Convergent preparation and photophysical characterization of dimaleimide dansyl fluorogens: Elucidation of the maleimide fluorescence quenching mechanism

Article abstract of DOI:10.1021/ja0738125

Dimaleimide fluorogens are being developed for application to fluorescent protein labeling. In this method, fluorophores bearing two maleimide quenching groups do not fluoresce until both maleimide groups have undergone thiol addition reactions with the C

Full text of DOI:10.1021/ja0738125

Published on Web 09/12/2007
Convergent Preparation and Photophysical Characterization
of Dimaleimide Dansyl Fluorogens: Elucidation of the
Maleimide Fluorescence Quenching Mechanism
Julia Guy,† Karine Caron,† Ste´phane Dufresne,† Stephen W. Michnick,‡
W.G. Skene,*,† and Jeffrey W. Keillor*,†,‡
Contribution from the Department of Chemistry and Department of Biochemistry, UniVersite´ de
Montre´al, C.P. 6128, Succursale centre-Ville, Montre´al, Que´bec, H3C 3J7, Canada
Received May 26, 2007; E-mail: w.skene@umontreal.ca; jw.keillor@umontreal.ca
Abstract: Dimaleimide fluorogens are being developed for application to fluorescent protein labeling. In
this method, fluorophores bearing two maleimide quenching groups do not fluoresce until both maleimide
groups have undergone thiol addition reactions with the Cys residues of the target protein sequence [J.
Am. Chem. Soc. 2005, 127, 559-566]. In this work, a new convergent synthetic route was developed that
would allow any fluorophore to be attached via a linker to a dimaleimide moiety in a modular fashion.
Series of dimaleimide and dansyl derivatives were thus prepared conveniently and used to elucidate the
mechanism of maleimide quenching. Intersystem crossing was ruled out as a potential quenching pathway,
based on the absence of a detectable triplet intermediate by laser flash photolysis. Stern-Volmer rate
constants were measured with exogenous dimaleimide quenchers and found to be close to the diffusion-
controlled limits, consistent with electron transfer being thermodynamically favorable. The thermodynamic
feasibility of the photoinduced electron transfer (PET) quenching mechanism was verified by cyclic
voltammetry. The redox potentials measured for dansyl and maleimide confirm that electron transfer from
the dansyl excited state to a pendant maleimide group is exergonic and is responsible for fluorescence
quenching of the fluorogens studied herein. Taking this PET quenching mechanism into account, future
fluorogenic protein labeling agents will be designed with spacers of variable length and rigidity to probe
the structure-property PET efficiency relationship.
the cellular localization of proteins at a whole genome scale,5
but there remain some limitations for this method. For GFP to
Introduction
Categorizing the huge number of gene products discovered
in this post-genomic era remains a monumental task. New
complementary protein labeling techniques are required to
accelerate and facilitate the process. The fluorescent labeling
of proteins has proven to be a powerful approach for following
the dynamic process of their synthesis and degradation, in
addition to determining their localization and protein-protein
interactions. Many labeling techniques have been developed that
are based on the reaction of fluorescent dyes, through succin-
imidyl ester or maleimide functional groups, with primary
amines or thiol groups exposed on the surface of a protein.1-3
However, the frequency of such functional groups on protein
surfaces renders these labeling techniques nonspecific. One
broadly applied method for labeling specific target proteins is
their genetic fusion to fluorescent proteins such as jellyfish green
fluorescent protein (GFP). This approach has been used to study
individual protein dynamics in living cells4 and to determine
function as a fluorophore, it must be properly folded into its
11-stranded â-barrel structure; however, maturation into a
soluble, fluorescent protein is not spontaneous and may be
problematic, especially at or above room temperature.6 GFP
fluorescence is also sensitive to the environment of its fusion
with test proteins and can be difficult to distinguish from the
autofluorescent background of living cells. GFP variants have
been produced through protein engineering that are able to
overcome some of these limitations, but the steric bulk and
metabolic stability of a 27 kDa â-barrel protein still represents
the potential for significant perturbation of the dynamics and
localization of test proteins.4,6
Alternative protein labeling strategies have been developed
that address the steric bulk and stability problems. For example,
Tsien and co-workers have elaborated on a method7 that makes
use of a short peptide sequence, optimized to form a â-turn
motif,8 that contains four Cys residues positioned to allow their
reaction with a biarsenical fluorogenic compound. The resulting
† Department of Chemistry.
‡ Department of Biochemistry.
(1) For example, see Haughland, R.P. In Handbook of Fluorescent Probes and
Research Chemicals, 5th ed.; Molecular Probes: Eugene, OR, 1992.
(2) Sipple, T. O. J. Histochem. Cytochem. 1981, 29, 314.
(3) Corrie, J. E. T. J. Chem. Soc., Perkin Trans. 1 1994, 2975.
(4) For a recent review, see: Zhang, J.; Campbell, R. E.; Ting, A. Y.; Tsien,
R. Y. Nat. ReV. Mol. Cell Biol. 2002, 3, 906.
(5) Huh, W.-K.; Falvo, J. V.; Gerke, L. C.; Carroll, A. S.; Howson, R. W.;
Weissman, J. S.; O’Shea, E. K. Nature 2003, 425, 686.
(6) Tsien, R. Y. Annu. ReV. Biochem. 1998, 67, 509.
(7) Griffin, B. A.; Adams, S. R.; Tsien, R. Y. Science 1998, 281, 269.
(8) Adams, S. R.; Campbell, R. E.; Gross, L. A.; Martin, B. R.; Walkup, G.
K.; Yao, Y.; Llopis, J.; Tsien, R. Y. J. Am. Chem. Soc. 2002, 124, 6063.
9
10.1021/ja0738125 CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 11969-11977
11969

A R T I C L E S
Guy et al.
Chart 1
biarsenical protein adduct is far more fluorescent than the initial
fluorogen. The probe peptide sequence is small (only 16 aa in
its typical application8), minimizing perturbation of the target
protein, and presents a hexapeptide 4-Cys motif (CCPGCC) that
confers enough specificity for the biarsenical fluorogen that it
may be used in ViVo.8 The toxicity of the arsenic derivatives is
the most obvious disadvantage of this method, although
protocols have been developed9 for the administration of
antidotes that render the method broadly applicable. Methods
that do not rely on such toxic metals would offer a comple-
mentary alternative.
evidence to unequivocally establish the fluorescence quenching
mechanism. Our devised route allows a variety of dimaleimide
moieties to be attached via a variable linker moiety to a number
of different flurophores. This convergent synthetic route permit-
ted the facile preparation of a series of derivatives of dimale-
imide dansyl fluorogens. These derivatives enabled us to
unequivocally elucidate the mechanism of maleimide quenching,
including the intermolecular quenching of dansylamide and
N-ethyldansylamide (1) by N-methylmaleimide (see Chart 1).
The application of this information in the design of successive
generations of fluorogenic protein labeling agents is also
discussed herein.
Maleimides are known to quench fluorescence in their
conjugated form.10 The conjugated CdC double bond of a
maleimide group is also known to undergo addition reactions
with significant selectivity for thiols. After this addition reaction,
the resulting thioalkyl succinimide groups no longer exhibit a
strong quench effect. Naphthopyranone,11,12 phthalimide,13
pyrene,14 and aminophenoxazone15 derivatives bearing a ma-
leimide group have thus been shown to undergo a dramatic
increase in fluorescence upon their reaction with thiols. We have
recently synthesized and characterized a first generation of
fluorogens that contain two maleimide groups, reasoning that
their latent fluorescence would only be realized upon its reaction
with 2 equiv of thiol.16-18 Furthermore, these maleimide groups
were separated by a precise distance on the fluorescent core,
allowing the fluorogens to react rapidly and specifically with
compounds presenting two sulfhydryl groups separated by the
appropriate distance. This was demonstrated for a dicysteine
mutant of the helical protein Fos, which bodes well for the
application of dimaleimide fluorogens to the labeling of proteins
bearing a short helical target sequence.
Experimental Section
Generalities. All starting materials were purchased from Sigma
Aldrich and used without further purification. Solvents were dried using
GlassContour System (Irvine, CA) columns. Reactions requiring
anhydrous conditions were carried out under a dry nitrogen atmosphere
employing conventional benchtop techniques. 13C and 1H NMR spectra
were recorded on AMXR400 and AMX300 spectrometers and were
referenced to the residual proton or 13C signal of the solvent. Mass
spectra were determined by FAB+ ionization on an AutoSpec Q
spectrometer at the Regional Mass Spectrometry Centre at the Universite´
de Montre´al. Melting points (uncorrected) were determined on a
Unimelt Tomas-Hoover melting point apparatus.
Synthesis.N-Ethyl-[1-(dimethylamino)-5-naphthalene]sulfonamide
(1): A solution of dansyl chloride (300 mg; 1.12 mmol; 1 equiv) in
dry CH2Cl2 (20 mL) was treated with ethylamine in THF (0.72 mL;
1.45 mmol; 1.3 equiv) followed by Et3N (0.31 mL; 2.22 mmol; 2 equiv)
under a nitrogen atmosphere. After the mixture was stirred for 2.5 h at
room temperature, the solvent was removed in Vacuo and the crude
product was purified by flash chromatography on silica gel (100%
CHCl3), leading to the desired product 1 (120 mg; 0.43 mmol; 39%)
as a green solid. Mp: 134-136 °C (lit.:21 126 °C). 1H NMR (CDCl3):
8.59 (d, J ) 8.5 Hz, 1H), 8.33 (d, J ) 8.7 Hz, 1H), 8.29 (dd, J ) 7.3
Hz, J ) 1.2 Hz, 1H), 7.64-7.54 (m, 2H), 7.23 (d, J ) 7.4 Hz, 1H),
4.70 (t, J ) 6.0 Hz, 1H, NH), 3.04-2.99 (m, 2H), 2.94 (s, 6H), 1.08
(t, J ) 7.2 Hz, 3H). 13C NMR (CDCl3): 152.8, 135.5, 131.3, 130.7,
130.6, 130.5, 129.3, 124.1, 119.6, 116.0, 46.3, 39.2, 16.0. MS: [M +
H]+: 279.12.
To date, the maleimide fluorescence quenching mechanism
has only been speculated. In fact, relatively few studies have
examined the mechanism of fluorescence quenching by male-
imides, and their conclusions differ greatly, ranging from
photoinduced electron transfer (PET)14 to rearrangement of
molecular orbitals.19,20 Since these conclusions were indirectly
derived without direct evidence for either mechanism, the
subsequent refinement of a method for labeling proteins using
dimaleimide fluorogens requires an understanding of the pho-
tophysical behavior of the maleimide group. Comprehension
of the mechanism of the quench process, before and after thiol
addition, is critical for the rational design of successive
generations of fluorogens that will, ideally, fluoresce very
weakly until their intense fluorescence is restored upon reaction
with a dicysteine target sequence.
A new modular synthetic route was therefore developed to
prepare a series of fluorophore-dimaleimides providing direct
3,5-Dimaleimidobenzoic Acid (2): A solution of 3,5-diaminobenzoic
acid (200 mg; 1.31 mmol; 1 equiv) and maleic anhydride (385 mg;
3.93 mmol; 3 equiv) in CHCl3 (12 mL) was heated to reflux for 20 h.
The resulting yellow precipitate was filtered, and acetic anhydride (10
mL) and sodium acetate (35.1 mg; 0.52 mmol; 0.4 equiv) were added.
The mixture was heated for another 1.5 h at 100 °C. The resulting
clear solution was poured on ice, and cold water (30 mL) was added.
A precipitate was formed after 10 min, and the mixture was vigorously
stirred for an additional hour. The precipitate was filtered and washed
with cold distilled water (100 mL) to give acid 2 (297 mg; 0.95 mmol)
(9) Griffin, B. A.; Adams, S. R.; Jones, J.; Tsien, R. Y. Methods Enzymol.
2000, 327, 565.
(10) Kokotos, G.; Tzougrakic, C. J. Heterocycl. Chem. 1986, 23, 87.
(11) Yang, J.-R.; Langmuir, M. E. J. Heterocycl. Chem. 1991, 28, 1177.
(12) Langmuir, M. E.; Yang, J.-R.; Moussa, A. M.; Laura, R.; LeCompte, K.
A. Tetrahedron Lett. 1995, 36, 3989.
(13) Corrie, J.E.T. J. Chem. Soc., Perkin Trans. I 1994, 2975.
(14) de Silva, A. P.; Gunaratne, N.; Gunnlaugsson, T. Tetrahedron 1991, 39,
5077.
(15) Cohen, B. E.; Pralle, A.; Yao, X.; Swaminath, G.; Gandhi, C. S.; Jan, Y.
N.; Kobilka, B. K.; Isacoff, E. Y.; Jan, L. Y. Proc. Natl. Acad. Sci. U.S.A.
2005, 102, 965.
1
as a beige solid in 72% yield. Mp: >180 °C. H NMR (DMSO-d6):
(16) Girouard, S. M.Sc. thesis, Universite´ de Montre´al, 2000.
(17) Houle, M.-H. M.Sc. thesis, Universite´ de Montre´al, 2002.
(18) Girouard, S.; Houle, M.-H.; Grandbois, A.; Keillor, J. W.; Michnick, S.
W. J. Am. Chem. Soc. 2005, 127, 559.
(20) Kanaoka, Y.; Machida, M.; Ando, K.; Sekine, T. Biochim. Biophys. Acta
1970, 207, 269.
(21) Kallmayer, H.-J.; Schwarz, P. Pharmazie 1989, 44, 119.
(19) Wu, C.-W.; Yarbrough, L. R.; Wu, F. Y.-H. Biochemistry 1976, 15, 2863.
9
11970 J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007

Dimaleimide Dansyl Fluorogens
A R T I C L E S
7.96 (d, J ) 2 Hz, 2H), 7.64 (t, J ) 2 Hz, 1H), 7.23 (s, 4H). 13C NMR
(DMSO-d6): 169.7, 166.0, 134.9, 132.4, 132.2, 128.6, 126.4. MS: [M
+ H]+: 313.05.
Vacuo, and the crude product was triturated with CH2Cl2 and filtered
to yield 201.1 mg (0.34 mmol; 78%) of the desired product 8 as a
white powder. Mp: 161-163 °C. 1H NMR (DMSO-d6): 8.58 (bt, J )
5.5 Hz, 1H, NH), 8.43 (d, J ) 8.6 Hz, 2H), 8.28 (d, J ) 8.7 Hz, 1H),
8.09 (dd, J ) 7.3 Hz, J ) 1.2 Hz, 1H), 7.94-7.90 (m, 1H), 7.79 (d, J
) 1.9 Hz, 2H), 7.54-7.45 (m, 3H), 7.15 (d, J ) 7.7 Hz, 1H), 7.07 (s,
4H), 3.34-3.30 (m, 2H), 2.96-2.93 (m, 2H), 2.83 (s, 6H). 13C NMR
(DMSO-d6): 170.9, 165.8, 152.6, 137.0, 136.7, 136.1, 133.3, 130.8,
130.3, 130.2, 129.6, 129.2, 128.9, 126.2, 124.9, 120.2, 116.4, 47.0,
46.3, 42.9. MS: [M + H]+: 588.15.
3,5-Dimaleimidobenzoyl Chloride (3): A stirred suspension of acid
1 (402 mg; 1.29 mmol; 1 equiv) in thionyl chloride (10 mL) was heated
to reflux until a clear solution had formed, after which the excess SOCl2
was evaporated in Vacuo and the resulting solid was dried under vacuum
to obtain 400 mg (1.21 mmol; 94%) of 3 as a beige solid which was
1
used without further purification. Mp: >180 °C. H NMR (DMSO-
d6): 7.96 (d, J ) 2 Hz, 2H), 7.64 (t, J ) 2 Hz, 1H), 7.22 (s, 4H). 13
C
NMR (DMSO-d6): 169.7, 166.0, 143.8, 134.9, 132.4, 132.1, 126.4.
N-[8-tert-Butoxycarbonylamino-3,6-dioxaoctyl]-1-dimethylamino-
5-naphthalene-sulfonamide (9): To a solution of dansyl chloride (405
mg; 1.50 mmol; 1 equiv) in dry CH2Cl2 (15 mL) was added a solution
of amine 5 (447 mg; 1.80 mmol; 1.2 equiv) and Et3N (0.42 mL; 3.00
mmol; 2 equiv) in CH3CN (10 mL) under a nitrogen atmosphere. After
stirring at room temperature for 3 h the organic layer was washed with
a saturated solution of NaHCO3 (3 × 40 mL), dried over MgSO4, and
concentrated in Vacuo to yield quantitatively compound 9 as a green
MS: [M + Na]+: 352.99.
2-tert-Butoxycarbonylaminoethylamine (4): A solution of di-tert-
butyldicarbonate (3.26 g; 15 mmol; 1 equiv) in CH2Cl2 (30 mL) was
added dropwise to a solution of ethylenediamine (10 mL; 150 mmol;
10 equiv) in CH2Cl2. The resulting reaction mixture was stirred at room
temperature for 20 h. The CH2Cl2 was then removed in Vacuo. The
residue was taken up in EtOAc (30 mL), washed with a saturated
solution of Na2CO3 (2 × 20 mL), dried over MgSO4, and concentrated
in Vacuo to furnish the desired product 3 (2.1 g; 13.4 mmol) as a white
1
foam. H NMR (CDCl3): 8.59 (d, J ) 8.5 Hz, 1H), 8.36 (d, J ) 5.9
Hz, 1H), 8.27 (dd, J ) 7.3 Hz, J ) 1.2 Hz, 1H), 7.60 (m, 2H), 7.24 (d,
J ) 7.5 Hz, 1H), 5.55 (bs, 1H, NH), 5.09 (bs, 1H, NH), 3.41 (m, 10H),
3.16 (m, 2H), 1.48 (s, 9H). 13C NMR (CDCl3): 159.6, 135.9, 131.2,
130.7, 130.5, 130.2, 129.2, 124.1, 119.8, 116.1, 80.2, 78.2, 71.1, 71.1,
70.9, 70.1, 46.3; 43.9, 41.1, 29.3. MS: [M + H]+: 482.23
1
foam with 89%. H NMR (CDCl3): 5.09 (bs, 1H, NH2), 3.13-3.08
(m, 3H), 2.74 (t, J ) 5.8 Hz, J2 ) 5.8 Hz, 2H), 1.38 (s, 9H). 13C NMR
(CDCl3): 156.3, 79.2, 43.3, 41.8, 28.4. MS: [M + H]+: 161.13.
N-8-tert-Butoxycarbonylamino-3,6-dioxaoctylamine (5): A solu-
tion of di-tert-butyldicarbonate (1.1 g; 5.00 mmol; 1equiv) in CH2Cl2
(5 mL) was added dropwise to a solution of 2,2′-(ethylenedioxy)bis-
(ethylamine) (7.32 mL; 50.0 mmol; 10 equiv) in CH2Cl2 (15 mL). The
resulting reaction mixture was stirred at room temperature for 21 h.
The CH2Cl2 was then removed in Vacuo. The white foam was taken
up in EtOAc (125 mL), washed with a saturated solution of Na2CO3
(3 × 50 mL), dried over MgSO4, and concentrated in Vacuo to furnish
the desired monoprotected amine 5 (1.0 g; 4.05 mmol) as a colorless
N-[8-(1-(Dimethylamino)-5-naphthalenesulfonamide)-3,6-diox-
aoctyl]-3,5-dimaleimidyl-1-benzamide (10): Under a nitrogen atmo-
sphere at room temperature were combined amide 9 (429 mg; 0.89
mmol; 1 equiv) and trifluoroacetic acid (10 mL). The reaction was
stirred for 3 h and concentrated in Vacuo to afford the corresponding
amine quantitatively as its trifluoroacetate salt, used immediately in
the following step. To a solution of acyl chloride 3 (323 mg; 0.98 mmol;
1.1 equiv) in dry CH2Cl2 (20 mL) was added the amine (543 mg; 0.89
mmol; 1 equiv) in CH2Cl2 followed by Et3N (0.62 mL; 4.46 mmol; 5
equiv) under a nitrogen atmosphere and stirred at room temperature
overnight. The organic layer was washed with a saturated solution of
NaHCO3 (2 × 30 mL) dried over MgSO4 and concentrated in Vacuo.
The crude product was purified by precipitation in CH2Cl2-Et2O. The
resulting solid was filtered and washed with Et2O to give the compound
10 as a beige solid in 44% yield. Mp: 118-120 °C. 1H NMR
(CDCl3): 8.54 (d, J ) 8.6 Hz, 1H), 8.30 (d, J ) 8.6 Hz, 1H), 8.22 (d,
J ) 7.3 Hz, 1H), 7.94 (d, J ) 1.8 Hz, 2H), 7.67-7.46 (m, 5H), 7.30
(bs, 1H, NH), 7.18 (d, J ) 7.5 Hz, 1H), 6.84 (s, 4H), 5.99 (t, J ) 5.8
Hz, 1H, NH), 3.63-3.41 (m, 10H), 3.14-3.10 (m, 2H), 2.89 (s, 6H).
13C NMR (CDCl3): 168.8, 165.5, 151.7, 136.2, 135.2, 134.2, 132.1,
130.2, 129.7, 129.5, 129.0, 128.1, 126.6, 125.3, 123.7, 123.1, 118.9,
115.1, 70.1, 69.7, 69.2, 45.3, 42.8, 39.9. MS: [M + H]+: 676.21.
N-[2-(1-Dimethylamino-5-naphtalene-sulfonyl)aminoethyl]-3-[3-
(2-carboxy)ethylthio succinimidyl]-5-maleimidyl-1-benzamide (11):
A solution of mercaptopropionic acid (6.10 µL; 0.07 mmol; 1 equiv)
in CH3CN (3 mL) was added dropwise to 8 (410.3 mg; 0.70 mmol; 10
equiv) dissolved in a CH3CN-DMSO (1:1, v/v, 20 mL) solution. The
resulting reaction mixture was stirred at room temperature overnight.
The CH3CN was removed in Vacuo, and Et2O (20 mL) was added to
the mixture to precipitate 11 and the excess of 8. The DMSO was
removed by filtration, and the resulting residue was triturated with
MeOH (20 mL). After filtration the MeOH was concentrated in Vacuo
to afford the desired compound 11 (40 mg; 0.07 mmol; 82%) as a
1
oil in 81% yield. H NMR (CDCl3): 5.26 (bs, 1H, NH), 3.60-3.58
(m, 4H), 3.52-3.48 (m, 4H), 3.37-3.28 (m, 2H), 2.87 (bs, 2H, NH2),
1.97-1.94 (m, 2H), 1.42 (s, 9H). 13C NMR (CDCl3): 156.9, 79.9, 74.1,
71.1, 71.0, 42.5, 41.1, 29.2. MS: [M + H]+: 249.18.
N-[2-tert-Butoxycarbonylaminoethyl]-3,5-dimaleimidobenza-
mide (6): To a solution of compound 3 (423 mg; 1.28 mmol; 1 equiv)
in dry CH2Cl2 (30 mL) were added amine 4 (272 mg; 1.70 mmol; 1.3
equiv) in CH2Cl2 (10 mL) and 0.37 mL of triethylamine (2.62 mmol;
2 equiv) under a nitrogen atmosphere. After the mixture was stirred at
room temperature for 2.5 h, the organic layer was washed with a
saturated solution of NaHCO3 (4 × 30 mL), dried over MgSO4, and
concentrated in Vacuo to give the desired compound 6 (557.3 mg; 1.23
mmol) as a brown solid in 94% yield. Mp: >180 °C. 1H NMR
(CDCl3): 7.83 (d, J ) 1.5 Hz, 2H), 7.58 (t, J ) 1.8 Hz, 1H), 7.44 (bs,
1H), 6.88 (s, 4H), 5.13 (bt, J ) 5.6 Hz, 1H, NH), 3.58-3.53 (m, 2H),
3.40-3.38 (m, 2H), 1.39 (s, 9H). 13C NMR (CDCl3): 168.9, 166.0,
157.4, 136.1, 134.4, 132.2, 125.4, 123.5, 79.8, 41.9, 40.1, 28.4. MS:
[M + H]+: 455.16.
N-(2-Aminoethyl)-3,5-dimaleimidobenzamide (7): Under a nitro-
gen atmosphere at room temperature was combined amide 6 (557.3
mg; 1.23 mmol; 1 equiv) and trifluoroacetic acid (20 mL). The reaction
mixture was stirred for 1.5 h, concentrated in Vacuo, triturated with
Et2O (20 mL), filtered, and dried to afford 7 quantitatively as its
1
trifluoroacetate salt (437 mg; 1.23 mmol). H NMR (CD3OD): 7.77
(d, J ) 1.8 Hz, 2H), 7.55 (t, J ) 1.8 Hz, 1H), 6.98 (s, 4H), 3.64 (t, J
) 5.8 Hz, 2H), 3.22 (bs, 2H, NH2), 3.17 (t, J ) 5.8 Hz, 2H). 13C NMR
(CD3OD): 171.7, 165.9, 136.2, 135.7, 133.2, 128.5, 125.5, 40.2, 38.5.
MS: [M + H]+: 355.10.
N-[2-(1-(Dimethylamino)-5-naphthalene-sulfonyl)aminoethyl]-
3,5-dimaleimidobenzamide (8): To a solution of amine 7 (205.9 mg;
0.44 mmol; 1 equiv) in dry CH3CN (30 mL) dansyl chloride (155 mg;
0.57 mmol; 1.3 equiv) in dry CH3CN (10 mL) was added followed by
Et3N (0.13 mL; 0.88 mmol; 2 equiv) under a nitrogen atmosphere and
stirred at room temperature overnight. The solvent was removed in
1
yellow solid. Mp: 155-157 °C. H NMR (CDCl3): 8.32 (d, J ) 8.3
Hz, 1H), 8.21-8.19 (m, 1H), 8.06-7.98 (m, 2H), 7.76 (s, 2H), 7.34-
7.30 (m, 4H), 6.92 (d, J ) 7.5 Hz; 1H), 6.72 (s, 2H), 3.31 (bs, 2H),
2.96-2.93 (m, 6H), 2.71-2.63 (m, 2H), 2.67 (s, 6H). 13C NMR
(CDCl3): 168.9, 165.5, 135.7, 135.5, 134.8, 134.3, 131.8, 130.0, 129.5,
129.3, 128.9, 128.0, 125.6, 124.1, 123.0, 119.0, 115.0, 77.4, 45.8, 45.2,
42.6, 39.9, 34.3, 26.7. MS: [M + H]+: 694.16.
N-[2-(1-(Dimethylamino)-5-naphthalene-sulfonyl)aminoethyl]-
3,5-di[3-(2-carboxy)ethyl thiosuccinimidyl]-1-benzamide (12): Com-
9
J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007 11971

A R T I C L E S
Guy et al.
pound 8 (125 mg; 0.21 mmol; 1 equiv) was dissolved in a DMSO-
DMF solution (5:1, v/v, 50 mL). Mercaptopropionic acid (46.2 µL;
0.53 mmol; 2.5 equiv) was added, and the mixture was stirred at room
temperature for 72 h. Distilled water (50 mL) and CHCl3 (75 mL) were
added, and the organic layer was washed with H2O (2 × 40 mL), dried
over MgSO4, filtered, and concentrated in Vacuo to afford 12 (125 mg;
0.213 mmol; 63%) as a yellow solid. Mp: 87-89 °C. 1H NMR
(acetone-d6): 8.52 (d, J ) 8.5 Hz, 1H), 8.33 (d, J ) 8.6 Hz, 1H), 8.25
(m, 1H, NH), 8.20 (d, J ) 7.3 Hz, 1H), 7.82 (d, J ) 1.7 Hz, 2H),
7.62-7.48 (m, 3H), 7.20 (d, J ) 7.5 Hz, 1H), 4.23 (dd, J ) 9.2 Hz, J
) 4.1 Hz, 2H), 3.47-3.42 (m, 4H), 3.17-2.96 (m, 8H), 2.83 (s, 6H),
2.76-2.70 (m, 4H). 13C NMR (acetone-d6): 176.6, 174.6, 173.5, 152.7,
136.6, 134.0, 130.7, 130.6, 130.4, 129.7, 128.9, 128.7, 126.4, 124.1,
120.0, 116.0, 45.5, 43.2, 40.8, 40.6, 36.9, 30.1. MS: [M + H]+: 800.17.
N-[8-(1-(dimethylamino)-5-naphthalenesulfonamide)-3,6-dioxaoc-
tan-1-yl]-3,5-di[3-(2-carboxy)ethyl)thiosuccinimidyl]-1-benzamide
(13): To a solution of 10 (100 mg; 0.15 mmol; 1 equiv) in CHCl3 (10
mL) was added the mercatopropionic acid (32 µL; 0.37 mmol; 2.5
equiv). After the mixture was stirred at room temperature for 23 h,
Et2O (10 mL) was added to the mixture. The resulting white precipitate
was filtered and washed with Et2O (2 × 5 mL) to afford 13 as a white
powder (105 mg; 0.12 mmol; 79%). Mp: 122-124 °C. 1H NMR
(DMSO-d6): 8.57 (m, 1H, NH), 8.33 (d, J ) 8.2 Hz, 1H), 8.18 (d, J
) 8.7 Hz, 1H), 8.01 (d, J ) 6.8 Hz, 1H), 7.77 (d, J ) 1.5 Hz, 2H),
7.51-7.39 (m, 2H), 7.33 (s, 1H), 7.10 (d, J ) 7.3 Hz, 1H), 4.02 (dd,
J ) 8.8 Hz, J ) 3.6 Hz, 2H), 3.37-3.16 (m, 18H), 3.01-2.91 (m,
2H), 2.86-2.77 (m, 6H), 2.57-2.50 (m, 6H). 13C NMR (DMSO-d6):
174.3, 172.5, 172.2, 163.9, 150.6, 135.0, 131.6, 128.7, 128.5, 128.4,
127.4, 126.8, 126.5, 124.8, 122.2, 118.3, 114.0, 68.9, 68.8, 68.2, 44.3,
41.5, 35.2, 33.4, 25.9. MS: [M + H]+: 888.22.
ments were done in a 1:4 methanol/ethanol glass matrix at 77 K,
exciting at the compound’s absorption maximum. The fluorescence
lifetimes were obtained via an Edinburgh Instruments time-resolved
FLS-920 fluorometer. The samples for lifetime measurements were
prepared at 10-5 M in anhydrous acetonitile and were thoroughly
deaerated with nitrogen prior to analysis. The lifetime decays (dansy-
lamide ) 14.6 ns; 12 ) 14.0 ns; 13 ) 17.1 ns) were fitted with a
signal variable monoexponential decay leading to g99% correlation
factors. Stern-Volmer fluorescence quenching was done according to
the following standard protocol: Fluorophore 1 was dissolved in
anhydrous and deaerated acetontrile, and its emission spectrum was
recorded by exciting at λ ) 335 nm. Known amounts of a deaerated
stock solution of quencher were added to the fluorophore sample. The
resulting fluorescence decrease was recorded and compared to the
unquenched fluorescence. Extreme care was taken to ensure the absor-
bance of the quencher was below 5% of the fluorophore absorbance at
the excitation wavelength used for the measurements. The Stern-
Volmer quenching constant (KSV) was taken from the linear portion of
the slope of relative fluorescence change as a function of the quenching
concentration. The change in the fluorophore’s fluorescence lifetime
with various maleimide quenchers was studied in a similar manner to
afford the Stern-Volmer constants. The subsequent dynamic quenching
rate constants (kq) for the fluorophore fluorescence deactivation with
the various maleimide quenchers were calculated by dividing the
measured values of KSV by the fluorescence lifetime of 1.
The triplet-triplet transient absorption spectrum was measured in
anhydrous acetonitrile using a Luzchem mini-LFP system excited at
355 nm with the third harmonic from a Continuum YAG:Nd Sure-lite
laser. The triplet quantum yields were measured by optically matching
samples within 5% at 355 nm relative to anthracene. The anthracene
triplet was monitored at 420 nm.24,25
Electrochemical Measurements. Cyclic voltammetric measure-
ments were performed on a Bio Analytical Systems EC Epsilon
potentiostat with a scan rate of 100 mV/s. The compounds were
dissolved in anhydrous and deaerated acetonitrile at 10-5 M along with
the addition of 0.1 M NBu4PF6. The redox potentials were measured
after deaerating the solution with nitrogen for 20 min. A glassy carbon
electrode and a platinum electrode were employed as working and
auxiliary electrodes, respectively. A saturated Ag/AgCl electrode was
used as the reference electrode.26
N-[2-tert-Butoxycarbonylaminoethyl]-3-[3-(2-carboxy)ethylthio-
succinimidyl]-5-maleimidyl-1-benzamide (14): Mercaptopropionic
acid (58.0 µL; 0.66 mmol; 1 equiv) in CHCl3 (10 mL) was added to a
solution of compound 6 (300 mg; 0.66 mmol; 1 equiv) in CHCl3 (100
mL). The mixture was stirred at room temperature for 4 h, and the
solvent was removed in Vacuo. The crude product was purified on a
silica preparative TLC plate (2 mm thickness; (EtOAc-Hex-AcOH,
7:3:0.5, v/v)), leading to 14 as a pink solid (16 mg; 0.03 mmol; 4.3%).
1
Mp: 67-69 °C. H NMR (acetone-d6): 8.00 (s, 1H), 7.81 (s, 1H),
7.63 (s, 1H), 7.03 (s, 2H), 4.14 (dd, J ) 9.2 Hz, J ) 3.9 Hz, 1H),
3.52-3.46 (m, 4H), 3.33-3.30 (m, 2H), 2.77-2.73 (m, 4H), 1.31 (s,
9H). 13C NMR (acetone-d6): 170.1, 135.5, 131.8, 128.0, 125.8, 125.6,
78.9, 40.8, 40.7, 37.0, 28.6, 27.5. MS: [M + Na]+: 583.15.
Results and Discussion
Synthesis. The convergent synthetic routes shown in Scheme
1 allowed the efficient preparation of all compounds by
straightforward procedures, involving typical protection/coupling/
deprotection strategies. The maleimide groups of compound 2
were introduced according to the typical two-step transformation
procedure10,18,27 in excellent yield. The activation of benzoic
acid derivative 2, prior to amide coupling, was effected
efficiently through the preparation28 of acid chloride 3, as a
stable synthetic intermediate in quantitative yield.
Dimaleimide fluorogens 8 and 10 were prepared by slightly
different routes, as shown in Scheme 1. First of all, commercially
available diamines were monoprotected as Boc derivatives 4
and 5 in excellent yield by adding a 10-fold excess of diamine
to Boc anhydride. The condensation of 3 with amine 4 according
to a literature procedure29 gave amide 6 in excellent yield, which
N-[2-tert-Butoxycarbonylaminoeth-1-yl]-3,5-di[3-(2-carboxy)eth-
ylthiosuccinimidyl]-1-benzamide (15): Mercaptopropionic acid (0.15
mL; 1.75 mmol; 2.5 equiv) was added to a solution of compound 6
(317 mg; 0.70 mmol; 1 equiv) in CHCl3 (25 mL). The mixture was
stirred at room temperature for 20 h, and the solvent was removed in
Vacuo. The resulting black oil was triturated with Et2O and filtered to
1
afford quantitatively 15 as a beige powder. Mp: >180 °C. H NMR
(acetone-d6): 8.18 (bs, 1H, NH), 7.91 (d, J ) 1.8 Hz, 2H), 7.47 (bs,
1H), 6.29 (bs, 1H, NH), 4.25 (dd, J ) 9.2 Hz, J ) 4.1 Hz, 2H), 3.53-
3.47 (m, 4H), 3.34-3.20 (m, 7H), 3.10-3.06 (m, 2H), 2.78-2.71 (m,
5H), 1.40 (s, 9H). 13C NMR (acetone-d6): 176.4, 174.3, 173.1, 165.7,
157.1, 137.1, 134.0, 128.6, 126.3, 78.9, 41.2, 40.8, 40.7, 37.0, 34.6,
28.6, 27.5. MS: [M + Na]+: 689.16.
Spectroscopic Measurements. Absorption measurements were done
on a Cary-100 spectrometer and fluorescence studies were carried out
on a Cary Eclipse and an Edinburgh Instruments FLS-920 fluorometer
after deaerating the samples thoroughly with nitrogen for 20 min.
Fluorescence quantum yields were measured at 10-5 M by exciting
the corresponding compounds that were optically matched at 335 nm
in spectroscopic grade acetonitrile and compared to dansylamide (φ511
nm ) 0.3722) excited at the same wavelength.23 The actinometer
absorbencies, and those of the compounds, were matched at the
excitation wavelength to within 5%. The phosphorescence measure-
(22) Li, Y.-H.; Chan, L.-M.; Tyer, L.; Moody, R. T.; Himel, C. M.; Hercules,
D. M. J. Am. Chem. Soc. 1975, 97, 3118.
(23) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991.
(24) Scaiano, J. C. CRC Handbook of Organic Photochemistry; CRC Press:
Boca Raton, FL, 1989.
(25) Carmichael, I.; Hug, G. L. J. Phys. Chem. Ref. Data 1986, 15, 1.
(26) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877.
(27) Reddy, P. Y.; Kondo, S.; Fujita, S.; Toru, T. Synthesis 1998, 999.
(28) Pickaert, G.; Cesario, M.; Ziessel, R. J. Org. Chem. 2004, 69, 5335.
9
11972 J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007

Dimaleimide Dansyl Fluorogens
A R T I C L E S
Scheme 1 a
a
(a) (i) Maleic anydride, CHCl3, reflux 20 h, (ii) Ac2O, NaOAc, 100 °C, 1.5 h, 72% for two steps; (b) SOCl2, reflux, 3.5 h, 98%; (c) Boc2O, CH2Cl2,
rt, 16 h, 89%; (d) 4, NEt3, CH2Cl2, rt, 3 h, 96%; (e) TFA, rt, 1.5 h, 100%; (f) dansyl chloride, NEt3, CH3CN, rt, 24 h, 78%; (g) 5, dansyl chloride, NEt3,
CH2Cl2, rt, 3 h, 100%; (h) (i) TFA, rt, 1.5 h, (ii) 3, NEt3, CH2Cl2, 24 h, 44% for two steps; (i) 3-mercaptopropionic acid, CH3CN/DMSO, rt, 24 h, 82%; (j)
12: mercaptopropionic acid, DMSO/DMF, rt, 72 h, 63%; 13: mercaptopropionic acid, CHCl3, rt, 24 h, 79%; (k) mercaptopropionic acid, CHCl3, rt, 4 h,
4.3%; (l) mercaptopropionic acid, CHCl3, rt, 20 h, 100%.
was subsequently deprotected30 to give primary amine 7,
quantitatively. This amine was then coupled with commercially
available dansyl chloride to give the desired fluorogen 8 in
excellent yield. For fluorogen 10, the relative order of addition
of the dimaleimide benzoyl moiety and dansyl chloride to the
diamine was inverted with respect to the synthesis of fluorogen
8. This was judged to be necessary to avoid potential degradation
of the maleimide groups in the presence of the very nucleophilic
PEG-diamine 5. For this reason, PEG-diamine 5 was first
coupled with dansyl chloride to give amide 9, quantitatively.
Removal of the Boc group30 from 9 followed by immediate
reaction of the resulting primary amine with acid chloride 3
afforded fluorogen 10 in moderate yield for the two steps.
The design of the synthetic route shown in Scheme 1 allows
dimaleimide fluorogens to be prepared simply, rapidly, and in
parallel. Furthermore, the alternate use of different activated
fluorophores, diaminobenzoic acid derivatives, or diamine
linkers permit, with relative ease, the facile variation of the color
of fluorophore, the distance between the maleimide groups, or
the nature of the spacer of the final fluorogen. These advantages
will be exploited in forthcoming work.
(29) Hidai, Y.; Kan, T.; Fukuyama, T. Chem. Pharm. Bull. 2000, 48, 1570.
(30) Jasys, V. J.; Kelbaugh, P. R.; Nason, D. M.; Phillips, D.; Rosnack, K. J.;
Saccomano, N. A.; Stroh, J. G.; Volkmann, R. A. J. Am. Chem. Soc. 1990,
112, 6696.
Fluorescence. Shown in Table 1 are the quantum yields
measured for reference compounds dansylamide and 1, for
9
J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007 11973

A R T I C L E S
Scheme 2
Guy et al.
Table 1. Molar Absorption Coefficients and Fluorescence
Quantum Yields Measured in Acetonitrile at λexc ) 335 nm
Table 2. Relative Fluorescence in a 1:4 Methanol/Ethanol Glass
Matrix at 77 K Compared to RT Solution, λexc ) 335 nm
a
ꢀmax
λem
(nm)
λ
em(77 K)
λ
em(298 K)
1
b
b
compound
(M-1 cm-
)
φfl
compound
(nm)
(nm)
φfl(rt)
φfl(77 K)
φIC
dansylamide
3200
3200
3700
3800
2700
2700
3000
511
511
511
525
458
511
525
0.37 c
0.35
0.04
0.03
0.04
0.22
0.11
1
6
8
12
462
490
462
462
511
490
511
511
0.35
-
0.04
0.22
1.0
-
0.20
0.77
0.65
-
0.16
0.53
a
1
8
10
11
12
13
a The emission measured at 77 K is 6.6-fold greater than that at rt;
however no value for the φfl(rt) is known. b The values of φIC were calculated
according to: φfl + φISC + φIC ≈ 1, where φISC ) 0 determined by LFP
and φfl (77K) - φfl (rt) ≈ φIC.
a Measured at 335 nm. b Quantum yields were calculated according to
the method shown in ref 23. c Value taken from ref 22.
Quenching Mechanism. It is evident from Table 1 that the
maleimide group efficiently quenches the dansylamide fluoro-
phore excited state. Even though quenching of the fluorophore
occurs in the presence of the maleimide, the mechanism by
which this occurs is not clearly understood. Knowledge of the
deactivation mechanism is therefore quintessential for the future
design of new generations of fluorescent protein labels of this
type.18 There are three plausible types of quenching modes
available for dansylamide deactivation in the presence of the
maleimide, these are: (1) Fo¨rster energy transfer, (2) Dexter
energy transfer, and (3) photoinduced electron transfer (PET).
To determine the type of fluorescence quenching, we first
examined the fluorescence deactivation of dansylamide. This
is of importance owing to the limited number of photophysical
studies of this fluorophore and the characterization of its excited
states.31-33
dimaleimide dansyl derivatives 8 and 10, and for their respective
dithiol adducts 12 and 13. Comparison of the quantum yields
of 8 and 10 to those of dansylamide and 1 shows the effect of
an intramolecular dimaleimide moiety. Fluorescence efficiency
is decreased by an order of magnitude, similar to what we
observed previously, for naphthalene and coumarin derivatives
bearing two maleimide groups attached directly to their fluo-
rescent cores.18 This quenching effect is of the same magnitude
as that measured for various fluorophores attached to a single
maleimide group, either directly or via an alkyl tether.11-14
Compound 1 was prepared to verify that the quench effect
observed in 8 and 10 was not due to conformational constraints
introduced by alkyl substitution on the dansylamide nitrogen.
The similarity of its quantum yield to that of the standard
dansylamide clearly shows that it is not simply the alkylation
of the dansylamide moiety of 8 and 10 that leads to their
quenched fluorescence, but rather the presence of at least one
maleimide group. Comparison of the quantum yields of 8 and
10 to those of their dithiol adducts 12 and 13 confirm that
fluorescence is restored to nearly the same efficiency as in the
standard compounds.
Given the complexity in assigning the quenching mechanism
without knowing the deactivation modes of the native fluoro-
phores, we first examined the fate of the excited state of
dansylamide and 1. There are generally three distinct pathways
for the singlet deactivation for such fluorophores, namely, (1)
fluorescence, (2) intersystem crossing (ISC), and (3) internal
conversion (IC). The reported34,35 low fluorescence quantum
yield (φfl ) 0.35) for dansylamide suggests that its deactivation
occurs predominately by nonradiative means such as ISC or
IC. Since the fluorophores studied herein are highly conjugated,
ISC to the triplet state is highly probable owing to the narrowing
of the singlet-triplet gap, which occurs with conjugated
compounds. The triplet manifold is additionally expected to be
populated by ISC occurring by slight spin orbit coupling induced
Previously, we suggested that the reaction of only one
maleimide group of a dimaleimide fluorogen is not sufficient
to restore its latent fluorescence.16-18 That is, only the dithiolated
adduct fluoresces strongly, after both maleimide groups have
undergone a thiol addition reaction. To verify this, authentic
monothiolated adduct 11 was synthesized and purified. Com-
parison of its quantum yields with that of 8 and 12 confirms
that the monothiolated derivative fluoresces as weakly as the
quenched dimaleimide derivative. All restoration of fluorescence
is realized upon reaction of the final maleimide group (Scheme
2). The successful design of future dimaleimide labeling agents
will depend upon our understanding of the reactivity of the
maleimide group (to be reported shortly) and the nature of its
quench mechanism (Vide infra).
(31) Metivier, R.; Leray, I.; Valeur, B. Photochem. Photobiol. Sci. 2004, 3, 374.
(32) Hebbink, G. A.; Klink, S. I.; Grave, L.; Alink, P. G. B. O.; Veggel, F. C.
J. M. v. Chem. Phys. Chem. 2002, 3, 1014.
(33) Holmes-Farley, S. R.; Whitesides, G. M. Langmuir 1986, 2, 266.
(34) Himel, C. M.; Mayer, R. T. Anal. Chem. 1970, 42, 130.
(35) Chen, R. F. Nature 1966, 209, 69.
9
11974 J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007

Dimaleimide Dansyl Fluorogens
A R T I C L E S
Figure 1. Relative phosphorescence spectra of 1 (9), 6 (O), 8 (0),
dansylamide (b) recorded at 77 K in a 1:4 methanol/ethanol glass matrix.
Inset: normalized fluorescence (open squares) measured at room temper-
ature, and phosphorescence (closed squares) of 1 measured at 77 K in a
1:4 methanol/ethanol glass matrix.
Figure 2. Transient absorption spectra recorded in acetonitrile, at 2.5 µs
after the excitation pulse at 355 nm, of anthracence (9) and 1 (0). Inset:
transient decay of anthracene monitored at 425 nm.
by the heavy atom effect of the heteroatoms.36 Evidence for
the triplet state can be obtained from phosphorescence measure-
ments at 77 K in a methanol/ethanol glass. Such measurements
produced a new long wavelength emission around 720 nm that
is consistent with the existence of a triplet state (Figure 1 and
Table 2). There was no emission shift observed among 1 and 6
and 8. This confirms the compounds have the same triplet energy
and that there is no conjugation between the maleimides and
the fluorophore. This notwithstanding, inconclusive evidence
for the fluorescence quenching of dansylamide by ISC is
obtained from the phosphorescence measurements because this
low temperature artificially populates the triplet state such that
its quantification is not possible. Alternatively, evidence for the
manifold shift to the triplet state by ISC is possible via laser
flash photolysis (LFP). The high degree of conjugation of
dansylamide should provide an intense signal visible around
400 nm by LFP. This expected signal would be similar to
anthracence and decay with similar first-order kinetics as those
shown for anthracene in Figure 2. However, direct excitation
of the fluorophore at 355 nm failed to produce any visible triplet
within the time frame of the experiment. This provides undeni-
able proof that dansylamide and 1 do not undergo singlet
deactivation by ISC. Furthermore, since Dexter energy transfer
occurs mostly by the triplet manifold, this cannot be the
mechanism responsible for efficient fluorophore quenching with
maleimides.
Having ruled out ISC, the predominant method of dansyla-
mide excited-state deactivation must therefore occur by the
nonradiative route of IC according to the following energy
conservation equation: φfl + φISC + φIC ≈ 1. Evidence that the
remaining 65% of the excited-state energy is dissipated by IC
is derived from low-temperature fluorescence measurements.
All normal modes of energy dissipation by bond rotation are
suppressed at 77 K, resulting in a fluorescence increase if IC
deactivation is present. Accurate values for the low-temperature
fluorescence yields are difficult because of the experimental
errors associated with calculating the weak fluorescence signal
at room temperature relative to the intense one at 77 K.
Figure 3. Relative fluorescence of 1 measured in a 1:4 methanol/ethanol
matrix at rt (9) and 77 K (0). Inset: Relative fluorescence of 12 measured
in a 1:4 methanol/ethanol matrix at rt (9) and 77 K (0).
Nonetheless, an approximately 4-fold fluorescence increase was
observed at 77 K, relative to the room-temperature emission,
for all compounds studied (Figure 3). This confirms that energy
dissipation occurs predominantly by IC and is consistent with
previous reports of similar dansylamides.37,38 IC in the dansyl
group is understood to occur by efficient deactivation involving
bond rotation of the heteroatomic groups bound to the naph-
thalene ring and is consistent with other substituted fluorophores
such as rhodamine and fluorescein.39 The hypsochromically
shifted emission observed at 77 K relative to room temperature
is from the highest energy rotamer that is exclusively formed
at this low temperature. The amount of IC for 1 can be calculated
from the two temperature fluorescence measurements according
to φfl (77 K) - φfl (rt) ) φIC. The same IC deactivation occurs
for both 1 and 12, which contains the maleimide covalently
linked to the dansylamide fluorophore (inset of Figure 3).
(37) Chen, H.; Jiang, Y. B. Chem. Phys. Lett. 2000, 325, 605.
(38) Takehira, K.; Suzuki, K.; Hiratsuka, H.; Tobita, S. Chem. Phys. Lett. 2005,
413, 52.
(36) Turro, N. J. Modern Molecular Photochemistry; University Science
(39) Perez Guarin, S. A.; Tsang, D.; Skene, W. G. New J. Chem. 2007, 31,
Books: Sausalito, CA, 1991.
210.
9
J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007 11975

A R T I C L E S
Guy et al.
Figure 5. Cyclic voltammetry of 1 (0) and 6 (9), measured in anhydrous
and deaerated dichloromethane, with Bu4NPF6 as the supporting electrolyte
and using a Ag/Ag+ reference electrode and a 100 mV/s scan speed.
Figure 4. Fluorescence quenching of the singlet excited-state of 1 with
varying amounts of N-methyl maleimide quencher, measured in deaerated
acetonitrile: 0 mM (9), 0.24 mM (0), 0.47 mM (O), 1.1 mM (2), 1.43
mM (1). Inset: Stern-Volmer plot showing the relative fluorescence
lifetime change without and with quencher as a function of N-methyl
maleimide concentration.
the dansylamide via electron transfer. The effective quenching
concentration for intramolecular quenching processes is 1 M
while the average singlet excited-state concentration produced
upon irradiation is ca. µM. Therefore, the maleimide quencher
concentration is approximately 6 orders of magnitude greater
than the dansylamide singlet excited-state concentration when
the two groups are tethered together such as in 8. This large
concentration leads to rapid and efficient intramolecular electron-
transfer quenching occurring in less than 100 ps.
The absence of a visible triplet by LPF implies deactivation
of the fluorophore’s singlet excited-state occurs preferentially
from the singlet manifold. The mechanism of maleimide
quenching must therefore occur from this manifold and can
involve either Fo¨rster energy transfer or PET. Given the
incompatibility of the maleimide absorption and the dansylamide
emission spectral overlap, Fo¨rster quenching is not a viable
mechanism. PET is therefore a feasible quenching mechanism.40
To confirm electron-transfer deactivation, the Stern-Volmer
quenching of 1 was investigated with various maleimides. This
method examines either the decrease in steady-state fluorescence
or fluorescence lifetimes with the addition of known amounts
of quencher and provides information regarding quenching
kinetics and ultimately the quenching mechanism. The typical
fluorescence behavior observed for all quenchers studied herein
is presented by the Stern-Volmer plot in Figure 4. The Stern-
Volmer (KSV) quenching rate constants are derived from slopes
of the quenching plots. The desired second-order quenching rate
constant (kq) is calculated from KSV by dividing by the
fluorescence lifetime of 1 (τ ) 22.8 ns), leading to kq. The
calculated values derived from the uniquely monoexponential
lifetimes for N-methyl maleimide (1.2 × 1010 M-1 s-1), 6 (8.5
× 109 M-1 s-1), and 15 (2.0 × 1010 M-1 s-1) are close to the
diffusion limit in acetontrile, which is consistent with a fast
quenching process such as electron transfer. The time-resolved
Stern-Volmer studies (inset Figure 4) further confirm the
quenching process is dynamic. This dynamic quenching mech-
anism is further supported by the absence of any change in the
absorption spectra when 1 and N-methyl maleimide are com-
bined. Such a change would otherwise indicate the formation
of a ground state complex leading to static quenching. The
observed dynamic quenching implies the quencher must diffuse
close to the fluorophore within the fluorescence lifetime of the
dansylamide in order to deactivate the singlet excited state.
Given that the quencher and the fluorophore are inherently
confined in close proximity, the quencher rapidly deactivates
Further evidence for electron-transfer quenching is derived
from the complementarity of the electron donor and acceptor
energy levels. The suitability of the dansylamide fluorophore
donor and the maleimide acceptor to undergo photoinduced
electron transfer is empirically calculated according to the
Rehm-Weller equation: ∆G° ) E(D+/D - A/A-) - ∆E(0,0)
eT
- wp.41 This relationship takes into account the capacity of 1
to donate its electron (Eox), the capacity of the maleimide group
to accept the transferred electron (Ered), and the reorganization
energy (wp). The oxidation and reduction potentials of 1 and
6, respectively, are therefore required to calculate the energetics
and hence the probability of PET. The oxidation potential of 1
(0.94 V) and the reduction potential of 6 (-1.03 V) were
consequently measured by cyclic voltammetry (Figure 5). The
energy gap (∆E(0,0)) between the ground and excited singlet
states of 1 must also be taken into account since PET involves
electron transfer from the excited state. The required value is
calculated from the intercept of the normalized plot of the
absorption and fluorescence spectra, according to Figure 6. The
reorganization energy can be ignored since it is a mere 5.5 kJ/
mol for acetonitrile and is a negligible amount compared to the
other thermodynamic parameters. The calculated values confirm
that the electron transfer is energetically favored (∆GeT ) -112
kJ/mol) and supports the proposed PET quenching mechanism.
The back electron transfer is, however, slightly more favorable
(-187 kJ/mol) and prevents separation of the charged species.
The steric confinement of the maleimide and the dansylamide
further prevents any charged species from diffusing apart.
Nonetheless, the complementarity of the donor and acceptor
(40) Ceroni, P.; Laghi, I.; Maestri, M.; Balzani, V.; Gestermann, S.; Gorka,
(41) Gilbert, A.; Baggott, J. Essentials of Molecular Photochemistry; CRC
M.; Vo¨gtle, F. New J. Chem., 2002, 26, 66.
Press: Boca Raton, FL, 1991.
9
11976 J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007

Dimaleimide Dansyl Fluorogens
A R T I C L E S
compatibility of their redox potentials and excited-state energies
with those of maleimide, as predicted by the Rehm-Weller
equation. For example, 7-methoxy coumarin43 (Eox ) 1.84 V;
∆E(0,0) ) 3.48 eV), F-Bodipy44 (Eox ) 0.95 V, ∆E(0,0) )
2.43 eV) and Rhodamine B45 (Eox ) 0.98 V, ∆E(0,0) ) 2.23
eV) are three fluorophores that span the emission spectrum from
blue to red. Comparison of their standardized potentials against
Ag/Ag+ (saturated) and spectroscopic properties with the
reduction potential of maleimide suggests that while the
quenching process would be highly efficient with the blue
coumarin, it would be less efficient with the red rhodamine.
Preliminary characterization of coumarin-18 and rhodamine46-
based fluorogens prepared in our laboratory confirms the
predictive power of our mechanism. Given the popularity of
maleimide-based fluorescent labeling techniques, the mecha-
nistic conclusions of the photophysical study presented herein
should prove to be broadly applicable as a predictive model in
the rational design of future labeling agents.
Figure 6. Normalized absorption (9) and fluorescence (0) of 1 in
acetonitrile. The fluorescence spectrum was recorded in deaerated aceto-
nitrile solution with λex ) 335 nm.
Conclusions
redox potentials, together with the measured rapid kq, provide
sound evidence of dansylamide-maleimide quenching by a PET
mechanism.
In conclusion, the establishment of a new, convergent
synthetic route for the preparation of dimaleimide fluorogens
permitted facile access to a large number of derivatives to be
assessed as potential fluorescent protein labeling agents. In this
study, dansyl fluorophore and dimaleimide quenchers were
prepared and used to elucidate the mechanism of the dimale-
imide quench phenomenon. The Stern-Volmer rate constants,
lack of triplet formation, and the compatible redox potentials
measured for the maleimides and for dansylamide herein provide
the first unequivocal proof of a PET quench mechanism for
the dimaleimide dansyl system, as was recently proposed for
an analogous monomaleimide pyrene system.42 Knowledge of
this quenching mechanism will impact the design of future
fluorogenic protein labeling agents. The ensuing predictive
model will be tested, and the application of the resulting com-
pounds to protein labeling will be reported in successive work.
Recently, Maeda et al. studied the quench mechanism of
maleimide groups attached to certain alkynyl pyrene derivatives
designed as biomolecular probes.42 Their Stern-Volmer study
of pyrene fluorescence quenching by exogenous maleimide pro-
vided rate constants (kq) that were greater than diffusion
controlled limits, leading them to propose a PET quenching
mechanism. Specifically, they suggested that reduction of the
maleimide group occurs by photoinduced oxidation from the
excited pyrene. Herein, we propose a similar mechanism for
the dansyl fluorophore, based on acceptable diffusion-controlled
Stern-Volmer rate constants, supported by our LFP observa-
tions and unequivocally confirmed by our measured redox
potentials.
These mechanistic conclusions have immediate consequences
on the rational design of future fluorogenic labeling agents. For
example, the efficiency of the PET-mediated quenching would
be enhanced by using relatively rigid spacers between the
fluorophore and dimaleimide moieties, whereas increasing the
length of the spacer would reduce the PET efficiency. The PET
mechanism is further evidenced by the similar photophysical
results obtained from relatively short PEG and ethylene spacers
for compounds 10 and 8, respectively. Significantly different
quenching rate constants would have been observed for these
two compounds if a dynamic contact-type quenching mechanism
was present. These two derivatives will ultimately further
provide valuable insight into the mechanism with low-temper-
ature studies and in viscous solvents. Furthermore, the PET
quenching mechanism allows one to predict which fluorophores
should be susceptible to maleimide quenching, owing to the
Acknowledgment. The authors thank Ste´phane Girouard,
Christophe Loup, and Isabelle Roy (Keillor group) for their
initial efforts on this project and Marie Bourgeaux and Sergio
Perez (Skene group) for their generous technical assistance. This
work was financially supported by the Canadian Institutes of
Health Research (CIHR), the Natural Sciences and Engineering
Research Council of Canada (NSERC), the University of
Montreal Centre for Self-Assembled Chemical Structures
(CSACS), and through additional equipment funding from the
Canada Foundation for Innovation (CFI). K.C. is also grateful
to NSERC for an Undergraduate Student Research Award.
JA0738125
(43) Seidel, C. A. M.; Schulz, A.; Sauer, M. H. M. J. Phys. Chem. 1996, 100,
5541.
(44) Goze, C.; Ulrich, G.; Ziessel, R. J. Org. Chem. 2007, 72, 313.
(45) Austin, J. M.; Harrison, I. R.; Quickenden, T. I. J. Phys. Chem. 1986, 90,
1839.
(42) Maeda, H.; Maeda, T.; Mizuno, K.; Fujimoto, K.; Shimizu, H.; Inouye,
M. Chem.sEur. J. 2006, 12, 824.
(46) Guy, J.; Keillor, J. W. Unpublished observations.
9
J. AM. CHEM. SOC. VOL. 129, NO. 39, 2007 11977