HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Tory, M. C. Articles by Merrill, A. R. Search for Related Content PubMed PubMed Citation Articles by Tory, M. C. Articles by Merrill, A. R. Social Bookmarking                      What's this?

J Biol Chem, Vol. 274, Issue 35, 24539-24549, August 27, 1999

Adventures in Membrane Protein Topology
A STUDY OF THE MEMBRANE-BOUND STATE OF COLICIN E1^*

Monica C. Tory and A. Rod Merrill

From the Guelph-Waterloo Centre for Graduate Work in Chemistry and Biochemistry, Department of Chemistry and Biochemistry, University of Guelph, Guelph, Ontario, N1G 2W1, Canada

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The molecular aggregate size of the closed state of the colicin E1 channel was determined by fluorescence resonance energy transfer experiments involving a fluorescence donor (three tryptophans, wild-type protein) and a fluorescence acceptor (5-(((acetyl)amino)ethyl)aminonaphthalene-1-sulfonic acid (AEDANS), Trp-deficient protein). There was no evidence of energy transfer between the donor and acceptor species when bound to membrane large unilamellar vesicles. These experiments led to the conclusion that the colicin E1 channel is monomeric in the membrane-bound closed channel state. Experiments were also conducted to study the membrane topology of the closed colicin channel in membrane large unilamellar vesicles using acrylamide as the membrane-impermeant, nonionic quencher of tryptophan fluorescence in a battery of single tryptophan mutant proteins. Furthermore, additional fluorescence parameters, including fluorescence emission maximum, fluorescence quantum yield, and fluorescence decay times, were used to assist in mapping the topology of the closed channel. Results suggest that the closed channel comprises most of the polypeptide of the channel domain and that the hydrophobic anchor domain does not transverse the membrane bilayer but nonetheless is deeply embedded within the hydrocarbon core of the membrane. Finally, a model is proposed which features at least two states that are in rapid equilibrium with each other and in which one state is more heavily populated than the other.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The cytotoxic function of the bactericidal protein colicin E1 is found in the ability of the protein to form voltage-gated, ion-conductive channels within the cytoplasmic membrane of susceptible cells. However, despite recent strides in the elucidation of the soluble structures of whole colicin (1) and various channel domain fragments (2-4), the current state of knowledge of the structure and function of membrane-associated colicins is modest at best. Colicin E1, secreted by Escherichia coli that possess the naturally occurring colE1 plasmid, consists of three functional domains: the translocation, receptor-binding, and channel-forming domains. Initially, the receptor-binding domain (5) interacts with the vitamin B₁₂ receptor of target cells (6). After recognition, the translocation domain interacts with the tolA gene product, which permits the translocation of colicin E1 across the outer membrane and into the periplasm (7). In the periplasm the channel domain undergoes a conformational change to an insertion-competent state and then inserts spontaneously into the cytoplasmic membrane, forming an ion channel. The channel translocates monovalent ions, resulting in the dissipation of the cationic gradients (H⁺, K⁺, Na⁺) of the cell, causing depolarization of the cytoplasmic membrane (8, 9). In an effort to compensate for the membrane depolarization effected by the colicin E1 channel, Na⁺/K⁺ ATPase activity is increased in the target cell, resulting in the consumption of ATP reserves, without concomitant replenishment (10). The final outcome is host cell death.

Treatment of colicin E1 with thermolysin (11) generates a compact fragment composed of the COOH-terminal channel-forming domain, or "channel peptide," which forms functional channels in osmotically shocked cells (10), membrane large unilamellar vesicles (LUVs),¹ (12) and within planar bilayers (13). Before insertion into model membrane systems, low pH-induced activation (14) is required.

Previous studies have generated evidence in support of both a monomeric and an oligomeric colicin E1 channel; however, the number of channel-forming domains required to form a functional colicin E1 channel has not been clearly established. In favor of a monomeric channel, Jacob et al. (15) demonstrated that cytotoxicity is first order with respect to colicin E1 concentration and proposed that a single channel is capable of killing a cell. In addition, the in vivo (16) and in vitro (17) efflux of K⁺ and intravesicular markers through colicin E1 channels also obey first-order kinetics. The implications of these kinetic studies were confirmed by Levinthal and co-workers (18), who demonstrated a 1:1 stoichiometry for active peptides forming channels in both small unilamellar vesicles and LUVs based on the discharge of a membrane potential in these target membrane systems. Also, steady-state conductance through colicin E1 channels is linear with respect to colicin E1 concentration (12). Finally, ultracentrifugation sedimentation equilibrium studies (19) and hydrodynamic Stokes radii comparison (20) indicated that colicin E1 is monomeric in solution over a broad pH range.

Despite this evidence in support of a monomeric channel, the ability of the colicin E1 channel to accommodate large organic molecules, including NAD⁺ (21), is not easily reconcilable with a monomeric structure. Carboxyl-terminal colicin E1 peptides, with as few as 88 amino acid residues, can form modestly functional channels in vitro (22); however, it is unlikely that a channel formed by such a short fragment would have a diameter sufficient to permit the passage of large organic molecules. More recently, the dimensions of the colicin I_a channel were mapped using a method based on the filling of the channel by nonelectrolytes from either side of the membrane. The channel was found to have cis and trans entrance diameters of 18 and 10 Å, respectively, with a constriction size of 7 Å (23). Notably, a channel with such a large lumen is not consistent with a two- or four-helix bundle assembled from a monomeric protein.

The colicin E1 channel peptide is a globular protein that consists of 10 -helices organized into three layers (3). Layer A is composed of helices 1, 2, and 10; layer B contains helices 8 and 9; and layer C contains helices 3, 4, 6, and 7. Helices 5_a and 5_b transverse part of the protein structure by running across the layers and are not included in any of the designated layers. The orientation of the channel domain helices in the membrane-bound state of the channel peptide has been the subject of a number of investigations. Hydrophobic helices 8 and 9 and the loop connecting them have been proposed to create a hydrophobic hairpin anchor for the induction of the membrane channel (24, 25). A model for colicin A has been proposed wherein most helices open like an "umbrella" on the surface of the membrane (26). An alternate model, called the "penknife model," has been suggested which proposes that helices 8 and 9 only partially penetrate into the membrane bilayer (27, 28).

In an earlier study, Palmer and Merrill (29) used nitroxide-labeled phospholipids as fluorescence quench probes to determine the membrane depths of Trp residues in single Trp mutant proteins of the colicin E1 channel peptide by a parallax analysis method (30). Recently, this experimental approach was extended in an investigation where a new shallower, nitroxide quencher (attached to the phospholipid headgroup) (31) was used as an additional quench species to probe Trp depths within the membrane bilayer. In the latter report, it was suggested that helices 8 and 9 are embedded in the membrane in a tilted fashion and do not protrude through the bilayer to the trans side. It was also found that the bilayer thickness could affect the structure of the colicin E1 channel domain in the membrane.

In the present investigation, one goal was to use fluorescence resonance energy transfer (FRET) in an attempt to reconcile the dispute over whether the colicin E1 channel is monomeric or oligomeric. A second objective was to use the water-soluble fluorescence quencher, acrylamide, to probe the solvent accessibility of Trp residues within the membrane-bound state of the colicin E1 channel domain. The second objective was rendered achievable by the use of single Trp mutant channel proteins of colicin E1 as characterized previously (31, 32). The fluorescence quench data were also correlated with various other Trp fluorescence parameters for the Trp mutant proteins, including fluorescence lifetimes, fluorescence quantum yields, and fluorescence emission maxima (_emmax) values.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Preparation and Purification of WT, Single Trp, and Trp Colicin E1 Channel Peptides-- The Trp-deficient (Trp) and single Trp mutant proteins of colicin E1 were prepared using site-directed mutagenesis to replace the three native Trp residues (W-424,² W-460, and W-495) with Phe residues, or leave/substitute a Trp residue at a single site, as described previously (14, 20). Colicin E1 was purified from a lex A strain of E. coli (IT3661) which contained the pSKE1 plasmid with the appropriate mutation (except for the WT protein) and the channel peptide isolated by digestion with thermolysin (32).

Preparation of LUVs-- LUVs composed of 60:40 (mol:mol) POPC:POPG (Avanti Polar Lipids, Birmingham, AL) were prepared by extrusion, using a hand-held extruder (Liposofast^TM, Avestin Inc., Ottawa, ON) according to a method described previously (29). Phospholipid concentration was determined using a microBartlett assay (33) following degradation of the phospholipids by ashing (34).

Preparation of AEDANS-labeled Trp Peptide-- The AEDANS-labeled Trp channel peptide was prepared, purified, and characterized according to methods recently developed in our laboratory (35, 36).

Spectroscopic Measurements-- Steady-state fluorescence measurements were performed using a PTI-Alphascan-2 spectrofluorometer (Photon Technologies Inc., South Brunswick, NJ) equipped with a thermostated cell holder. Fluorescence spectra were obtained using 295-nm excitation light, which selectively excites the Trp residue(s) of single Trp or WT channel peptides. The excitation and emission slit widths were 2 nm and 4 nm, respectively, and the temperature was maintained at 20 °C. For all fluorescence measurements, a wedge depolarizer (Oriel Corp., Stratford, CT; 95% transmittance) was placed on the exit side of the excitation monochromator, and fluorescence emission was detected at right angles. Steady-state Trp fluorescence emission was scanned from 303 to 450 nm upon 295-nm excitation. Corrections were made for the appropriate blanks and for the wavelength-dependent bias of the optical and detection systems. The fluorescence emission maxima values were determined by calculating the first derivative of the smoothed, corrected emission spectra for all samples and recording the wavelength where the derivative was zero.

FRET Measurements of Membrane-bound Donor and Acceptor Channel Peptide-- Titrations were performed according to one of two methods. In one method, AEDANS-labeled Trp channel peptide in LUVs was titrated with WT channel peptide while monitoring the AEDANS fluorescence emission. In the other method, WT channel peptide in LUVs was titrated with AEDANS-labeled Trp channel peptide, and the fluorescence emission of the Trp residues of WT channel peptide was monitored.

Samples were prepared in quartz cuvettes, by first preparing 0.24 µM AEDANS-labeled Trp channel peptide (or 0.24 µM WT channel peptide) in 10 mM dimethylglutarate, 100 mM NaCl, pH 4.0, followed by the addition of 400 µg of 60:40 (mol:mol) POPC:POPG LUVs. After the addition of LUV, the samples were incubated for 5 min with magnetic stirring to allow protein-LUV binding to reach equilibrium. Stirring was discontinued during spectroscopic measurements to minimize background signal. A control sample was prepared to test for energy transfer in the absence of LUVs. Blanks to correct for LUV-induced scatter contained LUVs and unlabeled Trp channel peptide. During the titration experiments, aliquots (5 µl each) of WT channel peptide (or AEDANS channel peptide) were added to the stirred samples, allowing 15 min of equilibration time between additions, prior to spectroscopic measurements.

Fluorescence Lifetime Measurements-- Time-resolved fluorescence measurements were performed using a PTI LaserStrobe model C-72 lifetime fluorometer (Photon Technology International). The excitation source was a pulsed nitrogen laser pumping a dye laser with a frequency doubler. The laser operated at 10 Hz, and the detection channel consisted of an emission monochromator with a stroboscopic detector. The data analysis was performed with a 1-to-4 exponential fitting program involving iterative reconvolution and minimization of chi-square. The excitation wavelength was 295 nm, and the emission was set at 340 nm with 1-nm band passes. The <_a> was calculated from the relationship <_a> = _ii/_i, and because _i = 1 (normalized preexponential values) then <_a> = _ii. The <_a> parameter was used in all cases except for the calculation of k_q from the Stern-Volmer constant (K_SV). The <_f> parameter was calculated from the equation: <_f> = _i²/_ii. The <_f> value was used only for the calculation of the k_q values from the K_SV data shown in Table I (37) where k_q = K_SV/<_f>.

Model for Peptide Association-- The degree of association of colicin E1 channel peptides within the membrane bilayer of LUVs was assessed by measuring FRET from a donor population (the three Trp residues in WT channel peptide) to an acceptor population (AEDANS-labeled Trp channel peptide). The derivation of the model for association of channel peptides (Equation 1), adapted from Veatch and Stryer (38), has been presented previously (36).

(Eq. 1)

Q_F/Q₀, the relative quantum yield of the donor species, is a ratio of the quantum yield of the donor species in the presence (Q_F) and absence (Q₀) of the acceptor species. E represents the efficiency of the FRET process, d is the distance separating the centers of the donor and acceptor chromophores, and n is the number of colicin E1 channel peptides that compose each functional colicin E1 channel.

Relative Quantum Yields (RQ_F)-- The relative Trp fluorescence quantum yields were measured for membrane-associated channel peptides by initially determining the quantum yield of the proteins in solution using N-acetyl-L-tryptophanamide as the quantum standard (Q_F = 0.14) in order to quantum calibrate the proteins as reported previously by our laboratory (14). This was conducted by exciting the samples at 295 nm (absorbance at the excitation wavelength for the various samples ranged between 0.03 and 0.05) while scanning the fluorescence emission from 305 to 450 nm in 0.5-nm increments (2-nm slits for both excitation and emission). To determine the RQ_F values for the membrane-bound peptides, an aliquot of the calibrated protein solution (quantum yield calibrated) was mixed with vesicles as described above. After a 15-min equilibration time, the Trp fluorescence from the proteoliposome solutions was scanned, and the quantum yield was calculated based on the previous determination for the membrane-free protein solution. The RQ_F was determined by a simple ratio, Q_F,mem/Q_F,sol where Q_F,mem is the Q_F value for the protein-lipid solution, and Q_F,sol is the Q_F for protein in solution.

Acrylamide Quenching in LUVs-- Samples were prepared by the addition of 60:40 DOPC:DOPG LUVs (40 µg) to the colicin E1 channel peptide (60 µg) in 20 mM dimethylglutarate, 130 mM NaCl, pH 4.0, in a 1-cm path length quartz cuvette (21 µM LUV; 1.27 µM colicin E1 channel peptide). The samples were incubated at 25 °C for 15 min to ensure complete binding of colicin E1 channel peptide to the membrane LUVs. A blank that did not contain protein was prepared to correct for the contribution of LUVs to the fluorescence emission. Aliquots of an acrylamide stock solution (4.0 M acrylamide in water) were added to the cuvette. Mixing was performed with a digital pipette set to 300 µl. Acquisition of fluorescence data was initiated immediately after the mixing because it was found that the signal was constant as a function of time after mixing. Fluorescence measurements were conducted with 295-nm excitation light, and the emission was monitored at 340. A 309-nm cutoff filter (Oriel Corporation, Stratford, CT) was placed in front of the emission monochromator entrance slit to minimize the contribution of LUV-induced light scattering to the fluorescence emission signal. The fluorescence signal was collected for 30 s, and the data were analyzed by integrating the area under the 30-s time trace. The contribution of the LUVs blank was automatically subtracted by the instrument/software. Quenching data were plotted as the ratio of fluorescence in the absence of quencher (F₀) to the intensity in the presence of quencher (F_i) against quencher concentration as described previously (39). The calculated slope was equated to dynamic parameters according to the modified Stern-Volmer equation: F₀/F_i = 1 + K_SV[Q], where K_SV is equal to k_qf (39, 40).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Colicin E1 Single Trp Mutant Channel Peptides-- The primary sequence for the colicin E1 thermolytic channel peptide (molecular weight 19,700; NH₂ terminus, Ile-345) is shown in Fig. 1A. The three naturally occurring Trp residues are located at positions 424, 460, and 495. 14 single Trp mutant peptides were chosen for investigation by Trp fluorescence spectroscopy in order to obtain site-specific topological information on various segments (helices) within the channel peptide upon the transition from the water-soluble to the membrane-bound state. The specific locale for the single Trp residues within these mutant peptides include: F355W, Y367W, F404W, F413W, Trp-424, F431W, F443W, Trp-460, Y478W, F484W, L492W, Trp-495, I499W, and Y507W. Phe-355 is located in the middle of helix 1 where this residue faces inward and is nestled between helices 5_b and 8. Tyr-367 is located at the NH₂ terminus of helix 2 and is largely buried at the interface among helices 8, 9, and 10 in the x-ray structure. Phe-404 is located in the loop region between helices 3 and 4 and is nearly fully exposed; Phe-413 is located in the middle of helix 4 and appears to be partially exposed to solvent and relatively close to Trp-495. Trp-424 is largely buried in the middle of helix 5_a, a small polar helix. Phe-431 is located in the middle of the small helix 5_b, a very flexible region of the channel peptide, and is mostly buried and inaccessible to the aqueous solvent. Phe-443 is located in the extended loop structure between helices 5_b and 6 and is close enough to Lys-470, which is a candidate for the source of the internal quenching seen in the F443W single Trp mutant (32), and the Trp-443 residue is highly exposed to the aqueous solvent (39). Trp-460 is located in a moderately polar region within the channel peptide, at the NH₂ terminus of helix 7. Tyr-478 is located at the NH₂ terminus of nonpolar helix 8 and is packed inside the protein in a largely nonpolar pocket that has Lys-470 at one end. Phe-484 is located in the middle of helix 8 and is well sequestered from solvent but may become exposed by any movement of helice 1 or 2. Leu-492 is located in the loop region between hydrophobic helices 8 and 9 and is located in a nonpolar environment. Trp-495 is situated at the NH₂ terminus of helix 9 and is only slightly exposed to the aqueous solvent. Moreover, Ile-499 is located in the middle of nonpolar helix 9 and is sandwiched between helices 4 and 6 in a hydrocarbon milieu. Finally, Tyr-507 is located near the COOH terminus of nonpolar helix 9 with its phenolic group largely buried and facing toward the center of the peptide core; it is the most inaccessible Tyr residue to the aqueous solvent (39). Furthermore, Trp-478, 484, 492, 495, 499, and 507 are located within the hydrophobic domain of the channel peptide, a nonpolar segment of the peptide consisting of a stretch of 35 amino acid residues; consequently, these Trp residues should report peptide structural changes involving the purported membrane anchor domain.

View larger version (54K): [in this window] [in a new window]   Fig. 1.   Panel A, primary sequence of the thermolytic colicin E1 channel peptide. Residue numbering corresponds to the numbering of whole colicin E1 (54). The locations of the -helices assigned from the crystal structure of the P-190 peptide (3) are indicated as shaded cylinders, and the hydrophobic -helical domain is underscored. The three naturally occurring Trp residues are identified with a filled circle above the residue number, and the sites where a Trp residue has been substituted into the peptide by site-directed mutagenesis to generate single Trp mutants are indicated by indole rings. Panel B, alignment of the helical segments in the (i) soluble peptide and (ii) the proposed transmembrane and translocating peptides for the membrane-bound state of the colicin E1 channel domain. The helices are designated as H1-H10, the trans-membrane segments as TM1-TM4, and the translocating segment as TLP. Helix 5 is shown as consisting of helices 5a and 5b, and the disposition and structure of helix 10 are not known in the membrane-bound peptide.

Helical Alignment of Water-soluble and Membrane-bound Channel Peptide-- The comparison of the helices in the water-soluble and the transmembrane (TM) and translocating peptides (TLP) of the membrane-bound channel peptide is shown in Fig. 1B, i-ii. The assignment of the helical segments for the water-soluble structure is based on the x-ray data (3), whereas the assignment of the TM and TLP segments within the membrane-associated channel domain is much less straightforward. The segments designated TM1, 2, 3, and 4 represent TM segments whose existence has been implied from various topological mapping experiments (see "Discussion"). The segment TLP refers to the translocating segment that has been located on the trans side of the membrane by biotinylation and epitope-mapping experiments (41-43). The membrane disposition of this peptide segment in the open channel state is highly controversial and is currently being investigated. Furthermore, the structure and disposition of helix 10 are currently not known in the membrane-bound peptide; consequently, it is shown as a single helix in Fig. 1Bii. A summary of the data presented in Fig. 1B is as follows. The open state of the colicin E1 channel is proposed to consist of four trans-membrane segments: TM1 (helix 1 and part of helix 2), TM2 (helices 6 and 7), TM3 (helix 8), and TM4 (helix 9). In addition, an unusually long segment of the peptide has been identified on the trans side of the membrane; this segment, TLP, is composed of helices 3, 4, 5_a, and 5_b.

Colicin E1 Channel Aggregate State-- Representative fluorescence emission spectra for the WT colicin E1 channel peptide in the presence of LUVs at various concentrations of acceptor are presented in Fig. 2, A and B. Titration of LUV-incorporated WT colicin E1 channel peptide with AEDANS-labeled Trp channel peptide did not cause appreciable quenching of the fluorescence emission of WT colicin E1 channel peptide as a result of intermolecular Trp-AEDANS FRET. While the fluorescence emission signal arising from the three naturally occurring Trp residues of WT channel peptide remained essentially unchanged during the titration, the signal arising from the AEDANS moiety of AEDANS-labeled Trp channel peptide increased as the titration progressed. This increase correlated with the change in concentration of AEDANS (as AEDANS-labeled Trp channel peptide), and consequently, the increase in AEDANS fluorescence emission signal (Fig. 2A) is not caused by FRET from the three Trp residues of WT channel peptide.

View larger version (40K): [in this window] [in a new window]   Fig. 2.   Panel A, titration of WT colicin E1 channel peptide in LUVs with AEDANS-labeled Trp-deficient channel peptide. The WT peptide (0.24 µM) was titrated with AEDANS-Trp peptide (0-0.25 µM) in 218 µM lipid and 100 mM NaCl, 10 mM dimethylglutarate buffer, pH 4.0 (AEDANS labeling stoichiometry of Trp peptide  1:1). The numbers (1-7) indicated on the figure refer to the following concentrations of AEDANS-Trp peptide (AEDANS:Trp peptide, 1:1 molar ratio): 1, 0 µM; 2, 0.042 µM; 3, 0.084 µM; 4, 0.13 µM, 5, 0.17 µM; 6, 0.21 µM; and 7, 0.25 µM. The donor Trp residues of the WT peptide were excited selectively using 295-nm light. In the case of an oligomeric channel, energy transfer from the donor (Trp) to the acceptor (AEDANS) chromophore is expected to be manifested as a decrease in the Trp fluorescence emission with a corresponding enhancement of the AEDANS fluorescence emission. The Trp fluorescence emission, however, remained relatively constant, indicating that the colicin E1 channel formed in LUVs is monomeric. Panel B, titration of AEDANS-labeled Trp-deficient colicin E1 channel peptide in LUVs with WT channel peptide. AEDANS-labeled Trp peptide (0.125 µM) was titrated with WT channel peptide (0-0.25 µM) in 105 µM lipid and 100 mM NaCl, 10 mM dimethylglutarate, pH 4.0, buffer. 11 spectra are overlaid in the figure with each spectrum representing a 0.025 µM increment increase in the WT peptide concentration. In the case of an oligomeric channel, energy transfer from the donor (Trp) to the acceptor (AEDANS) chromophore is expected to be manifested as an increase in the AEDANS fluorescence emission. The AEDANS fluorescence, however, remained largely unchanged, indicating that the colicin E1 channel formed in LUVs is monomeric.

The degree of quenching of the intrinsic fluorescence emission of WT channel peptide was assessed by comparing the area of the fluorescence emission peak arising from the donor species (Trp) in the presence of the highest concentration of acceptor (AEDANS) to the area of the donor fluorescence peak in the absence of acceptor. The mean quenching of donor fluorescence was only 6.4 ± 13% and did not show an increasing trend with the addition of acceptor.

In contrast, Steer and Merrill (36) demonstrated that FRET occurred within a dimeric unfolding intermediate of the colicin E1 channel peptide, using the same donor-acceptor pair. The observation of FRET between the Trp-AEDANS donor-acceptor pair (36) serves as a convenient positive control for this investigation. The absence of intermolecular Trp-AEDANS FRET under conditions where the channel peptide readily associates with membrane bilayers indicates that adjacent channel peptides are greater than 50 Å apart. A separation distance as large as 50 Å is not consistent with the formation of a multimeric channel. To confirm that the absence of energy transfer was indicative of a monomeric colicin E1 channel, the FRET data were plotted according to the channel peptide association model (Equation 1). The data did not fit a model with greater than one subunit (data not shown). The absence of FRET in this series of experiments supports previous evidence for a monomeric colicin E1 channel.

Results from the titration of AEDANS-labeled Trp channel peptide in LUVs with WT channel peptide also support a monomeric colicin E1 channel. The fluorescence emission arising from LUV-associated AEDANS-labeled Trp channel peptide remained unchanged as additional donor (WT) channel peptide was added to the system (Fig. 2B). The absence of Trp-AEDANS FRET indicates that the colicin E1 channel is composed of a single channel peptide within the membrane bilayer.

_emmax Values of Membrane-bound Channel Domain-- The _emmax values for membrane-associated WT and 14 single Trp mutant proteins of the colicin E1 channel peptide are shown in Table I. The _emmax values for the WT and a subset of the soluble single Trp mutant proteins (11 in total) were reported previously by our laboratory (39). The _emmax values shown in Table I range from 325.1 nm (W-495) to 337.5 nm (W-404). Eight of the 14 membrane-associated single Trp mutant proteins, along with the WT protein, possessed _emmax values that were greater than 330 nm. The remaining six mutant proteins exhibited _emmax values that were less than 330 nm but greater than 325 nm. These data are in contrast with the _emmax values for the soluble mutant proteins; in the latter situation, the values were generally more blue-shifted, i.e. shifted to shorter wavelength values. This indicates that the channel peptide is in a more polar (with mobile polar groups) environment upon membrane association (closed channel state in the absence of a membrane potential). This notion can be illustrated further by considering the average _emmax for the WT and 11 single Trp mutant proteins in the soluble as opposed to the membrane-bound forms (mean _emmax of 330.9 ± 3.7 and 327.3 ± 5.8 nm for membrane and soluble forms, respectively). Interestingly, the WT protein showed an 8-nm red shift in its _emmax value upon membrane association, reflecting a change in the chemical environment surrounding the bound peptide (Table I, Ref. 39, and Fig. 3).

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Change in the fluorescence emission maximum upon membrane association of the colicin E1 channel peptide. The emmax parameter was determined by the difference between the emmaxmem (membrane-bound protein) and the emmaxsol (soluble protein) for the Trp fluorescence emission of the corresponding protein. A positive value indicates a red shift in the fluorescence emission maximum upon membrane association, and a negative value indicates a blue shift. The emmax values for the membrane-bound peptides were taken from the data in Table I, and the emmax values for the soluble peptides were determined previously (39). The emmax data were acquired as described under "Experimental Procedures" and in the legend to Table I.

The WT peptide possesses three Trp residues at 424, 460, and 495 and thus reports on more global changes occurring within the COOH-terminal two-thirds of the protein. A plot showing the divergence of the _emmax values (_emmax = _emmax_mem values  _emmax_sol values) between the soluble and membrane-bound channel peptide states is shown in Fig. 3. There are six proteins that show large red shifts in _emmax values upon membrane binding (_emmax values between 5 and 11 nm). Each of these proteins (WT, W-367, W-413, W-460, W-492, and W-507) contains a single Trp residue (except for the WT protein) at various sites within the primary sequence of the channel peptide (Fig. 1A). In the soluble peptide, Trp-367 is located in the middle of helix 2 in a relatively polar environment surrounded by Gly-364, Lys-366, Ser-368, and Met-379. It has a blue-shifted _emmax value (320 nm, Ref. 39) and an intermediate quantum yield (0.18, Table I). The former can be explained by its relative inaccessibility to the aqueous solvent (39), whereas the latter is likely caused by the close proximity of this residue to the sulfur atom of Met-370. Trp-413 is found in the center of helix 4 in a nonpolar pocket lined with Ile-412, Phe-413, and Leu-416. This residue is part of the pH-activated trigger mechanism of the channel peptide proposed recently by our laboratory (14). Trp-460 is a naturally occurring residue that is found in a loop region between helices 6 and 7, possesses a relatively high Q_F,sol (0.27, Ref. 14), and is enveloped in a nonpolar niche within the soluble peptide. Trp-492 is also found in a loop region between hydrophobic helices 8 and 9 and is mostly buried in a nonpolar region of the soluble peptide. Trp-507 is situated at the COOH terminus of nonpolar helix 9 where it is buried in the core of the peptide (39) but possesses a low quantum yield (0.05, Table I) probably because of the proximity of the thiol group of Cys-505. Mutant proteins W-355, W-404, W-424, W-478, W-484, and W-495 show smaller, less significant red shifts in _emmax values upon membrane binding (Fig. 3). Two mutant proteins, W-431 and W-499, exhibit significant blue shifts in the _emmax values upon membrane association; both of these Trp residues are located in the center of helices (helices 5_b and 9, respectively; Fig. 1A).

Trp Quantum Yields of Membrane-bound Channel Peptides-- The fluorescence quantum yields for the soluble forms of nine of the single Trp mutant (and WT) proteins were measured previously using N-acetyl-L-tryptophanamide as the quantum standard (14). The quantum yield for the membrane-associated mutant proteins are shown in Table I. The Q_F value for the soluble WT protein, which possesses Trp-424, 460, and 495, represents the average of the Q_F values for the three single Trp mutant proteins (14). However, this is clearly not the case for the membrane-bound proteins. The WT has a Q_F value of 0.19, whereas the average of W-424, W-460, and W-495 is 0.36. This indicates that there is significant structural rearrangement of the WT peptide upon membrane association which is only observed for the single Trp mutants because Trp to Trp FRET can still occur within the WT peptide. It would appear that the Q_F value of the membrane-bound WT peptide reflects this possibility for the W-424 bound protein, suggesting that Trp to Trp FRET occurs from Trp-460 and 495 to Trp-424. The membrane-bound Q_F values for the WT and single Trp mutants ranged from 0.17 to 0.51. Trp residues exhibiting high Q_F values include: (Q_F,mem  0.30) Trp-355, 404, 413, 431, 443, 460, 495, and 507. In addition, the RQ_F values were calculated for the WT and 12 of the membrane-associated single Trp mutant proteins, and these are listed in Table I. Mutant proteins W-355, W-404, W-443, and W-507 showed a large increase (>2 fold) in RQ_F upon membrane binding, whereas only W-484 exhibited a significant decrease in its RQ_F value (0.7 ± 0.03, Table I). The remaining seven mutant proteins (W-367, W-413, W-424, W-431, W-460, W-478, and W-495; Table I) showed little or no change in their RQ_F values. Consequently, from the data in Table I, the Trp residues within the membrane-bound mutant channel peptides which may be considered as being located within a hydrophobic environment include Trp-355, 404, 413, 431, 443, 460, 495, and 507. From this subset of Trp residues, a few can be discounted based on the following observations. Trp-495 is located within a nonpolar environment in the soluble peptide, and its Q_F does not change upon membrane binding; hence, it is difficult to assess whether any change in environment has occurred for Trp-495. The increase in RQ_F values for Trp-404, 443, and 507 may partly be the result of alleviation of internal quenching mechanisms upon the transition of the soluble peptide to the membrane-bound state based on previous observations for the x-ray structure of the soluble peptide (3, 14, 32, 35, 38, 39). However, all of these Trp residues possess relatively high Q_F values. Therefore, these data imply that Trp-355, 404, 413, 431, 443, 460, and 507 are at least partially embedded within the membrane bilayer.

Acrylamide Quenching of Trp Fluorescence-- The Stern-Volmer quenching data for the WT, N-acetyl-L-tryptophanamide standard, and 14 single Trp mutant proteins are shown in Table I. The Stern-Volmer quench constant (K_SV) values ranged from 0.93 M¹ (W-507) to 3.95 M¹ (W-413). However, because the K_SV values are composite constants that possess the parameter of the <_f> (where K_SV = k_q<_f>) a better measure of the collisional rate between the Trp chromophore in the channel peptide and the aqueous quencher, acrylamide, is the bimolecular quench constant, k_q. The second-order bimolecular quench constant (k_q) values ranged from 0.24 M¹ ns¹ (W-507) to 0.95 M¹ ns¹ (W-413). Consequently, the channel proteins were divided into two classes according to their accessibility to acrylamide as a water-soluble, fluorescence quencher. Group I consisted of WT, W-413, W-424, W-478, W-484, and W-495 proteins that showed k_q values greater than 0.5 M¹ ns¹. Group II is comprised of W-355, W-367, W-404, W-431, W-443, W-460, W-492, W-499, and W-507 proteins that were less accessible to the aqueous quencher and, hence, had k_q values that ranged from 0.24 to 0.46 M¹ ns¹ (Table I). The acrylamide quenching data for the soluble form of 11 single Trp mutant proteins was reported previously (39). Generally, as expected, both of these quench parameters, K_SV and k_q, were lower for the membrane-bound compared with the soluble form of the proteins (Table I, Fig. 4, and Ref. 39), although changes in <_f> upon the transition from solution to membrane-bound form may result in a change in K_SV without a concomitant alteration in k_q (this was observed for W-418, W-424, and W-484 proteins; Table I, Figs. 4 and 5, and Ref. 39).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Change in the average bimolecular quench constant, kq, upon membrane binding of the colicin E1 channel peptide. The kq value represents the difference between the kq,sol (soluble protein) and kq,mem (membrane-bound protein). A positive kq value indicates a reduced accessibility of the Trp residue in the protein upon membrane association, and a negative value represents an enhanced accessibility. The kq data for the membrane-bound peptides were taken from Table I; the kq data for the soluble proteins were taken from previous data (39). The kq data were acquired as described under "Experimental Procedures" and in the Table I legend.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   Change in the amplitude average fluorescence lifetime upon membrane binding of the colicin E1 channel peptide. The <a> parameter was determined by subtracting the <a>sol value for the soluble peptide from the <a>mem value for the membrane-bound peptide. A positive <a> value indicates that the <a> increased for the membrane-bound peptide relative to the soluble peptide. The <a> data for the membrane-bound peptides were taken from Table I, and the <a> data for the soluble proteins were taken from previous data (39). The <a> data were obtained as described under "Experimental Procedures" and in the Table I legend.

The change in k_q values of the various single Trp channel peptides, and hence the extent of solvent accessibility, upon the binding of the channel peptides to a LUV membrane surface is shown in Fig. 4 (k_q = k_q,sol  k_q,mem). Notably, a positive k_q indicates a reduction in the solvent accessibility of the Trp residue(s) being probed or, in other words, a protection of the Trp residue within the channel peptide from the aqueous solvent as provided by the membrane bilayer structure. It is obvious from the data in Fig. 4 that generally there is a reduction in aqueous solvent exposure of the Trp residues upon association of the channel protein with membranes (k_q > 0 for the WT and all single Trp mutant peptides except W-413, W-424, W-484, and W-495). The largest degree of protection of the channel peptide provided by the membrane bilayer upon the solution to membrane transition occurred with Trp residues 367, 404, and 443. This indicated that at these sites along the channel peptide there is significant interaction, and possibly immersion of Trp residues, within the membrane bilayer. Trp residues at 355, 431, 460, 492, and 499 were protected to a lesser extent (Fig. 4). Trp residues at 413, 478, 484, 495, and 507 showed little or no significant change in solvent exposure upon channel protein associating with a membrane surface. Trp-424 was the only site that showed a significant increase in solvent exposure upon peptide binding to LUVs. Notably, the WT protein k_q value was equal to the average of the three single Trp mutant protein k_q values and showed a modest decrease in solvent accessibility upon membrane binding, reflecting the average of W-424, 460, and 495 (Fig. 4).

Fluorescence Lifetimes-- The amplitude average fluorescence lifetime (<_a>) and the intensity average lifetime (<_f>) were calculated from the measured decay time components (37) for the membrane-associated channel peptides and are shown in Table I. The intensity weighted fluorescence lifetimes <_f> were used for the calculation of the k_q values only. The <_a> values can be more readily correlated with Trp photophysical processes and thus were used to aid in the interpretation of the topological data for the channel peptide. The <_a> values ranged from 2.54 ns (W-355 and 507) to 4.24 ns (W-460) and exhibited changes within a large number of peptides upon the soluble to membrane transition (14, 39; Fig. 5) but generally did not show the range reported previously for the soluble channel peptides (39). The change in decay times (<_a>) upon peptide binding to membrane LUVs is plotted in Fig. 5 (<_a> = <_a>_mem  <_a>_sol). A positive <_a> reflects an increase in the Trp fluorescence lifetime of the respective channel peptide upon binding to the membrane surface. Upon membrane association, seven peptides showed a significant increase in <_a> values (<_a> 0; WT, W-355, W-404, W-413, W-431, W-460, and W-507). Five single Trp mutant proteins, W-367, W-424, W-443, W-484, and W-492, showed a large decrease in <_a> values upon binding to a bilayer surface (Fig. 5), whereas W-495 and W-499 showed modest decreases in <_a> values when membrane-bound. W-478 exhibited little or no significant change in <_a> when converting from the water-soluble state to the membrane-bound form.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Channel Aggregate State by FRET Analysis-- The absence of FRET in this series of experiments is further evidence that the colicin E1 channel formed by the channel-forming domain in LUVs is monomeric. Other experiments that support a monomeric channel model include the original cytotoxicity experiments on colicin E1 in which colicin E1 was observed to act according to a "one-hit" mechanism, demonstrating that a single protein molecule could kill a cell (15). In addition, kinetic data indicate that channel formation is first-order with respect to either colicin E1 channel peptide or colicin E1 concentration (12), suggesting that a single molecule forms each channel.

One piece of evidence which stands in opposition of a monomeric colicin E1 channel is the observation that large organic cations, including NAD⁺, can pass through the channel (21) and that the hydrated channel dimensions include a 7 Å constriction (23) as determined with nonelectrolytes that fill the channel lumen. It is presently not clear how a 178-190-residue monomeric peptide can form a channel with a lumen of such large dimensions unless most, if not all, of the polypeptide chain is involved in forming the channel structure.

It is important to recognize that multimeric channels that are composed either exclusively of donor channel peptides or exclusively of acceptor channel peptides may be formed. Multimeric channels of this type would be spectroscopically silent, yielding FRET data that would be indistinguishable from FRET data that corresponded to the formation of monomeric channels. The model of Veatch and Stryer (38) (Equation 1) accounts for the probability of forming oligomers composed either exclusively of donor channel peptides or exclusively of acceptor channel peptides. It is highly improbable that the absence of FRET in this experiment was the result of every one of the channels formed arising from a combination of monomers of the same population.

The absence of FRET between the donor- and acceptor-labeled peptides indicates that the distance separating the two chromophores was greater than the distance over which significant FRET occurs, approximately 50 Å in this system (R_o = 22 Å for the Trp and AEDANS pair (35, 44). Notably, this dimension is the distance that separates the donor and acceptor chromophores and not the distance separating the peptides. The three Trp residues of WT channel peptide serve as the donor chromophores, and so we would expect, based on the x-ray structure for the soluble peptide (3), that there is a reasonable distribution of the three donor chromophores throughout the peptide. The acceptor chromophore, however, is an extrinsic fluorescent probe (AEDANS) that is covalently attached to the lone Cys residue, Cys-505, of the Trp channel peptide. Theoretically, it is possible for an oligomeric colicin E1 channel to form such that the AEDANS label of the Trp channel peptide is directed outward from the rest of the channel complex. In an arrangement of that type, the distance between the donor Trp residues and the acceptor AEDANS moiety could, conceivably, be large enough to prevent efficient FRET, resulting in a second type of spectroscopically silent oligomer. The magnitude of the Stokes radius of soluble colicin E1 channel peptide, 21-23 Å (20), however, essentially precludes an arrangement of monomers, within an oligomeric colicin E1 channel, which would have a donor-acceptor distance sufficient to prevent energy transfer.

In conclusion, Trp-AEDANS FRET was not observed for WT channel peptide and AEDANS-labeled Trp channel peptide incorporated within membrane bilayers. Previously, FRET had been observed (36) for a dimeric unfolding intermediate of colicin E1 channel peptide, indicating that the donor-acceptor pair was suitable for measuring the association of colicin E1 channel peptides. The absence of FRET for the membrane-associated channel peptides provides further evidence that the colicin E1 membrane-bound closed channel is monomeric.

Topology of the Membrane-bound Closed Channel State-- It is generally believed that the Trp _emmax parameter is the best gauge of the chemical environment of a Trp residue within a protein (14, 44-46). The _emmax data for the contingent of single Trp mutant and WT proteins strongly suggest that the polarity of the Trp residues (at 11 of the 14 different Trp sites within the peptide) increases when the protein associates with the membrane surface. This suggests that the protein unfolds and spreads out on the surface of the membrane, which is in agreement with the previous proposals for colicin A (26-28) E1 (29, 31, 47-49) and I_a (25, 41-43). The _emmax values for W-355 and W-367 single Trp proteins indicate that both helices 1 and 2 are membrane-associated under the conditions used in this study but that the corresponding Trp residues are in a more polar environment than those found within the hydrophobic anchor domain of the channel protein (Table I). Previously, Trp-355 and 367 were determined by depth-dependent fluorescence quenching analysis to reside at moderately buried positions within the membrane bilayer (29, 31). Upon membrane binding, W-367, 413, 460, 492, and 507 exhibit a red shift in their _emmax values, indicating that these residues are likely interfacial and are not buried deeply within the membrane bilayer. It appears that both Trp-495 and 499, found on the same side of helix 9, are located in highly nonpolar environments, probably within the bilayer hydrocarbon core, when the channel domain is membrane-bound (Table I). However, Trp-495 is more accessible to aqueous quencher than Trp-499. This may reflect the orientation of the hydrophobic anchor within the membrane because Trp-499 is in the middle of helix 9, whereas Trp-495 is near the amino terminus of that helix. This implies that helix 9 is relatively deeply embedded within the membrane in the closed state of the colicin channel. Trp residues at positions 492 and 507 define the ends of helix 9, and both Trps report a more polar environment. Helix 8 is also submerged within the bilayer as reported by Trp-478 and 484, which exhibit relatively blue-shifted _emmax values (Table I) and are located on opposite faces of this nonpolar helix. A complication to the interpretation of Trp locations within the closed state of the colicin channel may entail a consideration of the lumen geography of the pore or partial pore structure in the absence of a membrane potential. It is conceivable that a given Trp residue may be deeply embedded within the membrane but be relatively accessible to the solvent via the channel lumen through the indole polar N-H group.

Acrylamide quenching experiments reported in this paper provide a means to probe the solvent accessibility of the Trp residues found at various sites throughout the closed channel structure. Acrylamide is believed to be superior to KI as a probe of the solvent exposure for a lipid-associated fluorophore because the former does not partition to a significant degree into the membrane bilayer and acts as a collisional fluorescence quencher (50). Notably, this approach does not simply provide a dipstick measure of the distance below the surface of a given Trp residue, as in the case where spin-labeled phospholipids were used as quenchers of indole fluorescence (29, 31), but also provides a means to probe any vestiges or nuclei of an embryonic channel lumen that exists as a precursor to the open channel state. Based on the overwhelming abundance of accumulated evidence, it seems most likely that the NH₂-terminal half of the channel peptide (closed channel state) is simply floating on the membrane surface with the amino-terminal helices (helices 1, 2, 3, and 4) oriented parallel to the lipid bilayer. However, the k_q data for helices 1 and 2 suggest that these two helices are appressed close to the surface of the membrane and are not fully exposed to the solvent (Table I and Fig. 4). An earlier report (47) indicated that there are a few protease accessible sites within this region of the membrane-associated peptide. This may simply reflect the dynamic nature of this association given that the digestion conditions used (47) were relatively harsh (1 h at 35 °C, high protease:protein mass ratio) and that not all of the protein was digested. Previous reports (20, 27, 28, 31, 41-43, 48) involving completely different approaches to probe the closed channel structure led to similar conclusions as to the dynamic nature and the existence of multiple states for membrane-bound colicin channel domains. In summary, the fluorescence quench data indicate that Trp-355, 367, 404, 431, 443, 460, 495, and 499 are partially embedded in the membrane bilayer and are protected from the aqueous solvent.

Trp-495 is deeply entrenched within the bilayer (29, 31), possesses a blue-shifted _emmax value, yet it is relatively accessible to acrylamide (k_q = 0.75 M¹ ns¹, Table I) and hence the aqueous solvent. This implies that the N-H group of its indole ring is partially exposed to the solvent in the closed channel structure. Unfortunately, it is not possible to discriminate between the presence of a membrane defect caused by protein insertion or whether this moderate level of solvent accessibility represents a primordial feature of the channel. However, the Trp-495 residue is not required for normal channel function because its replacement with Phe does not alter its channel properties (51).

The plots shown in Figs. 3-5 indicate that there are significant structural changes within the channel peptide upon the transition from the water-soluble structure to the membrane-bound state. The data in these figures indicate that the changes are not localized to a few sites within the peptide structure but are global in nature. Interestingly, the largest changes in the RQ_F values (membrane-bound peptides) were observed for single Trp mutant proteins that exhibited low Q_F values in solution (Table I and Ref. 39). This may indicate that the structural changes induced within the channel peptide upon membrane association cause an alleviation of the internal quenching mechanism(s) inherent to the soluble protein. However, the Q_F,mem values do indicate that the following Trp residues are likely membrane embedded: Trp-355, 404, 413, 431, 443, 460, 495, and 507. Unfortunately, the Q_F data were not determined for two additional Trp mutants that possess a Trp residue within the hydrophobic anchor domain of the channel peptide, W-492 and W-499. It is anticipated that these residues would also possess high Q_F,mem values similar to Trp-495 and 507.

Presently, there is no theoretical basis for assigning membrane location and/or depths to Trp residues based on the <_a> values for the single Trp mutant peptides. It is known that the Trp fluorescence lifetime is sensitive to a number of factors and that it provides a "fingerprint" for a given Trp residue in a specific chemical environment (14, 44, 46, 52). There is a correlation between the solvent dielectric constant and the magnitude of the <_a>; as the dielectric constant decreases the fluorescence lifetime generally increases. However, there are several interacting factors within protein structure, and also membrane bilayer structure, which complicate the interpretation of changes in <_a> values. It is curious that no Trp residue for any of the mutant proteins exhibited a high (5-6 ns) <_a> value, which might have been expected for a Trp residue buried in the hydrocarbon core of the membrane bilayer. This may indicate that no Trp residue for the closed channel state of the WT or any of the single Trp mutant proteins is completely buried within the membrane bilayer or that the direct interpretation of the <_a> value is complicated by the heterogeneous solvent system consisting of numerous interfacial regions that is part of a protein-liposome complex.

Although at present it is difficult to correlate membrane location of Trp residues within proteins from the measured <_a> values, it is clear that changes invoked in the <_a> values for the various mutant peptides reflect changes in the chemical environment around the indole chromophore (52). Therefore, the data in Fig. 5 indicate that the Trp residues in the WT, W-355, W-404, W-413, W-431, W-460, and W-507 peptides are likely located in a more nonpolar environment, whereas W-367, W-424, W-443, W-484, and W-492 are found in a less polar environment upon the solution to membrane transition of the channel peptide. Small changes (decreases) in <_a> values were observed for W-495 and W-499 peptides, and no change was seen for W-478 channel peptide (Fig. 5). Although not definitive concerning the disposition of the Trp residues at or below the membrane bilayer surface, these data strongly suggest that there are significant structural rearrangements occurring within the channel peptide when it binds to the membrane surface.

Topological Models of Colicin E1 Channel-- Simplified models of the closed channel structure as well as the more speculative open channel state are shown in Fig. 6. The former model (Fig. 6A) incorporates previous data from a large number of studies involving a wide array of experimental techniques and approaches to the elucidation of membrane protein topology. Important components of the model for the closed state include a monomeric species that features the partial immersion of helices 1 and 2 into the bilayer and the flickering of the hydrophobic domain (helices 8 and 9) between states 1 and 2 as suggested by Kienker et al. (41) except that state 2 is not fully trans-membrane. Depth-dependent quenching data suggest that state 2 is the preferred orientation and is the most heavily populated state for the closed channel structure (29, 31). The imposition of a membrane potential induces the formation of a mature, open channel by the translocation of the gating peptide (42, 51, 53) comprised of helices 5-7 from the soluble structure. However, recent data employing engineered epitopes suggest that colicin Ia is capable of translocating a much larger segment of hydrophilic peptide than previously believed possible and that the COOH-terminal 85 residues (residues 542-626 of Ia) functions as a protein translocator (43). It now seems likely that the gating peptide is part of a larger protein segment that moves across the membrane bilayer in response to an imposed membrane potential (41, 43). How this "dragon's tail" arranges itself to form an ion-conducting pathway through the membrane bilayer is currently beyond comprehension. However, considering that each channel entity is composed of a single colicin polypeptide, it seems reasonable to suggest that the channel makes use of nearly all of the polypeptide sequence. This channel (helix bundle) structure undoubtedly must involve sufficient polypeptide length to form an ion channel with a lumen of the physical dimensions as shown previously (21) with a 7 Å pore (23). Notably, amid a blurred picture for the open state of the colicin channels, one issue is clear, that the open channel model for the colicins, including the E1 protein, requires more detailed experimental evidence on the specific topological arrangement of the protein in the membrane bilayer in the presence of a membrane potential. Presently, this model provides a working hypothesis that can be subjected to the rigors of membrane biochemical/biophysical experiments.

View larger version (56K): [in this window] [in a new window]   Fig. 6.   Models for the membrane-bound states of the colicin E1 channel peptide. Panel A, the closed state of the colicin E1 channel and (panel B) a four-helix bundle in the membrane in the presence of a membrane potential. The models are based on data from photolabeling (51); biotinylation (41, 42, 53); epitope mapping (43); saturation mutagenesis (55); protease accessibility (28, 47); depth-dependent fluorescence quenching (29, 31); cysteine-scanning mutagenesis (56); FRET (27); CD, Fourier transform infrared radiation, and differential scanning calorimetry (48, 49); time-resolved spin labeling (57); and acrylamide quenching and other fluorescence data presented in this work. The initial surface-bound structure adopts at least two states with state 2 being the most populated. The channel is fully assembled in the presence of a membrane potential, which involves insertion of helices 1 and 2 (TM1), the translocating peptide (TLP, helices 3-5b), and the amphipathic TM2 (helices 6 and 7), with the hydrophobic domain (helices 8 and 9) providing the anchor for the peptide to the membrane. This results in an unusual channel structure that must form a lumen with the dimensions suggested by previous experiments (21, 23). The predicted trans-membrane helices in the channel domain, formed upon membrane association, are designated as TM, and the circles indicate helical segments lying on the surface of the bilayer parallel to the plane of the membrane.

	ACKNOWLEDGEMENTS

We thank Dr. Irwin Tessman for providing the E. coli strain IT3661 and Jill Duncan for purifying the colicin E1 channel peptides.

	FOOTNOTES

^* This work was supported by the Medical Research Council of Canada (to A. R. M.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed. Tel.: 519-824-4120, ext. 3806; Fax: 519-766-1499; E-mail: merrill@chembio.uoguelph.ca.

² The notation W-### designates the Trp mutant colicin E1 channel peptide that possesses a single Trp residue at that designated site (###) according to the numbering of the whole colicin E1 protein sequence. The notation Trp-### refers to the specific Trp residue within the channel peptide sequence.

	ABBREVIATIONS

The abbreviations used are: LUV, large unilamellar vesicle; _i, normalized preexponential value; , wavelength; AEDANS, 5- (((acetyl)amino)ethyl)aminonaphthalene-1-sulfonic acid; k_q, average bimolecular quench constant; k_q, change in average bimolecular quench constant; _emmax, fluorescence emission maximum wavelength; _emmax, change in fluorescence emission maximum; , fluorescence lifetime; <_a>, amplitude average lifetime; <_a>, change in amplitude average fluorescence lifetime; <_f>, intensity average lifetime; <_i>, fluorescence lifetime species where _i = 1, 2, or 3 for a given fluorophore; F_i, fluorescence intensity in the presence of quencher; F₀, fluorescence intensity in the absence of quencher; FRET fluorescence resonance energy transfer, K_SV, Stern-Volmer quench constant; POPC, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine; POPG, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylglycerol; Q_F, fluorescence quantum yield; Q₀, Q_F of donor species in the absence of acceptor species; RQ_F, relative quantum yield; TLP, translocating peptide(s); TM, trans-membrane; Trp, Trp-deficient peptide; WT, wild type.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. White, A. A. Musse, J. Wang, E. London, and A. R. Merrill Toward Elucidating the Membrane Topology of Helix Two of the Colicin E1 Channel Domain J. Biol. Chem., October 27, 2006; 281(43): 32375 - 32384. [Abstract] [Full Text] [PDF]

				A. A. Musse, J. Wang, G. P. deLeon, G. A. Prentice, E. London, and A. R. Merrill Scanning the Membrane-bound Conformation of Helix 1 in the Colicin E1 Channel Domain by Site-directed Fluorescence Labeling J. Biol. Chem., January 13, 2006; 281(2): 885 - 895. [Abstract] [Full Text] [PDF]

				A. A. Musse and A. R. Merrill The Molecular Basis for the pH-activation Mechanism in the Channel-forming Bacterial Colicin E1 J. Biol. Chem., June 27, 2003; 278(27): 24491 - 24499. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Tory, M. C. Articles by Merrill, A. R. Search for Related Content PubMed PubMed Citation Articles by Tory, M. C. Articles by Merrill, A. R. Social Bookmarking                      What's this?