Enhanced Green Fluorescent Protein (GFP)
fluorescence after polyelectrolyte caging
Alberto Diaspro and Silke Krol
IFOM-LAMBS/MicroScoBio, Department of Physics, University of Genoa,
Via Dodecaneso 33, 16146 Genoa, Italy.
[email protected],it
Barbara Campanini
Department of Biochemistry and Molecular Biology, University of Parma, 43100, Parma, Italy.Italy.
Fabio Cannone and Giuseppe Chirico
CNR-INFM, Department of Physics, University of Milano-Bicocca, Piazza della Scienza 3, 20125 Milan, Italy.
Abstract: Discovery of Green Fluorescent Protein (GFP) constituted an
important improvement for living cell studies on submicron resolution
allowing in vivo fluorescence labeling. We studied the photo-physical
properties of single GFP molecules incorporated in a charged
polyelectrolyte environment by means of single molecule spectroscopy. The
fluorescence characteristics change dramatically in terms of photo-stability,
lifetime and blinking behavior so that the proteins scale up to quantum dots.
The reported results highlight interesting applications in the design of
fluorescent markers and in the development of optical data storage
architectures.
©2006 Optical Society of America
OCIS codes: (160.2540) Fluorescent and luminescent materials; (170.2520) Fluorescence
microscopy; (190.4180) Multiphoton processes; (999.9999) Green fluorescent protein.
References and links
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
O. Shimomura, “The discovery of aequorin and green fluorescent protein,” J. Microscopy 217, 3-15 (2005).
A.B. Cubitt, R. Heim, S.R. Adams, A.E. Boyd, L.A. Gross, R.Y. Tsien, “Understanding using and
improving green fluorescent protein,” Trends in Biochem. Sci. 20, 448-455 (1995).
J. Lippincott-Schwartz, E. Snapp, A. Kenworthy, “Studying protein dynamics in living cells,” Nat. Rev.
Mol. Cell. Biol. 2, 444-456 (2001).
M. Zimmer, “Green fluorescent protein, Application structure and related photophysical behaviour.” Chem
Rev. 102, 759–782 (2002).
G. Jung, J.Wiehler, A. Zumbusch, “The Photophysics of GFP, influence of the Key Amino Acids at
Positions 65 203 and 222.” Biophys. J. 88, 1932-1947 (2005).
G.H. Patterson, J. Lippincott-Schwartz, “A photoactivatable GFP for selective photolabeling of proteins
and cells.” Science. 297, 1873-1877 (2002).
M. Schneider, S. Barozzi, I.Testa, M. Faretta, A. Diaspro A., “Two-photon activation and excitation
properties of PA-GFP in the 720-920-nm region.” Biophys J. 89, 1346-1352 (2005).
M. Andresen, M.C. Wahl, A. Stiel, F. Gräber, L.V. Schäfer, S.Trowitzsch, G. Weber, C. Eggeling, H.
Grubmüller, S.W. Hell, S. Jakobs, “Structure and mechanism of the reversible photoswitch of a fluorescent
protein.” Proc. Natl. Acad. Sci. USA 102 (37), 13070 – 13074 (2005).
A. Zumbusch, G. Jung “Single Molecule Spectroscopy of the GFP, A Critical Assessment.” Single Mol. 1,
261-270 (2000).
G. Chirico, A. Diaspro, F. Cannone, M. Collini, S. Bologna, V. Pellegrini, F. Beltram, “Selective
Fluorescence Recovery after Bleaching of Single E2GFP Proteins induced by two-photon excitation.”
Chem. Phys. Chem. 6, 328-335 (2005).
B.B. Cormack, R.H. Valdivia, S. Falkow, “FACS-optimized mutants of green fluorescent protein (GFP).”
Gene. 173, 33–38 (1996).
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9815
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
A. Diaspro, D. Silvano, S. Krol, O. Cavalleri, A. Gliozzi, “Single living cell encapsulation in nanoorganized polyelectrolyte shells.” Langmuir. 18, 5047-5050 (2002).
M. Chanana, A. Gliozzi, A. Diaspro, I. Chodnevskaja, S. Huewel, V. Moskalenko, K. Ulrichs, H.J. Galla, S.
Krol, “Interaction of polyelectrolytes and their composites with living cells.” Nano Lett.. 5, 2605-2612
(2005).
G. Malengo, R. Milani, F. Cannone, S. Krol, A. Diaspro, G. Chirico, “High sensitivity optical microscope
for single molecule spectroscopy studies.” Rev. Sci. Instrum. 75, 2746-2751 (2004).
L.A. Deschenes, D.A. Vanden Bout, “Single molecule photobleaching, increasing photon yield and survival
time through suppression of two-step photolysis.” Chem. Phys. Lett. 365, 387-395 (2002).
W. van Sark et al. “Photooxidation and Photobleaching of Single CdSe/ZnS QDs probed by RoomTemperature Time-Resolved Spectroscopy.” J. Phys. Chem. B. 105, 8281-8284 (2001).
G. Decher, “Fuzzy nanoassemblies, Toward layered polymeric multicomposites.” Science. 277, 1232-1237
(1997).
E. Donath, G.B. Sukhorukov, F. Caruso, S.A. Davis, H. Mohwald, “Novel hollow polymer shells by
colloid-templeted assembly of polyelectrolytes” Angew.Che.Int.Ed. 16, 2202-2205 (2000).
A. Diaspro, S. Krol, O. Cavalleri, D. Silvano, A. Gliozzi, “Microscopical Characterization of Nanocapsules
Templated on Ionic Crystals and Biological Cells Toward Biomedical Applications.” IEEE Trans.
Nanobiosci. 3, 110-115 (2002).
G. Chirico, F. Cannone, S. Beretta, G. Baldini, A. Diaspro, “Single Molecule Studies by Means of the TwoPhoton Fluorescence Distribution.” Micr. Res Tech. 55, 359-364 (2001).
G. Chirico, F. Cannone, S. Beretta, A. Diaspro, B. Campanini, S. Bettati, R. Ruotolo, A. Mozzarelli,
“Dynamics of green GFPmut2 in solution on spin-coated glasses and encapsulated in wet silica gels.”
Protein Science. 11, 1152–1161 (2002).
A. Diaspro, G. Chirico, M. Collini, “Two-photon fluorescence excitation and related techniques in
biological microscopy.” Q. Rev. Biophys. 38(2), 97-166 (2005).
F. Cannone, M. Caccia, S. Bologna, A. Diaspro, G. Chirico, “Single Molecule Spectroscopic
Characterization of GFP-Mut2 Mutant for Two-Photon Microscopy Applications.” Micr. Res. Tech. 65,
186-193 (2004).
J.M. Tsay, S. Doose, F. Pinaud, S. Weiss, “Enhancing the Photoluminescence of Peptide-Coated
Nanocrystals with Shell Composition and UV Irradiation.” J. Phys. Chem. B. 109, 1669-1674 (2005).
G.H. Patterson, D.W. Piston, “Photobleaching in two-photon excitation microscopy.” Biophys. J. 78, 21592162 (2000).
G. Chirico, F. Cannone, G. Baldini, A. Diaspro, “Two-photon thermal bleaching of single fluorescent
molecules.” Biophys. J. 84, 588-598 (2003).
M. Ormo, A.B. Cubitt, K. Kallio, L.A. Gross, R.Y. Tsien, S.J. Remington S.J. “Crystal structure of the
Aequorea victoria green fluorescent protein.” Science 273, 1392–1395 (1996).
H.M. Berman et al., “The Protein Data Bank” Nucleic Acids Research 28, 235-242 (2000).
The stabilizing and protecting effect in the polyelectrolyte matrix was also supported by chemically induced
denaturation experiments. For this purpose, urea and guanidine hydrochloride were added to the proteins
embedded in the polyelectrolyte layers and the disappearance of the fluorescence signals with time was
recorded. For both agents a prolongation of the protein unfolding in comparison to wet silica gel-enveloped
GFPmut2 was found (data not shown).
1. Introduction
The discovery of the Green Fluorescent Protein (GFP) from the jellyfish Aequorea Victoria
has had major impacts on microscopy, live cell imaging, cell biology and research relating to
its photophysics [1]. GFP can be expressed by cells as a fusion protein connected to specific
cellular structures and spatially visualized by means of three-dimensional optical microscopy
methods [2]. Several mutants have been designed to improve or change its photo-physical
features, like fluorescence emission (brightness), bleaching stability, photo-activation, photoswitching or pH sensitivity [3-8]. Unfortunately, there remains one major drawback, its low
photo-stability [9], even if in some special cases it is possible to recover the fluorescence of
the protein after bleaching [10].
In the following we will introduce a new approach in which single GFPmut2 [11]
molecules were incorporated into differently charged polyelectrolyte biocompatible matrices,
previously used to construct polyelectrolyte shells on living yeast cells [12] or nanogold
particles [13]. The photo-stability in terms of photon number released before bleaching and
other photo-physical parameters of such a caged GFPmut2 fluorescent protein were measured
at single molecule level [14]. Results show a final number of photons released before
bleaching only known for fluorescent dyes under vacuum [15] and quantum dots (QDs) in air
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9816
[16]. Layer-by-layer (Lbl) technique [17] was used to deposit stepwise, in a self-assembly
manner, oppositely charged polymers (PSS, poly-(styrenesulfonate sodium salt) and PAH,
poly-(allylamine hydrochloride)) on a charged template, forming a sort of capsule [18, 19].
Such a capsule is also used as a container for testing the photophysical properties of GFPmut2
proteins caged in differently charged environments. The motivation for studying GFP photophysics in polyelectrolyte matrices come from the sensitivity of GFP fluorescence dynamics
details on the environment. It is known that GFP photophysics is mainly determined by the
switching between a neutral (protonated state of the chromophore) and anionic (deprotonated
state) state [9, 10]. This balance is largely affected by the environment of GFP. The use of
polyelectrolytes to shield GFP from free diffusing charges resulted, as demonstrated in this
communication, in a successful method to improve the GFP photo-stability. The results
obtained also point out to the potential properties of a new class of fluorescent molecules
made by polyecletrolyte encapsulation of single fluorescent proteins.
2. Methods
2.1 Sample preparation
The polyelectrolytes (PEs) poly-(styrenesulfonate sodium salt) (PSS, MW 70 kDa, Aldrich,
Milan, Italy) and poly-(allylamine hydrochloride) (PAH, MW 15 kDa, Aldrich, Milan, Italy)
were solved in 0.5 M NaCl to a final concentration of 2 mg/mL in Milli-Q grade pure water.
As template an amorphous calcium carbonate material with a round appearance (∼ 4.7 μm in
diameter) was used. Each PE self-assembly step was performed by incubation of the core in
the supersaturated polyion solution (5 min for PAH, 10 min for PSS) followed by
centrifugation and two washing steps in 0.5 M NaCl [12]. The encapsulation procedure
generally started with PAH as first layer because with zeta-potential the amorphous carbonate
material revealed a negative surface charge. Six or eight polyelectrolyte layers were absorbed.
Due to the even number of deposited layers the outermost one was always PSS. GFPmut2 is a
triple GFP mutant (S65A, V68L, S72A) with enhanced fluorescence emission in comparison
to the wild type and an increased protein yield due to a more efficient folding [11].
GFPmut2 proteins were bound to the capsule in three different ways, namely:
1) the amorphous core was incubated for 10 min in the GFPmut2 solution (21 mg amorphous
spheres in 0.5 ml 0.5 M NaCl + 10 μl GFPmut2 solution (c(GFPmut2)=1 nM; pH= 6.8)).
Then the layer deposition was carried out onto the GFP-doped core;
2) PE capsule preparation was performed like described above and as last step the PSSterminated capsule was stored for 10 min in a highly diluted GFPmut2 solution (0.5 ml 0.5 M
NaCl + 10 μl GFPmut2 solution (c(GFPmut2)=1 nM; pH.= 6.8)) attaching the protein with a
positive net charge to the outermost layer by means of electrostatic interaction;
3) GFPmut2 was embedded within the PE layer by co-attachment of a GFPmut2-doped PAH
solution (0.5 ml PAH solution + 20 μl GFPmut2 solution (c(GFPmut2)=1 nM; pH.= 6.8)). For
this last configuration one can distinguish two different situations, namely: I) GFPmut2 is
embedded in the 3rd layer of a capsule with 6 layers ((PAH/PSS)3); II) GFPmut2 is enveloped
in the 5th layer of a 8 layers capsule ((PAH/PSS)4).
Additional to the above-described GFPmut2 entrapment into layers on colloidal particles,
the single proteins were also investigated in two-dimensional polyelectrolyte layers. In this
case, GFPmut2-doped polyion solutions (PAH or PSS) were spin-coated on cleaned glass
coverslips [20]. Real-time fluorescence images revealed that the initial diffusion of the
molecules stops after one hour. Afterwards no diffusion was observable over a 24 h time
period.
For all the experiments, GFPmut2 concentration was chosen to be low enough allowing
the observation of a reasonable number of single molecules [14, 20].
2.2 Single molecule microscopy
Single molecule fluorescence excitation and detection was achieved by using two-photon
excitation utilizing an ultrafast mode-locked Ti:sapphire laser (Tsunami 3960, Spectra
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9817
Physics, CA), (100 fs pulsewidth, 80 MHz repetition frequency) coupled to an inverted
microscope (TE300, Nikon, Florence, Italy) (oil immersion objective, 100X, 1.4 NA).
Fluorescence emission was selected using by a dichroic beam splitter and a band pass filter
(515/30) (Chroma Inc., Brattelboro, VT). Two avalanche photodiode (APD) detectors
(SPCM-AQ-151, EG & E, Salem, MA) and a three-dimensional nano-piezo stage (Physik
Instrumente, Milan, Italy) were used [14]. The acquisition of the fluorescence images (30 x 30
µm) with 1 ms residence time per pixel took around 1 s. The excitation intensity for imaging
was 9.5 kW cm-2 (25 µW). Analysis was performed using a home-coded software allowing
photon counting histograms and lifetime pixel-by-pixel computing. A PCI-Board for TimeCorrelated Single Photon Counting (TCSPC) (Time Harp 200, PicoQuant, Berlin, Germany)
was employed. The analysis of fluorescence profiles was performed by Gaussian fit.
2.3 Single molecule spectroscopy
Fluorescent parameters at single molecule level measurements were carried out through a
series of Z-stacks (2.8 µm from the coverslip glass surface) considering the plan,
perpendicular to the optical axis, were a maximum section of the nanocapsule is observed. In
this way interactions between glass and the fluorophore can be excluded. The bleaching
behavior also serves as a marker to distinguish single proteins from clusters. Photo-physical
characterization of single GFPmut2 proteins was achieved by measuring the brightness ε, the
lifetime, and the bleaching time tBleaching for a single GFPmut2 in the anionic state (λexc(TPE) =
880 nm, λem = 508 nm) [14]. Therefore, the fluorescence versus time and the lifetime of single
GFPmut2 were collected by focusing the beam laser on a fluorescence spot, i.e. a single
protein. The time in which the fluorescence intensity diminishes to background level was
defined as bleaching time: tBleaching. The duration for a lifetime measurement, τ, is < tBleaching
(typically 10 s at 9.5 kW cm-2).
3. Results and discussion
Different sets of experiments have been performed on the caged GFPmut2 proteins. Figure 1
sketches the different caging conditions occurring during the experiments.
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9818
Fig. 1. GFPmut2 configurations in the capsule A), B) and in spin coated polyelectrolyte layers
(red for PSS, blu for PAH). See Methods.
The first set of experiments was related to co-adsorption of GFPmut2 with the polycation,
PAH (poly-(allylamine hydrochloride)) during the layer self-assembly process onto
amorphous calcium carbonate as template. GFPmut2 is mainly located within the specific
layer. From the fluorescence features of the detected single molecules one can distinguish
three main “internal” configurations, denoted as A(1), B and C. Apart from co-adsorption
steps, capsules, ending with a negative PSS layer, were immersed in a solution containing
GFPmut2 proteins. In this case, GFPmut2 molecules attach to the outermost layer and are
either in contact to the negative polyelectrolyte or the environmental solution. Two further
fluorescent families, i.e., A(2) and S(1), can be identified as “external” configuration. As second
set of experiments, in order to identify the influence of the polyelectrolyte charge, GFPmut2
proteins were solved in either the polyanion or the polycation solution and spin-coated onto
glass substrate. In a third set of experiments, the amorphous calcium carbonate core [19] was
doped with GFP or the protein was entrapped in wet silica gel [21]. Two new families were
formed, namely: S(2) in the core and S(3) in wet silica gel. In all these experiments the proteins
were sparsely distributed and single molecule characterization was performed under twophoton excitation regime [14, 22], see Fig. 2.
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9819
Fig, 2. (Left) GFPmut2 molecules (bright green spots) incorporated in a polyelectrolyte matrix
onto amorphous calcium carbonate (dark sphere). The field of view is 3.6 x 3.6 μm (Central)
shows the brightness and (Right) is the lifetime image of (Left) panel. So, the left picture is a
fluorescence image where each pixel corresponds to the average photons counted per ms. The
pixel size of the image of the left panel (50x50 nm) is different from that of the lifetime and
brightness images (144 x 144 nm). This enhancement in terms of visualization results in a
much blurred image on central and right panels [14].
Fluorescence parameters, for the different families sketched in Fig. 1 and described in the text,
are reported in Table 1.
Table 1. Fluorescence parameters for different configurations.a
ε [KHz] (1)
τ
[ns]
TBleaching [s] (1)
Quantum yield
(%)
External
Configuration
A(2)
S(1)
Internal
Configuration
A(1)
B
C
Amorphous
Core
S(2)
Wet silica gel (3)
S(3)
Spin coated
PSS -GFP
PAH-GFP
Quantum Dots
17.0 ± 4.2
14.2 ± 3.5
3.4 ±0.2
2.8 ±0.2
250 ± 18
75 ± 11
-
16.8 ± 4.1
10.8 ± 3.2
7.5 ± 2.8
3.5 ± 0.4
2.1 ± 0.3
2.5 ± 0.2 (2)
255 ± 13
154 ± 12
108 ± 17
-
14.2 ± 3.3
2.8 ± 0.4
73 ± 5
-
14.0 ± 3.0
2.9 ± 0.2
75 ± 12
33
16.8 ± 7.0
10.7 ± 3.0
-
3.3 ± 0.5
2.2 ± 0.3
-
253 ± 19
158 ± 11
150 (4)
68
39
19-28 (5)
a
Notes: (1) - Brightness and bleaching time were determined with an excitation intensity of 9.5 Kw/cm2; (2) Fluorescence lifetime decay is fitted with two exponential functions: τ1 = 3.36 ns 50 weight (wt) % and τ1 = 2.1 ns 50
weight (wt) %; (3) - see [21, 23]; (4) - see [16]; (5) – see [24].
Here, the values of lifetime and brightness are not directly taken on the images. After
performing a first fast scan we focus on the each spot, by finding the maximum output, and
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9820
acquire fluorescence versus time in order to record the brightness, its fluctuation, the protein
excited state lifetime and the fluorescence blinking up to the bleaching time. The search of the
maximum output is performed in few seconds [14]. Usually, about 700 single molecules were
used to determine the average values and standard deviations. In order to classify entrapped
GFP in families both brightness and bleaching time can be used. However, since bleaching
time difference among the families is more pronunced we have used bleaching time to classify
the molecules in the different families. The quantum yield values have been measured on bulk
samples and not on single molecules. So it was not possible to give a measure of quantum
yield on each single protein belonging to the various families. Number of photons released
before bleaching as indicator of photo-stability are reported in Fig. 3 for some of the detected
families.
Fig. 3. Number of collected photons before bleaching versus the excitation intensity. A(2)
outside attached to PSS layer ( ); A(1) family (●) and B family (O) in the layers; silica gel (S(3)
in table 1) (◊); agarose gel (▼) as in [9].
Variations in excitation energy have a significant effect on the photo-stability: bleaching
time under two-photon excitation scales as I-2.5 [25, 26]. Nevertheless, with the embedding of
proteins in polyelectrolyte matrices as well as using low excitation energy we were able to
count up to 9 photons/ms for longer than 9 minutes at 6.6 kW/cm2. With about 7 * 106
collected photons at around 3.8 kW/cm2 the photon release of stabilized GFPmut2 is only
slightly smaller than the number of photons released by QDs 25 at 20 kW/cm2. The
−0.5
dependence of the total photon number detected per molecule scales as I
therefore
suggesting that the highest excitation power is not actually the best for fluorescence
measurements on single molecules.
Furthermore, the time needed to bleach the GFPmut2 irreversibly, Tbleach, was used as a
good parameter to quantify the protein stability since it also largely affects the overall
brightness of the caged proteins. A single step switch on/switch off mechanism under highenergy excitation serves as a proof that single molecules were measured, see Fig. 4.
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9821
Fig. 4. Fluorescence trace of proteins belonging to family A(1) as function of the excitation
energy at 6.8 kW/cm2 (red), 9.5 kW/cm2 (blue), and 12 kW/cm2 (green). Enlargement of the red
trace shows the short switch-off time, inset (A). The frequency of blinking in dependence of
the intensity reveals very low values, inset (B). The red solid line is the linear fit: ν= A + B
[Intensity] of the data with the slope B = 2.3 ± 0.5 cm2 Hz/kW.
A relevant result in terms of photo-stability is that the measured bleaching time for the caged
proteins with 4 min is higher than that for CdSe/ZnS QDs in air with 2.5 or 3.5 min [16].
GFPmut2, trapped in gel or in the amorphous carbonate core, exhibits fluorescent
properties similar to those of S(1) family (Table 1). For polyelectrolyte-GFPmut2 spin-coated
layers, we observed a more stable species in the PSS environment (PSS/GFP), similar to the
A(1) and A(2) families related to GFPmut2 molecules trapped within the nanocapsule (Table 1).
Moreover, we found a less stable species in the spin coated PAH environment (PAH/GFP)
and similar to the species B observed in the nanocapsule layer (Table 1). These similarities
point out that family A(1) can be related to proteins incorporated in the PSS layer. As a proof
of concept, the B family for within the polyelectrolyte layer has a fluorescent performance
close to the PAH/GFP spin coated ones. For inside the layer, a less stable population, named
C, was observed. We assume that this family is populated by proteins trapped in a sort of
“nowhere land” between the PAH and the PSS layer.
Comparing the results for GFPmut2 in contact to nanocapsule layers with those of spin
coated and uniquely interacting with a single polyion (PAH or PSS) component, the best
stabilization was achieved by incorporation in a negative environment. However, also caging
in positive polyelectrolyte produced a significant improvement of the fluorescence properties
in comparison to wet silica gel or amorphous calcium carbonate entrapment, see Table 1 and
Fig. 3. In the last case, it can be assumed that proteins were in water-filled but comparably
wide pores with respect to the protein size.
In comparison to QD the caged-GFPmut2 shows some clear advantages. For example,
clear spectral differences were observed for the time evolution of QD emission, even for dots
of the same batch in the same atmosphere, resulting in a 30 nm blue shift [16]. In the cagedGFPmut2 case, the spectral emission is stable with time. Besides, the number of photons
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9822
emitted before bleaching from caged-GFPmut2 under two-photon excitation regime is
particularly interesting. It was shown that using lower excitation energy one could prolong the
bleaching time. This condition shifted the bleaching time in the range of several minutes. Due
to the stabilization of the proteins against irreversible photo-destruction for a long time a high
number of photons can be collected in dependence of the excitation energy and of the
stabilization method being used as shown in Fig. 3. Another interesting feature is the blinking
behavior, Fig. 4, of the GFPmut2 proteins as compared to QDs high frequency blinking [16]
that constitutes a problem in some applications. For the caged GFPmut2, at all measured
excitation intensities, the maximum blinking frequency, ν, is <<<1 Hz, Fig. 4, and the switchoff and switch-on time (Δtoff and Δton) are 950 ± 50 ms and 13 ± 3 min respectively. Because
the Δton and ν have a quadratic dependence on the excitation intensity, we found that at 1,3
kW/cm2 the Δton value for the GFPmut2 caged in PSS is about twenty times larger than the
one measured for cadmium selenide nanocrystals [16, 24]. Moreover, the quantum yield of the
GFPmut2 in PSS and PAH layers resulted larger than the one measured for CdSe/ZnS [24]
and for GFPmut2 in gel [4, 21], see Table 1.
Now, the main point of a possible explanation of the results, even if at a speculative
stage, can be related to the consideration of the electrostatic interactions between the outer
barrel of the GFP and the polyelectrolyte, and of the known enhancement effect of rigidity on
the GFP chromophore. We are thinking about a tight wrapping of the polyelectrolyte chains
around the protein. Such a charged environment should be responsible for a support and
stabilization of the protein structure as well as of the fluorophore inside the β-barrel structure
of GFPmut2 [27, 28].
A
B
C
Fig. 5. Proposed model for the GFP interaction with PSS polyelectrolyte [27]. (a) (deepview of
Swiss-PDBview) [29]. The positively charge amino acids (red) are on a rim around the
fluorophore (yellow). (b) Sketch of GFP charge distribution. (c) Sketch of GFP in a PSS (blue)
charged environment. A rim of positive charges near the fluorophore induces a tightening and
stiffening of the beta barrel, a better protection and an increased time before bleaching. (Sketch
originally elaborated by S.Krol).
GFP charge distribution (Figs. 5(a) and 5(b)) presents a rim of positive charges (red) around
the fluorophore (yellow) [28]. The interaction between these and the negative charges of PSS
might induce a tightening and stiffening of the beta-barrel near the fluorophore (Fig. 5(c)).
This more rigid environment would lead to an increased quantum yield [2] and also to a
protection of the protein against bleaching and denaturation [29].
4. Conclusions
Summarizing the results, we conclude that a close wrapping of the green fluorescent protein in
polyelectrolyte chains could improve their fluorescence properties (photon release, blinking
behavior, photobleaching) reaching an order of magnitude known before only for dyes under
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9823
vacuum [15] and for QDs [16]. This points out to the possibility of realizing a new class of
fluorescent markers as potential competitors of QDs. The proposed model can also serve for
designing more efficient GFP in terms of stability. Moreover, the results reported address a
potential application of single caged-GFPmut2 proteins as fluorescent markers in biological
systems, based on the high stability in combination with the other photo-physical parameters
and with biocompatibility. Interesting developments could also come from the utilization of
switchable GFP variants. In this case, the enhanced properties due to polyelectrolyte caging
can be exploited in the developments of single-protein based optical memory arrays.
Acknowledgments
This research was partially supported by FIRB and PRIN grants of the Ministry of Research
and Education of Italy to GC and AD. LAMBS-MicroScoBio was supported by IFOM (FIRC
Institute of Molecular Oncology, Milan, Italy). SK was supported by the European
Community BARP+ grant contract. We thank BP Cormack (Stanford University School of
Medicine, Stanford, CA) and A Mozzarelli (University of Parma, Italy) for GFPmut2
availability.
#74009 - $15.00 USD
(C) 2006 OSA
Received 11 August 2006; revised 12 September 2006; accepted 13 September 2006
16 October 2006 / Vol. 14, No. 21 / OPTICS EXPRESS 9824