1. Structural Biology and Molecular Biophysics
Download icon

A processive rotary mechanism couples substrate unfolding and proteolysis in the ClpXP degradation machinery

  1. Zev A Ripstein  Is a corresponding author
  2. Siavash Vahidi  Is a corresponding author
  3. Walid A Houry
  4. John L Rubinstein  Is a corresponding author
  5. Lewis E Kay  Is a corresponding author
  1. University of Toronto, Canada
  2. The Hospital for Sick Children Research Institute, Canada
Research Article
Cite this article as: eLife 2020;9:e52158 doi: 10.7554/eLife.52158
6 figures, 3 videos, 2 tables, 6 data sets and 1 additional file


Figure 1 with 5 supplements
Functional and structural characterization of ClpXP from N. meningitidis.

(A) Schematic representation of the ClpXP degradation machinery. The overall positions of the substrate (orange), ClpX unfoldase (green), and the ClpP protease (blue) are shown; (B) GFP-SsrA degradation by ClpXP is ATP-dependent. The degradation rate of GFP-SsrA is monitored by measuring loss of fluorescence. The components (ClpX, ClpP, ATP) included for each measurement are denoted on the plot. All measurements included GFP-SsrA and were performed in triplicate on WT ClpXP; (C) ClpXP complex formation monitored by size exclusion chromatography (SEC). SEC profiles of isolated ClpP (blue trace) and ClpX (green trace) are consistent with their expected molecular weights. SEC profile of a 2:1 ClpX6:ClpP14 mixture (red trace) incubated for 10 min in the presence of MgATP shows the formation of doubly- and singly-capped ClpXP complexes. The running buffer for all traces contained 2 mM MgATP. SDS-PAGE gel of SEC fractions (fractions within the region delineated by black bracket) shows that they contain both ClpP and ClpX; (D) Overall architecture of the ClpXP complex as established by cryo-EM. The ClpP double ring component is shown in dark blue, with the exception of a pair of opposing protomers shown in light blue shade to delineate the IGF loop binding site that is located at the interface between protomers. Cutaway side (E) and top (F) views of Conformations A and B of ClpX resulting from focused classification and local refinement. The IGF loops that bind to ClpP, and the presence of substrate (orange) in the axial channel of ClpX, are highlighted. Each of the ClpX protomers is labeled. In panel (F), the engagement of the substrate (orange) with five of the six pore loops is indicated (red semi-circle).

Figure 1—figure supplement 1
Sequence map of NmClpX and SEC analysis of substrate binding by NmClpXP.

(A) NmClpX primary sequence domain map. Important structural and functional motifs are denoted. (B) NmClpXP binds a SsrA-tagged GFP. SEC profiles for ClpXP incubated with GFP-SsrA (top panel) and GFP-SsrA alone (bottom panel) measured using absorbance at 280 nm (black trace) and GFP fluorescence (green trace - λex:480 nm, λem:508 nm). The region used for SDS-PAGE analysis is highlighted by * (Figure 1C).

Figure 1—figure supplement 2
Cryo-EM image processing.

(A) Representative cryo-EM micrograph of NmClpXP bound to GFP-SsrA. Example complexes are circled in white. (B) Representative 2D class averages of NmClpXP with GFP-SsrA. (C) Image processing workflow used to obtain final maps.

Figure 1—figure supplement 3
Cryo-EM map validation.

(A to C) Fourier Shell Correlation (FSC) plots (left) and orientation distribution (right). Resolution values reported are for FSC = 0.143.

Figure 1—figure supplement 4
Examples of regions of the atomic models built into the experimental cryo-EM maps.
Figure 1—figure supplement 5
Experimental density maps and models for the substrate pore-1 loop interaction interface.
Figure 2 with 2 supplements
The interaction interface between ClpX and ClpP.

Protomer X1 occupies the US position in Conformation A and protomer X6 the LS position in Conformation B. (A) Cutaway density map of the overall architecture of the ClpXP interaction interface. (B) Models for the ClpP apical surface and the ClpX IGF loops (residues 265–275). The empty IGF-binding pocket (dotted oval) resides clockwise to the X5 protomer in both conformations. In Conformation A, an IGF loop was not built into the map for protomer X6 due to weak loop density (yellow dotted line). (C) Model in map fit for an IGF loop (grey), surrounded by regions of ClpP comprising the IGF binding site (blue). (D) Interaction of the ClpP N-terminal gates with ClpX. The X2 protomer moves away from the gates, while the X5 protomer directly contacts the gate of the ClpP protomer to which its IGF loop is bound, shown in (E). Inset shows the details of this interaction, with model in map fit for the ClpP β-hairpin. (F) Positions of ClpX protomers relative to ClpP. Images were generated by fitting all ClpP protomers to a common protomer and displaying the corresponding ClpX protomer. In both conformations, ClpX adopts a spiral arrangement relative to ClpP (dotted red line). In Conformation A, the X2 protomer is located at the top of the spiral, while in Conformation B the X1 protomer occupies the top position, ~7 Å higher than its position in Conformation A. The ClpX protomers rotate and translate relative to ClpP; protomers X2 and X3 sit nearly atop their ClpP protomers, while X4 and especially X5 show large deviations from this position (vertical lines), with X5 sitting overtop the ClpP axial pore. (G) View down the channel formed by the N-terminal ClpP β-hairpins (left); density and models for the most flexible and rigid β-hairpins (center and right, respectively).

Figure 2—figure supplement 1
IGF loop flexibility.

(A and B) Large domains of ClpX were aligned and models for the IGF loops are shown. (C and D) Unsharpened density map for Conformations A and B looking at protomer X6.

Figure 2—figure supplement 2
The ClpP apical loops extend upwards from the ring surface.

(A) Unsharpened density map of the ClpP N-terminal β-hairpins when bound to ClpX. (B) Cryo-EM map of apo ClpP from N. meningitidis showing that the apical loops no longer extend ‘up’. (C) ClpX-bound and apo forms of ClpP overlaid. There is no noticeable density for the turn of the ClpP N-terminal β-hairpin in the apo form, and the β-hairpins point inwards, constricting the axial pore of ClpP. (D) Fourier Shell Correlation (FSC) plot for Apo-NmClpP. (E) Closeup view of the N-terminal β-hairpins when bound to ClpX. (F) Density map of the β-hairpins in the ClpXP complex, colored by local resolution.

Figure 3 with 1 supplement
Substrate engagement by the ClpX pore loops.

(A and B) Pore-1 loop residues grip the substrate, as observed in this view looking down the axial channel. The substrate is modeled as poly-Ala. (A) In Conformation A, the X1 protomer is disengaged from the substrate in the US position (red arrow) and shows noticeable flexibility in its pore-1 loop. (B) In Conformation B, the X1 protomer contacts the substrate but the X6 protomer has disengaged into the LS position (red arrows). (C and D) Model of substrate and interacting residues, along with schematic, viewed perpendicular to the axial channel. (C) In Conformation A, only the Tyr153 residue of the pore-1 loop makes significant contacts with the substrate; the five protomers (X2, X3, X4, X5, X6) form a downward spiral surrounding the backbone of the substrate chain, extending to the apical surface of ClpP. (D) In Conformation B, in addition to the five Tyr153 residues from the five gripping protomers (X1, X2, X3, X4, X5), more contacts are formed: Ser197 and Ile198 from the pore-2 loop of the X1 protomer and His230 from the RKH loop of the X5 protomer. (E) Substrate-induced conformational changes of the X6 pore-1 loop upon releasing client. When the X6 protomer is substrate engaged its pore-1 loop forms a lasso like conformation, with Tyr153 and Val154 both pointing toward the axial channel (left). When disengaged, a part of the pore-1 loop becomes an α-helix, with Val154 and Tyr153 no longer pointing outwards (right). (F) Architecture of the additional contacts made by the pore-2 and RKH loops from the X1 and X5 protomers, respectively.

Figure 3—figure supplement 1
RKH loop positions.

(A and B) Unsharpened maps for both conformations A and B of ClpX are shown, with RKH loops highlighted in color (coloring done by a radius of 10 Å around residues 220–240).

Figure 4 with 3 supplements
Nucleotide occupancy and interactions with ClpX.

(A and B) Model of ClpX in conformations A and B, looking into the ClpX pore. Nucleotides are shown and color-coded, with ATP red and ADP purple, and bind between the large and small domains (boxed). Inset shows a schematic of the protomer positions and nucleotide occupancies. Conformation A has two bound ADP, in the X1 (dark blue, US in Conformation A in this representation) and X6 (yellow, LS in Conformation B in this representation) protomers, with the X1 protomer disengaged from the substrate and away from the axial pore (US position). In Conformation B only a single ADP is bound, protomer X1 is reengaged with substrate and X6 has disengaged. (C) Nucleotide-binding sites of Conformation A, with key interacting residues shown along with the experimental density maps corresponding to bound nucleotide. In the ADP-bound sites, both the arginine finger (R306) and the sensor-II arginine (R369) have moved away, while in the ATP bound sites they form close contacts with the β and γ-phosphates. (D) Nucleotide binding in Conformation B. As in Conformation A, R306 and R369 have moved away from the ADP, while in the X5 protomer only the arginine finger has moved away, while the sensor-II transitions closer to the γ-phosphate.

Figure 4—figure supplement 1
Relative orientation between large and small domains of ClpX.

(A and B) Residues 308–343 are shown. Note that protomers X1 and X6 in Conformations A and B, respectively, are displaced from substrate. These show among the largest differences in domain orientations.

Figure 4—figure supplement 2
Local resolution maps of ClpX.

(A and B) Density maps colored by local resolution for Conformations A and B. In Conformation A the US protomer is highlighted with a dashed line.

Figure 4—figure supplement 3
ATP-binding pocket densities.

Experimental density maps and models are shown for the ATP-binding pockets of all protomers. Two numerical thresholds are shown to highlight differences in density.

Translocation model for ClpXP.

(A) Schematic of the ClpX-nucleotide-dependent conformational cycle. The transition from Conformation A to Conformation B is mediated through nucleotide exchange in the US position (1st to 2nd ring), with ATP hydrolysis and phosphate release in the ATP* position restoring the A conformation from the B state (2nd to 3rd ring) so that the cycle can repeat. In the new Conformation A (3rd ring), protomer X6 (yellow) adopts the US conformation and the X1 protomer (previous US) moves to the top of the spiral. In this way each protomer of the complex cycles through all positions as ATP is hydrolyzed in successively new ATP* protomer states, requiring the hydrolysis of seven ATPs. (B) Hand-over-hand mechanism of pore-loop mediated translocation. In the first A to B transition depicted in (A) X1 (US protomer) reengages at the top of the spiral and the LS pore one loop (X6 in Conformation B) disengages from the bottom of the spiral. As this is repeated the pore loops move ‘up’ the substrate and pull it into the degradation chamber. (C) IGF loop movements accompanying the hand-over-hand movements of ClpX. Shown in the small circles are the seven IGF-loop-binding sites in ClpP with a ‘+’ sign indicating engagement with an IGF loop of a given ClpX protomer. The binding sites are color coded according to the color of the protomer that interacts with the corresponding site. In conformation A, the IGF loop of X6 is unbound or only weakly bound, while in Conformation B it is bound. A new Conformation A is generated when the IGF loop from the ATP* protomer unbinds. It subsequently reengages with ClpP in a new Conformation B (third and fourth rings in (A)). In this way, the IGF loops ‘step’ around the ClpP ring, with each complete rotation requiring the hydrolysis of seven ATPs. (D) Schematic of the ClpX-ClpP interactions during seven ATP hydrolysis steps that lead to a net rotation of ClpX protomers by 60o. Because of the six- to seven-fold symmetry mismatch a total of forty-two ATPs must be hydrolyzed to return the protomers back to their original starting positions in the cycle (i.e. ring 1). Each of the seven ring structures corresponds to Conformation A, with the red arrows indicating the subsequent step of the IGF loop of the ATP* protomer along the cycle. The US protomer in each step is highlighted by the dashed trace.

Appendix 1—figure 1
Peptidase rate measurements of NmClpP as a function of pH, monitored using the fluorogenic substrate PKM-AMC.


Video 1
Overview of the substrate-bound ClpXP complex (Conformation B).
Video 2
Hand-over-hand translocation cycle of the ClpXP complex.

Interpolation between states A → B → A → B. Side and top views of the complex are shown (top), along with the pore-1 loop Tyr153 and substrate (bottom left) as well as IGF loops (bottom right).

Video 3
ClpX precession on the ClpP apical surface.

Repeated interpolation between states reveals a precession of the ClpX ring as its IGF loops are placed successively in different ClpP binding sites. The position of the empty binding pocket on ClpP undergoes a complete cycle around the complex for every 60o rotation of the ClpX ring relative to the ClpP ring.


Table 1
Cryo-EM data acquisition, processing, atomic model statistics, and map/model depositions.
A. Cryo-EM data acquisition and image processing.
Data collection
NmClpXP + GFP-SsrAapo-NmClpP
Electron MicroscopeTitan KriosTecnai F20
CameraFalcon 3ECK2 Summit
Voltage (kV)300200
Nominal Magnification75,00025,000
Calibrated physical pixel size (Å)1.061.45
Total exposure (e/Å2)42.735
Exposure rate (e/pixel/s)0.85
Number of frames3030
Defocus range (μm)0.9 to 1.70.8 to 3.3
Image processing
Motion correction softwarecryoSPARC v2cryoSPARC v2
CTF estimation softwarecryoSPARC v2cryoSPARC v2
Particle selection softwarecryoSPARC v2cryoSPARC v2
Micrographs used2680122
Particle images selected466,549100,132
3D map classification and refinement softwarecryoSPARC v2cryoSPARC v2
B. Map and model statistics.
EM mapsNmClpXP with SymmetryFocused NmClpX
Conformation A
Focused NmClpX
Conformation B
Particle images contributing to maps377,234110,696178,44850,403
Applied symmetryD7C1C1D7
Applied B-factor (Å2)−116.2−84.1−88.1−229.7
Global resolution (FSC = 0.143, Å)
Model BuildingNmClpX Conformation A focusedNmClpX Conformation B focusedNmClpP with
Conformation A combined
Conformation B combined
Modeling softwareCoot, Phenix, Rosetta
Number of residues19122018267145834689
RMS bond length (Å)0.00430.00370.00390.00440.0038
RMS bond angle (°)1.040.930.911.000.92
Ramachandaran outliers (%)
Ramachandran favored (%)98.3397.3295.6897.0796.39
Rotamer outliers0.
C-beta deviations00000
MolProbity score0.881.
EMRinger score2.783.425.6
Map-Model CC_mask0.750.780.82
Ligand2 ADP
4 Mg-ATP
5 Mg-ATP
4 Mg-ATP
5 Mg-ATP
C. Residues excluded in atomic models*.
ClpX protomerNmClpXP
Conformation A combined
Conformation B combined
X1 (Chain A)1–62, 102–109, 149–156, 191–204, 224–236, 263–264, 273–282, 413–4141–62, 228–233, 275–278, 413–414
X2 (Chain B)1–62, 102–111, 192–198, 226–235, 275–279, 413–4141–62, 226–233, 275–279, 413–414
X3 (Chain C)1–62, 192–198, 225–232, 275–279, 413–4141–62, 192–198, 228–233, 274–278, 413–414
X4 (Chain D)1–62, 103–106, 193–198, 226–234, 276–280, 413–4141–62, 193–198, 229–233, 413–414
X5 (Chain E)1–62, 193–198, 229–233, 277–281, 413–4141–62, 193–198, 277–279, 413–414
X6 (Chain F)1–62, 193–201, 225–234, 259–287, 413–4141–62, 191–204, 227–233, 275–282, 413–414
ClpP subunitsNmClpXP
Conformation A combined
Conformation B combined
Chain H1–5, 16–18, 199–2041–5, 16–18, 199–204
Chain I1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain J1–5, 17, 199–2041–5, 17, 199–204
Chain K1–5, 199–2041–5, 199–204
Chain L1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain M1–5, 13–19, 199–2041–5, 13–19, 199–204
Chain N1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain O1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain P1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain Q1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain R1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain S1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain T1–5, 16–17, 199–2041–5, 16–17, 199–204
Chain U1–5, 16–17, 199–2041–5, 16–17, 199–204
D. Deposited maps and associated coordinate files.
MapsEMDB codeAssociated PDB ID
NmClpXP Conformation AEMD-211876VFS
NmClpXP Conformation BEMD-211946VFX
NmClpXP D7EMD-21195
  1. *Protomer X1 is US in Conformation A and protomer X2 is LS in Conformation B.

Key resources table
Reagent type
(species) or resource
DesignationSource or referenceIdentifiersAdditional
Gene (Neisseria meningitidis)clpXUniProtKB - Q9JYY3
Gene (Neisseria meningitidis)clpPUniProtKB - Q9JZ38
Strain, strain background (Escherichia coli)BL21(DE3)Sigma-AldrichCMC0016Chemically competent cells
Recombinant DNA reagentpET28-NmClpPSynthetic
UniProtKB - Q9JYY3Plasmid containing ClpP
Recombinant DNA reagentpET28a-NmClpXSynthetic
UniProtKB - Q9JZ38Plasmid containing ClpX
Recombinant DNA reagentpET28a-
(Ripstein et al., 2017)Plasmid containing GFP-SsrA
Chemical compound, drugPKM-AMCGenScriptFluorogenic peptide for protease assays
Software, algorithmEPUThermo Fischer ScientificEM imaging software
Software, algorithmcryoSPARC v2(Punjani et al., 2017)RRID:SCR_016501EM reconstruction software
Software, algorithmUCSF Chimera(Pettersen et al., 2004)RRID:SCR_004097Molecular Visualization
Software, algorithmUCSF ChimeraX(Goddard et al., 2018)RRID:SCR_015872Molecular Visualization
Software, algorithmCoot(Emsley and Cowtan, 2004)RRID:SCR_014222Protein Model Building Software
Software, algorithmPhyre2(Kelley et al., 2015)RRID:SCR_010270Protein Model Building Software
Software, algorithmRosetta(Wang et al., 2015)RRID:SCR_015701Protein Model Building Software
Software, algorithmPhenix(Adams et al., 2010)RRID:SCR_014224Protein Model Building Software
Software, algorithmMolprobity(Arendall III et al., 2010)RRID:SCR_014226Protein Model Evaluation Software
Software, algorithmEMRinger(Barad et al., 2015)Protein Model Evaluation Software

Data availability

CryoEM maps and models have been deposited in the EMDB and PDB.

The following data sets were generated
    1. Ripstein ZA
    2. Vahidi S
    3. Houry WA
    4. Rubinstein JL
    5. Kay LE
    (2020) RCSB Protein Data Bank
    ID 6VFS. Maps NmClpXP Conformation A.
    1. Ripstein ZA
    2. Vahidi S
    3. Houry WA
    4. Rubinstein JL
    5. Kay LE
    (2020) RCSB Protein Data Bank
    ID 6VFX. Maps NmClpXP Conformation B.
    1. Ripstein ZA
    2. Vahidi S
    3. Houry WA
    4. Rubinstein JL
    5. Kay LE
    (2020) Electron Microscopy Data Bank
    ID EMD-21187. Maps NmClpXP Conformation A.
    1. Ripstein ZA
    2. Vahidi S
    3. Houry WA
    4. Rubinstein JL
    5. Kay LE
    (2020) Electron Microscopy Data Bank
    ID EMD-21194. Maps NmClpXP Conformation B.

Additional files

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)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)