A red-shifted, fast-relaxing azobenzene photoswitch for visible light control of an ionotropic glutamate receptor

Article abstract of DOI:10.1021/ja408104w

The use of azobenzene photoswitches has become a dependable method for rapid and exact modulation of biological processes and material science systems. The requirement of ultraviolet light for azobenzene isomerization is not ideal for biological systems due to poor tissue penetration and potentially damaging effects. While modified azobenzene cores with a red-shifted cis-to-trans isomerization have been previously described, they have not yet been incorporated into a powerful method to control protein function: the photoswitchable tethered ligand (PTL) approach. We report the synthesis and characterization of a red-shifted PTL, L-MAG0₄₆₀, for the light-gated ionotropic glutamate receptor LiGluR. In cultured mammalian cells, the LiGluR+L-MAG0₄₆₀ system is activated rapidly by illumination with 400-520 nm light to generate a large ionic current. The current rapidly turns off in the dark as the PTL relaxes thermally back to the trans configuration. The visible light excitation and single-wavelength behavior considerably simplify use and should improve utilization in tissue.

Full text of DOI:10.1021/ja408104w

Communication
pubs.acs.org/JACS
A Red-Shifted, Fast-Relaxing Azobenzene Photoswitch for Visible
Light Control of an Ionotropic Glutamate Receptor
Michael A. Kienzler,†,‡,§ Andreas Reiner,†,§ Eric Trautman,† Stan Yoo,† Dirk Trauner,*,‡
and Ehud Y. Isacoff*,†
†Department of Molecular and Cell Biology, University of California, Berkeley, California 94720, United States
‡Department of Chemistry and Center for Integrated Protein Science, Ludwig-Maximilians-Universitat, Butenandtstraße 5-13,
̈
Munich 81377, Germany
S
* Supporting Information
ABSTRACT: The use of azobenzene photoswitches has
become a dependable method for rapid and exact
modulation of biological processes and material science
systems. The requirement of ultraviolet light for
azobenzene isomerization is not ideal for biological
systems due to poor tissue penetration and potentially
damaging effects. While modified azobenzene cores with a
red-shifted cis-to-trans isomerization have been previously
described, they have not yet been incorporated into a
powerful method to control protein function: the
photoswitchable tethered ligand (PTL) approach. We
report the synthesis and characterization of a red-shifted
PTL, L-MAG0460, for the light-gated ionotropic glutamate
receptor LiGluR. In cultured mammalian cells, the LiGluR
+L-MAG0460 system is activated rapidly by illumination
with 400−520 nm light to generate a large ionic current.
The current rapidly turns off in the dark as the PTL relaxes
thermally back to the trans configuration. The visible light
excitation and single-wavelength behavior considerably
simplify use and should improve utilization in tissue.
he development of synthetic photoswitches has been a
boon to researchers in the material and biological sciences
T
due to the precise spatiotemporal control that light allows.1
Azobenzenes in particular have proved to be robust photo-
switches that tolerate significant chemical modifications. For
applications in neuroscience, we have developed a two-
component approach that we refer to as chemical optoge-
netics.2,3 Therein, a photoswitchable tethered ligand (PTL),
usually derived from an azobenzene, is covalently attached to a
protein to enable its reversible activation or block in response
to flashes of light.4,5 PTLs, for example when attached to
neurotransmitter-gated ion channels and receptors, can be used
to manipulate neuronal signaling in living cells and organisms.
While they can be used like microbial opsins, which function as
retinal-dependent pumps or ion channels,6,7 to excite and
inhibit neuronal firing with light, they offer the unique
additional advantage of targeting native transmitter systems
that control synaptic strength and plasticity, which are thought
to be key for circuit function and memory formation.8
Figure 1. Photoswitchable tethered ligands enable ligand-gated ion
channels to be controlled by light. (a) Cartoon showing light-gating,
with a shorter wavelength inducing isomerization to cis, followed by
ligand binding and channel opening, while a longer wavelength
reverses the process. This can also occur thermally (kBT). (b) Modular
design of maleimide−azobenzene−glutamate (MAG) photoswitches
exemplified by regular L-MAG0 (1). (c) Structure of the red-shifted
PTL L-MAG0460 (2).
are composed of three parts, maleimide−azobenzene−gluta-
mate (MAG), such as L-MAG0 (Figure 1b, 1), which bind
covalently at their maleimide end to an engineered cysteine
We focus here on a family of PTLs that use the excitatory
neurotransmitter glutamate as a ligand to control ionotropic or
metabotropic glutamate receptors (Figure 1a).9,10 These PTLs
Received: August 9, 2013
Published: October 30, 2013
© 2013 American Chemical Society
17683
dx.doi.org/10.1021/ja408104w | J. Am. Chem. Soc. 2013, 135, 17683−17686

Journal of the American Chemical Society
Communication
introduced into the ligand binding domain of a homotetrameric
kainate receptor GluK2 (iGluR6) to generate the light-activated
“LiGluR”.9,11 Irradiation with 380 nm light isomerizes the
azobenzene core from the more stable trans state to the
metastable cis state. This docks the glutamate into its binding
pocket and opens the channel. Irradiation at ∼500 nm reverses
the processes and closes the channel.11 Even in cases where
labeling, photoswitching, or the ligand efficacy remain
submaximal, LiGluR constitutes a powerful tool for in vivo
studies, as low-affinity kainate receptors like GluK2 are not fully
activated in many physiological conditions either (see SI).
LiGluR has been used to evoke patterns of action potentials in
neurons,12 reproducibly inject calcium into glial cells and
chromaffin cells to evoke transmitter release,13,14 excite specific
cells for neural circuit analysis in vivo,15 and restore a retinal
light response and visual behavior to mice blinded by
photoreceptor cell degeneration.16
acceptor red-shift, but with minimal variation to the structure
so that it would resemble L-MAG0 (1) as closely as possible.
Modification of the acetamido-group in the 4′-position to an
electron-donating tertiary amine, in conjunction with the 4-
acetamido group, was expected to yield a sufficiently strong
push−pull system.
The readily available industrial dye “disperse red 1” (3,
Scheme 1) was chosen as our azobenzene precursor. Using an
a
Scheme 1. Synthesis of L-MAG0460 (2)
Two properties of L-MAG0 (1) and other “regular”
azobenzene switches impose limitations for neuroscience
applications. First, the most efficient isomerization of 1 from
trans to cis is triggered by illumination at 380 nm, i.e., in the
UV-A range. Ultraviolet light is problematic in biological
systems for several reasons: prolonged exposure can be
damaging, it penetrates mammalian tissue poorly, and it does
not transmit through the lens of the human eye, for which
visual restoration depends on detection in the visible range
(∼400−680 nm). The second property is the bistability of the
“regular” azobenzene core of L-MAG0 (1). Although the cis
isomer thermally relaxes to the lower energy trans form over
time, the half-life is on the order of minutes to hours. For
neuroscience applications, this requires a second aligned light
source at the longer wavelength to rapidly pump the switch
back into the trans state. For application to vision restoration,
this comparatively slow thermal relaxation prevents the system
from mimicking the spontaneous turn-off of the native
rhodopsin of the photoreceptor cell. The requirement for a
second wavelength to return from the metastable cis state also
means that a substantial part of the spectrum has an impact on
the PTL, reducing room for optical reporters with which one
would want to combine the light-gated system. These problems
would all be overcome with a modified azobenzene core that is
“activated” at one wavelength in the visible part of the spectrum
and then rapidly relaxes back to the “inactive” trans state in the
dark.
The literature on azobenzenes suggests increasing electron
density as a strategy to red-shift the absorption and lower the
energy barrier for the cis-to-trans isomerization.17,18 Both effects
are achieved in the “amino” and “push−pull” azobenzenes.
Push−pull azobenzenes are so-called because one benzene ring
is ortho or para substituted with an electron-donating group,
while the other ring is substituted with an electron withdrawing
group. This both red-shifts the absorbance of the trans isomer
by ∼100 nm and greatly decreases the thermal stability of the
cis isomer. Initially, these concepts were applied to azobenzene
cross-linkers of peptide helices.19−22 Additionally, diffusible
photochromic ligands modified to be push−pull have been
made as reversibly caged potassium channel blockers23−25 and
AMPA receptor agonists,26 but substituted azobenzenes have
not yet been tested as PTLs. Potential problems include steric
hindrance to operation of the tethered ligand and slowing of
thermal relaxation due to binding interaction of the PTL ligand
headgroup. With this in mind, we designed the compound L-
MAG0460 (Figure 1c, 2) to obtain the advantage of the donor−
a
Reagents and conditions: (a) CBr4, PPh3, CH2Cl2, 0 °C (90%); (b)
potassium phthalimide, DMF, 90 °C; (c) H2NNH2, EtOH, 60 °C
(68%, two steps); (d) Na2S, 1,4-dioxane, H2O, 90 °C (82%); (e)
FmocCl, DIPEA, CH2Cl2 (46%); (f) 5, HOBt, EDCl, DIPEA, MeCN
(53%); (g) piperidine, THF; (h) 1 M LiOH(aq), (40%, two steps); (i)
N-(methoxycarbonyl) maleimide, NaHCO3 H2O, THF, 0 °C; (j) HCl,
EtOAc (90%, two steps).
Appel reaction, we converted the hydroxyl group on 3 into a
bromide and in a subsequent two-step Gabriel synthesis to a
free amine. The selective reduction of the aromatic nitro group
with sodium sulfide yielded azoaniline 4. Protection of the
more nucleophilic primary amine in 4 with a fluorenylmethy-
loxycarbonyl (Fmoc) group, followed by amide coupling with
known pyroglutamate derivative 5,9 yielded the advanced
intermediate 6. Deprotection of the primary amine was
followed by hydrolysis of the pyroglutamate with concomitant
saponification of the ethyl ester, and then installation of the
maleimide. Finally, acidic removal of the tert-butyloxycarbonyl
(Boc) protecting group provided L-MAG0460 (2) as its
hydrochloride salt.
As expected, changing the 4-acetamido group to a dialkyl-
amine in 2 resulted in a strong (∼100 nm) red-shift of the
absorption, with a maximum at 462 nm in aqueous solution
(Figure 2b). To characterize the photoswitching properties of 2
and to test its applicability as a PTL, we labeled the ionotropic
glutamate receptor GluK2. For attachment we chose the
position L439C where regular L-MAG0 (1) acts as a highly
efficient cis agonist.9,11,27,28 GluK2(439C) receptors expressed
17684
dx.doi.org/10.1021/ja408104w | J. Am. Chem. Soc. 2013, 135, 17683−17686

Journal of the American Chemical Society
Communication
Figure 3. (a) Activation by 445 nm light and fast thermal relaxation in
the dark (τmean = 0.71 s (n = 3)). (b) Activation by 445 nm light,
unfiltered white light and 500 nm light.
We wondered whether the broad action spectrum of 2 in the
blue/green region, in combination with the minimal absorption
at higher wavelengths (Figure 2b) would enable activation by
white light. Indeed, we found that the unfiltered polychromatic
output of our Xe-lamp light source gave a high level of
activation, close to the maximal activation observed with
monochromatic 445 nm light (Figure 3b). A final important
feature is that LigluR+2 maintains fidelity over hundreds of
switching cycles (Figure S2).
Thus, LiGluR+2 acts as a single-wavelength photoswitch that
supports repeated activation with pulses of light and can be
activated by either narrow bands of light or light sources with
broad emission spectra. It needs no illumination to be
deactivated and deactivates rapidly enough to match the time
course of many signaling events. These properties make 2 a
versatile PTL, which maintains the same molecular specificity as
1 for genetically engineered glutamate receptors.
In summary, we report what is to our knowledge the
synthesis of the first PTL with a red-shifted azobenzene core. L-
MAG0460 (2) operates as a potent photoswitch for LiGluR that
is made from the calcium permeant, excitatory kainate receptor
GluK2. It generates large and reproducible currents when
irradiated with blue-green light and spontaneously turns off in
the dark. This fast, spontaneous deactivation is an important
functional feature that could be generally applicable to PTLs
with low-affinity ligands. The red-shift and single-wavelength
dependence should enable deeper penetration in brain tissue
while minimizing phototoxicity and achieving control with only
one light source. An exciting future application will be to build
on the demonstration that LiGluR+1 restores light sensitivity to
the retina in a mouse disease model that causes photoreceptor
degeneration.16 The broad excitation at visible wavelengths,
coupled to the spontaneous turning off in the dark, suggests
that L-MAG0460 (2), which has an activation spectrum similar
to blue cone photoreceptors (S-cones), is the better photo-
switch for this application. Finally, due to the inherent
modularity in the synthesis of MAG-type PTLs, other variants
with this red-shifted azobenzene core should be straightforward
Figure 2. (a) Representative action spectrum showing the activation of
LiGluR+2 with visible light. (b) Comparison of the absorption spectra
of 1 and 2 (in phosphate-buffered saline, pH 7.4) and the LiGluR+2
action spectrum (as in (a)) obtained from voltage-clamp recordings
(mean SEM, n = 4).
on the surface of HEK293 cells were covalently labeled by
incubation with the red-shifted L-MAG0460 (2) and excess
reagent was removed by washing. Patch clamp electro-
physiology was used to assess the photoinduced activation of
LiGluR with covalently tethered 2.
Illumination with blue light readily induced channel opening,
resulting in large inward currents (Figure 2a). This
demonstrates that excitation of L-MAG0460 (2) significantly
populates the cis configuration and that the structural
modification at the 4′-position of the azobenzene does not
impede ligand-binding or the closure of the ligand binding
domain that is required for the channel to activate and open.
Varying the irradiation wavelength allowed us to record an
action spectrum, which shows a broad plateau between 440 and
480 nm, along with some activation by green light. The degree
of LiGluR+2 activation across the spectrum closely follows the
absorption spectrum of the trans form (Figure 2b). The
exception to this agreement was at the peak at 445 nma
difference that may reflect the polarity of the local environment
surrounding L-MAG0460 (2) once it has covalently bound to
LiGluR. Polar solvents often red-shift the absorbance of
azobenzenes,17 as we indeed observed with L-MAG0460 (Figure
S1).
In stark contrast to regular L-MAG0 (1), which in the dark
remains in its cis configuration for extended times (τ = 25.5
min11), red-shifted L-MAG0460 (2) relaxes back to trans to
close the channel on a much shorter time scale when the light is
turned off (τmean = 0.71 s (n = 3), Figure 3a). This shows that
the thermal relaxation of 2 bound to the glutamate receptor is
>2200 times faster compared to 1.
17685
dx.doi.org/10.1021/ja408104w | J. Am. Chem. Soc. 2013, 135, 17683−17686

Journal of the American Chemical Society
Communication
(16) Caporale, N.; Kolstad, K. D.; Lee, T.; Tochitsky, I.; Dalkara, D.;
Trauner, D.; Kramer, R.; Dan, Y.; Isacoff, E. Y.; Flannery, J. G. Mol.
Ther. 2011, 19, 1212.
(17) Knoll, H. In CRC Handbook of Organic Photochemistry and
Photobiology, 2nd ed.; CRC Press: Boca Raton, FL, 2003; p 89.1.
(18) Bandara, H. M. D.; Burdette, S. C. Chem. Soc. Rev. 2012, 41,
1809.
to synthesize. Indeed, since PTLs attach to the surface of the
signaling protein, it makes sense that they would be tolerant to
substitutions that can tune their spectral and kinetic properties,
offering a direct and rational approach to modifying protein
function. This will provide a valuable expansion of the chemical
optogenetic toolset for the remote control of ion channels and
cellular signaling cascades.
(19) Beharry, A. A.; Sadovski, O.; Woolley, G. A. J. Am. Chem. Soc.
2011, 133, 19684.
(20) Chi, L.; Sadovski, O.; Woolley, G. A. Bioconjugate Chem. 2006,
17, 670.
(21) Sadovski, O.; Beharry, A. A.; Zhang, F. Z.; Woolley, G. A. Angew.
Chem., Int. Ed. 2009, 48, 1484.
(22) Samanta, S.; Beharry, A. A.; Sadovski, O.; McCormick, T. M.;
Babalhavaeji, A.; Tropepe, V.; Woolley, G. A. J. Am. Chem. Soc. 2013,
135, 9777.
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures, characterization for compounds,
methods of receptor transfection, and illumination protocols.
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
S
(23) Mourot, A.; Kienzler, M. A.; Banghart, M. R.; Fehrentz, T.;
Huber, F. M. E.; Stein, M.; Kramer, R. H.; Trauner, D. ACS Chem.
Neurosci. 2011, 2, 536.
(24) Banghart, M. R.; Mourot, A.; Fortin, D. L.; Yao, J. Z.; Kramer, R.
H.; Trauner, D. Angew. Chem., Int. Ed. 2009, 48, 9097.
(25) Fortin, D. L.; Banghart, M. R.; Dunn, T. W.; Borges, K.;
Wagenaar, D. A.; Gaudry, Q.; Karakossian, M. H.; Otis, T. S.; Kristan,
W. B.; Trauner, D.; Kramer, R. H. Nat. Methods 2008, 5, 331.
(26) Stawski, P.; Sumser, M.; Trauner, D. Angew. Chem., Int. Ed.
2012, 51, 5748.
AUTHOR INFORMATION
■
Corresponding Authors
dirk.trauner@lmu.de
ehud@berkeley.edu
Author Contributions
§M.A.K. and A.R. contributed equally.
Notes
The authors declare no competing financial interest.
(27) Numano, R.; Szobota, S.; Lau, A. Y.; Gorostiza, P.; Volgraf, M.;
Roux, B.; Trauner, D.; Isacoff, E. Y. Proc. Natl. Acad. Sci. U.S.A. 2009,
106, 6814.
ACKNOWLEDGMENTS
■
(28) Reiner, A.; Isacoff, E. Methods Mol. Biol. 2013, in press.
This work was supported by a Deutsche Forschungsgemein-
schaft grant (FOR 1279 to D.T.) and postdoctoral fellowship
(DFG RE3101/1 to A.R.), the National Institutes of Health
Nanomedicine Development Center for the Optical Control of
Biological Function (PN2EY018241 to D.T. and E.Y.I.), and an
NRSA postdoctoral fellowship (1F32EY022840 to M.A.K.).
REFERENCES
■
(1) Szymanski, W.; Beierle, J. M.; Kistemaker, H. A.; Velema, W. A.;
Feringa, B. L. Chem. Rev. 2013, 113, 6114.
(2) Fehrentz, T.; Schonberger, M.; Trauner, D. Angew. Chem., Int. Ed.
̈
2011, 50, 12156.
(3) Kramer, R. H.; Mourot, A.; Adesnik, H. Nat. Neurosci. 2013, 16,
816.
(4) Banghart, M.; Borges, K.; Isacoff, E.; Trauner, D.; Kramer, R. H.
Nat. Neurosci. 2004, 7, 1381.
(5) Gorostiza, P.; Isacoff, E. Y. Science 2008, 322, 395.
(6) Zhang, F.; Vierock, J.; Yizhar, O.; Fenno, L. E.; Tsunoda, S.;
Kianianmomeni, A.; Prigge, M.; Berndt, A.; Cushman, J.; Polle, J.;
Magnuson, J.; Hegemann, P.; Deisseroth, K. Cell 2011, 147, 1446.
(7) Miesenbock, G. Annu. Rev. Cell. Dev. Biol. 2011, 27, 731.
̈
(8) Feldman, D. E. Annu. Rev. Neurosci. 2009, 32, 33.
(9) Volgraf, M.; Gorostiza, P.; Numano, R.; Kramer, R. H.; Isacoff, E.
Y.; Trauner, D. Nat. Chem. Biol. 2006, 2, 47.
(10) Levitz, J.; Pantoja, C.; Gaub, B.; Janovjak, H.; Reiner, A.;
Hoagland, A.; Schoppik, D.; Kane, B.; Stawski, P.; Schier, A. F.;
Trauner, D.; Isacoff, E. Y. Nat. Neurosci. 2013, 16, 507.
(11) Gorostiza, P.; Volgraf, M.; Numano, R.; Szobota, S.; Trauner,
D.; Isacoff, E. Y. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 10865.
(12) Szobota, S.; Gorostiza, P.; Del Bene, F.; Wyart, C.; Fortin, D. L.;
Kolstad, K. D.; Tulyathan, O.; Volgraf, M.; Numano, R.; Aaron, H. L.;
Scott, E. K.; Kramer, R. H.; Flannery, J.; Baier, H.; Trauner, D.; Isacoff,
E. Y. Neuron 2007, 54, 535.
(13) Li, D. D.; Herault, K.; Isacoff, E. Y.; Oheim, M.; Ropert, N. J.
Physiol. 2012, 590, 855.
(14) Izquierdo-Serra, M.; Trauner, D.; Llobet, A.; Gorostiza, P.
Biochim. Biophys. Acta 2013, 1830, 2853.
(15) Wyart, C.; Del Bene, F.; Warp, E.; Scott, E. K.; Trauner, D.;
Baier, H.; Isacoff, E. Y. Nature 2009, 461, 407.
17686
dx.doi.org/10.1021/ja408104w | J. Am. Chem. Soc. 2013, 135, 17683−17686