ClpL is an autonomous disaggregase.

(A) Domain organizations of ClpL, ClpG, ClpB. All AAA+ proteins consist of two AAA domains (AAA1, AAA2), a coiled-coil middle-domain (M) and diverse N-terminal domains (N). ClpG additionally harbors a disordered C-terminal extension (CTE). (B) Luciferase disaggregation activity (% refolding of aggregated Luciferase/min) of Lm ClpL was determined (left). Relative Luciferase disaggregation activities of indicated disaggregation systems were determined (right). KJE: DnaK/DnaJ/GrpE. The disaggregation activity of Lm ClpL was set to 1. (C) GFP disaggregation activities (% refolding of aggregated GFP/min) of indicated disaggregation systems were determined. (D) Occurrence of ClpB and ClpL disaggregases in selected Gram-positive bacteria. Loss of heat resistance upon clpL deletion in bacteria harboring solely ClpL is indicated. n.d.: not determined. (E) Relative Luciferase and MDH disaggregation activities of L. gasseri (Lg) ClpL. Disaggregation activities of Lm ClpL were set to 1. Standard deviations are based on at least three independent experiments (B,C,E).

Specific molecular features separate ClpL from ClpB/DnaK.

(A) Incubation of Luciferase-YFP at 46°C only leads to unfolding of the Luciferase moiety and the formation of mixed aggregates including folded YFP. Unfolding of YFP during disaggregation of aggregated Luciferase-YFP reports on threading power. (B) Aggregated Luciferase-YFP was incubated in presence of indicated disaggregation machineries and YFP fluorescence was recorded. Initial YFP fluorescence was set at 100%. (C/D) Melting temperatures of ClpL, Lm ClpB and Lm DnaK were determined in presence of indicated nucleotides by SYPRO®Orange binding (C) or nanoDSF (D). Standard deviations are based on at least three independent experiments (C,D).

The ClpL NTD is essential for disaggregation activity.

(A) Disaggregation activities of ClpL and ClpL-ΔN towards aggregated Luciferase and MDH (each: % regain enzymatic activity/min) were determined. The activity of ClpL was set to 1. (B) ATPase activities of ClpL and ClpL-ΔN were determined. The ATPase activity of ClpL was set to 1. (C) E. coli ΔclpB cells harboring plasmids for constitutive expression of Luciferase and IPTG-controlled expression of indicated disaggregases were grown at 30°C to mid-logarithmic growth phase and heat shocked to 46°C for 15 min. Luciferase activities prior to heat shock were set to 100%. The regain of Luciferase activity was determined after a 120 min recovery period at 30°C. 10/100/500: μM IPTG added to induce disaggregase expression. p: empty vector control. (D) E. coli ΔclpB cells harboring plasmids for expression of indicated disaggregases were grown at 30°C to mid-logarithmic growth phase and shifted to 49°C. Serial dilutions of cells were prepared at the indicated time points, spotted on LB plates and incubated at 30°C. 10/100/500: μM IPTG added to induce disaggregase expression. p: empty vector control. (E) Domain organization of the ClpL-ClpB chimera LN-ClpB*. ClpB* harbors the K476C mutation in its M-domain, abrogating ATPase repression. (F) Luciferase disaggregation activities (% refolded Luciferase/min) of indicated disaggregation machineries were determined. The disaggregation activity of ClpL was set to 1. Standard deviations are based on at least three independent experiments (A,B,C,F).

Molecular basis of ClpL NTD binding to protein aggregates.

(A) AlphaFold2 model of ClpL NTD. The colour code depicts the calculated confidence of the prediction (pLDDT). Residues that potentially participate in the formation of small hydrophobic cores (α1/2 (cyan) and α2/β1,2 (magenta)) are indicated. (B) Secondary structure of ClpL-NTD as determined by NMR using secondary chemical shifts (Cα, Cβ). Secondary structure elements determined from NMR and from the AlphaFold prediction are indicated below the histogram. The predicted α1-helix only transiently forms in isolated solution context, which is confirmed by further NMR analysis (Fig. S5E). (C) Composition of ClpL NTDs. The frequencies (%) of individual amino acids represent the ratio of the number of a particular residue and the total length of respective NTDs (L. monocytogenes, Staphylococcus aureus, S. pneumoniae, Lactobacillus plantarum, Oenococcus oeni, Lactobacillus rhamnosus, Streptococcus suis). The average frequency of each amino acid in the total bacterial proteomes is given as reference (Bogatyreva et al, 2006). (D) Localization of patches A-E, consisting of aromatic and N/Q residues are indicated. (F/G) Luciferase and MDH disaggregation activities (% refolded enzyme/min) of ClpL WT and indicated patch mutants were determined. The disaggregtion activity of ClpL was set to 1. (G) ClpG, ClpL and indicated ClpL mutants were incubated with aggregated MDH in presence of ATPγS. The extend of aggregate binding was determined by co-sedimentation upon centrifugation. Standard deviations are based on at least three independent experiments (E,F,G).

Multiple ClpL NTDs are required for disaggregation activity.

(A) Varying the ratio of LN-ClpB* and ΔN-ClpB* leads to formation of mixed hexamers with diverse numbers of NTDs. (B) Luciferase disaggregation activities (% refolded Luciferase/min) of mixed LN-ClpB*/ΔN-ClpB* hexamers were determined and compared with curves calculated from a model (black to grey), which assumes that a mixed hexamer only displays disaggregation activity if it contains the number of NTDs indicated. Mixing ratios are indicated as number of ΔN-ClpB* in a hexamer. (C/D) Luciferase disaggregation activities of LN-ClpB* (C) and ClpL (D) were determined in absence and presence of an excess of isolated NTD as indicated. Disaggregation activities determined in NTD absence were set as 100%. Standard deviations are based on at least three independent experiments (C/D).

Stabilizing ClpL ring dimers strongly reduces disaggregation activity.

(A) AlphaFold2 model of Lm ClpL ring dimers. Positions of individual domains are indicated. (B) Negative stain EM of Lm ClpL. 2D class averages revealing single ring hexamers or heptamers, ring dimers and tetramers of rings are indicated. The scale bar is 100 nm. (C) Populations of diverse ClpL assembly states based on 2D class averages were determined for Lm ClpL WT and indicated mutants and Lg ClpL. (D) obtained for the WT and the F444S mutant. (D/E) Disaggregation activities of ClpL WT and indicated M-domain (M) mutants towards aggregated Luciferase (D) and MDH (E) (each: % regain enzymatic activity/min) were determined. The activity of ClpL was set to 1. (F) Model of ClpL ring dimers. Positions of T355 residues in interacting M-domains are depicted. Crosslinking of ClpL T355C was achieved by DTT removal and further incubation at RT as indicated. The formation of crosslinked ClpL T355C dimers was monitored by SDS-PAGE and Coomassie-staining. Addition of 10 mM DTT reversed the disulfide bonds. (G) Luciferase disaggregation activities of reduced ClpL WT and oxidized ClpL WT or ClpL T355C were determined in absence of DTT (-DTT). Oxidized variants (WT and T355C) were additionally preincubated with 10 mM DTT for 30 min and tested for disaggregation activity in presence of DTT (+DTT). The factor of increase in disaggregation activity upon reduction (+DTT) is indicated (right). Standard deviations are based on at least three independent experiments (D/E/G), standard deviations for the activity gain factors have been propagated from disaggregation activity standard deviations.

strains and plasmids used in this study

Sequence alignment of Escherichia coli ClpB, Staphylococcus aureus ClpC, Pseudomonas aeruginosa ClpGGI and Listeria monocytogenes ClpL. The domain organization is indicated. Similar and identical residues are highlighted in light and dark blue.

ClpL is a potent, stand-alone disaggregase. (A) Disaggregation of aggregated Luciferase was monitored by turbidity measurements in presence of indicated chaperones. KJE: DnaK/DnaJ/GrpE. Initial turbidity was set at 100%. (B) Relative Luciferase disaggregation activities (% turbidity/min) were determined for indicated chaperone combinations. The disaggregation activity of ClpL was set to 1. (C) α-Glucosidase disaggregation activities (% turbidity/min) were determined in presence of indicated chaperones. Initial turbidity of α-Glucosidase aggregates was set to 100%. Disaggregation activity of ClpL was set to 1. (D) Left: MDH disaggregation activity of ClpL (% refolded MDH/min). Right: MDH disaggregation activities of indicated chaperones were determined. The activity of ClpL was set to 1. (E) Disaggregation activities of indicated chaperone combinations towards aggregated Luciferase, MDH, GFP and α-Glucosidase were determined. The activity of ClpL was set to 1. n.d.: not determined. Standard deviations are based on at least three independent experiments (B-E).

ClpL disaggregation activity relies on ATP-fueled substrate threading. (A) Domain organization of ClpL. Pore loop (Y170, Y504) and Walker B (E197, E530) residues of AAA1 and AAA2 domains are indicated. (B) ClpL ATPase activities were determined at 0.125 and 1 μM protein concentrations. (C) ATPase activities of ClpL WT and indicated mutants were determined at 1 µM protein concentration. The ATPase activity of ClpL WT was set to 1. (D) Luciferase disaggregation activities (% refolded Luciferase/min) of ClpL WT and indicated mutants were determined. The activity of ClpL WT was set to 1. Standard deviations are based on at least three independent experiments (B-E).

The ClpL NTD is crucial for disaggregation activity. (A) E. coli dnaK103 cells harboring plasmids for IPTG-controlled expression of indicated disaggregases were grown at 30°C to mid-logarithmic growth phase and shifted to 49°C. Serial dilutions of cells were prepared at indicated time points, spotted on LB plates and incubated at 30°C. 100/500: μM IPTG added to induce disaggregase expression. p: empty vector control. (B-C) Production levels of disaggregases (ClpB, DnaK, ClpG, ClpL, ClpL-ΔN) in E. coli ΔclpB (B) and dnaK103 (C) mutant cells. Cells were were grown at 30°C for 1.5 h and disaggregase expression was induced by addition of indicated IPTG concentrations (μM) for 2 h. Total cell extracts were prepared and levels of disaggregases were determined SDS-PAGE followed by Coomassie-staining. ClpL levels were additionally determined by western blot analysis using ClpL-specific antibodies. p: empty vector control.

Structural analysis of ClpL NTD. (A) 1H,15N-HSQC of ClpL NTD with resonance assignment labeled at each peak. Several residues exhibit dispersed peak position, indicative of folding. (B) 1H,1H-2D NOESY in D2O of ClpL NTD confirming NOEs between aromatic ring protons of the hydrophobic core formed by α2 and the β-sheet. NOEs of the second predicted core between α1 and α2 are absent, indicating that there is no interaction between α1 and α2. (C) Selected strips of the 13C-edited 3D NOESY-HSQC to illustrate that the predicted hydrophobic core between α2 and the β-sheet is present. Multiple cross-peaks (NOEs) are present between relevant residues (Y36, V38, L43, F48, Y51). (D) Selected strips of the 13C-edited 3D NOESY-HSQC illustrating the absence of NOEs expected to between residues of the second hydrophobic core between α1 and α2. For these residues (F19, F23, A34, Y36) only intra-residue or sequential NOEs are visible (except for Y36, which has NOEs to residues of the first hydrophobic core). Thus, despite transient formation of the α1-helix, there is no or only transient interactions between α1 and α2. (E) 15N spin relaxation data of ClpL NTD. Top: Longitudinal relaxation rates, middle: transverse relaxation rates, bottom: Heteronuclear NOEs. All three measurements show that the N-terminal region up to residue R35 is highly flexible, whereas the region of the hydrophobic core between α2 and the β-sheet higher heteronuclear NOE values and longer transverse relaxation rates indicating a higher degree of order. The altogether low heteronuclear NOE values, rarely above 0.6 indicate that the secondary structure elements (α1 vs. α2/β-sheet) are mobile with respect to each other in the ps-ns time scale. (F) Sequence alignment of ClpL NTDs. Similar and identical residues are highlighted in light and dark blue. The positions of secondary structural elements and of patches (A-E) composed of Y, F, N and Q residues are indicated.

Mutant analysis of ClpL NTD. (A/D) ATPase activities of indicated ClpL mutants determined at 0.125 and 1 μM protein concentration. The ATPase activities of ClpL WT were set to 1. (B/C) Luciferase and MDH disaggregation activities (% refolded enzyme/min) of ClpL WT and indicated mutants were determined. The disaggregation activity of ClpL WT was set to 1. (E/F) Luciferase and MDH disaggregation activities (% refolded enzyme/min) of ClpL WT, LN-ClpB* and LN-ClpB* mutants were determined. The disaggregation activity of ClpL WT was set to 1. (G) ATPase activities of ClpL WT, LN-ClpB* and LN-ClpB* mutants were determined at 0.125 and 1 μM protein concentration. The disaggregation activity of ClpL WT was set to 1.

Multiple ClpL NTDs are required for disaggregation activity. (A) Luciferase disaggregation activities of indicated LN-ClpB* disaggregase mixtures were determined (DWB: ATPase deficient LN-ClpB*-E279A/E678A, ΔN: ΔN –ClpB*). Protein concentrations (μM) are indicated. The disaggregation activity of 0.3 μM LN-ClpB* was set to 1. (B) Luciferase disaggregation activities of LN-ClpB* were determined in absence and presence of ΔN-ClpB* excess as indicated. The disaggregation activity of LN-ClpB* was set to 1. (C) Luciferase disaggregation activities (% refolded Luciferase/min) were determined for mixtures of LN-ClpB* and ΔN-ClpB* and were set as 100% for non-mixed LN-ClpB* (0.3 μM). Mixing ratios are indicated as final concentration of LN-ClpB* in 0.3 μM mixed hexamers. As control disaggregation activities of lower LN-ClpB* concentrations were determined (+ buffer). (D) Luciferase disaggregation activities (% refolded Luciferase after 120 min) of mixed LN-ClpB*/ΔN-ClpB* hexamers were determined and compared with curves calculated from a model (black to grey), which assumes that a mixed hexamer only displays disaggregation activity if it contains the number of NTDs indicated. Mixing ratios are indicated as number of ΔN-ClpB* in a hexamer. (E) Luciferase disaggregation activities (% refolded Luciferase after 120 min) were determined for mixtures of LN-ClpB* and ΔN-ClpB* and were set as 100% for non-mixed LN-ClpB* (0.3 μM). Mixing ratios are indicated as final concentration of LN-ClpB* in 0.3 μM mixed hexamers. As control disaggregation activities of lower LN-ClpB* concentrations were determined (+ buffer).

ClpL rings interact in a M-domain dependent manner. (A) Oligomeric states of ClpL WT and indicated mutants were determined in presence of ATP by size exclusion chromatography. ATPase-deficient ClpB-E279A/E678A (DWB) served as reference for hexameric rings. Elution fractions were analyzed by SDS-PAGE. Positions of peak fractions of a protein standard are indicated. (B) Model of Lm ClpL highlighting the residues E352 and F354, which are crucial for M-domain mediated ring interactions. (C) Negative stain EM of indicated ClpL mutants. The insets show a magnification of a representative grid area. The scale bar is 100 nm. (D) Melting temperatures (TM) of ClpL WT and indicated deletion mutants were determined in presence of ATPγS by nanoDSF or SYPRO®Orange binding. (E) Populations of diverse ClpL assembly states were determined based on 2D class averages for indicated ClpL mutants. (F) ATPase activities of indicated ClpL M-domain mutants were determined at 0.125 and 1 μM protein concentration. The ATPase activities of ClpL WT were set to 1. Standard deviations are based on at least three independent experiments (D/F).

Stabilizing ClpL ring dimers by disulfide crosslinking. (A) ATPase activities of indicated ClpL WT and T355C were determined at 0.125 and 1 μM protein concentration. The ATPase activities of ClpL WT were set to 1. (B) Luciferase disaggregation activities (% refolded enzyme/min) of ClpL WT and T355C were determined. The disaggregation activity of ClpL WT was set to 1. (C) Negative stain EM of ClpL T355C in oxidized and reduced states in presence of ATP. The insets show a magnification of a representative grid area. The scale bar is 100 nm. (D) YFP unfolding rates (% YFP fluorescence/min) during disaggregation of aggregated Luciferase-YFP by ClpL WT and oxidized ClpL WT or ClpL T355C were determined in absence of DTT (-DTT). Oxidized variants (WT and T355C) were additionally preincubated with 10 mM DTT for 30 min and tested for YFP unfolding activity in presence of DTT (+DTT). The factor of increase in YFP unfolding activity upon reduction (+DTT) is indicated (right). Standard deviations are based on at least three independent experiments (A/B/D), standard deviations for the activity gain factors have been propagated from disaggregation activity standard deviations.