Synthesis and spectroscopic studies of model red fluorescent protein chromophores.

Article abstract of DOI:10.1021/ol0200403

[reaction: see text]. Here we describe the synthesis and spectroscopic characterization of two compounds designed to model the chromophore in DsRed, a red fluorescent protein. Comparison with model green fluorescent protein (GFP) chromophores indicates th

Full text of DOI:10.1021/ol0200403

ORGANIC
LETTERS
2002
Vol. 4, No. 9
1523-1526
Synthesis and Spectroscopic Studies of
Model Red Fluorescent Protein
Chromophores
Xiang He, Alasdair F. Bell, and Peter J. Tonge*
Department of Chemistry, State UniVersity of New York at Stony Brook,
Stony Brook, New York 11794-3400
peter.tonge@sunysb.edu
Received February 19, 2002
ABSTRACT
Here we describe the synthesis and spectroscopic characterization of two compounds designed to model the chromophore in DsRed, a red
fluorescent protein. Comparison with model green fluorescent protein (GFP) chromophores indicates that the additional conjugation in the
DsRed models can account, in part, for the red-shifted absorption and emission properties of DsRed compared to those of GFP. In contrast
to the GFP models, the DsRed models are fluorescent with quantum yields of 0.002−0.01 in CHCl3.
DsRed is a recently cloned bright red fluorescent protein from
coral (Discosoma sp.) with great potential as a complement
to the widely utilized Aequorea Victoria green fluorescent
protein (GFP)1 due to its significantly red-shifted excitation
and emission maxima (λex 558 nm, λem 583 nm).2 Although
DsRed and GFP share only 23% sequence identity, X-ray
crystallography reveals that DsRed also forms the â-can
structural motif observed for GFP.3,4 In addition, like GFP,
the chromophore in DsRed is formed by the intramolecular
reaction of three amino acid residues (Q66Y67G68). How-
ever, while both the DsRed and GFP chromophores share
the same 4-hydroxybenzylideneimidazolinone core, the chro-
mophore in DsRed differs from that in GFP by the presence
of an acylimino group at the 2-position of the imidazolinone
ring (Figure 1). The acylimino group is formed by oxidation
of the Q66 CR-N bond, and the resulting additional
conjugation in the chromophore is believed to partly account
for the red-shifted spectroscopic properties of DsRed com-
pared to those of GFP.3-6 To study the role of the protein
(1) Tsien, R. Y. Annu. ReV. Biochem. 1998, 67, 509-544.
(2) Matz, M. V.; Fradkov, A. F.; Labas, Y. A.; Savitsky, A. P.; Zaraisky,
A. G.; Markelov, M. L.; Lukyanov, S. A. Nat. Biotechnol. 1999, 17, 969-
973.
(3) Wall, M. A.; Socolich, M.; Ranganathan, R. Nat. Struct. Biol. 2000,
7, 1133-1138.
(4) Yarbrough, D.; Wachter, R. M.; Kallio, K.; Matz, M. V.; Remington,
S. J. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 462-467.
Figure 1. Structure of the DsRed chromophore, HBMPI 1,
HBMPDI 2, and HBDI.
10.1021/ol0200403 CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/09/2002

matrix in governing the spectral properties of the DsRed
chromophore, we have synthesized 4-hydroxybenzylidene-
1-methyl-2-propenyl-imidazolinone (1, HBMPI) and 4-hy-
droxybenzylidene-1-methyl-2-penta-1,3-dien-1-yl-imidazoli-
none (2, HBMPDI) (Figure 1) and performed an initial
characterization using absorption, fluorescence, and Raman
spectroscopies.
Scheme 2. Synthesis of HBMPDI 2
Due to the susceptibility of acylimines to nucleophilic
attack,7,8 we have initially focused on DsRed model com-
pounds containing olefinic substituents on the imidazolinone
ring. Two such compounds (1 and 2) were prepared via
Erlenmeyer azlactone synthesis as shown in Schemes 1 and
2, respectively.9,10
Scheme 1. Synthesis of HBMPI 1
20% overall yield. 4-Hydroxybenzylidene-1,2-dimethyl-
imidazolinone (HBDI, Figure 1) was also isolated in high
yield from the synthesis of 1. Similarly, HBMPDI 2 was
synthesized using the same strategy with 2,4-hexadienoic acid
as the initial acylating group. However, the 2-methyl-
substituted byproduct, observed in the synthesis of 1, was
not formed, presumably due to steric hindrance in the acetyl
exchange reaction. Compound 2 was obtained in 25% overall
yield. The absorption spectra of 1 and 2 are shown in Figures
2 and 3, respectively.
For compound 1, glycine was first acylated with crotonic
acid under alkaline conditions.11 Reaction of N-crotonylg-
lycine with 4-hydroxybenzaldehyde and anhydrous sodium
acetate in acetic anhydride provided the corresponding
intermediate azlactone 3 together with a byproduct having a
methyl group in place of the crotonyl group at the imida-
zolinone C2 position. The formation of the byproduct is
postulated to result from replacement of the N-crotonyl group
with an acetyl group derived from the solvent.
1H NMR analysis indicated that the product from the first
step of the synthesis contained more than 60% of the desired
azlactone, and this was used for the next step of the reaction
without purification. Subsequent treatment of 3 with me-
thylamine in the presence of K2CO3 afforded HBMPI 1 in
(5) Baird, G. S.; Zacharias, D. A.; Tsien, R. Y. Proc. Natl. Acad. Sci.
U.S.A 2000, 97, 11984-11989.
Figure 2. Absorption spectra for the cationic ([[[, 1 M HCl),
neutral (s[s[, 20 mM sodium acetate pH 5.5), and anionic (s
s, 1 M KOH) forms of HBMPI 1.
(6) Gross, L. A.; Baird, G. S.; Hoffman, R. C.; Baldridge, K. K.; Tsien,
R. Y. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 11990-11995.
(7) Kupfer, R.; Meier, S.; Wurthwein, E. U. Synthesis 1984, 688-690.
(8) Malassa, I.; Matthies, D. Chemiker-Zeitung 1987, 111, 253-261.
(9) Buck, J. S.; Ide, W. S. Organic Syntheses; Wiley: New York, 1943;
Collect. Vol. II, p 55-56.
(10) Kojima, S.; Ohkawa, H.; Hirano, T.; Maki, S.; Niwa, H.; Ohashi,
M.; Inouye, S.; Tsuji, F. I. Tetrahedron Lett. 1998, 39, 5239-5242.
(11) Sheehan, J. C.; Duggins, W. E. J. Am. Chem. Soc. 1950, 72, 2475-
2477.
Due to the extension of the conjugated system compared
to HBDI (λmax ) 368 and 425 nm for neutral and anionic
forms, respectively), the absorption maxima of the neutral
1524
Org. Lett., Vol. 4, No. 9, 2002

Figure 3. Absorption spectra for the cationic ([[[, 1 M HCl),
neutral (s[s[, 20 mM sodium acetate pH 5.5), and anionic (s
s, 1 M KOH) forms of HBMPDI 2.
forms of 1 and 2 are red-shifted to 401 and 421 nm,
respectively (Table 1). Similarly, for the anionic forms, the
λ
max of 1 is shifted to 460 nm, while the λmax of 2 is shifted
to 482 nm. The observation that λmax of anionic 1 is red
shifted 35 nm compared to that of anionic HBDI is in good
agreement with the 34 nm red shift predicted on the basis
of ZINDO calculations by Gross et al.6 for the addition of
one ethylenic group to the GFP chromophore. It is worth
noting that the anionic form of 1 has a λmax similar to that
of alkaline-denatured DsRed (λmax 452 nm). Gross et al. have
proposed that alkaline denaturation of DsRed results in
hydrolyis of the acylimino group and cleavage of the peptide
backbone.6 The product resulting from acylimine hydrolysis
is expected to have a carbonyl substituent at the imidazoli-
none C2 position, consistent with the observed similarity
between λmax for alkaline-denatured DsRed and the anionic
form of 1.
Figure 4. Raman spectra of HBDI, HBMPI 1, and HBMPDI 2 in
sodium acetate buffer (pH 5.5, 20 mM) obtained using 100 mW of
752 nm laser excitation.15
of 0.0005-0.0008 in H2O, consistent with the studies by
McCapra et al. on a related GFP model compound.13 In
CHCl3, the quantum yields rise to 0.002 and 0.01 for the
neutral forms of HBMPI and HBMPDI, respectively, similar
to the values observed for the model GFP chromophores
bearing a phenyl substituent on the imidazolinone ring.14
Most relevant for studies of DsRed is the anionic form of
compound 2, which has an excitation maximum at 482 nm
and an emission maximum at 565 nm. While the quantum
yields for 1 and 2 are still significantly lower than observed
for DsRed (0.7),6 the similarity in λem for the anionic form
of 2 and the DsRed protein (λem 583 nm) indicates that the
additional conjugation in 2 compared to that in the GFP
model HBDI can account for a large part of the red shift
observed between GFP and DsRed. However, the Stokes shift
in compound 2 is 83 nm, whereas it is only 25 nm in the
protein. The smaller Stokes shift for the protein compared
to the models indicates that less energy is needed to
reorganize the environment to accommodate the excited-state
structure in the protein than in the solvent.
Table 1. Spectroscopic Data of HBMPI (1), HBMPDI (2), and
DsRed
1
2
DsRed
neutral
anion
neutral
anion anion
solvent
H2O CHCl3 H2O
H2O
421
523
CHCl3 H2O
H2O
558
583
λmax (nm) 401
λem (nm) 460
400
472
460
530
425
531
482
565
φa
0.0005 0.002 0.0006 0.0008 0.01 0.0008 0.7
a Quantum yield (φ).12
The newly synthesized DsRed model chromophores have
(12) White, C. E.; Argauer, R. J. Fluorescence Analysis; A Practical
Approach; Dekker: New York, 1970.
(13) McCapra, F.; Razavi, Z.; Neary, A. P. J. Chem. Soc., Chem.
Commun. 1988, 790-791.
(14) You, Y. J.; He, Y. K.; Borrows, P. E.; Forrest, S. R.; Petasis, N.
A.; Thompson, M. E. AdV. Mater. 2000, 12, 1678-1681.
interesting fluorescence properties as listed in Table 1. While
HBDI, the model GFP chromophore, is not detectably
fluorescent at room temperature, both HBMPI and HBMPDI
have measurable fluorescence spectra with quantum yields
Org. Lett., Vol. 4, No. 9, 2002
1525

Raman spectroscopy has provided valuable information
on the structure of the chromophore in GFP model com-
pounds and proteins.15-18 To provide a basis for interpreting
the Raman spectra of wild-type and mutant DsRed proteins,
we have undertaken a Raman analysis of compounds 1 and
2.
1143 cm-1 band in HBDI, which corresponds to an imida-
zolinone C-C single-bond stretching mode, is shifted to 1148
and 1156 cm-1 for 1 and 2, respectively, indicating the
strengthening of the single bonds in the conjugated system.
In conclusion, chromophores 1 and 2 provide very useful
models for improving our understanding of the structure and
optical properties of the chromophore within DsRed. While
some differences are expected between model and protein
data due to the use of an ethylenic substituent, the results
suggest that a significant portion of the red shift in absorption
and emission properties between GFP and DsRed can be
accounted for by the extended conjugation in the DsRed
chromophore. Like GFP, the DsRed protein matrix plays a
major role in preventing nonradiative deactivation of the
chromophore’s excited state.
Figure 4 contains the Raman spectra of 1 and 2 together
with that of HBDI, a GFP model compound that has been
studied in detail using ab initio calculations and isotopic
labeling.19 While there are many similarities in the spectra
of the three compounds, reflecting the 4-hydroxybenzylide-
neimidazolinone structure common to each molecule, two
important differences can be observed in the model com-
pound series. Specifically, the 1567 cm-1 band in neutral
HBDI is shifted to 1530 and 1515 cm-1 in neutral 1 and 2,
respectively. This band has been assigned in HBDI to a mode
delocalized over the imidazolinone ring and exocyclic Cd
C bond.19 The decrease in wavenumber for this band is
consistent with increased delocalization as conjugation
increases through the series of model compounds. In HBMPI
and HBMPDI, the equivalent mode may also involve
contributions from the olefinic substituents. In addition, the
Acknowledgment. This work was supported in part by
NIH Grants AI44639 and GM63121. In addition, this
material is based upon work supported in part by the U.S.
Army Research Office under Grant DAAG55-97-1-0083.
A.F.B. is an American Heart Association postdoctoral fellow.
The NMR facility at SUNY at Stony Brook is supported by
grants from NSF (CHE9413510) and NIH (1S10RR554701).
(15) Bell, A. F.; He, X.; Wachter, R. M.; Tonge, P. J. Biochemistry 2000,
39, 4423-4431.
(16) Esposito, A. P.; Schellenberg, P.; Parson, W. W.; Reid, P. J. J. Mol.
Struct. 2001, 569, 25-41.
(17) Schellenberg, P.; Johnson, E.; Esposito, A. P.; Reid, P. J.; Parson,
W. W. J. Phys. Chem. B 2001, 105, 5316-5322.
(18) Tozzini, V.; Nifosi, R. J. Phys. Chem. B 2001, 105, 5797-5803.
(19) He, X.; Bell, A. F.; Tonge, P. J. J. Phys. Chem. B, submitted for
publication.
Supporting Information Available: Procedures for the
preparation of 1 and 2 and related analytical data. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL0200403
1526
Org. Lett., Vol. 4, No. 9, 2002