A semisynthetic fluorescent sensor protein for glutamate

Article abstract of DOI:10.1021/ja3002277

We report the semisynthesis of a fluorescent glutamate sensor protein on cell surfaces. Sensor excitation at 547 nm yields a glutamate-dependent emission spectrum between 550 and 700 nm that can be exploited for ratiometric sensing. On cells, the sensor displays a ratiometric change of 1.56. The high sensitivity toward glutamate concentration changes of the sensor and its exclusive extracellular localization make it an attractive tool for glutamate sensing in neurobiology.

Full text of DOI:10.1021/ja3002277

Communication
pubs.acs.org/JACS
A Semisynthetic Fluorescent Sensor Protein for Glutamate
Matthias A. Brun, Kui-Thong Tan,† Rudolf Griss, Anna Kielkowska, Luc Reymond, and Kai Johnsson*
Institute of Chemical Sciences and Engineering, Institute of Bioengineering, National Centre of Competence in Research Chemical
́
́ ́
Biology, Ecole Polytechnique Federale de Lausanne (EPFL), Lausanne, Switzerland
S
* Supporting Information
ABSTRACT: We report the semisynthesis of a fluores-
cent glutamate sensor protein on cell surfaces. Sensor
excitation at 547 nm yields a glutamate-dependent
emission spectrum between 550 and 700 nm that can be
exploited for ratiometric sensing. On cells, the sensor
displays a ratiometric change of 1.56. The high sensitivity
toward glutamate concentration changes of the sensor and
its exclusive extracellular localization make it an attractive
tool for glutamate sensing in neurobiology.
he amino acid glutamate is the prevalent neurotransmitter
T
in the vertebrate nervous system. It is used at well over
90% of the synapses in the human brain and influences
essentially all forms of behavior, including consciousness,
sensory perception, motor control, and mood.1 Further,
glutamate is involved at most synapses that are modifiable,
that is, that are capable of adapting to changing patterns of
stimuli by enhancing or reducing the efficiency of synaptic
transmission.2 These processes are thought to be responsible
for high-order brain functions, such as learning and memory.
Three fluorescent sensor proteins for investigating the role of
glutamate in neurobiology have been developed so far.3−5
However, the modest performance of these sensor proteins has
limited their use. Here we present a semisynthetic fluorescent
sensor protein for glutamate which shows higher sensitivity
toward glutamate concentration changes and operates at longer
wavelengths than the previously reported sensors.
Figure 1. A Snifit for glutamate. (A) Design principle of the
semisynthetic glutamate sensor protein. The protein part of the sensor
is a fusion protein of SNAP-tag, CLIP-tag, and the glutamate binding
protein iGluR5-S1S2 (iGluR5). The active semisynthetic sensor
protein is obtained by labeling SNAP-tag with a molecule containing
a Cy5 fluorophore (red star) and a terminal tethered glutamate
analogue (gray ball), and by labeling CLIP-tag with a DY-547
fluorophore (green star). Glutamate (pink ball) at appropriate
concentrations displaces the intramolecular ligand. (B) Structure of
BG-PEG11-Cy5-glutamate 1 and the corresponding control molecule 2
for sensor protein labeling.
Semisynthetic fluorescent sensor proteins (Snifits),6,7 are
fusion proteins consisting of SNAP-tag,8 CLIP-tag,9 and an
analyte-binding protein (Figure 1A). SNAP- and CLIP-tag are
labeled with a synthetic fluorescent ligand and a second
synthetic fluorophore, respectively. The ligand binds to the
binding protein in an intramolecular fashion and thereby keeps
the sensor protein in a closed conformation. Free analyte can
compete for binding to the binding protein and can shift the
equilibrium to the open conformation. This shift can be
sensor development. Second, it is known that the stereo-
selective functionalization of the γ-carbon of the glutamate side
chain does not significantly perturb its affinity toward
iGluR5,12,13 suggesting an attachment point for the required
synthetic tether. We therefore prepared the tethered glutamate
analogue 1 (Figure 1B, Scheme S1−S3) that contains a Cy5
fluorophore and an O6-benzylguanine (BG) group for reaction
with SNAP-tag. To confirm that our tethered glutamate
derivative 1 binds to S1S2, we prepared a derivative without
BG (Figure S1A) and tested its binding to purified recombinant
S1S2 of iGluR5 by using fluorescence polarization (Figure S2).
Assuming a similar effective molarity of the intramolecular
ligand as in our previous work,6,7 the measured Kd value of 7
1 μM should enable the tethered glutamate 1 to keep the
sensor in a closed confirmation in the absence of glutamate.
Next, we constructed the glutamate sensor protein
SNAP_PP15_CLIP_iGluR5-S1S2 (abbreviated as Snifit-
detected by a change in the Forster resonance energy transfer
̈
(FRET) efficiency between the two fluorophores.
For the construction of a glutamate sensor protein based on
the Snifit sensor concept, we have chosen the ionotropic
glutamate receptor 5 (iGluR5) as the binding protein for two
reasons. First, due to the modular construction of ionotropic
glutamate receptors, it is possible to express its glutamate
binding domain S1S2 as a soluble protein in bacteria while
conserving both the high affinity and specificity toward
glutamate.10,11 The possibility to characterize the soluble
binding protein in vitro before its use on cell surfaces facilitates
Received: January 9, 2012
Published: April 23, 2012
© 2012 American Chemical Society
7676
dx.doi.org/10.1021/ja3002277 | J. Am. Chem. Soc. 2012, 134, 7676−7678

Journal of the American Chemical Society
Communication
iGluR5). Snifit-iGluR5 is a linear fusion protein of SNAP-tag,
CLIP-tag, and the glutamate binding protein iGluR5-S1S2, with
a polyproline linker inserted between SNAP-tag and CLIP-tag.
We have shown in our previous work that the inclusion of a
rigid polyproline linker between SNAP- and CLIP-tag leads to
greatly improved sensor proteins in terms of their maximum
ratio changes.7 The polyproline linker increases the distance
between the two fluorophores in the open state but not in the
closed state of the sensor protein, which leads to a net increase
of the sensor’s dynamic range.7 The fusion protein was
successfully expressed and purified from Escherichia coli Rosetta
gami and labeled with the molecules BG-PEG11-Cy5-glutamate
1 and CLIP-Surface 547 (Figure S1B), which is a substrate for
labeling CLIP-tag with the fluorescent dye DY-547 (Figure S3).
We found that the addition of 1 mM glutamate led to an
increase of fluorescence emission of DY-547 and to a decrease
of Cy5 fluorescence, indicating decreased FRET efficiency
(Figure 2A). This behavior of Snifit-iGluR5 confirms the
confirmed the presence of the sensor protein on the cell surface
(Figure 3). The use of membrane impermeable dyes results in
Figure 3. Fluorescence labeling of Snifit-iGluR5 on HEK 293T cells.
SNAP_PP15_CLIP_iGluR5-S1S2 was labeled on the cell surface of
HEK 293T cells with the dyes DY-547 (A) and DY-647 (B) via their
corresponding O2-benzylcytosine- (BC) and BG-derivatives (Figure
S1). (C) Overlay of DY-547 channel, DY-647 channel, and
transmission channel. The white scale bar corresponds to 20 μm.
Images were taken using a Leica SP5 WLL confocal microscope.
an exclusive labeling of Snifit-iGlurR5 on the cell surface, which
should facilitate the specific sensing of secreted neuro-
transmitters.
Next, we investigated the response of Snifit-iGluR5 labeled
with BG-PEG11-Cy5-glutamate 1 and with CLIP-Surface 547
upon perfusion of the cells with 1 mM glutamate and found a
reversible ΔRmax of 1.56 0.05 (Figure 4A, Figure S6). This
Figure 2. Snifit-iGluR5 as a sensor protein for glutamate. (A) Emission
spectra of Snifit-iGluR5 at low and high concentrations of glutamate
(10 nM, black; 10 mM, white). The addition of glutamate leads to an
increase in the DY-547/Cy5 emission ratio with a maximum ratio
change of 1.93
0.11. (B) Fluorescence titration curve of Snifit-
iGluR5. Shown is the ratio of fluorescence donor (570 nm) and
acceptor emission (670 nm) obtained by titrating the glutamate sensor
protein with glutamate. The data are represented as the mean
standard deviation of triplicates. Data are fitted according to a single-
site binding isotherm.
general validity of our sensor concept.7 The maximum ratio
change ΔRmax of Snifit-iGluR5 amounts to 1.93
0.11. By
fitting the glutamate titration data to a single binding isotherm,
comp,glutamate
we obtained a Kd
= 12 6 μM (Figure 2B). Snifit-
iGluR5 labeled with BG-PEG11-Cy5 2 and CLIP-Surface 547
did not show any ratio change upon addition of glutamate
(Figure S4, Scheme S4), supporting the hypothesis that the
ratio change in Snifit-iGluR5 labeled with 1 is due to a specific
displacement of the intramolecular ligand by free glutamate.
Snifit sensor proteins are very variable in terms of the
employed fluorophore.6,7 To prove this versatility also for our
Snifit-iGluR5, we labeled the sensor protein with BG-PEG11-
Cy5-glutamate 1 and with BC-Cy7 (Figure S1D) which is a
substrate for labeling CLIP-tag with the near-infrared dye Cy7.
The combination of the two dyes Cy5 and Cy7 also led to a
functional sensor protein, albeit with a slightly decreased ΔRmax
Figure 4. The sensor protein Snifit-iGluR5 on the surface of HEK
293T cells. (A) Time course of the intensity ratio of donor emission vs
acceptor emission upon addition and removal of 1 mM glutamate. The
red bar indicates the time span of perfusion with glutamate. (B) Time
course of perfusion of labeled Snifit-iGluR5 with increasing
concentrations of glutamate. The red bar indicates the time span of
perfusion with glutamate. Concentrations from left to right: 100 nM, 1
μM, 5 μM, 10 μM, 50 μM, 100 μM, 1 mM. (C) Glutamate titration
curve of Snifit-iGluR5 on the extracellular surface of HEK 293T cells.
value is more than 20% larger than those reported for the
previously described genetically encoded glutamate sensor
proteins.3−5 In contrast, control experiments using the sensor
protein labeled with BG-PEG11-Cy5 and CLIP-Surface 547 did
not result in any intensity ratio change (Figure S7). We then
also determined the ΔRmax for cells expressing the sensor
protein at different densities on their cell surface in order to
investigate whether the ΔRmax would depend on the expression
level of the sensor protein. We found that the ΔRmax remains
constant for a wide range of different expression levels (Figure
S8A). Only for cells that express the sensor protein at a level
of 1.37
0.04 Figure S5). For this reason, and since
microscopes equipped with Cy7 filters are not very common,
we decided to use the DY-547/Cy5 FRET pair for all further
experiments.
We then performed the semisynthesis of our Snifit-iGluR5
on the surface of mammalian cells. To this end, we linked
Snifit-iGluR5 to a truncated PDGF receptor (pDisplay) that
then displays the sensor on the extracellular surface. The
labeling of HEK 293T cells expressing Snifit-iGluR5 with
SNAP-Surface 647 (2 μM) and CLIP-Surface 547 (10 μM)
7677
dx.doi.org/10.1021/ja3002277 | J. Am. Chem. Soc. 2012, 134, 7676−7678

Journal of the American Chemical Society
Communication
that approaches the detection limit, a decreased ΔRmax was
observed (Figure S8A). The reduced ΔRmax at very low
expression levels can be explained by the increasing relative
contribution of background fluorescence. Nevertheless, our
analysis shows that the Snifit-iGluR5 also performs reliably
even at the detection limit of the used fluorophores. This
property of our sensor protein might be advantageous in
experimental settings where a high expression level of sensor
protein is problematic.
AUTHOR INFORMATION
Corresponding Author
kai.johnsson@epfl.ch
■
Present Address
†Department of Chemistry, National Tsing Hua University,
Hsinchu 30013, Taiwan.
Notes
The authors declare no competing financial interest.
To test the selectivity of Snifit-iGluR5 for glutamate, we
measured the response of our sensor protein toward other
physiologically relevant amino acids and neurotransmitters.
Sequential perfusion of cells expressing Snifit-iGluR5 on the
cell surface with aspartate, GABA, glycine, and glutamate only
resulted in a detectable ratio change when cells were perfused
with glutamate, demonstrating the specificity of the sensor
(Figure S9).
ACKNOWLEDGMENTS
■
This work was supported by funding from the Swiss National
Science Foundation and EPFL. The authors thank Dr. M. L.
Mayer for kindly providing us with the plasmid encoding the
iGluR5-S1S2 sequence and A. Masharina for technical
assistance.
REFERENCES
■
We then determined the affinity of our Snifit for glutamate in
(1) Watkins, J. C.; Evans, R. H. Annu. Rev. Pharmacol. Toxicol. 1981,
21, 165−204.
comp,glutamate
single cell measurements and found a Kd
μM (Figure 4B,C).
of 15
5
(2) Kandel, E. R. Science 2001, 294 (5544), 1030−1038.
(3) Okumoto, S.; Looger, L.; Micheva, K.; Reimer, R.; Smith, S.;
Frommer, W. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (24), 8740−8745.
(4) Hires, S.; Zhu, Y.; Tsien, R. Proc. Natl. Acad. Sci. U.S.A. 2008, 105
(11), 4411−4416.
As in our previous study,7 the Kdcomp,glutamate does not change
for different cells expressing the sensor protein at varying
concentrations on the cell surface (Figure S8B). The
biologically relevant glutamate concentrations can vary
significantly in the neurological system. For example, the
glutamate concentration in the synaptic cleft lies at the high
nanomolar level at resting state, but reaches a very short-lived
millimolar concentration burst following an action potential.14
Around astrocytes, the extracellular glutamate concentration is
(5) Hirose, K.; Namiki, S.; Sakamoto, H.; Iinuma, S.; Iino, M. Eur. J.
Neurosci. 2007, 25 (8), 2249−2259.
(6) Brun, M. A.; Tan, K. T.; Nakata, E.; Hinner, M. J.; Johnsson, K. J.
Am. Chem. Soc. 2009, 131 (16), 5873−5884.
(7) Brun, M. A.; Griss, R.; Reymond, L.; Tan, K. T.; Piguet, J.; Peters,
R. J.; Vogel, H.; Johnsson, K. J. Am. Chem. Soc. 2011, 133 (40),
16235−42.
estimated to be around 5 μM.15 Accordingly, the measured
comp,glutamate
(8) Keppler, A.; Gendreizig, S.; Gronemeyer, T.; Pick, H.; Vogel, H.;
Johnsson, K. Nat. Biotechnol. 2003, 21 (1), 86−89.
(9) Gautier, A.; Juillerat, A.; Heinis, C.; Correa, I.; Kindermann, M.;
Beaufils, F.; Johnsson, K. Chem. Biol. 2008, 15 (2), 128−136.
(10) Kuusinen, A.; Arvola, M.; Keinanen, K. EMBO J. 1995, 14 (24),
6327−6332.
Kd
of 15 μM for our Snifit-iGluR5 on cells is in an
appropriate range to study glutamate transients in neurological
systems.
Another critical parameter of a glutamate sensor is its
temporal resolution. The published activation and deactivation
rates of iGluR5 upon binding and unbinding of glutamate are in
the millisecond range and similar binding and unbinding rates
could be expected for Snifit-iGluR5.16 In our perfusion
experiments (Figure 4), the sensor closes and opens within
seconds. However, these rates reflect the performance of our
perfusion system which only allows the exchange of the bath
solution within 3 to 4 s (Figure S10). A detailed analysis of the
kinetics of our glutamate sensor will require a different
experimental setup and will need to be investigated in future
experiments.
In conclusion, we have developed a novel semisynthetic
sensor protein for glutamate based on our recently developed
Snifit sensor concept. The observed maximum ratio change for
Snifit-iGluR5 on cell surfaces is higher than that of any other
existing glutamate biosensor. Further, the exclusive labeling of
Snifit-iGluR5 on the cell surface with far-red fluorophores
should facilitate applications on tissue samples or in vivo as well
as its use in combination with other biosensors (multiplexing).
These properties should make Snifit-iGluR5 a very useful tool
to study the glutamate dynamics in the nervous system.
(11) Chen, G. Q.; Gouaux, E. Pro. Natl. Acad. Sci. U.S.A. 1997, 94
(25), 13431−13436.
(12) Volgraf, M.; Gorostiza, P.; Numano, R.; Kramer, R.; Isacoff, E.;
Trauner, D. Nat. Chem. Biol. 2006, 2 (1), 47−52.
(13) Volgraf, M.; Gorostiza, P.; Szobota, S.; Helix, M. R.; Isacoff, E.
Y.; Trauner, D. J. Am. Chem. Soc. 2007, 129 (2), 260−261.
(14) Bergles, D. E.; Diamond, J. S.; Jahr, C. E. Curr. Opin. Neurbiol.
1999, 9 (3), 293−8.
(15) McLamore, E. S.; Mohanty, S.; Shi, J.; Claussen, J.; Jedlicka, S.
S.; Rickus, J. L.; Porterfield, D. M. J. Neurosci. Methods 2010, 189 (1),
14−22.
(16) Dingledine, R.; Borges, K.; Bowie, D.; Traynelis, S. F.
Pharmacol. Rev. 1999, 51 (1), 7−61.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthetic procedures, characterizations of the Snifit-iGluR5 in
vitro and on the cell surface; additional figures. This material is
available free of charge via the Internet at http://pubs.acs.org.
7678
dx.doi.org/10.1021/ja3002277 | J. Am. Chem. Soc. 2012, 134, 7676−7678