Mycosamine orientation of amphotericin B controlling interaction with ergosterol: Sterol-dependent activity of conformation-restricted derivatives with an amino-carbonyl bridge

Article abstract of DOI:10.1021/ja051597r

Amphotericin B (AmB 1) is known to assemble together and form an ion channel across biomembranes. The antibiotic consists of mycosamine and macrolactone moieties, whose relative geometry is speculated to be determinant for the drug's channel activity and

Full text of DOI:10.1021/ja051597r

Published on Web 07/12/2005
Mycosamine Orientation of Amphotericin B Controlling
Interaction with Ergosterol: Sterol-Dependent Activity of
Conformation-Restricted Derivatives with an Amino-Carbonyl
Bridge
Nobuaki Matsumori, Yuri Sawada, and Michio Murata*
Contribution from the Department of Chemistry, Graduate School of Science, Osaka UniVersity,
Toyonaka, Osaka 560-0043, Japan
Received March 12, 2005; E-mail: murata@ch.wani.osaka-u.ac.jp
Abstract: Amphotericin B (AmB 1) is known to assemble together and form an ion channel across
biomembranes. The antibiotic consists of mycosamine and macrolactone moieties, whose relative geometry
is speculated to be determinant for the drug’s channel activity and sterol selectivity. To better understand
the relationship between the amino-sugar orientation and drug’s activity, we prepared conformation-restricted
derivatives 2-4, in which the amino and carboxyl groups were bridged together with various lengths of
alkyl chains. K+ influx assays across the lipid-bilayer membrane revealed that ergosterol selectivity was
markedly different among derivatives; short-bridged derivative 2 almost lost the selectivity, while 3 showed
higher ergosterol preference than AmB itself. Monte Carlo conformational analysis of 2-4 based on NOE-
derived distances indicated that the amino-sugar moiety of 2 comes close to C41 because of the short
bridge, whereas those of 3 and 4 are pointing outward. The mutual orientation of the amino-sugar moiety
and macrolide ring is so rigid in derivatives 2 and 3 that these conformations should be unchanged upon
complex formation in lipid membranes. These results strongly suggest that the large difference in sterol
preference between derivatives 2 and 3 is ascribed to the different orientation of amino-sugar moieties.
These findings allowed us to propose a simple model accounting for AmB-sterol interactions, in which
hydrogen bonding between 2′-OH of AmB and 3â-OH of ergosterol plays an important role.
Chart 1. Structures of AmB 1, Ergosterol, and Cholesterol
Introduction
For over 40 years, amphotericin B (AmB, 1) (Chart 1) has
been a standard drug for treatment of deep-seated systemic
fungal infections, which are a continuing and growing concern
for the management of immunocompromised patients infected
by HIV.1-3 Due to lack of better alternatives coupled with the
infrequency of occurrence of opportunistic resistance in fungal
strains,4 the clinical importance of AmB has even been
enhanced. It is now widely accepted that AmB in phospholipid
membrane associates with sterols to form barrel-stave type pores,
where the AmB molecules are arranged in a quasi-parallel
orientation, their polyhydroxy side pointing inward to constitute
the pore lining and their lipophilic heptaene part directing
outward to interact with the membrane interior (Figure 1).5,6
The presence of such molecular assemblages in cell membrane
of fungi increases ion permeability and alters membrane
potentials, ultimately leading to the cell death.6-9 The pharma-
cological action of AmB is based on its selective toxicity against
eukaryotic microbes over mammalians. It has been repeatedly
shown that ergosterol-containing fungal membranes are more
sensitive to AmB than cholesterol-containing mammalian
membranes.8,10,11 Consequently, the concentration of AmB
(1) Gallis, H. A.; Drew, R. H.; Pckard, W. W. ReV. Infect. Dis. 1990, 12, 308-
329.
(2) Hartsel, S. C.; Bolard, J. Trends Pharmacol. Sci. 1996, 12, 445-449.
(3) Ablordeppey, S. Y.; Fan, P. C.; Ablordeppey, J. H.; Mardenborough, L.
Curr. Med. Chem. 1999, 6, 1151-1195.
(4) Vanden-Bossche, H.; Dromer, F.; Improvisi, I.; LozanoChiu, M.; Rex, J.
H.; Sanglard, D. Med. Mycol. 1998, 36, 119-128.
(8) Vertut-Croquin, A.; Bolard, J.; Chabbert, M.; Gary-Bobo, C. Biochemistry
1983, 22, 2939-2944.
(5) Andreoli, T. E. Ann. N.Y. Acad. Sci. 1974, 235, 448-468.
(6) De-Kruijff, B.; Demel, R. A. Biochim. Biophys. Acta 1974, 339, 57-70.
(7) Cohen, B. E. Biochim. Biophys. Acta 1986, 857, 117-122.
(9) Van-Hoogevest, P.; De-Kruijff, B. Biochim. Biophys. Acta 1978, 511, 397-
407.
(10) Milhaud, J.; Hartman, M. A.; Bolard, J. Biochimie 1989, 71, 49-56.
9
10.1021/ja051597r CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 10667-10675
10667

A R T I C L E S
Matsumori et al.
Chart 2. Structures of Intramolecular Bridged Derivatives 2-4
and Open-Bridged 5
cholesterol. So far, however, there has been no experimental
evidence accounting for the sterol selectivity in terms of the
conformation of AmB.
In this report, to gain insight into the conformation-activity
relationship, particularly for the glycosidic bond, we prepared
conformationally restricted derivatives of AmB 2-4 (Chart 2),
where intramolecular carboxyl and amino groups are bridged
together with various lengths of alkyl chains. These bridges
greatly restrict the rotational freedom around glycoside bonds
and may desirably allow for a single conformation. If this is
the case, it should be possible to deduce the active conformation
of the amino-sugar of AmB upon complexation in the ion
channel; the bridges of derivatives 2-4 are expected to minimize
the conformational difference between a sugar orientation in
solution and that in the channel complex. Consequently, 2-4
showed markedly variable ergosterol preference in K+ influx
assays, which can be interpreted with respect to the alteration
of the amino-sugar position induced by alkyl bridges. On the
basis of these results, we discuss the relationship between the
ergosterol selectivity and orientation of the amino-sugar moiety.
Figure 1. Hypothetical structure of a “barrel-stave” ion channel formed
by 8 AmBs and 8 sterols. Yellow and blue molecules represent AmB and
sterol, respectively.5,6
necessary to elicit the lethal permeability is higher with
cholesterol than with ergosterol.
Extensive experimental efforts, ranging from spectroscopic
measurements12 to electrophysiological examinations, have so
far been made to comprehensively understand the molecular
basis of the trans-membrane pore formed by AmB. However,
two conventional methods for structural biology, X-ray crystal-
lography and liquid NMR, are hardly applicable to membrane-
bound entities, hence leaving the molecular architecture of the
AmB’s ion-permeable channel unelucidated. During the past
decade, various computational approaches have also been carried
out pursuing structural factors involved in the channel formation
of AmB and its specificity toward ergosterol.13 Despite the
potentiality of these approaches, further accumulation of
experimental data on the AmB complex is essential to improve
the accuracy of computer simulations. The overall structure of
AmB comprises two rigid fragments, a macrolide ring and a
mycosamine sugar moiety, which are linked by a rotatable
â-glycosidic bond. Their relative position is thought to partly
control AmB-AmB and AmB-sterol interactions, where inter-
and intramolecular hydrogen bonds should play a crucial
role.13d,h Resat et al. hypothesized that the relative orientation
of the amino sugar with respect to the macrolide portion
influences the selective toxicity of the drug.13g In their model,
the channel structure was stabilized by sterol molecules, which
reside close to AmB by their hydrophobic interactions, subse-
quently resulting in the higher selectivity of ergosterol over
Results
Preparation of Conformationally Restricted AmB Deriva-
tives. We prepared three conformationally constrained deriva-
tives of AmB (2, 3, and 4), where carboxyl and amino groups
of AmB are intramolecularly linked with C4, C6, and C8 alkyl
chain bridges, respectively. These derivatives were expected to
adopt different relative orientations between the amino-sugar
and macrolide ring. This molecular design is rational in the light
of previous reports on the structure-activity relationship; the
presence of the amino group is reportedly essential for the
antibiotic activity, and the protection of the free carboxyl group
by esterification or amidation has little influence on the
membrane activity or sterol selectivity, which suggests that the
relatively rigid conformation of the macrolide ring is hardly
affected by charge on the carboxylic acid.14 The synthetic route
of derivatives 2-4 is represented in Scheme 1. Amino alcohols
with n-butyl, hexyl, and octyl chains were first protected with
an Fmoc group, and then converted to aldehydes by SO3-
pyridine or Swern oxidation. The aldehydes were coupled with
the amino group of AmB by the reductive amino-alkylation with
use of NaBH3CN. Deprotection of the Fmoc group followed
by intramolecular amide formation in the presence of PyBOP
(11) Vertut-Qroquin, A.; Bolard, J.; Gary-Bobo, C. M. Biochem. Biophys. Res.
Commun. 1984, 125, 360-366.
(12) Recent examples are: (a) Fujii, G.; Chang, J. E.; Coley, T.; Steere, B.
Biochemistry 1997, 36, 4959-4968. (b) Gago, M.; Koper, R.; Gruszecki,
W. I. Biochim. Biophys. Acta 2001, 1511, 90-98.
(13) Selected publications are: (a) Baginski, M.; Borowski, E. THEOCHEM
1997, 389, 139-146. (b) Baginski, M.; Bruni, P.; Borowski, E. THEOCHEM
1994, 311, 285-296. (c) Mazerski, J.; Borowski, E. Biophys. Chem. 1996,
57, 205-217. (d) Baginski, M.; Gariboldi, P.; Bruni, P.; Borowski, E.
Biophys. Chem. 1997, 65, 91-100. (e) Baginski, M.; Resat, H.; McCam-
mon, J. A. Mol. Pharmacol. 1997, 52, 560-570. (f) Silberstein, A. J.
Membr. Biol. 1998, 162, 117-126. (g) Resat, H.; Sungur, F. A.; Baginski,
M.; Borowski, E.; Aviyente, V. J. Comput.-Aided Mol. Des. 2000, 14, 689-
703. (h) Baginski, M.; Resat, H.; Borowski, E. Biochim. Biophys. Acta
2002, 1567, 63-78. (i) Mariusz Baran, M.; Mazerski, J. Biophys. Chem.
2002, 95, 125-133. (j) Langlet, J.; Berges, J.; Caillet, J.; Demaret, J. P.
Biochim. Biophys. Acta 1994, 1191, 79-93. (k) Sternal, K.; Czub, J.;
Baginski, M. J. Mol. Model. 2004, 10, 223-232.
(14) (a) Cheron, M.; Cybulska, B.; Mazerski, J.; Grzybowska, J.; Czerwinski,
A.; Borowski, E. Biochem. Pharmacol. 1988, 37, 827-836. (b) Mazerski,
J.; Bolard, J.; Borowski, E. Biochim. Biophys. Acta 1995, 1236, 170-176.
(c) Jarzebski, A.; Falkowski, L.; Borowski, E. J. Antibiot. 1982, 35, 220-
229. (c) Mazerski, J.; Bolard, J.; Borowski, E. Biochim. Biophys. Acta 1995,
1236, 170-176.
9
10668 J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005

Mycosamine Orientation of Amphotericin B
A R T I C L E S
Scheme 1. Preparation of 2-5a
Figure 2. Concentration-effect relationship of K+ influx activities for AmB
(1) and its analogues 2-5 in ergosterol-containing liposomes. Liposomes
were composed of egg-phosphatidylcholine (PC) and ergosterol at the ratio
of 9:1, and lipid concentration (eggPC plus ergosterol) in a liposome
suspension was 12 mM. The y axis denotes the ratio of permeabilized
liposomes, which was determined from a decrease in peak area at δ 1.2 in
31P NMR spectra.
a Reagents: (a) Fmoc-OSu, pyridine, MeOH; (b) SO3-pyridine, tri-
ethylamine, DMSO; or (COCl)2, DMSO, triethylamine, CH2Cl2; (c) AmB,
NaBH3CN, DMF-MeOH; (d) piperidine, DMF-MeOH; (e) PyBOP, HOBt,
N-ethyldiisopropylamine, DMF; (f) propionaldehyde, NaBH3CN, DMF-
MeOH; (g) aminopropane, PyBOP, HOBt, N-ethyldiisopropylamine, DMF
(for product yield in each step, see Experimental Section).
Table 1. K+ Influx Activity Using Liposomes
EC50 (sample/lipid)
×
106
sterol-free
liposomes
cholesterol-containing
liposomes
ergosterol-containing
liposomes
compound
as a condensation reagent afforded bridged derivatives 2-4.
We also attempted to prepare a derivative with a propyl chain
bridge, but the objective compound could not be obtained. This
indicates the spatial distance between the carboxyl and amino
groups is beyond the reach of the maximum length of the amino-
propyl chain, ca. 0.5 nm, or the compound might be decom-
posed, once formed, due to the strain of the short bridge. To
evaluate the influence of the alkyl chains substituted at the amino
and carboxylic acid groups, we also prepared 5, an open-bridged
congener of 3. This compound was obtained by the reductive
aminoalkylation of AmB’s amino group with propanal, followed
by the amidation of the carboxyl group with n-propylamine
(Scheme 1).
K+ Flux Assay Using Liposomes. We then assessed the K+
flux activity of these derivatives across lipid bilayer membrane
using a method reported by Herve et al.15,16 In this assay, K+
concentration was higher outside with trans-membrane pH
gradient (pH 5.5 inside and pH 7.5 outside). Once ion channels
form, K+ influx into liposomes induces H+/K+ exchange in the
presence of proton-transporter FCCP, leading to a rise in inner
pH. The increase in inner pH was monitored by chemical shift
changes of phosphate anions in 31P NMR; a downfield signal
at δ 3.1 stems from phosphate ions entrapped in liposomes
where inside pH rose due to the channel formation, while an
upfield signal at δ 1.2 is due to phosphate in intact liposomes.
Signals of phosphate ions outside the liposomes were quenched
by paramagnetic Mn2+ ions. Accordingly, channel activity
of the drugs can be estimated from the decrease in peak intensity
at δ 1.2 or increase at δ 3.1. The spectral features of 31P NMR
1 (AmB)
2 (n ) 4)
3 (n ) 6)
4 (n ) 8)
5 (n ) 3,3)
47
61
60
180
53
43
86
58
110
62
11
50
5.8
27
27
observed in this assay are described in recent publica-
tions.17-22
Figure 2 shows the concentration-effect relationships for 1-5
in ergosterol-containing liposomes. In this figure, the vertical
axis denotes the reduction in the signal intensity at δ 1.2, which
is equivalent to the increase in the ratio of permeabilized
liposomes. Sigmoidal responses were observed for all of the
compounds tested. From these curves, the molar ratio of AmB
derivative/lipid at which 50% of liposomes were permeabilized
was determined as EC50. Table 1 lists the EC50 values of AmB
and its derivatives in sterol-free, cholesterol-containing and
ergosterol-containing liposomes. Although all of the compounds
elicited membrane-permeabilizing actions, only derivative 3
showed a higher influx activity than that of AmB in ergosterol-
containing liposomes. Interestingly, the effect of cholesterol is
only marginal or rather inhibitory for all of the compounds
tested. Recently, we have demonstrated that cholesterol prevents
ion channel formation by membrane-bound AmB using similar
preparations of liposomes.19 These data further support the idea
(17) Matsumori, N.; Yamaji, N.; Matsuoka, S.; Oishi, T.; Murata, M. J. Am.
Chem. Soc. 2002, 124, 4180-4181.
(18) Yamaji, N.; Matsumori, N.; Matsuoka, S.; Oishi, T.; Murata, M. Org. Lett.
2002, 4, 2087-2089.
(19) Matsuoka, S.; Murata, M. Biochim. Biophys. Acta 2002, 1564, 429-434.
(20) Matsuoka, S.; Murata, M. Biochim. Biophys. Acta 2003, 1617, 109-115.
(21) Matsuoka, S.; Matsumori, N.; Murata, M. Org. Biomol. Chem. 2003, 1,
3882-3884.
(15) Herve, M.; Cybulska, B.; Gary-Bobo, C. M. Eur. Biophys. J. 1985, 12,
121-128.
(16) Herve, M.; Debouzy, J. C.; Borowski, E.; Cybulska, B.; Gary-Bobo, C.
M. Biochim. Biophys. Acta 1989, 980, 261-272.
(22) Matsumori, N.; Eiraku, N.; Matsuoka, S.; Oishi, T.; Murata, M.; Aoki, T.;
Ide, T. Chem. Biol. 2004, 11, 673-679.
9
J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005 10669

A R T I C L E S
Matsumori et al.
Table 2. 1H NMR Chemical Shifts for 1-5 in DMSO-d6 at 25 °C
(500 MHz)a
position
1
2
3
4
5
2
2.16
4.06
2.18
4.06
2.16
4.06
n.d.
3.52
n.d.
n.d.
2.16
4.05
n.d.
3.51
n.d.
n.d.
3.10
3.44
n.d.
4.22
n.d.
2.14
4.05
n.d.
3.49
n.d.
n.d.
3.10
3.45
n.d.
4.24
n.d.
3
4
1.32, 1.38 n.d.
3.51 3.54
1.26, 1.59 n.d.
1.24, 1.56 n.d.
3.09
3.48
1.31, 1.54 1.32, 1.49 n.d.
4.26
1.53
5
6
7
8
3.11
3.37
3.07
3.45
9
10
11
12
14
15
16
17
18
19
20
21
32
33
34
35
36
37
38
39
40
1′
4.17
n.d.
4.22
n.d.
1.09, 1.86 1.03, 1.83 1.12, 1.85 1.06, 1.85 1.06, 1.86
3.99
1.93
4.18
3.99
2.05
4.17
4.00
1.90
4.32
4.01
1.85
4.01
4.02
1.86
4.23
1.54, 2.05 1.73, 1.83 1.50, 1.71 1.45, 1.88 1.42, 1.96
Figure 3. Sterol selectivity of compounds 1-5 expressed as a ratio of
EC50 in sterol-free over that in sterol-containing liposomes (EC50 values
are listed in Table 1). The higher index of ergosterol preference is obviously
seen for 3 and 4, while cholesterol has little effect on the activity of all of
the compounds.
4.40
5.97
6.08
6.19
5.38
2.28
3.09
1.73
5.22
1.11
0.91
1.04
4.61
3.74
2.81
3.21
3.25
1.13
4.25
5.75
6.16
6.07
5.47
2.28
3.10
1.73
5.18
1.11
0.91
1.03
4.36
3.69
2.40
3.01
3.18
1.13
4.23
5.90
6.07
6.15
5.43
2.27
3.09
1.72
5.20
1.12
0.92
1.03
4.36
3.69
2.39
2.98
3.12
1.12
4.27
5.90
6.08
6.16
5.43
2.27
3.07
1.73
5.21
1.12
0.91
1.03
4.25
3.76
2.24
2.91
3.07
1.12
4.33
5.93
6.07
6.15
5.42
2.28
that cholesterol is not involved in formation of the channel
complex.19,23-26
3.06
1.70
5.21
Figure 3 depicts ratios of the activity in sterol-containing
membranes as compared to sterol-free membrane. Although the
ergosterol preference in membrane permeability was somewhat
observed in all of the derivatives prepared, the extent is markedly
different among them. It is noteworthy that the ergosterol
selectivity of compound 3 markedly exceeds that of AmB itself,
whereas that of 2 was only marginal. The higher index of
ergosterol selectivity for 3 as compared to 5 should be due to
the fixation of the amino-sugar position in 3. The drastic
difference in sterol selectivity among bridged derivatives 2-4
is also likely to be attributable to changes in the orientations of
the amino-sugar moiety induced by the different length of bridge
chains.
1.09
0.90
1.02
4.23
2′
3.75
3′
2.28
4′
2.96
5′
3.04
6′
1.12
1′′
2′′
3′′
4′′
5′′
6′′
7′′
8′′
amide
2.71, 2.92 2.65, 2.76 2.56
1.77
1.77
2.69, 3.66 1.21
1.46
2.44, 2.64
1.47
1.46
1.21
n.d.
n.d.
n.d.
n.d.
0.87
2.92, 3.07
1.42
2.74, 3.43 n.d.
0.84
1.74
3.00
7.95
7.96
8.00
7.94
NOE-Constrained Conformational Analysis. To address
geometrical arrangements of the amino-sugars with respect to
the macrolide portion in the bridged derivatives, we conducted
conformational analyses using NMR data in DMSO solutions.
If the structures of the derivatives are well rigidified by the
intramolecular bridges, the solution structures should reproduce
conformations occurring in the channel complex. Table 2 lists
a n.d.: not determined due to signal overlapping.
interests were focused on the location of amino-sugar moieties,
this approximation in macrolide rings would not cause any
significant errors.
Distance restraints for 2-5 were derived from 2D NOESY
spectra in DMSO. The total number of inter-proton distance
restraints derived from the NMR data was 22, 24, 14, and 20
for 2, 3, 4, and 5, respectively. On the basis of these conforma-
tional constraints, molecular modeling studies were performed
using the Monte Carlo conformation search algorithm in Macro-
Model 8.5 package.29 Figure 4 shows superposition of 20 lowest
energy conformations around sugar portions of AmB derivatives
2-5, in conjunction with the crystal structure of N-iodoacet-
ylated AmB.28 As expected, derivatives 2 and 3 have well
converged conformers probably due to constraints of the short
bridges. A slight but obvious difference in orientation of the
amino-sugar portion with respect to the macrolide ring was seen
between 2 and 3, which is probably due to the difference in the
length of bridges by two methylenes. Indeed, a short bridge of
2 appears to force the amino-sugar to locate closer to the C41
1
the assignments of H NMR of AmB27 and its derivatives in
DMSO-d6. The chemical shifts in the macrolide portions of the
derivatives coincided well with those of AmB within 0.1 ppm
except for the C16-C21 portion where the change in space
arrangement of amino-sugar portions is likely to influence the
chemical shifts. Spin-coupling patterns of the macrolide rings
were also similar among the derivatives (data not shown). These
observations indicate that no major change occurs in the
macrolide conformation upon derivatization. Due to the possible
inflexibility of the conjugated double bonds, the macrolide ring
of AmB and derivatives was treated as a motion-restricted part,
in which (30° allowance from the crystal structure28 was given
to each C-C single bond upon calculations. Because our
(23) Cotero, B. V.; Rebolledo-AntUnez, S.; Ortega-Blake, I. Biochim. Biophys.
Acta 1998, 1375, 43-51.
(24) Dufourc, E. J.; Smith, I. C. P.; Jarrell, H. C. Biochim. Biophys. Acta 1984,
776, 317-329.
(25) Ruckwardt, T.; Scott, A.; Scott, J.; Mikulecky, P.; Hartsel, S. C. Biochim.
Biophys. Acta 1998, 1372, 283-288.
(28) Ganis, P.; Avitabile, G.; Mechlinski, W.; Schaffner, C. P. J. Am. Chem.
Soc. 1971, 93, 4560-4564.
(29) Mohamadi, F.; Richards, N. G.; Guida, W. C.; Liskamp, R.; Lipton, M.;
Canfield, C.; Chang, G.; Hendrickson, T.; Still, W. C. J. Comput. Chem.
1990, 11, 440-467.
(26) Cuntinho, A.; Prieto, M. Biophys. J. 1995, 69, 2541-2557.
(27) McNamara, C. M.; Box, S.; Crawforth, J. M.; Hickman, B. S.; Norwood,
T. J.; Rawlings, B. J. J. Chem. Soc., Perkin Trans. 1 1998, 83-88.
9
10670 J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005

Mycosamine Orientation of Amphotericin B
A R T I C L E S
Figure 4. (a-d) Ensembles of 20 lowest energy conformations for 2-5, generated by Monte Carlo algorithm restrained with distance constrains from
2D-NOESY in DMSO-d6. Models are represented around polar head regions of molecules, and superposed for a minimum rmsd. (e) The structure of AmB
from the crystal data of its N-iodoacetyl derivative is shown for comparison.28 (f-i) Side views of the lowest energy conformers for (a-d), respectively. (j)
Side view of (e).
carbonyl carbon, thus leading to the conformation that the plane
of amino-sugar is nearly parallel to the macrolide plane. This
alteration is evidenced by an NOE between H-2′ and amide
proton, which is seen in 2 but not in 3. In contrast, derivatives
4 and 5 yield fluctuated conformations for amino-sugar seg-
ments, probably due to a more flexible long bridge and an open
chain structure, respectively.
Biological Activities. We then evaluated the membrane
permeabilizing activity in biological systems. The derivatives
2-4 induced hemolysis against 1% human erythrocytes at the
9
J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005 10671

A R T I C L E S
Matsumori et al.
Table 3. æ and ψ Angles (deg) of the Most Stable Conformers for
Derivatives 2-4a
are thought to form an intermolecular hydrogen bond with the
complementary carboxyl and amino groups of the neighboring
AmB molecules, thus facilitating AmB-AmB interactions.
Conversely, the closed conformation is deduced to be respon-
sible for formation of an intramolecular hydrogen bond between
the carboxyl and the amino groups. The æ and ψ angles of the
most stable conformers for derivatives 2-4 in our experiments
(Table 3) are not coincident with Baginski’s prediction. The
present study shows that, as opposed to drastic conformational
changes, rather minute conformational alteration can be at-
tributable to the sterol preference of AmB.
compound
æ
ψ
2 (n ) 4)
3 (n ) 6)
4 (n ) 8)
-86.1
-89.4
-85.5
122.1
153.8
160.5
The sterol dependence in membrane permeabilizing activity
of AmB has been hitherto explained as a direct interaction with
different binding affinity between the drug and sterols.13j,16 The
affinity is generally believed to be due to two types of
interactions; one is a hydrogen bond between 3â-OH of sterol
and the amino-sugar moiety of AmB either directly or mediated
by water molecules,13h-j,16 and the other is van der Waals force
between AmB heptaene chromophore and hydrophobic steroid
rings.13b,j,31 The higher affinity of AmB toward ergosterol over
cholesterol is often explained by the presence of an additional
conjugated double bond, which allows the molecular proximity
necessary to maximize the van der Waals interactions with
AmB.13j,31 Regarding the interaction of the mycosamine moiety
and 3â-OH of sterols, Baginski et al. have proposed a model
based on molecular dynamics simulations that 3′-NH2 of AmB
forms a hydrogen bond with 3â-OH of sterols,13h whereas
Mazerski et al. postulated a hydrogen bonding between 2′-OH
and 3â-OH of ergosterol.13i A close examination of the
conformations calculated for 3 suggests that the interaction
between the 3′-NH2 of AmB and the 3â-OH of sterols is less
probable because the amino group is faced away from the
heptaene portion, with which hydrophobic sterol molecules are
thought to interact. In contrast, the position of 2′-OH in 3 seems
to allow for formation of the hydrogen bond with 3â-OH of
the sterol molecule, thus supporting Mazerski’s notion. Note
that 2′-OH of compound 2 is more distant from the heptaene
chromophore, and is thus seemingly unfavorable in interacting
with sterol (as discussed later).
The biological activities of derivatives do not completely
correlate with the results of ion-channel activity using artificial
liposomes. It is difficult to completely account for the factors
responsible for the discrepancy because there are significant
differences between biological and model membranes: integral
membrane proteins, lipid compositions, sterol contents, mem-
brane charges, and the existence of cell wall in fungi. Particu-
larly, in the antifungal test, the activity appears to decrease as
the bridge chain lengthens. This is probably because polysac-
charides of the cell wall form a strong diffusion barrier against
hydrophobic molecules, which may cause difference in pen-
etrability through the cell wall according to hydrophobicity of
the bridged derivatives. More detailed investigations on biologi-
cal activities of these derivatives are now in progress.
a The æ and ψ angles of the crystal structure of N-iodoacetylated AmB
are -87.6° and 142.0°, respectively.28
concentration of 5.5, 2.2, and 2.2 µM, respectively, which is
comparable to AmB (4.8 µM).17 Because derivative 5 was
decomposed in the conditions, we were unable to determine its
hemolytic activity. The derivatives also revealed significant
antifungal action against Aspergillus nigar, although less
effective than AmB; the minimum amounts to show inhibitory
zone on the culture of A. nigar are 10, 20, 50, and 10 µg per
paper disk for derivatives 2-5, respectively (AmB showed
inhibitory zone at 5 µg).17
Discussion
The present results clearly demonstrate the importance of
orientation of a mycosamine moiety relative to a macrolide ring
in membrane permeabilizing activity of AmB. Derivative 2, in
which the amino-sugar adopts a parallel orientation with respect
to the macrolide ring (Figure 4), shows less ergosterol selectiv-
ity, while derivative 3, in which the plane of the sugar is
somewhat twisted with respect to the macrolide, has the highest
preference for ergosterol among the agents tested. Much less
stimulatory effect by ergosterol for an open-bridged congener
5 rules out the possibility that N-alkyl substitution and/or amide
formation at C41 augment ergosterol preference of AmB. The
remarkable difference between 2 and 3 in sterol-dependent
activity (Figure 3) discloses that a minute conformational
alteration in the sugar part dramatically affects the sterol
recognition by AmB; the conformational analysis reveals that
the difference in ψ angles linkage between 2 and 3 is only 30°
(Table 3). The possible conformations calculated in an aqueous
environment converged well for 2 and 3, which is indicative of
the conformational rigidity. In addition, the macrocycles formed
by AmB and bridges are reasonably deduced to expose to
external water when bound to lipid bilayer, and macrocycles
of these sizes are known to be able to adopt a well-defined 3-D
structure even in water media if the macrocycles are sufficiently
rigid.30 It is, therefore, quite likely that the conformations of 2
and 3 are closely similar to those in Figure 4 when bound to
lipid bilayer membrane.
Positions of the amino-sugar are defined by æ and ψ dihedral
angles, where æ is C1′-O-C19-C18 and ψ is C2′-C1′-
O41-C19. Baginski et al. have proposed that the arrangements
between the amino-sugar moiety and the macrolide ring are
largely dominated by two conformers: open (æ is ca. -150°
and ψ is ca. 180°) and closed (æ is ca. -60° and ψ is ca. 180°)
forms.13h In the open conformer, the amino and carboxyl groups
To better understand the mechanism of AmB’s channel
formation, the significance of AmB-phospholipid interaction
as well as AmB-sterol should be appreciated.32 Recent mo-
lecular dynamics simulations indicated the presence of electro-
static interaction between the amino-sugar of AmB and phos-
(30) Reid, R. C.; Kelso, M. J.; Scanlon, M. J.; Fairlie, D. P. J. Am. Chem. Soc.
2002, 124, 5673-5683.
(31) Charbonneau, C.; Fournier, I.; Dufresne, S.; Barwicz, J.; Tancre`de, P.
Biophys. Chem. 2001, 91, 125-133.
9
10672 J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005

Mycosamine Orientation of Amphotericin B
A R T I C L E S
Figure 6. Possible interaction between AmB and ergosterol. Modeling
experiments have revealed that 2′-OH of mycosamine of 3 is closer to 3-OH
of ergosterol in the distance of 0.28 nm, which is the ideal length for
hydrogen bonding, in comparison with that of 2, 0.35 nm (see text for
details).
sugar moiety of derivative 3, which is oriented so as to interact
with sterol hydroxyl group, attracts a sterol molecule to the
heptaene region close enough for the van der Waals contact.
Ergosterol bearing the higher affinity for AmB heptaene can
stay close to 3 and reinforce the channel structure. However,
weakly interacting cholesterol molecules, even if they are drawn
close to the heptaene region of 3 by the hydrogen bond between
the amino-sugar and 3â-OH of cholesterol, might be easily
replaced by phospholipid molecules (Figure 5), thus resulting
in little or no effect on the channel activity.
Figure 5. Schematic diagram for sterol selectivity in derivatives 2 and 3.
(a) For 2, 3â-OH of sterol molecules are less interactive with 2′-OH of the
amino-sugar moiety due to the distant location of the OH, and therefore
sterol exerts little or no effect on the channel activity. (b) In derivative 3,
sterols can interact better due to the twisted orientation of the amino-sugar.
An AmB-ergosterol complex with the greater hydrophobic interaction is
more stable than AmB-cholesterol.
phate of phospholipid,13k and very recently we have also
succeeded in the direct observation of the close contacts between
phosphate groups of phospholipid and AmB’s polar part using
a solid-state NMR technique.33 Besides, it has been revealed
that the length and/or saturation of phospholipid acyl chains
also influence the channel activity of AmB, further indicating
the importance of hydrophobic interactions between acyl-chain
of phospholipid and AmB.20,32d Those results suggest the
similarity in interaction modes between AmB-sterol and
AmB-phospholipid. With this in mind, we hypothesize that
both phospholipids and sterols can interact with the amino-sugar
and heptaene regions of AmB as interaction sites, thus being
rather competitive than cooperative upon complexation with
AmB as implied in a previous report.34 If this is the case, our
data can be interpreted as in Figure 5; the amino-sugar
orientation in 2 is unfavorable to form hydrogen bonding
between 2′-OH of AmB and 3â-OH of sterols, and therefore
sterol molecules cannot come close to 2, thus making little or
no effect on ion channel activity of 2. In contrast, the amino-
The greater ergosterol selectivity of 4 than AmB is possibly
ascribable to the conformation of the amino-sugar moiety, which
is very close to that of 3 in the most stable conformer (Table
3). In contrast to its EC50 value in ergosterol-containing
membrane, which is comparable with the other congeners (Table
1), 4 reveals weak activity for sterol-free membrane (Table 1).
This may be due to its bulkier N-substituent, which possibly
interferes with electrostatic interaction between 4 and phospho-
lipid, and thus destabilizes the channel assemblage. However,
4 can accommodate ergosterol in the place of phospholipid
because, upon interacting with the sterol, 2′-OH facing down-
ward is not hindered by the N-substituent.
In the present study, we experimentally demonstrated that
the relative positions of the mycosamine moiety have a strong
influence on the sterol selectivity in channel activity of AmB
by using conformation-restricted derivatives of AmB. Because
the ion channel property of 3 is similar yet superior to that of
AmB itself, the conformation of 3 is thought to reflect the active
conformation of AmB in the channel complex to a considerable
extent. The vertical position between AmB and ergosterol is
reportedly restricted due to hydrophobic interaction.13i Our
preliminary simulations for their bimolecular fitting based on
van der Waals force also suggested the formation of stable
complexes at a certain vertical position, although other inter-
molecular affects were not taken into account. As illustrated in
Figure 6, when ergosterol is placed at this level, the distance
between the 2′-OH of derivative 3 and 3-OH of ergosterol
matches the stable hydrogen-bond length of 0.28 nm. On the
(32) (a) Ruckwardt, T.; Scott, A.; Scott, J.; Mikulecky, P.; Hartsel, S. C. Biochim.
Biophys. Acta 1998, 1372, 283-288. (b) Cotero, B. V.; Rebolledo-Antunez,
S.; Ortega-Blake, I. Biochim. Biophys. Acta 1998, 1375, 43-51. (c) Paquet,
M. J.; Fournier, I.; Barwicz, J.; Tancre`de, P.; Auger, M. Chem. Phys. Lipids
2002, 119, 1-11. (d) Min˜ones, J., Jr.; Dynarowicz-Latka, P.; Conde, O.;
Min˜ones, J.; Iribarnegaray, E.; Casas, M. Colloids Surf., B 2003, 29, 205-
215. (e) Fournier, I.; Barwicz, J.; Tancrede, P. Biochim. Biophys. Acta 1998,
1373, 76-86.
(33) Matsuoka, S.; Ikeuchi, H.; Matsumori, N.; Murata, M. Biochemistry 2005,
44, 704-710.
(34) Min˜ones, J., Jr.; Conde, O.; Min˜ones, J.; Rodriguez-Patino, J. M.; Seoane,
R. Langmuir 2002, 18, 8601-8608.
9
J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005 10673

A R T I C L E S
Matsumori et al.
stirred for 23 h, the mixture was poured into diethyl ether to form
precipitate. The precipitate was filtered over Celite, washed with diethyl
ether, and extracted with CHCl3-MeOH. The extract was purified by
HPLC to afford 2 (1.3 mg, 2.5%). HPLC conditions: column, COS-
MOSIL 5C18-AR-II Φ 4.6 × 150 mm, flow rate, 0.5 mL/min, mobile
phase, MeOH-5 mM ammonium acetate (pH 5.3) changing linearly
from 70:30 to 100:0 for 30 min, retention time, 14.0 min. ESI-MS m/z
contrary, the 2′-OH of 2 resides beyond the hydrogen-bonding
limit with the same ergosterol position, thus supporting the
notion in Figure 5. Although further experimental data should
be necessary to elucidate the molecular interaction between
AmB and ergosterol, the present results may provide clues to
understand the structure basis for the drug’s selective toxicity,
which may lead to the design of new antibiotics with improved
medicinal properties and less side effects.
1
977.3 (M + H)+. H NMR (500 MHz, DMSO-d6): see Table 2.
Derivative 3. N-(6-Aminohexyl)AmB 8b was obtained by the
procedure reported previously.17 To a solution of 8b (21 mg, 20 µmol)
in DMF (2 mL) were added diisopropylethylamine (6.9 µL, 40 µmol),
1-hydroxybenzotriazole (3.7 mg, 24 µmol), and PyBOP (10.4 mg, 20
µmol) sequentially. After being stirred for 22 h, the mixture was poured
into diethyl ether to form precipitate. The precipitate was filtered on
Celite, washed with diethyl ether, extracted with CHCl3-MeOH, and
purified by HPLC to afford 3 (1.3 mg, 2.5%). HPLC conditions were
the same as those for 2, retention time: 16.5 min. ESI-MS m/z 977.3
Experimental Section
Materials and Methods. Amphotericin B (AmB), egg phosphati-
dylcholine, amino-alcohols, cholesterol, and ergosterol were from
Nakarai Tesque. Deuterated solvents were purchased from Merck. All
other chemicals were obtained from standard venders and used without
further purification. NMR spectra were recorded on a JEOL GXS-500
spectrometer. ESI-MS spectra were measured on an LCQ-deca (Thermo
Finnigan). HPLC was performed on a Shimadzu LC-10ADvp with SPD-
M10Avp photodiode array detector. Thin-layer chromatography (TLC)
was performed on a glass plate precoated with silica gel (E. Merck
Kieselgel 60 F254). Column chromatography was performed with silica
gel 60 (E. Merck, particle size 0.063-0.200 mm, 60-230 mesh).
N-(9-Fluorenylmethoxycarbonyl)-4-amino-1-butanol (6a). To a
stirred solution of 4-amino-1-buthanol (342 µL, 3.71 mmol) in MeOH
(30 mL) were added 9-fluoromethyl succinimidyl carbonate (1000 mg,
2.96 mmol) and then pyridine (800 µL, 9.90 mmol). After being stirred
at 23 °C for 23 h, the mixture was diluted with water (100 mL) and
extracted with ethyl acetate. The organic layer was dried over MgSO4
and concentrated in vacuo to give 6a as white amorphous (920 mg,
100%). Rf 0.70 (CHCl3:MeOH ) 10:1). 1H NMR (500 MHz, CDCl3):
δ 7.75 (d, 2H, J ) 7.5 Hz), 7.57 (d, 2H, J ) 7.5 Hz), 7.38 (t, 2H, J )
7.5 Hz), 7.30 (t, 2H, J ) 7.5 Hz), 4.85 (brs, 1H), 4.39 (d, 2H, J ) 6.5
Hz), 4.20 (t, 1H, J ) 6.5 Hz), 3.66 (m, 2H), 3.23 (m, 2H).
1
(M + H)+. H NMR (500 MHz, DMSO-d6): see Table 2.
N-(9-Fluorenylmethoxycarbonyl)-8-amino-1-octanol (6c). 8-Amino-
1-octanol, a known compound, was prepared by a modification of
standard procedures. Briefly, 1,8-octanediol was treated with tosyl
chloride, triethylamine, and (dimethylamino)pyridine in CH2Cl2 to give
a mono-tosylated compound. The mono-tosylate was mixed with
sodium azide in DMF to furnish 8-azido-1-octanol, which was then
reduced by hydrogenation with Pd-C in MeOH to afford 8-amino-1-
octanol. To a stirred solution of 8-amino-1-octanol (123 µL, 0.85 mmol)
in MeOH (8 mL) were added 9-fluoromethyl succinimidyl carbonate
(357 mg, 1.06 mmol) and then pyridine (185 µL, 2.29 mmol). After
being stirred at 23 °C for 22 h, the mixture was diluted with water
(100 mL) and extracted with ethyl acetate. The organic layer was dried
over MgSO4, and concentrated in vacuo, yielding 6c as white amor-
phous (175 mg, 52%). Rf 0.26 (CHCl3:MeOH ) 10:1). 1H NMR (500
MHz, CDCl3): δ 7.75 (d, 2H, J ) 7.5 Hz), 7.57 (d, 2H, J ) 7.5 Hz),
7.38 (t, 2H, J ) 7.5 Hz), 7.30 (t, 2H, J ) 7.5 Hz), 4.71 (br, 1H), 4.38
(d, 2H, J ) 7.0 Hz), 4.20 (t, 1H, J ) 7.0 Hz), 3.62 (m, 2H), 3.17 (m,
2H).
N-(9-Fluorenylmethoxycarbonyl)-4-aminobutanal (7a). A CH2-
Cl2 solution (5 mL) of DMSO (177 µL, 2.4 mmol) and a CH2Cl2
solution (10 mL) of 6a (150 mg, 0.48 mmol) were added sequentially
to a solution of oxalyl chloride (104 µL, 1.2 mmol) dissolved in CH2-
Cl2 (5 mL) at -78 °C. After being stirred for 30 min at -78 °C, the
solution was treated with triethylamine (409 µL, 2.88 mmol) and stirred
for 10 min at 0 °C. The resulting solution was diluted with water and
extracted with diethyl ether. The organic layer was dried over MgSO4,
and concentrated in vacuo, yielding 7a as white amorphous (130 mg,
N-(9-Fluorenylmethoxycarbonyl)-8-amino-1-octanal (7c). To a
DMSO solution (2.5 mL) of 6c (85 mg, 0.23 mmol) were added
sequentially triethylamine (600 µL, 3.5 mmol) and SO3-Py (400 mg,
1.9 mmol). After being stirred for 30 min at 23 °C, the solution was
diluted with water and extracted with diethyl ether. The organic layer
was dried over MgSO4, and concentrated in vacuo, yielding a crude
mixture of 7c (crude 73 mg), which was used without further
purification.
1
94%). Rf 0.53 (CHCl3:MeOH ) 10:1). H NMR (500 MHz, CDCl3):
δ 9.77 (s, 1H), 7.75 (d, 2H, J ) 7.5 Hz), 7.57 (d, 2H, J ) 7.5 Hz),
7.38 (t, 2H, J ) 7.5 Hz), 7.30 (t, 2H, J ) 7.5 Hz), 4.80 (br, 1H), 4.39
(d, 2H, J ) 6.5 Hz), 4.20 (t, 1H, J ) 6.5 Hz), 3.21 (m, 2H), 2.48 (m,
2H), 1.82 (m, 2H).
N-{N-(9-Fluorenylmethoxycarbonyl)-8-aminooctyl}-AmB (8c). A
solution of 7c (13 mg, 36 µmol) and AmB (50 mg, 54 µmol) in DMF-
MeOH (4:3, 7 mL) was stirred for 2 h, and NaBH3CN (14 mg, 222
µmol) was added to the solution. After being stirred overnight, the
solution was poured into diethyl ether (100 mL) to form a yellow
precipitate, which was filtered over Celite and washed with diethyl
ether. The precipitate was extracted with CHCl3-MeOH-H2O 10:6:1
and purified by SiO2 column chromatography with the same solvent
system to afford 8c (28 mg, 41%) as a yellow solid. Rf 0.53 (CHCl3:
MeOH:H2O ) 10:6:1). ESI-MS m/z 1273.5 (M + H)+.
Derivative 4. 8c (28 mg, 28 µmol) was dissolved in DMF-MeOH
(1:1, 2 mL) and mixed with piperidine (500 µL, 5 mmol). After the
mixture was stirred for 30 min, diethyl ether was added to the solution
to form a yellow precipitate. The precipitate was isolated on Celite,
washed with diethyl ether, and extracted with CHCl3-MeOH to give
crude N-(8-aminooctyl)-AmB 9c (20 mg), which was used for next
reaction without further purification. To a solution of 9c (16 mg) in
DMF (2 mL) were added diisopropylethylamine (6.6 µL, 38 mmol),
1-hydroxybenzotriazole (3.7 mg, 24 mmol), and PyBOP (10.4 mg, 20
mmol) sequentially. After being stirred for 21 h, the products were
precipitated with diethyl ether (100 mL). The precipitate was extracted
and purified with the same method as those for derivative 2, HPLC
N-{N-(9-Fluorenylmethoxycarbonyl)-4-aminobututyl}-AmB (8a).
A solution of 7a (13 mg, 43 µmol) and AmB (50 mg, 54 µmol) in
DMF-MeOH (6:4, 10 mL) was stirred for 2 h, and then NaBH3CN
(14 mg, 222 µmol) was added to the solution. After being stirred
overnight, the solution was poured into diethyl ether (100 mL) to form
a yellow precipitate, which was filtered over Celite and washed with
diethyl ether. The precipitate on the Celite were extracted with CHCl3-
MeOH-H2O 10:6:1 and purified by column chromatography on SiO2
with the same solvent system to afford 7a (23 mg, 44%) as a yellow
solid. ESI-MS m/z 1217.7 (M + H)+.
Derivative 2. To a solution of 7a (23 mg, 19 µmol) in DMF-MeOH
(1:1, 2 mL) was added piperidine (500 µL, 5 mmol). After the mixture
was stirred for 30 min, diethyl ether was added to the solution to form
a yellow precipitate. The precipitate was filtered over Celite and washed
with diethyl ether. The product was extracted with CHCl3-MeOH to
give crude N-(4-aminobutyl)-AmB 8a (16 mg), which was used for
next reaction without further purification because of instability. To a
solution of N-(4-aminobutyl)-AmB (16 mg) in DMF (2 mL) were added
diisopropylethylamine (6 µL, 35 µmol), 1-hydroxybenzotriazole (3 mg,
20 µmol), and PyBOP (8.6 mg, 17 mmol) sequentially. After being
9
10674 J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005

Mycosamine Orientation of Amphotericin B
A R T I C L E S
1
retention time: 20.4 min. ESI-MS m/z 1033.8 (M + H)+. H NMR
(500 MHz, DMSO-d6): see Table 2.
Liposome Preparation. Large unilamellar vesicles (LUV) were
prepared according to methods reported by Herve et al.15 Briefly, egg
phosphatidylcholine (52.0 µmol) or egg phosphatidylcholine-sterol (46.8
and 7.2 µmol) was dissolved in CHCl3 (2 mL), and the mixture was
evaporated to a thin film in a round-bottom flask. After being dried in
vacuo for over 8 h, pH 7.5 buffer (0.726 mL) containing KH2PO4 (0.4
mM) and EDTA (1 mM) was added to the flask. The lipid mixture
was suspended in the buffer by vortexing and sonication. The resultant
suspension was frozen at -20 °C and thawed at 50 °C three times.
The liposome solution thus obtained was passed repeatedly through a
membrane filter (pore size, 0.2 µm, 19 times) with a Liposofast
apparatus (AVESTIN), and then diluted with K2SO4 (0.4 M, 3.63 mL)
to give 12 mM LUV solution.
K+ Flux Assays Using 31P NMR. For K+ flux assays, the LUV
suspension prepared above was adjusted to pH 7.5 with 10 M KOH. A
750 µL aliquot of the LUV suspension was transferred to a 1.5 mL
Eppendorf tube and mixed with 2 µL of 10 mM carbonyl cyanide-p-
trifluoromethoxyphenyl hydrazone (FCCP) ethanol solution and 2 µL
of a DMSO solution of drugs. After incubation for 6 h at room
temperature, a portion of 550 µL was transferred to a 5-mm NMR glass
tube and mixed with 4.4 µL of 100 mM MnCl2. 31P NMR spectra were
recorded at 23 °C on a JEOL GSX-500 spectrometer (31P at 202.35
MHz) with 1H-broad band decoupling. The assay was generally repeated
more than three times for each sample with good reproducibility. The
concentration-activity relationship shown in Figure 2 was depicted
by varying sample-lipid molar ratios. The y-axis is a ratio of a decrease
in peak area at δ 1.2. Sigmoid curves were fitted using Origin 6.1,
from which EC50 values were determined.
Determination of Hemolytic Activity. Freshly collected human
blood was centrifuged for 5 min at 1000g, and separated erythrocytes
were washed twice by suspending in PBS buffer (pH 7.4). Packed
erythrocytes were then resuspended with PBS to 100-fold volume of
the original blood. Gradually, diluted drug solutions (DMSO solution,
2 µL) were added to the erythrocyte suspensions (198 µL), and the
suspensions were incubated with gently shaking at 37 °C. After 18 h,
the suspensions were spun and the absorbance of the supernatant was
determined at 490 nm by micro-plate reader (Molecular Devices). For
a positive control, water was used instead of PBS buffer. As a negative
control, 2 µL of DMSO was added instead of sample solution. From
dose-response curves, the dosage that led to 50% hemolysis (EC50)
was determined.
Antifungal Assay. Aspergillus nigar was cultured in a GP liquid
medium (2% glucose, 0.2% yeast extract, 0.5% polypeptone, 0.05%
MgSO4, and 0.1% KH2PO4) at 25 °C for 2 days. An aliquot of the
broth was then spread onto a GP agar plate. The drugs dissolved in
DMSO were spotted on paper disks of 8 mm in diameter. As a control,
a disk containing only DMSO was also prepared. These paper disks
were then placed on an agar plate containing A. nigar mycelia. After
cultivating at 25 °C for 2 days, the diameter of the inhibitory zone on
each paper disk was measured.
N-Propyl-AmB (10). A solution of propionaldehyde (5 µL, 68 µmol)
and AmB (50 mg, 54 µmol) in DMF-MeOH (4:3, 7 mL) was stirred
for 2 h at room temperature, and then NaBH3CN (14 mg, 222 µmol)
was added to the solution. The mixture was agitated overnight at room
temperature, and the products were precipitated with diethyl ether (100
mL). The precipitate was extracted and purified in the same manner as
that of 8a to afford N-propyl-AmB (24 mg, 46%) as a yellow solid. Rf
0.54 (CHCl3:MeOH:H2O ) 10:6:1). ESI-MS m/z 988.7 (M + Na)+.
Derivative 5. To a solution of 10 (24 mg, 25 µmol) in DMF (2 mL)
were added propylamine (3.1 µL, 38 µmol), diisopropylethylamine (9.2
µL, 52 µmol), 1-hydroxybenzotriazole (4.9 µg, 30 µmol), and PyBOP
(13.6 mg, 25 µmol) sequentially. The mixture was stirred for 20 h at
room temperature and precipitated with diethyl ether (100 mL). The
precipitate was filtered on Celite, washed with ether, and purified by
HPLC to give 5 as yellow powders. HPLC conditions were the same
as those for 2, retention time: 19.5 min. ESI-MS m/z 1007.4 (M +
1
H)+. H NMR (500 MHz, DMSO-d6): see Table 2.
NMR Measurements. The antibiotic solutions were prepared under
dry argon in DMSO-d6 at 3-5 mM concentration. All NMR spectra
were recorded at 25 °C on a JEOL GSX500 spectrometer (1H 500
MHz). Spectra were processed using Alice2 V.4.1 (JEOL DATUM)
software. Homonuclear two-dimensional spectra COSY, TOCSY (HO-
HAHA), and NOESY were recorded with a 1.5 s recovery delay in the
phase-sensitive mode using the States method as data matrixes of 512
(F1) × 1024 (F2) complex data points. Mixing times of 80 ms for
TOCSY and 300 ms for NOESY spectra were used. The spectral width
in both dimensions was 5000 Hz.
The data were apodized with shifted square sine-bell window
functions in both F1 and F2 dimensions after zero-filling in the F2
dimension to obtain a final matrix of 512 (F1) × 2048 (F2) real data
points. Chemical shifts were referenced to the solvent chemical shift
(DMSO-d5 (1H), 2.49 ppm).
Experimental NMR Restraints. All of the interproton-distance
restraints between non J-coupled protons are derived from the two-
dimensional homonuclear NOESY experiments. Interproton restraints
were classified into three categories. Upper bounds were fixed at 2.8,
3.4, and 4.0 Å for strong, medium, and weak correlations, respectively.
A lower bound was fixed at 1.8 Å, which corresponds to the sum of
the hydrogen van der Waals radii. Pseudo atom corrections of the upper
bounds were applied for distance restraints involving the unresolved
methylene and methyl protons (+1 Å). For stereospecifically assigned
diastereotopic methylene protons, the interproton distances were applied
to each proton according to the NOE peak intensities. When possible,
H-C-C-H dihedral angles were restrained to dihedral domains
3
according to the different JHH coupling constants measured using
Karplus dihedral relations with (30° allowance.
Structure Calculations. Models were calculated using the Macro-
Model software version 8.529 installed on RedHat Linux 8 OS. Initial
atomic coordinates and structure files for each derivative were generated
step by step from the crystal data of N-iodoacetyl AmB.28 For each
molecule, the macrolide rings were treated as semirigid group, in which
(30° allowance from the crystal structure28 was given to each C-C
single bond upon calculations. The sampling of the conformational
space was performed following a Monte Carlo Multiple Minimum
method.35 The MMFF force field36 implemented in the MacroModel
was used. The non-bonded van der Waals’ interactions were represented
by a simple repulsive quadratic term. The experimental distance re-
straints were represented as a soft asymptotic potential, and electrostatic
interactions were ignored. 5000 steps of MCMM were carried out for
each molecule. Continuum solvation models for water using a General-
ized Born/Surface Area (GB/SA)37 were applied though the calculations.
Acknowledgment. We are grateful to Ms. Rie Masuda, Prof.
Tohru Oishi, and Dr. Shigeru Matsuoka in our laboratory for
invaluable discussions. This work was supported by Grant-In-
Aids for Scientific Research on Priority Area (A) (No. 12045243)
and for Young Scientists (B) from the Ministry of Education,
Sciences, Sports, Culture, and Technology, Japan, and by a grant
from CREST, Japan Science and Technology Corp.
Supporting Information Available: NMR spectra and NOE
correlations for 2-4. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA051597R
(35) Kolossvary, I.; Guida, W. C. J. Comput. Chem. 1999, 20, 1671-1684.
(36) Halgren, T. A. J. Comput. Chem. 1999, 20, 730-748.
(37) Qiu, D.; Shenkin, P. S.; Hollinger, F. P.; Still, W. C. J. Phys. Chem. A
1997, 101, 3005-3014.
9
J. AM. CHEM. SOC. VOL. 127, NO. 30, 2005 10675