Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Mol Biol. Author manuscript; available in PMC Feb 17, 2013.
Published in final edited form as:
PMCID: PMC3269524
A Central Swivel Point in the RFC Clamp Loader Controls PCNA Opening and Loading on DNA
Miho Sakato,1 Mike O’Donnell,2 and Manju M. Hingorani1*
1Department of Molecular Biology and Biochemistry Department, Wesleyan University, Middletown CT 06459
2The Rockefeller University and Howard Hughes Medical Institute, New York, NY 10021
* Corresponding author, Phone: (860) 685-2284, Fax: (860) 685 2141 ; mhingorani/at/
Replication Factor C (RFC) is a 5-subunit complex that loads PCNA clamps onto primer-template DNA (ptDNA) during replication. RFC subunits belong to the AAA+ superfamily, and their ATPase activity drives interactions between the clamp loader, clamp, and ptDNA, leading to topologically linked PCNA•ptDNA. We report the kinetics of transient events in S. cerevisiae RFC catalyzed PCNA loading, including ATP-induced RFC activation, PCNA opening, ptDNA binding, ATP hydrolysis, PCNA closing and PCNA•ptDNA release. This detailed perspective enables assessment of individual RFC-A, -B, -C, -D, -E subunit functions in the reaction mechanism. Functions have been ascribed to RFC subunits previously, based on steady state analysis of ‘arginine finger’ ATPase mutants; however, pre-steady state analysis provides a different view. The central subunit, RFC-C, serves as a critical swivel point in the clamp loader. ATP binding to this subunit initiates RFC activation and the clamp loader adopts a spiral conformation that stabilizes PCNA in a corresponding open spiral. The importance of RFC subunit response to ATP binding decreases as: RFC-C > RFC-D > RFC-B, with RFC-A being unnecessary. RFC-C-dependent activation of RFC also enables ptDNA binding, leading to formation of the RFC•ATP•PCNAopen•ptDNA complex. Subsequent ATP hydrolysis leads to complex dissociation, with RFC-D activity contributing the most to rapid ptDNA release. The pivotal role of the RFC-BCD subunit ATPase core in clamp loading is consistent with the similar central location of all three ATPase-active subunits of the E. coli clamp loader.
Keywords: DNA replication, RFC clamp loader, PCNA clamp, ATPase kinetic mechanism, Arginine finger
The eukaryotic clamp loader RFC loads PCNA clamps onto DNA where they serve as processivity factors for replicative polymerases, and modulate the functions of many other proteins that work on DNA 1,2. The structure and function of clamp and clamp loader proteins is highly conserved through evolution, reflecting their critical role in DNA metabolism. PCNA is a homo-trimeric ring with a central pore that accommodates duplex DNA 3,4. RFC is a hetero-pentameric complex, comprising one large (RFC-A) and four small subunits (RFC-B, -C, -D, -E) arranged in the form of a claw that binds PCNA at the base and primer-template DNA (ptDNA) within the central chamber (nomenclature note: RFC-A, -B, -C, -D, -E are RFC-1, -4, -3, -2, -5, respectively) 5. During the loading reaction, the clamp must be opened at an inter-subunit interface to enable entry of DNA, and then closed to form a topological link between the two macromolecules (Fig. 1). The clamp loader catalyzes this reaction by bringing the clamp and DNA substrates together in an appropriate configuration for loading, and then releasing the linked clamp•DNA product. The ATPase activity of the clamp loader fuels this mechanical work.
Fig. 1
Fig. 1
RFC structure and kinetic mechanism. (A) A computationally derived model of RFC bound to ATPγS and open PCNA 33. (B) Schematic of the ATPase sites at RFC subunit interfaces. Each subunit, except RFC-E, contains an active site to which a neighboring (more ...)
RFC subunits belong to the AAA+ family of ATPases—proteins that utilize free energy of the ATPase reaction to manipulate other macromolecules 6-8. The crystal structure of S. cerevisiae RFC complex shows that the subunits comprise three core domains (Fig. 1): the C-terminal domains pack against each other in a ring that holds the complex together at one end; the N-terminal domains form ATPase modules that are splayed in a spiral tracking the pitch of double-stranded DNA (note: the larger RFC-A subunit contains additional domains, including a C-terminal domain that packs against RFC-E) 5. These structural features appear strongly conserved among bacterial and eukaryotic clamp loaders 5,9-11. Four catalytic ATPase sites are located at the RFC-E/D (D site), RFC-D/C (C site), RFC-C/B (B site) and RFC-B/A (A site) subunit interfaces. One subunit binds ATP, providing the Walker A and B motifs required for phosphate and Mg2+ binding, respectively, and its neighbor provides the ‘arginine finger’ motif comprising SRC residues. The SRC Arg is analogous to the Arg residue in GTPase-activating proteins (GAPs) that inserts into the active site of Ras and related GTP-binding proteins. It facilitates nucleotide hydrolysis by stabilizing the transition state, and is also in position to couple nucleotide binding/hydrolysis/product release into conformational changes in the protein 12,13. In the case of clamp loaders, it has been reported that the SRC motif is not necessary for ATP binding, but is essential for ATP hydrolysis and clamp loading 14-17. In RFC, the RFC-E Arg contributes to the RFC-D ATP binding site at the RFC-E/D interface, and so on for the rest of the subunits, as shown in Fig. 1 (ESRC in site D, DSRC in site C, CSRC in site B, BSRC in site A); RFC-A does not have an SRC motif and the ATP binding site in RFC-E is not competent for hydrolysis 5,18.
This is the second of two reports tackling the question of how S. cerevisiae RFC uses ATP to load PCNA onto ptDNA. In the first study we focused on establishing the order and rates of transient events in the reaction, including PCNA opening and closing, ptDNA binding and release, and ATP hydrolysis and phosphate (Pi) release (see accompanying paper). Global analysis of the pre-steady state data yielded a model kinetic mechanism of this eukaryotic clamp loader, and enabled us to address specific questions about the role of individual RFC subunits in the reaction. In this study we focused on determining how ATP binding and hydrolysis by each subunit drives PCNA loading. The approach was to mutate SRC Arg residues to Ala (SAC) and assess the impact on the reaction mechanism. The results from pre-steady state analysis of mixed wild type/mutant complexes reveal the workings of each RFC subunit, and show how function and mechanistic significance depend on subunit position in the clamp loader complex.
Our report largely revises an earlier assignment of specific functions to the RFC subunits 14. Significant differences between the two studies are: (a) the earlier study was based on equilibrium and steady state measurements, which provide limited information on the kinetic mechanism and, correspondingly, and the role of individual subunits therein; (b) the RFCSAC complexes analyzed in the earlier study contained an N-terminal truncated version of RFC-A (ΔN283), which exhibits significantly lower stoichiometric activity than the full-length wild type protein for reasons as yet unclear 19.
Full-length wild type (SRC) and mixed wild type (SRC)/mutant (SAC) RFC complexes were purified from E. coli and tested for nucleotide binding activity. RFC-AllSAC exhibits similar ATPγS binding affinity and stoichiometry as wild type RFC, binding up to 5 ATPγS molecules in the presence of PCNA or both PCNA and ptDNA 19 (supplementary Fig. S1). Thus, mutation of the SRC motif to SAC does not appear to disrupt overall structural integrity and nucleotide binding by RFC. The mutations are, however, likely to disrupt RFC response to ATP binding as well as ATP hydrolysis, and a previous study has shown that RFC-AllSAC (RFC-A ΔN283) cannot load PCNA on ptDNA 14. To understand exactly how SAC mutations disrupt the loading mechanism, we analyzed their effects on the kinetics of distinct ATP-coupled steps in the reaction.
RFC-C activity is required for PCNA opening
Stabilization of the PCNA clamp in an open state requires ATP binding to RFC, but not ATP hydrolysis. ATP binding induces conformational changes that activate RFC and are coupled to PCNA opening, prior to ptDNA binding and ATP hydrolysis (Fig. 1; see accompanying paper, Table 1). We utilized a FRET-based assay with PCNA-WCAEDANS to assess the effect of SAC mutations on this early step in the reaction (PCNA-WCAEDANS: W185 changed to Trp; K107 changed to Cys and labeled with AEDANS). Clamp opening results in lower FRET efficiency between the trptophan-AEDANS donor-acceptor pair positioned across the inter-subunit interface, and clamp closure results in corresponding increase in FRET efficiency 20. Mixing wild type RFC and PCNA-WCAEDANS with ATP results in complete PCNA opening within ~ 3 s, as shown in Fig. 2A, i (FRET efficiencies were calculated 21 as described in the accompanying paper and normalized to the initial value of each kinetic trace to facilitate comparison between different RFC proteins). With RFC-AllSAC, PCNA-WCAEDANS FRET remains unchanged over time, indicating that the quadruple SAC mutant cannot open PCNA (ii). The RFC-DSAC (C site) mutation has the same outcome, indicating that subunit C plays a crucial role in PCNA opening (iv). Of the remaining mutants, RFC-ESAC (D site) (iii) and RFC-CSAC (B site) (v) exhibit fractional PCNA opening, while RFC-BSAC (A site) (vi) activity is comparable to wild type RFC. These results are summarized in Fig. 5A with PCNA opening kinetic traces overlaid for all RFC complexes. The ranking of subunit importance in ATP-binding induced RFC activation and PCNA opening is as follows: C > D > B > A.
Fig. 2
Fig. 2
The RFC-C site is critical for PCNA opening and loading on DNA. Opening/closing of PCNA clamp is measured by change in FRET between tryptophan (donor; λEX = 290 nm) and AEDANS (acceptor; λEM > 450 nm) at the PCNA-WCAEDANS inter-subunit (more ...)
Fig. 5
Fig. 5
Overlay of kinetic data for wild type and mutant RFC, and a model for RFC subunit contribution to PCNA loading. Kinetic traces are shown for (A) PCNA opening (Δt: 3 s), (B) PCNA closing (Δt: 3 s), (C) DNA binding/release in the presence (more ...)
The next step was to assess the effect of RFCSAC mutations on PCNA closure, which occurs toward the end of the reaction after ptDNA binding and ATP hydrolysis. Sequential mixing experiments were performed in which RFC and PCNA-WCAEDANS were pre-incubated with ATP for varying times to allow clamp opening (Δt: 0.02 – 3 s), then mixed with excess ptDNA, and the change in FRET monitored over time (FRET efficiencies were calculated 21 as described in the accompanying paper and normalized to the 3 s value from each PCNA opening kinetic trace to facilitate comparison between the different RFC proteins). At short Δt, initial FRET efficiency decrease (PCNA opening) and subsequent increase (PCNA closing) were detected (Fig. 2B, i), reflecting the ordered series of events—PCNA opening, ptDNA binding, then PCNA closure—in the reaction catalyzed by wild type RFC. With increasing Δt more PCNA was opened during the pre-incubation period, thus more closure was detectable on addition of ptDNA. At Δt: 3 s, all PCNA was opened prior to ptDNA addition and only PCNA closing was observed at a rate of ~ 7 s-1. There is almost no change in PCNA-WCAEDANS FRET efficiency with the RFC-AllSAC mutant, consistent with its inability to open PCNA (ii). Interestingly, addition of ptDNA to RFC-DSAC (C site) appears to force some PCNA opening, and this fraction of open PCNA can be closed as well (iv). This result implies a less critical role for subunit C after the PCNA opening step. A similar finding holds for RFC-ESAC (D site) (iii). Since the rate of PCNA closure by these two mutants is limited by slow opening, we cannot resolve whether RFC-DSAC and RFC-ESAC mutations affect the PCNA closing step directly. In case of RFC-CSAC (B site) (v) and RFC-BSAC (A site) (vi), the fraction of PCNA opened during pre-incubation is closed on ptDNA binding, but at 2–3-fold slower rate than wild type RFC. These results are summarized in Fig. 5B with kinetic traces at Δt: 3 s overlaid for all RFC complexes. No single subunit appears to have a predominant role in PCNA closure, though their contribution can accelerate this step in the reaction.
RFC-C activity is important for ptDNA binding and RFC-D activity for ptDNA release
Since rapid ptDNA binding to RFC triggers a burst of ATP hydrolysis followed by PCNA closure, if SAC mutations affect RFC-ptDNA interaction, the rates of these subsequent steps can be affected as well. In order to investigate this possibility, we measured ptDNA binding and release as reported by fluorescence of TAMRA dye conjugated to the 3’ primer end of the 40/65-nucleotide ptDNA substrate. Sequential mixing experiments were performed in which RFC and PCNA were pre-incubated with ATP for varying times to allow PCNA opening (Δt: 0.02 – 3 seconds), then mixed with ptDNATAMRA, and the change in fluorescence intensity monitored over time. As shown by the rise in ptDNATAMRA fluorescence (signal normalized to the initial value), wild type RFC binds ptDNA rapidly (5 × 107 M-1 s-1; see accompanying paper, Table I), and the bound fraction increases with Δt as more RFC•ATP•PCNAopen complex is formed during pre-incubation (Fig. 3, i). As ATP is hydrolyzed the complex dissociates at a limiting rate of ~ 8 s-1, shown by the decrease in ptDNATAMRA fluorescence (note: the signal remains higher than baseline level due to excess RFC in the reaction that can bind released ptDNATAMRA; see accompanying paper, Fig. 3). RFC-AllSAC shows no significant binding to ptDNA (ii). Both RFC-ESAC (D site) (iii) and RFC-DSAC (C site) (iv) bind ptDNA at a slow rate, apparently limited by the fractional PCNA opening that occurs on addition of DNA to the reaction for these two mutants (Fig. 2B, iii, iv). The RFC-CSAC (B site) (v) and RFC-BSAC (A site) (vi) mutants bind ptDNA at rates comparable to wild type RFC, in line with their ability to bind and open PCNA. The relative extent of ptDNA binding (Fig. 3; maximum bound fraction compared to wild type RFC) correlates with that of PCNA opening and closing (Fig. 2; maximum open and closed fractions) for all SAC mutants, except RFC-ESAC (iii), where the high ptDNATAMRA signal suggests exceptional affinity for ptDNA. These results are summarized in Fig. 5C with kinetic traces at Δt: 3 s overlaid for all RFC complexes in the presence of PCNA.
Fig. 3
Fig. 3
RFCSAC mutants exhibit defects in RFC•ATP•PCNA•ptDNA association and dissociation. ptDNA binding to RFC in the presence of PCNA is measured by change in fluorescence intensity of 3’ primer TAMRA-labeled ptDNA (λ (more ...)
All the SAC mutants exhibit significantly lower (or negligible) complex dissociation rates compared to wild type RFC (Fig. 3), which might be due to slower PCNA closure, PCNA•ptDNA release, and/or ptDNA release steps. In order to parse these possibilities, we performed the above DNA binding experiments in the absence of PCNA. As shown in the accompanying article, ATP-induced RFC activation occurs more slowly without PCNA, thus longer pre-incubation of RFC and ATP is necessary to achieve maximal ptDNA binding (Δt ≥ 8 s). ptDNA also dissociates faster from RFC in the absence of PCNA, thus the maximum bound fraction is lower than when PCNA is present. The rate of ptDNA binding to ATP-activated wild type RFC is the same with or without PCNA (5 × 107 M-1 s-1). The data are shown in Fig. 5D with ptDNA binding/release kinetic traces at Δt: 8 s overlaid for all RFC complexes in the absence of PCNA. The mutants bind ptDNA at rates comparable to wild type RFC. RFC-ESAC (D site), RFC-CSAC (B site), and RFC-BSAC (A site) also bind ptDNA to the same extent as RFC. Notably, the maximum fraction of ptDNA bound to RFC-DSAC (C site) is only about half that of RFC (despite a slightly slower rate of dissociation), suggesting that RFC-C activity plays a greater role in ptDNA binding to the clamp loader than the other subunits. Another notable result is the minimal dissociation of ptDNA from RFC-ESAC, which suggests that RFC-D has an important role in releasing ptDNA following ATP hydrolysis. This finding also explains the highly stable RFC-ESAC•ATP•PCNA•ptDNA complex observed in Fig. 3, iii (see also Fig. 5C). The RFC-CSAC mutant (B site) also has a slightly reduced ptDNA release rate compared to wild type RFC (Fig. 5D), which, along with slower PCNA closing (Fig. 2B, v), can explain slower dissociation of the RFC-CSAC•ATP•PCNA•ptDNA complex (Fig. 3, v; Fig. 5C). Finally, RFC-BSAC (A site) releases ptDNA as fast at wild type RFC (Fig. 5D); therefore, the relatively slow rate of RFC-BSAC•ATP•PCNA•ptDNA complex dissociation (Fig. 3, vi; Fig. 5C) is likely due to another step in the reaction such as slow PCNA closure (Fig. 2B, vi), which is proposed as a rate-limiting step in our model mechanism (see accompanying paper, Table I).
All ATPase active sites in RFC influence ptDNA-induced rapid ATP hydrolysis and Pi release
Once the RFC•ATP•PCNAopen•ptDNA intermediate complex is formed, ATP is hydrolyzed rapidly (20 – 50 s-1), followed by PCNA closure and release of products, including PCNA•ptDNA, phosphate (Pi) and ADP (see accompanying paper) 19. We measured the kinetics of one of these final steps, Pi release, for all RFC mutants, in order to assess how the loss of individual subunit ATP sensing and hydrolysis activities affects the ATPase activity of RFC as a whole. Pre-steady state Pi release kinetics were measured with the fluorophore-labeled MDCC-PBP reporter protein that binds Pi rapidly and with high affinity (kon = 1.4 × 107 M-1 s-1; KD = 0.1 μM) 22. Sequential mixing experiments were performed in which RFC and PCNA were pre-incubated with ATP for varying times to allow clamp opening (Δt: 0.02 – 3 s), then mixed with excess ptDNA and MDCC-PBP, and the change in MDCC-PBP fluorescence monitored over time. The signal was converted to Pi concentration using a calibration curve and plotted versus time 19. Fig. 4, i shows a burst of Pi release catalyzed by wild type RFC, at an apparent rate limited to ~ 10 s-1 and an amplitude indicating rapid ATP hydrolysis by three subunits (see accompanying paper, Table I). The steady state ATPase rate is 1.6 μM s-1 (kcat = 1.1 s-1, assuming 3 active ATPase sites per RFC at 0.5 μM in the reaction). All the SAC mutants exhibit lower burst amplitude and/or slower rates, and RFC-AllSAC has no detectable activity (Fig. 4; ;5E5E at Δt: 3s). This result means that all four nucleotide binding sites in RFC contribute to the burst, indirectly by sensing ATP binding or directly by catalyzing rapid ATP hydrolysis. RFC-DSAC (iv) shows almost no burst phase, consistent with the severe defect in PCNA opening and ptDNA binding steps (preceding ATP hydrolysis) caused by disrupting the RFC-C active site (Fig. 5). RFC-ESAC (iii) has lower Pi release burst amplitude and rate, and ~ 2-fold slower steady state rate (0.9 μM s-1) than wild type RFC, probably due to defective PCNA opening and ptDNA binding (preceding ATP hydrolysis) and ptDNA release (following ATP hydrolysis) caused by disrupting the RFC-D active site (Fig. 5). RFC-BSAC (vi) has almost the same Pi release profile as RFC-ESAC, but in this case the underlying reasons are different. Disruption of the RFC-A active site does not affect PCNA opening or ptDNA binding (Fig. 5); thus, lower burst amplitude and rate, and slower steady-state activity of RFC-BSAC (0.95 μM s-1), reflect defects in steps following ATP hydrolysis, such as PCNA closing and complex dissociation. In case of RFC-CSAC (v) all steps preceding and following ATP hydrolysis are moderately affected upon disruption of the RFC-B active site, which is consistent with its intermediate Pi release kinetics profile (Fig. 5).
Fig. 4
Fig. 4
Disruption of the RFC-C ATPase active site abrogates rapid ATP hydrolysis and Pi release. RFC-catalyzed Pi release is measured by change in fluorescence intensity of MDCC-PBP reporter. Pre-incubation of RFC, PCNA with ATP (Δt: 0.02 – 3 (more ...)
The eukaryotic RFC clamp loader comprises five subunits with four ATPase active sites located at the subunit interfaces. In each site, one subunit presents most of the conserved ATP binding and catalytic residues (e.g., Walker A, B motifs), while a neighboring subunit presents an arginine (SRC motif) that serves as a sensor of ATP binding and facilitates ATP hydrolysis. Biochemical studies of bacterial and eukaryotic clamp loaders have shown that ATP (ATPγS) binding is sufficient to form a protein-DNA complex containing an open clamp 17,20,23-26, and ATP hydrolysis leads to release of the clamp closed around DNA 19,25,27-29. In the accompanying paper we described key rate-determining events in the reaction: (a) ATP binding-induced conformational changes in RFC that enable PCNA opening and ptDNA binding, and (b) ATP hydrolysis-induced conformational changes that enable PCNA closure and PCNA•ptDNA release. Here we describe the role of each RFC subunit in these rate-determining steps, further elucidating the workings of this critical protein machine in DNA replication.
We chose to mutate the SRC motif in each subunit, expecting that this change would allow ATP binding but not hydrolysis, based on past reports 14,15. Moreover, the possibility of disrupting the effects of ATP binding and hydrolysis on RFC conformation provided a means to specifically investigate the ATP coupling mechanism 13. Finally, one mystery driving our study was that the PCNA clamp is closed in the crystal structure of S. cerevisiae RFC-AllSQC•ATPγS•PCNA complex 5, despite evidence that ATPγS supports PCNA opening by S. cerevisiae as well as P. furiosus RFC 20,26. Since the crystallized RFC had SRC motifs in all subunits mutated to SQC, we hypothesized that it was unable to respond to ATP binding and undergo the conformational changes required for PCNA opening. We changed the SRC arginines in RFC-E, -D, -C, and -B to alanine and assessed the impact by pre-steady state analysis of ATP binding-, hydrolysis- and Pi release-coupled steps in the PCNA loading reaction. Broadly, we found that RFC-C (DSRC) response to ATP binding is key to assembly of the protein-DNA complex, and RFC-D (ESRC) response to ATP hydrolysis is key to its disassembly; RFC-B (CSRC) activity contributes modestly to complex assembly, and RFC-A (BSRC) activity appears unnecessary except for slightly accelerating complex disassembly and catalytic turnover. The differing impact of SAC mutant subunits on the rate limiting steps matches their clamp loading phenotype exactly, with RFC-DSAC (C site) and RFC-ESAC (D site) unable to load PCNA, RFC-CSAC (B site) exhibiting a moderate defect, and RFC-BSAC (A site) indistinguishable from wild type RFC 14. The same is true for Walker A mutants of RFC, in which the conserved lysine required for ATP hydrolysis is mutated to glutamate 30.
Fig. 5 provides a summary of data on intermediate steps in the reaction for wild type and mutant RFC complexes. Disruption of SRC arginine in the subunit C ATPase active site (DSAC) abrogates PCNA opening completely, despite the fact that this mutation does not disrupt ATPγS binding to RFC (supplementary Fig. S1). Indeed, RFC-AllSAC binds two additional ATPγS molecules in the presence of PCNA, up to the maximum of five. This result suggests that the mutant clamp loader still binds PCNA, which in turn promotes ATP binding, but it does not undergo ATP binding-induced activation coupled with PCNA opening. Our conclusion is in accord with the RFC-AllSQC•ATPγS•PCNA structure in which all nucleotide-binding sites are occupied with ATPγS, but the clamp is not open 5. Notably, only RFC-A, -B, and -C subunits contact the closed clamp in the structure, suggesting that RFC-AllSQC cannot progress from this initial interaction to one in which RFC-D and -E also contact the clamp, as required for its opening 31; a similar proposal was made based on structural analysis of an Archaeoglobus fulgidus RFC sub-complex 32. The impact of disrupting the ATPase sites varies as RFC-C > RFC-D > RFC-B > RFC-A, from no opening to full opening activity (Fig. 5A). We propose that ATP binding at the RFC-D/C interface initiates conformational changes that propagate through RFC-D and -E and create a compatible binding surface for the open clamp 33. Thus, the RFC-AllSQC•ATPγS•PCNA structure is an early intermediate in the RFC activation pathway, wherein initial (partial) interaction between RFC and closed PCNA promotes additional ATP binding and ‘twisting’ of RFC until the proteins are mutually trapped in a right-handed spiral. This conclusion is in accord with the marked acceleration of ATP-induced RFC activation in the presence of PCNA (see accompanying paper) 19. A previous study had concluded that all RFCSAC mutants, including RFC-AllSAC (RFC-A ΔN283) can open PCNA, based on fluorescent labeling of Cys81 at the PCNA inter-subunit interface that is exposed on clamp opening 14. We suggest that this assay could not resolve the substantive differences in PCNA opening kinetics because of prolonged incubation with excess flurophore (30 seconds) during which even a transiently open clamp could be labeled. The same study also showed that the ATPase activity of triple RFCSAC (RFC-A ΔN283) mutants containing active RFC-C or RFC-D subunits was stimulated by PCNA. This result complements our finding that RFC-C and RFC-D response to ATP binding is critical for PCNA opening.
Only the RFC-DSAC (C site) mutation reduces ptDNA binding to RFC (Fig. 5D). This finding is consistent with equilibrium binding measurements that could not detect interaction between ptDNA and RFC-DSAC (RFC-A ΔN283), while the other mutants had similar or slightly lower affinity than wild type RFC 14. We suggest that the ATP-induced conformational changes initiating at subunit C optimize RFC contacts with ptDNA, and that no other subunit has a predominant role in creating/maintaining the helical dsDNA binding track along the interior of RFC. In the presence of PCNA, the effects of SAC mutations on ptDNA binding kinetics appear more significant (Fig. 5C); however, apparent slow ptDNA binding to RFC-ESAC (D site) and RFC-CSAC (B site) could be simply due to slow PCNA opening and/or slow RFC activation that occur prior to ptDNA binding and, conversely, rapid ptDNA binding to RFC-BSAC (A site) is consistent with its wild type-like PCNA opening activity. Only in the case of RFC-DSAC (C site), the apparent slow rate can be attributed definitively to defective PCNA opening/RFC activation (Fig. 5A) as well as defective ptDNA binding (Fig. 5D).
The kinetic model developed in the accompanying article shows that rapid ATP hydrolysis after ptDNA binding is followed by a relatively slow step at or prior to PCNA closure, PCNA•ptDNA and Pi dissociation (Fig. 1). We have proposed that this step involves relaxation of RFC back to an inactive conformation that has low affinity for open PCNA and ptDNA. Pi release kinetics indicate that all four active sites contribute to the ATPase activity, directly by catalysis or by allosteric effects on other sites, since the pre-steady state burst amplitude and rates are lower for all mutants compared to wild type RFC (Fig. 5E). Note that for the mutants, burst amplitude is not a direct measure of the number of ATP molecules hydrolyzed, since these complexes exhibit sub-stoichiometric activity prior to the ATP hydrolysis step. The steady state ATPase rates of wild type and mutant complexes appear quite similar, but that is not surprising since they all bind ptDNA, which alone can stimulate the rate to ~ 2 μM s-1 19. In case of RFC-DSAC (C site) the apparent rates of PCNA closure, PCNA•ptDNA and Pi release are slow, but we cannot resolve if RFC-C contributes specifically to these late steps in the reaction, because of its essential role in the early steps of RFC activation, PCNA opening (Fig. 5A) and ptDNA binding (Fig. 5D). The RFC-D subunit plays a significant role in the early steps as well, but we have direct evidence that disruption of ATP hydrolysis at this site (ESAC) also slows down ptDNA release (Fig. 5D). The PCNA closure, PCNA•ptDNA and Pi release kinetics of RFC-CSAC confirm a modest role for RFC-B in the reaction, starting with the PCNA opening step. Unlike the other subunits, ATP binding to RFC-A has no detectable impact on the early steps, but the RFC-BSAC mutation does lower the rates of PCNA closure, PCNA•ptDNA and Pi release slightly, suggesting that the site contributes to RFC catalytic turnover. Analysis of A. fulgidus RFC also indicates that the ATPase activity of the large subunit (equivalent to RFC-A) is involved in catalytic turnover 17,32.
Analysis of the pre-steady state burst of ATP hydrolysis and Pi release by wild type RFC reveals that at least 3 ATPase sites hydrolyze ATP rapidly, and at the same apparent rate, once the RFC•ATP•PCNAopen•ptDNA complex is formed (see accompanying paper) 19. It is possible that the fourth site hydrolyzes ATP at a slower rate that is difficult to distinguish from steady state, or that some aspect of our assay conditions masks the activity of this site. The mutant data indicate that ATP binding and hydrolysis by the RFC-BCD core is necessary and sufficient to complete the PCNA loading reaction; therefore, we propose that ptDNA triggers ATP hydrolysis at these three active sites to initiate complex dissociation. The RFC-A subunit makes extensive contacts with PCNA 5; thus, the ATPase activity of this site (perhaps at a slower rate) could further assist in PCNA closure, PCNA•ptDNA release and catalytic turnover. This interpretation is consistent with an earlier study of RFC Walker A site mutants, which also showed that only the RFC-BCD core ATPase activity is required for PCNA loading 30. A comparison of the effects of SRC mutations in E. coli γ complex 16,34 reveals significant similarities with the RFC clamp loader mechanism described here. In this case, the 3 central γ subunits possess ATPase active sites while δ’ and δ do not (δ’, γ1, γ2, γ3, δ are positional equivalents of RFC-E, -D, -C, -B, -A) 10. Both δ’SAC1 site) and γSAC2 and γ3 sites) mutants suffer significant loss in β clamp binding affinity and clamp opening, consistent with the significant loss in PCNA clamp opening by RFC-ESAC (D site) and RFC-DSAC (C site) mutants. The γSAC mutants (γ2 and γ3 sites) also lose affinity for DNA, and although the contribution(s) of γ2 and/or γ3 to DNA binding cannot be parsed, this result is also consistent with our finding that RFC-DSAC (C site) has lower ptDNA binding activity than wild type RFC. Importantly, the γ complex SRC mutant study revealed that the 3 ATPase active sites located in the clamp loader core are necessary and sufficient for loading β onto ptDNA, which mirrors our finding about the BCD subunit core in RFC 16.
In conclusion, we have found that RFC subunit function varies with position in the clamp loader complex (Fig. 5F). RFC-C serves as a critical swivel point in the center of RFC, and its response to ATP binding propagates conformational change towards the peripheral subunits. As a result, RFC adopts an active spiral form creating an extensive binding interface at which all subunits, including RFC-D and -E, contact an open PCNA clamp. RFC-C dependent RFC activation also facilitates interaction with ptDNA. Subsequent rapid ATP hydrolysis by at least three subunits, likely the RFC-BCD core, prompts relaxation of RFC back into a form with low affinity for ptDNA and PCNA. At this stage the conformational change appears to propagate in reverse, i.e., from the peripheral RFC-D and -E subunits. RFC-A involvement in these ATP-coupled conformational dynamics appears less significant, suggesting that the work done by the D/E arm of the RFC spiral is more important than the A/B arm while loading PCNA on DNA.
RFC and PCNA, PCNA-WCAEDANS, and MDCC-PBP proteins were prepared as described in the accompanying paper. For RFC complexes with mutant SRC motifs, two pLANT2/RIL-RFC-A/E plasmids containing full-length RFC-A were constructed by replacing a DraI fragment of pLANT2/RIL-RFC-HKΔN-A/E and pLANT2/RIL-RFC-HKΔN -A/ESAC 14 with a DraI fragment coding for the N-terminal 282 amino acids of RFC-A subunit from pLANT2/RIL-RFC-A/HKE 35. Co-expression with pET11a-RFC-BSAC/C/D, pET11a-RFC-B/CSAC/D or pET11a-RFC-B/C/DSAC plasmids yielded RFC complexes with individual SAC mutant subunits (B: Arg157; C: Arg160; D: Arg183; E: Arg184). ptDNA substrates were prepared as described in the accompanying paper.
Kinetic measurements of PCNA opening/closing, ptDNA binding/release, and phosphate release by the mixed wild-type/mutant RFCSAC complexes were performed and the data analyzed as described in the accompanying paper.
Supplementary Material
This work was supported by National Institutes of Health grants R15 GM094047 (M.M.H) and GM 38839 (M. O’D).
PCNAproliferating cell nuclear antigen
RFCreplication factor C
ptDNAprimer-template DNA
Piinorganic phosphate
TAMRA5-(and 6-) carboxytetramethylrhodamine
MDCC-PBP7-diethylamino-3-((((2-maleimidyl)ethyl)amino)carbonyl) coumarin-labeled phosphate binding protein
FRETFörster resonance energy transfer

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1. Johnson A, O’Donnell M. Cellular DNA replicases: components and dynamics at the replication fork. Annu Rev Biochem. 2005;74:283–315. [PubMed]
2. Moldovan GL, Pfander B, Jentsch S. PCNA, the maestro of the replication fork. Cell. 2007;129:665–79. [PubMed]
3. Gulbis JM, Kelman Z, Hurwitz J, O’Donnell M, Kuriyan J. Structure of the C-terminal region of p21(WAF1/CIP1) complexed with human PCNA. Cell. 1996;87:297–306. [PubMed]
4. Krishna TS, Kong XP, Gary S, Burgers PM, Kuriyan J. Crystal structure of the eukaryotic DNA polymerase processivity factor PCNA. Cell. 1994;79:1233–43. [PubMed]
5. Bowman GD, O’Donnell M, Kuriyan J. Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature. 2004;429:724–30. [PubMed]
6. Erzberger JP, Berger JM. Evolutionary relationships and structural mechanisms of AAA+ proteins. Annu Rev Biophys Biomol Struct. 2006;35:93–114. [PubMed]
7. Hanson PI, Whiteheart SW. AAA+ proteins: have engine, will work. Nat Rev Mol Cell Biol. 2005;6:519–29. [PubMed]
8. Iyer LM, Leipe DD, Koonin EV, Aravind L. Evolutionary history and higher order classification of AAA+ ATPases. J Struct Biol. 2004;146:11–31. [PubMed]
9. Simonetta KR, Kazmirski SL, Goedken ER, Cantor AJ, Kelch BA, McNally R, Seyedin SN, Makino DL, O’Donnell M, Kuriyan J. The mechanism of ATP-dependent primer-template recognition by a clamp loader complex. Cell. 2009;137:659–71. [PMC free article] [PubMed]
10. Jeruzalmi D, O’Donnell M, Kuriyan J. Crystal structure of the processivity clamp loader gamma (gamma) complex of E. coli DNA polymerase III. Cell. 2001;106:429–41. [PubMed]
11. O’Donnell M, Kuriyan J. Clamp loaders and replication initiation. Curr Opin Struct Biol. 2006;16:35–41. [PubMed]
12. Scheffzek K, Ahmadian MR, Kabsch W, Wiesmuller L, Lautwein A, Schmitz F, Wittinghofer A. The Ras-RasGAP complex: structural basis for GTPase activation and its loss in oncogenic Ras mutants. Science. 1997;277:333–8. [PubMed]
13. Chen B, Sysoeva TA, Chowdhury S, Guo L, De Carlo S, Hanson JA, Yang H, Nixon BT. Engagement of arginine finger to ATP triggers large conformational changes in NtrC1 AAA+ ATPase for remodeling bacterial RNA polymerase. Structure. 2010;18:1420–30. [PMC free article] [PubMed]
14. Johnson A, Yao NY, Bowman GD, Kuriyan J, O’Donnell M. The replication factor C clamp loader requires arginine finger sensors to drive DNA binding and proliferating cell nuclear antigen loading. J Biol Chem. 2006;281:35531–43. [PubMed]
15. Johnson A, O’Donnell M. Ordered ATP hydrolysis in the gamma complex clamp loader AAA+ machine. J Biol Chem. 2003;278:14406–13. [PubMed]
16. Snyder AK, Williams CR, Johnson A, O’Donnell M, Bloom LB. Mechanism of loading the Escherichia coli DNA polymerase III sliding clamp: II. Uncoupling the beta and DNA binding activities of the gamma complex. J Biol Chem. 2004;279:4386–93. [PubMed]
17. Seybert A, Wigley DB. Distinct roles for ATP binding and hydrolysis at individual subunits of an archaeal clamp loader. Embo J. 2004;23:1360–71. [PubMed]
18. Schmidt SL, Pautz AL, Burgers PM. ATP utilization by yeast replication factor C. IV. RFC ATP-binding mutants show defects in DNA replication, DNA repair, and checkpoint regulation. J Biol Chem. 2001;276:34792–800. [PubMed]
19. Chen S, Levin MK, Sakato M, Zhou Y, Hingorani MM. Mechanism of ATP-driven PCNA clamp loading by S. cerevisiae RFC. J Mol Biol. 2009;388:431–42. [PMC free article] [PubMed]
20. Zhuang Z, Yoder BL, Burgers PM, Benkovic SJ. The structure of a ring-opened proliferating cell nuclear antigen-replication factor C complex revealed by fluorescence energy transfer. Proc Natl Acad Sci U S A. 2006;103:2546–51. [PubMed]
21. Alley SC, Abel-Santos E, Benkovic SJ. Tracking sliding clamp opening and closing during bacteriophage T4 DNA polymerase holoenzyme assembly. Biochemistry. 2000;39:3076–90. [PubMed]
22. Brune M, Hunter JL, Howell SA, Martin SR, Hazlett TL, Corrie JE, Webb MR. Mechanism of inorganic phosphate interaction with phosphate binding protein from Escherichia coli. Biochemistry. 1998;37:10370–80. [PubMed]
23. Hingorani MM, O’Donnell M. ATP binding to the Escherichia coli clamp loader powers opening of the ring-shaped clamp of DNA polymerase III holoenzyme. J Biol Chem. 1998;273:24550–63. [PubMed]
24. Hingorani MM, Coman MM. On the specificity of interaction between the Saccharomyces cerevisiae clamp loader replication factor C and primed DNA templates during DNA replication. J Biol Chem. 2002;277:47213–24. [PMC free article] [PubMed]
25. Gomes XV, Burgers PM. ATP utilization by yeast replication factor C. I. ATP-mediated interaction with DNA and with proliferating cell nuclear antigen. J Biol Chem. 2001;276:34768–75. [PubMed]
26. Miyata T, Suzuki H, Oyama T, Mayanagi K, Ishino Y, Morikawa K. Open clamp structure in the clamp-loading complex visualized by electron microscopic image analysis. Proc Natl Acad Sci U S A. 2005;102:13795–800. [PubMed]
27. Anderson SG, Thompson JA, Paschall CO, O’Donnell M, Bloom LB. Temporal correlation of DNA binding, ATP hydrolysis, and clamp release in the clamp loading reaction catalyzed by the Escherichia coli gamma complex. Biochemistry. 2009;48:8516–27. [PMC free article] [PubMed]
28. Ason B, Handayani R, Williams CR, Bertram JG, Hingorani MM, O’Donnell M, Goodman MF, Bloom LB. Mechanism of loading the Escherichia coli DNA polymerase III beta sliding clamp on DNA. Bona fide primer/templates preferentially trigger the gamma complex to hydrolyze ATP and load the clamp. J Biol Chem. 2003;278:10033–40. [PubMed]
29. Turner J, Hingorani MM, Kelman Z, O’Donnell M. The internal workings of a DNA polymerase clamp-loading machine. Embo J. 1999;18:771–83. [PubMed]
30. Schmidt SL, Gomes XV, Burgers PM. ATP utilization by yeast replication factor C. III. The ATP-binding domains of Rfc2, Rfc3, and Rfc4 are essential for DNA recognition and clamp loading. J Biol Chem. 2001;276:34784–91. [PubMed]
31. Yao NY, Johnson A, Bowman GD, Kuriyan J, O’Donnell M. Mechanism of proliferating cell nuclear antigen clamp opening by replication factor C. J Biol Chem. 2006;281:17528–39. [PubMed]
32. Seybert A, Singleton MR, Cook N, Hall DR, Wigley DB. Communication between subunits within an archaeal clamp-loader complex. EMBO J. 2006;25:2209–18. [PubMed]
33. Tainer JA, McCammon JA, Ivanov I. Recognition of the ring-opened state of proliferating cell nuclear antigen by replication factor C promotes eukaryotic clamp-loading. J Am Chem Soc. 2010;132:7372–8. [PMC free article] [PubMed]
34. Paschall CO, Thompson JA, Marzahn MR, Chiraniya A, Hayner JN, O’Donnell M, Robbins AH, McKenna R, Bloom LB. The E. coli clamp loader can actively pry open the beta-sliding clamp. J Biol Chem 2011 [PMC free article] [PubMed]
35. Finkelstein J, Antony E, Hingorani MM, O’Donnell M. Overproduction and analysis of eukaryotic multiprotein complexes in Escherichia coli using a dual-vector strategy. Anal Biochem. 2003;319:78–87. [PubMed]