Internal Electrostatic Potentials in Bilayers: Measuring and Controlling
Dipole Potentials in Lipid Vesicles
J. Craig Franklin and David S. Cafiso
From the Department of Chemistry and Biophysics Program at the University of Virginia, Charlottesville, Virginia 22901 USA
vesicle systems formed from phosphatidylcholine. As a result of the membrane dipole potential, the binding and translocation
rates for oppositely charged hydrophobic ions are dramatically different. These differences were analyzed using a simple
electrostatic model and are consistent with the presence of a dipole potential of approximately 280 mV in phosphatidylcholine.
Phloretin, a molecule that reduces the magnitude of the dipole potential, increases the translocation rate of hydrophobic cations,
while decreasing the rate for anions. In addition, phloretin decreases the free energy of binding of the cation, while increasing
the free energy of binding for the anion. The incorporation of 6-ketocholestanol also produces differential changes in the binding
and translocation rates of hydrophobic ions, but in an opposite direction to those produced by phloretin. This is consistent with
the view that 6-ketocholestanol increases the magnitude of the membrane dipole potential. A quantitative analysis of the binding
and translocation rate changes produced by ketocholestanol and phloretin is well accounted for by a point dipole model that
includes a dipole layer due to phloretin or 6-ketocholestanol in the membrane-solution interface. This approach allows dipole
potentials to be estimated in membrane vesicle systems and permits predictable, quantitative changes in the magnitude of the
internal electrostatic field in membranes. Using phloretin and 6-ketocholestanol, the dipole potential can be altered by over 200
mV in phosphatidylcholine vesicles.
The binding and translocation rates of hydrophobic cation and anion spin labels were measured in unilamellar
Several distinct electrostatic potentials can be defined in lipid
bilayers that arise from a number of sources. For example,
transmembrane potentials result from charge separation
across the membrane, and membrane surface potentials re-
sult from charge that is bound to the membrane-solution in-
terface (1). Transmembrane potentials are known to regulate
the activity of certain ion channels, presumably as a result of
conformational changes in these proteins that are electrically
active. Although they have not yet been characterized, these
electrically active structural transitions must involve the
movement of protein charges or dipole moments within the
membrane interior so that conformational free energies are
dependent upon the transmembrane electric field. Surface
potentials represent a potential difference between the mem-
brane interface and the bulk aqueous phase. These potentials
are usually much smaller than transmembrane potentials, and
are typically on the order of a few tens of mV in biological
membranes; nonetheless, they appear to have an important
role in controlling the binding of charged proteins to mem-
branes. Indeed, surface potentials may control the activity of
enzymatic components of second-messenger systems, such
as protein kinase C, by regulating the binding of these pro-
teins to the membrane interface (2). In addition to trans-
membrane and surface potentials, lipid bilayers also possess
an internal potential termed a dipole potential. This potential
is much larger than either the surface or the transmembrane
Receivedforpublication 9December 1992 andinfinalfonn 12March 1993.
Address reprint requests to David S. Cafiso.
1993 by the Biophysical Society
potentials, and appears to be on the order of 300 mV hy-
drocarbon positive in phosphatidylcholine vesicles. The ex-
istence ofthe dipole potential has been known for some time,
both from electrical measurements in monolayers (3) and
work in model bilayers. In bilayers, it is believed to account
for the permeability differences of certain organic cations
and anions (4-10). Unlike the surface potential, the dipole
potential is independent ofionic strength; hence, it is thought
to result from oriented dipoles in the membrane interface.
The molecular source of the dipole potential has not been
clearly identified; however, carbonyl oxygens and/or water
at the membrane interface appear to be the most likely
sources of this potential. (11-13) A major role for water is
suggested by recent electrostatic calculations (14), as well as
by an apparent connection between the hydration pressure
and the dipole potential (12).
The magnitude of the dipole potential suggests that it
should be important in a number of processes. For example,
conformational changes in membrane proteins that result in
the movement of charge or dipole moments through the
membrane interface will be affected by the dipole potential.
The insertion of a-helical segments of proteins into mem-
branes should also be modulated by the dipole potential, and
this interaction may be an important energetic term affecting
the insertion ofpeptides or signal sequences into membranes
(15). The dipole potential is also expected to slow the trans-
location of positively charged segments of membrane pro-
teins during biosynthesis. As a result, the dipole potential
might account for the presence ofpositively charged residues
in stop-transfer sequences (16). Finally, virtually all efficient
protonophores are hydrophobic weak acids that shuttle
across bilayers in a negatively charged form (17). The lack
of any efficient weak base protonophores is likely the result
of the large the dipole potential, since these compounds
would need to shuttle charge in a positively charged form.
In spite of its likely importance, the role of the dipole
potential in controlling protein interactions with membranes
or its affect on membrane protein conformation has not been
demonstrated. This is a result of the fact that experimental
systems to examine these protein-membrane electrostatic in-
teractions have not been explored. The majority of studies on
dipole potentials in membranes have been carried out in pla-
narbilayer systems, and, while these systems provide a facile
means to monitor ion conduction, they do not readily permit
the measurement of events such as protein binding or the
characterization of protein structural changes. However,
membrane dipole potentials can also be estimated in lipid
vesicles where structural studies on proteins are also possible
An estimate ofthe dipole potential in vesicles can be made
by examining the membrane permeability and binding of
certain organic ions termed hydrophobic ions. Shown in Fig.
1 are the calculated free energy profiles for transferring the
hydrophobic cation tetraphenylphosphonium, #4P', and the
anion, tetraphenylborate 44B-, from solution across a lipid
bilayer in the absence of a surface potential. This energy is
estimated by adding together AGBOm,,the electrostatic Born-
Image energy, AGHYdO the (attractive) hydrophobic energy,
and AGDipole the energy of interaction of the ion with the
membrane dipole field (10).
AGO =AGBOm+AGOjPOIe + AGOYdrO
Distance from Bilayer Center (A)
ions tetraphenylborate (44B-) and tetraphenylphosphonium (44P+). These
energy profiles are calculated by adding together terms due to the electro-
static charging or Born-Image energy, the hydrophobic binding energy, and
the energy due to the dipole potential. The dipole potential makes these two
profiles dramatically different, a feature that accounts for the increased
binding and increased translocation rates of04B-when compared to
These profiles were determined as described previously (10), where the
intrinsic membrane dipoles have a strength of0.85 D with the density ofthe
lipid. The magnitude of the neutral binding energy was taken as -6.8 kcall
mol, with an ion radius of 4.2 and 4.0 A for 44B- and 44P+, respectively.
Free energy profiles across lipid bilayers for the hydrophobic
The large size of these ions lowers AGBOm,,, which makes
them quite membrane permeable. As a result of their hy-
drophobicity, the free energy profiles also show a minimum
within the membrane solution interface, and they are exper-
imentally observed to bind to a low dielectric domain of the
membrane (11). The differences between energy curves for
the hydrophobic cation and anion depicted in Fig. 1 are a
direct consequence of the dipole potential. As is readily ap-
preciated from Fig. 1, the dipole potential creates dramatic
differences in the translocation and binding of hydrophobic
ions, by altering the free energy barrier in the middle of the
bilayer and the depth of the free energy minima at the in-
terfaces. In fact, otherwise similar cations and anions are
observed experimentally to have as much as a 106 times
difference in their transport rate constant (18). By modeling
the experimental binding and translocation rates of hydro-
phobic ions, an estimate of the dipole potential is possible.
A number of compounds are known to modify the dipole
potential of bilayers. For example, phloretin is a compound
that appears to reduce the magnitude of the dipole potential,
and it has been studied in planar bilayer systems (19-21).
Previous work carried out in this laboratory demonstrated
that phloretin produces changes in the translocation rates of
hydrophobic ions in lipid vesicles that are consistent with the
magnitude of its molecular dipole moment (22). In these
vesicle systems the membrane concentration ofphloretin can
be easily determined, which makes a quantitative evaluation
of its action possible. In addition to changes in the translo-
cation rates of hydrophobic ions, changes in the dipole po-
tential should also produce changes in the binding of hy-
drophobic ions (10), however, these binding changes have
not been measured.
Recently, several newly developed spin-labeled hydro-
phobic anion probes were described that allow the direct
measurement of hydrophobic anion transport and binding in
vesicle systems (23). These probes provide an excellent com-
plement to previously developed positively charged phos-
phonium probes (24). In the present study, both the trans-
location rates andbinding free energies ofoppositely charged
spin-labeled hydrophobic ion probes (I-III below) are ex-
amined in vesicles. In the context of a simple model, we
demonstrate that these probes provide an estimate of the
membrane dipole potential. In addition, the effects of phlo-
retin and 6-ketocholestanol on both the binding and transport
rates of these probes are examined. Ketocholestanol is an
agent that increases monolayer surface potentials, and it is
expected to increase the membrane dipole potential (25). The
changes produced by these agents on the binding and trans-
port ofhydrophobic ions are shown to be consistent with the
effects of adding the molecular dipole moment of these mol-
ecules to the membrane-solution interface. The use of these
compounds allows the dipole potential to be lowered or
raised over a wide magnitude and provides an excellent ex-
perimental methodology to examine the effects ofdipole po-
tentials on membrane protein structure and binding.
Franklin and Cafiso
4. Liberman, E. A., and V. P. Topaly. 1969. Permeability of bimolecular
phospholipid membranes for fat-soluble ions. Biophysics. 14:477-487.
5. LeBlanc, 0. H. 1969. Tetraphenylborate conductance through lipid bi-
layer membranes. Biochim. Biophys. Acta. 193:350-360
6. Hladky, S. B. 1974. The energy barriers to ion transport by nonactin
across thin lipid membranes. Biochim. Biophys. Acta. 352:71-85.
7. Andersen, 0. S., and M. Fuchs. 1975. Potential energy barriers to ion
transport within lipid bilayers. Studies with tetraphenylborate. Biophys.
8. Pickar, A. D., and R. Benz. 1978. Transport of oppositely charged li-
pophilic probe ions in bilayer membranes having various structures.
J. Membr. Biol. 44:353-376.
9. Hladky, S. B., and D. A. Haydon. 1983. Membrane conductances and
surface potential. Biochim. Biophys. Acta. 318:464-468.
10. Flewelling, R. F., and W. L. Hubbell. 1986. The membrane dipole po-
tential in a total membrane potential model. Applications to hydropho-
bic ion interactions with membranes. Biophys. J. 49:541-552.
11. Ellena, J. F., R. N. Dominey, S. J. Archer, Z-C. Zu, and D. S. Cafiso.
1987. The localization of hydrophobic ions in phospholipid bilayers
using 'H nuclear Overhauser effect spectroscopy. Biochemistry. 26:
12. Simon, S. A., and T. J. McIntosh. 1989. Magnitude of the solvation
pressure depends on dipole potential. Proc. Natl. Acad. Sci. USA. 86:
13. Gawrisch, K., D. Ruston, J. Zimmerberg, V. A. Parsegian, R. P. Rand,
and N. Fuller. 1992. Membrane dipole potentials, hydration forces, and
the ordering of water at membrane surfaces. Biophys. J. 61:1213-1223.
14. Sheng, C., and G. Vanderkooi. 1992. Molecular origin of the internal
dipole potential in lipid bilayers: calculation of the electrostatic po-
tential. Biophys. J. 63:935-941.
15. Cafiso, D. S. 1991. Lipid bilayers: membrane-protein electrostatic in-
teractions. Curr. Opin. Struct. Biol. 1:185-190.
16. Dalbey, R. E. 1990. Positively charged residues are important deter-
minants of membrane protein topology. Trends Biochem. Sci. 15:253-
17. McLaughlin, S., and J. Dilger. 1980. The transport of protons across
membranes by weak acids. Physiological Rev. 60:825-863.
18. Flewelling, R. F., and W. L. Hubbell. 1986. Hydrophobic ion interac-
tions with membranes. Thermodynamic analysis of tetraphenylphos-
phonium binding to vesicles. Biophys. J. 49:531-540.
19. Andersen, 0. S., Finkelstein, A., Katz, I., and A. Cass. 1976. Effect of
phloretin on the permeability of thin lipid membranes. J. Gen. Physiol.
20. Melnik, E., Latorre, R., Hall, J., and D. Tosteson. 1977. Phloretin in-
duced changes in ion transport across lipid bilayer membranes. J. Gen.
21. Reyes, J., Greco, F., Motais, R., and R. Latorre. 1983. Phloretin and
phloretin analogues: mode of action in planar bilayers and monolayers.
J. Membr. Biol. 72:93-103.
22. Perkins, W. R., and D. S. Cafiso. 1987. A procedure using voltage-
sensitive spin-labels to monitor dipole potential changes in phospho-
lipid vesicles: the estimation ofphloretin-induced conductance changes
in vesicles. J. Membr: Biol. 96:165-173.
23. Franklin, C., D. Cafiso, R. Flewelling, and W. Hubbell. 1993. Probes
ofmembrane electrostatics. Synthesis and voltage-dependent partition-
ing of negative hydrophobic ion spin-labels in lipid vesicles. Biophys.
24. Cafiso, D. S., and W. L. Hubbell. 1981. EPR determination of mem-
brane potentials. Annu. Rev. Biophys. Bioeng. 10:217-244.
25. Simon, S. A., T. J. McIntosh, A. D. Magid, and D. Needham. 1992.
Modulation of the interbilayer hydration pressure by the addition of
dipoles at the hydrocarbon/water interface. Biophys. J. 61:786-799.
26. Singleton, W. S., Gray, M. S., Brown, M. L., and J. L. White. 1965.
Chromatographically homogeneous lecithin from egg phospholipids.
J. Am. Oil. Chem. Soc. 42:53-56.
27. Bartlett, G. R. 1959. Phosphorous assay in column chromatography.
J. Biol. Chem. 243:466-468.
28. Cafiso, D. S., and W. L. Hubbell. 1982. Transmembrane electrical cur-
rents of spin-labeled hydrophobic ions. Biophys. J. 39:263-272.
29. Bechinger, B., and J. Seelig. 1991. Interaction of electric dipoles with
phospholipid head groups. AH-2 and P-31 NMR study ofphloretin and
phloretin analogs in phosphatidylcholine membranes. Biochemistry. 30:
30. Ketterer, B., B. Neumcke, and P. Lauger. 1971. Transport mechanism
of hydrophobic ions through lipid bilayer membranes. J. Membr. Biol.
31. McConnell, H. M. 1991. Structures and transitions in lipid monolayers
at the air-water interface. Annu. Rev. Physical Chem. 42:171-195.
32. Ellena, J. F., S. J. Archer, R. N. Dominey, B. D. Hill, and D. S. Cafiso.
1988. Localizing the nitroxide group offatty acid and voltage-sensitive
spin-labels in phospholipid bilayers. Biochim. Biophys. Acta. 940:63-