If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported in part by a Canadian Institutes of Health Research grant. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ Contributed equally to this work. § Supported by a National Sciences and Engineering Council of Canada Postgraduate Scholarships A fellowship.
Clp ATPases are unique chaperones that promote protein unfolding and subsequent degradation by proteases. The mechanism by which this occurs is poorly understood. Here we demonstrate that the N-terminal domain of ClpX is a C4-type zinc binding domain (ZBD) involved in substrate recognition. ZBD forms a very stable dimer that is essential for promoting the degradation of some typical ClpXP substrates such as λO and MuA but not GFP-SsrA. Furthermore, experiments indicate that ZBD contains a primary binding site for the λO substrate and for the cofactor SspB. Removal of ZBD from the ClpX sequence renders the ATPase activity of ClpX largely insensitive to the presence of ClpP, substrates, or the SspB cofactor. All these results indicate that ZBD plays an important role in the ClpX mechanism of function and that ATP binding and/or hydrolysis drives a conformational change in ClpX involving ZBD.
Clp ATPases are a unique group of ATP-dependent chaperones associated with the disassembly of protein complexes (
). They are members of the AAA family of proteins, which are ATPases associated with a variety of cellular activities. The key feature of this family is a highly conserved AAA module of about 230 amino acids that is present in one or more copies in each protein. Each module can typically be divided into two subdomains: one αβ and one all α (
). The unifying structural feature of these AAA proteins is the arrangement of the subunits into ring-shaped hexameric or heptameric complexes.
The main Clp ATPases in the Escherichia coli cytoplasm are ClpA, ClpB, ClpX, and HslU. They function to unfold proteins or to disaggregate protein aggregates. They all have N-terminal domains followed by AAA modules, except for HslU, which has an insertion in the AAA module that is thought to correspond to the N-terminal domains in the other Clps (
). ClpA substrates include RepA, HemA, N-end rule proteins, and ClpA itself. Known ClpX substrates include λO, MuA, Mu repressor, UmuD′, σS, and Phd. Some of these proteins, such as λO and RepA, are recognized by the chaperones mainly through their N termini, whereas others, like MuA and Mu repressor, seem to be recognized mainly through their C termini. More recently, a proteomics approach revealed more than 50 potential ClpXP substrates in E. coli (
). The importance of the N-terminal domain of Clp ATPases in these processes has been rather enigmatic. The N terminus of ClpB from Thermus thermophilus has been shown not to be essential for chaperone activity in vitro (
The importance of the N-terminal domain of ClpX is evident from its conservation across all sequenced genomes. The domain is present in almost all known clpX sequences (Table I). Here we demonstrate that the N-terminal domain of ClpX is a C4-type zinc binding domain (ZBD)
that forms a very stable constitutive dimer, which is necessary for binding of λO and SspB to ClpX. In addition, we show that, unlike the full-length chaperone, the presence of ClpP, substrates, or cofactors does not significantly affect the ATPase activity of ClpX lacking the zinc binding domain.
Table IConservation of ClpP and ClpX across different genomes
Subcloning and Mutagenesis—All proteins were expressed from the pProEX HTb vector (Invitrogen) except for MBP-ZBD[1–60], which was expressed from pMAL C2 vector (New England Biolabs). Point mutations were introduced using the QuikChange system (Stratagene) according to manufacturer's protocol. All constructs were verified by DNA sequencing.
pET9aClpP overexpressing untagged ClpP was a gift from Dr. John Flanagan (Brookhaven National Laboratory). pRLM266 overexpressing λO protein was a gift from Dr. Roger McMacken (Johns Hopkins University). pMK951 overexpressing MuA protein was a gift from Dr. George Chaconas (University of Western Ontario).
Protein Purification and Peptide Synthesis—All tagged proteins were purified according to the manufacturer's protocols, and, if required, the tag was removed using tobacco etch virus protease. ClpP (
) were purified according to published protocols. Protein concentrations were determined using either the Bradford assay or the 5,5′-dithiobis(nitrobenzoic acid) assay to measure cysteine content. The SsrA peptide was synthesized using Fmoc (N-(9-fluorenyl)methoxycarbonyl) solid phase chemistry.
Analysis of ZBD Zinc Content—Zinc content was analyzed either by inductively coupled plasma atomic emission spectroscopy (ICP-AES), using a PerkinElmer Life Sciences Optima 3000 spectrometer or by a colorimetric assay. For ICP-AES, protein samples were dialyzed against two changes of Buffer A (50 mm Hepes, pH 7.8, and 150 mm NaCl) containing 10–20 mg/ml Chelex resin (Bio-Rad). The emission of 19 metals was monitored. In the colorimetric assay, ZBD was diluted to 5 μm in Buffer B (25 mm TrisHCl, pH 8, and 150 mm NaCl) containing 100 μm 4-(2-pyridylazo)resorcinol (PAR) at room temperature and then p-hydroxymercuribenzoate (PMB) was titrated into the solution. A sample was removed after each addition to monitor formation of PAR2·Zn+2 at 500 nm. The average of the Zn+2 concentration, determined from a standard curve, released after four molar equivalents of PMB was divided by the protein concentration in the assay to determine the moles of Zn+2 metal bound per mole of protein monomer.
Mu Phage Assay—Mu phage particles were isolated and purified from HM8305 F' Mucts62 E. coli strain, according to published protocols (
). The strain was a gift from Dr. George Chaconas (University of Western Ontario). Mu phage plaque formation at 37 °C in MC4100 clpX::KanR strain transformed with different ClpX mutants was measured. The strain was a gift from Ariane Toussaint (Université Libre de Bruxelles). All experiments were carried out in triplicate.
Measurement of GFP-SsrA Expression and Stability in Vivo—E. coli strains MC4100 and MC4100 clpX::KanR were transformed with pGFP or pGFP-SsrA (ChloramphenicolR) (
), and the ΔclpX strains were subsequently transformed with pProEX HTb (AmpR) plasmid encoding clpX or clpXΔZBD genes. Cultures were grown at 37 °C in LB medium with appropriate antibiotic selection until A600 of ∼0.4, and the expression of GFP and GFP-SsrA was induced using 0.2% arabinose. Samples were taken at 20, 40, and 60 min post-induction and diluted to an A600 of ∼0.3 using Buffer C (25 mm TrisHCl, pH 7.5, and 150 mm NaCl). Fluorescence was measured by setting excitation at 395 nm and monitoring emission from 410–600 nm. The emission was normalized to the peak at 456 nm, which is dependent on the density of the cells.
Coexpression of H6-ZBD[1–60] and ClpX—The plasmids pET28b and pET3a expressing H6-ZBD and ClpX, respectively, were transformed into BL21(DE3) cells, and the coexpression of the two proteins was induced with isopropyl-1-thio-β-d-galactopyranoside. ClpX was also expressed alone as a control. Cells from 50 ml of culture were lysed by sonication in Buffer D (25 mm TrisHCl, pH 7.5, 500 mm NaCl, 10% glycerol, 5 mm β-mercaptoethanol, and 10 mm imidazole) with 1 mg/ml lysozyme after 20 min of incubation on ice. Purification was carried out according to the manufacturer's protocol. Bound proteins were eluted with Buffer D plus 250 mm imidazole. Proteins were then separated on SDS-PAGE.
Cross-linking of ZBD[1–60]—Cross-linking was carried out in Buffer E (25 mm Hepes, pH 7.9, 150 mm NaCl, 100 μm ZnCl2, and 1 mm dithiothreitol) using glutaraldehyde at a final concentration of 0.05% and a final protein concentration of 1 μm. Samples were removed at different time points, and the reaction was quenched by adding TrisHCl, pH 8, to a final concentration of 0.5 m.
Sedimentation Equilibrium Analysis of ZBD[1–60]—Sedimentation equilibrium experiments and analysis were performed at the Analytical Centrifugation Facility at the Ontario Cancer Institute (Toronto). ZBD[1–60] was exchanged into Buffer F [25 mm Hepes, pH 7.9, 150 mm NaCl, and 2 mm tri(2-carboxyethyl)phosphine] and diluted to concentrations of 100, 166, and 233 μm. Experiments were performed at four different speeds at 20 °C.
Circular Dichroism of ZBD—CD measurements were obtained using an AVIV Circular Dichroism spectrometer Model 62ADS or a Jasco J-810. Proteins were diluted into Buffer G (25 mm TrisHCl, pH 8, 150 mm NaCl, 100 μm ZnCl2, and 1 mm dithiothreitol).
Fitting the Denaturation Curves—To analyze the data from thermal and chemical denaturation experiments, a model where the native ZBD dimer (N2) unfolds directly to two unfolded monomers (U) was used. Analysis was carried out according to Bowie and Sauer (
); however, a global fit to the Gibbs-Helmholtz equation, shown in Equation 1, was done to obtain Tm, ΔH(Tm), and ΔCp, where Tm is the temperature at which KU = [U]2/[N2] = 1.
Degradation Assays—Assays were typically carried out by preincubating a degradation mixture consisting of 1.5 units of creatine kinase, 16 mm creatine phosphate, 3 mm ATP, 4 μm substrate protein, 1.2 μm ClpP in Buffer H (25 mm Hepes, 5 mm MgCl2, 5 mm KCl, 0.03% Tween 20, and 10% glycerol) at 37 °C for 3 min. Subsequently, 1 μm ClpX or ClpX mutant was added to the reaction, and the incubation was continued at 37 °C. All concentrations are those of monomers. Samples were taken at the indicated time points, and the degradation was stopped by addition of 4× Laemmli buffer and boiling. Proteins were then separated on SDS-PAGE, and the intensities of the protein bands (IPB) were quantified using ImageQuant 5.0. The initial rate (V0) was determined from the slope of the plot of ln(IPB) versus time for the first 5 min of the reaction. Degradation of GFP-SsrA was also monitored by fluorescence using a Fluorolog spectrofluorometer (Jobin Yvon) with the excitation wavelength set at 395 nm and the emission wavelength set at 509 nm. For the GFP-SsrA competition assays, SsrA peptide, λO, MuA, or SspB were preincubated with the degradation mixture prior to the addition of ClpX.
ELISA Assay—The wells of a 96-well plate were coated with λO, MuA, GFP-SsrA, or BSA (100 μl, 20 μg/ml) in 20 mm Na2CO3, pH 9, for 1 h at 37 °C. The wells were then washed with Buffer I (25 mm TrisHCl, pH 7.5, 150 mm NaCl, and 0.1% Tween 20) and incubated with 200 μl of 5% milk protein in Buffer I for an additional hour at 37 °C. The wells were then washed with Buffer J (25 mm Hepes, pH 7.5, 150 mm KCl, 25 mm NaCl, 10 mm MgCl2, 2.5% glycerol, 0.1 mm EDTA, and 0.1% Tween 20) plus 1 mm dithiothreitol and incubated with increasing amounts of ZBD[1–60] for 1 h (total volume 100 μl). This step and all subsequent steps were performed at room temperature. Wells were washed with Buffer J (200 μl, three times), incubated with 100 μl of a 1:3000 dilution of anti-ClpX serum for 1 h, and washed and then horseradish peroxidase-conjugated Protein A (1:10,000 dilution, 100 μl) in Buffer J was added to the wells and incubated for 1 h. The wells were washed, and 100 μl of 3,3′,5,5′-tetramethylbenzidine liquid substrate (T0440; Sigma) was added. The reaction was allowed to proceed for 30 min, and the absorbance was measured at 650 nm using a SPECTRAmax 340PC plate reader (Molecular Devices).
ATPase Activity Using NADH—ATPase activity of ClpX was measured using a coupled assay (
). Assays were typically carried out by preincubating 1.0 μm ClpX or ClpX mutant, 0.2 mm NADH, 3.0 mm phosphoenolpyruvate, 4.7 units of pyruvate kinase, and 7.4 units of lactate dehydrogenase in Buffer H at 37 °C for 3 min. 1.2 μm ClpP, 4 μm GFP-SsrA, 4 μm λO, or 0.33 μm SspB were included as indicated. Subsequently, 5 mm ATP was added to the reaction, and the change in absorbance at 340 nm was measured for 20 min at 37 °C. The rate of ADP formation was calculated assuming a 1:1 correspondence between NADH oxidation and ATP regeneration and an extinction coefficient of 6230 m–1 cm–1 (
The N Terminus of ClpX Is a Dimer Containing a C4-type Zinc Binding Motif—Purified ClpX (10 μg in 140 μl) was subjected to limited proteolysis using sub-stoichiometric amounts of trypsin. The resulting peptides were analyzed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry. At very low trypsin quantities (0.1 ng), a 7-kDa peptide that mapped to the N-terminal 60 amino acids of ClpX was detected (numbering according to Swiss-Prot P33138). This region appeared to be very stable and was not degraded with higher protease quantities. Our results are generally consistent with earlier observations made by Singh et al. (
). Subsequently, the N terminus of ClpX was cloned and purified, and its ability to bind Zn+2 was characterized using ICP-AES and a colorimetric assay (Fig. 1A). Both techniques clearly demonstrated that the N terminus of ClpX spanning residues 1–60 bound Zn+2 in a ratio of one Zn+2 ion per protein monomer. We call this domain ZBD[1–60] for zinc binding domain of ClpX spanning residues 1 to 60. Only the first four cysteines (Cys-14, Cys-17, Cys-36, and Cys-39) are involved in chelating the Zn+2, because mutants lacking the fifth cysteine (Cys-43) were still able to bind Zn+2 in one to one molar ratio, whereas the C36A mutant resulted in the formation of a destabilized domain that did not contain Zn+2 (Fig. 1A). These results, as well as previous observations (
), clearly suggest that the highly conserved CX2CX18–40CX2C motif in ZBD of ClpX is a C4-type zinc finger.
ZBD[1–60] migrated as a 28-kDa protein on a gel filtration column (not shown). This is much larger than a monomer, because the predicted molecular mass of ZBD[1–60] is 7.046 kDa. Consequently, the oligomeric state of ZBD[1–60] was further investigated using several techniques. In an initial approach, chemical cross-linking was performed using glutaraldehyde. It was found that a cross-linked dimer could be detected in as little as 15 s, and higher oligomers were not detected even after 30 min (Fig. 1B). This oligomerization was judged to be specific at the protein concentration used (1 μm), because denaturation of ZBD[1–60] either by chemical or heat treatment (Fig. 1B) yielded no detectable dimeric species even after 30 min. ZBD[1–60] was further characterized by sedimentation equilibrium analytical ultracentrifugation at higher concentrations (100, 166, and 233 μm). The data was best fit to a single species with an apparent molecular mass of 14.94 kDa (Fig. 1C). This value is very close to the expected molecular mass of a dimer. The dimerization of ZBD[1–60] is not mediated by Cys-43, because ZBD[1–41] also formed a dimer as determined by gel filtration (not shown). ZBD[1–53], which has the domain boundaries identified by Singh et al. (
Furthermore, ZBD was able to force the dimerization of known monomers. Two fusion proteins, ZBD[1–60]-GFP (N-terminal fusion with green fluorescent protein) and MBP-ZBD[1–60] (C-terminal fusion with maltose binding protein), migrated as dimers when examined by size exclusion chromatography (Fig. 1D). Both GFP and MBP migrated as monomers. These experiments, together with the cross-linking and ultracentrifugation data, clearly demonstrate that the N-terminal domain of ClpX is not a monomer as published previously (
To demonstrate that this domain exists as a dimer in the context of the full-length chaperone, ClpX and His-tagged ZBD[1–60] were coexpressed and purified on a Ni-NTA column. ClpX co-purified on Ni-NTA only when coexpressed with His-tagged ZBD, whereas it did not bind to the resin nonspecifically in the absence of H6-ZBD[1–60] (Fig. 2A). However, when purified, untagged ZBD[1–60], which is already dimeric, was mixed with purified His-tagged ClpX, which forms mixed oligomers, and then passed over the Ni-NTA column, no interaction between the two proteins was observed (Fig. 2B). H6-ClpX bound specifically to the column, whereas untagged ZBD[1–60] did not bind to the column. Similarly, Singh et al. (
) had shown that the N-terminal domain of ClpX does not interact with the rest of the chaperone as determined by gel filtration. Hence, in the coexpression experiment, untagged ClpX must be interacting with His-tagged ZBD[1–60] through the dimerization of newly synthesized H6-ZBD[1–60] with the N-terminal domains of newly synthesized ClpX. This surprising observation of the dimerization of the ClpX N terminus leads us to propose that the hexameric ClpX could possibly function as a trimer-of-dimers.
Stability of the ZBD Dimer—CD spectra of ZBD[1–60] were obtained for the native, denatured, and renatured protein (Fig. 3A). The spectrum of the renatured protein overlays perfectly with that of the native protein, indicating that thermal denaturation is completely reversible. Thermally and chemically induced unfolding of ZBD[1–60], as monitored by CD at 220 nm, was reversible and occurred in a cooperative fashion, with a single transition region (Fig. 3, B and C). The stability of ZBD[1–60] was decreased in the presence of EDTA (Fig. 3B) indicating that Zn+2 is a structural metal necessary for the stability of this small domain. Furthermore, the denaturation profile exhibited a concentration dependence as expected for a protein oligomer (compare 3 to 30 μm in Fig. 3B). This result further establishes ZBD[1–60] as a dimer. The single transition observed upon thermal and chemical denaturation, and the concentration dependence of the denaturation process, all indicate that the dimer is present in the unfolding transition region. The monomer is not populated to a great extent in that region, because the free energy of unfolding, ΔGU, in the transition region of the thermal denaturation is the same at the two different concentrations (Fig. 3D). Therefore, the denaturation data were fit to a two state model where the native ZBD[1–60] dimer (N2) unfolds directly to two unfolded monomers (U). Based on this model and the fit of the data in Fig. 3D to the Gibbs-Helmholtz equation (see “Experimental Procedures”), the thermodynamic parameters for ZBD[1–60] stability at 25 °C were found to be ΔGU = 9.9 Kcal mol–1, ΔHU = 16.5 Kcal mol–1, ΔSU = 22 cal mol–1 K–1, and ΔCp = 0.41 Kcal mol–1 K–1. These values demonstrate that ZBD[1–60] is a very stable dimer that can withstand thermal stress and is ideally suited to be a domain in a protein involved in heat stress response.
ZBD Is Essential for ClpX Function in Mu Phage Replication—Replicative transposition during the lytic cycle of bacteriophage Mu requires the function of the ClpX chaperone (
). In an E. coli strain lacking the chromosomal clpX gene, the last step in transposition, which is the remodeling of the MuA transposase complexes, is blocked, because MuA remains bound to DNA as a tetramer (
), and, hence, the formation of plaques is prevented. A clpX knockout strain was transformed with plasmids expressing wild-type and mutant versions of the ClpX protein, and the ability to form plaques was assessed (
). It was found that plaque formation in the wild-type strain was comparable with that of the knockout strain expressing wild-type ClpX from a plasmid (Fig. 4A). Interestingly, ClpX lacking ZBD[1–60] (ClpXΔZBD) could not support the lytic growth of Mu phage even at high concentrations of phage. The inability to complement the knockout phenotype was observed with ZBD[1–60] alone and with an ATPase deficient mutant, ClpX (K124Q); this mutation in the Walker A motif prevents ATP hydrolysis and renders the chaperone inactive. The expression levels and solubility of ClpX and the different mutants were similar in these cells (not shown).
The Role of ZBD in GFP-SsrA Degradation in Vivo—GFP-SsrA was expressed in wild-type cells, as well as in cells deleted of clpX gene expressing ClpX or ClpXΔZBD from a plasmid. After 20, 40, and 60 min of induction of GFP-SsrA, a fluorescence emission scan of the intact cells was obtained. In the clpX knockout cells, the peak because of GFP-SsrA fluorescence is evident with peak maximum at 508 nm (Fig. 4B, top right). This characteristic peak is absent in the wild-type strain or in the clpX knockout strains expressing either ClpX or ClpXΔZBD. In all four strains, untagged GFP was expressed to similar extents, and its levels were not reduced by the expression of ClpX (insets in Fig. 4B). Thus, ClpXΔZBD is active in degrading GFP-SsrA in vivo although it cannot function in disassembling MuA-DNA complexes. Hence, ZBD could be required for binding and/or unfolding of some substrates such as MuA but not GFP-SsrA. This possibility is investigated further.
ZBD Modulates Substrate Degradation by ClpXP—To further investigate the role of ZBD in ClpX function, the degradation reactions of GFP-SsrA, MuA, and λO were carried out at 37 °C in the presence of ClpP, ATP, an ATP regenerating system, and ClpX or ClpXΔZBD (Fig. 5A). At different time points, samples were withdrawn from the degradation assay, boiled in Laemmli buffer, and separated on SDS-PAGE. The degradation of GFP-SsrA was also monitored by fluorescence (Fig. 6). These experiments revealed several unexpected results.
The ClpP-dependent degradation of GFP-SsrA can be mediated by ClpX lacking ZBD as determined by quantifying the bands on the gel or by fluorescence (Fig. 5A, right panel and Fig. 6, E–H, 0 μmcurves). This is in agreement with our in vivo observations (Fig. 4B), as well as with the observations of Singh et al. (
), who used enzymatically truncated ClpX. This indicates that the primary binding site for the SsrA tag is in the AAA domain of ClpX and not in the ZBD. However, the degradation of GFP-SsrA by ClpXΔZBD/ClpP was 33% less efficient than by ClpXP; this could reflect that ZBD further enhances the unfoldase activity of ClpX because of the presence of a secondary binding site for GFP-SsrA on ZBD. It should be noted that none of the ClpX constructs were able to degrade GFP-SsrA(DD) or GFP-SsrA in the presence of ADP or absence of ATP (data not shown).
On the other hand, ZBD seems to be absolutely essential for the binding and/or unfolding of λO, because ClpXΔZBD/ClpP was not able to degrade λO (Fig. 5A, right panel). This has been observed before with ClpX in which the first four cysteines involved in chelating Zn+2 were mutated to serines (
). Furthermore, ClpXΔZBD was six times less efficient than ClpX in mediating the degradation of MuA (Fig. 5A). This result is consistent with the inability of ClpXΔZBD to promote Mu phage replication (Fig. 4A). Hence, λO and MuA seem to represent a class of ClpX substrates that, unlike GFP-SsrA, requires the full-length chaperone to be efficiently degraded.
To demonstrate that ZBD is directly involved in substrate binding, ELISA analysis was performed (Fig. 5B). GFP-SsrA, MuA, and λO, as well as the nonspecific protein BSA, were bound to the ELISA wells. ZBD[1–60] was added to the wells at increasing concentrations. The amount of bound ZBD[1–60] was detected by first incubating with rabbit anti-ClpX antibodies, followed by incubating with Protein A conjugated to horseradish peroxidase. As can be seen in Fig. 5B, ZBD[1–60] bound with similar affinity to λO and MuA but did not interact significantly with GFP-SsrA or BSA. Thus, ClpX contains at least two different binding sites; one is present in the ZBD and is required for binding λO and MuA, and one is within the AAA domain and is essential for binding GFP-SsrA.
To further characterize the substrate binding sites in ClpX, the degradation of GFP-SsrA was performed in the presence of different concentrations of λO, MuA, SsrA peptide, or the cofactor SspB. Each competition assay included ClpP, ATP, an ATP regenerating system, ClpX, or ClpXΔZBD, and the degradation of GFP-SsrA was monitored by fluorescence (Fig. 6). λO, MuA, and the SsrA peptide competed in a concentration-dependent manner for the ClpX-mediated degradation of GFP-SsrA (Fig. 6, A–C), whereas SspB enhanced the degradation of GFP-SsrA as expected (Fig. 6D) (
). However, for ClpXΔZBD, only the SsrA peptide was able to significantly compete for GFP-SsrA degradation (Fig. 6G), further proving that the ZBD is not required for SsrA binding or recognition and that the binding of the SsrA tag occurs in the AAA domain of ClpX. On the other hand, the presence of 50 μm λO, which is at 12.5 times excess over GFP-SsrA, was insufficient to slow down the GFP-SsrA degradation by ClpXΔZBD (Fig. 6E). Similar curves were obtained if λO was added 2 min after the start of the reaction (not shown). Hence, ZBD seems to be essential for binding λO. It should be noted that λO at 2 μm was able to slow the degradation by ClpXP of 4 μm GFP-SsrA (Fig. 6A) indicating that the full-length chaperone has higher affinity for λO than GFP-SsrA. MuA was also able to compete with GFP-SsrA but much more weakly in the case of ClpXΔZBD as compared with ClpX (Fig. 6, F versus B). This suggests that ZBD in ClpX significantly increases the binding affinity of MuA to the chaperone.
The presence of as little as 0.05 μm of the cofactor SspB increased GFP-SsrA degradation by ClpXP (1 μm), and the degradation rate of GFP-SsrA was saturated between 0.33 and 10 μm SspB (Fig. 6D). On the other hand, SspB slowed down, rather than enhanced, the degradation of GFP-SsrA by ClpXΔZBD/ClpP (Fig. 6H). Hence, the ZBD is essential for the enhancing activity of SspB, which is probably because of the direct binding between SspB and ZBD. For the case of ClpXΔZBD, SspB would directly bind the SsrA tag of GFP-SsrA (
) but would not be able to transfer the substrate to the chaperone because of the absence of ZBD. Therefore, the presence of SspB would protect GFP-SsrA from degradation by ClpXΔZBD/ClpP as observed (Fig. 6H). This effect is not seen in vivo in Fig. 4B, probably because the levels of SspB are much lower than those of overexpressed ClpXΔZBD. Because both SspB and λO require the ZBD for binding to ClpX, it is likely that both proteins bind at the same site(s) or at very close site(s) on ZBD considering the small size of this domain.
ZBD Modulates the ATPase Activity of ClpX—The ATPase activity of ClpX can be stimulated by substrates (
). The addition of ClpP in the absence of substrates or cofactor lowered the ATPase activity to 218 pmol min–1 μg–1. In the absence of ClpP, the ATPase activity of ClpX increased upon addition of 4 μm λO, GFP-SsrA, or GFP-SsrA with 0.33 μm SspB to 944, 574, and 1160 pmol min–1 μg–1, respectively (Fig. 7). The presence of ClpP lowered those ATPase levels to 835, 328, and 932 pmol min–1 μg–1, respectively (Fig. 7). Because the Kd for GFP-SsrA binding to ClpX hexamer is estimated to be around 3 μm (
), and because λO binds ClpX tighter than GFP-SsrA (see above), the concentrations of λO and GFP-SsrA were increased to 20 μm (5x) to saturate the binding between the substrates and ClpX in the presence of ClpP. The ATPase activity of ClpX did not change significantly in the presence of increased λO concentration (759 pmol min–1 μg–1) possibly indicating that the lower λO concentration was already saturating; however, in the presence of increased GFP-SsrA, the activity increased back to levels similar to those in the absence of ClpP (568 pmol min–1 μg–1).
The ATPase activity of ClpXΔZBD alone was 209 pmol min–1 μg–1 (Fig. 7), which is only slightly lower than that of ClpX. However, the presence of λO, GFP-SsrA, or GFP-SsrA with SspB did not significantly change the ATPase activity of ClpXΔZBD, and the presence of ClpP did not inhibit the ATPase activity. The presence of five times the substrate concentration also did not significantly affect the ATPase activity of ClpXΔZBD. Based on these results, ZBD seems to be required for the stimulation or inhibition of the ClpX ATPase activity. This effect is probably mediated by a conformational change in ClpX involving ZBD.
It is interesting to note that the ATPase activity of ClpXΔZBD and ClpX in the presence of GFP-SsrA and ClpP correlate with the rate of degradation of GFP-SsrA. GFP-SsrA is degraded about one-third slower by ClpXΔZBD/ClpP than by ClpXP (1.8 × 10–3versus 2.7 × 10–3 s–1; see Fig. 5A), and the ATPase activity of ClpXΔZBD is about one-third lower than that of ClpX under the same conditions (237 versus 323 pmol min–1 μg–1; see Fig. 7). However, this correlation does not extend to other substrates such as λO. This could reflect the difference in binding to ClpX by GFP-SsrA and λO.
We have demonstrated that the N terminus of ClpX is a constitutive dimer that contains a C4-type zinc binding region. Cross-linking, ultracentrifugation, gel filtration, and thermal melts all support the conclusion that ZBD is a very stable dimer. The ZBD dimer is reminiscent of the pseudo-2-fold symmetry observed in the ClpA and ClpB N termini (
), which can be considered a pseudo-dimer. The N-terminal domain of ClpA has also been observed to contain a Zn+2 binding site in its first half consisting of two histidines, a glutamine, and a water molecule; however, there is only one Zn+2 bound per pseudo-dimer (
Pull-down experiments of coexpressed ClpX and ZBD suggest that ZBD is also a dimer in the context of the full-length ClpX chaperone. Hence, ClpX possibly functions as a trimer-of-dimers. On the other hand, the N termini of ClpA and ClpB have been shown to form monomeric domains (
). Because the pseudo-dimer domains of ClpA and ClpB are ∼150 amino acids, whereas the ZBD monomer of ClpX is less than half that size (60 amino acids), the protein density above the AAA hexamer of ClpA or ClpB is twice as much as that of ClpX. This could indicate that the N-terminal domain movements in ClpA and ClpB are different from those in ClpX; consequently, the ATP-dependent conformational changes that occur in ClpX might be different from those that occur in ClpA or ClpB implying a different mechanism of function.
In vitro and in vivo experiments indicate that ZBD is required for ClpX chaperone function for some substrates such as λO and MuA but not GFP-SsrA (Figs. 4, 5, 6). λO is a highly positively charged protein, with arginines or lysines constituting 16% of all its residues, resulting in a pI of 9.4. MuA has a similar character; 14.7% of its residues are lysines and arginines, and it has a pI of 8.9. On the other hand, SsrA (AANDENYALAA) is a very hydrophobic tag with a pI of 3.7. In E. coli proteins, the average R+K content is about 10% (refer to EBI Proteome analysis). Hence, λO and MuA may be characteristic of a typical substrate for which ClpX requires the function of the ZBD.
Substrates, cofactors, and interacting partners modulate the ATPase activity of ClpX (Fig. 7). Removal of the ZBD essentially abolishes this modulation. It is reasonable to speculate that the ATP-dependent movement of the ZBD causes the stimulation in the ATPase activity of ClpX. This could be either because of the direct binding of substrates and cofactors to ZBD such as that observed for λO and SspB or as a result of as yet unknown secondary effect because of the binding on the AAA domain, as for GFP-SsrA. The ATPase activity of ClpA and ClpB is also affected by the removal of their N-terminal domains (
). Thus, it appears that the N-terminal domains of the Clp chaperones play an essential role in regulating ATP hydrolysis in the presence of substrates and cofactors, reflecting an ATP-dependent conformational change involving these domains. Mapping out those conformational changes will be critical to understanding the unfoldase activity of this group of proteins.
We thank Brian Cox for help with mass spectrometry, Alan Davidson for the use of his CD instrument, Charlie Deber for use of his peptide synthesizer, Deborah Zamble for help in carrying out the ICP-AES experiments, George Chaconas for help in the Mu phage assay, Ipein Chua, Tanya Cordeiro, Khadija Mahmood, and Jamie Snider for help in cloning some of the constructs, and other members of the Houry laboratory for critical discussions.