Cooperative action of Escherichia coli ClpB protein and DnaK chaperone in the activation of a replication initiation protein.

The Escherichia coli molecular chaperone protein ClpB is a member of the highly conserved Hsp100/Clp protein family. Previous studies have shown that the ClpB protein is needed for bacterial thermotolerance. Purified ClpB protein has been shown to reactivate chemically and heat-denatured proteins. In this work we demonstrate that the combined action of ClpB and the DnaK, DnaJ, and GrpE chaperones leads to the activation of DNA replication of the broad-host-range plasmid RK2. In contrast, ClpB is not needed for the activation of the oriC-dependent replication of E. coli. Using purified protein components we show that the ClpB/DnaK/DnaJ/GrpE synergistic action activates the plasmid RK2 replication initiation protein TrfA by converting inactive dimers to an active monomer form. In contrast, Hsp78/Ssc1/Mdj1/Mge1, the corresponding protein system from yeast mitochondria, cannot activate the TrfA replication protein. Our results demonstrate for the first time that the ClpB/DnaK/DnaJ/GrpE system is involved in protein monomerization and in the activation of a DNA replication factor.

The Escherichia coli protein ClpB is a member of the universally conserved Hsp100/Clp protein family (1,2). Based on the growth defect phenotype of E. coli clpB mutants, it has been proposed that ClpB is involved in thermotolerance (3,4). In agreement with this, thermally aggregated proteins were shown to be disaggregated in wild-type cells but not in a clpB mutant strain upon a shift to lower temperature (5,6). Recently, the cooperative action of ClpB with the DnaK, DnaJ, and GrpE chaperone system has been reported. It has been shown that ClpB cooperates in vitro with the DnaK chaperone system in the reactivation of chemically and heat-denatured substrate proteins (7)(8)(9)(10).
Other members of the Hsp100/Clp protein family, namely the E. coli proteins ClpA, ClpX, and ClpY, have been shown to be involved in the ATP-dependent protein degradation through a transient association with the catalytic protease components ClpP and ClpQ (ClpP for ClpX and ClpA; ClpQ for ClpY) (11). In the absence of ClpP, ClpX and ClpA display chaperone activities (12,13). Several substrates have been identified both for the chaperone and proteolytic activities of these proteins (1, 2, 14 -18).
Previous work has shown that the broad-host-range plasmid RK2 can replicate and maintain itself in a wide variety of Gram-negative bacteria (19). This property has made RK2 a particularly interesting system for studying fundamental questions about the cellular processes involved in DNA replication. Only two factors required for RK2 replication are encoded by the plasmid DNA: (i) the replication origin (oriV) sequence consisting of five iterons (17-bp iterated sequences) (20) and four DnaA boxes (9-bp DnaA protein binding sites) (21), and (ii) the plasmid replication initiation protein TrfA (22)(23)(24). Other of the proteins required for plasmid DNA replication are supplied by the bacterial host (25). It has been demonstrated in a purified system that the DnaA initiation protein, DnaB helicase, DnaC, DNA gyrase, DnaG primase, SSB, HU, and Pol III proteins of E. coli are required for RK2 replication (26). The binding of the TrfA protein to the iteron sequences is the most crucial step during RK2 replication initiation. Only the monomeric form of the TrfA protein can bind to the iteron sequences and initiate DNA replication (27). The binding of TrfA in the presence of HU results in strand opening in the AT-rich oriV region, specifically at the four 13-mers present within this region (21). This open structure serves as the entry site for the DnaB helicase complex, thus leading to initiation of DNA replication (26).
The TrfA initiation protein exists in two forms, 44-kDa and 33-kDa, as a result of an internal translational start in the trfA gene (22)(23)(24). It has been shown that either protein can function in DNA replication in E. coli. However, the 44-kDa TrfA is indispensable during RK2 replication initiation in Pseudomonas aeruginosa (24,28,29). Although the purified TrfA protein is found largely in the dimeric form, it binds iteron sequences as a monomer (30 -32). The involvement of the chaperone protein ClpX in the activation of TrfA for iteron binding and consequently for initiating oriV replication has been previously studied in detail (33). It has been shown that the ClpX chaperone protein can convert TrfA dimers into active monomers in vitro.
In previous work, evidence was obtained that in addition to ClpX other chaperone(s) may also play a role in TrfA activation in E. coli (33). Here, we demonstrated that ClpB in combination with DnaK chaperone system plays a major role in the replication of the RK2 plasmid. Our work is in contrast with studies with plasmid P1 where it was shown in vitro that the replication initiation protein of this plasmid can be activated to a monomer form by either the DnaK chaperone machinery alone or the ClpA protein alone (12,34).

MATERIALS AND METHODS
Bacterial and Yeast Strains, Proteins, Plasmids, and Reagents-The E. coli strains used in this study were: MC1000 and SG20250, the parent strains of MC1000 (clpB::kan), and SG22100 (clpB::kan), respectively (3,35). Published protocols were utilized for the purification of E. coli chaperone proteins ClpB (36), DnaK, DnaJ, and GrpE (37). The yeast chaperones Hsp78, Ssc1, Mdj1, and Mdg1 were purified as described (38,39). Purification of the E. coli DnaA, DnaB, and DnaC proteins were carried out following the established protocols described previously (26). The 33-kDa histidine-tagged version of the RK2 plasmid replication initiation protein TrfA was purified as previously described (27). The histidine-tagged monomeric mutant form of the 33-kDa TrfAG254D/S267L was purified as described (40). This protein, has been used in previous RK2 biochemical analyses (21,26,33,(41)(42)(43)(44)(45) and is known to initiate replication in vitro with kinetics similar to those of the largely dimeric wild-type protein. Unlike the wild-type, the mutant protein is present largely in the form of the monomer, which is the active form of TrfA for binding to the iterons at the RK2 replication origin (27). Plasmids pTJS42 and pKD19L1 carrying the minimal RK2 origin region and plasmid pBSoriC carrying the oriC DNA origin fragment have been previously described (41,46). HU, SSB, and DNA gyrase proteins were obtained from Enzyco, Inc., creatine kinase, bo-vine serum albumin (Fraction V), creatine phosphate, and rNTPs from Sigma, dNTPs from Pharmacia, [methyl-3 H]dTTP from ICN Radiochemicals, and goat anti-rabbit IgG from Bio-Rad.
RK2 oriV and E. coli oriC DNA Replication Using Crude Extracts-Preparation of E. coli crude extracts (FII) and reaction conditions were essentially as previously described (47). The standard reactions contained 300 ng of plasmid template and histidine-tagged versions of wild-type 33-kDa TrfA or the largely monomeric preparations of the 33-kDa TrfA mutant protein TrfA 254D/267L (42) at the indicated concentrations. For oriC-dependent reactions, the pBSoriC template was used, and the TrfA protein was omitted. Incubation was at 32°C for 60 min. Reactions were stopped by trichloroacetic acid precipitation and total nucleotide incorporation (pmol) was measured by liquid scintillation counting following filtration onto Whatman GF/C glass fiber filters.
Glycerol Gradient Centrifugation-To analyze the monomeric and dimeric forms of TrfA protein, glycerol gradient centrifugation was performed essentially as described in (48). Reaction mixtures (60 l) with proteins in buffer were assembled on ice and contained 40 mM Hepes/KOH, pH 8.0, 25 mM Tris/HCl, pH 7.4, 80 g/ml bovine serum albumin, 4% sucrose, 4 mM dithiothreitol, 11 mM magnesium acetate, and 2 mM ATP. The mixture was incubated for 10 min at 32°C and then loaded onto a 3.2-ml linear 15-35% (v/v) glycerol gradient in buffer (50 mM Hepes, pH 8.0, 10 mM MgCl 2 , 200 mM KCl, 4 mM dithiothreitol) and centrifuged at 46,000 rpm for 24h at 2°C in a Beckman SW60 rotor. Fractions were collected from the top of the tube and analyzed by SDS/PAGE, followed by protein transfer and immunoblotting with rabbit antisera against the TrfA protein. Bound rabbit antibody was detected by a colorimetric reaction with an horseradish peroxidase-conjugated goat anti-rabbit IgG.

The ClpB Chaperone Plays a Role in RK2 DNA Replication-
Previously published in vitro results demonstrated that the plasmid RK2 replication initiation protein TrfA can be activated by a chaperone (33). ClpX, the molecular chaperone in the ATP-dependent reaction, converts TrfA dimers to active monomers. In agreement, the mutant protein TrfA 254D/267L, which exists mainly in monomeric form does not require the presence of ClpX for replication activity. Because we were not able to observe a significant effect of the deletion of the clpX gene on the maintenance of the RK2 plasmid in E. coli, we previously proposed that other chaperone system(s) are also involved in TrfA activation (33). Based on the previous analysis of RK2 DNA replication in protein extracts prepared from E. coli strains carrying mutations in different chaperones (33), ClpB was found to be the most attractive candidate.
The effect of ClpB on the initiation of DNA replication was tested using a bacterial crude extract replication system (see "Materials and Methods"). Extracts obtained either from wildtype E. coli cells or the clpB deletion mutant were used in two different replication reactions. The first one, oriV-dependent, utilizes the plasmid RK2 minimal replication origin and the plasmid replication initiation protein TrfA. The second one, oriC-dependent, represents DNA replication from the E. coli minimal replication origin dependent on the presence of the DnaA initiation protein. oriV DNA replication activity in crude FIG. 1. ClpB chaperone plays a role in oriV DNA replication. In vitro replication reactions using oriV (pKDL19) and oriC (pBSoriC) as DNA templates were carried out in bacterial crude Fraction II extracts. A, extracts prepared from an E. coli clpB deletion mutant and its wild-type isogenic strain have been used to test oriC-and oriV-dependent replication. The wild-type His6-TrfA (200 ng) was added to the oriV-dependent reaction. Relative DNA synthesis was calculated as a percentage of nucleotide incorporation obtained in standard reactions using extract prepared from wild-type E. coli (240 pmol for oriC and 120 pmol for oriV). B, the wild-type His6-TrfA (200 ng) was used in the oriVdependent replication mixture supplemented with various amounts of ClpB protein as indicated. C, various amounts of ClpB protein were added to the oriC-dependent replication mixture as indicated. Reactions shown in B and C were carried out in extracts prepared from an E. coli clpB deletion mutant under the conditions described under "Material and Methods." Fraction II extracts prepared from the clpB-deficient strain was significantly lower (80% reduction) when compared with the same reaction carried out in extracts prepared from wildtype bacteria (Fig. 1A). In contrast, the absence of ClpB in the crude extract had only a limited effect on oriC DNA replication (25% reduction).
We also analyzed the effect of the addition of purified ClpB protein to the oriV and oriC replication reactions using crude Fraction II extracts prepared from the E. coli clpB-deficient strain (Fig. 1, B and C). In the presence of ClpB, the level of oriV-dependent replication increased 2-fold (Fig. 1B). In contrast, oriC DNA replication was somewhat diminished (Fig.  1C). The reduction in oriC replication observed in these experiments may be due to the nonspecific disaggregation of particular replication machinery components. A similar nonspecific decrease in the DNA replication has been observed during experiments with the monomeric form of the TrfA protein (Fig.  2B). Overall the results indicate that plasmid RK2 DNA replication but not bacterial DNA replication is activated by the ClpB chaperone.
ClpB Is Involved in the Activation of the Dimeric Wild-type TrfA but Not Monomeric Mutant-Using the oriV-dependent in vitro replication system and utilizing crude Fraction II extracts prepared from the E. coli clpB mutant, we investigated the replication activity of the 33-kDa dimeric wild-type TrfA protein and the TrfA monomeric mutant TrfA 254D/267L. After purification, the 33-kDa TrfA wild-type protein is found mainly in a dimeric form. However, it is the TrfA monomeric form that binds iterons within the RK2 replication origin and initiates DNA replication (30 -32). The His6-TrfA 254D/267L mutant protein contains two amino acid replacements resulting in a high copy RK2 mutant phenotype in vivo. The mutant His6-TrfA 254D/267L protein is fully functional in vivo (40) and in vitro (21,26,33,(41)(42)(43)(44)(45), and is present largely in the monomeric form (27). It was found that the replication activity of wild-type TrfA in crude extracts prepared from the E. coli clpB mutant strain was significantly reduced in comparison to the activity of monomeric TrfA 254D/267L (33). Experiments performed in the course of this work are consistent with the previous observation. The addition of the purified ClpB chaperone to the replication mixture containing dimeric TrfA protein increased the level of replication 3-fold as compared with the reaction without ClpB (Fig. 2A). The observed level of replication activity of wild-type TrfA in the absence of ClpB protein is due to activation by the ClpX chaperone that is present in the bacterial extract (33). Also, possibly other not yet identified chaperones that are present in the extract, bring about a limited amount of activation of the wild-type TrfA protein. In contrast, the TrfA 254D/267L-dependent replication was not activated by the addition of ClpB. Instead, a slight reduction of DNA synthesis was observed (Fig. 2B). The experiments with both wild-type TrfA and monomeric TrfA 254D/ 267L were performed at low and high TrfA concentrations (Fig.  2). The results obtained were very similar, showing that ClpBdependent stimulation of replication is observed only with wild-type TrfA regardless of its concentration.
Activation of TrfA Requires Both ClpB and the DnaK/DnaJ/ GrpE Chaperone Machinery-The crude Fraction II extract used for oriC and RK2 replication contains not only DNA replication proteins but also other cellular components including ClpX and other molecular chaperones. To investigate whether ClpB acts by itself or in combination with other intracellular factors, we measured the activation of TrfA in a system reconstituted with purified proteins. It was previously published that the replication activity of TrfA can be determined by analyzing FI* formation, an extensively unwound covalently closed circular DNA replication intermediate (26). In a control reaction when TrfA 254D/267L was used, FI* formation was easily detectable in the absence of chaperone proteins (Fig. 3,  lane 2). The addition of dimeric wild-type TrfA did not result in template conversion to the FI* form (Fig. 3, lane 3). Surprisingly, the addition of the ClpB chaperone to this reaction mixture did not facilitate FI* formation indicating that additional cellular components are required to activate TrfA.
Recently, it has been demonstrated that ClpB cooperates with the DnaK/DnaJ/GrpE chaperone machinery in the refolding of stable protein aggregates (7,8). These results raised the possibility that ClpB might cooperate with the DnaK, DnaJ, and GrpE chaperones in plasmid RK2 replication activation. The addition of purified ClpB, DnaK, DnaJ, and GrpE chaperones to the reaction mixture containing wild-type TrfA, oriV template, and the other required components resulted in FI* FIG. 2. ClpB increases the replication activity of dimeric wildtype TrfA but not monomeric TrfA 254D/267L. A, the wild-type His6-TrfA in the low amount of 200 ng (filed circles) or the high amount of 2 g (open circles) was tested for replication activity in crude Fraction II extract prepared from the E. coli clpB mutant. Reactions were carried out as described under "Materials and Methods" except that the extracts were supplemented with the indicated amounts of ClpB. Stimulation of DNA synthesis in the reactions was calculated relative to nucleotide incorporation obtained in the reaction without ClpB protein i.e. 50 pmol with wild-type His6-TrfA at 200 ng or 169 pmol with wild-type His6-TrfA at 2 g. The data has been normalized so that 1 represents the level of DNA synthesis measured in the absence of ClpB. B, the monomeric mutant His6-TrfA 254D/267L in the low amount of 50 ng (filed squares) and the high amount of 500 ng (open triangles) was tested as described for wild-type His6-TrfA. Stimulation of DNA synthesis was calculated relative to nucleotide incorporation obtained in the reactions without ClpB protein, i.e. 290 pmol with His6-TrfA 254D/ 267L at 50 ng or 600 pmol and with His6-TrfA 254D/267L at 500 ng).
formation, presumably because of activation of the TrfA protein (Fig. 3, lane 6). The addition of ClpB alone or the DnaK, DnaJ, and GrpE chaperones alone was not sufficient for TrfA activation (Fig. 3, lanes 4 and 5).
We characterized in detail the requirements for individual chaperone machinery components in the TrfA activation reaction. Experiments with omissions of one or two chaperone proteins were performed (Fig. 4). Clearly, the full system, using all four chaperones was the most efficient (Fig. 4, lanes 6 and  11). FI* formation was slightly reduced when GrpE was omitted from the reaction (Fig. 4, lane 5). The absence of DnaJ resulted in a large reduction of FI* formation but did not abolished it (Fig. 4, lane 4). In contrast, the presence of both DnaK (Fig. 4, lane 3) and ClpB (Fig. 3, lane 5) is indispensable for TrfA activation. In the absence of both DnaJ and GrpE, reduced TrfA activation was observed (Fig. 4, lane 8). These results clearly show that ClpB and DnaK are the most crucial components; however, TrfA activation is most efficient when the ClpB, DnaK, DnaJ, and GrpE proteins are all present.
We also tested whether the E. coli DnaK/DnaJ/GrpE chaperone machinery affects RK2 replication in crude extracts. The addition of antibodies against DnaK substantially reduced the RK2 replication using wild-type TrfA and bacterial crude extracts (Fig. 3B). In contrast the same reaction with monomeric TrfA 254D/267L was not significantly affected (data not shown). Addition of DnaK protein substantially compensates the inhibitory effect of DnaK antibody (Fig. 3B).
Activation of TrfA Is Specific for Bacterial Chaperone Proteins-RK2 is a broad-host-range plasmid able to transfer and replicate its DNA in a wide spectrum of Gram-negative bacteria (19). In contrast, RK2 DNA can be transformed to, but is not stably maintained in, Gram-positive bacteria and Saccharomyces sp. (50). Therefore, it is an attractive candidate system for investigating the substrate specificity of various chaperones from various organisms. Purification of the Hsp78, Ssc1, Mdj1, and Mge1 yeast mitochondrial homologs of E. coli ClpB, DnaK, DnaJ, and GrpE proteins, allowed us to address the question regarding host specificity for the chaperone-mediated activation of TrfA protein. We analyzed FI* formation after addition of Saccharomyces cerevisiae chaperone proteins to the reaction with wild-type TrfA. The results presented in Fig. 5, showed that in contrast to the E. coli system, the yeast Hsp78/Ssc1/ Mdj1/Mge1 homologous system was not sufficient for TrfA activation (compare lanes 3 and 4). In contrast to the TrfA activation, both the bacterial and yeast chaperone proteins were capable of reactivation of chemically denatured luciferase (data not shown).
The ClpB/DnaK/DnaJ/GrpE Chaperone System Activates TrfA by Converting Inactive Dimers to Active Monomers-The experiments using the analysis of FI* formation clearly showed that the ClpB/DnaK/DnaJ/GrpE system activates RK2 DNA replication initiated by the largely dimeric TrfA protein. Moreover, it was shown that only dimers of wild-type TrfA, but not monomeric form (TrfA 254D/267L) require chaperone activation. Using glycerol gradient sedimentation we analyzed the TrfA quaternary structure before and after incubation with ClpB, DnaK, DnaJ, and GrpE. In control experiments, in the absence of chaperone proteins, TrfA was detected in fractions corresponding to the position of the dimeric form of the protein (approximate molecular mass of 74.8 kDa) (Fig. 6A). When TrfA was incubated in the presence of the ClpB/DnaK/DnaJ/GrpE FIG. 3. Both ClpB and the DnaK/DnaJ/GrpE chaperone machinery are required for TrfA activation. A, TrfA replication activity was determined using a reconstituted purified protein system and a helicase-dependent DNA unwinding assay with the supercoiled plasmid DNA template pTJS42 as described under "Materials and Methods." Formation of FI* (marked by a black arrow), an extensively unwound covalently closed circular DNA form, depends upon TrfA-dependent origin opening and helicase loading. All reactions contained E. coli DnaA, DnaB, DnaC, gyrase, SSB, HU, and creatine kinase. Proteins His6-TrfA (1 g), His6-TrfA 254D/267L (500 ng), ClpB (2.5 g), DnaK (2.2 g), DnaJ (400 ng), and GrpE (800 ng) were added as indicated. B, the effects of the addition of polyclonal DnaK antibodies (serum) on oriV DNA replication in crude extract prepared from wild-type E. coli was determined to assess the role of the DnaK protein in TrfA activation. A wild-type His6-TrfA (2 g) was used in the experiment. Reactions were carried out under the conditions described under "Materials and Methods" except that anti-DnaK antibodies and purified DnaK protein (1.5 g) were added to the replication reactions when indicated. chaperones prior to centrifugation, the TrfA protein was found in fractions corresponding to the size of a monomer (approximate molecular mass of 37.4 kDa) since very little, if any, TrfA protein was detected in the fraction corresponding to that of the TrfA dimer (Fig. 6B). The conversion of TrfA dimers to monomers appears to be very efficient. After sedimentation, any residual protein in the bottom of each tube was solubilized with 10% SDS to analyze for the presence of aggregated proteins. Insignificant amounts of pelleted protein were found (data not shown). DISCUSSION A surprising observation made during the past few years is that many of the plasmid-encoded proteins that specifically recognize the origin of replication bind to their corresponding DNA sites as monomer (51). Yet, the purified protein is found mainly in the inactive dimeric form. The pioneering studies of Wickner et al. (34) demonstrated in vitro that the DnaK/DnaJ/ GrpE chaperone machinery is capable of monomerization of the dimeric RepA replication protein of plasmid P1. These studies were motivated by earlier in vivo observations, demonstrating that mutations in either dnaK, dnaJ, or grpE resulted in plasmid P1 instability (52,53). Subsequent work demonstrated that in addition to DnaK/DnaJ/GrpE the ClpA chaperone alone is capable of monomerizing the P1 RepA protein (12). Our previous studies had uncovered an analogous reaction, i.e. the monomerization of the RK2 replication protein. In our original study the ClpX chaperone protein was shown to be sufficient for the monomerization of the TrfA replication protein (33). An accompanying result in this study was the fact that the clpX deletion mutation was unaffected in the maintenance of the RK2 plasmid. The work described here was motivated not only by this observation, but also by our previous demonstration of a substantial reduction in TrfA activity in replication assays performed in crude extracts prepared from an E. coli clpB mutant strain (33). Interestingly, suggestions of the possible involvement of ClpB in DNA replication come also from investigations of the bacteriophage Mu replication system. 1 Our results clearly showed that indeed ClpB is involved in TrfA activation. ClpB was sufficient to stimulate RK2 replication in crude extracts prepared from clpB mutant that also contained other chaperones ( Figs. 1 and 2). Experiments with a fully reconstituted system, where all components were added in purified form (Figs. 3A and 4), revealed that in addition to ClpB other factors are required in TrfA activation. We found that, in addition to ClpB, the DnaK, DnaJ, and GrpE chaperones are necessary for this reaction. The TrfA activation was shown to be the consequence of the conversion of its dimeric form to an active monomeric one. In agreement with this conclusion the TrfA 254D/267L mutant did not require the presence of either ClpB or the DnaK/DnaJ/GrpE chaperones for activity in the in vitro replication system.
Our results showing that these two chaperone systems cooperate are consistent with recently published observations. It was demonstrated that in vitro E. coli ClpB together with DnaK, DnaJ, and GrpE facilitate the disaggregation of stable protein aggregates (7)(8)(9)(10). The exact role of particular proteins in this bi-chaperone system as well as molecular events in the reaction are not clear. A sequential mechanism with ClpB initially binding to the substrate protein, followed by Hsp70 protein action, was postulated (7). The function of ClpB has been defined to some extent. Based on the growth defect phenotype of E. coli clpB mutants, it has been proposed that ClpB is a factor involved in thermotolerance (3,4). Thermally aggregated proteins were shown to be dissolved in wild-type cells but not in a clpB mutant strain upon shifting the cells to lower temperature (5,6).
Our observation that the ClpB/DnaK/DnaJ/GrpE bi-chaperone system plays a role in plasmid RK2 DNA replication initiation defines a novel function of ClpB. Thus, the ClpB/DnaK/ DnaJ/GrpE system is involved not only in solubilization and refolding of protein aggregates induced by heat or other stress conditions (6), but also, like certain other chaperone systems (12,13,33,34,37,51,53), it is involved in the activation of a DNA replication factor. Detailed analysis showed that all four proteins from the bi-chaperone system are needed for efficient TrfA activation; however, limited activation could still be obtained when GrpE and/or DnaJ were omitted (Fig. 4). Thus, the ClpB and DnaK chaperones are the key elements of the activation system. Interestingly, in contrast to P1 plasmid replication, the Hsp70 machinery alone is not able to activate the RK2 1 T. Baker, personal communication.
FIG. 6. The ClpB/DnaK, DnaJ, GrpE chaperone system activates TrfA by converting inactive dimers to active monomers. Glycerol gradient analytical centrifugation was performed as described under "Materials and Methods" after incubation of dimeric wild-type His6-TrfA (10 g) with (B) or without (A) ClpB (15 g), DnaK (11 g), DnaJ (2 g), and GrpE (4 g). After centrifugation, portions of collected fractions were analyzed by SDS-PAGE electrophoresis followed by Western blot analysis with polyclonal anti-TrfA antibodies. the arrow indicates the position of the DnaK monomer (70 kDa) used as a standard during sedimentation analysis. A wild-type His6-TrfA protein as a molecular weight marker is shown on the far right lane. replication initiation protein. Also, TrfA could not be converted to the active form solely by the action of ClpB. This reveals an important difference between the ClpB and ClpX chaperone function.
The ClpB/DnaK/DnaJ/GrpE system displays specificity for the RK2 plasmid replication system. We did not observe a significant effect of ClpB on replication initiation from oriC. Previous genetic and biochemical evidence had suggested that molecular chaperones could be involved in bacterial chromosomal DNA replication. For example, it had been demonstrated in vitro that either DnaK or GroEL/ES chaperones, in the presence of ATP, can protect the DnaA protein from aggregation. Furthermore, DnaA aggregates could be dissociated by DnaK, thus allowing the initiation of oriC DNA replication in vitro (54,55). In our experiments we observe only limited reduction in oriC-dependent DNA replication in crude Fraction II extracts prepared from the E. coli clpB mutant. This suggests that the ClpB chaperone is not an essential factor for DnaA protein activation. Further investigations must be performed to verify this preliminary conclusion.
It is likely that the TrfA protein, as the replication initiation factor of the broad-host-range plasmid RK2, requires the chaperone system for activation in diverse bacterial species. It has been proposed that TrfA activation in so many distantly related species can be performed by redundant chaperone systems (33). We further speculate that in E. coli ClpB-dependent TrfA activation plays a more important role than ClpX-dependent activation since replication in FII extract prepared from a clpB-deleted strain is more impaired than that from a clpX deletion strain (33). The chaperone requirements for activation of TrfA possibly vary depending on the bacterial species. On the other hand homologous chaperone systems isolated from diverse organisms can perform the same biological reactions utilizing the same substrates. For example, heat-or chemically denatured firefly luciferase is a substrate for reactivation by the E. coli ClpB/DnaK/DnaJ/GrpE, yeast mitochondrial Hsp78/ Ssc1p/Mdj1p/Mge1p chaperones as well as cytosolic Hsp104/ Ssa1p/Ydj1p (8,39,49). Our results show that TrfA is activated by bacterial but not by yeast chaperones. Interestingly, RK2 DNA can be transferred to, but not maintained, in Grampositive bacteria and Saccharomyces sp. (50).