Mechanism of opening a sliding clamp

Abstract

Clamp loaders load ring-shaped sliding clamps onto DNA where the clamps serve as processivity factors for DNA polymerases. In the first stage of clamp loading, clamp loaders bind and stabilize clamps in an open conformation, and in the second stage, clamp loaders place the open clamps around DNA so that the clamps encircle DNA. Here, the mechanism of the initial clamp opening stage is investigated. Mutations were introduced into the Escherichia coli β-sliding clamp that destabilize the dimer interface to determine whether the formation of an open clamp loader-clamp complex is dependent on spontaneous clamp opening events. In other work, we showed that mutation of a positively charged Arg residue at the β-dimer interface and high NaCl concentrations destabilize the clamp, but neither facilitates the formation of an open clamp loader-clamp complex in experiments presented here. Clamp opening reactions could be fit to a minimal three-step 'bind-open-lock' model in which the clamp loader binds a closed clamp, the clamp opens, and subsequent conformational rearrangements 'lock' the clamp loader-clamp complex in a stable open conformation. Our results support a model in which the E. coli clamp loader actively opens the β-sliding clamp.

Figure 1.

High-resolution structures of clamp loaders and clamps from different species depicting conformational states that may exist in the clamp opening pathway. (Left panel) Structures of the E. coli γ₃δδ’ clamp loader along with the β-clamp illustrate conformations that may exist prior to clamp loader–clamp binding (9,48). Amino acid residues in β that were mutated to Cys for fluorescent labeling are shown in spheres and labeled with arrows. (Center panel) The clamp loader initially binds the clamp to form a closed complex that may be similar in structure to the closed S. cerevisiae RFC–PCNA complex (44). The Rfc1 subunit (yellow) makes the most extensive interactions between the clamp loader and the clamp in the closed complex. Two of the subunits (Rfc2, green, and Rfc5, red) do not contact the surface of the clamp at all. (Right panel) After binding the clamp loader, the clamp either spontaneously opens or is actively opened by the clamp loader, to form an open clamp loader–clamp complex that may resemble the bacteriophage T4 gp44/62-gp45 clamp loader–clamp complex (30). In this open complex, each of the clamp loader subunits contacts the surface of the clamp such that the interaction surface is much greater than in the closed clamp loader–clamp complex.

Figure 2.

Equilibrium binding to form open clamp loader–clamp complexes. Relative AF488 fluorescence is plotted as a function of γ_cx concentration for solutions of β-S109C/Q299C-(AF488)₂ (black circles) and β-S109C/Q299C/R103S-(AF488)₂ (grey triangles) in assay buffer containing (A) 50 mM NaCl or (B) 500 mM NaCl. Final concentrations of β clamps were 10 nM. Data were fit to a quadratic equation (solid line through data points) assuming a two-state binding model (indicated in the diagram) to calculate apparent dissociation constants, K_d,app. Assay buffer contained 20 mM Tris–HCl pH 7.5, 8 mM MgCl₂, 5 mM DTT, 40 μg/ml BSA, 0.1 mM EDTA and 4% glycerol. Note that NaCl quenches the fluorescence of AF488 in unbound β-clamps in a concentration-dependent manner so that the magnitude of the signal change is greater in the assay containing 500 mM NaCl. Increased ionic strength likely increases the fraction of AF488 molecules that are stacked and quenched when the clamp is closed.

Figure 3.

Clamp loader–clamp binding/opening reactions at 20°C. The increase in AF488 fluorescence that occurs when γ_cx binds β-(AF488)₂ to form an open complex was measured as a function of time when a solution of γ_cx and ATP (0.5 mM) in stopped-flow assay buffer was added to a solution of β-S109C/Q299C-(AF488)₂ (dark colors) or β-S109C/Q299C/R103S-(AF488)₂ (light colors) and ATP (0.5 mM) in stopped-flow assay buffer. The concentration of γ_cx was varied and the concentration of β was held constant at 20 nM. Representative reactions are shown that contain 10 nM (blue), 20 nM (magenta), 40 nM (green) or 160 nM (black/grey) γ_cx and (A) 50 mM NaCl or (B) 500 mM NaCl.

Figure 4.

Clamp loader–clamp dissociation reactions at 20°C. The decrease in fluorescence that occurs when β-(AF488)₂ dissociates from γ_cx was measured in assays containing 40 nM γ_cx, 10 nM β-S109C/Q299C-(AF488)₂ (cyan) or β-S109C/Q299C/R103S-(AF488)₂ (magenta), 400 nM unlabeled β, and 0.5 mM ATP in buffer with (A) 50 mM NaCl or (B) 500 mM NaCl. Data were fit to the sum of two exponentials (black lines through traces), and rate constants and fractional amplitudes derived from fits are given in Table 2.

Figure 5.

Clamp loading on DNA bound by SSB at 20°C. The decrease in AF488 fluorescence that occurs when clamps are closed on DNA was measured in stopped-flow reactions in which a solution of γ_cx, β-(AF488)₂, and ATP were mixed with a solution of SSB-bound DNA and excess unlabeled β. DNA is symmetrical, as illustrated, with two 30-nt single-stranded 5′ DNA overhangs and a 30-nt duplex. Final concentrations were 20 nM β-S109C/Q299C-(AF488)₂ or β-S109C/Q299C/R103S-(AF488)₂, 20nM γ_cx, 200 nM unlabeled β, 160 nM DNA, 960 nM SSB and 0.5 mM ATP in stopped-flow assay buffer containing 50 mM NaCl.

Figure 6.

Clamp loader binding to singly AF488-labeled clamps affects AF488 fluorescence. Binding reactions were measured side-by-side in stopped-flow experiments for β-S109C/Q299C-(AF488)₂, β-S109C-AF488, and β-Q299C-AF488. Fluorescence intensity is plotted relative to the signal for unbound clamp in each case, but note the absolute fluorescence for singly-labeled clamps is greater than for the doubly-labeled clamp because there is no AF488-dimer quenching. (A) Clamp binding for β-S109C/Q299C-(AF488)₂ (black) versus β-S109C-AF488 (grey) is shown. (B) Clamp binding for β-S109C/Q299C-(AF488)₂ (black) versus β-Q299C-AF488 (grey) is shown. (C) Rates of γ_cx binding to β-S109C-AF488 were measured in reactions containing 20 nM β-S109C-AF488, 0.5 mM ATP, and 20 (magenta), 40 (light blue), 80 (green), 160 (grey) and 320 (dark blue) nM γ_cx in stopped-flow assay buffer containing 50 mM NaCl. (D) Time courses from panel C were fit to double exponentials, and observed rate constants for the rapid (black diamonds) and slow (grey circles) phases obtained from the fits are plotted as a function of γ_cx concentration.

Figure 7.

Clamp loader binding to β-S109C-AF488 and opening β-R103C/Q299C-(AF488)₂ were fit to the model shown in the scheme where γ complex (γ_cx) binds a closed β-clamp (β^c) to form a closed complex, the clamp opens (β^o) to form an open complex, and conformational rearrangements occur to form a ‘locked open’ clamp (β^oL). (A) Clamp binding reactions containing 40 (yellow), 80 (lavender), and 160 nM (green), 320 (light blue), 640 (orange) and 1280 nM (grey) γ_cx are shown. (B) Clamp opening reactions containing 40 (yellow), 80 (lavender), 160 nM (green), 250 nM (blue), 500 nM (red) and 1800 nM (tan) γ_cx are shown. Black lines through the data are the result of the fit of the model using KinTek Explorer with rate constants shown in the scheme.