In Vitro Analysis of the Stop-transfer Process during Translocation across the Cytoplasmic Membrane of Escherichia coli *

In this study, using a derivative of proOmpA containing an artificial stop-transfer sequence (proOmpA2xH1), we analyzed the process of stop-transfer during translocation across the cytoplasmic membrane of Escherichia coli. ProOmpA2xH1 did not interfere with the transit of wild-type proOmpA. When proOmpA2xH1 was anchored in the membrane, membrane-inserted SecA was deinserted with the reversion of the inverted topology of SecG. Cross-linking experiments revealed that the anchored proOmpA2xH1 that does not interact with either SecY or SecA. These results, taken together, suggest that proOmpA2xH1 leaves the translocation pathway by means of a specific interaction between the stop-transfer sequence and the translocational channel.

In Escherichia coli, the translocation of proteins across the cytoplasmic membrane is mediated through protein-protein interactions involving a set of Sec proteins (1)(2)(3)(4). Secretory proteins are synthesized with an NH 2 -terminal signal sequence that targets them for export. The amino-terminal positive charges and central hydrophobic region of the signal sequence are important for the interaction of preproteins with SecA (5), a translocation ATPase (6). Upon the binding of ATP and a preprotein, SecA changes its conformation (7), and a portion of SecA is inserted deep into the cytoplasmic membrane (8,9), which can be detected as the appearance of a protease-resistant ϳ30-kDa fragment (8). The hydrolysis of ATP results in deinsertion of the SecA (10). Coupled with this insertion-deinsertion cycle, the membrane topology of SecG becomes inverted (11). SecG, SecE, and SecY are components of the membraneembedded translocase (3) that provides a proteinaceous channel for the transit of preproteins (12).
The transit of preproteins across the membrane is stopped by particular amino acid sequences termed "stop-transfer sequence" (13). Although the hydrophobicity of the stop-transfer sequence is important for its function, the details of the mechanism of the stop-transfer process remains unclear. One possibility is that the stop-transfer sequence specifically interacts with the translocase via its hydrophobic region, and induces a conformational change of the translocase (14). Another possi-bility is that the stop-transfer sequence simply leaves the translocase as a consequence of the interaction with the hydrophobic core of the lipid bilayer (15).
Previously, we showed that the short hydrophobic segments in the mature region of proOmpA interact with the membrane translocase during polypeptide transit across the cytoplasmic membrane of E. coli (16), which causes a discontinuous mode of polypeptide translocation (17). These segments are less hydrophobic than the threshold required for the cessation of polypeptide transit. However, when such a short hydrophobic segment was duplicated, the resultant segment showed substantial stop-translocation activity (16). In this study, using a derivative of proOmpA containing the stop-transfer segment, we analyzed the process of stop-transfer. Based on the results obtained, we propose that the hydrophobic stop-transfer segment leaves the translocation pathway by means of a specific interaction between the hydrophobic segment and the translocation channel, which triggers the stop-transfer.
Materials-Na 125 I (100 mCi/ml) was purchased from ICN. IODO-GEN was from Pierce. ATP, AMP-PNP, 1 creatine kinase, and creatine phosphate were obtained from Boehringer Mannheim. Proteinase K was purchased from Merck. Restriction enzymes were from Takara Shuzo Co. Sephadex G-50 (medium) and protein A-Sepharose CL-4B were purchased from Pharmacia Biotech Inc. Ni 2ϩ -NTA-agarose was from QIAGEN. A polyclonal anti-SecG antibody was kindly provided by Drs. K.-i. Nishiyama and H. Tokuda. A horseradish peroxidase-conjugated anti-rabbit antibody was from Bio-Rad.
Everted membrane vesicles for in vitro translocation were isolated from E. coli K003 (Lpp Ϫ , ⌬uncB-C-Tn10) as described previously (25). SecA was purified as described previously (5). The wild-type and mutant proOmpAs were purified as described by Crooke et al. (26) from E. coli RK4788 harboring the plasmids encoding them, and the disulfide bridge of proOmpAL43 was formed according to the method of Uchida et al. (17). SecB was purified as described by Weiss et al. (27). All proOmpA derivatives used for the in vitro translocation reactions were synthesized in vitro in the presence of EXPRE 35 S 35 S protein labeling mix (NEN Life Science Products), and partially purified as described previously (17).
Construction of ProOmpA Derivatives-Plasmids pOA43, pTDL43, and pTDL43C302Q were prepared as described previously (16). To add a His 6 tag to the carboxyl terminus of proOmpA2xH1, oligonucleotide-directed mutagenesis was performed according to the method of Kunkel (28). For the detection of mutation, an AflII site was also deleted. Plasmid pTD-2xH1 (16) was transformed into E. coli CJ236, and then the uracil-containing single-stranded phagemid DNA was isolated. Using a primer (5Ј-ACCAGACGAGAACTTAATGGTGATGGTGATGG-TGAGCCTGCGGCTGAG-3Ј; the mutated AflII site is underlined), the complementary DNA strand was synthesized, and the resulting doublestranded DNA was transformed into JM109. Plasmid pTD-2xH1-His6 was isolated on the basis of a lack of the AlfII site. Plasmids pKS-ompA and pKS-2xH1 were prepared by the insertion of a 1.3-kilobase pair EcoRI-HindIII fragment of pOA13 and pTD-2xH1, respectively, into the corresponding site of pSTV28.
Cell Fractionation-Cells of RK4788 harboring pKS-ompA or pKS-2xH1 were grown in LB medium, and at A 600 ϭ 0.6 the cells were harvested. Half of the cells were precipitated with trichloroacetic acid to give a "whole cell" sample. The other half were pelleted, and then inner membrane vesicles were isolated according to the method of Yamada et al. (25).
Purification of His 6 -tagged ProOmpA2xH1-KI297 cells harboring pST30 and pTD-2xH1-His 6 were cultured until the mid-log phase in LB medium supplemented with 50 g/ml ampicillin, 10 g/ml chrolamphenicol, and 0.4% glucose. Isopropyl-1-thio-␤-D-galactopyranoside was added to a final concentration of 2.5 mM to induce His 6 -tagged proOmpA2xH1 as well as to overexpress Syd, which causes a severe secretion defect (23). After 2-3 h, the cells were collected and subjected to N-lauroylsarcosine solubilization as described by Crooke et al. (26). The final pellet was suspended in 20 mM Tris-Cl (pH 8.0), 6 M urea, 0.5 M NaCl, and 5 mM imidazole with continuous vortexing, followed by centrifugation to remove cell debris. The soluble fraction was applied to a Ni 2ϩ -NTA-agarose column and washed with a washing buffer (20 mM Tris-Cl (pH 8.0), 6 M urea, 0.5 M NaCl and 60 mM imidazole). His 6tagged proOmpA2xH1 was then eluted with an elution buffer (20 mM Tris-Cl (pH 8.0), 6 M urea, 0.5 M NaCl, and 1 M imidazole). The protein eluted was precipitated with 7.5% trichloroacetic acid, washed with acetone, and then dissolved in 50 mM potassium phosphate (pH 7.5) and 6 M urea.
Membrane Insertion of 125 I-Labeled SecA-Purified SecA was iodinated according to the method of Economou and Wickner (8) with a minor modification, as described below. Na 125 I and SecA were incubated on ice for 20 min in a tube that had been coated with IODOGEN. The reaction mixture was transferred to another tube containing 2 mM dithiothreitol to stop iodination, and then applied to a Sephadex G-50 column to remove Na 125 I. To prepare 125 I-SecA-bound membrane vesicles, 4 nM 125 I-SecA and 200 g/ml urea-washed membrane vesicles were incubated for 15 min on ice in 100 l of a reaction buffer (50 mM Tris-HCl (pH 8.0), 50 mM KCl, 5 mM MgCl 2 , and 0.2 mg/ml bovine serum albumin). The membrane vesicles were isolated and then resuspended in the same buffer. 125 I-SecA-bound membrane vesicles, 50 g/ml SecB, 5 mM phosphocreatine, creatine kinase (10 g/ml), and 20 g/ml proOm-pAL43 or proOmpA2xH1 were preincubated in the reaction buffer for 2 min at 37°C (800 l each). One hundred l was removed as a control at time 0, and then insertion of SecA was initiated by the addition of 1 mM ATP, and 100-l aliquots were removed as indicated. The aliquots were digested with proteinase K (final, 0.1 mg/ml) for 20 min on ice. The proteins were precipitated with trichloroacetic acid, and then analyzed by SDS-polyacrylamide gel electrophoresis, followed by fluorography and quantification.
Inversion of the Membrane Topology of SecG-Membrane topology analysis of SecG was performed according to the method of Nishiyama et al. (11) with minor modifications. The translocation mixture (total, 120 l) comprised 0.2 mg/ml membrane vesicles, 60 g/ml SecA, 50 g/ml SecB, 25 g/ml proOmpA derivative, 1 mM ATP, 1 mM MgSO 4 , 10 mM creatine phosphate, and 100 g/ml creatine kinase in 50 mM potassium phosphate, pH 7.5. Translocation mixtures containing proOm-pAL43 or proOmpA2xH1 were incubated at 37°C for the indicated times, and then 10 mM AMP-PNP and 10 mM MgSO 4 were added. After 5 min of incubation at 37°C, the reaction mixtures were chilled on ice for 2 min, and then aliquots (20 l) were treated with 20 l of proteinase K at the indicated concentrations on ice for 30 min. The samples were precipitated with trichloroacetic acid, washed with acetone, and then analyzed by SDS-polyacrylamide gel electrophoresis, followed by immunoblotting with an anti-SecG antibody.
In Vitro Translocation of ProOmpA Derivatives-The translated 35 Slabeled proOmpA derivatives were subjected to in vitro translocation essentially according to the method of Uchida et al. (17). The translocation mixture (25 l) comprised a 35 S-labeled proOmpA derivative, membrane vesicles (5 g of protein), 2 mM ATP (or 10 M ATP) or 10 mM AMP-PNP, 5 mM MgSO 4 , 50 mM potassium phosphate (pH 7.5), 15 g/ml SecB, and 40 g/ml SecA. A cold proOmpA derivative was added to a final concentration of 5 g/ml, if necessary. The ATP concentration was maintained through regeneration with 10 mM creatine phosphate and 10 g/ml creatine kinase. After a 10-min incubation at 37°C, the mixture was treated with proteinase K (200 g/ml) for 20 min at 0°C, subjected to trichloroacetic acid precipitation, and then analyzed by SDS-polyacrylamide gel electrophoresis, followed by fluorography.
Conjugation of APDP-A proOmpA derivative, proOmpAL43C302Q, was prepared as described previously (16). ProOmpA2xH1C, in which Ser-233 in proOmpA2xH1 was replaced with a cysteine residue, was also constructed as described above. 35 S-Labeled proOmpA derivatives were conjugated with APDP (N-[4-(p-azidosalicylamido)butyl]-3Ј-(2Јpyridyldithio)propionamide) via Cys-260 or Cys-233 according to the method of Joly and Wickner (12). Briefly, APDP was dissolved in dimethyl sulfoxide (3 mg in 50 l), and then diluted 200-fold with 100 mM potassium phosphate (pH 7.5). To 70 l of the 35 S-labeled proOmpA dissolved in 6 M urea and 100 mM potassium phosphate (pH 7.5) was added 100 l of diluted APDP. After 1 h at room temperature, the proteins were precipitated with trichloroacetic acid and then washed with acetone twice, and the washed precipitate was dissolved in 6 M urea and 50 mM potassium phosphate (pH 7.5).
Immunoprecipitation of Cross-linked Products-After the in vitro translocation reaction and photolysis as described previously (16), samples were solubilized in a solution comprising 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, and 1% SDS, diluted 33-fold with dilution buffer consisting of 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 100 M 5,5Ј-dithiobis(2nitrobenzoic acid), and 0.1% Triton X-100, and then mixed with an appropriate antibody, anti-SecY or anti-SecA. After incubation at 4°C for 1.5 h, the antigen-antibody complexes were precipitated with protein A-Sepharose CL-4B. The resin was washed with dilution buffer, and then incubated in the sample buffer for electrophoresis at 37°C for 5 min to recover the bound proteins.
Determination of Translocation ATPase Activity-An enzyme-coupled spectrophotometric assay involving pyruvate kinase and lactate dehydrogenase was performed as described (29). The cuvette was maintained at 37°C, and contained 1 ml of 50 mM potassium phosphate (pH 7.5), 150 mM NaCl, 2 mM MgSO 4 , 3 mM phosphoenolpyruvate, 0.25 mM NADH, 1 mM ATP, 5 mM KCN, 10 units of pyruvate kinase and 15 units of lactate dehydrogenase, a specified amount of urea-treated membrane vesicles containing overproduced SecY/E/G, and 20 g/ml SecB. The assay was started by the addition of SecA (10 g). ProOmpA derivatives were subsequently added to a final concentration of 20 g/ml. Oxidation of NADH was continuously monitored at 340 nm with a Shimadzu UV-3000 spectrophotometer, and the amount of ATP hydrolyzed was calculated by using a value of 6.22 for the millimolar absorption coefficient of NADH. The values were corrected for background ATP hydrolysis.
Alkali Fractionation-After the formation of in vitro translocation intermediates as described above, an aliquot of the reaction mixture was removed, without the proteinase K treatment, and centrifuged to isolate membranes. The membranes were suspended in the original volume of ice-cold 0.2 M Na 2 CO 3 (pH 11), vortexed, and then incubated for 30 min on ice. The solution was centrifuged to separate the soluble and membrane-associated fractions. Each fraction was precipitated with trichloroacetic acid, and then analyzed by SDS-polyacrylamide gel electrophoresis, followed by fluorography.

Cellular Localization of ProOmpA2xH1
-We previously demonstrated that short hydrophobic segments in the mature domain of proOmpA are transiently arrested in the Sec machinery during the polypeptide movement across the cytoplasmic membrane of E. coli in vitro (16). To confirm that the hydrophobicity of the segments is important for this transient arrest, we duplicated a hydrophobic segment referred to as H1 (residues 233-237), creating proOmpA2xH1. The resulting segment (2xH1; residues 229 -237) comprising 8 hydrophobic amino acids stopped polypeptide translocation and acted as a stable membrane anchor (16). The extent of translocation arrest was similar (92%) when the in vitro translocation reaction was performed in the presence of a proton motive force (data not shown).
To determine whether or not proOmpA2xH1 is located in the cytoplasmic membrane not only in vitro but also in vivo, we prepared inner membrane vesicles from spheroplasts of RK4788 cells (⌬ompA) harboring a low copy plasmid encoding proOmpA2xH1 or wild-type proOmpA without the induction of the proteins. As shown in Fig. 1, most proOmpA2xH1 was partitioned in the inner membrane fraction (84%) (lane 6), while little, if any, wild-type proOmpA was associated with the inner membrane (lane 4). These results suggest that 2xH1 acts as a membrane anchor in vivo as well as in vitro.
ProOmpA2xH1 Does Not Interfere with the Translocation of Wild-type ProOmpA-As a first step to understand the process of membrane anchoring of proOmpA2xH1, we examined whether or not the translocation channel is occupied by an anchored 2xH1 segment. For this purpose, wild-type 35 S-proOmpA was subjected to in vitro translocation in the presence of excess cold proOmpA2xH1 or wild-type proOmpA. As shown in Fig. 2, excess wild-type proOmpA completely blocked the translocation of wild-type 35 S-proOmpA, whereas excess proOmpA2xH1 only delayed the translocation. These results suggest that the translocation-interrupted proOmpA2xH1 had left the translocation apparatus, and that the mechanism of membrane anchoring of proOmpA2xH1 is somehow different from that of the translocation of wild-type proOmpA.
Stop-transfer Process Induces the Deinsertion of SecA-To investigate the process of membrane anchoring in more detail, we examined how the presence of a stop-transfer sequence affects the insertion-deinsertion cycle of SecA. SecA changes its conformation in the presence of ATP and a preprotein (7). This conformational change causes SecA to become inserted deep into the membrane, and the inserted portion of a 30-kDa fragment is resistant to protease treatment. Upon ATP hydrolysis, SecA changes its conformation again and thereby becomes deinserted from the membrane (8). As shown in Fig. 3, in the presence of looped proOmpAL43, whose translocation is interrupted by the intramolecular disulfide-bridged loop comprising 43 amino acid residues (17), the 30-kDa fragment was formed in an ATP-dependent manner, and its amount increased with time, reaching steady state after 15 min. This time course is similar to that of wild-type proOmpA (10). When the membrane insertion was conducted in the presence of proOmpA2xH1, accumulation of the membrane-inserted form of SecA was completed within 5-6 min, and subsequently deinsertion was observed. SecA-proOmpA2xH1 interaction is maintained through 20 min, as revealed by the sensitivity of SecA to protease V8 (7), indicating that the result is not for the aggregation of proOmpA2xH1 (data not shown).
ATP Hydrolysis by SecA during Translocation-The ATPase activity of SecA is remarkably enhanced by the membrane translocase, acidic lipids (30), and preproteins (31). This en-hanced ATPase activity is referred to as translocation ATPase activity. We next examined the effect of proOmpA2xH1 on the ATPase activity of SecA. ProOmpAL43 enhanced the ATPase activity of SecA to a similar extent to that observed for wildtype proOmpA (Fig. 4). This may imply that enhanced ATP hydrolysis continues even when polypeptide transition is blocked by an intramolecular disulfide-bridged loop. Similar results were observed when translocation was blocked by a derivative of proOmpA carrying a side chain of 20 amino acid residues via a cysteine residue (32). Although proOmpA2xH1 also enhanced the ATPase activity of SecA, the enhancement lasted for only approximately 100 s. The lack of enhancement after 100 s is not due to the inactivation of SecA because the addition of wild-type proOmpA caused reactivation of ATP hydrolysis.
Topology Inversion of SecG-We next examined how the presence of a stop-transfer sequence influences the topology inversion of SecG (11). SecG possesses two transmembrane FIG. 1. Localization of proOmpA2xH1. Everted membrane vesicles prepared from spheroplasts were layered on a continuous gradient formed from 30 -44% (w/w) sucrose in 50 mM potassium phosphate (pH 7.5), and then centrifuged at 60,000 ϫ g for 15 h at 4°C. Isolated inner membranes were diluted with 50 mM potassium phosphate (pH 7.5), and then centrifuged at 150,000 ϫ g for 2 h. The pellet containing inner membranes were resuspended in the original volumes, and the same volumes of whole cell samples and inner membrane fractions were subjected to SDS-polyacrylamide gel electrophoresis, and immunoblotted with an anti-OmpA antibody. Whole cell samples (lanes 1, 3, and  5) and inner membrane fractions (lanes 2, 4, and 6) of RK4788 (⌬ompA) harboring a vector plasmid (Control) or a plasmid encoding wild-type proOmpA (proOmpA) or proOmpA2xH1 (2xH1). The bands indicated by an arrowhead correspond to the mature form of the proOmpA derivatives.
FIG. 2. ProOmpA2xH1 does not interfere with in vitro translocation of wild-type proOmpA. In vitro translocation of 35 S-proOmpA was performed in the presence of 5 g/ml unlabeled proOmpA or proOmpA2xH1. Samples were then treated with proteinase K, recovered by trichloroacetic acid precipitation, and then analyzed by SDSpolyacrylamide gel electrophoresis, followed by fluorography and quantification.

FIG. 3. A stop-transfer sequence induces the deinsertion of
SecA. Membrane insertion of 125 I-SecA with wild-type proOmpAL43 (Ⅺ) or proOmpA2xH1 (f) was performed as described under "Experimental Procedures." Aliquots of the samples were digested with proteinase K, and then proteins were precipitated with trichloroacetic acid and analyzed by SDS-polyacrylamide gel electrophoresis, followed by fluorography and quantification of 30-kDa fragments of SecA. regions, and its amino-and carboxyl-terminal regions are located in the periplasm. In everted membrane vesicles, proteinase K cleaves sites close to the NH 2 terminus of 11.4-kDa SecG, producing a 9-kDa fragment. This fragment is immunodetected by anti-SecG, which recognizes the carboxyl-terminal 16 amino acid residues of SecG, because the COOH-terminal region is located inside the vesicles. In the presence of ATP and proOmpA, on the other hand, no fragment is detected by the antibody because the COOH-terminal region is cleavable by proteinase K as a consequence of the inversion of the membrane topology of SecG (11). Consistent with the previous results (11), the 9-kDa protease-resistant fragment was formed when SecA was incubated with everted membrane vesicles in the absence of a preprotein (Fig. 5A). When translocation was conducted in the presence proOmpAL43 and blocked by the addition of AMP-PNP at 3 min, the 9-kDa fragment was not formed (Fig. 5B). Although the previous study showed that SecG is almost completely cleaved by proteinase K under the translocation conditions (11), we found that only a small amount of SecG was cleaved in the presence of proOmpAL43. A similar result was obtained when wild-type proOmpA was used (data not shown). Perhaps the COOH-terminal region of SecG is present on the outside of membrane vesicles, but is not completely exposed and therefore is inaccessible to proteinase K. However, it is obvious that SecG changes its conformation in both preprotein-and ATP-dependent manners, as revealed by the changes in the proteolytic patterns. This conformational change was maintained for at least 15 min. When translocation was conducted in the presence of proOmpA2xH1 and terminated at 3 min, the proteolytic pattern was essentially the same as that observed in the presence of proOmpAL43 (Fig.  5C). In contrast, when translocation was stopped at 15 min, the proteolytic pattern was remarkably different from that observed in the presence of proOmpAL43, and rather similar to that observed in the absence of the preprotein. These results, taken together, suggest that proOmpA2xH1 containing a stoptransfer sequence is translocated through the normal translocation pathway at an early stage, but leaves the pathway at a later stage of translocation.
Nearest Neighbor of the Anchored 2xH1 Segment-To directly demonstrate that proOmpA2xH1 does not interact with the translocation machinery at a later stage of translocation, we performed a cross-linking experiment using a photoactivable and reducible cross-linker, APDP. We previously used this reagent and demonstrated that a translocation intermediate of a looped proOmpA interacts with SecY and partly with SecA (16). We replaced a serine residue (Ser-233) in the middle of the 2xH1 segment with a cysteine residue (proOmpA2xH1C) (Fig. 6A), and then APDP was attached to this cysteine residue. After photolysis, SecA and SecY were immunoprecipitated with anti-SecA and anti-SecY, respectively, and the precipitates were analyzed by SDS-polyacrylamide gel electrophoresis in the presence of dithiothreitol to cleave the cross-linked proteins. As shown in Fig. 6B, a significant amount of cross-linked 35 S-proOmpA2xH1 was not immunoprecipitated by either anti-SecA (lane 4) or anti-SecY (lane 5). This is not due to the inability to photo-cross-link the derivative of proOmpA2xH1 with a Cys residue in the 2xH1 region with APDP to the Sec machinery. When the deinsertion of SecA was blocked by the addition of AMP-PNP, as described in Fig. 7B, prior to photolysis, a significant amount of cross-linked 35 S-proOmpA2xH1 was immunoprecipitated by either anti-SecA (lane 2) or anti-SecY (lane 3). These results suggest that the 2xH1 segment transiently interacts with the Sec machinery. Once anchored, the 2xH1 region could not interact with the Sec machinery.
Alkali Extraction of the Membrane-anchored ProOmpA2x-H1-To assess the state of proOmpA2xH1 anchored in the membrane, membrane proteins were extracted with 0.2 M Na 2 CO 3 (pH 11) after completion of the translocation reaction. This alkali treatment extracts all peripheral membrane proteins and uninserted polytopic membrane proteins in mammalian cells (33,34) and in bacterial cells (22,35). As shown in Fig. 7A, 60% of the anchored proOmpA2xH1 was partitioned into the alkali soluble fraction, whereas 17% of the translocation intermediate of the looped proOmpA was observed in the same fraction (Fig. 7A). When AMP-PNP was added to inhibit the deinsertion of SecA from the membrane prior to alkali FIG. 4. Effect of proOmpA2xH1 on the ATP activity of SecA. The hydrolysis of ATP was started by the addition of proOmpA derivatives. At the indicated times (arrows), 40 g/ml wild-type proOmpA was added to the assay mixture. The amount of ATP hydrolyzed in the absence of proOmpA derivatives was subtracted.

FIG. 5. Conformational change of SecG coupled with translocation.
Reaction mixtures (120 l) without (A) or with proOmpAL43 (B) or proOmpA2xH1 (C), were incubated at 37°C for the indicated times, and after proteinase K treatment, SDS-polyacrylamide gel electrophoresis and immunoblotting with the anti-SecG antibody were performed as described under "Experimental Procedures." Membrane vesicles that had not been treated with proteinase K were also examined (Proteinase K, Ϫ). extraction, both proOmpAL43 and proOmpA2xH1 were almost exclusively recovered in the membrane fraction (Fig. 7B). When NaN 3 , an inhibitor of SecA (36), was included in the translocation assay mixture, the association of proOmpA derivatives with the membrane was not observed (data not shown), suggesting that the membrane association was mediated by SecA. These results suggest that proOmpA2xH1 is released from the membrane-embedded Sec machinery as the deinsertion of SecA occurs. DISCUSSION In this study we examined the stop-transfer process of a proOmpA derivative containing a stop-transfer sequence (proOmpA2xH1). The use of an in vitro translocation system enabled us to analyze the translocation process at defined stages of polypeptide transfer. The major conclusion of this study is that when the transfer of a polypeptide is halted by a stop-transfer sequence, the preprotein leaves the functional translocation site, which allows the passage of the following preprotein molecule. Cross-linking analysis confirmed that the membrane-anchored hydrophobic segment is located outside the translocation channel. Our data also showed that the cessation of translocation causes deinsertion of SecA and the return the inverted topology of SecG to the original one. Alkali extraction experiments revealed that the state of proOmpA2xH1 anchored in the membrane is different from that of the translocation intermediate formed as a consequence of the presence of an intramolecular loop. When deinsertion of SecA was interrupted by AMP-PNP, proOmpA2xH1, as well as the looped proOmpA, was not extracted on alkali treatment.
It has been proposed that secretory proteins are translocated across the cytoplasmic membrane through a proteinaceous channel (12). When a segment with moderate hydrophobicity enters the translocation channel, the passage of the polypeptide is interrupted and the preprotein is excluded from the the channel. There are two possible explanations for this mechanism. First, translocation arrest may be triggered by an interaction between the hydrophobic stop-transfer segment and the hydrocarbon of the phospholipid, and then the hydrophobic segment spontaneously leaves the translocase. Second, the translocation apparatus itself may recognize the hydrophobic segment, and decide to continue its transfer or to integrate it into the membrane. In the former case, the stop-transfer process is dependent on the affinity of the hydrophobic segment with the hydrophobic core of the lipid bilayer. If this is the case, it is reasonable to assume that the anchored segment would be spontaneously integrated into the lipid bilayer regardless of whether SecA is deinserted or not. However, alkali extraction experiments in the presence of AMP-PNP revealed that the anchored segment remains in the translocase unless SecA is deinserted, suggesting that the polypeptide is not spontaneously integrated into the membrane. Therefore, the second of the above two alternatives seems to be more likely.
We recently demonstrated that the short hydrophobic segments within the mature region of a preprotein promote translocation stalling. This event seems to be at least partly due to an interaction between the hydrophobic stretch and the translocase (16). Here, we propose that the hydrophobic segment itself functions as a kind of "integration signal" for Sec-dependent insertion of membrane proteins, such as type I membrane protein (having only a single transmembrane segment and its carboxyl terminus located on the cytoplasmic side of the membrane). The entrance of a stop-transfer sequence of a preprotein into the translocation channel may trigger the release of SecA from the translocase and thereby allows the preprotein to leave the active translocation site. This indicates that the discrimination between secretory proteins and membrane proteins is conducted by the translocase that can estimate the hydrophobicity of the polypeptide chain of preproteins. The results described in this study are consistent with this proposal. A, in vitro translocation of proOmpA derivatives was performed in the presence of everted membrane vesicles, and after the membranes had been isolated, the samples were treated with 0.2 M Na 2 CO 3 . Supernatant (S) and membrane (P) fractions were obtained by centrifugation, and then analyzed by SDS-polyacrylamide gel electrophoresis and fluorography. B, after the formation of in vitro translocation intermediates as described above, 10 mM MgSO 4 and 10 mM AMP-PNP were added to block the SecA movement prior to alkali extraction. The bands indicated by an arrowhead correspond to the mature form of the proOmpA derivatives.