Biophysical Chemistry 103 (2003) 179–190
0301-4622/03/$ - see front matter
2002 Elsevier Science B.V. All rights reserved.
PII: S0301-4622 Ž02 .00255-7
Kinetic isotope effects of proton transfer in aqueous and methanol
containing solutions, and in gramicidin A channels
Anatoly Chernyshev , Regis Pomes , Samuel Cukierman *a b,c a,´ `
Department of Physiology Loyola University Medical Center, 2160 South First Ave, Maywood, IL 60153, USAa
Structural Biology and Biochemistry, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canadab
Department of Biochemistry, University of Toronto, Toronto, Canadac
Received 23 August 2002; received in revised form 31 August 2002; accepted 31 August 2002
Abstract
The electrochemical conductivities of HCl and DCl were measured in: H O and D O; in methanol and fully2 2
deuterated methanol; and in water–methanol solutions. The single channel conductances to H (g ) and D (g ) inq qH D
various gramicidin A (gA) ion channels incorporated in glycerylmonooleate planar bilayers were also measured.
Kinetic isotope effects (KIE) were estimated from the ratio of conductivity measurements. In 1 and 5 M HCl aqueous
solutions and in 1 M HClq3.7 M methanol, the KIE (f1.35) is not different from values previously determined in
dilute acid solutions. This suggests that the mobility of protons in those solutions is largely determined by proton
transfer. In 10 M HCl, however, where the mobility of protons is likely to be determined by hydrodynamic diffusion,
the measured KIE is considerably larger (1.47). Possible causes for this effect are discussed. The KIE of proton
conductivities in 5 and 50 mM HCl in methanol and d-methanol is f1.15. This is considerably smaller than the ratio
between conductivities of 5 mM KCl in methanol and d-methanol (1.24). The KIE values (1.22–1.37) for g in gAH
channels in 1 M HCl are significantly larger than for other monovalent cations and consistent with H transfer.q
Methanol reduces g in gA channels. The KIE of this effect is not different from the one measured in the absenceH
of methanol. Possible mechanisms for the methanol-induced block of H conductivities in solution and gA channelsq
are discussed.

2002 Elsevier Science B.V. All rights reserved.
Keywords: Water wire; Ionic permeability; Grotthuss mechanism; Single channel conductance; Proton transfer in deuterated
methanol; Deuterium oxide
1. Introduction
The conductivity of protons in water is larger
than of any other ion. This high conductivity
cannot be explained by the hydrodynamic mobility
*Corresponding author. Tel.: q1-708-216-9471; fax: q1-
708-216-6308.
E-mail address: scukier@lumc.edu (S. Cukierman).
of a solvated proton w(H O) for examplex. Aq3
proton transfer mechanism that became known as
Grotthuss’s could account for the relatively high
mobility or conductivity of protons in water w1,2x.
Consider a set of water molecules interconnected
via hydrogen bonds (H-bonds). In a classical
Grotthuss’s mechanism, the mobility of protons
would occur in two distinct (hop and turn) steps

180 A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
w1–5x. The hopping step consists of a proton
transfer between (H O) and an adjacent H O. Asq3 2
the proton hops along water molecules, the dipole
moments of waters rotate in approximately the
same direction of proton hopping. If another proton
is to be transferred in the same direction as the
previous one, waters must rotate back (turn step)
to their initial configuration w1,2,5x. Historically,
the turn step (water rotation) has been considered
the limiting step for proton mobility in bulk water.
Agmon w6,7x has argued that the rate-limiting step
in proton transfer between water molecules cannot
be the rotation of bulk water molecules. Instead,
it was proposed that the rate-limiting step of proton
transfer in bulk water is the disruption of one H-
bond between waters in the first and second
solvation shells of (H O) w6–9x.q3
The production of ATP in all cells is ultimately
driven by a translocation of protons through a
membrane protein. Consequently, understanding
the mechanisms by which protons are transferred
inside proteins is a significant challenge in biology.
In particular, it is of interest to elucidate the
mechanisms by which protons are transferred along
an approximately unidimensional chain of water
molecules interconnected via H-bonds (water or
proton wires) w3,4x. It has been demonstrated that
water wires are present inside the cavities of
various proteins involved in bioenergetic processes
w10–13x. It is possible that proton transfer in water
wires follows hop-and-turn steps similar to a Grot-
thuss mechanism in which the turn step appears to
be rate limiting w3,4,14–16x.
The structures of bioenergetic proteins are
extremely complex, and the properties of proton
transfer in these proteins cannot be directly meas-
ured at the single molecular level. On the other
hand, it is possible to study proton transfer in a
membrane protein (gramicidin A, gA) that forms
ion channels in lipid bilayers. gA is a highly
hydrophobic pentadecapeptide secreted by Bacillus
brevis w17x. It consists of an alternating sequence
of D- and L-amino acids w18x that defines a right-
handed b helix in lipid bilayers w19–21x. The6.3
side chain residues of the gA channel are in contact
with the hydrophobic lipid environment while the
carbonyls and amides face the hydrophilic pore of
gA. Each gA molecule resides in one monolayer
of a lipid bilayer. The association via six H-bonds
between the amino termini of two gA’s located in
opposite monolayers forms an ion channel that is
selectively permeable to monovalent cations. The
functional gA channel is f25 A long, and its˚
hydrophilic pore (f4 A in diameter) contains a˚
single water wire comprised of seven to nine water
molecules w22,23x. The lifetime of the gA channel
is determined by the dissociation rate of gA mon-
omers in bilayers.
In our laboratory, the amino termini of two gA
molecules have been linked to a dioxolane group
w24–26x. The presence of two chiral carbons in
the dioxolane linker defines the SS and RR dias-
tereoisomers of dioxolane-linked gA channels (for
the sake of simplicity, these channels will be
referred to as the SS and RR channels). There are
significant differences between the proton transfer
properties in native gA, SS and RR channels
w24,25,27–30x. Those distinct properties make
these molecules interesting models to probe the
relationships between structure and function of
proton transfer in proteins.
The conductivity of protons in some alcohols,
and methanol in particular, is considerably larger
than of other monovalent cations w31–33x. In
analogy to what had been proposed for water, this
‘extra’ proton conductivity led to the suggestion
that protons could also be transferred between
methanol molecules by a Grotthuss-like mecha-
nism w31,32x. Our particular interest on methanol
effects on proton transfer relates to the fact that
this molecule could fit inside the pore of gA
channels. Thus, it was of interest to probe if and
how methanol modulates H transfer in gA chan-q
nels. We have demonstrated that, indeed, methanol
caused a significant attenuation of proton currents
in SS channels w34x. These experimental results
were consistent with a model in which one meth-
anol molecule is present between water molecules
in the water wire of the SS channel (andyor at the
channelysolution interface), and somehow attenu-
ates proton transfer through the channel. Quite
interestingly, longer chain alcohols like ethanol or
propanol, that are not likely to fit inside the pore
of gA channels, do not attenuate proton currents
in the various gA channels studied (Godoy and
Cukierman unpublished) w34x. The mechanisms,

181A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
by which proton currents in the SS channel are
attenuated by methanol are not known.
Consequently, our experimental proposals in this
study were: (1) to measure the kinetic isotope
effect (KIE) of proton transfer in native gA, SS
and RR channels. Are these KIE consistent with a
proton transfer mechanism in water? (2) To further
our understanding of the mechanisms by which
methanol could attenuate H conductivity in waterq
and in various gA channels, KIE values were also
measured in HCl or DCl solutions in the presence
of CH OH (or CD OD). Previous measurements3 3
of KIE for proton conductivities have apparently
been limited to dilute aqueous solutions of HCl
w35–39x. In order to properly evaluate the KIE for
proton transfer in various gA channels under sev-
eral of our experimental conditions, it became
necessary to perform conductivity measurements
in several aqueous and methanol solutions. To our
knowledge, several electrical conductivity meas-
urements and associated KIE values in various
solutions are presented here for the first time.
2. Material and methods
2.1. Bilayers
Planar lipid bilayers were formed from a decane
solution (f60 mgyml) of glycerylmonooleate
(GMO, NuCheck Co., Elysian, MN; Sigma, St.
Louis, MO). Bilayers were formed across a 150-
mm diameter hole on a polystyrene partition sep-
arating two aqueous compartments. The formation
of the lipid bilayer was monitored by visual inspec-
tion and capacitance measurements. Experiments
were performed at room temperature (23–25 8C).
2.2. Solutions
The values of single channel conductances to
protons (g , measured in picosiemens, pS) wereH
measured in 0.05, 1 and 5 M HCl (or DCl) in
H O (or D O). Experiments were also performed2 2
in 1 M HCl (or DCl) solutions with 15% vyv of
CH OH (or CD OD). The final concentration of3 3
methanol or fully deuterated methanol was f3.7
M. HCl and CH OH (HPLC grade) were obtained3
from Fisher Scientific (Pittsburgh, PA). DCl
(99.7% D) and D O (99.9% D) were purchased2
from Aldrich (Milwaukee, WI). CD OD (99.8%3
D) was purchased from Alfa Aesar (Ward Hill,
MA).
2.3. Measurements of solution conductivities
Electrical conductivities of freshly prepared
solutions were measured at 24 8C with a YSI-3200
conductivity meter (Yellow Spring Instruments,
Yellow Springs, OH). Equivalent conductivities
wL, mSy(cm M), where mS is milisiemensx are
reported in this study.
2.4. gA channels
The synthesis, purification and characterization
of dioxolane-linked gA channels were previously
described w26,27x. The native gA channels used in
this study were purchased from Fluka (Milwaukee,
WI). gA channels were added from a methanol
stock solution (f10 M) that was routinelyy8
stored at fy15 8C.
2.5. Single channel current measurements
Proton currents through a single channel mole-
cule were measured by voltage clamping the lipid
bilayer using an Axopatch 200B (Axon Instru-
ments, Union City, CA). For native gA channels,
a constant 50 mV DC voltage step was applied
across the membrane. For the covalently linked
gA dimers, voltage clamp ramps from 0 to f100
mV were applied in f5 s. Values for g inH
picosiemens (pS) were measured by regression
analysis of the linear portion (usually from 0 to
f75 mV) of I–V plots. pClamp (Axon Instru-
ments) was used for applying voltages and record-
ing single channel currents. At least five distinct
single channel measurements were obtained from
at least two distinct lipid bilayers in each experi-
mental condition. Experimental points in this study
are shown as mean"S.E.M. In most plots, the
error bars of the experimental points are smaller
than the size of the symbols.

182 A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
Fig. 1. Top panel: equivalent electrical conductivities (L) of
HCl (circles and squares) or DCl (triangles and diamonds) in
water. Squares and diamonds are from published data w35,36x
while circles and triangles represent our measurements. Bottom
panel: ratios (R) between L values of HCl and DCl solutions
(open symbols); and L yL (filled symbols). See text forH D
detailed description and sources of some of the experimental
points.
3. Results and discussion
3.1. KIE in aqueous solutions
The upper panel in Fig. 1 shows the equivalent
conductivities (L) of HCl and DCl solutions at
various concentrations. The squares and diamonds
were obtained from available data w35,36x. Circles
and triangles represent our own measurements.
The lower panel in Fig. 1 shows the ratios (R)
L yL (open symbols) and L yL (filledHCl DCl H D
symbols). Diamonds were calculated from squares
and triangles at a few selected concentrations of
HCl and DCl (upper panel of Fig. 1). The open
circles in the bottom panel of Fig. 1 are from our
measurements (calculated from circles and trian-
gles in the top panel of Fig. 1). The inverted open
triangle is from measurements by Baker and La
Mer w37x. The inverted filled triangle is the KIE
for proton transfer after correction for L . ThisCl
correction consisted in measuring the conductivity
of a 10 mM KCl solutions and subtracting it from
L of a 10 mM HCl solution w37x. In a similarHCl
way, the filled triangle is also the KIE for proton
transfer after correction for L using 17 mM KClCl
w38x. The open square was measured in 50 mM
HCl, and the corresponding filled square is the
KIE for proton transfer after direct measurements
of the transference numbers for H and D w39x.q q
The dashed line in this graph represents the ratio
between the shear viscosities of D O and H O at2 2
25 8C w36,39,40x. Table 1 lists some of our
measurements of L and L at various acidHCl DCl
concentrations (see also Fig. 1). Also shown is
the L in H O and D O (3 M KCl). The KIEKCl 2 2
values for diluted concentrations of HCl up to 5
M are within the range of 1.32–1.36. In 10 M
HCl, however, the KIE increases to 1.47. The ratio
between L in H O and D O is 1.17, and thisKCl 2 2
value is consistent with the ratio between shear
viscosities of deuterium oxide and water (Fig. 1).
The ratio between the electrical conductivities
of dilute solutions of various alkalines in H O and2
D O is approximately 1.20 w38,41,42x. In concen-2
trated 3 M KCl, this ratio is 1.17 (Table 1). These
numbers are similar to the ratio between the
viscosities of D O and H O at room temperature2 2
(1.22) w36,39,40x, and in agreement with the idea
that the mobility of these ions is determined in
part by the frictional hindrance of the solvent. By
contrast, the ratio between the electrical conductiv-
ities of HCl and DCl solutions is considerably
larger (f1.35, Fig. 1). The significantly larger
KIE for H conductivity in water in relation toq
other ions suggests that proton mobility in a wide
range of acid concentrations is not determined by
the hydrodynamic diffusion of (H O) .q3
Apparently, measurements of KIE were previ-
ously limited to dilute solutions of HCl (see Fig.
1). In this study, KIE values for L in 1, 5, andHCl
10 M acid solutions were determined. The ratios

183A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
Table 1
Equivalent conductivities of HCl, DCl and KCl solutionsa
0.05 M HCl 1 M HCl 5 M HCl 10 M HCl 3 M KCl
or DCl or DCl or DCl or DCl
H O2 377.6 322.8 157.0 69.74 85.37
D O2 279.6 244.2 118.9 47.51 73.20
KIE 1.35 1.32 1.32 1.47 1.17
Measurements performed at 24.0 8C (L in mS cm M ).a y1 y1
between L and L at 1 and 5 M concentra-HCl DCl
tions (1.32, Table 1) are not very different from
measurements in dilute acid solutions (Fig. 1,
bottom panel). Thus, it is possible that a substantial
fraction of proton mobility in 5 M HCl and DCl
is still likely to be determined by proton transfer.
However, in 10 M HCl, the KIE value is consid-
erably larger (1.47) than in dilute acid solutions
and in 3 M KCl (1.17).
The transfer of protons between water molecules
is obliterated as wHClx increases w43,44x. It has
been proposed that the high mobility (or conduc-
tivity) of H in water is a consequence of anq
almost isoenergetic equilibrium between (H O )q9 4
and (H O ) w6–9,45–47x. Experimental condi-q5 2
tions that disrupt this equilibrium would attenuate
proton conductivity in water w6,7x. In particular,
the structure of solvated protons in concentrated
acid solutions is quite different from dilute solu-
tions w48,49x. In 2 molal HCl for example, the
ratio between the mobilities of H and Cl is 6.5,q y
and in 10 molal solutions, this ratio decreases to
f2.5 w49x. This attenuation is caused by a signif-
icant reduction of H mobility w43x. As wHClxq
increases, the relative contribution of proton trans-
fer to L will also decrease, and at very concen-H
trated HCl solutions, it is likely that the mobility
of protons is determined by the hydrodynamic
diffusion of clusters of solvated protons
w29,43,44,48,49x. Consequently, the expectation
was that in 10 M HCl the KIE would be close to
the ratios between the viscosities of deuterium
oxide and water andyor between the L valuesKCl
in concentrated 3 M KCl in H O and D O. Instead,2 2
a KIE of 1.47, which is considerably larger than
those ratios has been measured (Fig. 1 and Table
1). Assuming that L in 10 M HCl is determinedH
essentially by the hydrodynamic diffusion of pro-
tonated water molecules, one explanation for the
larger KIE in 10 M HCl is that the size or volume
of protonated water(s) in a D O cluster is larger2
than in H O. Stronger H bonds between D O2 2
molecules compared to H O w6,48,49x could2
explain differences between the sizes of water
clusters. In itself, stronger D-bonds between water
molecules could also hamper the mobility of clus-
ters of protonated water(s).
3.2. KIE in gA channels in aqueous solutions
In Fig. 2, representative single channel H andq
D currents vs. transmembrane voltage (I–Vq
plots) for the SS and RR channels in 1 M HCl
and DCl solutions are shown. The single channel
conductances in this figure are: g (883 pS, SS;H
376 pS, RR); g (697 pS, SS; 288 pS, RR). Fig.D
3 shows single channel recordings of native gA
channels at a transmembrane voltage of 50 mV. In
this figure, g and g for single channel openingsH D
were f762 pS and f600 pS, respectively. A
summary of measurements of g and g in 0.05,H D
1 and 5 M HCl and DCl solutions for the various
gA channels is reported in Table 2.
The KIE values for H transfer in various gAq
channels at several concentrations of HCl and DCl
are shown in Table 3. The KIE values of Hq
transfer in the various gA channels varied from
1.22 in native gA channel at 50 mM HCl and DCl
to 1.37 for the RR channel in 5 M HCl and DCl
(overall average: 1.31"0.02). These results are in
agreement with determinations of KIE previously
performed in native gA channels only w51x.
The single channel conductances of native gA
channels to alkaline metals in H O and D O were2 2
previously determined w42,51x. The ratios between
these single channel conductances in H O and2

184 A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
Fig. 2. I–V (picopamperes, pA vs. milivolt, mV) plots of
H or D currents of single SS and RR channels. The singleq q
channel recordings were low-pass Bessel filtered at 0.5 kHz
and digitized at 5 kHz. The small downward deflections in each
recording are unresolved (due to filtering) channel closures.
Fig. 3. H or D currents in native gA channels at an appliedq q
membrane potential ofq50 mV. Channel recordings were low-
pass Bessel filtered at 100 Hz and digitized at 1 kHz. Channel
openings are represented by upward deflections of the current
trace. The top and bottom recordings have three and four dis-
tinct channel openings, respectively.
Table 2
Single channel conductances (pS) to H or D of gA channels in various solutions (mean"S.E.M., n)q q
GA SS RR
50 mM HCl 53.1"0.7 (27) 108.0"3.2 (18) –
1 M HCl 746.5"4.7 (12) 891.3"7.7 (7) 366.9"7.6 (16)
5 M HCl 2612.3"26.3 (19) 1849.7"18.4 (4) 1559.1"70.7 (10)
50 mM DCl 43.5"0.6 (20) 79.7"1.4 (26) –
1 M DCl 588.7"8.5 (12) 679.3"4.7 (19) 278.3"12.0 (11)
5 M DCl 1926.3"26.2 (7) 1417.1"25.1 (12) 1140.3"37.9 (9)
1 M HClqCH OH3 515.9"5.5 (9) 551.3"11.2 (10) 264.2"12.4 (12)
1 M DClqCD OD3 380.3"4.0 (17) 422.3"11.8 (9) 201.2"5.0 (25)
D O were between 1.03 for Li and 1.16 forq2
Cs . For Na , whose permeation in native gAq q
channels is limited by the permeability of water
molecules in a single file diffusion mechanism
w22x, that ratio was 1.11. Notice that because in

185A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
Table 3
Kinetic isotope effects of solution conductivities and single
channel conductances in HCl
50 mM 1 M 5 M
Solutions 1.35 1.32 1.32
gA 1.22"0.01 1.27"0.01 1.36"0.02
SS 1.36"0.01 1.31"0.01 1.31"0.01
RR – 1.32"0.03 1.37"0.02
Table 4
Equivalent conductivities of HCl, DCl and KCl in methanol containing solutionsa
5 mM HCl 5 mM KCl 50 mM HCl 1 M HCl
or DCl or DCl or DCl
CH OHb3 129.74 87.82 103.78
CD ODb3 111.42 70.78 90.14
KIE 1.16 1.24 1.15
3.7 M CH OH3 243.00
3.7 M CD OD3 181.00
KIE 1.34
Measurements performed at 24.0 8C (L in mS cm M ).a y1 y1
100% methanol solutions.b
the present experimental conditions gA channels
are selective only to H (Cl does not permeateq y
the channel), the ratio between g values in HClH
and DCl is indeed the real KIE for proton transfer
between the AgyAgCl electrodes located on dif-
ferent compartments across gA channels, and does
not need to be corrected for the Cl conductivityy
(see Fig. 1 above and related text). That the KIE
values for H transfer in native gA, and in the SSq
and RR channels are significantly larger than the
ratios for alkalines (especially the one measured
for Na w42,50,51x) suggests that it is likely thatq
a proton transfer mechanism is operating inside
these channels and not the hydrodynamic flow of
(H O) . An alternative interpretation would beq3
that the rate limiting step for proton transfer in gA
channels is not inside the channel (or at
membrane-channelysolution interface) but in bulk
solution. However: (1) g is strongly modulatedH
by the nature of the lipid bilayer
w24,25,30,52,53,54x; (2) the activation energies for
g in various gA channels (;28 kJymol) areH
significantly different from the activation energy
of H conductivity in bulk solution w52x; and (3)q
Native gA, SS and RR channels have different
susceptibilities to methanol blockade of proton
currents (see below). Taken together, these results
indicate that the limiting step for H transfer is inq
the gA channel or at the membrane-channely
solution interface rather than in bulk solution.
3.3. KIE in methanol and in methanolywater
mixtures
Table 4 shows measurements of L in HCl, DCl
and KCl solutions in methanol. The ratios between
L values in methanol and d-methanol solutions (5
and 50 mM HCl) are f1.15. Interestingly, Table
4 also shows that the ratio between the conductiv-
ity measurements of 5 mM KCl in methanol and
d-methanol (1.24) is larger than the ratios of
L in methanol and d-methanol.HCl
Two distinct experimental observations led to
the proposal that protons could be transferred
between methanol molecules: (1) in pure methanol
(as in water) there is an extra conductivity of HCl
in relation to LiCl, KCl and NaCl w31,33,53x. We
have now confirmed this for KCl and extended
the measurements to deuterated methanol solutions
(Table 4). Notice that the difference between
L and L wthe ‘extra’ proton conductivity,HCl KCl
f41 mSy(cm M)x is approximately the same in
both CH OH and CD OD. (2) L has an anom-3 3 HCl
alous mole fraction dependence on methanol in
waterymethanol solutions. L is attenuated asHCl
the methanol mole-fraction in aqueous solutions
increases from 0 to 80%. However, above 80%
there is a significant and continuous increase in
L w31,32,53x.HCl
The ratio between L in methanol and d-KCl

186 A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
Fig. 4. I–V plots of H and D currents of single SS and RRq q
channels in 1 M HClq3.7 M CH OH (top recordings for each3
panel) and in 1 M DClq3.7 M CD OD (bottom recordings3
for each panel). The single channel recordings were low-pass
Bessel filtered at 0.5 kHz and digitized at 5 kHz.
methanol provides an indication of how the relative
mobilities of K and Cl are affected in theseq y
different solutions. The fact that: (1) in methanol
and d-methanol solutions there is an ‘extra’ con-
ductivity of HCl in relation to KCl solutions; and
(2) the KIE of proton conductivity in methanol is
considerably smaller than the ratio of L inKCl
methanol and d-methanol provides additional sup-
port for proton transfer between methanol mole-
cules. Interestingly, H transfer in methanol isq
considerably less sensitive to HyD substitution
than in water. The reason for this effect is not
known.
The ratio between the conductivities of 1 M
HCl and DCl solutions in 3.7 M methanol (and d-
methanol) is 1.34 (Table 4, last column). The
equilibrium constant of the reaction (H O) qq3
CH OHlH Oq(CH OH ) is 0.23 w54x. Thus,q3 2 3 2
CH OH is a poorer base than H O. Assuming that3 2
a given (H O) has the same probability of beingq3
solvated by either H O or CH OH, the energeti-2 3
cally favored pathway for H transfer in metha-q
nolywater mixtures is between (H O) and H Oq3 2
w31,32,55x. The KIE for H transfer in 3.7 Mq
methanol solutions (1.34) is close to that measured
in its absence (1.32, Table 1 and Fig. 1) and
considerably larger than in pure methanol solutions
(see above). This suggests that in 3.7 M methanol,
protons are transferred mainly along chains of
molecules containing only water w55x. Because in
methanolywater mixtures there is a reduction in
the number of pathways containing water mole-
cules only, proton conductivity in these solutions
is significantly attenuated in relation to pure water.
3.4. KIE in gA channels in aqueous solutions
containing methanol
Fig. 4 shows I–V plots for the SS and RR
channels in 1 M HClq3.7 M CH OH (upper3
traces in both top and bottom panels), and in 1 M
DClq3.7 M CD OD (lower traces in top and3
bottom panels). In Fig. 5, representative recordings
of native single gA channels in solutions identified
at the top of each panel are shown. The attenuation
of H or D currents by methanol or d-methanolq q
in gA channels has the following characteristics
(Tables 2 and 5):
1. Quigley et al. w34x reported that in 1 M HCl,
3.7 M CH OH attenuated g by f37% in the3 H
SS channel. This observation has now been
independently confirmed for the SS channel
(38% attenuation in g ). Moreover, methanolH
also attenuated g in the RR and native gAH
channels by 28 and 31%, respectively (Table
5). Because attenuations of g are stronger thanH
the attenuation of L in 1 M HCl solutionsHCl
(Table 5), it is likely that methanol partitions in
the pore of gA channels or at the membrane-
channelysolution interfaces and causes an
‘extra’ reduction of H currents than in bulkq
solution w34x.

187A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
Fig. 5. H or D currents in native gA channels at an applied membrane potential of q50 mV. Channel recordings were low-passq q
Bessel filtered at 100 Hz and digitized at 1 kHz. Channel openings are represented by upward deflections of the current trace. Five
and six distinct channel openings are seen in the top and bottom recordings, respectively.
Table 5
Kinetic isotope effects of conductivities in solutions and in Gramicidin A channels in 1 M HClq3.7 M Methanol
Solution gA SS RR
L yL or g ygHCl,methanol HCl H,methanol H 0.77 0.69"0.01 0.62"0.02 0.72"0.03
L yL or g ygDCl,D-methanol DCl D,D-methanol D 0.74 0.65"0.02 0.62"0.02 0.72"0.02
L yL orHCl,methanol HCl,D-methanol 1.34 1.36"0.01 1.31"0.01 1.31"0.04
g ygH,methanol D,D-methanol
2. D currents in various gA channels were atten-q
uated by CD OD by approximately the same3
ratio as H currents by CH OH. d-Methanolq 3
attenuated g (1 M DCl) by 38, 28 and 35% inD
the SS, RR and native gA channels, respectively.
3. The attenuation of g (or g ) by methanol isH D
larger in the SS channel, followed by gA and
RR. This strengthens the hypothesis w34x that
the blockade of proton currents in gA channels
is significant, and cannot be simply explained
by the attenuation of proton conductivity in bulk
solution. It is possible that these gA channels
have distinct partition coefficients for methanol.
Moreover, because the difference between the
various gA channels relates to the absence of
the dioxolane linker as in native gA, or to the
different conformations of the linker in the SS
and RR channels, it may well be that the

188 A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
residence time (or permeability) of methanol in
the pore is modulated by interactions with the
dioxolane.
4. The KIE’s for proton transfer in methanol solu-
tions are 1.36 (gA) and 1.31 (SS and RR
channels). On average, these values are not
different from proton transfer in waterymethanol
solutions (Table 3).
In principle, two distinct and non-mutually
exclusive mechanisms could account for a
decreased g in various gA channels in the pres-H
ence of methanol w28,34x: (a) Assume that one
methanol molecule resides inside the pore of the
SS channel and shuttles protons between waters
by a hop-turn mechanism. As a consequence, the
transfer of protons in the channel occupied by a
methanol molecule would occur with a decreased
efficiency (smaller g ). Distinct mechanisms couldH
account for this effect. For example, if the proton-
ation of a water molecule by the H released fromq
an adjacent methanol, andyor the protonation of
methanol by a H released from an adjacent waterq
molecule is significantly slower than the H trans-q
fer between two adjacent water molecules (see
above), then g would be attenuated. AnotherH
possibility could be that the reorientation step of
the methanol molecule inside the channel is sig-
nificantly slower than the reorientation of water
molecules. In these cases, the amplitude of single
channel H currents would dwell between an openq
state (no methanol inside the pore), a closed state
(channel closure) and an intermediary level of
current between the fully closed and open states
(methanol is inside the pore and transfers H withq
a decreased efficiency in relation to water); (b) A
distinct possibility is that while methanol is inside
the channel, H transfer in the channel is com-q
pletely blocked. In this case, the amplitude of the
single channel proton current would dwell between
an open state (no methanol inside the pore), a
closed state (channel closure) and a fully blocked
state (methanol inside the pore). The fully blocked
and closed states cannot be discriminated in elec-
trical recordings of single channel H currents.q
The shortest time resolution of single channel Hq
currents in planar bilayers is of the order of tens
of ms w28x. If it is assumed that either mechanism
A (intermediary open current level between open
and closed states caused by methanol in the pore)
or B (complete blockade of the open state by
methanol) occurs in a time scale considerably
shorter than tens of microseconds, it will not be
possible to detect these phenomena in the record-
ings of single channel H currents. The measure-q
ments of KIE in gA channels in water–methanol
solutions were undertaken with the aim of discrim-
inating between possible blockade models of Hq
or D currents by methanol. Even though the KIEq
for the attenuation of proton currents by methanol
in gA channels did not permit a distinction
between mechanisms A and B above, the possibil-
ity that the KIE for H transfer with a methanolq
molecule in a single file of water molecules in the
channel may have the same value as in its absence
cannot be eliminated.
In summary, the novel results and conclusions
in this study were: (1) in 10 M HCl the KIE for
proton transfer is substantially larger than in more
dilute acid solutions. It is possible that there are
significant differences between the structures of
solvated H and D ; (2) the KIE for protonq q
conductivity in pure methanol and in mixtures of
waterymethanol solutions were measured. While
in waterymethanol mixtures the KIE values are
consistent with proton transfer occurring between
water molecules, the KIE values measured in pure
methanol solutions suggest that a proton transfer
mechanism does indeed occur between methanol
molecules. (3) KIE values measured in various
HCl solutions in several gA channels are similar
to that measured in HCl solutions, and considera-
bly larger than with other monovalent cations. This
suggests that proton transfer occurs between water
molecules in the various gA channels.
Acknowledgments
This work was supported in part by NIH (GM-
59674).
References
w1x J.D. Bernal, R.W. Fowler, A theory of water and ionic
solution, with particular reference to hydrogen and
hydroxyl ions, J. Chem. Phys. 1 (1933) 515–548.

189A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
w2x B.E. Conway, Proton solvation and proton transfer
processes in solution, in: J.O.M. Bockris, B.E. Conway
(Eds.), Modern Aspects of Electrochemistry, Butter-
worths, London, 1964, pp. 43–148.
w3x J.F. Nagle, H.J. Morowitz, Molecular mechanisms for
proton transport in membranes, Proc. Natl. Acad. Sci.
USA 75 (1978) 298–302.
w4x J.F. Nagle, S. Tristam-Nagle, Hydrogen bonded chain
mechanisms for proton conduction and proton pumping,
J. Membr. Biol. 74 (1983) 1–14.
w5x V.H. Danneel, Notiz uber ionengeschwindigkeiten, Z.¨
Elektrochem. 11 (1905) 249–252.
w6x N. Agmon, Hydrogen bonds, water rotation and proton
mobility, J. Chim. Phys. 93 (1996) 1714–1736.
w7x N. Agmon, The Grotthuss mechanism, Chem. Phys.
Lett. 244 (1995) 456–462.
w8x T.J.F. Day, U.W. Schmitt, G.A. Voth, The mechanism of
hydrated proton transfer in water, J. Am. Chem. Soc.
122 (2000) 12027–12028.
w9x U.W. Schmitt, G.A. Voth, The computer simulation of
proton transfer in water, J. Chem. Phys. 111 (1999)
9361–9381.
w10x L. Baciou, H. Michel, Interruption of the water chain
in the reaction center from rb. sphaeroides reduces the
rate of the proton uptake and of the second electron
transfer to Q_B, Biochemistry 34 (1995) 7967–7972.
w11x M. Branden, H. Sigurdson, A. Namslauer, R.B. Gennis,
P. Adelroth, P. Brzezinski, On the role of the K-proton
transfer pathway in cytochrome c oxidase, Proc. Natl.
Acad. USA. 98 (2001) 5013–5018.
w12x H.J. Sass, G. Buldt, R. Gassenich, et al., Structural
alterations for proton translocation in the M state of
wild-type bacteriorhodopsin, Nature 406 (2000)
649–653.
w13x H. Luecke, B. Schobert, H.T. Richter, J.P. Cartailler,
J.K. Lanyi, Structure of bacteriorhodopsin at 1.55 A˚
resolution, Proc. Natl. Acad. USA. 291 (1999) 899–911.
w14x R. Pomes, B. Roux, Structure and dynamics of a proton`
wire: a theoretical study of H translocation along theq
single-file water chain in the gramicidin A channel,
Biophys. J. 71 (1996) 19–39.
w15x R. Pomes, B. Roux, Free energy profiles for Hq`
conduction along hydrogen-bonded chains of water
molecules, Biophys. J. 75 (1998) 33–40.
w16x R. Pomes, B. Roux, Molecular mechanism of Hq`
conduction in the single-file water chain of the grami-
cidin channel, Biophys. J. 82 (2002) 2304–2316.
w17x R.E. Koeppe II, O.S. Andersen, Engineering the gram-
icidin channel, Annu. Rev. Biophys. Biomol. Struct. 25
(1996) 231–258.
w18x R. Sarges, B. Witkop, The structure of valine- and
isoleucine-gramicidin A, J. Am. Chem. Soc. 87 (1965)
2011–2019.
w19x A.S. Arseniev, I.L. Barsukov, A.L. Bystrov, V.L. Lonize,
Y.A. Ovchinnikov, Proton NMR study of gramicidin A
transmembrane ion channel. Head-to-head right handed,
single stranded helices, FEBS Lett. 186 (1985)
168–174.
w20x R.R. Ketchem, W. Hu, T.A. Cross, High resolution of
gramicidin A in a lipid bilayer by solid-state NMR,
Science 261 (1993) 1457–1460.
w21x D.W. Urry, Gramicidin A transmembrane channel: a
proposed p(L,D) helix, Proc. Nat. Acad. Sci. USA 68
(1971) 672–676.
w22x A. Finkelstein, O.S. Andersen, The gramicidin A chan-
nel: a review of its permeability characteristics with
special reference to the single-file aspect of transport,
J. Membr. Biol. 39 (1981) 155–171.
w23x D.G. Levitt, S.R. Elias, J.M. Hautman, Number of water
molecules coupled to the transport of sodium, potassium
and hydrogen ions via gramicidin nonactin or valino-
mycin, Biochem. Biophys. Acta 512 (1978) 436–451.
w24x S. Cukierman, E.P. Quigley, D.S. Crumrine, Proton
Conduction in gramicidin A and in its dioxolane-linked
dimer in different lipid bilayers, Biophys. J. 73 (1997)
2489–2502.
w25x E.P. Quigley, P. Quigley, D.S. Crumrine, S. Cukierman,
The conduction of protons in different stereoisomers of
dioxolane-linked gramicidin A channels, Biophys. J. 77
(1999) 2479–2491.
w26x C.J. Stankovic, S.H. Heinemann, J.M. Delfino, F.J.
Sigworth, S.L. Schreiber, Transmembrane channels
based on tartaric acid-gramicidin A hybrids, Science
244 (1989) 813–817.
w27x K.M. Armstrong, E.P. Quigley, P. Quigley, D.S. Crum-
rine, S. Cukierman, Covalently linked gramicidin chan-
nels: effects of linker hydrophobicity and alkaline metals
on different stereoisomers, Biophys. J. 80 (2001)
1810–1818.
w28x S. Cukierman, Flying protons in linked gramicidin A
channels, Isr. J. Chem. 39 (1999) 419–426.
w29x S. Cukierman, Proton mobilities in water and in different
stereoisomers of covalently linked gramicidin A chan-
nels, Biophys. J. 78 (2000) 1825–1834.
w30x C.M.G. Godoy, S. Cukierman, Modulation of proton
transfer in the water wire of dioxolane-linked gramicidin
channels by lipid membranes, Biophys. J. 81 (2001)
1430–1438.
w31x B.E. Conway, J.O.M. Bockris, H. Linton, Proton con-
ductance and the existence of the H O ion, J. Chem.q3
Phys. 24 (1956) 834–852.
w32x T. Erdey-Gruz, Transport Phenomena in Aqueous Solu-´
tions, John Wiley, New York, 1974.
w33x E. Grunwald, C.F. Jumper, S. Meiboom, Kinetics of
proton transfer in methanol and the mechanism of the
abnormal conductance of the hydrogen ion, J. Am.
Chem. Soc. 84 (1962) 4664–4671.
w34x E.P. Quigley, A. Emerick, D.S. Crumrine, S. Cukierman,
Proton current attenuation by methanol in a dioxolane-
linked gramicidin A dimer in different lipid bilayers,
Biophys. J. 5 (1998) 2811–2820.
w35x R.C. Weast, Handbook of Chemistry and Physics, CRC
Press, Boca Raton, 1989.

190 A. Chernyshev et al. / Biophysical Chemistry 103 (2003) 179–190
w36x O.E. Frivold, O. Hassel, E. Hetland, The electric con-
ductivity of solutions of HCl in ordinary water and of
DCl in heavy water, Avh. Norske Videnskaps-Akademi
I. Mat.-Naturv 9 (1943) 3–11.
w37x W.N. Baker, V.K. La Mer, The conductance of potassium
chloride and of hydrochloric–deuterochloric acid in
H O–D O mixture. The viscosity of H O–D O, Chem.2 2 2 2
Phys. 3 (1935) 406–410.
w38x G.N. Lewis, T.C. Doody, The mobility of ions in H H O,2 2
J. Am. Chem. Soc. 55 (1933) 3504–3506.
w39x L.G. Longsworth, D.A. MacInnes, Transference num-
bers and ion mobilities of some electrolytes in deuterium
oxide and its mixtures with water, J. Am. Chem. Soc
59 (1937) 1666–1670.
w40x D. Eisenberg, W. Kauzmann, The Structure and Prop-
erties of Water, Oxford University Press, New York,
1969.
w41x C.G. Swain, D.F. Evans, Conduction of ions in light
and heavy water at 25 8C, J. Am. Chem. Soc. 88 (1966)
383–390.
w42x R.H. Tredgold, R. Jones, A study of gramicidin using
deuterium oxide, Biochim. Biophys. Acta 550 (1979)
543–545.
w43x S. Lengyel, J. Giber, J. Tamas, Determination of ionic´
mobilities in aqueous hydrochloric acid solutions of
different concentration at various temperatures, Acta
Chim. Hung. 32 (1962) 429–436.
w44x D.A. Lown, H.R. Thirsk, Proton transfer conduction in
aqueous solution. Part 1. Conductance of concentrated
aqueous alkali metal hydroxide solutions at elevated
temperatures and pressures, Trans. Faraday Soc. 67
(1971) 132–148.
w45x M. Tuckerman, K. Laasonen, M. Sprik, M. Parinello,
Ab initio molecular dynamics simulation of the solva-
tion and transport of H O and OH ions in water, J.q y3
Phys. Chem. 99 (1995) 5749–5752.
w46x R. Vuilleumier, D. Borgis, Quantum dynamics of an
excess proton in water using an extended empirical
valence-bond Hamiltonian, J. Phys. Chem. B 102
(1998) 4261–4264.
w47x R. Vuilleumier, D. Borgis, Transport and spectroscopy
of the hydrated proton: a molecular dynamics study, J.
Chem. Phys. 111 (1999) 4251–4266.
w48x N. Agmon, Structure of concentrated HCl solutions, J.
Phys. Chem. A 102 (1998) 192–199.
w49x Y. Kameda, O. Uemura, The intramolecular structure of
oxonium ion in concentrated aqueous deuterochloric
acid solutions, Bull. Chem. Soc. Jpn. 65 (1992)
2021–2028.
w50x M. Akeson, D.W. Deamer, Proton conductance by the
gramicidin water wire. Model for proton conductance
in the F F ATPases?, Biophys. J. 60 (1991) 101–109.0 1
w51x O.S. Andersen, Ion movement through gramicidin A
channels. Studies on the diffusion-controlled association
step, Biophys. J. 41 (1983) 147–165.
w52x A. Chernyshev, S. Cukierman, Thermodynamoc view of
activation energies of proton transfer in various grami-
cidin A channels, Biophys. J. 82 (2002) 182–192.
w53x Shedlovsky. The Structure of Electrolyte Solutions, W.J.
Hamer, New York: John Wiley, 1959, Chapter 17..
w54x L.S. Guss, I.M. Kolthoff, The distribution of protons
between water and other solvents, J. Am. Chem. Soc
62 (1940) 1494–1496.
w55x N. Agmon, S.Y. Goldberg, D. Huppert, Salt effect on
transient proton transfer to solvent and microscopic
proton mobility, J. Mol. Liquids 64 (1995) 161–195.