The immunological dissection of the Escherichia coli molecular chaperone DnaJ

FEMS Microbiology Letters 196 (2001) 19^23

www.fems-microbiology.org

The immunological dissection of the Escherichia coli molecular chaperone DnaJ Salma Al-Herran, William Ashraf * Department of Biomedical Sciences, University of Bradford, Bradford BD7 1DP, UK Received 6 September 2000; received in revised form 3 November 2000; accepted 8 January 2001

Abstract In Escherichia coli, one of the main molecular chaperones is DnaJ (hsp40), which mediates in a variety of highly conserved cellular processes including protein-folding reactions and the assembly/disassembly of protein complexes. DnaJ is characterised by the presence of four distinct domains: the J-domain, glycine/phenylalanine-rich (G/F), cysteine-rich (Zn-finger) and C-terminal regions. Truncated DnaJ polypeptides (DnaJ 1^108, DnaJ v1^108, DnaJ v1^199) representing these domains were over-produced and used as a source of immunogens for the generation of sequence-specific polyclonal antibodies. Epitope mapping was achieved by Western blotting, which demonstrated the presence of antibodies directed against these domains. These characterised affinity-purified antibodies were then used to assess the role of DnaJ in the protection of firefly luciferase from irreversible heat-inactivation. In this study we have demonstrated the involvement of J-, G/F and Zn-finger domains in the protection of luciferase from heat-inactivation. The C-terminal region had only partial involvement in luciferase protection. ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : DnaJ; Escherichia coli; Luciferase

1. Introduction In Escherichia coli the main heat shock proteins, DnaK (hsp70) and DnaJ (hsp40), are highly conserved molecular chaperones and together with GrpE (hsp20) form the `DnaK chaperone machine' [1]. This machine mediates in a variety of highly conserved cellular processes including protein-folding reactions, the assembly/disassembly of protein complexes and the protection of proteins from thermal denaturation. DnaJ is a dimeric chaperone, which binds to nascent polypeptides emerging from ribosomes aiding correct protein folding, and in addition preventing the aggregation of denatured proteins. DnaK binds to DnaJ-substrate forming a stable ternary complex. DnaJ, possibly via the Nterminal QKRAA sequence [2], then stimulates the weak ATPase activity of DnaK and on addition of GrpE the substrate is released. DnaJ will also bind to folded polypeptides, such as c32 and bacteriophage RepA protein (for reviews see [3^5]).

* Corresponding author. Tel : +44 (1274) 233589; Fax: +44 (1274) 309742; E-mail : [email protected]

The hsp40 protein family, including DnaJ, are characterised by the combination of four distinct domains : the J-domain, glycine/phenylalanine-rich (G/F), cysteine-rich (Zn-¢nger) and C-terminal regions (Fig. 1). All hsp40 proteins contain the highly conserved J-domain, corresponding to a 70-amino acid residue present at the amino-terminal portion of the E. coli DnaJ protein [6]. The J-domain is separated by a 30^40-amino acid glycine/phenylalaninerich (over 40% glycine and 15% phenylalanine) G/F region which is thought to act as a £exible spacer that separates the J-domain from other parts of the protein [7]. A recent study, involving spectroscopic analysis of 5-hydroxytryptophan-labelled proteins, has demonstrated that the J-domain is a single well-ordered domain, whilst the G/F region shows a lack of well-ordered structure [8]. The third additional conserved domain is the cysteine-rich region that corresponds to a 150-amino acid stretch at the centre of the protein that has the four repeated sequence motifs, CXXCXGXG, where X denotes a charged or polar amino acid residue. It has been shown that each molecule of E. coli DnaJ binds two zinc ions that are tetrahedrally co-ordinated by the sulfur atoms of four cysteine residues. The cysteine-rich domain, is in fact two Zn ¢ngers that may form a pocket with a hydrophobic core. The

0378-1097 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 0 1 ) 0 0 0 2 8 - 3

FEMSLE 9797 8-3-01

20

S. Al-Herran, W. Ashraf / FEMS Microbiology Letters 196 (2001) 19^23

solution structure of this region has been elucidated by nuclear magnetic resonance and contains a novel V-shaped fold with an extended L-hairpin topology [9]. The C-terminal, representing approximately one third of DnaJ, does not contain any consensus motifs and this region is the least conserved and believed to be involved in substrate binding [10,11]. The DnaK/DnaJ/GrpE chaperone machine suppresses aggregation of both prokaryotic and eukaryotic polypeptides, thus promoting protein folding (for review see [12]). This chaperone machine can also protect partially aggregated enzymes, such as E. coli RNA polymerase, DNA polymerase II and ¢re£y luciferase. DnaJ alone can also protect luciferase and rhodanase from thermal denaturation [13,14]. In this study, characterised sequence-speci¢c polyclonal anti-DnaJ antibodies (Fig. 1) were employed as inhibitors to identify which domains on DnaJ are responsible for its ability to protect and reactivate heat-inactivated luciferase. Previous studies have used randomly generated monoclonal antibodies in order to gain functional understanding of E. coli c32 [15] and DnaK [16]. However, our immunological dissection has been programmed to ensure that antibodies are directed against all DnaJ domains.

(w/v) sucrose. The suspension was then frozen in liquid nitrogen and stored at 380³C until required.

2. Materials and methods

Western blotting was used to epitope map a¤nity-puri¢ed sequence-speci¢c antibodies directed against DnaJ 1^ 108, DnaJ v1^108 and DnaJ v1^199 and also polyclonal antibodies against full-length DnaJ. SDS^PAGE analysis of proteins was carried out using 12.5% (w/v) polyacrylamide gels, which were then electroblotted onto nitrocellulose. Immunological detection of DnaJ polypeptides was achieved using anti-DnaJ antibodies and a horseradish peroxidase conjugate anti-rabbit IgG (Dako) system [18].

2.1. E. coli strains, plasmids and growth conditions E. coli strains and plasmids used in this study are listed in Table 1 [11,17]. Cells were grown aerobically at 37³C in Luria broth (LB) supplemented with 100 Wg ml31 ampicillin. Full-length DnaJ was over-produced from E. coli B178 pMOB45-dnaJ+12 grown in LB supplemented with 30 Wg ml31 chloramphenicol. Truncated DnaJ polypeptides were over-produced using plasmids pDW19-dnaJ12, pAED4-dnaJv1^108 and pAED4-dnaJv1^199 (derivatives of the pET vector) after CaCl2 transformation into competent E. coli BL21 (DE3). Expression of dnaJ was induced at OD600 0.1 by the addition of 50 WM isopropyl L-D-thiogalactopyranoside. Cultures were harvested at OD600 0.8 by centrifugation (7000Ug for 10 min at 4³C) and then resuspended in 50 mM Tris^HCl (pH 8.0), 10%

2.2. Puri¢cation of DnaJ and truncated proteins Protein puri¢cation required a three-step process involving Q-Sepharose0 HP (Pharmacia Amersham Biotech), hydroxyapatite (Bio-Rad) and P11-cellulose phosphate (Whatman) column chromatography. The method was as previously described in [17] except that the ammonium sulfate precipitation step was omitted. 2.3. Generation and puri¢cation of anti-DnaJ antibodies Antisera were raised in female New Zealand white rabbits by administration of 100 Wg puri¢ed DnaJ protein and truncated proteins. The procedure was repeated at regular intervals over a 6-week period. Antibodies were puri¢ed using a Sepharose^DnaJ a¤nity column generated from 1 ml Hi Trap0 NHS-activated a¤nity column according to the manufacturer's instructions (Pharmacia Amersham Biotech). A¤nity-puri¢ed antibodies were evaluated by Western blotting and then stored at 320³C until required. 2.4. Epitope mapping

2.5. Protection of luciferase from irreversible heat-inactivation Fire£y luciferase (Sigma) was stored at 5 Wg ml31 in 1 M (pH 7.4) glycylglycine and kept in aliquots at 320³C. The assay was adapted from a previously published method [19]. Luciferase (250 ng) was inactivated by incubation at 42³C for 10 min in the presence of 1 Wg DnaK, 600 ng DnaJ and 400 ng GrpE. The reactivation mixture (50 Wl)

Table 1 Bacterial strains used in this study Strain

Genotype

Plasmid

Source/reference

E. coli B178

araD139, v(argF-lac) U169, rpsL 150, relA1, deoC1, ptsF25, rpsR, £bB301 F3 dcm, omp,T hsd (rB mB ), gal V (DE3) as above as above as above

pMOB45-dnaJ

[17]

none pAED4-dnaJv1^108 pAED4-dnaJv1^199 pDW19-dnaJ12

Stratagene [11] [11] [11]

E. E. E. E.

coli coli coli coli

BL21 (DE3) BL21(DE3)-DnaJ v1^108 BL21(DE3)-DnaJ v1^199 BL21(DE3)

FEMSLE 9797 8-3-01

S. Al-Herran, W. Ashraf / FEMS Microbiology Letters 196 (2001) 19^23

21

Fig. 1. Structural features of E. coli DnaJ and epitope mapping summary. Adapted from [20].

contained bu¡er A (20 mM Tris^HCl (pH 7.5), 100 mM KCl, 40 mM NaCl, 10 mM MgCl2 , 0.05 mM EDTA, 5 mM DTT). The protection assay was initiated by the addition of 4 mM ATP and then 1 Wl of each reaction was diluted into 60 Wl of luciferase assay mixture (50 Wl of luciferase assay substrate solution (Promega) and 10 Wl of lysis bu¡er reagent (Promega) containing 6 mg ml31 bovine serum albumin (BSA ; Sigma)). Activity was measured using a 1250 Luminometer (LKB-Wallac). Relative luciferase activity was calculated by taking as 100% the activity exhibited by control non-heat-treated luciferase. The results presented are the means of at least three independent experiments. Puri¢ed anti-DnaJ, anti-DnaJ 1^108, anti-DnaJ v1^108 and anti-DnaJ v1^199 were used as inhibitors of the luciferase protection. The procedure was carried out as detailed above except that DnaJ was pre-incubated in the presence of antibody molecules (1:10 molar ratio) for 90 min at 24³C prior to the heat-inactivation. Puri¢ed preimmune sera were also included as a control. 2.6. Protein determination Protein quanti¢cation was performed using the Bradford reagent (0.01% w/v Coomassie brilliant blue, 8.5% v/v phosphoric acid and 5% v/v methanol) supplied by Bio-Rad and used according to the manufacturer's instructions. Samples were measured for protein using an Ultrospec II (LKB, Biochrom) spectrophotometer at 595 nm. Protein concentration was calculated from a standard curve using BSA as a standard.

3. Results and discussion To investigate the functional roles of the conserved DnaJ domains, a series of a¤nity-puri¢ed sequence-specific antibodies were used to provide an alternative, sensitive and speci¢c means of understanding the function. These polyclonal antibodies were used as inhibitors to assess the role of de¢ned DnaJ domains in the protection of luciferase from heat-inactivation. A summary of the epitope mapping results is detailed in Fig. 1. For the sake of brevity the Western blotting data has not been presented. Western blot analysis con¢rmed the production of antibodies against full-length DnaJ and DnaJ polypeptide sequences. Anti-DnaJ v1^108 antibodies only produced a weak signal against DnaJ v1^199 protein and failed to recognise the DnaJ 1^108 (J-domain and G/F domain). This result strongly supports the notion that antibodies raised against DnaJ v1^108 were more speci¢c for the 109^199-amino acid sequence (90 amino acids) which contains the Zn-¢nger-like domain. From these results we speculate that the Zn-¢nger-like domain is potentially immunodominant. Anti-DnaJ v1^199 recognised all DnaJ polypeptides except for DnaJ 1^108. Anti-DnaJ 1^108 Table 2 Epitope mapping results for anti-DnaJ sequence-speci¢c antibodies Sequence-speci¢c antibody Anti-DnaJ 1^108 Anti-DnaJ v1^108 Anti-DnaJ v1^199

FEMSLE 9797 8-3-01

Maximum number of amino acids recognised 108 90 176

Domains J- and G/F Zn-¢ngers C-terminal region

22

S. Al-Herran, W. Ashraf / FEMS Microbiology Letters 196 (2001) 19^23

Fig. 2. Antibody inhibition of the protection of luciferase from irreversible heat-inactivation. Control reactivation reaction mixture (50 Wl) containing bu¡er A (20 mM Tris^HCl (pH 7.5), 40 mM NaCl, 10 mM MgCl2 , 0.05 mM EDTA, 5 mM DTT) with 250 ng of ¢re£y luciferase (Sigma) with or without 600 ng of DnaJ, 1 Wg of DnaK, and 400 ng of GrpE (F). After inactivation at 42³C for 10 min, the reaction was incubated at 24³C with 4 mM ATP. At the times indicated, 1-Wl aliquots were withdrawn, diluted 1:50 and analysed for luciferase activity as described in Section 2. Inhibition experiments were conducted as above, except that the DnaJ was pre-incubated in the presence of antibody molecules (DnaJ:antibody, 1:10 molar ratio) for 90 min at 24³C prior to the heat-inactivation of luciferase. Anti-DnaJ (a), anti-DnaJ 1^108 (P) anti-DnaJ v1^108 (S), anti-DnaJ v1^199 (b). The results presented are the means of at least three independent experiments. Please note that the data points for anti-DnaJ and anti-DnaJ v1^108 partially obscure each other on the scale used.

only recognised DnaJ 1^108 and full-length DnaJ. The maximum number of amino acids and domains recognised by a¤nity-puri¢ed anti-DnaJ sequence-speci¢c antibodies is detailed in Table 2. Antibody titration assays (data not shown) against full-length DnaJ showed that in all samples, each antibody was sensitive at s 1037 dilution. Luciferase was used as a model for functional studies because (i) its refolding does not require GroEL/GroES chaperones and (ii) the luminescence-based enzyme assay is extremely sensitive, thus allowing for refolding studies in small volumes. The in£uence of the Hsp70 chaperone machine on the thermal protection of luciferase function is shown in Fig. 2. Protection from heat-inactivation was only demonstrated when all components of the DnaK chaperone machine were present with luciferase, as previously reported in [13]. A¤nity-puri¢ed antibodies were then used to investigate the involvement of de¢ned DnaJ domains, the J-domain and the G/F region, Zn-¢nger-like and substrate-binding domains, in the protection of luciferase from heat-inactivation. Control experiments with af¢nity-puri¢ed antibodies from pre-immune sera did not inhibit the luciferase assay. Fig. 2 shows almost complete inhibition of the reactivation of heat-inactivated luciferase with full-length antiDnaJ (please note that the data points for anti-DnaJ and anti-DnaJ v1^108 partially obscure each other on

the scale used). This result was examined in more detail through the application of characterised anti-DnaJ sequence antibodies (see Fig. 1). Anti-DnaJ 1^108, antiDnaJ v1^108 and anti-DnaJ v1^199 sequence-speci¢c antibodies are directed against the highly conserved J-domain, Zn-¢nger-like domain and putative substrate-binding domain, respectively. Antibodies against DnaJ 1^108 almost completely inhibited the reactivation of denatured luciferase identifying this region as necessary for functioning of the Hsp70 chaperone machine. This supports results from a previous study, which demonstrated the role of the J-domain in the stimulation of the ATPase activity DnaK [11]. However, a more recent study has shown that the J-domain by itself neither stimulates ATPase activity of DnaK nor supports the replication of bacteriophage V DNA [21]. Antibodies against DnaJ v1^108, which were only directed against the Zn-¢nger domains, blocked the function of this domain in our protection assay. This result is in good agreement with a study which demonstrated that hdj2 (homologous to DnaJ), but not hdj1 (lacking the Zn-¢nger domains), together with hsc70 refolded luciferase e¤ciently [22]. In a previous study using a DnaK monoclonal antibody only 25% ATPase inhibition was observed at a stoichiometry of 1:10 [16]. Therefore, we propose that antibodies directed against the DnaJ N-terminal domains are powerful inhibitors (from Fig. 2 almost 100% inhibition was achieved at a stoichiometry of 1:10) of Hsp70 chaperone machine function. The most probable mechanism will involve inhibition of DnaJ-mediated stimulation of DnaK ATPase activity [23]. Whilst the possibility of steric hindrance and/or e¡ects cannot be entirely discounted we believe this unlikely as only one antibody, anti-DnaJ 1^ 108, inhibits DnaJ^DnaK complex formation as determined in enzyme-linked immunosorbent assays. For example, the binding of anti-DnaJ v1^108, as demonstrated by 100% inhibition of luciferase refolding does not sterically hinder DnaJ^DnaK interaction (data not shown). Antibodies against the C-terminal domain (anti-DnaJ v1^199) did not have a dramatic inhibitory e¡ect on the protection of luciferase compared with the other antibodies tested. The results show that in the presence of C-terminal antibodies there is a delay in reactivation, only achieving 80% reactivation after 45 min, in contrast with the control, which achieved 80% after 30 min. Previous studies have reported that only full-length DnaJ can cooperate with DnaK and GrpE in refolding denatured luciferase [14]. Our results are somewhat at variance with this hypothesis and may be due to the fact that our antibodies did not block all of the available epitopes contained within the C-terminus, some of which maybe necessary for chaperone machine function. In summary, our functional studies, using characterised sequence-speci¢c antibodies, demonstrate pivotal roles for the J, G/F and Zn-¢nger-like domains of DnaJ in the protection of luciferase from thermal denaturation. Pre-

FEMSLE 9797 8-3-01

S. Al-Herran, W. Ashraf / FEMS Microbiology Letters 196 (2001) 19^23

vious work on DnaJ has mainly concentrated on the J-domain [7,11,24] with the Zn-¢nger domain receiving less attention [14], however, this work also supports the view for a signi¢cant role for the latter in molecular chaperone function. Recent studies in E. coli have demonstrated the presence of a novel Hsc66^Hsc20 (J-type accessory protein) chaperone system which has been shown to refold denatured luciferase [25]. Therefore, it would be interesting to conduct a complementary immunological dissection of Hsc20 using the antibodies generated in this study.

[8]

[9]

[10]

[11]

Acknowledgements We wish to thank Dr. D. Wall (Stanford, USA) for the generous gifts of DnaJ over-producing strains. W.A. is grateful for grants received from the Nu¤eld Foundation and Society for General Microbiology. S.A.H. is also grateful to the British Council (Kuwait) for ¢nancial assistance. We are also grateful to Mr. P. Thompkins (University of Bradford) for critically reviewing this manuscript.

[12] [13]

[14]

[15]

[16]

References [17] [1] Georgopoulos, C. (1992) The emergence of the chaperone machines. Trends Biochem. Sci. 17, 295^299. [2] Auger, I. and Roudier, J. (1997) A function for the QKRAA amino acid motif mediating binding of DnaJ to DnaK (implications for the association of rheumatoid arthritis with HLA-DR4). J. Clin. Invest. 99, 1818^1822. [3] Morimoto, R., Tissieres, A. and Georgopoulos, C. (Eds.) (1994) The Biology of Heat Shock Proteins and Molecular Chaperones. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [4] Georgopoulos, C., Liberek, K., Zylicz, M. and Ang, D. (1994) Properties of the heat shock proteins of Escherichia coli and the autoregulation of the heat shock response. In: The Biology of Heat Shock Proteins and Molecular Chaperones (Morimoto, R., Tissieres, A. and Georgopoulos, C., Eds.), pp. 209^249. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [5] Cheetham, M.E. and Caplan, A.J. (1998) Structure, function and evolution of DnaJ: conservation and adaptation of chaperone function. Cell Stress Chaperone 3, 28^36. [6] Georgopoulos, C. and Welch, W.J. (1993) Role of the major heat shock proteins as molecular chaperones. Annu. Rev. Cell Biol. 9, 601^634. [7] Szyperski, T., Pellecchia, M., Wall, D., Georgopoulos, C. and Wuthrich, K. (1994) NMR structure determination of the Escherichia coli DnaJ molecular chaperone : secondary structure and backbone fold of the N-terminal region (residues 2^108) containing the highly

[18] [19]

[20]

[21]

[22]

[23]

[24]

[25]

23

conserved J-domain. Proc. Natl. Acad. Sci. USA 91, 11343^ 11347. Greene, M.K., Steede, N.K. and Landry, S.J. (2000) Domain-speci¢c spectroscopy of 5-hydroxytryptophan-containing variants of Escherichia coli DnaJ. Biochim. Biophys. Acta 1480, 267^277. Martinez-Yamout, M., Legge, G.B., Zhang, O., Wright, P.E. and Dyson, H.J. (2000) Solution structure of the cysteine-rich domain of the Escherichia coli chaperone protein DnaJ. J. Mol. Biol. 300, 805^818. Cyr, D.M. (1997) Hsp40 (DnaJ-related) proteins : an overview. In: Guidebook to Molecular Chaperones and Protein-folding Catalysts (Gething, M.J., Ed.). Oxford University Press, Oxford. Wall, D., Zylicz, M. and Georgopoulos, C. (1994) Identi¢cation of the conserved domain of the Escherichia coli DnaJ protein that stimulates DnaK's ATPase activity and required for V DNA replication. J. Biol. Chem. 269, 5446^5451. Hendrick, J.P. and Hartl, F.U. (1993) Molecular chaperone functions of heat shock proteins. Annu. Rev. Biochem. 162, 349^384. Schroder, H., Langer, T., Hartl, F.U. and Bukau, B. (1993) DnaK, DnaJ, GrpE form a cellular machine capable of repairing heat-induced protein damage. EMBO J. 12, 4137^4144. Szabo, A., Korrszun, R., Hartl, F.U. and Flanagan, J. (1996) A zinc ¢nger-like domain of the molecular chaperone DnaJ is involved in binding to denatured protein substrates. EMBO J. 15, 408^417. Lesley, S.A., Thompson, N.E. and Burgess, R.R. (1987) Studies of the role of the Escherichia coli heat shock regulatory protein c32 by the use of monoclonal antibodies. J. Biol. Chem. 262, 5404^5407. Krska, J., Elthon, T. and Blum, P. (1993) Monoclonal antibody recognition and function of a DnaK (Hsp70) epitope found in Gram-negative bacteria. J. Bacteriol. 175, 6433^6440. Zylicz, M., Yamamoto, T., McKiittrick, N., Snell, S. and Georgopoulos, C. (1985) Puri¢cation and properties of the DnaJ replication protein of Escherichia coli. J. Biol. Chem. 260, 7591^7598. Harlow, E. and Lane, D. (1988) Antibodies: a Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Wickner, S., Gottesman, S., Skowyra, D., Hoskins, J., McKenney, K. and Maurizi, M.R. (1994) A molecular chaperone, ClpA, functions like DnaK and DnaJ. Proc. Natl. Acad. Sci. USA 91, 12218^12222. Al-Herran, S. and Ashraf, W. (1998) Physiological consequences of the over-production of Escherichia coli truncated molecular chaperone DnaJ. FEMS Microbiol. Lett. 162, 117^122. Karzai, A.W. and McMachen, R. (1996) A bipartite signalling mechanism involved in DnaJ-mediated activation of the Escherichia coli DnaK protein. J. Biol. Chem. 271, 11236^11246. Terada, K., Kanazawa, M., Bukau, B. and Mori, M. (1997) The human DnaJ homologue dnaJ2 facilitates mitochondrial protein import and luciferase refolding. J. Cell Biol. 139, 1089^1095. Mayer, P.M., Schroder, H., Rudiger, S., Paal, K., Laufer, T. and Bukau, B. (2000) Mutlistep mechanism of substrate binding determines chaperone activity of Hsp70. Nat. Struct. Biol. 7, 587^593. Huang, K., Flanagan, J.M. and Prestegard, J.H. (1999) The in£uence of C-terminal extension on the structure of the `J-domain' in E. coli DnaJ. Protein Sci. 8, 203^214. Silberg, J.J., Ho¡, K.G. and Vickery, L.E. (1998) The Hsc66^Hsc20b chaperone system in Escherichia coli: chaperone activity and interactions with DnaK^DnaJ^GrpE system. J. Bacteriol. 180, 6617^6624.

FEMSLE 9797 8-3-01