Trigger Factor Peptidyl-prolyl cis/trans Isomerase Activity Is Not Essential for the Folding of Cytosolic Proteins in Escherichia coli*
Received for publication, December 12, 2003, and in revised form, January 15, 2004 Published, JBC Papers in Press, January 16, 2004, DOI 10.1074/jbc.M313635200
Gu¨ nter Kramer‡§, Holger Patzelt‡, Thomas Rauch, Thorben A. Kurz, Sonja Vorderwu¨ lbecke¶, Bernd Bukau, and Elke Deuerling储
From the Zentrum fu¨ r Molekulare Biologie (ZMBH), Universita¨t Heidelberg, Im Neuenheimer Feld 282, D-69120 Heidelberg, Germany
The ribosome-associated Trigger Factor (TF) cooper- ates with the DnaK system to assist the folding of newly synthesized polypeptides inEscherichia coli. TF unifies two functions in one to promote proper protein folding in vitro. First, as a chaperone it binds to unfolded pro- tein substrates, thereby preventing aggregation and supporting productive folding. Second, TF catalyzes the cis/transisomerization of peptidyl-prolyl bonds, which can be a rate-limiting step in protein folding. Here, we investigated whether the peptidyl-prolyl cis/trans isomerase (PPIase) function is essential for the folding activity of TFin vitro andin vivo by separating these two TF activities through site-directed mutagenesis of the PPIase catalytic center. Of the four different TF variants carrying point mutations in the PPIase do- main, only the exchange of the conserved residue Phe- 198 to Ala (TF F198A) abolished the PPIase activity of TF toward both a tetrapeptide and the model protein sub- strate RNase T1in vitro. In contrast, all other activities of TF F198A tested were comparable with wild type TF.
TF F198A retained a similar binding specificity toward membrane-bound peptides, assisted the refolding of de- natured D-glyceraldehyde-3-phosphate dehydrogenase in vitro, and associated with nascent polypeptides in an in vitro transcription/translation system. Importantly, expression of the TF F198A encoding gene comple- mented the synthetic lethality of⌬tig⌬dnaKcells and prevented global protein misfolding at temperatures be- tween 20 and 34 °C in these cells. We conclude that the PPIase activity is not required for the function of TF in folding of newly synthesized proteins.
Trigger Factor (TF)1was first discovered inEscherichia coli as a protein that triggered the translocation of the precursor of
pro-OmpA into membrane vesicles (1). Although this activity indicated a secretion-specific chaperone function for TF,E. coli cells depleted for TF were subsequently shown to lack any secretion defect (2). Later, TF was rediscovered as a ribosome- associated chaperone and peptidyl-prolyl cis/trans isomerase (PPIase) (3, 4).
TF is located on the large ribosomal subunit near the peptide exit channel and binds to virtually all nascent polypeptides (4 – 6). It was therefore proposed that TF is the first chaperone that interacts with nascent chains and assists cotranslational protein folding. Genetic evidence indicates that TF cooperates with the DnaK system to ensure proper folding of cytosolic proteins.E. colicells deleted for the TF-encodingtiggene show no growth defects at temperatures between 15 and 42 °C, and mutants lacking thednaKgene are viable between 20 °C and 37 °C. However, the combined deletion of the tig and dnaK genes causes aggregation of at least 340 species of newly syn- thesized cytosolic proteins and synthetic lethality at 37 °C (7–
9). In vitro TF chaperone activity prevents aggregation and promotes refolding of denatured rabbit D-glyceraldehyde-3- phosphate dehydrogenase (GAPDH) (10).
In contrast to DnaK, TF, in addition to its chaperone activity, can efficiently catalyze cis/trans isomerization of peptidyl- prolyl peptide bonds in chromogenic tetrapeptides and the RNase T1 protein substratein vitro(PPIase activity) (3, 4, 11).
For RNase T1, the slow isomerization is the rate-limiting step in the folding process (12). Unlike other prolyl isomerases, TF binds with high affinity to unfolded RNase T1, and this chap- erone function is the prerequisite for the excellent catalysis of folding (11). There are a few other examples of PPIases that also display a chaperone activity, including the periplasmic PPIase SurA fromE. coliand the mammalian FKBP52. How- ever, in these proteins the two activities are localized in differ- ent domains (13, 14). TF is different, because both activities involve the same substrate binding pocket in the central PPI- ase domain. A similar finding was only observed for the periplasmic chaperone and PPIase FkpA fromE. coli(15).
TF has a modular structure with functions assigned to two of the three domains. The N-terminal domain (aa 1–144) medi- ates ribosome binding. The central domain (aa 145–247) har- bors the peptide binding site and the catalytic PPIase activity, both of which involve the same binding pocket (16 –18). This domain has homology to PPIases of the FK506-binding protein (FKBP) family and mediates substrate binding specificity (16, 19).
An important open question is whether the activity of TF as a ribosome-associated folding factor relies on both its chaper- one and PPIase activity. To separate the PPIase and chaperone activities of TF, we constructed TF mutant proteins with spe- cific alterations in the PPIase domain. Of four different TF
* This work was supported by the grants from the Deutsche Forsch- ungsgemeinschaft (to B. B. and E. D.), the Human Frontier Science Program (to E. D.), and the Fellowship of the Boehringer Ingelheim Fonds (to T. R.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ These authors contributed equally to this work.
§ Present address: Whitehead Institute for Biomedical Research, Cambridge, MA 02142.
¶Present address: Ciphergen Biosystems GmbH, Hannah-Vogt- Str.1, 37085 Goettingen, Germany.
储To whom correspondence should be addressed. Tel.: 49-6221- 546870; Fax: 49-4221-545894; E-mail: e.deuerling@zmbh.uni- heidelberg.de.
1The abbreviations used are: TF, Trigger factor; aa, amino acid(s);
FKBP, FK506-binding protein; GAPDH,D-glyceraldehyde-3-phosphate dehydrogenase; IPTG, isopropyl-1-thio--D-galactopyranoside; PPIase, peptidyl-prolylcis/transisomerase; RCM-RNase T1, reduced and car- boxymethylated RNase T1.
© 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.
This paper is available on line at http://www.jbc.org
Konstanzer Online-Publikations-System (KOPS) 14165 URL: http://www.ub.uni-konstanz.de/kops/volltexte/2007/3709/
URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-37092
variants, the mutant protein TF F198A fully retains its chap- erone activity but lacks the catalytic PPIase function. This TF mutant provided a unique tool to investigate the importance of the PPIase activity for TF function in promoting the folding of newly synthesized proteinsin vivo.
EXPERIMENTAL PROCEDURES
Growth Conditions—Strains were grown in LB (Luria Broth) me- dium containing IPTG as indicated and supplemented with ampicillin (100g/ml), tetracycline (5 g/ml), or kanamycin (40 g/ml) when appropriate.
Strains and Plasmids—E. colistrains were derivatives of MC4100.
Strain GK2, carrying a chromosomal deletion of the entirednaKopen reading frame and expressing wild type levels of DnaJ, was constructed as follows. ThednaKgene of MC4100 was replaced with a kanamycin resistance cassette using plasmid pKD4 as PCR-template and the prim- ers dnaK-5⬘(5⬘-CAGACTCACAACCACATGATGACCGAATATATAGT- GGAGACGTTTAGGTGTAGGCTGGAGCTGCTTCG-3⬘) and dnaK-3⬘
(5⬘-TTCCGCTGTTTTGGAAACGCCTAAAATCTCGTAATAATCTTGC- TTAGCCATATGAATATCCTCCTTAG-3⬘) as described (20), resulting in strain GK1. Subsequently, the kanamycin resistance cassette was removed using plasmid pCP20 (20), resulting in strain GK2.
Mutations in thetiggene were created using the QuikChange site- directed mutagenesis kit (Stratagene) and pDS56-tig-His6as template (21), resulting in pDS56-tig178, pDS56-tig198, pDS56-tig221, pDS56- tig224, and pDS56-tig233, respectively. Plasmid pTrc-tig and pTrc- tig198are derivatives of pTrc99B (22). Utilizing PCR,tigwas amplified using pDS56-tigas template and primer 5⬘-tig (5⬘-CAAGTTTCAGTT- GAAACCACTC-3⬘) and 3⬘-tig-BamHI (5⬘-GGCCGGATCCTTACGCCT- GCTGGTTCATCAGCTC-3⬘), gel purified, and subsequently digested with BamHI. Plasmid pTrc99B was digested with NcoI, blunt-ended using T4-DNA-polymerase (New England Biolabs, Inc.), then digested with BamHI, gel purified, and ligated with the BamHI-digestedtig fragment, resulting in pTrc-tig. pTrc-tig198 was constructed by ex- changing the StuI fragment of pTrc-tigwith the corresponding StuI fragment of pDS56-tig198.
In Vivo Complementation Analysis—P1 lysates, P1 transductions, and disruption of chromosomal genes were done as described (20, 23).
GK2 was transformed using plasmid pTrc-tig or pTrc-tig198, respec- tively. The deletion oftig, encoding TF, was introduced by P1 transduc- tion using a P1 lysate derived from a strain carrying⌬tig::kangeneti- cally linked tozba-3054::Tn10, conferring tetracycline resistance as described (7). After P1 transduction, cells were selected on LB medium supplemented with 10 M IPTG and 5 g/ml tetracycline. Subse- quently, tetracycline-resistant clones were tested for cotransduction of
⌬tig::kan by growth on LB plates containing 10 M IPTG and 40
g/ml kanamycin.
Protein Purification—TF and TF variants were purified as C-termi- nally His6-tagged proteins by nickel nitrilotriacetic acid affinity chro- matography and anion exchange chromatography (21).
PPIase Tetrapeptide Assay—The assay was performed essentially as described (24). To start a reaction, 84Msuccinyl-Ala-Phe-Pro-Phe- para-nitroanilide (Bachem Biochemica GmbH), 0.08 mg/ml chymotryp- sin (Sigma; C-7762) and 100 nMTF or TF variant were mixed in 35 mM
HEPES-KOH, pH 7.6, at 10 °C. Absorption at 395 nm was followed in a photometer (Shimatsu). The first 30 s of the assay were treated as dead time in which the large excess of the pre-existingtransisomer is cleaved by chymotrypsin. Data after 30 s were fitted to a single-exponential equation (GraFit program). The catalytic efficiency of TF and TF vari- ants was corrected for spontaneous isomerization by the equation kcat/Km⫽(kexp⫺ksp)/cTF, wherekexpis the experimental rate constant, kspis the constant of the spontaneous isomerization reaction without TF, andcTFis the concentration of TF or TF variants, respectively.
RNase T1 Refolding Assay—RNase T1 fromAspergillus oryzae(Sig- ma; R-1003) was denatured, reduced, and carboxymethylated using dithiothreitol as the reducing agent according to the protocol (25).
Protein refolding was induced by dilution to final concentrations of 1M reduced and carboxymethylated RNase T1 (RCM-RNase T1) and 1.6M
NaCl, respectively, in a spectrofluorometer (PerkinElmer Life Sciences;
LS 55). The concentration of TF variants was 0.2M.
Screening of Membrane-bound Peptides—Cellulose membrane- bound 13-meric peptides scanning the sequence of thecI protein with an overlap of 10 amino acids were incubated with 500 nMTF or TF variants for 30 min at 25 °C. Bound chaperone was electrotransferred by fractionated semi-dry blotting to a polyvinylidene difluoride mem- brane and detected using TF-specific antisera (16, 26).
Prevention of GAPDH Aggregation—Aggregation of denatured
GAPDH from rabbit muscle (Sigma; G-2267) was measured according to the method described (10). 125MGAPDH was denatured overnight at 4 °C. Upon 50-fold dilution of the denatured enzyme into GAPDH buffer (0.1Mpotassium phosphate, pH 7.5, 1 mMEDTA, 5 mMdithio- threitol), aggregation was monitored by 90° light scattering at 620 nm in a spectrofluorometer (PerkinElmer Life Sciences; LS 55) in the presence or absence of 1 or 2MTF or TF variants.
GAPDH Refolding Assay—Refolding of denatured GAPDH was per- formed as described (10). Denatured enzyme was 50-fold diluted into GAPDH buffer (see above) containing 2.5MTF or TF variant, kept at 4 °C for 30 min, and for a further 240 min at 25 °C. At different time points within these 240 min, an aliquot was taken, and GAPDH activity was measured. To analyze time points, a 200-l aliquot was mixed with 800l of GAPDH buffer to final concentrations of 667M-nicotina- mide adenine dinucleotide (-NAD; Sigma; N-0632) and 667 M DL- glyceraldehyde-3-phosphate (GAP; Sigma G-5251). Absorbance of syn- thesized NADH was followed in a spectrometer (Shimatsu) at 340 nm.
Data were fitted to a single exponential equation (GraFit program).
Relative activity of refolded GAPDH was calculated by dividing the rate constants by the rate constant determined for the same amount of non-denatured GAPDH.
Miscellaneous—In vitro transcription/translation assays were per- formed in an E. coli-based transcription/translation system isolated from a⌬tigstrain (7, 27). Isolation of aggregates was performed as described previously (28).
RESULTS
Mutations in the TF PPIase Domain—Based on a study on human FKBP12 in which residues that are essential for its catalytic PPIase activity (29) were identified, we designed mu- tations in the homologous PPIase domain of TF. We exchanged the conserved residues Glu-178 (Glu-37 in FKPB12), Phe-198 (Trp-59 in FKBP12), Tyr-221 (Tyr-82 in FKBP12), and Phe-233 (Phe-99 in FKBP12) in the TF PPIase domain to generate the TF E178V, TF F198A, TF Y221F, and TF F233L mutants, respectively. As illustrated by a homology model of the TF PPIase domain, these amino acids localize within or adjacent to the central substrate binding pocket (Fig. 1). The TF mutant proteins were analyzed for their structural integrity by circular dichroism measurements, proteinase K digestion, and determi- nation of their ability to associate withE. coliribosomes. In these assays, all mutant proteins behaved similarly to wild type TF (data not shown).
TF F198A Is Deficient in PPIase Activity—We tested the TF mutants for their catalytic PPIase activity toward two sub- strates, a chromogenic tetrapeptide and the model protein sub- strate RNase T1. These assays may give different results, be- cause the binding affinity of TF for protein substrates is much higher than that for peptides (11, 16). Small alterations in the binding pocket may have a more pronounced effect on the affinity of TF for peptides. This phenomenon has also been observed for other PPIases (30).
The tetrapeptide assay detects PPIase activity by measuring FIG. 1.Homology model of the TF PPIase domain (aa 145–242) based on the structure of FKBP12 from yeast (16).The structure is illustrated using the WebLabViewer program. The backbone of the protein is shown as aribbon(green), and the conserved amino acid side chains forming the binding pocket asballs andsticks. Thered balls represent oxygen atoms, and all others represent carbon atoms. Amino acid residues depicted inyellowwere exchanged in this study. Single letter amino acid abbreviations (Ffor Phe,Yfor Tyr, andEfor Glu) are used.
thetransisomer-specific proteolytic cleavage of the substrate succinyl-Ala-Phe-Pro-Phe-4-nitroanilide, which contains a mix- ture ofcisandtransPhe-Pro bonds (24). The TF variants TF F198A and TF F233L were inactive (⬍1% of TF wild type activity; Fig. 2A), even when added in a 20-fold higher concen- tration than wild type TF (data not shown). TF Y221F showed a residual activity of ⬃15% compared with wild type TF, whereas TF E178V was slightly more active than wild type TF, as observed earlier (31). Importantly, none of the TF variants was degraded by chymotrypsin during the assay as detected by Western blotting (data not shown).
Refolding of RCM-RNase T1 is limited by the slowtransto cisisomerization of peptidyl-prolyl bonds at Pro-39 and Pro-55 (12). RCM-RNase T1 is unfolded at 15 °C and pH 8 under low salt conditions (0.4MNaCl) but folds back to an enzymatically
active form under 2M salt conditions. The refolding of RCM- RNase T1 upon dilution into 2Msalt buffer can be followed by measuring intrinsic tryptophan fluorescence. In this assay, only TF F198A exhibited no significant catalytic activity. In contrast to the tetrapeptide assay, the variants TF Y221F and TF F233L were active in the refolding of RNase T1, although less than wild type TF, and TF E178V displayed wild type activity (Fig. 2B). Together, these findings show a consistent lack of PPIase activity for the TF F198A mutant.
TF and TF F198A Show Similar Binding Specificities—It is possible that mutations in the substrate binding pocket of the PPIase domain may alter the substrate binding specificity of TF and, thereby, prevent TF mutant proteins from acting in the prolyl isomerization of the two substrates tested. The TF bind- ing specificity can be investigated by using libraries consisting of 13-meric membrane-bound peptides (16). To check whether the PPIase deficiency of TF F198A could be due to a shift in binding specificity, we investigated its recognition of different amino acid stretches in a model substrate. We incubated pep- tide libraries, scanning the sequence of thecI protein with either TF or TF F198A to equilibrium. Subsequently, bound chaperone was electrotransferred to a polyvinylidene difluoride membrane and immunodetected. As shown in Fig. 2C, both TF proteins recognized the same peptides and, therefore, dis- played comparable binding specificities. Although the signal intensities for TF and TF F198A on the peptide libraries were similar (Fig. 2C), TF and TF F198A may differ to some extent, because this method does not quantitatively measure the af- finity for peptides. Other TF mutants behaved similarly (data not shown).
TF F198A Is Active as a Chaperone in Vitro—TF efficiently prevents the aggregation and promotes the refolding of chem- ically denatured GAPDH. The refolding of GAPDH does not depend on the isomerization of peptidyl-prolyl bonds. There- fore, this assay selectively monitors the chaperone, but not the PPIase activity, of TF (10).
To analyze the chaperone activity of the TF F198A, we de- natured GAPDH overnight in 3MGdnHCl. Upon 50-fold dilu- tion, GAPDH aggregated rapidly. Stoichiometric amounts of TF efficiently prevented the GAPDH aggregation, as no in- crease of light scattering signal at 620 nm was observed. All TF mutant proteins, including the PPIase deficient variant TF F198A, showed a similar activity as wild type TF (Fig. 3A).
GAPDH activity can be monitored by following the reduction of NAD⫹to NADH at 340 nm in a photometer. This assay was used to determine the folding kinetics of denatured GAPDH in the presence of TF and TF variants. At different time points after 50-fold dilution of the denatured enzyme, we determined GAPDH activity. In the absence of TF, GAPDH remained en- zymatically inactive over a 4-h time period. In contrast, the presence of wild type TF restored up to 80% of non-denatured GAPDH activity, similar to results reported earlier (10). All TF variants revealed a chaperone activity comparable with wild type TF (Fig. 3B).
Mutations in the TF PPIase Domain Do Not Abolish Interac- tion with Nascent Polypeptides—Ribosome association and in- teraction with nascent polypeptides are crucial for TF function in vivo(5). To characterize the ability of TF PPIase mutant proteins to bind nascent polypeptide chains, we performed cross-linking analysis in anE. coli-basedin vitrotranscription/
translation system derived from TF-deficient cells (27). We generated arrested nascent polypeptide chains of isocitrate dehydrogenase (ICDH), which is a natural substrate of TF (9).
Translation was carried out in the presence of [35S]methionine to label the nascent polypeptides and exogenously added TF or TF variants. Upon addition of the chemical cross-linker disuc- FIG. 2.PPIase activity and binding specificity of TF and TF
variants.A, isomerization of a chromogenic tetrapeptide was moni- tored by an increase of the absorption signal at 395 nm. Specific PPIase activitieskcat/Kmare 0.52 mM⫺1s⫺1and 0.76 mM⫺1s⫺1for TF wild type (circles) and TF E178V (squares), respectively. TF Y221F (diamonds) has⬃15% of wild type activity (kcat/Km⫽0.080 mM⫺1s⫺1), whereas in the presence of TF F198A (triangles;kcat/Km⬍0.001 mM⫺1s⫺1) and TF F233L (hexagons;kcat/Km⫽0.004 mM⫺1s⫺1) isomerization kinetics are like those observed in the absence of TF variants (crosses).B, PPIase activity of TF and TF variants toward the model protein substrate RCM-RNase T1. RCM-RNase T1 refolding is monitored by an increase of intrinsic tryptophan fluorescence at 320 nm after excitation at 268 nm. Because refolding of RCM-RNase T1 is a complex process (12), no refolding rates could be determined. Symbols are the same as forpanel A. C, binding of TF wild type and TF F198A to libraries of membrane- bound peptides scanning the sequence of thecI protein.
cinimidyl suberate (DSS), all TF proteins revealed cross-link- ing products of⬃70 and 90 kDa with a similar cross-linking efficiency (Fig. 4). We conclude that none of the TF mutant proteins, including TF F198A, are impaired in the association with nascent polypeptides.
The PPIase Activity of TF Is Not Essential for Its in Vivo Function—Analyses described above revealed that TF F198A behaves similarly to wild type TF with regard to its chaperone functionin vitrobut is deficient in PPIase activity. Thus, this protein provided us a unique opportunity to determine the functional importance of TF PPIase activity in the chaperoning of nascent polypeptidesin vivo.
We analyzed whether the PPIase activity of TF is crucial for complementing the lethality of⌬tig⌬dnaKcells. We therefore introduced a plasmid expressing either TF wild type or TF F198A under the control of an IPTG-inducible promoter into
⌬dnaK cells. Subsequently, we performed phage P1 cotrans- duction experiments in which a Tn10::Tet-selective marker placed close to the⌬tig::kanallele was transduced into⌬dnaK cells. In the presence of 10MIPTG, tetracycline, and kana- mycin-resistant⌬tig⌬dnaKcotransductants expressing either plasmid-encoded TF F198A or TF wild type protein were found at 30 °C on LB media with a similar cotransduction frequency (91 and 88%, respectively). No cotransductants could be ob- served in the absence of IPTG in the growth media. Growth analysis revealed that ⌬dnaK⌬tig cells complemented with genes encoding either wild type TF or TF F198A show a very similar IPTG concentration-dependent growth behavior on LB/
IPTG-plates at temperatures between 20 and 34 °C,i.e.within the growth range of cells lacking DnaK (Fig. 5A). Thus, the PPIase-deficient TF F198A variant is as efficient as wild type TF in complementing the⌬tig⌬dnaKsynthetic lethality.
To analyze protein folding in these strains, we grew the cells
at 30 and 34 °C to logarithmic phase in the presence of 10M IPTG and subsequently isolated aggregated proteins. The cel- lular levels of wild type TF and TF F198A expressed from plasmid in⌬dnaK⌬tigcells were similar (Fig. 5B) and slightly higher as compared with the expression of chromosomally en- coded TF in⌬dnaKcells (data not shown). As reported earlier, only mild folding defects were detectable in⌬dnaKcells (Fig.
5C), in contrast to the massive protein aggregation observed in cells lacking both DnaK and TF (9). The presence of TF F198A compensated for the loss of TF in⌬dnaK⌬tigcells as efficiently as wild type TF. Neither the pattern nor the total amount of aggregated proteins isolated from these strains differed signif- icantly from the aggregated proteins found in⌬dnaKcells at 30 (Fig. 5C) and 34 °C (data not shown). We conclude that the chaperone activity of TF F198A is sufficient to compensate for the loss of TFin vivo.
DISCUSSION
The mechanism by which ribosome-associated TF assists the folding of newly synthesized polypeptides is not understood. In particular, it is unclear whether the activity of TF in the folding of newly synthesized proteins relies on its PPIase or its chap- erone activity. Both activities colocalize in the central PPIase domain of TF. To separate these activities, we genetically en- gineered four different TF mutant proteins with exchanges of amino acid residues in or adjacent to the substrate binding pocket of TF.
Replacing Glu-178 with Val or Tyr-221 with Phe had little effect on TF PPIase activityin vitro. The mutant protein TF F233L did not catalyze the isomerization of a tetrapeptide but promoted the peptidyl-prolyl isomerization-limited refolding of RNase T1. This TF variant may thus be deficient only in the correct binding of peptide substrate rather than in the PPIase activity itself. Our results are in agreement with earlier find- ings by Fischer and co-workers, who found that the exchange of Phe-233 to Tyr resulted in the loss of TF PPIase activity toward tetrapeptidesin vitro(31). In addition, when the F233Y mutant was used to complement a⌬tigstrain, a reduced survival rate on plates after storage for 25 days at 6 °C was observed as compared with wild type TF (32). This might suggest an in- volvement of TF in cell viability at very low temperatures.
However, the interpretation of this finding is difficult, because 6 °C is below the physiological growth temperature ofE. coli.
Moreover, no furtherin vitroandin vivocharacterizations of this TF mutant protein, such as an investigation of its sub- strate binding properties, had been performed.
It was reported earlier that the replacement of Phe-198 by either His or Trp in TF decreases the catalytic efficiency of the PPIase to ⬍0.3% compared with the wild type protein (31).
However, these variants were unstable and were only charac- terized regarding their PPIase activity toward tetrapeptidesin vitro. To obtain a stable PPIase-deficient TF, we exchanged Phe-198 with Ala based on a corresponding mutation in the homologous human FKBP12 that resulted in a stable but in- active enzyme (29). In contrast to all other TF variants tested here, TF F198A was deficient in PPIase activity toward both tetrapeptide and protein substrate. We characterized the TF F198A variant extensively. In addition to its PPIase activity, we tested the variant for its binding specificity, chaperone properties, and ability to associate with ribosomes and nascent polypeptidesin vitro. In all these assays, TF F198A behaved similarly to wild type TF.
InE. coli, the folding of newly synthesized cytosolic proteins critically depends on the function of ribosome-associated TF and the cooperating DnaK chaperone system. Cells lacking TF are viable but show increased levels of DnaK (9). The simulta- neous deletion of both chaperone genes,tiganddnaK, leads to FIG. 3.Chaperone activity of TF and TF variants.Symbols are
the same as in Fig. 2.A, prevention of GAPDH aggregation by TF and TF variants. Aggregation of GAPDH is followed by an increase of the light scattering signal at 620 nm after 50-fold dilution of the denatured enzyme (final concentration, 2.5M). The addition of 1MTF or TF variants significantly inhibits aggregation (open symbols). In the pres- ence of 2 M TF or TF variants, aggregation is almost completely prevented (filled symbols).B, refolding of denatured GAPDH is moni- tored by measuring the enzymatic activity at different time points after 50-fold dilution (final concentration, 2.5M) in the absence or presence of 2.5MTF or TF variants.
a massive aggregation of proteins, including many essential species, and is accompanied by cell death (7, 8). Thein vivo analysis of TF F198A provided evidence that the PPIase func- tion of TF is neither essential for complementing the lethality of⌬tig⌬dnaK cells nor essential for the productive folding of newly synthesized proteins at all temperatures tested. Like wild type TF, the TF F198A variant efficiently suppressed protein aggregation in⌬tig⌬dnaKcells. We cannot exclude the possibility that some substrates do depend on TF PPIase ac- tivity change, however, these protein aggregates were not de- tected by our method. Moreover, we cannot rigorously rule out a minor residual PPIase activity of TF F198A at 30 °Cin vivo, because all in vitro PPIase assays had to be conducted at temperatures below 30 °C due to the fast isomerization of pep- tidyl-prolyl bonds at room temperature and above. However, in an independent study, we provided evidence that TF is func-
tionalin vivoeven in the absence of the entire PPIase domain.2 Taken together, we conclude that thede novofolding of newly synthesized polypeptides, in the vast majority, does not require the ribosome-associated PPIase activity of TF inE. coli. How- ever, we do not exclude the possibility that the PPIase activity of other cytosolic PPIases found inE. colimight play a role in the folding of newly synthesized polypeptides.
What is the role of TF PPIase activity? At least two hypoth- eses are conceivable. First, the PPIase function might only be required for specific substrates or under specific environmental conditions not tested here. Very recently, it has been reported that the maturation of the extracellular protease SpeB ofStrep-
2G. Kramer, A. Rutkowska, R. Wegrzyn, H. Patzelt, T. A. Kurz, F.
Merz, T. Rauch, S. Vorderwu¨ lbecke, E. Deuerling, and B. Bukau, sub- mitted for publication.
FIG. 4. Interaction of TF and TF variants with nascent polypeptides.
Arrested 35S-labeled nascent isocitrate dehydrogenase (ICDH1–173) is synthe- sizedin vitroin a TF-deficient transcrip- tion/translation system supplemented with TF or TF variants. After cross-link- ing with disuccinimidyl suberate (DSS), added where indicated, and sucrose cush- ion centrifugation, ribosome-nascent chain complexes are coimmunoprecipitated to identify cross-links of TF and TF variants (lanes 3,6,9,12,and15).Bracketsindi- cate cross-linking products.
FIG. 5. Complementation analysis of the synthetic lethality of
⌬dnaK⌬tigcells.A, growth analysis of
⌬dnaK⌬tig cells expressing plasmid-en- coded TF wild type (pTrc-tig) or TF F198A (pTrc-tig198). Serial dilutions of cells were spotted on LB plates containing 10
MIPTG and incubated for 24 h at 30 or 34 °C or for 48 h at 20 °C. Cells grew similar at 20 and 50MIPTG (data not shown).B, cells grown in the presence of 10MIPTG at 30 °C were analyzed for their TF levels by Western blotting. C, aggregates were isolated from cells grown at 30 °C in LB medium supplemented with 10MIPTG and subsequently ana- lyzed by SDS-PAGE and Coomassie Blue staining.
tococcus pyogenesis influenced by the PPIase activity of its TF (33). However, a data base search revealed that no SpeB hom- olog exists inE. coli. Moreover, cells lacking TF do not show any secretion defect (2). We instead favor a second hypothesis according to which the FKBP-like hydrophobic substrate bind- ing pocket of TF provides a tool for the recognition of hydro- phobic stretches in unfolded proteins, as is typical for a chap- erone. This property is independent of the catalytic activity of TF, which has no functional importance forE. coliunder the conditions tested here. Two findings support this hypothesis. (i) Members of the FKBP family show PPIase-independent chap- erone activity (15). (ii) There is no ribosome-associated PPIase known in eukaryotic cells. Our results now provide evidence that the only known ribosome-associated PPIase activity in prokaryotes, based on TF, is not crucial for the folding of newly synthesized proteins inE. coli.
Acknowledgments—We thank the members of the Bukau Laboratory for discussions and S. Wilbanks and R. Wegrzyn for comments on the manuscript.
REFERENCES
1. Crooke, E., and Wickner, W. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 5216 –5220
2. Guthrie, B., and Wickner, W. (1990)J. Bacteriol.172,5555–5562
3. Stoller, G., Ruecknagel, K. P., Nierhaus, K. H., Schmid, F. X., Fischer, G., and Rahfeld, J.-U. (1995)EMBO J.14,4939 – 4948
4. Hesterkamp, T., Hauser, S., Lu¨ tcke, H., and Bukau, B. (1996)Proc. Natl. Acad.
Sci. U. S. A.93,4437– 4441
5. Kramer, G., Rauch, T., Rist, W., Vorderwu¨ lbecke, S., Patzelt, H., Schulze- Specking, A., Ban, N., Deuerling, E., and Bukau, B. (2002)Nature419, 171–174
6. Valent, Q. A., Kendall, D. A., High, S., Kusters, R., Oudega, B., and Luirink, J.
(1995)EMBO J.14,5494 –5505
7. Deuerling, E., Schulze-Specking, A., Tomoyasu, T., Mogk, A., and Bukau, B.
(1999)Nature400,693– 696
8. Teter, S. A., Houry, W. A., Ang, D., Tradler, T., Rockabrand, D., Fischer, G., Blum, P., Georgopoulos, C., and Hartl, F. U. (1999)Cell97,755–765 9. Deuerling, E., Patzelt, H., Vorderwu¨lbecke, S., Rauch, T., Kramer, G., Schaf-
fitzel, E., Mogk, A., Schulze-Specking, A., Langen, H., and Bukau, B. (2003) Mol. Microbiol.47,1317–1328
10. Huang, G. C., Li, Z. Y., Zhou, J. M., and Fischer, G. (2000)Protein Sci.9, 1254 –1261
11. Scholz, C., Stoller, G., Zarnt, T., Fischer, G., and Schmid, F. X. (1997)EMBO J.16,54 –58
12. Schmid, F. X. (1993)Annu. Rev. Biophys. Biomol. Struct.22,123–143 13. Pirkl, F., Fischer, E., Modrow, S., and Buchner, J. (2001)J. Biol. Chem.276,
37034 –37041
14. Behrens, S., Maier, R., de Cock, H., Schmid, F. X., and Gross, C. A. (2001) EMBO J.20,285–294
15. Ramm, K., and Pluckthun, A. (2001)J. Mol. Biol.310,485– 498
16. Patzelt, H., Rudiger, S., Brehmer, D., Kramer, G., Vorderwulbecke, S., Schaf- fitzel, E., Waitz, A., Hesterkamp, T., Dong, L., Schneider-Mergener, J., Bukau, B., and Deuerling, E. (2001)Proc. Natl. Acad. Sci. U. S. A.98, 14244 –14249
17. Stoller, G., Tradler, T., Rucknagel, K. P., Rahfeld, J.-U., and Fischer, G. (1996) FEBS Lett.384,117–122
18. Hesterkamp, T., and Bukau, B. (1996)FEBS Lett.385,67–71 19. Callebaut, I., and Mornon, J. P. (1995)FEBS Lett.374,211–215
20. Datsenko, K. A., and Wanner, B. L. (2000)Proc. Natl. Acad. Sci. U. S. A.97, 6640 – 6645
21. Hesterkamp, T., Deuerling, E., and Bukau, B. (1997)J. Biol. Chem.272, 21865–21871
22. Amann, E., Ochs, B., and Abel, K. J. (1988)Gene69,301–315
23. Miller, J. H. (1972)Experiments in Molecular Genetics, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
24. Fischer, G., Bang, B., Ludwig, B., Mann, K., and Hacker, J. (1992)Mol.
Microbiol.6,1375–1383
25. Mu¨ cke, M., and Schmid, F. X. (1992)Biochemistry31,7848 –7854
26. Ru¨ diger, S., Germeroth, L., Schneider-Mergener, J., and Bukau, B. (1997) EMBO J.16,1501–1507
27. Schaffitzel, E., Ru¨ diger, S., Bukau, B., and Deuerling, E. (2001)Biol. Chem.
382,1235–1243
28. Mogk, A., Tomoyasu, T., Goloubinoff, P., Ru¨diger, S., Ro¨der, D., Langen, H., and Bukau, B. (1999)EMBO J.18,6934 – 6949
29. DeCenzo, M. T., Park, S. T., Jarrett, B. P., Aldape, R. A., Futer, O., Murcko, M. A., and Livingston, D. J. (1996)Protein Eng.9,173–180
30. Scholz, C., Schindler, T., Dolinski, K., Heitman, J., and Schmid, F. X. (1997) FEBS Lett.414,69 –73
31. Tradler, T., Stoller, G., Ru¨ cknagel, K. P., Schierhorn, A., Rahfeld, J.-U., and Fischer, G. (1997)FEBS Lett.407,184 –190
32. Schiene-Fischer, C., Habazettl, J., Tradler, T., and Fischer, G. (2002)Biol.
Chem.383,1865–1873
33. Lyon, W. R., and Caparon, M. G. (2003)J. Bacteriol.185,3661–3667
