Pro-OmpA derivatives with a His6 tag in their N-terminal "translocation initiation domains" are arrested by Ni2+ at an early post-targeting stage of translocation.

We examined in vitro translocation of pro-OmpA derivatives with a His6 tag at various positions in their mature proteins and with a c-Myc tag at their C termini across inverted membrane vesicles of Escherichia coli. Those with a His6 tag in the N-terminal region of the mature domain, which corresponds to the "translocation initiation domain" proposed previously (Andersson, H., and von Heijne, G. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 9751-9754), could not be translocated in the presence of 100 micron Ni2+, while OmpA derivatives with a His6 tag in the middle of or at the C terminus did not show such Ni2+ sensitivity. The inhibitory action of Ni2+ on pro-3His-OmpA' (with a His6 tag after the third amino acid of the mature OmpA-c-Myc region) translocation was exerted only during early events, after which it became ineffective. The inhibition point of Ni2+ was suggested to lie between membrane targeting and exposure of the signal cleavage site to the periplasm since the unprocessed and membrane-bound form of pro-3His-OmpA' was accumulated by the addition of Ni2+. The Ni(2+)-"trapped" precursor was released from its translocation block by 30 mM histidine, which should compete with the His6 tag on the precursor protein for formation of a Ni2+ chelating complex. We propose that Ni2+ confers a reversible positive charge effect on the His6-tagged initiation domain of the pro-OmpA derivatives and inhibits an early event(s) of protein translocation, such as presentation of the precursor to the membranous part of the translocase. This system will be useful in dissecting early events of the protein translocation pathway.

tag in the middle of or at the C terminus did not show such Ni 2؉ sensitivity. The inhibitory action of Ni 2؉ on pro-3His-OmpA (with a His 6 tag after the third amino acid of the mature OmpA-c-Myc region) translocation was exerted only during early events, after which it became ineffective. The inhibition point of Ni 2؉ was suggested to lie between membrane targeting and exposure of the signal cleavage site to the periplasm since the unprocessed and membrane-bound form of pro-3His-OmpA was accumulated by the addition of Ni 2؉ . The Ni 2؉ -"trapped" precursor was released from its translocation block by 30 mM histidine, which should compete with the His 6 tag on the precursor protein for formation of a Ni 2؉ chelating complex. We propose that Ni 2؉ confers a reversible positive charge effect on the His 6tagged initiation domain of the pro-OmpA derivatives and inhibits an early event(s) of protein translocation, such as presentation of the precursor to the membranous part of the translocase. This system will be useful in dissecting early events of the protein translocation pathway.
Translocation across the cytoplasmic membrane is the first step of protein targeting to the cell surface in bacterial cells. This complex biochemical reaction involving topological change of molecules has been analyzed by combined approaches of genetics and biochemistry in Escherichia coli (for reviews, see Refs. [1][2][3][4]. The biochemical studies, notably purification and reconstitution of protein translocation machinery, have revealed key players of the translocation, translocation ATPase (SecA), a secretory protein-specific chaperone (SecB), and an integral membrane component (SecY-SecE-SecG complex) (5)(6)(7)(8)(9). From in vitro analyses using inverted bacterial plasma membrane vesicles, several subprocesses in the protein translocation reaction can be envisaged: 1) recognition of preproteins by chaperones (like SecB) that retain "translocation-competent conformation" of the secretory protein precursors, 2) targeting of the preprotein-SecB complex to SecA bound to the high affinity site of the plasma membrane, 3) ATP binding-dependent partial insertion of the precursors into a translocation channel, and 4) ATP hydrolysis-coupled and ⌬-dependent bulk protein translocation (9 -13). Translocating secretory proteins are surrounded by SecA and SecY, but not by lipid molecules (14). Recently, Economou and Wickner (15) found that the movement of the secretory protein is coupled with insertion and de-insertion of a 30-kDa segment of SecA. Deep insertion of SecA into the membrane is also detected in vivo (16).
During the course of these analyses, several systems have been developed to trap translocation intermediates during post-translational protein translocation. Except for the cases of kinetic trapping of intermediates by low ATP concentration (12) and formation of a disulfide bond loop of precursor protein in the absence of proton-motive force (11), most of the methods rely on some "tightly folded" structures that block further penetration of the preproteins into the translocase. For instance, translocation of epitope-tagged preprotein was blocked by epitope-specific antibody (13). Covalent attachment of stable structures such as bovine pancreas trypsin inhibitor (12) or methotrexate-binding dihydrofolate reductase (14) moieties to the precursor protein also generates translocation intermediates. But, in all of these cases, the blockades were exerted during the events that occur in the middle of translocation of the bulk of the mature domain.
In the cotranslational translocation system in the eukaryotic endoplasmic reticulum, the ribosome-nascent chain complex offers an ideal experimental tool to define various translocation intermediate states, including those in quite early stages in translocation (17). On the other hand, early biochemical events in the bacterial post-translational system have only insufficiently been investigated due to the lack of convenient methods to accumulate "early intermediates." Mutants affected in the translocation processes provide useful clues about the early events in vivo. Especially the prl mutations in secY and secE loci, which broaden the specificity of signal sequence recognition, suggested a direct interaction between signal sequence and SecY/SecE, the main membranous subunits of the E. coli translocase. Silhavy and co-workers (18,19) found a striking clustering of prlA mutations in the first periplasmic domain and in the seventh and tenth transmembrane domains, which they proposed are essential for Se-cY's recognition of signal sequence and SecE. Our isolation of cold-sensitive and dominant sec mutations in secY suggested that the region C-terminal to transmembrane domain 8 is important for translocation facilitation and that the fourth cytoplasmic region is required for interaction with SecE (20 -22). To obtain an integrated picture of translocation in molecular terms, it is essential to analyze the nature of the early interaction between precursors and the SecY-SecE-SecG complex on the membrane in vitro. More specifically, it is highly desired to devise a new method to "trap" translocation intermediates in the early stages in vitro.
In this report, we exploited the technique of hexahistidine tagging to use His 6 -tagged precursor proteins for easy purification as well as for generation of a new type of translocation intermediates. We found that the pro-OmpA derivatives with a His 6 tag in their N-terminal regions of their mature proteins could not be translocated in the presence of a low concentration of Ni 2ϩ . Ni 2ϩ acted only on pro-OmpA derivatives with a His 6 tag in the N-terminal region of the mature sequence. This inhibition occurs only at an early stage(s) of the translocation reaction and can be released by adding histidine, which competes for chelating Ni 2ϩ with the His 6 tag in the preprotein. Ni 2ϩ did not inhibit, but rather enhanced, membrane association of His 6 -tagged OmpA, suggesting that it acts just after the membrane targeting step. This system will be suitable for dissecting the early events in bacterial protein translocation. To construct pTYE005, a His 6 fusion vector, the "His 6 oligonucleotides" 5Ј-GGAATTCATCGAAGGCCGTCACCATCACCATCACCACAT-CGATGG-3Ј and 5Ј-CCATCGATGTGGTGATGGTGATGGTGACGG-CCTTCGATGAATTCC-3Ј were annealed, digested with EcoRI and ClaI, and cloned into pBluescript SK(Ϫ) digested with the same enzymes. For the construction of a c-Myc fusion vector, pTYE006, the "c-Myc oligonucleotides" 5Ј-CCATCGATGAAGAACAGAAACTCATC-TCCGAAGAGGACCTGCTGCGCAAACGTTAAGGTACCC-3Ј and 5Ј-GGGTACCTTAACGTTTGCGCAGCAGGTCCTCTTCGGAGATGAGT-TTCTGTTCTTCATCGATGG-3Ј were annealed and cloned between the ClaI and KpnI sites of pBluescript SK(Ϫ). pTYE007, a His 6 -c-Myc fusion vector, was then constructed by ligating a 1.12-kb ClaI-ScaI fragment of pTYE005 and a 1.84-kb ClaI-ScaI fragment of pTYE006.
The His 6 and c-Myc oligonucleotides are designed to encode IEGRHH-HHHH (factor Xa site followed by a His 6 tag) and EEQKLISEEDLL-RKR-ocher (c-Myc monoclonal antibody 9E10 epitope (24)), respectively. When these double-strand oligonucleotides were cloned into pBluescript SK(Ϫ) as described above, they did not disrupt the lacZ␣ open reading frame, and the tags were encoded on another reading frame. Therefore, the lacZ assay can be used for cloning an exogenous DNA fragment into these three vectors.
To construct the OmpA-His-Myc-expressing plasmid, a 1.23-kb SspI-PstI fragment of pRD87 covering the ompA open reading frame was cloned into pTYE007 to obtain pTYE008. A 0.25-kb BglII-EcoRI fragment of pTYE008 was replaced with a 0.15-kb fragment of the 3Јterminal region of the ompA gene amplified by polymerase chain reaction with the primers 5Ј-AAAGGTATCCCGGCAGAC-3Ј and 5Ј-GGAATTCAGCCTGCGGCTGAGTTAC-3Ј and digested with BglII and EcoRI. The resulting plasmid, pTYE009, encoded an in-frame fusion between OmpA and the His 6 -c-Myc tag. A 2.94-kb EcoRI-ScaI fragment from pTYE009 was ligated with a 1.16-kb EcoRI-ScaI fragment from pTYE006 to yield pTYE018, encoding OmpA-c-Myc (abbreviated as OmpAЈ) fusion protein. To insert a His 6 tag at various locations in the OmpA mature domain, pTYE018 was mutagenized with the mutagenic primers described below. 5Ј-GTAGCGCAGGCCGCTCCGAAACAC-CATCACCATCACCATATCGATAACACCTGGTACACTGG-3Ј was used for the construction of pTYE050, encoding 3His-OmpAЈ, which has a His 6 -Ile insert after the third amino acid of the OmpA mature sequence; 5Ј-GATAACACCTGGTACCACCATCACCATCACCATAGTA-CTGGTGCTAAACTG-3Ј for pTYE086, encoding 8His-OmpAЈ, which has His 6 -Ser after the eighth amino acid; 5Ј-TCCCAGTACCATGAT-CACCATCACCATCACCATAGTACTGGTTTCATCAAC-3Ј for pTYE-098, encoding 20His-OmpAЈ, which has His 6 -Ser after the 20th amino acid; and 5Ј-CGTTTATGGTAAAAACCACCATCACCATCACCAT-GTCGACACCGGCGTTTCTCC-3Ј for pTYE112, encoding 114His-OmpAЈ, which has His 5 -Val after His-114. Mutations were confirmed by the presence of new restriction sites, underlined in the oligonucleotides (ClaI, ScaI, ScaI, and SalI, respectively). To construct SecB-overproducing plasmid pTYE025, a 1.24-kb BamHI-PvuII secB fragment derived from pAK330 (43) was subcloned into pBluescript KS(Ϫ) digested with BamHI and EcoRV.
SecA was prepared from CU148/pKY173, 2 a SecA overproducer, by a combination of Matrex gel red A dye binding column and DEAE-Sepharose Fast Flow chromatography. SecB was overproduced in TYE126 (JM109(DE3)/pTYE025) from the T7 promoter and purified as described (26) up to the Q-Sepharose column step, and the sample was further purified by butyl-Sepharose column chromatography. 35 S-Labeled pro-OmpA derivatives were prepared by in vitro transcription and translation using appropriate plasmids, E. coli S130, and [ 35 S]Met (27). Proteins were then precipitated with 5% trichloroacetic acid (final concentration) and dissolved in HU buffer (50 mM HEPES/KOH, pH 8.0, 1 M NaCl, 8 M urea, 10 mM ␤-mercaptoethanol).
Purification of Pro-OmpA Derivatives-TYE055 harboring pTYE009 or pTYE050 was cultured until mid-log phase in 2.2 liters of LB medium, 0.4% glucose. Isopropyl-␤-D-thiogalactopyranoside was added at a final concentration of 1 mM to induce His 6 -tagged OmpA derivatives as well as to overproduce Syd, which causes a severe secretion defect (28). After 2-3 h, cells were harvested and lysed by sonication in urea lysis buffer (50 mM sodium phosphate, pH 8.0, 1 M NaCl, 8 M urea, 10 mM ␤-mercaptoethanol) supplemented with 0.5 mM phenylmethylsulfonyl fluoride, 2 M pepstatin A, and 2 M leupeptin. Cell debris and membranes were removed by two consecutive centrifugations at 4500 ϫ g for 10 min and at 100,000 ϫ g for 60 min at 4°C. The soluble fraction was loaded onto a Ni 2ϩ -NTA-agarose column and washed with urea lysis buffer at 4°C. His 6 -tagged protein was eluted with a 0 -150 mM imidazole gradient, and eluate fractions were examined by Western blotting with anti-c-Myc monoclonal antibody. Proteins in the peak fractions were precipitated with 5% trichloroacetic acid (final concentration) and dissolved in 50 mM Tris-HCl, pH 8.0, 8 M urea, 10 mM ␤-mercaptoethanol. The sample was loaded onto a Hi-Trap Q column and eluted with a 0 -150 mM NaCl gradient at room temperature. Purified pro-OmpA derivatives were again precipitated with 5% trichloroacetic acid, dissolved in HU buffer, dispensed into small aliquots, and 1 The abbreviations used are: NTA, nitrilotriacetic acid; kb, kilobase(s); INV, inverted inner membrane vesicle(s); TPCK, L-1-tosylamido-2-phenylethyl chloromethyl ketone; PAGE, polyacrylamide gel electrophoresis. 2 C. Ueguchi, unpublished results. stored at Ϫ80°C. Iodination of Pro-OmpA Derivatives-100 g of a pro-OmpA derivative in HU buffer was trichloroacetic acid-precipitated and redissolved in 200 l of 0.1 M sodium phosphate, pH 7.0, 0.15 M NaCl, 8 M urea. Two IODO-BEADs (Pierce), 5 l of carrier-free Na 125 I (1.9 MBq), and 2 l of 1 mM unlabeled NaI were added to the solution. After a 10-min incubation at room temperature, the beads were removed, and dithiothreitol at a final concentration of 10 mM was added to terminate iodination. Iodinated protein was precipitated with 10% trichloroacetic acid (final concentration) and redissolved in 200 l of HU buffer, and its radioactivity and protein concentration were determined.
In Vitro Translocation Assay-An in vitro translocation assay was performed in 25 l of standard assay buffer (50 mM HEPES/KOH, pH 8.0, 50 mM KCl, 5 mM MgCl 2 , 0.1 mg/ml bovine serum albumin) as described (14) with the following modifications. 1) 1.6 mg/ml INV or 1 mg/ml 6 M urea-extracted INV was used; 2) membrane and soluble factors were premixed, and the translocation reaction was started by adding pro-OmpA derivatives; and 3) the translocation reaction was terminated at 15 min unless otherwise mentioned. Reactions without an ATP/ATP regeneration system and sodium succinate were used as negative controls. After the translocation reaction, two 10-l aliquots were withdrawn and subjected to a 10-min incubation in the presence or absence of 0.25 mg/ml TPCK-treated trypsin on ice, followed by a further 10-min incubation on ice with a 2-fold weight of chicken egg trypsin inhibitor. In the reactions with radioactive substrates, samples were subjected to SDS-PAGE, and radioactivities of pro-OmpA and mature OmpA derivatives were measured by the use of a combination of a Fuji BAS2000 analyzer and a PDI image analyzer. In the reactions with nonradioactive substrates, the OmpA species were visualized by Western blotting and quantified by a PDI image analyzer.
Binding of Pro-3His-OmpAЈ and Pro-OmpA-HisЈ to INV-The binding of pro-OmpA derivatives to INV was examined as described (9). Substrate (1 g) was mixed with 100 l of standard reaction mixture containing 1 mg/ml 6 M urea-extracted INV with appropriate supplements. After incubation for 15 min at 0°C, an 80-l aliquot was loaded on 160 l of 20% (w/v) sucrose in standard assay buffer and centrifuged at 12,100 ϫ g for 20 min to fractionate into soluble (upper phase) and membrane (pellet) fractions. Relative amounts of total, unbound, and bound pro-OmpA were determined by Western blotting. Distribution of SecA and SecY was examined similarly. In the case of the translocation assay of prebound preprotein, binding reactions were performed in 175 l with 0.35 g of 125 I-pro-3His-OmpA, and 150-l portions were subjected to centrifugation.
Protein Techniques-For Western blotting, proteins were transferred to Immobilon P polyvinylidene difluoride membrane (Millipore Corp.) and decorated with the appropriate primary antibodies. Signals were visualized by horseradish peroxidase-conjugated secondary antibodies and the ECL Western blotting detection system (Amersham Corp.) or by alkaline phosphatase-conjugated secondary antibody and a nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate system.
SecA ATPase was assayed as described in Ref. 7. Protein concentration was determined by the Bio-Rad protein assay solution with bovine ␥-globulin as a standard.

RESULTS
Purification of Precursors of OmpA Derivatives-As an application of the His 6 tag method (29,30), we constructed His 6and c-Myc-tagged OmpA derivatives as described under "Experimental Procedures." Among several in vivo conditions we examined to accumulate unprocessed precursor proteins, the tight inhibition of protein export in the secY24 mutant upon overexpression of the syd gene (28) was most effective. In our system, Syd and a precursor protein were overproduced from two compatible plasmids under the regulation of lac promoter. Plasmid-encoded proteins were detected by anti-c-Myc monoclonal antibody (24) during purification and in the in vitro translocation assay. We expressed and purified two types of His 6 -and c-Myc-tagged OmpA precursors, pro-OmpA-HisЈ with a His 6 tag at the C terminus of the OmpA sequence and pro-3His-OmpAЈ with a His 6 tag in the N-terminal region of the mature domain (Fig. 1), as described under "Experimental Procedures" in detail (Fig. 2, A and B). The proteins were kept in buffers containing 8 M urea to maintain their translocation competence in vitro. We used a DEAE-Sepharose or Hi-Trap Q anion-exchange column after the Ni 2ϩ -NTA-agarose column to remove residual contaminations and small amounts of mature proteins. Typically, 7.5 mg of OmpA precursor with a purity of Ͼ95% was obtained from 2.2 liters of culture. Purified precursor was translocated into E. coli wild-type INV in an ATP-dependent manner (Fig. 2C). In the following experiments, translocation of these purified preproteins was assayed either by Western blotting with anti-c-Myc monoclonal antibody or by quantification of the radioactivity of the 125 I-labeled preproteins after SDS-PAGE.
Effects of Ni 2ϩ on Translocation of His 6 -tagged OmpA Precursors-We examined the possible effects of Ni 2ϩ on translo-  (Fig. 3A, upper and middle panels, lanes 8 -11). In contrast, translocation of pro-3His-OmpAЈ was impaired by only 20 M NiCl 2 (Fig. 3A, lower panel,  lane 9). A similar NiCl 2 sensitivity of translocation was also demonstrated for purified pro-3His-OmpAЈ (Fig. 3B). Even the signal sequence cleavage of the precursor did not occur in the presence of NiCl 2 (Fig. 3B, lane 6). Ni 2ϩ seemed to be responsible for this inhibition because similar inhibitory effects were observed for NiSO 4 . We also tested the ability of purified pro-OmpA derivatives to activate SecA translocation ATPase in the presence of Ni 2ϩ . In the presence of 50 g/ml INV, 0.3 mg/ml pro-3His-OmpAЈ activated SecA ATPase by 6.4-fold. 120 M NiCl 2 lowered this ATPase activation to 1.5-fold. Up to 400 M Ni 2ϩ only slightly affected the lipid ATPase and intrinsic (uncoupled) ATPase activities of SecA (28 and 35% inhibition, respectively). Therefore, the inhibitory action of Ni 2ϩ on translocation of pro-3His-OmpAЈ was accompanied by inhibition of SecA translocation ATPase.
The above results suggested that the inhibition of translocation by Ni 2ϩ was dependent on the position of the His 6 tag in the OmpA protein. We constructed a series of OmpA derivatives with a His 6 tag at various positions in the mature domain of the OmpA protein (Fig. 1). The Ni 2ϩ sensitivities of their translocation were compared using in vitro translated preproteins (Fig. 4). Translocation of pro-3His-OmpAЈ (Fig. 4A, å) and pro-8His-OmpAЈ (ࡗ) with His 6 tags after the third and eighth amino acids of the mature domain, respectively, was completely inhibited by 120 M Ni 2ϩ (Fig. 4A). The same concentration of Ni 2ϩ did not inhibit translocation of pro-OmpAЈ, pro-OmpA-HisЈ, or pro-114His-OmpAЈ. Pro-20His-OmpAЈ showed an intermediate sensitivity to Ni 2ϩ (Fig. 4A, ϫ). Although 400 M Ni 2ϩ somewhat inhibited translocation of even pro-OmpA-HisЈ and pro-114His-OmpAЈ, but not of pro-OmpAЈ, we did not pursue this Ni 2ϩ effect further. We conclude that 20 -100 M Ni 2ϩ affects translocation of only precursor proteins with a His 6 tag in the N-terminal region of the mature domain.
The spectrum of the positional effect in the His 6 tag and the Ni 2ϩ -dependent translocation block (Ni 2ϩ inhibition at 120 M; summarized in Fig. 4B) reminds us of the results of Andersson and von Heijne (31), who found that the first 30 amino acid residues of the mature domain of the secretory precursor are particularly sensitive to introduction of positive charges (6 consecutive lysines). They proposed to call this region the "translocation initiation domain." The inhibitory effect of Ni 2ϩ on translocation of N-terminally His 6 -tagged pro-OmpAЈ may also result from introduction of positive charges in the translocation initiation domain by chelating Ni 2ϩ . Consistent with this notion, protein translocation of pro-3His-OmpAЈ in the presence of 80 M NiCl 2 was restored by adding NTA (Fig. 3A,  lower panel, compare lanes 6 and 10), which should interact with the His 6 tag with high affinity through chelating Ni 2ϩ (30). Because an NTA molecule has two minus charges at pH 8.0, at which we performed the translocation assay, the antagonistic activity of NTA on Ni 2ϩ inhibition may simply be explained by its charge neutralization effect. Still other explana- tions remain. For instance, a steric effect caused by introduction of Ni 2ϩ in the His 6 tag portion may prevent the conformational change required for the molecular movement through the translocation channel.
Ni 2ϩ Affects Only Early Event(s) in Translocation-We next investigated the time course of the Ni 2ϩ effects on translocation of pro-3His-OmpAЈ. NiCl 2 at a final concentration of 100 M was added at various time points to the reaction mixture, and translocation of 125 I-labeled pro-3His-OmpAЈ was allowed for a total of 25 min (Fig. 5, scheme I). As a control time course (Fig.  5, scheme II), the amounts of 3His-OmpAЈ already translocated at each time point of Ni 2ϩ addition were measured by immediately chilling the reaction mixture.
Although it has not been established to what extent the in vitro translocation reaction using the E. coli INV system is synchronized and how long it takes for a single precursor molecule to complete translocation, the following considerations will be useful for interpretation of the data obtained. If one assumes that a single cycle reaction occurs synchronously and that an inhibitory action is exerted at a specific substep of the reaction, addition of the inhibitor prior to the inhibition step in the scheme I reaction completely blocks the final yield of translocation, whereas after the inhibition step, the inhibitor is totally ineffective in lowering the final yield. As shown in Fig.  5A (E), the total amount of translocation (scheme II) increased steadily up to the 25-min incubation period examined. The translocation yields in the scheme I reaction (q) were significantly higher that those in scheme II, except for the 0-and 1-min time points. The effectiveness of the Ni 2ϩ inhibition after its addition is shown in Fig. 5B. It was found that the inhibitory action was gradually lost during the course of this in vitro reaction. This indicates that Ni 2ϩ does not inhibit the reaction uniformally at every step or, at least, the final step of translocation. Rather, the inhibition point(s) should be located early in the translocation pathway. This interpretation is also supported by the fact that Ni 2ϩ blocked the signal cleavage of pro-3His-OmpAЈ, which occurs early in translocation (12). The lack of a clear "cutoff" point in Fig. 5B may be due to asynchrony in the initial process as well as to possible random slowing down during late steps of translocation. Therefore, we conclude that Ni 2ϩ acts early in the translocation event(s).
Ni 2ϩ Does Not Inhibit Recruitment of Precursor to INV-Several events are assumed to occur early in translocation, including interaction of a soluble translocation precursor with a secretory protein-specific chaperone (SecB) and targeting of the precursor to the inner membrane. We observed that the purified pro-OmpA derivatives were eluted from the gel filtration column as a high molecular mass form with SecA in the presence of SecA and SecB irrespective of the presence of Ni 2ϩ (data not shown). This suggested that Ni 2ϩ did not affect SecA/ SecB recognition of pro-OmpA derivatives. Next, we addressed precursor recruitment to INV. We incubated pro-OmpAЈ derivatives with INV in translocation reaction mixture on ice for 15 min and isolated membranes by centrifugation through a sucrose cushion. SecY, a marker of INV, was quantitatively recovered in the pellet under these conditions, and distribution of SecA was unchanged throughout the present experiments (Fig.  6, second and third rows). Interestingly, Ni 2ϩ did not inhibit, but rather enhanced, the recovery of pro-3His-OmpAЈ in the membrane fraction (Fig. 6, lanes under pro-3His-OmpAЈ ϩ Ni 2ϩ ). A slight increase in recovery in the membrane fraction, but not as significant as that of pro-3His-OmpAЈ, was observed in the case of pro-OmpA-HisЈ (Fig. 6, lanes under pro-OmpA-HisЈ ϩ Ni 2ϩ ). Ni 2ϩ may enhance the membrane targeting of

FIG. 5. Ni 2؉ affects only early step(s) of translocation reaction.
Two types of translocation reactions were performed as summarized above the graph. In reaction scheme I, a 150-l reaction was started at 37°C by adding 0.6 g (14.6 pmol) of 125 I-labeled pro-3His-OmpAЈ. At the indicated times, a 10-l aliquot was withdrawn and transferred to a new tube containing 1 l of 1.1 mM NiCl 2 and further incubated for a total of 25 min. Another 150-l reaction was carried out as described above, and a 10-l aliquot was cooled in ice water to measure the amounts of translocated 3His-OmpAЈ at each time point (scheme II). After the reaction, samples were trypsin-treated and subjected to SDS-PAGE. A, translocation efficiencies of the two reaction schemes at each time point are plotted. q, scheme I; E, scheme II. Translocation efficiency is expressed by the amount of trypsin-resistant pro-OmpA and mature OmpA derivatives in a 25-l standard reaction for comparison with other experiments. B, efficiency of Ni 2ϩ inhibition after its addition was calculated from the data at each time point in A by the following formula: % inhibition efficiency ϭ (((scheme I value) Ϫ (scheme II value))/((scheme I value at 25 min) Ϫ (scheme II value))) ϫ 100. pro-3His-OmpAЈ or stabilize the membrane-bound state of the preprotein. Under the standard translocation conditions at 37°C, the 3His-OmpAЈ species was mostly associated with INV as a mature form. In the presence of Ni 2ϩ , the protein was also recovered in the membrane fraction, but as its precursor form (data not shown).
We then subjected the membrane-targeted pro-OmpA preparation to the translocation reaction. Pro-3His-OmpAЈ that had been targeted to the membrane in the presence of Ni 2ϩ was competent for the subsequent translocation. This translocation was Ni 2ϩ -sensitive as shown in Fig. 7. Slight translocation of pro-3His-OmpAЈ was noted in the presence of Ni 2ϩ ; possibly a small fraction of pro-3His-OmpAЈ had escaped from the Ni 2ϩsensitive step during sample manipulations. These results imply that NiCl 2 inhibits a step that occurs subsequent to the recruitment of the precursor to the membrane. It might result in the apparent accumulation of pro-3His-OmpAЈ in INV through stabilizing or "locking" the membrane-bound state of the precursor.
Effect of Ni 2ϩ Is Reversible-We demonstrated that Ni 2ϩ interferes with an early but post-targeting event(s) of translocation of the N-terminally His 6 -tagged precursor. We then addressed the reversibility of this inhibition. NTA or histidine canceled the inhibitory effect of Ni 2ϩ when added to the reaction mixture prior to addition of the precursor proteins. Histidine is a competitor of His 6 in forming a chelating complex with Ni 2ϩ . In the experiment shown in Fig. 8, we first incubated the translocation mixture with 100 M NiCl 2 for 4 or 10 min to block the reaction completely (Fig. 8, Ⅺ) and then added 30 mM histidine (final concentration). Histidine restored translocation (Fig. 8, å and f, respectively). The histidine-released translocations occurred almost at the same apparent rate as the control translocation reaction. Closer inspections revealed that the release of inhibition occurred without any time lag, while a lag period of 0.5-2 min was usually observed before onset of the increase in the amount of trypsin-protected full-length OmpA derivatives in the normal translocation reaction (Fig. 8, E). The simplest interpretation of these results is that a combination of Ni 2ϩ and N-terminally His 6 -tagged preprotein acts as a trap of an intermediate state(s) on the membrane, such that complex formation with soluble Sec factors and targeting to INV are skipped when released from this trap.
We do not believe, however, that the apparent translocation rates shown in Figs. 5 and 8 represent the kinetics of translocation of individual precursor molecules. They must represent a sum of the heterogeneous population at different stages of the reactions. The fact that the Ni 2ϩ trap only shortened the initial lag period but did not enhance the apparent translocation rate upon release may suggest that there are multiple "bottleneck" processes in vitro and that some of them occur after the Ni 2ϩsensitive step. Although we need a further investigation of the molecular nature of the Ni 2ϩ -trapped intermediate, it is a new type of "reversible" inhibition of an early event of translocation in the bacterial system. DISCUSSION In this study, we made use of the His 6 tag method developed by Bush et al. (29) not only to purify chemical amounts of E. coli pro-OmpA derivatives, but also to dissect their translocation across INV of E. coli in vitro. Our system using a combination of secY24 mutation and overexpression of syd (28) will be useful to accumulate bacterial precursor proteins in E. coli cells. We found that a His 6 tag introduced into the N-terminal region of the OmpA mature domain confers Ni 2ϩ sensitivity to its translocation. Ni 2ϩ inhibits only the early step(s) of the translocation reaction, which is after the association of precursor with the membrane, but before the signal cleavage. Inhibition can be released by addition of histidine, which breaks the His 6 ⅐Ni 2ϩ chelating complex.
It is likely that the effect of Ni 2ϩ is due to introduction of positive charges to the N-terminal mature region of the precursor protein. The position-specific Ni 2ϩ effects are difficult to explain in terms of nonspecific jamming of the translocation machinery, damage to the ⌬ H ϩ generating system by heavy metal ion, or the molecular size of the His 6 ⅐Ni 2ϩ chelating complex. The fact that Ni 2ϩ inhibition was observed only when the His 6 tag was positioned within the first 20 residues or so of the mature domain of pro-OmpA suggests that the chelating complex affects some specific event(s) where the translocase interacts with this particular N-terminal region of the precursor protein. Toxicity of positively charged amino acids in the N-terminal mature domain has been reported in several secre- tory proteins in vivo and in vitro (32)(33)(34)(35). Systematic insertion of Lys 6 after the first or second transmembrane domain of leader peptidase indicates that the 30 -40 amino acid residues following the signal sequence or the signal anchor sequence form a special domain that cannot tolerate the positive charges (31). Andersson and von Heijne (31) termed this region the translocation initiation domain. The location of the His 6 tag that confers Ni 2ϩ -sensitive translocation on OmpA coincided well with the translocation initiation domain. The antagonistic effect of NTA can be explained by its minus charges and ability to form a ternary complex with the His 6 motif and Ni 2ϩ , although we do not have direct evidence of the translocation of this ternary complex. Another possibility might be that the N-terminal region of the mature protein is quite sensitive to a conformational constraint. For instance, the initial formation of the hairpin loop structure in the N terminus of preproteins could be important for translocation initiation, and the local conformation of the His 6 ⅐Ni 2ϩ complex may prevent the formation of such a structure.
Previously, the effect of positive charges in the translocation initiation domain had been discussed from the point of view of the electrochemical nature of the membrane, such as interaction of the N-terminal region of preprotein with acidic phospholipids and determination of its orientation according to the "positive inside rule" (31)(32)(33)(34)(35). However, the existence of positive charges themselves in this domain, but not charge balance flanking the signal sequence core, is essential for inhibition (34). Both in the bacterial plasma membrane and in the eukaryotic endoplasmic reticulum, which have homologous translocation machinery, compelling evidence suggests that proteinaceous pores composed of SecY-SecE-SecG or Sec61␣-Sec61␤-Sec61␥ complexes lead preproteins across the membrane (14, 36 -39). A tempting and serious possibility is that the translocation initiation domain directly interacts with some part of the translocase and that this interaction itself is sensitive to positively charged amino acids in this domain.
The existence of prlA and prlG mutants is one piece of strong evidence for direct interaction between membranous translocase subunits and precursor protein (2,18,19). Certain prlA mutants can also translocate mutant preprotein with positive charges in the translocation initiation domain in vivo (33). We found that pro-3His-OmpAЈ translocation in vitro was less sensitive to Ni 2ϩ when INV prepared from the prlA3 mutant was used. 3 We suppose that Ni 2ϩ affects translocation of N-terminally His 6 -tagged pro-OmpA through the step of precursor protein recognition by the SecY-SecE-SecG complex. A similar signal sequence recognition event by translocase is also proposed in the mammalian Sec61 system, as revealed in a reaction in the absence of signal recognition particle (40). Actually, translocation across the endoplasmic reticulum is also sensitive to positives charges in the N-terminal portion of the mature protein albeit its lower sensitivity compared with the prokaryotic system (41).
While signal sequence recognition by SecY could be regarded as essentially a proofreading activity that rejects nonfunctional precursor proteins (18), the fact that histidine addition can restore translocation without a short lag may indicate that a translocon-associated precursor can be reactivated on site, i.e. no rejection occurs on the membrane. We detected efficient cross-linking between pro-3His-OmpAЈ on the membrane with SecA irrespective of the existence of Ni 2ϩ , 3 supporting this notion. Interaction of SecY(-SecE-SecG) with the signal sequence and the N-terminal portion of the mature protein may be required for some intrinsic mechanism of translocation, such as gate opening of the translocation channel (18,40). It is interesting to point out that the prlA3 and other prlA alleles that suppress the translocation defects caused by the basic amino acids (33) or the His 6 ⅐Ni 2ϩ conjugate (this study) reside in the first periplasmic domain of the SecY protein (18,42). The idea that this region of SecY acts to accept or reject the early mature part (3) is reasonable in terms of topological consideration of SecY and preprotein; this SecY domain may recognize preprotein during or after the insertion of its hairpin loop structure composed of the signal sequence and the translocation initiation domain. But our results imply that the Ni 2ϩinhibited precursor can still remain associated with the membrane. On the other hand, the His 6 tag portion of the Ni 2ϩtrapped precursor on the membrane should not be completely buried in the translocation machinery. It is accessible to exogenous histidine added from the cytoplasmic surface. Either the intermediate-bearing translocation channel may be open to the cytoplasmic side, or this intermediate is in a fast equilibrium between a membrane-embedded state and a water-accessible state.
Since most translocation intermediate traps developed so far confer a translocation block at its middle or late stages (11)(12)(13)(14), the His 6 ⅐Ni 2ϩ -tagged precursor system is a novel tool for analyzing initial translocation events. It will be useful for investigating the nature of signal recognition by the SecY-SecE-SecG complex through in vitro analysis of translocon mutants, especially prlA mutants. Also in vitro characterization of the secY mutants with lowered secretory efficiency (22) with the His 6 ⅐Ni 2ϩ -tagged precursor system will be promising in identifying secY mutants with a deficiency in the early events.