Solvatochromism of the green fluorescence protein chromophore and its derivatives

Article abstract of DOI:10.1021/ja063128a

The solvatochromic behavior of the green fluorescence protein (GFP) chromophore (p-hydroxybenzylideneimidazolone, p-HBDI) and its derivatives (p-methoxybenzylideneimidazolone, p-MeOBDI, and N-methyl-p-hydroxybenzylideneimidazolonium iodide, p-HBDIMe+) was

Full text of DOI:10.1021/ja063128a

Published on Web 08/25/2006
Solvatochromism of the Green Fluorescence Protein Chromophore and Its
Derivatives
Jian Dong, Kyril M. Solntsev,* and Laren M. Tolbert*
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332-0400
Received May 4, 2006; E-mail: solntsev@gatech.edu; tolbert@chemistry.gatech.edu
Green fluorescent protein (GFP) and its mutants have revolution-
ized molecular biology.1 Despite the effort toward expanding the
class of GFP through gene shuffling techniques, successful examples
invariably incorporate the â-barrel to isolate the chromophore. The
â-barrel, presumably by restricting the conformational space of the
chromophore, reduces the rate of radiationless decay, leading to
ca. 104 higher fluorescence quantum yields. At the same time, the
â-barrel can apparently tolerate a high degree of disorder within
the protein, as judged by analysis of the numerous crystal structures,
as well as a variety of π systems. The â-barrel obviously plays an
enormous role in the photophysics of GFP; its most obvious effect
is to solvate both the chromophore and its conjugate base. In this
work we consider the effect of the â-barrel on the optical properties
of the GFP chromophore (p-hydroxybenzylideneimidazolone, p-
HBDI) experimentally by selective variation of the protonation state
of the latter and its microenviroment.
It is known that a careful pH titration indicates three forms of
p-HBDI: a (presumably nitrogen-) protonated cationic form C
below pH ) 2, the neutral form N, and the anionic form A above
pH ) 10.2 The absorbance spectrum of wild-type (wt)-GFP
demonstrates the presence of the two forms of the chromophore
with absorbance bands at 398 (Ap band) and 477 nm (Bp band,
Figure 1), their ratio very weakly depending on pH. It has been
noted that the high-energy wt-GFP absorbance band corresponds
to the absorbance of the p-HBDI cation in water (Figure 1).
Nevertheless, most of the current literature assigns band Ap to the
absorbance of the neutral species, not the cation, because no
evidence of the latter was found in the Raman and resonance Raman
spectra of wt-GFP and its S65T mutant,3 and the imidazolone
nitrogen atom is not sufficiently basic to be protonated in the
protein. As a reference point, the analogous spectra of p-HBDI at
different pH were used.4,5 Despite this observation, the possibility
Figure 1. Absorption spectra of different protonation states of p-HBDI,
p-MeOBDI, and p-HBDIMe+ in various solvents. For comparison we
present the absorption spectra of wt-GFP at pH 7 in the “p-HBDI C” graph
of cationic involvement in the absorbance spectra of the wt-GFP
cannot be excluded.6,7
as a black dash-dotted line. Absorbance of p-HBDIMe+ in water at pH
Only scattered experimental treatments of solvatochromism in
p-HBDI and its derivatives are known.2,8-10 A very weak solvent
dependence of the N form absorption was reported,8 while the
absorbance of the C form exhibited a blue shift with polarity
increase.9 The hypsochromic shift of the A form was correlated
with either solvent polarity8 or acidity2,10 increase. At the same time,
it was found that mutations influence the position of the absorption
bands of the chromophore to a much greater extent than simple
solvent variations.11 In all cases, no quantitatiVe analysis was made.
To more clearly understand the role of the solvent and protein in
mediating the photophysics, and to correlate the absorbance bands
of the chromophore in bulk solvents and in protein, we have
undertaken a careful evaluation of the solvent effects on the various
acid/base forms of p-HBDI and its new derivatives, O- and
N-methylated analogues. The selective blocking of the acid-base
groups in the chromophore may give insight into the relative
7.6 is presented in the bottom graph as a black dotted line; the absorption
peak at 486 nm is due to the zwitterion (Z) absorption. The intensity of the
latter is scaled to fit the graph.
contribution of the various protonated form of the chromophore to
the ground-state behavior of the protein.
The GFP chromophore p-HBDI was synthesized according to
ref 12. Its O-methylated derivative, p-MeOBDI, was synthesized
by a slightly modified procedure using p-methoxybenzaldehyde.
The N-methylated derivative, p-HBDIMe+, was synthesized by
methylation of p-HBDI with methyl trifluoromethanesulfonate.13
Similar to p-HBDI, both new compounds had negligible fluores-
cence in bulk solvents at room temperature.
The C forms of p-HBDI and p-MeOBDI were generated by
adding trifluoromethanesulfonic acid to their neutral solutions, while
the A forms of p-HBDI were formed by adding KOH/crown ether
complex. In aqueous solutions, the pH was controlled to ensure
9
12038
J. AM. CHEM. SOC. 2006, 128, 12038-12039
10.1021/ja063128a CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Solvatochromic Parameters (in 103/cm) Used in
Multivariable Regression Fits of Absorption Data for Neutral,
Cationic, and Anionic Forms of Various Chromophores According
to Eq 1; pKa Values Are Also Presented13
misassigned and the Ap band corresponds to the cation, leaving
the 477 nm absorption to the zwitterion.7 However, as we have
discovered, the red-shifted absorbance maximum of the Z form does
not match this value in any solvent, while the p-HBDI A absorption
in some basic solvents fits the 477 nm wt-GFP absorption quite
well (Figure 1). Thus, there is no reason to discount the common
assignment of Ap as the neutral form. To explain the existing
discrepancy between the experimental data on neutral species
absorption in p-HBDI and red-shifted wt-GFP, a combination of
mainly structural factors (small breakdown of planarity), with a
small contribution from polarization effects of the environment,19,20
must be taken into account. Equally important could be the presence
in GFP and all its fluorescent mutants of the obligatory positively
charged Arg96. Calculations indicate20 that Arg96 red-shifts the
absorption spectrum of the chromophore by ca. 40 nm, presumably
by lowering the energy of the charge-transfer excited state.
Nevertheless, this is the first experimental verification of this
rationale.
The complex solvatochromic behavior of the GFP chromophore
and its derivatives requires both polar and acid/basic properties of
the environment to be taken into account. However, no solvent
effect serves to “turn on” the fluorescence, which remains subject
to stereoelectronic effects. This stark contrast between solution and
protein-constrained behavior suggests that there is something truly
unusual about this chromophore, which is the subject of continued
investigation in many laboratories.
b
compound
ν0
p
a
b
Ra
pKa
p-HBDI N
p-HBDI C
p-HBDI A
p-MeOBDI N
p-MeOBDI C
p-HBDIMe + C
28.3
24.5
22.9
27.9
24.0
23.0
-0.71
1.2
-1.4
-0.6
0.8
-0.17
-0.7
0.94
0.90
0.89
0.87
0.95
0.92
8.53
2.36
0
1.7
0
0.5
0.53
-0.18
1.8
0
2.76
6.4
2.2
-1.3
1.9
a Correlation coefficient. b Measured in MeOH/H2O 1:1 v/v.
complete (de)protonation.13 The basicity of the imidazole nitrogen
increased slightly upon methylation of the hydroxy group in
p-HBDI, while the acidity of the hydroxy group increased by more
than 2 pKa units upon methylation of the imidazole nitrogen (Table
1). Titration of p-HBDIMe+ in the region pH ) 4-7.4 demon-
strated13 a clear equilibrium between C and its conjugate base, the
zwitterion Z that absorbs at 486 nm in water (Figure 1). This is the
first experimental observation of the ground-state GFP chromophore
zwitterion that allows the investigation of the long-debated p-HBDI
tautomer.6,7,14,15 In solutions with pH > 8.0, the Z form decomposed
rapidly,13 probably by hydrolysis of the imidazolone ring. For the
same reason, we could not detect absorbance of the Z in nonaqueous
solvents in the presence of base; however, trace amounts of Z were
observed in several neutral solvents.
Each of these forms showed a complex solvatochromic behavior
(see Figure 1). The absorption maxima of the N forms of the
molecules had the weakest solvent dependence, while those of C
and especially A forms varied over a very broad range. To analyze
the solvatochromic behavior of the N, C, and A forms of the
chromophores, we used the Kamlet-Taft multivariant approach,16
which correlates the spectral shift ν of the solute with the solvent
parameters that are responsible for its acidic (R), basic (â), and
polar solvating (π*) properties:
Acknowledgment. We thank the National Science Foundation
(CHE-0456892) for financial support.
Supporting Information Available: Synthesis of p-MeOBDI and
p-HBDIMe+, experimental data, and fitting procedures. This material
is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Zimmer, M. Chem. ReV. 2002, 102, 759-781.
(2) Bell, A. F.; He, X.; Wachter, R. M.; Tonge, P. T. Biochemistry 2000, 39,
4423-4431.
ν (1000/cm) ) ν0 + pπ* + aR + bâ
(1)
(3) Bell, A. F.; Stoner-Ma, D.; Wachter, R. M.; Tonge, P. J. J. Am. Chem.
Soc. 2003, 125, 6919-6926.
(4) He, X.; Bell, A. F.; Tonge, P. T. J. Phys. Chem. B 2002, 106, 6056-
We17 and others18 have successfully used this approach for
various hydroxyaromatic compounds. It allows a straightforward
separation of selective (H-bonding) and nonselective (dipole-dipole
interaction) solvation. Here we report that the solvatochromic
behavior of p-HBDI and its derivatives is governed by both polar
and acid/base properties of the solvents. The magnitudes and
directions of the solvatochromic shifts strongly depend on the
protonation state of the solute (Table 1). For N forms of p-HBDI
and p-MeOBDI, an increase in all solvent parameters induced a
bathochromic shift. As with the naphthols,17 the presence of a
hydroxy group in p-HBDI makes the solvent basicity (selective
solvation) as important as the polar solvation factor. The p-HBDI
cationic form blue-shifts as solvent polarity and basicity are
increased, while for the anionic form the polarity and acidity of
the solvent work in opposite directions! The magnitude of the solute
parameters p can be related to the relative dipole moments of the
molecules, while proton susceptibility parameters a and b reflect
the relative proton basicity and acidity of the chromophore. Our
analysis clearly demonstrates an increase in dipole moment of the
chromophores upon ionization and shows their amphoteric behavior.
p-HBDIMe+ C shows the same absorption pattern for the p-HBDI
C, validating the protonation site in p-HBDI.
6066.
(5) Schellenberg, P.; Johnson, E.; Esposito, A. P.; Reid, P. J.; Parson, W. W.
J. Phys. Chem. B 2001, 105, 5316-5322.
(6) El Yazal, J.; Prendergast, F. G.; Shaw, D. E.; Pang, Y.-P. J. Am. Chem.
Soc. 2000, 122, 11411-11415.
(7) Voityuk, A. A.; Michel-Beyerle, M.-E.; Ro¨sch, N. Chem. Phys. 1998,
231, 13-25.
(8) Niwa, H.; Inouye, S.; Hirano, T.; Matsuno, T.; Kojima, S.; Kubota, M.;
Ohashi, M.; Tsuji, F. I. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 13617-
13622.
(9) Voityuk, A. A.; Kummer, A. D.; Michel-Beyerle, M.-E.; Ro¨sch, N. Chem.
Phys. 2001, 269, 83-91.
(10) Yampolsky, I. V.; Remington, S. J.; Martynov, V. I.; Potapov, V. K.;
Lukyanov, S.; Lukyanov, K. A. Biochemistry 2005, 44, 5788-5793.
(11) Jung, G.; Wiehler, J.; Zumbusch, A. Biophys. J. 2005, 88, 1932-1947.
(12) Kojima, S.; Ohkawa, H.; Hirano, T.; Maki, S.; Niwa, H.; Ohashi, M.;
Inouye, S.; Tsuji, F. I. Tetrahedron Lett. 1998, 39, 5239-5242.
(13) See Supporting Information for more details.
(14) Weber, W.; Helms, V.; McCammon, J. A.; Langhoff, P. W. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 6177-6182.
(15) Scharnagl, C.; Raupp-Kossmann, R. A. J. Phys. Chem. B 2004, 108, 477-
489.
(16) Kamlet, M. J.; Abboud, J.-L. M.; Abraham, M. H.; Taft, R. W. J. Org.
Chem. 1983, 48, 2877-2887.
(17) (a) Solntsev, K. M.; Huppert, D.; Tolbert, L. M.; Agmon, N. J. Am. Chem.
Soc. 1998, 120, 7981-7982. (b) Solntsev, K. M.; Huppert, D.; Agmon,
N. J. Phys. Chem. A 1998, 102, 9599-9606.
(18) (a) Mijin, D. Z.; Uscumlic, G. S.; Perisic-Janjic, N. U.; Valentic, N. V.
Chem. Phys. Lett. 2006, 418, 223-229. (b) Reta, M.; Cattana, R.; Silber,
J. J. J. Solut. Chem. 2001, 30, 237-252. (c) Magnes, B.-Z.; Pines, D.;
Strashnikova, N.; Pines, E. Solid State Ionics 2004, 168, 225-233.
(19) Marques, M. A. L.; Lo´pez, X.; Varsano, D.; Castro, A.; Rubio, A. Phys.
ReV. Lett. 2003, 90, 258101.
We note, however, that at no polarity/acidity extreme did the
absorption of the p-HBDI neutral correspond to those in the protein.
We are tempted to conclude that the wt-GFP absorption was
(20) Laino, T.; Nifosi, R.; Tozzini, V. Chem. Phys. 2004, 298, 17-28.
JA063128A
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12039