C O M M U N I C A T I O N S
charged groups on the gA derivatives.12 Thus, under conditions of
high ionic strength, the local ion concentration near the opening of
the gA pore should be similar to the bulk ion concentration of the
recording electrolyte, regardless of the charge on the group presented
on the gA derivative. The results shown in Figure 2B support this
hypothesis since they demonstrate that the conductance values of the
charged derivatives of gA approach the conductance value of native
(uncharged) gA with increasing ionic strength of the recording
electrolyte. The results in Figure 2B also suggest that as the ionic
strength increases, the Coulombic effects on conductance relative to
gA are outweighed by other factors introduced by the functional groups
and linkers attached to gA. These non-Coulombic effects on the
conductance appear to correlate most strongly with the steric bulk
introduced by the functional groups attached to the C-terminus of gA
(Figure 1D). Similar conclusions were drawn for another gA derivative
described by Apell et al.12a
at pH 9, with high conductance observed when gA-P was at the
entrance of the pore (Figure 4B). Since these charged derivatives
of gA do not cross the membrane by themselves (see Figure S2),
the rectifying behavior of these gA derivatives was stable over
several hours in a membrane. Additionally, we demonstrated the
capability to controllably “turn on” the rectifying behavior of a
channel in situ by adding the enzyme alkaline phosphatase (which
cleaves the phosphate group off gA-P to produce an uncharged gA
derivative)11c to only one side of the membrane in the bilayer setup.
Before addition of this enzyme, this setup contained only nonrec-
tifying homodimeric channels comprised of gA-P (see Figure S3).
The possibility of changing the conductance of gA derivatives in
real time in response to a variety of external stimuli9,11 may present
opportunities for sensing and for controlling the “turn on” or “turn
off” rectifying properties of these channels.
These results demonstrate the modular and tunable nature of a
gA platform for generating ionic diodes in a membrane. The dimeric
characteristic of gramicidin-based channels makes it possible to
access each half of the channel independently in order to incorporate
the channel asymmetry required for the desired rectification.
Thus, the modular nature of gramicidin channels makes it
possible to tune the diode-like behavior of these pores. Addition-
ally, the possibility of changing the conductance properties of gA
derivatives in the presence of external stimuli9-11 may be advanta-
geous for manipulating the rectifying behavior of gA-based ionic
diodes after incorporation of these pores into a nanofluidic device.
Finally, since these gA-based diodes are the smallest fluidic circuit
elements reported to date, they may represent an important step
toward the realization of miniaturized devices that rely on control
of ionic flow within fluidic networks.
Figure 3. Schematic illustration of homodimeric and heterodimeric
gramicidin channels in a membrane. (A) Nonrectified conductance of
monovalent cations through a symmetric gA channel. (B) The charged
groups on an asymmetric channel comprising gA-P and gA-NMe3 presum-
ably have different effects on the local concentration of cations near the
pore,10 resulting in diode-like rectifying behavior.
In order to engineer rectifying behavior in gA channels, we
generated heterodimeric channels by adding gA-NMe3 to one side
(the cis side) of the membrane and either gA-T or gA-P to the
other side (the trans side) of the membrane (Figure 3B). For these
asymmetric channels, we expected the positive and negative charges
on the gA derivatives to affect (depending on the sign and
magnitude of the charge) the local concentration of cations near
the pore on either side of the membrane. This effect would result
in an overall decrease in single channel conductance of ions from
the cis side of the membrane relative to the conductance of ions
originating from the trans side of the membrane. Figure 4A shows
that asymmetric channels comprising gA-NMe3 and gA-T exhibited
at pH 7 a (3.4 ( 0.1)-fold difference in single channel conductance
of monovalent cations across the membrane in one direction relative
to the conductance in the other direction (i.e., with higher
conductance observed when gA-T was at the entrance of the pore).
To demonstrate the modular nature of these gA-based diodes, we
found that replacing gA-T with gA-P (to generate asymmetric
channels comprising gA-NMe3 and gA-P) resulted in a (4.8 ( 0.1)-
fold difference in single-channel conductance across the membrane
Acknowledgment. This work was partially supported by the
NSF (CHE-0847530 and CBET-0449088). M.X.M. acknowledges
support from UC TSR&TP and CSGC Graduate Fellowships.
Supporting Information Available: Synthesis, experimental meth-
ods, and I-V curves and traces of single channel recordings. This
References
(1) (a) Vlassiouk, I.; Siwy, Z. Nano Lett. 2007, 7, 552. (b) Vlassiouk, I.; Kozel,
T.; Siwy, Z. J. Am. Chem. Soc. 2009, 131, 8211. (c) Kalman, E.; Sudre,
O.; Vlassiouk, I.; Siwy, Z. Anal. Bioanal. Chem. 2009, 394, 413. (d) He,
Y.; Gillespie, D.; Boda, D.; Vlassiouk, I.; Eisenberg, R.; Siwy, Z. J. Am.
Chem. Soc. 2009, 131, 5194.
(2) Xu, J.; Lavan, D. A. Nat. Nanotechnol. 2008, 3, 666.
(3) Xu, J.; Sigworth, F. J.; LaVan, D. A. AdV. Mater. 2010, 22, 120.
(4) (a) Han, J. H.; Kim, K. B.; Kim, H. C.; Chung, T. D. Angew. Chem., Int.
Ed. 2009, 48, 3888. (b) Yameen, B.; Ali, M.; Neumann, R.; Ensinger, W.;
Knoll, W.; Azzaroni, O. Nano Lett. 2009, 9, 2788.
(5) Alcaraz, A.; Ramirez, P.; Garcia-Gimenez, E.; Lopez, M. L.; Andrio, A.;
Aguilella, V. M. J. Phys. Chem. B 2006, 110, 21205.
(6) Maglia, G.; Heron, A. J.; Hwang, W. L.; Holden, M. A.; Mikhailova, E.;
Li, Q.; Cheley, S.; Bayley, H. Nat. Nanotechnol. 2009, 4, 437.
(7) Hille, B. Ionic Channels of Excitable Membranes, 2nd ed.; Sinauer
Associates: Sunderland, MA, 1992.
(8) Andersen, O. S.; Koeppe, R. E., II; Roux, B. In Biological Membrane Ion
Channels; Chung, S.-H., Andersen, O. S., Krishnamurthy, V., Eds.;
Springer: New York, 2007; Chapter 2.
(9) Blake, S.; Capone, R.; Mayer, M.; Yang, J. Bioconjugate Chem. 2008, 19, 1614.
(10) (a) Capone, R.; Blake, S.; Restrepo, M. R.; Yang, J.; Mayer, M. J. Am.
Chem. Soc. 2007, 129, 9737. Majd, S.; Yusko, E. C.; MacBriar, A. D.;
Yang, J.; Mayer, M. J. Am. Chem. Soc. 2009, 131, 16119.
(11) (a) Blake, S.; Mayer, T.; Mayer, M.; Yang, J. ChemBioChem 2006, 7, 433.
(b) Macrae, M. X.; Blake, S.; Mayer, T.; Mayer, M.; Yang, J. Proc. SPIE
2009, 7397, 739709. (c) Macrae, M. X.; Blake, S.; Jiang, X.; Capone, R.;
Estes, D. J.; Mayer, M.; Yang, J. ACS Nano 2009, 3, 3567.
(12) (a) Apell, H. J.; Bamberg, E.; Alpes, H.; La¨uger, P. J. Membr. Biol. 1977,
31, 171. (b) Veatch, W.; Stryer, L. J. Mol. Biol. 1977, 113, 89. (c) Cifu,
A. S.; Koeppe, R. E.; Andersen, O. S. Biophys. J. 1992, 61, 189. (d) Matile,
S.; Som, A.; Sorde´, N. Tetrahedron 2004, 60, 6405.
Figure 4. Single channel conductance (γ) profiles of asymmetric channels.
(A) I-V curves for an asymmetric channel consisting of gA-NMe3 and
gA-T in recording electrolyte (100 mM CsCl, 1 mM HEPES, pH 7). (B)
I-V curves for an asymmetric channel consisting of gA-NMe3 and gA-P
in recording electrolyte (50 mM CsCl, 0.5 mM Tris, pH 9). The bilayers
were composed of DiPhyPC lipids. pS ) picosiemens.
(13) (a) Reiss, P.; Al-Momani, L.; Koert, U. ChemBioChem 2008, 9, 377. (b)
Woolley, G. A.; Zunic, V.; Karanicolas, J.; Jaikaran, A. S.; Starostin, A. V.
Biophys. J. 1997, 73, 2465.
(14) The charge on phosphate is expected to be roughly -1.5 at pH 7 and -2 at pH 9.
JA909876H
9
J. AM. CHEM. SOC. VOL. 132, NO. 6, 2010 1767