Structure of the human CTF18−RFC clamp loader bound to PCNA
- Bioscience Program, Division of Biological and Environmental Sciences and Engineering, King Abdullah University of Science and Technology, Saudi Arabia
- Department of Chemistry, The Pennsylvania State University, United States
- KAUST Center of Excellence for Smart Health, King Abdullah University of Science and Technology, Saudi Arabia
Figures
Figure 1 with 4 supplements
Cryo-EM structure of the human CTF18–RFC–PCNA complex in the presence of ATP.
(a) Domain organization of the CTF18–RFC subunits forming the RFC pentamer. The dotted line above the Ctf18 subunit highlights the regions that are observed in the cryo-EM map. (b) Two views of the cryo-EM map of the CTF18–RFC–PCNA complex, coloured by subunits. The less sharp definition of the map region corresponding to the AAA+ domain of the Ctf18 subunit supports a relatively high mobility of this subunit. (c) Molecular model of the CTF18–RFC–PCNA complex in ribbon representation. (d) Structures of the human (left) and Saccharomyces cerevisiae (right) RFC–PCNA complex, which present the RFC pentamer in an autoinhibited conformation analogous to the one captured in the CTF18–RFC–PCNA complex.
Figure 1—figure supplement 1
SDS–PAGE gel (4–20%) of the purified human clamp loader CTF18–RFC WT and CTF18Δ165–194–RFC.
Bands corresponding to the different subunits are labelled. The molecular weight standard is shown.
-
Figure 1—figure supplement 1—source data 1
TIFF file containing uncropped SDS–PAGE gel image indicating the relevant bands.
- https://cdn.elifesciences.org/articles/103493/elife-103493-fig1-figsupp1-data1-v1.tiff
-
Figure 1—figure supplement 1—source data 2
TIFF file containing uncropped and unlabelled SDS–PAGE gel image.
- https://cdn.elifesciences.org/articles/103493/elife-103493-fig1-figsupp1-data2-v1.tiff
Figure 1—figure supplement 2
Cryo-EM of the CTF18–PCNA complex in the presence of ATP.
(a) Representative electron micrograph acquired on a Falcon 4i electron detector in counting mode, and representative 2D class averages. (b) Angular distributions of projections. (c) Gold-standard Fourier shell correlation for the reconstruction of the full complex after post-processing, and resolution estimation using the 0.143 criterion. (d) Cryo-EM map of the complex, colour-coded by local resolution.
Figure 1—figure supplement 3
Workflow of cryo-EM image processing and 3D reconstruction of CTF18–RFC–PCNA complex in the presence of 0.5 mM ATP and without Mg2+ (Dataset 1).
Figure 1—figure supplement 4
Interfaces between RFC1 or Ctf18 and RFC2, and RFC1 or Ctf18 and RFC3.
(a) Comparison of the AAA+ module interfaces. Key interactions observed with RFC1 are lost with Ctf18. (b) Comparison of the collar module interface. Far fewer interactions are observed with Ctf18 compared to RFC1. Salt bridges are shown in turquoise and h-bonds in green.
Figure 2
Structure of the Ctf18 subunit.
(a) Sequence of the human Ctf18 subunit, highlighting the secondary structural elements as assigned in the cryo-EM model. The atypical PCNA-interacting motif (PIP-box) and conserved DEXX box are boxed. Residues that are not modelled in the cryo-EM structure are labelled as X in grey. Below the Ctf18 sequence to the left, the sequences of the N-terminus of human and other lower eukaryotes Ctf18, together with the secondary prediction by AlphaFold (Jumper et al., 2021), are shown, highlighting the conservation of the b-hairpin fold. Below, the cryo-EM β-hairpin fold model is overlaid onto the corresponding motifs predicted by AlphaFold for Drosophila or Saccharomyces cerevisiae Ctf18. Below the Ctf18 sequence to the right, the alignment of the C-terminal region of human and S. cerevisiae Ctf18, which interacts with the Ctf8/Dcc1 subunits, is shown. Asterisks denote conserved residues. (b) Superposition of the human Ctf18 and RFC1 (PDB: 6VVO) subunit structures, with subdomains labelled. (c) Structure of the Ctf18 subunit, in ribbon representation, within the CTF18–RFC–PCNA complex. The Ctf18 subdomains are colour-coded as in panel (a). The other complex subunits are shown as surfaces.
Figure 3
Structural details of the Ctf18–RFC5 and CTF18–RFC–PCNA interactions.
(a) Model of the RFC5 collar domain and Ctf18 N-terminal b-hairpin in ribbon representation. The inset shows the interface in stick representation, with hydrogen bonds shown as green dotted lines. (b) Map region around the Ctf18 N-terminal b-hairpin. (c) Model of the PCNA homotrimer bound to the four CTF18 interacting regions. To the right, the structure of the yeast RFC–PCNA complex with an open clamp, highlighting the engagement of all five RFC subunits to PCNA, required for clamp opening. (d) Model of the regions of the CTF18–RFC–PCNA interactions, with CTF18–RFC subunit residues shown as sticks and PCNA as surface or ribbon. Polar interactions are shown as black dotted lines. The amino acid sequence encompassing the interacting motifs in the various subunits is shown.
Figure 4 with 1 supplement
Analysis of the autoinhibited conformation of the CTF18 pentamer and nucleotide coordination.
(a) Ribbon model of the CTF18–RFC pentamer (PCNA hidden). Nucleotide molecules are shown as sticks, with the corresponding cryo-EM density. (b) Surface representation of (top) the CTF18–RFC pentamer and (bottom) the yeast RFC pentamer bound to open PCNA and primer–template DNA. PCNA is hidden in both structures. The overtwisted CTF18–RFC pentamer cannot accommodate DNA within the inner chamber. (c) Model of the ATP active sites in the CTF18–RFC–PCNA complex, showing disruption of the interactions mediated by the arginine fingers. Below, insets show the nucleotide-binding site of subunit RFC4 in either the human CTF18–RFC (left) or human RFC (right) models, highlighting the structural incompatibility with nucleotide binding in the CTF18–RFC structure. Polar interactions with ATPgS in human RFC4 are indicated as cyan dotted lines. (d) Detail of the inner chamber region of the CTF18–RFC pentamer, showing the b-hairpin (E-plug) of RFC5 protruding in the region predicted to be occupied by DNA. The DNA (grey ribbon) is modelled as in the yeast RFC–DNA–PCNA complex. (e) The structures of the RFC1 subunit from the yeast RFC–DNA–PCNA complex (Gaubitz et al., 2022) and the Ctf18 subunit from the CTF18–RFC–PCNA complex are aligned to highlight conservation of the separation pin. Side chains of residues implicated in this motif in yeast RFC, along with the corresponding amino acids in CTF18–RFC, are shown.
Figure 4—figure supplement 1
Walker A sequence and structure.
(a) Walker A sequence comparison between CTF18 and hRFC (PDB: 6VVO). (b) The upper panels illustrate the conformation of the Walker A motifs in Ctf18 and human RFC1 (PDB: 6VVO).
Figure 5 with 2 supplements
Cryo-EM structure of the human CTF18–RFC–PCNA complex in the presence of ATP and Mg2+.
(a) Two views of the cryo-EM map, coloured by subunits. (b) Molecular model in ribbon representation. The inset shows a map and model of the ATP molecule bound to the RFC2 subunit, with a discernible Mg2+ ion.
Figure 5—figure supplement 1
Cryo-EM of the CTF18–PCNA complex in the presence of ATP and Mg2+.
(a) Representative electron micrograph acquired on a Falcon 4i electron detector in counting mode, and representative 2D class averages. (b) Angular distributions of projections. (c) Gold-standard Fourier shell correlation for the reconstruction of the full complex after focused refinement, and resolution estimation using the 0.143 criterion. (d) Cryo-EM map of the complex, colour-coded by local resolution.
Figure 5—figure supplement 2
Workflow of cryo-EM image processing and 3D reconstruction of CTF18–RFC–PCNA complex in the presence of 0.5 mM ATP and 5 mM Mg2+ (Dataset 2).
Figure 6 with 7 supplements
FRET assays to monitor loading of PCNA onto P/T junctions by human clamp loader complexes.
(a) Schematic representation of the FRET pair and experiment to monitor loading of Cy5-PCNA onto 5′ddPCy3/T DNA substrates engaged by RPA. The front and back faces of PCNA are displayed in green and grey, respectively. When loaded onto a 5′ddPCy3/T DNA substrate, the Cy5 label on the back face of PCNA is oriented towards the Cy3 label near the blunt duplex end of the 5′ddPCy3/T DNA substrate, yielding a FRET signal (b–d) data. Each normalized EFRET trace is the mean of at least three independent traces with the SEM shown in grey. Each respective trace is fit to its corresponding minimal kinetic model and the times required for the trace to reach 50% (t0.50), 90% (t0.90), and 95% (t0.95) of the maximal value (i.e., 1.0) are reported. Data for loading of PCNA by RFC, CTF18, and CTF18Δ165–194–RFC are displayed in panels b, c, and d, respectively.
Figure 6—figure supplement 1
Workflow of cryo-EM image processing and 3D reconstruction of CTF18–RFCΔ165–194–PCNA complex in the presence of 0.5 mM ATP and without Mg2+ (Dataset 3).
Relion 5.0 was used for image processing and 3D reconstruction.
Figure 6—figure supplement 2
2D classification of CTF18 WT and CTF18Δ165–194–RFC.
(a) The 2D classification of CTF18–RFC WT reveals the formation of the CTF18–RFC–PCNA complex with clear structural integrity (515,195 particles in all the showed 2D classes). (b) The 2D classification of CTF18Δ165–194–RFC mutant highlights the impact of β-hairpin deletion on complex stability. The presence of a large majority of particles of PCNA alone (108,439 particles) indicates that the absence of the β-hairpin contributes to the increased instability of the complex. Only the 2D classes squared in red represent the CTF18Δ165–194–RFC–PCNA complex (14,616 particles).
Figure 6—figure supplement 3
P/T DNA substrate utilized in the FRET studies.
The sequences and lengths of the double- and single-stranded DNA regions are indicated. The size of the double-stranded DNA (dsDNA) region (29 bp) is in agreement with the requirements for assembly of a PCNA ring onto DNA by RFC (Hedglin et al., 2013; Hedglin and Benkovic, 2017; Hedglin et al., 2017). The single-stranded DNA (ssDNA) region accommodates 1 RPA heterotrimer (Kim et al., 1994; Kim et al., 1992; Kim and Wold, 1995). RPA prevents loaded PCNA from sliding off the ssDNA end of the substrate (Hedglin and Benkovic, 2017). When pre-bound to neutravidin, the biotin attached to the 5′-end of the primer strand of the substrate prevents loaded PCNA from sliding off the dsDNA end. The primer strand is terminated at the 3′-end with a dideoxy C nucleotide (shown) and, hence, cannot be extended by a DNA polymerase.
Figure 6—figure supplement 4
FRET assays to monitor loading of PCNA onto P/T junctions by human clamp loader complexes.
(a) Schematic representation of the FRET pair and experiment to monitor loading of Cy5-PCNA onto 5′ddPCy3/T DNA substrates engaged by RPA. The front and back faces of PCNA are displayed in green and grey, respectively. When loaded onto a 5′ddPCy3/T DNA substrate, the Cy5 label on the back face of PCNA is oriented towards the Cy3 label near the blunt duplex end of the 5′ddPCy3/T DNA substrate, yielding a FRET signal (b–d) data. Each trace is the mean of at least three independent traces with the SEM shown in grey. The time trajectories of I563 (magenta) and I665 (cyan) are displayed in the top panels and the corresponding EFRET (mustard) is displayed in the bottom panels. The time at which a clamp loader complex is added (t = 60 s) is indicated by a red arrow. Changes in I563 and I665 are indicated in the top panel by magenta and cyan arrows, respectively. The I563, I665, and the corresponding EFRET values observed prior to the addition of clamp loader complexes represent the complete absence of interactions between 5′ddPCy3/T·RPA complexes and Cy5-PCNA. These values are each fit to a flat line that is extrapolated to the axis limits where the Y-intercept of the fit for the EFRET is equivalent to YMin (indicated). The EFRET values observed over the last 60 s of the plateaus for the observed EFRET increases are each fit to a flat line that is extrapolated to the axis limits where the Y-intercepts are equivalent to YMax (indicated). Data for loading of PCNA by RFC, CTF18, and CTF18Δ165–194–RFC are displayed in panels b, c, and d, respectively.
Figure 6—figure supplement 5
The minimal kinetic model for the increase observed in the normalized EFRET trace for CTF18 is a single-exponential rise.
The normalized EFRET trace is displayed in the top panel (in yellow) and fit to a single-exponential rise. The trace is the mean of at least three independent traces with the SEM shown in grey. The residuals from the corresponding standard curve fitting of normalized EFRET trace are displayed in the bottom panel (in orange).
Figure 6—figure supplement 6
The minimal kinetic models for the increase observed in the normalized EFRET traces for RFC and CTF18Δ165–194–RFC are double-exponential rises.
The normalized EFRET traces are displayed in the top panels (in yellow) and fit to a kinetic model (indicated). Each trace is the mean of at least three independent traces with the SEM shown in grey. The respective residuals from the standard curve fittings of the corresponding normalized EFRET traces are displayed in the bottom panels (in orange). Standard curve fittings of the normalized EFRET traces for RFC and CTF18Δ165–194–RFC to double-exponential rises are shown in a and b, respectively. Standard curve fittings of the normalized EFRET traces for RFC and CTF18Δ165–194–RFC to single-exponential rises are shown in c and d, respectively.
Figure 6—figure supplement 7
Direct comparisons of the minimal kinetic models for the increase observed in the normalized EFRET traces for RFC (black), CTF18–RFC (green), and CTF18Δ165–194–RFC (purple).
For perspective, dashed, flat grey lines at 0.5, 0.9, and 0.95. The time courses to 350 and 150 s are displayed in (a) and (b), respectively.
Figure 7 with 1 supplement
Primer extension assay with Pol ε and CTF18–RFC WT or CTF18△165–194–RFC.
(a) Time course reactions on M13mp18 single-strand DNA with 40 nM wild-type CTF18–RFC or CTF18Δ165–194–RFC. Reactions were performed as described in Materials and methods. (b) Quantification of the band intensities from the primer extension assay shows that the rate of increase in intensity, derived from the initial linear portion of the reaction time before saturation, is 4142 ± 95 a.u./min for CTF18–RFC and 2389 ± 143 a.u./min for CTF18Δ165–194–RFC. This indicates that the mutant slows Pol ε synthesis by 42%. The experiment was conducted three times. Dots in (b) represent mean values from the three replicas.
-
Figure 7—source data 1
TIFF file containing uncropped agarose gel image indicating the relevant bands.
- https://cdn.elifesciences.org/articles/103493/elife-103493-fig7-data1-v1.zip
-
Figure 7—source data 2
TIFF file containing uncropped and unlabelled agarose gel image.
- https://cdn.elifesciences.org/articles/103493/elife-103493-fig7-data2-v1.zip
Figure 7—figure supplement 1
Primer extension assays with Pol ε and CTF18 WT and CTF18Δ165–194–RFC.
CTF18–RFC or CTF18Δ165–194–RFC concentration titration reactions on M13mp18 single-strand DNA. Reactions were running for 10 min at 37°C.
-
Figure 7—figure supplement 1—source data 1
TIFF file containing uncropped agarose gel image indicating the relevant bands.
- https://cdn.elifesciences.org/articles/103493/elife-103493-fig7-figsupp1-data1-v1.zip
-
Figure 7—figure supplement 1—source data 2
TIFF file containing uncropped and unlabelled agarose gel image.
- https://cdn.elifesciences.org/articles/103493/elife-103493-fig7-figsupp1-data2-v1.zip
Figure 8
Deletion of the N-terminal β-hairpin in the Ctf18 large subunit slows down PCNA loading and reduces DNA synthesis by Pol ε.
(a) The structure of the RFC module of CTF18–RFC is stabilized by the presence of a β-hairpin in the Ctf18 subunit (amino acids 165–194). Thus, wild-type CTF18–RFC loads PCNA onto DNA efficiently and stimulates DNA synthesis by Pol ε interacting with the polymerase. (b) The stability of the CTF18–RFC–PCNA complex is diminished by the deletion of the β-hairpin. Consequently, a CTF18–RFC deletion mutant lacking the β-hairpin (CTF18Δ165–194–RFC) shows slower loading of PCNA and lower activity in stimulating primer synthesis by Pol ε.
Videos
Video 1
3DVA analysis of the human CTF18–RFC–PCNA complex in the presence of ATP.
The video shows morphing transitions across the first three principal components obtained by 3D variability analysis (3DVA). Three volumes per component are morphed, illustrating continuous conformational flexibility within the complex. The dominant motion is localized to the AAA+ domain of the CTF18 subunit, consistent with reduced local resolution in this region and supporting a model in which this domain.
Additional files
-
Supplementary file 1
Cryo-EM data collection, model refinement, and validation statistics.
- https://cdn.elifesciences.org/articles/103493/elife-103493-supp1-v1.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/103493/elife-103493-mdarchecklist1-v1.docx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Structure of the human CTF18−RFC clamp loader bound to PCNA
eLife 13:RP103493.
https://doi.org/10.7554/eLife.103493.4
