Chemistry and Physics of Lipids 186 (2015) 68–78

Contents lists available at ScienceDirect

Chemistry and Physics of Lipids

journal homepage: www.elsevier.com/locate/chemphyslip

Interactions of the antimalarial amodiaquine with lipid model

membranes
Rafael P. Barroso, Luis G.M. Basso, Antonio J. Costa-Filho *
Laboratório de Biofísica Molecular, Departamento de Física, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto. Av. Bandeirantes, 3900, 14040-901
Ribeirao Preto, SP, Brazil

A R T I C L E I N F O A B S T R A C T

Article history: A detailed molecular description of the mechanism of action of the antimalarial drug amodiaquine (AQ) is
Received 12 March 2014 still an open issue. To gain further insights on that, we studied the interactions of AQ with lipid model
Received in revised form 24 December 2014 membranes composed of dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphosphatidylserine
Accepted 26 December 2014
(DPPS) by spin labeling electron spin resonance (ESR) and differential scanning calorimetry (DSC). Both
Available online 30 December 2014
techniques indicate a coexistence of an ordered DPPS-rich domain with a disordered DPPC-rich domain
in the binary DPPC/DPPS system. We found that AQ slightly lowered the melting transition temperatures
Keywords:
associated to both domains and significantly increased the enthalpy change of the whole DPPC/DPPS
Amodiaquine
DSC
phase transition. DSC and ESR data also suggest that AQ increases the number of DPPC molecules in the
ESR DPPC-rich domains. AQ also causes opposing ordering effects on different regions of the bilayer: while
Antimalarial the drug increases the ordering of the lipid acyl chains from carbon 7 to 16, it decreases the order
Spin label parameter of the lipid head group and of carbon 5. The gel phase was mostly affected by the presence of
AQ, suggesting that AQ is able to influence more organized lipid domains. Moreover, the effects of AQ and
cholesterol on lipid acyl chain ordering and mobility were compared at physiological temperature and, in
a general way, they are similar. Our results suggest that the quinoline ring of AQ is located completely
inside the lipid bilayers with its phenol ring and the tertiary amine directed towards the head group
region. The nonspecific interaction between AQ and DPPC/DPPS bilayers is a combination of electrostatic
and hydrophobic interactions.
ã 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction low or absent, the therapeutic efficacy of artesunate–amodiaquine

Malaria is still one of the most infectious diseases affecting the
developing world. According to the World Health Organization
(WHO), 655,000 people die every year due to malaria, out of which
86% are children under five years of age (WHO, 2011). One of the
major barriers on fighting malaria is the worldwide drug
resistance. Malaria parasites have developed resistance to all
current drugs. To overcome drug resistance, WHO recommends the
use of combined therapies with two or more drugs, in particular,
the combination of amodiaquine (AQ) and artesunate (WHO,
2010). AQ is a 4-aminoquinoline antimalarial drug widely used in
the past and whose chemical structure is similar to that of
chloroquine (Olliaro et al., 1996 O’Neill et al., 1998). Several
countries chose AQ as the first-line drug in combination with
artesunate (WHO, 2010). In areas where artesunate resistance is
* Corresponding author. Tel.: +55 163315 3665; fax: +55 16 33715381.
E-mail address: ajcosta@ffclrp.usp.br (A.J. Costa-Filho).
is highly correlated with the efficacy of AQ alone (Sá et al., 2009).
The mechanism of action of 4-aminoquinolines is not
completely understood, but it is believed that is related to
hemoglobin degradation in the digestive vacuole of the parasite
(Pussard and Verdier, 1994; O’Neill et al., 1998; Fitch, 2004).
Hemoglobin is degraded by several enzymes of the parasite
resulting in toxic heme metabolites. A process of detoxification
converts heme into an insoluble crystal called hemozoin or
“malaria pigment”, which participates in the parasite life cycle
(Sherman, 1977). A complex formed by 4-aminoquinolines with
heme (ferriprotoporphyrin IX) inhibits conversion of heme to
hemozoin, thus killing the parasite. Although AQ forms complexes
with ferriprotoporphyrin IX (Blauer et al., 1993), its complete
mechanism of action, at a molecular level, remains unclear. In
particular, there is not much information on how the drug
overcomes the physical barrier represented by the cell membrane.
Membrane lipids are associated with vital cell functions.
Besides being the structural blocks of cell membranes and the
sources of metabolic energy, lipids play a central role in the
regulation of innumerous cellular processes (Escribá, 2006;

http://dx.doi.org/10.1016/j.chemphyslip.2014.12.003
0009-3084/ ã 2014 Elsevier Ireland Ltd. All rights reserved.
 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
Bittman, 2004). Changes in the structure and lipid composition of
membranes seem to be related to the development of several
diseases (Lee, 2004). Although drug targets are usually proteins
(Cohen, 2002; Overington et al., 2006), drugs can also bind to
membrane lipids. Non-specific interactions of antiarrhythmics
with myocardial membranes, for instance, are related to their
mode of action (Gourley, 1971). Amlodipine, an antihypertensive
drug, also interacts non-specifically with membranes (Mason et al.,
1991). In a previous work, the non-specific interaction of the
antimalarial drug primaquine with biological membranes was
suggested to represent an additional route in its overall mode of
action and/or could be associated to its adverse effects (Basso et al.,
2011). The importance of the non-specific drug-membrane
interactions as a molecular event can be enhanced by a new
therapeutic approach called membrane-lipid therapy. It is based on
the binding of drugs to lipids, thus modulating the structure of
membranes (Escribá, 2006). To understand the overall mechanism
of action of AQ, studies of AQ-membrane interaction are important
since the molecular targets of this drug, either proteins or lipids
from membranes, are not completely known yet.
A combination of techniques is frequently used to study drug-
membrane interactions (Seydel and Wiese, 2002). In particular,
electron spin resonance (ESR) and differential scanning calorime-
try (DSC), among several available techniques, have provided
important contributions (Zhao et al., 2007; Budai et al., 2003;
Könczöl et al., 2005; Pentak et al., 2008; Hamilton et al., 1991;
Basso et al., 2011). Changes in membrane properties induced by
drugs or other molecules are detected by lineshape alterations in
the ESR spectra or by changes in heat exchange in DSC experiments
(Duarte et al., 2008; Basso et al., 2011; Couto et al., 2011; Dyszy
et al., 2013).
In this paper, interactions of the antimalarial drug AQ with lipid
membrane models were investigated by ESR and DSC. The
membrane models were constituted by an equimolar mixture of
dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphospha-
tidylserine (DPPS), which are representative lipids of the parasite
membrane (Hsiao et al., 1991). The effects of AQ on the melting
behavior of DPPC/DPPS bilayers were monitored by DSC. Using spin
labels attached to different positions along the lipid acyl chains and
to the head group, we investigated the alterations induced by AQ
on the ordering and dynamics of different regions of lipid bilayers
by ESR. We also compared the effects caused by AQ with the effects
caused by cholesterol on the structural dynamics of the binary
system at physiological temperature. A possible drug location in
the lipid bilayer was proposed.
2. Material and methods
2.1. Materials
The lipids 1,2-dipalmitoyl-sn-glycero-3-phosphocholine
(DPPC), 1,2-dipalmitoyl-sn-glycero-3-phospho-L-serine (DPPS),
cholesterol, and the spin labels 1-palmitoyl-2-stearoyl-
(n-doxyl)-sn-glycero-3-phosphocholine (n-PCSL, where n = 5, 7,
10, 12, 14, and 16) and 1,2-dioleoyl-sn-glycero-3-phospho(tempo)
choline (DOPTC) were obtained from Avanti Polar Lipids, Inc.
(Alabaster, AL). The antimalarial drug AQ and the Hepes buffer
(4-(2-hydroxyethyl)-1-piperizineethanesulfonic) were purchased
from Sigma Chem., Co. (St. Louis). All reagents were used without
further purification.
2.2. Methods
2.2.1. Lipid dispersion preparation
A lipid film was formed from a chloroform: methanol (2:1)
solution of DPPC:DPPS (1:1) or DPPC:DPPS:cholesterol (10:10:1),
69
and 0.5 mol% of spin label for ESR experiments, by drying under a
stream of N2 and left under reduced pressure for 2 h to remove all
traces of organic solvent. The film was then hydrated with 10 mM
Hepes, at pH 7.4, to a lipid concentration of 20 mM, heated at 60
C
for 5 min, and vortexed. Part of the lipid dispersion was diluted
with Hepes buffer to a final lipid concentration of 10 mM and
another part was diluted with a 2 mM drug buffered solution to
final lipid and drug concentrations of 10 and 1 mM, respectively.
For ESR measurements with cholesterol-containing samples, a
lipid film was formed from stock solutions of pure DPPC, DPPS,
and cholesterol and the dried film was hydrated with Hepes to a
final concentration of 10 mM of DPPC and DPPS and 1 mM of
cholesterol.
2.2.2. Electron spin resonance
ESR measurements were performed with a JEOL FA-200
X-band spectrometer (JEOL Ltd., Tokyo, Japan). Temperature was
controlled by a circulator bath that pumps fluid through an
aluminum radiator attached to the ESR resonant cavity as
described by Alaouie and Smirnov (2006). The samples were
placed into 1 mm inner diameter glass capillaries and then
concentrated by centrifugation. The glass capillary was
accommodated within a quartz tube. The acquisition conditions
were: field modulation frequency, 100 kHz; field modulation
amplitude, from 1 to 2 G; sweep width, 100 G; microwave power,
10 mW. Data presented here are an average of duplicate experi-
ments and the uncertainties are the respective standard
deviations.
2.2.3. Nonlinear least-squares (NLLS) simulations of ESR spectra
NLLS analyses of the ESR spectra were performed using the
program Multicomponent (https://sites.google.com/site/alten-
bach/labview-programs/epr-programs/), a LabVIEW-based ver-
sion (Altenbach, 2014) of the traditional scheme of Freed et al. for
fitting ESR spectra of nitroxide spin probes (Schneider and Freed,
1989; Freed, 1976). The software calculates an ESR spectrum by
solving the stochastic Liouville equation, allowing for the
quantification of the structural dynamics of spin-labeled
biomolecules in terms of motional rates and orientational
ordering imposed by the local environment around the nitroxide
radical.
The dynamics of spin-labeled lipids in lipid bilayers are better
described by an axially symmetric rotational diffusion tensor,
whose principal components, R// and R?, represent the rotational
diffusion rates around axes parallel and perpendicular, respec-
tively, to the mean symmetry axis for the rotation, zR (Schneider
and Freed, 1989). For n-PC spin labels (n = 5, 7, 10, 12, 14, and 16), zR
is defined along the symmetry axis of the lipid, i.e., the fully
extended direction of the acyl chains (see Fig. S1B of the
Supporting information). For DOPTC, zR is defined to be collinear
to both the N-C4 and the N—O bonds of the tempo-choline group
(Ge and Freed, 1998; Ge and Freed, 2003), which makes it also
parallel to the magnetic xm axis, in which the magnetic frame is
defined (xm points along the N—O bond, zm lies along the 2pz axis
of the nitrogen atom, and ym is perpendicular to both – see
Fig. S1A). We found in our simulations that the spectra are not very
sensitive to R//. Thus, to reduce the number of fitting parameters,
qffiffiffiffiffiffiffiffiffiffiffiffiffi
N  R== =R? was fixed to 1 and R ¼ R2? R== was varied instead. The
3
mobility of the lipid acyl chains and the head groups was then
reported as R?.
The calculated ESR spectra of the spin-labeled lipids are the
result of the so-called MOMD (microscopic order-macroscopic
disorder) model that takes into account the tendency of the spin
labels to become partially ordered with respect to a local director
zD (defined as the normal vector to the bilayer surface), but these
 70 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
[(Fig._1)TD$IG]
(a)
Fig. 1. DSC traces of samples with 10 mM DPPC/DPPS in Hepes buffer with (gray) and without (black) AQ. Inset: zoom of the calorimetric DSC peaks.
local directors are randomly oriented (macroscopically disor-
dered) in the sample (Meirovitch et al., 1984). The microscopic
molecular ordering of the spin labels is characterized by the order
parameters S0 and S2, which are manifested through an orienting
potential on the lipid head groups and the acyl chains that restricts
the range of orientations of the spin probes (Schneider and Freed,
1989; Budil et al., 1996). S0 is a measure of the inclination of zR with
respect to the local director, zD, whereas the nonsymmetric order
parameter S2 gives the deviation from cylindrical symmetry of the
molecular alignment of zR relative to zD. The larger the S0, the
larger is the alignment, and the more restricted is the spin label
motion.
The parameters involved in the simulations were: g-tensor (gxx,
gyy, gzz) and hyperfine-tensor (Axx, Ayy, Azz) components, rotational
diffusion rates (R?, R//), orienting potential coefficients (c20, c22),
and whenever needed, homogeneous lorentzian or Gaussian
linewidths. Seed values of the magnetic tensors were obtained
from Earle et al. (1994), but small variations in their values were
allowed during the fitting process. The strategy of the simulation
follows that of Basso et al. (2011).
As for the uncertainties of the parameters from the NLLS
analyses, we found an estimated error of 0.01 for S0 and 5%
for R? of n-PC spin labels, and of 0.006 for S0 and 3% for R? of
DOPTC. This means that we found good solutions for the
stochastic Liouville equation with different sets of c20, c22 (only
used in the simulations of DOPTC and 16-PCSL), and R?. In
general, good agreement between the experimental and
theoretical spectra is obtained whenever S0 and R? are both
simultaneously increased within the uncertainties. Overall, the
general trends observed in the ordering and dynamics of the
spin labels for different sets of (S0, R?) values and for different
lipid sample preparations were all consistent, even though the
differences in S0 and mostly in R? might be within the
uncertainty of the corresponding parameters.
2.2.4. Differential scanning calorimetry
DSC traces were obtained by heating the samples from 5 to
70
C at a scan rate of 20
C/h with the Microcalorimeter VP-DSC
(Microcal, Northampton, MA) or the calorimeter Nano-DSC-II
(Calorimetry Sciences Corporation, Lindon, Utah, USA). Baseline
subtractions and peak integrals were carried out using MicroCal
Origin software.
(b)
3. Results and discussions
3.1. Calorimetric results show AQ-bilayer interaction
The effect of AQ on lipid membranes was investigated by
determining the thermal behavior of a DPPC/DPPS mixture. Fig. 1
shows the DSC traces of DPPC/DPPS dispersions with and without
AQ. Upon heating, the AQ-free samples exhibited a nonideal lipid
mixing, as also reported by other authors (Sevcsik et al., 2008),
characterized by a broad melting transition. Nonideal lipid mixing
means that the lipids are not randomly distributed in the
membrane but organized in clusters or domains. In this situation
the bilayer displays lateral heterogeneity. This lateral phase
separation was previously observed in DPPC/DPPS mixtures (Luna
and McConnell, 1977; Gordon et al., 2006). DSC traces of AQ-free
and AQ-containing samples presented two calorimetric peaks at
temperatures T1 and T2 (T1 < T2, see the inset of Fig. 1). We suggest
that these peaks could be attributed to the melting of lipids in two
different domains, probably DPPC-rich and DPPS-rich domains.
The thermograms also displayed very low enthalpic pre-transition
peaks that are neither well defined nor reproducible and thus will
not be discussed. The enthalpy change (DH) of the whole transition
and the values of T1 and T2 are presented in Table 1. As observed in
Fig. 1, incorporation of AQ into the DPPC/DPPS lipid mixture
increases the peak intensity at T1, slightly lowers both T1 and T2,
and significantly increases DH by (20  2)% on average (Table 1). In
a previous work, a decrease of the transition temperature was also
observed with primaquine, another aminoquinoline antimalarial
drug, in DMPC bilayers (Basso et al., 2011). However, a much
smaller amount of primaquine (lipid/drug molar ratio 91:1)
was required to produce a similar effect than that caused by AQ
(lipid/drug molar ratio 10:1). On the other hand, while the
transition enthalpy of DMPC bilayers remained unchanged in the
Table 1
Calorimetric parameters of the phase transition of the binary DPPC/DPPS
membranes obtained from the DSC traces.
Sample DH (kcal/mol) T1 (
C) T2 (
C)
DPPC/DPPS 1 17.4 43.23 44.39
2 18.9 43.19 44.40
DPPC/DPPS with AQ 1 21.3 43.00 44.20
2 22.2 43.00 44.12
 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78 71
[(Fig._2)TD$IG]

Fig. 2. Experimental (black) and best-fit (gray) ESR spectra of the head group spin label DOPTC and the acyl chain n-PCSL (n = 5, 7, 10, 12, 14, and 16) in DPPC/DPPS membranes
in the absence (left) and in the presence (right) of AQ. (a) One-component nonlinear least squares fits of DOPTC, 5-, 7-, 10-, and 12-PCSL ESR spectra. (b) Spectra of 14- and
16-PC spin labels were simulated with two components. The arrows show the second, more mobile, and less populated component. The spectra were acquired at 36
C.
72 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
presence of primaquine, AQ induced a large increase in DH
of DPPC/DPPS membranes, suggesting different drug-lipid
interactions and thus different effects on membranes.
Our DSC results show a clear interaction between AQ and the
lipid membranes, since AQ changes the shape of the thermograms,
decreases the melting temperatures and increases the transition
enthalpy of the binary system. These results could indicate a
penetration of AQ into the DPPC/DPPS bilayers, since molecules
that penetrate lipid bilayers usually lower the gel-to-fluid phase
transition (Jain and Wu, 1977). As mentioned before, the narrow
DSC peak at T1 (Fig. 1) can be due to the melting of either DPPC-rich
or DPPS-rich domains. Using infrared spectroscopy, Schultz et al.
showed the coexistence of DPPC-rich and DPPS-rich aggregates in a
binary DPPC/DPPS membranes and also showed that DPPC
molecules melt at a temperature (45.2
C) lower than that of
DPPS (47.8
C) in this lipid system (Schultz et al., 2009). Thus, we
propose that the peak at T1 in our DSC thermograms might be due
to the melting of DPPC in DPPC-rich domains, whereas the peak at
T2 could be related to the melting of DPPS in DPPS-rich domains.
Furthermore, lipids do not melt independently of each other but
rather simultaneously in clusters of a certain amount of lipids
(Heimburg, 2007). Thus, the melting of lipids is a cooperative
phenomenon and the cooperativity is proportional to the number
of lipids in a cluster or domain. The larger the cooperativity, the
narrower and more intense the transition peak. The intensity
increase of the DSC peak at T1 caused by AQ binding to the binary
system is therefore likely due to an increase of the number of lipids
in the domains melting at T1, that is, the size of the DPPC-rich
domains increases. Since there is a direct correlation between
membrane curvature and both enthalpy changes (Yokoyama et al.
(2013)) and lateral phase separation (Duwe et al., 1990; Sackmann,
2006), the aforementioned effects of AQ on the lipid mixture could
be explained by an increase in DPPC/DPPS membrane curvature.
3.2. ESR results suggest AQ penetration into the bilayer
Changes in the order and dynamics of the membrane caused by
AQ were investigated by means of ESR spin labeling technique at
physiological temperature. The hydrophobic region of the bilayer
was monitored by using spin labels located at different depths
along the lipid acyl chain. Fig. 2a and b shows the experimental
and simulated ESR spectra of the head group spin label DOPTC
and the acyl chain spin labels n-PCSL (n = 5, 7, 10, 12, 14, and 16)
acquired at 36
C in the absence and in the presence of AQ.
Generally, the experimental and best-fit theoretical spectra were
in very good agreement and considerable changes were observed
in all ESR spectra upon incorporation of AQ. Except for the 14- and
the 16-PCSL signals (Fig. 2b), all other ESR spectra were simulated
with only one component (one spin label population). The
two-component feature in the 14- and 16-PCSL spectra will be
discussed below.
Fig. 3 presents the order parameter (S0) profile obtained from
the NLLS simulations of the ESR spectra displayed in Fig. 2. Overall,
NLLS fits show that AQ increases the ordering and slightly
decreases the rotational mobility of the whole hydrophobic region
of the lipid bilayers, except for the DOPTC and for 5-PCSL, in which
an opposite effect was observed: for example, AQ slightly
decreases S0 (from 0.711 to 0.692 – Table S2) and significantly
increases R? (from 3.31 to 4.36  107 s1 – Table S2) for 5-PCSL.
Interesting features were reported by some of the spin labeled
lipids and will be discussed in more details.
NLLS simulations of the ESR spectra of the head group spin label
DOPTC provided information on the dynamics and on the
orientation of the nitroxide radical at the bilayer surface. With a
positive value for S0 = 0.518 and a negative value for the nonaxial
order parameter S2 = –0.084, the orientation of the trimethyl
[(Fig._3)TD$IG]
1.0
DPPC/DPPS - comp 1
DPPC/DPPS - comp 2
0.8
DOPTC
DPPC/DPPS/AQ - comp 1
DPPC/DPPS/AQ - comp 2
Order parameter
0.6
0.4
0.2
0.0
HG 4 6 8 10 12 14 16
n-PCSL
Fig. 3. Order parameter profile obtained from non-linear least-squares simulations
of DOPTC and n-PCSL (n = 5, 7, 10, 12, 14, and 16) 36-
C ESR spectra in the absence
(full circle) and in the presence (open triangle) of AQ. HG stands for head group.
ammonium (TMA) group of the DOPTC in the binary DPPC/DPPS
system is almost similar to that of the dipalmitoylphosphatidyl
tempo (2,2,6,6,-tetramethyl-1-oxy) choline (DPPTC) in a pure DPPC
dispersion in the gel phase (S0 = 0.50 and S2 = –0.22 at 25
C)
(Ge and Freed, 1998) or to that of 4-O-(1,2-dipalmitoyl-sn-glycero-
3-phospho) 4-hydroxy-2,2,6,6-tetramethyl-piperidine-1-oxy
(DPP-Tempo) in dioleoylphosphatidylcholine (DOPC) dispersions
at –11
C (S0 = 0.516 and S2 = –0.238) in the presence of excess water
(Ge and Freed, 2003). This means that the principal diffusion axis
(zR), which coincides with xm (parallel to the N—O bond – see
Section 2.2.3), is preferentially aligned along the local director of
the bilayer surface. The main difference in the orientation
of the TMA group of DOPTC in DPPC/DPPS membranes compared
to that of the other spin labels (DPPTC in DPPC or DPP-Tempo
in DOPC) is the lower preference of either xR or yR axes to be
oriented perpendicular to director, due to the low value of S2
(–0.084). To make this clear, we decomposed the spherical tensor
components S0 and S2 in their Cartesian counterparts (Ge and
p
Freed, 2003): Szz = S0 = 0.518, Sxx = 1/2( 3/2 S2 – S0) = –0.310, and
1 p
Syy = – /2( 3/2 S2 + S0) = –0.207. As discussed before, the high
positive value of Szz means a strong preference for zR to be aligned
along the normal to the bilayer, whereas negative and close values
of Sxx and Syy indicate an almost equal preference for xR and yR
diffusion axes to be oriented tangential to the bilayer. The addition
of AQ to DPPC/DPPS samples changes the ordering and dynamics of
DOPTC, which means that AQ binds to and interacts with the lipid
head group of the bilayers. Interestingly, AQ significantly decreases
both the perpendicular rotational mobility R? (from 3.02 to
2.45  107 s1) and the order parameter S0 (from 0.518 to 0.452) of
the spin label. The latter result implies that lipid-AQ interaction
increases the average angular amplitude of the wobbling motion of
the nitroxide moiety (the larger S0, the smaller the amplitude). In
terms of the Cartesian components, Sxx = –0.291, Syy = –0.160, and
Szz = 0.452, we can also infer that AQ binding to membranes
induces zR to be much less oriented along the local director,
whereas xR and yR axes retain their orientation tangential to the
bilayer normal, but yR becomes less oriented.
Another interesting feature observed in Fig. 2a is that the
spectrum of 12-PCSL from the AQ-free sample presented a much
broader line than would be normally expected for this spin label.
This clearly indicates strong spin–spin interactions that likely arise
from cluster formation of 12-PCSL molecules in the DPPC/DPPS
membranes. Partial segregation of spin labels was also noted, for
instance, by Fajer et al. for n-PCSL in the DPPC gel phase in a
 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
saturation transfer ESR study (Fajer et al., 1992) and by Earle et al.
in a rigid-limit 250 GHz ESR study of n-PCSL incorporated in DPPC
and in palmitoyl-oleoyl-PC (POPC) lipid vesicles (Earle et al., 1994).
In fact, the best least-square fits to 12-PCSL ESR spectra were
achieved by allowing for a Heisenberg exchange coupling in the
simulations, which yielded a vHE = 4.9  107 s1. Interestingly,
addition of AQ into the preformed 12-PCSL-clustered DPPC/DPPS
lipid vesicles abolished the spectral broadening, thus suggesting
disruption of those clusters. Direct lipid-AQ interactions might
promote a better miscibility of the 12-PCSL molecules in the binary
system.
A distinct behavior was observed in the spectra of 14- and
16-PCSL (Fig. 2b), which exhibited additional spectral features (see
the low-field and high-field lines) characteristic of a second spin
population, indicating the coexistence of two types of domains in
the 1:1 DPPC/DPPS lipid bilayers, which is in agreement with our
DSC results. A two-component system was also observed, for
instance, by Basso et al. in the rippled phase of dimyristoyl-PC
(DMPC) lipid bilayers (Basso et al., 2011) and by Chiang et al. in a
ternary DPPC/dilauroyl-PC/cholesterol (DPPC/DLPC/Chol) lipid
mixture (Chiang et al., 2004). Our result shows that ESR is very
sensitive to detect the heterogeneity of different micro-environ-
ments in a binary system. Thus, an extra ESR signal with different
ordering and rotational dynamics was added to the fitting process.
From the NLLS fits, we found that the most populated component
of the two spin labels presented the higher order parameter: S0 of
0.367 for 14-PCSL (87% population) and of 0.343 for 16-PCSL (93%
population), S0 values of the less ordered domain were 0.196 for
14-PCSL (13% population) and 0.117 for 16-PCSL (7% population).
As shown in Fig. S2 of the Supporting information, the theoretical
16-PCSL spectrum of the ordered domain (component 1) in the
binary system is very similar to the experimental spectrum of this
same spin label in pure DPPS vesicles at the same temperature.
Similarly, the theoretical spectrum of the less ordered component
resembles, despite the difference in intensity, that of the 16-PCSL
incorporated in pure DPPC lipid bilayers at the same temperature
(Fig. S2). Thus, the coexistence of the ordered and disordered
regions in the binary DPPC/DPPS lipid system is most likely due to
the coexistence of DPPS-rich (ordered) and DPPC-rich (disordered)
domains.
The slight differences in the spin population values of the
ordered and disordered domains for 14- and 16-PCSL suggest that
these spin-labeled lipids have different partition coefficients KP
between those domains. Since the mole fraction of 14-PCSL in the
less ordered state is higher than the population of 16-PCSL in this
same domain (0.13 and 0.07, respectively), 14-PCSL must partition
better in the disordered domains than 16-PCSL. In fact, a slighter
preference of 14-PCSL for the disordered phase compared to
16-PCSL in a two-phase coexisting region of a ternary system has
been reported (Swamy et al., 2006).
Upon AQ addition to the lipid samples, changes in the lineshape
of the 14- and 16-PCSL are observed (Fig. 2b). The nonlinear least-
squares fits show that binding of AQ to the DPPC/DPPS membranes
increases the ordering (S0: from 0.367 to 0.413 for 14-PCSL and from
0.343 to 0.379 for 16-PCSL) and decreases the perpendicular
rotational diffusion rate (R?: from 6.76 to 6.31 107 s1 for 14-PCSL
and from 2.09 to 1.78  108 s1 for 16-PCSL) of the DPPS-rich
domains (Table S2). The overall result is an increased packing of the
bilayer center of the DPPS-rich domain, whereas almost no change
was observed on ordering and dynamics of the center of the
hydrophobic core of the DPPC-rich domains (Table S2). However,
both 14- and 16-PCSL spectra of the AQ-containing samples show
an increase of the spin-labeled lipid population of the less
ordered DPPC-rich domain (from 0.13 to 0.28, for 14-PCSL, and
from 0.07 to 0.17, for 16-PCSL). The redistribution of the spin label
molecules between the two domains upon AQ addition might be a
73
result of two main effects: (1) changes in the equilibrium constant
between the DPPC-rich and DPPS-rich domains and/or (2) changes
in the partition coefficient of the spin labels between the two
regions. We explored these two possibilities as follows. Firstly, we
defined the partition coefficient KP as the ratio of spin-label
concentration in the ordered (Lo) and disordered (Ld) states as in
Swamy et al. (2006):
PðLo Þ ½mol%ofLd 
KP ¼ x
PðLd Þ ½mol%ofLo 
where P(Ld) and P(Lo) are the populations of the spin labels
incorporated in the disordered (Ld) and ordered (Lo) domains,
respectively, obtained from the NLLS fits of the ESR spectra, and the
second term on the right hand side of the equation represents the
percentage of lipids in the DPPC-rich (Ld) and DPPS-rich (Lo)
domains. KP would be unity if spin labels partition equally into
each domain. If we further suppose that the relative populations
(percentage of lipids) of the ordered and disordered domains are
characteristic of the binary system and thus do not change with the
addition of 0.5 mol% of the spin label (this is reasonable since the
acyl chain of spin-labeled lipids are almost identical to that of DPPC
and DPPS), the ratio K between KP values for 14- and 16-PCSL can be
directly calculated as the ratio of the spin label populations
obtained from the NLLS fits:
K 14PC ½PðLo Þ=PðLd Þ14PC
P
K¼ ¼
K 16PC
P
½PðLo Þ=PðLd Þ16PC
Thus, K gives a good estimate of the relative partition coefficient
of both spin labels into the Ld and Lo phases. We found
that K = 0.504 for the AQ-free sample and K = 0.527 for the
AQ-containing sample, that is, the difference is within the
uncertainty of the spin label populations obtained from NLLS
simulations, which is 3%. This result suggests that the partition-
ing of the spin labels in the ordered and disordered domains does
not depend on the AQ binding to membranes at 36
C (at least on
the specific conditions of our experiments). Consequently, changes
on spin label population upon AQ binding reflect changes on the
equilibrium constant between DPPC-rich and DPPS-rich domains
in the binary system. As mentioned before, our DSC experiments
suggest an increase of the cooperativity of the DPPC-rich domains
upon AQ incorporation (see Section 3.1). Our ESR results also
support this hypothesis. Binding of AQ to DPPC/DPPS membranes
could shift the equilibrium constant between DPPC-rich and
DPPS-rich domains towards the DPPC-rich one. As a result, there
may be an increase in the number of lipids in the DPPC-rich
clusters, which can contribute to the increased cooperativity of the
phase transition of this domain.
Taken together, the ESR/NLLS results suggest that AQ inserts
into the DPPC/DPPS membranes and changes the molecular force
field, i.e., the orienting potential, in the headgroup and in the acyl
chain regions. We suggest that both the phenol ring and the
protonated tertiary amine of AQ (at pH 7.4, AQ is mostly
monoprotonated) are most likely located in the headgroup
region of the lipid bilayer. This can enhance the hydrogen bond
network of that region (due to additional hydrogen bonds between
AQ and the lipid head groups) and also decrease electrostatic
repulsion among PS in DPPS-rich domains. Direct AQ-head group
interaction via the aforementioned groups also changes the
orientation of the TMA group of the DOPTC spin label, which
impacts the membrane orienting potential. Additionally, except for
the local environment around carbon 5, the increased S0 values of
almost the entire acyl chain region indicate an increased lateral
packing density of the lipids induced by AQ, which can enhance
van der Waals interactions between lipids. This increased bilayer
 74 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
ordering is consistent with a higher transition enthalpy, since
more heat is required to break molecular interactions and melt the
lipids. The slight reduction of S0 with a remarkably increase of
rotational mobility around carbon 5 – as shown by NLLS
simulations of 5-PCSL spectra – can be explained by the insertion
of the quinoline ring of AQ into the lipid matrix. With the quinoline
ring located into the hydrophobic core of the lipid bilayer, but close
to the polar–apolar interface, and the phenol ring OH and the
tertiary amine directed towards the head group, a free volume
could be created around the very first carbon atoms of the lipid acyl
chains, which can explain the markedly increase of the rotational
degree of freedom of 5-PCSL. Insertion of the quinoline structure
into lipid matrix may also enhance van der Waals attractions
between lipids, which can contribute to increase the lateral
packing. In fact, insertion of both unprotonated and protonated
forms of the quinoline ring of another 4-aminoquinoline
antimalarial drug, chloroquine, into the apolar region of an anionic
micelle has been also proposed (Santos et al., 2014; Usman and
Siddiq, 2013).
To gain further insights on the hypothesis of AQ penetration
into the membrane, as strongly suggested by our results, we also
compared the effect of AQ on acyl chain order and mobility with
those provoked by cholesterol. It is well known that cholesterol is
located within the lipid bilayers (Huang, 1977; Ahmad and Mellors,
1978; Subczynski et al., 1992) and, as will be discussed further,
cholesterol is also soluble in PS bilayers.
Fig. 4 shows the ESR spectra of DOPTC and n-PCSL (n = 5, 7, 10,
12, 14, and 16) in DPPC/DPPS (1:1) and DPPC/DPPS/Cholesterol
(1:1:0.1) model membranes and Table S2 summarizes the
[(Fig._4)TD$IG]
Fig. 4. Experimental (black) and best-fit (gray) of ESR spectra of the head group spin label DOPTC and the acyl chain n-PCSL (n = 5, 7, 10, 12, 14, and 16) in either DPPC/DPPS
(left) or DPPC/DPPS/Chol (right) membranes. Spectra of 14- and 16-PC spin labels were simulated with two components. All the spectra were acquired at 36
C.
dynamics and order parameters obtained from the NLLS
simulations. As before, the spectra of 14- and 16-PCSL were
simulated with two components, with the greater contribution
(most populated) coming from the ordered, DPPS-rich domains,
for both spin labels (Table S2).
Incorporation of 5 mol% of cholesterol into the binary
DPPC/DPPS system causes a rigidifying effect on the whole acyl
chain of the lipids (Table S2). Cholesterol also provokes a
depth-dependent effect on the perpendicular rotational diffusion
rate of the lipids: a slight reduction of R? from carbon 5 to carbon
7 and then a prominent increase of R? from carbon 12 to the end of
the lipid acyl chain (Table S2). This depth-dependent effect of
cholesterol on the structural dynamics of lipid acyl chains in lipid
bilayers is well known (Costa-Filho et al., 2003; Mainali et al.,
2013). Besides that, cholesterol causes no effect on the ordering of
the head groups (S0 = 0.52; S2 = –0.11), but slightly reduces the
rotational mobility (from 3.16 to 3.02  107 s1), as monitored by
DOPTC spin label. Despite the small difference in R? (4.4%), this
variation is higher than the uncertainty calculated for this
parameter (3% – see Section 2.2.3). Also, the observed trend of
reduced head group mobility in cholesterol-containing samples is
obtained in different lipid preparations: there is a consistent
decrease of the ratio between the intensity of the low-field and the
center-field lines of the DOPTC ESR spectrum in the presence of
cholesterol.
Changes in lipid order parameters can be used to infer how deep
foreign molecules go inside the lipid bilayers (Biaggi et al., 1997). To
gain further insights into the localization of AQ inside DPPC/DPPS
bilayers, we plotted the variation of S0 induced by AQ or cholesterol
 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
[(Fig._5)TD$IG]
Fig. 5. Variation of the order parameter, S0, relative to the pure lipid bilayer as a
function of nitroxide position, n, of n-PCSL in DPPC/DPPS membranes in the
presence of AQ or cholesterol. Temperature: 36
C.
(Chol) relative to the pure lipid bilayer as a function of the labeled
carbon atoms of the lipid acyl chain (Fig. 5). AQ and Chol affect the
order parameter along the whole acyl chain with Chol effects
decreasing from n = 7 down the chain, whereas AQ-induced
alterations oscillate in the same region. Furthermore, the increase
in S0 at carbons 14 and 16 caused by AQ in the more ordered
domain is larger than that caused by cholesterol (component
1 – Fig. 5). Unlike Chol, AQ also causes a significant perturbation of
the headgroup organization, which indicates a distinct mode of
organization of the drug in the bilayer.
The previous experiments were carried out with AQ addition to
preformed liposomes (see Section 2), which means that only the
external part (the outer leaflet) of the bilayers was available to
interact with AQ. However, cholesterol was not incorporated into
the bilayers by the same procedure, but rather by preparing the
liposome as a lipid film constituted of DPPC, DPPS, and cholesterol.
Thus, as both the inner and outer monolayers were available to
cholesterol, the effect of AQ on the lipid ordering and mobility
might be even stronger than that caused by cholesterol. To test this
hypothesis, we qualitatively studied the effect of AQ on the acyl
chain ordering and dynamics when the drug was also allowed to
interact with the inner monolayers of the liposomes. In this
condition, we compared the effect of AQ and Chol on the lineshape
of 14-PCSL spectra at temperatures above and below the phase
transition temperatures (above T2 and below T1) of the binary
[(Fig._6)TD$IG]
(a)
Fig. 6. (a) ESR spectra of 14-PCSL in DPPC/DPPS membranes in the absence and in the presence of AQ or Chol, in the gel phase (left) and fluid phase (right). (b) DSC traces of
10 mM DPPC/DPPS in Hepes buffer with (gray) and without (black) cholesterol.
75
system (Fig 6a). Above the phase transition, there is no change in
the lineshape of the spectra due to addition of AQ. However,
cholesterol slightly decreases the ratio between the high-field and
the center-field lines, as indicated by the arrow in the figure, which
means that 5 mol% of Chol slightly decreases the mobility and/or
increases the ordering of the region monitored by 14-PCSL. On the
other hand, below the phase transition, both AQ and cholesterol
broaden the spectrum relative to the pure lipid mixture. This
means that both molecules increase the bilayer ordering, but the
effect caused by AQ is stronger than that of cholesterol, which
corroborates our initial hypothesis. The lineshape alterations in the
14-PCSL spectrum caused by incorporation of AQ into the inner and
outer monolayers (Fig. 6a) are even more pronounced than those
caused by incorporation of the drug only into the outer monolayer
(see Fig. 2b). The effect of cholesterol on the heat capacity of a
DPPC/DPPS/Chol lipid dispersion was also similar to the effect
caused by AQ, though there was no change in the transition
enthalpy (Fig. 6b). Despite the low miscibility of cholesterol in
phosphadityl serines (PS) bilayers reported by several authors
(Bach and Wachtel, 2003), the effect of limiting amounts of
cholesterol on the thermotropic behavior of binary phospholipid
mixture indicated that anionic phospholipids exhibit greater
affinity for cholesterol than do the zwitterionic phospholipids
PC and PE (Van Dijck, 1979). McMullen et al. (2000) also showed
that cholesterol exhibits greater miscibility in anionic PG than
in uncharged glycolipid bilayers. A personal communication
published in the McMullen et al. paper (McMullen et al., 2000)
reports an up-to 67 mol% cholesterol solubility in 1-palmitoyl-
oleoyl-PS bilayers. Cholesterol incorporation, up to 15 mol%, in the
positively charged DODAB bilayers was also reported (Benatti et al.,
2007). We used a cholesterol concentration of 5 mol% of the total
lipid and, considering its reported higher affinity to PS compared
to PC, we suggest that cholesterol partitions better in PS-rich
domains than in PC-rich ones.
The incorporation of cholesterol in the bilayer is related to
hydrophobic effects between the cholesterol and the lipid bilayer.
Since AQ is a hydrophobic drug, and it is worth noting that
its hydrophobicity has been linked to its antimalarial activity
(Elslager et al., 1964), we expect that hydrophobic effects
between AQ and the bilayers could also take place. However,
besides the hydrophobic effect, electrostatic forces between AQ
and the negatively charged DPPC/DPPS bilayers may occur, since at
the pH studied, about 98% of AQ was in the monoprotonated form
(Naisbitt et al., 1997). To verify the role of the bilayer charge in the
interaction with AQ, we measured the effect of the same amount of
AQ on the ESR spectrum of 12-PCSL in DPPC/DPPS bilayers with
(b)
 76 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
[(Fig._7)TD$IG]
Fig. 7. Ratio between the amplitudes of the low- and central-field lines (h+1/h0) of
12-PCSL in 10 mM of DPPC and different amounts of DPPS in the absence (closed
square) and in the presence (open square) of AQ. Temperature: 36
C.
increasing amounts of DPPS (Fig. 7). Although variations can be
observed even in the absence of DPPS, which means that AQ also
interacts with pure DPPC bilayers, it is only with 50% that we
observed significant changes in the ratio h+1/h0, a parameter that
increases to unity with increasing disordering. This indicates that
the bilayer surface charge has a crucial role in the AQ-bilayer
interaction.
3.3. Possible role of drug-membrane interactions to the antimalarial
mechanism of action
The amphiphilic or hydrophobic character of many pharmaco-
logically active compounds confers them different types of
association, such as self-association and specific and/or non-
specific binding to proteins and lipid membranes. Usually, the
plasma membrane is one of their sites of action and, even if the
target is intracellular, the interaction with this first barrier plays an
important role (Schreier et al., 2000). Since several antimalarials
are amphiphilic, such as AQ, chloroquine, primaquine, quinacrine,
mefloquine, and quinine, drug-membrane interactions may
contribute to their whole mechanism of action. Indeed, inter-
actions between halofantrine and lipid bilayers of PC and PE were
observed by DSC measurements (Lim and Go, 1999). Halofantrine
caused a broadening of the phase transition of PC bilayers and a
decrease in both enthalpy and transition temperature. A similar
effect of halofantrine was observed with PE. Our results also
showed a decrease in the transition temperature caused by AQ. In
another study, the effects caused by halofantrine on the phase
transition of DPPC bilayers were compared with those caused
by quinine and mefloquine (Lim and Go, 1995). Similar to
halofantrine, the latter drugs also interacted with DPPC bilayers,
changing the phase transition and decreasing the transition
temperature. Pajeva et al. showed that quinacrine was also able
to interact with membranes, since the drug caused changes in
the thermotropic behavior of DPPS bilayers by decreasing the
transition enthalpy (Pajeva et al., 1996). Primaquine-membrane
interactions were shown by a combination of pressure perturba-
tion calorimetry, DSC and ESR (Basso et al., 2011). It was shown that
primaquine stabilizes the fluid phase of different lipid model
membranes, causing a local disorder on the acyl chains of the
lipids. This fluidizing effect seems to be due to the interaction of
primaquine with the lipid head groups, causing a less dense gel
phase. Interactions of mefloquine, quinacrine, chloroquine, and
quinine with DPPC bilayers were investigated by 2H and 31P NMR
(Zidovetzki et al., 1989). Although chloroquine did not seem to
affect the bilayer structure of DPPC, a weak chloroquine–DPPC
interaction could not be disregarded by the authors. Their results
also pointed that quinacrine interacts only with the surface of the
DPPC bilayers, whereas both quinine and mefloquine penetrate
into the membrane. A further investigation of DPPC–chloroquine
interactions using a fluorescent probe inserted into the bilayer
showed that chloroquine rigidified DPPC membranes (Ghosh et al.,
1995). Our results also show a rigidifying effect on the hydrophobic
core of the bilayers caused by AQ.
It is believed that the heme is the target of both AQ and
chloroquine, since these drugs prevent heme detoxification.
The parasite protects itself against the toxicity of heme by
polymerizing it to insoluble hemozoin. As weak bases, AQ and
chloroquine accumulate in the acidic environment of the parasite's
food vacuole and inhibit heme polymerization. After exiting the
food vacuole, the heme that is not polymerized is degraded by
glutathione. Ginsburg et al. found that AQ and chloroquine
competitively inhibit the degradation of heme by glutathione,
allowing it to accumulate into the membranes (Ginsburg et al.,
1998). Heme, which is released as ferriprotoporphyrin IX (FPIX) in
the presence of oxygen (Jacob and Winterhalter, 1970), disrupts the
membrane barrier properties and disturbs the ion homeostasis in
the cells, which kills the parasites. Famin et at. showed that the
inhibition constant of glutathione-mediated destruction of heme
in ghost membranes by AQ is 25% greater than by chloroquine
(Famin et al., 1999). It was supposed that this difference might be
related to different locations of AQ, chloroquine, and FPIX within
the membranes. It was proposed that chloroquine is located
mainly in the head group region of the lipid bilayer, but not
intercalated into the phospholipids (Ginsburg and Demel, 1983;
Kursch et al., 1983 Zidovetzki et al., 1989). However, AQ is much
more hydrophobic than chloroquine (the partition coefficient of
AQ in octanol is 359% greater than that of chloroquine, according to
Bray et al., 1996) and thus it may intercalate in the bilayer, as we
showed in the present paper. In this way, the non-specific
interaction between lipids and AQ, as other antimalarial drugs,
may have an important role in their overall and complex
mechanism of action.
4. Conclusions
We showed that AQ is able to interact non-specifically with
DPPC/DPPS lipid bilayers affecting the gel phase more than the
lipid fluid phase. AQ seems to influence more organized
lipid domains, since the fluid phase of the model membrane
investigated was not affected. This AQ-bilayer interaction is driven
by a combination of electrostatic forces and hydrophobic effects.
Firstly, the approximation between the drug and the bilayer
surface is driven by electrostatic forces between the positively
charged amino group of AQ and the negatively charged PS
headgroups. Once close the bilayer, hydrophobic effects also
take place. Our results indicate that AQ binding to the binary
DPPC/DPPS system shifts the equilibrium constant between
the disordered DPPC-rich and the ordered DPPS-rich domains
 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
towards the DPPC-rich one. We suggest that AQ is located
completely inside the bilayer with the positively charged amino
group and the OH of the phenol group pointing outwards to the
polar headgroups of the bilayer.
Conflict of interest
The authors declare that there are no conflicts of interest.
Transparency document
The Transparency document associated with this article can be
found in the online version.
Acknowledgments
This research was supported by USP, FAPESP (Grants No.
2010/17662-8, 2011/1755-1 and 2012/20367-3) and CNPQ. The
authors thank Professor P. Ciacanglini and Professor M. T. Lamy for
allowing access to the DSC equipments.
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.
chemphyslip.2014.12.003.
References
C. Altenbach, 2014. LabVIEW programs for the analysis of EPR Data https://sites.
google.com/site/altenbach/Home.
Ahmad, P., Mellors, A., 1978. Nuclear magnetic resonance studies in liposomes: effects
of steroids on lecithin fatty acyl chain mobility. J. Membr. Biol. 41, 235–247.
Alaouie, A.M., Smirnov, A.I., 2006. Ultra-stable temperature control in EPR
experiments: thermodynamics of gel-to-liquid phase transition in spin-labeled
phospholipid bilayers and bilayer perturbations by spin labels. J. Magn. Reson.
182, 229–238.
Bach, D., Wachtel, E., 2003. Phospholipid/cholesterol model membranes: formation
of cholesterol crystallites. Biochim. Biophys. Acta 1610, 187–197.
Basso, L.G.M., Rodrigues, R.Z., Naal, R.M.Z.G., Costa-Filho, A.J., 2011. Effects of the
antimalarial drug primaquine on the dynamic structure of lipid model
membranes. Biochim. Biophys. Acta 1808, 55–64.
Benatti, C.R., Epand, R.M., Lamy, M.T., 2007. Low cholesterol solubility in DODAB
liposomes. Chem. Phys. Lipids 145, 27–36.
Biaggi, M.H., Riske, K.A., Lamy-Freund, M.T., 1997. Melanotropic peptides-lipid
bilayer interaction: comparison of the hormone alpha-MSH to a biologically
more potent analog. Biophys. Chem. 67, 139–149.
Bittman, R., 2004. The 2003 ASBMB-Avanti Award in lipids address: applications of
novel synthetic lipids to biological problems. Chem. Phys. Lipids 129, 111–131.
Blauer, G., Akkawi, M., Bauminger, E.R., 1993. Further evidence for the interaction of
the antimalarial drug amodiaquine with ferriprotoporphyrin IX. Biochem.
Pharmacol. 46, 1573–1576.
Bray, P.G., Hawley, S.R., Mungthin, M., Ward, S.A., 1996. Physicochemical properties
correlated with drug resistance and the reversal of drug resistance in
Plasmodium falciparum. Mol. Pharmacol. 50, 1559–1566.
Budai, M., Szabó, Z., Szogyi, M., Gróf, P., 2003. Molecular interactions between
DPPC and morphine derivatives: a DSC and EPR study. Int. J. Pharm. 250,
239–250.
Budil, D.E., Lee, S., Saxena, S., Freed, J.H., 1996. Nonlinear-least-squares analysis of
slow-motion EPR spectra in one and two dimensions using a modified
Levenberg–Marquardt algorithm. J. Magn. Reson. Ser. A 120, 155–189.
Chiang, Y.W., Shimoyama, Y., Feigenson, G.W., Freed, J.H., 2004. Dynamic molecular
structure of DPPC–DLPC–cholesterol ternary lipid system by spin-label electron
spin resonance. Biophys. J. 87 (4), 2483–2496.
Cohen, P., 2002. Protein kinases – the major drug targets of the twenty-first
century? Nat. Rev. Drug Discov. 1, 309–315.
Costa-Filho, A.J., Shimoyama, Y., Freed, J.H., 2003. A 2D-ELDOR study of the liquid
ordered phase in multilamellar vesicle membranes. Biophys. J. 84 (4),
2619–2633.
Couto, S.G., Cristina Nonato, M., Costa-Filho, A.J., 2011. Site directed spin labeling
studies of – dihydroorotate dehydrogenase N-terminal extension. Biochem.
Biophys. Res. Commun. 414, 487–492.
Duarte, E.L., Oliveira, T.R., Alves, D.S., Micol, V., Lamy, M.T., 2008. On the interaction
of the anthraquinone barbaloin with negatively charged DMPG bilayers.
Langmuir 24, 4041–4049.
Duwe, H.P., Kaes, J., Sackmann, E., 1990. Bending elastic moduli of lipid bilayers:
modulation by solutes. J. Phys. 51, 945–961.
77
Dyszy, F., Pinto, A.P.A., Araújo, A.P.U., Costa-Filho, A.J., 2013. Probing the interaction
of brain fatty acid binding protein (B-FABP) with model membranes. PLoS One 8,
e60198.
Earle, K.A., Moscicki, J.K., Ge, M., Budil, D.E., Freed, J.H., 1994. 250-GHz electron spin
resonance studies of polarity gradients along the aliphatic chains in
phospholipid membranes. Biophys. J. 66, 1213–1221.
Elslager, E.F., Gold, E.H., Tendick, F.H., Werbel, L.M., Worth, D.F., 1964. Amodiaquine
N-oxides and other 7-chloro-4-aminoquinoline N-oxides. J. Heterocycl. Chem. 1,
6–12.
Escribá, P.V., 2006. Membrane-lipid therapy: a new approach in molecular
medicine. Trends Mol. Med. 12, 34–43.
Fajer, P., Watts, A., Marsh, D., 1992. Saturation transfer continuous wave saturation,
and saturation recovery electron spin resonance studies of chain-spin labeled
phosphatidylcholines in the low temperature phases of dipalmitoyl
phosphatidylcholine. Biophys. J. 61, 879–891.
Famin, O., Krugliak, M., Ginsburg, H., 1999. Kinetics of inhibition of glutathione-
mediated degradation of ferriprotoporphyrin IX by antimalarial drugs.
Biochem. Pharmacol. 58, 59–68.
Fitch, C.D., 2004. Ferriprotoporphyrin IX phospholipids, and the antimalarial actions
of quinoline drugs. Life Sci. 74, 1957–1972.
Freed, J.H., 1976. Theory of slow tumbling ESR spectra for nitroxides. In: Berliner, L.J.
(Ed.), Spin Labeling: Theory and applications. Academic Press, New York, pp.
53–132 (Chapter 3).
Ge, M., Freed, J.H., 1998. Polarity profiles in oriented and dispersed
phosphatidylcholine bilayers are different: an electron spin resonance study.
Biophys. J. 74, 910–917.
Ge, M., Freed, J.H., 2003. Hydration, structure, and molecular interactions in the
headgroup region of dioleoylphosphatidylcholine bilayers: an electron spin
resonance study. Biophys. J. 85 (6), 4023–4040.
Ginsburg, H., Demel, R.A., 1983. The effect of ferriprotoporphyrin IX and chloroquine
on phospholipid monolayer and the possible implications to antimalarial
activity. Biochim. Biophys. Acta 732, 316–319.
Ginsburg, H., Famin, O., Zhang, J., Krugliak, M., 1998. Inhibition of glutathione-
dependent degradation of heme by chloroquine and amodiaquine as a
possible basis for their antimalarial mode of action. Biochem. Pharmacol. 56,
1305–1313.
Gordon, V.D., Beales, P.A., Zhao, Z., Blake, C., Mackintosh, F.C., Olmsted, P.D., Cates, M.E.,
Egelhaaf, S.U., Poon, W.C.K., 2006. Lipid organization and the morphology of
solid-like domains in phase-separating binary lipid membranes. J. Phys. Condens.
Matter 18, L415–L420.
Ghosh, A.K., Basu, R., Dey, S., Das, S., Nayak, N.P., Barat, B., Nandy, P., 1995. Lipid-
disordering effect of aspirin on the liposomal membrane of dipalmitoyl
phosphatidyl choline – a fluorescence anisotropy study. Colloid Surf. B 4, 309–311.
Gourley, D.H.R.,1971. Interaction of Drugs with Cells, Springfield.
Hamilton, K.S., Barber, K.R., Davis, J.H., Neil, K., Grant, C.W.M., 1991. Phase behaviour
of amphotericin B multilamellar vesicles. Biochim. Biophys. Acta-Biomembr.
1062, 220–226.
Heimburg, T., 2007. Thermal Biophysics of Membranes. Wiley-VHC, Weinheim.
Hsiao, L.L., Howard, R.J., Aikawa, M., Taraschi, T.F., 1991. Modification of host cell
membrane lipid composition by the intra-erythrocytic human malaria parasite
Plasmodium falciparum. Biochem. J. 274, 121–132.
Huang, C.H., 1977. A structural model for the cholesterol–phosphatidylcholine
complexes in bilayer membranes. Lipids 12, 348–356.
Jacob, H., Winterhalter, K., 1970. Unstable hemoglobins: the role of heme loss in
Heinz body formation. Proc. Natl. Acad. Sci. U. S. A. 65, 697–701.
Jain, M.K., Wu, N.M., 1977. Effect of small molecules on the dipalmitoyl lecithin
liposomal bilayer: III. Phase transition in lipid bilayer. J. Membr. Biol. 34,
157–201.
Könczöl, F., Farkas, N., Dergez, T., Belágyi, J., LÅrinczy, D., 2005. Effect of tetracaine on
model and erythrocyte membranes by DSC and EPR. J. Therm. Anal. Calorim. 82,
201–206.
Kursch, B., Lüllmann, H., Mohr, K., 1983. Influence of various cationic amphiphilic
drugs on the phase-transition temperature of phosphatidylcholine liposomes.
Biochem. Pharmacol. 32, 2589–2594.
Lee, A.G., 2004. How lipids affect the activities of integral membrane proteins.
Biochim. Biophys. Acta 1666, 62–87.
Lim, L.Y., Go, M.L., 1995. A differential scanning calorimetry study of the interaction
of the antimalarial agent halofanthrine with dipalmitoyl phosphatidyl choline
bilayers. Chem. Pharm. Bull. 43, 2226–2231.
Lim, L.Y., Go, M.L., 1999. The antimalarial agent halofantrine perturbs
phosphatidylcholine and phosphatidylethanolamine bilayers: a differential
scanning calorimetric study. Chem. Pharm. Bull. 47, 732–737.
Luna, E.J., McConnell, H.M., 1977. Lateral phase separations in binary mixtures of
phospholipids having different charges and different crystalline structures.
Biochim. Biophys. Acta 470, 303–316.
Mainali, L., Hyde, J.S., Subczynski, W.K., 2013. Using spin-label W-band EPR to study
membrane fluidity profiles in samples of small volume. J. Magn. Reson. 226,
35–44.
Mason, R.P., Rhodes, D.G., Herbette, L.G., 1991. Reevaluating equilibrium and kinetic
binding parameters for lipophilic drugs based on a structural model for drug
interactions with biological membranes. J. Med. Chem. 34, 869–877.
McMullen, T.P., Lewis, R.N., McElhaney, R.N., 2000. Differential scanning
calorimetric and Fourier transform infrared spectroscopic studies of the effects
of cholesterol on the thermotropic phase behavior and organization of a
homologous series of linear saturated phosphatidylserine bilayer membranes.
Biophys. J. 79, 2056–2065.
78 R.P. Barroso et al. / Chemistry and Physics of Lipids 186 (2015) 68–78
Meirovitch, E., Nayeem, A., Freed, J.H., 1984. Analysis of protein-lipid interactions
based on model simulations of electron spin resonance spectra. J. Phys. Chem.
88 (16), 3454–3465.
Naisbitt, D.J., Ruscoe, J.E., Williams, D., O'Neill, P.M., Pirmohamed, M., Park, B.K., 1997.
Disposition of amodiaquine and related antimalarial agents in human
neutrophils: implications for drug design. J. Pharmacol. Exp. Ther. 280, 884–893.
O’Neill, P.M., Bray, P.G., Hawley, S.R., Ward, S.A., Park, B.K., 1998. 4-Aminoquinolines
– past, present, and future: a chemical perspective. Pharmacol. Ther. 77, 29–58.
Olliaro, P., Nevill, C., LeBras, J., Ringwald, P., Mussano, P., Garner, P., Brasseur, P., 1996.
Systematic review of amodiaquine treatment in uncomplicated malaria. Lancet
348, 1196–1201.
Overington, J.P., Al-Lazikani, B., Hopkins, A.L., 2006. How many drug targets are
there? Nat. Rev. Drug Discov. 5, 993–996.
Pajeva, I.K., Wiese, M., Cordes, H.P., Seydel, J.K., 1996. Membrane interactions of
some catamphiphilic drugs and relation to their multidrug-resistance-
reversing ability. J. Cancer Res. Clin. 122, 27–40.
Pentak, D., Sułkowski, W.W., Sułkowska, A., 2008. Calorimetric and EPR studies of
the thermotropic phase behavior of phospholipid membranes. J. Therm. Anal.
Calorim. 93, 471–477.
Pussard, E., Verdier, F., 1994. Antimalarial 4-aminoquinolines: mode of action and
pharmacokinetics. Fundam. Clin. Pharmacol. 8, 1–17.
Sá, J.M., Twu, O., Hayton, K., Reyes, S., Fay, M.P., Ringwald, P., Wellems, T.E., 2009.
Geographic patterns of Plasmodium falciparum drug resistance distinguished by
differential responses to amodiasackquine and chloroquine. Proc. Natl. Acad.
Sci. U. S. A. 106, 18883–18889.
Sackmann, E., 2006. Thermo-elasticity and adhesion as regulators of cell membrane
architecture and function. J. Phys. Condens. Matter 18, R785–R825.
Santos, M.S., Del Lama, M.P.F.M., Ito, A.S., Naal, R.M.Z.G., 2014. Binding of chloroquine to
ionic micelles: effect of pH and micellar surface charge. J. Lumin. 147, 49–58.
Schneider, D.J., Freed, J.H., 1989. Calculating slow motional magnetic resonance
spectra: a user’s guide. In: Berliner, L.J., Reuben, J. (Eds.), Biological Magnetic
Resonance, vol. 8. Plenum Publisher Corporation, New York, pp. 1–76.
Schreier, S., Malheiros, S., de Paula, V., 2000. Surface active drugs: self-association
and interaction with membranes and surfactants: physicochemical and
biological aspects. Biochim. Biophys. Acta 1508, 210–234.
Sevcsik, E., Pabst, G., Richter, W., Danner, S., Amenitsch, H., Lohner, K., 2008.
Interaction of LL-37 with model membrane systems of different complexity:
influence of the lipid matrix. Biophys. J. 94, 4688–4699.
Seydel, J., Wiese, M., 2002. Drug-membrane interactions: analysis, drug
distribution, modeling. Wiley-VHC, Weinheim.
Sherman, I.W., 1977. Amino acid metabolism and protein synthesis in malarial
parasites. Bull. World Health Organ. 55, 265–276.
Schultz, Z.D., Pazos, I.M., McNeil-Watson, F.K., Lewis, E.N., Levin, I.W., 2009.
Magnesium-induced lipid bilayer microdomain reorganizations: implications
for membrane fusion. J. Phys. Chem. B 113, 9932–9941.
Subczynski, W.K., Markowska, E., Gruszecki, W.I., Sielewiesiuk, J., 1992. Effects of
polar carotenoids on dimyristoylphosphatidylcholine membranes: a spin-label
study. Biochim. Biophys. Acta 1105, 97–108.
Swamy, M.J., Ciani, L., Ge, M., Smith, A.K., Holowka, D., Baird, B., Freed, J.H., 2006.
Coexisting domains in the plasma membranes of live cells characterized by
spin-label ESR spectroscopy. Biophys. J. 90 (12), 4452–4465.
Usman, M., Siddiq, M., 2013. Surface and micellar properties of chloroquine
diphosphate and its interactions with surfactants and human serum albumin. J.
Chem. Therm. 58, 359–366.
Van Dijck, P.W., 1979. Negatively charged phospholipids and their position in the
cholesterol affinity sequence. Biochim. Biophys. Acta 555, 89–101.
WHO, 2010. Global report on antimalarial drug efficacy and drug resistance:
2000–2010, Drug Resistance Updates.
WHO, 2011. World Malaria Report 2011.
Yokoyama, H., Ikeda, K., Wakabayashi, M., Ishihama, Y., Nakano, M., 2013. Effects of
lipid membrane curvature on lipid packing state evaluated by isothermal
titration calorimetry. Langmuir 29, 857–860.
Zhao, L., Feng, S.-S., Kocherginsky, N., Kostetski, I., 2007. DSC and EPR investigations
on effects of cholesterol component on molecular interactions between
paclitaxel and phospholipid within lipid bilayer membrane. Int. J. Pharm. 338,
258–266.
Zidovetzki, R., Sherman, I.W., Atiya a De Boeck, H., 1989. A nuclear magnetic
resonance study of the interactions of the antimalarials chloroquine quinacrine,
quinine and mefloquine with dipalmitoylphosphatidylcholine bilayers. Mol.
Biochem. Parasitol. 35, 199–207.