The effects of different anionic polymers on the kinetic
properties of ionic channels formed by neutral gramicidin A (gA) and
its positively charged analogs
gramicidin-tris(2-aminoethyl)amine (gram-TAEA) and
gramicidin-ethylenediamine (gram-EDA) in a bilayer lipid membrane were
studied using a method of sensitized photoinactivation. The addition of
Konig's polyanion caused substantial deceleration of the
photoinactivation kinetics of gram-TAEA channels, which expose three
positive charges to the aqueous phase at both sides of the membrane. In
contrast, channels formed of gram-EDA, which exposes one positive
charge, and neutral gA channels were insensitive to Konig's polyanion.
The effect strongly depended on the nature of the polyanion added,
namely: DNA, RNA, polyacrylic acid, and polyglutamic acid were
inactive, whereas modified polyacrylic acid induced deceleration of the
channel kinetics at high concentrations. In addition, DNA was able to
prevent the action of Konig's polyanion. In single-channel
experiments, the addition of Konig's polyanion resulted in the
appearance of long-lived gram-TAEA channels. The deceleration of the
gram-TAEA channel kinetics was ascribed to electrostatic interaction of
the polyanion with gram-TAEA that reduces the mobility of gram-TAEA
monomers and dimers in the membrane via clustering of channels.
 |
INTRODUCTION |
Studies of how polyelectrolytes interact with
membranes are important for understanding and improving the process of
gene delivery. Since the demonstration that cationic liposomes can enhance transfection (Felgner et al., 1987
), substantial efforts have
gone into characterizing interaction between DNA and cationic membranes
(Kinnunen et al., 1993; Koiv et al., 1994; Mitrakos and Macdonald,
1996, 1998; Hirsch-Lerner and Barenholz, 1998; Meidan et al., 2000;
Subramanian et al., 2000; Safinya, 2001).
Membrane ion channels can act as exquisitely sensitive sensor elements
that report on polyelectrolyte/membrane interactions. There is a
substantial body of evidence demonstrating modulation of the activity
of certain pore-forming proteins by polyanions, e.g., by DNA sequences
(Wright and Harding, 2000; Tosteson et al., 2001), and by Konig's
polyanion (Konig et al., 1982; Colombini et al., 1987; Benz et al.,
1988; Tedeschi et al., 1987; Tedeschi and Kinnally, 1987; Mannella and
Guo, 1990; Mirzabekov et al., 1993). Transport of polyelectrolytes, in
particular of nucleic acid fragments, through proteinaceous pores in
membranes has also been reported (Kasianowicz et al., 1996; Szabo et
al., 1997; Hanss et al., 1998; Akeson et al., 1999; Meller et al.,
2000, 2001; Henrickson et al., 2000; Vercoutere et al., 2001).
The gramicidin channel is a structurally well-defined and functionally
well-characterized system (Woolley and Wallace, 1992; Busath, 1993;
Koeppe and Andersen, 1996; Andersen et al., 1999) that can be
engineered to act as a sensor for a variety of purposes, e.g., pH
measurement (Borisenko et al., 2002) and avidin/streptavidin detection
(Cornell et al., 1997, 1999; Suarez et al., 1998; Rokitskaya et al.,
2000a; Futaki et al., 2001). Previously, we used
O-pyromellitylgramicidin (OPg), a gramicidin derivative with three
negative charges at the C-terminal end, to investigate the interaction
of polycations with membranes (Krylov et al., 1998, 2000).
Because the interaction of polyanions with membranes is expected to be
influenced by the concentration and charge density of cationic species
in the membrane (Safinya, 2001), we prepared gramicidin analogs bearing
one (gramicidin-EDA) and three (gramicidin-TAEA) positive charges and
studied the effect of various synthetic and natural polyanions on the
behavior of the channels formed by the gramicidin analogs in planar
bilayer lipid membranes using the methods of sensitized
photoinactivation and single-channel recording.
| |
MATERIALS AND METHODS |
Peptide synthesis
Gramicidin-ethylenediamine (gram-EDA, Fig.
1) was prepared as described previously
(Jaikaran et al., 1995). Gramicidin-tris(2-aminoethyl) amine
(gram-TAEA, Fig. 1) was synthesized in a similar fashion. Commercial
gramicidin D (20 mmol) was esterified (1 h, 4°C) with p-nitrophenyl chloroformate (200 mmol) in dry
tetrahydrofuran (2 ml) containing triethylamine (TEA) (100 ml). Then
tris(2-aminoethyl) amine in a 150-fold molar excess was
added, and the reaction mixture was stirred for 20 min at room
temperature. The solution then was filtered, then dried on a
rota-evaporator. The product was then separated by gel-filtration using
LH-20 in methanol and by ion exchange chromatography with an AG-MP 50 column (Weiss and Koeppe, 1985). Finally, a thin-layer
chromatography separation (chloroform/methanol/water, 65:25:4)
was applied to give gram-TAEA (0 < Rf < 0.29), which was characterized
by ultraviolet and mass spectrometry. MS (maldi positive): the
major peak was 2054, corresponding to gramA-TAEA. Other peaks,
corresponding to gramB-TAEA and gramC-TAEA, were of significantly lower
intensity.
View larger version (14K):
[in this window]
[in a new window]
|
FIGURE 1
Chemical structure of (A)
gramicidin-ethylenediamine (gram-EDA), (B)
gramicidin-tris(2-aminoethyl)amine (gram-TAEA),
(C) a repeat unit of Konig's polyanion, and
(D) a repeat unit of modified polyacrylic acid.
|
|
Polyanions synthesis
Konig's polyanion (the copolymer of styrene, maleic acid, and
methacrylic acid, Fig. 1) was synthesized according to Konig et al.
(1977). The monomeric composition of the final product was estimated
using infrared and ultraviolet spectroscopy, and the molecular weight
of the polymer was evaluated using gel-permeation chromatography in
ethanol with Waters polyethylene oxide-markers as molecular
weight standards.
The copolymer of acrylic acid with stearylacrylate (the modified
polyacrylic acid, Fig. 1) was synthesized by esterification of
polyacrylic acid by octadecyl alcohol in dioxane in the presence of
dicyclohexylcarbodiimide. Polyacrylic acid, 100 mg, (MW 6700; 1.39 mmol
of acrylic units) were dissolved in 5 ml dry dioxane at 80°C and 300 mg of octadecyl alcohol (1.11 mmol) (Sigma, St. Louis, MO) and 70 mg
dicyclohexyl carbodiimide (0.35 mmol) (Aldrich, Milwaukee, WI) were
added. The mixture was incubated at 60°C for 5 days. The solution was
cooled and precipitated octadecyl alcohol was separated by filtration.
The filtrate was collected and mixed with 4 ml 0.1 M KOH solution in
methanol to obtain potassium polyacrylate, which is insoluble in
organic solvents. The mixture was then mixed with hexane, and
polyacrylate was allowed to precipitate overnight at room temperature.
The polymer was collected by centrifugation, washed with hexane, and
dried in vacuo. Then the polymer was dissolved in 0.1 M borate buffer,
pH 8.5, and the solution was centrifuged to separate the polymer from
admixtures of water-insoluble octadecyl alcohol. The solution was then
acidified with HCl to pH 2.5-3.0 and dialyzed extensively against
distilled water to obtain the acidic form of the polymer, which is
practically insoluble in water. The final product was lyophilized and
the amount of the attached ester groups was estimated using alkaline
titration. For this purpose, 10 mg of the polymer was dissolved in 2 ml
0.09 M NaOH and the content of carboxylic groups was estimated using a
pH-stat (Radiometer, Copenhagen, Denmark) titration. The apparent degree of modification was ~18%, i.e., each polymer chain contained ~16 octadecyl radicals.
Single-channel measurements
Peptides (~10 nM in methanol) were added to membranes formed
from diphytanoyl-phospatidylcholine/decane (50 mg/ml). Lipid bilayers
were formed across an ~100-µm hole in a polypropylene pipet tip by
painting a solution of lipid in decane. The pipet tip was mounted in a
Teflon cell through a small hole in the back face. The front face of
this cell had a removable circular glass window. Silver/silver chloride
electrodes were placed in the pipet tip and a cylindrical well drilled
from the top of the cell. Symmetrical buffered (0.5 mM BES, pH 7) KCl
(0.1 M) solutions with and without Konig's polyanion were used. All
measurements were made at room temperature.
The current through lipid bilayers containing the gramicidin derivative
was measured, and the voltage was set using an Axopatch 1D patch-clamp
amplifier (Axon Instruments, Union City, CA) controlled by
Synapse (Synergistic Research Systems) software. Single-channel events were recorded for a period of several hours for each set of
experimental conditions. Data were filtered at 100 Hz, sampled at 1 kHz, stored directly to disk, and analyzed using Synapse and Igor
software (Wavemetrics, Inc., Lake Oswego, OR). The mean lifetimes and current amplitudes were determined by fitting appropriate functions to corresponding histograms using the program Mac-Tac (Version 2.0, Instrutech Corp., Port Washington, N.Y.).
Photoinactivation measurements
Bilayer lipid membranes (BLMs) were formed from a 2% solution
of diphytanoylphosphatidylcholine (Avanti Polar Lipids, Alabaster, AL)
by the brush technique (Mueller et al., 1963) on a 0.8-mm diameter hole
in a Teflon partition separating two compartments of a cell containing
aqueous solutions of 50 mM (unless otherwise stated) KCl (Fluka, Buchs,
Switzerland), 10 mM MES (Sigma), 10 mM TRIS (Sigma) and 0.1 mM EDTA at
pH 7.0. Gram-TAEA, gram-EDA, or gramicidin A (Sigma) (Cifu et al.,
1992) was added from stock solutions in ethanol to the bathing
solutions at both sides of the BLM and routinely incubated for 15 min
with constant stirring. Polyanions were added to both compartments of
the cell unless otherwise stated. Polyacrylic acids (MW 5000 and
250,000) were from Aldrich, poly-L-glutamic acid (MW
15,000) from Serva (Heidelberg, Germany). Calf thymus DNA (single- and
double-stranded) and DNA from salmon testes were from Sigma. RNA from
tobacco mosaic virus was a generous gift of Prof. D. B. Zorov,
Moscow State University. Poly-L-lysine hydrobromide (MW
7500) was from Sigma. Experiments were carried out at room temperature
(20-22°C). Aluminum trisulfophthalocyanine (AlPcS3) was from Porphyrin Products
(Logan, UT). AlPcS3 was added to the bathing
solution at the trans-side (the cis-side is the front side with respect to the flash lamp). The electric current (I) was recorded under voltage-clamp conditions. The
currents were measured by means of a U5-11 amplifier (Moscow,
Russia), digitized by using a LabPC 1200 (National Instruments, Austin, TX) and analyzed using a personal computer with the help of WinWCP Strathclyde Electrophysiology Software designed by J. Dempster (University of Strathclyde, UK). Ag-AgCl electrodes were placed directly into the cell and a voltage of 30 mV (unless otherwise stated)
was applied to the BLM. The value of the current was usually ~1 µA,
which corresponded to 3 × 106 conducting
channels in the bilayer. BLMs were illuminated by single flashes
produced by a xenon lamp with flash energy of ~400 mJ/cm2 and flash duration < 2 ms. A glass
filter cutting off light with wavelengths < 500 nm was placed in
front of the flash lamp. To avoid electrical artifacts, the electrodes
were covered by black plastic tubes.
In the presence of a photosensitizer (e.g., aluminum phthalocyanine or
Rose Bengal, the dyes that sensitize singlet oxygen formation with high
quantum yield), irradiation of BLM with visible light is known to
decrease the gramicidin-mediated transmembrane current, I
(Strassle and Stark, 1992; Rokitskaya et al., 1993). The decrease in
the current is believed to result from damage to tryptophan residues of
gramicidin (tryptophan oxidation or peptide fragmentation) (Strassle
and Stark, 1992; Kunz et al., 1995) caused by reactive oxygen species
that are generated upon excitation of a photosensitizer (Rokitskaya et
al., 1996, 2000b). If BLM is illuminated with a single flash of visible
light, I is a monoexponential function of time (Rokitskaya
et al., 1996): I(t) = (I0 
I
)exp(t/
) + I, where
I0,
I, and are the initial current
before illumination, the stationary level of the current established as
a result of relaxation after the flash, and the characteristic time of
photoinactivation, respectively. The relative amplitude of
photoinactivation, 
, is defined as = (I0 I)/I0.
Because the light-induced decrease in the gramicidin-mediated current
is due to the reduction of the number of open channels (Rokitskaya et
al., 1993; Kunz et al., 1995), is equal to the damaged part of
gramicidin channels. It should be noted that, in the presence of
polyanions, the flash-induced decrease in the current was recorded
10-15 min after the addition of polyanions, when a new steady-state
level of the current was reached.
| |
RESULTS |
To study channel kinetics of gram-TAEA, one of the positively
charged analogs of gramicidin A, we applied a method of sensitized photoinactivation previously developed by Rokitskaya et al. (1996). It
is shown in Fig. 2 that a visible light
flash induced an irreversible decrease in the gram-TAEA-mediated
current across BLM, if a photosensitizer (aluminum
trisulfophthalocyanine) was added to the bathing solution. The time
course of the current decrease (below called the kinetics of
photoinactivation) was fitted well by a monoexponential curve similarly
to the data obtained earlier for gramicidin A, with the characteristic
time of photoinactivation (, the exponential factor of the curve)
being close to that of gA. The addition of Konig's polyanion at one
side of the BLM slightly increased the characteristic time of
photoinactivation (from 0.60 to 0.67 s), and the result was
independent of the side of the polyanion addition with respect to the
photosensitizer side (trans). In contrast, the addition of
the polyanion at both sides of the BLM produced a marked increase in
up to 8.5 s (Fig. 2, curve 3). Figure 2 illustrates
the data of typical measurements performed with a single membrane. In
the presence of Konig's polyanion (curve 3) the kinetics of
the current decrease displayed pronounced deviations from a
monoexponential curve not seen in the control kinetics (curve
1). Subsequent addition of 0.1 mM
poly-L-lysine to the bathing solutions completely
reversed the polyanion effect on ( = 0.61 s, data not
shown). In contrast to the photoinactivation kinetics of gram-TAEA
possessing three positive charges, that of neutral gA was completely
insensitive to additions of Konig's polyanion (data not shown).
View larger version (14K):
[in this window]
[in a new window]
|
FIGURE 2
The kinetics of the decrease in the gram-TAEA-mediated
current (I) through BLM in the presence of 1 µM
AlPcS3 (at the trans side) after a flash of
visible light (at zero time) in the control (curves 1 in
panels A and B), and in the presence of
0.06 µg/ml Konig's polyanion at the cis side of BLM
(curve 2 of panel A), and at both sides
of BLM (curves 3 of panels A and
B). The experimental data points were plotted as
(I/I0) 100% in panel
A and replotted as (I I)100%/(I0 I) versus the time (logarithmic scale)
in panel B, where I0 is the
initial current and I is the stationary
value of the current after a flash. Data were fitted with single
exponentials (solid curves) with characteristic times
= 0.60 s (curve 1), = 0.67 s
(curve 2), and = 8.5 s (curve
3). In the case of curve 3, double-exponential
fitting I = I0 + 1exp(t/1) + 2exp(t/2) (panel
B, dashed curve) was also calculated with the
following parameters: 1 = 70%,
1 = 10.8 s, 2 = 30%,
2 = 1.5 s. The initial value of the current
(I0) was ~1 µA.
|
|
It is seen from Fig. 2 A that Konig's polyanion not only
slowed down the kinetics of gram-TAEA photoinactivation, but also decreased its relative amplitude. The reduction of the
photoinactivation amplitude can be accounted for by quenching of
reactive oxygen species by the polyanion. As has been shown by
Rokitskaya et al.(1996), the characteristic time and the amplitude of
photoinactivation represent two independent parameters; for instance,
variation of the photosensitizer concentration alters markedly the
amplitude of photoinactivation, but does not change its characteristic
time. The present study deals with the effect of polyanions on the
characteristic time of photoinactivation, .
As described in the Appendix, the dependence of on the
concentration of the channel former, plotted as 1/ versus square
root of the BLM conductance (
), can be used
to calculate the rate constants of the channel formation
(KR) and termination
(KD). Figure
3 presents this dependence for gram-TAEA.
Using the data of Fig. 3 and Eq.10 of the Appendix, we obtain
KR = (3.6 ± 1.5) * 1012
mol
1s1cm2
and KD = (0.9 ± 0.1)
s1 (T = 23°C).
View larger version (11K):
[in this window]
[in a new window]
|
FIGURE 3
(A) The dependence of the characteristic
time of gram-TAEA photoinactivation () measured in the absence of
Konig's polyanion on the BLM current at 50 mV. The current was varied
by increasing the concentration of Gram-TAEA in the bathing solution.
(B) The dependence of 1/ versus square root of the
BLM conductance, which was used in calculation of the kinetic constants
of Gram-TAEA. The buffer solution contained 100 mM KCl.
|
|
The value of the gram-TAEA single-channel conductance (
) needed for
the estimations of the rate constants was taken from single-channel
recordings of gram-TAEA shown in Fig. 4.
The predominant value of the conductance () was determined to be 10 pS (0.1 M KCl, DPhPC) and the channel lifetime was 1.0 s. The
single-channel current-amplitude histogram indicated the existence at
least of two open sublevels, as was observed for other gramicidin
analogs with modified C-terminus (Woolley et al., 1995). The value of the single-channel lifetime of gram-TAEA is close to the
(1/KD) value calculated from the
photoinactivation data, thus demonstrating the compatibility of the two
methods. The addition of 100 ng/ml Konig's polyanion did not
significantly alter the average single-channel conductance as seen from
the corresponding amplitude histogram (Fig.
5). However, this addition led to the
appearance of a subset of channel events of much longer duration (see
Fig. 5 A, recording 2). These events manifested
themselves in the time histogram as a small shoulder at long time
durations (Fig. 5 C). These channel events also exhibited
fast flickering (Fig. 5 A).
View larger version (13K):
[in this window]
[in a new window]
|
FIGURE 4
(A) Current trace, (B)
single-channel current amplitude histogram, and (C)
duration histogram of gram-TAEA. The currents were obtained at +100 mV.
Several hundred single channels were characterized.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
FIGURE 5
(A) Current traces, (B)
single-channel current amplitude histogram, and (C)
duration histogram of gram-TAEA in the presence of 100 ng/ml of
Konig's polyanion. The two rows of current traces show two
representative parts of the single-channel recording. Several hundred
single channels were characterized.
|
|
The effect of Konig's polyanion depended on the number of positive
charges on the gramicidin molecules. In fact, the addition of 0.06 µg/ml Konig's polyanion at both sides of the membrane produced only
small effects on for gram-EDA having a single positive charge at pH
7 (Fig. 6). Even at high polyanion
concentrations (3 µg/ml, see of Fig. 6, curve 3), increased only twofold (from 0.49 to 0.85 s).
View larger version (17K):
[in this window]
[in a new window]
|
FIGURE 6
The kinetics of the decrease in the
gramicidin-EDA-mediated current (I) through BLM in the
presence of 1 µM AlPcS3 after a flash of visible light
(at zero time) in the control (curve 1), and in the
presence of 0.06 µg/ml Konig's polyanion (curve 2),
and 3 µg/ml Konig's polyanion (curve 3) at both sides
of BLM. The experimental data points were plotted as
(I/I0) 100% in panel
A and replotted as (I I)100%/(I0 I) versus the time (logarithmic scale)
in panel B where I0 is the
initial current and I is the stationary
value of the current after a flash. Data were fitted with single
exponentials with characteristic times = 0.49 s
(curve 1), = 0.58 s (curve
2) and = 0.85 s (curve 3). The
initial value of the current (I0) was
~1 µA.
|
|
In Fig. 7 the characteristic time of
gram-TAEA photoinactivation (curve 1) calculated from the
single-exponential approximation of the kinetics is plotted versus the
concentration of Konig's polyanion. It is seen that the dependence of
on the concentration includes three regions. At low concentrations,
the polyanion did not produce any effect on . Upon exceeding a
certain (threshold) concentration, began to grow progressively and
finally reached the maximum value. Further raising of the polyanion
concentration led to a gradual decrease in . By contrast, only a
slight increase in of gram-EDA was detected upon increasing the
concentration of Konig's polyanion (Fig. 7, curve 2).
View larger version (19K):
[in this window]
[in a new window]
|
FIGURE 7
The dependences of the characteristic time () of
gram-TAEA (curve 1) and gram-EDA (curve
2) photoinactivation measured in the presence of Konig's
polyanion on its concentration. The polyanion was added at both sides
of the BLM. The initial value of the current for each measurement of
was ~1 µA. The values of were obtained from single
exponential approximation.
|
|
It is reasonable to expect that the effect of the polyanion on is
dependent on the ionic strength of the bathing solution. Figure
8 shows the dependence of on the
concentration of potassium chloride in the bathing solution at a
constant concentration of the polyanion (note that, except for
single-channel experiments and the data of Fig. 3, the results
presented in all the other figures are obtained at 50 mM KCl). It is
seen that the increase in the ionic strength led to the reduction of
the polyanion effect on .
View larger version (11K):
[in this window]
[in a new window]
|
FIGURE 8
The dependence of the characteristic time of gram-TAEA
photoinactivation () measured in the presence of 0.06 µg/ml
Konig's polyanion on the concentration of KCl in the bathing solution.
Polyanion was added at both sides of the BLM. The initial value of the
current for each measurement of was approximately 1 µA. The
values of were obtained from single exponential approximation.
|
|
We tested different other polyanions in the system with gram-TAEA,
namely double-stranded and single-stranded DNA, RNA from tobacco mosaic
virus, polyglutamic acid, and polyacrylic acids of different molecular
weights (5000 and 250,000). None of these increased the value of .
It was found that modified polyacrylic acid, however, was effective,
although at very high concentrations (Fig.
9). The decelerated kinetics of the
current photoinactivation also had pronounced deviations from the
monoexponential curve in the presence of high concentrations of
modified polyacrylic acid as it was in the case of Konig's polyanion
(Fig. 9 A, curve 2). The concentration
dependence of the effect of modified polyacrylic acid is presented in
Fig. 9 B.
View larger version (14K):
[in this window]
[in a new window]
|
FIGURE 9
(A) The kinetics of the decrease in the
gram-TAEA-mediated current (I) through BLM in the
presence of 1 µM AlPcS3 after a flash of visible light
(at zero time) in the control (curve 1), and in the
presence of 140 µg/ml modified polyacrylic acid (curve
2) at both sides of BLM. The experimental data points were
fitted with single exponentials (solid curves) with
characteristic times = 0.60 s (curve 1),
and = 6.5 s (curve 2). In the case of
curve 2, double exponential fitting
I = I0 + 1exp(t/1) + 2exp(t/2) (dashed
curve) was also calculated with the following parameters:
1 = 52%, 1 = 10.4 s,
2 = 48%, 2 = 1.5 s. The
data presented as (I I)100%/(I0 I) versus the time (logarithmic
scale), where I0 is the initial current and
I is the stationary value of the current
after a flash. (B) The dependence of the characteristic
time of gram-TAEA photoinactivation () measured in the presence of
modified polyacrylic acid on its concentration. The polyacrylic acid
was added at both sides of the BLM. The initial value of the current
for each measurement of was approximately 1 µA. The values of were obtained from single exponential approximation.
|
|
As mentioned above, the addition of DNA or polyacrylic acid did not
increase the value of even at concentrations that were much higher
than the effective concentrations of Konig's polyanion. However, it
should be noted that the subsequent addition of Konig's polyanion
after the addition of these polyanions did not lead to the change in
the value of (data not shown). These data showed that, although DNA
did not change , it did interact with the membrane containing
gram-TAEA.
It should be pointed out that the addition of Konig's polyanion to the
bathing solutions (marked by the arrow in Fig.
10) led to an increase in the current
through BLM mediated by gram-TAEA. It should be noted in connection
with this that the photoinactivation kinetics (Figs. 2, 3, 6-9) were
recorded 10-15 min after the addition of polyanions when a new
steady-state level of the current was established. No change in the
current was observed in the case of usual gramicidin, gA (data not
shown).
View larger version (12K):
[in this window]
[in a new window]
|
FIGURE 10
Effect of the addition of Konig's polyanion (5 ng/ml)
at the moment marked by the arrow on the gram-TAEA-mediated current
through BLM.
|
|
| |
DISCUSSION |
As shown earlier (Rokitskaya et al., 1996, 1997), the
characteristic time of gramicidin channel photoinactivation () is
determined by the time of establishment of the new equilibrium between
transmembrane dimers and monomers of gramicidin after irreversible
inactivation of a part of them. The dependence of on the BLM
conductance induced by gram-TAEA (Fig. 3) enabled us to estimate the
rate constants of channel formation and termination for this gramicidin analog. These values (see Results) are close to the kinetic constants of gA known from the literature (KR = 4.6 * 1013
mol1s1cm2
and KD = 0.5 s1 for DPhPC at 26°C (Rokitskaya et al.,
1996); or KR = 20 * 1013
mol1s1cm2
and KD = 1.6 s1 for dioleoyllethicin at 25°C (Bamberg and
Lauger, 1974); KR = 3 * 1013
mol1s1cm2
and KD = 1.5 s1 for dioleoyllethicin at 21-23°C
(Zingsheim and Neher, 1974), though the
KR value of gram-TAEA is noticeably
lower than that of gA, which may be due to the increased polarity of
the C-terminus of the gram-TAEA peptide hindering the process of
formation of the transmembrane dimer (the channel state).
The polyanion/gram-TAEA system studied here can be considered as the
inverted one to the polylysine/OPg system examined in our previous
works (Krylov et al., 1998, 2000). Actually, it is the interaction
between polyelectrolytes and the charged gramicidin with the signs of
the charges switched. Several lines of evidence that will be discussed
in detail below confirm the similarity of these two systems, namely: 1)
the polyelectrolyte effects on the photoinactivation kinetics of
charged gramicidins were pronounced only when the additions of the
polyelectrolytes were made at both sides of the BLM (i.e., under the
symmetric conditions), 2) both polyelectrolyte effects were prevented
by increasing the ionic strength, and 3) both effects exhibited
the bell-shaped dependence on the concentration of the polyelectrolytes.
It has been suggested previously that the increase in the
characteristic time of photoinactivation observed upon binding of polylysine to OPg is due to formation of aggregates of OPg-polylysine clusters, i.e., is related to sequestering of OPg molecules into
domains floating in the matrix of neutral phosphatidylcholine molecules
(Krylov et al., 2000). The domain formation apparently results in
reduction of lateral and rotational mobility of gramicidin molecules in
a lipid bilayer, which manifests itself in the lengthening of the
channel lifetime. The interaction of polyanions with positively charged
gramicidin also can be described by this model. In fact, as seen from
Fig. 5 C, channel openings of increased duration occurred
very rarely after the addition of 100 ng/ml Konig's polyanion. However, under the conditions of measuring the integral current provided by simultaneous functioning of ~106
channels, the addition of Konig's polyanion at the same concentration resulted in the predominance of channels with a longer lifetime (Fig.
2). The dependence of the polyanion effect on the gram-TAEA concentration can be explained in terms of the above-mentioned model,
in particular, by assuming aggregation of gram-TAEA channels induced by
adsorption of polyelectrolyte chains on the membrane surface. Formation
of large aggregates of gram-TAEA channels is expected to result in
stabilization of the channels, i.e., in the lengthening of the channel
lifetime, due to reduction of the peptide mobility in the aggregates.
According to the data of 2H-NMR spectroscopy
(Mitrakos and Macdonald, 1998), binding of anionic polyelectrolytes
causes deceleration of lateral diffusion and ensemble fluctuations of
cationic amphiphiles in lipid bilayers. The requirement of adding
Konig's polyanion at both sides of BLM to produce a marked effect on
gram-TAEA photoinactivation kinetics indicates that molecules residing
in both leaflets of the membrane are involved in the interaction with
the polyanion, which results in domain formation.
The interaction of Konig's polyanion with lipid bilayers containing
positively charged gramicidin is of electrostatic character, which is
supported by the removal of the polyanion effect caused by polylysine,
and by the dependence of the polyanion effect on the ionic strength of
the bathing solution as evidenced by the data presented in Fig. 8. The
interaction does not occur with neutral gA and is hardly detected in
the case of gram-EDA having a single positive charge (Fig. 6).
Mechanisms of segregation of lipid domains induced by interaction of
polyelectrolytes with membranes containing charged lipids have been
discussed in the literature (Denisov et al., 1998; May et al., 2000).
It has been argued in (Hartmann et al., 1978; Sackmann et al., 1984;
May et al., 2000) that binding of a multicharged basic protein to
negatively charged lipid provokes a local change in line tension of BLM
(i.e., elastic deformation of a lipid bilayer), which becomes the
driving force of domain formation in membranes. Considering the effect
of Konig's polyanion in terms of the latter model, it can be assumed
that, following the step of electrostatic adsorption, Konig's
polyanion can interact hydrophobically with neutral lipids of the
membrane, thus altering its line tension. This two-step hypothesis
allows us to explain the absence of the effect of other polyanions on
gramicidin photoinactivation kinetics and the prevention of the effect
of Konig's polyanion by the other polyanions. It was noted in the
Results that a series of other polyanions studied here, except for
modified polyacrylic acid, did not produce deceleration of gram-TAEA
channel kinetics, though these polyanions prevented the effect of
Konig's polyanion. Most likely, DNA and other polyanions actually bind
to gram-TAEA on the membrane surface, which prevents the subsequent
binding of Konig's polyanion, but does not cause the aggregation of
gram-TAEA channels and thereby does not bring about an increase in .
The reason might be that binding of DNA and other polyanions, except for modified polyacrylic acid, does not lead to changes in the line
tension of lipid bilayers.
It can be supposed that the presence of abundant hydrophobic groups in
polyelectrolyte molecules, such as Konig's polyanion and modified
polyacrylic acid, provides their additional interaction with membrane
lipids, leading to modulation of the line tension. This hypothesis is
favored by the data of Tribet (1998) demonstrating that incorporation
of hydrophobic groups (e.g., palmitoyl, cholesteroyl) provokes binding
of charged polymers to bilayers formed of neutral lipids. Actually, it
has been found by Maltseva et al. (2002) that Konig's polyanion, in
contrast to unmodified polyacrylic acid, can adsorb on a neutral lipid
membrane, thereby making it negatively charged. It is also relevant to
this point that, according to a series of studies, myristoylation or
palmitoylation of a number of basic proteins substantially increases
their membrane affinity (Hancock et al., 1990; Sigal et al., 1994; Kim
et al., 1994; Buser et al., 1994; McLaughlin and Aderem, 1995;
Swierczynski and Blackshear, 1996; Murray et al., 1997, 1998; Bahr et
al., 1998; Victor and Cafiso, 1998; Victor et al., 1999; Arbuzova et al., 2000).
As seen from Fig. 7, there was a maximum in the concentration
dependence of the polyanion effect on the characteristic time of gram-TAEA photoinactivation. That is, further increasing the concentration of added Konig's polyanion led to partial reversal of
the polyanion-induced increase in . The similar phenomenon was
observed in our study of polylysine interaction with OPg channels (Krylov et al., 2000). This reversal most probably results from the
electrostatic repulsion of polymer loops under the conditions of
polymer competition for gram-TAEA charges, which leads to loosening of
charged domains by including additional molecules of neutral lipid into
them, i.e., to partial recovery of homogeneity in lateral distribution
of two components (neutral and charged, see Macdonald et al., 1998). It
is worth noting that disappearance of ordered domains in mixed bilayers
formed of neutral and cationic lipids was observed earlier at high
concentrations of added DNA (Hirsch-Lerner and Barenholz, 1998;
Subramanian et al., 2000) and some other polyanions (Mitrakos and
Macdonald, 1997).
The above consideration implies the existence of at least two
populations of gram-TAEA channels in BLM, namely free channels and
those arranged in domains. In fact, a single exponential did not
suffice to describe the photoinactivation kinetics in the presence of
Konig's polyanion. Under these conditions the kinetics was fitted well
by a sum of two exponential components, with the characteristic time of
one of them being close to that of the control. The exponential factor
of the slower component supposedly corresponds to the channel lifetime
in the domain phase.
In conclusion, it should be pointed out that certain advantages of our
model system, including the simplicity and reliability of current
measurements, make it useful to study mechanisms of protein-mediated
domain formation in membranes. These mechanisms are of particular
interest now in view of their involvement in processes of lipid raft
formation and signal transduction (Brown and London, 2000; Simons and
Toomre, 2000, and references in both).
According to the analysis performed by Rokitskaya et al. (1996),
the flash-induced decay of the gramicidin-mediated current, here called
the kinetics of photoinactivation, represents the relaxation process
after a concentration jump, i.e., the gramicidin monomer-dimer
equilibration. The following formalism, based mainly on the
consideration of Bamberg and Lauger (1973), describes the kinetics of
the current transient after a flash of light.
We are grateful to Dr. T. I. Rokitskaya for helpful discussions.
This work was partially supported by the grants 00-04-48299 and
00-03-33172 from the Russian Foundation for Basic Research.
Address reprint requests to Y. N. Antonenko, Belozersky Institute,
Moscow State University, Moscow 119899, Russia. Tel.: +70-95-939-5360;
Fax: +70-95-939-3181; E-mail: antonen{at}genebee.msu.su.
TOP
ABSTRACT
INTRODUCTION
the structure in a lipid environment.
Nat. Struct. Biol.
6:609-612
