An amphotericin B-fluorescein conjugate as a powerful probe for biochemical studies of the membrane

Article abstract of DOI:10.1002/anie.200460489

Not only skin deep: A fluorescein-amphotericin B conjugate with a new piperazine linker (see structure) was prepared and used as a probe of the biomembrane and the mechanism of action of amphotericin B both in vivo and in liposomal experiments. The amphotericin B analogue localizes at the membrane of yeast cells but is internalized by mammalian cells.

Full text of DOI:10.1002/anie.200460489

Angewandte
Chemie
Bioorganic Chemistry
the primary cause of amphotericin-B-induced cell death
remains elusive.[4] Recent attempts at mechanistic inquiry
have involved use of the tools of synthesis[5] and metabolic
engineering.[6] There is a need for the design and synthesis of
amphotericin B probes with which to test the various existing
hypotheses.[7] Herein, we describe the use of a readily
accessible piperazine linker as a synthetic anchor for the
introduction of reporter groups such as fluorescein to
amphotericin B. We used the resulting probe to conduct
several in vivo and in vitro comparative studies involving
mammalian and fungal cells, as well as vesicles (Liposomes).
Several of the observations documented herein are of
particular significance: 1) the amphotericin–fluorescein con-
jugate is rapidly internalized in mammalian cells, but no such
uptake occurs in fungal cells; 2) in contrast, the conjugate is
localized in the fungal membrane, as observed by epifluo-
rescence microscopy; 3) despite the fact that liposome studies
show the amphotericin conjugate to lead to rapid K+ efflux,
the compound is not lethal to yeast over a wide range of
concentrations; 4) the conjugate was found to be uniformly
distributed throughout the membrane in exponentially grow-
ing and budding yeast. The various intriguing properties of
this probe may facilitate further studies aimed at gaining an
understanding of the difference between the behavior of
amphotericin towards mammalian cells and that towards
fungal cells, and may already provide an insight into the
origins of the severe side-effects of the drug.
An Amphotericin B–Fluorescein Conjugate as a
Powerful Probe for Biochemical Studies of the
Membrane**
Andreas Zumbuehl, Damien Jeannerat, Scott E. Martin,
Marc Sohrmann, Pasquale Stano, Tamas Vigassy,
Daniel D. Clark, Stephen L. Hussey, Mathias Peter,
Blake R. Peterson, Ernö Pretsch, Peter Walde, and
ErickM. Carreira*
Dedicated to Professor Peter B. Dervan
Amphotericin B is an antimycotic agent extracted and
isolated from a soil streptomycete[1] and is currently in clinical
use against chronic fungal infections. An enormous amount of
data on the biological activity of this polyketide has been
accumulated, which underlines the potency and effectiveness
of the medicament, as well as its importance in membrane
research.[2] Even so, after five decades of intensive research
the details of the mechanism of action of amphotericin B
remain far from completely elucidated.[3] It is generally
accepted that amphotericin B induces leakage of electrolytes
and small molecules from exposed cells. However, a definitive
answer to the question of whether such electrolyte efflux is
Despite the density of functionality of amphotericin B, it
only has two addressable reactive centers that allow selective
conjugation: the primary amine function found in the pendant
mycosamine moiety and the carboxylic acid group in the
amphoteronolide fragment (Scheme 1). Analysis of prior
work in this area[8] suggests that conjugates anchored through
the mycosamine moiety must contain a linker that preserves a
protonatable amine function to allow observation of the fast
kinetics of K+ release in liposomes. Murata and co-workers[5c]
recently published a study in which a linker was introduced
[*] A. Zumbuehl, Prof. Dr. E. M. Carreira
Laboratorium für Organische Chemie
ETH Hönggerberg, HCI H335
8093 Zürich (Switzerland)
Fax: (+41)1-632-1328
E-mail: carreira@org.chem.ethz.ch
Dr. D. Jeannerat
DØpartment de Chimie Organique
UniversitØ de Gen›ve
1211 Gen›ve 4 (Switzerland)
Dr. S. E. Martin, D. D. Clark, Dr. S. L. Hussey, Prof. Dr. B. R. Peterson
Departement Materialwissenschaft
ETH Hönggerberg
8093 Zürich (Switzerland)
Dr. M. Sohrmann, Prof. Dr. M. Peter
Institut für Biochemie
ETH Hönggerberg
8093 Zürich (Switzerland)
P. Stano, Prof. Dr. P. Walde
Department of Chemistry
The Pennsylvania State University
Pennsylvania 16802 (USA)
Dr. T. Vigassy, Prof. Dr. E. Pretsch
Laboratorium für Organische Chemie
ETH-Hönggerberg
8093 Zürich (Switzerland)
[**] We would like to thank Prof. Dr. Josef Brunner for helpful
discussions. E.M.C. is grateful for generous support from the
Eidgenössische Technische Hochschule (ETH), Zürich (TH-
Gesuch) and from F. Hoffmann LaRoche. B.R.P. acknowledges
support from the National Institutes of Health, USA (Grant no.
RO1-CA83831).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Scheme 1.
Angew. Chem. Int. Ed. 2004, 43, 5181 –5185
DOI: 10.1002/anie.200460489
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5181

Communications
onto amphotericin B (1) through reductive amination of an
Quantitative deprotection (piperidine/DMSO) of 8 to
form the free amine 9 prepared the compound for conjuga-
tion.[13] Treatment of 9 with the Fmoc-protected 6-amino-
hexanoic acid N-hydroxysuccinimidyl ester gave 10, which
was deprotected (piperidine, DMSO, quantitative yield) then
allowed to react with fluorescein isothiocyanate (Hünigꢀs
base, DMF, 66% yield) to form the amphotericin B–fluo-
rescein conjugate 11.[14–16]
aldehyde. Such a strategy suffers from the disadvantage that a
second reductive alkylation can ensue, a process that can be
difficult to control.[9] Our attempts to carry out reductive
monoalkylation at the amphotericin B mycosamine function
proved the process to be capricious; production of a mixture
of secondary and tertiary amines could not be avoided, and
separation of these compounds was far from simple. We
sought a novel alternative conjugation strategy and reasoned
that a linker programmed to undergo double reductive
alkylation would completely obviate the problems outlined
above. Moreover, the tertiary amines 2 so produced would
retain their basicity and could be protonated under physio-
logical conditions.[8]
There is no precedence for the use of such a linker with
amphotericin B, or any other biomolecule. We developed an
easy, straightforward synthetic route to attach a piperazine
linker to amphotericin B, which allowed subsequent conjuga-
tion (Scheme 2). Coupling of 6-N-trifluoroacetamide-hexa-
noyl chloride (3)[10] with 3-pyrroline (4)[11] gave 5, which was
converted into the corresponding Fmoc-protected amide 6.
Oxidative cleavage of 6 afforded dialdehyde 7[12]. This
aldehyde readily underwent double reductive amination
with the primary amine group of amphotericin B to form 8
in 63% yield.
To evaluate the new linker system, the free amine 9 and
the amphotericin B–fluorescein conjugate 11 were examined
thoroughly in both liposomal and cellular assays. For each of
the two compounds we determined the minimal inhibitory
concentration (MIC) required to inhibit growth of Saccha-
romyces cerevisiae. Native amphotericin B (1; MIC = 1 mm)
and the aminohexanoyl piperazinyl conjugate 9 (MIC =
1.6 mm) are equally active, which indicates the usefulness of
the new linker system. The fluorescein–amphotericin B con-
jugate 11 is not toxic in the measured range of up to 500 mm.
We adapted a vesicle assay to our purposes.[17,18] Large
unilamellar vesicles (LUVs) made from POPC (1-palmitoyl-
2-oleoyl-sn-glycero-3-phosphocholine) or from POPC
admixed with sterols (mimicking conditions in natural
biomembranes[19]) were prepared from a thin film of the
lipid(s) in a mixture containing KCl (150 mm) and 2-[4-(2-
hydroxyethyl)-1-piperazinyl]ethanesulfonic acid (HEPES)
buffer (5 mm) at pH 7.4.[20] The unilamellar vesicles
were dialyzed against NaCl to create an ion gradient (K+
inside, Na+ outside). The amphotericin B conjugates
were added to the vesicle suspension to form ion
channels leading to K+ efflux, which was measured
with freshly prepared valinomycin-based K+-selective
electrodes.[21,22]
Native amphotericin B (1), the conjugate with the
piperazine linker 9, and the fluorescein conjugate 11 all
induced immediate total release of trapped K+ from
ergosterol-containing vesicles (Figure 1). Treatment of
cholesterol-containing vesicles with the new conjugates
9 and 11 resulted in a slower release of ions than
treatment with amphotericin B (1), which indicates that
the piperazine-linked conjugates have a preference for
ergosterol-containing rather than cholesterol-containing
vesicles.[23] K+ release from sterol-free vesicles was slow,
as we expected based on literature data for amphoter-
icin B.[2] The fluorescein–amphotericin B conjugate 11
induced rapid K+ release from vesicles while having no
toxic effect on yeast cell growth. This observation is
consistent with previous proposals that channel forma-
tion is not sufficient to induce cell death.[4]
We tested fluorescein–amphotericin B conjugate 11
in a variety of other in vivo experiments. A cholesterol-
derived synthetic membrane anchor has recently been
developed that allows the uptake of antifluorescein
antibodies and the associated Protein A into mamma-
lian cells.[24] Since amphotericin B is known to insert into
membranes,[1] we tested the fluorescein–amphotericin B
conjugate 11 in the same assay.
Scheme 2. a) 3-pyrroline (4), pyr, CH2Cl2, 3 h, 69%; b) K2CO3, MeOH, H2O, 13 h, 82%;
c) FmocOSu, DMF, pyr, 12 h, 77%; d) K2OsO3·2H2O, acetone, H2O, tBuOH, 16 h,
66%; e) NaIO4, Si gel, CHCl3, H2O, 0.5 h; f) amphotericin B, DMF, NaCNBH3, HCl,
13 h, 63% over 2 steps; g) piperidine, DMSO, quantitative; h) N-Fmoc-6-NH-
(CH2)5CO2Su, iPr2NEt, DMF, 38%; i) piperidine, DMSO, quantitative; j) fluorescein iso-
thiocyanate, iPr2NEt, DMF, 66%. Pyr, pyridine. DMF=dimethylformamide, DMSO=
dimethyl sulfoxide, Fmoc=9-fluorenylmethoxycarbonyl, Su=succinimide.
To investigate the recruitment of 11 both by
Saccharomyces cerevisiae FY250 and by Jurkat human
T cells, the cells were incubated with 11 for 1 h at 378C
5182
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 5181 –5185

Angewandte
Chemie
mammalian cells. The generally held view
is that, despite the inherent preference of
amphotericin B for channel formation in
yeast membranes as a result of ergosterol
stabilization, the compound retains the
ability to form channels competitively in
mammalian membranes. Model studies with
vesicles have shown that approximately 16
amphotericin B molecules are required for
channel formation.[28] In the light of the
observations described above for Jurkat
cells, we propose that transport of the
molecules across mammalian cell mem-
branes into the interior depletes amphoter-
icin B at the membrane and thus diminishes
channel formation. Oxidative (free radical)
Figure 1. K+ efflux from vesicles prepared from pure POPC (black), POPC admixed with
30 mol% cholesterol (red), and POPC admixed with 13 mol% ergosterol (blue). The con-
centration of amphotericin B (1, a) and its conjugates 9 (b) or 11 (c) after addition (as a
DMSO solution) to the stirred vesicle suspension was 5 mm. The total concentration of
lipids (POPC+sterol) was 1 mm in 5 mm HEPES buffer at pH 7.4.
(Jurkat cells) or 308C (S. cerevisiae) and the cultures were
washed with media (Phosphate–buffered saline (PBS) for the
yeast cells and RPMI media for the Jurkat cells, see
Supporting Information) and resuspended. Subsequent anal-
ysis by flow cytometry revealed that both yeast and Jurkat
lymphocytes recruit 11, but the compound has a slight
preference for the mammalian cell line.[25]
activity inside the cell may be the source of the observed side-
effects of the drug.[2e,f]
We next conducted experiments aimed at establishing the
avidity of antifluorescein IgG for cellular plasma membranes
incubated with 11. We employed nonfluorescent rabbit IgG
complexed with red fluorescent conjugates of commercially
available Protein A (PrA) from Staphylococcus aureus in a
previously described assay.[26,27] Flow cytometry allowed
quantification of the recruitment of antifluorescein IgG
bound to Alexa Fluor 633-conjugated PrA by Jurkat lym-
phocytes and yeast.[25] Jurkat cells were treated with 11 for 1 h
and then with IgG for 4 h at 378C. The cells were washed with
5-aminofluorescein prior to measurement to ensure that all
the noninternalized IgG was removed. Yeast cells were
treated similarly but were not washed with 5-aminofluores-
cein. The results of these experiments lead to the conclusion
that 11 associates with both yeast and Jurkat human T cells in
a dose-dependent manner, which is in line with the results of
the studies described above. More importantly, treatment
with the amphotericin conjugate leads to a significanct
increase in the uptake of IgG by Jurkat lymphocytes. No
internalization of 11 or its complex with antifluorescein IgG
by yeast cells was observed.
To investigate the nature of the interaction of 11 with the
cellular targets in the presence of IgG in greater detail, cells
incubated with 11 were examined by epifluorescence micros-
copy (see Figure 2). Uptake into Jurkat lymphocytes and IgG
recruitment by yeast were assayed as described above, except
that Alexa Fluor 594-conjugated Protein A was substituted
for Alexa Fluor 633-conjugated Protein A. Both cell lines
showed red and green fluorescence. It is particularly interest-
ing that the amphotericin B–conjugate 11 is localized at the
membrane in yeast, while the micrograph of the mammalian
cell lines reveals that the amphotericin conjugate has been
taken up by the cells. The fact that the fluorescein–ampho-
tericin B conjugate 11 is taken up into mammalian cells leads
us to hypothesize about the mechanism of action of ampho-
tericin B with reference to its reported side-effects on
Figure 2. Epifluorescence and differential interference contrast (DIC)
microscopy of compound-treated Jurkat lymphocytes (a) and Saccharo-
myces cerevisiae (b) 30 min after IgG addition. Green fluorescence:
lex =480 nm, lem =535 nm; red fluorescence: lex =560 nm,
lem =645 nm.
We carried out additional in vivo experiments on yeast
cells to demonstrate the utility of conjugate 11 as a probe for
investigation of biological phenomena at the membrane.
Previous work has suggested that ergosterol-rich plasma
membrane domains localize in areas of cell polarization.
Polyene macrolide filipin was used as a probe because it is
known to associate with ergosterol.[29,30] Localized fluores-
cence regions were observed in S. cerevisiae at the tip of
mating projections formed in response to pheromones.[31] In
previous studies with the fission yeast S. pombe, ergosterol
appeared to accumulate at cell-division-associated regions
and at the growing poles.[32] We therefore wanted to
investigate whether 11 distributes uniformly throughout the
plasma membrane of S. cerevisiae or accumulates specifically
in polarized regions of the cells. We analyzed cells during
exponential growth, when the cells polarize towards the newly
forming daughter bud (Figure 3a and c), and in the course of
the formation of mating projections in response to phero-
mones (Figure 3b and d). In both cases, fluorescein–ampho-
tericin conjugate 11 appeared to be distributed uniformly
throughout the plasma membrane of each yeast cell. Ampho-
Angew. Chem. Int. Ed. 2004, 43, 5181 –5185
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5183

Communications
transmembrane transport or as an anchor for selective
mediation of the delivery of toxins to fungal cells.
Received: April 28, 2004
Revised: June 24, 2004
Keywords: antifungal agents · bioorganic chemistry ·
.
biophysics · fluorescent probes · synthetic methods
[1] a) T. H. Sternberg, E. T. Wright, M. Oura, Antibiot. Annu. 1955–
1956, 566 – 573; b) B. A. Steinberg, W. P. Jambor, L. O. Suydam,
Antibiot. Annu. 1955–1956, 574 – 578; c) W. Gold, H. A. Stout,
J. F. Pagano, R. Donovick, Antibiot. Annu. 1955–1956, 579 – 586;
d) J. Vandeputte, J. L. Wachtel, E. T. Stiller, Antibiot. Annu.
1955–1956, 587 – 591.
[2] a) G. Medoff, J. Brajtburg, G. S. Kobayashi, Annu. Rev. Phar-
macol. Toxicol. 1983, 23, 303 – 330; b) J. Bolard, Biochim.
Biophys. Acta 1986, 864, 257 – 304; c) J. Brajtburg, W. G.
Powderly, G. S. Kobayashi, G. Medoff, Antimicrob. Agents
Chemother. 1990, 34, 183 – 188; d) F. C. Szoka, Jr., M. Tang, J.
Liposome Res. 1993, 3, 363 – 375; e) S. C. Hartsel, C. Hatch, W.
Ayenew, J. Liposome Res. 1993, 3, 377 – 408; f) J. Bolard, V. Joly,
P. Yeni, J. Liposome Res. 1993, 3, 409 – 427; g) S. C. Hartsel, J.
Bolard, Trends Polym. Sci. 1996, 17, 445 – 449; h) I. Al-Mohsen,
W. T. Hughes, Ann. Saudi Med. 1998, 18, 28 – 38; i) B. E. Cohen,
Int. J. Pharm. 1998, 162, 95 – 106; j) I. M. Hann, H. G. Prentice,
Int. J. Antimicrob. Agents 2001, 17, 161 – 169; k) S. B. Zotchev,
Curr. Med. Chem. 2003, 10, 211 – 223.
[3] a) W. Huang, Z. Zhang, X. Han, J. Tang, J. Wang, S. Wang, S.
Dong, E. Wang, Biophys. J. 2002, 83, 3245 – 3255; b) B. V.
Cotero, S. Rebolledo-Antunez, I. Ortega-Blake, Biochim. Bio-
phys. Acta 1998, 1375, 43 – 51; c) G. Fujii, J.-E. Chang, T. Coley,
B. Steere, Biochemistry 1997, 36, 4959 – 4968; d) M. Baginski, E.
Borowski, J. Mol. Struct. (Theochem) 1997, 389, 139.
[4] a) C. C. Hsu-Chen, D. S. Feingold, Nature 1974, 251, 656 – 659;
b) M. L. Sokol-Anderson, J. Brajtburg, G. Medoff, J. Infect. Dis.
1986, 154, 76 – 83; c) W. H. Beggs, Antimicrob. Agents Chemo-
ther. 1994, 38, 363 – 364.
[5] a) S. D. Rychnovsky, Chem. Rev. 1995, 95, 2021 – 2040; b) B. N.
Rogers, M. E. Selsted, S. D. Rychnovsky, Bioorg. Med. Chem.
Lett. 1997, 7(24), 3177 – 3182; c) N. Matsumori, N. Yamaji, S.
Matsuoka, T. Oishi, M. Murata, J. Am. Chem. Soc. 2002, 124,
4180 – 4181; d) N. Yamaji, N. Matsumori, S. Matsuoka, R. Oishi,
M. Murata, Org. Lett. 2002, 4, 2087 – 2089; e) S. Matsuoka, N.
Matsumori, M. Murata, Org. Biomol. Chem. 2003, 1, 3882 – 3884;
f) N. Matsumori, R. Masuda, M. Murata, Chem. Biodiversity
2004, 1, 346 – 352.
[6] a) P. Caffrey, S. Lynch, E. Flood, S. Finnan, M. Oliynyk, Chem.
Biol. 2001, 8, 713 – 723; b) B. Byrne, M. Carmody, E. Gibson, B.
Rawlings, P. Caffrey, Chem. Biol. 2003, 10, 1215 – 1224; c) J. F.
Aparicio, P. Caffrey, J. A. Gil, S. B. Zotchev, Appl. Microbiol.
Biotechnol. 2003, 61, 179 – 188.
Figure 3. Localization of 11 in the plasma membrane of yeast cells.
The cell outlines were observed by phase contrast microscopy (a, b),
and the localization of 11 by fluorescence microscopy (c, d). The figure
shows exponentially growing cells (a, c) and cells forming mating pro-
jections after exposure to mating pheromone (b, d).
tericin B is expected to undergo pronounced complex for-
mation with ergosterol[33] so these observations suggest that
ergosterol-rich domains might not accumulate significantly in
distinct regions of the plasma membrane. In support of this
hypothesis, recent data suggests that filipin itself might
perturb the local composition of cellular membranes.[34]
Further work is required to determine whether filipin and
amphotericin B bind to the same molecular structures in vivo,
and to answer the question of whether ergosterol-rich
membrane domains are polarized during yeast cell growth.
In conclusion, we have established a new piperazine linker
system for bioconjugation through amines that allows the
synthesis of amphotericin B conjugates on a multigram scale.
This linking strategy is suitable for bioconjugation in a
number of other biologically important systems for which
functionalization through a primary amine is required and
may be used to obtain molecules with interesting new
properties. A fluorescein conjugate was prepared as a bench-
mark case. This conjugate displayed preserved K+ efflux-
inducing properties in an LUV assay but completely lacks
toxic effects on yeast cell growth. These results imply that ion-
channel formation in artificial vesicles and the lethal effect
induced by amphotericin B are separate phenomena. We also
demonstrated that amphotericin B–fluorescein conjugate 11
can function as a useful probe of biological activity. Our
studies revealed that 11 associates with both yeast (Saccha-
romyces cerevisiae FY250) and Jurkat human T cells in a
dose-dependent manner but that it is taken up by mammalian
cells whilst it remains localized in the yeast membrane.
Additional investigations with this probe in yeast cells
revealed a uniform distribution of the compound throughout
the plasma membrane, which opens up the possibility that
ergosterol might not accumulate in polarized regions of the
cells as suggested previously. Additional applications can be
envisioned for this new fluorescein–amphotericin B conju-
gate, for example, it could be used as a means for studying
[7] a) J. Krüger, E. M. Carreira, Tetrahedron Lett. 1998, 39, 7013 –
7076; b) J. Tholander, E. M. Carreira, Helv. Chim. Acta 2001, 84,
613 – 622.
[8] M. Chꢁron, B. Cybulska, J. Mazerski, J. Grzybowska, A. Czerwin-
ski, E. Borowski, Biochem. Pharmacol. 1988, 37, 827–836.
[9] J.-P. Salvi, N. Walchshofer, J. Paris, Tetrahedron Lett. 1994, 35,
1181 – 1184.
[10] J. S. Clark, P. A. Hodgson, M. D. Goldsmith, L. J. Street, J.
Chem. Soc. Perkin Trans. 1 2001, 24, 3312 – 3324.
[11] J. S. Warmus, G. J. Dilley, A. I. Meyers, J. Org. Chem. 1993, 58,
270 – 271.
[12] M. Daumas, Y. Vo-Quang, L. Vo-Quang, F. Le Goffic, Synthesis
1989, 64 – 65.
5184
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 5181 –5185

Angewandte
Chemie
[13] The importance of structural assignment for any subsequent
mechanistic investigation led us to confirm the success of the
conjugation reaction by using a spectral aliasing technique to
acquire high-resolution HSQC and HMBC spectra (see refs. [15,
16]). All but a few signals in the double-bond regions of such 1H,
13C, DEPT, COSY, HSQC, and HMBC spectra were unambig-
uously assigned to 8.
[14] For the synthesis of an NBD (7-Nitro-2-oxa-1,3-diazol-4-yl)–
amphotericin conjugate with an amide linker, see: N. O.
Petersen Spectros. Int. J. 1983, 2, 408 – 414. No characterization
data were given for this conjugate; to the best of our knowledge
it has not been used for biological studies.
[15] a) D. Jeannerat, Magn. Reson. Chem. 2002, 41, 3; b) D.
Jeannerat, D. Ronan, Y. Baudry, A. Pinto, J.-P Saulnier, S.
Matile, Helv. Chim. Acta, in press ; c) D. Jeannerat, D. Ronan, Y.
Baudry, A. Pinto, J.-P Saulnier, S. Matile, Helv. Chim. Acta, in
press; d) D. Ronan, Y. Baudry, D. Jeannerat, S. Matile, Org. Lett.
2004, 6, 885 – 887; e) http://rmn.unige.ch/simplealias. We applied
the most practical implementation (see ref. [15c]), which allows
the acquisition conditions of heteronuclear experiments (HSQC
and HMBC) to be optimized by copying and pasting a list of
carbon chemical shifts into an internet base program (see
ref. [15e]). See the Supporting Information for further details.
[16] a) A. Aszalos, A. Bax, N. Burlinson, P. Roller, C. McNeal, J.
Antibiot. 1985, 38, 1699 – 1713; b) P. Sowinski, J. K. Pawlak, E.
Borowski, P. Gariboldi, Magn. Reson. Chem. 1992, 30, 275 – 279;
c) P. Sowinski, J. K. Pawlak, E. Borowski, T. Iwashita, J. Antibiot.
1985, 38, 175 – 180; d) J. M. Brown, P. J. Sidebottom, Tetrahedron
1981, 37, 1421 – 1428.
[17] a) P. van Hoogevest, B. de Kruijff, Biochim. Biophys. Acta 1978,
511, 397 – 407; b) D. S. Orlov, T. Nguyen, R. I. Lehrer, J. Micro-
biol. Methods 2002, 49, 325 – 328.
[18] The use of external Ca2+ instead of Na+ ions (as reported
previously) seemed to block the pores and yielded no results.
[19] J. Barwicz, P. Tancrꢂde, Chem. Phys. Lipids 1997, 85, 145 – 155.
[20] V. Dorovska-Taran, R. Wick, P. Walde, Anal. Biochem. 1996,
240, 37 – 47.
[21] L. A. R. Pioda, V. Stankova, W. Simon, Anal. Lett. 1969, 2, 665 –
674. For improved ruggedness, ion-selective membranes were
used that had a higher than usual (see ref. [22]) content of PVC
and a lipophilic salt was added to compensate for the increased
resistance.
[22] E. Bakker, P. Bühlmann, E. Pretsch, Chem. Rev. 1997, 97, 3083 –
3132.
[23] This result may imply a beneficial increase in the therapeutic
effect of amphotericin B.
[24] S. L. Hussey, E. He, B. R. Peterson, J. Am. Chem. Soc. 2001, 123,
12712 – 12713.
[25] For flow cytometry data, see the Supporting Information.
[26] PrA is a 57-kDa protein that binds to the invariant Fc fragment
of rabbit-derived IgG proteins with a dissociation constant of
approximately 60 nm. See ref [24].
[27] L. Cedergren, R. Andersson, B. Jansson, M. Uhlen, B. Nilsson,
Protein Eng. 1993, 6, 441 – 448.
[28] G. Fujii, J.-E. Chang, T. Coley, B. Steere, Biochemistry 1997, 36,
4959 – 4968.
[29] K. Takeo, J. Gen. Microb. 1985, 131, 309 – 316.
[30] W. Drabikowski, E. Lagwinska, M. G. Sarzala, Biochim. Bio-
phys. Acta 1973, 291, 61 – 70.
[31] M. Bagnat, K. Simons, Proc. Natl. Acad. Sci. USA 2002, 99,
14183 – 14188.
[32] V. Wachtler, S. Rajagopalan, M. K. Balasubramanian, J. Cell Sci.
2003, 116, 867 – 874.
[33] I. Gruda, P. Nadeau, J. Brajtburg, G. Medoff, Biochim. Biophys.
Acta 1980, 602, 260 – 268.
[34] J. Valdez-Taubas, H. R. B. Pelham, Curr. Biol. 2003, 13, 1636 –
1640.
Angew. Chem. Int. Ed. 2004, 43, 5181 –5185
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5185