Selective monosaccharide transport through lipid bilayers using boronic acid carriers

Article abstract of DOI:10.1021/ja961264h

Twenty-one boronic acids were studied for their ability to transport saccharides in and out of liposomes. The rates of liposome efflux were determined using an enzymatic assay, whereas the influx studies used a radiotracer method. All boronic acids examined, except those that were highly hydrophilic, facilitated monosaccharide transport. The order of transport selectivity was sorbitol > fructose > glucose. The disaccharides maltose and sucrose were not transported to any significant degree. Facilitated transport was demonstrated with a variety of liposome types, including multilamellar and unilamellar vesicles with anionic or cationic polar lipid additives. Transport mechanism studies included the accumulation of structure-activity data, as well as systematic investigations of various environmental changes such as pH, added salt, membrane potential, and temperature. Overall, the evidence is strongly in favor of a membrane carrier mechanism. The boronic acid combines reversibly with a diol group on the monosaccharide to produce a tetrahedral, anionic boronate, which is the major complexed structure in bulk, aqueous solution. At the bilayer surface, the tetrahedral boronate is in equilibrium with its neutral, trigonal form, which is the actual transported species. At low carrier concentrations, a first-order dependence on carrier was observed indicating that the transported species was a 1:1 sugar-boronate. At higher carrier concentrations the kinetic order approached 2, suggesting the increased participation of a 1:2 sugar-bisboronate transport pathway. The effect of boronic acids on liposomal bilayer fluidity was probed by fluorescence spectroscopy using appropriate reporter molecules. Adding cholesterol to the liposome membranes reduced translational fluidity by 'packing and ordering' the bilayer. Addition of lipophilic arylboronic acids (either free or complexed with monosaccharides) induced a similar but smaller effect.

Full text of DOI:10.1021/ja961264h

J. Am. Chem. Soc. 1996, 118, 11093-11100
11093
Selective Monosaccharide Transport through Lipid Bilayers
Using Boronic Acid Carriers
Pamela R. Westmark, Stephen J. Gardiner, and Bradley D. Smith*
Contribution from the Department of Chemistry and Biochemistry, UniVersity of Notre Dame,
Notre Dame, Indiana 46556
ReceiVed April 17, 1996X
Abstract: Twenty-one boronic acids were studied for their ability to transport saccharides in and out of liposomes.
The rates of liposome efflux were determined using an enzymatic assay, whereas the influx studies used a radiotracer
method. All boronic acids examined, except those that were highly hydrophilic, facilitated monosaccharide transport.
The order of transport selectivity was sorbitol > fructose > glucose. The disaccharides maltose and sucrose were
not transported to any significant degree. Facilitated transport was demonstrated with a variety of liposome types,
including multilamellar and unilamellar vesicles with anionic or cationic polar lipid additives. Transport mechanism
studies included the accumulation of structure-activity data, as well as systematic investigations of various
environmental changes such as pH, added salt, membrane potential, and temperature. Overall, the evidence is strongly
in favor of a membrane carrier mechanism. The boronic acid combines reversibly with a diol group on the
monosaccharide to produce a tetrahedral, anionic boronate, which is the major complexed structure in bulk, aqueous
solution. At the bilayer surface, the tetrahedral boronate is in equilibrium with its neutral, trigonal form, which is
the actual transported species. At low carrier concentrations, a first-order dependence on carrier was observed
indicating that the transported species was a 1:1 sugar-boronate. At higher carrier concentrations the kinetic order
approached 2, suggesting the increased participation of a 1:2 sugar-bisboronate transport pathway. The effect of
boronic acids on liposomal bilayer fluidity was probed by fluorescence spectroscopy using appropriate reporter
molecules. Adding cholesterol to the liposome membranes reduced translational fluidity by “packing and ordering”
the bilayer. Addition of lipophilic arylboronic acids (either free or complexed with monosaccharides) induced a
similar but smaller effect.
Introduction
Artificial sugar receptors can be divided into two groups,
those that mimic natural binding systems in the sense that they
utilize noncovalent bonding interactions such as hydrogen
bonding and those that use non-natural bonding interactions.5
If the goal is to develop an artificial receptor that operates in
water, then an approach based primarily on hydrogen bonding
is likely to be very challenging due to competitive bonding with
the bulk water. As a consequence, a number of groups have
chosen to pursue systems that use non-natural bonding inter-
actions. Boronic acids are of particular interest as they form
reversible covalent complexes with sugars in aqueous solution.6-10
Association constants are typically between 10 and 104 M-1. A
range of boronic acid receptors have been developed with the
capability of molecular functions such as selective binding,6
allosteric binding,7 reaction catalysis,8 chemosensing,7,9 and
membrane transport.10 This report focuses on membrane
The selective transport of saccharides, such as glucose, across
cell membranes is a ubiquitous cellular activity. A typical
human red cell, for example, contains around 200 000 glucose
transport molecules.1 It is well recognized that defects in these
transport processes can lead to serious medical problems.2 As
a consequence, the mechanisms of biotic sugar transport systems
are currently under active investigation.3 While much structural
data has been accumulated, the kinetic picture is still largely
unknown.4 A recent review concluded that, “Virtually nothing
is known at present about the molecular mechanism of glucose
transport or the transport of any solute across the membrane”.3a
As part of an interdisciplinary research effort to understand
carbohydrate recognition, chemists have started designing
artificial sugar receptors. There are two primary goals for this
work. The first is to treat the artificial systems as simplified,
small molecular weight models whose properties can be probed
in a systematic manner. This provides an experimentally
accessible way of exploring some of the fundamental questions
in sugar recognition. Alternatively, there is the possibility of
developing the artificial sugar binders into practical molecular
devices.
(5) Recent reports of artificial sugar receptors that use noncovalent
interactions: (a) Cotero´n, J. M.; Hacket, F.; Schneider, H.-J. J. Org. Chem.
1996, 61, 1429-1435. (b) Jime´nez-Barbero, J.; Junquera, E.; Mart´ın-Pastor,
M.; Sharma, S.; Vicent, C.; Penade´s, S. J. Am. Chem. Soc. 1995, 117,
11198-11204. (c) Bonar-Law, R. P.; Sanders, J. K. M. J. Am. Chem. Soc.
1995, 117, 259-271. (d) Anderson, S.; Neidlein, U.; Gramlich, V.;
Diederich, F. Angew. Chem., Int. Ed. Engl. 1995, 34, 1596-1600. (e)
Inouye, M.; Miyake, T.; Furusyo, M.; Nakazumi, H. J. Am. Chem. Soc.
1995, 117, 12416-12425.
X Abstract published in AdVance ACS Abstracts, November 1, 1996.
(1) Widdas, W. F. Biochim. Biophys. Acta 1988, 947, 385-404.
(2) White, M. K.; McCubrey, J. A. Int. J. Oncol. 1995, 7, 701-712.
(3) (a) Mueckler, M. Eur. J. Biochem. 1994, 219, 713-725. (b) Baldwin,
S. A. In Molecular Aspects of Transport Proteins; De Pont, J., Ed.;
Elsevier: New York, 1992; Chapter 6. (c) Quiocho, F. A. Curr. Opin. Struct.
Biol. 1991, 1, 922-933.
(4) (a) Schirmer, T.; Keller, T. A.; Wang, Y.-F.; Rosenbusch, J. P. Science
1995, 267, 512-514. (b) Vyas, M. N.; Vyas, N. K.; Quiocho, F. A.
Biochemistry 1994, 33, 4762-4768.
(6) Wulff, G. Pure Appl. Chem. 1982, 54, 2093-3000.
(7) James, T. D.; Samankumara Sandanayake, K. R. A.; Shinkai, S.
Supramol. Chem. 1995, 6, 141-157.
(8) Rao, G.; Phillip, M. J. Org. Chem. 1991, 56, 1505-1512.
(9) Czarnik, A. W. Acc. Chem. Res. 1994, 27, 302-308.
(10) (a) Smith, B. D. Supramol. Chem. 1996, 7, 55-60 and references
cited therein. (b) Smith, B. D. In Chemical Separations with Liquid
Membranes; Bartsch, R. A., Way, J. D., Eds.; ACS Symposium Series
Number 642; American Chemical Society: Washington, DC, 1996.
S0002-7863(96)01264-4 CCC: $12.00 © 1996 American Chemical Society

11094 J. Am. Chem. Soc., Vol. 118, No. 45, 1996
Westmark et al.
Figure 1. Sugar transport through a lipid bilayer mediated by a boronic
acid carrier; S ) sugar, B ) boronic acid, SB ) sugar-boronate
complex.
Figure 2. Amount of glucose transported as a function of time. (a)
Glucose influx from external solution containing 50 mM glucose into
empty liposomes as monitored by radiotracer assay. Transport (at pH
7.5) induced by carriers: 2 (3, 2 mM); 4 (1, 5 mM); 0 (4, 5 mM); b
(8, 5 mM). Liposome composition DPPC:C:PA, 20:15:2; total lipid
concentration 9.4 mM, T ) 25 °C. (b) Glucose efflux from liposomes
containing 300 mM glucose as monitored by enzyme assay. Transport
(at pH 7.5) induced by carriers: 2 (3, 1 mM); 4 (1, 5 mM); 0 (4, 5
mM); b (8, 5 mM). Liposome composition DPPC:C:PA, 20:15:2; total
lipid concentration was 0.75 mM, T ) 25 °C, PL ) phospholipid,
reproducibility (10%.
transport, where boronic acids are one of only two classes of
artificial compounds that are known to transport sugar deriva-
tives through bilayer membranes.11 Over the past few years,
our group, along with others, has been examining facilitated
membrane transport using boronic acid carriers.10 Initially, the
study focused on ways to transport hydrophilic diol-containing
compounds through bulk, liquid organic membranes and more
recently through polymer-supported liquid membranes.12 We
also examined transport through lipid bilayer membranes and
discovered that boronic acids can facilitate the efflux of glucose
and ribonucleosides from liposomes.13,14 In this paper, we
describe in full detail our efforts to facilitate the transport of
glucose and other saccharides. In short, we have found that
boronic acids can selectively transport monosaccarides in and
out of liposomes. In addition, we have elucidated the transport
mechanism.
direct measure of the rate of sugar efflux since the enzymes
were unable to penetrate the liposomes, and the rates of sugar
efflux were slower than the enzymes turnover rates. This proved
to be a straightforward way of monitoring sugar efflux in real
time. However, there were a few drawbacks, namely, that some
boronic acids inhibited the enzymatic assays making them
prohibitively slow, and the low sensitivity required that un-
naturally high concentrations of sugar be encapsulated inside
the liposomes.
The enzyme assays were not practical for the influx experi-
ments, therefore a radiotracer method was developed. Empty
liposomes with a trace of 14C-labeled DPPC incorporated in
the phospholipid membrane were incubated with a sugar solution
Transport Assays
A schematic description of monosaccharide transport through
a lipid bilayer mediated by a boronic acid carrier is shown in
Figure 1. Both liposome efflux and liposome influx experiments
were conducted. Usually a lipid mixture of dipalmitoylphos-
phatidylcholine (DPPC), cholesterol (C), and egg phosphatidic
acid (PA) was used in a ratio of 20:15:2.15 Compared to
unsaturated phospholipids, saturated DPPC forms liposomes that
are less permeable to small hydrophilic molecules.16 In addition,
the presence of large amounts of cholesterol within the
membrane is also known to decrease permeability.16 Thus, we
expected the liposomes used in this study to be quite stable
and relatively impermeable.
Sugar-encapsulated unilamellar vesicles (80 nm diameter)
were prepared by the rapid extrusion technique and found to
be essentially impermeable to sugar leakage over many hours.17
Sugar escape from the liposomes was detected spectrometrically
using enzymatic assays which produced an NADPH absorption
at 340 nm. Generally, the rate of NADPH production was a
3
that contained trace amounts of H-labeled sugar and boronic
acid carriers. Every 10 min, an aliquot of the mixture was
withdrawn and passed through a column of Sephadex gel which
separated the liposomal component from the external solu-
3
tion. The ratio of H to 14C associated with the liposomal
fraction was determined from scintillation counting and con-
sidered as a measure of the amount of sugar transported into
the liposomes.
A significant difference between these two transport assays
is that the efflux experiment results in uphill transport because
of the destructive nature of the enzyme assay, whereas the influx
experiment simply monitors passive equilibration. Thus, efflux
continues until the liposomes are effectively empty of saccha-
ride. Influx, on the other hand, reaches a steady-state position
where the electrochemical potential inside the liposome equals
that outside (in the absence of a secondary energy source, such
as a pH gradient, the final internal and external saccharide
concentrations will be equal).
(11) For a remarkable example of artificial channel-mediated glucose
transport, see: Granja, J. R.; Ghadiri, M. R. J. Am. Chem. Soc. 1994, 116,
10785-10786.
(12) Paugam, M.-F.; Bien, J. T.; Smith, B. D.; Chrisstoffels, S.; de Jong,
F.; Reinhoudt, D. N. J. Am. Chem. Soc. In press.
(13) Westmark, P. R.; Smith, B. D. J. Am. Chem. Soc. 1994, 116, 9343-
9344.
(14) Westmark, P. R.; Smith, B. D. J. Pharm. Sci. 1996, 75, 266-269.
(15) For discussions on the properties of the bilayer membranes formed
by the lipid composition used in this work, see: (a) McMullen, T. P. W.;
McElhaney, R. N. Biochim. Biophys. Acta 1995, 1234, 90-98. (b) Murari,
R.; Murari, M.; Baumann, W. J. Biochemistry 1986, 25, 1062-1067. (c)
Mayer, L. D.; Bally, M. B.; Hope, M. J.; Cullis, P. R. Biochim. Biophys.
Acta 1985, 816, 294-302B.
(16) (a) Liposomes, a Practical Approach; New, R. R. C., Ed.; IRL
Press: Oxford, 1990. (b) Lasic, D. D. Liposomes; From Physics to
Applications; Elsevier: Amsterdam, 1993.
Carrier Effectiveness
The discussion on carrier effectiveness will focus on glucose
transport in and out of liposomes; however, it should be noted
that similar results were obtained with other monosaccharides
(e.g., fructose, sorbitol). The efflux experiments were conducted
with a wide range of boronic acids, whereas the influx
experiments surveyed a smaller number of carriers. In both
cases, the qualitative order for transport enhancement was the
same (Figure 2). Examination of the glucose efflux data for
carriers 1-20, shown in Table 1, reveals a general pattern that
initial transport rates increased as the acidity and lipophilicity
(17) MacDonald, R. C.; MacDonald, R. I.; Menco, B. P. M.; Takeshita,
K.; Subbarao, N. K.; Hu, L. Biochim. Biophys. Acta 1991, 1061, 297-303.

Monosaccharide Transport through Lipid Bilayers
J. Am. Chem. Soc., Vol. 118, No. 45, 1996 11095
Table 1. Initial Rates of Glucose Efflux at pH 7.5, pKa Values,
and Substituent Hydrophobicity Constants For Various Boronic
Acid Derivatives
init rate
of effluxa
sum of
boronic acid, R-B(OH)2
phenylboronic acid, 1
(3,5-dichlorophenyl)boronic acid, 2
(rel rate) pKa π valuesb
0.4 (1)
6 (15)
8.9c
7.4d
7.2d
0.0
1.42
1.76
[3,5-bis(trifluoromethyl)phenyl]boronic 12 (30)
acid, 3
diphenylborinic acid, 4
(4-methoxyphenyl)boronic acid, 5
(3-methoxyphenyl)boronic acid, 6
(4-tert-butylphenyl)boronic acid, 7
(4-carboxyphenyl)boronic acid, 8
1-butylboronic acid, 9
<10-3
6.2d
2.64
0.3 (0.8) 9.3c
0.3 (0.8) 8.7d
-0.02
-0.02
1.98
9.3d
8.4e
Figure 3. Efflux profiles for liposomes containing 300 mM of 2
sorbitol, 0 fructose, or 9 glucose after treatment with phenylboronic
acid (1, 1 mM) at pH 7.5. Liposome composition DPPC:C:PA, 20:15:
2; total lipid concentration 0.75 mM, T ) 25 °C, PL ) phospholipid,
reproducibility (10%.
12 (30)
<10-3
-4.36
1.2 (0.3) 10.4f
boric acid, 10
<10-3
9.0f
(4-bromophenyl)boronic acid, 11
(4-fluorophenyl)boronic acid, 12
(2-methylphenyl)boronic acid, 13
(3-methylphenyl)boronic acid, 14
(4-methylphenyl)boronic acid, 15
(3,5-dimethylphenyl)boronic acid, 16
3-pyridylboronic acid, 17
18
2.8 (7)
8.6d
0.86
0.14
0.56
0.56
0.56
1.12
0.5 (1.3) 9.1d
0.3 (0.8) 9.7c
0.5 (1.4) 9.0c
0.5 (1.4) 9.3c
. maltose g sucrose (Figure 3).19 This transport order matches
the known order of sugar-boronate complex stabilities (boronic
acids favor complexation with compounds containing vicinal
cis-diols) and is strong evidence in favor of a complexation
mediated transport phenomenon.7,10 To demonstrate the extent
of the transport selectivity, competitive influx and efflux
experiments were conducted.
1 (2.8)
9.1d
4.0g
<10-3
1.5 (3.8)
13 (30)
25 (60)
19
20
a Values in nmol of glucose/µmol of PL/min ((15%) induced by 1
mM of boronic acid, PL ) phospholipid; total lipid concentration was
0.75 mM. b Sum of the hydrophobicity constants for all aryl substituents
except the boronic acid (ref 18a). c Reference 18b. d Determined by the
method described in ref 18c. e Reference 18d. f Reference 18e. g Ref-
erence 18f.
The glucose influx experiment described in Figure 2a was
repeated except that the “empty” liposomes were incubated with
solutions containing equimolar amounts of glucose and sucrose.
Each transport experiment involved two concurrent incubations;
3
in one case the glucose included trace H-glucose and in the
3
other case the sucrose contained trace H-sucrose. When the
of the boronic acids increased.18 The dependence on boronic
lipophilicity is most apparent when the fluxes for carriers 18-
20 are compared. The carriers with more lipophilic ester groups
liposome/glucose/sucrose incubation mixture was treated with
boronic acid 1 the amount of glucose associated with the
liposomes increased over time in a similar fashion to that shown
in Figure 2a. On the other hand, negligible sucrose was
incorporated into the liposomes (data not shown). When the
nontransporting boronic acid 8 was used, or when boronic acid
was omitted, there was no evidence of glucose or sucrose influx.
These results are consistent with the selective delivery of glucose
(contains a vicinal cis-diol) in the presence of weakly com-
plexing sucrose (no vicinal cis-diol).
Efflux experiments were conducted with liposomes that
encapsulated equal amounts of sorbitol and sucrose. As shown
in Figure 4, addition of boronic acid 7 to the liposome mixture
resulted in greatly enhanced sorbitol efflux, whereas the sucrose
remained untouched (i.e., effectively complete transport selec-
tivity in favor of sorbitol). This efflux result suggests that
boronic acids may have hitherto unconsidered therapeutic
properties. For example, boronic acids provide a conceptually
interesting approach to treating polyol (e.g., sorbitol) accumula-
tion in ocular tissues; a problem that is associated with various
types of diabetic complications.20 Our results suggest that
boronic acids may be able to relieve polyol buildup by
selectively facilitating cellular efflux.
were clearly superior transport agents. The dependence on
carrier acidity was harder to demonstrate unambiguously because
of the inherent difficulty in delineating the dependence on
lipophilicity. Nonetheless, a qualitative survey of Table 1 shows
that the more acidic carriers were generally more effective.
Perhaps the clearest indication of the importance of carrier Lewis
acidity is gained by comparing the transport rates for the equally
lipophilic tolyl isomers 13-15. The sterically encumbered ortho
derivative 13 had a little over half the transport ability of the
meta and para isomers 14 and 15, respectively. This is
consistent with a model where complexation between the sugar
and the boronic acid is a prerequisite for transport.
Transport Mechanism
Transport Selectivity
After successful membrane transport was demonstrated, the
first mechanistic question was to prove it was due to a specific
sugar complexation event and not to a boronic acid-induced
A series of noncompetitive efflux experiments showed a sugar
transport enhancement order of sorbitol > fructose > glucose
(18) (a) Chu, K. C. In Burger’s Medicinal Chemistry, Part I. The Basis
of Medicinal Chemistry, 4th ed.; Wolff, M. E., Ed.; Wiley: New York,
1979; p 401. (b) Torsell, K. In Progress in Boron Chemistry; Steinberg,
H., McCloskey, A. L., Eds.; Pergamon: New York, 1964; p 385. (c) Albert,
A.; Serjent, E. P. The Determination of Ionization Constants, 3rd ed.;
Chapman and Hill: London, 1984. (d) Soundararajan, S.; Badawi, M.;
Kohlrust, C. M.; Hageman, J. H. Anal. Biochem. 1989, 178, 125-134. (e)
Babcock, L.; Pizer, R. Inorg. Chem. 1980, 19, 56-61. (f) Fischer, F. C.;
Havinga, E. Recl. TraV. Chim. Pays.-Bas. 1974, 93, 21-25.
(19) Since boronic acids are also known to form covalent complexes
with R-hydroxy acids, we searched for evidence of facilitated transport.
Using the UV absorption assay that was developed for nucleoside transport,14
we found that boronic acids induced large enhancements in the rates of
efflux and influx of (p-nitrophenyl)glucurinide (data not shown). Therefore,
it appears that in addition to monosaccharides and ribonucleosides boronic
acids can also be used to transport certain types of R-hydroxy acids.
(20) (a) Burg, M. B.; Kador, P. F. J. Clin. InVest. 1988, 81, 635-640.
(b) Kicic, E.; Palmer, T. N. Biochim. Biophys. Acta 1994, 1226, 213-218.

11096 J. Am. Chem. Soc., Vol. 118, No. 45, 1996
Westmark et al.
Figure 4. (a) Sorbitol efflux from liposomes filled with 150 mM
sorbitol and 150 mM sucrose after treatment with 9 (7, 1 mM), or 0
(no boronic acid). (b) Sucrose efflux from the same liposome prepara-
tion after treatment with 9 (7, 1 mM) and then eventually lysing with
Triton-X. Liposome composition DPPC:C:PA, 20:15:2; total lipid
concentration 0.75 mM, pH ) 7.5, T ) 25 °C, PL ) phospholipid,
reproducibility (10%.
Figure 5. Glucose (50 mM) influx induced by 1 (1 mM) but inhibited
by competitive-binding sugars (50 mM): 0 (no sugar); b (sucrose);
× (gluconic acid). Radiotracer assay using liposomes of composition
DPPC:C:PA, 20:15:2; total lipid concentration 9.4 mM, pH ) 7.5, T
) 25 °C, PL ) phospholipid, error bars reflect the maximum deviation
over three experiments.
increase in bilayer permeability. Our initial approach was to
examine the efflux of various marker compounds that were not
expected to be transported under the conditions used to induce
monosaccharide transport. In addition to the negligible sucrose
transport already discussed, boronic acids 1-4 were unable to
induce the efflux of anionic monosaccharide derivatives iso-
citrate, glucose-6-phosphate, and fructose-6-phosphate (as proved
by enzymatic assays), as well as the anionic carboxyfluoroscein,
calcein (fluorescence assay), and arsenazo III (spectrometric
assay).16a The structure of the carrier was another variable that
was changed in order to gain mechanistic insight. For example,
neither excess boric acid nor the extremely hydrophilic boronic
acid 8 were capable of inducing monosaccharide influx or efflux.
Overall, the data are strongly in favor of a transport carrier
mechanism mediated by reversible sugar-boronate complex-
ation.
conjugate base 23, shown in eq 2 (which is perhaps part of an
ion pair)? Indirect evidence was gained by conducting transport
experiments under conditions that were expected to facilitate
one of the transport pathways. It was felt that this would be an
informative approach because one transport mechanism forms
a neutral intermediate (22), whereas the alternative involves an
ionic species (23).
A more difficult mechanistic issue concerned the molecular
details of the transport phenomenon. The problem was ad-
dressed by accumulating a range of kinetic and structure-
activity evidence. The kinetic order was determined for glucose
efflux and influx using carriers 1, 3, and 9. At low carrier
concentrations, a first-order dependence was observed, indicating
that the transported species had a 1:1 stoichiometry. At higher
carrier concentrations the carrier kinetic order approached 2,
suggesting the increased participation of a 1:2 glucose-
bisboronate pathway.21 Since transport systems are kinetically
related to enzyme reactions it seemed appropriate to gather
inhibition data.22 In this case, efflux and influx rates were
determined in the presence of “inhibitors” (highly hydrophilic
boronate binding compounds such as anionic sugars that
compete for the boronate and lower its effective concentration).
These inhibition studies were again in favor of a complexation-
mediated transport event. For example, glucose influx rates
were determined using carrier 1 in the presence of gluconic acid
or sucrose (Figure 5). The sucrose was found to have negligible
inhibitory effect on facilitated glucose transport, whereas the
gluconic acid was a strong transport inhibitor. This is in
agreement with the poor boronate complexation ability of
sucrose and the strong boronate complexation ability of gluconic
acid (contains several acyclic vicinal diols).10
First, transport was examined as a function of pH and added
counterion. The effect of pH was probed by determining the
rates of efflux and influx for samples where the pH of the
internal solution was held constant at pH 7.5 and the external
solution pH was either 6.5, 7.5, or 8.5 (control fluorescence
experiments using encapsulated carboxyfluoroscein showed
essentially no decay of the pH gradient over the time frame of
the experiment). The resulting pH effect was dependent on the
identity of the boronic acid. Rates of efflux and influx were
generally a maximum at a pH value just below the pKa of the
boronic acid. This is evidence against 23 as the transported
species, because the amount of tetrahedral complex should
increase with pH. Transport rates were found to be rather
insensitive to cation or anion lipophilicity. Addition of 5 mM
sodium perchlorate or tetramethylammonium chloride to the
external solution had essentially no effect on transport rates
induced by carrier 1. The more lipophilic tetrabutylammonium
chloride had no effect on influx and increased efflux 2- to 4-fold.
Various neutral ionophores were examined for their abilities to
affect the rates of efflux and influx. Both natural ionophores
such valinomycin and gramicidin, as well as artificial ionophores
such as 18-crown-6 and [2.2.2]-cryptand, were examined with
salts such as sodium or potassium chloride (75 mM) included
inside or outside the liposomes. At low ionophore concentra-
tions (1 mM for the artificial ionophores and 1 µM for natural
At the molecular level, the bilayer transport mechanism
reduces to two limiting possibilities. Is the transported species
the neutral trigonal boronate ester 22, shown in eq 1, or the
(21) The association constants for glucose and phenylboronic acid to
give 1:1 and 1:2 tetrahedral boronates in aqueous solution are K1:1 ) 110
M-1 and K1:2 ) 770 M-2. Lorand, J. P.; Edwards, J. O. J. Org. Chem.
1959, 24, 769-774.
(22) Stein, W. D. Transport and Diffusion Across Cell Mebranes;
Academic: San Diego, CA, 1986.

Monosaccharide Transport through Lipid Bilayers
J. Am. Chem. Soc., Vol. 118, No. 45, 1996 11097
A comparative survey was made of transport with other types
of liposomes. Liposomes were prepared by methods known to
produce multilamellar vesicles (MLVs), large unilamellar
vesicles (LUVs), and small unilamellar vesicles (SUVs).16
A
full characterization of each vesicle system was not possible,
so only the glucose efflux per phospholipid content was
determined. The observed order of glucose efflux rates from
the different liposomes in the presence of 1 was LUV > SUV
> MLV, which matches the literature order of relative liposome
permeability.16
The effect of membrane surface charge was examined.
Changing the anionic lipid component (DPPC/C/x, 20:15:2,
where x ) phosphatidic acid) to a cationic additive such as
(S)-(2,3-dioleoxypropyl)trimethylammonium chloride (DOTAP),
stearyltributylphosphonium bromide, or dodecyltrimethyl-
ammonium bromide produced little change in glucose efflux
rates, suggesting that the transported species is insensitive to
membrane charge (i.e., the transported species is the neutral
trigonal boronate ester 22).
Figure 6. Covalent complexation between a diol and 4 in aqueous
solution only producing the chelated tetrahedral borinate anion 24 and
not the monodentate trigonal ester 25.
ionophores), there was little change in rates of influx or efflux.
At higher ionophore concentrations (5 mM for the artificial
ionophores and 10 µM for natural ionophores), the rates did
increase by a factor of 2-4. These results are interpreted as
evidence against 23 as a major transport species, because an
anionic boronate would be expected to be far more sensitive to
changes in cation lipophilicity and/or membrane potential.
Various carriers were examined that had structural features
that should have facilitated the formation and transport of
anionic boronate 23. The highly acidic 3-pyridylboronic acid
(17), which exists as a zwitterion at neutral pH18f and the cationic
N-methylnicotinamide derivative (21) which supplies a co-
valently attached cation to counter a putative anionic boronate,
were both unable to induce glucose influx. In addition,
diphenylborinic acid (4) was also incapable of facilitating
glucose influx or efflux. The same negative result was obtained
with ribonucleoside transport.14 Because 4 contains two non-
labile B-C bonds and is relatively acidic (pKa ) 6.2), it can
only form the anionic tetrahedral complex (24) when it combines
with a diol in aqueous solution (Figure 6). The monodentate
borinate ester 25 does not occur to any great extent in water.
Previous work with liquid organic membranes has shown that
under conditions where an ion pair transport mechanism is
operating, the highly lipophilic 24 is readily transported.23
Taken together, these negative results are strong indirect
evidence against 23 as the transported species.
Membrane Fluidity
The final issue concerned the effect of the boronic acids on
the bilayer structure. Did the boronic acids alter the bilayer
fluidity? Fluorescence spectroscopy using appropriate probe
molecules is a versatile means of examining the fluidity of a
bilayer membrane.25 Time-averaged fluorescence experiments
were conducted using two well-known fluorescent probes.
Rotational fluidity was examined using 1-[(4-trimethyl-
ammonium)phenyl]-6-phenyl-1,3,5-hexatriene 4-toluenesulfonate
(TMA-DPH), and translational fluidity was investigated using
1-hexadecanoyl-2-(1-pyrenedecanoyl)-sn-glycerol-3-phospha-
tidylcholine (PyDPPC).25 TMA-DPH intercalates into bilayer
membranes in a parallel orientation relative to the other polar
lipids with its cationic ammonium group in the polar head group
region. The degree in which its rotational freedom is hindered
is reflected by the magnitude of its steady-state anisotropy. A
decrease in anisotropy reflects an increase in rotational fluidity.26
With PyDPPC, the ratio of excimer to monomer emission (I ′/
I) is known to increase as membrane translational fluidity
increases.27
Fluorescence spectroscopy was used to characterize the
membrane fluidity of liposomes that did not contain cholesterol
(DPPC/C/PA, 20:0:2). Both probes showed clear evidence for
a sharp gel to liquid phase transition around 42 °C. As expected,
incorporating >30% cholesterol into the liposome membranes
(DPPC/C/PA, 20:15:2) eliminated this phase transition.27 At
25 °C, the TMA-DPH anisotropy was essentially unchanged
upon cholesterol addition, whereas the I '/I ratio for PyDPPC
was significantly lower. The cholesterol had reduced transla-
tional fluidity by “packing” the membrane.
The following experiments were performed at 25 °C. Lipo-
somes containing either TMA-DHP or PyDPPC were added to
solutions of glucose (100 mM in pH 7.5 buffer), gluconic acid
(100 mM in pH 7.5 buffer), or pH 7.5 buffer alone. To each
solution was then added boronic acid 1 in incremental amounts
up to 5 mM. The change in probe response was monitored as
Transport as a Function of Membrane Composition
Most transport experiments were conducted with a lipid
mixture of DPPC/C/PA in a ratio of 20:15:2; however, other
compositions were examined. Replacing the DPPC with
unsaturated lecithin resulted in increased membrane fluidity and
a consequent increase in permeability (both background leakage
and facilitated transport rates went up). When the amount of
cholesterol was systematically varied (DPPC/C/PA, 20:x:2), the
transport flux did not show a linear change. Additions up to
10% cholesterol produced slight increases in transport rates,
which then dipped when the cholesterol levels reached 20%,
only to increase again when the cholesterol level reached 40%.24
Increasing the temperature from 22 to 50 °C resulted in a five-
fold increase in glucose influx (liposomes contained DPPC/C/
PA, 20:15:2). This correlates with an increase in membrane
fluidity over this temperature range (there is no gel to liquid
phase transition for this bilayer composition).15
(25) (a) Slavik, J. Fluorescent Probes in Cellular and Molecular Biology;
CRC: Boca Raton, FL, 1994. (b) Haughland, R. P. Molecular Probes
Handbook of Fluorescent Probes and Research Chemicals; Molecular
Probes: Eugene, OR, 1992.
(26) Prendergast, F. G.; Haugland, R. P.; Callahan, P. J. Biochemistry
1981, 20, 7333-7337.
(27) (a) Galla, H.-G.; Hartman, W. Chem. Phys. Lipids 1980, 27, 199-
219. (b) Mulders, F.; Van Langen, H.; Van Ginkel, G.; Levine, Y. K.
Biochim. Biophys. Acta 1986, 859, 209-218.
(23) Morin, G. T.; Hughes, M. P.; Paugam, M.-F.; Smith, B. D. J. Am.
Chem. Soc. 1994, 116, 8895-8901.
(24) Cholesterol does not “pack” a bilayer in a uniform manner. At
intermediate cholesterol levels, inhomogeneous pockets are formed, which
may explain the nonlinear correlation between cholesterol level and bilayer
permeability. For a discussion, see: ref 16a, p 19.

11098 J. Am. Chem. Soc., Vol. 118, No. 45, 1996
Westmark et al.
amined was sorbitol > fructose > glucose, which matches the
known order of sugar-boronate complexation stabilities. The
disaccharides maltose and sucrose were not transported to any
significant degree. The evidence is strongly in favor of a carrier
mechanism. The boronic acid combines reversibly with a diol
group on the monosaccharide to produce the tetrahedral boronate
23 which is the major complexed structure in bulk, aqueous
solution. At the bilayer surface, the tetrahedral boronate is in
equilibrium with its neutral, conjugate acid (22), which is the
actual transported species. Transport mediated by reversible
Figure 7. Change in excimer/monomer ratio (I ′/I) upon addition of
phenylboronic acid (1) to PyDPPC-containing liposomes incubating
in solutions of × (gluconic acid, 100 mM); b (glucose, 100 mM); 0
(buffer control). All samples were buffered at pH 7.5; T ) 25 °C. (a)
Liposome composition DPPC:C:PA, 20:15:2; (b) liposome composition
DPPC:C:PA, 20:0:2.
a function of boronic acid addition. Regardless of sample
composition, the TMA-DPH anisotropy did not change upon
boronic acid addition. The PyDPPC response, however,
depended on both the composition of the liposomes as well as
the external solution. When the liposomes contained cholesterol
(DPPC/C/PA, 20:15:2), the I ′/I ratio remained low and invariant
upon boronic acid addition (Figure 7a). However, in certain
cases, the liposomes that did not contain cholesterol (DPPC/
C/PA, 20:0:2) showed a dose dependent response. When the
liposomes were in buffer or glucose solution, the I ′/I ratio
decreased as boronic acid was added. When the liposomes were
in gluconic acid solution, the I ′/I ratio did not change (Figure
7b). These fluorescent reponses did not vary with time.
These results can be explained in the following way. Adding
cholesterol to liposome membranes reduces translation fluidity
by “packing and ordering” the bilayer. Addition of lipophilic
arylboronic acids induces a similar but smaller effect. If the
membrane already incorporates large amounts of cholesterol,
then these “packing” effects have already occurred and the
addition of boronic acid induces no observable change in
fluidity. In the presence of glucose, the boronic acid forms a
trigonal phenylboronate ester, such as 26,28 which can also
intercalate into the membrane (the first step toward membrane
transport) and decrease translational fluidity. In the presence
of gluconic acid, however, the analogous gluconic acid phenyl-
boronate ester (27)29 is still anionic and does not penetrate the
negatively charged membrane. Thus, there is no change in
membrane fluidity.
formation of a trigonal boronate ester has been reported before
with bulk, liquid organic membranes.10 Precedence for this
mechanism occurring in bilayer membranes is the fact that weak
organic acids and bases can permeate bilayers via their chemi-
cally neutral forms, even under conditions where the neutral
forms are present in minor amounts.30
At low carrier concentrations, a first-order dependence on
carrier was observed indicating that the transported species was
a 1:1 sugar-boronate such as 26.28 Liposome leakage experi-
ments with aryl glycosides such as 28 (whose lipophilicity is
similar to 26) found that these compounds diffuse unassisted
through a lipid bilayer at a rate that is close to the facilitated
rates recorded in Table 1. This suggests that attachment of one
lipophilic boronate group to a monosaccharide provides enough
transient lipophilicity to induce partitioning into a bilayer. At
higher carrier concentrations, the carrier kinetic order ap-
proached 2, indicating the increased participation of a 1:2
sugar-bisboronate transport pathway.21
It is possible that boronic acids may be useful as selective
bilayer transport agents. In addition to the application already
discussed concerning polyol efflux, boronic acids should be
capable of delivering a wide range of hydrophilic diol-containing
compounds, such as monosaccharides, ribonucleosides, and
R-hydroxy acids, into cells. The area of drug delivery is a long
term goal and many obstacles need to be overcome. An
interesting short-term application is the possibility of using
boronic acids to load cells or vesicles with hydrophilic drugs
or biochemical reagents such as fluorescent probes.
Experimental Section
Materials. Reagents were obtained from the following sources: all
lipids (Avanti Polar Lipids) except phosphatidic acid (Sigma); all
enzymes (Sigma); all enzyme substrates (Sigma); (R)-1,2-di[1-14C]-
palmitoyl-phosphatidylcholine (14C-DPPC, Amersham); 3H-sucrose
3
(fructose-1-3H(N)) and H-glucose (Sigma); Triton-X-100 (Aldrich);
Sephadex G-50 (Pharmacia); dialysis tubing (Spectrum); Fluorodyne
scintillation fluid (Life Sciences); all fluorescent probes (Molecular
Probes), (p-nitrophenyl)glucurinide and phenyl R-D-glucoyranoside
(Sigma). UV spectra were obtained on a Perkin Elmer Lambda 2
instrument. Fluorescence data was obtained on a Perkin Elmer LS 50B
luminescence spectrometer.
Conclusion
Lipophilic boronic acids are capable of greatly facilitating
the transport of monosaccharides through lipid bilayers. The
order of transport enhancement for the monosaccharides ex-
Boronic Acids. Boronic acids 1-6 and 8-16 were obtained
(28) For simplicity, boronate ester 26 is drawn as the pyranose isomer;
however, it is well-known that boronic acids usually form complexes with
monosaccharides in their furanose configuration. Norrild, J. C.; Eggert, H.
J. Am. Chem. Soc. 1995, 117, 1479-1485.
(29) Structure 27 is postulated on the basis of the following work. (a)
Ferrier, R. J. AdV. Carbohydr. Chem. 1978, 35, 31-80. (b) van Duin, M.;
Peters, J. A.; van Bekkum, H. Tetrahedron 1985, 41, 3411-3421. van Duin,
M.; Peters, J. A.; Kieboom, A. P. G.; van Bekkum, H. Carbohydr. Res.
1987, 162, 65-78.
commercially (Aldrich or Lancaster). Compounds 713 and 1718f were
(30) (a) Eastman, S. J.; Hope, M. J.; Cullis, P. R. Biochemistry 1991,
30, 1740-1745. (b) Madden, T. D.; Harrigan, P. R.; Tai, L.; Bally, L. O.;
Mayer, T. E.; Redelmeier, H. C.; Loughrey, H. C.; Tilcock, C. P. S.;
Reinisch, L. W.; Cullis, P. R. Chem. Phys. Lipids 1990, 53, 37-46. (c)
Rottenberg, H. In Methods in Enzymology; Fleischer, S., Fleischer, B., Eds.;
Academic Press: London, 1989; Vol. 172, p 62-85.

Monosaccharide Transport through Lipid Bilayers
J. Am. Chem. Soc., Vol. 118, No. 45, 1996 11099
prepared by standard literature procedures where the appropriate bromo
precursor was treated with butyllithium followed by trimethyl borate.
Liposome Transport Assays. (A) Glucose Efflux. The rate of
glucose efflux was determined by the method of Kinsky.31 Aliquots
of the above liposome preparation (usually 40 µL, final concentration
of phospholipid plus cholesterol was 0.75 mM) encapsulating 300 mM
of glucose were added to two 1 mL cuvettes. One cuvette contained
the complete assay solution of 35 mM sodium phosphate, pH 7.5, 75
mM NaCl, 2 mM MgCl2, 2.5 units of enzyme (hexokinase/glucose-6-
phosphate dehydrogenase), 1 mM ATP, and 0.5 mM NADP. The other
cuvette contained the same reagents except for the NADP and was a
control solution used to correct for background absorption at 340 nm.
The boronic acids were added as small aliquots of concentrated aqueous
or DMSO solutions (the final concentration of DMSO was almost
always less than 5%), and the absorbance at 340 nm was monitored
over time. The initial efflux rates were determined from the amount
of glucose transported after the first 10 min. There was no glucose
consumption if the glucose was inside the liposomes and boronic acids
were absent, proving that the enzymes were unable to penetrate the
liposomes. Facilitated efflux generally occurred over a period of hours,
whereas enzyme turnover rates were much faster.31 If the glucose was
not encapsulated within the liposome, then it was totally consumed by
the coupled enyzme system within 1 min. The total amount of glucose
encapsulated inside the liposomes was determined by lysing the
liposomes with 100 µL of 10% Triton-X-100 solution at the end of
each assay.
(B) Glucose-6-phosphate Efflux. The procedure was the same as
that for glucose efflux except the liposomes were prepared with 300
mM glucose-6-phosphate as the encapsulated marker compound, and
ATP and hexokinase were deleted from the assay.
(C) Fructose Efflux. The procedure was the same as that for
glucose efflux except the liposomes were prepared with 300 mM
fructose as the encapsulated marker compound and the enzyme
phosphoglucoisomerase (5.0 units/mL) was added to the mixed enzyme
system.32
(D) Fructose-6-phosphate Efflux. The procedure was the same
as that for glucose efflux, except the liposomes were prepared with
300 mM fructose-6-phosphate as the encapsulated marker compound
and the enzyme phosphoglucoisomerase (5.0 units/mL) was added to
the mixed enzyme system, while ATP and hexokinase were deleted.
(E) Sorbitol Efflux. The procedure was similar to that for glucose
efflux, except the liposomes were prepared with 300 mM sorbitol as
the encapsulated marker compound and the enzyme was sorbitol
dehydrogenase (2 units/mL).
(A) Ester Boronic Acids 18-20. These compounds were prepared
by alkylating (4-carboxyphenyl)boronic acid (8) using the following
general procedure. Potassium hydrogen carbonate (2 mmol) was added
to a solution of 8 (0.6 mmol) in DMF (30 mL). The mixture was
stirred at reflux for 2 h. The appropriate alkylating agent (dimethyl
sulfate, benzyl bromide, or 4-tert-butylbenzyl bromide, 0.7 mmol) was
added, and the mixture was heated overnight at 60 °C. After standard
workup, the crude product was purified by column chromatography.
For 18: 1H NMR (CD3COCD3 plus 1 drop of D2O) 3.87 (s, 3H), 7.96
(s, 4H) ppm; 13C NMR (CD3COCD3 plus 1 drop of D2O) 52.3, 116.0,
129.0, 132.3, 135.0, 167.5 ppm; HRMS (positive FAB with glycerol
matrix) [M + 57]+ ) 237.0934, calcd for M ) C8H9BO4 237.0936.
For 19: 1H NMR (CD3COCD3) 5.37 (s, 2H), 7.32-7.51 (m, 5H), 7.97
(d, 2H), 8.02 (d, 2H) ppm; HRMS (positive FAB with glycerol matrix)
[M + 57]+ ) 313.1284, calcd for M ) C14H13O4B 313.1262. For 20:
1H NMR (CD3COCD3) 1.29 (s, 9H), 5.32 (s, 2H), 7.42 (s 4H), 7.97 (s,
4H). Due to extensive fragmentation, HRMS could not be performed.
(B) 3-(N-Methylnicotinamidophenyl)boronic Acid (21). Nicotinic
acid (1.82 mmol) and (3-aminophenyl)boronic acid (1.82 mmol) were
coupled together using N,N′-carbodiimidazole (1.82 mmol) in THF. A
standard workup produced 3-(nicotinamidophenyl)boronic acid as a
white solid (yield 49%): 1H NMR (CD3SOCD3 plus 1 drop of D2O)
7.33 (t, 1H), 7.54 (m, 2H), 7.79 (d, 1H), 8.00 (s, 1H), 8.27 (d, 1H),
8.72 (d, 1H), 9.06 (s, 1H) ppm; 13C NMR (CD3COCD3 plus 1 drop of
D2O) 123.4, 124.5, 126.9, 128.6, 129.3, 130.9, 131.6, 136.4, 138.4,
148.9, 152.4, 165.1 ppm; HRMS (positive FAB with glycerol matrix)
[M + 57]+ ) 299.1213, calcd for M ) C15H16N2BO4 299.1206. A
solution of 3-(nicotinamidophenyl)boronic acid (0.33 mmol) in acetone
was treated with iodomethane (1.0 mmol). After 3 h a precipitate had
formed, which was filtered to give 21 as its iodide salt (yield 25%):
1H NMR (CD3COCD3) 4.43 (s, 3H), 7.39 (t, 1H), 7.63 (d, 1H), 7.88
(d, 1H), 8.01 (s, 1H), 8.12 (s, 1H), 8.29 (t, 1H), 9.02 (d, 1H), 9.13 (d,
1H), 9.51 (s, 1H) ppm; 13C NMR (CD3COCD3 plus 1 drop of D2O)
49.4, 133.5, 136.9, 137.9, 139.1, 141.6, 147.7, 154.3, 154.5, 155.9,
158.0, 171.7 ppm; HRMS (positive FAB with glycerol matrix) [M +
56]+ ) 313.183, calcd for M ) C16H19N2BO4 313.1363.
Liposome Preparation.16 A chloroform solution of dipalmit-
oylphosphatidylcholine (DPPC, 2 µmol), cholesterol (C, 1.5 µmol), and
egg phosphatidic acid (PA, 0.2 µmol) was evaporated using a rotary
evaporator (<30 °C), and the lipid film was dried under vacuum for 1
h (longer drying times had no apparent effect on observed transport
results). The liposomes were dispersed in 200 µL of marker solution
(depending on the purpose of the experiment, the solution was either
water or 35 mM sodium phosphate buffer with 75 mM NaCl at pH
7.5) with the aid of a Vortex mixer. Pyrex beads were added before
vortexing to facilitate removal of the lipid film from the sides of the
flask. The resulting opaque dispersion was frozen in an ethanol/dry-
ice bath and then allowed to thaw in a water-bath at 37 °C. This
freeze-thaw cycle was repeated 10 times. The resulting mixture was
extruded, at room temperature, 29 times through a 19 mm polycarbonate
filter (Nucleopore) with 100 nm diameter pores using a hand-held Basic
LiposoFast device purchased from Avestin, Inc., Ottawa, Canada.17 The
extrusion device consisted of a machined housing that secured the
polycarbonate filter between two 0.5 mL syringes. The liposomal
mixture was forced back and forth from one syringe to the other. An
odd number of passages ensured that the mixture always ended up in
the receiving syringe. This method has been shown to produce LUVs
with a mean diameter of 80 nm.17 To separate untrapped marker
compound, the mixture was dialyzed against NaCl solution (0.15 M)
for at least 2 h using dialysis tubing of 15 000 MW cut-off. Ocas-
sionally, Sephadex G-50 minispun columns were used to rapidly
separate untrapped marker from the liposomes.16 The amount of
phospholipid in the liposomal fraction after separation was determined
using the Stewart assay and the amount of encapsulated glucose by
the enzymatic assay decribed below.16 The encapsulation volume was
typically around 1.0 µL/µmol of lipid.
(F) Sucrose Efflux. The procedure was the same as that for fructose
efflux except the liposomes were prepared with 300 M sucrose as the
encapsulated marker compound and invertase (1.0 units/mL) was added.
Note that 2 equiv of NADPH are produced for each equivalent of
sucrose consumed.
(G) Isocitrate Efflux. The procedure was the same as that for
glucose efflux except the liposomes were prepared with 300 mM
isocitrate as the encapsulated marker compound, the assay enzyme was
isocitrate dehydrogenase (1.5 units/mL), and the ATP was deleted.33
(H) Arsenazo III Efflux. Liposomes were prepared with 3 mM
arsenazo III as the encapsulated marker compound. The assay utilizes
the 560 nm to 606/660 nm spectral shift of arsenazo III that occurs
when the dye binds calcium ions. Both the normal assay with arsenazo
inside and calcium outside the liposomes, as well as the reverse assay
with arsenazo and calcium inside and EDTA outside, were performed.
The literature method was followed, which monitored the change in
absorbance at 660 nm.16
(I) Carboxyfluoroscein/Calcein Efflux.16 Liposomes were prepared
with carboxyfluoroscein (CF, 100 mM in 10 mM sodium phosphate
buffer, pH 7.7) and glucose (300 mM) as the encapsulated marker
compounds. CF fluorescence is linearly proportional to concentration
between 3 and 30 µM. Above 30 µM self-quenching becomes
dominant. To monitor CF efflux, an aliquot of the liposome preparation
(40 µL) was added to a glass fluorescence cuvette containing 1 mL of
(31) Kinsky, S. C. In Methods in Enzymology; Fleischer, S., Packer, L.,
Eds.; Academic Press: London, 1974; Vol. 32, p 501-512.
(32) Beutler, H.-O. In Methods of Enzymatic Analysis, 3rd ed.; Berg-
meyer, H. U., Ed.; Verlag: Weinheim, 1983; Vol. 6, pp 321-327.
(33) Crowe, L. M.; Womersley, C.; Crowe, J. H.; Reid, D.; Appel, L.;
Rudolph, A. Biochim. Biophys. Acta 1986, 861, 131-140.

11100 J. Am. Chem. Soc., Vol. 118, No. 45, 1996
Westmark et al.
the complete glucose assay solution as well as the appropriate boronic
acid. The change in CF fluorescence at 520 nm (excitation at 470
nm) was monitored for 90 min before the liposomes were lysed by
addition of Triton-X-100 (300 µL of a 10% solution). The calcein
assay followed the same procedure except the calcein was encapsulated
at pH 7.3 and the fluorometer settings were slightly different (excitation
at 491 nm and emission at 511 nm).
(J) Radiotracer Assay for Glucose Influx. A stock solution
containing sucrose (50 mM) and glucose (50 mM, doped with 8 µCi/
mL of 15.2 Ci/mmol 3H-glucose) was prepared. Empty liposomes
containing 70 mM sodium phosphate at pH 7.5 were prepared as
described above, except the phospholipid lipid layer was doped with
14C-DPPC (112 mCi/mmol) at a ratio of 0.25 µCi/10 µmol of DPPC.
The glucose influx experiment was initiated when 475 µL of liposome
solution, 475 µL stock solution, and 50 µL of boronic acid solution
(aqueous or DMSO) were combined at 25 °C. Every 10 min, a 100
µL sample was withdrawn and spun-down through a 1 mL Sephadex
G-50 column.16 The liposome fraction was added to 10 mL of
scintillation fluid, and the 3H and 14C activity levels were counted using
a scintillation counter. Each influx experiment was repeated with
independent solutions. Sucrose influx was determined using the same
method except the sucrose was doped with tracer and the glucose was
not.
mM NaCl, pH 7.5). The G value (correction for instrument polarization
response which causes the ratio of IV/IH to deviate from unity) was
calculated. After each addition of boronic acid (10 µL of concentrated
boronic acid solution), the polarization and anisotropy values were
determined. The excitation wavelength was 360 nm, and the emission
wavelength was 427 nm, with slit widths of 4 and 3.5, respectively.
Perkin Elmer StdPol. Software was used to obtain IV(V) (emission
intensity of vertically polarized light parallel to the plane of excitation),
IV(H) (emission intensity of horizontally polarized light perpendicular
to the plane of excitation), polarization, and anisotropy values.
(B) Fluorescence Assay for Translational Fluidity.27 The lipo-
somes were prepared as described above, except that PyDPPC (2 mol
% for liposomes without cholesterol and 5 mol % for liposomes with
cholesterol) was added to the starting phospholipid solution in
chloroform. The final concentration of phospholipid in the 3 mL
fluorescence cell was 3.5 µM. The excitation wavelength was 320 nm;
emission wavelengths were 397 and 479 nm.
Acknowledgment. We thank Professor R. Bretthauer for
allowing us to use his radiochemical facilities. We are grateful
to Jeffrey Bien, Jennifer Riggs, Linbee Valencia, and Carol
Monahan for assistance. We thank two referees for their careful
and thoughtful reviews. This work was supported by the
National Science Foundation (CHE96-27099). B.D.S. acknowl-
edges a Cottrell Scholar award of Research Corporation.
Membrane Fluidity Studies. (A) Fluorescence Assay for Rota-
tional Fluidity Using TMA-DPH.26 A liposome mixture containing
only buffer solution was prepared using the procedure described above
and diluted to a concentration of 345 µM phospholipid. TMA-DPH
(0.5 mol % probe to phospholipid) was added, and the mixture was
immediately vortexed and then allowed to incubate for 18 h. The
fluorescence assay used a 3 mL cuvette that was stirred magnetically
at 25 °C. The cuvette was filled with probe-containing liposomes (58
µM phospholipid) in buffer solution (35 mM sodium phosphate, 75
Supporting Information Available: Representative raw
data for transport kinetic order, pH effects, counterion effects,
and membrane composition (9 pages). See any current mast-
head page for ordering and Internet access instructions.
JA961264H