Tải bản đầy đủ - 0 (trang)


Tải bản đầy đủ - 0trang


E.K. Schmitt and C. Steinem



G / pS











Frequency / %
















Frequency / %


G / pS













G / pS

Figure 10. A) Single channel conductance values

G of gramicidin D in a DPhPC nano-BLM on a

porous alumina substrate with a mean pore diameter of 280 nm. Three different cations with a concentration of 500 mM were added to both sides of

the porous substrate. B) Histogram analysis of the

observed conductance states of the 2,2’bipyridine-functionalized peptidic ion channel in

nano-BLMs up to 1000 pS (3556 events, bin

width: 5 pS, normalized to all events (4524)) in

the absence of NiCl2. C. Histogram analysis of the

conductance states of the 2,2’-bipyridinefunctionalized peptidic ion channel in the presence of 2.5-5 μM NiCl2 up to a conductance of

1000 pS (2029 events, bin width: 5 pS, normalized to all events (2189)). The solid lines are the

results of fitting four Gaussian distributions (dotted lines) to the histograms.58

Ion Channels and Transporters in Pore-Suspending Membranes


of (11 ± 2) pS (Fig. 10 A). The observed conductivities for the different alkali cations are in good agreement with the selectivity series of gramicidin, which is known to be Cs+ > K+ > Li+.

(b) Alamethicin

Full functionality of a membrane system is only given if the

lateral mobility of the components is proven. Gramicidin can be

observed in a functional state even in a solid supported membrane

system, in which one leaflet is attached to the substrate as demonstrated by Cornell and coworkers.54,55 In contrast, one alamethicin

helix, a peptaibol from Trichoderma viridae, spans the entire lipid

bilayer and forms voltage gated ion channels with multiple conductance levels. This behaviour can be explained by the formation

of transmembrane aggregates according to the barrel stave mechanism.56 The uptake and release of alamethicin helical monomers

from a channel aggregate accounts for the observed multi-state

conductance levels.57 The formation of such helix bundles requires

the lateral movement of the individual helices. We analyzed

whether such helix bundles are formed, if alamethicin is reconstituted into nano-BLMs. The nano-BLMs were bathed in 500 mM

KCl and alamethicin was added only to the cis side from an ethanolic solution, resulting in a final concentration of about 100 nM,

while applying a holding potential of Vh = +70 mV. The voltagedependent activation of single alamethicin channels with up to five

conductance states was observed.23 The different conductance

states are defined by the number of monomers making up the pore

forming aggregate. Hence, from the results one can conclude that,

indeed, monomeric alamethicin helices can move laterally within

the nano-BLM to form conducting oligomers.

(c) 2,2’-Bipyridine-Functionalized Peptidic Ion Channels

Based on this knowledge, it was the idea to control the helix

bundle formation of amphipathic helices in nano-BLMs. Thus, we

synthesized an Į-helical amphipathic peptide with the sequence

H2N-(LeuSerSerLeuLeuSerLeu)3CONH2, to which a bipyridine

moiety with a spacer of 7 carbon atoms was attached to the Nterminus,58 in order to answer the question if it is possible to control the conductance behaviour of the peptide by complexation of


E.K. Schmitt and C. Steinem

the bipyridine residues with Ni2+. The peptide sequence itself59 had

already been shown to insert into lipid bilayers as an amphipathic

Į-helix, forming cation selective ion channels with an effective

diameter of 0.8 nm under voltage control.59-61 From the effective

diameter and computer simulations it was concluded that a hexameric bundle is formed in lipid bilayers.60,62 The synthetic peptide

was inserted into nano-BLMs from an ethanolic solution and its

channel properties were characterized by single channel recordings

under voltage clamp conditions in the absence and presence of

NiCl2. Characteristic rectangular current traces were observed either in the absence or in the presence of Ni2+.58 We attribute this

step-like current increase and decrease to the assembly and disassembly of a helix bundle composed of a defined number of amphipathic peptide helices. Altogether, four different opening levels

were discernable in the current range of around 25 pA (O1), 35 pA

(O2), 45 pA (O3) and 70 pA (O4). Channel events of one opening

level were often observed in bursts, which lasted several seconds.

Event histograms were used to determine the conductance states

(Fig. 10 B), resulting in four Gaussian distributions with G1 = (131

± 19) pS, G2 = (181 ± 20) pS, G3 = (234 ± 20) pS, and G4 = (374 ±

70) pS. Twenty one percent of all events did not fall into these

conductance levels but were larger than 600 pS. Taking the fact

that the lowest conductance state is attributed to six peptide monomers forming a helix bundle, as proposed by DeGrado and

coworkers,59,60 the higher conductance states would be a result of a

continuous increase in the number of helices participating in the

bundle, up to nine helices.

In the presence of Ni2+, the analysis of the conductance states

indeed revealed a considerable change. NiCl2 was added in a concentration of 2.5-5 μM to both sides of the membrane and voltage

clamp experiments were again performed. The majority of events

were found in a current range of 15-110 pA. A histogram analysis

demonstrated that the conductance states are in the same range as

those observed in the absence of Ni2+ (Fig. 10C). However, their

distribution is significantly different. Four Gaussian functions

were fitted to the data with the assumption that the mean conductance states do not change in the presence of Ni2+, thus keeping these four parameters constant. It is obvious that the first and second

conductance levels are very prominent, while the third and fourth

ones are greatly diminished. The relative area of the first and se-

Ion Channels and Transporters in Pore-Suspending Membranes


cond conductance states equals 78 % in the presence, and only 43

% in the absence, of NiCl2. Only 7 % of all events were found with

a conductance level above 600 pS. From these findings, it can be

concluded that the complexation of bipyridine by Ni2+ indeed results in a considerable confinement of the observed multiple conductance states. Assuming that the lowest conductance state is a

result of the formation of a six helix bundle, helix bundles with six

and seven helices are preferentially formed in the presence of Ni2+,

while higher-order aggregates become less likely.


Protein Channels

Relatively large transmembrane proteins such as Į-hemolysin63

and the glutamate receptor19 have been reconstituted into membranes on orifices in either polycarbonates or silicon, with diameters greater than 1 μm. We raised the question of how the pore size

of the porous material limits the incorporation and the functionality of complex integral membrane proteins in nano-BLMs. To address this question, the functional insertion of the outer membrane

protein F (OmpF) from E. coli and of the eukaryotic protein connexon 26 (Cx26) was investigated in nano-BLMs on porous alumina with pore diameters of 60 nm by observing the ion channel

activity in voltage clamp experiments, as detailed below.

(i) Outer Membrane Protein F

The porin OmpF is a well characterized protein in terms of

structure64,65 and channel activity. It is composed of 16 antiparallel

aligned E-sheets (E-barrel), connected by amino acid sequences

building up a water-filled pore. Three of these monomeric units

with a molecular weight of 37.1 kDa and a length of 5 nm are arranged around a three-fold molecular axis. One protein covers an

area of roughly 80 nm2. In a nano-BLM, it is surrounded by just a

few thousand lipids. Ion selectivity and conductivity of the channel

is assumed to be a result of the constriction zone66 that is formed

by loop 3, which folds into the barrel at approximately half the

height of the channel. By voltage clamp experiments, we were able

to prove that the protein is fully functional in nano-BLMs. The insertion of the porin results in a three-step increase and decrease in

current due to the stepwise opening and closing of each subunit


E.K. Schmitt and C. Steinem

pore.26 Point histogram analysis of the current traces allowed for

the determination of the three different conductance levels with G1

= (1700 r 80) pS, G2 = (3360 r 80) pS and G3 = (5060 r 50) pS. In

addition, we investigated whether we are also able to monitor the

reported fast kinetics and subconductance states. Also, these characteristics of the porin pore were found in nano-BLMs, indicating

that the close proximity of immobilized lipid bilayers on the pore

rims does not influence the ion channel activity.26 Nestorovich et

al.67 reported that the E-lactam antibiotic ampicillin is capable of

blocking the OmpF channel by moving in and out of the channel.

This process takes place on a rather fast time scale. It turned out

that the time resolution of our system is not sufficient to fully resolve every single ampicillin blockade as was demonstrated in the

work of Nestorovich et al.67 However, blockades of ampicillin

were clearly detected as downward spikes of the ion flow during

the opening of one monomer.26 These results indicate that the porous substrate underneath the membrane does not prevent the full

functionality of a large transmembrane protein, even though one

membrane, spanning a 60 nm pore of the porous alumina substrate,

is composed of only roughly 4000 lipids.

(ii) Connexon 26

Connexins are members of a multigene family of membranespanning proteins that form intercellular channels, which are composed of two hexameric hemichannels, called connexons (Cx).

These intercellular channels organize into gap junctional plaques

and span the extracellular space/matrix of adjacent cells, thus allowing a passive exchange of small molecules up to about 1 kDa.68

By the formation of gap junctions, two membranes can be electrically connected. As a prerequisite for the electrical coupling, connexons need to be reconstituted into a membrane system that is

accessible from both sides. Nano-BLMs can provide the required

accessibility and, at the same time, the stability to connect a membrane such as the plasma membrane of a cell via the formation of

gap junctions. To elucidate the functionality of a connexon in

nano-BLMs, connexon 26 (Cx26) isolated from baculovirusinfected Sf9 cells was reconstituted. This was achieved by adding

Cx26 dissolved in a 1 % (v/v) octyl-polyoxyethylene (o-POE) detergent solution to the cis side of the Teflon cell, leading to a final

Ion Channels and Transporters in Pore-Suspending Membranes


concentration of 0.7-2.7 ng mL-1. Since the insertion of a Cx26 oligomer is a random process, activity is observed sometimes already after several minutes, but in other cases only after hours. By

means of voltage clamp experiments, information about the conductance states of Cx26 can then be gathered. If a hemichannel

was inserted, current traces were recorded and different holding

potentials ranging between Vh = +150 mV and -150 mV were applied. It turned out that the hemichannels exhibit a pure Ohmic behaviour with a mean conductance of (33 ± 3) pS (Fig. 11 A).

This result is in good agreement with conductance values obtained from the same hemichannel inserted into planar bilayers on

microstructured glass supports, a method provided by the company

Nanion (Munich, Germany).69 A statistical analysis of all single

channel events reveals one prominent conductance state with a

mean conductance of G1 = (34 ± 8) pS. In addition, but less frequently, larger conductance states were observed with two distributions at G2 = (70 ± 8) pS and G3 = (165 ± 19) pS. Other groups

found similar single channel conductance values for reconstituted

hemichannels, being in the range of 35-316 pS.70,71 This result indicates that even Cx26 hemichannels can be inserted into nanoBLMs in a functional manner.



Nano-BLMs with membrane resistances in the range of gigaohms

prove to be a suitable model system for the investigation of peptides and large transmembrane proteins on a single channel level,

as outlined in Section III. In contrast, pore-spanning lipids bilayers

generated by vesicle fusion on porous alumina exhibit resistances

two to three orders of magnitude lower than those of nano-BLMs.

Still, ion channel activity can be measured by using EIS. This

technique allows detecting conducting channel proteins in porespanning bilayers in an integral manner and, thus, overcomes the

need for gigaohm seals. Here, we present the monitoring of ion

channel activity, namely of OmpF and gramicidin, by means of



E.K. Schmitt and C. Steinem


I / pA 6



-150 -100






Vh / mV





Frequency / %








G / pS



Figure 11. A) Cx26 current-voltage relationship of averaged current steps ranging

between 0-63 pS (2169 events). The hemichannel shows a linear I-Vh-dependence

with a main conductance of (33 ± 3) pS. B) Event-histogram of the observed conductance states of Cx26 in nano-BLMs (2691 events, bin width: 2 pS). The solid

line is the results of fitting three Gaussian distributions to the histogram. Three

main conductance states are assigned to G1 = (34 ± 8) pS, G2 = (70 ± 8) pS, and G3

= (165 ± 19) pS.69

Ion Channels and Transporters in Pore-Suspending Membranes



Reconstitution of OmpF

In Section II, the electrochemical characteristics of pore-spanning

membranes investigated by EIS were discussed in detail. The data

reveal that the absolute value of the impedance |Z| at frequencies

< 101 Hz is solely governed by the resistive properties of the bilayer and is thus frequency independent. Thereby, performing EIS

experiments at a fixed frequency of 10 Hz or less allows following

the membrane resistance Rm without taking the capacitive properties into account. Membrane-integrated, active ion channels produce a change in membrane conductivity, which is readily detected

as a change in the absolute value of the impedance |Z| at the chosen

frequency as a function of time. Prior to this time resolved experiment, impedance spectra need to be recorded to determine the frequency range in which the membrane resistance Rm is well defined.

As mentioned in Section III, reconstitution of transmembrane

proteins from an aqueous solution requires that the molecules are

solubilised in and added from a detergent solution. Detergents are

amphiphiles and known to interact with lipid bilayers. Studies on

liposomes revealed that detergents at concentrations considerably

below their critical micellar concentration (CMC) partition into the

bilayers and permeabilise them.72,73 Hence, prior to protein addition, we first investigated the influence of the detergent o-POE on

Rm, which is widely used to solubilise membrane proteins. Figure

12 represents |Z|(1 Hz) of a pore-spanning bilayer as a function of

time monitored in 0.1 M NaCl. It is obvious that a final concentration of 0.0002% (v/v) o-POE, which is four orders of magnitude

below its CMC, is sufficient to permeabilise the bilayer such that

|Z|(1 Hz) decreases by two orders of magnitude. However, by exchanging for detergent-free electrolyte solution, the original membrane resistance is recovered. Thus, the interaction of o-POE in

sub-CMC concentrations with pore-spanning membranes is fully

reversible. Then, OmpF in o-POE (0.0002% (v/v) o-POE, 0.5 nM

OmpF) was added to the pore-spanning lipid bilayers, again resulting in a significant decrease in membrane resistance. Yet, rinsing

with electrolyte solution resulted in only partial recovery of the

membrane resistance. The net drop in bilayer resistance Rm is a

result of the formation of OmpF pores in the pore-suspending

|Z|(1 Hz) / W


E.K. Schmitt and C. Steinem

OmpF in o-POE


















Figure 12. Time-resolved change of the magnitude of the impedance |Z| at a fixed

frequency of 1 Hz of a pore-suspending membrane after the addition of o-POE with

a final concentration of 0.0002 % (v/v) and rinsing with buffer, followed by the

addition of OmpF dissolved in o-POE to the cis side, leading to a final protein concentration of 0.5 nM in 0.0002 % (v/v) o-POE. Electrolyte: 0.1 M NaCl.35

membrane, demonstrating that the proteins are functionally inserted.

Based on this result, we raised the question of how many

OmpF molecules contribute to the decrease in membrane resistance. In order to quantify this, the change in conductivity ǻG is

calculated from the bilayer resistance before (R0) and after (Rm) the

insertion of OmpF. According to Eq. (5):




Rm R0

Gm  G0


ǻG equals 19.7 μS. Taking the conductivity of 535 pS of a

single OmpF trimer in 0.1 M NaCl into account,18 approximately

106 proteins cause the monitored decrease in bilayer resistance.

This calculation demonstrates that a significant number of mole-

Ion Channels and Transporters in Pore-Suspending Membranes


cules need to be incorporated in order to detect active ion channels

in pore-spanning membranes.

These facts raise the question, why only one or a few protein

molecules insert into nano-BLMs, whereas 106 are incorporated

into pore-suspending lipid bilayers. Both the total membrane area

and the final protein concentration in the measuring chamber cannot explain the observed difference. On the one hand, the total

membrane area of pore-spanning membranes is approximately

twofold smaller, which in theory decreases the probability of an

insertion event. On the other hand, even though the concentration

of OmpF added to pore-spanning membranes is at most 300-fold

higher, it cannot explain the incorporation of 106 times more molecules compared to nano-BLMs. Thus, another parameter seems to

limit the insertion of protein molecules and detergents into nanoBLMs. Presumably, detergent molecules and proteins dissolve in

the residues of the organic solvent, which is required for preparing

nano-BLMs and is known to remain in the membrane.29 Thus, the

molecules might lose their functionality in these membrane areas.

In contrast, pore-suspending membranes are nominally free of organic solvent and all incorporated molecules are fully functional.

The comparison of the insertion of OmpF in nano-BLMs versus pore-spanning membranes reveals that these two membrane

systems address different experimental needs. While porespanning membranes cannot be used for single channel recordings,

they are suited for monitoring membrane resistances by EIS and

prove to be highly sensitive to the incorporation of vast numbers of

membrane soluble molecules. Indeed, the use of this high insertion

efficiency could be envisioned to be applied for monitoring lipid

bilayer interactions of membrane modulators, which cause permeabilisation. Furthermore, these insertion rates in pore-spanning bilayers hold the advantage of introducing membrane proteins in

such numbers that allow for integral measurements of their channel activity. However, the interactions of detergent molecules that

are always present when working with solubilised membrane proteins are considerable, and their impact on the electric properties of

the bilayers is significantly disrupting. An alternative route to reconstitute ion channels in pore-spanning membranes prevents the

use of detergents and involves the use of proteoliposomes as described in detail in the following paragraph.


E.K. Schmitt and C. Steinem


Analysis of Gramicidin D Activity

We aimed to reconstitute ion channels in pore-spanning membranes by fusion of proteoliposomes on functionalized porous substrates. Gramicidin D turned out to be a good test case for the development of the principle strategy.

(i) Channel Activity of Gramicidin D Reconstituted into

Pore-Spanning Membranes

Large unilamellar vesicles composed of DOPC doped with 2

mol% gramicidin D were prepared and the conformation of gramicidin D was analyzed by circular dichroism (CD) spectroscopy,

verifying that it forms the conducting conformation.29 To generate

pore-suspending membranes with reconstituted active gramicidin

channels, we followed the strategy of first generating a poresuspending bilayer composed of DPhPC/DOPC (6:4) from vesicle

spreading before adding gramicidin D-doped DOPC vesicles to the

system for 8–12 hours. Porous alumina substrates with partially

opened pore bottoms were used, as the lipid membranes form considerably faster on these substrates (see Section II). If the alumina

oxide resistance Rox is known, model (2’) can be applied for analysis. For large Rox, the overall resistance Ra resembles Rm,o (Eq. 3).

This allows to directly link the change in overall resistance to a

change in Rm,o generated by a conducting ion channel. Throughout

the preparation steps, a buffer solution containing TMA was used.

TMA is too large to diffuse through the gramicidin channel74 and,

thus, the membrane resistance is supposed to remain unaffected

upon insertion of the channel active peptide. To verify the functional transfer of gramicidin D into the pore-spanning membranes,

its channel activity was monitored by adding monovalent cations

(LiCl, NaCl, and KCl) to the TMA-containing buffer solution into

the upper cis compartment of the liquid cell (Fig. 1 B). The distinct

selectivity for alkali cations and its concentration dependence are

the main characteristics of the gramicidin channel. Impedance

spectra were recorded before and after increasing the ion concentration to follow the electrical response of the membrane system.

The entire experiment with various cations was performed on one

membrane preparation, as the electrical properties of porespanning membranes depend on the underlying alumina substrate,

Tài liệu bạn tìm kiếm đã sẵn sàng tải về


Tải bản đầy đủ ngay(0 tr)