Caged glutathione - Triggering protein interaction by light

Article abstract of DOI:10.1002/anie.201108073

Light, GSH, action! Glutathione (GSH) fulfills a universal role as redox factor, scavenger of reactive oxygen species, and as an essential substrate in the conjugation, detoxification, and reduction reactions catalyzed by glutathione S-transferase (GST). A photoactivatable glutathione allows the GSH-GST network to be triggered by light. GST fusion proteins can be assembled in situ at variable density and structures by laser-scanning activation. Copyright

Full text of DOI:10.1002/anie.201108073

.
Angewandte
Communications
DOI: 10.1002/anie.201108073
Chemical Biology
Caged Glutathione – Triggering Protein Interaction by Light**
Volker Gatterdam, Tatjana Stoess, Clara Menge, Alexander Heckel,* and Robert Tampꢀ*
The pseudotripeptide glutathione (GSH), consisting of glu-
tamate, cysteine, and glycine, is essential for the regulation of
the redox environment and detoxification in eukaryotic
cells.[1] The ratio of reduced GSH and oxidized GSSG
controls the redox potential in cellular compartments.[2]
Furthermore, GSH acts as an important redox scavenger of
reactive oxygen species (ROS) in aging and host–pathogen
interactions, causing oxidative stress involved in many
diseases, such as Alzheimerꢀs, Parkinsonꢀs, liver disease,
sickle-cell anemia, parasitic diseases,[3] AIDS, cancer,[4] car-
diac diseases, and diabetes.[5] Furthermore, the nucleophilic
thiol of GSH is used by glutathione S-transferase (GST)
enzymes to covalently bind xenobiotics, such as drugs, toxins,
or ROS.[6] GSH conjugates can then be secreted by the
glutathione S-conjugate transporters for detoxification.[7]
GSH recognition by GST is widely exploited in life science,
for example, protein purification,[8] high-throughput protein–
protein interaction screens,[9] or protein immobilization.[10]
Herein, we describe an approach to trigger the generic
GSH–protein interaction by light. An optimal compatibility
with cells or even animals is given by the usage of wavelengths
in the visible spectrum, which do not harm the target. Light
can be regulated very precisely, thus making a spatial,
temporal, and dosage control possible. In recent years,
many photoactivatable approaches have been realized.[11]
Important examples are caged biotin,[12] caged O6-benzylgua-
nine,[13] and photoactivatable trisNTA,[14] but also light-
controlled applications[15] such as peptide synthesis,[16] gene
regulation,[17] and structured cell adhesion[18] are emerging
fields.
Photoactivatable glutathione was designed by modifica-
tion of the amino and carboxyl groups that strongly impact
the recognition by GST and a number of other GSH-binding
proteins. The photoactivatable nitrophenylpropyl (NPP)
protection group was chosen.[19] Based on the X-ray structure
of GST with bound GSH, the amino and carboxyl moiety of
the g-l-glutamyl residue as well as the carboxyl group of the
glycine were identified as putative interaction sites[20] that are
available for chemical modification (Figure 1).
Figure 1. a) X-ray structure of the human glutathione S-transferase
M2-2 with bound GSH (PDB: 1XW5).[21] b) The GSH binding pockets
of the homodimer are in the gap between both monomers. Dotted red
lines illustrate the interaction network between GST and GSH. The SH
function of the cysteine is freely accessible. c) Synthesis and structure
of photoactivatable GSH derivatives GSHNPP. TMS-Cl=trimethylsilyl
chloride; TCEP=tris(2-carboxyethyl)phosphine hydrochloride.
[*] V. Gatterdam,[+] Prof. Dr. R. Tampꢀ
Institute of Biochemistry, Biocenter
We synthesized a doubly and a singly caged GSH,
GSHNPP2 and GSHNPP, respectively (see the Supporting
Information). The former displayed very low solubility in
aqueous solution, and in particular after coupling to fluoro-
phores. Therefore, GSHNPP2 was applied only in protein
interaction studies at interfaces. Protection at the C-terminus
of glycine was also realized but is not described herein. The
sulfhydryl group of GSH was used for covalent modification
with various fluorophores and functionalization of interfaces.
These modifications should not interfere with the GST
binding, because they are used for coupling to xenobiotics
in the cellular environment. The expression of GST fusion
proteins is well-established[8] and allows almost any protein of
interest to be fused to GST.
Cluster of Excellence Frankfurt (CEF)
Goethe University Frankfurt
Max-von-Laue-Strasse 9, 60438 Frankfurt am Main (Germany)
E-mail: tampe@em.uni-frankfurt.de
Homepage: http://www.biochem.uni-frankfurt.de
T. Stoess,[+] C. Menge, Prof. Dr. A. Heckel
Frankfurt Institute for Molecular Life Sciences
Cluster of Excellence Frankfurt (CEF)
Goethe University Frankfurt
Max-von-Laue-Strasse 9, 60438 Frankfurt am Main (Germany)
E-mail: heckel@uni-frankfurt.de
[+] Both authors contributed equally to this work.
[**] This work was supported by the DFG through the Cluster of
Excellence Macromolecular Complexes (EXC 115) and by the
Goethe University Frankfurt. A generous donation of silyl protecting
group precursors by the Wacker Company and the GST-eGFP vector
by T. Nuutinen and Prof. J. Syvꢁoja (University of East Finland) is
gratefully acknowledged.
We first investigated the specificity and rate of the
deprotection reaction of GSHNPP by reverse-phase HPLC.
The photoreaction was triggered at 366 nm using an LED
(140 mWcmÀ2). For sensitive detection, the fluorophore
ATTO565 (as maleimide derivate) was covalently coupled
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201108073.
3960
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3960 –3963

Angewandte
Chemie
to the cysteine of GSHNPP. This fluorescent GSHNPP (10 nmol
in TBS buffer) was illuminated for different periods. After
60 s, no starting material and only minute amounts of side
products were detected (Figure 2). The deprotection follows
Figure 2. Time-dependent photoactivation of ATTO565-labeled GSHNPP
monitored by reverse-phase HPLC. a) Time series of the deprotection
reaction followed at 565 nm. Caged and uncaged GSHNPP elute at 15.0
and 8.5 min, respectively. The identity of the photoproduct was
Figure 3. Light-triggered GSH-GST interaction detected using FRET.
*
confirmed by ESI-MS. b) The starting material ( ) is consumed at
a) Complex formation with GSH and GST was followed using GST-
eGFP (10 nm) as FRET donor. Increasing concentrations of ATTO565-
labeled GSHNPP as FRET acceptor were added before (see the Support-
ing Information) and after photoactivation. The donor was excited at
450 nm. Photoactivation leads to a dramatic decrease of donor
fluorescence and increase of the acceptor fluorescence, demonstrating
complex formation. b) The donor fluorescence was corrected towards
dilution and inner filter effects with appropriate experiments. The
inverted and normalized change in donor fluorescence at 511 nm is
plotted against the concentration of FRET acceptor and fitted with
Equation (1) (see the Supporting Information), yielding a KD of
a rate of (0.060Æ0.013) sÀ1 and the rate of formation of the uncaged
À1
*
product ( ) is (0.071Æ0.016) s
.
a monoexponential decay in combination with a monoexpo-
nential growth of the product. In the range of error, the rate
constant for the consumption of the starting material
((0.060 Æ 0.013) sÀ1, t1/2 = (11.5 Æ 3.1) s), and formation of
the product ((0.071 Æ 0.016) sÀ1, t1/2 = (9.8 Æ 2.8) s) are iden-
tical, indicating a one-step deprotection without rate-limiting
intermediates. We also examined the long-term stability by
30 minute illumination. No major photo-damage of the
uncaged product or bleaching of the fluorophore was
observed. In conclusion, we demonstrate a fast, one-step
photoactivation of GSHNPP without intermediate compounds.
To follow the GSH-GST interaction in real time, we
established a Fçrster resonance energy transfer (FRET) assay
using GST C-terminally fused to enhanced green fluorescent
protein (eGFP) and ATTO565-labeled GSHNPP as the donor–
acceptor pair. GST-eGFP (10 nm in TBS buffer) was incu-
*
(53Æ2) nm after photoactivation ( ). A very low affinity interaction
*
was observed before photoactivation ( ). c) An extensive spectral
change was observed before (black) and after photoactivation (at
366 nm, red) of a mixture of GST-eGFP (10 nm) and labeled GSHNPP
(500 nm). d) Interaction of GST-eGFP (10 nm) and labeled GSHNPP
(500 nm) followed by FRET in real time. The acceptor fluorescence at
592 nm was recorded by excitation of the donor at 366 nm, which also
triggers the photoactivation process. CPS=counts per second.
We next studied the in situ assembly of GST fusion
proteins on GSHNPP- and GSHNPP2-functionalized surfaces by
light. Light-directed protein organization was performed by
the use of a quartz glass mask covered with a chrome pattern
of different grid sizes. The mask was placed on top of the
GSHNPP- or GSHNPP2-functionalized glass slides and illumi-
nated at 366 nm (Xe/Hg lamp, 200 W) for 3 minutes. During
structured illumination, the functionalized interfaces were
incubated with GST-eGFP (500 nm). Confocal laser scanning
microscopy (CLSM) was used to monitor the process of
photoactivation and assembly. Defined structures of GST-
eGFP at the photopattern areas could be observed (Fig-
ure 4a). Superb contrast over large areas reflects the high
specificity of the photoactivation of GSHNPP. However, owing
to the hydrophobicity of doubly caged GSHNPP2, a dramatic
increase of unspecific protein binding was detected at the
GSHNPP2-functionalized interfaces (see the Supporting Infor-
mation). Despite the fast on and off binding kinetics, the
protein patterns at photoactivated GSHNPP interfaces were
stable over several days and could be reused for further
bated with caged and uncaged (ATTO-labeled) GSHNPP
,
respectively. Titration with labeled GSHNPP showed no
significant FRET (see the Supporting Information). How-
ever, illumination of labeled GSHNPP leads to a significant
FRET demonstrated by an increase and decrease of the
acceptor and donor fluorescence, respectively (Figure 3a,b).
The KD of non-illuminated ATTO565-labeled GSHNPP was
(53 Æ 2) nm, which indicates a higher affinity than GSH.[22]
Results were confirmed by validation of the acceptor
fluorescence, which yielded in a similar value of KD.[23] The
light-triggered GSH-GST interaction was followed in real
time by FRET. GST-eGFP (10 nm) and ATTO565-labeled
GSHNPP (500 nm) were illuminated at 366 nm. After photo-
activation, an increase and decrease of acceptor and donor
fluorescence, respectively, was detected (Figure 3c). Notably,
the reaction rate is limited by the diffusion of GSHNPP
through the very small illumination spot in the cuvette
(Figure 3d) and therefore not comparable to the true rate of
reaction.
Angew. Chem. Int. Ed. 2012, 51, 3960 –3963
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3961

.
Angewandte
Communications
Patterns can already be
detected after two iterations
(82 ms). The observed lines
have a full width at half
maximum (FWHM) of
500 nm (Figure 4c). Even
faster writing times can be
achieved by downsizing the
structures. Higher iteration
numbers lead to saturation
and blurring of the struc-
ture.
We then followed the
binding
of
GST-eGFP
(500 nm) in real time. Acti-
vation with the 405 nm laser
was performed by a single
iteration (light flash) fol-
lowed by immediate moni-
toring of the GST-eGFP
fluorescence (Figure 4d).
The exponential binding
kinetics revealed
a
kon
value of 0.2 minÀ1. Thus,
sequential writing and par-
allel readout of complex
structures were realized.
We finally used the 405 nm
laser to assemble GST
fusion proteins in freely
designed patterns (Fig-
ure 4e). By use of several
GST fusion proteins, a mul-
tiprotein surface[10] with
defined areas can be gener-
ated under light control.
Therefore, protein com-
plexes and the GSH-GST
interaction networks[24] can
be controlled by light.
Figure 4. In situ assembly of GST fusion proteins triggered by light. a) Protein-resistant PEGylated interfaces
are functionalized with GSHNPP and illuminated at 366 nm by mask lithography. Binding of GST-eGFP
(500 nm) was analyzed after washing with TBS buffer by confocal laser scanning microscopy. A 488 nm argon
laser (10% max. power, 30 mW) was used for excitation. A patterning with well-defined edges and high
contrast was observed. An intensity profile was taken across the indicated area. b) Different lateral densities
of GST fusion proteins were generated by in situ patterning using increased illumination power of the 405 nm
UV diode laser (100% max. power, 20 mW). The structures were visualized with a 488 nm argon laser. c) To
determine the spatial resolution, a single line was written with different iterations (41 ms for one scan,
0.29 mmꢂ58.0 mm). The first protein pattern could already be seen after two iterations (82 ms). Increasing the
number of iterations leads to a blurring of the structure owing to the Gaussian illumination distribution.
d) In situ protein assembly followed in real time. A GSHNPP-functionalized protein-resistant PEGylated interface
was illuminated with a 405 nm laser in the highlighted spot for less than 10 s. Monitoring with a 488 nm laser
revealed the real-time binding kinetics of GST-eGFP (500 nm in TBS buffer). Analysis with a monoexponential
À1
*
growth of GST-eGFP fluorescence ( ) revealed an association rate constant of (0.21Æ0.01) min (t1/2
=
*
(3.3Æ0.1) min). represent the fluorescence of GST-eGFP (500 nm) present in solution. e) Free design of
protein arrays using the 405 nm laser with one iteration only (<1 min). The white bar indicates 20 mm in all
images.
In conclusion, we intro-
duced a new photoactivat-
able capturing tag based on
the GST-GSH interaction.
experiments. It was possible to dissociate the protein pattern
by incubation with release buffer (10 mm glycine, pH 2.0) in
less than 10 seconds. After equilibration of the recovered
slides in TBS buffer, an identical pattern of GST-eGFP could
be restored. Apart from mask lithography, in situ patterning
with a scanning laser was also conducted. Freely designed
regions were illuminated with the 405 nm diode laser and the
assembly of GST-eGFP was followed by CLSM (lex/em = 488/
520 nm). Different protein densities could be achieved by
varying the illumination power (Figure 4b).
To investigate the spatial resolution, the interface was
illuminated by a point laser source. The numerical aperture of
the objective lens and the laser wavelength determine the
focal volume. Diffraction-limited lines (0.29 mm thick and
58 mm in length) were scanned by the 405 nm laser in 41 ms.
The compound GSHNPP can be functionalized in a versatile
way on the free cysteine without affecting the GST inter-
action. Functionalization with bulky fluorophores and cova-
lent immobilization was demonstrated without interfering
with the complex formation. The deprotection with light is
very fast and complete in less than a millisecond to second
range. The system has a very good temporal resolution. This is
an advantage in terms of photo damage in vitro and in vivo.
Furthermore, a good spatial resolution was achieved by the
use of a CLSM, allowing the illumination and reading of
0.5 mm thick lines, which is limited by the optical resolution.
Received: November 16, 2011
Published online: March 5, 2012
3962 www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 3960 –3963

Angewandte
Chemie
628; e) M. Goeldner, R. Givens, Dynamic Studies in Biology,
Wiley-VCH, Weinheim, 2005.
[12] M. C. Pirrung, C. Y. Huang, Bioconjugate Chem. 1996, 7, 317 –
321.
[13] S. Banala, A. Arnold, K. Johnsson, ChemBioChem 2008, 9, 38 –
41.
Keywords: glutathione S-transferase ·
light-triggered chemical biology · photoactivatable compounds ·
photoactivation · protein immobilization
.
[1] J. D. Hayes, L. I. McLellan, Free Radical Res. 1999, 31, 273 – 300.
[2] a) H. F. Gilbert, Methods Enzymol. 1984, 107, 330 – 351; b) F. Q.
Schafer, G. R. Buettner, Free Radical Biology and Medicine
2001, 30, 1191 – 1212; c) S. Dimmeler, A. M. Zeiher, Cell 2007,
130, 401 – 402.
[3] R. L. Krauth-Siegel, H. Bauer, R. H. Schirmer, Angew. Chem.
2005, 117, 698 – 724; Angew. Chem. Int. Ed. 2005, 44, 690 – 715.
[4] K. J. Ritchie, S. Walsh, O. J. Sansom, C. J. Henderson, C. R. Wolf,
Proc. Natl. Acad. Sci. USA 2009, 106, 20859 – 20864.
[5] G. Wu, Y. Z. Fang, S. Yang, J. R. Lupton, N. D. Turner, J. Nutr.
2004, 134, 489 – 492.
[14] a) C. Grunwald, K. Schulze, A. Reichel, V. U. Weiss, D. Blaas, J.
Piehler, K. H. Wiesmꢁller, R. Tampꢂ, Proc. Natl. Acad. Sci. USA
2010, 107, 6146 – 6151; b) M. Bhagawati, S. Lata, R. Tampꢂ, J.
Piehler, J. Am. Chem. Soc. 2010, 132, 5932 – 5933.
[15] T. Fehrentz, M. Schçnberger, D. Trauner, Angew. Chem. 2011,
123, 12362 – 12390; Angew. Chem. Int. Ed. 2011, 50, 12156 –
12182.
[16] a) S. Fodor, J. Read, M. Pirrung, L. Stryer, A. Lu, D. Solas,
Science 1991, 251, 767 – 773; b) S. Li, D. Bowerman, N. Mar-
thandan, S. Klyza, K. J. Luebke, H. R. Garner, T. Kodadek,
J. Am. Chem. Soc. 2004, 126, 4088 – 4089.
´
[6] V. B. Djordjevic, Int. Rev. Cytol. 2004, 237, 57 – 89.
[17] a) A. Deiters, Curr. Opin. Chem. Biol. 2009, 13, 678 – 686; b) X.
Tang, I. J. Dmochowski, Mol. BioSyst. 2007, 3, 100 – 110.
[18] S. Petersen, J. M. Alonso, A. Specht, P. Duodu, M. Goeldner, A.
del Campo, Angew. Chem. 2008, 120, 3236 – 3239; Angew. Chem.
Int. Ed. 2008, 47, 3192 – 3195.
[19] S. Walbert, W. Pfleiderer, U. E. Steiner, Helv. Chim. Acta 2001,
84, 1601 – 1611.
[20] A. E. Adang, J. Brussee, A. van der Gen, G. J. Mulder, Biochem.
J. 1990, 269, 47 – 54.
[21] S. Raghunathan, R. J. Chandross, R. H. Kretsinger, T. J. Allison,
C. J. Penington, G. S. Rule, J. Mol. Biol. 1994, 238, 815 – 832.
[22] I. Jakobson, M. Warholm, B. Mannervik, J. Biol. Chem. 1979,
254, 7085 – 7089.
[23] S. F. Martin, M. H. Tatham, R. T. Hay, I. D. W. Samuel, Protein
Sci. 2008, 17, 777 – 784.
[7] A. Pompella, A. Visvikis, A. Paolicchi, V. De Tata, A. F. Casini,
Biochem. Pharmacol. 2003, 66, 1499 – 1503.
[8] D. B. Smith, K. S. Johnson, Gene 1988, 67, 31 – 40.
[9] a) H.-S. Ro, B. H. Koh, S. O. Jung, H. K. Park, Y.-B. Shin, M.-G.
Kim, B. H. Chung, Proteomics 2006, 6, 2108 – 2111; b) J.-W. Jung,
S.-H. Jung, H.-S. Kim, J. S. Yuk, J.-B. Park, Y.-M. Kim, J.-A. Han,
P.-H. Kim, K.-S. Ha, Proteomics 2006, 6, 1110 – 1120.
[10] C. M. Kolodziej, C.-W. Chang, H. D. Maynard, J. Mater. Chem.
2011, 21, 1457 – 1461.
[11] a) A. del Campo, D. Boos, H. W. Spiess, U. Jonas, Angew. Chem.
2005, 117, 4785 – 4791; Angew. Chem. Int. Ed. 2005, 44, 4707 –
4712; b) J. M. Alonso, A. Reichel, J. Piehler, A. del Campo,
Langmuir 2008, 24, 448 – 457; c) G. Mayer, A. Heckel, Angew.
Chem. 2006, 118, 5020 – 5042; Angew. Chem. Int. Ed. 2006, 45,
4900 – 4921; d) G. C. R. Ellis-Davies, Nat. Methods 2007, 4, 619 –
[24] G. MacBeath, S. L. Schreiber, Science 2000, 289, 1760 – 1763.
Angew. Chem. Int. Ed. 2012, 51, 3960 –3963
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3963