Catechols as membrane anion transporters

Full text of DOI:10.1021/ja809733c

Published on Web 02/03/2009
Catechols as Membrane Anion Transporters
Sofya Kostina Berezin and Jeffery T. Davis*
Department of Chemistry and Biochemistry, UniVersity of Maryland, College Park, Maryland 20742
Received December 15, 2008; E-mail: jdavis@umd.edu
Compounds that transport ions across cell membranes may
function as sensors, signal transducers, or antimicrobials. Synthetic
transporters may also help us understand how natural systems move
ions across hydrophobic barriers.1 There is an interest in identifying
compounds that transport anions, especially Cl-, across lipid
membranes.2 Reports that catechol 1 binds Cl- in organic solvents
prompted us to use this group to form new anion transporters.3,4
While bacterial siderophores and synthetic analogues use catechol-
ates to move Fe3+ across membranes,5 this paper is the first report,
to our knowledge, of catechols facilitating transmembrane transport
of anions.
intravesicular pH is likely caused by ligand-facilitated NO3- efflux
coupled with cotransport or diffusion of H+ out of the liposome.
We report that bis-catechol 3 is an anion transporter whose
activity depends on the catechol’s substitution and amphiphilicity.
We also describe a liposomal assay that allows one to readily
measure anion transport selectivity. This assay shows that transport
facilitated by 3 follows the Hofmeister sequence,6 wherein anions
that are easier to dehydrate are made more permeable to the
membrane by this bis-catechol.
Figure 1. NO3- transport assays: y-axis “Ratio” in (a-c) refers to the ratio
of HPTS fluorescence emission at 510 nm (I0/I1), where I0 is excitation at
460 nm (basic form of HPTS) and I1 is excitation at 403 nm (acidic form
of HPTS).8 Intravesicular 100 mM NaNO3, 10 mM phosphate buffer (pH
) 6.4), extravesicular 75 mM Na2SO4, 10 mM phosphate buffer (pH )
6.3), ratio of EYPC: 2-7 ) 10:1, solutions of amphiphiles 2-7 in MeOH
were injected at t ) 25 s, and aqueous 10% Triton-X was injected at t )
500 s. Data for (a) 2,3-catechol analogues 2-5, (b) 3 vs 6, and (c) 3 vs 7.
(d) Schematic showing the experimental assay.
Chart 1. Structures of Catechol 1 and Bis-catechols 2-7
Analogues 2-5 showed major differences in their ion transport
activity (Figure 1). Bis-catechol 3, with a medium-length alkyl
chain, was the most active of the 2,3-catechols, indicating that
transport activity depends on the compound’s ability to partition
into the membrane. Compound 5 with the longer C17 chain likely
aggregates in water, limiting its partitioning into the liposomes.
Compound 2, with the shortest chain, was the least active analogue.
Presumably, the most active analogue 3 does not aggregate too
much in water and is hydrophobic enough to partition into
liposomes. The bis-catechol’s alkyl chain length was not the only
structural determinant for transport function. Figure 1b/c shows that
the catechol’s substitution pattern is also crucial for ion transport.
Both 2,3-O-methyl analogue 6 and 3,4-substituted catechol 7 were
inactive, indicating that 2,3-OH diols are essential for membrane
transport activity. Simple changes in structure can clearly result in
profound differences in function.7
In the course of determining whether intra- and extravesicular
anions influence the membrane activity of 3, we devised a method
to measure the selectivity of ligand-mediated anion transport.9 This
assay relies on incorporating the salt of a weak acid, such as NaN3
(pKA of HN3 ) 4.7), into the liposomes.10 In the presence of a
transmembrane anion and/or pH differential, azide is protonated
and the resulting neutral hydrazoic acid readily diffuses out of the
vesicle. We then monitored intravesicular pH as a function of the
extravesicular anion after addition of transporter 3.
ESI-MS analysis showed that a dimer, 12•Cl-, was the major
complex formed when TBA+Cl- was mixed with excess catechol
(1) (Figure S2). Based on this finding we attached two catechols,
as either 2,3-dihydroxy or 3,4-dihydroxy benzoates, to a TREN
scaffold.7 An alkyl amide group was linked to TREN’s third
position. Syntheses of analogues 2-7 in Chart 1 are described in
the Supporting Information (SI). The membrane transport activity
of these amphiphiles was first evaluated using an assay with the
pH-sensitive HPTS dye.8 EYPC liposomes filled with NaNO3 and
HPTS were suspended in a buffer containing Na2SO4 as an external
electrolyte. Fluorescence measurements monitored changes in
intravesicular pH after adding analogues 2-7. Figure 1a indicates
that intravesicular alkalinization occurred upon addition of 2-5.
Because sulfate is poorly membrane permeable, the increase in
9
2458 J. AM. CHEM. SOC. 2009, 131, 2458–2459
10.1021/ja809733c CCC: $40.75
2009 American Chemical Society

C O M M U N I C A T I O N S
anion permeation in the CFTR chloride channel shows a weak
dependence on the anion’s hydration energy (Table 1).11,13 We
observed a nonlinear dependence of the rate constant for anion
transport on the concentration of transporter 3 (Figure S11). This
result indicated that anion transport became more efficient when
bis-catechol 3 self-associates in the membrane. We propose that
self-association of 3 provides pathways that increase anion perme-
ability across the bilayer without requiring complete dehydration
of the transported anion.
In conclusion, we have shown that anion permeation across a
phospholipid bilayer can be catalyzed by amphiphilic bis-catechols
such as 3. This anion transport process, which depends significantly
on both the catechol’s substitution pattern and its amphiphilicity,
follows a Hofmeister sequence that shows that permeability depends
on the ion’s hydration energy. Future efforts will include incorpo-
rating selectivity filters into these catechol transporters to help
overcome this Hofmeister bias in an effort to make Cl- selective
transporters.14 We also believe that the liposomal assay described
in Figure 2 will be useful for those interested in determining the
selectivity and mechanism for other synthetic and natural trans-
membrane ion transporters.
Figure 2. Anion transport assays. Solution of 3 in MeOH was injected at
t ) 25 s, ratio EYPC/3 ) 10:1, aqueous 10% Triton-X was injected at t )
-
1400 s. Extravesicular 100 mM Na+ A- (A- ) Cl-, Br-, NO3-, I-, ClO4
,
N3-) or 75 mM Na2SO4, 10 mM phosphate (pH ) 7.15); intravesicular
100 mM NaN3, 10 mM phosphate (pH ) 5.5).
When strong electrolytes were encapsulated within liposomes
no pH changes were observed in the absence of 3 because of the
poor membrane permeability of those ions. In contrast, when
liposomes loaded with NaN3 were exposed to a transmembrane
anion and pH gradient, the intravesicular pH immediately rose due
to outward diffusion of the neutral acid HN3 (the starting I0/I1 ratio
of 3.2 indicates an intravesicular pH near 8.5).8 Addition of
amphiphile 3 then enabled a compensating influx of extravesicular
anions back into the liposome. The consequent decrease in
intravesicular pH shown in Figure 2 reflects the receptor-facilitated
change in the chemical gradients of the N3- anion and the particular
extravesicular anion. The time dependence of this change in
intravesicular pH can be described by first-order kinetics, allowing
one to determine the rate constant for ligand-mediated anion influx,
kAnion, and the turnover number, nAnion, the number of anions
transported per liposome per second (Table 1).
Acknowledgment. We thank the U.S. Department of Energy
for support.
Supporting Information Available: Experimental details and
spectra. This material is available free of charge via the Internet at
http://pubs.acs.org.
References
(1) Reviews: (a) Fyles, T. M. Chem. Soc. ReV. 2007, 36, 335. (b) Sakai, N.;
Mareda, J.; Matile, S. Mol. BioSystems 2007, 3, 658.
(2) Review: Davis, A. P.; Sheppard, D. N.; Smith, B. D. Chem. Soc. ReV. 2007,
36, 348 Examples: (a) Schlesinger, P. H.; et al. J. Am. Chem. Soc. 2002,
124, 1848. (b) Koulov, A. V.; et al. Angew. Chem., Int. Ed. 2003, 42, 4931.
(c) Sessler, J. L.; et al. Angew. Chem., Int. Ed. 2005, 37, 5989. (d)
Santacroce, P. V.; et al. J. Am. Chem. Soc. 2007, 129, 1886. (e) Gale,
P. A.; Garric, J.; Light, M. E.; McNally, B. A.; Smith, B. D. Chem.
Commun. 2007, 1736.
Table 1. Anion Transport Rates (kAnion), Turnover Numbers (n),
and Differences in Activation Energy for Transmembrane
Transport by Bis-catechol 3 Relative to Cl- (∆∆Gq)a
(3) (a) Smith, D. K. Org. Biomol. Chem. 2003, 1, 3874. (b) Winstanley, K. S.;
Sayer, A. M.; Smith, D. K. Org. Biomol. Chem. 2006, 4, 1760.
(4) For catechol-like receptors that bind anions: Miyaji, H.; Sessler, J. L. Angew.
Chem., Int. Ed. 2001, 40, 154.
(5) (a) Raymond, K. N.; Dertz, E. A.; Kim, S. S. Proc. Natl. Acad. Sci. U.S.A.
2003, 100, 3584. (b) Apostol, M.; et al. Angew. Chem., Int. Ed. 2005, 44,
2580.
(6) Hofmeister, F. Arch. Exp. Pathol. Pharmacol. 1888, 24, 247.
(7) Smith, D. K. et al. found that some receptors containing two catechols are
better chloride binders than unfunctionalized catechol 1; see ref 3b.
(8) (a) Kano, K.; Fendler, J. H. Biochim. Biophys. Acta 1978, 509, 289. (b)
Sidorov, V.; Kotch, F. W.; Lam, Y.-F.; Kuebler, J. S.; Davis, J. T. J. Am.
Chem. Soc. 2003, 125, 2840. See Supporting Information for more
information on how the HPTS fluorescence ratio (I0/I1) relates to intrave-
sicular pH.
kAnion
s-1 × 103
1.29 ( 0.01
4.40 ( 0.07
11.3 ( 0.10
36.2 ( 0.80
69.5 ( 1.9
n,
∆∆Gq
∆∆Gq
∆∆Ghydr
CFTR
kJ·mol-1
Anion
s-1
kJ·mol-1
kJ·mol-1
Cl-
Br-
44
150
384
1230
2400
0
0
0.49
0.89
1.72
N/A
0
26
41
64
133
2.9
5.3
8.1
9.7
-
NO3
I-
-
ClO4
a The last two columns in the table list differences in activation
energy for anion permeation relative to Cl- in the CFTR channel
(∆∆Gq
values from ref 11) and differences in anion hydration
CFTR
energy relative to Cl- (∆∆Ghydr values from ref 12). See SI for more
detailed explanation of the ∆∆G values.
(9) For another approach toward determining anion transport selectivity by
using the HPTS dye, see: Gorteau, V.; Bollot, G.; Mareda, J.; Matile, S.
Org. Biomol. Chem. 2007, 5, 3000.
(10) The use of NaN3 as a protonophore to induce pH gradients across cell
membranes has been described: (a) Garland, P. B.; Downie, J. A.; Haddock,
B. A. Biochem. J. 1975, 152, 547. (b) Reid, R. J.; Loughman, B. C.;
Ratcliffe, R. G. J. Exp. Botany 1985, 36, 889.
Table 1 shows that the selectivity of ionic permeability across
the membrane in the presence of 3 follows a Hofmeister sequence
with kAnion decreasing in the order ClO4 > I- > NO3 > Br-
>
-
-
(11) Linsdell, P.; et al. J. Gen. Physiol. 1997, 110, 355.
Cl-. This weak dependence of transport rates on the anion’s
hydration energy indicates that the anions only need to be partly
dehydrated to pass across the membrane in the presence of bis-
catechol 3. This selectivity pattern for the synthetic transporter 3,
where anion transport rates correlate with the anion’s hydration
energy, is also seen for some of Nature’s anion transporters. Thus,
(12) Markus, Y. Ion Properties; Marcel Dekker, New York, 1997; p 56.
(13) Smith, S. S.; Steinle, E. D.; Meyerhoff, M. E.; Dawson, D. C. J. Gen.
Phyiol. 1999, 114, 799.
(14) For recent examples of overcoming the Hofmeister bias, see: (a) Fowler,
C. J.; et al. J. Am. Chem. Soc. 2008, 130, 14386. (b) Sisson, A. L.; Clare,
J. P.; Davis, A. P. Chem. Commun. 2005, 5263.
JA809733C
9
J. AM. CHEM. SOC. VOL. 131, NO. 7, 2009 2459