The N Terminus of ClpB from Thermus thermophilus Is Not Essential for the Chaperone Activity

Abstract

ClpB from Thermus thermophilusbelongs to the Clp/Hsp100 protein family and reactivates protein aggregates in cooperation with the DnaK chaperone system. The mechanism of protein reactivation and interaction with the DnaK system remains unclear. ClpB possesses two nucleotide binding domains, which are essential for function and show a complex allosteric behavior. The role of the N-terminal domain that precedes the first nucleotide binding domain is largely unknown. We purified and characterized an N-terminal shortened ClpB variant (ClpBΔN; amino acids 140–854), which remained active in refolding assays with three different substrate proteins. In addition the N-terminal truncation did not significantly change the nucleotide binding affinities, the nucleotide-dependent oligomerization, and the allosteric behavior of the protein. In contrast casein binding and stimulation of the ATPase activity by κ-casein were affected. These results suggest that the N-terminal domain is not essential for the chaperone function, does not influence the binding of nucleotides, and is not involved in the formation of intermolecular contacts. It contributes to the casein binding site of ClpB, but other substrate proteins do not necessarily interact with the N terminus. This indicates a substantial difference in the binding mode of κ-casein that is often used as model substrate for ClpB and other possibly more suitable substrate proteins. ClpB from Thermus thermophilusbelongs to the Clp/Hsp100 protein family and reactivates protein aggregates in cooperation with the DnaK chaperone system. The mechanism of protein reactivation and interaction with the DnaK system remains unclear. ClpB possesses two nucleotide binding domains, which are essential for function and show a complex allosteric behavior. The role of the N-terminal domain that precedes the first nucleotide binding domain is largely unknown. We purified and characterized an N-terminal shortened ClpB variant (ClpBΔN; amino acids 140–854), which remained active in refolding assays with three different substrate proteins. In addition the N-terminal truncation did not significantly change the nucleotide binding affinities, the nucleotide-dependent oligomerization, and the allosteric behavior of the protein. In contrast casein binding and stimulation of the ATPase activity by κ-casein were affected. These results suggest that the N-terminal domain is not essential for the chaperone function, does not influence the binding of nucleotides, and is not involved in the formation of intermolecular contacts. It contributes to the casein binding site of ClpB, but other substrate proteins do not necessarily interact with the N terminus. This indicates a substantial difference in the binding mode of κ-casein that is often used as model substrate for ClpB and other possibly more suitable substrate proteins. ClpB from Thermus thermophilus is a member of the AAA protein superfamily that is important for a variety of biological activities (1Neuwald A.F. Aravind L. Spouge J.L. Koonin E.V. Genome Res. 1999; 9: 27-43Google Scholar). Despite their different cellular functions AAA proteins (ATPases associated with a variety of cellularactivities) employ a general mechanism. They mediate the assembly and disassembly of large protein complexes that are involved in processes like DNA replication, vesicle transport, or organelle biogenesis. The AAA protein superfamily comprises the Clp/Hsp100 proteins with its members ClpA, ClpX, and ClpY on the one hand and ClpB on the other. ClpA, ClpX, and ClpY (HslU) interact with cellular peptidases to form ATP-dependent proteases. The role of these Clp proteins is the unfolding of substrates and the delivery of unfolded polypeptides to the protease subunit. In contrast ClpB from Escherichia coli and T. thermophilus and theSaccharomyces cerevisiae homologue Hsp104 do not appear to bind to cellular proteases. Instead they interact with the DnaK/Hsp70 chaperone system to assist the disaggregation of protein aggregates (2Glover J.R. Lindquist S. Cell. 1998; 94: 73-82Google Scholar, 3Mogk A. Tomoyasu T. Goloubinoff P. Rudiger S. Roder D. Langen H. Bukau B. EMBO J. 1999; 18: 6934-6949Google Scholar, 4Motohashi K. Watanabe Y. Yohda M. Yoshida M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7184-7189Google Scholar, 5Ben Zvi A.P. Goloubinoff P. J. Struct. Biol. 2001; 135: 84-93Google Scholar, 6Zolkiewski M. J. Biol. Chem. 1999; 274: 28083-28086Google Scholar). Clp proteins are further classified according to the number of nucleotide binding domains (NBD), 1The abbreviations used are: NBD, nucleotide binding domain; DTE, dithioerythritol; MANT, 2′(3′)-O-N′-methylanthraniloyl; N-domain, N-terminal domain; C-domain, C-terminal domain; FITC, fluorescein isothiocyanate; BSA, bovine serum albumin; LDH, lactate dehydrogenase; MOPS, 3-(N-morpholino)propanesulfonic acid which are 220–250 amino acids in length and show a high level of sequence homology. Class two proteins contain only one NBD whereas the class one members ClpA and ClpB contain two NBDs. In ClpB and Hsp104 ATP hydrolysis at both NBDs was shown to be necessary for the chaperone activity of the Clp proteins (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar, 8Parsell D.A. Kowal A.S. Lindquist S. J. Biol. Chem. 1994; 269: 4480-4487Google Scholar). Structural information on the class two protein HslU (ClpY) and the hexamerization domain (D2) of N-ethylmaleimide-sensitive fusion protein give insight into the domain architecture of AAA proteins (9Bochtler M. Hartmann C. Song H.K. Bourenkov G.P. Bartunik H.D. Huber R. (2000) Nature. 2000; 403: 800-805Google Scholar, 10Sousa M.C. Trame C.B. Tsuruta H. Wilbanks S.M. Reddy V.S. McKay D.B. Cell. 2000; 103: 633-643Google Scholar, 11Lenzen C.U. Steinmann D. Whiteheart S.W. Weis W.I. Cell. 1998; 94: 525-536Google Scholar). These crystal structures show ring-shaped oligomers composed of six monomers. The AAA modules of both proteins show a similar overall structure, consisting of the core NBD and a C-terminal, mostly helical, domain. The core NBD contains the Rossman fold including the phosphate binding loop of ATP- and GTP-binding proteins, which consists of five central β-sheets flanked by α-helices. In the case of HslU an additional helical domain is inserted in the NBD that is not present in other Clp proteins. In ClpBEco the C-terminal domain of the second AAA cassette was shown to be important for oligomerization, ATPase activity, and chaperone function (12Barnett M.E. Zolkiewska A. Zolkiewski M. J. Biol. Chem. 2000; 275: 37565-37571Google Scholar). Isolated C-domains of ClpA, ClpX, and ClpY, which associate with peptidases, were found to interact with different substrate proteins in vivo. Therefore the C-terminal domains were proposed to be sensor and substrate discrimination domains although the C-terminal domains show only little sequence homology between different Clp proteins (13Smith C.K. Baker T.A. Sauer R.T. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 6678-6682Google Scholar). The class one Clp proteins have, in addition, an N-terminal domain that precedes the first NBD. The ClpA and ClpB mRNAs of E. coli contain internal translation initiation sites and are expressed in vivo as two gene products: the full-length proteins and shortened versions lacking the N-terminal domain (14Squires C.L. Pedersen S. Ross B.M. Squires C. J. Bacteriol. 1991; 173: 4254-4262Google Scholar, 15Seol J.H. Yoo S.J. Kim K.I. Kang M.S. Ha D.B. Chung C.H. J. Biol. Chem. 1994; 269: 29468-29473Google Scholar). The N-terminal domains are supposed to consist of two ∼75-amino acids-long repeats that can fold independently (16Lo J.H. Baker T.A. Sauer R.T. Protein Sci. 2001; 10: 551-559Google Scholar, 17Tek V. Zolkiewski M. Protein Sci. 2002; 11: 1192-1198Google Scholar). For ClpAEco it was shown that the N-terminal domain is associated only weakly with the rest of the molecule, and a mutant lacking the N-terminal domain showed a reduced unfolding activity depending on the substrate protein tested (16Lo J.H. Baker T.A. Sauer R.T. Protein Sci. 2001; 10: 551-559Google Scholar, 18Singh S.K. Rozycki J. Ortega J. Ishikawa T. Lo J. Steven A.C. Maurizi M.R. J. Biol. Chem. 2001; 276: 29420-29429Google Scholar). Recently an interaction of the N terminus with the proposed ClpA modulator protein ClpS was reported (19Dougan D.A. Reid B.G. Horwich A.L. Bukau B. Mol. Cell. 2002; 9: 673-683Google Scholar). Similar results were reported for the N-domain of ClpBEco, which did not interact strongly with the shortened ClpB variant (17Tek V. Zolkiewski M. Protein Sci. 2002; 11: 1192-1198Google Scholar). The N-terminally shortened ClpB from E. coli was shown to be inactive as a chaperone but retained its oligomerization capacity and ATPase activity in vitro (12Barnett M.E. Zolkiewska A. Zolkiewski M. J. Biol. Chem. 2000; 275: 37565-37571Google Scholar). The shortened protein version, however, can restore bacterial thermo-tolerance in vivo (20Clarke A.K. Eriksson M.J. J. Bacteriol. 2000; 182: 7092-7096Google Scholar). In this study we focused on the function of the N-terminal domain of ClpB from T. thermophilus and its interactions with substrate proteins. Assigning the individual functions to the different ClpB domains can help to reveal the currently unknown mechanism of protein disaggregation by the ClpB/DnaK system. The DNA fragment encoding ClpBΔN (amino acids 140–854) was generated by PCR using the pRS-ClpB plasmid as a template (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). The forward primer was GACTGCTAAGCATATGGTGCAGACGGAACACGCGGAAAGC, and the reverse primer was (T7 terminator) CTAGTTATTGCTCAGCGGTGGC. The PCR product was subcloned between the NdeI and EcoRI restriction sites of NdeI/EcoRI digested pET27b(+) vector (Novagen). The integrity of the sequence generated by PCR amplification was verified by DNA sequencing. ClpB and ClpBΔN were expressed in E. coli BL21(DE3) cells. The purification was performed according to the protocol for the wild type (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). DnaK, DnaJ, and GrpE from T. thermophilus were purified as described previously (21Klostermeier D. Seidel R. Reinstein J. J. Mol. Biol. 1999; 287: 511-525Google Scholar). Protein concentration was measured using the method of Ehresmann et al. (22Ehresmann B. Imbault P. Weil J.H. Anal. Biochem. 1973; 54: 454-463Google Scholar). If not noted otherwise all protein concentrations refer to monomers. CD spectra were recorded with a Jasco-J710 spectropolarimeter in 0.02-cm cuvettes at a scan rate of 20 nm/min, 0.2-nm resolution, 1-nm bandwidth, a time constant of and a of 20 were using an at (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). The were the ClpB and the were with the using J. P. A. Scholar). The of by were to a Reinstein J. B.M. Scholar). of ClpB to was using an at The was and the was of protein were to in DTE, the the of as the for the was was for at in DTE, BSA, For the was into DTE, BSA, ClpB or ClpBΔN and DnaK GrpE The refolding was measured at in with a of product the activity in the Y. D. C. T. Seidel R. Reinstein J. J. Mol. Biol. 2001; Scholar). from was used as described previously K. Watanabe Y. Yohda M. Yoshida M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7184-7189Google with the was at in the were or ClpBΔN and DnaK GrpE to refolding at activity was at the time at in the using the from B. was at in the K. Watanabe Y. Yohda M. Yoshida M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7184-7189Google Scholar). were or ClpBΔN and DnaK GrpE to refolding at activity was at the time at in the 20 using the ATP hydrolysis was measured with a H. using the (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). ClpB or ClpBΔN was at with different concentrations of in DTE, 20 LDH, For casein stimulation κ-casein was to a concentration of The rate were from the the first of The were with using the shown binding (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). ATP hydrolysis was measured in a H. using the (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). ClpB or ClpBΔN was at in with of or were were performed using a with a system (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). ClpB or ClpBΔN was with a rate of 20 DTE, The from the not can be to an The for the ClpB and ClpBΔN oligomers were from their by The of ClpB was by a from to the for the were to the for the were to and the for the oligomers were to ClpB from T. thermophilus a high sequence homology to the E. coli protein K. Watanabe Y. Yohda M. Yoshida M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7184-7189Google Scholar). It contains a by a to the of ClpBEco, it is used as an internal translation (14Squires C.L. Pedersen S. Ross B.M. Squires C. J. Bacteriol. 1991; 173: 4254-4262Google Scholar, Kim K.I. A.L. Ha D.B. Chung C.H. J. Biol. Chem. Scholar). in E. coli form of the was found in study the of the ClpB from T. thermophilus we and purified a protein variant lacking the first amino which to the N-terminal domain of ClpBEco The of the N-terminal on the of the protein was by CD The CD of ClpBΔN is similar to the of full-length It the two at and which are for proteins. This indicates that the of the N-terminal domain did not influence the of the protein significantly and that both proteins are The oligomerization behavior of ClpB and ClpBΔN was by at with the full-length ClpBΔN a similar oligomerization behavior at the protein concentration ATP and the formation of oligomers to whereas the binding of and of the influence of the nucleotide and the of the a between the two proteins. The results are in In the of ATP and both proteins form oligomers ClpB, In the of and high concentrations the is the ClpB, ClpBΔN more oligomers in the of and to with ClpB ClpB, These that the N terminus is not essential for the oligomerization of ClpB in the protein concentration of ClpB and of the oligomerization of ClpB and ClpBΔN depending on nucleotide and The for ClpB are and the for ClpBΔN are in The different concentrations of and the The give the of a for the according to the for ClpB are from S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google in a of the oligomerization of ClpB and ClpBΔN depending on nucleotide and The for ClpB are and the for ClpBΔN are in The different concentrations of and the The give the of a for the according to the for ClpB are from S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google study the influence of the N terminus on nucleotide binding we the binding of and the nucleotide and for ClpB and ClpBΔN at T. Scholar). The of the were by and the for the were by (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). The results are in for ClpB and of and the nucleotide to ClpB and ClpBΔN at The of the were by and for the were by of ATP and was as binding of the ATP hydrolysis of the proteins and the of the of and ATP for ClpB are from S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google in a of and the nucleotide to ClpB and ClpBΔN at The of the were by and for the were by of ATP and was as binding of the ATP hydrolysis of the proteins and the of the of and ATP for ClpB are from S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google the of the N terminus did not influence the nucleotide binding The proteins for and the binding for is The and ATP are one of the The for are similar and ATP is with of for the full-length protein and the ATPase activity the used is the for ATP a and can be (7Schlee S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). the influence of the N-domain on the ATPase activity, were performed at the rate the ATP were for ClpB and ClpBΔN in the ATP from to ClpB of ATP hydrolysis and ATPase stimulation in the of ClpBΔN a ATPase activity which is weakly by The were with the behavior of the proteins The of ClpB is by κ-casein whereas the of ClpBΔN is only by The in the of κ-casein is in a similar for the two proteins, which be the ATPase of the at the The allosteric behavior is not by the N-terminal which is by the similar for the two proteins the ATP hydrolysis and in the of casein for both This that the N-domain is not involved in intermolecular important for the allosteric of the ATPase activity using the with ClpB and ClpBΔN were performed at In the ATP used to a is for both proteins the rate the ATP casein are and in the of casein are in The were using the binding in a with ClpB and ClpBΔN were performed at In the ATP used to a is for both proteins the rate the ATP casein are and in the of casein are in The were using the binding In the we the the stimulation of ClpBΔN by κ-casein is of a of casein binding or of reduced of casein binding and ATP We used casein as an interaction of we tested this casein is to the ATPase of ClpB and ClpBΔN similar to ClpB and ClpBΔN were at with concentrations of in the of ATP The ATPase activity of ClpB is and the ATPase of ClpBΔN is only by This stimulation is similar to the stimulation by κ-casein Therefore the ATPase activities of ClpB and ClpBΔN to as they to The binding of ClpB and ClpBΔN to was by the change chaperone addition addition of ClpB the is more addition of ClpBΔN and a The addition of ClpBΔN is to the of protein and does not of a was not of the of casein and the binding of It however, that ClpBΔN with a with ClpB and that the reduced ATPase stimulation of ClpBΔN by casein is with a reduced protein was used as a reduced binding and ATPase stimulation of and κ-casein are general of the ClpB form we used which was shown to the ATPase of ClpBEco and Hsp104 (17Tek V. Zolkiewski M. Protein Sci. 2002; 11: 1192-1198Google Scholar, D.A. J.R. M.S. Lindquist Mol. Cell. 2002; 9: Scholar). ClpB and ClpBΔN were at with concentrations of in the of ATP ClpB and ClpBΔN to In both proteins the ATPase activities are by a of in the of This indicates that the ATPase activity of ClpBΔN can be and that both proteins bind the reduced binding for casein is with a of function we tested the refolding capacity of ClpBΔN in cooperation with the DnaK chaperone system. LDH, and were used as model substrate proteins for the chaperone of the substrates the were the The and substrate proteins were in the of ClpB and the DnaK ClpBΔN and the DnaK the DnaK system or from B. can be by the chaperone system at for at The refolding to a full-length ClpB of the activity was ClpBΔN of the wild type activity in this The of ClpB is large as shown by the refolding of the DnaK system which is of the activity in the of In the of did not to a from B. the second substrate protein is more for at of the activity was ClpB show a similar refolding activity of of the refolding at The of ClpB is the DnaK system can to of the at the refolding in activity was is a substrate of the system (2Glover J.R. Lindquist S. Cell. 1998; 94: 73-82Google Scholar, 6Zolkiewski M. J. Biol. Chem. 1999; 274: 28083-28086Google Scholar, S. Groemping Y. Herde P. Seidel R. Reinstein J. (2001) J. Mol. Biol. 2001; 306: 889-899Google Scholar). We used as the substrate protein for the system of by ClpB DnaK system. was into ClpB and the DnaK system ClpBΔN and the the DnaK system or was at was for by in the of For refolding the was into refolding In the of the DnaK system activity was a of In the of ClpB or ClpBΔN and the DnaK system the of and to a significantly refolding of the activity showed that The activity of in a of product Y. D. C. T. Seidel R. Reinstein J. J. Mol. Biol. 2001; Scholar). The functions of the different domains of Clp proteins by a number of the role of the N terminus was of an internal translation initiation site and the two of ClpB and ClpA present in E. coli (16Lo J.H. Baker T.A. Sauer R.T. Protein Sci. 2001; 10: 551-559Google Scholar, 17Tek V. Zolkiewski M. Protein Sci. 2002; 11: 1192-1198Google Scholar). Similar to results for ClpBEco and ClpAEco the of ClpBΔN from T. thermophilus is not by the N-terminal we show that the of the first amino acids did not the nucleotide binding of the which can be by interactions of the N-terminal domain with other domains of the molecule, with the two NBDs. interactions between the N-domain and the shortened ClpB were in the E. coli protein (17Tek V. Zolkiewski M. Protein Sci. 2002; 11: 1192-1198Google Scholar). In addition the N-terminally shortened protein showed a similar oligomerization behavior depending on and the In the of ATP and the proteins into which are at of these oligomers K. Yoshida M. J. Biol. Chem. 2002; Scholar). Therefore the N-domain does not appear to intermolecular necessary for the This a of the N terminus at the of the a of the N-domain was for ClpA from E. coli S.K. Rozycki J. Ortega J. Ishikawa T. Lo J. Steven A.C. Maurizi M.R. J. Biol. Chem. 2001; 276: 29420-29429Google Scholar). The similar allosteric behavior in ATPase activity of ClpB and ClpBΔN that the N terminus is not involved in both proteins used in this study similar This that the N-domain is not involved in intermolecular that allosteric The ATPase activity of ClpBΔN at a function of the In addition the formation of oligomers by ClpBΔN as from to the ATPase ClpB a stimulation of the ATPase activity by κ-casein with which is in with results from (12Barnett M.E. Zolkiewska A. Zolkiewski M. J. Biol. Chem. 2000; 275: 37565-37571Google Scholar, S.K. Kim K.I. J.H. K. A. Ha D.B. Chung C.H. J. Biol. Chem. Scholar). We were to that this reduced stimulation of the ATPase activity is of a reduced binding for casein by ClpBΔN and does not a of The of the N terminus does not in a stimulation of the ATPase activity as ClpB and ClpBΔN the reduced stimulation of the ATPase activity is to a used model substrate K. Watanabe Y. Yohda M. Yoshida M. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 7184-7189Google Scholar, M.E. Zolkiewska A. Zolkiewski M. J. Biol. Chem. 2000; 275: 37565-37571Google Kim K.I. A.L. Ha D.B. Chung C.H. J. Biol. Chem. Scholar). chaperone activity, the of the N-terminal truncation remained The form was to restore thermo-tolerance in in vivo (20Clarke A.K. Eriksson M.J. J. Bacteriol. 2000; 182: 7092-7096Google Scholar). In the N-terminal shortened variant of ClpBEco was reported to be inactive in vitro in a refolding (12Barnett M.E. Zolkiewska A. Zolkiewski M. J. Biol. Chem. 2000; 275: 37565-37571Google Scholar). We the in vitro with ClpBΔN We refolding capacity for and and a reduced refolding for The reduced refolding activity for one substrate protein is in with with ClpAEco (16Lo J.H. Baker T.A. Sauer R.T. Protein Sci. 2001; 10: 551-559Google which show that the form can interact with and substrate proteins at rate with the full-length Therefore we that the N-terminal domain from ClpB is in general not essential for disaggregation and refolding by the system. It the ATPase activity and an function the interaction with proteins or only be necessary for a of substrate proteins. to this class of substrates that possibly a The of casein that were used as model substrate for ClpB are not for other substrate proteins. This is in with different of in the N-domain of ClpBEco the ATPase which were reported V. V. Zolkiewski M. J. Mol. Biol. 2002; Scholar). were by but all tested to and were to bind the N terminus of ClpAEco was shown to interact with the substrate modulator ClpS (19Dougan D.A. Reid B.G. Horwich A.L. Bukau B. Mol. Cell. 2002; 9: 673-683Google casein interact with the binding site of a modulator at the N terminus of ClpB by an The of the used model substrate casein on ClpB ATPase activity not be for other substrate proteins. In as casein is not a protein and does not for of This an for the study of ClpB/DnaK disaggregation The substrates are in general not which the of of proteins from different of substrate proteins are possibly including to the of protein aggregates as by Goloubinoff and S. A.P. Bukau B. Goloubinoff P. (2000) J. Biol. Chem. 2000; 275: Scholar). In the between the role of the N terminus from ClpB from T. thermophilus and E. coli are of the different or of the individual remains to be The with a of substrate proteins including nucleotide binding and hydrolysis that the N terminus of ClpB is not essential for in We are to S. for and