Friction-mediated dynamic disordering of phospholipid membrane by mechanical motions of photoresponsive molecular glue: Activation of ion permeation

Article abstract of DOI:10.1021/ja3074424

A water-soluble photoresponsive molecular glue, Azo-¹⁸Glue, consisting of a photochromic azobenzene core and two adhesive dendritic wedges with a total of 18 peripheral guanidinium ion (Gu⁺) pendants tightly adheres to the surface of a phospholipid membrane, even in buffer, via a multivalent salt-bridge formation with phosphate anions. A photomechanical motion of adhering Azo-¹⁸Glue possibly gives rise to dynamic structural disordering of the phospholipid membrane and activates transmembrane ion permeation. In sharp contrast, no activation of ion permeation results when poorly adhesive Azo-⁶Glue carrying only six Gu⁺ pendants is used in place of Azo-¹⁸Glue.

Full text of DOI:10.1021/ja3074424

Communication
pubs.acs.org/JACS
Friction-Mediated Dynamic Disordering of Phospholipid Membrane
by Mechanical Motions of Photoresponsive Molecular Glue:
Activation of Ion Permeation
Yushi Suzuki, Kou Okuro,* Tadashi Takeuchi, and Takuzo Aida*
Department of Chemistry and Biotechnology, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo
113-8656, Japan.
S
* Supporting Information
ABSTRACT: A water-soluble photoresponsive molecular
glue, Azo-18Glue, consisting of a photochromic azobenzene
core and two adhesive dendritic wedges with a total of 18
peripheral guanidinium ion (Gu+) pendants tightly adheres
to the surface of a phospholipid membrane, even in buffer,
via a multivalent salt-bridge formation with phosphate
anions. A photomechanical motion of adhering Azo-18Glue
possibly gives rise to dynamic structural disordering of the
phospholipid membrane and activates transmembrane ion
permeation. In sharp contrast, no activation of ion
permeation results when poorly adhesive Azo-6Glue
carrying only six Gu+ pendants is used in place of
Azo-18Glue.
riction is a force resisting the relative motion of two
surfaces in contact sliding against each other. Not only in
F
macroscopic but also molecular-scale events, friction is often
crucial for mechanical information transfer between two
objects. Examples of friction-mediated interfacial phenomena
in biology include muscle contraction,1 selectin-mediated
internalization of a flowing leukocyte,2 and operation of
biomolecular machines.3 We envisioned if such a molecular-
scale friction, intentionally generated by a tailored adhesive
motif, might noncovalently give rise to a structural perturbation
of cell membranes and therefore alter cellular activities. As a
proof of concept, we designed Azo-18Glue (Figure 1) that
consists of a photochromic azobenzene core and two sticky
dendritic wedges carrying 18 peripheral guanidinium ion (Gu+)
pendants, and explored the possibility that its “photo-
mechanical motion” might dynamically alter physical properties
of a natural phospholipid membrane.
Figure 1. Schematic structures of photoresponsive molecular glues
Azo-6Glue and Azo-18Glue together with that of fluorescently labeled
molecular glue FITC-9Glue. Red circles represent adhesive guanidi-
nium ion (Gu+) pendants.
Recently, we developed water-soluble dendritic molecular
glues having multiple peripheral Gu+ pendants.4 Gu+ is known
to bind oxyanions by the formation of a salt bridge via
hydrogen-bonding and electrostatic interactions.5 The Gu+-
appended dendritic molecular glues were designed with such an
expectation that a multivalency, given to the salt-bridge
formation by the dendritic scaffold, could amplify the adhesivity
toward target objects.4a In fact, the dendritic molecular glues, in
aqueous media, can tightly adhere to the surfaces of proteins
Photoisomerization of azobenzene, at a single-molecule level, is
known to occur almost instantaneously in a subpicosecond
range.6 Although such an abruptness in motion might be much
tempered at the adhesive points due to a conformational
flexibility of the dendritic architecture, we found that the
photomechanical motion of Azo-18Glue possibly activates ion
permeation across a phospholipid membrane.
Photoresponsive Azo-18Glue (Figure 1) was synthesized by
condensing azodianiline with a dendron carrying azide groups,
4a,b
4c
(CO2 ) and clay nanosheets (SiO−). In the present work,
Azo-18Glue (Figure 1) was designed in such a way that its two
adhesive dendritic wedges, upon photoisomerization of the
azobenzene core, can change their relative locations (Figure 2).
−
Received: July 28, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja3074424 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Figure 3. Adhesion of fluorescently labeled molecular glue
FITC-9Glue onto a phospholipid membrane in HEPES buffer (10
mM HEPES, 100 mM NaCl, pH 7.0). (a) Fluorescence microscopy
(λext = 488 nm) of giant EYPC liposomes (140 μM) containing
FITC-9Glue (25 nM). (b) Flow cytometry histograms (λext = 488 nm)
of giant EYPC liposomes (140 μM) before (blue) and after (green)
mixing with FITC-9Glue (25 nM).
the Gu+ pendants could adhere more tightly than FITC-9Glue
to the EYPC liposomes.
Photoisomerization of the azobenzene core of Azo-18Glue
was confirmed by electronic absorption spectroscopy (Figure
S3, Supporting Information [SI]).7 When a HEPES buffer
solution of Azo-18Glue was irradiated at 365 nm for 2 min in
the presence of EYPC liposomes (∼100 nm), an absorbance at
385 nm due to the trans-form of the azobenzene core10
decreased (Figure S3b, SI, green curve).7 Successive photo-
irradiation at 546 nm allowed resultant Azo-18Glue to recover
the 385-nm absorbance almost completely (Figure S3b, SI, red
curve).7 The spectral change profiles thus observed were
virtually identical to those of Azo-18Glue without EYPC
liposomes (Figure S3a, SI,),7 indicating that adhesion onto
the liposomal surface hardly affects the isomerization profile of
Azo-18Glue.
Figure 2. Schematic representations of possible effects of photo-
responsive molecular glue Azo-18Glue on a phospholipid membrane.
(a) Static adhesion effects of trans-Azo-18Glue (left) and cis-Azo-18Glue
(right). In either case, structural disorder is unlikely for the
phospholipid membrane due to a thermodynamic equilibration. (b)
Dynamic effects of trans-to-cis (left) and cis-to-trans (right) photo-
isomerization events of adhering Azo-18Glue. In either case, a dynamic
structural disorder, allowing for transmembrane ion permeation (see
Figure 4a), can be expected for the phospholipid membrane by an
abrupt photomechanical motion of Azo-18Glue.
Of particular interest, the photoisomerization of Azo-18Glue
is accompanied by ion permeation across the liposomal
membrane (Figure 4a). The transmembrane ion permeation
was traced by using EYPC liposomes (35 μM) including
pyranine in their interior and Azo-18Glue (1.5 μM) in both the
interior and exterior liposomal environments. Pyranine is a pH-
responsive fluorescent dye, which emits fluorescence at 510 nm
upon photoexcitation at 460 and 405 nm.11 The fluorescence
intensity upon 460-nm excitation (I460) is known to change
with pH, whereas that excited at 405 nm (I405) remains intact to
followed by deprotection of a ‘click’ reaction product between
the resultant dendritic azide with a dendron carrying an alkyne
focal core and Boc-protected Gu+ pendants.7 Fluorescently
labeled FITC-9Glue (Figure 1) carrying nine Gu+ pendants was
synthesized by the isothiocyanate-mediated addition of FITC
(fluorescein isothiocyanate) onto a dendron carrying a focal
amino group and Boc-protected Gu+ pendants.7 An aqueous
dispersion of egg-yolk phosphatidylcholine (EYPC) liposomes
in HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid) buffer ([HEPES] = 10 mM, [NaCl] = 100 mM)8 was
prepared according to a literature method9 by subjecting a
HEPES buffer dispersion of crude EYPC liposomes to freeze−
thaw cycles followed by repeated extrusion processes.7 For
fluorescence microscopy and flow cytometry, additional
freeze−thaw cycles were conducted subsequently to obtain
giant EYPC liposomes.9c When mixed with FITC-9Glue
(Figure 1; 25 nM) in HEPES buffer, the EYPC liposomes
(140 μM), as observed by fluorescence microscopy, turned
fluorescent upon photoexcitation at 488 nm (Figure 3a). In
flow cytometry (λext = 488 nm; Figure 3b), the EYPC
liposomes treated with FITC-9Glue displayed a monodisperse
histogram (green) that shifted completely from that of the
original EYPC liposomes (blue). From this histogram along
with a fluorescence spectral feature of the mixture, roughly 17%
of added FITC-9Glue molecules were estimated to attach to the
EYPC liposomes.7 Therefore, it is most likely that a multivalent
salt-bridge formation operates properly even in buffer. Now
one can expect that Azo-18Glue having a larger number (18) of
a pH variation.11 Thus, the pH change at the liposomal interior
11
can be evaluated from I460/I405
.
By using NaOH, the initial
pH values of the inner and outer environments of the
liposomes were adjusted to 7.0 and 8.4, respectively, in order
to generate a gradient of [OH−]. The resultant dispersion,
located in a fluorescence spectrometer, was continuously
exposed over a period of 5 min to an external UV light (100-
W xenon light source) at 365 nm for executing the trans-to-cis
isomerization of Azo-18Glue. Then, the sample was subjected to
fluorescence spectroscopy as for the internal pyranine dye by
photoexcitation at λext = 460 and 405 nm to record I460/I405 for
obtaining a time-dependent pH-change profile at the liposomal
interior. As shown in Figure 4a (red circles), the ratio I460/I405
after the 5-min photoirradiation of Azo-18Glue at 365 nm was
greater than that before the irradiation, indicating an increment
of pH at the liposomal interior (blue-shaded areas). Such a
light-driven internal pH jump could be repeated multiple times.
Dynamic light scattering (DLS) analysis (Figure 4b) revealed
that the liposomes, even after 15 min photoirradiation at 365
nm, maintain their morphological and dimensional integrities
B
dx.doi.org/10.1021/ja3074424 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
most likely due to poor adhesivity of Azo-6Glue onto the
liposomal surface.
We consider that the light-driven transmembrane ion
permeation, thus observed in Figure 4a (red circles), is most
likely due to dynamic structural disordering of the phospholipid
membrane by a photomechanical motion of adhering
Azo-18Glue (Figure 2b). Another possibility to consider is
that the adhesion of cis-Azo-18Glue just statically deteriorates
the phospholipid membrane for inducing the ion permeation.
However, this possibility was excluded by means of
fluorescence depolarization.13 Thus, by using DPH-embedded
EYPC liposomes (DPH: 1,6-diphenyl-1,3,5-hexatriene), we
investigated if the liposomal membrane might alter its fluidity
by the interaction with cis-Azo-18Glue.14 Photoexcited DPH at
336 nm is known to fluoresce at 450 nm. If the membrane
turns more fluid upon interaction with cis-Azo-18Glue,
embedded DPH could rotate more vigorously. As a possible
consequence of this, its polarized fluorescence, observable upon
photoexcitation with a 336-nm plane-polarized light,14 may be
depolarized to some extent. However, as shown in Figure 5,
Figure 4. Transmembrane ion permeation profiles of pyranine-
containing EYPC liposomes (35 μM; initial inner pH = 7.0) at 25 °C
in HEPES buffer (10 mM HEPES, 100 mM NaCl, pH 8.4) containing
Azo-18Glue (1.5 μM) or Azo-6Glue (4.5 μM) ([Gu+] = 27 μM). (a)
Figure 5. Static adhesion effects of Azo-18Glue on the fluidity of a
phospholipid membrane. Fluorescence depolarization profiles at 25 °C
of DPH (70 nM), embedded in an Azo-18Glue-attached EYPC
liposomal membrane (35 μM), relative to that without Azo-18Glue, in
HEPES buffer (10 mM HEPES, 100 mM NaCl, pH 7.0) containing
1.5 μM trans- (blue bar) or cis-rich12 Azo-18Glue (red bar).
Changes in relative fluorescence intensity I460/I405 at 510 nm (λext
=
460 and 405 nm for I460 and I405, respectively) upon on−off
photoirradiation of Azo-18Glue (red circles) and Azo-6Glue (blue
circles) with a 365-nm external UV light (blue-shaded areas).
Enhanced I460/I405 is the consequence of an increment of the inner
pH by OH− permeation, possibly by photomechanically induced
disordering of the liposomal membrane (Figure 2b). Note that the
sample in yellow-shaded areas was exposed to visible excitation lights
for fluorescence spectroscopy, so that cis-Azo-18Glue likely isomerized
back to its trans-form. This isomerization may also give rise to
disordering of the liposomal membrane (Figure 2b). (b) DLS
histograms of EYPC liposomes before (green) and after (red) 15
min photoirradiation of Azo-18Glue with a 365-nm UV light. (c) Flow
cytometry histograms (λext = 488 nm) of pyranine-containing giant
EYPC liposomes before (green) and after (red) 10-min photo-
irradiation of Azo-18Glue with a 365-nm UV light.
neither trans- nor cis-Azo-18Glue on the EYPC liposomes
altered the fluorescence anisotropy15 of embedded DPH.
Regardless of its configuration, adhering Azo-18Glue does not
statically deteriorate the phospholipid membrane (Figure 2a).
Finally, we point out that Figure 4a, upon closer look, shows
a continuous increase in internal pH for a short period (yellow-
shaded areas) after stopping the 365-nm UV irradiation. This
observation seems a bit intriguing but is consistent with our
claim, considering that the sample at this stage was kept
exposed to visible excitation lights for fluorescence spectrosco-
py, wherein cis-Azo-18Glue may be tempted to isomerize
backward. As discussed in Figure 2b, this isomerization can also
generate a photomechanical force for disordering the
phospholipid membrane.
(green histogram; before photoirradiation, red histogram; after
photoirradiation). By flow cytometry (Figure 4c, λext = 488
nm), we confirmed that the individual liposomes, after
continuous UV irradiation for 10 min, display an enhanced
fluorescence emission due to the interior pyranine dye (green
histogram; before photoirradiation, red histogram; after photo-
irradiation), again indicating the occurrence of light-driven
transmembrane permeation of OH− into the liposomal interior.
Lower-generation Azo-6Glue (Figure 1) presumably bears a
smaller conformational freedom than Azo-18Glue, thereby a
mechanical force may transmit more efficiently from the
azobenzene core to the adhesive points. However, photo-
isomerization of Azo-6Glue (4.5 μM) led to little activation of
transmembrane ion permeation (Figure 4a, blue circles). This is
In conclusion, we demonstrated that Azo-18Glue (Figure 1),
upon photoisomerization, can activate ion permeation across a
natural phospholipid membrane by photoisomerization.
Through detailed investigations, the activated ion permeation
is most likely due to dynamic structural disordering of the
phospholipid membrane, induced by a “photomechanical
motion” of adhering Azo-18Glue (Figure 2). This mechanism
is essentially different from that reported for an assembly of a
lipid molecule covalently attached to a photochromic module,16
where a photoisomerized lipid with a poor molecular packing
nature likely deteriorates the lipid membrane statically. The
concept of “photoresponsive molecular glues” is universal and
C
dx.doi.org/10.1021/ja3074424 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
(15) r = (IVV − G × IVH)/(IVV + 2G × IVH), where I represents
fluorescence intensity, subscripts V and H denote vertical and
horizontal orientations of the excitation and emission polarizers,
respectively, and G is given by IHV/IHH, which accounts for the relative
instrument sensitivity toward vertically and horizontally polarized
lights.
(16) (a) Bisby, R. Biochem. Biophys. Res. Commun. 2000, 276, 169−
173. (b) Hamada, T.; Sato, Y.; Yoshikawa, K.; Nagasaki, T. Langmuir
2005, 21, 7626−7628.
potentially applicable to live cells, because neither genetic nor
chemical modification of native cells is required.17
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthesis of Azo-18Glue, Azo-6Glue, and FITC-9Glue; NMR
and MALDI-TOF mass spectral data, flow cytometry,
fluorescence microscopy, electronic absorption spectroscopy,
and related experimental procedures. This material is available
free of charge via the Internet at http://pubs.acs.org.
(17) Examples of chemically modified ion channels with photo-
chromic molecules: (a) Bartels, E.; Wassermann, N. H.; Erlanger, B. F.
Proc. Natl. Acad. Sci. U.S.A. 1971, 68, 1820−1823. (b) Lien, L.;
Jaikaran, D.; Zhang, Z. H.; Woolley, G. A. J. Am. Chem. Soc. 1996, 118,
12222−12223. (c) Banghart, M.; Borges, K.; Isacoff, E.; Trauner, D.;
AUTHOR INFORMATION
■
Kramer, R. H. Nat. Neurosci. 2004, 7, 1381−1386. (d) Kocȩ r, A.;
Corresponding Author
Walko, M.; Meijberg, W.; Feringa, B. L. Science 2005, 309, 755−758.
(e) Loudwig, S.; Bayley, H. J. Am. Chem. Soc. 2006, 128, 12404−
12405. (f) Volgraf, M.; Gorostiza, P.; Numano, R.; Kramer, R. H.;
Isacoff, E. Y.; Trauner, D. Nat. Chem. Biol. 2006, 2, 47−52.
okuro@macro.t.u-tokyo.ac.jp (K.O.); aida@macro.t.u-tokyo.ac.
jp (T.A.)
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We acknowledge the Center for NanoBio Integration, the
University of Tokyo. This work was partially supported by
Grant-in-Aid for Research Activity Start-up (23850007) to K.O.
REFERENCES
■
(1) (a) Huxley, A. F.; Niedergerke, R. Nature 1954, 173, 971−973.
(b) Huxley, H.; Hanson, J. Nature 1954, 173, 973−976.
(2) (a) Setiadi, H.; Sedgewick, G.; Erlandsen, S. L.; McEver, R. P. J.
Cell Biol. 1998, 142, 859−871. (b) Wymer, N.; Gerasimov, O.;
Thompson, D. Bioconjugate Chem. 1998, 9, 305−308.
(3) (a) Fenton, W. A.; Horwich, A. L. Q. Rev. Biophys. 2003, 36,
229−256. (b) Silflow, C. D.; Lefebvre, P. A. Plant Physiol. 2001, 127,
1500−1507.
(4) (a) Okuro, K.; Kinbara, K.; Tsumoto, K.; Ishii, N.; Aida, T. J. Am.
Chem. Soc. 2009, 131, 1626−1627. (b) Okuro, K.; Kinbara, K.;
Takeda, K.; Inoue, Y.; Ishijima, A.; Aida, T. Angew. Chem., Int. Ed.
2010, 49, 3030−3033. (c) Wang, Q.; Mynar, J. L.; Yoshida, M.; Lee,
E.; Lee, M.; Okuro, K.; Kinbara, K.; Aida, T. Nature 2010, 463, 339−
343.
(5) (a) Chen, H.; Neerman, M. F.; Parrish, A. R.; Simanek, E. E. J.
Am. Chem. Soc. 2004, 126, 10044−10048. (b) Wender, P. A.; Kreider,
E.; Pelkey, E. T.; Rothbard, J.; VanDeusen, C. L. Org. Lett. 2005, 7,
4815−4818.
(6) Tamai, N.; Miyasaka, H. Chem. Rev. 2000, 100, 1875−1890.
(7) See Supporting Information.
(8) (a) Sakai, N.; Kamikawa, Y.; Nishii, M.; Matsuoka, T.; Kato, T.;
Matile, S. J. Am. Chem. Soc. 2006, 128, 2218−2219. (b) Izzo, I.; Licen,
S.; Maulucci, N.; Autore, G.; Marzocco, S.; Tecilla, P.; De Riccardis, F.
Chem. Commun. 2008, 2986−2988.
(9) (a) Hope, M.; Bally, M.; Mayer, L.; Janoff, A.; Cullis, P. Chem.
Phys. Lipids 1986, 40, 89−107. (b) Hope, M.; Bally, M.; Webb, G.;
Cullis, P. Biochim. Biophys. Acta 1985, 812, 55−65. (c) Pick, U. Arch.
Biochem. Biophys. 1981, 212, 186−194.
(10) Ulysse, L.; Cubillos, J.; Chmielewski, J. J. Am. Chem. Soc. 1995,
117, 8466−8467.
(11) Clement, N. R.; Gould, J. M. Biochemistry 1981, 20, 1534−1538.
(12) There exist experimental difficulties in determining the
isomerization degree of Azo-18Glue. However, its non-dendritic
version was reported to furnish the trans/cis ratio of 25/75 at the
photostationary state. For details, see: Gorostiza, P.; Volgraf, M.;
Numano, R.; Szobota, S.; Trauner, D.; Isacoff, E. Y. Proc. Natl. Acad.
Sci. U.S.A. 2007, 104, 10865−10870.
(13) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed;
Kluwer Academic: New York, 1999; pp 353−381.
(14) Bothun, G. D.; Knutson, B. L.; Strobel, H. J.; Nokes, S. E.
Langmuir 2005, 21, 530−536.
D
dx.doi.org/10.1021/ja3074424 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX