Surface active benzodiazepine-bromo-alkyl conjugate for potential GABA A-receptor purification

Article abstract of DOI:10.1039/c1ob05210a

A conjugable analogue of the benzodiazepine 5-(2-hydroxyphenyl)-7-nitro- benzo[e][1,4]diazepin-2(3H)-one containing a bromide C₁₂-aliphatic chain (BDC) at nitrogen N1 was synthesized. One-pot preparation of this benzodiazepine derivative was ac

Full text of DOI:10.1039/c1ob05210a

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 5737
www.rsc.org/obc
PAPER
Surface active benzodiazepine-bromo-alkyl conjugate for potential
GABAA-receptor purification†
A. V. Turina,a G. J. Quinteros,b B. Caruso,a E. L. Moyanob and M. A. Perillo*a
Received 8th February 2011, Accepted 10th May 2011
DOI: 10.1039/c1ob05210a
A conjugable analogue of the benzodiazepine 5-(2-hydroxyphenyl)-7-nitro-benzo[e][1,4]diazepin-
2(3H)-one containing a bromide C12-aliphatic chain (BDC) at nitrogen N1 was synthesized. One-pot
preparation of this benzodiazepine derivative was achieved using microwave irradiation giving 49%
yield of the desired product. BDC inhibited FNZ binding to GABAA-R with an inhibition binding
constant Ki = 0.89 mM and expanded a model membrane packed up to 35 mN m-1 when penetrating in
it from the aqueous phase. BDC exhibited surface activity, with a collapse pressure p = 9.8 mN m-1 and
2
˚
minimal molecular area Amin = 52 A /molecule at the closest molecular packing, resulted fully and
~
non-ideally mixed with a phospholipid in a monolayer up to a molar fraction x 0.1. A
=
geometrical–thermodynamic analysis along the p–A phase diagram predicted that at low xBDC (<0.1)
and at all p, including the equilibrium surface pressures of bilayers, dpPC–BDC mixtures dispersed in
water were compatible with the formation of planar-like structures. These findings suggest that, in a
potential surface grafted BDC, this compound could be stabilize though London-type interactions
within a phospholipidic coating layer and/or through halogen bonding with an electron-donor
surface via its terminal bromine atom while GABAA-R might be recognized through the CNZ
moiety.
The integral membrane receptor GABAA (GABAA-R) is a
1. Introduction
pentameric proteinthat acts as a ligand-gated chloride channel and
is the site of action of a variety of pharmacologically important
drugs including benzodiazepines.7 GABAA-R has been previously
purified by ligand affinity chromatography8 as well as by means of
immunochemical techniques.9 The development of a new purifica-
tion method as a succedaneum of traditional inmunoprecipitation
procedures to avoid the requirement of monoclonal antibodies
against specific epitopes of the GABAA-R subunit requires the
immobilization of a ligand molecule in a solid support that should
be easily concentrated and recovered after sedimentation.
Benzodiazepines (1,4-benzodiazepin-2-ones, BZDs) i.e. clon-
azepam (CNZ), diazepam (DZ), flunitrazepam (FNZ) (Scheme
1a) are drugs, pharmacologically grouped within the minor tran-
quillizers and widely used as anxiolytics, hypnotics and anticonvul-
sants. They stimulate inhibitory neurotransmission throughout the
brain by enhancing the activity of the neurotransmitter g-amino
butyric acid (GABA). BZDs are not active per se but potentiate
the GABA-mediated chloride conductances acting at a binding
site located at GABAA-R. The BZD binding site is different but
allosterically related with the GABA binding site.10
Purification of membrane bound proteins has always been a
difficult challenge. It is a necessary step prior to many basic and
applied studies. As an example, sometimes the isolation of proteins
in a pure state may be required for their reconstitution in a model
membrane. This would allow the evaluation – under controlled
conditions – of the modulation induced by environmental prop-
erties such as composition, curvature, molecular packing, etc, on
the conformation–activity relationship of an intrinsic membrane
protein.1–5 Protein purification is also required for the development
of sensors or coated nanoparticles based on surface-grafted
proteins through non-covalent immobilization which would be
of major interest in bioanalytical sciences associated with the
development of “lab-on-chips” systems.6
aBiof´ısica-Qu´ımica, Ca´tedra de Qu´ımica Biolo´gica, Departamento de
Qu´ımica, Facultad de Ciencias Exactas, F´ısicas y Naturales, Universidad
Nacional de Co´rdoba, Av. Ve´lez Sarsfield 1611, 5016, Co´rdoba, Argentina.
E-mail: mperillo@efn.uncor.edu; Fax: +54-351-4334139; Tel: +54-351-
4344983 int 5
In a previous work we reported the synthesis and charac-
terization of an amphipathic CNZ (5-(2-hydroxyphenyl)-7-nitro-
benzo[e][1,4]diazepine-2-(3H)-one) derivative capable of binding
at the BZD site of the GABAA-R through its hydrophilic end and
to anchor in a biomembrane by means of its hydrophobic end,
represented by a hydrocarbon chain attached at the N1 position.11
bINFIQC-Departamento de Qu´ımica Orga´nica, Facultad de Ciencias
Qu´ımicas, Universidad Nacional de Co´rdoba. Av. Haya de la Torre s/n
Ciudad Universitaria, X5000HUA, Co´rdoba, Argentina. E-mail: lauramoy@
dqo.fcq.unc.edu.ar; Tel: 54-351-4334171 and 54-351-4334168
† Electronic supplementary information (ESI) available: The 1H-NMR
and 13C-NMR spectra and K vs. BDC molar fraction plot. See DOI:
10.1039/c1ob05210a
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5737–5747 | 5737
©

View Article Online
Scheme 1 (a) General chemical structure and substituents of the BZDs used and synthesized in the present paper. ∑ In [3H]FNZ one of the H in R2 is
substituted by tritium. (b) Representation of the most abundant GABAA-R isoform (subunit composition a1b2g2)37 inserted into SMs. The GABAA-R is
a pentamer composed of 5 subunits in a circular arrangement around a central axis. Each subunit contains a large extracellular N terminus, four a helix
transmembrane domains, and a small extracellular C terminus. The N terminus forms a sandwich of b sheets, which bears the binding sites for GABA and
BZDs.37–39 In the scheme the ligand binding sites for GABA (between a1b2 subunits) and for BZDs or BDC (between a1g2) are indicated. (c) Depiction of
a solid substrate coated with a mixed phospholipid–BDC monomolecular layer binding purified GABAA-R reconstituted in a micellar system. Coating
may be achieved by the transference of the monolayers from the air–water interface to the solid substrate by the Langmuir–Blodgett technique.33
The design of this BZD derivative was based on theoretical
studies on the BZD structure–activity relationship indicating that
high affinity BZD analogues interacted with three GABAA-R
electrophilic groups located at C7, C2 and the iminic nitrogen N4.
It is known that the N4 interaction is improved by the presence of
halogens in C2¢ and when the phenyl ring is rotated in the direction
that confers greater coplanarity between the phenyl ring and the
plane C1¢–C5 = N4.12 Consequently, some chemical modifications
at the N1 position were not expected to affect (at least chemically)
the BZD capacity to interact with the receptor binding site and
our results confirmed this hypothesis.11 So, the N1-octadecyl-CNZ
was not only an active ligand at GABAA-R but also was able to be
spontaneously inserted in phospholipidic Langmuir films packed
within a wide range of lateral surface pressures. Although it was
partially excluded from the film at high molecular packings, in
low proportions it could remain in the film up to the equilibrium
surface pressures of natural membranes bilayers.
In spite of these promising results, the synthetic procedure
applied gave rise to a methyl N-methylene glycinate as a secondary
product; hence, to obtain the alkyl-CNZ with purity higher than
95%, the reaction mixture required a tedious purification proce-
dure by TLC. Furthermore, a longer alkyl chain would confer on
the CNZ derivative a longer surface for van der Waals dispersion
interactions which would improve its membrane stability.
(2-hydroxyphenyl)-7-nitro-benzo[e][1,4]diazepin-2(3H)-one sub-
stituted at the N1 position with a dodecyl aliphatic chain contain-
ing a bromine atom at the end of the chain, which was denominated
Bromo-dodecylclonazepam (BDC). The ability of this compound
to interact and stabilize at the membrane–water interface was
studied in monomolecular layers at the air–water interface by
the Wilhelmy plate method. Molecular parameters determined
from surface pressure–area isotherms were used to predict not
only the type of self-assembled structures that BDC would be able
to form in aqueous dispersions but also the molecular features
that would contribute to enhancing the stability of this compound
in the typical planar configuration that characterizes membrane
bilayers. The presence of the bromine atom at the end of the alkyl
chain provides this new derivative with the potential of interacting
through halogen-bonding to appropriate electron donors. Hence,
the possibility that this BDC may become surface-grafted and find
several potential applications within the area of material science
is discussed.
2. Experimental procedures
2.1 Materials
DZ and CNZ were kindly supplied by Products La Roche
(Co´rdoba, Argentina). FNZ labeled with tritium (3H) at the N1
position ([3H]FNZ) was purchased from New England Nuclear
Chemistry (E.I. DuPont de Nemours & Co. Inc., Boston, MA,
USA). Dipalmitoylphosphatidylcholine was obtained from Avanti
Polar Lipids (Alabaster, Al, USA). Other drugs and solvents were
These previous findings motivated the development of a new
and more effective synthetic procedure which was applied to the
synthesis of a new CNZ derivative with a longer alkyl chain.
So, in the present paper we describe the synthesis, the
GABAA-R binding kinetics and biophysical properties of a 5-
5738 | Org. Biomol. Chem., 2011, 9, 5737–5747
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 1 Experimental conditions of BDC (3) synthesis
Entry
Heating
Solvent
Base
t (min)
T/◦C
% 3
1
conventional
conventional
conventional
MW (CV)a
MW (CV)
MW (CV)
MW (CV)
MW (CV)
MW (CV)
MW (OV)b
MW (OV)
MW (OV)
MeOH
DMF
DMF
DMF
DMF
DMF
DMF
DMF
—
NaOMe
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
1440
360
600
5.0
65
80
153
80
80
80
80
100
80
80
80
3
0
0
25
40
22
0
30
3
2
3
4
5
7.5
6
12.5
22.5
7.5
7.5
7.5
7.5
7.5
7
8
c
9
K2CO3
10
11
12
DMF
—
—
K2CO3
K2CO3
K2CO3
49
11
2
c
80
a Closed-vessel conditions. b Open-vessel conditions. c Solvent free reactions: substrates were mixed with silica gel (100 mg 3/100 mg silica).
of analytical grade. Merck silica gel 60 was used for filtration and
Aldrich silica gel 60 F254 were used for preparative TLC. The
chemical structures of the BZDs used are depicted in Scheme 1a.
mixture was then purified by chromatographic column using
hexane/ethyl acetate as eluents to obtain the desired product.
The highest yield obtained through this methodology (49%) was
achieved after 7.5 min irradiation in an open-vessel system at
80 ◦C (see below and Table 1 for further explanation).
2.2 NMR spectroscopy
The characterization of compound 3 was done by 1H NMR, 13
C
All starting materials were commercially available and solvents
NMR, HSQC, HMBC and COSY using CDCl3 as solvent. Full
spectra are provided as supporting information.†
1
were distilled and dried before use. H and 13C NMR spectra
were recorded in a Bruker FT-400 (Rheinstetten, Germany) (1H at
400 MHz and 13C at 200 Hz) spectrometer using CDCl3. Chemical
shifts are reported in parts per million (ppm) downfield from TMS.
5-(2-Chlorophenyl)-6-nitro-1-bromo-dodecyl-1,3-dihydro-2H-1,
1
4-benzodiazepin-2-one (3, BDC). H NMR (400 MHz, CDCl3),
d (ppm): 1.43 (m, 18H), 1.85 (q, J = 6.96 Hz, 2H), 3.41 (t, J =
6.88 Hz, 2H), 3.78 (m, 1H), 4.36 (m, 1H), 4.95 (d, J = 10.72 Hz,
1H), 3.81 (d, J = 10.72 Hz, 1H), 7.30 (m, 1H), 7.46 (m, 2H), 7.57
(d, J = 9.1 Hz, 1H), 7.66 (m, 1H), 7.94 (d, J = 2.64 Hz, 1H),
8.36 (d, J = 2.54 Hz y J = 9.08 Hz, 1H). 13C (200 MHz, CDCl3),
d (ppm): 26.91, 28.15, 28.44, 28.72, 29.16, 29.35, 29.41, 33.78,
47.49, 57.17, 122.57, 124.65, 126.00, 127.44, 130.38, 131.75,
132.97, 137.11, 139.20, 143.26, 147.10, 167.98, 168.52.
2.3 Procedure for preparation of CNZ derivative
In the first approach 0.2 mL of a methanolic solution of sodium
methoxide (1.060 M) was added to a solution of CNZ (1)
(0.152 mmol) in 25 mL of anhydrous methanol. The mixture
was refluxed for a period of 15 min. After this time, 0.297 mmol
of 1,12-dibromododecane (2) was added and subsequently re-
fluxed for 24 h (Table 1, entry 1). After cooling, methanol
was evaporated and the crude product was filtered using silica
gel. This mixture was then purified by chromatographic column
using hexane/ethyl acetate as eluents. Following this protocol,
the desired 1-(12-bromododecyl)-5-(2-chlorophenyl)-7-nitro-1H-
benzo[1,4]diazepin-2-one (3, BDC) was obtained with a very low
yield (3%).
2.4 Monolayer studies
2.4.1 Surface pressure–area and surface potential–area com-
pression isotherms. Monomolecular layers were prepared and
monitored essentially as described previously.14,15 Experiments
were performed at room temperature. The surface pressure (p,
Wilhelmy plate method via a platinized-Pt plate), surface potential
(DV, vibrating plate method) and the area enclosing the monolayer
(A) were automatically measured with a Minitrough II (KSV,
Helsinki, Finland). The Teflon trough used had 24,075 mm2
total area. Bidistilled water (230 mL total volume) was used as
subphase. Lipid monolayers were formed by spreading, on the air–
water interface, between 30–80 mL of a 2 : 1 chloroform–methanol
solution containing 1 mg mL-1 of: a) the newly synthesized
CNZ alkyl derivative (BDC), b) a pure saturated phospholipid
(dipalmitoylphosphatidylcholine, dpPC) or c) a binary mixture
dpPC–BDC at a molar fraction of BDC (xBDC) varying between 0
and 1. The p–A and DV–A compression isotherms were recorded
continuously, at a compression rate of 5 mm min-1. The surface
isotherm experiments were repeated at least twice.
In the second methodology, 0.148 mmol of 1, 0.222 mmol of
2 and 0.445 mmol of K2CO3 in 10 mL of anhydrous DMF were
◦
heated for 6 h in an oil bath at 80 C (Table 1, entry 2) and for
10 h at 153 ◦C (Table 1, entry 3). Unfortunately, product 3 (or
BDC) was never formed and a mixture of undesired products was
obtained.
In the third methodology, 0.360 mmol of 1, 0.538 mmol of 2
and 1.160 mmol of K2CO3, were dissolved in 3.3 mL of DMF. The
mixture was subjected to microwave irradiation13 in quartz closed-
vessel and open-vessel systems for several minutes (5–22.5 min).
A CEM Discover LabMate IntelliVent (USA, Matthews, North
Carolina) reactor operating at 2.45 GHz with 300 W maximum
power and thick-walled was used for MW irradiation. The runs
were done in a temperature controlled mode and the irradiation
power was set within the 40–300 W range. Temperatures were
measured by means of an infrared sensor. In the case of closed-
vessel experiments the pressure was set at 40 psi. The reaction
2.4.2 Compressional modulus and critical packing parameter
calculation. The compressional modulus (K) was calculated
according to eqn (1).
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5737–5747 | 5739
©

View Article Online
lipophilic characteristics, BZDs interact not only with the receptor
(specific interaction) but also, through hydrophobic interactions,
with the membrane as a whole by a partitioning process (non-
specific interactions, NS) locating mainly at the polar head group
region of the biomembranes.20–23 Total binding (TB) and non-
specific interactions (NB) are experimentally accessible but specific
binding to the receptor (B) has to be calculated as a difference
between TB and NB. The latter are irreversible. So, only the
specific binding of [3H]FNZ can be displaced from the receptor
with high concentrations of a non-labeled ligand such as DZ and
the remaining non-specifically bound [3H]FNZ represents NS. TB
is measured in the presence of the same concentration of [3H]FNZ
but in the absence of non-labeled ligand.
Thus, two sets of samples were prepared to measure the both NB
and TB radiolabeled ligand bound to the membrane, respectively.
The incubation system (230 mL final volume) contained: the SM
suspension at a final total protein concentration of 0.25 mg mL-1
and 0.5–12 nM [3H]FNZ (minimum specific activity 81.4 Ci
mmol-1). Then, 100 mM NaCl–50 mM Tris–HCl pH 7.4 buffer,
containing (NB samples set) or not (TB samples set) 9.4 mM DZ
were added.
⎛
⎜
⎜
⎞
⎟
⎠
p
dp
dA
⎟
K = − A
.
⎟
(1)
(
)
p
⎟
⎜
⎝
The Critical Packing Parameter (Pc) was calculated according
to the Israelachvili theory16 by eqn (2).
v
(2)
Pc =
ao ⋅lc
The average molecular area (a0) was experimentally determined
from p–A isotherms, and optimal values for the hydrocarbon
volume (v) and chain length (lc) were calculated from eqn (3) and
(4), according to Perillo et al.,14 and refs. therein:
~
v
(27.4 + 26.9.n)n
(3)
(4)
=
ch
~
lc 1.5 + 1.265n
=
where n and nch are the number of methylene groups in the
hydrocarbon chain and the number of chains per molecule,
respectively. The molecular area (a0) is a function of the interfacial
free energy hence, it was assigned the actual mean molecular
areas17 at specific values of surface pressure taken from the p–A
isotherm of each monolayer at the different compositions studied.
Pc values for the binary dpPC–BDC mixtures were calculated from
the weighted mean values of (v) and (lc) of individual components.
Samples were incubated at 4 ◦C in the dark for 1 h and then
filtered through SS filters Whatman GF/B type with a Bran-
del automatic filtration apparatus (Brandel, Gaithersburg, MD,
USA). Then, filters were rinsed, dried in the air and placed in vials
containing 2.5 mL of scintillation liquid (25% v/v Triton X-100,
0.3% w/v diphenyloxazole in toluene). The retained radioactivity
was measured with a scintillation spectrometer Rackbeta 1214
(Pharmacia-LKB, Finland) at a 60% efficiency for tritium. Specific
binding (B) was calculated as the difference between TB and NB
determined in the absence and in the presence of DZ, respectively.
Eqn (5) was fitted to the saturation curves (specific binding (B)
versus free [3H]FNZ concentration), by a non-linear regression
analysis performed by a computer-aided least squares method:17
2.4.3 Penetration of BDC in monomolecular layers of dpPC at
the air–water interface. The aim of this experiment was to deter-
mine the maximum value of p that allowed drug penetration in
the monolayer (pcut-off). These experiments were done in a circular
Teflon trough (4.5 cm diameter and 0.5 cm depth). Between 5 and
30 mL of a solution of dpPC in 2 : 1 chloroform : methanol were
spread on an aqueous surface (bidistilled water) and about 5 min
were allowed for solvents evaporation. Monolayers were prepared
at constant surface area but at different initial surface pressures
(pi). The temporal variation of p induced by the BDC penetration
into the monolayer after the injection of an ethanolic solution
of the derivative in the subphase was measured until reaching a
plateau (pmax). The values of Dp = pmax–pi were plotted against pi
and a straight line was fitted to the experimental data. The pcut-off
was determined from the intersection of the regression line with
the abscissa axis.
Bmax .F
B =
(5)
Kd + F
where F is the free [3H]FNZ concentrations, B and Bmax are the
ligand concentration-dependent and the maximal specific binding
activities, respectively, and Kd is the equilibrium dissociation
constant. Protein concentration was determined by the method
of Lowry.24
2.5 Synaptosomal membrane preparations
Synaptosomal membranes (SM) were obtained from bovine cere-
bral brain cortex. Meninges were eliminated, the cortex dissected
and the SM was purified essentially according to the method
of Enna and Snyder, modified by Perillo and Arce,18 lyophilized
and stored at -20 ◦C. Immediately before used, membranes were
resuspended in 50 mM pH 7.4 Tris–HCl buffer containing 100 mM
NaCl at a final total protein concentration of 0.25 mg mL-1. This
SM suspension was used as membrane receptor preparation and
GABAA-R source in the experiments that followed.
2.7 [3H]FNZ displacement curve.
The aim of this experiment was to assess the ability of the newly
synthesized BDC to displace [3H]FNZ from its binding site at the
GABAA-R in SM. The experiments were carried out as described
in Section 2.6 but using a constant concentration (2 nM) of
radioligand ([3H]FNZ) and in the presence of 0–50 mM BDC (final
concentrations). The [3H]FNZ concentration was chosen because
of its equivalence with the thermodynamic dissociation constant
(Kd) experimentally determined.
2.6 Binding experiments
Binding was performed essentially as described previously.19 The
BZDs [3H]FNZ and DZ were used as labeled and non-labeled
ligands, respectively. Both compete for their reversible interaction
with the same binding site at GABAA-R (Scheme 1b). Due to their
The IC50 values were determined from the % [3H]FNZ bound
vs. log BDC concentration (nM) plot, as the BDC concentration
required to displace the 50% of [3H]FNZ bound; Ki was calculated
by means of the Cheng and Prusoff equation (eqn (6)),17
5740 | Org. Biomol. Chem., 2011, 9, 5737–5747
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
reactions compared to conventional synthetic methods. For the
advantages of the microwave chemistry, we applied this method-
ology to the N-alkylation of 1. Thus, when a mixture of 1, 2 and
K2CO3 in DMF was irradiated at 80–100 ◦C in open- and closed-
vessel systems, yields of 3 were notably increased up to 49% at 80 ◦C
and irradiation time of 7.5 min (Table 1, entries 5 and 10). The
increase of the reaction time at constant temperature, or vice versa,
did not increase the amount of 3 probably because this compound
decomposed under these severe conditions. Taking into account
the advantages of carrying out the microwave transformations in
dry media we decided to eliminate DMF from the reaction in an
environmentally benign approach. Thus, solid reagents and base
were mixed and subsequently irradiated. Analysis of the crude
indicated the presence of mainly CNZ 1 (>90%) and traces of
N-alkylated product (Table 1, entries 9, 11 and 12). This result
could be explained by the reflection of penetrating microwaves by
the reaction mixture in the absence of DMF and/or the inefficient
stirring of the reagents during the irradiation process. When silica
was used as inorganic support to improve the dispersion of active
sites and facilitate the workup, conversion of 1 was poor (<5%)
and a very small amount of product was obtained (Table 1, entries
9 and 12). On the other hand, it is known from the literature
that there is a dependence of the yield of the products on the
use of closed-vessel or open-vessel microwave heating. When we
compared both types of microwave processing, the formation of
3 was slightly favored in open-vessel systems (Table 1, compare
entries 5 and 10 or 9 and 12). Thus, N-alkylation of 1 requires
the removal of the generated HBr to drive these reactions to
completion and this process could be favored in open systems.
Detailed spectra are provided as supporting information.†
IC50
Ki =
[[3 H]FNZ]
(6)
1+
Kd
2.8 Critical micellar concentration determination
The critical micellar concentration (CMC) of BDC value was
estimated as described previously.25 Briefly, BDC was dispersed in
water at different final concentrations (between 0 and 0.016 mM)
in the presence or in the absence (Control) of the hydrophobic dye
bromothymol blue (BTB) (80 mM final concentration). Samples
were incubated at room temperature during 10 min and then the
absorbance at 437 nm was measured with a Beckman DU 7500
(Fullerton, CA, USA) spectrophotometer. The absorbance values
were plotted against the concentration of BDC (mM), and the
CMC was estimated as the concentration of amphiphile at which
the plot of absorbance vs. concentration suffered an abrupt slope
change.
2.9 Statistical analysis
Binding data were statistically analyzed using a two-tailed Stu-
dent’s t-test for independent samples. P < 0.05 was considered as
statistical significant. Regression analysis was done by the least
squares method.26
3. Results and discussion
3.1 Synthesis of the probe precursor
The CNZ derivative (3, BDC) synthesis was performed using
CNZ (1) as a GABAA-R ligand and 1,12-dibromododecane (2) as
hydrophobic rest under different conditions (Scheme 2 and Table
1). First, the reaction involved the formation of the sodium salt
intermediate which then reacted with the dibromo derivative 2 in
refluxing methanol to give the expected product in very low yields
(3%). Several attempts to increase the formation of 3 were carried
out however, the resulting yield could not be improved.
3.2 [3H]FNZ binding experiments
The aim of this experiment was to obtain the kinetic parameters
necessary for the inhibition binding constant calculation (eqn (5)),
as well as to evaluate the quality of the purified synaptosomal
membranes as GABAA-R source. Bmax reflects the amount of
GABAA-R in SM and depends on the animal source (i.e. rat,
monkey, chicken, and cow) and physiological state as well as the
type of sub-cellular fractionation applied to obtain SMs. Kd is a
measurement of the ligand–receptor binding affinity and from its
value it would be possible to infer about the maintenance of the
conformational stability of the receptor protein.
In these experiments, the specific binding of [3H]FNZ was
calculated from the total minus non-specific binding curves (Fig.
1, inset) as was described in the Materials and methods section.
Fig. 1 shows a typical saturation curve. Eqn (4) was fitted to
the experimental data and the kinetic parameters of the specific
binding were determined as Kd = 2.27 0.16 mM and Bmax
=
Scheme 2 Synthesis of BDC (3).
1715 37 fmol x (mg protein)-1. These values are in agreement
with previous reports.19
In view of the previous results, we decided to apply the
methodology described in the regioselective aminoethylation of
1,4-benzodiazepinones.27 Thus, we tried the N-alkylation of 1
using K2CO3 as a Brønsted–Lowry base and DMF as solvent
at different temperatures and reaction times. Unfortunately, this
change of reaction conditions did not lead to the expected product.
The use of microwave energy to heat chemical reactions has been
shown to dramatically reduce reaction times, increase product
yields and enhance product purities by reducing unwanted side
3.3 [3H]FNZ displacement experiments
The ability of BDC to displace [3H]FNZ was evaluated by means
of a radioligand competition assay. The results, depicted in Fig. 2,
show that the 20–80% displacement of [3H]FNZ specifically bound
was achieved at BDC concentrations between 0.4 and 7.4 mM.
However, BDC was unable to displace 100% of the total [3H]FNZ
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5737–5747 | 5741
©

View Article Online
whole absorbance spectrum of BTB, which exhibited a maximum
at 432 nm in water, suffered a red-shift (lmax = 437 nm) as
well as an hyperchromic concentration dependent effect in the
presence of aromatic molecules such as terpenes.25 The results,
depicted in Fig. 3, showed that BDC induced an increase in
BTB absorbance at 437 nm. The absorbance of BTB at 437 nm
increases as a function of BDC concentration and at CMCBDC
it reached a maximum. This indicated the appearance of a new
hydrophobic phase in the system compatible with the formation
of self-assembled structures of BDC with an estimated CMCBDC
value of approximately 0.0105 mM. Note that the change in the
slope of the non-specific binding curve occurs at a concentration
equivalent to the CMCBDC
.
Fig. 1 Binding of [3H]FNZ to synaptosomal membranes from bovine
brain cortex. The DZ and protein concentration used were 9.24 mM and
0.25 mg protein mL-1, respectively. ( ) Specific binding. The inset shows
᭹
the total (᭡) and the non-specific (ꢀ) [3H]FNZ binding curves.
Fig. 3 BDC self-assembly in aqueous medium. Absorbance of BTB at
437 nm with respect to the control in water (A437 nm), as a function of the
BDC concentration (mM). The CMCBDC value (pointed to by the arrow)
was estimated from the intersection of both regression lines.
Fig. 2 BDC induced displacement of the [3H]FNZ bound to synaptoso-
mal membranes. The [3H]FNZ and protein concentration used were 2 nM
and 0.25 mg protein mL-1, respectively. The inset shows the non-specific
[3H]FNZ binding. The arrow points to CMCBDC value. Points above this
concentration, where BDC is expected to be self-aggregated, were not
considered for the fitting to eqn (6). Other details were described in the
Experimental procedures section. Data shown are the mean s.e.m. of
duplicates.
3.5 Monomolecular layers at the air–water interface
3.5.1 Surface pressure–molecular area isotherms. Surface
pressure–area isotherms are shown in Fig. 4. In Fig. 4a, an
isotherm of pure dpPC is depicted as a reference. Dashed
specifically bound to SM due to its self-aggregation. Assuming
that BDC induces a competitive inhibition of [3H]FNZ binding,
a IC50 = 1.68 mM value was determined from data shown in Fig.
2 and the inhibition binding constant (Ki = 0.89 mM) of BDC
was calculated from eqn (6). The non-specific [3H]FNZ binding
curve (inset in Fig. 2) showed a non-linear behavior indicating that
some artefact was affecting the availability of ligand molecules
to interact with SM, explaining the need for large amounts of
BDC to displace [3H]FNZ from its binding site in the GABAA-R.
Due to the relatively long alkyl chain, the BDC molecule has an
amphipathic structure and the self-association of BDC molecules
into micelles or other self-assembling structure could explain the
non-specific curve shape. To investigate this hypothesis the critical
micellar concentration of BDC (CMCBDC) was determined.
lines indicate the determination of the collapse pressure (pc
=
63.2 mN m-1); the molecular area at the closest packing also
2
˚
known as minimal molecular area (Mma = 42.3 A ) and the
transition surface pressure (pT = 6 mN m-1) corresponding to the
liquid expanded/liquid condensed bidimensional phase transition
characteristic of dpPC.
BDC was able to form monomolecular layers at the air–water
interface both alone (Fig. 4a) as well as in mixtures with dpPC
at various molar fractions (x) (Fig. 4b). From the analysis of the
p–A compression isotherm of BDC (Fig. 4a), a collapse pressure
2
pc = 9.8 mN m-1 and Mma = 52 A were determined and no
˚
bidimensional transitions were observed. If compared with dpPC,
BDC monolayer exhibited a lower stability as reflected by its lower
pc value which may be due to the shorter length of its hydrocarbon
chain that provides a low London dispersion energy stabilizing the
aggregate which, at high p near pc, is not enough to counterbalance
the repulsive interactions between the polar head groups. This
leads to an inefficient molecular packing explaining the fact that,
in spite of having only one hydrocarbon chain and not two as in
3.4 BDC self-assembly
In order to confirm the self-assembling capability of BDC pre-
dicted from its molecular structure, the variation in the absorbance
of BTB was evaluated as a function of the BDC concentration. The
5742 | Org. Biomol. Chem., 2011, 9, 5737–5747
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 4 Behavior of dpPC/BDC binary mixtures at the air–water interface. a) p–Mma compression isotherms of pure dpPC and pure BDC. Collapse
pressure (pc), transition pressure (pT) and minimal molecular area (Amin) determination exemplified over a dpPC p–A compression isotherm. b) Surface
pressure–Mma isotherms of dpPC–BDC binary mixtures. c) Compressional modulus at 6 and 35 mN m-1, as a function of monolayer composition. d)
Surface potential–Mma and e) m^–Mma compression isotherms of dpPC–BDC binary mixtures. Dipole moments (m^) were calculated according to eqn
(6) from DV values taken from panel (d) and were plotted as a function of Mma. f) Resultant dipole moment at 6 ( ) and 35 (᭢) mN m-1, as a function
᭹
of monolayer composition. Small numbers in panels b, d and e refer to xBDC
.
the case of dpPC, BDC exhibits a minimal molecular area higher
than that of the phospholipid.
the compressibility modulus K (Fig. 4c) in conditions of full
miscibility at all compositions (6 mN m-1) or of compositionally-
dependent miscibility (35 mN m-1). At 6 mN m-1, K exhibited
a tendency to increase continuously as a function of xBDC up to
xBDC = 0.6 reflecting both the higher coherence of BDC films with
respect to the liquid expanded phase of dpPC (corresponding to
xBDC = 0) as well as full miscibility of both compounds at this
lateral pressure. Above xBDC = 0.6, K values did not change and
the curve reached a plateau. At 35 mN m-1, K decreased abruptly
~
The BDC–dpPC mixtures exhibited full miscibility up to xBDC
=
0.1 (Fig. 4b). Above xBDC = 0.1, mixtures showed composition-
dependent surface pressure transitions (not related with the
bidimensional surface pressure of dpPC, which occurs at a lower
p). Upon further compression these monolayers collapse at a
~
composition-independent p p of pure dpPC (Fig. 5), suggesting
=
c
that those transitions may consist of a partial collapse of the
monolayer. This also suggested a partial miscibility of both
components. Further analysis was done through a phase diagram
(Fig. 5).
The effect of the proportion of BDC in the mixtures with
dpPC on the elasticity of monolayers was analyzed through
~
as a function of xBDC up to xBDC 0.1 and at higher xBDC showed a
=
dispersion in the data set around the line drawn to follow the eye
that reinforced the different miscibility regimes exhibited by the
phase diagram (the miscibility of BDC in dpPC is partially lost at
xBDC ≥ 0.1).
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5737–5747 | 5743
©

View Article Online
different electrostatic contributions. In the case of uncharged
molecules such as dpPC (a zwitterionic phospholipid), and BDC
(uncharged at the pH assayed) the main contribution arises from
the resultant dipole moments of the monolayer components and
those of the water hydration network (water molecules oriented
at the polar headgroup). In turn, contributions from monolayer
components include the polar headgroup and hydrocarbon chains
contributions. A general decrease in the magnitude of m^ was
observed upon compression due to molecular reorientations at
the surface (Fig. 4e). While the whole plot of m^ vs. Mma at xBDC
=
0.1 was above that of the pure dpPC, it was observed a decreasing
trend in m^ vs. Mma plots up on increasing xBDC. In conjunction
with the information taken from the p–Mma curves, this can be
interpreted as an indication of a monolayer compositional change
as a consequence of its destabilization at high surface pressures.
Fig. 4f shows that, after an initial increase of m^, within the p range
exhibiting total mixing between dpPC and BDC, m^ decreased as a
function of xBDC. At high surface pressure (35 mN m-1), m^ values
were lower than those at 6 mN m-1, with a decreasing tendency at
xBDC > 0.1 even more noticeable than at 35 mN m-1.
Fig. 5 Surface pressure–composition phase diagram overlapped with
critical packing parameter (Pc) for dpPC–BDC mixtures. The phase
diagram (p–xBDC) was constructed using p values ( ), typical pure dpPC
᭹
c
bidimensional phase transition pressures ( ) and the phase transitions
᭺
pressures (ꢀ) or partial collapse pressures (᭢), identified as slope changes
in Fig. 4b detected trough minima in the K vs. Mma plot. Numbers on the
graph refer to Pc values calculated for each mixture at the specified surface
pressure. Note that above the line joining the phase transition points
(᭢) (at xBDC > 0.1) Pc values would be meaningless due to immiscibility.
Predicted type of self-assembled structures formed were indicated by the
dark shadowed area with 0.5 < Pc < 1 in the case of bilayers and a
light shadowed area with Pc £ 0.5, for micelles. The inset shows the Mma
corresponding to those surface pressures at which bidimensional phase
transitions and collapses occur, at each of the xBDC studied. (ꢀ) Mma
at the monolayer collapse point (᭢ and ꢀ) other bi-dimensional phase
transitions observed in mixtures.
3.5.3 Phase diagrams of dpPC–BDC binary mixtures and pre-
diction of their self-assembling structures in water. From minimal
values observed in K–Mma curves (see supporting information†),
the p–x phase diagrams were constructed (Fig. 5). Lines in Fig.
5 indicate the presence of p–x states at which phase transitions
occur and delimit regions of a single bidimensional phase that
may coexist with an already collapsed phase.
A single monolayer phase can be found at low pressures within
the whole compositional range. This monolayer shows a pT similar
~
to that of the pure dpPC up to xBDC 0.1 (hollow circles) which can
=
be associated with a bidimensional phase transition, and collapses
-1
3.5.2 Surface electrostatics. The surface potential DV–area
isotherms (Fig. 4d) showed that DV increased upon compression
and decreased as a function of BDC in the mixtures. At the
closest packing DVdpPC = 750 mV and DVBDC = 340 mV while
the mixtures showed intermediate values. Although the DV–area
isotherm for dpPC showed in Fig. 4 exhibits higher DV values
compared to most reports that have appeared lately,28–30 its shape
as well the shape and p values of the p–area isotherm measured
simultaneously in the same monolayer are correct. Hence, on a
relative basis, the effect of BDC on the electrostatics of its binary
mixtures with dpPC is reliable.
~
at pc 63 mN m (pure dpPC). At xBDC ≥ 0.1 a partial collapse
=
of the monolayer occurs (hollow circles). This is supported by the
inspection of the p–A isotherm shape at xBDC = 0.1 which shows
that the xBDC-dependent collapse point decreases up to the point
where pc equaled the pc value of pure BDC. At higher pressures
a second collapse point is observed (xBDC = 0.1 and 0.3) with
a composition-dependent p, probably consisting of a remaining
dpPC–BDC mixture stable at the air–water interface.
Finally, at xBDC < 0.1 pc increases (black circles), then BDC
seems totally miscible in dpPC exhibiting a bidimensional phase
transition between 5–8 mN m-1.
Surface potential is a measure of the electrostatic field gradient
perpendicular to the membrane interface and thus varies con-
siderably with the molecular surface density and with changes in
orientation accompanying the monolayer compression. Taking the
dielectric constant of the medium as unity, the molecular dipole
moment can be calculated according to eqn (7):
Critical packing parameter values (Pc) allow the prediction of
the type of self-assembling structures that would be form when
an amphipathic substance is dispersed in water. Pc values for the
binary dpPC-BDC mixtures were calculated by eqn (2)–(4) using
Mma values taken from the isotherms shown in Fig. 4b. These
Pc values were superimposed on the phase diagram (Fig. 5). Pc
< 0.5 predicts the self-assembling into micelles, 0.5 < Pc < 1
predicts the self-association into bilayer phases which would lead
to the formation of vesicles. Pc >1 would correspond to phases
with negative curvature and although they are not predicted in
Israelachvili’s theory, they are compatible with inverted vesicles,
such as with hexagonal II or cubic phases.14,31 According to this
interpretation given to data in Fig. 5, within the region located
at the left side of the p–xBDC phase space (at low xBDC (<0.1)
and at all p above pT), the resulting Pc values were compatible
with the formation of bilayers. At the bottom within the whole
p
A
(7)
DV =12. m⊥ +y0
where p = 3.1416, A (molecular area) and V (surface potential)
2
are expressed in A .molecule-1 and mV, respectively, m^ (expressed
˚
in debye units, mD) is the apparent (resultant) perpendicular
dipole moment of the molecule and y0 is the electrostatic potential
difference at the interface with respect to bulk subphase solution
in ionized monolayers and modified by the ionic double layer
(for uncharged molecules y0 = 0).14,15 The parameter m^ contains
5744 | Org. Biomol. Chem., 2011, 9, 5737–5747
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 6 Mean molecular area (a, c and e) and surface potential per unit of molecular surface density (b, d and f) vs. composition plots at constant surface
pressures. Mma values (a, c and e) and DV/molec (b, d and f) values at different molar fractions of BDC and at constant p = 6 (a,b), 18.5 (c,d) or 35
mN m-1 (e,f). Dashed lines represent the behavior expected for ideal mixing or complete segregation of the components. Deviations from this behavior
are more clearly reflected in Mma–x plots.
35 mN m-1 the monolayer phase exists up to xBDC 0.6 and at
~
range of xBDC, Pc values predicted the formation of micelles
(light shadowed area) or bilayers (dark shadowed area). The
region above the line joining the black triangles had Pc values
which, although compatible with a bilayer (Pc > 0.5), should
be considered meaningless due to the partial collapse. Hence,
to obtain stable planar membranes at a p compatible with the
equilibrium surface pressure of biomembranes (near 30–35 mN
m-1) the molar faction of BDC should not exceed 0.1.
=
higher proportions of BDC a discontinuity was shown in Mma
plot (there is no data at these p and xBDC, as shown in Fig. 4b
and 5). Negligible expansion was observed with respect to the
ideality in Mma (Fig. 6e) and in DV x molec-1 (Fig. 6f) only at
xBDC £ 0.2.
The analysis of the variation of the Mma (mean molecular
area) or the DV/molec (surface potential per unit of molecular
surface density) with the mole fraction (xBDC) at constant packing
conditions (p = 6, 18.5 and 35 mN m-1) is shown in Fig. 6. Straight
dotted lines joined the values corresponding to the molecular
parameters of pure dpPC and pure BDC. Experimental points
lying on this line would represent states with either an ideal
mixing or immiscibility behavior of components in the monolayer.
The compositional dependence or independence, respectively, of
the collapse pressure might help to discriminate between both
behaviors. Mma (Fig. 6a, 6c, 6e) as well as DV/molec (Fig. 6b, 6d,
6f) decreased as a function of xBDC. At 6 mN m-1, Mma showed
positive deviations from ideality, suggesting repulsive interactions
or other steric restrictions to the ideal molecular packing between
the monolayer components. Deviations from ideality in Mma and
those in DV/molec were small or negligible, respectively. These
results indicate that the permanence of BDC at high proportions
in the monolayer at low p disrupts the lateral organization of the
phospholipids (Mma is expanded) without affecting significantly
the coherence of the molecular orientation in the monolayer
(small changes in DV/molec). At 18.5 and 35 mN m-1, Mma
and DV/molec vs. xBDC plots decreased discontinuously up to the
abscissa axis and reached the zero value. The behavior in the Mma–
xBDC plot at 18.5 mN m-1 was similar to that at 6 mN m-1. This
surface pressure (18.5 mN m-1) was still within the miscibility
region up to xBDC = 0.6 according to the phase diagram. Strong
positive deviation from ideality was observed in Mma (Fig. 6c)
while DV exhibited a negligible positive deviation (Fig. 6d). At
3.5.4 BDC penetration in phospholipids monolayers. The abil-
ity of BDC to penetrate in monomolecular layers of dpPC from
the aqueous subphase was evidenced by the p increase at constant
area at different initial surface pressures up to a pcut-off = 34.8 0.8
mN m-1. This value was determined by extrapolating the plot of
Dp versus pi to Dp = 0 (Fig. 7). On the other hand, the smooth
variation of the monolayer compressibility (K) with xBDC also
showed that BDC was stable in the film at low molecular packings
(low p). Above 34.8 mN m-1 BDC did not penetrate in the film
from the subphase (Dp = 0) while K at high p (Fig. 4c) suffered a
discontinuous variation with xBDC at xBDC >0.2, and at xBDC > 0.3
the BDC–dpPC mixture partially collapsed at p £ 20 mN m-1 (Fig.
5). Taken together these results suggested that it was possible to
stabilize small amounts of BDC in highly packed monolayers (at
lateral surface pressures compatible with the equilibrium pressure
of bilayers – ca. 30–35 mN m-1, ref. 32), independently of the
direction from which the drug accessed the model membrane.
In general terms, the drug access from the aqueous phase to the
membrane can be considered as part of a partition process and
a kind of non-specific interaction. Thus, the results presented in
this section suggest the possibility of using BDC as a precursor
to develop a molecular probe to allow the evaluation of the
molecular organization of SM within the molecular surroundings
of GABAA-R. Similar to what was proposed for OC,11 the probe
would partition in the membrane with a preferential binding to
the receptor through the CNZ moiety and then to the membrane
though the hydrocarbon tail. For this it will be required to modify
BDC by the covalent attachment of an environmentally sensitive
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5737–5747 | 5745
©

View Article Online
Table 2 Physico-chemical characteristics of OC and BDC derivatives as
compared with dpPC all at the closest packing
Property (units)
dpPC
OC
BDC
alkyl chain length
16
63
42
8
18.8
49
12
9.8
52
pc (mN m-1)
2
˚
Mma (A /molec)
DVpc (V)
0.750
835
213
0.75
bilayer
0.296
385
55
0.35
micelle
0.340
462
31.2
0.51
micelle
m^,pc (mD)a
Kc (mN m-1)
b
Pc
Predicted
self-assembling
structurec
xmax for 0.5 < Pc <
1 at p = 30 mN
m-1d
—
0.3
0.1
Fig. 7 BDC penetration in monomolecular layers of dpPC at different
initial molecular packings. The line represents the fit of a straight line to the
experimental points by regression analysis by the least squares method. The
pcut-off (mN m-1)
Ki (nM)
—
—
13
176
34.8
890
p
cut-off value, indicated by the arrow, represents the maximal p allowing the
a Calculated as indicated in Section 2.4.2. b Calculated at pc of BDC. c Type
of structure of self-assembling formed upon dispersed in water. d Highest
molar fraction allowing self-assembly into bilayers at a surface pressure of
30 mN m-1.
drug penetration and monolayer deformation. Regression line is defined
by the equation: Dp = a + bpi, where a is the ordinate (16.7 0.8 mN m-1)
and b the slope (-0.48 0.04 mN m-1). At Dp = 0, pi equals the pcut-off
=
34.8 mN m-1.
to packing defects introduced by the bromine atom at the methyl
end but improved its ability to penetrate into dpPC monolayers
showing a pcut-off = 34.8 0.8 mN m-1, even at a lateral surface
pressure compatible with the equilibrium pressure of bilayers.
On the other hand, the calculated inhibition binding constant
of BDC was Ki = 0.89 mM, indicating a very low affinity of BDC
to the benzodiazepines binding site in the GABAA-R. The marked
amphipathic structure of the BDC molecule dictated its tendency
to self-association in aqueous media with a CMCBDC = 0.0105 mM
and could explain the shape of the non-specific binding curve as
well as the large amount of BDC necessary to displace [3H]FNZ
from the GABAA-R. The main physico-chemical characteristics
of both OC and BDC derivatives are shown in Table 2.
This finding may contribute to the use of BDC as a probe
to evaluate the molecular organization of SM in the molecular
surroundings of GABAA-R, similar to the OC molecule. In
addition, in spite of the relatively low affinity exhibited by BDC
dispersed in water it would be an interesting tool to develop a
binding-based precipitation method to purify GABAA-R based on
BDC coated microparticles (Scheme 1c). This system may be pro-
posed as a useful succedaneum of traditional immunoprecipitation
procedures to avoid the requirement of monoclonal antibodies
against specific epitopes of GABAA-R subunits.34 The method of
preparation of the ligand-coated particle would be fast, the particle
would be recoverable, permitting repeated micro or large scale one-
step purification of GABAA-R without antibody requirements.
Under those conditions, with the alkyl chain immobilized at a solid
substrate, the self-aggregation of BDC into micelles or other highly
curved structured would be prevented and the specific binding
interaction of the CZ moiety to the BZD site in the GABAA-R
would be improved. Considering our results, the microparticles
of alkylated glass could be coated with a monomolecular layer
of a dpPC–BDC mixture xBDC £ 0.1 to ensure the homogeneous
distribution of BDC molecules on the particle surface and free
access of BDC to the receptor site. Furthermore, advantage can
be taken of the presence of a Br atom at the methyl end of
BDC to stabilize this compound to a surface of appropriate
chemical composition through halogen bonding interactions.
label (fluorescent or paramagnetic) in its molecular structure,
preferably at the hydrocarbon chain.
On the other hand, the objective, as proposed in the present
paper, of immobilizing BDC at the surface of microparticles
should be accomplished in a two-step procedure. BDC molecules
will be firstly incorporated in a monomolecular layer at the air–
water interface (1st step) which will be subsequently transferred to
a solid substrate by the Langmuir–Blodgett technique (2nd step).33
In this case London-type dispersion interaction between BDC and
phospholipid hydrocarbon chains (which can be considered a kind
of non-specific interaction) would contribute to the stabilization of
BDC at the film and will reduce its exchange (partitioning) between
Langmuir–Blodgett films and the nanoparticles containing the
solubilized GABAA-R (Scheme 1c).
4. Conclusions
In the present paper we described the synthesis, the biophysical
characterization and the GABAA-R binding kinetics properties of
a new clonazepam derivative (BDC).
The goal of this work was the use of microwave radiation as a
non-conventional energy source showing better selectivity towards
the desired product. This methodology increased the yield of the
desired product (BDC) in very short reaction times and facilitated
the product purification.
In a previous work, we obtained a clonazepam derivative
bearing an 8 carbon alkyl chain attached al N1 position (octyl
clonazepam, OC). This molecule was able to interact with the
GABAA-R and stabilize at the membrane–water interface but with
a pcutoff = 13 5 mN m-1. Above 13 mN m-1, OC not only did not
penetrate in the film from the subphase, but also from the p–Mma
isotherm analysis and the phase diagram, K at high p suffered a
discontinuous variation with xOC at xOC > 0.2, suggesting that it
was not possible to stabilize high amounts of OC in highly packed
monolayers independently of the direction from which the drug
reached the model membrane. The new derivative BDC, was less
stable at the air–water interface (lower pc) than OC probably due
5746 | Org. Biomol. Chem., 2011, 9, 5737–5747
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Halogen bonding is the non-covalent interaction between halogen
atoms (Lewis acids) and neutral or anionic Lewis bases. There
is a close similarity with hydrogen bonding. Although halogen-
bonding interactions such as the iodine–oxygen interaction were
known in the early 1950s, the study of the halogen bond and
its application in supramolecular chemistry, crystal engineering,
and materials science has resumed in recent years.35,36 This would
further enable us to design and manipulate novel halogen-bonded
functional supramolecular materials based on BDC that might
allow useful applications beyond our immediate interest in the
GABAA-R purification system.
3 D. A. Kelkar and A. Chattopadhyay, Biochim. Biophys. Acta,
Biomembr., 2007, 1768, 1103–1113.
4 M. O. Jensen and O. G. Mouritsen, Biochim. Biophys. Acta, Biomembr.,
2004, 1666, 205–226.
5 A. G. Lee, Biochim. Biophys. Acta, Biomembr., 2004, 1666, 62–87.
6 A. Pallandre, B. de Lambert, R. Attia, A. M. Jonas and J. L. Viovy,
Electrophoresis, 2006, 27, 584–610.
7 W. Sieghart, Trends Pharmacol. Sci., 1992, 13, 446–450.
8 S. M. Dunn and R. P. Thuynsma, Biochemistry, 1994, 33, 755–763.
9 K. Kannenberg, R. Baur and E. Sigel, J. Neurochem., 1997, 68, 1352–
1360.
10 A. K. Mehta and M. K. Ticku, Brain Res. Rev., 1999, 29, 196–217.
11 A. V. Turina, B. Caruso, G. I. Yranzo, E. L. Moyano and M. A. Perillo,
Bioconjugate Chem., 2008, 19, 1888–1895.
12 G. H. Loew, J. R. Nienow and M. Poulsen, Mol. Pharmacol., 1984, 26,
19–34.
13 A. Bangham, M. Hill and N. Miller, Methods in Membrane Biology,
ed. E. D. Korn, Plenum Press, New York, 1974.
14 M. A. Perillo, A. Polo, A. Guidotti, E. Costa and B. Maggio, Chem.
Phys. Lipids, 1993, 65, 225–238.
Abbreviations
BDC
BZD
bromo-dodecylclonazepam
benzodiazepines
critical micellar concentration
clonazepam
Correlation Spectroscopy H1–H1
dimethylformamide
dipalmitoylphosphatidylcholine
diazepam
flunitrazepam
gamma aminobutyric acid
GABAA receptor
Heteronuclear Single Quantum Correlation
Heteronuclear Multiple Bond Correlation
nuclear magnetic resonance spectroscopy
Nuclear Overhauser Effect
octadecylclonazepam
15 G. Gaines, Insoluble Monolayers at Liquids-Gas Interfaces, Interscience
Publishers, New York, 1966.
CMC
CNZ
COSY
DMF
dpPC
DZ
16 J. N. Israelachvili, Intermolecular and Surface Forces, Academic Press,
New York, 1989.
17 H. I. Yammamura, S. J. Enna and J. K. Michael, Neurotransmiter
Receptor Binding, Raven, New York, 1978.
18 M. A. Perillo and A. Arce, J. Neurosci. Methods, 1991, 36, 203–208.
19 M. A. Perillo, D. A. Garcia, R. H. Marin and J. A. Zygadlo, Mol.
Membr. Biol., 1999, 16, 189–194.
FNZ
20 M. A. Perillo, D. A. Garcia and A. Arce, Mol. Membr. Biol., 1995, 12,
GABA
GABAA-R
HSQC
HMBC
NMR
NOE
OC
217–224.
21 D. A. Garcia and M. A. Perillo, Biomed. Chromatogr., 1997, 11, 343–
347.
22 D. A. Garcia and M. A. Perillo, Biochim. Biophys. Acta, Biomembr.,
1999, 1418, 221–231.
23 D. A. Garcia and M. A. Perillo, Biophys. Chem., 2002, 95, 157–164.
24 O. H. Lowry, N. J. Rosebrough, A. L. Farr and R. J. Randall, J. Biol.
Chem., 1951, 193, 265–275.
25 A. V. Turina, M. V. Nolan, J. A. Zygadlo and M. A. Perillo, Biophys.
Chem., 2006, 122, 101–113.
SEM
SM
TLC
standard error of the mean
synaptosomal membranes
thin layer chromatography
26 R. Sokal and F. Rohlf, Introduction to Biostatistics, W. H. Freeman &
Company, New York, 1987.
27 J. K. Mishra, J. Srinivasa Rao, G. Narahari Sastry and G. Panda,
Tetrahedron Lett., 2006, 47, 3357–3360.
28 M. Winterhalter, H. Burner, S. Marzinka, R. Benz and J. J. Kasianow-
icz, Biophys. J., 1995, 69, 1372–1381.
Acknowledgements
The present work was partially financed by grants from CON-
ICET, SECyT-Universidad Nacional de Co´rdoba and ANPCyT
from Argentina. B.C. is a Ph.D. student of the Doctorado en
Ciencias Biolo´gicas of the Universidad Nacional de Co´rdoba.
G.J.Q. is a Ph.D. student of the Doctorado en Ciencias Qu´ımicas
of the Universidad Nacional de Co´rdoba. G.J.Q. and B.C. are
fellowship holders and. A.V.T, E.L.M, and M.A.P are Career
Investigators from CONICET.
29 V. L. Shapovalov, Thin Solid Films, 1998, 327-329, 599–602.
30 K. Sabatini, J. P. Mattila, F. M. Megli and P. K. Kinnunen, Biophys. J.,
2006, 90, 4488–4499.
31 M. A. Perillo, N. J. Scarsdale, R. K. Yu and B. Maggio, Proc. Natl.
Acad. Sci. U. S. A., 1994, 91, 10019–10023.
32 D. Marsh, Biochim. Biophys. Acta, 1996, 1286, 183–223.
33 E. M. Clop, P. D. Clop, J. M. Sanchez and M. A. Perillo, Langmuir,
2008, 24, 10950–10960.
34 M. J. Duggan, S. Pollard and F. A. Stephenson, J. Biol. Chem., 1991,
266, 24778–24784.
35 P. Metrangolo, H. Neukirch, T. Pilati and G. Resnati, Acc. Chem. Res.,
2005, 38, 386–395.
36 J. Xu, X. Liu, J. Kok-Peng Ng, T. Lin and C. He, J. Mater. Chem., 2006,
16, 3540–3545.
References
37 S. W. Baumann, R. Baur and E. Sigel, J. Biol. Chem., 2001, 276, 36275–
1 R. Sogaard, T. M. Werge, C. Bertelsen, C. Lundbye, K. L. Madsen, C.
H. Nielsen and J. A. Lundbaek, Biochemistry, 2006, 45, 13118–13129.
2 J. A. Lundbaek, P. Birn, A. J. Hansen, R. Sogaard, C. Nielsen, J.
Girshman, M. J. Bruno, S. E. Tape, J. Egebjerg, D. V. Greathouse,
G. L. Mattice, R. E. Koeppe, 2nd and O. S. Andersen, J. Gen. Physiol.,
2004, 123, 599–621.
36280.
38 Y. Muroi, C. Czajkowski and M. B. Jackson, Biochemistry, 2006, 45,
7013–7022.
39 J. E. Baenziger and P. J. Corringer, Neuropharmacology, 2011, 60, 116–
125.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 5737–5747 | 5747
©