Ion channel topography of the neuronal nicotinic acetylcholine receptor: Pharmacochemical approaches

Article abstract of DOI:10.1023/A:1015017407412

Forty-three bisammonium ganglionic blockers were synthesized to study the structure of the ion channel of nicotinic acetylcholine receptor. The conformational parameters of these blockers were studied, and their effects toward the ganglionic transmission in situ on the sympathetic feline superior cervical ganglia and in vitro on the parasympathetic guinea-pig small intestine ganglia were determined. A model of the binding site for the bisammonium ganglionic blockers in the neuronal ion channel was proposed.

Full text of DOI:10.1023/A:1015017407412

Russian Journal of Bioorganic Chemistry, Vol. 28, No. 2, 2002, pp. 116–127. Translated from Bioorganicheskaya Khimiya, Vol. 28, No. 2, 2002, pp. 135–146.
Original Russian Text Copyright © 2002 by Gmiro, Brovtsyna, Serdyuk, Lukomskaya.
Ion Channel Topography of the Neuronal Nicotinic Acetylcholine
Receptor: Pharmacochemical Approaches
1
V. E. Gmiro*, N. B. Brovtsyna**, S. E. Serdyuk***, and N. Ya. Lukomskaya**
*Institute of Experimental Medicine, Russian Academy of Medical Sciences,
Kamennoostrovskii pr. 69/71, St. Petersburg, 197022 Russia
**Sechenov Institute of Evolutionary Physiology and Biochemistry, Russian Academy of Sciences,
pr. Morisa Toreza 44, St. Petersburg, 194223 Russia
***Donetsk Medical University, pr. Il’icha 16, Donetsk, 340003 Ukraine
Received February 5, 2001; in final form, March 16, 2001
Abstract—Forty-three bisammonium ganglionic blockers were synthesized to study the structure of the ion
channel of nicotinic acetylcholine receptor. The conformational parameters of these blockers were studied, and
their effects toward the ganglionic transmission in situ on the sympathetic feline superior cervical ganglia and
in vitro on the parasympathetic guinea-pig small intestine ganglia were determined. A model of the binding site
for the bisammonium ganglionic blockers in the neuronal ion channel was proposed.
Key words: conformational analysis, ganglionic blockers, ion channel, neuronal nicotinic acetylcholine recep-
tor, parasympathetic ganglion, structure–activity relationship, sympathetic ganglion
1
INTRODUCTION
logue to another [1]. However, within the type (2)
+
series, a minimal change in the structure of R1R2R3N
A successful search for new medicines depends
largely on information about the biological target
affected by the drug. In particular, a receptor binding
site, an ion channel, an acetylcholine receptive mem-
brane, or intracellular structures can be such targets for
the compounds blocking the transmission of nervous
impulses through autonomic ganglia.
+
group, for example, a replacement of Et2MeN by a
+
Et3N group, may result in an abrupt change in activity
+
[3]. The authors speculated that, unlike the Et3N group,
+
a more bulky Et2MeN group cannot penetrate into the
For ganglion blockers of the general formula
channel and, hence, the blocking activity of the type (2)
+
+
[Me3N–(CH2)5–NR1R2R3] · 2Br–,
(1)
compounds is controlled by the steric parameters of the
+
varied R1R2R3N group [4]. If this group can penetrate
a close correlation between the ganglion-blocking
effect and the channel-blocking activity was found [1].
Hence, the ganglion-blocking activity can be used to
into the channel, a possibility of the formation of a
tighter double-point fixation of the bis-cationic mole-
determine the channel-blocking activity. This approach cule arises, and the blocking activity of the compound
was successfully applied in the study of specific struc-
tural features of the channel of neuronal nicotine ace-
tylcholine receptor using compounds of the general for-
mula
is dramatically increased.
Thus, there is the virtual possibility of using the data
on the structure–activity relationship in the series of
ganglion blockers for the evaluation of the specific
structural features of the biological target affected by
them, namely, a neuronal ion channel of the nicotinic
acetylcholine receptor. This was the goal of our work.
+
+
[Et3N–(CH2)5–NR1R2R3] · 2Br–.
(2)
In this case, the results of the analysis of the ganglion-
blocking activity [2] agree well with the data on their
channel-blocking activity [3].
RESULTS AND DISCUSSION
Among type (1) compounds, the ganglion-blocking
activity gradually changes when turning one homo-
Compounds (1) and (2) were synthesized according
to the following general scheme:
1
Please address correspondence to g2119@online.ru.
1068-1620/02/2802-0116$27.00 © 2002 MAIK “Nauka /Interperiodica”

ION CHANNEL TOPOGRAPHY
117
+
+
Table 1. The properties and activity of compounds of the general formula [Et3N–(CH2)5–NR1R2R3] · 2Br– (2)
Activity***
+
Population,
ΣPl, %**
toward para-
sympathetic
ganglion
–NR1R2R3
Number
Empirical formula Yield, %*
Mp, °C
toward sympa-
thetic ganglion
+
(I)
C15H36Br2N2
C14H34Br2N2
51
46
257–259
263–264
100
100
7.70
5.60
25.80
16.60
–NMe2Et
+
–NMe3
(II)
+
N
(III)
C16H36Br2N2
40
256–257
95.5
3.70
11.0
Me
+
N
O
(IV)
(V)
C16H36Br2N2O
C16H38Br2N2
C17H38Br2N2
49
50
79
246–247
256–257
257–258
92.4
54
3.20
3.10
1.50
16.20
16.90
23.20
Me
+
–NMeEt2
+
N
(VI)
49.4
Et
+
N
O
(VII)
C17H38Br2N2O
38
242–243
47.3
1.30
6.90
Et
+
N
(VIII)
(IX)
(X)
C18H40Br2N2
C16H38Br2N2
C18H40Br2N2
C17H40Br2N2
48
50
48
72
256–257
179–180
249–250
257–258
8.1
0.16
0.20
0.17
0.12
0.42
0.39
1.24
0.22
Pri
+
–NMe2Prn
6.0
+
N
0.00
0.00
Et
+
(XI)
–NEt3
N+
(XII)
C18H38Br2N2
C18H40Br2N2
56
60
261–262
202–204
0.00
0.00
0.08
0.07
0.32
0.18
+
N
(XIII)
Prn
– +SMe2
(XIV)
(XV)
C13H31Br2NS
C11H28Br2N2
C13H32Br2N2
C15H36Br2N2
20
36
50
67
156–157
181–183
176–178
159–160
100
100
100
79.6
1.60
0.70
1.80
1.10
7.30
1.80
6.20
4.40
+
–NH3
+
–NHMe2
+
(XVI)
(XVII)
–NHEt2
+
HN
(XVIII)
C15H34Br2N2
47
156–157
100
1.10
4.00
+
–NH2But
(XIX)
(XX)
C15H36Br2N2
C15H36Br2N2
57
56
229–230
184–185
100
1.10
1.00
1.27
2.90
+
–NHMePri
98.5
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

118
GMIRO et al.
Table 1. (Contd.)
Activity***
+
Population,
ΣPl, %**
toward para-
sympathetic
ganglion
–NR1R2R3
Number
Empirical formula Yield, %*
Mp, °C
toward sympa-
thetic ganglion
+
–NH2Pri
(XXI)
C14H34Br2N2
C17H40Br2N2
C16H38Br2N2
69
27
43
196–197
209–210
219–220
100
0.60
0.18
0.11
0.97
1.05
0.38
+
–NHPri2
(XXII)
(XXIII)
41.6
87.2
+
–NHMeBut
NH
Me2
(XXIV)
C13H32Br2N2O
57
167–168
–
0.01
0.03
O
* The yield of the analytically pure product from starting 5-bromopentyltriethylammonium bromide is shown.
** An overall population of productive conformations (for a section 8.3 × 6.1 Å in size).
*** Activity (relative to that of Hexamethonium bromide, which is taken as 1). The Hexamethonium bromide activity on the feline upper
cervical sympathetic ganglion corresponded to ED = 0.035 µM and on the guinea-pig small intestine parasympathetic ganglion to K
50
d
4.5 µM.
+
+
+
R1R2R3N
R3N + Br–(CH2)5–Br
(3) ‡) R = Me (4)
[R3N–(CH2)5–Br] Br–
[R3N–(CH2)5–NR1R2R3] 2Br–
(5) ‡) R = Me
·) R = Et
(1) R = Me
(2) R = Et
·) R = Et
We developed a short two-step synthetic scheme of of 8.3 × 6.1 Å [7]. Similar parameters were also
the target compounds. A proper solvent ensured the obtained for the case when isolated neurons were used
precipitation of monoquaternary ammonium salts (5), in the calculations; their channel-forming activity was
as it formed, and, therefore, mono rather than bisqua- determined according to the inhibition of the trans-
ternary derivatives were formed at the first stage.
membrane current resulting from the ionophoretic
application of acetylcholine [3]. The analysis of the
structure–activity relationship for another series of the
type (I) bications allowed us to postulate the presence
of another stricture, narrower than the first, in the ion
channel pore [8].
The properties and the pharmacological activity of
(1) and (2) are given in Tables 1 and 2. Activities of all
the studied compounds are expressed in relative units,
with ÖD50 of Hexamethonium {hexane-1,6-bis(trime-
thylammonium) bromide, a reference ganglion blocker
[5]} being taken as 1.0. The procedure for the determi-
nation of the ganglion blocking activity is described in
the EXPERIMENTAL section.
+
Varying R1R2R3N groups in descending order of
their populations in productive conformations, which
reflect the ability of the group within the bicationic
molecule to penetrate into the stricture 8.3 × 6.1 Å in
size, are plotted in Fig. 1 in regular abscissa intervals.
As the Y-axis, relative values of the ganglion-blocking
activity are plotted, with the activity of (I) being taken
as 100%. One can see that generally a population
decrease (i.e., the ability to penetrate into the stricture
of the mentioned size) is accompanied by a decrease in
Since (1) and (2) have a flexible molecular structure
and can exist in numerous conformations, conforma-
tional computing was performed according to the pro-
cedure described in [6]. For the analysis of the struc-
ture–activity relationship, stable conformers were
sorted out, whose energy did not exceed 3 kcal/mol.
The parameters of the stricture in the ion channel pore
were determined by the correlation between the popu-
lation of productive conformations (i.e., the ability of
molecules to pass through the channel cross section
approximated with a rectangle) and the compound
activity. The computing demonstrated that the maxi-
mum correlation coefficients (0.97 for sympathetic and
+
+
activity. For example, (XI) (R1R2R3N = Et3N) is prac-
tically inactive, which completely correlates with the
fact that it does not penetrate into the stricture (its pop-
ulation is 0%). If the conformational flexibility of two
+
ethyl radicals in the Et3N group of this compound is
0.80 for parasympathetic ganglia) correspond to the restricted by closing a cycle to give (VI), the population
rectangular profile of the stricture with rectangle sides will be increased from 0 to 49.4% and the activities
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

ION CHANNEL TOPOGRAPHY
119
+
+
Table 2. The properties and activity of compounds of the general formula [Me3N–(CH2)5–NR1R2R3] · 2Br– (1)
Activity***
+
Yield,
%*
Number
–NR1R2R3
Empirical formula
Mp, °C
**
ΣP1
***
ΣP2
toward
toward para-
sympathetic sympathetic
ganglion
ganglion
+
(XXV)
C8H22Br2N2
C9H24Br2N2
C10H26Br2N2
C10H26Br2N2
C11H28Br2N2
C12H30Br2N2
C12H30Br2N2
43
53
80
51
65
33
70
208–209 100
185–187 100
201–203 100
169–171 100
191–192 100
221–223 100
0.00
0.00
0.03
0.05
–NH3
+
(XXVI)
(XXVII)
(XXVIII)
(XXIX)
(XXX)
0.20
0.44
–NH2Me
+
33.8
57.2
100
100
79.6 100
199–200 100 100
1.00
1.22
–NHMe2
+
2.70
3.20
–NH2Et
+
–NH2Pri
7.70
22.40
26.90
19.60
+
–NH2But
4.20
+
(XXXI)
221–223
3.70
–NHEt2
+
HN
(XXXII)
C12H28Br2N2
48
4.00
–
+
–NHMePri
(XXXIII)
(XXXIV)
(XXXV)
C12H30Br2N2
C12H30Br2N2
C14H34Br2N2
77
72
54
210–212
252–253
222–223
98.5 100
87.2 100
41.6 100
7.70
5.50
9.40
13.90
20.00
31.00
+
–NHMeBut
+
–NHPri2
+
(XXXVI)
C14H32Br2N2
62
225–226
32.9 100
4.20
11.80
NH2
+
(XXXVII)
(XXXVIII)
C11H28Br2N2
C12H30Br2N2
74
50
288–289 100
269–271 100
100
100
1.40
1.90
1.60
7.10
–NMe3
+
–NMe2Et
+
N
(XXXIX)
C13H30Br2N2
79
271–272
95.5 100
4.00
6.30
Me
+
(XL)
C13H32Br2N2
C13H32Br2N2
68
55
250–252
263–265
50.4 100
12.6 100
2.60
3.90
9.40
6.00
–NMeEt2
+
–NMe2Pri
(XLI)
Me
+N CH2
Me
(XLII)
C17H32Br2N2
61
41
189–190
158–160
15.6 100
0.8
1.1
4.00
20.0
Pri
NHMeP
(XLIII)
C12H30Br2N2O
–
–
O
* The yield of analytically pure product from starting 5-bromopentyltriethylammonium bromide is shown.
** An overall population of productive conformations for a section of 8.3 × 6.1 Å in size. Productive conformations are those penetrating
through a section of 8.3 × 6.1 Å in size.
*** Activity (relative to that of Hexamethonium bromide, which is taken as 1). The Hexamethonium bromide activity on the feline upper cer-
vical sympathetic ganglion corresponded to ED = 0.035 µM and in guinea-pig small intestine parasympathetic ganglion to K 4.5 µM.
50
d
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

120
GMIRO et al.
100
90
80
70
60
50
40
30
20
10
0
3
1
2
(I)
(II)
+
(III)
+
–N
(IV)
(V)
(VI) (VII) (VIII) (IX)
(X)
(XI) (XII) (XIII)
+
+
+
+
+
–N
+
+
+
–N
+
+
i
–NMe2Et –NMe3
–NMeEt2
–NMe2Pr
–NEt3
–N
O
–N
–N
O
–N
–N+
Me
Me
Et
Et
Pri
Et
Pr n
Fig. 1. The correlation between (curve 1) the relative population of productive conformations of the compounds of the general for-
+
+
1
2
3
–
mula [Et N–(CH ) –NR R R ] · 2Br and (curve 2) their relative blocking activity toward the sympathetic and(curve 3) parasym-
3
2 5
+
1
2 3
pathetic ganglia. The population and activity of (I) are taken as 100%. The structures of the NR R R groups are given below the
plot.
toward sympathetic and parasympathetic ganglions
will be increased 13- and 100-fold, respectively. The
addition of one ëç2 unit into the N-ethylpyrrolidinium
group of (VI) by the replacement of the ethyl with i-Pr
(VIII) or n-Pr (XIII) group or by extending the cycle
from a five- to a six-membered cycle in (X) caused, in
The activity curve for parasympathetic ganglia
(Fig. 1) correlates with the population curve to a lesser
extent than for sympathetic ganglia and demonstrates
an activity growth in the series of compounds from (IV)
to (VI) instead of the expected decrease, with (VI)
being the most active. This specific feature probably
all cases, a decrease in both population (from 49.4 to involves the “trap” effect, recently demonstrated for
8.1, 0, and 0%, respectively) and activity (9-, 21-, and (VI) [9].
10-fold for sympathetic and 56-, 60-, and 18-fold for
In addition to the compounds considered above, we
parasympathetic ganglia, respectively). A minimally
analyzed the properties and activities of (XV)–(XXIII),
+
possible reduction of Et3N group (by one ëç2 unit) in
(XI), i.e., the transition to analogue (V) with the
the compounds that lack quaternary ammonium group
in the varied amino group, to elucidate whether the qua-
ternary structure (and, hence, a stable positive charge of
the nitrogen atom) is obligatory for displaying a high
ganglion-blocking activity by a substance. Despite the
formal similarity in the (+)-charge on the nitrogen
atom, quaternary ammonium and nonquaternary amino
+
Et2MeN group results in the growth of population from
0 to 54% and a 26- and 82.5-fold increase in the activity
toward the sympathetic and parasympathetic ganglia,
respectively.
The nonpenetrating N-ethylpiperidinium group in
(X) becomes a penetrating one when one cyclic meth-
ylene unit is replaced by oxygen [see (VII)], i.e., by the
replacement of piperidine by the morpholine cycle.
Oxygen can affect binding (e.g., due to the formation of
a hydrogen bond); however, conformational calcula-
tions clearly indicate that the steric factor is decisive:
the population is increased from 0% for (X) to 47.3%
for (VII), and this is accompanied by a 5–8-fold growth
in the activity toward the ganglia of both types.
groups are fundamentally different. In ammonium
+
groups of type R3NH containing a nonquaternary nitro-
gen atom, positive charge on the nitrogen atom is
formed due to the nitrogen protonation, the extent of
which depends on the basic properties of the amino
group: at physiological pH value (7.4) and the basicity
of amino groups of the compounds studied (pKb of
about 10–12), the equilibrium between the nonproto-
+
nated (R3N) and protonated (R3NH) forms is shifted
The data presented in Table 1 and Fig. 1 indicate that
the general factor that determines the activity of the
type (2) compounds is steric parameters of their mole-
cules, which can be quantitatively deduced from the
population of their productive conformations.
toward the latter form by more than 99% [10]. In the
case of a quaternary ammonium group of the general
+
formula R4N, all four C–N bonds are covalent, which
excludes the dissociation and formation of nonproto-
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

ION CHANNEL TOPOGRAPHY
121
+
nated tertiary amine R3N. This is the cause of the dra-
matic difference in the biological properties of quater-
nary group-containing and quaternary group-lacking
amines, since the stable positive charge of the quater-
nary ammonium group hinders the molecule penetra-
tion through the blood-brain barrier and biological
membranes, whereas the molecules with nonquaternary
ammonium groups can pass through biological mem-
branes in the form of uncharged (deprotonated) amine
[11].
blockers of the åÂ3N series pass through the first stric-
ture (population 100%). To explain this fact, we previ-
ously presumed that the ion channel had the second
stricture, more narrow than the first one, which steri-
cally hinders the penetration of these bications [8]. In
this case, the harder the varied cationic group penetra-
tion through the second stricture, the higher the block-
ing activity of this series of compounds. The parame-
ters of the channel stricture were calculated, for which
the correlation coefficient between the population of
penetrating conformations and the value inverse to
activity value was the highest. The highest correlation
coefficients, 0.91 for sympathetic and 0.96 for para-
sympathetic ganglia, corresponded to the stricture of
the rectangular profile of 6.4 × 5.5 Å in size [13].
One can see from Table 1 that the replacement by
hydrogen of one of three alkyl substituents in the qua-
ternary nitrogen atom always results in a decrease in
activity. It is probable that the second nucleophilic
binding site, situated deeper in the channel and binding
+
the penetrating R1R2R3N groups (the first site binds the
+
The compounds of the Me3N series of the general
+
+
+
nonpenetrating Et3N group) fixes stronger quaternary
formula (1) with H3N- and MeNH2 groups (XXV) and
(XXVI) are low active due to insufficient hydrophobic-
ity preventing an effective binding (the population of
productive, i.e., nonpenetrating conformations is 0%).
and weaker nonquaternary ammonium groups. This
may imply that the hydrogen bond with the nucleo-
philic site, in principle possible for the nonquaternary
analogues bearing the protonated nitrogen, does not
considerably contribute to the binding by the second
nucleophilic site. The fact that the activities of bications
+
+
The analogues bearing Me2NH (XXVII) and EtNH2
groups (XXVIII) were substantially more active than
(XXV) and (XXVI), since some of their conformers
cannot pass through the second stricture (population of
+
+
containing åÂ2NH- and Me2S groups are practically
the same, although the formation of hydrogen bond is
possible only in the first case [cf. (XVI) and (XIV) in
Table 1] supports this statement. The removal of charge
from the nitrogen atom by N-oxide formation
33.8 and 57.2%, respectively). Other compounds of the
+
Me3N series (1) are completely retarded before the sec-
ond stricture (population 100%) and displayed a rela-
tively high activity [except for (XLII) containing a
dimethylbenzylammonium group, whose configuration
may not be complementary to the surroundings of the
anionic binding site]. Some differences in the activities
of these compounds may be due to such factors as
hydrophobicity and steric correspondence to the bind-
ing sites.
[(XVI)
(XXIV)] causes the disappearance of
activity (pKa values for N-oxides of aliphatic tertiary
amines are within the range of 4–5 [12]; i.e., they are
virtually nonprotonated and lack a positive charge in
aqueous solutions).
Two factors may determine the presence of the sec-
ond positive charge as a condition for a high activity of
ganglion blockers. First, a possibility of the more effi-
cient two-point fixation arises. Second, the repulsion of
two positive charges results in the completely extended
conformation of the polymethylene chain between
nitrogen atoms and this contributes to the two-point fix-
ation at two nucleophilic sites.
+
As mentioned above, (XXIV) containing both Et3N
and Me2NO groups in one molecule is practically inac-
tive because of the lack of a positive charge on the N-
oxide group that possesses a low pKa value. In this con-
nection, note that the replacement of the positively
+
+
Of the bications of the general formula (2) with a
charged PriåÂNH group in the åÂ3N series com-
+
+
nonquaternary ammonium group, (XVI) with åe2NH
pounds, unlike those of the Et3N series, by an
uncharged PriMeNO (XXXIII)
group is most active, because it seems to be most com-
plementary to the surroundings of the second binding
site. The correlation coefficient of this group of com-
pounds was the highest for the cross section of 8.3 ×
6.1 Å: it is equal to 0.91 for sympathetic and 0.80 for
parasympathetic ganglia, which implies a crucial role
of steric factors for blockers with both quaternary and
nonquaternary ammonium groups.
(XLIII) does not
cause a complete loss of the blocking activity: the activ-
ity decreases 7-fold for sympathetic ganglia (7.7 to 1.1)
and remained approximately equal to that for Hexame-
thonium and was even increased 1.4-fold (from 13.9 to
20.0) for parasympathetic ganglia (Table 2). This may
+
be explained by the ability of the åÂ3N group in
+
A high ganglion-blocking activity is characteristic
(XLIII), unlike the Et3N group in (XXIV), to penetrate
+
+
through the first stricture and come to the second stric-
ture where it ensures a rather tight binding to the corre-
of bications of both Et3N- and åÂ3N series (1) (Table 2),
although both ammonium groups of many highly active
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

122
GMIRO et al.
on the calculation of the correlation coefficients
(‡)
(b)
between the population of productive conformations
and ganglion-blocking activity allowed the conclusion
that the ion channel of the neuronal nicotinic acetylcho-
line receptor has a funnel shape and has two narrow
spots in the ganglion blocker binding sites that are
approximated as rectangles of 8.3 × 6.1 Å and 6.4 ×
5.5 Å in size (Fig. 2).
It is interesting to compare our data with those
obtained by other researchers. It was shown that ion
channels of nicotinic acetylcholine receptors of fish
electric organ and frog muscle are funnel-shaped [14–
18]. The narrow diameter of the muscular acetylcholine
receptor channel, calculated according to the ability of
organic cations to penetrate through the channel (with-
out taking into account their conformations), was esti-
mated at 8.4 Å [19].
+
I
I
I
+
I
–
+
–
–
8.3 × 6.1 Å
–
8.3 × 6.1 Å
6.4 × 5.5 Å
7.5 Å
7.5 Å
II
II
II
+
–
II
–
6.4 × 5.5 Å
–
–
(c)
(d)
+
I
I
–
I
–
+
I
8.3 × 6.1 Å
–
–
8.3 × 6.1 Å
7.5 Å
Nicotinic acetylcholine receptors in neurons of rat
upper cervical sympathetic ganglia consist of α3, α5,
α7, β2, β4 [20, 21], and α4-1 subunits [22] forming
various receptor subtypes. Subunits α3 and β4 are most
important for neurons of this type [23]. The presence of
α3, α5, and α7 subunits in the rat superior cervical gan-
glion was recently shown by immunochemical methods
[24].
7.5 Å
II
II
+
+
–
–
II
6.4 × 5.5 Å
II
+
–
–
6.4 × 5.5 Å
Fig. 2. The binding models of (a, b, c, and d) variants of
Studies using methods of photolabeling and
mutagenesis led to the conclusion that three rings
formed by the amino acid residues of M2 segments of
different subunits responsible for cation binding inside
the channel are present in the nicotinic acetylcholine
receptor [25, 29]. In these rings, the corresponding
amino acid residues are situated in the same positions
and exposed inside the channels. The term rings has
been applied due to the fact that, in the cross section
plane perpendicular to the channel axis, similar amino
acids of different subunits are localized along the ring.
It is assumed that these rings play a decisive role in the
interaction with bivalent cations [15, 25–27] and block-
ers of the open channel [28, 29]. For the homogeneous
α7 receptor with the determined pentamer structure
[31], amino acid residues forming its selective filter
were identified [30].
Previously, we found that two populations of nico-
tinic acetylcholine receptors are present in the rat supe-
rior cervical ganglion, whose constants of blocker bind-
ing to the open channel differ fourfold [32]. Apparently,
this fact reflects a difference in the subunit composition
of nicotinic receptors of two types localized in the same
neuron [15].
+
bisammonium blockers of the general formulae [Me N–
3
+
+
1
2
3
–
(CH ) –NR R R ]
· 2Br (1) and [Et N–(CH ) –
2 5
3 2 5
+
1
2
3
–
NR R R ] · 2Br (2) with nucleophilic sites inside the ion
channel of the neuronal nicotinic acetylcholine receptor: (a)
the intrachannel orientation of blocker (2) with a constant
+
Et N group retaining in the area of the first constriction
3
(designated as I) and a varied cationic group capable of pen-
etration to the second constriction of the ion channel (des-
ignated as II); (b) the orientation of blocker (2) with a con-
+
stant Et N group and with the second cation group either
3
+
equal or exceeding in size the Et N group; (c), two types of
3
the possible orientation of molecule (1) inside the ion chan-
+
nel with the constant Me N group located in the area of the
3
second constriction and the varied cation group, lower in
+
size than Me N, capable of either penetrating through the
3
second constriction or retained in the region of the first con-
striction; (d), two types of the possible orientation of mole-
+
cule (1) inside the ion channel with the constant å N
3
group, both cation groups of which are bigger in size than
that of the second constriction.
sponding channel site. A relatively high activity of
(XLIII) toward ganglia of both types implies their
binding with nucleophilic sites of both strictures. The
selective action of (XLIII) toward the parasympathetic
ganglia may be due to the specific binding in the very
region of the first stricture.
Numerous experiments showed that amino acid res-
idues localized in the M2 segment of nicotinic acetyl-
choline receptors are responsible for various pharmaco-
logical and functional properties, such as binding to
noncompetitive blockers, desensitization, and ion chan-
nel conductivity [25–29]. Amino acid sequences of M2
segments of some subunits of nicotinic acetylcholine
receptors are shown in Table 3. The M2 segment has an
Thus, analysis of the structure–activity relationship
of the compounds of general formulas (1) and (2) based α-helical structure (3.6 aa per one turn). Amino acid
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

ION CHANNEL TOPOGRAPHY
123
Table 3. Amino acid sequences of M2 segments of nicotinic acetylcholine receptors [20–23]
(α Torp.Cal.)* 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261
Refe-
rences
Subunits
α TC
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21
E
E
Q
E
E
E
Q
E
E
E
E
E
E
E
E
E
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
K
M
M
C
M
M
M
C
T
T
S
L
L
L
T
L
L
V
V
L
L
L
L
L
L
L
L
S
I
S
S
S
S
S
F
N
S
S
S
S
S
T
S
S
S
V
A
V
V
V
A
V
V
V
V
V
V
V
V
V
V
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
V
L
L
V
L
S
L
V
Q
Q
L
L
Q
Q
L
L
L
L
L
L
L
L
T
T
T
A
T
T
T
S
V
V
I
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
L
L
L
L
L
L
L
L
L
L
L
L
M
L
L
L
L
L
F
L
L
L
F
L
L
L
L
L
L
L
L
L
V
L
L
L
V
L
L
L
L
V
L
V
L
L
V
L
I
V [19, 25]
A [19, 25]
A [19, 25]
β TC
S
I
A
A
A
S
L
I
γ TC
T
S
S
I
δ TC
A
S
I
V
V
V
V
V
V
V
V
V
V
V
V
F
T
I
S
[19, 25]
[19]
[19]
[19]
[19]
[38]
[38]
[38]
[38]
[39]
[38]
[38]
[38]
α mouse
β mouse
γ mouse
δ mouse
α2 rat
α3 rat
α4 rat
α5 rat
α7 rat
β2 rat
β3 rat
β4 rat
T
G
T
S
I
V
A
A
S
S
I
T
A
A
S
L
V
I
A
A
C
C
C
C
G
C
S
T
I
I
T
T
T
S
I
T
T
T
T
T
T
T
T
I
T
T
T
E
A
S
V
V
I
I
S
I
I
S
I
T
I
S
I
I
S
S
V
I
M
L
T
S
I
A
S
T
I
I
E
S
M
T
C
A
I
* At the bottom line, the numbers of amino acid residues 1–2l in M2 segments of the of α-subunit of nicotine acetylcholine receptor
sequence of Torpedo californica (TC) electric organ are given.Amino acid residues forming the threonine, serine, inner leucine, and outer
leucine rings (positions 4, 8, 11, 15, and 18) are given in bold (vertical lines).
** Subunits α, β, γ, and δ comprise the muscle nicotinic acetylcholine receptor (studied for Torpedo californica [19, 25] and for mice [19]);
subunits α2–α7 and β2–β4 comprise the neuronal nicotinic acetylcholine receptors (studied for rats [38, 39]).
residues in positions 4, 8, 11, 15, and 18 determining those obtained in this work, and found a close correla-
the ion conductivity and binding of noncompetitive tion between these two models (Fig. 3). The energy
blockers form the corresponding rings: threonine, profile for our bicationic blockers passing through the
serine, equatorial leucine, valine, and outer leucine channel region where the threonine and serine rings are
(Table 3). A combination of mutagenesis methods with situated had several minima that demonstrated ligand–
experiments on binding of competitive blockers in the receptor interaction. The global minimum corre-
region of the threonine and serine rings showed that the sponded to the state in which both cationic groups of
threonine ring is the narrowest part of the channel [26, the blocker simultaneously interacted with the serine
33, 34].
These facts allowed us to postulate that it is the thre-
onine ring that corresponds to the narrowest stricture in
and threonine rings [36]. Binding of this type is shown
in Fig. 4.
Thus, the proposed topological model of binding
our model (6.4 × 5.5 Å). The closeness of the distances sites for the bicationic blockers based on the analysis of
(7.2 Å) between the nitrogen atoms in the tested com- conformation–biological activity relationship is in
pounds and between the neighboring turns of the α good agreement with the data on the structure and func-
helix suggests that the serine ring corresponds to the tion of the nicotinic acetylcholine receptor [20, 27].
larger stricture in our model (8.3 × 6.1 Å). Oxygen The model sets the following demands of the effective
atoms of hydroxyl groups in the threonine and serine noncompetitive blockage of neuronal nicotinic acetyl-
rings serve as nucleophilic binding sites for the blocker choline receptors: (1) the blocker must meet steric hin-
cation groups.
drances that impede its further penetration into the
channel in the region of the serine or threonine rings;
and (2) the blocker must be rather complementary to
the binding sites to enable effective hydrophobic and
ion-dipole interactions of both ammonium groups with
the corresponding nucleophilic sites of channel.
The data on the binding sites of the channel blocker
in the ion channel of acetylcholine receptor were the
basis for the design of a hypothetic three-dimensional
model of the ion channel of nicotinic acetylcholine
receptor of the fish Torpedo californica electric organ
[35]. We reproduced this model, compared the sizes of
The approach used in this work opens some possi-
the strictures for the serine and threonine rings with bilities for target-directed design of new blockers of
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

124
GMIRO et al.
In conclusion, we can mention that the use of the
(a)
pharmacochemical method is interesting for two major
reasons. First, it accelerates the search for new highly
effective drugs. For example, compound (XXXV), con-
+
+
taining Me3N- and Pri2 HN groups at terminal positions
on the pentamethylene chain, is considerably more
active than the known bicationic ganglion blockers.
Second, the structure and topography of the biological
model are inevitably perceived in the course of the
study of the structure–activity relationship. It is impor-
tant that, when using the pharmacochemical approach,
we obtain data characterizing a native functioning
receptor in its actual physiological medium.
EXPERIMENTAL
(b)
For the determination of the ganglion-blocking
activity, suberyldicholine dichloride, 5-Methylfurme-
thide, and Hexamethonium bromide from Sigma were
used. Other chemicals were synthesized at the Institute
of Experimental Medicine (St. Petersburg).
Methods of conformational analysis. Theoretical
conformational analysis was performed when consider-
ing the internal rotation around all simple bonds of the
trialkylammonium group and the bond of the pentame-
thylene chain closest to the nitrogen atom. The energy
of conformations was computed at the atom–atom
approximation (using the molecular mechanics
method) as a sum of noncovalent atom interactions, tor-
sion energy of bonds, the energy of elastic deforma-
tions of the carbon and nitrogen valent angles, and the
energy of electrostatic interactions. The computing was
performed on an IBM PC/AT using a universal program
for conformational calculations of organic molecules
[37]. For constructing the spatial molecule models,
their most favorable stable conformations were visual-
ized using the program Desk Top Molecular Modeling
(Oxford Electronic Publishing). In order to determine
the minimal conformer profile, its bar model was
rotated around the nitrogen-linking axis until its projec-
tion on the display was inscribed into the rectangular
frame of a minimal square.
Fig. 3. A model of the channel of nicotinic acetylcholine
receptor in the membrane cross section: (a) in the area of the
serine ring, 8.3 × 6.1 Å in size and (b) in the area of the thre-
onine ring, 6.4 × 5.5 Å in size [35, 13].
Serine ring
8.3 × 6.1 Å
Threonine
ring
Pharmacological Methods: Determination
of Ganglion-Blocking Activity
6.4 × 5.5 Å
(a) The determination of the ganglion-blocking
activity in the experiments with the cat superior
cervical sympathetic ganglion [38]. After the cats
were anesthesized by ether, a mixture of urethane
(250 mg/kg) and chloralose (50 mg/kg, intravenously)
was administered, and a contraction of the nictitating
membrane in isotonic regime was registered (the mus-
cle load was 7 g). The nictitating membrane contraction
was evoked by repetitive supramaximal electrical stim-
ulation of the preganglionic trunk with a frequency of
+
Fig. 4. The position of bicationic blocker [Me N–(CH ) –
3
2 5
+
–
NMeEt ] · 2Br (XL) inside the ion channel of the nicotinic
2
acetylcholine receptor.
neuronal nicotinic acetylcholine receptors. The calcu-
lated sizes of the channel strictures can be useful for
further experimental and theoretical studies of neuronal 5 Hz; the duration of the stimulus was 1 ms. The gan-
nicotinic acetylcholine receptors.
glion-blocking effect was determined by a relaxation of
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

ION CHANNEL TOPOGRAPHY
125
the nictitating membrane caused by the intraarterial evaporated, and the residue was crystallized from an
injection of ganglion-blocking compounds. The com- ethanol–ethyl acetate mixture to give 34 g (59%) of the
pounds were injected using a thin polyethylene cannule target compound, mp 137–139°ë. Found, %: C 33.51,
introduced into the central terminus of the lingual H 6.65, Br 55.54, Br (ion) 27.44.Calc. for C8H19Br2N,
artery, the external carotid artery being tied up. Under
these conditions, the injected compounds were headed
directly to the ganglion. This helped escape side effects
(the fall of blood pressure, change of breathing, etc.) on
ganglionic transmission and enabled more accurate
correlation of the activities of the compounds under
study. As a control of the state of the health of the ani-
mal, the blood pressure was registered in the femoral
artery in all the experiments. This control was neces-
sary, since a considerable decrease in blood pressure
was sometimes observed in the experiments; it changed
the blood supply of ganglions and may affect their sen-
sitivity to pharmacological agents. To prevent the blood
clotting in cannule, 300 U/kg of heparin were injected
intravenously.
%: C 33.21, H 6.57, Br 55.36, Br (ion) 27.68.
(5-Bromopentyl)triethylammonium
bromide
[(5b) R = Et]. A mixture of 1,5-dibromopentane (46 g,
0.2 mol), triethylamine (10.1 g, 0.1 mol), and benzene
(100 ml) was refluxed for 10 h. The solution was
decanted from a solid precipitate; the precipitate was
refluxed with acetone (50 ml) and filtered. The acetone
filtrate was kept for about 10 h at –17°ë, filtered, the fil-
trate was evaporated, and the residue was crystallized
twice from an acetone–ether mixture to give 7 g (21%)
of a highly hydroscopic target compound; mp 86–
88°ë. Found, %: C 33.90, H 7.84, Br 47.99, Br (ion)
24.47. Calc. for C11H25Br2N, %: C 39.87, H 7.55, Br
48.33, Br (ion) 24.17.
General procedure for the preparation of asym-
metrical bisammonium derivatives of types (1) and
(2). A mixture of the respective 5-(trialkylammo-
nio)pentyl bromide of the general formula (5)
(0.01 mol), amine of the general formula R1R2R3N
(0.03 mol), and propanol (15 ml) was refluxed or (in the
case of low-boiling amines Me2NH or Me3N) heated in
a sealed glass ampoule for 3–5 h at 100–110°ë. Then,
The dose, which caused the relaxation of the nicti-
tating membrane by 50% (ÖD50) for Hexamethonium
bromide, was determined in each experiment and com-
pared with that for each ganglion-blocking compound.
This allowed the determination of the compound activ-
ity relative Hexamethonium bromide for the same ani-
mal and, thus, the weakening of the influence of fluctu-
ations in the individual sensitivity toward the ganglion-
blocking agents. The comparisons were based on molar
doses of ganglion blockers. The ÖD50 value for Hexam-
ethonium bromide was 0.035 ± 0.001 µmol (an average
value from 120 experiments). All compounds dissolved
in isotonic solution were injected into the artery in a
constant volume of 0.2 ml. The standard interval
between injections was 15–20 min.
(method A) the solvent and an excess of amine (if
+
R1R2R3N is a quaternary group) were removed, and the
target bisammonium derivative was isolated by crystal-
lization from propanol or isopropanol. In the case of a
+
nonquaternary R1R2R3N group (method B), potassium
carbonate (0.01 mol) was added to the reaction mixture
after the reaction completion, the mixture was refluxed
for 10–15 min, and the solvent and amine were evapo-
rated.A solution of 48% HBr (3 ml) in propanol (20 ml)
was added to the residue, the mixture was evaporated,
the residue was refluxed in propanol (15 ml), a hot solu-
tion was filtered, the filtrate was evaporated, and the tar-
get product was isolated by crystallization from pro-
panol or isopropanol. The data on the elementary anal-
yses corresponded to the calculated ones.
(b) Determination of ganglion-blocking activity
in experiments with the parasympathetic ganglion
of guinea-pig small intestine [38]. A contraction of an
isolated piece of the guinea-pig small intestine placed
in the Tyrode solution (37°ë) was evoked by the selec-
tive ganglionic agonist suberyldicholine dichloride.
The dissociation constant (Kd) was determined accord-
ing to the shift of curves in the coordinates logarithm of
concentration–suberyldicholine dichloride effect in the
presence of a ganglion blocker. The lack of antagonism
with the selective muscarine agonist 5-Methylfurme-
thide was a criterion for the lack of the ganglion effect
on the signal transfer from the nerve to a smooth mus-
cle. The Hexamethonium bromide Kd was also deter-
mined in each experiment for the subsequent compari-
son with that of the ganglion blocker under study.
Examples
1-(Triethylammonio)-5-(trimethylammonio)pen-
tane dibromide (II) was obtained by the interaction of
(5a) with triethylamine; the mixture was refluxed for
5 h and treated using method A followed by crystalli-
zation from isopropanol. The yield was 1.8 g (46%), mp
263–264°ë.
(5-Bromopentyl)trimethylammonium bromide
[(5a) R = Me]. 1,5-Dibromopentane (46 g, 0.2 mol)
was added to a solution of trimethylamine (11.8 g,
1-(Trimethylammonio)-5-(isopropylammonio)pen-
0.2 mol) in benzene (120 ml). The mixture was kept for tane dibromide (XXIX) was obtained by the interaction
6 days at room temperature, and the precipitate was fil- of (5a) with isopropylamine; the mixture was refluxed
tered and dissolved in methanol (40 ml). Acetone for 5 h and treated using method B followed by crys-
(400 ml) was added to the resulting solution, the mix- tallization from propanol. The yield was 1.9 g (55%),
ture was kept for 10 h at –15°ë, filtered, the filtrate was mp 191–192°ë.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

126
GMIRO et al.
8. Gmiro, V.E., Brovtsyna, N.B., Gorbunova, O.B., Ser-
1-(Triethylammonio)-5-(dimethylammonio)pen-
dyuk, S.E., and Lukomskaya, N.Ya., Dokl. Ross. Akad.
Nauk, 1995, vol. 341, pp. 699–705.
tane dibromide (XVI) was obtained by the interaction
of (5b) with dimethylamine in a sealed glass ampoule
for 3 h at 100–110°ë; the mixture was treated using
method B followed by crystallization from a 1 : 3 iso-
propanol–acetone mixture. The yield was 1.9 g (50%),
mp 176–178°ë.
9. Kurennyi, D.E., Derkach, V.A., Gmiro, V.E., and
Skok, V.I., Neirofiziologiya, 1988, vol. 20, pp. 826–829.
10. Albert, A. and Serjeant, E., Ionization Constants, Lon-
don: Methuen, 1962. Translated under the title Kon-
stanty ionizatsii kislot i osnovanii, Moscow: Khimiya,
1964.
1-(Triethylammonio)-5-(dimethylaminooxidonio)-
pentane dibromide (XXIV). A mixture of 5-(dimethyl-
aminopentyl)triethylammonium bromide (2.9 g,
0.01 mol) and 20% hydrogen peroxide (5 ml) was kept
for 24 h at room temperature. The solution was evapo-
rated, and the residue was twice crystallized from iso-
propanol to give 2.2 g (57%) of the target compound,
mp 167–168°C.
11. Hughes, R.A., Physiol. Behav., 1982, vol. 29, pp. 67–71.
12. Comprehensive Organic Chemistry: The Synthesis and
Reactions of Organic Compounds, Barton, D.H.R. and
Wallace, U.D., Eds., London: Pergamon, 1979. 6 Vols.,
vol. 2: Nitrogen Compounds, Carboxylic Acids, and
Phosphorus Compounds. Translated under the title
Obshchaya organicheskaya khimiya, Moscow: Khimiya,
1982.
1-(Triethylammonio)-5-(dimethylsulfonio)pentane
dibromide (XIV). A solution of MeSNa (1.05 g,
0.015 mol) in ethanol (10 ml) was added to a solution
of 5-(bromopentyl)triethylammonium bromide (5b) in
ethanol (10 ml), the mixture was kept for 24 h at room
temperature and filtered. A solution of methyl bromide
(8 g) in ethanol (50 ml) was added to the filtrate, the
13. Brovtsyna, N.B., Tikhonov, D.B., Gorbounova, O.B.,
Gmiro, V.E., Serduk, S.V., Lukomskaya, N.Ya., Magaza-
nik, L.G., and Zhorov, B.S., J. Membr. Biol., 1996,
vol. 152, pp. 77–87.
14. Changeux, J.P., Galzi, J.L., Devillers-Thiery, A., and
Bertrand, D., Quarterly Rev. Biophys., 1992, vol. 25,
pp. 395–432.
mixture was kept for 5 days at room temperature and 15. Vidal, C. and Changeux, J.P., Neural Information Pro-
cessing Systems, 1996, vol. 11, pp. 202–207.
evaporated, and the residue was crystallized from ace-
tone and recrystallized from isopropanol to give 1.2 g 16. Unwin, N., J. Mol. Biol., 1993, vol. 229, pp. 1101–1124.
(20%) of the target compound, mp 156–157°C.
17. Unwin, N., Nature, 1995, vol. 373, pp. 37–43.
18. Miyazawa, A., Fujiyoshi, Y., Stowell, M., and
Unwin, N.J., Mol. Biol., 1999, vol. 288, pp. 765–786.
ACKNOWLEDGMENTS
19. Cohen, B.N., Labarca, C., Davidson, N., and Lester, H.A.,
J. Gen. Physiol., 1992, vol. 100, pp. 373–400.
This work was supported by the Russian Foundation
for Basic Research, project no. 99-04-49815.
20. Lindstrom, J., Ion Channels, 1996, vol. 4, pp. 377–450.
21. Lindstrom, J., Peng, X., Kuryatov, A., Lee, E.,
Anand, R., Gerzanich, V., Wang, F., Wells, G., and Nel-
son, M., Ann. N.Y. Acad. Sci., 1998, vol. 841, pp. 71–86.
REFERENCES
22. Rust, G., Burgender, J.M., Lautenburg, T.E., and Cache-
lin, A.B., Eur. J. Neurosci., 1994, vol. 6, pp. 478–485.
1. Skok, V.I., Selyanko, A.A., Derkach, V.A., Gmiro, V.E.,
and Lukomskaya, N.Ya., Neirofiziologiya, 1984, vol. 16,
pp. 54–61.
23. Rust, G., Burgunder, J.M., Lauterburg, T.E., and Cache-
lin, A.B., Eur. J. Neurosci., 1994, vol. 6, pp. 478–485.
24. Skok, M.V., Voitenko, L.P., Voitenko, S.V., Lykh-
mus, E.Y., Kalashnik, E.N., Litvin, T.I., Tzartos, S.J.,
and Skok, V.I., Neuroscience (Oxford), 1999, vol. 93,
pp. 1427–1436.
25. Imoto, K., Bush, H., Sakmann, B., Mishina, M.,
Konno, T., Nakai, J., Bujo, H., Mori, Y., Fukuda, K., and
Numa, S., Nature, 1988, vol. 335, pp. 645–648.
2. Brovtsyna, N.B., Gmiro, V.E., Gorbounova, O.B., Ser-
dyuk, S.S., and Lukomskaya, N.Ya., Membr. and Cell
Biol., 1996, vol. 10, pp. 59–73.
3. Zhorov, B.S., Brovtsyna, N.B., Gmiro, V.E., Luko-
mskaya, N.Ya., Serdyuk, S.E., Potapyeva, N.N., Maga-
zanik, L.G., Kurenniy, D.E., and Skok, V.I., J. Membr.
Biol., 1991, vol. 121, pp. 119–132.
26. Imoto, K., Konno, T., Nakai, J., Wang, F., Mishina, M.,
and Numa, S., FEBS Lett., 1991, vol. 289, pp. 193–200.
4. Gmiro, V.E., Lukomskaya, N.Ya., and Serdyuk, S.E.,
Dokl. Akad. Nauk SSSR, 1987, vol. 292, pp. 746–750.
27. Bertrand, D., Galzi, J.L., Devillers-Thiery, A., Ber-
trand, S., and Changeux, J.P., Curr. Opin. Cell Biol.,
1993, vol. 5, pp. 688–693.
28. Leonard, R.L., Labarca, C.G., Charnett, P., David-
son, N., and Lester, H., Science, 1988, vol. 242,
pp. 1578–1581.
29. Revah, F., Galzi, J.L., Giraudat, J., Haumont, P.Y., Lede-
rer, F., and Changeux, J.P., Proc. Natl. Acad. Sci. USA,
1990, vol. 87, pp. 4675–4679.
5. Gyermek, L., Pharmacology of Ganglionic Transmis-
sion, Kharkevich, D.A., Ed., Berlin: Springer, 1980,
pp. 63–121.
6. Zhorov, B.S., Brovtsyna, N.B., Gmiro, V.E., Luko-
mskaya, N.Ya., Serdyuk, S.E., Potap’eva, N.N., and
Magazanik, L.G., Biol. Membr., 1989, vol. 6, pp. 1069–
1084.
7. Brovtsyna, N.B., Gmiro, V.E., Gorbunova, O.B., Ser- 30. Corringer, P.J., Bertrand, S., Galzi, J.L., Devillers-Thi-
dyuk, S.E., and Lukomskaya, N.Ya., Biol. Membr., 1996,
vol. 13, pp. 57–70.
ery, A., Changeux, J.P., and Bertrand, D., Neuron, 1999,
vol. 22, pp. 831–843.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002

ION CHANNEL TOPOGRAPHY
127
31. Drisdel, R.C. and Green, W.N., J. Neurosci., 2000, 36. Tikhonov, D.B., Potap’eva, N.N., Gmiro, V.E., Zho-
vol. 20, pp. 133–139.
rov, B.S., and Magazanik, L.G., Biol. Membr., 1996,
vol. 13, pp. 185–195.
32. Kertser, S., Bobryshev, A., Voitenko, S., Gmiro, V.,
Brovtsyna, N., and Skok, V., J. Membr. Biol., 1998,
vol. 163, pp. 111–118.
37. Zhorov, B.S. and Brovtsyna, N.B., Membrane Biol.,
1993, vol. 135, pp. 19–26.
33. Charnet, P., Labarca, C., Leonard, R.J., Vogelaar, N.J.,
Czyzyk, L., Gouin, A., Davidson, N., and Lester, H.A.,
Neuron, 1990, vol. 2, pp. 87–95.
34. Villarroel, A., Herlitze, S., Koenen, M., and Sak-
mann, B., Proc. R. Soc. London, B, 1991, vol. 243,
pp. 69–74.
38. Boulter, J., O’Shea-Greenfield,A., Duvoisin, R.M., Con-
nolly, J.G., Wada, E., Jensen, A., Gardner, P.D., Bal-
livet, M., Deneris, E.S., NcKinnon, D., Heinemann, S.,
and Patrick, J., J. Biol. Chem., 1990, vol. 265, pp. 4472–
4482.
39. Seguela, P., Wadiche, J., Dineley-Miller, K., Dani, J.A.,
and Patrick, J.W., J. Neurosci., 1993, vol. 13, pp. 596–
604.
35. Furois-Corbin, S. and Pullman, A., Biochim. Biorhys.
Acta, 1989, vol. 984, pp. 339–350.
RUSSIAN JOURNAL OF BIOORGANIC CHEMISTRY Vol. 28 No. 2 2002