A novel strategy for bioconjugation: Synthesis and preliminary evaluation with amphotericin B

Article abstract of DOI:10.1039/b701953j

A novel linker strategy is presented based on a double reductive amination of a dialdehyde to the amine of the amphotericin B mycosamine sugar and the biological activity of a series of conjugates is compared to the native amphotericin B. This journal is

Full text of DOI:10.1039/b701953j

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
A novel strategy for bioconjugation: synthesis and preliminary evaluation with
amphotericin B†
Andreas Zumbuehl,a Pasquale Stano,b Marc Sohrmann,c Mathias Peter,c Peter Waldeb and Erick M. Carreira*a
Received 7th February 2007, Accepted 6th March 2007
First published as an Advance Article on the web 23rd March 2007
DOI: 10.1039/b701953j
A novel linker strategy is presented based on a double
reductive amination of a dialdehyde to the amine of the
amphotericin B mycosamine sugar and the biological activity
of a series of conjugates is compared to the native ampho-
tericin B.
report the synthesis of additional conjugates using a piperazine
as a point of attachment, and we introduce a related conju-
gation strategy involving a 4-carboxypiperidinyl linker, which is
Despite the fact that is was discovered well over 40 years ago,
the mode of action of the polyene macrolide antimycotic agent
amphotericin B remains enigmatic, puzzling biochemists and
pharmacologists alike.1,2
Although there is clear evidence that its incorporation into
biological and abiological membranes results in electrolyte efflux,
there persists an ongoing debate as to the causal relationship of
this observed effect on cell death. Indeed, numerous investigations
reveal contradicting behaviour between model studies and those
in vivo.3
We have been interested in developing biochemical probes based
on amphotericin B (1, Fig. 1) as a means of studying several
hypotheses concerning the mode of action of this important
antifungal agent and to facilitate the study of processes believed to
occur at the membrane. We have initiated a program in the design
and synthesis of tailor-made amphotericin B conjugates bearing
reporter groups that permit insight into the fate of amphotericin
in cells and liposomes.4,5 In our preliminary disclosure, we
documented the preparation of an amphotericin B–fluorescein
conjugate which exhibited interesting properties when employed
as a probe in the fungal membrane. In this communication we
Scheme 1 Synthesis of aminohexanoyl piperazinyl amphotericin B.
Fig. 1 Structure of the antifungal agent amphotericin B.
aLaboratorium fu¨r Organische Chemie, ETH Ho¨nggerberg, 8093, Zu¨rich,
Switzerland. E-mail: carreira@org.chem.ethz.ch
Fig. 2 K+ release from POPC based liposomes measured via potentiom-
etry. Substrate was added externally to lipid vesicles at a molar ratio
of 1 : 200 (reagent : total lipid), solid line: POPC alone, dashed line:
POPC admixed with 30 mol% cholesterol; dotted line: POPC admixed
with 13 mol% ergosterol.
bDepartement Materialwissenschaft, ETH Ho¨nggerberg, 8093, Zu¨rich,
Switzerland
cInstitut fu¨r Biochemie, ETH Ho¨nggerberg, 8093, Zu¨rich, Switzerland
† Electronic supplementary information (ESI) available: Experimental
procedures and spectroscopic data. See DOI: 10.1039/b701953j
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1339–1342 | 1339
©

View Article Online
Fig. 3 Piperazinoyl-linked amphotericin B conjugates with reporter groups biotin (5), fluorescein (6), cholesterol (7), and photo-crosslinking diazirane
(8), along with dimer 9.
prepared through a short convenient sequence of reactions from
commercially available 3-cyclopentene-1-carboxylic acid and 1-
Fmoc-1,6-diaminohexane. Furthermore a preliminary validation
on the potential utility of one of these conjugates is carried in yeast;
we also document ion-flux studies in POPC vesicles (liposomes).
Earlier structure–activity relationship studies on amphotericin
B underlined the importance of the basic amine for the mechanism
of action. Simple acylations of this functional group have led
to compounds with significantly lowered biological activity.6
As such, this delimits the types of functionality that can be
relied upon to prepare conjugates. Two different solutions to this
Fig. 4 Amide-linked biotin-amphotericin B.
This journal is The Royal Society of Chemistry 2007
1340 | Org. Biomol. Chem., 2007, 5, 1339–1342
©

View Article Online
problem have appeared recently. Murata and co-workers
have reported the monoalkylation of amphotericin mycosamine
sugar through reductive amination of an aldehyde.7 We have
documented the use of dialdehydes in a double reductive
alkylation reaction of amphotericin mycosamine to afford novel
piperazinyl linked structures (Scheme 1). This latter approach
is convenient in that the formation of dialkylated products
produced when monoaldehydes are employed are avoided, thus
giving products that are easier to purify.
In order to evaluate the new linker we determined the minimal
inhibitory concentration (MIC) required to inhibit growth of
Saccharomyces cerevisiae. Amphotericin B (1) completely stopped
cell growth when added at a concentration of 1 lM. Conju-
gate 4 retained activity and was toxic above a threshold of
1.6 lM.
We also prepared large (100 nm) unilamellar vesicles (LUVs)
and directly measured the induced K+ permeability of the LUVs
with a K+-selective electrode. This assay contrasts with that
recently employed, which monitors the pH-sensitive shift corre-
sponding to the 31P NMR signal of a membrane probe.6 For such a
commonly employed assay, amphotericin B-induced leakage of K+
ions from liposomes triggers a counter H+ ion influx, levelling out a
transmembrane pH gradient. We reasoned that the direct analysis
of K+ efflux via a K+-selective electrode would have the benefit of
fast on-line measurements with increased accuracy as well as have
the advantage of maintaining the same physiologically-relevant
pH throughout the efflux experiment. This is especially important
since amphotericin B–liposome interactions are known to be very
sensitive to pH change.8
Biotin conjugate 5 was selected for thorough examination in
both liposomal (K+ efflux) and cellular assays. For comparison
purposes, amide-linked conjugate 10 was prepared (Fig. 4).
Both biotin-amphotericin B conjugates showed a K+ release pat-
tern similar to native amphotericin B (1) (Fig. 5). The piperazinyl-
linked conjugate 5 however, retained a high toxicity against fungal
cells (MIC = 20 lM) whereas the amide linked conjugate 10
showed a fivefold drop in activity (MIC = 100 lM) compared to 5.
This feature of probe 5 permits its use for subsequent investigations
of channel-forming properties in the membrane.11
As an alternative approach we have examined the use of a 3-
cyclopentene-1-carboxylic acid derivative, providing a diverse set
of building blocks for the construction of amphotericin conjugates.
The new linker strategy also allows for simple synthetic variations
of the binding motive. The commercially available cyclopentene
carboxylic acid is converted into the corresponding acid chloride
(oxalyl chloride, DMF) and treated with monoprotected 1,6-
diaminohexane in pyridine to afford a cyclopentenyl amide
which following dihydroxylation furnishes 11. Oxidation of this
diol to the dialdehyde and double reductive amination with
amphotericin B (1) installed the new linker. After deprotection
the diaminohexanoyl pyridyl amphotericin B 12 was isolated in
91% yield (Scheme 2).
In the vesicle assay that we have adapted for our purposes,9
LUVs from POPC or from POPC with admixed sterols (mimicking
conditions in natural biomembranes10) were prepared in a KCl
solution and the vesicles were dialyzed against NaCl in order to
create an ion gradient (K+ inside, Na+ outside). Freshly prepared
valinomycin-based K+-selective electrodes were characterized and
employed as described earlier.4c
The parent amphotericin B (1) triggered an immediate release
of K+ ions from liposomes containing cholesterol or ergosterol
(Fig. 2). Conjugate 4 showed an efflux pattern that was very similar
to native amphotericin B in the case of ergosterol containing
vesicles. However, a difference was found when comparing the
efflux from cholesterol and non-sterol containing liposomes.
The new conjugate 4 induced a clear distinction between the
three vesicle systems, a fact that may show beneficial influence
on the selectivity of the drug for fungal versus mammalian
cells.
We set out to establish the utility of 4 by preparation of
several amphotericin B conjugates incorporating typical affinity-
(5), fluorescent- (6), photolabelling- (8) groups (Fig. 3) as well
as amphotericin–cholesterol conjugate 7 and dimer 9. Piperazine-
linked amphotericin analog 4 was allowed to react with biotin N-
hydroxysuccinimide ester to give 5, with fluorescein isothiocyanate
to furnish 6, with cholesterol chloroformate to provide 7, and
with the N-hydroxysuccinimide ester of 4-(3-trifluoromethyl-3H-
diazirin-3-yl)-benzoate ester to afford diazirane 8. When 4 was
condensed with an activated form of glutaric acid dimer 9 was
isolated in 24% yield from amphotericin B, which is a considerable
improvement over the yield that had been previously described for
related dimeric compounds.7
Scheme 2 Synthesis of diaminohexyl pyridyl amphotericin B.
Both conjugates 4 and 12 bearing the novel linkers retained
the activity and were toxic above a threshold of 1.6 lM. The
piperidyl conjugate 12 not only showed remarkably reduced
release kinetics but a positive change was also found in the
increased selectivity for cholesterol-over ergosterol-containing
vesicles (Fig. 6).
In conclusion, we have established a new linker for bioconjuga-
tion through amines that allows for the synthesis of amphotericin
B conjugates in multi-gram scale. In a broader sense, the linking
strategy we document is suitable for bioconjugation in a number
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1339–1342 | 1341
©

View Article Online
Acknowledgements
We would like to thank Professor Dr Josef Brunner for kindly
providing us with the active ester of the diazirine photo-crosslinker.
We are grateful for generous support from ETH Zu¨rich in the form
of a TH-Gesuch.
Notes and references
1 (a) G Medoff, J Brajtburg and G. S Kobayashi, Annu. Rev. Pharmacol.
Toxicol., 1983, 23, 303–330; (b) J Bolard, Biochim. Biophys. Acta, 1986,
864, 257–304; (c) J Brajtburg, W. G Powderly, G. S. Kobayashi and G.
Medoff, Antimicrob. Agents Chemother., 1990, 34(2), 183–188; (d) F. C.
Szoka, Jr. and M. Tang, J. Liposome Res., 1993, 3(3), 363–375; (e) C.
Hatch, W. Ayenew and S. C. Hartsel, J. Liposome Res., 1993, 3(3),
377–408; (f) J. Bolard, V. Joly and P. Yeni, J. Liposome Res., 1993, 3(3),
409–427; (g) S. C. Hartsel and J. Bolard, Trends Pharmacol. Sci., 1996,
17, 445–449; (h) I. Al-Mohsen and W. T. Hughes, Ann. Saudi Med.,
1998, 18(1), 28–38; (i) B. E. Cohen, Int. J. Pharm., 1998, 162, 95–106;
(j) I. M. Hann and H. G. Prentice, Int. J. Antimicrob. Agents, 2001, 17,
161–169; (k) S. B. Zotchev, Curr. Med. Chem., 2003, 10, 211–223.
2 For a recent review of the biology and chemistry of amphotericin, see:
D. M. Cereghetti and E. M. Carreira, Synthesis, 2006, 914–942.
3 (a) C. C. Hsu-Chen and D. S. Feingold, Nature, 1974, 251, 656–659;
(b) J. Brajtburg, J. Medoff and M. L. Sokol-Andersen, J. Infect. Dis.,
1986, 154, 76–83; (c) W. H. Beggs, Antimicrob. Agents Chemother., 1994,
38, 363–364.
Fig. 5 K+ release from POPC based liposomes measured via potentiom-
etry. Substrate was added externally to lipid vesicles at a molar ratio
of 1 : 200 (reagent : total lipid), solid line: POPC alone, dashed line:
POPC admixed with 30 mol% cholesterol; dotted line: POPC admixed
with 13 mol% ergosterol.
4 (a) J. Kru¨ger and E. M. Carreira, Tetrahedron Lett., 1998, 39, 7013–
7016; (b) J. Tholander and E. M. Carreira, Helv. Chim. Acta, 2001,
84(3), 613–622; (c) A. Zumbuehl, P. Stano, M. Sohrmann, T. Vigassy,
E. Pretsch, M. Peter, P. Walde, D. Jeannerat and E. M. Carreira, Angew.
Chem., Int. Ed., 2004, 43, 5181–5185.
5 For additional recent studies with amphotericin, see: (a) A. Zumbuehl,
P. Stano, D. Heer, P. Walde and E. M. Carreira, Org. Lett., 2004,
6, 3683–3686; (b) V. Paquet and E. M. Carreira, Org. Lett., 2006, 8,
1807–1809; (c) J. M. Manthorpe, A. M. Szpilman and E. M. Carreira,
Synthesis, 2005, 3380–3388; (d) N. Matsumori, Y. Sawada and M.
Murata, J. Am. Chem. Soc., 2006, 128, 11977–11984.
6 M. Che´ron, B. Cybulska, J. Mazerski, J. Grzybowska, A. Czerwinski
and E. Borowski, Biochem. Pharmacol., 1998, 37(5), 827–836.
7 N. Matsumori, N. Yamaji, S. Matsuoka, T. Oishi and M. Murata, J. Am.
Chem. Soc., 2002, 124, 4180–4181.
Fig. 6 K+ release from POPC based liposomes measured via potentiom-
etry. Substrate was added externally to lipid vesicles at a molar ratio
of 1 : 200 (reagent : total lipid), solid line: POPC alone, dashed line:
POPC admixed with 30 mol% cholesterol; dotted line: POPC admixed
with 13 mol% ergosterol.
8 A. W. Norman, R. A. Demel, B. de Krujiff, B., W. S. M. Geurts
vanKessel and L. L. M. van Deenen, Biochim. Biophys. Acta, 1972,
290, 1–14.
9 (a) P. van Hoogevest and B. de Kruijff, Biochim. Biophys. Acta, 1978,
511, 397–407; (b) D. S. Orlov, T. Nguyen and R. I. Lehrer, J. Microbiol.
Methods, 2002, 49, 325–328.
10 J. Barwicz and P. Tancre`de, Chem. Phys. Lipids, 1997, 85, 145–155.
11 However, we believe that although LUVs can serve as excellent model
systems in a narrowly defined area we suggest caution in drawing a
direct correlation between liposomal and cell studies. In fact, in a
previous paper we documented a case with a conjugate where potassium
efflux from LUV is induced by a non-toxic compound up to 500 lM4c.
of other biologically important systems wherein functionalization
through a primary amine is required yielding molecules with
interesting new properties. A biotin conjugate was prepared as a
benchmark case which displayed comparable biological activity to
the parent compound and preserved K+ efflux-inducing properties
in a LUV assay. The new linker strategy and various prepared
amphotericin B conjugates will be useful in shedding some more
light on the mechanism of action of this important antifungal
drug.
1342 | Org. Biomol. Chem., 2007, 5, 1339–1342
This journal is
The Royal Society of Chemistry 2007
©