Kyle Simonetta        Stamp

The Mechanism of ATP-Dependent Primer-Template Recognition by a Clamp Loader Complex

Kyle R. Simonetta, Steven L. Kazmirski, Eric R. Goedken, Aaron J. Cantor, Brian A. Kelch, Randall McNally, Steven N. Seyedin, Debora L. Makino, Mike O’Donnell and John Kuriyan

Cell Volume 137, Issue 4, 15 May 2009, Pages 659-671 doi:10.1016/j.cell.2009.03.044 (local copy)

Abstract / Figures from the paper / Supplemental Material / Coordinates / References


Clamp loaders load sliding clamps onto primer-template DNA. The structure of the E. coli clamp loader bound to DNA reveals the formation of an ATP-dependent spiral of ATPase domains that tracks only the template strand, allowing recognition of both RNA and DNA primers. Unlike hexameric helicases, in which DNA translocation requires distinct conformations of the ATPase domains, the clamp loader spiral is symmetric and is set up to trigger release upon DNA recognition. Specificity for primed DNA arises from blockage of the end of the primer and accommodation of the emerging template along a surface groove. A related structure reveals how the ψ protein, essential for coupling the clamp loader to single-stranded DNA-binding protein (SSB), binds to the clamp loader. By stabilizing a conformation of the clamp loader that is consistent with the ATPase spiral observed upon DNA binding, ψ binding promotes the clamp-loading activity of the complex.

Figures from the paper.

Click on the small image to get a bigger one.

Figure 1: Structure of the clamp loader:DNA complex (A) Schematic diagram of the clamp loader cycle. (B) Unbiased electron density for the DNA, calculated using a model at a stage prior to the inclusion of DNA and improved using density modification (Terwilliger, 2000). Contour lines at 1.2 standard deviations above the mean are shown in blue. The phosphate groups in the final DNA model are shown as spheres. The DNA interacting helices of the clamp loader are shown in yellow in this and subsequent figures. (C) The structure of wild type γ complex bound to primer-template DNA (left panel) and a schematic representation (right panel). The contacts between the clamp loader and the template strand are restricted primarily to the template strand, which is shown outlined in yellow.

Figure 2: DNA recognition (A) Diagram showing the ATPase subunits of the clamp loader and DNA duplex. The DNA interacting helices are shown in yellow. The three rotation axes that relate the B subunit to the C subunit, the C subunit to the D subunit, and the D subunit to the E subunit are shown in blue, red and green, respectively. The three axes are nearly coincident with each other and with the axis of the DNA duplex (not shown). (B) Schematic diagram of contacts with the primer-template DNA. (C) Expanded view of a domain 1:DNA interaction, highlighting hydrogen bonding interactions between the DNA and the protein. (D) The sidechain of Tyr 316 blocks the path of the primer strand by stacking on the last nucleotide base of the primer.

Figure 3: Symmetry in the AAA+ spiral and interfacial ATP coordination (A) The left panel shows the B, C, D, and E subunits (domain 1 only) of the γ complex, in a view that is orthogonal to that shown in Figure 2A. The middle panel shows a symmetrized version of the RFC clamp loader. In this model, the A subunit is in the same position, docked on the PCNA clamp, as in the crystal structure of the mutant RFC:PCNA complex (Bowman et al., 2004). The B, C, D and E subunits are positioned by applying the transformation that relates one subunit to the next one in the γ complex. The right panel shows the actual positions of the RFC subunits in the crystal structure of the mutant RFC:PCNA complex. (B) Coordination of the ATP analog bound to the B subunit by the arginine finger presented by the C subunit. Only the AAA+ modules (domains 1 and 2) are shown (left). A schematic representation of this interaction is also shown (right). (C) An expanded view of the coordination of ADP•BeFi3 bound to B is shown on the left. A similar view of ATPγS bound to the A subunit of the mutant RFC complex is shown on the right (Bowman et al., 2004). The arginine finger in each of the subunits of the mutant RFC complex is replaced by glutamine. A modeled arginine sidechain at the glutamine position is shown in grey, and it is positioned to coordinate the γ-phosphate of ATP, as do the actual arginine fingers in the γ complex. Each of the ATP binding sites in the γ complex has essentially the same configuration of sidechains shown here (see Figure S3). This symmetry is absent in the structure of the mutant RFC complex, in which only the A and C sites display tight coordination of ATP.

Figure 4: The exit channel for the template strand overhang (A) The structure of the clamp loader is shown, with the E(δ) subunit removed to reveal a tunnel leading through the collar, indicated by red spheres. In the expanded view on the right, sidechains presented by the collar domain of the A(δ) subunit and that interact with DNA are shown. Two sidechains that line the collar tunnel are also shown. (B) Fluorescence anisotropy data for fluorescently labeled primer-template DNA binding to the wild type γ complex and five mutants are shown. Mutation of residues that line the collar tunnel (R299A and R307A; see panel A) does not affect DNA binding affinity while mutation of residues that are in the observed exit path (R252A, R248A, and K313A) reduces DNA binding affinity. (C) Dissociation constants of mutant clamp loaders for DNA. The inset shows a zoomed in region of the chart. Error bars are the standard error of the fit.

Figure 5: Recognition of RNA-DNA hybrids (A) The DNA strand of an RNA:DNA hybrid ((Fedoroff et al., 1993); PDB code 124D) is aligned on the template strand in the crystal structure of the clamp loader. (B) The RNA and DNA strands of an RNA:DNA hybrid, aligned as in (A) are shown. The RNA strand (orange) is accommodated without steric clash because the clamp loader only engages the template strand. (C) The structure of an RNA:DNA hybrid, with its DNA strand aligned on the template strand in the crystal structure as in (A) and (B), is shown. Note that the RNA strand of the aligned hybrid duplex preserves the interaction with the separation pin (Compare with Figure 2D).

Figure 6: Binding of the ψ-peptide to the clamp loader collar (A) Isothermal titration calorimetry data for the binding of the ψ-peptide to the clamp loader complex. The calorimetric titration of 100 μM wild-type ψ-peptide into 10 μM of the clamp loader complex (left) and the ψ-peptide with Trp 17 mutated to Ser (right) are shown. The Trp 17 mutation leads to a 55-fold decrease in binding affinity. (B) Fluorescence anisotropy data for the binding of fluorescently labeled DNA to the clamp loader in the absence of the clamp. DNA binding in the presence (blue) and absence (red) of 10μM ψ-peptide are shown. Error bars are the standard deviation of individual readings. The values of the KD of DNA binding are 0.38±0.03μM and 18±3μM in the presence and absence of ψ-peptide, respectively. (C) The crystal structure of the ψ-peptide bound to the clamp loader collar. The clamp loader collar domains are shown as surface representations. The ψ-peptide is shown in red. (D) Close up view of the ψ-peptide interactions with the collar domains of the B(γ), C(γ), and D(γ) subunits, with hydrophobic sidechains shown as spheres. The C-terminal tail of B(γ), which forms a short anti-parallel β-sheet with the ψ-peptide, is shown as a blue ribbon.

Figure 7: ψ binding links SSB to the clamp loader and breaks symmetry in the collar (A) The collar domain of the B(γ) subunit undergoes a conformational change upon the binding of DNA by the clamp loader. Alignment of the collar domains of the apo E. coli clamp loader (Jeruzalmi et al., 2001) onto the DNA bound clamp loader reveals close overlap of the collar domains, with the exception of the B(γ) collar domain (dark blue) which undergoes a rotation of ~10° toward the AAA+ spiral in the DNA bound complex (shown in light blue). (B) The collar domains of the B(γ) and C(γ) subunits of the DNA bound complex are overlayed, revealing a difference in the orientation of the AAA+ domains of these subunits with respect to the collar. The AAA+ module of the B subunit (shown in light blue) rises up towards the collar domain, forming a tight interaction, whereas the C subunit (shown in red) is in an extended conformation. (C) The location of the ψ peptide on the clamp loader positions the χ:ψ assembly for interaction with SSB bound to the single stranded template exiting the clamp loader. The χ:ψ assembly (Gulbis et al., 2004) is positioned at the C-terminal end of the ψ-peptide bound to the clamp loader. The γ subunit binds the C-terminal tail of SSB. The 5’ template overhang of the DNA (green spheres) exits the clamp loader and wraps around SSB (Raghunathan et al., 2000).

Supplemental Material

Click on the small image to get a bigger one.

Figure S1: The A(γ) subunit is disengaged from DNA A view of the clamp loader:DNA complex that is orthogonal to the view shown in Figure 1C (left) and a schematic representation of the same view. The AAA+ module of the A(γ) subunit (shown in purple) is disengaged from the DNA and “swung out” from the spiral formed around the DNA by the other AAA+ modules.

Figure S2: The protein:DNA interactions are highly symmetric (A) The interface between Domains 1 from the C(γ) and B(γ) subunits is shown, along with the phosphate groups (sticks) in the template strand backbone that these domains interact with, as well as the nucleotide analogs bound to each subunit (spheres). (B) Three pairs of subunits (E/D, D/C, and C/B) are shown superimposed and in the same orientation as in (A). Note the close overlap of the protein structures, the ATP analogs, and the phosphate groups.

Figure S3: Comparison of ATP analog coordination at the three ATP binding sites. The coordination of ADP•BeF3 at the three interfacial binding sites in the clamp loader is shown. An overlay of these three sites was constructed by aligning Domain 1 from the subunits to which the analog is bound. Note the resulting overlap in positions of the arginine finger residues that are contributed by the adjacent subunits in each interface. Electron density from the final model for the ATP analog bound at the B:C interface is also shown.

Figure S4: FRET assay for clamp opening and closing by clamp loader mutants. (A) Clamps labeled with FRET donor/acceptor pairs on either side of a dimer interface were used to follow the open or closed state of the clamp (See Supplemental Experimental Procedures). The schematic diagram shows the expected decrease in FRET as the clamp is opened by the clamp loader, followed by an increase in FRET upon ATP hydrolysis and release of the clamp. The experiment is done using a clamp that is labeled with donor/acceptor pairs as indicated (Goedken et al., 2005; Goedken et al., 2004). ATP is added to solutions containing the clamp and various forms of the clamp loader, followed by primer-template DNA and then by excess ADP, which resets the system. (B) Fluorescence spectra are shown for wild type and one mutant (K313A) clamp loader. The ratio of fluorescence at the donor and acceptor wavelengths are shown in the insets for the four different conditions indicated. (C) The effects of mutations on clamp opening and closing are reported by comparing changes in FRET for mutant clamp loaders to the results obtained for the wild type clamp loader. The decrease in FRET upon addition of ATP is set to 100% for the wild type clamp loader. This decrease in FRET is the same for each of the mutant proteins (shown in the bar diagrams to the left in red), which are therefore not impaired in their ability to bind to and open the clamp. The increase in FRET upon addition of DNA reflects the hydrolysis of ATP and the release of the closed clamp, and is set to 100% for the wild type clamp loader. Mutations in the collar tunnel (R299A and R307A) have no significant effect on DNA-stimulated clamp release (shown on the right in green). In contrast, mutations in the observed exit channel (R252A, R248A, and K313A) lead to a reduction in the efficiency of clamp release.

Figure S5: Difference electron density for the ψ-peptide bound to the clamp loader:DNA complex An unbiased difference electron density map, calculated prior to the inclusion of the ψ-peptide in the model, is shown. Contour lines at 2.5 standard deviations above the mean value of electron denisty are shown in green. The final model for the ψ-peptide is shown in red. Note the clear electron density features surrounding Trp 7 (left panel) and Trp 17 (right panel), which allowed unambiguous determination of the sequence register of the ψ-peptide in the density.

Figure S6: Ψ binding is inconsistent with an inherent symmetry in the clamp loader collar (A) The apo structure of the E. coli clamp loader (Jeruzalmi et al., 2001) is shown with the AAA+ modules shown in a surface representation and the collar domains shown as ribbons. (B) The B(γ):C(γ) collar interaction is shown as in A (top) and the C(γ):D(γ) collar interaction is shown in the same orientation (bottom). The B(γ):C(γ) and C(γ):D(γ) collar interactions are overlayed (right), revealing almost perfect overlap and the inherent symmetry in the collar domains of the γ subunits in the apo complex. (C) Left: The ψ-peptide bound to the clamp loader collar. Middle: An expanded view of ψ-peptide binding to the B(γ), C(γ), and D(γ) collar domains in the DNA bound clamp loader structure, in which the inherent symmetry in the γ collar domains has been broken by a rotation of the B(γ) collar domain. Right: A model of the ψ-peptide bound to a collar in which the B(γ) collar domain (grey) has been positioned relative to the C(γ) collar as in the apo clamp loader structure. Note the steric clashes between the ψ-peptide and the B(γ) collar. (D) The same as (C) except from a side view with the collar domain of the C(γ) subunit removed.

Figure S7: Recognition of DNA structures with 3’ overhangs (A) A schematic representation of the modeling procedure used in this analysis. The primer-template junction was removed from the crystal structure and rotated by 180° around an axis perpendicular to the helical axis. This results in an inversion of the polarity of the template strand such that the 3’ end of the template is now at the “top.” The phosphate groups of this template strand running in the opposite direction could now be aligned with the positions of the phosphate groups in the template from the structure to generate a model of the clamp loader which interacts with a template running in the opposite direction. (B) The phosphate groups of the template strand with reversed polarity are shown in red, and they overlap closely with the phosphate groups of the original template strand (green). (C) Both strands of the DNA duplex with reversed polarity are shown, aligned as in (B). The clamp loader subunits now track the major groove rather than the minor groove. The "primer" strand, shown in orange, is positioned such that it terminates within the inner chamber of the clamp loader, near the collar subunit of the A subunit, which has been removed for clarity. The altered conformation of the "primer strand" is accommodated by the clamp loader, which does not interact with it. The "template" strand is positioned such that the 3' end is located at the gap between the A and E subunits. A 3' overhang extending from the reversed polarity "template" would interact with the A subunit. This model explains why replacement of the A subunit alone is sufficient to enable DNA with reversed polarity to be recognized by the clamp loader (Ellison and Stillman, 2003).

Figure S8: Ψ-peptide binding induces the collar conformational change in the apo clamp loader (A) The orientation of the B(γ) collar domains from the three clamp loader complexes in the crystal structure of the ψ-peptide complex in the absence of DNA (cartoon, light blue) are shown relative to the position of the B(γ) collar domain (dark blue) in the apo clamp loader (Jeruzalmi et al., 2001). (B) Difference electron density for the ψ-peptide bound to one of the clamp loaders in the asymmetric unit is shown (green contour lines at 3.0 standard deviations above the mean). The final ψ-peptide model from the clamp loader:DNA:ψ-peptide structure, positioned by aligning the collar domains of the C(γ) subunits, is shown (sticks). (C) and (D) are the same as (B) for the other two clamp loaders in the asymmetric unit.

Table S1: Data Processing

Table S2: Data Processing

Table S3: Data Processing

Table S4: Refinement Statistics

Table S5: Data statistics for the wild-type Apo γ complex:ψ-peptide structure

Table S6: Refinement statistics for the wild-type Apo γ complex:ψ-peptide structure


Coordinates in the Protein Data Bank:
3GLF clamp loader/DNA complex
3GLG mutant γT157A clamp loader/DNA complex
3GLH clamp loader/ψ-peptide complex
3GLI clamp loader/ψ-peptide/DNA complex


Bowman, G.D., O'Donnell, M., and Kuriyan, J. (2004). Structural analysis of a eukaryotic sliding DNA clamp-clamp loader complex. Nature 429, 724-730.

Ellison V., and Stillman, B. (2003). Biochemical characterization of DNA damage checkpoint complexes: clamp loader and clamp complexes with specificity for 5' recessed DNA. PLoS Biol 1, E33.

Fedoroff, O., Salazar, M., and Reid, B.R. (1993). Structure of a DNA:RNA hybrid duplex. Why RNase H does not cleave pure RNA. J Mol Biol 233, 509-523.

Goedken, E.R., Kazmirski, S.L., Bowman, G.D., O'Donnell, M., and Kuriyan, J. (2005). Mapping the interaction of DNA with the Escherichia coli DNA polymerase clamp loader complex. Nat Struct Mol Biol 12, 183-190.

Goedken, E.R., Levitus, M., Johnson, A., Bustamante, C., O'Donnell, M., and Kuriyan, J. (2004). Fluorescence measurements on the E.coli DNA polymerase clamp loader: implications for conformational changes during ATP and clamp binding. J Mol Biol 336, 1047-1059.

Gulbis, J.M., Kazmirski, S.L., Finkelstein, J., Kelman, Z., O'Donnell, M., and Kuriyan, J. (2004). Crystal structure of the chi:psi sub-assembly of the Escherichia coli DNA polymerase clamp-loader complex. Eur J Biochem 271, 439-449.

Jeruzalmi, D., O'Donnell, M., and Kuriyan, J. (2001). Crystal structure of the processivity clamp loader gamma (gamma) complex of E. coli DNA polymerase III. Cell 106, 429-441.

Raghunathan, S., Kozlov, A.G., Lohman, T.M., and Waksman, G. (2000). Structure of the DNA binding domain of E. coli SSB bound to ssDNA. Nat Struct Biol 7, 648-652.

Terwilliger, T.C. (2000). Maximum-likelihood density modification. Acta Crystallogr D Biol Crystallogr 56, 965-972.