Strain-Dependent Recognition of a Unique Degradation Motif by ClpXP in Streptococcus mutans

Regulated proteolysis in bacteria is an important biological process that maintains protein homeostasis. ClpXP, an intracellular proteolytic complex, is the primary protease that is responsible for protein turnover. While the substrates for ClpXP were identified in Escherichia coli, the substrates for vast majority of bacteria are currently unknown. In this study, we identified a unique substrate for ClpXP-mediated degradation in Streptococcus mutans, a dental pathogen. We also found that a small motif composed of 3 amino acids is sufficient for ClpXP-mediated degradation. Identification of this motif will clearly help us to understand the pathogenesis of this organism and other related pathogens.

(UA159) strain to those of its isogenic ⌬clpX-mutant (IBSJ4) strains. Crude cellular lysates from UA159 and IBSJ4 were prepared from mid-exponential-phase cultures and subjected to proteomic analysis. Approximately 560 well-resolved protein spots could be detected in the pI range of 5.0 to 8.2 by silver staining (see Fig. S1 in the supplemental material). Comparison of the proteomes from UA159 and IBSJ4 revealed that at least 53 proteins had altered levels of expression, with a fold change value of 3.0 (P value of Յ0.05). Among the differentially expressed proteins, 27 were upregulated and 26 were downregulated in the IBSJ4 (⌬clpX) strain relative to those in the UA159 strain (see Table S2 in the supplemental material). We were able to excise only 37 of 53 protein spots. We then determined the identities of those selected spots by mass fingerprinting after in-gel trypsin digestion. We could determine the identity of only 17 spots unambiguously (Table S3). Among the upregulated proteins, the spot that showed the highest fold change value (7.9ϫ) corresponded to an autolysin (AltA) encoded by SMU.689. Two other protein spots corresponding to DivIVA (SMU.557) and SsbA (SMU.1859) were also significant since the relative abundances of these two spots were above 3-fold compared to that of the wild-type strain. Since the peptide coverage for these three proteins was relatively high, we considered these proteins to be putative targets for ClpXP. Proteins that were downregulated in the IBSJ4 strain included UvrA (SMU.1851) and LacA (SMU.1496). The relative abundance differences for these two proteins were above 3-fold. As expected, we found ClpX only in the proteome of the wild-type strain and not in the IBSJ4 proteome.
ClpX recognizes both DivIVA and SsbA. Other than the SsrA-tagged substrates, very little is known about the identity of the substrates that are recognized by ClpX in S. mutans. Our proteomic study indicated that at least three proteins, AltA, DivIVA, and SsbA, could be targeted by ClpXP-mediated degradation, probably by recognizing certain motifs present in the proteins. To identify the motifs that are recognized by ClpX, we chose to study DivIVA and SsbA further, as these proteins are relatively small (31 and 18 kDa, respectively). We first cloned the genes encoding the respective proteins in a shuttle expression vector, pIB190, which expresses the proteins with an N-terminal 6ϫ histidine (6ϫHis) tag. These constructs were then introduced into UA159, IBSJ4 (ΔclpX), and IBS512 (ΔclpP) cells. We then measured the levels of SsbA and DivIVA proteins in these strains by Western blot assays using anti-His antibody. We also used anti-enolase antibody to ensure that equal amounts of proteins were loaded into all lanes. As shown in Fig. 1, both the DivIVA protein and the SsbA protein were barely visible in the wild-type strain, whereas they were both accumulated in the ΔclpX-and cloned in pIB190 vector in N-terminal histidine-tagged form and were transformed in the wild-type (WT) UA159 strain and its isogenic ⌬clpX and ⌬clpP mutant strains. After cell lysis, equal amounts of protein (~20 g) were loaded in all lanes. The amounts of SsbA and DivIVA present in each strain were evaluated using anti-His antibody. The blots were developed with a chemiluminescent kit (ECL Plus; Thermo Scientific) by exposure to X-ray film and then photographed. Endogenous enolase was used as a loading control. Experiments were repeated at least three times, and the respective areas of a representative blot are shown.
ΔclpP-deficient strains. Taken together, our results suggest that both DivIVA and SsbA are the bona fide substrates for ClpX.
ClpXP recognizes three residues at the C terminus of the SsbA. To identify the sequences recognized by ClpXP, we focused on the C-terminal domain, since both the substrates were tagged at the N terminus with 6ϫHis residues. Furthermore, E. coli ClpXP is known to recognize various C-terminal tags (16). Since the tags that are recognized by E. coli ClpXP are small (three to six residues in length), we attached the last six residues of the DivIVA and SsbA to the green fluorescent protein (GFP), which we had previously found to be suitable for proteolytic studies (22). Addition of the six residues (KLNINE) derived from DivIVA to GFP did not render the protein degradable by ClpXP, whereas the addition of six residues (DDDLPF) derived from SsbA made the GFP susceptible to ClpXP-mediated degradation (Fig. 2). Deletion of the last six residues from SsbA rendered the protein refractory to degradation (Fig. S2). On the other hand, when the last six residues were deleted from DivIVA, ClpXP was still able to degrade DivIVA, suggesting that last six residues are not required for ClpXP recognition (Fig. S2).
To determine the minimum length of the motif that is recognized by S. mutans ClpXP, we made several sequential deletion derivatives of the six-residue tags derived from SsbA and added these tags to GFP. As shown in Fig. 2, addition of just the last three residues (LPF) to GFP rendered the protein susceptible to degradation by S. mutans ClpXP (please see Fig. S3). To understand the importance of each residue and its position in the motif for ClpXP recognition, we made several single amino acid substitutions in the LPF sequence. As shown in Fig. 3, it appears that for accurate ClpXP recognition, all three positions are very important. Thus, our results indicate that S. mutans ClpXP recognizes the C-terminal LPF motif present in the target substrate.
SsbA is not degraded in a clpX-deficient strain. To gain insight into the mechanism of ClpXP function, we performed a time course experiment to examine the stability of SsbA in the wild-type strain and the clpX-deleted strain under nonstressed conditions. As shown in Fig. 4, SsbA was rapidly degraded in the wild-type strain. In fact, we found only trace amounts of SsbA immediately before the addition of chloramphenicol and they remained the same up to the end of the time course (1 h). On the other hand, the amount of SsbA was constant in the clpX mutant throughout the time course of the experiment (Fig. 4A). Therefore, our results suggest that SsbA is highly unstable in S. mutans; however, when clpX was absent, SsbA was not degraded and accumulated in the cell. A similar observation was also made when GFP-LPF was used as the substrate (Fig. S4).

SsbA is not recognized by other Clp ATPases in S. mutans.
We recently found that SsrA-tagged proteins in S. mutans are degraded by ClpCP and ClpEP, in addition to being degraded by ClpXP (23). We therefore wanted to study whether ClpC or ClpE could also recognize SsbA protein for degradation. For this, we introduced the plasmid carrying SsbA with the N-terminal histidine (His) tag into the IBSJ2 (ΔclpC) and IBSJ5 (ΔclpE) strains (25). Expression of the SsbA protein was then evaluated in these strains at the mid-logarithmic-growth phase by Western blotting. As shown in Fig. 5, neither ClpC nor ClpE recognized SsbA in S. mutans. Thus, SsbA is specifically targeted by ClpXP-mediated degradation and not by ClpCP or ClpEP proteolytic complexes in S. mutans. S. mutans ClpXP also degrades SsbB containing the LPF motif. Since ClpXP recognizes the C-terminal LPF motif on the target substrates, we scanned the S. mutans genome for the presence of the LPF motif at the C terminus. We found only two additional proteins, namely, SsbB (SMU.1967) and a hypothetical protein encoded by SMU.394, with the LPF motif in the S. mutans UA159 genome. To verify whether ClpXP recognizes the SsbB protein, we cloned SMU.1967 in plasmid pIB190 so that SsbB could be expressed as N-terminal histidine-tagged protein. The construct was then introduced into the wild-type (UA159) and ΔclpX (IBSJ4) strains. As shown in Fig. 6, SsbB, like SsbA, was rapidly degraded in the wild-type strain, whereas SsbB was highly stable in the ΔclpX mutant strain. Thus, S. mutans ClpXP also recognizes the SsbB protein in addition to the SsbA protein.

Recognition of the LPF motif by ClpXP is strain dependent. We previously
showed that ClpXP could degrade GFP with the AVAA tag at the C terminus in vitro. To verify whether ClpXP is able to recognize the LPF motif in vitro, we used GFP with the LPF tag (GFP-LPF) as the substrate. We also included GFP and GPF-AVAA as negative and positive controls, respectively. As expected, ClpP alone did not show any degradation of any of the substrates (Fig. 7C) and ClpXP showed rapid degradation of GFP-AVAA (Fig. 7A). Half of the GFP-AVAA substrates were degraded at 10 min after ClpXP addition, and complete degradation was observed within 40 min ( Fig. 7A and B). On the other hand, we observed no degradation of GFP-LPF even after prolonged (up to 2 h) incubation. This result suggests that either the in vitro condition is not optimal for the recognition of the LPF motif or an accessory factor such as an adaptor protein is needed for the recognition.
The S. mutans genome is composed of roughly 1,500 core genes and about 4,000 pan-genes (26). While the clpX and clpP genes are considered part of the core genome, there is a pronounced strain-to-strain variability present among various S. mutans strains. Thus, we wanted to evaluate whether recognition of the LPF motif by ClpXP in S. mutans is a universal phenomenon or is strain dependent. For this, we included five additional strains (GS-5, N3209, OMZ175, UA130, and V403); the complete genome sequences for all, except UA130, are in the public domain. These strains were transformed with SsbA and GFP constructs and were evaluated for ClpXP-mediated degradation. As shown in Fig. 8A, LPF was degraded only in N3209 and UA130. On the other hand, as shown in Fig. 8B, the LPF motif was not recognized in strains GS-5, OMZ175, and V403. To confirm that ClpXP is functional in the latter strains, we introduced GFP-AVAA in GS-5, OMZ175 and V403 strains. As shown in Fig. 8B, GFP-AVAA was rapidly degraded in all three strains, indicating that ClpXP is functional in these strains. Together, these data suggest that the LPF motif is recognized by some but not all S. mutans strains.
The LPF motif is also not recognized by ClpXP in other bacteria. To determine whether ClpXP from other organisms recognizes SsbA or the last six residues of SsbA, we first selected E. coli. We introduced either an SsbA protein-encoding plasmid or GFP-DDDLPF into wild-type E. coli as well as clpX-and clpP-deficient mutants of E. coli and measured the amounts of SsbA or GFP accumulation in the cell. As shown in Fig. 9A, we observed that GFP, GFP-DDDLPF, and SsbA accumulated in comparable amounts in all the strains, suggesting that neither SsbA nor the GFP-DDDLPF tag was recognized by E. coli ClpXP.
We then extended the study to Streptococcus gordonii, also an oral bacterium. To this end, we wanted to construct both clpP-and clpX-deficient derivatives of S. gordonii. After repeated attempts, we were unable to generate a clpP-deficient S. gordonii. However, we were successful in generating a clpX-deficient derivative. We introduced the SsbA-encoding plasmid into both the wild-type and the clpX-deficient S. gordonii strains and measured the SsbA accumulation by Western blotting. As shown in Fig. 9B, we did not observe any significant difference in the SsbA amounts in these strains, suggesting that S. gordonii ClpXP does not recognize S. mutans SsbA and hence the LPF motif. Similarly, we found that S. sangunis ClpXP does not degrade the LFP motif (Fig. 9C). Among the species tested, our data clearly suggest that the LPF motif is selectively recognized by S. mutans ClpXP.

DISCUSSION
Regulated proteolysis is an essential physiological process responsible for cellular protein homeostasis and protein quality control in bacteria. Key to this important process is the involvement of the ClpXP proteolytic complex. As is the case in other well-studied bacteria such as E. coli and Bacillus subtilis, ClpXP is the major proteolytic complex involved in protein quality control in S. mutans (19,22,27). As expected from its role in regulated proteolysis, inactivation of either clpP or clpX leads to pleotropic effects, including defects in biofilm formation, bacteriocin production, and competence development (19,27). While we know the importance of ClpXP in cellular physiology, the substrates recognized by this proteolytic complex are largely unknown. In this study, we used two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) to identify putative substrates that are recognized by ClpX. We found both up-and downregulated proteins by proteomic analysis. While the exact reason for the presence of downregulated proteins in the ClpX-deficient mutant is not known, we speculated that some of the upregulated proteins are the substrates for ClpX. Surprisingly, among the 17 spots that we identified by liquid chromatography-mass spectrometry (LC-MS), none were identified by our previous proteomic study of ClpP-dependent expression Experiments were repeated at least three times, and a representative blot is shown. (19). The probable reasons could be attributable to differences in growth conditions, incomplete identification of the protein spots, and redundant activity of some of the Clp ATPases. Nevertheless, using tagged proteins in an in vivo degradation assay, we confirmed the identities of two proteins that are bona fide ClpX substrates.
ClpXP-mediated substrate degradation often requires short peptide motifs that are recognized by the protease complex (1). Among these motifs, the best-studied one is encoded by an SsrA tag, which generally adds an~11-residue peptide with either AVAA or ALAA at the C terminus. The SsrA tag is highly conserved in bacteria and is recognized by ClpX in many bacterial species, including S. mutans. Although several ClpX substrates have been identified in other bacteria, including B. subtilis, the short degradation motifs (three residues) that are recognized by ClpX have been documented only in E. coli (15,28). In that bacterium, ClpX can recognize five types of motifs: three located at the N terminus and two at the C terminus of the target proteins (15). In this study, we identified a three-residue-long C-terminal motif (LPF) that is recognized by S. mutans ClpX. Interestingly, it was previously reported that ClpC in B. subtilis also recognizes a short motif (LCN) located at the C terminus of SpoIIB protein (29); however, no such short motif is reported for ClpX ATPase. To the best of our knowledge, LPF is the first such short motif identified in any organism other than E. coli that is solely recognized by ClpX ATPase.
GPF is a folded protein whose last three residues are LYK and is not degraded by ClpXP. We found that addition of LPF to GFP leads to complete degradation of the protein and that three residues are sufficient for the recognition and unfolding by ClpX ATPase. When any of these residues were replaced with other residues, ClpX was unable to recognize the GFP. Among the replacements that we tested, the only substitution in the LPF motif that was tolerated by ClpX was the last residue, phenylalanine, replaced with tyrosine. Since these 2 amino acids share similar structural physiochemical properties, it was not surprising that ClpX could also recognize LPY. The mere presence of a phenylalanine or tyrosine residue along with proline (PF or PY) at the C terminus of a protein is also not sufficient for the recognition. However, we have not extensively and systematically tested all the residues for ClpXP recognition. We also found that the LPF motif is not recognized by any other Clp ATPases (Fig. 3), suggesting that LPF is specific for ClpXP.

FIG 8 Recognition of the LPF motif in various S. mutans isolates. Panel A shows the strains that successfully recognized the LPF motif for degradation, whereas panel B shows the strains that failed to recognize the motif. GFP construct was used as a negative control, whereas GFP-AVAA was used as a positive control to demonstrate that ClpXP is functional in those strains that do not degrade LPF. Experiments were repeated at least twice, and a representative blot is shown.
In the S. mutans core genome, only three proteins contain the LPF motif and one protein contains the LPY motif. We tested two of the LPF-containing proteins, SsbA and SsbB; both are single-strand DNA binding proteins. These two proteins are highly similar to the SsbA and SsbB proteins from Streptococcus pneumoniae. The SsbA protein is similar in size to the well-characterized SSB protein from E. coli (SsbEc). The SsbB protein, in contrast, is a smaller protein that is specifically induced during natural transformation in S. pneumoniae and has no counterpart in E. coli (30); the amount of SsbB is about 20 times more than the amount of SsbA during competence in S. pneumoniae (30). Even after overexpression, we could barely detect the SsbA (or SsbB) protein in the wild-type S. mutans strain, indicating rapid degradation of the proteins. It is not clear why these key proteins are rapidly degraded in S. mutans, but the reason  Table 1)

. The levels of His-SsbA, GFP-DDDLPF, and GFP expression were analyzed by Western blotting using anti-His antibody. (B) Recognition of the LPF motif in S. gordonii. His-SsbA was also expressed in S. gordonii DL-1 and its isogenic clpX mutant strains. The level of SsbA expression was measured by Western blot analysis using anti-His antibody. The GFP-AVAA construct was used as a positive control to demonstrate that ClpXP is functional in DL-1. Experiments were repeated at least three times, and a representative blot is shown. Enolase was used as a loading control only for the lower panel. (C) Recognition of the LPF motif in S. sanguinis. GFP, GFP-LPF, and GFP-AVAA constructs were introduced into S. sanguinis SK-36. Protein expression was measured using anti-GFP antibody. Enolase was used as a loading control. Experiments were repeated at least twice, and a representative blot is shown.
could be that the proteins are required to maintain stringency of competence development. Furthermore, since these proteins are not degraded in all S. mutans strains, how the competence is regulated in these strains remains unclear. We speculate that, in strains where SsbA/B and DivIVA are rapidly degraded, either additional regulators are involved that help to modulate the expression of ClpXP or some key protein (adaptor; see below) controls the function of the ClpXP complex. During competence development, the abundance of these regulators or of the adaptors is altered and the Ssb and DivIVA proteins become transiently available.
The other LPF-containing protein is SMU.394, which encodes a nucleoid-associated protein of the EbfC family (31). The role of this protein in S. mutans is currently unknown, but its homolog in Borrelia burgdorferi is involved in virulence gene regulation (31). A search for LPF-containing protein in the S. mutans pan-genome also identified another small (~25-kDa) protein of the Abi C family (GenBank accession number AFM81543). Abi C family proteins are involved in abortive phage infections (32). The Abi C protein that we found is present in about 25 S. mutans strains, including GS-5, but the exact function of this protein is currently unknown. On the other hand, the LPY motif-containing protein that we identified is an ABC transporter protein (SMU.459) that might be involved in amino acid (cysteine) transport and is well conserved across the S. mutans strains.
Recently (in 2010), Niu and colleagues identified another motif, VSA, which is present internally in the IrvR repressor (24). Upon autocatalytic cleavage of IrvR, the VSA motif is amenable to degradation. However, those authors found that simply adding the VSA motif does not render the substrate sensitive to ClpXP-mediated degradation; the additional three residues are required for complete degradation (24). Furthermore, ClpC also recognizes this short six-residue motif that is not specific for ClpX recognition. We have found that DivIVA is also a bona fide substrate for ClpX recognition. However, in this case, ClpX appears to recognize some motifs that are not located at the C terminus. This is because deletion of the last six residues (KLNINE) from DivIVA did not protect the substrate from ClpXP-mediated degradation. Furthermore, addition of the KLNINE motif to GFP did not make the protein sensitive to ClpXP-mediated degradation. DivIVA is a conserved protein in Gram-positive bacteria that localizes at the pole and division sites (33). In S. pneumoniae, DivIVA is also phosphorylated by a eukaryoticlike Ser/Thr kinase (StkP) (34). Protein phosphorylation often plays a role in degradation. For example, autophosphorylation of a eukaryotic-like tyrosine kinase in E. coli stabilizes the protein and protects it from degradation (35). In B. subtilis, protein phosphorylation leads to degradation of certain proteins by Clp proteolytic complexes (36). At present, we do not know whether protein phosphorylation plays a role in the degradation of DivIVA by ClpXP in S. mutans. However, we are in the process of evaluating the role of Ser/Thr phosphorylation in substrate degradation by ClpXP. It is also not clear how the level of DivIVA is modulated in the cell during cell division, since it is an important protein required for cell division.
We found that other streptococci such as S. sanguinis and S. gordonii do not recognize the LPF tag. Surprisingly, we also found that the LPF tag is recognized by a subset of S. mutans strains but not by all isolates. While the exact reason for this strain-specific phenomenon is under investigation, we can speculate with respect to two different scenarios. First, it is possible that the LPF tag is not directly recognized by the ClpXP complex and that the tag is instead brought to the complex by an adaptor protein. To identify the putative adaptor protein, we compared the genome sequences of two strains that recognize the tag (UA159 and N3209) with the genome sequences of the two strains that do not (GS-5 and OMZ175). We were unable to find any strain-specific protein that is present in UA159 and N3209 but absent in GS-5 and OMZ175. The absence of a strain-specific protein does not rule out the possibility of involvement of an adaptor protein for several reasons. For example, the adaptor protein may be encoded by GS-5 and OMZ175 but may not be expressed in the strains. Alternatively, the adaptor may be encoded and expressed but may carry some mutations that render the protein nonfunctional. Another possibility is the presence of an inhibitory protein in the strains that do not recognize the tag because the protein sequesters the LPF tag such that ClpXP is unable to access the tag. In fact, the charged C-terminal tail of Ssb proteins is known to recruit at least 12 proteins that are involved in chromosome stability (37). Interestingly, our in vitro assay using purified ClpXP was unable to degrade the LPF-tagged substrates. It is also possible that the substrate used for in vitro studies was isolated from E. coli and therefore was folded in such a way that ClpXP could not recognize the LPF tag. On the other hand, it is also possible that the adaptor protein was absent in the in vitro reactions. We attempted to supplement the in vitro assay with crude lysate prepared from the stationary-phase UA159 culture but were unable to observe LPF degradation. This negative result suggests, among many other possibilities, either that the adaptor protein was not very stable or that it was not present in high abundance. An extensive in-depth analysis involving both genetics and biochemical approaches is necessary to identify the adaptor protein required for ClpXP activity in S. mutans.

MATERIALS AND METHODS
Bacterial growth conditions. E. coli strains were grown in Luria-Bertani (LB) medium supplemented with erythromycin (Er; 500 g/ml), ampicillin (Amp; 100 g/ml), and kanamycin (Kan; 50 g/ml) (whenever required). S. mutans (UA159 and its derivatives) strains were routinely grown in Todd-Hewitt medium supplemented with 0.2% yeast extract (THY medium). Whenever necessary, 5 g/ml erythromycin or 400 g/ml kanamycin was added to the THY medium.
Plasmid construction. To express SsbA and SsbB in S. mutans as N-terminal histidine-tagged variants, plasmids pIBP1 and pIBP72, respectively, were constructed by cloning ssbA and ssbB (using UA159 genomic DNA as a template) in the pIB190 vector (38). To express GFP with various tags at the C-terminal end, pIBP6 (GFP alone), pIBP7 (GFP-DDDLPF), pIBP57 (GFP-DDLPF), pIBP58 (GFP-DLPF), pIBP59 (GFP-LPF), and pIBP5 (GFP-KLNINE) were generated. Plasmids pIBP68, pIBP69, and pIBP70 were constructed by site-directed mutagenesis using pIBP59 as a template. To check the stability of DivIVA and DivIVAΔ6 (without the last six residues), pIBP2 and pIBP72, respectively, were constructed using pIB190. All pIB190-specific derivatives contained a His tag (6ϫ) at the N-terminal end. All plasmids were transformed in various S. mutans strains by natural transformation. Transformants were verified by PCR using vector (pIB190)-specific primers (pIB190F and pIB190R). Primers used for constructing these plasmids are listed in Table S1 in the supplemental material.
Construction of the clpX deletion in Streptococcus gordonii. Deletion of clpX in S. gordonii was performed using a Cre-loxP protocol as described previously (39,40). Briefly, the S. gordonii clpX gene fragment was cloned in pGEM-T Easy vector. Using the inverse PCR method, an internal region of the clpX gene was deleted and a kanamycin resistance cassette (Kan r ) with modified loxP sites (lox71-Kan-lox66) was inserted within the deleted gene to generate plasmid pIBP73. Using pIBP73, the clpX gene disrupted by the Kan r cassette was amplified with primers (SGordClpXF and SGordClpXR) and introduced into S. gordonii by natural transformation. The transformants were selected on THY plates containing kanamycin, and the disruption of the clpX gene was verified by PCR.
Two-dimensional gel electrophoresis (2-DGE) and spot identification. S. mutans strains (UA159 and IBSJ2) were grown to the mid-exponential-growth phase, and the cells were harvested by centrifugation. The cell pellets were lysed in 300 l each of osmotic lysis buffer containing 10ϫ nuclease stock, phosphatase inhibitor stocks (I and II), and 100 mg of washed glass beads as described by Jazwinski (41). The samples were subjected to vortex mixing (5 min), sonicated (5 min), and subjected to vortex mixing again. Protein concentrations of the supernatants were determined using the bicinchoninic acid (BCA) assay. The samples were adjusted to a concentration of 1 mg/ml in urea sample buffer before loading was performed (42). 2-DGE was performed according to the carrier ampholine method of isoelectric focusing (IEF) with tropomyosin (molecular weight [MW] of 33,000 and pI 5.2) as an internal standard (42). A 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) procedure was carried out as the second dimension, and staining was performed with a special silver stain that is compatible with mass spectroscopy. Two independent samples were run for each strain. The differences among the spots were calculated from spot percentages (density of the individual spots divided by total density of all measured spots). Spots showing at least a 1.7-fold difference between the strains with a P value of Յ0.05 (Student's t test) were extracted, and their identity was determined by electrospray ionization (ESI) mass spectrometry. The sample preparations and the 2-DGE were carried out at Kendrick Laboratories.
Protein extraction and Western blot analysis. For streptococcal protein extraction, a culture that had been grown overnight was subcultured in fresh THY medium with a 5% inoculum and was allowed to grow to late log phase (approximate optical density at 600 nm [~OD 600 ], 0.7). Cultures (50 ml) were harvested by centrifugation, resuspended in 1 ml of buffer A (HEPES-NaOH [pH 7.5], 300 mM NaCl, 5% glycerol), and lysed by bead beating. The suspension was centrifuged, the supernatant (500 l) was collected, and the total protein concentration was measured by the Bradford method (43). Equal amounts of protein (~20 g) were loaded in all lanes and subjected to SDS-PAGE. The gels were electroblotted on a polyvinylidene difluoride (PVDF) membrane. Western blotting was performed using a standard protocol. Mouse anti-His antibody (Qiagen) or mouse anti-GFP antibody (Sigma) was used as the primary antibody and goat anti-mouse IgG peroxidase (Sigma) as a secondary antibody. Blots were developed using ECL Plus reagent (Thermo Scientific). Enolase was used as a loading control, and the same membrane was used to measure enolase levels. For this, the PVDF membrane was stripped with stripping buffer (Tris-buffered saline containing 2% SDS and 100 mM ␤-mercaptoethanol) and was reprobed with anti-enolase antibody. To isolate total protein from E. coli, a saturated culture that had been grown overnight was inoculated in LB and grown to mid-log phase (OD 600 , 0.5). Cultures (10 ml) were centrifuged, and cell pellets were lysed by sonication. Equal amounts of protein were loaded in all lanes for SDS-PAGE, and Western blotting was performed as described above.
Time course assay for protein stability. To measure the stability of SsbA and GFPs in S. mutans, UA159 and clpX mutant strains carrying either pIBP1 or pIBP6 were grown overnight and subcultured into fresh THY medium at 37°C with appropriate antibiotics. When the cultures reached the late-log phase (~OD 600 , 0.7), chloramphenicol (20 g/ml) was added to inhibit protein synthesis. At different time intervals, 10-ml aliquots were recovered for total cellular lysate preparation and Western blot analyses.
Protein purification for in vitro studies. To purify native ClpP, the open reading frame was cloned in the pET3a vector. Protein expression was induced in the E. coli Bl21(DE3) strain using 1 mM IPTG (isopropyl-␤-D-thiogalactopyranoside). After induction, the cells pellets were lysed in buffer B (25 mM HEPES-NaOH [pH 7.5], 250 mM NaCl, 5% glycerol) by sonication. Clarified cell lysate was loaded onto an SP HQ column (GE Healthcare) and subjected to anion exchange chromatography. The protein was eluted using a salt gradient of 250 to 1,000 mM NaCl, and the purity of the recovered protein was assessed by SDS-PAGE. S. mutans ClpX was purified as described by Tao and Biswas (22).
To purify GFP and GFP-DDDLPF, pIBP12 and pIBP13 were constructed using pETDuet1 expression vector. Protein expression was induced in the E. coli Bl21(DE3) strain by the use of 1 mM IPTG. Recombinant proteins were purified using nickel-nitrilotriacetic acid (Ni-NTA) (Novagen) affinity chromatography and dialyzed against buffer C (25 mM HEPES/KOH [pH 7.5], 200 mM KCl, 10% glycerol). The purity of the proteins was assessed by SDS-PAGE. GFP-AVAA was purified as described by Tao and Biswas (22).