Portal de Periódicos da CAPES

Refolded Outer Membrane Protein A of Escherichia coliForms Ion Channels with Two Conductance States in Planar Lipid Bilayers

2000; Elsevier BV; Volume: 275; Issue: 3 Linguagem: Inglês

10.1074/jbc.275.3.1594

1083-351X

Ashish Arora, Dennis Rinehart, Gábor Szabó, Lukas K. Tamm,

Bacterial Genetics and Biotechnology

Outer membrane protein A (OmpA), a major structural protein of the outer membrane of Escherichia coli, consists of an N-terminal 8-stranded β-barrel transmembrane domain and a C-terminal periplasmic domain. OmpA has served as an excellent model for studying the mechanism of insertion, folding, and assembly of constitutive integral membrane proteinsin vivo and in vitro. The function of OmpA is currently not well understood. Particularly, the question whether or not OmpA forms an ion channel and/or nonspecific pore for uncharged larger solutes, as some other porins do, has been controversial. We have incorporated detergent-purified OmpA into planar lipid bilayers and studied its permeability to ions by single channel conductance measurements. In 1 m KCl, OmpA formed small (50–80 pS) and large (260–320 pS) channels. These two conductance states were interconvertible, presumably corresponding to two different conformations of OmpA in the membrane. The smaller channels are associated with the N-terminal transmembrane domain, whereas both domains are required to form the larger channels. The two channel activities provide a new functional assay for the refolding in vitro of the two respective domains of OmpA. Wild-type and five single tryptophan mutants of urea-denatured OmpA are shown to refold into functional channels in lipid bilayers. Outer membrane protein A (OmpA), a major structural protein of the outer membrane of Escherichia coli, consists of an N-terminal 8-stranded β-barrel transmembrane domain and a C-terminal periplasmic domain. OmpA has served as an excellent model for studying the mechanism of insertion, folding, and assembly of constitutive integral membrane proteinsin vivo and in vitro. The function of OmpA is currently not well understood. Particularly, the question whether or not OmpA forms an ion channel and/or nonspecific pore for uncharged larger solutes, as some other porins do, has been controversial. We have incorporated detergent-purified OmpA into planar lipid bilayers and studied its permeability to ions by single channel conductance measurements. In 1 m KCl, OmpA formed small (50–80 pS) and large (260–320 pS) channels. These two conductance states were interconvertible, presumably corresponding to two different conformations of OmpA in the membrane. The smaller channels are associated with the N-terminal transmembrane domain, whereas both domains are required to form the larger channels. The two channel activities provide a new functional assay for the refolding in vitro of the two respective domains of OmpA. Wild-type and five single tryptophan mutants of urea-denatured OmpA are shown to refold into functional channels in lipid bilayers. outer membrane protein diphytanoylphosphatidylcholine picosiemens maltophorin ferric hydroxamate uptake receptor ferric enterobactin receptor The outer membrane of Gram-negative bacteria serves as a molecular sieve, resisting the entry of noxious compounds, while at the same time allowing the uptake of essential nutrients. Molecules up to about 600 Da in size are taken up by a set of more or less substrate-specific porins and transporters. Outer membrane proteins were among the first integral membrane proteins for which crystal structures have been solved by x-ray diffraction. In contrast to most proteins of the inner membrane, which are of the helical bundle type, the common structural motif of the outer membrane proteins is the β-barrel, i.e.an antiparallel β-sheet that closes on itself. Porins with 16 (OmpF, OmpC, PhoE (phosphophorin)),118 (LamB), or 22 (FhuA, FepA) antiparallel β-strands have been described (1.Cowan S.W. Schirmer T. Rummel G. Steiert M. Gosh R. Pauptit R.A. Jansonius J.N. Rosenbusch J.P. Nature. 1992; 358: 727-733Crossref PubMed Scopus (1334) Google Scholar, 2.Schirmer T. Keller T.A. Wang Y.-F. Rosenbusch J.P. Science. 1995; 267: 512-514Crossref PubMed Scopus (534) Google Scholar, 3.Ferguson A.D. Hofmann E. Coulton J.W. Diederichs K. Welte W. Science. 1998; 282: 2215-2220Crossref PubMed Scopus (668) Google Scholar, 4.Buchanan S.K. Smith B.S. Venkatramani L. Xia D. Esser L. Palnitkar M. Chakraborty R. van der Helm D. Deisenhofer J. Nat. Struct. Biol. 1999; 6: 56-63Crossref PubMed Scopus (492) Google Scholar). The outer walls of their structures are generally hydrophobic (to match the lipid bilayer), and they have water-filled central pores of variable sizes through which substrates are taken up. Whereas the porins are usually trimeric, the iron-siderophore transporters, FhuA and FepA, are monomeric. The substrate specificity of these proteins arises from specific residues in the pore, special peptide segments (aromatic "greasy slide" in LamB), or entire protein domains (periplasmic "cork" in FhuA). When incorporated into black lipid membranes, porins exhibit single channel activities that strongly depend on the particular porin, the applied transmembrane potential, and the type and concentration of the electrolyte in the environment. For example, OmpF of Escherichia coli has a conductance of 0.8 nanosiemens in 1 m NaCl, whereas the major porin from Rhodopseudomonas blastica, which is of similar size, exhibits a conductance of 3.9 nanosiemens in 1m KCl (5.Phale P.S. Schirmer T. Prilipov A. Lou K-L. Hardmeyer A. Rosenbusch J.P. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 6741-6745Crossref PubMed Scopus (87) Google Scholar, 6.Schmid B. Maveyraud L. Krömer M. Schulz G.E. Protein Sci. 1998; 7: 1603-1611Crossref PubMed Scopus (43) Google Scholar) OmpA is another major outer membrane protein of E. coli. Although the mass of OmpA is similar to that of many other porins, OmpA consists of two separate domains. The N-terminal domain is integrated into the membrane in the form of a small β-barrel of only eight antiparallel β-strands. In addition, a C-terminal globular domain of approximately 150 residues extends into the periplasm. Unlike the porins, OmpA most likely exists as a monomer in the outer membrane. The crystal structure of the N-terminal domain of OmpA has recently been solved by x-ray diffraction (7.Pautsch A. Schulz G.E. Nat. Struct. Biol. 1998; 5: 1013-1017Crossref PubMed Scopus (419) Google Scholar). Compared with the larger porins, the barrel of this "miniporin" is very tight, with mostly hydrophilic residues and a few aqueous cavities closely packed in the lumen of the barrel. Providing structural stability to the cell appears to be one of the main functions of OmpA. This is probably accomplished by linking through the C-terminal domain the outer membrane to the periplasmic peptidoglycan. Additionally, OmpA mediates bacterial conjugation and functions as a receptor for various bacteriophages. OmpA has also been reported to form channels or pores in lipid bilayers, although this aspect of the protein is somewhat controversial. Saint et al. (8.Saint N. De E. Julien S. Orange N. Molle G. Biochim. Biophys. Acta. 1993; 1145: 119-123Crossref PubMed Scopus (60) Google Scholar) measured single channel conductance events in "solvent-free" planar bilayers on the order of 180 pS at 100 mV and in 0.25 m KCl. Based on osmotic swelling experiments with reconstituted OmpA proteoliposomes, Sugawara and Nikaido (9.Sugawara E. Nikaido H. J. Biol. Chem. 1992; 267: 2507-2511Abstract Full Text PDF PubMed Google Scholar, 10.Sugawara E. Nikaido H. J. Biol. Chem. 1994; 269: 17981-17987Abstract Full Text PDF PubMed Google Scholar) concluded that OmpA forms a diffusion channel of about 10 Å in diameter. However, only 2–3% of all OmpA molecules were in the open conformation in their preparation, and vesicles containing open channels could be separated from closed channel vesicles by density gradient centrifugation. These experiments indicated that the open and closed forms represent two different, relatively stable conformations of OmpA. The crystal structure of the N-terminal domain of OmpA, which was obtained from crystals formed from an OmpA fragment that was solubilized in the detergent C8E4, indicated no obvious aqueous pore, as the water-filled cavities were not connected in that structure (7.Pautsch A. Schulz G.E. Nat. Struct. Biol. 1998; 5: 1013-1017Crossref PubMed Scopus (419) Google Scholar). Based on their structure, these authors therefore questioned the ability of OmpA to form ion- or solute-conducting pores. OmpA has also served as an excellent model to study the folding and membrane insertion of a constitutive integral membrane protein (11.Kleinschmidt J.H. Tamm L.K. Biochemistry. 1996; 35: 12993-13000Crossref PubMed Scopus (150) Google Scholar, 12.Kleinschmidt J.H. Tamm L.K. Biochemistry. 1999; 38: 4996-5005Crossref PubMed Scopus (95) Google Scholar, 13.Kleinschmidt J.H. den Blaauwen T. Driessen A.J.M. Tamm L.K. Biochemistry. 1999; 38: 5006-5016Crossref PubMed Scopus (124) Google Scholar). In these studies, OmpA was refolded from 8 m urea by rapid dilution into a solution containing preformed lipid vesicles (or detergent micelles). OmpA seems to be ideally suited for these studies partly because it is a relatively small, yet polytopic integral membrane protein and also because as a β-barrel membrane protein it has a sequence of alternating hydrophobic and hydrophilic residues. This property makes the membrane-spanning sequences on average less hydrophobic than those of helical bundle membrane proteins, which therefore cannot be completely solubilized and unfolded in urea. However, a good functional assay to monitor the refolding of OmpA has been lacking up until now. Because the formation of ion channels by OmpA has not been universally accepted and to develop a practical assay for the refolding of OmpA, we re-investigated the channel activity of OmpA in a well defined reconstituted lipid bilayer system. We found that both native OmpA and OmpA that was refolded in the detergent C8E4forms two types of ion-conducting channels in planar bilayer membranes. The more frequent smaller channels exhibited a conductance of 50–80 pS and the less frequent larger channels a conductance of 260–320 pS in 1m KCl at a 100 mV membrane potential. The smaller but not the larger conductance state was also observed when the N-terminal transmembrane domain, i.e. a fragment comprising residues 1–176, was refolded and incorporated into planar lipid bilayers. Five single tryptophan mutants of OmpA exhibited small and large channel conductances similar to those of wild-type OmpA. OmpA and single tryptophan mutants of OmpA were expressed in the OmpA-deficient E. coli strain MC4100rh− as described previously (13.Kleinschmidt J.H. den Blaauwen T. Driessen A.J.M. Tamm L.K. Biochemistry. 1999; 38: 5006-5016Crossref PubMed Scopus (124) Google Scholar). These proteins were purified from the outer membranes by urea extraction and ion-exchange chromatography of the unfolded proteins in 8 m urea using a Q-Sepharose Fast Flow column (11.Kleinschmidt J.H. Tamm L.K. Biochemistry. 1996; 35: 12993-13000Crossref PubMed Scopus (150) Google Scholar). The transmembrane domain of OmpA Trp-7, Trp-7-(1–176), was generated from the single tryptophan mutant Trp-7 by utilizing site-directed mutagenesis to convert the codon for proline 177 to a stop codon. Briefly, the OmpA gene was transferred from pET1102 (13.Kleinschmidt J.H. den Blaauwen T. Driessen A.J.M. Tamm L.K. Biochemistry. 1999; 38: 5006-5016Crossref PubMed Scopus (124) Google Scholar) into pAlter-1 (Promega, Madison, WI). The mutagenic primer 5′-CAGGGCGAAGCAGCTTAAGTAGTTGCTCCGGC-3′, which also contained a unique site for AflII and the commercial Amp selection primer (Promega), were used to introduce a stop codon at position 177. The mutagenized gene was reintroduced into pET1102 for expression in MC4100rh−. The correct sequence was verified by sequencing. Trp-7-(1–176) was purified using essentially the same protocol as for OmpA, except that a final gel filtration step on a Superdex-75 HR column (Amersham Pharmacia Biotech) in 20 mmpotassium phosphate, pH 7.3, 50 mm NaCl, containing 8m urea, was added to separate the transmembrane domain from residual impurities. A sample of wild-type OmpA that was purified in its native form was a kind gift of Dr. Hiroshi Nikaido (University of California, Berkeley). This protein was derived from an OmpF- and OmpC-deficient K12-derivative E. coli strain (HN705) and was purified from the outer membrane by detergent extraction and repeated gel filtration over a Sephacryl S-300 column in dodecylmaltoside (14.Sugawara E. Steiert M. Rouhani S. Nikaido H. J. Bacteriol. 1996; 178: 6067-6069Crossref PubMed Google Scholar,15.Sugawara E. Nikaido H. Biophys. J. 1997; 72: A138Google Scholar). Native OmpA was suspended in 10 mm Tris-Cl buffer, pH 7.5, containing 0.4 m NaCl, 1 mmdithiothreitol, 1 mm EDTA, and 0.1% dodecylmaltoside. For single channel measurements, these samples were diluted into a 20 mm micellar solution of tetraoxyethylene mono-n-octylether (C8E4, Bachem, Philadelphia, PA) to a concentration of 80 μg/ml, i.e. the same as that used for the refolded proteins that were obtained by the urea purification method. Refolding of OmpA and its derivatives was carried out as described in more detail by Kleinschmidt et al. (16.Kleinschmidt J.H. Wiener M.C. Tamm L.K. Protein Sci. 1999; 8: 2065-2071Crossref PubMed Scopus (120) Google Scholar). Briefly, 5 μl of a 4 mg/ml solution of unfolded OmpA in 15 mm Tris-Cl, pH 8.5, containing 8 m urea was diluted 50-fold into a 20 mm solution of C8E4 in 2 mm sodium borate, pH 10.0, containing 0.4 mmEDTA. The mixture was incubated overnight at 40 °C to ensure complete refolding of the protein. To remove misfolded protein aggregates the samples were centrifuged at 14,000 rpm in a table-top centrifuge (Eppendorff, Rexdale, Ontario) for 15 min prior to the addition to the planar membranes. SDS-polyacrylamide gel electrophoresis was performed according to the method of Laemmli (17.Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (207516) Google Scholar). For checking the purity of the proteins, samples were diluted (1:1, v/v) with treatment buffer (0.125m Tris-Cl, 4% SDS, 20% glycerol, 10% β-mercaptoethanol, 0.2% bromphenol blue), boiled at 100 °C for 5 min, and run on 12% polyacrylamide gels for the full-length OmpA proteins or 14% polyacrylamide gels for the transmembrane fragment. For measuring protein refolding by the gel-shift assay, the refolded samples were either boiled or incubated at 40 °C for 5 min and run on a 10–15% gradient polyacrylamide gel. Protein bands were visualized by staining with Coomassie Brilliant Blue R-250. Planar lipid bilayers were formed from a solution of 17 mg/ml diphytanoylphosphatidylcholine (DPhPC, Avanti Polar Lipids, Birmingham, AL) in n-decane (Aldrich) using the method of Mueller et al. (18.Mueller P. Rudin D.O. Tien H.T. Wescott W.C. Circulation. 1962; 26: 1167-1171Crossref Google Scholar) with some modifications (19.Busath D. Szabo G. Nature. 1981; 294: 371-373Crossref PubMed Scopus (76) Google Scholar). The lipid solution was painted on a 500-μm hole in a Teflon partition separating two 1.5-ml compartments, which were filled with KCl buffer (1 m KCl, 10 mm Tris, pH 7.1). The compartments were connected to the recording system through two chlorided silver electrodes, one of which (the front,cis side) was grounded, whereas the other (the rear,trans side) was connected to a custom designed trans-impedance amplifier. The painted DPhPC/n-decane bilayer membranes were tested for integrity by checking the reflectance optically and also by their resistance and capacitance. After the bilayers were formed, 5 μl of 80 μg/ml of OmpA in C8E4 micelles were added to the ciscompartment and stirred. Conductance measurements were made after about 10 min of equilibration with a 100-mV potential applied to thetrans compartment. The current across the bilayer was recorded on magnetic medium. For analysis, the signal was filtered at 100 Hz with an 8 pole low-pass filter and digitized at 1 KHz (LABMAN, G. Szabo and C. Q. Ye, University of Virginia). Single channel conductance events were analyzed using TRANSIT (Baylor School of Medicine, Houston, TX), and IGOR (Wavemetrics, Portland, OR) software packages. Single channel conductance events were identified automatically in most traces. In a few traces where baseline noise was unusually large, it was difficult to identify transitions reliably by automated analysis. In such cases conductance steps were analyzed interactively using IGOR. The interactive and automated procedures yielded the same results, in terms of conductance levels and single channel histograms in all traces that were analyzed by both methods. The data were averaged from three to seven independent recordings, which all lasted for several minutes for each protein and condition. A typical trace of a single channel recording of OmpA that was purified in its native form from the outer membranes of anE. coli K12-derivative strain by detergent extraction is shown in Fig. 1. In this (and all subsequent) experiment(s) OmpA in C8E4 micelles was added to the cis compartment next to DPhPC/n-decane bilayers and equilibrated, and a 100-mV potential was applied to the trans compartment. Two types of unitary conductance increases, indicated by downward deflections, were immediately induced. When in control experiments the detergent C8E4 alone was added in an amount equivalent to four times of that used in the mixed detergent/OmpA micelles, no channel activity was observed for several minutes (trace labeled C8E4 in Fig. 1). Three different conductance states are evident in the recording of native OmpA. In state I, all channels are closed, exhibiting a baseline conductance. In state II, a single small channel is open, exhibiting a conductance of the order of 50 pS, and in state III, a large channel with a conductance of the order of 320 pS is open. The trace begins at time A in state II with a single small channel open. At B, a large channel opens (state III) and closes again at C (state II). At D, the small channel closes to the baseline conductance state I. Between D and E, a small channel opens and closes. At E, a large channel opens directly from the baseline conductance state (state I → state III) but closes only to the open state of small channel at F (state III → state II), which closes at G to the baseline (state II → state I). A small channel opens again at H (state I → state II) and converts into a large channel at I (state II → state III), which closes directly to the baseline at J (state III → state I). A histogram of the single channel conductance levels of native OmpA is shown in the lower part of Fig. 1. The distribution of 413 total observed events shows three well separated peaks. Small channels of 40–60 pS conductance (mean 52 pS) accounted for 55% of all events. Two additional peaks corresponding to the state II ↔ state III (mean 268 pS) and state I ↔ state III (mean 317 pS) transitions are also evident. One might ask whether the coexistence of small and large channels of OmpA reflects two separate populations of molecules that co-exist in planar bilayers or whether the observed channels originate from a single population with interchanging conformations. The following observations support the latter possibility: first, we never observed the simultaneous opening of two large channels or two small channels; and second, the large channels open and close either from/to the baseline or from/to the open state of a small channel, but small channels only open and close from/to the baseline and never from the open state of a large channel. Therefore, it appears that the small channels are kinetic precursors of the larger channels. OmpA is one of only a few proteins that can be refolded into preformed lipid bilayers. One goal of this study was to determine whether refolded OmpA exhibits the same single channel activity as native OmpA. A convenient assay to follow the refolding of OmpA is to monitor an increase in the mobility on polyacrylamide gels of the refolded relative to the unfolded proteins (20.Schweizer M. Hindennach I. Garten W. Henning U. Eur. J. Biochem. 1978; 82: 211-217Crossref PubMed Scopus (156) Google Scholar). This shift in mobility is only expressed in samples that are not boiled prior to electrophoresis. Upon denaturation by heat (or urea), OmpA runs on SDS gels as a single band that corresponds to an apparent molecular weight of 35 kDa, whereas native or refolded OmpA runs as a single band with an apparent molecular weight of 30 kDa (Fig.2). Refolding of OmpA into lipid bilayers or detergent micelles can also be demonstrated by monitoring a change of the intrinsic Trp fluorescence (11.Kleinschmidt J.H. Tamm L.K. Biochemistry. 1996; 35: 12993-13000Crossref PubMed Scopus (150) Google Scholar, 16.Kleinschmidt J.H. Wiener M.C. Tamm L.K. Protein Sci. 1999; 8: 2065-2071Crossref PubMed Scopus (120) Google Scholar). When we incorporated refolded OmpA in C8E4 into planar lipid bilayers at the same concentrations as native OmpA, we again found unitary conductance values of similar magnitude as for the small and large channels of native OmpA. However, in contrast to the native protein, interconversions between the two conductance states were much rarer. In most cases, a single bilayer recording contained either small or large channels. About 80–85% of all recordings displayed predominantly small channels and about 15–20% predominantly large channels. Small and large channel recordings are displayed in Figs.3 and 5, respectively. Because the data of the two types of channels are from different recordings (but still from the same refolded OmpA stocks), separate histograms are shown for the small and large channels in Figs. 4and 5, respectively. There are precedents for the observation that single channels occur in two different but well defined conformations with relatively rare interconversions. For example, gramicidin A is known to form "mini-channels" and the interconversion frequency between the main and mini-channels is rather low (19.Busath D. Szabo G. Nature. 1981; 294: 371-373Crossref PubMed Scopus (76) Google Scholar). Analyzing 1173 small OmpA channel events from seven independent experiments, we found that their single channel conductance values ranged from 40 to 95 pS and followed an approximately Gaussian distribution with a mean value of 66 pS and a standard error of 15 pS. In all experiments, the membranes were stable, and the channel activity lasted for the duration of 15–30-min experiments. The large channels of refolded OmpA were relatively noisy in the open state (Fig. 5,top trace). Despite this noise, we were able to measure the distribution of the large channel events of the refolded OmpA as shown in the lower left panel of Fig. 5. The mean large channel conductance of refolded OmpA was 261 pS (212 events), i.e.similar to the conductance change associated with the small-to-large channel conversion of native OmpA (see Fig. 1).Figure 3Single channel recordings of small channels of refolded OmpA, its single Trp mutants, and the N-terminal transmembrane domain (residues 1–176) of the Trp-7 mutant in planar bilayers. Experimental conditions are the same as those described in the legend to Fig. 1. The trace of a control experiment with a misfolded, water-collapsed OmpA (Trp-57 mutant) is shown at the bottom of the figure.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 5Single channel recordings and histograms of large channels of refolded OmpA and some of its single Trp mutants in planar lipid bilayers. Top, four representative recordings showing large channels formed after incorporation of refolded OmpA and single Trp mutants. Experimental conditions are the same as described in the legend to Fig. 1. Bottom,histograms showing the distribution of the large channel openings and closings of refolded OmpA and the single Trp mutants Trp-7, Trp-15, and Trp-143. The number of evaluated events and mean channel conductances for each protein are described under "Results."View Large Image Figure ViewerDownload Hi-res image Download (PPT)Figure 4Histograms showing the distribution of small channels of refolded OmpA, its single Trp mutants, and the N-terminal transmembrane domain (residues 1–176) of the Trp-7 mutant in planar bilayers. Channel activities were recorded at 100 mV in 1 m KCl buffer. The number of evaluated events and mean channel conductances for each protein are described under "Results."View Large Image Figure ViewerDownload Hi-res image Download (PPT) OmpA contains five tryptophans that are each located on a different transmembrane strand of the N-terminal β-barrel. In previous work we have used single Trp mutants of OmpA to monitor the site-specific rates of polypetide translocation and thus were able to dissect the folding pathway of this protein in lipid bilayers at unprecedent resolution (13.Kleinschmidt J.H. den Blaauwen T. Driessen A.J.M. Tamm L.K. Biochemistry. 1999; 38: 5006-5016Crossref PubMed Scopus (124) Google Scholar). These mutants each had four of the five native tryptophans replaced by phenylalanines. To support our earlier conclusion, which was based on the absence of a perturbation of a phage-binding epitope, that these mutations did not significantly perturb the overall structure and function of OmpA, we wanted to know whether these changes had an effect on the single channel conductance of refolded OmpA. Fig.2 shows that all five single Trp mutants that were used in our previous study (13.Kleinschmidt J.H. den Blaauwen T. Driessen A.J.M. Tamm L.K. Biochemistry. 1999; 38: 5006-5016Crossref PubMed Scopus (124) Google Scholar) refolded into C8E4 micelles as determined by the gel shift assay. Representative recordings of small channels of each of these refolded mutant proteins are shown intraces 2–6 of Fig. 3. Although there are clear differences in open channel noise levels and average open times, similar single channel conductance levels were observed for all mutants and these were not much different from those of the wild-type protein. The Trp-7 mutant exhibited well defined channels, which, however, were slightly smaller in conductance than the wild-type channels. Analyzing 217 events yielded a mean channel conductance of 51 pS (Fig. 4). The channels of the Trp-15 mutant were typically open for several seconds and closed only for very short durations. The 136 counted events exhibited a mean channel conductance of 54 pS (Fig. 4). A different pattern was observed with the Trp-57 mutant. The channels were either closed or open for several seconds, but while in their open state, they showed fast fluctuations which may correspond to rapid opening and closing events. For 647 events counted, the mean channel conductance was 67 pS. The channels formed by the Trp-102 and Trp-143 mutants displayed a similar behavior, but were open most of the time (Fig. 3). The 624 events counted for Trp-102 had a conductance of 59 pS, whereas the mean channel conductance of Trp-143 was 65 pS, counted for 114 events (Fig. 4). Thus, the mean channel conductance levels were about 60–70 pS for the refolded wild-type, Trp-57, Trp-102, and Trp-143 proteins and about 50–60 pS for native OmpA and the refolded Trp-7 and Trp-15 proteins. Large channels were also observed for all refolded single Trp mutants. For example, Trp-15 exhibited a mean conductance of 246 pS (1040 events) and Trp-143 a mean conductance of 270 pS (365 total >150 pS events; Fig. 5). In addition, relatively rare conductance steps of intermediate size (mean 191 pS) were observed with this mutant. Trp-7, Trp-57, and Trp-102 also showed intermediate size single channel conductance levels (only Trp-7 is shown as an example in Fig. 5). Channels of the order of 198 pS conductance (98 events), 155 pS (42 events), and 161 pS (70 events) were observed for the Trp-7, Trp-57, and Trp-102 mutants, respectively. Fig. 3 (bottom trace) shows the recording of a further control that was carried out with the Trp-57 mutant protein. In this case, the protein was deliberately misfolded. It is well known that OmpA hydrophobically collapses, exhibits a very different CD spectrum, and eventually aggregates when diluted from an 8 m urea solution into an aqueous buffer lacking detergent micelles or lipid bilayers (16.Kleinschmidt J.H. Wiener M.C. Tamm L.K. Protein Sci. 1999; 8: 2065-2071Crossref PubMed Scopus (120) Google Scholar, 23.Surrey T. Jähnig F. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 7457-7461Crossref PubMed Scopus (207) Google Scholar). When an equivalent amount of water-collapsed OmpA was added to the planar bilayers, neither small nor large single channel events could be observed. This experiment demonstrates that the channels observed in the upper traces of Figs. 3 and 5 are because of functionally refolded OmpA. Because OmpA is a two domain protein consisting of an N-terminal transmembrane β-barrel domain and a periplasmic C-terminal domain, it is interesting to ask whether the transmembrane domain alone exhibits a similar channel activity as the full-length protein. To address this question we expressed the N-terminal domain (residues 1–176) of the Trp-7 mutant protein, designated Trp-7-(1–176). Trp-7-(1–176) was correctly targeted to the outer membrane of E. coli, which gives a first indication that this domain may fold and assemble normally in the outer membrane. The purified fragment was then refolded in C8E4 and analyzed by its migration on SDS g