examined by diluting active replication forks into reaction mixtures that contain no additional template but in which the concentrations of all the enzyme components except the one under investigation have been kept constant. After dilution, the reaction proceeds unchanged if the enzyme acts processively, whereas it stops or slows if the enzyme acts distributively.
DNA Polymerase III (pol III) from E. coli is a single protein of molecular weight 130 kDa (130,000 grams per mole). It is also referred to as polC, dnaE, or the alpha subunit. Though the molecule has DNA polymerase activity by itself, polIII works to replicate DNA in the bacterial cell in conjunction with other proteins. This multi-protein complex is referred to as the pol III holoenzyme (Figure P). The proteins (called subunits) that associate with pol III in the holoenzyme perform several functions. The most interesting subunit is called beta, which forms a donut shaped ring around the DNA and helps to anchor the holoenzyme to the DNA during replication (Figure Q). By acting as a sliding "clamp", beta helps the holoenzyme to replicate long stretches of DNA without "falling off" the strand (this is called processivity). Pol III holoenzyme directs both leading and lagging strand synthesis simultaneously by virtue of having two polymerase subunits. The Table summarizes the pol III subunits, subassemblies, and their functions:
|alpha||DNA polymerase||core (there are two cores per DNA polymerase III holoenzyme)|
|epsilon||3'-to-5' exonuclease (editing exonuclease)|
|theta||stimulates 3'-to-5' exonuclease|
|tau||dimerizes cores, activates DnaB helicase activity|
|gamma||binds ATP||gamma complex (clamp loader), uses ATP energy when loading beta onto primed DNA.|
|delta prime||stimulates clamp loading|
|chi||interacts with SSB to allow removal of DnaG primase from primer|
|beta||sliding clamp. The beta subunit can be loaded onto DNA by the clamp loader (gamma complex) in an ATP-dependent reaction (Figure G1). (The clamp loader also unloads clamps!) Beta cannot be loaded onto linear DNA , covalently closed circular DNA, or single-stranded circular DNA, but it can be loaded onto nicked circles, gapped circles, and primed single-stranded circles; that is, clamp loader requires and recognizes a 3'-hydroxyl-terminus (primer-terminus). Once loaded onto a nicked circle, beta stays associated with the DNA. However, linearization of the nicked circle with a restriction endonuclease releases beta from the DNA; that is, beta is a sliding clamp. It can slide along double-stranded DNA (or DNA-RNA in double-stranded form), but cannot slide on single-stranded DNA or single-stranded DNA coated with SSB.|
In 1981, Arthur Kornberg's laboratory developed a crude soluble enzyme system capable of replicating oriC plasmids in a DnaA- and oriC-dependent manner. In addition to DNA polymerase III (the replicative enzyme), the following proteins are also required to replicate an oriC plasmid:
|DnaA||DnaA is an origin-binding protein. It binds cooperatively to the four 9-bp repeats in oriC. The initial closed complex contains the origin DNA wrapped around an assembly of 10-20 monomers of DnaA complexed with ATP. An open complex forms when the three AT-rich 13-bp repeats in oriC unwind as a consequence of the DNA wrapping around the assembly of DnaA. oriC plasmids must be negatively supercoiled in order to form open complexes (relaxed oriC plasmids can only form closed complexes). DnaA then guides the DnaB (helicase) hexameric protein from a DnaB-DnaC complex in solution to its places around each strand.|
|DnaB (helicase)||DnaB (helicase) unwinds DNA strands using ATP energy and moves processively (i. e., does not leave the DNA until replication is finished; it encircles the DNA strand) in the 5'-to-3' direction along DNA. DnaA together with the use of ATP energy is required to load DnaB (helicase) onto DNA in the form of a DnaB-DnaC complex. After loading DnaB onto the replication fork, DnaC is released from the DnaB-DnaC complex and leaves the DNA. DnaB helicase activity is stimulated more than 10-fold by making contact with DNA polymerase III holoenzyme.|
|DnaC||DnaC forms a complex with DnaB. It is required for loading DnaB onto DNA.|
|DnaG (primase)||All known DNA polymerases require a pre-exisitng "primer" from which to begin replication. E. coli uses an RNA primer which is synthesized by an enzyme called DnaG (primase). DnaG makes RNA primers (about 10 nucleotides long) that are used by DNA pol III holoenzyme to start DNA synthesis. DnaG acts distributively (does not remain associated with DNA). It drops off DNA after primer synthesis, then reloads onto DNA a second or so later by protein-protein interactions with DnaB to synthesize the next primer on the lagging strand. There are many places along the template where RNA primers can be made by primase.|
|SSB||SSB (single-strand binding protein) does not itself unwind DNA, but binds to and stabilizes unwound single-stranded DNA.|
|Gyrase (topo II)||Double-stranded DNA is unwound by DnaB helicase so that each strand can be replicated into new daughter double-stranded molecules. However, the unwinding activity of helicases at one position along a DNA molecule causes the "downstream" portion of the same DNA molecule to become OVERWOUND (remember class demonstration). Unless this overwinding is relieved by an enzyme called gyrase (topoisomerase II), DNA replication will soon stop. Gyrase uses ATP energy to introduce negative supercoiling into the DNA. The negative supercoiling compensates for the positive supercoiling generated during DNA replication. Gyrase is not limited to the site of DNA replication, but can act at any position on the DNA molecule. Gyrase can be considered as the SWIVEL for replicating molecules.|
|DNA polymerase I||Required to remove RNA primers by simultaneous action of 5'-to3' exonuclease and DNA polymerase (nick translation).|
|DNA ligase||Required to join Okazaki fragments together, uses NAD+ as energy cofactor.|
The initiation of DNA replication at oriC is shown in Figure H3.
The leading strand is synthesized continuously as shown in Figure H4.
The lagging strand is synthesized discontinuously as Okazaki fragments. Figure U shows the primase-to-polymerase switch at the replication fork. In panel (A) the hexameric DnaB helicase encircles the lagging strand, and the two core polymerases are held to the two strands of DNA by beta clamps. The tau dimer binds two cores, one gamma complex clamp loader, and contacts the DnaB helicase (this contact between tau and DnaB stimulates helicase activity). The requirement for primase to contact DnaB for activity localizes primers near the replication fork. Primase remains at the primed site through interaction with SSB. In panel (B) the gamma complex opens the beta ring in response to binding ATP. Upon binding to SSB through the chi subunit of the gamma complex, the primase-to-SSB contact becomes destabilized and dissociates from DNA. Panel (C) illustrates that the released primase can recycle to prime a new Okazaki fragment, consistent with the distributive action of primase in a replication fork system. The mechanism by which DNA polymerase acts on the lagging strand is also shown in the figure. Due to the antiparallel structure of duplex DNA, the polymerase on the lagging strand extends DNA in a direction opposite the polymerase on the leading strand. However, the lagging strand polymerase is connected to the leading strand polymerase and thus moves with the replisome. This results in a DNA loop on the lagging strand that grows larger as the lagging core polymerase draws DNA up through it until it finishes an Okazaki fragment, whereby it "bumps" into the fragment it synthesized previously. This is shown in panel (A). Upon completion of an Okazaki fragment, the lagging strand polymerase releases from DNA by severing its tie to the beta clamp, leaving the beta ring behind on the finished fragment as in panel (B). The DNA polymerase then reassociates with a new beta clamp assembled onto an upstream primer by the gamma complex (panel C).
The following experiment shows that DNA pol III* is released from beta at the termination of DNA synthesis, but beta is retained on the DNA (Figure G2). Other experiments like this one have shown that beta binds tighter to core pol III than to gamma complex in the presence of primed DNA; that is, gamma complex cannot unload beta until polymerization is finished and pol III core is released. In fact, DNA pol III* replicates a primed template to the last nucleotide before it is released from the grip of beta (Figure G3).
As shown in Figure H5, when the synthesis of the new Okazaki fragment terminates by colliding with the downstream Okazaki fragment, the RNA primer that was used to initiate the downstream Okazaki fragment is removed by the 5'-to-3' exonuclease activity of DNA polymerase I with simultaneous filling of the gap with deoxyribonucleotides (nick translation). Finally, adjacent Okazaki fragments are joined together by DNA ligase using NAD+ as the energy cofactor.
The terminus (ter) of DNA replication is opposite the origin of replication on the circular E. coli chromosome. Ter is a "trap" (Figure U1): replication forks enter, but do NOT leave, this region spanning 450 kb (1 kb = 1,000 bp). There are six ter sites in this region. A protein called Tus binds to the ter sites, and this binding stops DnaB (helicase). Leading strand synthesis terminates one nucleotide away from bound Tus. Ter sites only work in one direction. For each round of replication, only one termination site is usually used. The fork that meets a termination site in its active orientation first is stalled and waits for the other fork to reach the same site. A replication fork halted at a properly oriented termination complex therefore appears to form a barrier to the opposing replication fork, ensuring that each segment of the chromosome is replicated exactly once. Interestingly, E. coli mutants lacking the ter region or having mutated Tus are viable, but segregate non-viable cells at high frequency.
Segregation (separation) of daughter circular DNA molecules requires the action of topoisomerases. Topoisomerase IV (parC and parD) mutants are defective for segregation; the chromosomes remain interlocked circles. A shift of temperature-sensitive topoisomerase IV mutants to elevated temperatures is lethal. The defect in chromosome separation does not appear to reflect a replication defect, consistent with the finding that replication requires topoisomerase II (gyrase; gyrA and gyrB) to act as a swivel.