|Home | About | Journals | Submit | Contact Us | Français|
ClpL, a member of the HSP100 family, is widely distributed in Gram-positive bacteria but is absent in Gram-negative bacteria. Although ClpL is involved in various cellular processes, such as the stress tolerance response, long-term survival, virulence, and antibiotic resistance, the detailed molecular mechanisms are largely unclear. Here we report that ClpL acts as a chaperone to properly fold CtsR, a stress response repressor, and prevents it from forming protein aggregates in Streptococcus mutans. In vitro, ClpL was able to successfully refold urea-denatured CtsR but not aggregated proteins. We suggest that ClpL recognizes primarily soluble but denatured substrates and prevents the formation of large protein aggregates. We also found that in vivo, the C-terminal D2-small domain of ClpL is essential for the observed chaperone activity. Since ClpL widely contributes to various cellular functions, we speculate that ClpL chaperone activity is necessary to maintain cellular homeostasis.
Bacteria are constantly exposed to stressful environments, such as exposure to free radicals, acidic or alkaline conditions, or high temperatures. Exposure to environmental stresses often causes the denaturation of cellular proteins that subsequently accumulate within the cell as protein aggregates. To encounter the detrimental effects of thermal and other stresses, bacteria transiently synthesize a highly conserved set of proteins with molecular chaperone or protease activities that are generally referred to as heat shock proteins (HSPs) (1, 2). HSPs are ubiquitous in bacteria, and depending on their molecular weights, HSPs are grouped into four major classes: the HSP100, HSP70, HSP60, and small HSP families (1, 3). The HSP100 subgroup, which is also known as the caseinolytic protease system (Clp), is typically an AAA+ (ATPases associated with a variety of cellular activities) superfamily protein that often forms complexes with a peptidase subunit, such as ClpP, for the proteolytic activity required for the removal of damaged and denatured proteins as well as protein-folding functions (4, 5). Clp proteases also play important roles in regulating various cellular functions, such as controlling growth at low or high temperatures, competence development, sporulation, and virulence (6–9).
According to the number of ATP-binding domains on the polypeptide chain, regulatory ATPase subunits can be grouped into class I (two ATP-binding domains, AAA-1 and AAA-2) and class II (one ATP-binding domain, AAA-1) (10). Clp proteins have also been categorized into various classes based on the length of the spacer sequence at the middle region, overall sequence similarity, and variation in the N- and C-terminal regions (10). Clp ATPases that interact with ClpP peptidase, such as ClpA and ClpX, encode a conserved tripeptide IGF motif that is present at a surface loop (11, 12). These ATPases are assembled into hexameric rings with a narrow pore in the center and use the energy generated by ATP hydrolysis to unfold target protein substrates and translocate them into the chamber of the associated barrel-like ClpP protease complex, where the peptide bonds are cleaved (13–16).
There are a few Clp ATPases that do not interact with ClpP peptidase; instead, they sometimes cooperate with the HSP70 system to function as a chaperone that can disaggregate and refold denatured proteins (17). In Escherichia coli, ClpB is one such Clp ATPase that does not interact with ClpP. ClpB contains two ATP-binding domains (AAA-1 and AAA-2) that are separated by a middle domain (M domain), forming a coiled-coil structure (18, 19). Like other Clp ATPases, ClpB also forms a hexameric-ring structure and cooperates with the DnaK chaperone system that includes DnaK, DnaJ, and GrpE (17, 19). ClpB has a remarkable ability to rescue proteins from an aggregated state by aiding in the disaggregation of denatured proteins. The complete refolding of the denatured proteins is dependent on the concerted effort of the cognate DnaK-DnaJ-GrpE system (20–22). ClpB alone may also suppress the aggregation of labile proteins with its monomeric or dimeric form (23). Despite a wealth of biochemical and structural information, the mechanistic aspects of this bichaperone system remain poorly understood. This is in part because ClpB and DnaK systems interact only transiently (24, 25) and recognize substrate proteins via sequential binding (26). Recently, it was suggested that the ClpB M domain determines the specificity of ClpB and DnaK cooperation, which is required for protein disaggregation and thermotolerance (27, 28).
In many low-GC Gram-positive bacteria, in addition to ClpB, another protein, ClpL, with homology to ClpB is also present (6, 29). In Streptococcus pneumoniae, a respiratory pathogen, ClpL has been shown to be involved in thermotolerance, acid tolerance, virulence regulation, and resistance to various antibiotics that target the cell wall (29–32).
Streptococcus mutans, a primary etiological agent of dental caries, possesses five proteins that belong to the Clp family, ClpB, ClpC, ClpE, ClpL, and ClpX (33). CtsR, a major repressor for the Clp ATPases (34, 35), mainly regulates the expression of stress response genes by recognizing a tandemly repeated heptanucleotidic sequence known as the CtsR box (35). We recently demonstrated that CtsR is accumulated in larger amounts in cells that do not have a functional ClpL protein (36). This finding was interesting since in most Gram-positive bacteria, CtsR is degraded by ClpCP (37). In this study, we showed that in the absence of ClpL, the CtsR protein accumulated in S. mutans cells largely as an inactive aggregated form. A direct protein-protein interaction between ClpL and CtsR was confirmed both in vitro and in vivo. We also observed that the presence of ClpL largely enhanced the refolding of urea-denatured CtsR in vitro. In vivo complementation showed that ClpL lacking the D2-small domain was no longer able to alleviate the CtsR aggregates. These results indicate that ClpL is crucial for the proper folding of the CtsR protein in S. mutans, even at ambient growth temperatures.
The S. mutans strains and plasmids used in this study are listed in Table 1. Escherichia coli strains were routinely grown in Luria-Bertani medium supplemented (when necessary) with 100 μg/ml ampicillin and/or 50 μg/ml kanamycin. S. mutans isolates were normally grown at 37°C in Todd-Hewitt medium (BBL, BD) supplemented with 0.2% yeast extract (THY medium). When necessary, 5 μg/ml erythromycin and/or 400 μg/ml kanamycin was included in THY medium.
For S. mutans protein extraction, unless otherwise stated, cultures grown overnight were inoculated into THY medium and grown to the exponential phase (optical density at 600 nm [OD600] of 0.4). A 10-ml aliquot was harvested by centrifugation, resuspended in 600 μl of phosphate-buffered saline (PBS), and homogenized with a bead beater (MP Biomedicals, LLC). The cell lysate was centrifuged at 18,000 × g for 10 min, and ~200 μl of the supernatant was carefully transferred into a new tube and kept as the soluble fraction. The remaining supernatant was removed, and the cell debris (pellet) together with silicon beads were washed three times with PBS, resuspended in 200 μl of PBS, and stored as the insoluble fraction. Both the soluble and insoluble fractions were added to protein sample buffer, boiled for 5 min, and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The gels were stained by Coomassie blue R250 or blotted onto polyvinylidene difluoride (PVDF) membranes. Western blot assays were carried out by using standard techniques. An antipolyhistidine (anti-His) monoclonal antibody (Sigma) was used as the primary antibody to detect His-tagged proteins. For Western blot experiments using the cell lysates as samples, the abundance of cellular enolase was chosen as an internal control. PVDF membranes were stripped with Tris-buffered saline (TBS) containing 2% SDS and 100 mM β-mercaptoethanol and reprobed with an anti-S. mutans enolase (anti-SmuEno) polyclonal antibody (Genscript). Western blots were developed with Pierce ECL Plus reagents (Thermo Scientific), and the fluorescent signals were detected with a Typhoon FLA9000 biomolecular imager (GE Healthcare). All Western blot experiments were repeated as least twice to confirm the results.
Plasmid pIB521 containing the PclpP-gusA fusion was previously used to measure clpP promoter activity (38). Plasmid pIB521 was linearized with BglI and transferred into S. mutans IBSJ3 cells via natural transformation, using a protocol described previously (39), to generate strain IBSJ9. A β-glucuronidase assay was performed as previously described (36). Briefly, the OD600 of an exponential-phase culture was recorded before cell harvesting. One milliliter of culture was harvested, washed in saline, and resuspended in 500 μl of Z buffer (60 mM Na2HPO4, 40 mM Na2HPO4, 10 mM KCl, 1 mM MgSO4, and 20 mM dithiothreitol [DTT]). Cells were homogenized by bead beating. Two hundred microliters of cell lysate was then transferred into a new tube, 100 μl of p-nitrophenyl-β-d-glucoside (4 mg/ml in Z buffer) was added, and the mixture was incubated in 37°C until a yellow color developed. The reaction was stopped by the addition of 200 μl of 1 M Na2CO3 to the mixture. The absorbance at 420 nm (A420) and the time period of the reaction in minutes (T) were noted. Gus activity was defined as (1,000 × A420)/(T × OD600) in Miller units (MU).
The open reading frame (ORF) of ctsR was amplified from UA159 genomic DNA with primers ctsR-duet-F (5′-CTTCATATGACGTCAAAAAATACTTCAG-3′) and ctsR-duet-R (5′-CTTCTCGAGTCATAGATGGTATCCTTTTCTATC-3′) and restricted with NdeI-XhoI. Similarly, the ORF of clpL was amplified from UA159 genomic DNA with primers clpL-duet-F (5′-CTTGGATCCGATGGCAAATTTTAATGGACGCG-3′) and clpL-duet-R (5′-CTTCTGCAGTTAAGCTTCTTCAATAATCAATTTGTC-3′) and restricted with BamHI-PstI. The restricted ctsR and clpL fragments were cloned into vector pETduet to create pIBJ33 and pIBJ34, respectively. The restricted ctsR fragment was then cloned into NdeI-XhoI-digested pIBJ34 to obtain pIBJ35. Plasmid pIBJ35 was transformed into E. coli BL21(DE3) cells and allowed to coexpress two proteins induced with isopropyl-β-d-thiogalactoside (IPTG).
His-CtsR expression was induced with 200 μg/liter anhydrotetracycline in E. coli DH5α with pIBC37. E. coli BL21(DE3) with either pIBJ33 or pIBJ34 was used for the overexpression of His-ClpL or CtsR-Stag induced by 1 mM IPTG. His-tagged proteins were purified with nickel-nitrilotriacetic acid (Ni-NTA) resin (Novagen), in accordance with the manufacturer's instructions. The protein was dialyzed overnight against a buffer containing 20 mM Tris-Cl (pH 7.4), 100 mM NaCl, and 1 mM DTT. Soluble CtsR-Stag protein was extracted from purified inclusion bodies with PBS containing 8 M urea. The protein extract was then dialyzed against PBS with urea in gradient concentrations, and the final protein solution contained 2 M urea. The purity of the proteins was >95%, as determined by SDS-PAGE analysis. Protein concentrations were estimated by the absorbance at 280 nm using a Nanodrop 2000c spectrophotometer (Thermo Scientific).
A pulldown assay for His-tagged proteins was performed according to standard methods. Briefly, 50 μg of His-ClpL and/or 5 μg of CtsR-Stag was added to 1 ml of binding buffer (20 mM sodium phosphate, 150 mM NaCl, 4 mM ATP [pH 7.6]) together with 30 μl of prewashed, settled Ni-NTA resin. The mixture was incubated at 4°C with rotation for 30 min; the resin was washed twice with PBS containing 30 mM imidazole, followed by washing twice with PBS containing 0.01% Triton X-100. Bound proteins were eluted with elution buffer (50 mM sodium phosphate, 300 mM NaCl, 250 mM imidazole [pH 7.4]) and subjected to SDS-PAGE analysis.
CtsR-Stag dissolved in PBS containing 2 M urea was used for a refolding assay by dialysis at 4°C. Urea (2 M) was preadded to all additives to prevent the precipitation of CtsR-Stag before dialysis. To initiate renaturation, 400 μg/ml CtsR-Stag was dialyzed at 4°C against a dialysis buffer containing 20 mM Tris-Cl (pH 7.4), 50 mM NaCl, 1 mM DTT, 10% glycerol, and 4 mM ATP in the presence or absence of His-ClpL at various concentrations. The dialyzed samples were centrifuged at 18,000 × g for 10 min, and the supernatant fractions were analyzed by SDS-PAGE. The efficiency of renaturation was determined by measuring the intensity of protein bands with ImageQuantTL software on a Coomassie blue-stained gel. The electrophoretic mobility shift assay (EMSA) used to verify the DNA-binding activity of refolded CtsR-Stag was performed as previously described (36). To test the potential for the folding of CtsR aggregates by ClpL, insoluble CtsR-Stag aggregates were incubated with His-ClpL in refolding buffer (20 mM Tris-Cl, 50 mM NaCl, 1 mM DTT, 10% glycerol, 4 mM ATP [pH 7.4]) for 4 h at 37°C. The mixture was then centrifuged at 18,000 × g for 10 min, and the supernatant fraction was separated by SDS-PAGE and transferred onto a PVDF membrane. The detection of CtsR-Stag was performed by Western blotting using a monoclonal anti-Stag antibody (Novagen).
A DNA fragment encoding the clpL ORF but lacking the 117 N-terminal amino acid residues (clpLΔN) was amplified from UA159 genomic DNA by using primers ClpLΔN-184km-F (5′-CACGGATCCATGCCTGTTCTGGTCGGTGATG-3′) and ClpL-184km-R (5′-CACGGTACCACAGCTTCTTCAATAATCAATTTGTC-3′). Similarly, a DNA fragment encoding the clpL ORF but lacking the 78 C-terminal amino acid residues (clpLΔC) was also amplified from UA159 genomic DNA by using primers ClpL-184km-F (5′-CACGGATCCATGGCAAATTTTAATGGACGC-3′) and ClpLΔC-184km-R (5′-CACGGTACCACGTGAGAGAATTCAATAACTGC-3′). The amplified fragments were cloned into vector pIB184Km (40), which contains the P23 promoter, to create pIBJ63 and pIBJ64, respectively. Strain IBSJ3/pIBJ1 was transformed with either pIBJ63 or pIBJ64, and the transformants were selected on THY agar plates containing both erythromycin and kanamycin. The presence of the complementary genes and his-ctsR on the selected transformants was confirmed by PCR.
Two compatible plasmids, one expressing the T18 fusion (pUT18; Euromedex) and the other expressing the T25 fusion (pKNT25; Euromedex), were chosen for our bacterial adenylate cyclase two-hybrid (BACTH) assays. DNA fragments carrying full-length clpL, clpL lacking the 117 N-terminal amino acids (ClpLΔN), or clpL lacking the 78 C-terminal amino acids (ClpLΔN) were amplified from UA159 genomic DNA and cloned into vector pUT18 to create pIBJ58, pIBJ59, and pIBJ60, respectively. Meanwhile, the DNA fragments carrying ctsR and clpL were amplified and cloned into vector pKNT25 to create pIBJ55 and pIBJ62, respectively. The following primers were used for the above-described gene amplifications: ClpL-T18-F (5′-CACGGATCCAATGGCAAATTTTAATGGACGC-3′), ClpL-T18-R (5′-CACGGTACCACAGCTTCTTCAATAATCAATTTGTC-3′), ClpLΔN-T18-F (5′-CACGGATCCACCTGTTCTGGTCGGTGATG-3′), ClpLΔC-T18-R (5′-CACGGTACCACGTGAGAGAATTCAATAACTGC-3′), CtsR-T25-F (5′-CACGGATCCAATGACGTCAAAAAATACTTCAG-3′), and CtsR-T25-R (5′-CACGGTACCACTAGATGGTATCCTTTTCTATC-3′).
Two compatible plasmids, a pUT18 derivative and a pKNT25 derivative, were cotransformed into E. coli indicator strain BTH101 by electroporation and screened on LB agar plates containing both ampicillin and kanamycin. The presence of the target genes in the selected transformants was verified by PCR. The confirmed bacterial cells were streaked onto LB agar plates containing antibiotics, IPTG (0.5 mM), and 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) (40 μg/ml) and grown at 30°C. If there is an interaction between the two proteins of interest, the colonies turn blue within 24 to 72 h, according to the manufacturer's instructions. Quantification of the functional complementation mediated by the interaction of two proteins of interest was achieved by measuring β-galactosidase (β-gal) activity.
E. coli BTH101 cells with plasmids of interest were inoculated into liquid LB medium containing ampicillin, kanamycin, and IPTG (0.5 mM). The culture was grown overnight at 37°C to reach the stationary phase. The OD600 of the culture grown overnight was recorded before harvesting. One milliliter of the culture grown overnight was centrifuged, and cell pellets were washed twice with PBS and resuspended in an equal volume of Z buffer. One hundred microliters of resuspended bacterial cells was diluted in 1 ml of Z buffer (dilution factor [DF] = 10). Afterwards, 100 μl of chloroform and 50 μl of 0.1% SDS were added and mixed well to permeabilize the cells. Two hundred fifty microliters of the mixture was then transferred into a new Microfuge tube and brought to 28°C, 50 μl of prewarmed o-nitrophenyl-β-galactoside (4 mg/ml in Z buffer) was added, and the mixture was incubated at 28°C until a yellow color developed. The reaction was stopped by the addition of 200 μl of 1 M Na2CO3. The A420 and the precise time period of the reaction in minutes (T) were recorded. β-Galactosidase activity was defined as (1,000 × A420 × DF)/(T × OD600) in Miller units.
Our previous study showed that the CtsR protein accumulated in large amounts in a clpL-deficient S. mutans strain (36). Further analyses suggested that the accumulated CtsR (His-CtsR) is present predominantly in the pellet fraction and not in the soluble fraction of the cell lysate from the ΔclpL strain (IBSJ3/pIBJ1). On the other hand, very little His-CtsR protein was found in the insoluble fraction in the wild-type background (i.e., UA159/pIBJ1) (Fig. 1A). To determine if the accumulation was specific to CtsR, we used HrcA, a transcriptional repressor also involved in the heat shock response, as a control. As shown in Fig. 1B, the amounts of HrcA protein in the wild type and the mutant were similar, suggesting that ClpL is not involved in protein accumulation. In contrast, CtsR was greatly accumulated in S. mutans ΔclpL cells, and most of the protein was present as an aggregated form, suggesting that the accumulated form may be functionally inactive.
To determine whether the accumulation of CtsR in ΔclpL cells correlates with the ability of CtsR to repress transcription, we used a reporter fusion strain that carries the PclpP-gusA construct in the chromosome. The clpP promoter is an authentic target for CtsR binding, and PclpP-gusA was successfully used to measure CtsR repressor activity (36, 38). Since CtsR is a repressor, small amounts of active CtsR would produce increased Gus activity from this promoter fusion. The PclpP-gusA fusion construct was introduced into wild-type S. mutans strain UA159 and ΔclpL mutant strain IBSJ3 to create strains IBS514 and IBSJ9, respectively, and the β-glucuronidase (GusA) activity from these reporter strains was measured (Fig. 1C). Surprisingly, we observed that the GusA activity was increased in IBSJ9 compared with that in IBS514, even though the amount of the CtsR protein was much larger in the ΔclpL mutant strain (Fig. 1A). These results suggest that although the CtsR protein accumulated in the cell, in the absence of ClpL, the accumulated CtsR remained as an inactive form.
While aggregated proteins are usually highly refractory to various cellular proteases, we wanted to investigate whether ClpL participates directly in the degradation of the CtsR protein or whether ClpL prevents the misfolding of CtsR and thereby reduces total aggregation in the cells. Sequence analysis suggested that ClpL does not harbor any peptidase-like domains, and unlike other Clp ATPase proteins (such as ClpC or ClpX), clpL does not encode the IGF motif that is required for interactions with ClpP (11). Therefore, it is unlikely that ClpL is directly involved in the degradation of native CtsR to control the protein level. We designed an in vitro degradation assay to test the proteolytic potential of ClpL. Purified His-CtsR (100 ng/ml) was incubated with His-ClpL (300 ng/ml) and the whole-cell lysate (500 μg/ml) from S. mutans UA159 in the presence of 4 mM ATP at 37°C. The mixture was then separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a polyvinylidene difluoride (PVDF) membrane. Both His-CtsR and His-ClpL were detected with an anti-His antibody. No obvious degradation of His-CtsR was observed after 30 min of incubation (data not shown). Therefore, we speculate that ClpL controls the CtsR level in S. mutans by preventing protein aggregation rather than the degradation of the natural-form CtsR protein.
If ClpL were to engage in the folding of the CtsR protein, one would expect ClpL to interact directly with CtsR. To explore the possibility of the interaction between ClpL and CtsR, we employed an in vitro pulldown assay with purified proteins to assess the affinity between these two proteins. N-terminally tagged ClpL (His-ClpL) and C-terminally tagged CtsR (CtsR-Stag) were expressed separately and purified from E. coli as bait and prey proteins. CtsR-Stag (dissolved in PBS containing 2 M urea) exhibited a strong binding affinity for His-ClpL, while it showed no binding to empty Ni-NTA resin (Fig. 2A).
The ClpL-CtsR interaction was also verified in vivo. For this, we first expressed His-ClpL alone in E. coli cells. As expected, the majority of the ClpL was in the soluble fraction when induced at 37°C. On the other hand, when CtsR-Stag alone was expressed in E. coli cells, the protein was always in the insoluble fraction, even when the cells were grown at a lower temperature. However, when both His-ClpL and CtsR-Stag were coexpressed in E. coli, we found that nearly all the His-ClpL protein became insoluble (Fig. 2B). These data indicate that the ClpL-CtsR interaction also exists in vivo. To further confirm the in vivo protein-protein interaction between ClpL and CtsR, a bacterial two-hybrid (BACTH; Euromedex) system was employed to test their binding affinity. Bacterial colonies that coexpressed ClpL-T18 and CtsR-T25 fusion proteins became light blue, while the control bacterial colonies remained white (Fig. 2C). The affinity of the interaction was quantified by using a β-galactosidase (β-gal) assay and is shown in Fig. 2D. These data suggest that ClpL interacts directly with CtsR; however, the interaction between fusion proteins appears not to be very strong, since the level of β-gal activity was low to moderate.
Since we determined the protein-protein interaction between ClpL and CtsR, we speculated that ClpL might have a chaperone activity that helps the folding of CtsR. Note that CtsR was hardly soluble when expressed in E. coli cells despite the addition of an N-terminal or a C-terminal tag and was mainly inactive when accumulated in the S. mutans ΔclpL strain. Urea was used to solubilize CtsR-Stag aggregates, and we examined the chaperone activity of ClpL. When CtsR-Stag was dialyzed alone, the soluble CtsR protein was hardly detected after the removal of urea (Fig. 3). The addition of sheared salmon sperm DNA, which was previously reported to help the refolding of HrcA (41), had little contribution to CtsR-Stag refolding (data not shown). However, when twice the molar amount of His-ClpL was supplemented during dialysis, the refolding efficiency of CtsR-Stag was attained to approximately 80% (Fig. 3B). This refolded CtsR protein, which was devoid of urea, was active and retained its target DNA-binding activity (Fig. 3C). To test whether ClpL helps the folding of CtsR before aggregate formation or can disaggregate once the protein aggregates are formed, we incubated His-ClpL with CtsR-Stag aggregates in the presence of ATP and without urea. No soluble CtsR-Stag was detected in the supernatant fraction after 4 h of incubation, suggesting that ClpL alone was not sufficient to resolubilize the CtsR aggregates (Fig. 3D). However, our data suggest that ClpL can fold CtsR in vitro as well as in vivo and prevent the accumulation of CtsR aggregates due to the proper folding of the protein in the cell.
Primary sequence analysis showed that S. mutans ClpL contained two highly conserved ATP-binding regions (AAA-1 and AAA-2 domains) and a D2-small domain at its C-terminal domain. In comparison with protein sequences of known Clp ATPases in S. mutans, we found that the domain organization of ClpL is similar to that of ClpB, which is widely present in both Gram-positive and Gram-negative bacteria (Fig. 4A). ClpB is self-assembled to form an oligomeric complex mainly as a hexamer but also exists as a monomer and dimer (18, 23). When we applied the purified ClpL protein onto a Superdex 200 10/300 GL size-exclusion column, the elution profile yielded two equal-intensity peaks that correspond to molecular masses of >600 kDa and ~470 kDa (data not shown). The latter size corresponds to a hexameric protein, and the larger size indicates protein aggregates. Treatment of the sample with TCEP [tris(2-carboxyethyl)phosphine] (a reducing agent) greatly reduced the intensity of the larger peak but did not abolish it (data not shown). Thus, it appears that ClpL, like ClpB, also exists as a hexamer and perhaps other higher-order oligomers. Because of the sequence similarity with ClpB and the presence of the C-terminal D2-small domain that forms a tight interface with the AAA-2 domain of the neighboring subunit of ClpB (18), we then wanted to explore the contribution of the N- and C-terminal domains of ClpL to oligomerization. Toward this end, truncated ClpL proteins with either N- or C-terminally deleted regions were evaluated for interactions by BATCH assays (Fig. 4B). As expected, the assay showed that the ClpL-T18 fusion interacted strongly with the ClpL-T25 fusion (Fig. 4C). We also observed that the C-terminal D2-small domain is essential for ClpL oligomerization, since the deletion of this domain in ClpL-T18 (ClpLΔC-T18) resulted in the loss of the interaction with ClpL-T25 (Fig. 4C). However, the deletion of the N-terminal 117 residues in ClpL-T18 (ClpLΔN-T18) also impaired its interaction with ClpL-T25 (Fig. 4C). This is in contrast to the ClpB protein, where the deletion of the N-terminal domain does not abolish oligomerization, as reported previously (42). Taken together, our data indicate that ClpL can form a hexamer and that both the N- and C-terminal domains contribute to protein oligomerization. We also observed that the deletion of the N- or C-terminal domain of ClpL also weakens the interaction with CtsR (Fig. 2C and andDD).
To further understand the contribution of the ClpL N-terminal and C-terminal domains to its chaperone activity, we expressed either ClpLΔN or ClpLΔC in ΔclpL mutant strain IBSJ3/pIBJ1. Our previous study showed that His-CtsR accumulation was prevented in IBSJ3/pIBJ1 when full-length ClpL was expressed in trans from a plasmid (36). We found that the expression of ClpLΔN in IBSJ3/pIBJ1 can also prevent the cellular accumulation of His-CtsR (Fig. 5). However, the level of His-CtsR accumulation was unchanged when ClpLΔC was expressed in IBSJ3/pIBJ1 (Fig. 5), suggesting that the deletion of the C-terminal D2-small domain in ClpL resulted in the loss of its chaperone activity. Thus, the C-terminal D2 domain appears to play an important role in protein folding.
ClpL is a unique member of the HSP100 family that does not encode the motif required for interactions with ClpP. This protein is found strictly in Gram-positive bacteria, including streptococci, and is involved in various cellular processes, such as the stress tolerance response, virulence, long-term cell survival, and antibiotic resistance (7, 30–32). Despite its importance, very little is known about the molecular functions of this protein. Our previous study indicated that CtsR, a major heat shock regulator, accumulated in the clpL-deficient strain but not in the clpP-deficient strain (36). Subsequently, we found that most of the accumulated protein was present in the pellet fraction. Since the expression level from a CtsR-repressed promoter (PclpP) was increased in the clpL-deficient strain, the accumulated CtsR protein seemed to be improperly folded in the mutant strain. A previous study reported that S. pneumoniae ClpL can refold in the presence of ATP urea-denatured rhodanese (a nonstreptococcal protein) in a dose-dependent manner (29), indicating that ClpL might have chaperone activity. Therefore, we hypothesized that ClpL might be involved in the folding of CtsR in the cell. Our study presented here conclusively showed that CtsR is a bona fide substrate for ClpL.
In wild-type S. mutans cells, the CtsR level is very low, even after the overexpression of CtsR from a multicopy plasmid, indicating that the protein is readily degraded in the cell. Our data showed that ClpL did not enhance the degradation of the CtsR protein in its native form; thus, we suggest that the degradation step is correlated with the protein-folding stage. This step may require the involvement of other accessory proteins. Since ClpL does not contain any known peptidase domain, it seems that other proteases are involved in substrate degradation. Previously, it was reported that a large portion of ClpL is membrane associated, suggesting that ClpL may cooperate with membrane-associated proteases (31). Genome analysis indicated that S. mutans encodes two major membrane-associated stress-related proteases, FtsH and HtpX. In E. coli and Bacillus subtilis, FtsH is induced by heat and osmotic shocks (43, 44). While FtsH is not essential for growth in species such as B. subtilis, Corynebacterium glutamicum, and Lactobacillus plantarum (44–46), FtsH is essential for some bacteria, including E. coli and Lactococcus lactis (47, 48). Our repeated attempts to inactivate ftsH in S. mutans were unsuccessful, suggesting that FtsH is also essential in this organism. The other stress-related protease, HtpX, is a zinc-dependent metalloprotease (49) and is very poorly characterized. When we inactivated htpX in S. mutans, we found that HtpX was not involved in the degradation of CtsR (data not shown). Thus, additional bioinformatic and biochemical analyses are required to identify the protease involved in CtsR degradation.
It appears that CtsR is not the only substrate that is recognized by ClpL. Several proteins, especially some high-molecular-weight proteins, were differentially expressed in the clpL-deficient strain compared to its parental strain (data not shown). Recently, it was demonstrated that a knockout mutant of clpL displays increased susceptibility to penicillin-induced lysis in S. pneumoniae (31). The reason for this increased susceptibility is due to reduced PBP2x protein, which is required for cell wall biosynthesis. That study proposed that ClpL has two functions: to “stabilize and reactivate” PBP2x under stressful conditions and to facilitate the translocation of PBP2x to the cell wall, by some unknown mechanism, which leads to wall thickening and penicillin resistance (31). ClpL probably plays a similar role in S. mutans as well, since we observed increased bacitracin sensitivity in the clpL-deficient strain (data not shown). Furthermore, we also found that several other cell wall-damaging agents, such as penicillin and cefixime, have a similar effect on the clpL mutant strain, which displayed increased susceptibility toward these antibiotics.
Although primary sequence analysis indicated that ClpL is analogous to the ClpB protein, a molecular chaperone which is present in both Gram-negative and Gram-positive bacteria, some important differences between these two proteins also exist. Compared to ClpL, ClpB possesses two additional domains, the N and M domains. Despite the wealth of biochemical and structural data, the exact functions of these domains are not fully understood. The N domain of ClpB appears to be dispensable for its oligomerization ability and in vivo chaperone activity (42, 50), but it can increase interactions with protein aggregates (51, 52). On the other hand, the M domain, consisting of four α helices, is very specific for ClpB function and forms a large coiled-coil structure that protrudes from the ClpB hexameric ring (18). The conserved helix 3 of the M domain is particularly required for the DnaK-dependent shuffling of aggregated proteins and not for soluble denatured substrates (53). Mutations of this domain result in a loss of DnaK/J-GrpE-dependent disaggregation activity while retaining the DnaK/J-GrpE-independent functions of ClpB (42, 53, 54). Since the M domain is absent from the ClpL protein, we speculate that ClpL does not cooperate with the DnaK system and therefore is not able to disaggregate preformed protein aggregates, and our observation that ClpL alone was not capable of resolubilizing preformed CtsR aggregates is consistent with this notion. Instead, we speculate that ClpL interacts primarily with the target substrates before the formation of protein aggregates and helps to refold the substrates. Our data indicated that ClpL, like ClpB, is present as a hexamer in solution (23, 42, 55). However, unlike in ClpB, both the N- and C-terminal domains are necessary for ClpL oligomerization, as demonstrated by our bacterial two-hybrid assays. Our data also indicated that the C-terminal D2-small domain is essential for preventing CtsR aggregation in vivo.
In conclusion, we provide evidences that ClpL functions as a chaperone to fold the CtsR repressor, an important endogenous protein, and prevents the formation of deleterious protein aggregates under ambient growth conditions. Therefore, the ClpL chaperone activity is not restricted to stress-induced conditions. The reason why ClpL plays a critical role in long-term survival (7) is because ClpL contributes overall to cellular homeostasis by preventing the accumulation of protein aggregates. Our in vitro assays suggest that other protein cofactors are not required for this folding activity. However, we cannot rule out the possible involvement of other proteins during the in vivo folding and/or degradation process. Because ClpL plays an important role in various cellular functions, detailed biochemical and structural characterizations are necessary to understand the molecular mechanism of this novel chaperone. This will shed further light on how Gram-positive bacteria maintain cellular homeostasis and respond to various stresses.
This work was supported in part by an NIDCR grant (DE021664) to I.B.
Published ahead of print 30 November 2012