Trigger Factor Retards Protein Export in Escherichia coli
2002; Elsevier BV; Volume: 277; Issue: 45 Linguagem: Inglês
10.1074/jbc.m205950200
ISSN1083-351X
AutoresHin C. Lee, Harris D. Bernstein,
Tópico(s)Toxin Mechanisms and Immunotoxins
ResumoTrigger factor (TF) is a ribosome-associated protein that interacts with a wide variety of nascent polypeptides inEscherichia coli. Previous studies have indicated that TF cooperates with DnaK to facilitate protein folding, but the basis of this cooperation is unclear. In this study we monitored protein export in E. coli that lack or overproduce TF to obtain further insights into its function. Whereas inactivation of genes encoding most molecular chaperones (including dnaK) impairs protein export, inactivation of the TF gene accelerated protein export and suppressed the need for targeting factors to maintain the translocation competence of presecretory proteins. Furthermore, overproduction of TF (but not DnaK) markedly retarded protein export. Manipulation of TF levels produced similar effects on the export of a cytosolic enzyme fused to a signal peptide. The data strongly suggest that TF has a unique ability to sequester nascent polypeptides for a relatively prolonged period. Based on our results, we propose that TF and DnaK promote protein folding by distinct (but complementary) mechanisms. Trigger factor (TF) is a ribosome-associated protein that interacts with a wide variety of nascent polypeptides inEscherichia coli. Previous studies have indicated that TF cooperates with DnaK to facilitate protein folding, but the basis of this cooperation is unclear. In this study we monitored protein export in E. coli that lack or overproduce TF to obtain further insights into its function. Whereas inactivation of genes encoding most molecular chaperones (including dnaK) impairs protein export, inactivation of the TF gene accelerated protein export and suppressed the need for targeting factors to maintain the translocation competence of presecretory proteins. Furthermore, overproduction of TF (but not DnaK) markedly retarded protein export. Manipulation of TF levels produced similar effects on the export of a cytosolic enzyme fused to a signal peptide. The data strongly suggest that TF has a unique ability to sequester nascent polypeptides for a relatively prolonged period. Based on our results, we propose that TF and DnaK promote protein folding by distinct (but complementary) mechanisms. inner membrane alkaline phosphatase β-lactamase inner membrane protein isopropylthiogalactoside maltose-binding protein outer membrane protein phosphoglycerate kinase ribose-binding protein signal recognition particle trigger factor hemagglutinin Protein biogenesis in Escherichia coli is aided by a diverse set of specialized factors that interact with nascent polypeptides chains. Several of these factors ("molecular chaperones") promote the folding of cytoplasmic proteins by binding promiscuously to exposed hydrophobic surfaces and preventing aggregation (reviewed in Ref. 1Frydman J. Annu. Rev. Biochem. 2001; 70: 603-647Crossref PubMed Scopus (944) Google Scholar). DnaK, a member of the highly conserved hsp70 family of chaperones, promotes protein folding at an early stage by binding to short, extended peptides in an ATP-dependent cycle that is regulated by two co-chaperones. Substrate folding progresses in a stepwise fashion through release and repeated rebinding. Proteins that have not completely folded after interaction with DnaK but that have acquired considerable secondary structure are subsequently transferred to GroEL, another ubiquitous chaperone. GroEL is a tetradecameric complex that also utilizes an ATPase cycle regulated by a co-chaperone to drive protein folding. By binding to exposed hydrophobic surfaces, molecular chaperones also "target" nascent and newly synthesized presecretory proteins to the inner membrane (IM)1 (2Bernstein H.D. Rapoport T.A. Walter P. Cell. 1989; 58: 1017-1019Abstract Full Text PDF PubMed Scopus (27) Google Scholar). Chaperones maintain presecretory proteins in a loosely folded conformation, which is required for their transport across the IM and to keep signal peptides accessible to gate open the channel that mediates the translocation reaction (the SecY complex or "translocon"). A subset of presecretory proteins are targeted to the E. coli IM by SecB, a nonessential "export-specific" chaperone found only in Gram-negative bacteria (3Kumamoto C.A. Beckwith J. J. Bacteriol. 1985; 163: 267-274Crossref PubMed Google Scholar). Like other members of the hsp70 family, DnaK promotes protein export and can at least partially substitute for SecB (4Wild J. Altman E. Yura T. Gross C.A. Genes Dev. 1992; 6: 1165-1172Crossref PubMed Scopus (171) Google Scholar,5Wild J. Rossmeissl P. Walter W.A. Gross C.A. J. Bacteriol. 1996; 78: 3608-3613Crossref Google Scholar). 2H.-Y. Qi, J. B. Hyndman, and H. D. Bernstein, manuscript submitted for publication. GroEL may also play a role in protein sorting (6Bochkareva E.S. Lissin N.M. Girshovich A.S. Nature. 1988; 336: 254-257Crossref PubMed Scopus (255) Google Scholar, 7Kusukawa N. Yura T. Ueguchi C. Akiyama Y. Ito K. EMBO J. 1989; 8: 3517-3521Crossref PubMed Scopus (178) Google Scholar). Despite a similarity in function, SecB differs considerably from DnaK and GroEL in its structure and mechanism of action. SecB is a homotetramer that interacts with ∼150-amino acid polypeptide segments via an ATP-independent kinetic partitioning mechanism (8Hardy S.J.S. Randall L.L. Science. 1991; 251: 439-443Crossref PubMed Scopus (186) Google Scholar). The x-ray structure of SecB suggests that substrates are accommodated in a long continuous channel on the surface of the tetramer (9Xu Z. Knafels J.D. Yoshino K. Nat. Struct. Biol. 2000; 7: 1172-1177Crossref PubMed Scopus (101) Google Scholar). Another factor, the signal recognition particle (SRP), specifically promotes the insertion of inner membrane proteins (IMPs) into the IM (reviewed in Ref. 10de Gier J.-W. Luirink J. Mol. Microbiol. 2001; 40: 314-322Crossref PubMed Scopus (82) Google Scholar). SRP is an essential ribonucleoprotein complex that, like DnaK and GroEL, is universally conserved. E. coliSRP, which is much smaller than its eukaryotic counterpart, consists of a single protein (Ffh) and a ∼100-nucleotide RNA (4.5 S RNA). Strictly speaking, SRP is not a chaperone in that it recognizes substrates with much greater specificity than factors like DnaK (2Bernstein H.D. Rapoport T.A. Walter P. Cell. 1989; 58: 1017-1019Abstract Full Text PDF PubMed Scopus (27) Google Scholar). Although mammalian SRP binds to both signal peptides of presecretory proteins and the initial transmembrane segments of polytopic membrane proteins (which often lack cleaved signal peptides) as they emerge from translating ribosomes, E. coli SRP binds primarily to the latter class of targeting signals (11Neumann-Haefelin C. Schäfer U. Müller M. Koch H.-G. EMBO J. 2000; 19: 6419-6426Crossref PubMed Scopus (110) Google Scholar, 12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar). SRP targets ribosome-nascent chain complexes to the IM and then releases its cargo upon interaction with a peripheral membrane receptor (FtsY) in a GTP-dependent reaction (13Miller J.D. Bernstein H.D. Walter P. Nature. 1994; 367: 657-659Crossref PubMed Scopus (187) Google Scholar, 14Valent Q.A. Scotti P.A. High S. de Gier J.-W.L. von Heijne G. Lentzen G. Wintermeyer W. Oudega B. Luirink J. EMBO J. 1998; 17: 2504-2512Crossref PubMed Scopus (245) Google Scholar). Like proteins targeted by chaperone-based mechanisms, SRP substrates are transported across the IM by the SecY complex. Trigger factor (TF), an abundant ∼50-kDa protein found in all eubacteria, is the least understood of all the E. colinascent chain binding factors. Although TF was first purified as a ribosome-associated factor that can stabilize the translocation-competent form of a secretory precursor in vitro (15Crooke E. Guthrie B. Lecker S. Lill R. Wickner W. Cell. 1988; 54: 1003-1011Abstract Full Text PDF PubMed Scopus (131) Google Scholar, 16Lill R. Crooke E. Guthrie B. Wickner W. Cell. 1988; 54: 1013-1018Abstract Full Text PDF PubMed Scopus (120) Google Scholar), subsequent experiments indicated that it is not essential for either cell viability or protein export in vivo (17Guthrie B. Wickner W. J. Bacteriol. 1990; 172: 5555-5562Crossref PubMed Google Scholar). In biochemical assays, TF can be cross-linked to virtually all nascent secretory and cytoplasmic proteins (18Valent Q.A. Kendall D.A. High S. Kusters R. Oudega B. Luirink J. EMBO J. 1995; 14: 5494-5505Crossref PubMed Scopus (238) Google Scholar, 19Hesterkamp T. Hauser S. Lütcke H. Bukau B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 4437-4441Crossref PubMed Scopus (204) Google Scholar) and exhibits ATP-independent chaperone-like activities (20Scholz C. Stoller G. Zarnt T. Fischer G. Schmid F.X. EMBO J. 1997; 16: 54-58Crossref PubMed Scopus (176) Google Scholar). Unlike typical molecular chaperones, however, TF does not appear to recognize exposed hydrophobic surfaces (21Patzelt H. Rüdiger S. Brehmer D. Kramer G. Vonderwülbecke S. Schaffitzel E. Waitz A. Hesterkamp T. Dong L. Schneider-Mergener J. Bukau B. Deuerling E. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 14244-14249Crossref PubMed Scopus (149) Google Scholar). TF is also homologous to FK506-binding proteins and displays a peptidyl-prolyl-cis-isomerase activity (19Hesterkamp T. Hauser S. Lütcke H. Bukau B. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 4437-4441Crossref PubMed Scopus (204) Google Scholar, 22Stoller G. Rücknagel K.P. Nierhaus K. Schmid F.X. Fischer G. Rahfeld J.-U. EMBO J. 1995; 14: 4939-4948Crossref PubMed Scopus (227) Google Scholar) of unknown significance. The first important insights into the function of TF in vivo emerged from the observation that null alleles of the trigger factor gene (tig) anddnaK are synthetically lethal (23Deuerling E. Schulze-Specking A. Tomoyasu T. Mogk A. Bukau B. Nature. 1999; 400: 693-696Crossref PubMed Scopus (409) Google Scholar, 24Teter S.A. Houry W.A. Ang D. Tradler T. Rockabrand D. Fischer G. Blum P. Georgopoulos C. Hartl F.U. Cell. 1999; 97: 755-765Abstract Full Text Full Text PDF PubMed Scopus (350) Google Scholar). A large number of cytosolic proteins aggregate in cells that lack both DnaK and TF. Deletion of tig also results in an increase in the number of nascent polypeptide chains that associate with DnaK. These results suggest that TF and DnaK "cooperate" in the folding of newly synthesized proteins, but the molecular basis for this cooperation is unclear. Because TF can be cross-linked to very short nascent chains, it may function upstream of DnaK. Intriguingly, eukaryotic cells contain a factor called the nascent chain-associated complex that can be cross-linked to nascent chains as short as 17 amino acids (25Wang S. Sakai H. Wiedmann M. J. Cell Biol. 1995; 130: 519-528Crossref PubMed Scopus (105) Google Scholar). Despite a complete lack of sequence homology, it is conceivable that nascent chain-associated complex and TF fulfill similar roles in protein folding. A recent study that we conducted on the mechanism by which different classes of proteins are routed into distinct targeting pathways led to the more extensive examination of TF function described in this report.In vitro cross-linking data suggested that E. coli SRP fails to recognize signal peptides because TF interacts with sequences near the N terminus of the mature region of presecretory proteins and occludes SRP binding (26Beck K., Wu, L.-F. Brunner J. Müller M. EMBO J. 2000; 19: 134-143Crossref PubMed Scopus (147) Google Scholar). Experiments using modified presecretory proteins, however, led to the conclusion that the targeting pathway of E. coli proteins under physiological conditions is dictated by the hydrophobicity of their targeting signals and not by the binding of TF (12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar). In the present study we initially set out to corroborate this conclusion. Surprisingly, during the course of our investigation we found that unlike typical molecular chaperones, TF impedes protein export. Our data suggest that TF sequesters both nascent presecretory and cytoplasmic proteins in a distinctive fashion. The results not only provide novel insights into protein targeting, but may also yield important clues about the relationship of TF and DnaK in protein folding. Rabbit polyclonal antisera were obtained from the following sources: 5 Prime → 3 Prime, Inc., Boulder, CO (AP, Bla, and CAT), New England Biolabs (maltose-binding protein, MBP), Covance (HA), Jon Beckwith (ribose-binding protein, RBP), Greg Phillips (OmpC and OmpF), P. C. Tai (OmpA), Don Oliver (SecA), and Tim Yahr (SecY). The bacterial strains used in this study and their genotypes are MC4100 (F− araD139 Δ(argF-lac)U169 rpsL150 relA1 thi fib5301 deoC1 ptsF25 rbsR), HDB37 (MC4100 araΔ714), HDB38 (HDB37tig::cat), HDB55 (MC4100secB::kan), HDB56 (MC4100tig::cat), and HDB57 (MC4100secB::kan tig::cat). HDB55 contains thesecB::kan allele from strain CK1953 (27Kumamoto C.A. Francetic O. J. Bacteriol. 1993; 175: 2184-2188Crossref PubMed Google Scholar). HDB56 is identical to MC4100tig::CamR (24Teter S.A. Houry W.A. Ang D. Tradler T. Rockabrand D. Fischer G. Blum P. Georgopoulos C. Hartl F.U. Cell. 1999; 97: 755-765Abstract Full Text Full Text PDF PubMed Scopus (350) Google Scholar), which was kindly provided by Ulrich Hartl. Medium preparation and basic bacterial manipulations were performed using standard techniques. Except as noted, all experiments were conducted at 37 °C. Selective media contained 100 μg/ml ampicillin and 30 μg/ml chloramphenicol as required. Plasmids pHL12, pJH28, and pTRC-ftsY (G385A) have been described previously (12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar, 28Ulbrandt N.D. Newitt J.A. Bernstein H.D. Cell. 1997; 88: 187-196Abstract Full Text Full Text PDF PubMed Scopus (293) Google Scholar). To construct plasmid pHL20, the dnaK gene was first amplified by PCR using the oligonucleotides 5′-TATTACAGAGCTCACAACCACATGATGA-3′ and 5′-ATCTCGTAATAAGCTTGCTTAGCCATCT-3′ and pS368 (28Ulbrandt N.D. Newitt J.A. Bernstein H.D. Cell. 1997; 88: 187-196Abstract Full Text Full Text PDF PubMed Scopus (293) Google Scholar) as a template. The amplified DNA was then digested with SacI andHindIII and cloned into the cognate sites of pTRC99a (Amersham Biosciences). All other E. coli genes were amplified by PCR using genomic DNA (strain LE392) as a template. Plasmid pJH42 was constructed by amplifying the tig gene using the oligonucleotides 5′-GTGACGGGCCTTTGTGCCAATTGAGCGCGTTAT-3′ and 5′-ACCTGTGCTTGCGGGGTAACAATTGACCGAGC-3′. The amplified DNA was digested with MunI and cloned into the EcoRI site of pBAD18 (29Guzman L.-M. Belin D. Carson M.J. Beckwith J. J. Bacteriol. 1995; 177: 4121-4130Crossref PubMed Scopus (3978) Google Scholar). An NheI-HindIII fragment of pJH42 was cloned into pBAD30 (30Newitt J.A. Bernstein H.D. J. Biol. Chem. 1998; 273: 12451-12456Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar) to generate pJH43. To construct pHL21 and pHL22, E171K and F233Y mutations were introduced into thednaK and tig genes in pHL20 and pJH42, respectively, using the QuikChange mutagenesis kit (Stratagene). The gene encoding AP (phoA) was amplified by PCR and cloned behind the lac promoter of pRB11 (12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar). A DNA fragment containing Plac -phoA andlacI q was then cloned into the AseI and XbaI sites of pACYC184 to make pJH48. The pgkgene was amplified and HA-tagged using the oligonucleotides 5′-TGCTAACCGAATGCTCTAGACGACGTTAGCTA-3′ and 5′-CCCAAGCTTACAGGCTCGCATAATCCGGCACATCGTACGGATAACACTTCTTAGCGCGCTCTTCGAGCATC-3′. The amplified DNA was digested with XbaI andHindIII and cloned into pTRC99a to produce pHL32. The tagged version of pgk was then reamplified using the oligonucleotides 5′-ATTCACCATGTCTGAGATCTTGATGACCGATC-3′ and 5′-CCCAAGCTTACAGGCTCGCATAATC-3′ and pHL32 as a template. To construct pHL33, the PCR product was digested with BglII andHindIII and cloned into a derivative of pMAL-p2x (New England Biolabs) containing a BglII site at position 1607. Plasmid pHL36 was constructed by cloning anEcoRI-HindIII fragment of RB11-OmpA-HA (12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar) containing an HA-tagged version of ompA into pTRC99a. An I8N mutation was then introduced into the ompA gene in pHL36 as described above to generate pHL37. For most experiments, cells were grown in M9 medium containing 0.2% glucose. For experiments using HDB37 and HDB38, cultures were grown in M9 containing 0.2% glycerol. Overnight cultures were washed and diluted into fresh medium atA 550 = 0.025. When cultures reachedA 550 = 0.2–0.3 (after 4–5 h), cells were either radiolabeled or an inducer was added to drive the expression of a plasmid-borne gene. The synthesis of TM-OmpA, wild-type DnaK, MBP, and TF was induced by the addition of 1 mm IPTG, 100 μm IPTG, 50 μm IPTG, and 0.2% arabinose, respectively, and cells were radiolabeled 1 h later. The synthesis of OmpA-HA was induced by the addition of 50 μm IPTG and cells were radiolabeled 30 min later. Likewise, the synthesis of dominant negative alleles of ftsY and dnaK were induced by the addition of 2 mm and 100 μmIPTG, respectively, and cells were radiolabeled 40 min later. In temperature shift experiments, cultures were shifted to 22 °C when they reached A 550 = 0.2–0.3 and radiolabeled 1 h later. In some experiments, chloramphenicol (0.3–30 μg/ml) was added to cultures 15 min prior to labeling or sodium azide (2 mm) was added 2 min prior to labeling. In all pulse labeling experiments, cultures were radiolabeled for 30 s with 30 mCi/ml Tran35S-label (ICN). In pulse-chase experiments, cultures were radiolabeled for 20 s. Cold cysteine and methionine (1 mm) were then added and incubation was continued for the indicated time. Proteins were precipitated with ice-cold trichloroacetic acid, and immunoprecipitations were performed essentially as described (30Newitt J.A. Bernstein H.D. J. Biol. Chem. 1998; 273: 12451-12456Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). The membrane insertion of newly synthesized TM-OmpA was assessed as previously described (12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar). The localization of OmpA was analyzed using an established gel mobility assay (31Freudl R. Schwarz H. Stierhof Y.-D. Gamon R. Hindennach I. Henning U. J. Biol. Chem. 1986; 261: 11355-11361Abstract Full Text PDF PubMed Google Scholar) with slight modifications. Pulse-labeled cells were first converted to spheroplasts (28Ulbrandt N.D. Newitt J.A. Bernstein H.D. Cell. 1997; 88: 187-196Abstract Full Text Full Text PDF PubMed Scopus (293) Google Scholar), and 2% Triton X-100 and a protease inhibitor mixture (Sigma) were then added. Following a 5-min incubation on ice, cell lysates were centrifuged for 5 min at 20,800 × g in a microcentrifuge, the supernatants were divided in half, and anti-OmpA was added. Antibody-antigen complexes were bound to protein A-agarose beads, the beads were washed as described (30Newitt J.A. Bernstein H.D. J. Biol. Chem. 1998; 273: 12451-12456Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar), and OmpA was eluted by incubating the samples at either 60 °C for 20 min or 99 °C for 5 min. Protein samples were analyzed by SDS-PAGE on 8–16% minigels (Novex). Radiolabeled proteins were visualized using a Fuji BAS 2500 PhosphorImager, and percent protein export was calculated using a previously described formula (32Siegel V. Walter P. J. Cell Biol. 1985; 100: 1913-1921Crossref PubMed Scopus (100) Google Scholar). Western blotting was performed as described (28Ulbrandt N.D. Newitt J.A. Bernstein H.D. Cell. 1997; 88: 187-196Abstract Full Text Full Text PDF PubMed Scopus (293) Google Scholar). Whereas SecB is normally required for the efficient export of many outer membrane proteins (OMPs) and a few periplasmic proteins such as MBP, we found that these proteins could be diverted into the SRP pathway simply by increasing the hydrophobicity of their signal peptides (12Lee H.C. Bernstein H.D. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 3471-3476Crossref PubMed Scopus (164) Google Scholar). Based on our results, we inferred that chaperones provide a default targeting mechanism for proteins that are bypassed by SRP and that TF does not play an active role in targeting pathway selection. To obtain direct evidence that TF is not involved in protein targeting, we examined the export of a variety of presecretory proteins in tig− strains. We predicted that proteins whose export is SecB-dependent in wild-type cells, for example, would remain SecB-dependent in the absence of TF. Contrary to our expectation, we found that the SecB requirement in the export of several different proteins was partially or completely suppressed by inactivation of the tig gene. Wild-type MC4100 and secB−, tig−, andsecB−/tig− derivatives (HDB55, HDB56, and HDB57, respectively), which all grew at essentially the same rate under our experimental conditions, were pulse-labeled and OmpA, OmpC, OmpF, or MBP were immunoprecipitated. In these experiments export efficiency was assessed by comparing the relative amounts of the precursor and mature forms of each protein. As observed in many other studies, >90% of all four proteins was exported in wild-type cells within the pulse labeling period (Fig. 1 A,lane 1). By contrast, the vast majority of each protein was retained in the cytoplasm of secB− cells (Fig.1 A, lane 2). Consistent with a previous study in which OmpA was used as a model presecretory protein (17Guthrie B. Wickner W. J. Bacteriol. 1990; 172: 5555-5562Crossref PubMed Google Scholar), all of the proteins were rapidly exported in tig− cells (Fig.1 A, lane 3). Interestingly, the small amount of pro-OmpA, pro-OmpC, and pro-OmpF observed in MC4100 was not detected in HDB56. This highly reproducible result (see Figs.2 A, 3, 4, and 6 D) raised the possibility that the OMP precursors were exported faster in the tig− strain than in the wild-type strain (see below). Surprisingly, >95% of each OMP and about half of the MBP appeared to be exported in HDB57 (Fig. 1 A, lane 4). To confirm that the presecretory proteins were properly translocated across the IM in the secB−/tig− strain, we assessed the localization of OmpA using an independent assay. Cells were pulse-labeled and OmpA was immunoprecipitated under nondenaturing conditions. Previous work has shown that integration of OmpA into the OM produces an altered conformation that is stable at moderate temperatures (31Freudl R. Schwarz H. Stierhof Y.-D. Gamon R. Hindennach I. Henning U. J. Biol. Chem. 1986; 261: 11355-11361Abstract Full Text PDF PubMed Google Scholar). This form of the protein (OmpA*) migrates rapidly on SDS-PAGE. The finding that only OmpA* was detected in HDB57 when the sample was heated at 60 °C (Fig. 1 B, lane 7) provides direct evidence that SecB substrates are efficiently exported in the secB−/tig− strain. In addition, Western blot analysis showed that tig inactivation did not suppress the SecB requirement by stimulating the production of SecY or SecA, two key components of the translocation machinery (Fig. 1 C). These results are remarkable in that the elimination of multiple chaperones and chaperone-like factors usually compounds, rather than ameliorates, export defects.Figure 2Disruption of tig enhances Bla export in SRP-deficient cells. A, MC4100, HDB55, HDB56, and HDB57 were transformed with pTRC-ftsY (G385A). Cultures were divided in half, and IPTG was added to one portion to induce expression of the mutant ftsY allele. Cells were pulse-labeled and OmpA and Bla were immunoprecipitated. B, sodium azide (NaN3) was added to MC4100 and HDB56 transformed with pTRC99a 2 min prior to labeling. Cells were pulse-labeled and incubated for the indicated chase period, and Bla was immunoprecipitated.View Large Image Figure ViewerDownload (PPT)Figure 3OmpA is not rerouted into the DnaK targeting pathway in tig− cells. MC4100, HDB55, HDB56, and HDB57 were transformed with pTRC99a or pHL21 (Ptrc -dnaK E171K). IPTG was added to induce expression of the mutant dnaK allele. Cells were pulse-labeled and OmpA was immunoprecipitated. Lanes 5–7were overexposed ∼10-fold.View Large Image Figure ViewerDownload (PPT)Figure 4Retardation of protein synthesis does not suppress the SecB requirement in protein export. MC4100 and HDB55 were pulse-labeled 15 min after the addition of the indicated amounts of chloramphenicol (CM), and OmpA was immunoprecipitated.Lanes 7 and 8 were overexposed ∼10-fold.View Large Image Figure ViewerDownload (PPT)Figure 6Overexpression of tigspecifically retards protein export. A, HDB37 (MC4100 Δara) was transformed with pJH42 (PBAD -tig) and pJH48 (encoding AP) (top panel), pJH42 alone (middle panel), or pJH43 (PBAD -tig) and pJH28 (encoding MBP) (bottom panel). Cultures were divided in half and arabinose was added to one portion (+Ara) to induce expression of tig. Cells were pulse-labeled and the indicated protein was immunoprecipitated. B, HDB38 (HDB37 tig−) was transformed with pJH42 or pHL22 (PBAD -tig F233Y) and arabinose was added to half of the cells as in A to induce overproduction of either wild-type TF (WT) or the TF F233Y mutant (F233Y). Cells were pulse-labeled and incubated for the indicated chase period, and OmpA was immunoprecipitated.C, HDB37 was transformed with pJH42 and pHL12 (encoding TM-OmpA) and arabinose was added to half of the cells as inA. Pulse-labeled cells were converted to spheroplasts and treated with proteinase K. TM-OmpA was immunoprecipitated along with chloramphenicol acetyltransferase (CAT), which served as a cytoplasmic marker. Proteinase K was added to the samples inlanes 1 and 3. D, MC4100 and HDB56 were transformed with pHL20 (Ptrc -dnaK). Cultures were divided in half and IPTG was added to one portion (+IPTG) to induce expression of dnaK. Cells were pulse-labeled and the indicated protein was immunoprecipitated.View Large Image Figure ViewerDownload (PPT) To test the possibility that tig inactivation suppresses the SecB requirement by relaxing the constraints on protein export and permitting the translocation of proteins that would normally remain in the cytoplasm, we examined the fate of an OmpA signal peptide mutant insecB−/tig− cells. MC4100 and HDB57 were transformed with a plasmid that encodes an HA-tagged version of wild-type OmpA (pHL36) or the OmpA I8N mutant (pHL37). The kinetics of export are monitored by pulse-chase analysis. As previously observed (33Goldstein J. Lehnhardt S. Inouye M. J. Biol. Chem. 1991; 266: 14413-14417Abstract Full Text PDF PubMed Google Scholar), only a tiny fraction of the OmpA I8N mutant was exported in MC4100 (Fig. 1 D, lanes 5–8). A slightly greater percentage of the protein was exported in HDB57, perhaps because presecretory proteins are targeted to the IM earlier intig− cells and therefore have a longer opportunity to interact effectively with the translocon (see below). In any case, the results imply that the absence of TF does not fundamentally alter the fidelity of the export process. We next used β-lactamase (Bla) as a model protein to examine the effect of tig inactivation on the fate of proteins whose export does not require SecB. Although the Bla targeting pathway(s) has never been clearly identified, previous studies have shown that a variety of physiological insults that increase the presence of abnormal proteins (and that likely induce the heat shock response) lead to Bla export defects (7Kusukawa N. Yura T. Ueguchi C. Akiyama Y. Ito K. EMBO J. 1989; 8: 3517-3521Crossref PubMed Scopus (178) Google Scholar, 34Izard J.W. Rusch S.L. Kendall D.A. J. Biol. Chem. 1996; 271: 21579-21582Abstract Full Text Full Text PDF PubMed Scopus (45) Google Scholar). These and other results (5Wild J. Rossmeissl P. Walter W.A. Gross C.A. J. Bacteriol. 1996; 78: 3608-3613Crossref Google Scholar) suggest that Bla is targeted by one or more heat shock-regulated chaperones such as DnaK or GroEL. Inactivation of the SRP pathway has often been observed to delay Bla export, but as previously proposed (35Poritz M.A. Bernstein H.D. Strub K. Zopf D. Wilhelm H. Walter P. Science. 1990; 250: 1111-1117Crossref PubMed Scopus (210) Google Scholar) this effect is probably an indirect consequence of perturbing the function of heat shock proteins. Analysis of Bla export in tig− cells yielded additional unexpected results. To inhibit Bla export, we first transformed MC4100, HDB55, HDB56, or HDB57 with pTRC-FtsY(G385A), a plasmid that contains the dominant negative ftsY G385A allele under control of the strong IPTG-inducible trc promoter. The SRP pathway was then inactivated by overexpression of the mutant ftsY allele and cells were pulse-labeled. In the absence of IPTG, Bla was exported rapidly in every strain (Fig. 2 A, lanes 1–4). As previously observed (29Guzman L.-M. Belin D. Carson M.J. Beckwith J. J. Bacteriol. 1995; 177: 4121-4130Crossref PubMed Scopus (3978) Google Scholar), the addition of IPTG slightly inhibited Bla export in MC4100 (Fig. 2 A, lane 5). Overexpression of ftsY G385A in HDB55, however, resulted in the retention of almost half of the pre-Bla in the cytoplasm (Fig.2 A, lane 6). Because Bla is not a SecB substrate, the stronger export bl
Referência(s)