Electrochimica Acta 138 (2014) 360–366

Contents lists available at ScienceDirect

Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta

Electrochemical impedance studies on the interaction of midazolam

with planar lipid bilayer
Kumaravel Mallaiya a,∗ , S. Rameshkumar a , S.S. Subramanian a , S. Ramalingam b ,
T. Ramachandran c
a
Department of Chemistry, PSG College of Technology, Peelamedu, Coimbatore 641 004, India
b
PSG Institute of Medical Science and Research, Coimbatore 641 004, India
c
Department of Chemistry, Amrita Vishwa Vidyapeetham University, Coimbatore 641 112, India

a r t i c l e i n f o a b s t r a c t

Article history: This work presents the non-specific action of midazolam (MDZ) on planar bilayer lipid membranes
Received 4 February 2014 (BLMs), studied by following the dose-dependent changes in the electrical characteristics of BLMs. MDZ is
Received in revised form 7 June 2014 found to interact with the polar regions or the hydrocarbon core of the membrane depending on whether
Accepted 7 June 2014
it exists in charged or neutral form. The extent of these interactions depends on the ionic strength of the
Available online 30 June 2014
medium supporting the membrane, which influences the proportions of the charged and neutral forms
of MDZ present in the medium. The observed electrical characteristics of the membranes are a conse-
Keywords:
quence of competitive effects brought about by these drug species on both regions of the membrane.
Midazolam
BLM
The drug is observed to impart a stabilizing effect at low doses and a destabilization above a saturation
Nonspecific action level at the polar regions of the lipids. These results demonstrate that the level of MDZ necessary for the
Impedance spectroscopy destabilization and fluidization effect on biomembranes is close to that obtained after a single dose, and
Capacitance that its continuous administration in the spinal cord is neurotoxic.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction following injection, at physiological pH, the imidazole ring closes,

Midazolam (MDZ) is an ultra short-acting tricyclic benzo-
diazepine used in anesthetic practice. When compared to the
traditional bicyclic benzodiazepines, it exhibits better local toler-
ance, faster onset of action, faster plasma clearance, and a shorter
half-life elimination (1.7-2.4 hours) with no active metabolites.
Presently, MDZ in the form of its hydrochloride is the most widely
used intravenous benzodiazepine [1–5].
Chemically, MDZ is 8-Chloro-6-(2-fluorophenyl)-1-methyl-4H-
imidazo-[1,5-a] [1,4]-benzodiazepine (C18 H13 ClFN3 ), and it differs
from most of the “traditional bicyclic benzodiazepines” in having a
nitrogen atom in its additional ring structure. This nitrogen atom
is not sufficiently basic to be protonated at physiological pH, but
it is basic enough to yield water-soluble salts when treated with
strong acids. Midazolam hydrochloride (MDZH+ Cl− ) displays a pH
dependent opening of the benzodiazepine ring below pH ≈ 4.0, but
Abbreviations: BLM, Bilayer Lipid Membrane; MDZ, Midazolam; MDZH+ Cl− ,
Midazolam Hydrochloride; FDD, Frequency Dependent Dispersion.
∗ Corresponding author.
E-mail addresses: mkvteam.research@gmail.com, mkv@che.psgtech.ac.in
(K. Mallaiya).
resulting in a highly lipid-soluble drug [3–6].
The exact mechanism and sites of action of benzodiazepines
responsible for their pharmacological effects are still under study.
Benzodiazepines are proposed to act on the subunits of GABAA
receptors at the postsynaptic neurons and to augment the activity
of the inhibitory neurotransmitter ␥-amino butyric acid [7–12].
The clinical use of these drugs involves a compromise between
their benefits and toxicities. Many studies using animal models
showed that spinally administered MDZ induces severe neurotox-
icities [13–17], but many reports contradict these results [18–24],
which leads to the present work.
Severe neurotoxicity has been reported with a single dose
of intrathecal MDZ administration in rabbits [13,14]. Chronic
subarachnoid administration of both preservative-free and com-
mercial preparations of MDZ in rats have been shown to be

http://dx.doi.org/10.1016/j.electacta.2014.06.038
0013-4686/© 2014 Elsevier Ltd. All rights reserved.
 K. Mallaiya et al. / Electrochimica Acta 138 (2014) 360–366
neurotoxic with significant decreases in cell number and a tendency
towards a higher mean cell volume, with a loss of small neurons
confirming neuronal death [15].
On the other hand, long-term intrathecal infusions of midazo-
lam in humans have been shown to cause no neurotoxicity [18].
After MDZ administration, sections of rat spinal cord did not show
any morphological changes suggestive of MDZ-induced neurotox-
icity [19]. MDZ had minimal neurotoxicity in rat spinal cords, only
at a higher dose of 3 mg/kg [20]. Reports continued to show no sig-
nificant differences in the histological changes in neural tissues,
despite repeated applications of MDZ in the subarachnoid space
[21].
MDZ, which induces anesthesia in humans at intravenous doses
of 0.3 mg/kg, did not anesthetize cats at doses of 20 mg/kg [22].
Spinally administered MDZ, even in large doses, did not cause
acute neurotoxicity or inflammation of the spinal cord of cats [23].
Addition of MDZ to human cerebrospinal fluid in saline neither
decreased the pH below 7.0 nor reduced transparency [24]. Hence,
the results of this in vitro study also suggested that clinically useful
doses of intrathecal or epidural MDZ are not neurotoxic.
Another interesting result was the finding that MDZ provides
protection against opioid (fentanyl) neurotoxicity [25]. Narcotics
in large doses cause brain damage in rats, which is attenuated by
naloxone, MDZ, and phenytoin, which are respectively a narcotic
antagonist, a sedative, and an antiepileptic drug [25].
Though receptor-specific interactions are present, the possi-
bility of site-nonspecific interactions on the lipid core cannot be
excluded [26,27]. Since the bilayer construct is a common fea-
ture for all types of cell membranes, nonspecific disturbances to
the membrane could occur to all cells, including target cells. Such
site-nonspecific interactions (drug-lipid interactions) may be the
cause of undesirable side effects such as neurotoxicities [26,27].
The diversity of the structures of molecules that can act as CNS
depressants led to the argument that anesthetic-receptor interac-
tions could also be nonspecific [27]. There has been a long-standing
controversy over whether membrane lipids or proteins are the tar-
gets for general anesthetics. Lipid/protein interface is shown to be
the anesthetic target, rather than protein or lipid alone [28].
The interaction of drugs with cell membranes can be studied
using model membrane systems that mimic the lipid bilayer archi-
tecture of the cell membranes such as lipid vesicles, bilayer lipid
membranes, supported bilayer lipid membranes etc., [29–33]. Of
these, the bilayer lipid membrane system is extensively used as
an experimental model for biomembranes [34,35]. Human myelin
consists of about 80% lipids and only 20% proteins [36], and hence
artificial lipid bilayers serve as the best models of nerve cell mem-
branes.
Only a few studies have been conducted on the nonspecific
interaction of benzodiazepines with model membranes [37–39],
and little work has been done on MDZ-BLM interaction. Benzodi-
azepines bind to specific receptors, but also interact nonspecifically
with the lipid part of the membrane. Flunitrazepam was found to
localize at the polar region of the bilayer [37]. The extent of interac-
tion and localization of benzodiazepines at the lipid-water interface
depends on the nature of the medium, the membrane composition
and the chemical structure of the drug [38]. On the other hand, ben-
zodiazepines were shown to localize at the hydrocarbon core of the
membrane [39]. The above two results give two different views on
the interaction of benzodiazepines with model membranes, and
the implication of these results will be discussed.
Studies on benzodiazepine partitioning between octanol and a
buffer at physiological pH showed that lipid solubility partly deter-
mines the extent of benzodiazepine distribution in vivo, which
in turn is a major determinant of the duration of clinical action
after a single dose [40]. Nonspecific interactions have important
consequences for a wide variety of biological functions. This type
361
of interaction may mediate several effects of benzodiazepines
observed at global concentrations above their affinity for their
specific receptors [40]. It is well known that the black lipid mem-
branes used as a model membrane in our studies can be very well
characterized using Electrochemical impedance spectroscopy (EIS)
[41–48]. Moreover, the impedance technique has provided a non-
invasive means of characterizing the electrical properties of many
systems [41,48]. This paper describes the real time characteriza-
tion of MDZ - bilayer membrane interactions using electrochemical
impedance spectroscopy.
2. Materials and methods
Midazolam in its hydrochloride form, obtained from Neon
Laboratories Ltd, India, was used for the studies. Egg lecithin
containing greater than or equal to 99% L-␣-Phosphatidylcholine
(Sigma Aldrich) was dissolved in chloroform (5 mg/mL) and used
as the BLM forming stock solution. The BLM-forming dispersion
was prepared by evaporating 100 ␮L of the stock solution in a
2 mL screw-cap tube under nitrogen atmosphere and dissolving the
resulting lipid film in 200 ␮L of n-decane (Merck, Germany).
The BLMs were formed using a chamber, constructed indige-
nously by fixing a 1mm-thick PMMA septum containing a 1 mm
diameter aperture between the 5 mL cavities drilled into two halves
of the PMMA block cut from an 85 × 60 × 22 mm block. Stirring
device, faraday cage and Ag/AgCl electrodes were fabricated fol-
lowing standard procedures [30–32,49–52,34]. A vibration isolated
platform (MINUSK USA) was used to arrest the floor vibrations.
The BLMs were formed by employing the standard procedure
proposed by H.T. Tien [30]. The required bathing solution (1 M or
0.1 M NaCl solution at pH = 7) was added to the BLM chamber after
preconditioning the aperture with about 2 ␮L of the dispersion,
and approximately 5 ␮L of the BLM forming solution was delivered
directly over the aperture. In a stabilization time of 30 minutes, at an
applied DC potential of 40 mV, the phospholipid molecules formed
a lipid bilayer membrane spontaneously through self assembly.
In the stabilization time, the capacitance of membrane increases
continuously due to the self assembly of phospholipid molecules
into bilayer phase by excluding solvent molecules and reaches a
steady value which is in the same order reported in the literature
[30,33,37,48,49].
The electrochemical impedance spectra of the BLMs were
recorded using a Potentiostat (PARSTAT 2273, Princeton Applied
Research–USA). After the stabilization of the membrane, the
impedance measurements were carried out in the frequency range
1 MHz to 10mHz at the open circuit potential by superimposing
a sinusoidal AC signal of small amplitude 25 mV. Data acquisition
was performed utilizing PowerSuite software and analyzed using
ZSimpWin 3.21 software.
After every drug dose, a stabilization period of 30 minutes was
allowed for the drug to equilibrate between the lipid bilayer and
aqueous phases. In accordance with earlier reports, each experi-
ment was repeated to a minimum of 5 to 7 times to confirm the
trend and to avoid errors.
3. Results and discussion
The dose-dependent changes induced by MDZ in the electri-
cal characteristics of the bare and drug-doped membranes were
studied and compared.
3.1. Electrical characterization of interaction of MDZ with BLM
The Nyquist plots obtained for bare and drug-doped BLMs in
1 M and 0.1 M NaCl bath solutions are shown in Fig. 1a and 1b
362 K. Mallaiya et al. / Electrochimica Acta 138 (2014) 360–366

Fig. 1. AC Impedance Spectroscopy of bare and drug-doped ((–䊉–) bare, (––) 10 ␮M, (–*–) 20 ␮M, (––) 40 ␮M, (–|–) 60 ␮M, (––) 80 ␮M, () 100 ␮M) BLM in (1a) 1 M
NaCl and (1b) 0.1 M NaCl bath solutions.

respectively. The observed curves are double semicircles, with their
centers on the real axes. The first semicircle observed at the high
frequencies represents the dispersion of impedance in which the
solution resistance dominates [48,53]. The diameter of the second
semicircle decreases with drug dose showing decrease in mem-
brane impedance or increase in membrane conductance with drug
dose, which is a characteristic property of the membrane to become
much leakier to smaller cations like Na+ , K+ due to fluidization of
the membrane in the presence of drugs [46]. The double semi circle
nature of the curves also indicates that the lipid bilayers are dielec-
tric layers with leakage [47,48]. The impedance spectra obtained
are not straight forward and the interpretation of data can be done
using an equivalent circuit [33]. As a rule, an equivalent circuit is
assembled from resistors and capacitors to represent the domi-
nant components of biomimetic membranes [54]. A biomimetic
membrane can be considered to consist of slabs with different
dielectric properties [54]. The flow of ions across each slab gives
rise to an ionic current. The ions can also accumulate at the bound-
ary between contiguous dielectric slabs and under AC conditions
the accumulation of ions at the boundary of the dielectric slabs
varies with time giving rise to a capacitive current [54]. Hence, each
slab can be simulated by a parallel combination of a resistor and a
capacitor, namely by an RC mesh, for ionic and capacitive currents
respectively. Thus, a commonly accepted -R(RC)(RC)- model (Fig. 2)
[50] is used to obtain the electrochemical impedance parameters
by fitting the experimental impedance data, where P and H stands
for polar groups and hydrocarbon tails of the bilayer phospholipid
membranes respectively.
The specific membrane capacitance (Cm ) and its thickness are
related by the expression
Cm = ␧o ␧/d (1)
Where ␧o is the permittivity of free space
(␧o = 8.854 × 10−14 F cm−1 )
␧ is the dielectric constant of the lipid bilayer (␧ = 2.05) [51].
The calculated thickness for the bare BLM is, d = 4.7 nm, which is
very close to twice the thickness of the lecithin monolayer (2.5 nm)
[52,34,53,54]. Thus, the lipid membrane formed is considered as
bilayer membrane.
In general, if two circuit elements have appreciably different
impedances, their overall impedance is controlled by the circuit
element of higher impedance, if they are in series and by the cir-
cuit element of lower impedance, if they are in parallel. In the
accepted equivalent circuit model for the membrane, there are
two RC meshes to represent the membrane-solution interface and
hydrocarbon core of the membrane of the membrane system and
one resistor connected in series to represent the resistance of bath
solution. Each circuit element has appreciably different impedance.
Of these three circuit elements, the overall impedance of the two RC
meshes depends on the impedance determined by the capacitive
element, since the impedance determined by the capacitive ele-
ment in the RC mesh is always less than resistance of the resistor
element in the RC mesh. The impedance of the capacitive element
in the RC mesh moves towards the resistance of the resistor in the
RC mesh with decreasing frequency and becomes close to that of
resistor at low frequency but still less than resistance of the resistor
in the RC mesh.
The capacitance (Cm ) and conductance (Gm ) of the bilayer phase
are expressed in terms of conductance (G) and capacitance (C) of
polar (P) and hydrocarbon (H) regions of the bilayer phase as fol-
lows [48].

 2     C 
CH GP ⁄2 + CP ⁄2 G2H + ␻2 CH CP ⁄2 P ⁄2 + CH
Cm = G 2  2 (2)
P ⁄2 + GH + ␻2 CP ⁄2 + CH

Fig. 2. An equivalent circuit for BLM Model.

 K. Mallaiya et al. / Electrochimica Acta 138 (2014) 360–366 363

Fig. 3. Bode Plots of lZl (black solid circles) and phase angle (green solid circles) against log(frequency) for a bare membrane in (a) 1 M and (b) 0.1 M NaCl bath solutions.

Table 1
Electrochemical Impedance parameters of bare and drug doped membranes in 1 M and0.1 M NaCl bath solutions.

S.No Concentration 1 M NaCl 0.1 M NaCl

Of MDZ M

Cp (F cm−2 ) CH (F cm−2 ) Rm 7 2

(X10 ˝ cm ) Cm (F cm−2 ) *␴ X10−3 Cp (F cm−2 ) CH (F cm−2 ) Rm 7 2
(X10 ˝ cm ) Cm (F cm−2 ) *␴ X10−3

1 0 85 0.365 4.39 0.351 9.03 46 0.305 5.71 0.292 9.12

2 10 83 0.377 4.13 0.362 9.81 44 0.311 4.40 0.299 9.09
3 20 80 0.385 3.71 0.370 9.52 40 0.322 4.08 0.306 8.15
4 40 77 0.396 3.44 0.383 8.59 38 0.360 3.64 8.344 8.77
5 60 76 0.451 3.39 0.432 9.77 38 0.376 3.09 0.352 9.21
6 80 76 0.463 3.08 0.445 8.91 37 0.381 2.98 0.362 8.72
7 100 76 0.477 2.84 0.459 7.88 38 0.388 2.73 0.367 8.33

*␴ - standard deviation values of specific membrane capacitance.

of the membrane than the hydrocarbon core, and where the ionized
G  G     C 2  form of midazolam interacts [55]. At the saturation concentration,
P ⁄2 + GH + ␻2 C 2 G
P ⁄2 +
GH P ⁄2 H
P ⁄2 GH the CP attains a constant value. From Fig. 4 it is also clear that the
Gm = G 2 C 2 increase in CH with drug dose is biphasic. The initial increase in the
P ⁄2 + GH + ␻2 P ⁄2 + CH core capacitance represents the decrease in membrane thickness,
(3) due to the interaction of ionized forms of MDZ at the polar regions
of the membrane and the second phase represents penetration of
the drug molecules into the lipid bilayer [56]. The total membrane
Thus, there is a frequency dependent dispersion in the mem- capacitance (Cm ), which has the contribution from both the bilayer
brane capacitance and conductance or impedance and the overall hydrophobic domain (CH ) and the two electrical double layers at the
impedance of the membrane system is determined by the differ- two membrane solution interfaces (CP ), increases with drug dose,
ent circuit elements depending on the frequency range of applied which can be explained based on discussions of Liviu Movileanu
AC signal. The bode plots for the variation total impedance of the
membrane and the phase angle of |Z| against the frequency of the
applied AC signal in 1 M and 0.1 M NaCl bath solutions are shown
in Fig. 3a and 3b respectively. At the highest range of frequencies,
the overall impedance |Z| of the bilayer membrane system is deter-
mined by the résistance of the bath solution (Rs ). Because at the
highest range of frequencies the resistance of the bath solution is
greater than 1/␻Cp and 1/␻CH . With further decrease in frequency
1/␻CH is comparable with RH and the bode plot tends to become
independent of applied frequencies. This corresponds to complete
control of total impedance of bilayer lipid membrane by RH .
The electrochemical impedance parameters for the bare and the
drug doped BLM in 1 M and 0.1 M NaCl bath solutions are given in
Table 1. The calculated standard deviation values are in the order
of 10−3 and all the values of capacitance lie between ␮-2␴ and
␮ + 2␴ (where ␮ is mean and ␴ is the standard deviation of the
experimental values of capacitance) showing good reproducibility
and reliability of the data. It is seen that the capacitance of the
polar region (CP ) decreases and reaches a constant value while the
capacitance of the hydrocarbon core increases with drug dose. This Fig. 4. Effect of MDZ concentration on the capacitance of core region of BLM in (䊉)
is because of higher relative permittivity (D = 60) of the polar region 1 M and (––) 0.1 M NaCl bath solutions.
364 K. Mallaiya et al. / Electrochimica Acta 138 (2014) 360–366

et al. on the interaction of flavonoid quercetin with planar BLM 3.2. Interaction of different forms of MDZ H+ Cl− with BLM
[46] as follows;
Since two capacitors CP and CH are in series MDZH+ Cl− is a salt of a weak base and a strong acid and is soluble
in water. MDZ itself is water soluble at pH ≤ 4, but becomes com-
1 1 2 pletely lipid soluble at physiological pH [57]. A complex equilibrium
= + (4)
Cm CH CP exists in solution among different species [58] as represented by
Accordingly equation (12)

CH CP MDZH+ Cl− + + −
solution  MDZH(aq)  [MDZH(..) Cl ] (12)
(ion-pair)
Cm = (5)
CP + 2CH
Thus, in a dilute solution, the possible species present are the
The capacitance of the hydrophobic domain is associated with solvated form of the drug, free ions and ion pairs.
the thickness and the dielectric coefficient of the phospholipids acyl The solubility profile of MDZH+ Cl− is affected by both pH and the
chain as presence of a common ion. The chloride ion present in the medium
exerts an influence on the ionization of MDZH+ Cl− as shown in
(␧0 ␧H ) equation (13). In a concentrated NaCl bath, ion pairs and undisso-
CH = (6)
dH ciated forms are present in higher proportions, whereas in a dilute
bath, there is higher proportion of free ions.

(13)

The dose-dependent drug-induced changes in the electrical

properties of BLMs depend greatly on the ionic strength of the
medium, which in turn depend on the proportion of the ionic
Where εH and dH are relative permittivity and thickness of
and neutral forms of MDZH+ Cl− present. The stability and tight-
the hydrocarbon core of the membrane respectively. When the
ening effect brought about by the reduction in the thickness of the
MDZ molecules are inserted within the hydrophobic core of the
bilayer at lower drug doses arise from the localization of the proton-
lipid bilayer, the capacitance of the membrane system changes due
ated form of MDZ (MDZH+ ) at the polar regions of the membrane.
to change in the capacitance of hydrophobic domain. Therefore,
The destabilization and fluidization of the interior structure of the
when the MDZ molecules are inserted into the lipid bilayer, eq. (5)
bilayer results from the action of the neutral and ion pair forms
changes as
of MDZH+ Cl− . The pure form of MDZ is thus, expected to localize
CP (CH + CMDZ ) still deeper into the membrane, causing greater fluidization of the
CMDZ
m = (7) membrane.
CP + 2 (CH + CMDZ )
The initial increase in the total membrane capacitance (Table 1)
CMDZ
m can be expressed as a function that also depends on the with drug dose is a consequence of the interaction of a greater
capacitance of the membrane in the absence of MDZ, as follows number of protonated drug molecules into the polar regions of the
    bilayer structure, since the bare membrane is freely exposed to the
CP CMDZ CP + 2CH medium. This interaction is higher in 0.1 M NaCl bath than in 1 M
CMDZ
m = Cm + X (8)
CP + CH CP + 2CH + 2CMDZ NaCl bath, which is due to the presence of more number of charged
form of the drug molecules at lower bath concentration. In general,
Where
the interaction of drug molecules with the BLMs depends on the
␧0 ␧MDZ lipophilicity of charged and neutral forms of the drug species. The
CMDZ = (9)
NMDZ VMDZ lipophilicity of charged and neutral forms of a compound is gen-
erally related to the intramolecular structural stabilization of the
NMDZ = average number of MDZ molecules per unit area of the
charge [52] and due to high lipophilicity, the MDZ is able to get
bilayer
partitioned into the BLM and this is the reason for accumulation of
VMDZ = the molecular volume of MDZ
MDZ in high lipid containing tissues [55].
␧MDZ = the dielectric coefficient of MDZ
The CMDZ will be smaller than the other two components
(CH ) and CP ) even at very high concentrations of MDZ, because 3.3. Effect of BLM surface charge on MDZ interaction
the cross sectional area of phospholipid occupied domain is
much higher than the cross-sectional area of the MDZ occupied Kotynska et al. have reported that the surface charge of the BLM
domain. in NaCl solution depends on both pH and the concentration of NaCl
In this condition, [59]. At neutral pH, most of the positive charges in the membrane
surface are covered by Cl− ions and thus the surface is negatively
CP + 2CH charged due to uncovered phosphate groups. The adsorption of Na+
≈1 (10)
CP + 2CH + 2CMDZ ions on the BLM surface partially neutralizes the negative charges
of phosphate groups and thus a net negative charge exists on the
and BLM surface [59]. As the NaCl concentration increases the surface
 CP CMDZ
 negative charge decreases [60] due to an increase in adsorption
CMDZ
m = Cm + (11) of Na+ ions at the BLM surface. Since the excess surface negative
CP + CH
charge of the BLM formed in 0.1 M NaCl bath is higher than that
Hence Cm increases with drug dose. formed in 1 M NaCl bath, the extent of interaction between the
 K. Mallaiya et al. / Electrochimica Acta 138 (2014) 360–366
protonated form of the drug molecules and the surface of BLM is
relatively higher in the former.
3.4. MDZ - BLM interaction and the correlation to its
neurotoxicity
D.A.Garcia and M.A. Perillo [37,38] showed that flunitrazepam
localizes at a region of the bilayer with a dielectric constant D = 60,
which corresponds to the dielectric constant of the polar regions.
The drug molecules are accommodated among lipid molecules by
an unlimited partitioning, which induce a constant increase in the
lipid molecular area and an increase in the curvature of the surface
until it becomes incompatible with a bilayer phase. Thus, when
flunitrazepam interacts nonspecifically with BLM, it becomes an
integral part of the bilayer and converts it into a non-bilayer phase
[37,38,42]. Since they are lipophilic, benzodiazepines are expected
to interact nonspecifically with biomembranes, reaching concen-
trations at least one order of magnitude higher than those in the
surrounding medium and affecting membrane structural proper-
ties [38,61]. Kurishingal et al. [39] and Mennini et al. [62] showed
that benzodiazepines are localized at the hydrocarbon core of the
membrane and are capable of inducing a decrease in membrane
order.
Based on our results, it can be stated that MDZ interacts non-
specifically in both regions of the bilayer, depending on whether it
is in the charged or neutral form. The increase in capacitance at the
initial dose should have been due to a tightening effect imposed
by the MDZH+ by localizing at the polar region of the bilayer and
inducing an increase in the area occupied by the lipid molecules
and a reduction in membrane thickness. The fluidizing effect is due
to the neutral and ion pair forms, which penetrate deep into the
hydrocarbon core structure.
Below the saturation concentration at the solution - membrane
surface interface, the stabilizing effects of the charged forms of
MDZ dominate. Above this concentration, the fluidizing effect of
the neutral form of the MDZ in the hydrocarbon core structure
and the destabilization on the polar regions by excessive partition-
ing of protonated forms dominate. Similar saturation trends have
been reported for the localization of anesthetic with planar bilayers
[63].
The existence of water molecules associated with polar head
groups of the lipid bilayer through hydrogen bonding has been
confirmed by various groups [50,61,64] and also have been
demonstrated using computer simulations [65,66]. These water
molecules, referred to as bound or electrostricted, are essential
to preserve the structural integrity of the bilayer [50,63,66,67].
They are only released from the membrane interface by exter-
nal stresses, such as chemical, mechanical, thermal or electrical.
Since MDZ interacts at both the polar and the hydrocarbon regions,
it should have some dislodging effects on the membrane-bound
water molecules.
The charged head groups of the bilayer attract counterions from
the bath, forming an electrical double layer. The drug molecules,
when added to the bath, first come into contact with the polar head
groups. The positively charged form of MDZ interacts electrostatic-
ally with the negatively charged polar groups of the phospholipids,
while the neutral forms of MDZ penetrate and become partitioned
into the membrane interior. In 0.1 M NaCl solution MDZ exists in
the ionized form relatively in larger proportion than in 1 M NaCl
solution and hence the tightening effect, due to ionized form of the
drug, is comparatively higher at lower doses of MDZ in 0.1 M NaCl
than in 1 M NaCl solution.
Thus, this study has shown that MDZ molecules localize into
membranes at different regions, depending on whether they are
charged or neutral species. It is evident that the increase in mean
cell volume of nerve cells after chronic administration of MDZ
365
might be due to swelling of the nerve cell membrane resulting from
excessive partitioning of the drug [15]. This study also confirms that
at low doses, MDZ imparts a stabilizing and tightening effect on
the membrane, and this could be the reason for the attenuation of
opioid-induced neurotoxicity in the presence of MDZ [25]. Severe
neurotoxic effects in the rabbit spinal cord were demonstrated even
in single doses of MDZ [13,15]. Since biological membranes are
highly ordered, are thinner, and have surfaces and interiors occu-
pied by proteins, the saturation limit of MDZ in biomembranes must
be much lower than that observed for BLMs in our studies. Further,
the freely exposed area of the lipids in a nerve cell is less than that
of the BLMs studied here. Thus, the saturation limit of MDZ in bio-
logical membrane architecture should lie close to concentrations
achieved by a single dose. This explains the reasons why some stud-
ies reveal the neurotoxic nature of MDZ, while others do not. The
results of this study confirm that spinal administration of MDZ is
highly neurotoxic on continuous and chronic administration.
The nonspecific interactions in biological membranes are
reported to induce conformational changes in membrane lipids,
and in turn in the associated proteins. These effects perturb the
permeability properties of membranes. As the biological system is
quite complex, the presence of proteins and other such molecules
might hinder the nonspecific interactions, but total elimination of
nonspecific interaction is not possible. The present study reveals
that there is a good correlation between clinical reports on MDZ
neurotoxicity and its nonspecific action on the lipid bilayer.
4. Conclusions
The charged and neutral forms of MDZ interact nonspecifically
in the polar regions and hydrocarbon core regions of the mem-
brane, respectively. The extent of the interaction depends on the
ionic strength of the medium supporting the membrane. At low
doses, the drug imparts a tightening effect on the polar regions and
a destabilization effect after reaching a saturation level, but a flu-
idization at the interior occurs from the very first dose. There is
a good correlation between clinical reports on MDZ-induced neu-
rotoxicity and its nonspecific action on the lipid bilayer. The level
of MDZ for the destabilization and fluidization effect on biomem-
branes could lie close to the level achieved by a single dose, and its
continuous administration in the spine is neurotoxic.
Acknowledgements
The authors gratefully acknowledge the research grant provided
by the All India Council for Technical Education (AICTE) to the
Department of Chemistry, PSG College of Technology, and the facil-
ities provided by the Principal and the Management of PSG College
of Technology.
References
[1] Roche Laboratories, Versed® (Midazolam Hydrochloride), prescribing informa-
tion, Nutley, NJ, (1997).
[2] J.G. Reves, R.J. Fragen, H.R. Vinik, D.J. Greenblatt, Midazolam: pharmacology
and uses, Anesthesiology 62 (3) (1985) 310–324.
[3] J.W. Dundee, N.J. Hallida, K.W. Harper, R.N. Brogden, Midazolam: A review of
its pharmacological properties and therapeutic uses, Drugs 28 (1984) 519–543.
[4] U. Khanderia, S.K. Pandit, Use of midazolam hydrochloride in anesthesia, Clin
Pharm. 6 (7) (1987) 533–547.
[5] J.H. Kanto, Midazolam: the first water soluble benzodiazepine: Pharmacology,
pharmacokinetics and efficacy in insomnia and anesthesia, Pharmacotherapy
5 (1985) 138–155.
[6] F. García-Pedrajas, J.L. Arroyo, Midazolam in anesthesiology, Rev Med Univ.,
Navarra 33 (4) (1989) 211–221.
[7] Ashok K. Mehta, Maharaj K. Ticku, An update on GABAA receptors, Brain Res
Brain Res Rev. 29 (1999) 196–217.
[8] C. Braestrup, M. Nielsen, Benzodiazepine receptors, Clin. Neuropharmacol 8
(1985) 52–57.
366 K. Mallaiya et al. / Electrochimica Acta 138 (2014) 360–366
[9] E. Siegel, A. Buhr, The benzodiazepine binding site of GABAA receptors, Trends
Pharmacol Sci 18 (11) (1997) 425–429.
[10] Q. Wang, Y. Han, H. Xue, Ligands of the GABAA Receptor Benzodiazepine Bind-
ing Site, CNS Drug Rev 5 (2) (1999) 125–144.
[11] W. Sieghart, Structure and pharmacology of ␥-aminobutyric acid A receptor
subtypes, Pharmacol Rev 47 (2) (1995) 181–234.
[12] J. Lim, G. Lim, B. Sung, S. Wang, J. Mao, Intrathecal midazolam regulates spinal
AMPA receptor expression and function after nerve injury in rats, Brain Res
1123 (1) (2006) 80–88.
[13] B. Ugur, K. Basaloglu, T. Yurtseven, U. Ates, O.N. Aydin, D. Ozenç, M. Yurtseven,
A. Gürel, Neurotoxicity with single dose intrathecal midazolam administration,
Eur J Anaesthesiol 22 (12) (2005) 907–912.
[14] S. Erdine, A. Yucel, S. Ozyalcin, E. Ozyuvaci, G.K. Talu, B. Ahiskali, H. Apak, N.
Savci, Neurotoxicity of midazolam in the rabbit, Pain 80 (1–2) (1999) 419–423.
[15] B.A. Svensson, M. Welin, T. Gordh Jr., J. Westman, Chronic subarachnoid
midazolam (Dormicum) in the rat. Morphologic evidence of spinal cord neu-
rotoxicity, Reg Anesth 20 (5) (1995) 426–434.
[16] J.M. Malinovsky, A. Cozian, J.Y. Lepage, J.M. Mussini, M. Pinaud, R. Souron,
Ketamine and midazolam neurotoxicity in the rabbit, Anesthesiology 75 (1)
(1991) 91–97.
[17] P. Bozkurt, Y. Tunali, G. Kaya, I. Okar, Histological changes following epidural
injection of midazolam in the neonatal rabbit, Paediatr Anaesth 7 (5) (1997)
385–389.
[18] S. Canavero, V. Bonicalzi, M. Clemente, No neurotoxicity from long-term (>5
years) intrathecal infusion of midazolam in humans, J Pain Symptom Manage
32 (1) (2006) 1–3, and the references therein.
[19] I.M. Schwieger, M. Jorge-Costa, G.P. Pizzolato, A. Forster, D.R. Morel, Intrathecal
MDZ reduces isoflurane MAC and increases the apnoeic threshold in rats, Can
J Anaesth 41 (2) (1994) 144–148.
[20] M. Carassiti, A. Ysasi, L.M. Gonzalo, Effects of the chronic administration of
midazolam on the rat hippocampus, Rev Med Univ Navarra 42 (1) (1998) 18–28.
[21] M. Bahar, M.L. Cohen, Y. Grinshpoon, U. Kopolovic, M. Herbert, D. Nass, M.
Chanimov, An investigation of the possible neurotoxic effects of intrathecal
midazolam combined with fentanyl in the rat, Eur J Anaesthesiol 15 (6) (1998)
695–701.
[22] J.D. Leah, R. Malik, D.R. Curtis, Actions of midazolam in the spinal cord of the
cat, Neuropharmacology 22 (12A) (1983) 1349–1356.
[23] T. Nishiyama, T. Matsukawa, K. Hanaoka, Acute phase histopathological study of
spinally administered midazolam in cats, Anesth Analg 89 (3) (1999) 717–720.
[24] T. Nishiyama, N. Sugai, K. Hanaoka, In vitro changes in the transparency and
pH of cerebrospinal fluid caused by adding midazolam, Eur J Anaesthesiol 15
(1) (1998) 27–31.
[25] E.H. Sinz, W.A. Kofke, R.H. Garman, Phenytoin, midazolam and naloxone pro-
tect against fentanyl-induced brain damage in rats, Anesth Analg 91 (6) (2000)
1443–1449.
[26] K. Uma Maheswari, T. Ramachandran, D. Rajaji, Interaction of cisplatin with
planar model membranes – dose dependent change in electrical characteristics,
Biochim Biophys Acta 1463 (2000) 230–240.
[27] I. Ueda, Molecular mechanism of anesthesia, Keio J Med 50 (1) (2001) 20–25.
[28] D. Pflugmacher, H. Sandermann Jr., The lipid/protein interface as a target site
for general anesthetics: a multiple-site kinetic analysis of synaptosomal Ca2+ -
ATPase, Biochim Biophys Acta 1415 (1998) 174–180.
[29] A. Gliozzi, d M. Robello, Model Systems: A strategy for Studying biological
membranes, Colloids and Surfaces 35 (1989) 135–150.
[30] H.Ti. Tien, in: Bilayer Lipid Membranes (BLM) – Theory and Practice, Marcel
Dekker Inc, New York, 1974.
[31] M.K. Jain, in: The Bimolecular Lipid Membrane: A System, Van Nostrand Rein-
hold, New York, 1972.
[32] W. Hanke, W.R. Schlue, in: Planar Lipid Bilayers-Methods and Applications,
Academic Press, 1993, pp. 1–78.
[33] H.T. Tien, A.L. Ottova, The lipid bilayer concept and its experimental realization:
from soap bubbles, kitchen sink, to bilayer lipid membranes, J. Membrane. Sci
189 (2001) 83–117.
[34] Xiaoquan Lu, Tianlu Liao, Lan Ding, Xiuhui Liu, Yan Zhang, Yina Cheng, Jie Du,
Interaction of Quercetin with Supported Bilayer Lipid Membranes on Glassy
Carbon Electrode, Int. J. Electrochem. Sci. 3 (2008) 797–805.
[35] J. Kotyńska, Z.A. Figaszewski:, Adsorption equilibria between liposome
membrane formed of phosphatidylcholine and aqueous sodium chloride
solution as a function of pH, Biochimica et Biophysica Acta 1720 (2005)
22–27.
[36] A. Fink, N. Sal-Man, D. Gerber, Y. Shai, Transmembrane domains interactions
within the membrane milieu: Principles, advances and challenges, Biochimica
et Biophysica Acta 1818 (2012) 974–983.
[37] D.A. Garcia, M.A. Perillo, Localisation of flunitrazepam in artificial membranes.
A spectrophotometric study about the effect the polarity of the medium exerts
on flunitrazepam acid-base equilibrium, Biochim Biophys Acta 1324 (1997)
76–84.
[38] D.A. Garcia, M.A. Perillo, Benzodiazepine localization at the lipid-water inter-
face: effect of membrane composition and drug chemical structure, Biochim
Biophys Acta 1418 (1999) 221–231.
[39] H. Kurishingal, P. Barain, C. Restall, Benzodiazepine-induced changes in
biomembrane fluidity, Biochem Soc Trans 20 (1992) 1575.
[40] D.J. Greenbalt, R.M. Arendt, D.R. Abernethy, H.G. Giles, E.M. Sellers, R.I. Shader,
In vitro quantitation of benzodiazepine lipophilicity: relation to in vivo distri-
bution, Br J Anaesth 55 (10) (1983) 985–989.
[41] M. Kates, in: T.S. Work, E. Work (Eds.), Techniques of Lipidology - Isola-
tion, Analysis and Identification of Lipids, North-Holland publishing Company,
Amsterdam, 1972.
[42] N.S. Radin, in: J.M. Lowenstein (Ed.), Methods in Enzymology, 14, Academic
Press, New York, 1969, pp. 245–254.
[43] L.A. Geddes, in: Electrodes and the measurement of bioelectric events, Wiley
Interscience, 1972, pp. 10–14.
[44] A.K. Corvington, in: “Ion-selective electrode methodology, 1, CRC press, Boca
Raton Florida, 1979.
[45] Axon Guide - Axon Instruments, Inc. Product information, January 2000.
[46] Liviu Movileanu, Ioana Neagoe, Maria Luiza Flonta, Interaction of the antiox-
idant flavonoid quercetin with planar lipid bilayers, International Journal of
Pharmaceutics 205 (2000) 135–146.
[47] M. Naumowicz, Z.A. Figaszewski, Impedance analysis of phosphatidylcholine
membranes modified with gramicidin D, Bioelectrochemistry 61 (2003) 21–27.
[48] Coster, H.G. L. Dielectric and electrical properties of lipid bilayers in relation to
their structure, in: A. Ottova-Leitmannova, H.T. Tien (Eds.), Planar lipid bilayers
(BLMs) and their applications Amsterdam, Elsevier (75-108).
[49] H.G.L. Coster, J.R. Smith, The molecular organization of bimolecular lipid mem-
branes. A study of the low frequency Maxwell-Wagner impedance dispersion,
Biochemica Biophysica Acta 373 (1974) 151–164.
[50] X. Han, Y. Tong, W. Huang, E. Wang: Study of the interaction between lan-
thanide ions and a supported bilayer lipid membrane by cyclic voltammetry
and ac impedance 523(2002) 136-141.
[51] A. Watt, K. Harlos, D. Marsh:, Charge-induced Tilt in ordered-phase phos-
phatidylglycerol bilayers. Evidence from X-ray diffraction, Biochim. Biophys.
Acta 645 (1981) 91–96.
[52] G. Favero, A. D’Annibale, L. Campanella, R. Santucci, T. Ferri, Membrane sup-
ported bilayer lipid membranes array: preparation, stability and ion-channel
insertion, Analytica Chimica Acta 460 (2002) 23–34.
[53] W. Romer, C. Steinem, Impedance analysis and single channel recordings on
nano-block lipid membranes based on porous alumina, Biophysical Journal 86
(2004) 955–965.
[54] R. Guidelli, L. Becucci, Applications of Electrochemistry and Nanotechnology
in Biology and Medicine II Modern Aspects of Electrochemistry, springer 53
(2012) 147–266.
[55] Daniel A. Garcı a, Marı a A. Perillo, Localization of flunitrazepam in artificial
membranes. A spectrophotometric study about the effect the polarity of the
medium exerts on flunitrazepam acid-base equilibrium, Biochimica et Biophys-
ica Acta (BBA) - Biomembranes 1324 (1997) 76–84.
[56] Tadayoshi Yoshida, Hiroshi Kamaya, Issaku Ueda, Giant planar lipid bilayer. III,
Maxwell—Wagner impedance dispersion and anesthetic effects upon interfa-
cial capacitance, Journal of Colloid and Interface Science 105 (1985) 129–135.
[57] R.I. Fryer, in: E. Costa (Ed.), The Benzodiazepines: From Molecular Biology to
Clinical Practice, Raven Press, New York, 1983.
[58] J.O.M. Bockris, A.K.N. Reddy, M. Gamboa-Aldeco, in: Modern Electrochemistry:
Ionics, Plenum Press, 1998.
[59] J. Kotyńska; Z.A. Figaszewski, Adsorption equilibria between liposome mem-
brane formed of phosphatidylcholine and aqueous sodium chloride solution
as a function of pH, Biochimica et Biophysica Acta(BBA)Biomembranes (1720)
0-27.
[60] Hideo Matsumura, Ken-ichi Watanabe, Kunio Furusawa, Flocculation behav-
ior of egg phosphatidylcholine liposomes caused by Ca2+ ions, Colloids and
Surfaces A: Physicochemical and EngineeringAspects 98 (1995) 175–184.
[61] S. Koronkiewicz, S. Kalinowski, K. Bryl, Changes of structural and dynamic prop-
erties of model lipid membranes induced by ␣-tocopherol: implication to the
membrane stabilization under external electric field, Biochim Biophys Acta
1510 (2001) 300–306.
[62] H. Heller, M. Schaefer, K. Schulten, Molecular dynamics simulation of a bilayer
of 200 lipids in the gel and in the liquid crystal phases, J Phys Chem 97 (1993)
8343.
[63] T. Mennini, A. Ceci, S. Caccia, S. Garattini, P. Masturzo, M. Salmona, Diazepam
increases membrane fluidity of rat hippocampus synaptosomes, FEBS Lett 173
(1) (1984) 255–258.
[64] T. Yoshida, H. Kamaya, I. Ueda, Giant Planar Lipid Bilayer, J Colloid Interface Sci
105 (1) (1985) 129–135.
[65] J.P. Dilger, S.G. McLauglin, T.J. McIntosh, S.A. Simon, The dielectric constant of
phospholipid bilayers and the permeability of membranes to ions, Science 206
(1979) 1196–1198.
[66] A. Pohorille, in: P.V.R. Schleyer, N.L. Allinger, T. Clark, J. Gasteiger, P.A. Kollman,
H.F. Schaefer III, P.R. Schreiner (Eds.), Phenomena at aqueous interfaces, The
Encyclopedia of Computational Chemistry, John Wiley & Sons, Chichester, 1998,
pp. 30–44.
[67] M.A. Wilson, A. Pohorille, Mechanism of unassisted ion transport across mem-
brane bilayers, J Am Chem Soc 118 (1996) 6580–6587.