Effect of sterol side chain on ion channel formation by amphotericin b in lipid bilayers

Article abstract of DOI:10.1021/bi500122c

Amphotericin B (AmB) is one of the most efficient antimycotic drugs used in clinical practice. AmB interacts with membrane sterols increasing permeability of fungal membranes; however, it is still unclear how AmB selectively recognizes the fungal sterol, ergosterol (Erg), over other sterols in cell membranes. In this study, we investigated the effect of an Erg side chain on AmB activity by testing a series of Erg analogues that shared the same alicyclic structure as Erg but varied in the side chain structure by using the K⁺ influx assay. The results clearly showed that the sterol side chain is essential for AmB selectivity toward Erg and for the activity of AmB-sterol ion channels. In agreement with our previous findings showing the direct interaction between the drug and Erg, these data suggested that AmB directly recognizes the sterol side chain structure, consequently promoting the formation of ion channels by AmB. Furthermore, the C24 methyl group and Δ22 double bond in the side chain of Erg are equally important for the interaction with AmB. Conformational analysis revealed that the C24 methyl group contributes to the interaction by increasing the van der Waals (VDW) contact area of the side chain, while the Δ22 double bond restricts the side chain conformation to maximize the VDW contact with the rigid AmB aglycone. This study provides direct experimental evidence of the mechanism of AmB selectivity toward fungal Erg.

Full text of DOI:10.1021/bi500122c

Article
pubs.acs.org/biochemistry
Effect of Sterol Side Chain on Ion Channel Formation by
Amphotericin B in Lipid Bilayers
Yasuo Nakagawa, Yuichi Umegawa, Tetsuro Takano, Hiroshi Tsuchikawa, Nobuaki Matsumori,*
and Michio Murata*
Department of Chemistry, Graduate School of Science, Osaka University, 1-1 Machikaneyama, Toyonaka, Osaka 560-0043, Japan
S
* Supporting Information
ABSTRACT: Amphotericin B (AmB) is one of the most
efficient antimycotic drugs used in clinical practice. AmB
interacts with membrane sterols increasing permeability of
fungal membranes; however, it is still unclear how AmB
selectively recognizes the fungal sterol, ergosterol (Erg), over
other sterols in cell membranes. In this study, we investigated
the effect of an Erg side chain on AmB activity by testing a
series of Erg analogues that shared the same alicyclic structure
as Erg but varied in the side chain structure by using the K+
influx assay. The results clearly showed that the sterol side
chain is essential for AmB selectivity toward Erg and for the activity of AmB-sterol ion channels. In agreement with our previous
findings showing the direct interaction between the drug and Erg, these data suggested that AmB directly recognizes the sterol
side chain structure, consequently promoting the formation of ion channels by AmB. Furthermore, the C24 methyl group and
Δ22 double bond in the side chain of Erg are equally important for the interaction with AmB. Conformational analysis revealed
that the C24 methyl group contributes to the interaction by increasing the van der Waals (VDW) contact area of the side chain,
while the Δ22 double bond restricts the side chain conformation to maximize the VDW contact with the rigid AmB aglycone.
This study provides direct experimental evidence of the mechanism of AmB selectivity toward fungal Erg.
theoretical data, including UV/CD spectroscopy,9−12 ion-flux
mphotericin B (AmB) is the most important representa-
Ative of polyene macrolide antibiotics, and it has remained assays,10,13,14 and computational analysis.15−18 Although it has
been assumed that π−π electronic interactions between the two
extra double bonds in Erg and the heptaene moiety of AmB
stabilize the drug-Erg complex,16 the details of their mutual
recognition are largely unknown. In particular, the role of the
side chain has not been fully investigated experimentally,
although some theoretical studies have suggested that the Δ22
double bond significantly contributes to AmB−Erg interaction
by enhancing the rigidity and/or electron density of the side
chain, while the methyl group at the C24 position is less
important.15,16,19 The lack of experimental data can be
attributed to the limited availability of structural derivatives of
Erg from commercial sources. Since AmB−sterol interactions
are known to depend heavily on membrane lipids,20,21 a choice
of phospholipid is important for examining the membrane
activity of AmB using liposomes; in the previous study, we have
demonstrated that 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-
choline (POPC) is a suitable lipid to clearly observe the sterol
specificity of AmB, while 1,2-dimyristoyl-sn-glycero-3-phospho-
choline (DMPC) membrane obscures the selectivity.20 In this
study, we synthesized a series of Erg analogues, which share the
common alicyclic structure but differ in the side chain
the drug of choice for the treatment of life-threatening systemic
fungal infections despite its serious side effects.1,2 It has long
been established that AmB forms ion-permeable channels in
fungal cell membranes in the presence of sterols,3 which disrupt
membrane functions by causing cation flux across the
membrane eventually leading to cell death.4 The selective
toxicity of AmB toward fungi is thought to be underlined by its
higher affinity to ergosterol (Erg)-containing fungal membranes
than to cholesterol(Cho)-containing mammalian mem-
branes.5,6 On the other hand, Burke’s groups has recently
suggested that the channel formation is a secondary mode of
action and that simply binding to Erg is the critical mechanism
of AmB antifungal activity, because AmB binding inhibits
essential Erg functions in yeast physiology, including vacuole
fusion, endocytosis, pheromone signaling, and the activity
control of membrane proteins.7,8 Further understanding of the
molecular mechanisms underlying AmB recognition of Erg
should help improve the efficacy and reduce the side effects of
AmB.
The molecular structure of Erg differs from that of
mammalian Cho by the presence of an additional methyl
group and a double bond in the side chain and another
unsaturated bond in the steroid nucleus. These minor structural
differences significantly increase Erg affinity to AmB compared
to that of Chothis was confirmed by experimental and
Received: January 28, 2014
Revised: April 24, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Article
Figure 1. Chemical structures of amphotericin B and sterol analogues used in this study. Analogues 2−6 share the alicyclic structure with Erg 1,
whereas Cho and brassicasterol (bottom) have a different alicyclic structure.
composition (Figure 1). Except for commercially available Erg
and 7-dehydrocholesterol (7-DHC), the other Erg derivatives
were obtained using novel synthetic schemes. Then we
compared the capability of these sterols to promote AmB-
induced ion flux into POPC liposomes.
vacuum for over 6 h to completely remove the solvent. Lipid
films thus obtained were hydrated with phosphate buffer (0.15
M potassium phosphate and 0.76 μM BCECF, pH 5.8) by
sonication and vortexing. The resultant suspension was frozen
at −20 °C and thawed at 60 °C five times to form large vesicles.
After the sizing of the liposomes by stepwise extrusion (19
times) through a polycarbonate filter (200 nm pore size) using
a syringe extruder Liposofast (AVESTIN Inc., Ottawa,
Canada), the resultant BCECF-entrapping liposomes yielding
mostly large unilamellar vesicles (LUVs) were separated from
the remaining BCECF by gel filtration using Sepharose 4B
(Sigma-Aldrich, St. Louis, MO) and 0.15 M potassium
phosphate buffer (pH 5.8). Lipid concentration in the
liposomes was determined by phospholipids C (Wako Pure
Chemical Industries, Ltd., Japan), and the liposome suspension
was diluted with the same buffer to 0.5 mM lipid concentration.
To monitor the potassium−proton exchange, a 50 μL aliquot of
1 mM FCCP was added to 5 mL of liposome suspension stored
at 6 °C. The suspension (200 μL) was added to a cuvette
containing 1.8 mL of K2HPO4 (0.15 M), which was positioned
in a fluorescence spectrometer (FP-6600, JASCO, Tokyo,
Japan). After addition of 10 μL of 32 μM AmB in DMSO (a
final concentration of AmB was 0.16 μM), the time-dependent
change in the liposomal pH was monitored by the increase in
the fluorescence intensity (excitation, 500 nm and emission,
535 nm) at 6 °C. The initial pH values of the liposome
suspension and liposome lumen were 7.8 and 5.8, respectively.
Valinomycin in aqueous solution (5 mM, 10 μL) was added to
MATERIALS AND METHODS
■
Materials. Erg analogues 22-dihydroergosterol (2), 24-
demethylergosterol (3), 24-dehydroergosterol (4), and desC5-
ergosterol (6) were synthesized from Erg (the detailed
synthesis protocol is provided as Supporting Information).
AmB, Cho, carbonyl cyanide-p-trifluoro-methoxy-phenyl hydra-
zine (FCCP), 2′,7′-bis(carboxyethyl)-4(5)-carboxyfluorescein
(BCECF), and valinomycin were purchased from Nacalai
Tesque (Kyoto, Japan). Erg was obtained from Tokyo Kasei
(Tokyo, Japan), 7-DHC was obtained from MP Biomedicals
Inc. (California, USA), and POPC was obtained from NOF
corp. (Tokyo, Japan). All other chemicals were obtained from
standard vendors and were of analytical grade.
K+ Flux Assays Using BCECF. K+ influx induced by AmB is
directly associated with H+-efflux from the liposomes and can
be monitored as pH change inside the liposomes by
fluorescence intensity of the liposome-entrapped pH sensitive
dye BCECF, as described in our previous report.22 POPC and
sterols, Erg 1, 22-dihydroergosterol 2, 24-demethylergosterol 3,
24-dehydroergosterol 4, 7-DHC 5, desC5-ergosterol 6, were
dissolved in CHCl3, and the solvent was evaporated to form
lipid films in a round-bottom flask, which was then left under a
B
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Article
a
Scheme 1. Reagents and Conditions
a
(a) Pd−C, AcOEt, room temperature (rt), 5 h; (b) LiAlH4, THF, reflux, 2 h; (c) TBAF, rt, overnight, then HPLC purification, 10% from 7; (d) O3,
pyridine, CH2Cl2, −78 °C, then Me2S, 86%; (e) 5-(isopentylsulfonyl)-1-phenyl-1H-tetrazole, KHMDS, THF, −78 °C → rt, overnight, 44% (E/Z =
6/1); (f) LiAlH4, THF, reflux, 2 h; (e) TBAF, THF, rt, overnight, then HPLC purification, 69% from 10; (h) LiAlH4,THF, reflux, 2 h, 56%; (i)
SO3−Py, Et3N, DMSO, CH2Cl2rt, 2 h, 93%; (j) 3-methyl-2-oxo-butylphosphonate, t-BuOK, THF, 0 °C → reflux, 96%; (k) CH3PPh3Br, t-BuOK,
THF, 0 °C → reflux, 91%; (j) TBAF, THF, rt, overnight, 85%; (m) CH3PPh3Br, t-BuOK, THF, 0 °C → rt, overnight, 96%; (n) LiAlH4, THF, reflux,
1.5 h; (o) TBAF, THF, rt, overnight, 71% for 2 steps.
obtain the fluorescence intensity corresponding to 100%
exchange (Fmax); the following changes in fluorescence intensity
(PRCG) method with 7000 maximum iterations. The resultant
structures were then minimized, and redundant conformers
were eliminated. Energy minimization was carried out by using
the truncated newton conjugate gradient (TNCG) method
with 7000 maximum iterations.
were normalized to Fmax
.
For the preincubation experiment, 200 μL of liposome
suspension was mixed with 10 μL of 32 μM AmB in DMSO
and 2.0 μL of 1 mM FCCP in EtOH and incubated for 3 h at 6
°C. The suspension was added to the cuvette containing 1.8
mL of K2HPO4 (0.15 M) positioned in the fluorescent
spectrometer. After the addition of liposome suspension, the
time course of pH change in the liposomes was monitored and
expressed as the percentage to that of Fmax as described above.
Surface Plasmon Resonance (SPR). The experiments
were performed as described in our previous report.23 Briefly,
the liposome solution was prepared with POPC (4 μmol) and
sterols (1 μmol) mixed in CHCl3. After removal of the solvent,
the film was hydrated with PBS buffer, vortexed, and sonicated,
yielding multilamellar vesicles. The suspension was subjected to
several cycles of freezing, thawing, and vortexing to obtain
LUVs, which were passed through a 100 nm polycarbonate
filter and diluted to the final lipid concentration of 0.25 mM.
SPR measurements were carried out at 25 °C using the CM5
sensor chip mounted in a BIAcore T200 analytical system
(BIAcore AB, Uppsala, Sweden). The experimental details are
provided as Supporting Information.
Molecular Modeling of the Sterol Side Chain. The
conformational search and multiple minimizations were carried
out using the Macromodel software version 9.3.5 installed in
the Windows 7 operating system. Sampling of the conforma-
tional space was performed with a mixed torsional/low-mode
sampling (MCMM/LMCS).24 The OPLS 2005 force field25
implemented in the Macromodel program was used for the
conformational searches in 5000 steps. Energy minimization
was carried out by using the Polak-Ribiere conjugate gradient
RESULTS
■
Synthesis of Sterol Analogues. Erg analogues 2, 3, 4, and
6 were synthesized via common intermediate 726 (Scheme 1).
The synthesis of 22-dihydroergosterol 2 commenced with the
direct hydrogenation of 7 to provide an inseparable mixture of
the products carrying the unsaturated (8) and saturated 6,7-
bond. Deprotection of the diene group and removal of tert-
butyldimethylsilyl (TBS) group gave a mixture of 22-
dihydroergosterol 2 and the saturated azo-adduct without
TBS, which could be readily separated by silica gel column
chromatography. The other analogues were successfully
synthesized from aldehyde 9 by the introduction of the
corresponding side chains. Julia-Kocienski coupling between
aldehyde 9 and 5-(isopentylsulfonyl)-1-phenyl-1H-tetrazole27
afforded 10 as an inseparable mixture of E/Z = 6/1. After
deprotection, HPLC purification yielded 24-demethylergosterol
3 as a single isomer. For the synthesis of 24-dehydroergosterol
4, although the Horner−Wadsworth−Emmons reaction
between aldehyde 9 and 3-methyl-2-oxo-butylphosphonate28
was unsuccessful, it occurred smoothly between phosphonate
and deprotected aldehyde 11 to produce the desired ketone 12
with good yield. Subsequent Wittig methylenation of ketone
12, followed by removal of the TBS group, yielded 24-
dehydroergosterol 4. DesC5-ergostrol 6 was synthesized by the
direct methylenation of aldehyde 9 and subsequent depro-
tection of the diene and hydroxy groups.
C
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Article
Figure 2. Effect of Erg analogues on AmB-induced K+ influx. Fluorescence intensity (F/Fmax) was monitored with the pH-sensitive dye BCECF. The
⧫
●
○
liposomes were prepared from POPC mixed with 5 mol % Erg (1) , 22-dihydroergosterol (2) , 24-demethylergosterol (3) , 24-
dehydroergosterol (4) , 7-DHC (5) , and desC5-ergosterol (6) . The results are presented at R(AmB/lipid) = 3.16 × 10−3. A: at 0 s, AmB in
DMSO (final DMSO concentration below 1%) was added to the liposomes. C: AmB was preincubated with the liposomes for 3 h before the
fluorescence recording, and then the pH gradient was formed by raising the pH of the liposome suspension. B and D: K+/H+ exchange rate (r)
obtained from the experiments A and C, respectively, and calculated by fitting to the exponential curve (F = A(1−exp(−t/T)); r immediately after
adding AmB was presented as ΔF = A/T (t = 0). The initial fluorescence values for 60 and 20 s were used for the fitting analysis of A and C data,
respectively. *The initial rate of the fluorescence change for Erg in panel D was too fast to obtain accurate values.
◊
▲
△
K+ Influx Assays. For evaluating the membrane-permeabi-
lizing activity of AmB, usual fluorescence assays such as dye-
leakage from liposomes cannot be used because the pore
formed by membrane-bound AmB and sterol is too small to
pass a dye molecule. We have devised a new fluorescence
method to measure K+ flux into liposomes by monitoring pH
change with a pH-dependent fluorescent dye, BCECF.22 In the
present study, we carried out this K+ influx measurement in two
different ways to investigate the effect of the sterol side chain
on the activityone focused on the early stage of AmB−sterol
complex formation (Figure 2A,B) and the other on the steady
state after 3 h incubation (Figure 2C,D). As shown in Figure
2A,C, any changes in the side chain structure significantly
reduced the K+ influx activity of AmB, clearly indicating that the
side chain in Erg (1) afforded the highest level of AmB-induced
membrane permeabilization among the tested sterols (Figure
2A,C). As seen in the results of the same flux assays with the
sterol content of 10 mol % in Supporting Information, the
relative potency of the sterol analogues in stimulating AmB-
induced K+ flux was not significantly changed at higher sterol
content in liposomes.
further reduction of K+ influx was observed for the liposomes
containing 5, which lacked both the 24-methyl group and the
Δ22-double bond. Interestingly, the replacement of the 24-
methyl group with exomethylene (4), which was expected to
enhance π−π electronic interaction with AmB, rather reduced
the activity to the level comparable to that of the 24-demethyl
derivative. In both experiments presented in Figure 2, almost
no K+/H+ exchange was observed for the membrane containing
desC5-ergosterol 6 that lacked a large part of the side chain
structure 1.
Figure 2D shows the K+-flux activity in the liposomes after 3
h incubation with AmB, which was similar to the results in
Figure 2B for all the derivatives (2−6), implying that the side
chain played roles both in the binding to AmB-containing
membrane and in the stabilization of the AmB−sterol complex
in the membrane; the former is clearly illustrated by the
exchange rates presented in Figure 2A,B, while the latter can be
better evaluated after longer incubation times, as shown in
Figure 2C,D. The detailed comparison, however, revealed
certain dissimilarity between A/B and C/D. The differences
between the flux-promoting activity of Erg and other sterols
were more pronounced after 3 h incubation compared to that
without incubation, suggesting that the side chain structure
contributed to the stabilization of the AmB−sterol complex
rather than to the binding of AmB to the membrane.
In the early stage experiments, K+/H+ exchange rates were
calculated from the changes in fluorescence intensity
immediately after AmB addition by curve fitting (Figure 2B).
It was found that 3 and 4, which lacked the 24-methyl group
and Δ22-double bond, respectively, showed significantly
reduced K+/H+ exchange compared to Erg (1). Consistently,
SPR-Based Affinity Assays. To further evaluate the impact
of the side chain on AmB-sterol interaction in comparison, we
D
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Article
carried out SPR experiments for Erg (1), its analogues 2, 3, 5,
and brassicasterol and Cho (Figure 1) using a sensor chip
coated with the sterol-containing POPC membranes. Since the
kinetic parameters in the association step of AmB such as the
association rate constant ka1 are known to be significantly
influenced by the experimental conditions, particularly the
concentration of AmB and the amount of membrane lipids
immobilized on the sensor chip,23,29 we examined the lifetime
of the AmB-sterol complex using the dissociation time constant
Td2 (Figure 3). The results appeared to be similar to those
experiments;26 this stable interaction between the sterol side
chain and the hydrophobic portion of AmB is plausibly caused
by their contact with larger areas of the molecular surface as
discussed below.
The ordering effect induced by sterols is another important
factor to influence the binding affinity of AmB to lipid bilayers.
Hsueh et al. measured the first moment of quadrupolar splitting
of POPC-d31 to evaluate the ordering of an acyl chain in
membrane containing 10% Cho, 7-DHC or Erg32 and reported
that these three sterols gave rise to a similar effect while Cho is
slightly higher than 7-DHC (5) or Erg, which was opposite to
that of AmB-affinity shown in Figure 3. AmB is known to have
higher affinity to ordered membrane,33,34 thus implying that the
affinity of AmB to sterol-containing membranes is due not to
membrane ordering but largely to AmB−sterol interaction.
The present results also clearly demonstrate that both methyl
groups at C24 and the double bond between C22 and C23 in
Erg are important for the affinity to AmB. Previous studies,
including the recent one by Neumann et al.,35 have suggested
that van der Waals (VDW) forces are essential for the
interaction of AmB with membrane sterols;14−18,36,37 therefore
the weak channel activity and low AmB affinity of demethyl
analogue 3 can be explained by a loss of the VDW contact area
in the side chain. However, the difference in the VDW force
alone seems insufficient to explain the lower activity of 2 and 4,
which have the same bulky side chains as Erg. Baginski’s group
has reported that Cho has higher conformational flexibility of
the side chain compared to Erg;38 therefore, it is not far-fetched
to assume that the conformation of the side chain is partially
responsible for the lower Cho affinity to AmB and less favored
VDW interaction with the rigid AmB aglycone. We compared
the conformational preferences of each sterol analogue. Figure
4 shows the superposition of sterol derivatives in a window of 5
kJ mol−1 calculated by a conformational search algorithm (see
Methods). The major difference resides in the C17−C20−
Figure 3. Lifetime of the AmB−sterol complexes evaluated by the time
constant of dissociation Td2 (experimental details and sensorgrams are
provided as Supporting Information). Bras (brassicasterol) and Cho
have a difference structure in their alicyclic portion from that of Erg.
obtained in the K+ influx assays (Figure 2B,D). Interestingly,
brassicasterol, which has the same side chain as Erg and the
same alicyclic structure as Cho showed weak affinity to AmB,
similar to that of 7-DHC (5), which has the opposite structural
relationships. Thus, their comparable lifetimes may indicate
that the higher affinity to AmB of Erg compared to Cho is
underlined by the additional unsaturated bond in the ring B as
well as by the difference in the side-chain structure. The data on
the K+ flux activity of brassicasterol and other sterol analogues
with different cyclic structures will be published elsewhere.
DISCUSSION
■
In order to investigate the effect of the sterol side chain on the
membrane-permeabilizing activity of AmB, we synthesized a
series of Erg derivatives with different side chain structures,
incorporated them in the liposomes, and analyzed them in the
K+ influx assay. The results clearly demonstrated that AmB-
induced ion-channel formation depended on the structure of
the sterol side chain. Given our previous data obtained by solid
state NMR showing that Erg directly bound to AmB in
hydrated lipid bilayers,20,26 it is reasonable to deduce that the
sterol side chain is directly involved in the interaction between
sterols and AmB, which consequently influences membrane
permeabilization by AmB.
The dependency of AmB binding affinity on the structure of
the sterol side chain indicates that upon the binding to the
membrane, AmB recognizes the sterol side chain residing deep
in the hydrophobic interior of the membrane bilayer. This is in
sharp contrast to the mode of action of other sterol-binding
natural products such as theonellamides and amphidinol 3,
which largely interact with the shallowly embedded 3-OH
group of the sterol A-ring and therefore show less dependency
on the side chain structure.30,31 A possible explanation for AmB
recognition of the side chain during the membrane binding is
that the drug enters into the membrane interior to interact with
the sterol side chain rather than shallowly embedded on the
membrane surface as revealed by our solid-state NMR
Figure 4. Structure superposition of sterol derivatives within a window
of 5 kJ mol−1 from the ground minimum (generated by conforma-
tional search followed by multiple minimization) (a−e). A total of 15,
27, 18, 7, and 30 conformers for 1, 2, 3, 4, and 5 were generated, and
each model was superposed for the minimum RMSD at the steroid
ring. (f) X-ray crystal structure of AmB.39
E
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Article
C22−C23 dihedral angle of the sterols bearing the C22−C23
unsaturated bond (Figure 4, 1, 3, and 4), which is around
−120°, while that of the saturated sterols (2 and 5) is around
60°. Due to this conformational difference, the saturated side
chains of 2 and 5 tend to have perpendicular orientation to the
plane of the steroid rings, which would decrease their affinity
for the rigid AmB aglycone (Figure 3).
Meanwhile, the lower activity of 4 may contradict the
previous data on the significance of π−π electronic interaction
between the AmB polyene chromophore and the sterol side
chain.16 The conformation in Figure 4 may provide an
explanation to its low activity; the conjugated diene moiety
forces s-trans conformation of the C23−C24 bond, and the
resultant orientation of the terminal portion, which is different
from that of 1, would hinder the VDW contact with AmB.
These results confirm that the VDW interaction between the
rigid AmB aglycone and the sterol side chain is critical for the
formation of functional AmB−sterol ion channels. The C24
methyl group contributes to the VDW contact of the side chain
with AmB, whereas the Δ22 double bond restricts the side
chain in the conformation suitable for interaction with AmB.
ACKNOWLEDGMENTS
■
Y.N. as a JSPS fellow is grateful to Japan Society for the
Promotion of Science. A research fellowship to T.T. from the
GCOE program, “Global Education and Research Center for
Bio-Environmental Chemistry,” is also acknowledged.
ABBREVIATIONS
■
AmB, amphotericin B; BCECF, 2′,7′-bis(carboxyethyl)-4(5)-
carboxyfluorescein; Bras, brassicasterol; Cho, cholesterol; 7-
DHC, 7-dehydrocholesterol; DMPC, 1,2-dimyristoyl-sn-glyc-
ero-3-phosphocholine; Erg, ergosterol; FCCP, carbonyl cya-
nide-p-trifluoro-methoxy-phenyl hydrazine; POPC, 1-palmito-
yl-2-oleoyl-sn-glycero-3-phosphocholine; SPR, surface plasmon
resonance; TBS, tert-butyldimethylsilyl
REFERENCES
■
(1) Gallis, H. A., Drew, R. H., and Pickard, W. W. (1990)
Amphotericin B: 30 years of clinical experience. Rev. Infect. Dis. 12,
308−329.
(2) Lemke, A., Kiderlen, A. F., and Kayser, O. (2005) Amphotericin
B. Appl. Microbiol. Biotechnol. 68, 151−162.
(3) Kruijff, B. D., and Demell, R. A. (1974) Polyene antibiotic-sterol
interactions in membranes of Acholeplasma laidlawii cells and lecithin
liposomes. III. Molecular structure of the polyene antibiotic-
cholesterol complexes. Biochim. Biophys. Acta 339, 57−70.
(4) Bolard, B. (1986) How do the polyene macrolide antibiotics
affect the cellular membrane properties? Biochim. Biophys. Acta 864,
257−304.
(5) Zotchev, S. B. (2003) Polyene Macrolide Antibiotics and their
Applications in Human Therapy. Curr. Med. Chem. 10, 211−223.
(6) Baginski, M., Czub, J., and Sternal, K. (2006) Interaction of
amphotericin B and its selected derivatives with membranes: molecular
modeling studies. Chem. Rec. 6, 320−332.
CONCLUSIONS
■
In this study, we performed ion-flux assays and SPR
measurements using Erg derivatives with various structures of
the side chain and showed that the 24-methyl group and the
Δ22 double bond in the side chain are required for the affinity
to AmB and the formation and functional activity of the sterol-
AmB ion channels. The present data imply that AmB directly
recognizes the sterol side chain, consequently promoting the
formation of ion channels with AmB. We also show that both
the C24 methyl group and the Δ22 double bond independently
influence AmB−sterol interaction. Conformational analysis of
different side chain structures suggests that the C24 methyl
group functions to increase the VDW contact area of the side
chain, and the Δ22 double bond restricts the side chain to the
conformation suitable for the VDW contact with AmB. The
findings of this study provide important insight for a better
understanding of bimolecular interaction between AmB and
sterols.
(7) Palacios, D. S., Dailey, I., Siebert, D. M., Wilcock, B. C., and
Burke, M. D. (2011) Synthesis-enabled functional group deletions
reveal key underpinnings of amphotericin B ion channel and antifungal
activities. Proc. Natl. Acad. Sci. U.S.A. 108, 6733−6738.
(8) Gray, K. C., Palacios, D. S., Dailey, I., Endo, M. M., Uno, B. E.,
Wilcock, B. C., and Burke, M. D. (2012) Amphotericin primarily kills
yeast by simply binding ergosterol. Proc. Natl. Acad. Sci. U.S.A. 109,
2234−2239.
(9) Charbonneau, C., Fournier, I., Dufresne, S., Barwicz, J., and
̀
Tancrede, P. (2001) The interactions of amphotericin B with various
sterols in relation to its possible use in anticancer therapy. Biophys.
Chem. 91, 125−133.
(10) Vertut-Croquin, A., Bolard, J., Chabbert, M., and Gary-Bobo, C.
(1983) Differences in the interaction of the polyene antibiotic
amphotericin B with cholesterol- or ergosterol-containing phospholi-
pid vesicles. A circular dichroism and permeability study. Biochemistry
22, 2939−2944.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental details of synthesis of Erg analogues and SPR
experiments. This material is available free of charge via the
Internet at http://pubs.acs.org.
(11) Clejan, S., and Bittman, R. (1985) Rates of amphotericin B and
filipin association with sterol. J. Biol. Chem. 260, 2884−2889.
(12) Chen, W. C., and Bittman, R. (1977) Kinetics of association of
amphotericin B with vesicles. Biochemistry 16, 4145−4149.
(13) Gary-Bobo, C. M. (1989) Polyene-sterol interaction and
selective toxicity. Biochemie 71, 37−47.
AUTHOR INFORMATION
■
Corresponding Authors
*(M.M.) Phone (+81)-6-6850-5774. Fax (+81)-6-6850-5774.
E-mail murata@chem.sci.osaka-u.ac.jp.
*(N.M.) Phone (+81)-6-6850-5790. Fax (+81)-6-6850-5789.
E-mail matsmori@chem.sci.osaka-u.ac.jp.
(14) Hsuchen, C., and Feingold, D. (1973) Selective membrane
toxicity of polyene antibiotics: study on lecithin membrane models
(liposomes). Antimicrob. Agents Chemother. 4, 309−315.
(15) Baginski, M., Resat, H., and Borowski, E. (2002) Comparative
molecular dynamics simulations of amphotericin B-cholesterol/
ergosterol membrane channels. Biochim. Biophys. Acta 1567, 63−78.
(16) Baran, M., Borowski, E., and Mazerski, J. (2009) Molecular
modeling of amphotericin B-ergosterol primary complex in water II.
Biophys Chem. 141, 162−168.
Funding
This study was partly supported by Grant-In-Aids for Scientific
Research (A) (No. 25242073) and for Young Scientists (B)
(No. 25750383) from MEXT, Japan, and also in part as “Lipid-
acting Structure Project (ERATO)” from Japan Science and
Technology Agency.
Notes
(17) Neumann, A., Baginski, M., and Czub, J. (2010) How do sterols
determine the antifungal activity of amphotericin B? Free energy of
The authors declare no competing financial interest.
F
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX

Biochemistry
Article
binding between the drug and its membrane targets. J. Am. Chem. Soc.
132, 18266−18272.
(18) Neumann, A., Czub, J., and Baginski, M. (2009) On the
possibility of the amphotericin B-sterol complex formation in
cholesterol- and ergosterol containing lipid bilayers: A molecular
dynamics study. J. Phys. Chem. B 113, 15875−15885.
(19) Baginski, M., Bruni, P., and Borowski, E. (1994) Comparative
analysis of the distribution of the molecular electrostatic potential for
cholesterol and ergosterol. J. Mol. Struct. Theochem. 311, 285−296.
(20) Matsumori, N., Tahara, K., Yamamoto, H., Morooka, A., Doi,
M., Ohishi, T., and Murata, M. (2009) Direct interaction between
amphotericin B and ergosterol in lipid bilayers as revealed by 2H NMR
spectroscopy. J. Am. Chem. Soc. 131, 11855−11860.
(21) Dufourc, E. J., Smith, I. C. P., and Jarrell, H. C. (1984)
Amphotericin and model membranes; the effect of amphotericin B on
cholesterol-containing systems as viewed by 2H-NMR. Biochim.
Biophys. Acta 776, 317−329.
binding between the drug and its membrane targets. J. Am. Chem. Soc.
132, 18266−18272.
(36) Kasai, Y., Matsumori, N., Umegawa, Y., Matsuoka, S., Ueno, H.,
Ikeuchi, H., Ohishi, T., and Murata, M. (2008) Self-assembled
amphotericin B is probably surrounded by ergosterol: bimolecular
interactions as evidenced by solid-state NMR and CD spectra.
Chem.Eur. J. 14, 1178−1185.
(37) Langlet, J., Berges, J., Caillet, J., and Demaret, J. P. (1994)
Theoretical study of the complexation of amphotericin B with sterols.
Biochim. Biophys. Acta 1191, 79−93.
(38) Czub, J., and Baginski, M. (2006) Comparative molecular
dynamics study of lipid membranes containing cholesterol and
ergosterol. Biophys. J. 90, 2368−2382.
(39) Jarzembska, K. N., Kaminski, D., Hoster, A. A., Malinska, M.,
Senczyna, B., Wozniak, K., and Gagos, M. (2012) Controlled
crystallization, structure, and molecular properties of iodoacetylam-
photericin B. Cryst. Growth Des. 12, 2336−2345.
(22) Takano, T., Konoki, K., Matsumori, N., and Murata, M. (2009)
Amphotericin B-induced ion flux is markedly attenuated in
phosphatidylglycerol membrane as evidenced by a newly devised
fluorometric method. Bioorg. Med. Chem. 17, 6301−6304.
(23) Mouri, R., Konoki, K., Matsumori, N., Ohishi, T., and Murata,
M. (2008) Complex formation of amphotericin B in sterol-containing
membranes as evidenced by surface plasmon resonance. Biochemistry
47, 7807−7815.
(24) Kolossvary, I., and Guida, W. C. J. (1999) Low-mode
conformational search elucidated: Application to C39H80 and flexible
docking of 9-deazaguanine inhibitors into PNP. Comp. Chem. 20,
1671−1684.
(25) Kaminski, G. A., Friesner, R. A., Tirado-Rives, J., and Jorgensen,
W. J. (2001) Evaluation and reparametrization of the OPLS-AA force
field for proteins via comparison with accurate quantum chemical
calculations on peptides. J. Phys. Chem. B 105, 6474−6487.
(26) Umegawa, Y., Nakagawa, Y., Tahara, K., Matsumori, N., Ohsishi,
T., and Murata, M. (2012) Head-to-tail interaction between
amphotericin B and ergosterol occurs in hydrated phospholipid
membrane. Biochemistry 51, 83−89.
(27) Fettes, A., and Carreira, E. M. (2003) Leucascandrolide A:
Synthesis and related studies. J. Org. Chem. 68, 9274−9283.
(28) Lundy, B. J., Popova, S. J., and May, J. A. (2011)
Enantioselective conjugate addition of alkenylboronic acids to
indole-appended enones. Org. Lett. 13, 4958−4961.
(29) Onishi, M., and Kaminori, H. (2013) High-Throughput and
sensitive assay for amphotericin B interaction with lipid membrane on
the model membrane systems by surface plasmon resonance. Biol.
Pharm. Bull. 36, 658−663.
(30) Espiritu, R. A., Matsumori, N., Murata, M., Nishimura, S.,
Kakeya, H., Matsunaga, S., and Yoshida, M. (2013) Interaction
between the marine sponge cyclic peptide theonellamide A and sterols
in lipid bilayers As viewed by surface plasmon resonance and solid-
2
state H nuclear magnetic resonance. Biochemistry 52, 2410−2418.
(31) Kasai, Y., Matsumori, N., Ueno, H., Nonomura, K., Yano, S.,
Murata, M., and Oishi, T. (2011) Synthesis of 6-F-ergosterol and its
influence on membrane-permeabilization of amphotericin B and
amphidinol 3. Org. Biomol. Chem. 9, 1437−1442.
(32) Hsueh, Y. W., Chen, M. T., Patty, P. J., Code, C., Cheng, J.,
Frisken, B. J., Zuckermann, M., and Thewalt, J. (2007) Ergosterol in
POPC membranes: Physical properties and comparison with
structurally similar sterols. Biophys. J. 92, 1606−1615.
(33) Bolard, J., Vertut-Croquin, A., Cnbulska, B. E., and Gary-Bobo,
C. M. (1981) Transfer of the polyene antibiotic amphotericin B
between single-walled vesicles of dipalmitoyl-phosphatidylcholine and
egg-yolk phosphatidylcholine. Biochim. Biophys. Acta 647, 241−248.
(34) Matsuoka, S., and Murata, M. (2002) Cholesterol markedly
reduces ion permeability induced by membrane-bound amphotericin
B. Biochim. Biophys. Acta 1564, 429−434.
(35) Neumann, A., Baginski, M., and Czub, J. (2010) How do Sterols
determine the antifungal activity of amphotericin B? Free energy of
G
dx.doi.org/10.1021/bi500122c | Biochemistry XXXX, XXX, XXX−XXX