A fluorogenic probe for the catalyst-free detection of azide-tagged molecules

Article abstract of DOI:10.1021/ja309000s

Fluorogenic reactions in which non- or weakly fluorescent reagents produce highly fluorescent products can be exploited to detect a broad range of compounds including biomolecules and materials. We describe a modified dibenzocyclooctyne that under catalyst-free conditions undergoes fast strain-promoted cycloadditions with azides to yield strongly fluorescent triazoles. The cycloaddition products are more than 1000-fold brighter compared to the starting cyclooctyne, exhibit large Stokes shift, and can be excited above 350 nm, which is required for many applications. Quantum mechanical calculations indicate that the fluorescence increase upon triazole formation is due to large differences in oscillator strengths of the S₀ S ₁ transitions in the planar C_2v-symmetric starting material compared to the symmetry-broken and nonplanar cycloaddition products. The new fluorogenic probe was successfully employed for labeling of proteins modified by an azide moiety.

Full text of DOI:10.1021/ja309000s

Article
pubs.acs.org/JACS
A Fluorogenic Probe for the Catalyst-Free Detection of Azide-Tagged
Molecules
Frederic Friscourt,† Christoph J. Fahrni,‡ and Geert-Jan Boons*,†
́
́
†Complex Carbohydrate Research Center, University of Georgia, 315 Riverbend Road, Athens, Georgia 30602, United States
‡School of Chemistry and Biochemistry and Petit Institute for Bioengineering and Bioscience, Georgia Institute of Technology, 901
Atlantic Drive, Atlanta, Georgia 30332, United States
S
* Supporting Information
ABSTRACT: Fluorogenic reactions in which non- or weakly
fluorescent reagents produce highly fluorescent products can
be exploited to detect a broad range of compounds including
biomolecules and materials. We describe a modified
dibenzocyclooctyne that under catalyst-free conditions under-
goes fast strain-promoted cycloadditions with azides to yield
strongly fluorescent triazoles. The cycloaddition products are
more than 1000-fold brighter compared to the starting
cyclooctyne, exhibit large Stokes shift, and can be excited
above 350 nm, which is required for many applications.
Quantum mechanical calculations indicate that the fluores-
cence increase upon triazole formation is due to large differences in oscillator strengths of the S0 ↔ S1 transitions in the planar
C2v-symmetric starting material compared to the symmetry-broken and nonplanar cycloaddition products. The new fluorogenic
probe was successfully employed for labeling of proteins modified by an azide moiety.
amine followed by cycloaddition with difluorinated cyclooctyne
(DIFO) conjugated to a fluorophore.15
INTRODUCTION
■
The selective and noninvasive labeling of biomolecules, ideally
in context of their endogenous environment, is a prerequisite
for a broad range of biological studies. As a genetically encoded
label, the green fluorescent protein (GFP) and its variants have
revolutionized cell biology by enabling the visualization of
target proteins within live cells, tissues, and whole organ-
isms.1−3 GFP is, however, not suitable for tagging metabolites
and post-translationally modified biomolecules such as lipids,
nucleic acids, and complex carbohydrates.
The bioorthogonal chemical reporter methodology, pio-
neered by Bertozzi and co-workers, is emerging as a versatile
method for labeling a wide variety of biomolecules.4−6 In this
approach, an abiotic chemical functionality (reporter) is
incorporated into a target biomolecule, which can be reacted
with a complementary bioorthogonal functional group linked to
a diverse set of probes, including biotin and fluorescent tags.
Organic azides are particularly versatile reporters due to their
small size and virtual absence in biological systems.7 They can
be conjugated by Staudinger ligation using modified
phosphines,8,9 copper(I)-catalyzed cycloaddition with terminal
alkynes (CuAAC),10−12 or by strain-promoted alkyne−azide
cycloaddition (SPAAC).13,14 The latter type of reaction is
attractive because it is fast and does not require a potentially
toxic metal catalyst, thereby offering opportunities for the
labeling of glycans, lipids and proteins in living cells, tissues or
whole organisms. For example, Bertozzi and co-workers have
shown that glycoconjugates of zebrafish embryos can be
visualized by metabolic labeling with N-azidoacetylgalactos-
We have demonstrated that 4-dibenzocyclooctynol (DIBO)
is an attractive reagent for SPAAC. It rapidly reacts with azido-
containing compounds and can be employed for tagging a wide
variety of biomolecules including cell surface glycoconjugates of
living cells.16 Dibenzylcyclooctynes can be readily derivatized,
which was exploited to influence their subcellular location via
incorporation of sulfates.17 Furthermore, the cycloadditions can
be accelerated by modifying the 8-membered ring of DIBO
with an sp2-hybridized atom.18−20 Dibenzocyclooctynes can
also be generated photochemically by short irradiation with UV
light of corresponding cyclopropenones, thereby making it
possible to label target substrates in a spatially and temporally
controlled manner.21 In addition to bioconjugation, SPAAC has
found wide utility in materials and surface chemistry.22
We envisaged that DIBO derivative 1, which contains a
reactive alkyne and an alkyne masked as a cyclopropenone,
would offer an attractive bifunctional reagent for tagging
biomolecules and materials under catalyst free conditions.23
Although the sequential SPAAC did not proceed satisfactorily,
we found that cycloaddition of 1 with azides yielded strongly
fluorescent triazoles that possess favorable photophysical
properties, such as exceptional brightness, a large Stokes shift,
and excitation above 350 nm. Quantum mechanical calculations
indicate that the fluorescence increase upon triazole formation
Received: September 10, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
is due to substantial differences in oscillator strengths of the
S0 ↔ S1 transitions in the planar C2v-symmetric Fl-DIBO (1)
compared to the symmetry-broken and nonplanar cyclo-
addition products. Compounds that become fluorescent upon
reaction with a chemical reporter have many attractive features
such as eliminating the need for probe washout, reducing
background labeling, offering opportunities for monitoring
biological processes in real time.24
constant of 0.019 M−1 s−1 (Supporting Information, Figure S1).
The reaction rate is somewhat lower than that of the parent
DIBO (0.057 M−1 s−1) indicating that the cyclopropenone
moiety does not induce additional ring strain.26
Photophysical Properties. An unexpected and exciting
observation was the strong fluorescence increase upon
conversion of the Fl-DIBO (1) probe to the corresponding
triazole products 8a−e (Figure 1b). Thus, excitation of Fl-
DIBO 1 at 420 nm produces only a weak emission band
centered around 574 nm with a low quantum yield of 0.2%
(Table 1 and Figure 1b). In contrast, triazole 8a, which was
obtained by cycloaddition of 1 with triethyleneglycol methyl
monoester azide, fluoresces strongly upon excitation at 370 nm
with a maximum intensity at 491 nm and a quantum yield of
11.9% (Table 1 and Figure 1b). Because of the large difference
in the extinction coefficients of Fl-DIBO (1) (ε420 nm = 190
M−1 cm−1) and the triazole product 8a (ε370 nm = 4300 M−1
cm−1), the apparent brightness, defined as the product of
quantum yield and molar absorption coefficient, is more than
1000-fold increased. Whereas many common fluorophores
including fluorescein, rhodamine, or BODIPY derivatives
exhibit small Stokes shifts that may cause reabsorption of
emitted photons, triazole 8a possesses an unusually large Stokes
shift of 7180 cm−1 (128 nm).
RESULTS AND DISCUSSION
■
Synthesis of 1 and Cycloaddition to Azides. The
preparation of target compound 1 commenced by Sonogashira
cross-coupling of 3-ethynylanisole (2) and 3-iodoanisole (3) to
give the symmetrical ethyne 4 in near quantitative yield
(Scheme 1). Partial hydrogenation of 4 using Lindlar’s catalyst
a
Scheme 1. Synthesis of Fl-DIBO
More detailed examination of compounds 8a−e revealed that
the nature of the triazole substituent does not substantially
affect the photophysical properties (Table 1 and Supporting
Information, Figures S8−S18). Only compound 8c showed a
significant red-shift of the emission maximum along with a
reduced quantum yield and shorter fluorescence lifetime,
presumably due to direct π-interaction of the anisole
substituent with the fluorophore core π-system. The radiative
deactivation rate constants of 8a−e, calculated from the
fluorescence quantum yields and lifetimes, are narrowly
distributed with an average of 1.13 0.16 × 107 s−1, suggesting
a single and uniform deactivation pathway.
Furthermore, compounds 8a−e showed a pronounced
negative solvatochromism, indicating a less polar excited state
compared to the ground state. For example, the emission
maximum of benzyl triazole 8b moved from 493 nm in toluene
to 482 nm in methanol (Supporting Information, Figure S7).
Finally, treatment of the triazole 8b (10−5 M in MeOH) with
glutathione (10 mM) for 2 h resulted in marginal bleaching.
Strong reducing agents such as DTT caused, however,
complete bleaching within 30 min, probably by addition of
DDT to the cyclopropenone moiety of 8b (Supporting
Information, Figure S6).
Protein Labeling. Next, Fl-DIBO (1) was examined as a
fluorogenic labeling reagent for proteins. For this purpose,
bovine serum albumin (BSA) was modified with an azide
moiety by reaction of the Cys34 sulfhydryl group with N-{2-[2-
(2-azidoethoxy)ethoxy]ethyl}-2-iodoacetamide to give 9 (Sup-
porting Information). A stock solution of Fl-DIBO (1) in
ethanol was then added to a solution of BSA-azide (9) in PBS
containing 1% SDS and the resulting reaction mixture was
incubated at 37 °C for 18 h. An aliquot was analyzed by SDS−
PAGE followed by fluorescence imaging and staining with
Coomassie Blue for protein detection (Figure 2, lane 1). As
expected, a strong fluorescent band was observed at 66 kDa. As
negative controls, Fl-DIBO was excluded (lane 3), or BSA-azide
replaced by unmodified BSA (lane 2) and, gratifyingly, the
fluorescent images showed an absence of background labeling.
As an additional control, BSA-azide was exposed to DIBO-
a
Reagents and conditions: (a) Pd(PPh3)4, CuI, (iPr)2NEt, THF,
reflux, 18 h, 97%; (b) H2, Lindlar′s catalyst, quinoline, hexane, rt, 20
min, 89%; (c) AlCl3, tetrachlorocyclopropene, CH2Cl2, −20 °C to rt, 4
h, H2O, 62%; (d) Br2, CH2Cl2, 0 °C, 2 h, 87%; (e) KOH, EtOH, rt, 18
h, 76%.
produced Z-alkene 5, and subsequent Friedel−Crafts alkylation
with tetrachlorocyclopropene in the presence of AlCl3
21,25
yielded the corresponding annulated dichlorocyclopropene
intermediate, which after in situ hydrolysis gave cyclo-
propenone derivative 6 in 62% overall yield. Selective
bromination of the bridging double bond provided dibromo-
cyclopropenone derivative 7, which upon treatment with
ethanolic KOH solution underwent a double-elimination
reaction to give the target compound 1. Despite the
considerable strain imposed by the central ring system,
compound 1 is stable in aqueous solution over a broad pH
range and inert toward treatment with a diverse set of
nucleophiles including glutathione (10 mM) (Supporting
Information, Figures S2−S5).
Triazoles 8a−e were obtained in near quantitative yield by
cycloaddition of 1 with the corresponding azide derivatives in a
mixture of dichloromethane and methanol (4:1, v/v) at room
temperature in the course of 2 h (Figure 1a). The progress of
the reaction of 1 with benzyl azide in a mixture of CDCl3 and
1
CD3OD (4:1, v/v) was monitored by H NMR by integration
of the benzylic proton signals, and yielded a second-order rate
B
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. (a) SPAAC between Fl-DIBO probe (1) and various azides; (b) various triazole products generated from the corresponding azides with Fl-
DIBO probe; absorption (solid blue trace) and emission (solid green trace, λexc = 420 nm) spectra of Fl-DIBO and absorption (dashed blue trace)
and emission (dashed green trace, λexc = 370 nm) spectra of triethyleneglycol triazole 8a recorded in MeOH. Inset images compare the visible
fluorescence emission of the Fl-DIBO probe (left cuvette) to the triazole product (8a) (right cuvette) under UV excitation (365 nm).
Table 1. Photophysical Properties of Fl-DIBO (1) and Triazoles 8a−e in Methanol at 298 K
compound
1
8a
8b
8c
8d
8e
e
λabs [nm]
420
363
364
366
360
362
a
εabs [M−1 cm−1
λem [nm]
]
190
4300
4600
489
1700
499
5000
6100
574
491
492
489
Stokes shift [cm−1
]
6390
0.002
7180
7020
0.12
7280
0.065
7450
7170
b
ΦF
0.119
0.087
8.81 0.01
0.99
0.12
c
f
g
h
Life time [ns]
kr [107 s−1
1.61 0.01
0.12
9.55 0.01
8.59 0.01
1.16
7.10 0.01
0.92
9.80 0.01
d
]
1.25
1.22
a
b
c
Determined at 420 nm for 1 and at 370 nm for 8a−e. Fluorescence quantum yield, quinine sulfate in 1.0 N H2SO4 as standard. Fluorescence
d
e
decay acquired at 490 nm and fitted to a monoexponential decay model. Radiative deactivation rate constant (kr = Φf/τf). Maximum of excitation
spectrum. Natural decay lifetime based on biexponential decay with τ1 = 0.51 ns (62%) and τ2 = 3.40 ns (38%). Natural decay lifetime based on
biexpoential decay with τ1 = 9.43 ns (88%) and τ2 = 2.41 ns (12%). Acquired at 500 nm.
f
g
h
FITC and after appropriate washing steps a fluorescent band
was observed at 66 kDa (lane 4). A similar reaction with
unmodified BSA (lane 5) showed faint background labeling,
thus highlighting superior properties of 1.27,28
Computational Studies. The unexpected and large
increase in quantum yield upon conversion of Fl-DIBO (1)
to the triazoles 8a−e prompted us to explore the underpinning
mechanism by quantum chemical calculations. The fluores-
cence switching of previously reported fluorogenic CuAAC
reagents24,29−31 can be rationalized by considering the relative
energy levels of the emissive singlet 1(π−π*)-excited state and a
quenching triplet 3(n−π*)-state,32 which is typically the lowest
excited state in azido-substituted fluorophores.33 Upon
3
conversion to the triazole, the quenching (n−π*)-state rises
above the emissive 1(π−π*)-state and the fluorescence is
switched on. A similar excited state inversion mechanism is also
responsible for the fluorogenic response of an alkyne-
substituted coumarin derivative.29 In this case, triazole
formation increases the charge-transfer character of the
1
emissive (π−π*)-state, which in polar solvents is sufficiently
3
stabilized to move below the quenching (n−π*)-state. Similar
to coumarins, excitation of the carbonyl lone-pair electrons in
3
Fl-DIBO (1) could yield an energetically low-lying (n−π*)-
state, which upon conversion to the triazoles might then rise
C
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Formation of azido-BSA and its labeling with Fl-DIBO probe. Purified azido-BSA (50 μM) and native BSA (50 μM) were labeled with 250
μM of Fl-DIBO (1) or DIBO-FITC for 18 h at 37 °C in PBS containing 10% EtOH and 1% SDS. The crude reaction mixtures were separated by
SDS−PAGE and the gel was analyzed by fluorescence imaging (top row; λexc = 365 nm; λdetec = 480 nm) and by Coomassie Blue stain to reveal total
protein content (bottom row).
a
Table 2. Computational Data for Fl-DIBO (1) and Triazole 8f (R = CH3)
b
d
d
cmpd
data
exp
state
TD-B3LYP/6-31+G(d) gasphase TD-B3LYP/6-31+G(d) PCM
TD-CAM-B3LYP/6-31+G(d) PCM
c
c
1
absorption [eV]
2.95
S1 (π−π*)
S2 (n−π*)
S1 (π−π*)
2.60 (0.014)
2.79 (0.0001)
1.78 (0.007)
2.19
2.60 (0.023)
3.24 (0.0002)
1.82 (0.011)
2.21
2.95 (0.023)
4.01 (0.0003)
2.06 (0.011)
2.50
emission [eV]
2.16
2.56
3.41
e
E00 [eV]
8f
absorption [eV]
S1 (π−π*)
S2(n−π*)
S1 (π−π*)
3.17 (0.093)
3.27 (0.0152)
2.13 (0.026)
2.65
3.23 (0.18)
3.68 (0.0123)
2.26 (0.058)
2.75
3.65 (0.23)
4.37 (0.028)
2.56 (0.071)
3.11
emission [eV]
2.53
3.06
e
E00 [eV]
f
MAE [eV]
0.36
0.30
0.08
a
b
c
All geometries were optimized at the B3LYP/6-31G(d) level of theory. Lowest energy absorption band maximum of 8a in methanol. Vertical
d
e
f
excitation energy; oscillator strength in parentheses. Polarized continuum model with methanol as solvent. Excited state equilibrium energy. Mean
absolute error.
1
above the emissive (π−π*)-level. To explore this scenario, we
calculated the excited state manifold of Fl-DIBO (1) and the
corresponding methyl-substituted triazole product 8f by TD-
DFT at the B3LYP/6-31+G(d)//B3LYP/6-31G(d) level of
theory and gauged the validity of the computational results
based on the experimental absorption and emission data. As
evident from the data compiled in Table 2, the vertical
excitation and emission energies were underestimated by an
average of 0.39 eV, a value that falls within the error margin of
previous benchmark tests of TD-DFT/B3LYP.34 Accounting
for solvent effects with a polarized continuum model (PCM)
yielded further improved values, and switching to the long-
range corrected hybrid functional CAM-B3LYP35 reproduced
the experimental data very well with a low average error of 0.08
eV. Although the vertical excitation energy is not a measurable
quantity as the electronic spectrum is further resolved into
vibrational bands, the experimental absorption maximum for
symmetry-allowed transitions lies usually close to the vertical
excitation energy,36 and thus may still serve as a meaningful
gauge. As illustrated with the electron density difference plots
and the state energy diagram in Figure 3, the lowest energy
absorption and emission bands correspond to (π−π*) states,
1
both for Fl-DIBO (1) and the triazole 8f. Furthermore, the
1(n−π*) state of Fl-DIBO (1) resides by more than 1 eV above
the emissive 1(π−π*) state, suggesting that nonradiative
deactivation via a 1(π−π*)−3(π−π*) route is unlikely
responsible for the low quantum yield. Within the C2v-
symmetric molecular framework, the two states have B2 and
B1 symmetry, respectively, thus rendering such a deactivation
pathway also symmetry-forbidden. Given the rigid molecular
architecture of Fl-DIBO (1), excited state deactivation
pathways that require large structural changes, for example,
via a conical intersection to a lower-lying triplet state, appear
also improbable. Rather, the low quantum yield of Fl-DIBO (1)
likely originates from the much weaker oscillator strengths and
lower energy associated with the S0 ↔ S1 transitions compared
to the triazoles 8. According to classical theory, the radiative
rate constant kr for emission is proportional to the product of
D
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1
1
Figure 3. Excited state manifold for (A) Fl-DIBO (1) and (B) the triazole 8f illustrating the relative energies of the (π−π*) and (n−π*) singlet
states in methanol (TD-DFT at the CAM-B3LYP/6-31+G(d)//B3LYP/6-31G(d) level of theory, including PCM solvent correction). The color
plots illustrate the total electron density difference between the respective ground and excited states (decreasing density shown in blue, increasing
density red; GS = ground state; S1, S2 = first and second excited singlet state).
the oscillator strength and the squared state energy.37 On the
basis of the data in Table 2, kr is expected to be approximately
13-times slower for 1 compared to 8f, a value that mirrors well
the experimental data (Table 1).
On the basis of an extensive set of structural and
computational data, it was recently suggested that the alkyne
bond angle in cyclooctyne derivatives may serve as a predictor
for their reactivity.38 Consistent with this proposal, the alkyne
bond angles in the ground state equilibrium geometry of Fl-
DIBO (1) are 156°, thus implying a reduced ring strain
compared to the more reactive biarylazacyclooctynone
derivative BARAC with alkyne bond angles of 153°.38
ring with a coumarin fluorophore.41 Although the reaction
product with 2-azidoethanol exhibited a 10-fold increased
emission compared to unreacted CoumBARAC, it exhibited a
low fluorescence quantum yield and required excitation at 300
nm, a wavelength regime that is incompatible with many
biological applications. In contrast, the cycloaddition product of
Fl-DIBO (1) is more than 1000-fold brighter compared to the
unreacted reagent, offers a large Stokes shift, and can be excited
above 350 nm, the typical cutoff wavelength of standard
fluorescence microscopes. Quantum mechanical calculations
indicate that the fluorescence increase upon triazole formation
is not due to a (n−π*)/(π−π*) inversion mechanism as found
in previous probes but rather related to the substantial
differences in oscillator strengths of the S0 ↔ S1 transitions
in the planar C2v-symmetric Fl-DIBO (1) compared to the
symmetry-broken and nonplanar cycloaddition products.
Because the fluorescence switching mechanism does not rely
on the alteration of excited state energy levels, the turn-on
response of Fl-DIBO (1) is very robust and virtually
independent from the polarity of the surrounding environment,
thus eliminating the potential for artifacts in complex biological
samples. Compared to previously reported fluorogenic click
reagents, Fl-DIBO (1) does not require a metal-catalyst for
bioconjugation while still maintaining rapid reaction kinetics.
Given these favorable properties, Fl-DIBO (1) represents a
significant advance and is expected to facilitate fluorescent
labeling of biomolecules and materials while eliminating the
problem of background labeling and need of washing steps. The
proposed fluorescent turn-on mechanism will offer unique
opportunities to design additional fluorogenic probes with
modified properties.
CONCLUSIONS
■
Bioorthogonal fluorogenic reactions in which non- or weakly
fluorescent reagents produce highly fluorescent products
greatly expand the possibilities for detecting a broad range of
compounds and are particularly advantageous for applications
in which probe washout is not possible or desirable.24
A
number of such probes have been developed by modifying
common fluorophore platforms such as anthracene, BODIPY,
or coumarins with an azide moiety as quencher.29−31,39,40 Upon
Cu(I)-catalyzed cycloaddition to an alkyne-tagged target
molecule, the quenching azide is converted to a triazole and
the fluorophore emission restored. While in principle the same
approach could be used for strain-promoted cycloadditions, the
corresponding tags would be inevitably much larger compared
to the rather inconspicuous propargyl groups used in CuAAC,
and the increased molecular size might adversely affect the
properties of smaller target molecules, especially with regard to
their biodistribution and biological activity. For this reason, the
preferred approach for developing a catalyst-free fluorogenic
click reagent is to utilize azide as the tagging moiety and to
integrate the strained and bulky cyclooctyne group into the
fluorophore structure. A first attempt toward this goal was
recently reported by Bertozzi and co-workers, who designed the
fluorogenic reagent CoumBARAC by fusing the cyclooctyne
EXPERIMENTAL SECTION
■
Bis(3-methoxyphenyl)acetylene (4). N,N-Diisopropylethyl-
amine (1.5 mL, 9.0 mmol) was added dropwise to a solution of 3-
ethynylanisole (0.45 mL, 3.6 mmol), 3-iodoanisole (0.36 mL, 3.0
mmol), tetrakis(triphenylphosphine)palladium(0) (170 mg, 0.15
mmol) and copper(I) iodide (60 mg, 0.3 mmol) in THF (15 mL).
E
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The reaction mixture was then refluxed overnight. After being cooled
to room temperature, all volatiles were removed under reduced
pressure. The residue was then purified by flash column chromatog-
raphy on silica gel using a mixture of hexane and ethyl acetate (8:1)
The reaction mixture was then stirred at room temperature overnight.
The solution was quenched with an aqueous solution of HCl (1N,
until pH≈6). The mixture was then extracted with CH2Cl2 (3 × 10
mL) and the combined organic extracts were dried over MgSO4,
filtered and concentrated under reduced pressure. The residue was
then purified by flash chromatography on silica gel (16 g) using a
mixture of 1% MeOH in CH2Cl2 affording pure cyclooctyne 1 (116
1
affording pure alkyne 4 as a colorless solid (694 mg, 97%): H NMR
(300 MHz, CDCl3): d 3.82 (s, 6H), 6.89 (dd, J = 7.8, 2.0 Hz, 2H),
7.06 (br s, 2H), 7.13 (d, J = 7.8 Hz, 2H), 7.25 (t, J = 7.8 Hz, 2H); 13C
NMR (75.5 MHz, CDCl3): d 55.44 (2 × CH3), 89.26 (2 × C), 115.15
(2 × CH), 116.49 (2 × CH), 124.31 (2 × C), 124.35 (2 × CH),
129.55 (2 × CH), 159.50 (2 × C); HRMS (m/z): [M+] calcd. for
C16H14O2, 238.0994; found, 238.0971.
1
mg, 76%) as a yellow solid: H NMR (300 MHz, CDCl3): δ 3.80 (s,
6H), 6.45 (d, J = 2.5 Hz, 2H), 6.63 (dd, J = 8.5, 2.5 Hz, 2H), 7.47 (d, J
= 8.5 Hz, 2H); 13C NMR (75.5 MHz, CDCl3): δ 55.81 (2 × CH3),
106.82 (2 × C), 113.70 (2 × CH), 114.80 (2 × CH), 125.38 (2 × C),
126.87 (2 × C), 136.76 (2 × CH), 146.48 (2 × C), 153.81 (CO),
163.39 (2 × C); HRMS (m/z): [M+H+] calcd. for C19H13O3,
289.0859; found, 289.0813.
General Procedure for Triazole Formation. A solution of Fl-
DIBO 1 (14.4 mg, 0.05 mmol) and the respective azide 10a, 10b
(BnN3), 10c, 10d, and 10e (0.1 mmol) in a mixture of CH2Cl2 and
methanol (4:1, 5 mL) was stirred at room temperature for 2 h. All
volatiles were removed under reduced pressure and the residue was
purified by flash chromatography on silica gel (7 g) using an
appropriate mixture of MeOH in CH2Cl2 affording pure triazoles 8a−
e, respectively (for product characterization see Supporting
Information).
(Z)-3,3′-Dimethoxystilbene (5). A solution of alkyne 4 (672 mg,
2.82 mmol), Lindlar’s catalyst (135 mg, 20% w/w) and quinoline (0.72
mL, 6.1 mmol) in hexane (17 mL) was stirred at room temperature
under 1 atm of H2 for 20 min. The mixture was then filtered through
Celite and washed with ethyl acetate (20 mL). The filtrate was then
concentrated under reduced pressure and the residue was purified by
flash column chromatography on silica gel using a mixture of hexane
and ethyl acetate (8:1) affording pure (Z)-alkene 5 as a colorless oil
(604 mg, 89%): 1H NMR (300 MHz, CDCl3): d 3.64 (s, 6H), 6.57 (s,
2H), 6.73 (dd, J = 7.9, 2.5 Hz, 2H), 6.80 (br s, 2H), 6.84 (d, J = 7.9
Hz, 2H), 7.14 (t, J = 7.9 Hz, 2H); 13C NMR (75.5 MHz, CDCl3): d
55.15 (2 × CH3), 113.41 (2 × CH), 113.93 (2 × CH), 121.62 (2 ×
CH), 129.31 (2 × CH), 130.46 (2 × CH), 138.65 (2 × C), 159.49 (2
× C); HRMS (m/z): [M+] calcd. for C16H16O2, 240.1150; found,
240.1109.
Kinetic Measurements. The rate measurements of cycloaddition
1
of cyclooctyne 1 with benzyl azide 10b were conducted by using H
NMR spectroscopy (Varian Mercury 300 MHz) at 25 °C. A 20 mM
solution of benzyl azide 10b (0.2 mL) in CDCl3:MeOD (4:1) was
added to a thermally equilibrated solution of cyclooctyne 1 (10 mM,
0.4 mL) in a mixture of CDCl3:MeOD (4:1), leading to a mixture of
both reactants in 1:1 ratio with a respective concentration of 6.66 mM.
Reactions were monitored by following the decay of characteristic
peaks of cyclooctyne 1 (d = 3.78 (2 × OCH3), 6.43 (2 × CHAr) and
6.63 (2 × CHAr) ppm) as well as the formation of characteristic
triazole peaks (d = 3.65 (OCH3), 3.89 (OCH3), 5.43 (CH2-triazole)
ppm). Consumption of starting materials followed a second-order
equation and the second-order rate constants were obtained by least-
squares fitting of the data to a linear equation (Figure S1).
Azido-BSA (9). A solution of N-{2-[2-(2-azidoethoxy)ethoxy]-
ethyl}-2-iodoacetamide 17 (50 mM, 0.6 mL) in a mixture of PBS
buffer (20 mM, pH 7.3) and CH3CN (1:1) was added to a solution of
BSA (1 mM, 1 mL) in PBS buffer (20 mM, pH 7.3). The reaction
mixture was shaken overnight at room temperature, dialyzed (MWCO
= 10 kDa) against a mixture of PBS buffer (20 mM, pH 7.3) and
CH3CN (4:1) and stored at 4 °C.
4,9-Dimethoxy-1H-dibenzo[a,e]cyclopropa[c]cycloocten-1-
one (6). A solution of aluminum chloride (400 mg, 3.0 mmol) and
tetrachlorocyclopropene (0.15 mL, 1.2 mmol) in CH2Cl2 (15 mL) was
stirred at room temperature for 20 min. The reaction mixture was then
cooled to −20 °C and a solution of (Z)-alkene 5 (240 mg, 1.0 mmol)
in CH2Cl2 (5 mL) was added. The reaction mixture was stirred at −20
°C for 1 h, was then allowed to warm to room temperature over a
period of 2 h and was stirred for an extra h at room temperature.
Water (20 mL) was then added and the reaction mixture was stirred
for an extra 30 min. The organic layer was then extracted with CH2Cl2
(3 × 20 mL), dried over MgSO4 and concentrated under reduced
pressure. The residue was then purified by flash chromatography on
silica gel (16 g) using a mixture of 2% MeOH in CH2Cl2 affording
1
pure cyclopropenone 6 (180 mg, 62%) as a yellow/orange solid: H
NMR (300 MHz, CDCl3): δ 3.82 (s, 6H), 5.99 (s, 2H), 6.60 (d, J =
2.5 Hz, 2H), 6.69 (dd, J = 8.4, 2.5 Hz, 2H), 7.44 (d, J = 8.4 Hz, 2H);
13C NMR (75.5 MHz, CDCl3): δ 55.66 (2 × CH3), 113.45 (2 × CH),
115.34 (2 × C), 121.75 (2 × CH), 131.87 (2 × CH), 136.13 (2 ×
CH), 140.01 (2 × C), 147.73 (2 × C), 152.39 (CO), 163.52 (2 ×
C); HRMS (m/z): [M + H+] calcd. for C19H15O3, 291.1016; found,
291.1011.
SPAAC with Azido-BSA. A solution of Fl-DIBO 1 (2.5 mM, 20
μL) in ethanol was added to a solution of azido-BSA 9 (50 μM, 180
μL) in PBS buffer (20 mM, pH 7.3, containing 1% SDS). The reaction
mixture was incubated at 37 °C overnight, separated by SDS−PAGE
(10% gradient) and the gel was analyzed by fluorescence imaging (λexc
= 365 nm; λdetec = 480 nm) and by Coomassie Blue stain to reveal total
protein content.
6,7-Dibromo-4,9-dimethoxy-6,7-dihydro-1H-dibenzo[a,e]-
cyclopropa[c]cycloocten-1-one (7). A solution of bromine (48 μL,
0.93 mmol) in CH2Cl2 (4 mL) was added dropwise to a solution of
cyclopropenone 6 (180 mg, 0.62 mmol) in CH2Cl2 (8 mL) at 0 °C.
The reaction mixture was allowed to warm to room temperature and
was stirred for 4 h. The mixture was then quenched with a saturated
aqueous solution of sodium thiosulfate (10 mL). The two phases were
then separated and the aqueous layer was further extracted with
CH2Cl2 (3 × 10 mL). The combined organic extracts were dried over
MgSO4, filtered and concentrated under reduced pressure. The residue
was then purified by flash chromatography on silica gel (16 g) using a
mixture of 2% MeOH in CH2Cl2 affording pure dibromo-cyclo-
propenone 7 (243 mg, 87%) as a colorless solid: 1H NMR (300 MHz,
CDCl3): δ 3.91 (s, 6H), 5.75 (s, 2H), 6.97 (d, J = 2.5 Hz, 2H), 7.05
(dd, J = 8.5, 2.5 Hz, 2H), 8.06 (d, J = 8.5 Hz, 2H); 13C NMR (75.5
MHz, CDCl3): δ 50.44 (2 × CH), 55.88 (2 × CH3), 114.14 (2 × CH),
115.71 (2 × C), 118.24 (2 × CH), 137.16 (2 × CH), 141.74 (2 × C),
142.96 (2 × C), 152.71 (CO), 162.72 (2 × C); HRMS (m/z): [M+]
calcd. for C19H14O379Br2, 447.9310; found, 447.9313.
Absorption and Fluorescence Measurements. UV−vis spectra
were recorded using Varian Cary Bio50 UV−vis spectrophotometer at
25
0.1 °C. Fluorescence spectra were recorded with a PTI
fluorimeter using a cuvette with 1 cm path length. All spectra were
corrected for the spectral response of the detection system and for the
spectral irradiance of the excitation source (via a calibrated
photodiode). For spectra, see Supporting Information Figures S8−
S13.
Quantum Yield Determination. Quantum yields were deter-
mined from the slope of the integrated fluorescence emission between
360 and 700 nm (excitation at 346 nm) versus absorbance using
quinine sulfate (Φf = 0.54 0.03) as fluorescence standard.42 For each
compound, four data points were acquired with absorbances ranging
between 0.05 and 0.5 (l = 10 cm).
Computational Studies. Quantum chemical calculations were
carried out with the Gaussian 09 suite of programs (Rev. C01). All
ground-state geometries were energy minimized by DFT with the
B3LYP hybrid functional and Pople’s 6-31G(d) split valence basis set
with added polarization functions. In case of compound 1, geometry
4,9-Dimethoxy-6,7-didehydro-1H-dibenzo[a,e]cyclopropa-
[c]cycloocten-1-one (1). A solution of potassium hydroxide (300
mg, 5.30 mmol) in ethanol (5 mL) was added to a solution of
dibromo-cyclopropenone 7 (240 mg, 0.53 mmol) in ethanol (25 mL).
F
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
optimizations were performed within the C2v point group. To ensure a
stationary point on the ground-state potential surface, the optimized
geometries were verified by a vibrational frequency analysis. Vertical
excitation energies were calculated at the ground-state equilibrium
geometry based on TD-DFT with the B3LYP or CAM-B3LYP35
hybrid functionals and the 6-31+G(d) basis set with added diffuse
functions. To estimate fluorescence emission energies, the geometry of
the lowest excited state was optimized by TD-DFT at the B3LYP/6-
31G(d) level of theory and the vertical emission energy was calculated
at the excited state geometry (TD-DFT with B3LYP/6-31+G(d)).
Bulk solvent effects were evaluated based on the polarizable
continuum model (PCM) using default parameters.43 For this
purpose, ground state geometries were energy minimized with PCM
equilibrium solvation (Tables S1 and S2), and the vertical excitation
energies were subsequently calculated with ground-state solvation
using TD-DFT in combination with B3LYP/6-31+G(d) or CAM-
B3LYP/6-31+G(d), respectively. Similarly, fluorescence emission
energies were obtained based on TD-DFT geometry optimizations
with state-specific PCM equilibrium solvation,44,45 followed by a TD-
DFT calculation with state-specific solvation using B3LYP/6-31+G(d)
or CAM-B3LYP/6-31+G(d), respectively. Total electron density
differences between the ground and excited states were visualized
with VMD46 based on Gaussian cube output files.
(12) Soriano Del Amo, D.; Wang, W.; Jiang, H.; Besanceney, C.; Yan,
A. C.; Levy, M.; Liu, Y.; Marlow, F. L.; Wu, P. J. Am. Chem. Soc. 2010,
132, 16893.
(13) Debets, M. F.; van Berkel, S. S.; Dommerholt, J.; Dirks, A. T. J.;
Rutjes, F. P. J. T.; van Delft, F. L. Acc. Chem. Res. 2011, 44, 805.
(14) Jewett, J. C.; Bertozzi, C. R. Chem. Soc. Rev. 2010, 39, 1272.
(15) Laughlin, S. T.; Baskin, J. M.; Amacher, S. L.; Bertozzi, C. R.
Science 2008, 320, 664.
(16) Ning, X.; Guo, J.; Wolfert, M. A.; Boons, G.-J. Angew. Chem., Int.
Ed. 2008, 47, 2253.
(17) Friscourt, F.; Ledin, P. A.; Mbua, N. E.; Flanagan-Steet, H. R.;
Wolfert, M. A.; Steet, R.; Boons, G.-J. J. Am. Chem. Soc. 2012, 134,
5381.
(18) Mbua, N. E.; Guo, J.; Wolfert, M. A.; Steet, R.; Boons, G.-J.
ChemBioChem 2011, 12, 1912.
(19) Debets, M. F.; van Berkel, S. S.; Schoffelen, S.; Rutjes, F. P. J. T.;
van Hest, J. C. M.; van Delft, F. L. Chem. Commun. 2010, 46, 97.
(20) Jewett, J. C.; Sletten, E. M.; Bertozzi, C. R. J. Am. Chem. Soc.
2010, 132, 3688.
(21) Poloukhtine, A. A.; Mbua, N. E.; Wolfert, M. A.; Boons, G.-J.;
Popik, V. V. J. Am. Chem. Soc. 2009, 131, 15769.
(22) Canalle, L. A.; van Berkel, S. S.; de Haan, L. T.; van Hest, J. C.
M. Adv. Funct. Mat. 2009, 19, 3464.
(23) Kii, I.; Shiraishi, A.; Hiramatsu, T.; Matsushita, T.; Uekusa, H.;
Yoshida, S.; Yamamoto, M.; Kudo, A.; Hagiwara, M.; Hosoya, T. Org.
Biomol. Chem. 2010, 8, 4051.
(24) Le Droumaguet, C.; Wang, C.; Wang, Q. Chem. Soc. Rev. 2010,
39, 1233.
(25) Tobey, S. W.; West, R. J. Am. Chem. Soc. 1964, 86, 1459.
(26) Dommerholt, J.; Schmidt, S.; Temming, R.; Hendriks, L. J.;
Rutjes, F. P. J. T.; van Hest, J. C. M.; Lefeber, D. J.; Friedl, P.; van
Delft, F. L. Angew. Chem., Int. Ed. 2010, 49, 9422.
(27) van Geel, R.; Pruijn, G. J. M.; van Delft, F. L.; Boelens, W. C.
Bioconjugate Chem. 2012, 23, 392.
(28) Yao, J. Z.; Uttamapinant, C.; Poloukhtine, A.; Baskin, J. M.;
Codelli, J. A.; Sletten, E. M.; Bertozzi, C. R.; Popik, V. V.; Ting, A. Y. J.
Am. Chem. Soc. 2012, 134, 3720.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthetic procedures, kinetic data (Figure S1), stability studies
(Figures S2−S6), solvatochromism (Figure S7), absorption and
fluorescence spectra (Figures S8−S13), time-dependent
fluorescence decay profiles (Figures S14−S18), computational
studies (Tables S1−S4) and NMR spectra of synthesized
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
(29) Zhou, Z.; Fahrni, C. J. J. Am. Chem. Soc. 2004, 126, 8862.
(30) Sivakumar, K.; Xie, F.; Cash, B. M.; Long, S.; Barnhill, H. N.;
Wang, Q. Org. Lett. 2004, 6, 4603.
Corresponding Author
gjboons@ccrc.uga.edu
(31) Sawa, M.; Hsu, T.-L.; Itoh, T.; Sugiyama, M.; Hanson, S. R.;
Vogt, P. K.; Wong, C.-H. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 12371.
(32) Lower, S. K.; Elsayed, M. A. Chem. Rev. 1966, 66, 199.
(33) Reiser, A.; Marley, R. Trans. Faraday Soc. 1968, 64, 1806.
(34) Silva-Junior, M. R.; Schreiber, M.; Sauer, S. P. A.; Thiel, W. J.
Chem. Phys. 2008, 129, 104103.
(35) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393,
51.
(36) Roos, B. In Theoretical and Computational Chemistry. Vol 16.
Computational Photochemistry; Olivucci, M., Ed.; Elsevier: Boston, MA,
2005; Vol. 16, p 317.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Financial support from the National Institutes of Health
(5R01CA088986, 5P41RR005351, and 8P41GM103390 to G.-
J.B.) and (R01GM067169 to C.J.F.) is gratefully acknowledged.
REFERENCES
■
(37) Turro, N. J. Modern Molecular Photochemistry; University
Science Books: Mill Valley, CA,1991.
(1) Crivat, G.; Taraska, J. W. Trends Biotechnol. 2012, 30, 8.
(2) Lippincott-Schwartz, J.; Patterson, G. H. Science 2003, 300, 87.
(3) Zhang, J.; Campbell, R. E.; Ting, A. Y.; Tsien, R. Y. Nat. Rev. Mol.
Cell Biol. 2002, 3, 906.
(38) Gordon, C. G.; Mackey, J. L.; Jewett, J. C.; Sletten, E. M.; Houk,
K. N.; Bertozzi, C. R. J. Am. Chem. Soc. 2012, 134, 9199.
(39) Devaraj, N. K.; Hilderbrand, S.; Upadhyay, R.; Mazitschek, R.;
Weissleder, R. Angew. Chem., Int. Ed. 2010, 49, 2869.
(40) Lemieux, G. A.; De Graffenried, C. L.; Bertozzi, C. R. J. Am.
Chem. Soc. 2003, 125, 4708.
(41) Jewett, J. C.; Bertozzi, C. R. Org. Lett. 2011, 13, 5937.
(42) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991.
(43) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999.
(44) Improta, R.; Barone, V.; Scalmani, G.; Frisch, M. J. J. Chem.
Phys. 2006, 125, 054103.
(45) Improta, R.; Scalmani, G.; Frisch, M. J.; Barone, V. J. Chem.
Phys. 2007, 127, 074504.
(46) Humphrey, W.; Dalke, A.; Schulten, K. J. Mol. Graphics Modell.
1996, 14, 33.
(4) Best, M. D.; Rowland, M. M.; Bostic, H. E. Acc. Chem. Res. 2011,
44, 686.
(5) Jing, C.; Cornish, V. W. Acc. Chem. Res. 2011, 44, 784.
(6) Sletten, E. M.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2009, 48,
6974.
(7) Debets, M. F.; van der Doelen, C. W. J.; Rutjes, F. P. J. T.; van
Delft, F. L. ChemBioChem 2010, 11, 1168.
(8) Schilling, C. I.; Jung, N.; Biskup, M.; Schepers, U.; Brase, S.
̈
Chem. Soc. Rev. 2011, 4840.
(9) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007.
(10) Meldal, M.; Tornoe, C. W. Chem. Rev. 2008, 108, 2952.
(11) Kennedy, D. C.; McKay, C. S.; Legault, M. C. B.; Danielson, D.
C.; Blake, J. A.; Pegoraro, A. F.; Stolow, A.; Mester, Z.; Pezacki, J. P. J.
Am. Chem. Soc. 2011, 133, 17993.
G
dx.doi.org/10.1021/ja309000s | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX