Ion channel formation from a calix[4]arene amide that binds HCl

Article abstract of DOI:10.1021/ja012338e

The ion transport activity of calix[4]arene tetrabutylamide 1,3-alt 2 was studied in liposomes, planar lipid bilayers, and HEK-293 cells. These experiments, when considered together with ¹H NMR and X-ray crystallography data, indicate that calix[4]arene tetrabutylamide 2 (1) forms ion channels in bilayer membranes, (2) mediates ion transport across cell membranes at positive holding potential, (3) alters the pH inside liposomes experiencing a Cl^- gradient, and (4) shows a significant Cl^-/SO₄^2- transport selectivity. An analogue, calix[4]arene tetramethylamide 1, self-assembles in the presence of HCl to generate solid-state structures with chloride-filled and water-filled channels. Structure-activity studies indicate that the hydrophobicity, amide substitution, and macrocyclic framework of the calixarene are essential for HCl binding and transport. Calix[4]arene tetrabutylamide 2 is a rare example of an anion-dependent, synthetic ion channel.

Full text of DOI:10.1021/ja012338e

Published on Web 02/16/2002
Ion Channel Formation from a Calix[4]arene Amide That Binds
HCl
Vladimir Sidorov,*,† Frank W. Kotch,† Galya Abdrakhmanova,‡ Robert Mizani,§
James C. Fettinger,† and Jeffery T. Davis*,†
Contribution from the Department of Chemistry and Biochemistry, UniVersity of Maryland,
College Park, Maryland 20742, Department of Pharmacology, 4000 ReserVoir Road,
Georgetown UniVersity, Washington, D.C., 20007, and Department of Biology, UniVersity of
Maryland, College Park, Maryland 20742
Received October 11, 2001
Abstract: The ion transport activity of calix[4]arene tetrabutylamide 1,3-alt 2 was studied in liposomes,
planar lipid bilayers, and HEK-293 cells. These experiments, when considered together with 1H NMR and
X-ray crystallography data, indicate that calix[4]arene tetrabutylamide 2 (1) forms ion channels in bilayer
membranes, (2) mediates ion transport across cell membranes at positive holding potential, (3) alters the
pH inside liposomes experiencing a Cl- gradient, and (4) shows a significant Cl-/SO42- transport selectivity.
An analogue, calix[4]arene tetramethylamide 1, self-assembles in the presence of HCl to generate solid-
state structures with chloride-filled and water-filled channels. Structure-activity studies indicate that the
hydrophobicity, amide substitution, and macrocyclic framework of the calixarene are essential for HCl binding
and transport. Calix[4]arene tetrabutylamide 2 is a rare example of an anion-dependent, synthetic ion
channel.
the properties of natural anion channels.7 Another rationale for
developing synthetic anion channels is that these compounds
Introduction
Ion transport across membranes is one of the most important
processes in living cells. Proteins that serve as ion channels or
carriers provide this crucial activity.1 The desire to understand
these proteins, along with a need for new antibiotics,2 has
spurred development of ion channel models.3 Although most
synthetic channels have been designed for cation transport, there
are some synthetic peptides that function as Cl--selective ion
channels.4 These peptides have sequences corresponding to the
transmembrane domains of anion channel proteins, such as the
cystic fibrosis transmembrane conductance regulator (CFTR).
There are fewer examples of nonpeptidic synthetic channels that
interact with anions. Most notably, Regen has proposed that
protonated amines in a spermine-steroid conjugate may facili-
tate Cl- movement across a membrane bilayer.5 Matile has
indicated that rigid-rod polyols can mediate anion transport.3d
Both these studies are encouraging for the development of new
synthetic anion channels. While there has been significant
progress in understanding the structure and function of natural
and artificial cation channels, much less is known about anion
channels.6 Studying appropriate models may provide insight into
may well control anion exchange processes in cells. For
example, synthetic peptides can restore Cl- transport in CFTR-
deficient cells.4e In this paper, we describe a calixarene that
forms a synthetic ion channel in a Cl- dependent fashion.
Both single molecule and self-assembly approaches have been
used to construct synthetic ion channels.3 Self-assembly is an
attractive approach for making pores within a membrane, as
functional structures can arise from simple building blocks.
Many synthetic cation channels have been made from com-
pounds that presumably form the active structures by self-
assembly. Nature certainly makes spectacular use of self-
assembly to form ion channels, as illustrated by the crystal
(3) For general reviews: (a) Gokel, G. W.; Murillo, O. Acc. Chem. Res. 1996,
29, 425. (b) Kobuke, Y. In AdVances in Supramolecular Chemistry; Gokel,
G. W., Ed.; JAI Press: Greenwich, CT, 1997; Vol. 4, pp 163-210. For
recent examples: (c) Sakai, N.; Brennan, K. C.; Weiss, L. A.; Matile, S.
J. Am. Chem. Soc. 1997, 119, 8726. (d) Weiss, L. A.; Sakai, N.;
Ghebremariam, B.; Ni, C.; Matile, S. J. Am. Chem. Soc. 1997, 119, 12142.
(e) Tedesco, M. M.; Ghebremariam, B.; Sakai, N.; Matile, S. Angew. Chem.,
Int. Ed. 1999, 38, 540. (f) Fyles, T. M.; Loock, D.; Zhou, X. J. Am. Chem.
Soc. 1998, 120, 2997. (g) Abel, E.; Maguire, G. E. M.; Meadows, E. S.;
Murillo, O.; Jin, T.; Gokel, G. W. J. Am. Chem. Soc. 1997, 119, 9061. (h)
Abel, E.; Maguire, G. E. M.; Murillo, O.; Suzuki, I.; De Wall, S. L.; Gokel,
G. W J. Am. Chem. Soc. 1999, 121, 9043. (i) Clark, T. A.; Buehler, L. K.;
Ghadiri, M. R. J. Am. Chem. Soc. 1998, 120, 651. (j) Otto, S.; Osifchin,
M.; Regen, S. L. J. Am. Chem. Soc. 1999, 121, 7276. (k) Baumeister, B.;
Sakai, N.; Matile, S. Angew. Chem., Int. Ed. 2000, 39, 1955. (l) Schrey,
A.; Vescovi, A.; Knoll, A.; Rickert, C.; Koert, U. Angew. Chem., Int. Ed.
2000, 39, 900. (m) Arndt, H.-D.; Knoll, A.; Koert, U. Angew. Chem., Int.
Ed. 2001, 40, 2076. (n) Bandyopadhyay, P.; Janout, V.; Zhang, L.; Regen
S. L. J. Am. Chem. Soc. 2001, 123, 7691. (o) Bandyopadhyay, P.; Janout,
V.; Zhang, L.; Sawko, J, A.; Regen S. L. J. Am. Chem. Soc. 2000, 122,
12888. (p) Ishida, H.; Qi, Z.; Sokabe, M.; Donowaki, K.; Inoue, Y. J. Org.
Chem. 2001, 66, 2978. (q) Gokel, G. W.; Ferdani, R.; Liu, J.; Pajewski,
R.; Shabany, H.; Uetrecht, P. Chem. Eur. J. 2001, 7, 33. (r) Kobuke, Y.;
Nagatani, T. J. Org. Chem. 2001, 66, 5094.
† Department of Chemistry and Biochemistry, University of Maryland.
‡ Department of Pharmacology, Georgetown University.
§ Department of Biology, University of Maryland.
(1) (a) Hille, B. Ionic Channels of Excitable Membranes, 2nd ed.; Sinauer:
Sunderland, MA, 1992. (b) Saier, M. H. J. Membr. Biol. 2000, 175, 165.
(2) (a) Fernandez-Lopez, S.; Kim, H. S.; Choi, E. C.; Delgado, M.; Granja, J.
R.; Khasanov, A.; Kraehenbuehl, K.; Long, G.; Weinberger, D. A.;
Wilcoxen, K. M.; Ghadiri, M. R. Nature 2001, 412, 452. (b) Porter, E. A.;
Wang, X. F.; Lee, H. S.; Weisblum, B.; Gellman, S. H. Nature 2000, 404,
565. (c) Hamuro, Y.; Schneider, J. P.; DeGrado, W. F. J. Am. Chem. Soc.
1999, 121, 12200. (d) Yamashita, K.; Janout, V.; Bernard, E. M.;
Armstrong, D.; Regen, S. L. J. Am. Chem. Soc. 1995, 117, 6249.
9
10.1021/ja012338e CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2267

A R T I C L E S
Sidorov et al.
Figure 1. Compounds used in this study.
structures of the mechanosensitive ion channel and the K+
channel.8
that connect the nucleosides to the calix[4]arene 1,3-alt scaffold.
We prepared compounds 1-4 shown in Figure 1 to determine
whether calix[4]arene 1,3-alt amides might self-assemble in the
presence of cations and/or anions. Herein, we report that calix-
[4]arene tetrabutylamide 2 binds HCl in solution, forms ion
channels in a lipid bilayer, and effectively transports HCl across
liposomal and cell membranes. We also present solid-state
evidence that the analogous calix[4]arene tetramethylamide 1
self-assembles in the presence of HCl to form ordered arrays
containing chloride-filled and water-filled channels. By compar-
ing ion binding and transport properties of calix[4]arene
tetrabutylamide 2 with those of its analogues 1, 3, and 4, we
find that the compound’s hydrophobicity, the amide’s substitu-
tion pattern, and the calixarene’s macrocyclic framework are
essential for mediating ion transport. The evidence points toward
a self-assembled channel being formed by calix[4]arene tet-
rabutylamide 2 in an anion-dependent process.
Expanding on the work of Shinkai9 and Hosseini,10 we
recently used cation-templated self-assembly to build tubular
constructs containing an ion-filled pore. Thus, a calix[4]arene
1,3-alt11 bearing four guanosine units formed micron-length rods
in the presence of NaBPh4.12 This controlled aggregation was
ascribed to the ability of Na+ to stabilize a hydrogen-bonded
G-quartet,13 a motif that bridged calixarenes into a tubular
structure. Calix[4]arenes are attractive scaffolds due to their
ready preparation and functionalization,11 their rigidity, and their
ability to bind charged14 and neutral species.15 Of relevance are
the findings that calix[4]arene-crown ethers16 and calix[4]-
resorcinarenes17 form synthetic cation channels in lipid mem-
branes.
While investigating self-assembly of our guanosine-calix-
arene conjugate,12 we focused attention on the secondary amides
Results and Discussion
(4) (a) Reddy, G. L.; Iwamoto, T.; Tomich, J. M.; Montal, M. J. Biol. Chem.
1993, 268, 14608. (b) Oblatt-Montal, M.; Reddy, G. L.; Iwamoto, T.;
Tomich, J. M.; Montal, M. Proc. Nat. Acad. Sci., U.S.A. 1994, 91, 1495.
(c) Wallace, D. P.; Tomich, J. M.; Eppler, J. W.; Iwamoto, T.; Grantham,
J. J.; Sullivan, L. P. Biochim. Biophys. Acta (Biomembranes) 2000, 1464,
69. (d) Mitchell, K. E.; Iwamoto, T.; Tomich, J.; Freeman, L. C. Biochim.
Biophys. Acta (Biomembranes) 2000, 1466, 47. (e) Broughman, J. R.;
Mitchell, K. E.; Sedlacek, R. L.; Iwamoto, T.; Tomich, J. M.; Schultz, B.
D. Am. J. Physiol.: Cell Physiol. 2001, 280, C451.
(5) (a) Deng, G.; Dewa, T.; Regen, S. L. J. Am. Chem. Soc. 1996, 118, 8975.
(b) Merritt, M.; Lanier, M.; Deng, G.; Regen, S. L. J. Am. Chem. Soc.
1998, 120, 8494.
(6) Ashcroft, F. M. Ion Channels and Disease; Academic Press: New York,
2000; pp 186-198.
(7) Anderson M. P.; Sheppard, D. N.; Berger, H. A.; Welsh, M. J. Am. J.
Physiol. 1992, 263, L1.
Calix[4]arene tetramethylamide 1,12 tetrabutylamide 2, and
octabutylamide 3 were prepared by coupling the appropriate
amine with calix[4]arene-1,3-alt tetraacid chloride.12 Synthesis
and characterization of 1-3 are detailed in the Experimental
Section.
Calix[4]arene tetrabutylamide 2 transports ions across the
bilayer without cation selectivity, whereas calix[4]arene
tetramethylamide 1 does not mediate ion transport. The
ability of 1 and 2 to mediate ion flux across a lipid bilayer was
studied fluorimetrically (Figure 2). A solution of 1 or 2 was
added to a suspension of EYPC large unilamellar vesicles
(LUVs) containing the pH-sensitive dye pyranine.18 A NaOH
solution was added to the extravesicular buffer to create a
gradient of approximately one pH unit. The resulting electro-
static potential, caused by proton efflux from the liposomes,
can be compensated by cation influx or anion efflux, as mediated
by the exogenous ligand.19 Ion transport across the membrane
was signaled by the increase in relative fluorescence that
accompanied the rise in intravesicular pH.
(8) (a) Chang, G.; Spencer, R. H.; Lee, A. T.; Barclay, M. T.; Rees, D. C.
Science 1998, 282, 2220. (b) Doyle, D. A.; Cabral, J. M.; Pfuetzner, R.
A.; Kuo, A. L.; Gulbis, J. M.; Cohen, S. L.; Chait, B. T.; MacKinnon, R.
Science 1998, 280, 69.
(9) Ikeda, A.; Shinkai, S. Chem. Commun. 1994, 2375.
(10) (a) Jaunky, W.; Hosseini, M. W.; Planeix, J. M.; De Cian, A.; Kyritsakas,
N.; Fischer, J. Chem. Commun. 1999, 2313. (b) Kleina, C.; Graf, E.;
Hosseini, M. W.; De Cian, A.; Fischer, J. Chem. Commun. 2000, 239.
(11) (a) Gutsche, C. D. Calixarenes. In Monographs in Supramolecular
Chemistry; Stoddart, J. F., Ed.; Royal Society of Chemistry: Cambridge,
1989. (b) Gutsche, C. D. Calixarenes Revisited. In Monographs in
Supramolecular Chemistry; Stoddart, J. F., Ed.; Royal Society of Chem-
istry: Cambridge, 1998.
(12) Sidorov, V.; Kotch, F. W.; El-Kouedi, M.; Davis, J. T. Chem. Commun.
2000, 2369.
(13) (a) Guschlbauer, W.; Chantot, J. F.; Thiele, D. J. Biomol. Struct. Dynam.
1990, 8, 491 (b) Wong, A.; Fettinger, J. C.; Forman, S. F.; Davis, J. T.;
Wu, G. J. Am. Chem. Soc. 2002, 124, 742-743.
(16) de Mendoza, J.; Cuevas, F.; Prados, P.; Meadows, E. S.; Gokel, G. W.
Angew. Chem., Int. Ed. 1998, 37, 1534.
(17) (a) Yoshino, N.; Satake, A.; Kobuke, Y. Angew. Chem., Int. Ed. 2001, 40,
457. (b) Tanaka, Y.; Kobuke, Y.; Sokabe, M. Angew. Chem., Int. Ed. 1995,
34, 693. (c) Wright, A. J.; Matthews, S. E.; Fischer, W. D.; Beer, P. D.
Chem. Eur. J. 2001, 7, 3474.
(14) (a) Bo¨hmer, V. Angew Chem., Int. Ed. Engl. 1995, 34, 713. (b) Beer, P.
D.; Gale, P. A. Angew. Chem., Int. Ed. 2001, 40, 487.
(15) (a) Arduini, A.; McGregor, W. M.; Paganuzzi, D.; Pochini, A.; Secchi, A.;
Ugozzoli, F.; Ungaro, R. Perkin Trans. 2 1996, 839. (b) Smirnov, S.;
Sidorov, V.; Pinkhassik, E.; Havlicek, J.; Stibor, I. Supramol. Chem. 1997,
8, 187.
(18) Kano, K.; Fendler, J. H. Biochim. Biophys. Acta 1978, 509, 289.
(19) For more on the pH-gradient transport assays see refs 3c-e.
9
2268 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

Ion Channel Formation from a Calix[4]arene Amide
A R T I C L E S
Table 1. Initial Pseudo-first Order Transport Rate Constants
(×10-3 s-1) for Intravesicular and Extravesicular Electrolyte
Exchange in the Presence of 5 µM of 2 or 3 and Different Salts
salt transporter
NaCl
KCl
CsCl
Na2SO
4
2a
3a
6.8b,c
6.4b,c
7.7b,c
5.4b,c
1.1b,d
9.1b,d
19b,c
12b,c
a 20 µL of 0.5 mM THF solution of the compound was added to 2 mL
of a 0.5 mM LUV suspension. b Rate constants were determined as described
in the Experimental Section. Values are not corrected for unmediated
electrolyte exchange. The accuracy of measurements was (15% for
experiments using the same set of liposomes. c 100 mM NaCl, 10 mM
phosphate buffer, pH 6.4 inside; 100 mM MCl, M ) Na, K, or Cs, 10 mM
phosphate buffer, pH 7.4 outside, after pH pulse. Note that values are not
directly comparable because of different ion gradients. d 100 mM Na2SO4,
10 mM phosphate buffer, pH 6.4 inside; 100 mM Na2SO4, 10 mM phosphate
buffer, pH 7.4 outside, after pH pulse.
observed in the course of the experiments. The slower phase of
fluorescence increase is due to the rise in intravesicular pH as
calixarene 2 transports ions across the EYPC membrane. It is
the rate constant for this second phase that we report in Table
1. Furthermore, negative controls shown in Figure 2 clearly
indicate that there is little fluorescence increase in the absence
of calixarene 2.
Calix[4]arene tetramethylamide 1 showed little transport
activity in the presence of extravesicular Na+, K+, or Cs+
chloride at ligand:lipid ratios of up to 1:20,23 whereas rapid
exchange of intra- and extravesicular electrolytes was observed
in the presence of calix[4]arene tetrabutylamide 2 at a ligand:
lipid ratio of 1:100 (Figure 2a). These fluorescence assays in
buffers containing Na+, K+, or Cs+ indicated that ion transport
mediated by 2 was nonselective toward the cation (Figure 2a,
Table 1). The reduced activity for tetramethylamide 1 can be
attributed to its lower hydrophobicity with respect to tetrabu-
tylamide 2.24
Figure 2. (A) Liposome transport assays with 2 in the presence of NaCl,
KCl, and CsCl. Suspensions of EYPC LUVs containing the pH-sensitive
dye pyranine in a phosphate buffer were used. The intravesicular solution
contained 10 mM sodium phosphate, pH 6.4, 100 mM NaCl and the
extravesicular solution contained 10 mM sodium phosphate, pH 6.4, 100
mM MCl (M ) Na, K, Cs). At t ) 0 s, 20 µL of a 0.5 mM solution of 2
in THF was added (top three curves) to give 1:100 2:lipid molar ratio or
20 µL of THF was added (bottom three curves). At t ) 60 s, 21 µL of a
0.5 M NaOH solution was added. At t ) 500 s, 40 µL of 5% Triton X-100
was added. Counting from top to bottom, traces 1 and 4 correspond to the
solutions containing 100 mM CsCl, traces 2 and 6 correspond to the
solutions containing 100 mM NaCl, and traces 3 and 5 correspond to
solutions containing 100 mM KCl buffer. (B) Schematic representation of
transport experiments. A pH gradient results from addition of extravesicular
NaOH solution. The charge caused by H+ efflux or OH- influx is
compensated by cation influx or anion efflux, as mediated by the exogenous
ligand such as 2. The increase in intravesicular pH, monitored by the
entrapped pH-sensitive dye, HPTS, reflects the electrolyte exchange rate.
Calix[4]arene tetrabutylamide 2 does not induce defects
into the bilayer membrane but provides selective transport
of Cl- over HSO4-. Calix[4]arene tetrabutylamide 2 may
transport ions across LUVs by inducing membrane defects, by
uniport of ions, or via cotransport of ion pairs (H+/M+ or OH-/
A- antiport and H+/A- or M+/OH- symport mechanisms are
possible). The lytic potential of 2 was evaluated using a dye
release assay. The self-quenching dye calcein loses 95% of its
fluorescence emission at concentrations above 100 mM.25
A
suspension of LUVs containing 120 mM calcein showed no
fluorescence enhancement upon dilution with an isoosmotic
buffer containing 2 (5-100 µM, 1-20 mol % ligand, relative
to total lipid concentration, data not shown). This calcein
experiment indicates that 2 does not induce membrane defects.
The polyanionic pyranine is known to adsorb strongly to
positively charged vesicles.18,20,21 Pyranine also adsorbs to some
extent to neutral liposomes such as EYPC.5b Our fluorescence
traces, which measure the ratio of deprotonated to protonated
forms of the dye (see Experimental Section), register a response
that is directly proportional to pH.22 In this way, dye adsorption
to the internal surface of the vesicles does not interfere with
the pH measurements. We can estimate the amount of dye
adsorbed to the outer surface of EYPC LUVs by analyzing the
burst response upon base addition (Figure 2a). This response
has never exceeded 15% of the total fluorescence change
Secondary amides can interact with anions and cations,
making it reasonable that calix[4]arene tetrabutylamide 2 could
(23) Initial pseudo-first-order rate constants determined for NaCl, KCl, and CsCl
in the presence of 25 µM of 1 were 2.5, 2.2, and 4.1 (× 10-3 s-1),
respectively (ligand: lipid ratio 1:20). Rate constants in the absence of 1
or 2 were in the range of 1.5-2.7 (× 10-3 s-1) for all three salts. No
transport was observed in the presence of 5 µM of 1(ligand:lipid ratio
1:100).
(24) The relative hydrophobicity of 1 and 2 was clearly different based on reverse
phase HPLC retention times (C-18, 1:1 CH3CN:0.1% aqueous TFA, flow
rate of 1 mL/min). Retention times were 4.8 and 54.1 min for 1 and 2,
respectively. For use of RPLC retention times as a measure of partition
coefficients: Kaliszan, R. in QuantitatiVe Structure-Chromatographic
Retention Relationships. 1987. pp 232-267, John Wiley & Sons: New
York.
(20) Clement, N. R.; Gould, J. M. Biochemistry 1981, 20, 1534.
(21) Dye adsorption can be minimized by incorporation of negatively charged
lipids into the liposomes. Such liposomes discriminate, however, in their
membrane permeability toward different monovalent cations (Papahad-
jopoulos, D. Biochim. Biophys. Acta 1971, 241, 254) and, therefore, may
cause misleading results in our transport assays.
(22) Venema, K.; Gibrat, R.; Grouzis, J.-P.; Grignon, C. Biochim. Biophys. Acta
1993, 1146, 87.
(25) New, R. R. C. in Liposomes: a Practical Approach; IRL Press: Oxford,
1990; pp 131-134.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2269

A R T I C L E S
Sidorov et al.
Calix[4]arene tetrabutylamide 2 mediates effective H+/
Cl- transport across bilayer membranes. With NMR evidence
that calix[4]arene tetrabutylamide 2 binds HCl in a hydrophobic
environment we asked whether 2 could transport HCl across
LUVs. Previously, we showed that a pH gradient across Cl--
filled vesicles was collapsed by addition of 2 (Figure 2a). We
studied the reverse situation by monitoring the intravesicular
pH of LUVs experiencing a Cl- gradient (Figure 5).
When a suspension of LUVs filled with saline phosphate
buffer (100 mM NaCl, 10 mM sodium phosphate, pH 6.4) and
suspended in isoosmotic Na2SO4 buffer (75 mM Na2SO4, 10
mM sodium phosphate, pH 6.4) was treated with 2, we observed
a rapid increase in intravesicular pH (Figure 5, top trace). The
pH reached its maximum ∼250 s after addition of 2 and held
constant until the liposomes were lyzed with detergent.28,29
These results are consistent with an H+/Cl- symport or Cl-/
OH- antiport process. In effect, compound 2 moves H+ along
with Cl- down the chloride gradient, consequently building a
pH gradient across the membrane. Therefore, NaCl-filled
vesicles lose Cl- and H+ upon addition of 2 and become more
alkaline.30,31
Figure 3. Transport assays with 2 in the presence of extravesicular buffers
containing NaCl (top trace) and Na2SO4 (lower trace). Suspensions of EYPC
LUVs containing the pH-sensitive dye pyranine in a phosphate buffer were
used. At t ) 20 s, 20 µL of a 500 µM of 2 in THF was added to 1.95 mL
of LUV suspension to give 1:100 2:lipid molar ratio. At t ) 60 s, 21 µL of
0.5 M NaOH was added to change the extravesicular pH from 6.4 to 7.4.
At t ) 500 s, 40 µL of 5% Triton X-100 was added. No significant transport
in the Na2SO4 buffer was observed even in the presence of up to 50 µM
of 2.
transport either ion type. Since the transport activity of calix-
[4]arene 2 was cation-independent (Figure 2a), we tested
whether ion transport mediated by 2 might be anion dependent.
Standard fluorescence assays were conducted with sodium
sulfate-filled LUVs suspended in an isotonic solution. Unlike
for chloride-containing solutions, experiments in sulfate buffers
revealed that the pH-induced ion flux across the membrane was
not mediated by 5-50 µM concentrations of 2 (1-10 mol %
The secondary amide NH and calix[4]arene framework
are essential for the HCl transport activity of 2. Calix[4]-
arene octabutylamide 3 was used as a negative control for
tetrabutylamide 2, as the lack of amide NH protons in 3 would
preclude Cl- binding. Instead, 3 should be a cation ionophore,
much like tetrakis(diethylcarbamoylmethoxy)-p-tert-butylcalix-
[4]arene-1,3-alt that coordinates cations in the solid-state and
solution.32 To confirm that 3 binds Na+, sodium picrate
extraction in CDCl3 was monitored by 1H NMR spectroscopy.
Complexation of the salt by 3 was evident. Two distinct
complexes were formed, corresponding to a 1:1 and a 2:1 Na+:3
stoichiometry (Figure 6). The 1:1 complex was generated by
adding 1 equiv of solid sodium picrate to a solution of 3. Binding
ligand, Figure 3). This pronounced Cl-/HSO4 transport
-
selectivity (Table 1) implies that 2 functions as a Cl- transporter.
1H NMR experiments show that calix[4]arene tetrabuty-
lamide 2 binds HCl. To obtain evidence for Cl- binding by 2,
1
we carried out H NMR experiments in CDCl3 using n-Bu4-
NCl and HCl as Cl- sources (Figure 4). Anion receptors usually
show chemical shift perturbations upon anion complexation,
particularly for amide NH protons that hydrogen bond to the
1
anion.26 Surprisingly, no changes in the H NMR spectrum of
(27) Attempts to distinguish H+/Cl- symport and Cl-/OH- antiport mechanisms
by carrying out transport assays on polarized vesicles failed. Electrical
silence of transport is usually ascribed to a symport mechanism. Electrical
current leading to a change in membrane potential is, however, possible
for both symport and antiport mechanisms. Although no apparent alteration
of transport rates due to the membrane potential was observed in the
presence of 2, monitoring the membrane potential revealed its rapid collapse
upon application of 2. Further transport occurred across the nonpolarized
the secondary amide 2 in CDCl3 were observed upon addition
of n-Bu4NCl (Figure 4b). This lack of spectral changes indicates
that 2 does not bind strongly to the “naked” Cl- anion, even in
this nonpolar solvent. Calix[4]arene tetrabutylamide 2 does,
however, bind HCl. Thus, shaking a CDCl3 solution of 2 with
aqueous HCl led to disappearance of the amide NH proton and
a significant upfield shift of the OCH2C(O) protons (Figure 4c).
The NMR spectral changes of 2 induced by HCl extraction were
not simply due to acid treatment, since washing a CDCl3 solution
of 2 with a H2SO4 solution did not cause any changes (Figure
4d).
These 1H NMR experiments indicate that calix[4]arene
tetrabutylamide 2 is a ditopic HCl receptor, since both H+ and
Cl- are needed to induce spectral changes. These NMR data
for 2 are consistent with the LUV transport results, where
efficient ion transport across the membrane was observed in
the presence of Cl- but not HSO4-. The combined fluorescence
and NMR results lead us to conclude that calix[4]arene
tetrabutylamide 2 supports ion transport across LUVs in an
anion-dependent fashion by either an H+/Cl- symport or a Cl-/
OH- antiport mechanism.27
membrane. Moreover, injection of
2 into a suspension of vesicles
experiencing a Cl- gradient (Na2SO4 inside, NaCl outside) resulted in a
potential with the negative pole inside the membrane. Apparently, H+ and
Cl- transport occur at different rates.
(28) A base-pulse assisted transport assay carried out on unequally loaded
vesicles (NaCl inside, Na2SO4 outside) revealed a ∼165% transport rate,
as judged from lysis of liposomes by Triton X-100.
(29) Formation of a pH gradient across the membrane arising from ionophore-
mediated Cl-/OH- exchange has been reported for Hg2+ and Cu+ salts:
(a) Karniski, L. P. J. Biol. Chem. 1992, 267, 19218, organotin, organolead,
and organomercury compounds: (b) Watling, A. S.; Selwyn, M. J. FEBS
Lett. 1970, 10, 139, (c) Selwyn, M. J.; Dawson, A. P.; Stockdale, M.; Gains,
N. Eur. J. Biochem. 1970, 14, 120. (d) Wieth, J. O.; Tosteson, M. T. J.
Gen. Physiol. 1979, 73, 765, and prodigiosins: (e) Sato, T.; Konno, H.;
Tanaka, Y.; Kataoka, T.; Nagai, K.; Wasserman, H. H.; Ohkuma, S. J.
Biol. Chem. 1998, 273, 21455.
(30) The proton regulatory function displayed by 2 is similar to that of the anion-
exchanging protein’s function in red blood cells: Jennings, M. L. J. Membr.
Biol. 1978, 40, 365.
(31) Preliminary experiments on inversely loaded vesicles (Na2SO4 inside, NaCl
outside, HPTS as dye) revealed acidification of the vesicular compartment
upon addition of 2. This result is qualitatively consistent with HCl influx.
The pH change for these inversely loaded vesicles was, however, lower
than that observed for NaCl-filled vesicles suspended in sulfate buffer
(Figure 5, top trace). This nonsymmetrical response in the two experiments
may be due to changes in osmotic pressure during the experiment or due
to Cl--mediated aggregation of 2 in the saline buffer. Such possibilities
will be investigated as part of our ongoing study.
(26) (a) Kavallieratos, K.; Bertao, C. M.; Crabtree, R. H. J. Org. Chem. 1999,
64, 1675. (b) Deetz, M. J.; Shang, M.; Smith, B. D. J. Am. Chem. Soc.
2000, 122, 6201.
(32) Beer, P. D., Drew, M. G. B.; Gale, P. A.; Leeson, P. B.; Odgen, M. I. J.
Chem. Soc., Dalton Trans. 1994, 3479.
9
2270 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

Ion Channel Formation from a Calix[4]arene Amide
A R T I C L E S
1
Figure 4. A series of H NMR spectra (400 MHz, CDCl3, 22 °C) of calixarene tetrabutylamide 2 in the presence of different species: (A) compound 2
alone (5 mM); (B) 2 with 1 equiv of Bu4NCl; (C) after washing CDCl3 solution of 2 with aqueous solution of 0.5 M HCl; (D) after washing CDCl3 solution
of 2 with aqueous solution of 0.5 M H2SO4.
this Na+ led to a symmetry loss for 3, indicated by doubling of
the H NMR peaks. Addition of further sodium picrate (up to
containing NaCl to one containing Na2SO4 (Figure 3, Table 1).33
Further evidence for different ion transport selectivity provided
by calixarene octabutylamide 3 and tetrabutylamide 2 was
obtained from an experiment with NaCl-loaded vesicles sus-
pended in isoosmotic Na2SO4 solution. Addition of 3 did not
result in formation of a pH gradient across the membrane (Figure
5, bottom trace), whereas addition of 2 to the identical system
had produced a significant pH gradient (Figure 5, top trace).
To determine whether chloride binding and transport is an
intrinsic property of calixarene 2 or a general property of
hydrophobic secondary amides, N-butyl-2-phenoxyacetamide 4
(Figure 1) was tested in the LUV transport assays. Compound
4 is a monomeric analogue of calixarene 2, bearing a secondary
amide that could bind Cl-, but lacking the calixarene’s
macrocycle. The standard fluorescence assay, even in the
1
10 equiv) resulted in a new 2:1 complex. Binding of the second
Na+ cation restored the ligand’s initial D2d symmetry, as seen
by the new set of single NMR signals for this 2:1 complex.32
In contrast, sodium picrate extraction by calix[4]arene tetra-
1
butylamide 2 was not detected by H NMR spectroscopy.
Ion transport assays revealed a high activity for octabutyla-
mide 3 in NaCl buffer (k ) 1.9 × 10-2 s-1, Table 1). Unlike 2,
calix[4]arene octabutylamide 3 discriminates between monova-
lent cations, showing a 3.5-fold drop in rate constant when
changing the extravesicular electrolyte from NaCl to CsCl (Table
1). Fluorescence assays with 3 in the presence of Na2SO4 loaded
vesicles suspended in isotonic buffer indicated that 3 still
mediated significant ion transport, albeit with a drop in activity
(k ) 9.1 × 10-3 s-1). In marked contrast, calix[4]arene
tetrabutylamide 2 had shown essentially complete loss of
transport activity when the buffer was changed from one
(33) The initial pseudo-first-order rate constant changed from 7.9 to 1.1 (× 10-3
2-
s-1) for 5 µM of 2 when changing from a Cl- to SO4 buffer. The latter
value is indistinguishable from the rate constant in the absence of 2.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2271

A R T I C L E S
Sidorov et al.
Figure 5. Changes in intravesicular pH upon addition of calixarene tetrabutylamide 2 and calixarene octabutylamide 3 to NaCl-loaded vesicles suspended
in a sulfate phosphate buffer (emission of HPTS is used to monitor the pH changes in LUV’s). At t ) 60 s, 20 µL of 500 µM of 2 (top trace) or 3 (bottom
trace) was added to 1.95 mL of a LUV suspension to give 1:100 ligand:lipid molar ratio. At t ) 500 s, 40 µL of 5% Triton X-100 was added. Intravesicular
pH values were obtained as a function of the ratio of HPTS emission intensities at 510 nm, when excited at 403 and 460 nm. The calibration was performed
by measuring the HPTS emission intensities and the pH values of a 4.70 nM HPTS solution in 10 mM phosphate buffer containing 100 mM NaCl. The
calibration equation obtained (pH ) 1.1684 × log(Io/I1) + 6.9807, r ) 0.998, where Io is the emission intensity with excitation at 460 nm and I1 is emission
intensity with excitation at 403 nm) was used to convert the emission intensities into actual pH values.
Figure 6. (A) Region of the 1H NMR spectrum (400 MHz, CDCl3, 22 °C) of calixarene octabutylamide 3 (5 mM) in CDCl3, (B) compound 3 after addition
of 1 equiv of NaPic, (C) compound 3 after addition of 2 equiv of NaPic. No change in this NMR spectrum was observed in the presence of up to 10 equiv
of solid NaPic.
presence of high concentrations of 4 (50 µM) showed no ion
transport across the LUV membrane. In addition, H NMR
spectroscopy indicated no HCl binding by 4 in CDCl3. These
controls with N-butyl-2-phenoxyacetamide 4 support the hy-
pothesis that the calixarene’s core is critical for the HCl binding
and ion transport properties of 2.
Calix[4]arene tetramethylamide 1 and HCl form channels
in the solid state. After failing to crystallize tetrabutylamide
2, we turned attention to the analogous calix[4]arene tetram-
ethylamide 1. Whereas 1 does not transport ions efficiently
across the LUV membrane, presumably due to insufficient
hydrophobicity, its mode of HCl binding should be similar to
1
9
2272 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

Ion Channel Formation from a Calix[4]arene Amide
A R T I C L E S
Table 2. Crystal Data and Structure Refinement for 1‚HCl‚3(H2O) and 1‚HCl‚H2O‚CH2Cl2 Complexes
1‚HCl‚3(H O)
1‚HCl‚H O‚CH Cl2
2
2
2
empirical formula
formula weight
temperature
C40H51ClN4O11
799.29
193(2) K
C41H49Cl3N4O9
848.18
193(2) K
crystal system
space group
orthorhombic
Fddd
monoclinic
P2(1)/n
unit cell dimensions
a ) 20.9026(7) Å, R) 90°
b ) 26.5756(9) Å, â) 90°
c ) 27.7713(10) Å, γ ) 90°
15426.9(9) Å3
16
a ) 10.1237(3) Å, R) 90°
b ) 22.7400(7) Å, â) 97.07°
c ) 17.7902(6) Å, γ ) 90°
4064.4(2) Å3
volume
Z
4
density (calculated)
absorption coefficient
data/restraints/parameters
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
1.375 mg/m3
1.384 mg/m3
0.166 mm-1
3404/3/326
1.060
R1 ) 0.0617, wR2 ) 0.1785
R1 ) 0.0851, wR2 ) 0.1958
0.286 mm-1
7147/0/723
1.092
R1 ) 0.0431, wR2 ) 0.1117
R1 ) 0.0570, wR2 ) 0.1196
that for 2. Different protocols afforded two types of 1‚HCl single
crystals whose structures were solved by X-ray crystallography
(see Table 2 for data). Both solid-state structures unequivocally
show that Cl- anion is hydrogen bonded to the calixarene’s
amide NH.
Crystals of 1‚HCl‚3(H2O), obtained by evaporation of a
dichloromethane/THF solution of 1 and HCl, have a structure
that consists of two-dimensional planar arrays of calixarenes
held together by NH‚‚‚Cl-‚‚‚HN hydrogen bonds in the x-
direction and by CdO‚‚‚H-O-H‚‚‚OdC hydrogen bonds in
the y-direction (Figure 7). A side view in the y-direction reveals
a chloride-filled channel, and a side view in the x-direction
reveals a water-filled channel.
Analysis of a second crystal (1‚HCl‚H2O‚CH2Cl2), grown by
layering aqueous HCl over a dichloromethane solution of 1,
revealed a structure where two molecules of calixarene 1 are
linked by oppositely directed N‚‚‚Cl-‚‚‚H-O(H)+-H‚‚‚OdC
bridges to give a dimer (Figure 8). Each molecule of calixarene
1 in 1‚HCl‚H2O‚CH2Cl2 retains a similar geometry as observed
in the structure of 1‚HCl‚3(H2O). The main difference in the
second structure is that a hydronium cation separates Cl- from
the neighboring calixarene.
Although the three-dimensional network formed by the two
1‚HCl complexes is different, both crystal structures indicate
that calixarene 1 uses its secondary amide groups to bind HCl.
These crystal structures are relevant since they indicate that
calix[4]arene 1,3-alt amides can self-assemble to give extended
channels held together by hydrogen bonds to bridging Cl-
anions and water.
Calix[4]arene tetrabutylamide 2 forms an ion channel.
Ligand-mediated ion transport across a lipid bilayer is usually
attributed to three mechanisms; defect induction, a carrier
process, or channel formation.3d,34 Dye-release experiments
showed that 2 does not induce membrane defects. Both carrier
and ion channel mechanisms for H+/Cl- transport are conceiv-
able for a ligand like calix[4]arene tetrabutylamide 2. A single
molecule of 2 is certainly too small to span the membrane. But
the calixarene’s 1,3-alt conformation makes channel-like struc-
tures possible, particularly in the presence of a bridging species.
Crystal structures of the 1‚HCl complexes show calixarenes held
together by amide‚‚‚chloride‚‚‚amide, amide‚‚‚water‚‚‚amide,
and amide‚‚‚chloride‚‚‚water‚‚‚amide hydrogen bonds. If HCl-
mediated self-assembly of 2 gives similar structures in lipid
membranes then such aggregates could act as ion channels. Even
the shortest hydrogen-bonded unit in these structures, a calix-
arene dimer of ∼25 Å (Figure 8), could conceivably span a
bilayer membrane.35,36
Since ion transport supported by 2 is not electrically silent,27
current across the bilayer can be detected using the voltage-
clamp technique, a technique that distinguishes between
channel and carrier mechanisms. Evidence for ion channels
formed by calix[4]arene tetrabutylamide 2 was obtained from
voltage-clamp experiments on black lipid membranes (BLMs)
(Figure 9).
The voltage-clamp assay was conducted using a setup with
two chambers, each filled with an electrolyte solution and
separated by a BLM. After voltage was applied across the
membrane, a solution of 2 was added to one of the chambers.
Subsequent current across the membrane was attributed to ion
transport mediated by 2, since control experiments with DMSO
solvent were negative. Typical current recordings are shown in
Figure 9a. The abrupt changes in current intensity (events) are
consistent with an ion channel mechanism. The relatively long-
lasting open states indicate that channels formed by calixarene
tetrabutylamide 2 are stable.
Analysis of 480 events gave a distribution of conductance
levels (50 pS-2 nS) centered near 100 pS (Figure 9b). We
attribute the various conductivity levels to formation of different
self-assembled channel structures.37 One explanation for dif-
ferent assemblies formed by 2 is competition of calix[4]arene
amides with water for Cl- binding, as revealed in the two
different crystal structures for 1‚HCl complexes (Figures 7 and
8). While one can only speculate about the structure of the active
species, the experiments summarized in Figure 9 clearly indicate
that calixarene 2 can form ion channels.
Calix[4]arene tetrabutylamide 2 forms voltage-dependent
channels in HEK-293 cell membranes. Patch-clamp recordings
in a whole-cell configuration (Figure 10a) revealed that extra-
cellular application of 2 to HEK-293 cells caused current across
(35) (a) Caffrey, M.; Feigenson, G. W. Biochemistry 1981, 20, 1949. (b) Lewis,
B. A.; Engelman, D. M. J. Mol. Biol. 1983, 166, 211.
(36) The active gramicidin dimer was estimated to have a length of 22 Å: (a)
Ketchem, R. R.; Hu, W.; Cross, T. A. Science 1993, 261, 1457. (b) Goulian,
M.; Mesquita, O. N.; Fygenson, D. K.; Nielsen, C.; Andersen, O. S.;
Libchaber, A. Biophys. J. 1998, 74, 328.
(37) (a) Renkes, T.; Schafer, H. J.; Siemens, P. M.; Neumann, E. Angew. Chem.,
Int. Ed. 2000, 39, 2512. (b) Fyles, T. M.; Knoy, R.; Mullen, K.; Sieffert,
M. Langmuir 2001, 17, 6669.
(34) Bell, T. W. Curr. Opin. Chem. Biol. 1998, 2, 711.
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2273

A R T I C L E S
Sidorov et al.
Figure 7. Top and side views of the hydrogen-bonded array found in the crystal structure of 1‚HCl‚3(H2O). The layer continues in the x and y directions
infinitely. The z-direction is composed of identical layers held together by crystal-packing forces. The Cl- anions are shown as green spheres. All molecules
in the structure are shown in a wire frame representation following the standard color scheme. Hydrogen bonds between the water molecules depicted as
tetrahedral (due to oscillation between two different binding modes) are omitted for clarity.
the cell membrane at holding potentials of +120 mV (3 of 15
cells actually responded at +80 mV, whereas the other cells
responded at +120 mV). Importantly, this current was observed
only when the cells were suspended in NaCl buffer. Replace-
ment of NaCl buffer with isoosmotic Na glutamate (Figure 10b)
or Na2SO4 (Figure 10c) resulted in no current across the
membrane at holding potentials between -120 and +120 mV.
These results lend further credence to the Cl--selectivity of ion
channels formed from calix[4]arene tetrabutylamide 2.
Calixarene tetrabutylamide 2 could be washed out from the
cell membrane. After 1.5 min of lavage, the current returned to
the baseline (data not shown) and no further response at +120
mV was observed without further addition of 2. These data
indicate that application of 2 did not damage the cell membrane.
Subsequent reapplication of 2 resulted in a shortening of the
delay prior to the response (Figure 10d). One explanation for
this phenomenon is that washing did not completely remove 2
from the cell membrane. Instead, inactive aggregates of 2 may
remain. These membrane-retained “seeds” may then accelerate
formation of new ion channels upon reapplication of the
compound.
Conclusions
Calix[4]arene 1,3-alt tetrabutylamide 2 binds HCl and forms
ion channels in lipid bilayer membranes. This is a rare example
of a synthetic, non-peptide ion channel showing an anion
dependence. We attribute the HCl binding and transport activity
of 2 to the calixarene’s three-dimensional structure, its hydro-
phobicity and the amide side chain’s substitution pattern.
Alteration of any of these features results in a selectivity switch
from anions to cations (compound 3) or to loss of transport
activity (compounds 1 and 4). Ion transport provided by 2
proceeds via either an H+/Cl- symport or a Cl-/OH- antiport
mechanism. Compound 2 can regulate H+ and Cl- gradients
across the membrane. Two crystal structures demonstrate that
the calix[4]arene 1,3-alt tetraamide motif is capable of forming
9
2274 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

Ion Channel Formation from a Calix[4]arene Amide
A R T I C L E S
Figure 8. Self-assembled dimer observed within the 1‚HCl‚H2O‚CH2Cl2 crystal structure. The CH2Cl2 molecules are omitted. The Cl- anions are shown
as green spheres. All molecules in the structure are shown in a wire frame representation following the standard color scheme.
Experimental Section
1H NMR spectra were recorded on a Bruker DRX400 instrument at
400.130 MHz. Chemical shifts are reported in ppm relative to the
residual solvent peak. The 13C NMR spectra were recorded at 100.613
MHz and chemical shift values are reported in ppm relative to the
solvent peak. Mass spectra were recorded on a JEOL SX-102A
magnetic sector mass spectrometer using fast atom bombardment.
Fluorimetric experiments were done on an SLM Aminco (Aminco
Bowman Series 2) Luminescence Spectrometer. Vapor pressure os-
mometry (VPO) was performed on a Wescor 5520, Vapro instrument.
Solution pH was monitored with an Orion pH-meter, model 420A, with
a Ag/AgCl pH-sensitive electrode. Chromatography was performed
using 60-200 mesh silica from Baker. Thin-layer chromatography was
performed on Kieselgel 60 F254 and Uniplate Silica Gel GF silica-
coated glass plates and visualized by UV. HPLC analysis was performed
on a Shimadzu LC-10AS liquid chromatograph with a Shimadzu SPD-
10 A UV-Vis detector set to 254 nm. A C-18 silica column was used
as a stationary phase, and 1:1 CH3CN:0.1% aqueous TFA was the
mobile phase. Chemicals and solvents were purchased from Sigma,
Fluka, Aldrich, or Acros. EYPC and cholesterol were purchased from
Avanti Polar Lipids.
Synthesis. Calix[4]arene Tetramethylamide 1. Calix[4]arene tet-
ramethylamide 1 was synthesized as described.12
1‚HCl crystals: (a) Calix[4]arene tetramethylamide 1 (15 mg) was
dissolved in 1.5 mL of THF:12 M HCl (95:5). The solution slowly
evaporated to give an oily precipitate that was dissolved in 2 mL of
CH2Cl2. Slow evaporation afforded crystals of 1‚HCl‚3(H2O). (b) Calix-
[4]arene tetramethylamide 1 (15 mg) was dissolved in 2 mL of CH2-
Cl2, and then 1 mL of 6 M aqueous HCl was layered over the organic
solution. A white emulsion in the organic layer occurred immediately.
Figure 9. (A) Representative current events across a BLM in voltage clamp
experiments after application of calixarene tetrabutylamide 2 (2.5 µM). The
voltage applied was -20 mV for all three recordings. (B) The distribution
of conductance ranges observed in voltage clamp experiments with 2 acting
as an ion channel.
(38) While mammalian cells have a resting potential of -40 to -100 mV (ref
1a, p 24), Streptococcus boVis, the bacteria that causes endocarditis, has a
transmembrane potential of up to +158 mV (positive pole inside) in the
presence of glutamine, rendering this bacteria potentially susceptible to 2.
On endocarditis, see: Kupferwasser, I.; Darius, H.; Muller, A. M.; Mohr-
Kahaly, S.; Westermeier, T.; Oelert, H.; Erbel, R.; Meyer, J. Heart 1998,
80, 276. On S. boVis, see: Cook, G. M.; Russell, J. B. FEMS Microbiol.
Lett. 1993, 111, 263.
self-assembled channels filled with Cl- anions and water.
Finally, compound 2 enables Cl--selective and voltage-depend-
ent ion transport across HEK-293 cell membranes. This last
property makes calix[4]arene 1,3-alt tetrabutylamide 2 a po-
tential antibiotic, capable of selective interactions with bacteria.38
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2275

A R T I C L E S
Sidorov et al.
Figure 10. Voltage clamp recordings in whole cell configuration (HEK-293 cells). (A) Current across membrane after application of 2 at different voltages.
From bottom to top, traces denote currents at -120, -80, -40, 0, 40, 80, 120 mV, respectively, with saline physiological buffer being used. (B) Current
across membrane in a sodium glutamate isoosmotic buffer at different voltages. The potential was varied over a range of -120 through +120 mV, with a
40 mV step. (C) Current across membrane in Na2SO4 isoosmotic buffer at different voltages. The potential was varied over a range of -120 through +120
mV, with a 40 mV step. (D) Current across the membrane at +120 mV in application/wash cycle: traces from bottom to top denote: current in absence of
2, current after first application of 2 (blue trace), followed by 1.5 min lavage, second application (red trace) and the third application (green trace). Trace
on top designates current after lavage without further application of 2. Baselines of the first and the last traces shifted for clarity, saline physiological buffer
used.
Crystals of 1‚HCl‚H2O‚CH2Cl2 formed at the aqueous-organic interface
after several days.
methylene chloride were added 2 mL of triethylamine (14.3 mmol,
5.5 equiv) and 3 mL of dibutylamine (17.8 mmol, 6.8 equiv). The
reaction mixture was stirred overnight at room temperature and then
washed with water and 0.1 N HCl, evaporated to dryness under reduced
pressure, and submitted to column chromatography (silica gel,
MeOH: CH2Cl2 6:94) to give 3, contaminated with its Na+ complex
(determined by 1H NMR). Repeated washing of this material with
deionized water gave 443 mg of 3 (0.40 mmol, 62%) as a white solid.
1H NMR (CDCl3, 25 °C) δ: 7.16 (d, J ) 8.1 Hz, 8 H, ArH), 6.67 (t,
J ) 7.7 Hz, 4 H, ArH), 4.41 (s, 8 H, ArOCH2CO), 3.56 (s, 8 H, ArCH2-
Ar), 3.41 (t, J ) 7.7 Hz, 8 H, ArOCH2CON(CH2)2), 3.23 (t, J ) 7.7
Hz, 8 H, ArOCH2CON(CH2)2), 1.63-1.52 (m, 16 H, ArOCH2CON-
(CH2CH2CH2CH3)2), 1.40-1.27 (m, 16 H, ArOCH2CON(CH2CH2CH2-
CH3)2), 0.95 (t, J ) 7.3 Hz, 12 H, ArOCH2CON(CH2CH2CH2CH3)2,
0.92 (t, J ) 7.3 Hz, 12 H, ArOCH2CON(CH2CH2CH2CH3)2). 13C NMR
(CDCl3, 25 °C) δ: 167.6, 155.39, 133.3, 129.5, 123.0, 71.7, 47.0, 45.5,
33.6, 31.2, 29.7, 20.2, 20.0, 13.9, 13.7. FAB MS ([M+H]+): 1101.7,
calcd for C68H101N4O8 1101.75.
Calix[4]arene tetrabutylamide 2. (a) To a suspension of tetrakis-
(carboxymethoxy)-p-H-calix[4] arene 1,3-alt12 (700 mg, 1.07 mmol)
in 30 mL of benzene was added 2 mL of thionyl chloride (27.4 mmol,
6.4 equiv per COOH group). The reaction mixture was stirred at reflux
for 2.5 h, the solvent was evaporated under reduced pressure, and
residual SOCl2 was removed by coevaporation with benzene. IR
spectroscopy indicated conversion of the carboxylic acid groups to acid
chlorides. The acid chloride was used immediately without purification.
(b) To a solution of the acid chloride in 30 mL of dry CH2Cl2 were
added 3 mL of triethylamine (21.5 mmol, 5 equiv) and 3 mL of
butylamine (30.4 mmol, 7.1 equiv). The reaction mixture was stirred
overnight at room temperature and washed with water and 0.1 N HCl,
and the resulting material was purified by column chromatography
(silica gel, MeOH:CH2Cl2 4:96) to give 814 mg (0.93 mmol, 87% yield)
of 2 as a white solid. 1H NMR (CDCl3, 25 °C) δ: 7.00 (d, J ) 7.5 Hz,
8 H, ArH), 6.73 (t, J ) 7.5 Hz, 4 H, ArH), 6.26 (t, J ) 6.2 Hz, 4 H,
CONH), 4.05 (s, 8 H, ArOCH2CO), 3.58 (s, 8 H, ArCH2Ar), 3.41 (m,
8 H, ArOCH2CONHCH2CH2CH3), 1.66 (m, 8 H, Ar-OCH2-CO-
NH-CH2-CH2-CH3), 1.54 (m, 8 H, ArOCH2CONHCH2CH2CH3),
1.06 (t, J ) 7.5 Hz, 12 H, ArOCH2CONHCH2CH2CH3). 13C NMR
(CDCl3, 25 °C) δ: 167.7, 154.3, 133.9, 131.2, 122.1, 70.1, 38.8, 36.2,
32.5, 20.2, 13.9. FAB MS ([M + H]+): 877.5, calcd for C52H69N4O8:
877.50.
N-Butyl-2-phenoxyacetamide 4. Phenoxyacetic acid (2 g, 13.15
mmol) was activated with SOCl2 (3 mL, 41.1 mmol, 3 equiv) in 50
mL of dry benzene using the procedure described for 2. To a solution
of the acid chloride in 30 mL of dry methylene chloride were added 3
mL of Et3N (21.5 mmol, 1.6-fold excess) and 3 mL of BuNH2 (30.4
mmol, 2.3-fold excess). The reaction mixture was stirred overnight at
room temperature, the solvent was evaporated to dryness under reduced
pressure, and the resulting oil was dried under high vacuum. Purification
by column chromatography (silica gel, MeOH:CH2Cl2 2:98 vv) gave
2.43 g (11.71 mmol, 89%) of 4 as a crystalline solid. 1H NMR (CDCl3,
Calix[4]arene Octabutylamide 3. Tetrakis(carboxymethoxy)-p-H-
calix[4]arene12 (423 mg, 0.65 mmol) was activated with thionyl chloride
as described for 2. To a solution of the acyl chloride in 30 mL of dry
9
2276 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002

Ion Channel Formation from a Calix[4]arene Amide
A R T I C L E S
25 °C) δ: 7.32 (dd, J ) 8.0, 7.2 Hz, 2 H, Ar-H), 7.02 (t, J ) 7.2 Hz,
1 H, ArH), 6.91 (d, J ) 8.0 Hz, 2 H, ArH), 6.57 (bs, 1 H, NH), 4.48
(s, 2 H, ArOCH2CO), 3.34 (q, J ) 6.8 Hz, 2 H, NHCH2CH2CH2CH3),
1.50 (m, 2 H, NHCH2CH2CH2CH3), 1.33 (m, 2 H, NHCH2CH2CH2-
CH3), 0.91 (t, J ) 7.2 Hz, 3 H, NHCH2CH2CH2CH3). 13C NMR (CDCl3,
25 °C) δ:168.0, 157.1, 129.7, 122.0, 114.5, 67.3, 38.7, 31.5, 19.9, 13.6;
MS (FAB) ([M]+): 207.3, calcd for C12H17NO2 207.13.
varied depending on the age of the vesicular solution and the actual
stock solution of liposomes used. Ratios between absolute values of
rate constants obtained from experiments on the same batch of
liposomes, however, did not vary significantly.
Calcein-Release Assay. Calcein-loaded vesicles (100 µL of the stock
solution) were suspended in 1.85 mL of isoosmotic buffer and submitted
to fluorescence analysis. Calcein emission was monitored at 520 nm
with excitation at 490 nm. During the experiment, 20 µL of a 500 µM
to 10 mM THF or THF/MeOH solution of 1 or 2 was added, followed
by injection of 0-21 µL of 0.5 M aqueous NaOH and 40 µL of 5%
aqueous Triton x100.
Liposome Preparation. EYPC HPTS-Containing LUVs. Egg yolk
L-R-phosphatidylcholine (EYPC, 60 mg, 79 µM) was dissolved in a
CHCl3/MeOH mixture, the solution was evaporated under reduced
pressure, and the resulting thin film was dried under high vacuum for
2 h. The lipid film was hydrated in 1.2 mL of phosphate buffer (10
mM sodium phosphate, pH ) 6.4, 75-100 mM MnX, M ) Na+, K+,
Cs+, X ) Cl-, SO42-) containing 10 µM HPTS (pyranine, 8-hydroxy-
pyrene-1,3,6-trisulfonic acid trisodium salt) for 2 h. The LUV suspen-
sion (1 mL) was submitted to high-pressure extrusion at room
temperature (21 extrusions through a 0.1 µm polycarbonate membrane
afforded LUVs with a diameter of 100 nm). The LUV suspension was
separated from extravesicular dye by size exclusion chromatography
(SEC) (stationary phase: Sephadex G-10, mobile phase: phosphate
buffer) and diluted with the same phosphate buffer to give a stock
solution with a lipid concentration of 11 mM (assuming 100% of lipid
was incorporated into liposomes).
Analysis of pH Changes in the Liposomes Experiencing a Cl-
Gradient. HPTS-loaded vesicles (100 µL of the stock solution), filled
with a saline phosphate buffer (10 mM sodium phosphate, pH 6.4, 100
mM NaCl), were suspended in 1.85 mL of an isoosmotic phosphate
sulfate buffer (10 mM sodium phosphate, pH 6.4, ∼75 mM Na2SO4)
and placed into a fluorimetric cell. HPTS emission at 510 nm was
monitored with excitation wavelengths at 403 and 460 nm simulta-
neously. During the experiment, 20 µL of a 0-500 µM THF solution
of the compound of interest was added through an injection port.
Intravesicular pH values were obtained as a function of the ratio of
HPTS emission intensities at 510 nm, when excited at 403 and 460
nm. The calibration was performed by measuring the HPTS emission
intensities and the pH values of a 470 pM HPTS solution in 10 mM
phosphate buffer containing 100 mM NaCl (pH was varied in the range
of 5.6 through 7.6 by addition of 0.5 M NaOH or 0.5 M HCl). The
calibration equation obtained (pH ) 1.1684 × log(Io/I1) + 6.9807, r
) 0.998, where Io is the emission intensity with excitation at 460 nm
and I1 is emission intensity with excitation at 403 nm) was applied to
convert the emission intensities into pH values. At the end of
experiment, the aqueous compartment of liposomes was equilibrated
with extravesicular solution by lysis of liposomes with detergent (40
µL of 5% aqueous Triton x100).
EYPC Calcein-Containing Vesicles. Calcein (224.1 mg, 0.36 mmol)
was suspended in 3 mL of phosphate buffer (100 mM NaCl, 10 mM
sodium phosphate, pH 6.4). The pH of the solution was adjusted to
6.4 with a 7.4 M NaOH-containing phosphate buffer, at which point
all the calcein was dissolved, to give a 120 mM solution of the dye.
The molality of the extravesicular buffer was adjusted to equal that of
the calcein-containing solution with 1 M NaCl. Osmomolality was
determined by VPO. A thin film was prepared from 60 mg of EYPC
(79 µM) as described above. The lipid film was hydrated in 1.2 mL of
the calcein buffer. During hydration, the suspension was submitted to
five freeze-thaw cycles (dry ice in acetone, water at room temperature)
and then submitted to high-pressure extrusion (21 extrusions through
a 0.1 µm polycarbonate membrane afforded LUVs with an average
diameter of 100 nm). Calcein-containing LUVs were separated from
extravesicular dye by SEC (stationary phase: Sephadex G-25, mobile
phase: extravesicular buffer with adjusted molality).
Electrophysiological Recordings. Patch-Clamp Experiments.
HEK 293 cells, plated in 100-mm dishes containing 10 mL of the
growth media, were maintained at 37 °C with 5% CO2 in the incubator.
Growth medium for HEK 293 cells was minimum essential medium
supplemented with 10% fetal bovine serum, 100 U/mL penicillin and
100 µg/mL streptomycin. Electrophysiological recordings were per-
formed in the whole-cell configuration of the patch-clamp technique
using a DAGAN 8900 amplifier (Dagan Corp., Minneapolis, MN). The
patch electrodes, pulled from borosilicate class capillaries, had a
resistance of 3-4 MΩ when filled with internal solution containing
(in mM): CsCl, 110; tetraethylammonium chloride, 20; MgATP, 5;
EGTA, 14; HEPES, 20 and titrated to pH 7.4 with CsOH. Ap-
proximately 90% of electrode resistance was compensated electroni-
cally, so that effective series resistance was less than 1 MΩ. HEK cells
were used for experiments 2 to 4 days after plating the cells on the
cover slips. Generation of the voltage-clamp protocols and data
acquisition were carried out using PCLAMP software (Axon instru-
ments, Inc., Burlingame, CA). Sampling frequency was 0.5-2.0 kHz,
and current signals were filtered at 10 kHz before digitization and
storage. All experiments were performed at room temperature (23-25
°C). Cells selected for recordings had a capacitance of 25-35 pF. Cells
plated on plastic cover slips (15 mm round thermanox, Nunc, Inc.,
Napierville, IL) were transferred to an experimental chamber mounted
on the stage of an inverted microscope (Diaphot, Nikon, Nagano, Japan)
and were bathed in a solution containing (in mM): NaCl, 137; CaCl2,
2; KCl, 5.4; HEPES, 10; glucose, 10; MgCl2, 1 (pH 7.4 adjusted with
NaOH). The experimental chamber was constantly perfused at a rate
of about 1 mL/min with a control bathing solution. In solutions buffered
on the cell under recording KCl was omitted and in some experiments
NaCl was replaced by isoosmotic amounts of Na2SO4 or Na glutamate.
Servo-controlled miniature solenoid valves were used for rapid switch-
ing between control and test solutions (delay in solution change was
Unequally Loaded Vesicular Suspensions (intra- and extrave-
sicular buffers are different). The molality of the extravesicular buffer
was adjusted to that of intravesicular buffer by addition of a 1 M
solution of the appropriate salt in phosphate buffer (molality of solutions
was monitored by VPO). The pH of extravesicular buffer was adjusted
to that of intravesicular buffer by addition of 1 M HCl or 1 M NaOH
solution (monitored by pH meter).
Fluorimetric Transport Assays. pH-Assisted Transport Assays.
Typically, 100 µL of HPTS-loaded vesicles (stock solution) was
suspended in 1.85 mL of the corresponding buffer and placed into a
fluorimetric cell. HPTS emission at 510 nm was monitored with
excitation wavelengths at 403 and 460 nm simultaneously. During the
experiment, 20 µL of a 0-10 mM THF solution of the compound of
interest was added through an injection port, followed by injection of
21 µL of 0.5 M aqueous NaOH. Addition of the NaOH caused a pH
increase of approximately 1 pH unit in the extravesicular buffer.
Maximal changes in dye emission were obtained at the end of each
experiment by lysis of the liposomes with detergent (40 µL of 5%
aqueous Triton x100). The final transport trace was obtained as a ratio
of the emission intensities monitored at 460 and 403 nm and normalized
to 100% of transport. Pseudo-first-order rate constants were calculated
from slopes of the plot of ln([H+ins] - [H+out]) versus time, where [H+
]
ins
and [H+out] are the intravesicular and extravesicular proton concentra-
tions. The [H+out] was assumed to remain constant during the experi-
ment, while [H+ins] values were calculated for each point from the HPTS
emission intensities according to the calibration equation pH ) 1.1684
× log(Io/I1) + 6.9807 (Figure 5). The absolute values for rate constants
9
J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002 2277

A R T I C L E S
Sidorov et al.
< 20 ms). Compound 2 was dissolved in DMSO, and DMSO was added
in an equal amount to the control and test solutions.
thanks the Dreyfus Foundation for a Teacher-Scholar Award.
We thank Prof. Neil Blough (University of Maryland) for use
of his fluorimeter and Prof. Martin Morad (Georgetown
University) for use of the electrophysiological setup. We thank
Prof. Marco Colombini (University of Maryland) for access to
the voltage-clamp setup, helpful discussions, and advice regard-
ing the cell experiments. V.S. and J.D. thank Prof. Steven Rokita
and Prof. Lyle Isaacs (University of Maryland) and Dr. Bruce
Moyer (ORNL) for helpful comments.
Voltage-Clamp Experiments. Bilayer membranes were made from
monolayers of diphytanoylphospatidylcholine (DPhPC) using the
method of Montal and Mueller.39 Recordings were made under voltage-
clamp conditions using calomel electrodes with saturated KCl bridges.
Solutions consisted of 1 M KCl, 1 mM MgCl2, 5 mM HEPES, pH 7.0.
The voltage was applied from the trans side of the lipid membrane.
Once calixarene tetrabutylamide 2 was incorporated into the lipid, the
current through the membrane was recorded using both a chart recorder
and directly digitized and stored as a computer file. The conductance
values were calculated by dividing the current by the applied voltage.
The events were grouped into conductance ranges of equal magnitude
for the purpose of showing the distribution of recorded conductances.
Supporting Information Available: Crystallographic data
including diffractometer and refinement data, final coordinates,
bond lengths and angles, hydrogen bond lengths and angles,
and anisotropic displacement values (PDF). Crystallographic
data is also available in cif format. This material is available
free of charge at http://pubs.acs.org.
Acknowledgment. This work was supported by the National
Institutes of Health and the Department of Energy (J.D.). J.D.
(39) Montal, M.; Mueller, P. Proc. Natl. Acad. Sci. U.S.A. 1972, 69, 3561.
JA012338E
9
2278 J. AM. CHEM. SOC. VOL. 124, NO. 10, 2002