Kuriyan Lab Logo        
Randall Mcnally

Analysis of the role of PCNA-DNA contacts during clamp loading

Randall McNally, Gregory D Bowman, Eric R Goedken, Mike O'Donnell and John Kuriyan

BMC Structural Biology 2010, 10:3 doi:10.1186/1472-6807-10-3 (local copy)

Abstract / Figures from the paper


Background: Sliding clamps, such as Proliferating Cell Nuclear Antigen (PCNA) in eukaryotes, are ring-shaped protein complexes that encircle DNA and enable highly processive DNA replication by serving as docking sites for DNA polymerases. In an ATP-dependent reaction, clamp loader complexes, such as the Replication Factor-C (RFC) complex in eukaryotes, open the clamp and load it around primer-template DNA.

Results: We built a model of RFC bound to PCNA and DNA based on existing crystal structures of clamp loaders. This model suggests that DNA would enter the clamp at an angle during clamp loading, thereby interacting with positively charged residues in the center of PCNA. We show that simultaneous mutation of Lys 20, Lys 77, Arg 80, and Arg 149, which interact with DNA in the RFC-PCNA-DNA model, compromises the ability of yeast PCNA to stimulate the DNA-dependent ATPase activity of RFC when the DNA is long enough to extend through the clamp. Fluorescence anisotropy binding experiments show that the inability of the mutant clamp proteins to stimulate RFC ATPase activity is likely caused by reduction in the affinity of the RFC-PCNA complex for DNA. We obtained several crystal forms of yeast PCNA-DNA complexes, measuring X-ray diffraction data to 3.0 Å resolution for one such complex. The resulting electron density maps show that DNA is bound in a tilted orientation relative to PCNA, but makes different contacts than those implicated in clamp loading. Because of apparent partial disorder in the DNA, we restricted refinement of the DNA to a rigid body model. This result contrasts with previous analysis of a bacterial clamp bound to DNA, where the DNA was well resolved.

Conclusion: Mutational analysis of PCNA suggests that positively charged residues in the center of the clamp create a binding surface that makes contact with DNA. Disruption of this positive surface, which had not previously been implicated in clamp loading function, reduces RFC ATPase activity in the presence of DNA, most likely by reducing the affinity of RFC and PCNA for DNA. The interaction of DNA is not, however, restricted to one orientation, as indicated by analysis of the PCNA-DNA co-crystals.

Figures from the paper.

Click on the small image to get a bigger one.

First figure from paper

Figure 1: Clamp loader structure and clamp loader cycle.
(A) The E. coli clamp loader, γ-complex, bound to primer-template DNA [30]. On left, the domain structure of the clamp loader subunits is noted.
(B) Schematic diagram of the clamp loader cycle.
Illustrator file, local only

First figure from paper

Figure 2: Model of RFC bound to PCNA and DNA.
(A) The RFC-PCNA-DNA model, which depicts the path of DNA bound to the clamp loader and through the clamp, combines features of a DNA-bound γ-complex structure [30] and an RFC-PCNA structure [31].
(B) The α-helices lining the hole of PCNA, with the positions of the nine arginines and lysines that reside on them indicated.
(C) Top and side views of the positions of the DNA molecule from the RFC-PCNA-DNA model relative to PCNA, with a model of PCNA opened by RFC overlaid upon the closed PCNA. The overlay of the open clamp shows that much of the PCNA-DNA interaction remains in place when the clamp is opened and pulled out of axis.
Illustrator file, local only

First figure from paper

Figure 3: Stimulation of DNA-dependent ATPase activity of RFC by mutant clamps.
(A-D) Results of enzyme-coupled ATPase assay (each reaction contained 50 nM RFC, 100 nM PCNA, 100 nM DNA, and 0.5 μM ATP; see Methods). The ATPase rate of each reaction is displayed relative to the ATPase rate of RFC in the presence of PCNA and DNA-30, which is scaled to a value of one. Error bars represent standard deviation of multiple trials. DNA-30 has 30 base pairs in the duplex region and a 10-base 5’ overhang on the primer strand; DNA-25 has 25 base pairs in the duplex region, etc. DNA-13/overhang is identical to DNA-13 but contains a 17-base 3’ overhang on the template strand.
(A) Wild- type PCNA stimulates RFC ATPase activity in the presence of DNA.
(B) The point mutations in PCNA that result in the largest effects on ATPase activity in the presence of DNA are R80A and R149A, which have deficiencies of only 20-25% relative to wild-type PCNA.
(C) Simultaneous mutation of PCNA residues K20A, K77A, R80A, and R149A results in an inability of PCNA to stimulate DNA-dependent ATPase activity.
(D) PCNA K20A/K77A/R80A/R149A is deficient in stimulating ATPase activity when the primer-template DNA constructs present are long enough to extend through the clamp during loading, but its activity approaches that of wild-type PCNA in the presence of short DNA that cannot reach into the clamp.
Illustrator file, local only

First figure from paper

Figure 4: Affinity of RFC and mutant PCNA for DNA as measured by fluorescence anisotropy.
(A) Fluorescence anisotropy binding curves for RFC titrated into a mixture of 1 μM PCNA (or PCNA K20A/K77A/R80A/R149A) and 100 nM TAMRA-labeled DNA-30 (30 base pairs in the duplex region) or DNA-13 (13 base pairs in the duplex region) in the presence of 1 mM ATP-γ-S. Curves represent best fit to a one-site binding equation (see Methods).
(B) Bar graph presenting dissociation constants of binding data from Figure 4A. Error bars represent standard error of the curve fit.
(C) The clamp loader cycle is broken into three parts; fluorescence anisotropy binding data suggests that step 2 of the cycle, the formation of the RFC-PCNA-DNA ternary complex, is hindered by PCNA K20A/K77A/R80A/R149A through the loss of compatibility between DNA and the surface of the clamp’s hole.
Illustrator file, local only

First figure from paper

Figure 5: X-ray derived model of PCNA bound to DNA.
(A) Crystals of single-chain PCNA-DNA. The DNA is labeled on the 5’ end of the primer strand with a Cy5 chromophore.
(B) Unbiased electron density for the DNA produced from single-chain PCNA-DNA crystal, calculated before DNA was placed into the model. 2Fo-Fc map is displayed at a contour level of 0.7σ. Positions of the DNA phosphates in the final model are shown as spheres.
(C) PCNA-DNA model derived from X-ray diffraction data. The DNA makes a ~40° angle with the central axis of PCNA.
Illustrator file, local only

First figure from paper

Figure 6: PCNA-DNA contacts in the X-ray derived PCNA-DNA model.
(A) Electrostatic surface representation of the center of PCNA; positive electrostatic surfaces are colored blue, negative surfaces are red, and neutral surfaces are white. Note the interactions made between the negatively charged backbone of DNA strand 1 (depicted as spheres) and the positively charged patch located in the center of PCNA formed by residues Lys 13, Arg 14, Lys 146, and Arg 149. The PCNA subunit in the foreground is removed for clarity.
(B) Contacts between DNA and the His 190 loop and Asn 84 loops are indicated.
(C) The PCNA-DNA model and an adjacent PCNA molecule in the crystal lattice (the adjacent PCNA molecule is rendered as a surface). The DNA end protruding from the distal face of the clamp forms a crystal contact.
(D) Close-up view of the positions of the DNA molecules from the RFC-PCNA-DNA model and the PCNA-DNA model relative to PCNA. The angles of the DNA molecules in the RFC-PCNA-DNA model and the PCNA-DNA model differ by ~26°. Note the positions of residues identified to interact with DNA from the RFC ATPase assay and the PCNA-DNA model. DNA phosphates are shown as spheres. A PCNA subunit is removed from the foreground for clarity.
(E) Overlay of PCNA-DNA model and β clamp-DNA structure [25], aligned on their respective clamps. The β clamp is removed for clarity. On the right, the angles of the two DNA molecules are compared relative to the rotation axis of the clamp to which they are bound.
Illustrator file, local only