The E. coli replicase, DNA polymerase III holoenzyme, contains two polymerases for replication of duplex DNA. The DNA strands are antiparallel requiring different modes of replicating the two strands: one is continuous (leading) while the other is discontinuous (lagging). The two polymerases within holoenzyme are generally thought to have asymmetric functions for
replication of these two strands. This report finds that the two polymerases have equal properties, both are capable of replicating the more difficult lagging strand. Asymmetric action is, however, imposed by the helicase that encircles the lagging strand. The helicase contact defines the leading polymerase constraining it to a subset of actions, while leaving the other to cycle on the lagging strand. The symmetric actions of the two polymerases free holoenzyme to assemble into the replisome in
either orientation without concern for a correct match to one or the other strand. Chromosomal replicases are highly processive multiprotein machines ( 16 ). Their processivity is rooted in a DNA sliding clamp protein that encircles DNA yet freely slides along it ( , ). The sliding clamp binds the polymerase acting as a mobile tether holding it to DNA for high processivity. These polymerases also require a multiprotein clamp loader that couples ATP hydrolysis to assembly of clamps onto DNA. In E. coli these activities are organized into one particle called DNA polymerase III holoenzyme (reviewed in ). The stoichiometry and general organization of these components within the holoenzyme has been determined ( ). The polymerase, called core, is composed of one subunit each of α (DNA polymerase), ε (3′-5′ exonuclelase) and θ. The β clamp is a dimer of identical subunits in the shape of a ring ( ). The clamp loader, the γ complex, contains 5 subunits (γδδ′χω) and is capable of rapid assembly of multiple β clamps onto DNA. These components are organized in the holoenzyme particle by the τ subunit, which binds two cores and one γ complex to form Pol III*. Each core polymerase within Pol III* binds a β clamp to form the holoenzyme (see Figure 1).  Figure 1Both Cores in Pol III* Are Lagging Strand Polymerases Show full caption Pol III holoenzyme was assembled onto an excess of gapped plasmid using 3H-Pol III* and 32P-βPK. (A) Idling polymerase in the presence of only two dNTPs. (B) Gaps in the DNA are filled upon adding the dNTPs. Reactions were gel filtered to resolve protein bound to DNA (fractions 10–15) from free protein (fractions 16–30). Squares32P-βPK; circles3H-Pol III*.
Two polymerase cores within one holoenzyme particle are consistent with the hypothesis that chromosomal replicases act in pairs for simultaneous replication of both strands of a duplex chromosome ( 33
Efficient in vitro replication of double-stranded DNA templates by a purified T4 bacteriophage replication system.
). In semidiscontinuous replication, the processes of replicating the leading and lagging strands is quite different. Replication of one strand is a continuous process while the other is discontinuous, being made as a series of Okazaki fragments. To accommodate this asymmetry, it is generally believed that the two polymerase cores are positioned in different chemical environments within the holoenzyme such that they have asymmetric roles, one having properties consistent with leading strand function and the other with properties for the lagging strand ( 9
Adenosine 5′-0-(3-Thiotriphosphate) can support the formation of an initiation complex between the DNA polymerase III holoenzyme and primed DNA.
, , ). Indeed, the single γ complex within the holoenzyme generates structural asymmetry. However, it is unknown whether the γ complex confers distinct properties onto the two core polymerases. Are the two cores within the holoenzyme functionally distinct, with one predestined for the leading strand, and the other for the lagging, as hypothesized? Discontinuous replication is achieved by a multistep mechanism. The processive polymerase-clamp fully extends a fragment, finishes DNA to a nick whereupon it “knows” it is complete, and disengages from DNA leaving the clamp behind. The core then reassociates with the next primed site upon which a β clamp has been preassembled. In contrast, the polymerase on the leading strand simply remains attached to its clamp for continuous extension, and therefore only performs a limited subset of actions needed on the lagging strand. This report examines the basis for asymmetric function in the E. coli replicase. The findings show that the two core polymerases in the holoenzyme are actually symmetric in function. Both have the properties required for the more difficult job of lagging strand replication. Further, the single γ complex acts equivalently with both cores in assembling β clamps onto DNA. Asymmetric function of the DNA polymerases is imposed from the outside by the helicase that unwinds DNA at the front of the E. coli replication fork. The DnaB helicase directly contacts the τ subunit of DNA polymerase III holoenzyme and holds the leading polymerase-clamp to DNA for continuous chain extension. However, the τ-to-DnaB contact does not prevent the other core from cycling on and off DNA during lagging strand replication. Hence, the DnaB–τ contact within this “replisome,” besides speeding the helicase ( ), generates asymmetry in replicase action on the two strands of duplex DNA. The fact that the two cores are not prededicated to one strand or the other, but are defined upon assembly of the holoenzyme into a replication fork, solves another problem. If the cores had been prededicated to leading and lagging strands, there would need to exist some mechanism to ensure proper assembly of the two polymerases onto their matching strands. Symmetric cores with redundant function make the assembly problem simple, reducing it to one in which the holoenzyme can assemble into the replication fork in either orientation. In addition, this report documents that the twin polymerases in one holoenzyme particle, when assembled with one core on each strand of duplex DNA, in fact do replicate both leading and lagging strands as hypothesized long ago. ResultsThe initial goal of this study was to study the functional asymmetry of the two polymerases within DNA polymerase III holoenzyme (Pol III holoenzyme). However, the two polymerases appeared symmetric in function, advancing the question of whether other proteins at the replication fork confer asymmetric function onto the two polymerases. Identification of a contact between the DnaB helicase and holoenzyme and study of the replisome on a synthetic replication fork revealed replisome composition, mechanism of assembly, and the underlying basis for asymmetric function in the holoenzyme. Pol III Holoenzyme Is Functionally SymmetricPol III holoenzyme contains two cores, but only a single γ complex (see Figure 1, diagrams). The experiment in Figure 1A was designed to determine whether the single γ complex could assemble clamps onto DNA for only one or for both cores. In Figure 1A32P-β and 3H-Pol III* were used to assemble the holoenzyme onto a molar excess of primed DNA. Gel filtration over a large pore resin was used to separate protein bound to DNA from protein that remains in solution. Pol III* and β were quantitated from their known specific activities. The results show that indeed two β clamps were bound to DNA for each Pol III*, consistent with the single γ complex loading a clamp onto DNA for each polymerase. These clamps are loaded on different primed templates each bound by a core polymerase, as linearization of the DNA does not lead to release of β as would occur if the ring were not held to DNA by the polymerase bound to a 3′ terminus (see Stukenberg and O'Donnell, 1995: Figure 7). Ability of γ complex to load β for both polymerases is demonstrated again later using a synthetic replication fork (in Figure 6). Figure 6One Pol III Holoenzyme Replicates Both Strands of Duplex DNA, and β Clamps Accumulate on the Lagging Strand during Replication Show full caption DnaB, Pol III*, and β were assembled onto replication fork DNA with EBNA1 blocking the end of the lagging strand. One core within Pol III* remained unbound to DNA by virtue of using limiting DNA. The reaction was gel filtered to remove excess free proteins. (A) Replication was initiated upon adding β and 3H-dNTPs in the presence (squares) or absence (circles) of primase. The alkaline gel analysis is shown at the right. (B) Replication was initiated upon adding 3H-β and dNTPs in the presence (squares) or absence (circles) of primase (plot to the left). At the indicated times, aliquots were quenched by removing ATP and analyzed for 3H-β on DNA by gel filtration. The plot to the right are gel filtration profile of reactions containing primase.
Figure 7Asymmetric Polymerase Action Is Determined by the Helicase Show full caption The two polymerase cores within the holoenzyme are shaded differently to show that whichever orientation the holoenzyme assumes, the resulting replisome is the same. Contact with the DnaB helicase holds the leading strand core to DNA, the lagging stand core remains free to cycle.
Next we determined the fate of the holoenzyme after replication by each of the core polymerases. The result (Figure 1B) shows that Pol III* fully dissociated from the DNA, leaving β on each of the completed DNA templates. If only one core polymerase (i.e. lagging strand half) had released the DNA, all of the Pol III* would have eluted in the excluded fractions, since one core would still remain associated with the template. These results indicate that Pol III holoenzyme is functionally symmetric. The single γ complex places β clamps onto DNA for both polymerases, and they both have the interesting property of disengaging their β clamp and DNA upon finishing a template. The τ Subunit of Pol III Holoenzyme Interacts with DnaB and PrimaseThe above results indicate that the default mode of the core polymerases in the holoenzyme is suited to the lagging strand. To gain further insight into the action of holoenzyme during replication, we examined it for interaction with other proteins that act at the replication fork using the surface plasmon resonance technique (SPR). In Figure 2A, DnaB was immobilized to a sensor chip and individual subunits of Pol III holoenzyme were passed over it. The only subunit with a detectable interaction was τ. We have confirmed the τ–DnaB interaction by gel filtration chromatography (data not shown). This interaction was also detected by another group as this work was in progress and was shown to be needed for efficient action of DnaB in unwinding duplex DNA ( ). Figure 2Interactions among Subunits Comprising the Replisome Show full caption (A) Surface plasmon resonance (SPR) analysis of interaction between immobilized DnaB and subunits of Pol III holoenzyme. The first arrow indicates time of subunit injection and the second arrow indicates termination of subunit injection. (B) SPR analysis of Pol III subunits passed over immobilized primase. (C) The τ subunit stimulates priming efficiency on ssDNA in the presence of DnaB. (D) SPR analysis of interaction between primase (immobilized) and DnaB.
In Figure 2B, a similar analysis was performed with primase attached to the sensor chip. The results show that τ and γ interact with primase, although the interaction with τ is much stronger. Does τ stimulate primase action? Primase is active on ssDNA in the presence of DnaB ( Tougu et al., 1994), and Figure 2C shows that τ stimulates this reaction. Stimulation is based in an increased frequency of primer initiation, rather than longer RNA chains (data not shown). Hence, at a moving replication fork, τ may help recruit primase. Interaction between helicase and primase has not been observed directly. In Figure 2D, direct DnaB–primase interaction was observed by SPR. The apparent Kd is weak, 8.5 μM, consistent with the inability to detect this interaction by conventional methods. Assembly of DnaB onto DNAInteractions among replication fork proteins suggest that a stabile replisome may be assembled onto DNA. As a first step in this assembly, the association of DnaB helicase onto a replication fork was studied (Figure 3). To aid the analysis, a protein kinase recognition sequence was engineered onto the N-terminus of DnaB and radiolabeled using γ32P-ATP and protein kinase. DnaB modified in this manner retained full ssDNA- dependent ATPase, full activity in stimulating primase action on ssDNA, and 80% activity in oriC-dependent plasmid replication (data not shown). Figure 3Assembly of DnaB Helicase onto DNA Show full caption The scheme at the top provides an explanation for the data shown below the scheme. In the scheme, the ssDNA binding sites in the DnaB hexamer are in the center of the ring. ATP binding opens an intraprotomer interface allowing access to ssDNA. 32P-DnaB was used to follow its interaction with a variety of DNA substrates in gel filtration analyses. 32P-DnaB on DNA elutes in fractions 10–15, free protein elutes later. (A) 32P-DnaB, in the presence of ATP, binds circular ssDNA (squares) but not circular dsDNA (circles). (B) DnaB binds circular ssDNA in the presence of AMPPNP (squares), but not in the absence of nucleotide (circles). (C) DnaB does not bind ssDNA precoated with SSB (circles), but after preloading DnaB on ssDNA, SSB does not displace it (squares). (D) Treatment of DnaB with ATP followed by hexokinase and glucose prevents assembly onto ssDNA (circles), but after preloading DnaB on ssDNA, removal of ATP does not release it (squares). (E) DnaB binds end-labeled oligo dT65 in the presence or absence of ATP (squares and diamonds, respectively). Circles show oligo dT65 in the absence of DnaB. (F) One 32P-EBNA1 dimer binds duplex DNA containing an EBNA1 binding site (squares). 32P-EBNA1 in the absence of DNA is shown by the circles. (G) In the absence of ATP, DnaB binds a circular nicked duplex with a 5′ tail containing a hairpin duplex (the EBNA1 binding site) (squares). EBNA1 prevents DnaB from binding DNA in the absence of ATP (circles). (H) Addition of DnaB followed by EBNA1 results in one each bound on DNA without ATP.
32 P-DnaB was incubated with ATP, and either circular ssDNA or dsDNA, then gel filtered. The result (Figure 3A) shows that ∼40 DnaB hexamers assembled onto ssDNA (fractions 10–14), while no detectable interaction with dsDNA was observed within the resolution of this assay (DnaB is large and thus is partly excluded in the absence of DNA). The experiments in Figure 3A were performed in the presence of ATP. The nonhydrolyzable analog, AMPPNP, also supported interaction of 32P-DnaB with circular ssDNA; however, omission of nucleotide did not, indicating that ATP binding is necessary and sufficient for DnaB to bind circular ssDNA (Figure 3B). SSB inhibited interaction of DnaB with ssDNA when added to DNA before DnaB, but not after DnaB (Figure 3C). 32 P-DnaB is retained on DNA during the 20 min gel filtration procedure even though the filtration buffer lacks nucleotide. Hence once DnaB is on DNA, ATP is no longer required for DNA binding. DnaB interaction with ssDNA was prevented by removing ATP with hexokinase and glucose prior to adding ssDNA (Figure 3D). This treatment does not remove DnaB preassembled onto ssDNA. Hence, ATP induces a transient change in DnaB necessitating the copresence of ssDNA for assembly onto DNA. Two recent electron microscopy studies show the DnaB hexamer is ring shaped ( , ). Binding of DNA through the DnaB ring is consistent with the small site size of 20 nucleotides/DnaB, an insufficient length to wrap around the hexamer ( 2
Interactions of Escherichia coli primary replicative helicase DnaB protein with single-stranded DNA.
). Further, images of two other ring-shaped helicases, RuvB ( ) and T7 g4p ( 7
T7 helicase/primase proteins form rings around single-stranded DNA that suggest a general structure for hexameric helicases.
), appear with DNA through the center. The results of Figure 3 can be explained by ssDNA binding sites sequestered on the inner surface of the DnaB ring which, upon ATP binding and ring opening, become available for interaction with circular ssDNA. This model predicts that DnaB will bind linear ssDNA in the absence of ATP. In Figure 3E, oligo dT65 was end labeled, mixed with DnaB, then gel filtered. The result shows that DnaB binds the linear ssDNA in the presence of ATP as well as in the absence of ATP, consistent with a ring that simply slides over an end while remaining closed in the absence of ATP. Next, this information was used to develop a replication fork structure that would allow DnaB to assemble without ATP and therefore would not unwind DNA (until ATP addition). A circular dsDNA template was constructed containing a 5′ tail with a terminal 18 bp duplex hairpin and an internal 31 nucleotide stretch of ssDNA. Consistent with a ring that slips onto DNA32P-DnaB bound the 5′-tailed circular duplex in the absence of ATP (Figure 3G). The 18 bp duplex at the 5′ end contains the binding site for EBNA1, the latent origin binding protein of EBV. 32P-DnaB was unable to bind the DNA when EBNA1 was added before DnaB (Figure 3G). In Figure 3H, side-by-side experiments were performed that contained EBNA1 and DnaB but differed in which was labeled with 32P. The result shows approximately one each of DnaB and EBNA1 on the DNA indicating they both bind their respective sites at the same time. Assembly of the ReplisomeRolling circle systems that synthesize both leading and lagging strands are well established in the T4 and E. coli systems ( ). The E. coli system utilizes duplex circular substrates with a 5′ ssDNA tail containing a primosome assembly site. Several proteins act upon this site to assemble DnaB onto ssDNA ( 16
). However, once DnaB is assembled on DNA, these factors are not required for replication fork propagation with primase and Pol III holoenzyme ( 23
The Escherichia coli preprimosome and DnaB helicase can form replication forks that move at the same rate.
). In Figure 4, a 5′-tailed fully duplex circular DNA was used to assemble DnaB, Pol III*, β, and primase onto the tail in the absence of ATP, and therefore in the absence of strand separation. dCTP and dATP were added to support Pol III holoenzyme assembly and to prevent the 3′-5′ proofreading nuclease from digesting the 3′ end. DnaB action is specific to rNTPs ( ), and therefore strand separation is held in check by their omission. To measure the stoichiometry of each component on DNA, the assembly reaction was repeated using either 32P-DnaB3H-β3H-Pol III*, or 3H-primase, followed by gel filtration (Figure 4). The results show that approximately one each of DnaB, β, and Pol III* assembled on the synthetic fork. 3H-primase did not efficiently assemble with the other proteins (Figure 4D), suggesting that the replisome contains helicase and holoenzyme, but lacks firm attachment of primase. Low stability of primase in the replisome is consistent with its distributive action in the E. coli and T7 systems ( 43
Coordinated leading- and lagging-strand synthesis at the Escherichia coli DNA replication fork. IV. Reconstitution of an asymmetric, dimeric DNA polymerase III holoenzyme.
, ). Helicase Stabilizes Pol III Holoenzyme on Completed DNAIt is interesting to note that Pol III* normally disengages from β upon completing a template (see Figure 1), yet Pol III* was retained on the completely duplex circle in Figure 4. Either the 5′ ssDNA tail stabilizes Pol III* on DNA or the τ-to-DnaB contact holds Pol III* to DNA. To distinguish these, the holoenzyme was assembled onto the 5′-tailed fully completed circular duplex DNA (containing only a nick) and then gel filtered (Figure 5A). The result shows that the ssDNA tail was insufficient to hold 3H-Pol III* to the DNA. In Figure 5B, this experiment was repeated in the presence of DnaB. Now 3H-Pol III* remained stably associated with the DNA. As the tailed duplex does not retain Pol III* in the absence of DnaB, but does in the presence of DnaB, it may be concluded that contact between DnaB and Pol III holoenzyme prevents release of the polymerase from completed DNA. The β subunit is essential for the DnaB-mediated stability of 3H-Pol III* (data not shown). The Two Polymerases in One Pol III* Catalyze Leading and Lagging Strand SynthesisFigure 5DnaB Helicase Stabilizes Polymerase on Completed DNA Show full caption 3H-Pol III* was assembled with β on limiting duplex circular DNA containing a 5′ ssDNA tail at the site of the nick in either the absence (A) or presence (B) of DnaB. Reactions were gel filtered to resolve 3H-Pol III* bound to DNA (fractions 10–15) from 3H-Pol III* in solution.
In Figure 4 and Figure 5, the replisome was assembled using limiting DNA to position only one core in Pol III* on a β clamp (leading). Is the second core active for lagging strand replication? The Pol III holoenzyme-DnaB complex was assembled onto limiting DNA, then gel filtered. Gel filtration removes excess Pol III* not bound to DNA thereby preventing extension of lagging strand fragments by Pol III* molecules other than the one already on DNA. After filtration, primase and β were added and replication was initiated. Most of the DNA templates (70%–80%) participated in rolling circle replication as determined from the amount of DNA remaining in the initial position on a native agarose gel (data not shown). Alkaline gel analysis resolves the continuous leading strand from the discontinuous fragments (Figure 6A). In the absence of primase, only leading strands were observed, but the complete system shows both leading and lagging strands. The relative amounts of leading and lagging strand synthesis were within 15%. Addition of 43
Coordinated leading- and lagging-strand synthesis at the Escherichia coli DNA replication fork. IV. Reconstitution of an asymmetric, dimeric DNA polymerase III holoenzyme.
). These results show that the two core polymerases within one molecule of Pol III* do indeed replicate both strands and confirmed that the single γ complex assembles β clamps onto DNA for both leading and lagging strands. Stoichiometric Use of β on the Lagging StrandBased on Pol III holoenzyme behavior on circular ssDNA templates, it is hypothesized that the lagging strand polymerase hops from one β clamp to another upon completing each Okazaki fragment, leaving the clamps behind on DNA ( , ). This stoichiometric use of β for each Okazaki fragment predicts that β clamps will accumulate on the lagging strand. This hypothesis was tested by gel filtering the replisome, adding 3H-β and primase, and initiating fork movement. At different times the reaction was halted by removing ATP/dATP with hexokinase and glucose. The aliquots were gel filtered to quantitate 3H-β accumulation on the DNA. β clamps slide off DNA ends and therefore EBNA1 was bound to the 5′ hairpin end of the lagging strand to block their escape. The results show that β clamps accumulate on DNA (Figure 6B). Accumulation of β requires primase, indicating that β accumulates specifically on the lagging strand. The replication fork advances at 600–800 nt/s ( 23
The Escherichia coli preprimosome and DnaB helicase can form replication forks that move at the same rate.
, 43
Coordinated leading- and lagging-strand synthesis at the Escherichia coli DNA replication fork. IV. Reconstitution of an asymmetric, dimeric DNA polymerase III holoenzyme.
). Assuming 70% utilization of input DNA, a rate of 720 nt/s is calculated from the observed DNA synthesis at 5 min (Figure 6A). In the first 30 s each fork should become extended ∼21 kb. The observed length of Okazaki fragments is 3 kb on average, for a total of ∼7 fragments on each active template, essentially the same number of β clamps on each active DNA (4.5 ÷ 0.7 = 6.4). After 5 min, forks achieve a length of about 216 kb, for an average of 72 Okazaki fragments. Approximately 24 β clamps are observed per active template (17 ÷ 0.7 = 24), suggesting that some (66%) clamps dissociate from DNA during the 5 min reaction. Yet, β clamps are exceedingly stable on DNA (half-life of 72 min; 45
Clamp loading, unloading and intrinsic stability of the PCNA, β, and gp45 sliding clamps of human, E.
). It may be presumed that γ complex, which both loads and unloads clamps from DNA, assists dissociation of β during ongoing fork movement ( , ). Clamp unloading during replication is consistent with a need to recycle β for reuse on the 10-fold greater amount of Okazaki fragments compared to β ( ). DiscussionThe Replisome Coordinately Replicates Both Strands of DNAIt has long been hypothesized that replicative polymerases act in pairs for simultaneous synthesis of both strands of duplex DNA ( 33
Efficient in vitro replication of double-stranded DNA templates by a purified T4 bacteriophage replication system.
). Studies in the T4, T7, and E. coli systems have shown that polymerase action is processive on both leading and lagging strands of a replication fork in vitro ( 32
Studies on the T4 bacteriophage DNA replication system.
, , ). In those studies, dilution of a preassembled replication fork did not significantly diminish replication of both strands, indicating that the polymerases act processively on both strands. However, it remained possible that each strand contained its own independent processive polymerase. For example, one Pol III holoenzyme on the lagging strand could conceivably use its two polymerases to “walk” from one primer to the next thereby achieving processivity. In this study, the two core polymerases of one Pol III holoenzyme were assembled onto the replication fork in stages to specifically place one core on the leading strand and the other core on the lagging strand. The result confirmed that one molecule of Pol III holoenzyme could indeed simultaneously replicate both strands of DNA (see Figure 6). Further, the single γ complex assembled β clamps onto DNA for both polymerases. Finally, β clamps accumulated on the lagging strand showing that β clamps are used stoichiometrically and that the single γ complex repeatedly loads β clamps onto the lagging strand. Helicase Determines Asymmetric Function in Pol III HoloenzymeDuring ongoing fork movement, the leading strand is synthesized continuously and the lagging strand is synthesized discontinuously as a series of fragments. Thus the leading strand core need simply be highly processive, a property conferred onto it by the β DNA sliding clamp. However, the lagging strand core needs to cycle off DNA each time it completes a lagging strand fragment, in order to associate with an upstream primed site for the next fragment. Previous studies indicate that this cycling on the lagging strand occurs by a mechanism in which the core retains tight affinity to the β clamp during fragment extension, but rapidly disengages its β clamp upon finishing a fragment (leaving the β clamp on DNA), followed by rapid reassociation with a new β clamp on the next primed site ( ). The different tasks of leading and lagging strand synthesis suggest that the two cores within Pol III holoenzyme are functionally distinct, one that is enabled to hop among β clamps (lagging) and the other that remains fast to its β clamp for continuous extension (leading). However, the two DNA polymerases in the holoenzyme are encoded by the same gene and thus are chemically equivalent. It has been hypothesized that they obtain their different properties for leading and lagging strands by being placed in different environments ( 9
Adenosine 5′-0-(3-Thiotriphosphate) can support the formation of an initiation complex between the DNA polymerase III holoenzyme and primed DNA.
, , ). In this view, the asymmetric function of the two cores is conferred by differential juxtaposition of the several accessory proteins relative to the two polymerases. The architecture of the holoenzyme, elucidated through its assembly from individual subunits, shows it has only one γ complex for the two cores ( ). Although the single γ complex may place the two cores in different environments for asymmetric function, this has not been tested. One line of evidence that suggested functional asymmetry was that ATPγS (hydrolyzed by the holoenzyme) supports approximately one-half the normal amount of β clamps assembled onto DNA ( 9
Adenosine 5′-0-(3-Thiotriphosphate) can support the formation of an initiation complex between the DNA polymerase III holoenzyme and primed DNA.
). On hindsight this may be explained by altering the γ complex catalyzed equilibrium of clamps on and off DNA such that fewer clamps are assembled on DNA using ATPγS than with ATP. This interpretation is consistent with ATPγS-mediated disassembly of Pol III holoenzyme from DNA ( ). ATPγS also supports a suboptimal level of PCNA loading in both the yeast and human polymerase δ holoenzyme, systems that lack a dimeric polymerase ( , 3
Saccharomyces cerevisiae replication factor C. II. Formation and activity of complexes with the proliferating cell nuclear antigen and with DNA polymerases δ and ε.
). This report shows that the two cores in the holoenzyme are not functionally distinct after all. Both are capable of recognizing completed DNA for disengaging from β. Further, the single γ complex assembles β onto DNA for both cores. Is there asymmetric function of the leading and lagging strand polymerases at the replication fork, and if so, where does it come from? This report reveals that the asymmetric function is imposed onto the holoenzyme by the DnaB helicase. The directionality of DnaB activity places it on the lagging strand ( ). Since the holoenzyme is functionally symmetric, either core can assemble onto either strand of the replication fork with the same consequence (see Figure 7). But the contact between DnaB and Pol III holoenzyme (mediated by τ) stabilizes the core on the continuous leading stand, leaving the core polymerase on the lagging strand free to cycle. The τ Subunit Is an “Organizing Center” of the ReplisomeThe τ subunit plays a central role in organization of the holoenzyme particle and of the replisome. τ cements together two core polymerases and one γ complex into a holoenzyme particle ( ), and it binds the DnaB helicase and primase. The interaction of τ with DnaB fulfills at least two functions: one, it speeds up helicase action at the replication fork ( ), and two, it prevents the leading strand polymerase from cycling off DNA (this study). The τ subunit also interacts with primase and stimulates the frequency of priming. τ also appears to prevent γ complex from removing β clamps from DNA during synthesis ( ). It seems likely that τ, as a replisome organizer, may signal primase to initiate primer synthesis on the lagging strand. Interaction of the Replicative Helicase, DnaB, with DNAThe helicases RuvB, T7 g4p, and E. coli DnaB form closed rings with a central cavity large enough to accommodate DNA ( , 7
T7 helicase/primase proteins form rings around single-stranded DNA that suggest a general structure for hexameric helicases.
San Martin et al., 1995; ). The RuvB holiday junction branch migrating helicase appears to encircle dsDNA in the electron microscope ( ). Similar studies of T7 g4p suggest that it encircles ssDNA ( 7
T7 helicase/primase proteins form rings around single-stranded DNA that suggest a general structure for hexameric helicases.
). This report provides biochemical evidence that the E. coli replicative helicase, DnaB, encircles ssDNA. DnaB requires ATP transiently to assemble onto circular ssDNA, but not a short linear ssDNA, suggesting that the DnaB ring uses ATP to open up and reclose around ssDNA. Inability to bind circular ssDNA without ATP suggests that the ssDNA binding sites are occluded, perhaps positioned on the inner surface of the ring, and therefore inaccessible unless it encircles the DNA. Generality of These Principles to Eukaryotes and Phage T4These replicative mechanisms of E. coli generalize to phage T4. In the T4 system, a gp44/62 complex assembles gp45 clamps onto DNA which in turn confers processivity to gp43 DNA polymerase ( ). Further, the T4 polymerase rapidly dissociates from its ring upon finishing a template, similar to the E. coli system ( 8
The rapid dissociation of the T4 DNA polymerase holoenzyme when stopped by a DNA hairpin helix a model for polymerase release following the termination of each Okazaki fragment.
). In E. coli, the three replicase components are held together by the τ subunit, which also mediates the helicase contact that enhances helicase speed and holds the polymerase to the continuous leading strand. Although the equivalent of the τ subunit for polymerase dimerization has yet to be found in T4, two cores with symmetrical function in the E. coli Pol III holoenzyme is consistent with the use in T4 of two identical gp43 polymerases at a replication fork. It is interesting to note that the polymerases of T4 and T7 interact with their respective helicases, and therefore may not require an intermediary τ-like protein ( , 29
Trypsin cleavage in the COOH terminus of the bacteriophage T4 gene 41 DNA helicase alters the primase-helicase activities of the T4 replication complex in vitro.
). Eukaryotes, from yeast to humans, also utilize similar replication strategies to E. coli and T4. The PCNA clamp is assembled onto DNA by the RFC clamp loader, and it provides processivity to Pol δ ( ). RFC, besides loading clamps onto DNA, also recycles PCNA from DNA ( 45
Clamp loading, unloading and intrinsic stability of the PCNA, β, and gp45 sliding clamps of human, E.
). As with T4, no τ-like protein has yet been identified to hold two eukaryotic polymerases together. However, the symmetrical action of the cores within Pol III holoenzyme, along with the fact that helicase imposes asymmetry for function at the replication fork, implies that the polymerases on the two strands of the eukaryotic fork need not be chemically different as often proposed (i.e., Pol ε and Pol δ), but may instead be of only one type (e.g., one Pol δ on each strand) with asymmetric action imposed by a helicase. Experimental ProceduresMaterialsRadioactive nucleotides, New England Nuclear; unlabeled nucleotides, Pharmacia; DNA modification enzymes, New England Biolabs. SPR buffer is 10 mM HEPES-NaOH (pH 7.4), 150 mM NaCl, 3.4 mM EDTA, and 0.005% Tween 20. Buffer B is 20 mM Tris–HCl (pH 7.5), 0.1 mM EDTA, 5 mM DTT, 5% glycerol, 40 μg/ml BSA, and 8 mM MgCl2. Buffer C is 20 mM Tris HCl (pH 8), 0.1 mM EDTA, 5 mM DTT, 20% glycerol, 5 mM MgCl2, and 50 mM NaCl. Buffer D is Buffer C with no NaCl, but containing 100 mM ATP. ProteinsTritiated Pol III Assemblies and PrimaseSubunits of Pol III holoenzyme were purified as described: α, ε, γ, τ, and SSB ( ), β ( ), δ and δ′ ( ), χ and ω ( ), θ ( ). Pol III* was constituted from pure subunits and purified ( ). Primase was purified as described ( 30
Primase, the DnaG protein of Escherichia coli.
). 3H-θ3H-β, and 3H-primase were prepared by reductive methylation ( 11 Kelman, Z., Naktinis, V., and O'Donnell, M. (1995a). Radiolabeling of proteins for biochemical studies. In Methods in Enzymology, DNA Replication, J. Campbell, ed. (Orlando, Florida: Academic Press) 262, 430–442.
). The specific activities were: 3H-θ, 15 cpm/fmol; 3H-β, 67 cpm/fmol; and 3H-primase, 190 cpm/fmol. The 3H-subunits retained at least 90% activity. 3H-Pol III*, labeled in θ, was constituted as described ( 39
Assembly of a chromosomal replication machine two DNA polymerases, a clamp loader and sliding clamps in one holoenzyme particle. V. Four different polymerase-clamp complexes on DNA.
); specific activity was 33.8 cpm/fmol. 32P-labeled β, DnaB, and EBNA1To label proteins with 32P, we engineered expression vectors containing a modified version of the gene for β, DnaB, and EBNA1 ( 11 Kelman, Z., Naktinis, V., and O'Donnell, M. (1995a). Radiolabeling of proteins for biochemical studies. In Methods in Enzymology, DNA Replication, J. Campbell, ed. (Orlando, Florida: Academic Press) 262, 430–442.
). βPK and EBNA1PK have the six amino acid recognition motif for cAMP-dependent protein kinase at their C-terminus. DnaB was cloned into the pHK vector ( ), which places a 23-residue extension containing a His Tag and a kinase motif onto the N-terminus. All proteins were phosphorylated as in Kelman et al., 1995b; NaF was deleted from the kinase buffer. Because DnaB has high ATPase activity, 1 mM CTP was included in the kinase reaction (CTP is hydrolyzed by DnaB, but does not interfere with the kinase). ATP was removed by gel filtration. Specific activities of βPK, EBNA1PK and DnaBPK were 20–65 cpm/fmol. Purification of DnaBPK was as follows: 12 liters of BL21(DE3)pLysS harboring pHK-DnaB were grown at 37°C in LB containing 100 μg/ml ampicillin and 25 μg/ml chloramphenicol. After reaching an OD595 of 0.5, IPTG was added to 1.5 mM. After 3 hr of induction cells were harvested by centrifugation (30 g paste) followed by heat lysis ( ). To the clarified supernatant (Frxn I, 445 mg) was added MgCl2 and ATP to final concentrations of 5 and 1 mM, respectively. The supernatant was treated with 0.2 g/ml of ammonium sulfate and the pellet (Frxn II) was dialyzed against 2 L of Buffer C. The dialyzed Frxn II (20 ml, 337 mg) was applied to a 5 ml ATP-Agarose column (Sigma, A-6888) equilibrated in Buffer C, followed by three washes: 20 ml Buffer C, 20 ml Buffer C containing 100 mM AMP, then 20 ml Buffer C. DnaBPK was eluted with 30 ml of Buffer C containing 2 M NaCl and 5 mM ATP. The eluted DnaBPK (21 mg in 13 ml) was dialyzed against Buffer D. Next, 7 mg of DnaBPK was applied to an 8 ml MonoQ column equilibrated in Buffer D, and eluted with a 100 ml linear gradient of 50 mM NaCl to 0.5 M NaCl in Buffer D. Fractions of 2.5 ml were collected. Fractions 37–40 were pooled (4 mg, 11 ml), dialyzed against Buffer C containing 5 μM ATP, aliquoted, and stored at −70°C. Wild-type DnaB, from N100 cells containing the DnaB overproducing plasmid, pRLM29 (a gift of Dr. Roger McMacken, Johns Hopkins University), was purified by the same procedure; 300 L of induced cells yielded 250 mg of pure protein. DNAsGapped plasmid was prepared by nicking pBS (Stratagene) with gpII (gift of Drs. P. Model and K. Horiuchi, Rockefeller University) and treating with exo III ( ). M13mp18 ssDNA was phenol extracted from pure phage ( 41 Turner, J., and O'Donnell, M. (1994). Cycling of E. coli DNA polymerase III from one sliding clamp to another: model for the lagging strand. In Methods in Enzymology, DNA Replication, J. Campbell, ed. (Orlando, Florida: Academic Press) 262, 442–449.
). Circular double strand DNAs containing 5′ “tails” (Tailed Form II, TFII) were prepared by annealing 250 pmol of oligonucleotide (Oligos Etc.) to 20 pmol of M13mp18 ssDNA. The primer was extended full circle by incubating 16 fmol of primed DNA with 62.5 U of Pol I Klenow fragment in 1.25 ml buffer containing 100 μM of each dNTP in 10 mM Tris–HCl (pH 7.5), 5 mM MgCl2, and 7.5 mM DTT. After 45 min, sufficient time for complete synthesis, the reaction was stopped with 20 ml 0.5 M EDTA, followed by phenol/chloroform extraction, and gel filtration on BioGel A15M equilibrated in 0.5 mM EDTA, and 300 mM sodium acetate (pH 5.2). Fractions containing DNA were quantitated from A260. Two different primer were used to make the TFII DNA, one was a 73-mer, which hybridized to 27 nucleotides of M13mp18 (map position 6808–6834), and a 5′ ssDNA tail of dT46 (referred to as “TFII”); the other was an 89-mer consisting of 18 hybridized nucleotides (map position 6625–6606) and a 5′ tail of dT31 followed by an 18 bp duplex (EBNA1 site) connected by a loop of 5 dT (referred to as “TFII-EBNA1”). Interaction of 3H-Pol III* and 32P-β with DNABoth reactions contained 600 fmol 3H-θ Pol III*, 3000 fmol pBS, 8 μg SSB and 1700 fmol 32P-β (as dimer). Protein and DNA were incubated in 100 μl Buffer B with 60 μM dGTP and dCTP, 0.5 mM ATP and 40 mM NaCl for 2 min at 37°C to allow clamp loading and polymerase idling. Then 60 μM dATP and dTTP were added to one reaction. Replication was for 30 s at 37°C, then 20 μl was removed for product analysis by agarose gel. The remaining 80 μl was applied to a 5 ml A-15M column at 4°C. For the idling reaction, the column was equilibrated in Buffer B with 125 mM NaCl and 60 μM dGTP and dCTP. For the replicated sample, the column buffer also contained 60 μM dATP and dTTP. Fractions of 200 ml were collected and proteins were quantitated from their known specific activities. Surface Plasmon ResonanceImmobilization of primase (3518 RU) was performed on the CM dextran matrix coated sensor chip CM5 (Pharmacia Biosensor) by carbodiimide coupling (30 μl of a 0.166 μM solution of DnaG at a flow rate of 5 μl/min in 50 mM sodium acetate [pH 4.5]). Immobilization of DnaB was performed similarly, except for use of 0.813 μM DnaB (as hexamer) in 10 mM sodium acetate (pH 4.5) (11043 RU). The sensor chip was washed with SPR buffer for 12 hr (final RU, 9235 RU). After each analysis, the surface was regenerated by a 10 min wash with SPR buffer without decreasing the capacity of the immobilized primase or DnaB. SPR analysis was performed by passing solutions (30 μl) of a (460 nM), θ (850 nM), ε (545 nM), δ (800 nM), δ′ (590 nM), χω (750 nM, as 1:1 heterodimer), γ (242 nM, as tetramer), τ (250 nM, as tetramer), and DnaB (1 pM, as hexamer) for 2 min followed by SPR buffer for 3 min. All proteins were dialyzed against SPR buffer to remove buffer-related artifacts. All subunits were tested for interaction with a sensor chip containing no protein. In each case, no signal was observed. General Priming AssayReactions contained 800 fmol of M13mp18 ssDNA, 14.4 pmol primase, 10 pmol DnaB, and 35 pmol τ (where added) in Buffer B with 1 μM ATP and 35 μM ATP, GTP, UTP, and 32P-CTP. The 280 μl reaction was incubated at 37°C, and 25 μl aliquots were removed at 0, 0.5, 1, 2, 3, 4, and 5 min. Reactions were quantitated by spotting on DE81 filters and by analysis on a 20% denaturing polyacrylamide gel. Analysis of DnaB–DNA ComplexesReactions were incubated at 30°C in 100 μl of Buffer B and analyzed by gel filtration on 5 ml A-15M columns equilibrated in Buffer B with 50 mM NaCl at 25°C. Reactions in Figure 3A contained 10 fmol DNA (ssM13mp18 or Form I pBS), 1.2 pmol 32P-DnaB, and 1.6 mM ATP or AMP–PNP. Reactions were 10 min. Preincubation of ssDNA with SSB was for 5 min, followed by 10 min with DnaB. In reactions using hexokinase (Sigma), glucose was added to 6.25 mM either after a 5 min preincubation of DnaB with ATP (after which time 10 fmol of ssDNA was added) or following a 10 min incubation of the ssDNA with DnaB and ATP. In Figure 3E, 15 pmol of 32P-dT65 was incubated with 20 pmol wild-type DnaB in the presence or absence of 1.6 mM ATP. In Figure 3F fmol TFII-EBNA1 DNA was incubated with 3 pmol 32P-EBNA1 dimer for 10 min at 30°C or 1.2 pmol 32P-DnaB with or without wt EBNA1, in the presence and absence of 1.6 mM ATP (as indicated). Replisome AssemblyFive parallel experiments were performed in which the only difference was the radioactive protein added in place of its wild-type counterpart. Each assay contained 1 pmol TFII-EBNA1 DNA, 6 pmol DnaB, 10 pmol primase, 6 pmol Pol III*, 6 pmol β, and 10 pmol EBNA1 in 100 μl Buffer B with 40 μM dCTP and dATP. After 10 min at 30°C, reactions were gel filtered over 5 ml A-15M columns equilibrated in Buffer B with 50 mM NaCl and 20 μM dCTP and dATP at 25°C. As one EBNA1 dimer binds each DNA molecule, the experiment using 32P-EBNA1 indicated a recovery of DNA to be at least 90%. Pol III* Stability on TFII DNATFII DNA (550 fmol) was incubated with 2000 fmol 3H-θ Pol III* and 3680 fmol β in 100 μl Buffer B with 40 μM dCTP and dATP for 10 min at 30°C in the presence or absence of 10 pmol DnaB. Reactions were gel filtered over 5 ml A-15M columns equilibrated in Buffer B with 50 mM NaCl and 20 μM dCTP and dATP at 25°C. Rolling Circle ReplicationTFII-EBNA1 DNA (1 pmol) was incubated with 20 pmol DnaB in 150 μl Buffer B at 30°C for 10 min. Then 14 pmol 32P-EBNA1 was added on ice, and after 5 min, 60 pmol β, 31.8 pmol Pol III*, and 40 μM dATP and dCTP were added. After 5 min at 30°C, the mixture was gel filtered on a 5 ml A15M column at 4°C in Buffer B with 50 mM NaCl, and 20 μM dATP and dCTP. Aliquots (15 μl) of each fraction were counted to quantitate the DNA (from 2P-EBNA1, assuming one EBNA1 per DNA). The peak fraction (450 fmol DNA) was divided into two tubes (75 μl each) and the following components were added to 150 μl: 3H-β (26.5 pmol), SSB (13.2 μg), ATP (1.2 mM), 4 rNTPs (200 μM each), 4 dNTPs (40 μM each), and when present, primase (24 pmol). Reactions were analyzed for accumulation of β on DNA and DNA synthesis. Accumulation of 3H-β was analyzed by shifting reactions to 30°C and removing 30 μl aliquots at 0.5, 1.5, 3, and 5 min. Aliquots were immediately transferred to ice and quenched with addition of sufficient hexokinase (0.5 U) and glucose (6.25 mM) to consume the ATP within 1 s. Each aliquot was gel filtered on a second A15M column. 32P-EBNA1 and 3H-β in each fraction was quantitated and the number of 3H-β clamps accumulated on DNA was determined from the ratio of 3H-β to 32P-EBNA1. DNA synthesis was monitored by removing 22 μl of each of the two reactions (+/− primase) and adding 3H-dTTP (to 1300 cpm/pmol) or 32P-dTTP (5000 cpm/pmol). Aliquots of 2 μl were withdrawn at 0.5, 1.5, 3, and 5 min and spotted on DE81 filters. For analysis by electrophoresis, 12 μl of each reaction was quenched by adding 2 μl of 0.5 M EDTA and 7 μl of 300 mM NaOH, 6 mM EDTA, 18% FICOL, 0.15% bromocresol green, and 0.25% xylene cyanol. Samples were analyzed on a 0.6% alkaline gel followed by autoradiography and analysis on a phosphoimager. AcknowledgementsThis work was supported by a grant from the NIH (GM38839). We are grateful to Dr. Susan Taylor for the cAMP dependent protein kinase. References
Article InfoPublication HistoryReceived in revised form: July 31, 1996 Received: June 12, 1996 IdentificationDOI: https://doi.org/10.1016/S0092-8674(00)80163-4 Copyright© 1996 Cell Press. Published by Elsevier Inc. User LicenseCreative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) |How you can reuse PermittedFor non-commercial purposes:
Not Permitted
Elsevier's open access license policy ScienceDirectAccess this article on ScienceDirect
Related ArticlesWhat is the basis for the difference in how the leading and lagging strands?The separated DNA strands form a replication fork, where both the DNA strands get replicated forming a lagging and leading strand. The major difference between a lagging and leading strand is that the lagging strand replicates discontinuously forming short fragments, whereas the leading strand replicates continuously.
What is the basis for the difference in how the leading and lagging strands of DNA molecules are synthesized a the origins of replication occur only at the 5 end?The fragments of the lagging strand are joined by DNA ligase. From the above information we have found that 'DNA polymerase can join new nucleotides only to the 3' end of the growing strand'. Hence, the correct answer is option (C).
What is the basis for the difference in how the leading and lagging strands of DNA molecules are synthesized quizlet?What is the basis for the difference in how the leading and lagging strands of DNA molecules are synthesized? DNA polymerase can join new nucleotides only to the 3′ end of a pre-existing strand, and the strands are antiparallel.
|
What is the basis for the difference in how the leading strand and the lagging strand of?
zusammenhängende Posts
Urheberrechte © © 2024 de.ketajaman Inc.
